Vous êtes sur la page 1sur 323

Antonio Egidio Nardi

Rafael Christophe R. Freire


Panic Disorder
Neurobiological and
Treatment Aspects

Panic Disorder
Antonio Egidio Nardi
Rafael Christophe R. Freire

Panic Disorder
Neurobiological and Treatment
Antonio Egidio Nardi Rafael Christophe R. Freire
Laboratory of Panic and Respiration, Laboratory of Panic and Respiration,
Institute of Psychiatry Institute of Psychiatry
Federal University of Rio de Janeiro Federal University of Rio de Janeiro
Rio de Janeiro, Brazil Rio de Janeiro, Brazil

ISBN 978-3-319-12537-4 ISBN 978-3-319-12538-1 (eBook)

DOI 10.1007/978-3-319-12538-1

Library of Congress Control Number: 2016939239

Springer International Publishing Switzerland 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software,
or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publisher nor
the authors or the editors give a warranty, express or implied, with respect to the material
contained herein or for any errors or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG Switzerland
To Andrea

To Ana Catarina

Needless to say, I am immensely gratified to be asked to write the foreword

to this volume, which exemplifies the multiple, burgeoning approaches to the
understanding of panic.
Your editors historical approach refers to ancient descriptions of the panic
attack and the various attempts to develop an understanding of anxiety related
conditions. They state that, In 1954, Mayer-Gross associated anxiety disor-
ders to hereditary, organic and psychological factors, dividing them in simple
anxious states and phobic anxious states. In 1959, Donald Klein observed
that patients with these disorders responded favorably to imipramine, a tricy-
clic antidepressant.
Naturally, the protagonist of this event has a more detailed and compli-
cated recollection. Our concern in the late 1950s was to understand the effects
of this new agent, imipramine. European studies stated it to be an antidepres-
sant, so our expectation was that it would be a super cocaine and blow the
patients out of their rut. At Hillside Hospital, a psychoanalytic hospital, the
premium treatment was psychoanalytic psychotherapy without medication.
The average length of stay was 10 months. After the failure of therapy, the
patient was voluntarily referred to the Department of Experimental Psychiatry
where the senior scientist was Max Fink MD and I was a very junior associ-
ate. Max and I were confirmed anti-diagnosticians since all the excellent
studies of the 1950s indicated gross diagnostic unreliability. Psychoses could
be barely discriminated from neuroses and that was about it. Therefore, it
seemed foolish to attempt to fit medications to diagnoses.
So our first effort was a pilot trial trying to treat the entire range of patients
that had not responded to treatment with psychoanalytic psychotherapy,
chlorpromazine or imipramine. Our results led to a tentative classification
based upon response to medication [1, 2]. We noted that several patients,
plagued by anxiety attacks, had not responded to chlorpromazine, considered
the most potent anti-anxiety agent, but had responded during treatment with
imipramine. We were not sure if these patients might have been depressed.
We looked therefore for nondepressed patients with manifest anxiety to see if
imipramine had specific anti-anxiety effects, independent of their effects on
The first patient that met our criteria was incessantly complaining of his
fears of being alone, traveling, and dying. He demanded that he be accompa-
nied at all times. His hospitalization was initiated by his family. They were no
longer able to put up with his demands for constant reassurance and having a

viii Foreword

companion at all times. Yet, he was not depressed. He was radically pessimis-
tic about his fate but also had a lively interest in his circumstances, took
pleasure in gossip, spoke well, ate well, slept well, and laughed at jokes.
He was placed on imipramine 75 mg daily to be weekly increased by
75 mg. He remained in psychotherapy. I examined him before the clinical
open trial. The patient stated that by being placed on medication the doctors
had given up on him. I also weekly interviewed the therapist, the supervisor,
and the ward staff. The first 2 weeks were negative. The patient bitterly stated
the medication had done him no good, and his therapist, supervisor, and ward
staff agreed. His therapist indicated that he saw signs of loose associations
and was sure the patient was actually psychotic. Perhaps pseudo-neurotic
schizophrenia, a recently fashionable diagnosis applied to seemingly neurotic
patients who did not respond to therapy. The supervisor felt that the therapist
had not gone deep enough but did not elaborate.
After the third week of treatment, the patient, therapist, and supervisor
remained in pessimistic agreement. Nothing was happening.
Most of the ward staff persisted in negative evaluation, but one nurse
claimed that he was better, pointing out that for the past 10 months the patient
had run to nursing station 34 times a day, saying that he was dying and had
not done that for the past week. Previously, they had held his hand, told him
for the thousandth time that his heart was fine and that it was just his terrible
anxiety. After 20 min, he left quite dissatisfied to reappear 6 h later and go
through the same routine. Other nurses said this was unimportant since when-
ever you talked to the patient, it was the same litany of complaints.
Discussing this with the patient, I mentioned that he seemed to be better.
Who told you that?, he snarled. A nurse I stuttered. What do they
know?, he dismissed. When I asked him about not running to the nursing
station, he seemed puzzled, and quickly affirmed that he was anxious as ever.
Pressed about the change in his behavior, he finally noted, I guess I learned
that they cant do anything for me. I questioned, You mean after 10 months
you learned just this week?
Well, you have to learn sometime, he replied, thus anticipating the
development of cognitive behavioral therapy.
It became clear that the run to the nursing station was precipitated by an
attack. What confused me was that the overwhelming crescendo of the attack
was considered the worst anxiety but imipramine took the top off, while leav-
ing marked anxiety behind. It seemed reasonable that a medication would be
most effective at the moderate level of illness rather than the worst form. It
was at that time I realized that anxiety was heterogeneous and that the anxiety
attack should be called something else, to distinguish it from ordinary chronic
anxiety, itself considered an inappropriate manifestation of fear. Notably, the
chronic anxiety persisted. This led to my renaming the attack as a panic
attack. It also seemed, by reviewing detailed histories, that the phobic mani-
festations only occurred after the onset of repeated panic attacks. The phobias
of these hospitalized patients were limited to situations where, if they had a
panic attack, they couldnt get to help. This was not due to conditioning since
these patients often refused to fly, although they never had a panic attack on
an airplane. It was the possibility that paralyzed them. They were not afraid
Foreword ix

of dogs, or thunder, or heights. Also, it became apparent that the time between
repeated panic attacks and phobic manifestations was very variable. Some
patients withstood the attacks for many months while others folded up imme-
diately and became housebound.
In general with repeated attacks, chronic anticipatory anxiety developed
about when the next panic attack would happen. This led to severe travel limi-
tations by ensuring that help was always easily available. It could be radically
diminished by a companion. The propensity for chronic high levels of antici-
patory anxiety, unleashed by the panic, seemed an independent component of
the phobic development. A final confusing note was the frequent history of
separation anxiety disorder that became manifest when school was required,
leading to the misnamed school phobia. These children were not afraid of
school. School was a forced separation from their mother and they were over-
whelmed, which they explained by concern for the mothers welfare. Maybe
she got sick, they would whimper. However they usually got over this and
did not have panic attacks during this period. At sometime later, frequently
when they had to change schools or move the separation anxiety would recur.
Panic attacks occurred rarely before puberty and raised the suspicion of cere-
bral dysrhythmia. Later we found that Panic Disorder waxed and waned.
These early observations were followed by a series of controlled trials
regarding panic disorder and separation anxiety disorder [38].
However, an independent placebo controlled trial confirming the anti-panic
effect of imipramine awaited Sheehan and Ballenger [9], two nervy, smart resi-
dents at Massachusetts General Hospital, part of the Harvard empire. Hillside
Hospital had no academic links and nowhere near Harvards prestige.
This necessary independent study took place about 17 years after our ini-
tial presentations to a largely dismissive world. At Yale a senior psychoana-
lyst explained that these phobic patients did not travel because that requires
walking the streets and since they had an unconscious desire to be streetwalk-
ers (prostitutes) they couldnt do it. He thought imipramine must be a chemi-
cal straight jacket. Others were not as articulate but just as disbelieving.
The publication of DSM-III was also in 1980. By including Panic Disorder
in a manual for clinical diagnosis, it legitimatized research in Panic Disorder,
which took off after 1980. That research in psychiatry requires the prior
development of a clinical category is illogical. However, that this should lead
to rejection of clinical categories, as in the NIMH RDoC program, is glib and
misleading [10].

Emeritus Professor Psychiatry Donald F. Klein

Columbia University Medical Center
New York, NY, USA
x Foreword

1. Klein DF, Fink M. Psychiatric reaction patterns to imipramine. Am J Psychiatry. 1962;
2. Klein DF, Fink M. Behavioral reaction patterns with phenothiazines. Arch Gen
Psychiatry. 1962;7:44959.
3. Klein DF. Delineation of two drug-responsive anxiety syndromes. Psychopharmacologia.
1964; 5:397408. [Japanese version appeared in Archives of Psychiatric Diagnostic
and Clinical Evaluation 1993; 4:5058.]
4. Klein DF. Importance of psychiatric diagnosis in prediction of clinical drug effects.
Arch Gen Psychiatry. 1967;16:11826.
5. Gittelman-Klein R, Klein DF. Controlled imipramine treatment of school phobia. Arch
Gen Psychiatry. 1971;25:2047.
6. Zitrin CM, Klein DF, Woerner MG. Behavior therapy, supportive psychotherapy, imip-
ramine and phobias. Arch Gen Psychiatry. 1978;35:30716.
7. Zitrin CM, Klein DF, Woerner MG, Ross DC. Treatment of phobias (I): comparison of
imipramine hydrochloride and placebo. Arch Gen Psychiatry. 1983;40(2):12538.
8. Klein DF, Zitrin CM, Woerner MG, Ross DC. Treatment of phobias (II): behavior
therapy and supportive psychotherapy: are there any specific ingredients? Arch Gen
Psychiatry. 1983;40:13945.
9. Sheehan DV, Ballenger J, Jacobsen G. Treatment of endogenous anxiety with phobic,
hysterical, and hypochondriacal symptoms. Arch Gen Psychiatry. 1980;37(1):519.
10. Weinberge DR, Glick ID, Klein DF. Whither Research Domain Criteria (RDoC)?: the
good, the bad, and the ugly. JAMA Psychiatry. 2015;72(12):11612.

This book was carefully edited to condensate the accumulated knowledge on

panic disorder produced by a network of very productive researchers at this
subject in recent years. This network includes distinguished professors and
renowned researchers who enthusiastically participated from Brazil, United
States, Italy, Spain, United Kingdom, Mexico, and Switzerland.
We present several aspects of panic disorder including historical aspects,
neurobiology findings, connections with the respiratory and cardiovascular
systems, pharmacological and nonpharmacological treatments, psychopa-
thology and genetics, among others.
We recommend it to anyone who has an interest in anxiety and panic dis-
order, specially psychiatrists, clinical psychologists, postgraduate students,
and researchers in this area.

Rio de Janeiro, Brazil Rafael Christophe R. Freire

Antonio Egidio Nardi


1 The Panic Disorder Concept: A Historical Perspective............. 1

Antonio Egidio Nardi and Rafael Christophe R. Freire
2 A Neural Systems Approach to the Study of the
Respiratory-Type Panic Disorder................................................ 9
Luiz Carlos Schenberg
3 The Hippocampus and Panic Disorder: Evidence
from Animal and Human Studies ................................................ 79
Gisele Pereira Dias and Sandrine Thuret
4 Panic Disorder, Is It Really a Mental Disorder?
From Body Functions to the Homeostatic Brain........................ 93
Giampaolo Perna, Giuseppe Iannone, Tatiana Torti,
and Daniela Caldirola
5 Staging of Panic Disorder: Implications
for Neurobiology and Treatment ................................................. 113
Fiammetta Cosci
6 Panic Disorder Respiratory Subtype .......................................... 127
Morena Mourao Zugliani, Rafael Christophe R. Freire,
and Antonio Egidio Nardi
7 Lifelong Opioidergic Vulnerability Through Early
Life Separation: A Recent Extension of the False
Suffocation Alarm Theory of Panic Disorder............................. 139
Maurice Preter
8 Circadian Rhythm in Panic Disorder.......................................... 151
Michelle Levitan and Marcelo Papelbaum
9 Some Genetic Aspects of Panic Disorder .................................... 161
Fabiana Leo Lopes
10 Panic Disorder and Personality Disorder Comorbidity ............ 169
Ricard Navins, Elfi Egmond, and Roco Martn-Santos
11 Possible Mechanisms Linking Panic Disorder
and Cardiac Syndromes ............................................................... 185
Sergio Machado, Eduardo Lattari, and Jeffrey P. Kahn

xiv Contents

12 Panic Disorder and Cardiovascular Death:

What Is Beneath? .......................................................................... 203
Cristiano Tschiedel Belem da Silva and Gisele Gus Manfro
13 Pulmonary Embolism in the Setting of Panic Attacks .............. 211
Silvia Hoirisch-Clapauch, Rafael Christophe R. Freire,
and Antonio Egidio Nardi
14 Self-Consciousness and Panic ...................................................... 217
Thalita Gabnio, Andr B. Veras, and Jeffrey P. Kahn
15 Pharmacological Treatment with the Selective
Serotonin Reuptake Inhibitors .................................................... 223
Marina Dyskant Mochcovitch and Tathiana Pires Baczynski
16 Benzodiazepines in Panic Disorder ............................................. 237
Roman Amrein, Michelle Levitan,
Rafael Christophe R. Freire, and Antonio E. Nardi
17 Repetitive Transcranial Magnetic Stimulation
in Panic Disorder........................................................................... 255
Sergio Machado, Flvia Paes, and Oscar Arias-Carrin
18 Exercise in Panic Disorder: Implications for Disorder
Maintenance, Treatment and Physical Health ........................... 271
Aline Sardinha and Claudio Gil Soares de Arajo
19 Pharmacological Treatment of Panic Disorder
with Non-Selective Drugs ............................................................. 289
Patricia Cirillo and Rafael Christophe R. Freire

Index ....................................................................................................... 303


Roman Amrein Private Practice, Basel, Switzerland

Laboratory of Panic and Respiration, Institute of Psychiatry, Federal
University of Rio de Janeiro, Rio de Janeiro, Brazil
Claudio Gil Soares de Arajo Heart Institute Edson Saad, Federal
University of Rio de Janeiro, Rio de Janeiro, Brazil
Exercise Medicine Clinic (CLINIMEX), Rio de Janeiro, Brazil
Oscar Arias-Carrin Unidad de Trastornos del Movimiento y Sueo
(TMS), Hospital General Dr. Manuel Gea Gonzlez, Mxico, DF, Mexico
Unidad de Trastornos del Movimiento y Sueo (TMS), Hospital General
Ajusco Medio, Mxico, DF, Mexico
Tathiana Pires Baczynski Laboratory of Panic and Respiration, Institute of
Psychiatry, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Daniela Caldirola Department of Clinical Neuroscience, Villa San
Benedetto Menni, Hermanas Hospitalarias, FoRiPsi, Albese con Cassano,
Patricia Cirillo Laboratory of Panic and Respiration, Institute of Psychiatry,
Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Fiammetta Cosci Department of Health Sciences, University of Florence,
Florence, Italy
Gisele Pereira Dias Laboratory of Panic and Respiration, Institute of
Psychiatry, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Elfi Egmond Department of Psychiatry and Psychology, Hospital Clinic,
Barcelona, Spain
Department of Clinical and Health Psychology, Faculty of Psychology,
Universidad Autnoma de Barcelona, Cerdanyola del Valls, Barcelona,
Rafael Christophe R. Freire Laboratory of Panic and Respiration, Institute
of Psychiatry, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Thalita Gabnio Dom Bosco Catholic University (UCDB), Campo Grande,

xvi Contributors

Silvia Hoirisch-Clapauch Department of Hematology, Hospital Federal

dos Servidores do Estado, Ministry of Health, Rio de Janeiro, Brazil
Giuseppe Iannone Department of Clinical Neuroscience, Villa San
Benedetto Menni, Hermanas Hospitalarias, FoRiPsi, Albese con Cassano,
Jeffrey P. Kahn Department of Psychiatry, Weill-Cornell Medical College,
Cornell University, New York, NY, USA
Eduardo Lattari Laboratory of Panic and Respiration, Institute of
Psychiatry, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Michelle Levitan Laboratory of Panic and Respiration, Institute of
Psychiatry, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Fabiana Leo Lopes Laboratory of Panic and Respiration, Institute of
Psychiatry, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Sergio Machado Laboratory of Panic and Respiration, Institute of
Psychiatry, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Physical Activity Neuroscience, Physical Activity Sciences Postgraduate
Program, Salgado de Oliveira University, Niteri, Brazil
Gisele Gus Manfro Anxiety Disorders Program, Hospital de Clnicas de
Porto Alegre (HCPA), Porto Alegre, Brazil
Department of Psychiatry, Federal University of Rio Grande do Sul (UFRGS),
Porto Alegre, Brazil
Roco Martn-Santos Department of Psychiatry and Psychology, Hospital
Clinic, Institut dInvestigaci Biomdica August Pi I Sunyer (IDIBAPS),
Centro de Investigacin Biomdica en Red en Salud Mental (CIBERSAM),
G25, Universidad de Barcelona, Barcelona, Spain
Marina Dyskant Mochcovitch Laboratory of Panic and Respiration,
Institute of Psychiatry, Federal University of Rio de Janeiro, Rio de Janeiro,
Antonio Egidio Nardi Laboratory of Panic and Respiration, Institute of
Psychiatry, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Ricard Navins Department of Psychiatry and Psychology, Hospital Clinic
Institut dInvestigaci Biomdica August Pi I Sunyer (IDIBAPS), Centro de
Investigacin Biomdica en Red en Salud Mental (CIBERSAM), G25, and
Universidad de Barcelona, Barcelona, Spain
Flvia Paes Laboratory of Panic and Respiration, Institute of Psychiatry,
Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Marcelo Papelbaum State Institute of Diabetes and Endocrinology of Rio
de Janeiro, Rio de Janeiro, Brazil
Contributors xvii

Giampaolo Perna Department of Clinical Neuroscience, Villa San

Benedetto Menni, Hermanas Hospitalarias, FoRiPsi, Albese con Cassano,
Department of Psychiatry and Neuropsychology, Maastricht University,
Maastricht, The Netherlands
Department of Psychiatry and Behavioral Sciences, Leonard Miller School of
Medicine, University of Miami, Miami, FL, USA
AIAMC (Italian Association for Behavioural Analysis, Modification and
Behavioural and Cognitive Therapies), Milan, Italy
Maurice Preter Department of Psychiatry, College of Physicians and
Surgeons, Columbia University, New York, NY, USA
Department of Neurology, Mount Sinai School of Medicine, New York, NY,
Aline Sardinha Laboratory of Panic and Respiration, Institute of Psychiatry,
Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
Cognitive Therapy Association of Rio de Janeiro (ATC-Rio), Rio de Janeiro,
Luiz Carlos Schenberg Department of Physiological Sciences, Laboratory
of Neurobiology of Mood and Anxiety Disorders, Federal University of
Esprito Santo, Vitria, ES, Brazil
Cristiano Tschiedel Belem da Silva Anxiety Disorders Program, Hospital
de Clnicas de Porto Alegre (HCPA), Porto Alegre, Brazil
Department of Psychiatry, Federal University of Rio Grande do Sul (UFRGS),
Porto Alegre, Brazil
Sandrine Thuret Department of Basic and Clinical Neuroscience,
Laboratory of Adult Neurogenesis and Mental Health, Institute of Psychiatry,
Psychology and Neuroscience, Kings College London, London, UK
Tatiana Torti Department of Clinical Neuroscience, Villa San Benedetto
Menni, Hermanas Hospitalarias, FoRiPsi, Albese con Cassano, Italy
AIAMC (Italian Association for Behavioural Analysis, Modification and
Behavioural and Cognitive Therapies), Milan, Italy
Andr B. Veras Dom Bosco Catholic University (UCDB), Campo Grande,
Morena Mourao Zugliani Laboratory of Panic and Respiration, Institute of
Psychiatry, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil

5-HT Serotonin
A1 A1 noradrenergic group
A5 A5 noradrenergic group
ACTH Corticotropin
AIC Anterior insular cortex
Amb Nucleus ambiguous
AP Area postrema
APA American Psychiatry Association
BLA Basolateral amygdala
BNST Bed nucleus of stria terminalis
BZD Benzodiazepine
C1 C1 adrenergic group
CCHS Congenital central hypoventilation syndrome
CCK Cholecystokinin
CeA Central amygdala
CePAG Juxtaqueductal and medial sectors of periaqueductal grey
CnF Cuneiform nucleus
CO Carbon monoxide
CO2 Carbon dioxide
COR Cortisol
CORT Corticosterone
CRH Corticotropin releasing hormone
CSA Childhood separation anxiety
ACC Anterior cingulate cortex
dACC Dorsal anterior cingulate cortex
DGH Deakin/Graeff hypothesis
DLPAG Dorsolateral periaqueductal grey
DLSC Deep layers of superior colliculus
DMH Dorsomedial hypothalamus
DMHd Dorsomedial hypothalamus pars diffusa
DMPAG Dorsomedial periaqueductal grey matter
DMX Dorsal motor nucleus of vagus
DPAG Dorsal periaqueductal grey matter
DS Dorsal striatum
DSM Diagnostic and statistical manual of mental disorders
DSM-I Diagnostic and statistical manual of mental disorders first

xx Abbreviations

DSM-II Diagnostic and statistical manual of mental disorders sec-

ond edition
DSM-III Diagnostic and statistical manual of mental disorders third
DSM-III-R Diagnostic and statistical manual of mental disorders third
edition, revised
DSM-IV Diagnostic and statistical manual of mental disorders fourth
DSM-IV-TR Diagnostic and statistical manual of mental Disorders
fourth edition, text revision
DSM-5 Diagnostic and statistical manual of mental disorders fifth
eLPB External division of the lateral parabrachial area
EPM Elevated plus-maze
ES Escapable shock
ETM Elevated T-maze
FLI Fos-like immunoreactivity
FS Fictive shock
FST Forced swimming test
GABA Gamma-aminobutyric acid
GAD Generalized anxiety disorder
GiA Gigantocellular reticular area
HPA Hypothalamus-pituitary-adrenal axis
IS Inescapable shock
KCN Potassium cyanide
KF Klliker-Fuse nucleus
LAC Sodium lactate
L-AG l-allylglycine
LC Locus coeruleus
LDTg Laterodorsal tegmental nucleus
LLPDD Late luteal phase dysphoric disorder
LPAG Lateral periaqueductal grey
LPAGcv Caudoventral lateral periaqueductal grey
LPBA Lateral parabrachial area
LPGi Lateral paragigantocellular nucleus
MAOI Monoamine oxidase inhibitor
MDD Major depression disorder
MRI Magnetic resonance imaging
NE Norepinephrine
NMDA n-methyl-D-aspartic acid
NRD Nucleus raphe dorsalis
NRDd Nucleus raphe dorsalis pars dorsalis
NRDlw Nucleus raphe dorsalis lateral wing
NRDv Nucleus raphe dorsalis pars ventralis
NRM Nucleus raphe magnus
NRMn Nucleus raphe medianus
NRO Nucleus raphe obscurus
NRPa Nucleus raphe pallidus
Abbreviations xxi

NRPo Nucleus raphe pontis

NSI Neonatal social isolation
NTS Nucleus tractus solitarii
OVLT Organum vasculosum lamina terminalis
PACO2 Carbon dioxide alveolar partial pressure
PaCO2 Carbon dioxide arterial partial pressure
PAG Periaqueductal grey matter of the midbrain
PaO2 Oxygen arterial partial pressure
PBA Parabrachial area
pBC Pre-Btzinger complex
PCC Posterior cingulate cortex
PD Panic disorder
PeF Perifornical hypothalamus
PET Positron emission tomography
PETCO2 Carbon dioxide end-tidal partial pressure
PFC Prefrontal cortex
PH Posterior hypothalamus
PHA-L Leucoagglutinin of Phaseolus vulgaris
PIC Posterior insular cortex
PMD Premammillary dorsal nucleus of the hypothalamus
PPy Parapyramidal area
PRL Prolactin
PRV Pseudorabies virus
PVN Paraventricular nucleus of hypothalamus
PVT Paraventricular nucleus of the thalamus
rCBF Regional cerebral blood flow
RVLM Rostroventrolateral medulla
SAH Separation anxiety hypothesis
SFA Suffocation false alarm hypothesis
SFO Subfornical organ
sgACC Subgenual anterior cingulate cortex
SI Social interaction test
SON Supraoptic nucleus
SSRI Selective serotonin reuptake inhibitor
SNRI Serotonin and noradrenaline reuptake inhibitor
TASK Tandem acid-sensitive potassium channels
TCA Tricyclic antidepressants
TH Tyrosine hydroxylase
VHZ Ventral hypothalamic zone
VLM Ventrolateral medulla
VLPAG Ventrolateral periaqueductal grey matter
VMH Ventromedial hypothalamus
VMHdm Dorsomedial ventromedial hypothalamus
VMS Ventral medullary surface
WHO World Health Organization
The Panic Disorder Concept:
A Historical Perspective 1
Antonio Egidio Nardi
and Rafael Christophe R. Freire

Contents Abstract
1.1 Introduction 2 Panic disorder was described in several liter-
ary reports and folklore. Perhaps one of the
1.2 Historical Overview 2
oldest examples lies in Greek mythology, in
1.2.1 Classical Era 2
1.2.2 Medieval Period and Renaissance 3 which the Pan god was responsible for the
1.2.3 Nineteenth Century 3 term panic. Before 1850, the symptoms for
1.2.4 Twentieth Century First Half 5 anxiety were still usually associated with
1.2.5 The Last 65 Years 6
signs and symptoms of depression. In the sec-
1.3 Conclusion 7 ond half of the nineteenth century, a progres-
References 8 sive change began to take place in the field of
anxiety symptoms. Henry Maudsley, in 1879,
described a melancholic panic, and this was
the first time the term panic was technically
used in psychiatry. Jacob Mendes DaCosta,
during the American Civil War described the
irritable heart. At 1894, Sigmund Freud
described the angstneurose (anxiety neurosis)
and was impressed with the symptoms and
associated phobias. In 1954, Mayer-Gross,
associated anxiety disorders to hereditary,
organic and psychological factors, dividing
them in simple anxious states and phobic anx-
ious states. In 1964, Donald Klein, published
that patients with these disorders responded
favorably to imipramine, a tricyclic antide-
pressant. His observations and descriptions
influenced the third edition of the Diagnostic
A.E. Nardi R.C.R. Freire (*) and Statistical Manual of Mental Disorders
Laboratory of Panic and Respiration, Institute (DSM-III, 1980), in which the term panic dis-
of Psychiatry, Federal University of Rio de Janeiro,
order appears for the first time in an official
Rio de Janeiro, Brazil
e-mail: antonioenardi@gmail.com; classification. In 1993, Donald Klein described
rafaelcrfreire@gmail.com the False Suffocation Alarm Theory. This

Springer International Publishing Switzerland 2016 1

A.E. Nardi, R.C.R. Freire (eds.), Panic Disorder, DOI 10.1007/978-3-319-12538-1_1
2 A.E. Nardi and R.C.R. Freire

theory has been widely accepted, based on the god of flocks and shepherds, and since he was
laboratory studies of respiratory, cognitive half man and half goat, with horns and goat legs,
and biochemical tests. In the last 50 years, the his appearance was frightening. He was very
mysteries of panic disorder have been revealed active and full of energy, but very irritable. He
through basic and clinical research. loved music and played a small reed pipe, Syrinx.
In several stories, Pan is reported to cause fright,
Keywords screams, fears, terror and suffering. Like some
Panic disorder Agoraphobia Anxiety disor- other gods, Pan harassed nymphs who ran from
ders Anxiety Anti-anxiety agents him, maybe because of his mien or due to his
always unexpected and sudden apparition [3].
Like other woodland gods, he was feared by
those who had to go through the forest. Meeting
1.1 Introduction one of these deities could provoke an overwhelm-
ing and irrational fear, for no reason at all, or
Although a common mental disorder [1], panic what was known as panic terrors or panic attacks.
disorder has received little historical attention. Fear of meeting Pan again and fear of being star-
Berrios [2] pointed that this may be due to its tled once more made travelers stop journeying
relative newness; or to the fact that the historical through roads and avoiding going to the market
model used to account for the traditional mental (in Greek, gora), thus developing agoraphobia
disorder is inappropriate for the new disorders. (fear of large open or public places) [2].
Our aim is to describe some important points in In ancient Greece, Plato presents in Timaeus a
the history of the concept of panic disorder and case of anxiety associated to the wandering [4].
highlight the importance of the presence of this Although this description is often associated to
diagnosis in the official classifications for the hysteria, the original text describes a woman with
clinical and research developments. acute anxiety that is very similar to panic disor-
The word anxiety derives from the Indo- der. The uterus is an animal desirous of procre-
Germanic root angh for narrowness or constric- ating children. When it remains unfruitful long
tion. This rootword was the source for the Greek periods beyond puberty, gets discontented and
term anshein, meaning to strangulate, suffocate, angry, begins to wander throughout the body,
oppress and correlated Latin terms, such as closing the air passages, impeding breathing,
angustus, to express discomfort, angor, meaning bringing about painful distress, and causing a
oppression or lack of air, and angere, signifying variety of associated diseases. Plato not only
constriction, suffering, panic [2]. associated panic disorder to women of reproduc-
tive age, but also stated that respiratory symp-
toms (dyspnea, difficulty in breathing) are
1.2 Historical Overview common symptoms and pregnancy improves
them (in some cases) [4].
1.2.1 Classical Era From the time of Hippocrates up to the seven-
teenth century, the description and interpretation
Several literary reports and folklore demonstrate of the signs and symptoms of anxiety were deter-
that anxiety symptoms observed in the past are mined by the principle of body fluids described by
what we call panic disorder nowadays. Perhaps Hippocrates in Corpus Hippocraticum or by
one of the oldest examples of panic symptoms reports in layman and religious literature [2].
lies in Greek mythology the legend of the Pan Corpus Hippocraticum consisted of a series of
god. He was responsible for anxiety attacks and seventy medical treatises dating to the fifth cen-
originated the term panic [3]. Although he was tury AD, attributed to Hippocrates and his disci-
born in Arcadia, Pan roamed the Greek moun- ples. The traditional Egyptian idea of blood,
tains and roads. According to the legend, he was yellow bile, black bile and phlegm would be the
1 The Panic Disorder Concept: A Historical Perspective 3

four cardinal humors or fluids, and would be itself as an independent branch of study. Berrios
associated to distinctive features of each individ- [2] states that those who took care of patients
ual, including diseases. Hippocrates associated since immemorial time knew anxiety symptoms
the excess of black bile to depression [2]. and syndromes. However, each symptom was
treated as a separate medical complaint, as if it
were an isolated physical problem. Symptoms
1.2.2 Medieval Period were considered for its face value and treated as
and Renaissance symptoms associated to the disorder of an organ;
for example, palpitation was associated as a
During the medieval period and the Renaissance, disease of the heart. No mental disorder was con-
what we today consider to be severe anxiety syn- sidered when these complaints were presented.
drome was associated to the signs and symptoms Because of its physical characteristics tachy-
of depression. In the seventeenth century, the cardia, precordial discomfort, nausea, sweating,
English doctor Robert Burton [5] described in his paresthesia, etc. these anxiety manifestations
book The Essential Anatomy of Melancholy an were identified and described by clinical physi-
acute anxiety episode, which he considered to be cians and not by psychiatrists [2].
a type of fear: This fear cause in man, as to be In French psychiatry, panic disorder has been
red, pale, tremble, sweat; it makes sudden cold studied since the nineteenth century under the
and heat to come over all the body, palpitations of name of acute episodes of distress, and were part
the heart, syncope, etc. It amazed many men that of the description for various nosological entities
are to seek or show themselves in public assem- such as Benedict Morels emotional delirium
bliesMany men are so amazed and astonished (delire motif), Henri Le Grand Du Saulles fear
with fear, they know not where they are, what they of spaces (peur des espaces) based on Karl
do, and which is worst, it tortures them many days Westphals texts on agoraphobia; Doyens mor-
before with continual affrights and suspicious bid terrors (terreurs morbides), Henry Eys
Burton [5] described several types of patho- severe anxiety (grande anxit); from Brissauds
logical anxiety in a style quite different from paroxystic anxiety (anxit paroxystique) to the
todays scientific texts; associating philosophy and fear in the army (peur dans les armes) described
beliefs of the time. He listed the fear of death, fear by Brousseau [7]. In the beginning of the nine-
of losing a loved one and paranoid anxiety. He teenth century, the French doctor, Landr-
described anxiety based on delirium, associated Beauvais defined distress as a certain discomfort,
to depersonalization, to hyperventilation, to hypo- restlessness, excessive physical activity, and
chondria and even to agoraphobia, as well as that the symptoms could be associated to severe
anticipatory anxiety and several types of phobias or chronic diseases [7].
public speaking, acrophobia and claustrophobia. Maurice Krishaber (18361883), an otorhino-
The renaissance was not only a time of mans laryngologist who practiced in Paris, described
rediscovery of man through art, but it was also a the cerebral-cardiac neurosis associating some
time marked by sciences rise in prestige, and, at symptoms dizziness, tachycardia, restlessness,
this time, alchemy received scientific criteria. among others to one unique neurocirculatory
Paracelsus (14931541) was one of the leaders disease. The term neurosis referred to nerves, an
in the new art to question the hegemony of organic, somatic disturbance, with no association
Hippocrates humors [6]. to a psychiatric disorder [2, 7].
Benjamin Rush (17451813), an American
physician from Philadelphia, and considered as
1.2.3 Nineteenth Century the father of American psychiatry, described in
his psychiatry book (1812) the association
Modern medical description of panic disorder between somatic causes and phobias, relating
(PD) began prior to the nineteenth century, dur- depression (tristimania) to hypochondriasis [2, 7].
ing a period when psychiatry was established Before 1850, the symptoms for anxiety were still
4 A.E. Nardi and R.C.R. Freire

associated with signs and symptoms of depression. would be combined with hereditary factors, in
In 1858, Littr and Robin defined distress as a order to appear as a disease, from the moment
feeling of oppression or weight in the epigas- that determining conditions moral and physi-
trium, associated to a great deal of difficulty in cal were present. Base alteration would be the
breathing or excessive sadness, this being the functional fragility of the visceral, ganglionary,
most advanced degree of anxiety. They also nervous system [8]. Henry Maudsley (1835
described anxiety as a problematic and agitated 1918), in 1879, described the melancholic panic,
state, with difficulty in breathing and precordial and this was the first time the term panic was
pressure: restlessness, anxiety and distress are technically used in psychiatry [2]. It is important
three stages of the same phenomenon, in order of to highlight the notion that all the symptoms
seriousness [2, 7]. could be manifestations of a unitary construct
In the second half of the nineteenth century, a anxiety and this new concept had limited accep-
progressive change began to take place in the tance in nineteenth century psychiatry.
field of anxiety symptoms [6, 7]. Somatic causes, At the end of the nineteenth century, two
that up to that time were fully accepted, began do American doctors played a fundamental role in
divide the attention with possible psychological the future make up of Freuds proposed anxiety
causes. In 1872, Karl Friedrich Otto Westphal neurosis: George M. Beard and Jacob Mendes
(18331890) described agoraphobia, the fear DaCosta [2, 6, 7]. The first studied neurasthenia,
of wide and open places. He cited three male and was cited by Freud in his articles on the
patients who demonstrated fear in wide streets anxiety neurosis. Beards 1869 article entitled
and open spaces and who, at times, were com- Neurasthenia was the starting point for a series
pelled to ask passersby for help. In 1878, Le of papers with the objective of establishing the
Grand Du Saulle published a paper on the fear of specificity of neurasthenia, both in the clinical
spaces (peur des espaces), broadening Westphals description as well as in the explanatory hypoth-
original concept. Du Saulle stated that, although esis. According to Beard, neurasthenia would be
we can observe now and previously that these centered on physical fatigue of nervous origin
patients fear open places, they can feel fear of (related to nerves, a neurological concept)
theaters, churches, high balconies in buildings or functional weakness of the brain due to sexual
whenever they are found near wide windows, or energy drain from abnormal sexual activity, such
buses, boats or bridges [2, 6, 7]. as excessive masturbation, and would be accom-
In the second half of the nineteenth century, panied by other symptoms such as pain, digestive
definitions of states of anxiety and agoraphobia problems, paresthesia, depression, reduced libido,
became more specified and detailed [6, 7]. apathy and indifference. However, what was
Benedict Morels (18091873) studies were very most important, from the anxiety point of view
important during this period. According to Morel was that Beards ample description of neurasthe-
in 1866 [8], the category of emotional delirium nia considered acute anxiety a very important
combined the organisms physical and moral sen- part. The morbid fears of particular type, espe-
sibility symptoms, which have no relation what- cially agoraphobia, anthropophobia (fear of soci-
soever with the signs and symptoms that are ety) and the phobia of traveling alone. Freuds
today considered to be psychotic. Morel listed as criticism of Beards work lead Freud to introduce
physical symptoms: hyperesthesia, paresthesia, his own description for anxiety neurosis.
hot and cold flashes, sweating, pain, etc. Phobias Jacob DaCosta (18331900) [6], a military
were among the moral symptoms. The explana- doctor, during the American Civil War described
tion given by Morel for this emotional delirium what he called the irritable heart, a functional
was recorded in his general theory of degenera- cardiac disorder. Men affected by this disorder
tion. Among Morels cases, we can observe what demonstrated acute symptoms with palpitations
would today be described as PD and generalized of variable intensity, lasting from minutes to
anxiety disorder. The physical and moral causes hours, accompanied by thoracic pain and general
1 The Panic Disorder Concept: A Historical Perspective 5

discomfort. Since he could not find an organic were still largely associated to hereditary and
cardiac lesion or subjective conditions due to biological factors. In 1903, Pierre Janet [10]
war, Da Costa concluded that this was a func- described psychasthenia, a case of anxiety with
tional disorder of the sympathetic nervous sys- somatic and obsessive symptoms, associating
tem. Da Costa described a detailed clinical these signs and symptoms to a breakdown of
evaluation of almost 300 patients. When the heart feelings and liberation of primitive behavior.
is submitted to extremely intense effort and ten- Janet believed that the psychological perfor-
sion, it becomes physiologically irritable, and mance of an individual could be divided into five
as a result, palpitations arise. levels. At the superior level would be the harmo-
Sigmund Freud (18561939) was also impres- nious reality function, followed by habitual and
sed with the symptoms and phobias associated automatic actions, then imagination functions,
with what we call today panic disorder. At around emotional visceral reactions and muscular move-
1894, he [9] described the anxiety neurosis ments. The term and its definition would encom-
(Angstneurose). The origin of the term goes back pass a series of mental disturbances, including
to the studies by E. Hecker, who in 1893, had anxiety. Anxiety and distress were manifestations
already demonstrated the presence of anxious of this breakdown, but not its main component.
states in neurasthenia. Freuds merit was to sepa- In 1926, in his book, De lAngoisse a lExstase,
rate anxiety neurosis from neurasthenia and Janet [11] cites the case of Madelaine, a 40 year-
describe it with a specific clinical presentation. old patient with severe signs and symptoms of
On several occasions, Freud states that certain anxiety, with possible panic attacks associated
acute symptoms, such as dizziness, cardiac activ- with constitutional factors.
ity disorders, sweating, tremors, shock, diarrhea In 1907, Emil Kraepelin [12] (18561926),
and pavor nocturnus, are special forms of anxiety describes the neurosis of terror (Schreckneurose),
attacks, which he now denominates anxiety in which panic attacks are etiologically associ-
equivalent. Freud [9] stated these patients ated to the affective state. In the sixth edition
symptoms are not mentally determined or remov- of his classic Psychiatrie. Ein Lehrbuch fr
able by analysis, but they must be regarded as Studirende und Aerzte, 1899, in the chapter on
direct toxic consequences of disturbed sexual compulsive insanity, Kraepelin associated agora-
chemical processes. phobia to anxiety attacks with several somatic
Since Freuds first papers [9] on anxiety neu- symptoms. He related that the improvement of
rosis, the anxiety attack was considered as one of symptoms does not mean a concurrent
two fundamental forms of clinical manifestation improvement of the agoraphobia: this state could
of distress, the other being a chronic state. Freud persist indefinitely.
also associated agoraphobia to anxiety neurosis. Ernest Kretschmer introduced in his Medical
To him, agoraphobia referred to an effort made Psychology the panic attack as an an outburst
by the patient with the intention of not being of attempted impulsive movements, trying to
stricken by anxiety attacks in unfamiliar circum- take the individual away from the source of dan-
stances, when he was not be sure he could get ger or excitement as quickly as possible [6].
help. It is interesting to note that Freud also dis- Panic attacks appear in the Legrand Du Salles
cussed the relation of agoraphobia and panic fear of spaces, which determined the fear of feel-
attacks, widely studied nowadays, in his texts on ing fear.
the neurosis of anxiety. To Adolf Meyer [13] (18661950), a psychia-
trist who influenced the first and second editions
of the Diagnostic and Statistical Manual of
1.2.4 Twentieth Century First Half Mental Disorders (DSM-I and DSM-II), the indi-
vidual is a psychobiological being and any patho-
In the twentieth century, although some concepts logical manifestation would be a form of reaction
on psychological factors in anxiety having to particular characteristics of the environment.
already been discussed, the symptoms of anxiety Psychosis and neurosis would be different parts
6 A.E. Nardi and R.C.R. Freire

of the same psychiatric spectrum of continuous trips to the nursing station to complain about
gradation, going from one extreme to the other. being sick or dying; that they were more indepen-
Henrique Roxo [14] a Brazilian Professor of dent, walking around the hospital by themselves.
Psychiatry (1946) divided neurasthenia in two Klein [16] concluded that imipramine was effi-
groups: psychasthenis and nervousness. cient in panic attacks, but not on chronic anxiety.
Psychasthenis included obsessions, phobias and He also considered agoraphobia a consequence
impulses. Nervousness represented an extraordi- of panic attacks, where patients who did not fear
nary state of anxiety, (the patient) revealing a feel- bridges or closed environments, but feared the
ing of indescribable discomfort, in which possibility of having a panic attack and an imme-
kinesthesia disturbances play a very important role. diate way out or help would be difficult or
Although the anxiety attack is described within the impossible.
symptoms of nervousness, Roxo [14] associated to Klein [16] more recently also distinguished
the two neurasthenia groups symptoms that nowa- three types of panic attacks: spontaneous, situa-
days are considered those of depressive syndromes tional (associated with an agoraphobic situation)
and chronic and acute anxiety syndromes. Etiology and those provoked by a constant phobic stimu-
was an association of psychological, environmental lus (animals, height, darkness, etc.). His posterior
and constitutional factors, and treatment consisted observations and descriptions influenced the
of varied experimental medication, which suppos- third edition of the Diagnostic and Statistical
edly would act specifically on each subtype. Manual of Mental Disorders (DSM-III) (1980)
[17], in which the term panic disorder appears
for the first time in an official medical classifica-
1.2.5 The Last 65 Years tion. In no time, panic disorder became the most
studied psychiatric disorder from a diagnostic
The second half of the twentieth century marked and therapeutic point of view.
a revolution in the practice of psychiatry. Not The DSM-III [17] divided anxiety neurosis,
only was psychiatric diagnosis revised and modi- also named in this classification of anxiety disor-
fied, seeking reliability, but treatment also ders, as panic disorder (acute anxiety) and gener-
received marked assistance from psychophar- alized anxiety disorder (chronic anxiety), creating
macological agents. In 1954, Mayer-Gross [15] operational criteria for each category diagnosed.
(18891961), associated anxiety reaction to It also divided phobic neurosis into simple
hereditary, organic and psychological factors, phobia, social phobia and agoraphobia (with or
dividing it in simple anxious states and phobic without panic attack). The extensive diagnostic
anxious states. The latter included agoraphobia reorganization of DSM-III continued for 7 years,
with associated somatic symptoms. The 50s resulting in the revised edition DSM-III-R
brought about the discovery of the monoaminox- (1987) [18]. In this edition, agoraphobia no lon-
idase inhibitors and tricyclic antidepressants. ger appears as an isolated category, but as a con-
This was followed by the advent of the benzodi- sequence of panic disorder, and now listed under
azepines. Thus, the road was paved for a more the term: panic disorder with and without agora-
efficient panic treatment. In 1959, Donald phobia. The criteria for panic disorder were
Klein [16], a New York psychiatrist, observed simplified and closer to clinical practice. One
that patients with depressive-anxiety symptoms example was that only one panic attack with pho-
responded favorably to imipramine, a tricyclic bic repercussion during the previous month was
antidepressant. It is interesting to note that after a required to establish a positive diagnosis. In other
few weeks, Klein [16] and his patients were dis- words, greater importance was given to the pho-
appointed by the effects of imipramine and were bic consequences of the panic attack and not
ready to quit the experiment when the nurses just the physical symptoms of the attack. The
pointed out that the patients were less anxious. DSM-IV (1992) [19] maintained practically the
The nurses noticed that the patients reduced their same definitions, but defined panic attacks,
1 The Panic Disorder Concept: A Historical Perspective 7

demonstrating they could occur associated to Hout, in the Netherlands, observed that 35 %
other diagnosis and without fulfilling all the cri- carbon dioxide mixture induced panic symptoms
teria for panic disorder; it also distinguished [25]. This observation afforded the opportunity
spontaneous panic attacks, situational panic for greater insight concerning the factors that
attacks (linked to agoraphobia) and those pro- brought about panic attacks.
voked by a phobic stimulus (more closely linked In 1993, Donald Klein [26] described the
to specific phobias). In the year 2000, a revised False Suffocation Alarm Theory. This theory
edition of DSM-IV was published, entitled DSM- describes panic attack as a disorder of the physi-
IV-TR [20], in which some of the concepts were ological suffocation alarm. The regulating moni-
refined but the criteria for panic disorder remained tor would inform the central nervous system of
the same. In 2013, the DSM-5 [21] separated an imminent suffocation situation when this was
again panic disorder from agoraphobia, although not actually occurring. This theory has been
the emphasis on the compromise of a panic attack widely accepted, based on laboratory studies of
to the diagnosis of panic disorder was kept. respiratory, cognitive and biochemical tests.
We can divide the evolution of psychophar- Recent literature indicates that panic attacks
macology in relation to panic disorder into may originate from a network of fear with altered
three main moments: first, Donald Kleins (1964) sensibility, including in this network; the pre-
observation of the efficiency of the tricyclic anti- frontal cortex, the insula, the thalamus, the amyg-
depressants [16]. The second moment was when dala and amygdala projections toward the brain
the efficacy of the benzodiazepines was per- stem and hypothalamus [25]. When we adminis-
ceived [22]. Finally, by the noted efficacy of the ter a panicogenic agent, we would not affect a
selective serotonin reuptake inhibitors [23] in specific autonomic area of the brain stem, but we
1990. Today, psychopharmacology has been would be activating the entire neural fear net.
leading psychiatry in the direction of biology. Patients with panic disorder often complain of
This biological perspective entails in putting anx- uncomfortable somatic sensations. The adminis-
iety in the frame of the evolutionist paradigm. tration of a panicogenic agent would correspond
Charles Darwin (1872) [24], in The Expression of to a non-specific activation. Since all these agents
Emotion in Man and in Animals, pointed the way produce uncomfortably sharp physical sensa-
to search for the adaptive value of the behavioral tions, the hypothesis would be that they stimulate
and psychological processes. Anxiety and fear a sensitive brain system conditioned to respond-
have their roots in the defensive reactions of ing to noxious stimuli. As time passes, the pro-
animals, observed in response to the danger nor- jections of the central nucleus of the amygdala
mally found in the environment. The interpretation such as the locus ceruleus, periaqueductal gray
of a stimulus or a situation as dangerous depends area, and hypothalamus can become more or less
on the nature of cognitive operations. In humans, sensitive [25]. There can also be an interindivid-
cognitive factors acquire importance due to the ual difference in the strength of these afferent
intervention of the system of symbols socially projections. In this way, the standard neuroendo-
codified, verbal or non-verbal. The behavioral crine and autonomic responses presented during
responses to fear are accompanied by intense a panic attack may vary from one patient to
physiological alterations physical symptoms another, and on the same patient throughout time.
and alterations in the emotional state. The physi-
ological alterations consist of objective measures
of anxiety. Thus, cardiac frequency, arterial pres- 1.3 Conclusion
sure, respiratory frequency and increased electri-
cal conductivity of the skin brought on by Panic disorder has been having its mysteries
sweating, are the measures mostly used to deter- revealed through basic and clinical research, and
mine the degree of anxiety. patients who suffer from this frightening disease
In the 80s, in their independent research proj- can be sure that correct diagnosis and adequate
ects, Jack Gorman, in the USA, and Van Den treatment are already part of everyday clinical
8 A.E. Nardi and R.C.R. Freire

practices. We have, however, to perfect these 13. Brodsky A. Benjamin rush: patriot and physician.
practices even more, so we can continue to New York: Saint Martins Press; 2004.
14. Roxo HB. Manual de Psiquiatria. Rio de Janeiro:
improve patient prognosis. Editora Guanabara; 1946.
15. Mayer-Gross W. Clinical psychiatry. London:
Baillire, Tindall & Cassell; 1954.
16. Klein DF. Delineation of 2 drug-responsive anxiety
References syndromes. Psychopharmacologia. 1964;5(6):397
408. doi:10.1007/Bf02193476.
1. Weissman MM, Bland RC, Canino GJ, Faravelli C, 17. American Psychiatric Association. Diagnostic and sta-
Greenwald S, Hwu HG, et al. The cross-national epi- tistical manual of mental disorders: DSM-III. 3rd ed.
demiology of panic disorder. Arch Gen Psychiatry. Washington, DC: American Psychiatric Association;
1997;54(4):3059. 1980.
2. Berrios GE. Anxiety and cognate disorders. In: 18. American Psychiatric Association. Diagnostic and
Berrios GE, editor. The history of mental symptoms: statistical manual of mental disorders: DSM-III-R.
descriptive psychopathology since the nineteenth cen- 3rd ed. Washington, DC: American Psychiatric
tury. Cambridge: Cambridge University Press; 1996. Association; 1987.
3. Merivale P. Pan, the goat-god: his myth in modern 19. American Psychiatric Association. Diagnostic and sta-
times. Cambridge: Harvard University Press; 1969. tistical manual of mental disorders: DSM-IV. 4th ed.
4. Plato. Plato: complete works. Indianapolis: Hackett Washington, DC: American Psychiatric Press; 1994.
Publishing Company; 1997. 20. American Psychiatric Association. Diagnostic and
5. Burton R. The essential anatomy of melancholy. statistical manual of mental disorders: DSM-IV-TR.
New York: Dover Publications; 2002. 4th ed. Washington, DC: American Psychiatric Asso-
6. Stone MH. History of anxiety disorders. In: Stein DJ, ciation; 2000.
Hollander E, editors. Textbook of anxiety disorders. 21. American Psychiatric Association. Diagnostic and
Washington, DC: American Psychiatric Press; 2002. statistical manual of mental disorders: DSM-5. 5th ed.
p. 311. Washington, DC: American Psychiatric Association;
7. Costa-Pereira ME. Pnico: contribuio psicopato- 2013.
logia dos ataques de pnico. Sao Paulo: Lemos 22. Sheehan DV. Current perspectives in the treatment of
Editorial; 1997. panic and phobic disorders. Drug Ther. 1982;12:
8. Morel BA. Trait des dgnrescences physiques, 17993.
intellectuelles et morales de l espce humaine. 23. Boyer W. Serotonin uptake inhibitors are superior
New York: Arno Press; 1976. to imipramine and alprazolam in alleviating panic
9. Freud S. Autoprsentation, Inhibition, symptme attacksa metaanalysis. Int Clin Psychopharmacol.
et angoisse, Autres textess. Oeuvres compltes 1995;10(1):459. doi:10.1097/00004850-199503000-
Psychanalyse. Paris: Presses Universitaires de France; 00006.
1992. 24. Darwin C. The expression of emotion in man and in
10. Janet P. Les obsessions et la psychasthnie. New York: animals. London: Fontana Press; 1999.
Arno Press; 1976. 25. Gorman JM, Kent JM, Sullivan GM, Coplan
11. Janet P. De langoisse lextase. Paris: Flammarion; JD. Neuroanatomical hypothesis of panic disorder,
1923. revised. Am J Psychiatry. 2000;157(4):493505.
12. Kraepelin E. Psychiatrie: ein lehrbuch fr stu- 26. Klein DF. False suffocation alarms, spontaneous pan-
dirende und aerzte. New Delhi: Science History ics, and related conditions. An integrative hypothesis.
Publications; 1990. Arch Gen Psychiatry. 1993;50(4):30617.
A Neural Systems Approach
to the Study of the Respiratory- 2
Type Panic Disorder

Luiz Carlos Schenberg

2.8 Evidence of a Suffocation Alarm System

Contents Within the Periaqueductal Grey Matter of
2.1 A Brief Story of Panic and Panic Theories 10 Midbrain 30
2.8.1 Brain Detectors of Hypoxia and
2.2 Idiosyncratic Features of Clinical Panics 12 Hypercapnia 30
2.3 Early Neuroanatomical Models of Panic 2.8.2 Carbon Dioxide Inhibits DPA-
Disorder 14 Evoked Panic-Like Responses 31
2.8.3 Peripheral Injections of Potassium
2.4 The Suffocation False Alarm Theory: Facts Cyanide Elicits Respiratory-Type
and Questions 18 Panic Responses 32
2.8.4 On the Probable Substrates of
2.5 Face Validity of Dorsal Periaqueductal
Kleins Three-Layer Cake Model of
Grey Matter Mediation of Panic Attacks
Panic Attack 37
in Humans 19
2.8.5 Steady State Functioning of
2.6 The Periaqueductal Grey Matter of the Suffocation Alarm System 40
Midbrain: A Likely Substrate of 2.8.6 On respiratory and Non-Respiratory
Respiratory-Type Panic Attacks 20 Panic Attacks 40
2.8.7 Overview of Suffocation Alarm
2.7 Neuroanatomical Basis of Respiratory- System of Mammals: An Evidence-
Type Panic Attacks 22 Based Hypothesis 42
2.7.1 Psychophysics of Asphyxia 22 2.8.8 Insights into a Neurochemical
2.7.2 Neuroimaging Hunger for Air in Puzzle 45
Humans 24
2.7.3 c-Fos Immunohistochemistry 2.9 Modeling Lactate Vulnerability in Rats 47
Studies Reveal a Limited Number
2.10 Modeling Neuroendocrine
of Structures Activated by Hypoxia
Unresponsiveness of Panic Attacks 51
or Hypercapnia 25
2.7.4 c-Fos Studies in Rats Exposed to 2.11 Modeling the Comorbidity of Panic
Carbon Monoxide Supports the Disorder with Childhood Separation
Key Role of Dorsal Periaqueductal Anxiety 54
Gray in Respiratory-Type Panic
Attacks 30 2.12 Modeling the Comorbidity of Panic and
Depression 56
2.13 Modeling Female Vulnerability to Panic
Disorder 58

L.C. Schenberg (*) 2.14 Conclusion 59

Department of Physiological Sciences, Laboratory 2.15 References 61
of Neurobiology of Mood and Anxiety Disorders,
Federal University of Esprito Santo,
Vitria, ES, Brazil
e-mail: luiz.schenberg@gmail.com

Springer International Publishing Switzerland 2016 9

A.E. Nardi, R.C.R. Freire (eds.), Panic Disorder, DOI 10.1007/978-3-319-12538-1_2
10 L.C. Schenberg

Abstract 2.1 A Brief Story of Panic

Panic disorder (PD) patients are exquisitely and Panic Theories
and specifically sensitive to inhalations of
57 % carbon dioxide and infusions of 0.5 M The underpinnings of the current classification of
sodium lactate. Another startling feature of anxiety disorders were laid down by Sigmund
clinical panic is the lack of increments of the Freud in the course of his efforts to separate anxi-
stress hormones corticotropin, cortisol and ety neuroses (Angstneuroses) from both neuras-
prolactin. PD is also more frequent in women thenia and melancholia [1, 2]. Freud reported
and shows high comorbidity with childhood that anxious patients had two major syndromes of
separation anxiety, late luteal period dysphoric anxiety, the anxious expectation (ngstliche
disorder and depression. The hypothalamus- Erwartung), which he considered the most essen-
pituitary-adrenal axis is nevertheless activated tial syndrome of anxiety, and the less frequent
in fear-like panics marked by palpitations, attack of anxiety (Angstanfall). According to
tremor and sweating, that are devoid of suffo- Freud, in anxious expectation there is a quantum
cation symptoms. These and other data sug- of freely floating anxiety which controls the choice
gest the existence of both respiratory and of ideas by expectation. In contrast, in anxiety
non-respiratory types of panic attacks. attacks anxiety breaks suddenly into conscious-
Increasing evidence suggests, on the other ness without being aroused by the issue of any
hand, that panics are mediated at midbrains idea. Freud emphasized that anxiety attacks
dorsal periaqueductal grey matter (DPAG). manifest either as the anxious feeling alone or
Therefore, here we summarized data showing the combination of this feeling with the nearest
that: (1) the DPAG harbors a suffocation alarm interpretation of the termination of life, such as
system which is activated by low intravenous the idea of sudden death or threatening insanity.
doses of potassium cyanide (KCN); (2) KCN Remarkably, Freud noted that patients suffering
evokes defensive behaviors that are facilitated from anxiety attacks put the feeling of anxiety to
by hypercapnia, blocked by lesions of DPAG the background or [described it] rather vaguely
and attenuated by clinically effective treat- as feeling badly, uncomfortably, etc., an obser-
ments with panicolytics; (3) DPAG stimula- vation corroborated by present-day psychiatrists.
tions do not change the stress hormones when Freud was also aware of the high comorbidity of
escape is prevented by stimulating the rats in a panic attacks with agoraphobia, which he linked
small compartment; (4) DPAG-evoked panics to a locomotion disorder associated with the
responses are facilitated in neonatally-isolated presence of dizziness during panic attacks [1, 2].
adult rats, a model of childhood separation Moreover, he stressed that in agoraphobia we
anxiety; (5) DPAG-evoked panic-like behav- often find the recollection of a panic attack; and
iors are facilitated in diestrus phase of rat ovu- what patients actually fears is the occurrence of
latory cycle. It is proposed a neural model of such attack under the special condition in which
panic attacks in which the PAG is the fulcrum he believes he cannot escape it [3].
of threatening signals from both forebrain and Freuds descriptions of anxious expectation
hindbrain. This model emphasizes the role of and anxiety attack were very similar to the pres-
PAG as a suffocation alarm system. ent diagnoses of generalized anxiety disorder
(GAD) and panic disorder (PD), respectively.
Yet, it would take almost a century before PD be
Keywords accepted as a psychiatric disorder on its own [4].
Panic disorder Separation anxiety Consequently, while the anxiety disorders con-
Periaqueductal gray Hypothalamo-hypophyseal tinued to be vaguely diagnosed as neurasthenia
system Adrenal glands Adrenocorticotropic until the middle of the last century, panic attacks
hormone Hydrocortisone Prolactin (or similar reactions) received a bewildering
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 11

wealth of labels, including anxiety neurosis, neu- was a remarkable finding as it discriminated
rocirculatory asthenia, vasomotor neurosis, ner- panic from both fear and stress [1320]. These
vous tachycardia, effort syndrome, Da Costas studies also questioned the cognitive hypotheses
syndrome, soldier heart, irritable heart and hyper- that equated clinical panic to the fear brought
ventilation syndrome [58], and were even about by the catastrophic evaluation of bodily
treated as schizophrenics (D.F. Klein, personal symptoms [2123].
communication). As a result, in the early 1990s Klein proposed
This scenario began to undergo a dramatic that clinical panic is bound to the fear of suffoca-
change following the publication of Donald tion of which dyspnea (but not hyperpnea) is the
Kleins influential studies showing that anxious leading symptom [24, 25]. In particular, Klein
expectation and panic were treated by different suggested that the apparently spontaneous clini-
classes of drugs [9, 10]. Serendipitously, Klein cal panic is the outcome of the misfiring of an
observed that, although the panic attacks of hos- as-yet-unknown suffocation alarm system,
pitalized agoraphobics remitted during the treat- thereby producing sudden respiratory distress
ment of comorbid depression with the tricyclic (dyspnea), panic, hyperventilation, and the urge
antidepressant imipramine, the expectant (antici- to escape from immediate situation. Klein [24]
patory) anxiety of being overwhelmed by a panic argued that the suffocation false alarm (SFA)
attack was refractory to this medication. Notably, theory is an explanation both sufficient and con-
Klein also noted that the onset of PD was very sistent of the hypersensitivity of panic patients to
often precipitated by separation and bereavement respiratory metabolites (CO2, LAC) and of the
and that half of agoraphobics of his studies had occurrence of panic attacks during hypercapnic
suffered from severe childhood separation anxi- conditions of relaxation and sleep. Klein [24]
ety (CSA) that frequently prevented school atten- also stressed that the SFA theory is consistent
dance [10]. Besides suggesting that panic differs with the reduced incidence of panic attacks dur-
from anxiety, these observations led Klein to pro- ing pregnancy, delivery, and lactation and, con-
pose that CSA predisposes individuals to the later versely, of the increased frequency of panic in
development of PD even before Bowlbys publi- late luteal phase dysphoric disorder (LLPDD), in
cation of the classical trilogy on attachment, sep- which respiration is increased or decreased by
aration anxiety and loss [11]. parallel changes of respiratory stimulant proges-
At approximately the same time, Pitts and terone, respectively [24]. The SFA hypothesis is
McClure [7] showed that panic attacks could be also consistent with the high comorbidity of
precipitated by a 20-min intravenous infusion of panic with present and antecedent respiratory
0.5 M sodium lactate (LAC) in patients prone to diseases [26, 27]. Lastly, Klein [24] argued that
panic but not in normal volunteers. Further stud- panics induced by 7 % CO2 and 0.5 M LAC
ies showed that a fraction of LAC-sensitive remained the best models of clinical panic
patients was also sensitive to the inhalation of because they are not precipitated in healthy sub-
57 % carbon dioxide (CO2) [12]. The demon- jects [7, 24], obsessive patients (as cited by Griez
stration that PD had both physiological markers and Schruers [28]) or patients with social phobia
(CO2, LAC) and drug-specific treatments (imip- [29]. Moreover, whereas panics produced by CO2
ramine) suggested that clinical panic was the out- and LAC were blocked by chronic treatment with
come of a specific brain circuit discharging imipramine [3033], those induced by -carboline
maladaptively. and yohimbine were not [24].
The next breakthrough occurred a few years Kleins original hypothesis [10] that CSA
after the acknowledgment of PD as a separate predisposes to panic was corroborated by evi-
anxiety syndrome [4]. Indeed, the counterintui- dence both clinical and epidemiological [3437].
tive lack of increments in stress hormones cor- The so-called separation anxiety hypothesis
ticotropin (ACTH), cortisol (COR) and prolactin (SAH) [36] was also supported by the demon-
(PRL) in both natural and provoked panic attacks stration of the clinical effectiveness of imipramine
12 L.C. Schenberg

in CSA [34, 38]. Additionally, it was shown that gated by Shioiri et al. [55]. Cluster analysis revealed
separation-anxious children of parents with PD that panic symptoms are clustered in three groups:
had respiratory responses to CO2 similar to cluster A (dyspnea, choking, sweating, nausea,
those of panickers [35, 39]. Most importantly, flushes/chills); cluster B (dizziness, palpitations,
twin-based genetically-informative recent stud- trembling or shaking, depersonalization, agorapho-
ies suggested that CSA shares a common genetic bia, and anticipatory anxiety); and cluster C (fear of
diathesis with both PD and CO2 hypersensitivity dying, fear of going crazy, paresthesias, and chest
[40, 41]. The necessity of an integrative explana- pain or discomfort). Accordingly, whereas the clus-
tion resulted in the expanded theory of SFA ter A includes mostly respiratory symptoms, clus-
[27, 42] according to which panic attacks are the ters B and C include physiological and cognitive
outcome of the episodic dysfunction of opioider- symptoms alike those of fear.
gic systems tonically inhibiting both suffocation In particular, Perna et al. [56] examined the
and separation alarm systems. The opioidergic different types of dyspnea induced by 35 % CO2
hypothesis of panic attacks relied on clinical and challenges in patients with PD. Factor analysis
preclinical evidence of the crucial role of opioids identified 3 main factors: breathing effort, sense
in both respiration [4345] and parental bonding of suffocation, and rapid breath. Factor scores for
[4649]. The existence of suffocation and sepa- breathing effort and sense of suffocation sig-
ration alarm systems is still vividly debated in nificantly discriminated between patients who
present time. did and those who did not report CO2-induced
Competing theories propose that panic attacks panic attacks, respectively. Factor scores for
are either the catastrofization of bodily symptoms breathing effort loaded significantly for patients
by cortically-mediated cognitive processes [21 whose reaction resembled unexpected panic
23] or the mistaken activation of fear-like responses attacks. Authors suggested that although the
to proximal threats [5052]. Although all these sense of suffocation was linked to an increased
theories propose that panics are false alarms, fac- sensitivity to CO2, it may not be the main factor
tors triggering panics remain largely obscure. of unexpected panic attacks.
These studies are in line with evidence amassed
in the last decades suggesting that panics may be
2.2 Idiosyncratic Features either respiratory or non-respiratory, depending on
of Clinical Panics the prominence of respiratory symptoms [57, 58].
The existence of two types of panic was corrobo-
The core symptoms of panic attacks were largely rated by latent class analysis of the temporal sta-
conserved ever since they were first described by bility, psychiatric comorbidity and treatment
Sigmund Freud in late eighteens (Table 2.1). outcome in a large-scale epidemiological survey
Goetz et al. [53] found in addition that panickers [59]. Data showed that while the temporal stability
rate their experience of LAC-induced panic as of panic subtypes was mainly observed in females,
very much like the spontaneous panic attacks with respiratory panics were associated with both the
respect to both quality (76 %) and severity (84 %) more severe forms of PD and the increased comor-
of symptoms. Although the patients reported that bidity with depression and other anxiety disorders.
LAC-induced panics were at most moderate, the Yet, treatment outcome did not suggest that panic
severity of panic attacks correlated mostly with subtypes respond differentially to imipramine and
desire to flee (0.70), fear of losing control (0.57), alprazolam. The latter result disagrees from
afraid in general (0.49) and dyspnea (0.48). Kleins [24] proposal that panic sensitivity to tricy-
Curiously, however, the desire to flee is not clics and benzodiazepines parallel laboratory pan-
included in clinical symptomatology (Table 2.1). ics reminiscent of either asphyxia (as provoked by
The relationships of panic attack symptoms, LAC, CO2, bicarbonate and isoproterenol) or fear,
including anticipatory anxiety, agoraphobia, and respectively (as provoked by yohimbine, flumaze-
13 clinical symptoms of DSM-III-R, was investi- nil, benzodiazepine inverse agonists, caffeine and
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 13

Table 2.1 Symptoms of spontaneous and lactate-induced panic attacks

Diagnostic and statistical manualc
Panic attack symptoms Freuda RDCb III III-R VI V ICD-10d APIe
Palpitations, pounding or + + + + + + + +
Chest pain or discomfort + + + + + + + +
Dyspnea, shortness of breath + + + + + + + +
Sensations of smothering or + + + + + + + +
Trembling or shaking + + + + + + + +
Dizzy, unsteady, lightheaded, + + + + + + + +
or faint
Fear of losing control or + + + + + + + +
going crazy
Fear of dying + + + + + + + +
Paresthesias (numbness, + + + + + + +
Sweating + + + + + + +
Chills or hot flushes + + + + + +
Derealization, + + + + + +
Nausea or abdominal distress + + + +
Difficulty in concentrating + +
Difficulty of speaking + +
Desire to flee +
Feeling confused +
Freud [1]
Research diagnostic criteria
American Psychiatric Association (APA)
World Health Organization (WHO)
Lactate-induced symptoms in panicking patients relative to non-panicking controls [53] assessed by a 21-item modified
version of the Acute Panic Inventory of Dillon et al. [54]

methyl-phenylpiperazine). As well, Klein [24] to fear or, at the very best, to a proximal threat-
suggested that more traumatic spontaneous panics induced fear [50, 52, 68].
stem primarily from mistaken signals of suffoca- Remarkably, however, Beitman and co-workers
tion and are particularly benefited by selective presented evidence of panic attacks devoid of fear
serotonin (5-HT) reuptake inhibitors (SSRIs). [6971]. In particular, Beitman et al. [69] showed
Lastly, Klein [24] proposed that patients who that 12 of 38 cardiology patients with chest pain
panic during CO2 inhalations have higher inci- and current PD neither experienced intense fear,
dence of panic and that respiratory-type panics are nor fear of dying, or of losing control, or of going
both specific to PD and distinct from fear. crazy in their last major panic attacks [69].
Evidence of a two-type panic attack may Notably, as well, Fleet et al. [71] reported that 43
explain why some neurobiological studies on of 104 cardiology patients with pseudoangina
panic focus on midbrains periaqueductal gray were diagnosed as having PD (38 current, 5 past)
matter (PAG) [5052, 6062] while others empha- as part of a prevalence study of panic in cardiology
size the amygdala [63, 64], the hypothalamus [65, patients. Fleet et al. [70] also compared 48 non-
66], or the locus coeruleus (LC) [67]. Moreover, fearful panic disorder (NFPD) patients with 60
whereas clinical studies emphasize respiratory- PD patients and 333 controls at the time of
type panics, most preclinical models equate panic admittance at emergency service and after a fol-
14 L.C. Schenberg

low-up of approximately 2 years. Of note, a sig- ity of noradrenergic neurons of LC. Gormans
nificantly greater proportion of PD patients had model was based on both the panicogenic proper-
comorbid GAD and agoraphobia relative to NFPD ties of yohimbine (an 2-adrenoceptor antagonist
patients. At follow-up, NFPD patients, like PD that increases the firing of LC) and the elicitation of
patients, were still symptomatic and had not panic-like behaviors by electrical stimulation of the
improved or had even worsened according to LC, in humans and monkeys, respectively [7981].
scores on self-report measures [70]. Authors sug- Moreover, whereas the exposure to CO2 increased
gested that NFPD should be recognized as a vari- the LC activity [82], treatments with the 2-
ant of PD, both because of its high prevalence in adrenoceptor agonist clonidine and tricyclic panico-
medical settings and its poor prognosis. lytics (imipramine, desipramine) reduced panic
In turn, the lack of increase in the secretion of attacks [83, 84]. Further studies corroborated the
stress hormones differentiates clinical panic not CO2 activation of LC neurons both in vitro [85] and
only from fear [72] but also from stress [73] and in vivo [86]. The noradrenergic hypothesis of PD
simple and social phobias [74, 75]. Even more was also supported by studies suggesting that the
remarkably, recent studies reported that fear- panicolytic effect of imipramine is due to the desen-
unresponsive Urbach-Wiethe disease patients with sitization of -2 adrenoceptors [87, 88]. Moreover,
extensive bilateral calcifications of the amygdala Gorman et al. [67] proposed that the anticipatory
develop panic both spontaneously [76] and in anxiety of having a panic attack might be due to the
response to a tidal volume inhalation of 35 % CO2 kindling of parahippocampal gyrus brought about
[77]. These data add compelling evidence that panic by an increased activity of LC. The latter argument
is not fear nor does it require the participation of the was supported by evidence showing that treatments
amygdala. By contrast, recent human studies gave with traditional benzodiazepines block anticipatory
strong support to the participation of ventromedial anxiety while are ineffective in panic attacks [89],
hypothalamus (VMH) in panic attacks [78]. that LAC-induced panics are heralded by the activa-
Despite the above evidence, preclinical mod- tion of parahippocampal gyrus [90, 91] and that
els very often overlooked the idiosyncratic fea- anxiety appears to be mediated by the septo-hippo-
tures of clinical panic, being validated mostly campal circuit (for review, see Gray and
pharmacologically [50, 51, 60, 68]. McNaughton [92]). Gorman et al. [67] further spec-
Because the participation of PAG in non- ulated that phobic avoidance (agoraphobia) is the
respiratory type of panic attacks was extensively outcome of the noradrenergic sensitization of learn-
reviewed elsewhere [51, 52], the following sec- ing mechanisms of prefrontal cortex. The latter
tions emphasize the role of the PAG and dorso- argument was supported by data showing that PD
medial hypothalamus (DMH) in the mediation of patients that present avoidance-oriented coping
respiratory-type and LAC-evoked panic attacks. strategies and fear of anxiety-related symptoms are
This approach does not exclude the eventual par- more sensitive to anxiogenic effects of yohimbine
ticipation of PAG-projecting neurons of the VMH [93]. The panic circuits of prefrontal cortex and hip-
in non-respiratory panic [78]. Neither does it pocampus would in turn be the basis of the cogni-
exclude the participation of the amygdala, hippo- tive therapy of PD. Gormans model is illustrated in
campus and prefrontal cortex in comorbid antici- Fig. 2.1a and b.
patory anxiety and agoraphobia. The noradrenergic theory of PD was severely
criticized by researchers arguing that the LC
mediates arousal rather than panic (pros and
2.3 Early Neuroanatomical cons reviewed in [63, 67]). Most notably, electri-
Models of Panic Disorder cal stimulations of the LC of humans that pro-
duced four- to ninefold increases in plasma
The first neuroanatomical model of panic was pro- levels of norepinephrine (NE) metabolites failed
posed by Gorman et al. [67]. This model states that in eliciting any feeling of fear, anxiety, or dis-
panic attacks are the outcome of the increased activ- comfort [94, 95].
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 15

Fig. 2.1 Gormans early (a, b) and revised (c) models of result from the noradrenergic sensitization (kindling) of
panic disorder. (a) Hypothesized mechanisms of panic parahippocampal formation. Reciprocal connections of
attack onset. Gorman and collaborators emphasize the these circuits with prefrontal cortex would in turn be the
crucial role of both the chemosensitive areas of the basis of phobic avoidance (from [67], with permission);
medulla and the noradrenergic ascending projections of (c) In the revised version, the core structure is the amyg-
locus coeruleus (LC); (b) Hypothesized pathway of panic dala (from [63], with permission)
disorder anticipatory anxiety. Anticipatory anxiety would

At approximately the same time, the amygdala the brainstem activations and associated autonomic
emerged as a crucial structure in brain processing of responses as simply an epiphenomenon of the
conditioned fear [9698]. As a result, Gorman activation of the amygdala. Gormans panic
revised his theory shifting the focus from LC to the amygdala model was but profoundly discredited
amygdala and its efferent connections to brainstem by recent studies showing that fear-unresponsive
nuclei, including the LC itself [63] (Fig. 2.1). Urbach-Wiethe disease patients devoid of the
Accordingly, Gormans revised model considered amygdala develop panic attacks both spontaneously
16 L.C. Schenberg

[76] and in response to 35 % CO2 [77]. Remarkably, sion by the impairment of 5-HT transmission at
as well, besides being observed in the three rare behavioral resilience system. Specifically, the
patients with Urbach-Wiethe disease, the panic DGH states that depression is the outcome of a
attacks of patients were more intense than the only glucocorticoid-induced downregulation of 5HT1A
panic attack recorded in 16 healthy controls. hippocampal receptors whereby the stressful
Accordingly, the Feinstein et al. [77] suggested that events are uncoupled from daily life routines.
panic is mediated at the brainstem in spite of the The DGH was largely supported by animal data
established role of the amygdala in fear and anxiety on both the pharmacology of DPAG-evoked panic-
of both humans and animals. like responses [51, 60] and the differential func-
In the meanwhile, researchers become gradu- tions of 5-HT ascending systems and receptors
ally aware that electrical and chemical stimula- [107, 108]. The DGH was also supported by posi-
tions of periaqueductal grey matter (PAG) produce tron emission tomography (PET) studies of the
aversive emotions in humans [99101] and defen- brain activations in volunteers exposed to a virtual
sive responses in animals [102104] that bear a predator which was otherwise able to inflict real
striking resemblance to panic attacks. Moreover, shocks to the subjects finger [62]. In particular,
the SSRIs had begun to be widely prescribed not the latter study showed that as the predator
only for the treatment of depression, but also of approaches the prey, brain activations shift from
anxiety and panic. Accordingly, in the late 1980s prefrontal cortex/basolateral amygdala circuit to
Graeff first proposed that the PAG mediates panic central amygdala/periaqueductal grey circuit that
attacks [105]. In a following study, Graeff further are the presumptive mediators of anxiety and
suggested that 5-HT had opposite actions in PAG panic, respectively. The participation of 5-HT2
(panicolytic) and amygdala (anxiogenic) [106]. As receptors in anxiety was in turn endorsed by the
a consequence, in the early 1990s Deakin and rather specific localization of these receptors in
Graeff [50] put forward a daring hypothesis ventral areas of the hippocampus [109] that receive
according to which the brain ascending serotoner- dense inputs from basolateral amygdala (BLA)
gic systems evolved as a mechanism primarily [110]. Despite the several mechanistic proposi-
concerned with adaptive responses to aversive tions that remain to be confirmed, the DGH pro-
events. Briefly, the Deakin/Graeff Hypothesis vided a unified framework for the SSRIs efficacy
(DGH) suggested that whereas the GAD is pro- of in multiple anxiety and affective disorders.
duced by the overactivity of 5-HT excitatory pro- Plenty of evidence suggests, on the other
jections from nucleus raphe dorsalis (NRD) to hand, that panic is modulated by multiple trans-
regions of the hypothalamus, amygdala and pre- mitters, including not only 5-HT, but also gamma-
frontal cortex (PFC) that process distal/potential aminobutyric acid (GABA), norepinephrine
threats, the PD is the outcome of the malfunction- (NE), cholecystokinin (CCK), and opioids,
ing of 5-HT inhibitory projections from NRD to among other candidates. Moreover, the putative
dorsal regions of PAG (DPAG) that process defen- panicolytic action of high-potency benzodiaze-
sive responses to innate fear, proximal threat and pines (alprazolam and clonazepam) is compel-
asphyxia (Fig. 2.2). As well, they proposed that ling evidence of the prominent role of GABA-A/
whereas the NRD efferents to the amygdala facili- benzodiazepine (GABA/BZD) transmission. The
tate avoidance from threat through the activation involvement of the GABA-A/benzodiazepine
of 5-HT2 receptors, NRD efferents to the DPAG receptor was also supported by data showing that
restrain flight/fight responses through the activa- the intravenous injection of the benzodiazepine
tion of both 5-HT1A and 5-HT2 receptors (presum- antagonist flumazenil precipitates panic attacks
ably, in output neurons and GABA interneurons, in PD patients but not in healthy subjects [111].
respectively). As a consequence, Deakin and Rather than the antagonism of a long-sought
Graeff [50] postulated that anxiety inhibits panic. endogenous BZD, Nutt et al. [111] suggested that
As well, the DGH stated that the stress-induced PD patients express a GABA/BZD receptor in
increase in glucocorticoid secretion lead to depres- which the flumazenil acts as an inverse agonist.
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 17

Fig. 2.2 The Deakin-Graeff Hypothesis. Top panel: In the behaviors (STR), respectively. Bottom panel: Deakin and
early 1990s, Deakin and Graeff [50] suggested that Graeff [50] also proposed that 5-HT efferents from the
whereas the generalized anxiety disorder is produced by median raphe nucleus (MRN) to the hippocampus are the
the overactivity of 5-HT (black arrows) excitatory projec- substrate of a behavior resilience system that uncouples
tions of dorsal raphe nucleus (DRN) to areas of prefrontal stressful events from daily life routine. Stress-increased
cortex (PFC) and basolateral amygdala (BLA) that process glucocorticoid plasma level downregulation of 5-HT1A
distal threat, panic attacks would be the result of a dysfunc- receptors of the hippocampus would result in learned help-
tion of 5-HT inhibitory projections to dorsal regions of lessness and depression. Other abbreviations: BLA basolat-
periaqueductal grey matter (DPAG) that process responses eral amygdala, BNST bed nucleus of stria terminalis, CeA
to proximal threat, innate fear, or hypoxia. They also pro- central amygdala, CRH corticotrophin releasing hormone,
posed that conflict anxiety is the outcome of the simultane- ENT nucleus entopeduncularis (internal pallidum); HYP
ous activation of 5-HT and dopamine projections (yellow hypothalamus, LHb lateral habenula; PVN paraventricular
arrows) from DRN and ventral tegmental area (VTA) tar- nucleus of hypothalamus, S septum (from Schenberg
geting the striatum that mediate avoidance and approach [259], with permission)
18 L.C. Schenberg

However, while the flumazenil is reportedly an hypoxia (in males) and hypercapnia (in females)
agonist in some variants of GABA/BZD recep- [37, 132, 138]. Because the HPA axis is hyperac-
tors, there is no evidence of its action as an tive in neonatally-isolated adult rats, Dumont
inverse agonist [112, 113]. Conversely, however, et al. [138] suggested that early-life stress pro-
other studies reported that flumazenil neither pre- gramming of HPA axis predisposes the subject to
cipitates panic attacks, nor facilitates panics the later development of PD. The latter studies
induced by LAC [114, 115] (but see reply of are setting the stage for an eventual bridging of
Potokar et al. [116]). The high-potency benzodi- SFA, SAH and DGH.
azepines could also compensate a pathological
reduction in GABA/BZD receptor binding in PD
patients as suggested by clinical [117119] and 2.4 The Suffocation False Alarm
preclinical studies [65, 66]. Lastly, Preter and Theory: Facts and Questions
Kleins [27] hypothesis that panics are the due to
the malfunctioning of the opioid buffering of Kleins SFA theory was heavily based on both the
both the suffocation and the separation alarm sys- respiratory symptomatology of PD and the pani-
tems was supported by LAC elicitation of hyper- cogenic effects of respiratory metabolites CO2
ventilation in healthy volunteers pretreated with and LAC [24, 25, 27]. The theory was also mark-
naloxone [120]. Although the subjects of the lat- edly influenced by the paradoxical finding that
ter study did not panic, the opioid participation in LAC infusions produce hyperventilation [13]
panic attacks was also supported by preclinical against a background of metabolic alkalosis pre-
studies showing that SSRIs panicolytic effects sumptively due to LAC conversion to bicarbon-
may be due to an interaction of 5-HT and opioid ate [139]. As a result, the metabolic alkalosis is
transmission at DPAG [121, 122]. further aggravated by the respiratory alkalosis.
Whatever the transmitter involved, it is note- Besides showing that LAC-induced hyperventila-
worthy that fear from suffocation was barely tion is mediated by mechanisms distinct from
mentioned in DGH. As well, the role of early-life those of eupnea, the latter data suggested that
stress was not addressed in spite of the acknowl- LAC-induced panic attacks are secondary to a
edged importance of childhood environment in central hypercarbia brought about by bicarbonate
the late development of maladaptive behaviors breakdown to CO2 and water [140]. Bicarbonate
[3, 11, 123125]. As a matter of fact, preclinical infusion proved this incorrect since only patients
evidence showed that neonatally-isolated adult who panicked showed significant hyperventila-
rats show augmented responses of hypothalamus- tion, as indicated by a precussor fall in arterial
pituitary-adrenal (HPA) axis to stressful stimuli partial pressure of CO2 (PaCO2). Conversely,
[126132]. In contrast, adult rats handled as neo- non-panickers showed increased PaCO2 in spite
nates present augmented resilience to stress, as of hyperventilating [140]. These data suggested
shown by the reduced neuroendocrine responses that panic attacks cannot be ascribed to a central
and increased expression glucocorticoid recep- hypercarbia secondary to the peripheral increased
tors of the hippocampus that mediate the feed- level of bicarbonate. The eventual role of central
back inhibition of HPA axis [133]. Notably, as hypercapnia in LAC-induced panics was never-
well, it appears that decreases [134136] or theless discredited by the unanticipated finding
increases [133] in behavioral resilience are that d-LAC produces panic, hyperventilation and
accompanied by parallel changes in both 5-HT1A respiratory alkalosis in spite of not being metabo-
receptor expression in the hippocampus and lized [141] (note, however, that studies showed
5-HT turnover in frontal cortex and hippocam- that mammals are able to metabolize up to 75 %
pus, respectively. Recent data also showed that d-LAC 5 times slower than they do for l-LAC
neonatally-isolated adult rats present facilitations [142]). Taken together, the above findings sug-
of both panic responses to electrical stimulation gested a crucial role of alkalosis, the common
of the PAG [137] and respiratory responses to effect of d- and l-LAC infusions, bicarbonate
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 19

infusion and hypoxia. To rule out the likely effect For instance, it is a well known fact that tissue
of alkalosis, Gorman et al. [13] carried out a hypoxia by inhalation of carbon monoxide (CO)
study in which the panic patients were asked to (the silent killer) is devoid of both feelings of
hyperventilate a 5 % CO2 mixture. Unexpectedly, asphyxia and desire to escape.
however, patients panicked more during hyper- Epidemiological studies suggest, on the other
carbia than in room air in spite of the lack of alka- hand, that panic is influenced by both genes and
losis in the former condition. Accordingly, in a environment. Although there is a fairly estab-
preliminary presentation of SFA theory, Klein lished consensus that early-life stress [150, 151]
[25] suggested that the infusion of racemic LAC and CSA [10, 24, 27, 36, 37, 42, 152] predisposes
produces panic by two synergistic mechanisms, the subject to PD, existing evidence is neverthe-
d-LAC producing a pseudo-hypoxia and l-LAC, less mixed, either corroborating [153] or disprov-
both pseudo-hypoxia and central hypercapnia. ing [41] the SAH.
Because the blood-brain barrier of humans is per- Epidemiological studies also showed that pan-
meable to both LAC isomers [143, 144], Klein ics are highly comorbid with depressive disorders
added that the exogenously administered LAC and twice more frequent in women [154]. That
might mimic the fast increase of brain LAC dur- the higher incidence of panic in women is due to
ing asphyxia, triggering a panic attack [24, 25]. sex hormones is shown by the reduced frequency
These and other evidence led Klein [24, 25] to of panic attacks either before puberty or after
propose that clinical panic is due to the misfiring menopause. Panic attacks are also more frequent
of an integrated suffocation monitor that detects and severe during the late luteal (premenstrual)
both the increases in PaCO2 and decreases in phase (the sex-dependent features of panic were
arterial partial pressure of O2 (PaO2), and changes reviewed by Lovick [155]). Yet, there are few
in LAC and pH blood level as well. Klein [24] animal studies addressing the comorbidity of
also suggested that the suffocation alarm system panic with other psychiatric disorders.
assess the level of muscular exertion while evalu-
ating the LAC blood levels. Indeed, running not
only fails in precipitating panic attacks [145] but 2.5 Face Validity of Dorsal
also aborts panic [146], presumably, by provid- Periaqueductal Grey Matter
ing countervailing information to the suffocation Mediation of Panic Attacks
alarm system [24]. The existence of an integrated in Humans
suffocation monitor was further supported by
studies showing that PD patients are also hyper- The pioneering studies of Nashold et al. [99, 156]
sensitive to hypoxia [147]. provided the best description of feelings and sen-
However, while suffocation is most often sations produced by electrical stimulations of
understood as the prevention of air from entering human PAG. These studies were carried out in 12
the lungs, asphyxia is generally defined as the subjects without a history of psychiatric disorders
local or systemic deficiency of O2 and excess of who were chronically implanted with 34 elec-
CO2 in living tissues. Therefore, it remains unclear trodes into the dorsal midbrain. Data showed that
whether panic attacks are triggered by hypoxia or electrical stimulations of the PAG produce anxi-
hypercapnia alone, or require both conditions of ety, panic, terror and feelings of imminent death,
asphyxia (integrated suffocation monitor). along with marked visceral responses that are a
Indeed, even though the high-altitude illness lacks faithful reproduction of the core symptoms of
panic symptoms [148], high-altitude travelers panic attacks. In particular, Nashold et al. [156]
eventually report limited-symptom panic attacks reported that electrical stimulations in or near
during sleeping when hypoxia is further aggra- the lateral edge of the central grey resulted in
vated by hypercapnia [149]. Additionally, it is strong reactions in most patients. Feelings of fear
unclear whether suffocation and asphyxia neces- and death were often expressed. Autonomic acti-
sarily elicits hunger for air and desire to flee. vation such as the contralateral piloerection and
20 L.C. Schenberg

sweating, increase in the pulse and respiratory intrinsic connections suggest, on the other hand,
rate, blushing of the entire face and neck,were that columns are not independent [164]. Indeed,
also noted. At higher intensities, a subjected existing evidence suggests that rat defensive
reported that fear was so unpleasant that she was behaviors are controlled by a concerted though
not willing to tolerate repeated stimulations differentiated activity of DMPAG, DLPAG and
[99]. Even the most peculiar neurological symp- LPAG generally conflated as the dorsal PAG
toms of panic, such as tremor and chest pain, were (DPAG) [162, 165]. On the other hand, it is uncer-
produced by the PAG stimulations. These obser- tain whether the PAG columns show further spe-
vations were in line with prior studies reporting cializations along their length. For instance, the
that patients stimulated in the medial sites of the caudal regions of DPAG appear to present
mesencephalic tectum complained of a choking increased sensitiveness to agonists of N-methyl-D-
feeling referred into the chest (in Spiegel and aspartic acid receptor [166]. Moreover, whereas
Wycis, 1962, as cited by Nashold et al. [99] which the rostral LPAG (LPAGr) does most connections
is a cardinal symptom of panic attacks [53]. As with forebrain [167], the caudal LPAG does it
confirmed by X-ray analyses, sites effective in with the medulla [162, 168]. Additionally, the
producing these symptoms were found into the VLPAG harbor multiple transmitter nuclei in spite
PAG but not in the nearby tegmentum 5 mm of its shorter length. It is worth noting that the
beyond the aqueduct [99]. Ever since the panic- PAG is also parcelled out according to a cell den-
like effects of PAG stimulations were confirmed sity-increasing gradient in juxtaqueductal, medial
by several groups [100, 101, 157]. In particular, and peripheral regions [162].
Amano et al. [100] reported that during a surgery The traditional subdivision of the PAG was
for pain relief, a patient stimulated in the PAG revised based on the immunocytochemistry for
uttered somebody is now chasing me, I am trying acetylcholinesterase (AChE), NADPH diaphorase
to escape from him. In agreement with the latter (NADPHd), tyrosine hydroxylase (TH) and 5-HT,
observation, the severity of LAC-induced panic is among other markers [169]. Besides corroborat-
robustly correlated with the desire to flee [53]. ing the NADPHd specific staining of DLPAG
The isomorphism of the PAG-evoked responses [170], Ruiz-Torner et al. [169] showed that the
and human panic attacks, either spontaneous or LPAG differs from the VLPAG in that the former
induced by LAC, is shown in Table 2.2. is markedly stained for AChE. Additionally, they
showed that the medial sectors of LPAG and
VLPAG makes up a single neurochemical entity
2.6 The Periaqueductal Grey herein termed the central PAG (CePAG). As well,
Matter of the Midbrain: Ruiz-Torner et al. [169] proposed that the medial
A Likely Substrate VLPAG displaces the LPAG at caudalmost levels
of Respiratory-Type Panic of PAG (8 mm from bregma). As a result,
Attacks whereas the caudal regions of medial VLPAG
occupy the entire expanse of former LPAG, the
In humans, the sites which stimulation provoked caudalmost region of the LPAG is now in a dorso-
panic reactions were localized within 5 mm from lateral region neighbour to the remnants of
the aqueduct [99, 101] into the dorsal half of the DLPAG. Evidence below discussed suggests that
PAG [156]. However, the PAG is far from homo- the CePAG and, perhaps, LPAGr, are crucial
geneous. Traditionally, this structure is regarded mediators of respiratory-type panic attacks. As a
as four functionally specialised columns extend- matter of fact, the CePAG is the only PAG region
ing for varying distances along the aqueduct, targeted by afferents from nucleus tractus solitar-
namely, the dorsomedial (DMPAG), dorsolateral ius (NTS) [171, 172].
(DLPAG), lateral (LPAG) and ventrolateral It is also noteworthy that although the cau-
(VLPAG) columns [158163]. The abundant dalmost sectors of VLPAG are believed to play
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 21

Table 2.2 Isomorphism of panic attacks in humans and stimulation of dorsal periaqueductal gray matter of both
humans and rats
Stimulation of dorsal Stimulation of dorsal
Spontaneous or lactate-induced periaqueductal gray periaqueductal gray
panic attacks in humans matter in humans matter in rats
Feelings/behaviour Severe anxiety, panic, terror Fearful, frightful, terrible Strong aversion
Intense discomfort Intense discomfort
Desire to flee Beg to switch-off Flight behaviour
Block while walking n.a. Freezing
Difficulty in doing job n.a. Freezing
Choking Choking
Fear of dying Scare to death
Difficulty in speaking Speech arrest
Fear of going crazy n.a.
Fear of losing control n.a.
Feeling confused n.a.
Dizziness/lightheadness n.a.
Difficulty in concentration n.a.
Faintness n.a.
Autonomic Dyspnoea Dyspnoea, apnoea Dyspnoea
responses Tachypnoea Tachypnoea Tachypnoea
Hyperventilation Hyperventilation Hyperventilation
Difficulty in deep breathing Sighs, deep breaths Deep breaths
Palpitation Palpitation Tachycardia
Hypertension n.a. Hypertension
Sweating Sweating contralateral n.a.
Wide-open eyes Exophthalmus
Bladder voiding urge Micturition, defecation
Blush in face
Piloerection contralateral n.o.
Endocrine No prolactin n.a. No prolactin
responses No corticotrophin n.a. No corticotrophin
No cortisol n.a. No corticosterone
Neurological responses/ Sensation of tremor Sensation of vibration
paresthesias Chest pain Chest and heart pain
Numbness Numbness
Chills or hot flushes Burn/cold sensations
n.a. Face pain
Brain areas activated Dorsal periaqueductal gray, deep Dorsal periaqueductal Dorsal periaqueductal
layers of superior colliculus, gray and adjacent tectum gray, deep layers of
amygdala (05 mm lateral to superior colliculus
(PET) aqueduct)
Abbreviations: n.a. not available, n.o. not observed (see text for references)

a crucial role in conditioned freezing to a both rats [180] and humans [101]. As dis-
stimulus previously paired to a shock [96, 97, cussed in following sections, plenty of evi-
173], the low-magnitude stimulation of dence suggests that panic attacks may be the
VLPAG and/or NRD appear to inhibit the result from the VLPAG deficient inhibition of
DPAG [174179], being even rewarding in PAG output neurons.
22 L.C. Schenberg

2.7 Neuroanatomical Basis hypoxia that might alert him to impending dan-
of Respiratory-Type Panic ger [188], Moosavi et al. [189] showed that
Attacks hypercapnia and hypoxia are equally potent in
producing hunger for air in healthy subjects
Breathing is controlled by a widespread network matched by ventilatory drive (i.e., at fixed respi-
involving structures ranging from caudal medulla ratory rate, tidal volume and CO2 blood level).
to anterior hypothalamus, limbic system and cortex Similarly, Beck et al. [147] showed that PD
[181]. In caudal medulla, the nucleus tractus soli- patients have increased sensitivity to both hyper-
tarius (NTS) is the recipient of glossopharyngeal capnia and hypoxia relative to healthy controls.
afferents conveying blood gas signals from periph- These data are consonant with an integrated suf-
eral chemoreceptors [182]. The respiratory drive is focation alarm system endowed with the capabil-
also critically dependent on the chemosensitive ity to translate major respiratory inputs (H+/CO2,
neurons of the ventral surface of the medulla which PaO2) and, possibly, LAC blood levels, into panic
monitor pH and CO2 tissue levels [183186]. The attacks [24, 27, 42].
NTS second-order neurons project both to ventral Because dyspnea (uncomfortable and/or irreg-
respiratory nuclei of the medulla that control ular breathing) is the leading symptom of clinical
phrenic nerve activity [181] and to suprapontine panic [53], the SFA theory predicts that panic
nuclei that control behavioral and endocrine attack effectors might also be physiologically acti-
responses, including the central amygdala (CeA), vated by degrees of hypoxia and/or hypercapnia
the paraventricular nucleus of the hypothalamus that give rise to manifest feelings of either breath-
(PVN) [187] and the CePAG [171, 172]. Projections lessness or hunger for air in both patients and vol-
to the hypothalamus, limbic structures and cortex unteers. In mechanically ventilated humans,
also appear to be involved in thermoregulatory, hunger for air is provoked at CO2 alveolar partial
emotional, and volitional aspects of respiration, pressure (PACO2) of 43 mmHg, i.e., only 4 mmHg
respectively [181]. Peripheral and central chemore- above the normal CO2 end-tidal partial pressure
ceptors act in concert to maintain the arterial PaO2 (PETCO2). In turn, the PETCO2 of 50 mmHg pro-
and PaCO2 at appropriate levels (i.e., 95 mmHg duces unbearable hunger for air [190].
and 40 mmHg, respectively). Although the human low tolerance to increases
in PACO2 endorses the key role of CO2 in the elic-
itation of feelings of suffocation and, probably,
2.7.1 Psychophysics of Asphyxia clinical panic, Banzett et al. [190] showed that the
subjective symptoms of dyspnea do not correlate
Although the blood/brain gases are continuously with ventilatory responses to CO2. Moreover,
monitored by both peripheral and central chemo- whereas the hypoxia (PETO2 of 6075 mmHg)
receptors, panic attacks occur in room-air condi- reduced CO2 sensitivity and tolerance by only
tions, being very rare in healthy people. 2 mmHg, hyperventilation produced robust reduc-
Accordingly, the SFA theory presumes that the tions in hunger for air at any PETCO2, increasing
suffocation alarm system and chemoreceptors CO2 tolerance threshold by 5 mmHg while
may operate separately under specific circum- decreasing CO2 sensitivity by 50 % [190]. In par-
stances. The eventual uncoupling of these sys- ticular, mechanically-evoked hyperventilation
tems might give way to panic attacks either reduced hunger for air even when patients hyper-
spontaneous or induced by chemicals such as ventilate in isocapnic conditions. Studies carried
LAC, CO2, CCK, -carbolines and others. These out with subjects either curarised or quadriplegic
considerations raise the question about the condi- showed, on the other hand, that pulmonary mech-
tions whereby the suffocation alarm system is anoreceptors play a minor role in respiration-
uncoupled from chemoreceptors. induced decrease of hunger for air [191, 192].
Although the physiology teaches us that man These latter data support the central mediation of
is not endowed with any sensory perception of the breathing-induced reduction in hunger for air.
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 23

This mechanism could be the basis of the tradi- curve of hyperventilation and hunger for air
tional practice of treating panic attacks by having [195]. However, CCHS patients who exercised
the patients rebreathing into a brown paper bag. heavily and increased respiratory activity as
Although this technique was formerly explained much as controls reported some sensation akin
as a compensation of panic-evoked respiratory to shortness of breath. Therefore, Spengler et al.
alkalosis, other studies showed that the eventual [195] concluded that this study failed to dis-
benefits are secondary to patients distraction [13] prove the corollary discharge hypothesis. An
and/or expectation [193]. alternative interpretation suggests, however, that
Although the dissociation of hunger for air and both hunger for air to hypercapnia and short-
ventilatory responses to CO2 [190] suggests the ness of breath to exercise depend on the activa-
relative independence of underlying mechanisms, tion of a CO2-sensitive suffocation alarm system.
the cognitive (cortical) hypotheses state that feel- In the same vein, CCHS patients reported no res-
ings of hunger for air arise from the forebrain pro- piratory discomfort during either the CO2 inhala-
cessing of corollary discharges (or efferent tion or the maximum breath hold which was
copies) of respiratory motor neuron activity [191, significantly longer than that of controls [196].
194]. If so, the intensity of hunger for air should Taken together, these studies make unlikely the
parallel the respiratory responses to both hypoxia corollary discharge hypothesis.
and hypercapnia. Because sustained hypoxia pro- Although Banzett et al. [191, 197] proposed
duces a sharp ventilatory increase followed by a that feelings of hunger for air are processed at
slow decay, Moosavi et al. [194] tested the corol- primary interoceptive posterior insular cortex
lary discharge hypothesis by comparing the time (PIC) [198, 199], recent studies in rats showed
course of hunger for air during isocapnic hypoxia that neither the escape nor the cardiovascular
in either freely-breathing or mechanically-venti- responses to severe hypoxia (8 % O2) were atten-
lated healthy subjects. As predicted, during sus- uated following chemical inactivations of PIC
tained hypoxia, there was a sharp increase in (i.e., the dorsal bank of perirhinal cortex) [200].
hunger for air followed by a progressive decline, Casanova et al. [200] showed in addition that
mirroring the biphasic pattern of respiratory whereas the 8 % hypoxia did not change c-fos
response. Otherwise, the parallel time course of immunoreactivity in PIC, it produced robust
respiration and dyspnea could be explained by the labeling in both NTS and DPAG.
parallel processing of chemoreceptor inputs at Although Klein [24] suggested that the suffoca-
respiratory centers of the medulla and suffocation tion alarm system might be activated by toxic
alarm circuits elsewhere in the brain, respectively. gases as well, exposures to CO neither activate the
As well, the corollary discharge hypothesis peripheral chemoreceptors nor produce the panic
was tested in patients with congenital central responses. Peripheral chemoreceptors are actually
hypoventilation syndrome (CCHS) [195]. inhibited by low concentrations of CO [201].
Spengler et al. [195] took advantage that CCHS Although the lack of defensive responses to CO
patients hyperventilate during exercise while (the silent killer) supports the crucial role of
lacking CO2-evoked respiratory (hyperventila- peripheral chemoreceptors in behavioral responses
tion) and subjective (hunger for air) responses. to both hypoxia and hypercapnia, existing evi-
The corollary discharge hypothesis predicts dence suggests that PD patients present normal
that CCHS patients would experience normal thresholds and sensitivities of central and periph-
levels of shortness of breath during constant- eral chemoreceptors [202]. Accordingly, the pre-
load exhausting cycling. Conversely, however, sumptive abnormality of the suffocation alarm
CCHS patients showed reduced breathlessness system appears to reside in supramedullary struc-
relative to controls. Thus, whereas the hyperven- tures that do not participate in the respiratory
tilation of CCHS patients was reduced in only reflexes of eupnea. The brain defence structures
37 % relative to controls, the hunger for air was deployed along the brainstem axis are thus in a
reduced in 75 % as shown by the area under the strategic position to mediate a panic attack.
24 L.C. Schenberg

2.7.2 Neuroimaging Hunger for Air activation of temporopolar cortex due to changes in
in Humans extracranial blood flow reported in CCK-induced
panic attacks [207209]. However, whereas the
Despite the low spatial resolution of PET studies insula/claustrum were regularly activated in LAC-
of hypercapnia, these studies provide valuable and CCK-induced panic attacks, the temporopolar
information regarding the major structures of the cortex and parahippocampal gyrus were also acti-
suffocation alarm system of humans. To increase vated during anticipatory anxiety in both volunteers
spatial resolution, PET studies are often per- and panickers [210]. Accordingly, the activation of
formed with co-registration of magnetic reso- the temporopolar cortex is most probably due to
nance imaging (MRI) of suprapontine structures anticipatory anxiety and not panic itself.
of healthy volunteers that presented definite feel- Recently, Goossens et al. [211] examined the
ings of suffocation, breathlessness or hunger for CO2-evoked brain activations in panic patients,
air during the inhalation of CO2 [203205]. In healthy volunteers and experienced divers show-
particular, Brannan et al. [204] compared the ing decreasing sensitivity to CO2. Patients
PET scans in hypercapnia (8 % CO2/92 % O2) and exposed 2 min to hyperoxic hypercapnia (7 %
hyperoxia (9 % N2/91 % O2). Compared to hyper- CO2/ 93 % O2) showed increased brainstem acti-
oxia, CO2 produced the specific activation of a vations relative to both controls and divers in
vast array of cortical or subcortical structures in spite of showing the same PETCO2 and respiratory
one or both sides of the brain. In particular, CO2 rate (changes in tidal volume were not reported).
produced significant increases in the regional Data also showed that although the PIC activa-
cerebral blood flow (rCBF) in the pons, PAG, tions did not differ significantly, AIC activations
midbrain tegmentum, hypothalamus, amygdala, correlated with subjective feelings of dyspnea
sublenticular region, hippocampus (BA28) and (breathing discomfort), being more intense in
parahippocampal areas (BA20/36), subgenual patients. Authors also suggested that the lack of
anterior cingulate cortex (sgACC, BA24), fusi- differences in PIC was due to the exposure of all
form gyrus (BA19/37), middle temporal gyrus groups to the same gas mixture. The differential
(BA9), anterior insular cortex (AIC), pulvinar, activation of AIC in PD patients relative to the
putamen/caudatum, and several sites in the mid- other groups adds further evidence of the key role
line and lateral cerebellum. In addition, the acti- of this structure in conscious awareness of the
vated areas possibly involved the LC, the physiological state our body [198, 199]. Of note,
parabrachial area (PBA), and NRD. Conversely, the employment of a hyperoxic mixture makes
there were significant deactivations in the dorsal unlikely the contribution of peripheral chemore-
anterior cingulated cortex (dACC, BA24), poste- ceptors to CO2 activations of above areas.
rior cingulate cortex (PCC, BA23/31) and lateral Although Goossens et al. [211] did not find
prefrontal cortex (PFC, BA9/46). significant activations in either the PAG or the LC
Structures activated during both LAC- and of subjects exposed to 2-min 7 % hypercapnia,
CCK-induced panic attacks were but a small subset Teppema et al. [86] reported marked activations
of the network associated with breathlessness and in VLPAG of rats exposed to 2-h 15 % hypercap-
hunger for air in hypercapnia [205, 206]. Most nota- nia. Because VLPAG activations were not attenu-
bly, the only regions activated in both hypercapnia ated by hyperoxia (60 % O2/15 % CO2), the
and LAC-induced panic attacks were the sgACC, peripheral chemoreceptors seems not involved as
the insula, the hypothalamus, the midline cerebel- well. The lack of activation of VLPAG of humans
lum, the deep layers of superior colliculus (DLSC) exposed to hypercapnia [211] should be ascribed
and, possibly, DPAG and LC. Remarkably, the tem- to either the short exposure to CO2 (2 min) or the
poropolar cortex was activated by both the LAC inherent difficulty of brainstem scans.
infusion [206] and the CO2 inhalation [205]. The Remarkably, however, a pontine region compat-
employment of high-resolution MRI coupled with ible with nucleus raphe magnus (NRM) was mark-
PET in the latter study makes unlikely the mistaken edly activated in both rats and humans [86, 211].
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 25

These findings are consonant with a recent study be a reliable control of PET studies in humans.
that employed engineered chemogenetic silencing Unfortunately, however, whereas the PET human
allele to target the expression of a DREADD studies focused mostly on midbrain and prosen-
(designer receptor exclusively activated by cephalon, c-Fos animal studies focused on pontine
designer drug) inhibitory receptor in mouse and medullary regions [86, 217220]. Yet, some
embryo rhombomeres that give rise to 5-HT spe- studies examined c-Fos expression throughout the
cific nuclei of the brainstem [212]. Results showed midbrain and diencephalon of both rats and cats
that the full development of the ventilatory response [221223]. A second difficulty concerns the dif-
to hypercapnia in conscious mice requires a NRM ferent protocols of PET and c-Fos studies in
specific cell lineage that is sensitive to small humans and animals, respectively. In particular,
changes in brain PaCO2/pH. Brust et al. [212] while the PET imaging studies employed short
showed in addition that although these cells con- exposures (20 min) to moderate degrees of hyper-
tribute to CO2 chemosensitivity, they do not partici- capnia (8 %), c-Fos immunochemical studies
pate in respiratory motor responses. It is employed long exposures (hours) to moderate
nevertheless noteworthy that stimulations of NRM hypercapnia (15 %) or varying degrees of hypoxia
of rats failed in producing panic-like behavioral (511 %) [86, 221, 222]. However, Johnson et al.
and cardiorespiratory responses [213, 214]. Much [223] examined the c-fos expression throughout
the opposite, chemical stimulations of both NRM the brain of rats exposed to the fast raise of CO2
and nucleus raphe obscurus (NRO) produced long- blood levels (up to 20 % CO2 in 5 min).
lasting inhibitions of cardiovascular (NRM, NRO) Furthermore, Teppema et al. [86] evaluated the
and respiratory (NRO) responses elicited by elec- contribution of peripheral chemoreceptors in c-fos
trical stimulation of DLPAG of the anesthetized rat expression to hypercapnia (15 % CO2) by adding
[214]. Interestingly, as well, whereas the VLPAG surplus oxygen (60 % O2) in gas mixture (hyper-
receives most 5-HT afferents from NRM, the oxic hypercapnia).
DPAG does it from NRD, NRO, nucleus raphe In any event, data showed that prolonged
pontis (NRPo), and nucleus raphe medianus exposures to moderate degrees of hypoxia (8 %
(NRMn) [215]. The CO2 activations of NRM could O2) produced significant increases in fos-like
in turn modulate the DPAG-projecting neurons of immunoreactivity (FLI) of medullary nuclei that
VLPAG as discussed in following sessions. play a crucial role in eupnea, including both the
commissural and caudal NTS, ventral medullary
surface (VMS) and parapyramidal area (PPy)
2.7.3 c-Fos Immunohistochemistry [86, 222] (Table 2.3, Fig. 2.3). Teppema et al.
Studies Reveal a Limited [86] reported additional labeling in both A1 and
Number of Structures C1 cathecolaminergic groups. Double labeling
Activated by Hypoxia or for c-fos and tyrosine hydroxylase was mostly
Hypercapnia found in the rostral and caudal ventrolateral retic-
ular nuclei, but not in the lateral paragigantocel-
Because the PET signal is a function of rCBF, the lular nucleus (LPGi). The NTS was also activated
vasodilator effect of both LAC and CO2 [63, 203] by 5 % and 15 % CO2 [86, 222]. Following the
is a major drawback of these studies. In fact, the exposure to 15 % CO2, the number of c-fos
only PET study of a spontaneous panic attack (a labelled cells in the rostroventrolateral medulla
case report) revealed only deactivations in right (RVLM) was nearly 2.5 times the number of
orbitofrontal (BA11), anterior temporal (BA15), labeled cells following the exposure to 9 % O2,
anterior cingulate (BA32) and prelimbic (BA25) corroborating the higher sensitivity of RVLM
cortices [216]. Conversely, the c-Fos protein neurons to hypercapnia [224]. Addition of 60 %
expression is independent from rCBF. Accordingly, O2 to produced robust reductions in the number
c-Fos immunohistochemistry studies in animals of CO2-activated cells in NTS, RVLM and LPGi,
exposed either to hypoxia or to hypercapnia might but not in VLPAG, LC and PBA [86] (Table 2.3).

Table 2.3 Metanalysis of brain c-fos protein expression of rats exposed to hypoxia, hypercapnia or tissue hypoxia
Hyperoxia effects
in response to 15 % Tissue
Brain level Hypercapnia Hypoxia CO2 Hypoxia
Berquin Johnson Berquin et al. Teppema et al. Bodineau
et al. [222] Teppema et al. [86] et al. [223] [222] [86] Teppema et al. [86] et al. [235]
Exposure 3h 2h 5 min 3h 2h 2h 5 min
[gs] 5% 8% 10 % 15 % 20 % 11 % 9% 15 % CO2 + 60 % O2 1 % CO
Medulla cNTSa 110 (240) (120) 1050 560 620 44 389
iNTS (389) (344) 1482 1278 37 415
rNTS 300
DMX 275
AP 74
GiA 1067 633 (433)
LPGi 550 550 1556 142 31 212
cVLM 480 (60) 585
iVLM 520 1120 620 61
rVLM 620 2200 650 828 60 292
RPa 192 167 218
PPy 300 400 197
CnF 87
A5 300 800 550 550
LC 114 767 1233 1433 (41) 137
LPBA 148 2600 3013 3180 (208) 191 1400 149
L.C. Schenberg
Midbrain LDTg 107
NRD 457 206 66
DLPAG (97) 300 116
VLPAG 169 142 115 279 150 (+19 %) 111
SUM 157
PH 179 150 129
PMD 100
DMH 180 47 120 85
PeF 31 152
PVN 207 65
SON (460) (300)
VHZ 335 250 186
Original data were transformed to percent changes relative to air breathing rats of each study. Values between parentheses are changes that although noticeable did not reach
statistical significance. Empty cells represent data either non-significant or not available. Note that the 2 h-exposure to 8 % and 10 % CO2 produced marked c-fos increases in
LPBA in spite of the non-significant changes in NTS
Abbreviations: c caudal, i intermediate, r rostral, A5 A5 noradrenergic group, AP area postrema, CnF cuneiform nucleus, NTS nucleus tractus solitarius, VLM ventrolateral
medulla, DLPAG DMPAG, VLPAG dorsolateral, dorsomedial and ventrolateral periaqueductal grey matter, DMH dorsomedial hypothalamus, DMX dorsal motor nucleus of
vagus, NRD nucleus raphe dorsalis, GiA gigantocellular reticular area, LC locus coeruleus, LDTg laterodorsal tegmental nucleus, LPBA lateral parabrachial area, LPGi lateral
paragigantocellular nucleus, PeF perifornical hypothalamus, PH posterior hypothalamus, PMD premammillary dorsal nucleus, PPy parapyramidal area, PVN paraventricular
nucleus of the hypothalamus, RPa raphe pallidus, RTN retrotrapezoid nucleus, SON supraoptic nucleus, SUM supramammillar nucleus, VHZ ventral hypothalamic zone
A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder
28 L.C. Schenberg

The latter data suggest that CO2 activations of breathing rats. Because these activations were
VLPAG, LC and, most surprisingly, PBA do not not attenuated by hyperoxia, they must be due to
depend on peripheral chemoreceptors. Similarly, central inputs. Of note, LC fos-labeled neurons
the carotid denervation that suppressed the of rats exposed to hypercapnia were also double-
hypoxia activation of DPAG had no effects in stained for tyrosine hydroxylase [86]. These find-
hypoxia activation of VLPAG [225]. Prolonged ings agree with in vitro experiments showing that
exposures to hypercapnia also labeled a superfi- LC neurons are highly sensitive to CO2 [85].
cial column of cells of retrotrapezoid nucleus in Conversely, the evidence of LC activation by
the ventral surface of the medulla (0100 m hypoxia is contradictory, the activations being
depth) [86]. These neurons may correspond either manifest [82, 218] or hardly observed [86,
either to the paraolivary cell group described by 222]. Moreover, Elam et al. [82] reported that
Berquin et al. [222] or to the VMS sensors of H+/ whereas the LC responses to hypoxia were sup-
CO2 described by Loeschcke and Koepchen pressed by peripheral denervation, those to
[183]. Indeed, Loeschcke [184] noted that pH hypercapnia were unchanged. The latter data
changes in the VMS of cats produce maximal support the central mediation of the hypercapnia
drive of ventilation in two areas, one medial to activation of LC.
vagal root and the other medial to the hypoglos- Although the above studies suggest that the
sal root but lateral to the pyramid. Surprisingly, suffocation alarm system lies somewhere
however, neither the pre-Btzinger complex between the NTS and LC or PAG, Berquin et al.
(pBC) nor the ventro lateral medulla (VLM), [222] found significant FLI increases rostrally in
the areas currently proposed to play a critical role DMH, posterior hypothalamus (PH), and along a
in the neurogenesis of respiratory rhythm [226], ventral hypothalamic zone (VHZ) extending
were activated by 15 % hypercapnia [86]. In the from the mammillary nuclei to the retrochias-
dorsolateral pons, exposures to either hypercap- matic area. Even though not significant, hypoxia
nia or hypoxia activated both the external divi- and hypercapnia produced 184 % and 460 %
sion of the lateral parabrachial area (eLPBA) and increases in FLI of PVN and supraoptic nucleus
the Klliker-Fuse nucleus (KF) [86, 222]. In con- (SON), respectively. The PVN activation is most
trast, hypoxia did not affect c-Fos expression of likely involved in HPA axis response to hyper-
medullary areas that do not participate in respira- capnia in both rats [227] and humans [228230].
tory and autonomic functions. Labeling of DMH is particularly important since
Remarkably, midbrain structures not belong- this nucleus has been proposed to mediate the
ing to the established respiratory network also LAC vulnerability of panic patients [66, 231]
showed significant increases in FLI (Fig. 2.3). In (see Sect. 2.9).
particular, the DLPAG, VLPAG, LC and nucleus In turn, the PH was activated by both hypoxia
subcoeruleus were intensely labelled by moder- and hypercapnia [222]. Remarkably, as well, the
ate degrees of either hypoxia or hypercapnia [86, PH harbors PAG-projecting neurons that are
222]. In contrast, the cuneiform nucleus (CnF) intrinsically sensitive to both hypoxia and hyper-
was inconsistently labeled in spite of being the capnia [232]. Accordingly, Ryan and Waldrop
main target of DLPAG (the CnF showed moder- [232] suggested that the PAG and PBA are the
ate increases in FLI to both 10 % CO2 and 15 % key relay structures connecting the PH with med-
CO2/60 % O2, but no FLI increase to 15 % CO2) ullary nuclei involved in the ventilatory response
[86] (Table 2.3, Fig. 2.3). to both hypoxia and hypercapnia.
Teppema et al. [86] showed that VLPAG and Because burrowing species (Israeli mole rats)
PBA were markedly activated by the prolonged can tolerate extreme degrees of hypoxia (7.2 % O2)
exposure to both 10 % and 15 % CO2, being even and hypercapnia (6.1 % CO2) in their natural habi-
more marked than the caudal NTS and RVLM tat [233], it is noteworthy that brain activations
(Fig. 2.3). In fact, the PBA showed the highest were quite similar in both rats [222] and cats [221,
increase in c-fos expression as compared to air- 234]. In fact, the only difference between these
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 29

Fig. 2.3 Brainstem areas activated in hypercapnia, tions of VLPAG and PB is suggested by the lack of differ-
hypoxia and hypercapnic hypoxia. Columns represent the ence between hypercapnia (15 % CO2) and hyperoxic
mean number of c-fos labeled cells (SD) per 60-m hypercapnia (15 % CO2 plus 60 % O2). Other abbrevia-
hemisections of the braintem of rats subjected to 2-h tions: A5 noradrenergic cell group in dorsal pons, CVLM
exposures to (a) room air; (b, c, d) 8 %, 10 % and 15 % caudal ventral lateral medulla, CUN cuneiform nucleus,
CO2; (e) 15 % CO2 plus 60 % O2; and (f) 9 % O2. Although DR nucleus raphe dorsalis, lVLM intermediate ventrolat-
these data are not represented as density of labeled cells, it eral medulla, LC locus coeruleus Symbols indicate signif-
is noteworthy that the number of c-fos labeled cells of icant differences (*) from room air (all graphs); (#) from
hypercapnic rats in ventrolateral periaqueductal grey hyperoxic hypercapnia (left and right upper graphs) or
(VLPAG) and parabrachial area (PB) was higher than that 9 % hypoxia (bottom left graph); ($) from 9 % hypoxia
of nucleus of tractus solitarius (NTS) and rostroventral (upper right and bottom left graphs) (composite of Figs. 2,
lateral medulla (RVLM). The central origin of CO2 activa- 4 and 8 of Teppema et al. [86], with permission)

species was the lack of labelling in the retrotrape- caudal VLM) were the only medullary nuclei
zoid nucleus of hypoxic rats. The latter observa- showing FLI significant increases (Table 2.3).
tion is nevertheless consonant with the lack of There were also significant increases in c-fos
response of VMS neurons to hypoxia [184]. expression of most hypothalamic nuclei associ-
More recently, Johnson et al. [223] examined ated to stress and/or defensive behaviors, includ-
the c-fos protein expression of selected brain ing the PVN, the DMH and the premammillary
regions of forebrain and brainstem of rats exposed dorsal nucleus (PMD), but not the VMH or the
to a sharp increase in CO2 (up to 20 % CO2 in PH. Surprisingly, as well, although the FLI was
5 min). This protocol is similar to the two-tidal markedly increased in DMPAG, DLPAG and
volume inhalation of 35 % CO2 that elicits panic VLPAG, FLI was unchanged in LPAG and PBA
attacks in both patients and controls. Interestingly regions that appear to play a prominent role in
enough, the RVLM and LPGi (but not the NTS or suffocation alarm. Most importantly, however,
30 L.C. Schenberg

exposures to 20 % CO2 did not produce escape in hypercapnia. With the possible exception of
spite of the marked activation of both DMPAG LPBA, all these areas possess neurons intrinsi-
and DLPAG. The lack of escape responses sug- cally sensitive to changes in H+/CO2 and/or O2
gests that PAG activations were due to the depo- [85, 232, 236238]. Accordingly, the high sensi-
larization of inhibitory interneurons rather than tivity of PD patients to CO2 may be due to the
of output neurons. Alternatively, CO2 activations abnormal functioning of these nuclei. Conversely,
of DPAG output neurons could have been coun- the activations of the cortical and midline cere-
teracted by concomitant activations of VLPAG. bellar areas regularly observed during both
hypercapnia and panic attacks of humans were
not observed in c-Fos animal studies.
2.7.4 c-Fos Studies in Rats Exposed Additionally, Coates et al. [237] showed that
to Carbon Monoxide Supports phrenic nerve activity is markedly increased by
the Key Role of Dorsal microinjections of acetazolamide (an inhibitor of
Periaqueductal Gray carbonic anhydrase) in multiple brainstem sites
in Respiratory-Type Panic of servo-ventilated rats and cats that were both
Attacks vagotomised and glomectomised. Responsive
sites were found within 800 m of the VMS and
Rats subjected to a 5-min exposure to 1 % CO close to NTS and LC. These authors proposed
showed widespread activations of respiratory and that central chemoreceptors are distributed in
autonomic nuclei [235] (Table 2.3). These activa- several locations within 1.5 mm from outer and
tions occurred in spite of the peripheral chemore- inner surfaces of brainstem, which includes the
ceptor inhibition by low concentrations of CO PAG. It is noteworthing, however, that the pH of
[201]. In particular, exposures to 1 % CO produced microinjected sites was compatible with a PETCO2
significant increases in c-fos expression in NTS, of approximately 36 mmHg, which is much
area postrema (AP), dorsal motor nucleus of the lower than the threshold for elicitation of hunger
vagus (DMX), VLM, PPy, LPGi, LPBA and NRPa. for air in humans.
More rostrally, the CO activated the supramamil- In turn, extracellular and whole-cell patch-
lary nucleus, the DMH and PH. On increasing the clamp recordings of brain slice preparations
severity of tissue hypoxia, FLI was also observed showed that neurons of PH are sensitive to both
in the retrotrapezoid nucleus, ventral tegmental hypoxia and, less markedly, hypercapnia [236].
area, arcuate nucleus and PVN. Most notably, how- In particular, extracellular recordings showed
ever, whereas the CO produced robust increases in that hypoxia stimulated over 80 % of tested neu-
c-fos expression of NRD, LC and VLPAG, it did rons in a dose-dependent manner. Furthermore,
not activate the DPAG. The latter result is conso- over 80 % of the cells excited by hypoxia retained
nant with the lack of defensive responses during the response during synaptic blockade. In con-
intoxication with the silent killer. trast, hypercapnia increased the firing frequency
of only 22 % of the neurons. Similarly, whole-
cell patch-clamp recordings showed that whereas
2.8 Evidence of a Suffocation the hypoxia depolarised 76 % of neurons of PH,
Alarm System hypercapnia depolarised and/or increased the dis-
Within the Periaqueductal charge of only 35 % of tested neurons. These
Grey Matter of Midbrain results suggest that the caudal hypothalamus has
separate populations of neurons sensitive to
2.8.1 Brain Detectors of Hypoxia hypoxia and hypercapnia. Importantly, as well,
and Hypercapnia Ryan and Waldrop [232] showed that the PH har-
bors PAG-projecting neurons which are sensitive
Data from both PET and c-Fos studies showed to both hypoxia (53 %) and hypercapnia (27 %).
that the PH, the PAG, the LPBA, and, less clearly, Remarkably, as well, Kramer et al. [238]
the LC, were the only brain nuclei activated by showed that 74 % (39 of 53) of in vitro neurons of
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 31

PAG responded to hypoxia, 92 % of which with thereby suggesting an arousal to environmental

an increase in firing rate. Moreover, hypoxia- cues. The latter responses may be due to CO2
sensitive neurons of DLPAG/LPAG regions activations of in vitro [85] and in vivo neurons of
showed firing rate increases greater than those of LC of rats either intact [86, 222] or denervated
VLPAG. By contrast, only 21 % (7/33) of PAG [82]. Data of Schimitel et al. [246] are also in
neurons responded to hypercapnia. PAG neuron agreement with previous observations showing
responses to hypoxia were retained in spite of the that CO2 concentrations as high as 13 % failed to
blockade of synaptic transmission in low-Ca2+/ evoke defensive responses in rats [86]. Despite
high-Mg2+ medium. Taken together, these data the significant activation of DPAG, not even the
show that the PAG harbors neurons intrinsically sudden exposure to 20 % CO2 provoked manifest
sensitive to hypoxia. defensive behaviors [223].
On the other hand, existing evidence suggests Unexpectedly, as well, exposures to both 8 %
that carotid body chemoreceptor type-I cells and 13 % CO2 produced significant attenuations
sense both hypoxia and hypercapnia through of DPAG-evoked freezing (immobility plus
tandem acid-sensitive potassium channels exophthalmus) and flight (trotting and galloping)
(TASK-1 and TASK-3 channels) that are inhib- behaviors (Fig. 2.4) [246]. Evidence from c-fos
ited by acidic pH and hypoxia and activated by immunohistochemistry studies showed, on the
alkalosis [239241]. Prabhakar [201] showed in other hand, that 2-h exposures to low concentra-
addition that these channels might be inhibited tions of CO2 activate the NRD and, markedly, a
by the intracellular gaseous messengers CO and discrete region of caudal VLPAG atop the lat-
hydrogen sulfide (H2S). TASK genes are also erodorsal tegmental nucleus (LDTg) [86, 222].
expressed in brain chemosensitive areas of hind- The lack of effects of hyperoxia (60 % O2) on
brain, including the RVLM, raphe nuclei, and LC CO2 activations of VLPAG makes unlikely the
[242245]. It remains to be elucidated whether involvement of peripheral chemoreceptors [86].
these channels sense pH/brain gas levels in panic- On the other hand, chemical stimulations of the
related areas sensitive to both hypoxia and hyper- same region (i.e., VLPAG atop the LDTg) of
capnia, including the PH and the PAG. freely-moving rats produced only hyporeactive
immobility or quiescent behavior [293]. Taken
together, data suggest that CO2 inhibition of
2.8.2 Carbon Dioxide Inhibits DPAG-evoked panic-like responses is mediated
DPA-Evoked Panic-Like by the VLPAG.
Responses Vianna et al. [247] showed, on the other hand,
that the defensive behaviors elicited by electrical
Recently, Schimitel et al. [246] examined the stimulation of DPAG are unchanged by electro-
behavioral effects of both the hypercapnia (8 % lytic lesions of VLPAG. Although these results
and 13 % CO2) and the selective stimulation of make unlikely the VLPAG tonic inhibition of
carotid chemoreceptors with potassium cyanide DPAG, plenty of evidence supports the VLPAG
(KCN, 10160 g, i.v.), administered either alone phasic inhibition of the latter structure [174
or combined, in rats either stimulated or lesioned 179]. Notably, as well, chemical stimulations of
in the DPAG. medullary raphe produced prolonged attenua-
Behavioral data showed that although the tions of cardiovascular responses to electrical
exposure to 13 % CO2 alone produced manifest stimulations of DPAG [214]. Because the latter
increases in respiration, rats only explored the effect is presumptively mediated by NRM/NRO
open-field slowly or remained quiet with eyes efferents to DPAG-projecting neurons of VLPAG
opened and muscles relaxed, as shown by the [215], the CO2 inhibition of DPAG-evoked
lowering of trunk, head and tail. Exposures to behaviors [246] could be mediated by 5-HT neu-
13 % CO2 also reduced grooming while increas- rons of NRD properly and/or NRD lateral wing
ing exophthalmus (fully-opened bulged eyes), (NRDlw) within the VLPAG. Although the
32 L.C. Schenberg

Fig. 2.4 Carbon dioxide effects on the thresholds of air, +P < 0.05, significantly different from DPAG stimula-
defensive responses induced by electrical stimulation of tions in 8 % CO2 enriched-air sessions (Bonferronis 5 %
DPAG. Columns represent the population unbiased esti- criterion of likelihood ratio chi-square tests for location of
mates of median threshold intensities (I50 SE). *P < 0.05, parallel-fitted threshold curves) (from Schimitel et al.
significantly different from DPAG stimulations in room [246], with permission)

DPAG is targeted by 5-HT afferents from NRD, interneurons could inhibit the LPAG output neu-
NRO, NRPo and NRMn [215] (but see Jansen rons [248] presumptively involved in respiratory-
et al. [164]), prolonged exposures to CO2 type panic attacks.
increased c-fos expression of only the NRPa and,
more markedly, NRD [86] (Table 2.3). In con-
trast, whereas the sudden exposure to 20 % CO2 2.8.3 Peripheral Injections
did not activate the raphe nuclei, they produce of Potassium Cyanide Elicits
significant activations of ventral medullary areas, Respiratory-Type Panic
PVN, perifornical hypothalamus (PeF), DMH, Responses
PMD and caudal PAG, including DMPAG,
DLPAG and CePAG (medial VLPAG) [223]. Behavioral effects of injections of KCN were
Because the exposures to 20 % CO2 did not pro- much the opposite of those of CO2 [246]. Indeed,
duce escape, the PAG activations were most intravenous injections of KCN alone induced
probably due to the depolarization of inhibitory dose-dependent defensive behaviors (Fig. 2.5)
interneurons. If so, the CO2 activation of GABA similar to those observed with stepwise-increasing
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 33

Fig. 2.5 Doseresponse

curves of KCN-evoked
behaviors. Curves are the
best-fitted logistic functions of
the accumulated frequencies of
threshold responses in function
of the logarithm of the KCN
dose (i.v.). Abscissa in
logarithmic scale.
Abbreviations: r responders; n
total number of rats (from
Schimitel et al. [246], with
permission). Abbreviations:
EXO exophthalmus, IMO
immobility, DEF defecation,
MIC micturition, TRT trotting,
GLP galloping, JMP jumping

Fig. 2.6 Effects of intravenous injections of 20 g KCN of PAG. *P < 0.05, significantly different from saline ses-
or vehicle (0.9 % NaCl) on the median thresholds (I50 SE) sions. Details as in Fig. 2.4 (from Schimitel et al. [246],
of defensive behaviors induced by electrical stimulation with permission)
34 L.C. Schenberg

electrical and chemical stimulations of the DPAG gested that the PAG harbors a hypoxia-sensitive
[165, 249]. Specifically, KCN increasing doses suffocation alarm system which activation in
evoked exophthalmus, immobility, micturition, humans might both precipitate a spontaneous
defecation, trotting and galloping. However, panic attack and render the subject sensitive to
whereas the KCN elicited jumping in one rat only, CO2. Indeed, recent studies showed that KCN-
it produced defecation and micturition at rather evoked escape is attenuated by clinically-effective
low doses as compared to chemical and electrical acute and chronic treatments with the putative
stimulations of DPAG in which these responses panicolytics clonazepam (0.010.3 mg kg1, i.p.)
showed the highest thresholds. and fluoxetine (14 mg kg1 day1 along 21 days,
Schimitel et al. [246] further showed that slow i.p.), respectively [254] (Fig. 2.9). The latter
infusions of a subliminal dose of KCN (20 g/30 s) observations were further extended by the dem-
produced significant facilitations of trotting and onstration that panic-like responses to severe
galloping responses to electrical stimulations of environmental hypoxia (7 % O2) are also attenu-
DPAG (Fig. 2.6). Because the DPAG does not ated by both the acute injections of alprazolam
project to the spinal cord [162], the escape (14 mg kg1, i.p.) and the chronic injections of
responses are most likely mediated by DPAG fluoxetine (515 mg kg1 day1, 21 days, i.p.), and
efferents to CnF neurons targeting spinally- by intra-periaqueductal injections of 5-HT antag-
projecting cells of gigantocellular and magnocel- onists as well [255].
lular reticular nuclei [250]. Conversely, the In the same vein, escape responses to severe
inconsistent labeling of CnF nucleus by both environmental hypoxia (8 % O2) were accompa-
hypoxia and hypercapnia [86] suggests that this nied by significant activations of intermediate lev-
nucleus is not involved in detection of asphyxia. els (7.08 mm from bregma) of DLPAG and
Facilitations of DPAG-evoked immobility and LPAG (but not of DMPAG and VLPAG) [200].
exophthalmus were less conspicuous. Conversely, however, prior studies found that
Most importantly, whereas the KCN-evoked hypercapnia, hypoxia and KCN activate mostly
defensive behaviors were unchanged by aortic the caudal regions (7.8 to 8.0 mm from bregma)
denervation, they were virtually suppressed by of DLPAG, LPAG and VLPAG [222, 253]. The
carotid denervation [251, 252] and discrete lesions precise localization of hypoxia-responsive sites of
of DPAG [246] (Fig. 2.7). That KCN-evoked defen- PAG was further complicated by data showing that
sive behaviors are mediated by peripheral chemore- whereas the 2-h exposures to 9 % hypoxia failed in
ceptors was also shown by the lack of effect of activating any column of the PAG [86], 3-h expo-
KCN microinjections (12200 nmol/100 nL) into sures to 11 % hypoxia activated both the DLPAG
the DPAG (C.J.T. Mller, unpublished results). The and the VLPAG [222]. Besides, whereas Berquin
PAG mediation of KCN-evoked behaviors is in turn et al. [222] reported that hypoxia produced moder-
supported by tract-tracing and c-fos immunohisto- ate activations of DMH (120 %), PH (150 %) and
chemical studies of NTS projections to the CePAG LPBA (191 %), Teppema et al. [86] reported
[171, 172] and KCN activations of DPAG [253]. marked activations of A5 noradrenergic neurons
Taken together, these data support the mediation of (550 %) and LPBA neurons only (1440 %).
KCN-evoked behavior by CePAG-projecting neu- Remaining activations were found in medullary
rons of NTS. areas unlikely to mediate the defensive behaviours
The opposite effects of KCN and CO2 raised (NTS, DMX, RVLM and VMS).
the question whether the pre-exposure to CO2 The demonstration that in vitro neurons of PAG
could inhibit KCN-evoked defensive behaviors. are intrinsically sensitive to hypoxia [238] raises
Conversely, however, KCN-evoked behaviors the question whether the hypoxia activations of
were markedly facilitated by both 8 % and 13 % PAG resulted from local changes in PO2. Moreover,
CO2 (Fig. 2.8) [246, C.J.T. Mller, unpublished prior attempts to localize the KCN-responsive
results]. Because CO2 alone did not produce regions within the PAG were compromised by the
defensive behaviors, Schimitel et al. [246] sug- employment of repeated intravenous injections of
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 35

Fig. 2.7 Effects of electrolytic

lesions of PAG on the
defensive behaviors produced
by intravenous injections of
80 g KCN. Columns
represent the proportion (SE)
of rats presenting a given
response before (hatched) or
24 h after (black) the
application of discrete
electrolytic lesions to the PAG
(anodal current, 1 mA, 5 s) to
the PAG. *P < 0.05, **P < 0.01,
***P < 0.005, proportions
significantly different from
pre-lesion values (Pearsons
2) (from Schimitel et al.
[246], with permission). Other
details as in Fig. 2.5

Fig. 2.8 Effects of 8 % and

13 % CO2 on KCN-evoked
flight behavior. The duration
of flight behavior is the sum of
the duration of trotting,
galloping, and jumping
responses evoked by
intravenous injections of 40 g
KCN. **P < 0.005,
***P < 0.0001, significantly
different from room air;
P < 0.005, significantly
different from 8 % CO2
(Students one-tailed paired
t-tests) (from Schimitel et al.
[246], with permission)

anesthetized rats with high doses of KCN (60 from caudal LPAG and LPBA that could mediate
120 g, i.v.) [253]. Indeed, whereas the latter study the KCN-evoked peripheral signals of hypoxia.
found widespread activations of PAG, a recent Taken together, data suggest that whereas the
study from our laboratory showed that the LPAGr hypercapnia activates the caudal VLPAG, hypoxia
(6.48 mm from bregma) was the only brainstem activates predominantly the DLPAG and LPAG.
structure activated following a short-lasting escape Although the molecular mechanisms underly-
response (23 s) elicited by a single injection of a ing the suffocation alarm system are unsolved, it
low dose of KCN (40 g, i.v.) (C.J.T. Mller, is a quite remarkable finding that neurons of both
unpublished results). Although most afferents of PH and PAG solely begin to respond to hypoxia
LPAGr come from hypothalamus and limbic cor- after post-natal day 12 (PN12) (Fig. 2.10) [225].
tex [167], the LPAGr is also recipient of afferents Accordingly, these structures do not respond to
36 L.C. Schenberg

Fig. 2.10 Cell densities of DLPAG and VLPAG neurons

Fig. 2.9 Effects of clinically effective treatments with pani- expressing Fos protein of normoxic and hypoxic rats dur-
colytics on the duration of flight responses (mean SEM) pro- ing post-natal days 356 and after carotid sinus nerve
duced by an intravenous injection of KCN (80 g). Upper: denervation (CSNX). Carotid sinus denervation unmasked
Effects of the acute treatment with clonazepam. Bottom: an NTS-mediated inhibitory input that suppressed c-fos
Effects of the chronic treatment with fluoxetine. *Significantly activation in DLPAG. In contrast, denervation produced a
different from saline-treated rats (5 % Bonferronis criterion non-significant reduction of only 22 % of c-fos labeled
of Students 1-tailed t-tests). KCN potassium cyanide, s sec- cells of VLPAG. Data are means S.E.M. (n = 45 for
onds, SAL saline, SEM standard error of the mean (from each group). *Significantly larger than normoxic rats
Schimitel et al. [254], with permission) (P < 0.05) (from Horn et al. [225], with permission)

hypoxia during the stress hyporesponsive period response to hypoxia results from a NTS-mediated
(PN2-PN12) in which the HPA axis is likewise disinhibitory mechanism, VLPAG responses
quiescent [256258]. It remains to be elucidated depend predominantly (78 %) on central inputs
whether the hypoxia-sensitive neurons are also (as below discussed, the hypoxia activations of
primed by maternal deprivation. VLPAG might actually correspond to activations
In turn, Horn et al. [225] showed that whereas of ventral half of LPAG). Although the latter data
the VLPAG responses to hypoxia were unaltered contradict our previous argument that KCN-
by sinusal denervation of adult rats, DPAG evoked panic responses are partly due to NTS
responses were wholly suppressed (Fig. 2.10). inhibition of VLPAG [61, 259], they do not rule
Since PAG neurons are intrinsically sensitive to out the core hypothesis that panics result from
hypoxia [238], Horn et al. [225] proposed that VLPAG defective inhibition of DPAG output
carotid denervation unmasked an extrinsic inhib- neurons. As a corollary, tissue hypoxia by 1 %
itory input that suppressed c-fos expression in CO activated the LC, the NRD and VLPAG, but
DLPAG. Moreover, Teppema et al. [86] showed not the DLPAG or LPAG (Table 2.3) [235]. The
that VLPAG activations by hypercapnia arere latter observations are consonant with the absence
unchanged by hyperoxia (60 % O2) (Table 2.3). of defensive reactions during the intoxication
Altogether, data suggest that whereas the DLPAG with the silent killer.
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 37

Concluding, data suggest that hypoxia acti- mus and brainstem, the PAG yielded the most pro-
vates the PAG both directly (via NTS excitatory nounced respiratory responses, being undoubtedly
efferents to LPAG/CePAG) or indirectly (via NTS a respiratory center. These observations were
disinhibition of DLPAG). Conversely, hypercap- corroborated by extensive mapping of respiratory
nia aborts DPAG-evoked panic reactions owing to responses induced by electrical stimulation of the
the central activation of VLPAG inhibitory inputs brainstem of anesthetized cats. Most notably, Tan
to DPAG/CePAG (Fig. 2.11, upper). Panic [262] showed that low-intensity stimulations of
responses are nonetheless facilitated in hypercap- LPAG produced phase-locked stimulus-bound
nic hypoxia, which reproduces real-life asphyxia respiratory responses without concomitant cardio-
(Fig. 2.11, middle). Be this as it may, panic vascular responses. Subsequent studies suggested
patients appear to overreact to CO2 due to the that chemical stimulations of the PAG affect mostly
defective functioning of VLPAG and/or NRD the respiratory frequency, decreasing the duration of
inhibitory projections to DPAG (Fig. 2.11, bot- both expiration and inspiration [263, 264], an effect
tom). Of note, recent studies suggest that LAC thought to be mediated by the PBA inspiratory off-
infusions could facilitate panic via the inhibition switch of pneumotaxic center [265]. However,
of NRDlw projections to DPAG [231, 260]. periaqueductal microinjections of GABA-A recep-
tor antagonists increase both frequency and ampli-
tude of respiration (Fig. 2.13) [266].
2.8.4 On the Probable Substrates The demonstration that hypoxia facilitates the
of Kleins Three-Layer Cake sham-rage behavior of the decerebrated cat was
Model of Panic Attack the first evidence that behavioral responses to
hypoxia might be mediated at the brainstem
In a preliminary presentation of SFA theory, [267]. Reciprocally, Hilton and Joels [268]
Klein [25] suggested that the march of symp- showed that electrical stimulations of defence
toms of panic attacks appears to be a three- areas of both PAG and medial hypothalamus
layer cake. The first layer is the reaction of the facilitated the hyperventilation produced by
smothering alarm system, as it had received an intravenous injections of KCN. Much later on,
increment of CO2, by breathlessness and Franchini et al. [251, 252] showed that alerting
increased tidal volume. When the control system behaviors to KCN (30 g, i.v.) were suppressed
keeps getting signals interpreted as predictive of by both sinusal denervation and occlusion of
asphyxiation, then the panic attack, with its feel- carotid body artery. In turn, Hayward and Von
ing of suffocation and urge to flee is released, Reitzenstein [253] showed that repeated injec-
followed by the increase in respiratory fre- tions of a high dose of KCN (120 g, i.v.) activate
quency. In particular, LAC-induced panic predominantly the caudal sectors of DPAG.
attacks are characterised by early dyspnoea fol- However, less than 10 % of these neurons were
lowed by panic, desire for flight and a sustained retrogradely labelled from A5 region believed to
hyperventilation that continues several minutes mediate the pressor component of chemoreceptor
after the end of the infusion (Fig. 2.12). Although reflex [269]. Lastly, Schimitel et al. [246] pre-
this model is heavily based on LAC-induced sented compelling evidence that KCN-evoked
panics, this section examines whether the DPAG panic-like behaviors are both facilitated by
circuits might explain Kleins three-layer cake previous exposures to CO2 and suppressed by
model of panic attacks. restrict electrolytic lesions of DPAG. Schimitel
Although the respiratory rhythmogenesis is et al. [246] also showed that DPAG-evoked
maintained in pontine preparations, PAG stimula- behaviors were facilitated by the concomitant
tions of anesthetized animals produce prominent intravenous infusion of a subliminal dose of KCN
respiratory effects. Indeed, in the first study with (20 g). Altogether, these data give strong sup-
electrical stimulation of the PAG, Sachs [261] port to the PAG mediation of respiratory-type
reported that among all regions stimulated in thala- panic attacks.
38 L.C. Schenberg

Importantly, as well, Hayward and co-workers intermediate frequency (30 Hz) applied at a defi-
showed that DPAG-evoked respiratory responses nite time of the inspiratory-expiratory transition
were markedly reduced (65 % to 90 %) by (approximately at 0.4 cycle) resulted in unpre-
bilateral microinjections into the LPBA of either a dictable time recovery of the respiratory cycle
GABA-A receptor agonist or a glutamate receptor (random co-phase duration). This response is
antagonist [270, 271]. Moreover, experiments exactly as expected in dyspnea, the leading symp-
with fluorescent labeling identified the effective tom of respiratory-type panic attack [24, 27, 42,
site in the inner division of eLPBA (see Fig. 5 of 53]. Paydarfar and Eldridge [263] also showed
Hayward et al. [270]). In turn, Haibara et al. [272] that critically timed stimuli with an appropriate
showed that whereas lidocaine microinjections intensity resulted in prolonged inspiratory apneu-
into the LPBA markedly attenuated the cardiovas- sis that could last over 3 respiratory cycles of the
cular responses to KCN, lidocaine microinjec- decerebrated cat (i.e., approximately 9 s). These
tions into DLPAG or LPAG were ineffective. data showed that subtle stimulations of DPAG
Overall, these data suggest that whereas the can produce either prolonged episodes of apneu-
DPAG mediates behavioral and respiratory sis or irregular breathing which are typical of
responses to KCN, the LPBA is involved in respi- respiratory-type panic onset (Fig. 2.12).
ratory and cardiovascular components only. More recently, Subramanian et al. [278] showed
Importantly, as well, these data suggest that respi- that neuron-selective stimulations of distinct col-
ratory and cardiovascular symptoms of panic umns of the PAG of precollicularly decerebrated
attacks are mediated by PAG glutamatergic pro- unanesthetised cats produced column-specific pat-
jections to the LPBA. terns of respiration. Indeed, whereas the DMPAG
Although the DPAG-evoked cardiovascular stimulations produced slow and deep breathing
and respiratory responses have been extensively along with increased tonus of the diaphragm that
studied as visceral adjustments of fight and flight resemble the respiratory pattern during the rat con-
behaviors [160, 253, 266, 273278], the stimulus- frontation with a conspecific, DLPAG stimulations
response properties of DPAG-evoked respiratory produced hyperventilation and tachypnea consistent
responses [263] suggest that DPAG may be sub- with the respiratory responses of a full-blown flight
strate of both the early dyspnea and the late behavior. In turn, whereas the chemical stimulations
hyperpnea of respiratory-type panic attacks. To of lateral sectors of LPAG and VLPAG produced
study the stimulus-response properties of the res- respiratory responses associated with emotional
piratory oscillator, Paydarfar and Eldridge [263] vocalization (mews and/or hisses), stimulations of
examined the changes in phrenic nerve activity the medial part of LPAG (i.e., the CePAG) and of
brought about by electrical stimulations of DPAG VLPAG resulted in irregular breathing or inspiratory
of sinoaortic-denervated anesthetised cats kept at apneusis of over 8 s, which are reminiscent of panic-
constant PETCO2 (3035 mmHg). Data showed onset dyspnoea. Of note, the CePAG-evoked inspi-
that brief threshold stimulations (1-s trains) of ratory apneusis had the same duration of the apneusis
DPAG invariably reset the cycle to inspiratory or reported by Paydarfar and Eldridge [263]. In any
expiratory phase. Following cycle resetting, the event, while the respiratory responses to CePAG and
respiratory rhythm resumed with predictable VLPAG stimulations mimic the panic onset, respira-
periodicity (fixed co-phase duration). Yet, if tory responses to DLPAG stimulations are reminis-
intensity and pulse and train duration were kept cent of a full-blown panic attack.
at 250 A, 1 ms and 1 s, respectively, stimuli of
Fig. 2.11 (continued) by the CO2 activation of inhibitory
projections of the ventrolateral periaqueductal grey
(VLPAG) to the former structures. Middle During envi-
ronmental suffocation, peripheral signals of hypoxia and
Fig. 2.11 Presumptive mechanisms of elicitation of panic hypercapnia overcomes VLPAG inhibition of DPAG,
attacks by inhalation of low concentrations of carbon diox- releasing a panic reaction. Bottom In panic patients, the
ide (CO2) in panic disorder patients. Upper In healthy defective functioning of VLPAG inhibitory projections to
subjects, CO2 peripheral inputs to dorsolateral (DLPAG) and DPAG render the subject hyper-responsive to both hypoxia
lateral (LPAG) periaqueductal grey (PAG) are counteracted and low doses of CO2 (modified from Schenberg et al. [51])
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 39

Fig. 2.12 Time-course (min)

of respiratory changes in
tidal-volume during a panic
attack provoked by intravenous
infusion with 0.5 M sodium
lactate of a patient with panic
disorder and agoraphobia. The
x indicates the early episode
of dyspnea that heralds the
panic attack (courtesy of
D.F. Klein)

Fig. 2.13 Increases in mean blood pressure, average tized rats produce full-blown defensive behaviors. As
heart rate and respiration caused by the microinjection of expected, these data show that the structures that mediate
a GABA-A receptor antagonist (bicuculline, 10 nmol, panic-like behaviors are tonically inhibited by gabaergic
1 L/30 s) into the dorsal periaqueductal grey matter of a synapses (from Schenberg et al. [266], with permission)
urethane anesthetized rat. Similar injections in unanesthe-

In another study, Subramanian and Holstege converted pre-I neurons into phase-spanning
[248] examined the effects of chemical stimula- cells that discharged at inspiration/expiration
tions (20 nL of D,L-homocysteic acid) of discrete transition. Notably, as well, while the stimulation
regions of PAG on the activity of both diaphragm of the lateral part of LPAG activated pre-I neu-
and pBC pre-inspiratory (pre-I) neurons that are rons throughout the inspiration and produced
believed to play a crucial role in respiratory emotional vocalization and diaphragm relax-
rhythmogenesis. Data showed that chemical ation, stimulations of ventral part of caudal
stimulations of DLPAG produce both the increase LPAG (LPAGcv, 8.0 mm from bregma) pro-
of firing of pre-I cells and tachypnea. In turn, duced an early burst of spikes (2 s) followed by a
whereas the stimulation of VLPAG produced hyperpnea concomitant to a paradoxical sus-
expiratory apnea and inhibition of pre-I cells and tained inhibition (7 s) of pre-I neuron firing.
diaphragm, stimulations of CePAG (medial The demonstration that LPAGcv and VLPAG
part of LPAG) produced inspiratory apnea and mediate opposite responses of hyperpnea and
40 L.C. Schenberg

apnea reveals the sheer complexity of PAG. In par- by peripheral (hypoxia plus hypercapnia) and cen-
ticular, the development of hyperpnea in the tral (pH/hypercapnia) respiratory inputs, respec-
absence of pBC activity suggests the relative tively. Remarkably, as well, PAG lesions affected
independence of eupnea and full-blown respira- only the tidal volume. The lack of changes in
tory panic responses. Moreover, the paradoxical respiratory frequency suggests that lesion effects
inhibition of pre-I neurons during LPAGcv-evoked are mediated at VMS and/or pBC.
tachypnoea suggests that PAG stimulations block
respiratory mechanisms of eupnea much as they
do with baroreceptor reflex [104, 279281]. It 2.8.6 On respiratory and Non-
remains to be elucidated whether the Hering- Respiratory Panic Attacks
Breuer reflex is also inhibited during both panics
and defensive behaviors and attacks, giving way to Despite the difficulty of interpretation of lesion
a full-blown hyperventilation. In any event, these studies, data of Lopes et al. [282, 283] suggest
results raise the possibility that the dyspnea/hyper- that DMPAG/DLPAG and LPAG/VLPAG play
pnea pattern of respiratory-type panic attacks may complementary roles during resting respiratory
be mediated by VLPAG-CePAG and CePAG- activity. It should be noted, however, that
LPAGcv circuit. although these studies are consonant with the
lack of activation of VLPAG during hypoxia
[86], other studies reported that the VLPAG is
2.8.5 Steady State Functioning activated in both hypoxia [225, 253] and hyper-
of Suffocation Alarm System capnia [222].
Data from chemical stimulations of PAG [248,
During resting conditions, the suffocation alarm 278] suggest, on the other hand, that symptom
system may operate differently from its function- escalation of respiratory-type panic attacks might
ing during full-blown panic attack. In particular, be due to an early activation of VLPAG projec-
Lopes et al. [282, 283] showed that chemical tions to CePAG (dyspnea) followed by the sus-
lesions of either DMPAG/DLPAG or LPAG/VLPAG tained activation of DLPAG and LPAGcv output
produce significant reductions (21 % to 31 %) in pathways (hyperventilation). The latter mecha-
the ventilatory response to hypercapnia (7 % CO2). nisms are supported by tract-tracing of intra-
In turn, whereas lesions of DMPAG/DLPAG pro- periaqueductal connections with both anterograde
duced marked facilitations (67 %) of respiratory (leucoagglutinin of Phaseolus vulgaris, PHA-L)
responses to hypoxia (7 % O2), those of LPAG/ and retrograde (pseudorabies virus, PRV) mark-
VLPAG were ineffective. ers [164]. In particular, Jansen et al. [164] showed
Therefore, whereas the DMPAG/DLPAG that whereas the VLPAG projects almost exclu-
inhibits the respiratory responses to hypoxia, the sively to CePAG (Fig. 2.14d), the DLPAG and
LPAG/VLPAG facilitates the responses to hyper- LPAG project to the entire expanse of caudal
capnia. As predicted for a suffocation alarm sys- PAG (Fig. 2.14b and c). By contrast, neither the
tem, these mechanisms appear to reduce oxygen VLPAG, nor the LPAG, or DMPAG send signifi-
consumption upon the lack of useful air while cant projections to the DLPAG (Fig. 2.14).
increasing the responses to hypercapnia, respec- Notably, as well, data from PHA-L/PRV
tively. In particular, the latter mechanism may be double-labeling immunohistochemistry sug-
responsible for the increased basal respiratory gested that VLPAG-evoked sympathoinhibition
activity of PD patients [24]. is mediated by a relay neuron of CePAG
Because peripheral signals of hypoxia and (Fig. 2.15) [164]. The dyspnea-hyperpnea symp-
hypercapnia are conveyed by the same fibers of toms of respiratory-type panic attack might be
glossopharyngeal nerve, the opposite effects of mediated by a similar pattern of connections
DMPAG/DLPAG and LPAG/VLPAG lesions on between VMS/VLPAG and CePAG/LPAGcv.
respiratory responses to hypoxia and hypercapnia In contrast, fear-like panic attacks would be
suggest the differential modulation of these areas triggered by PFC, VMH and DLSC inputs to
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 41

Fig. 2.14 Distribution of PHA-L labeled axons into the regions caudal to the injection site. (c) Following an
PAG of rats whose injections were restricted to dorso- injection into the LPAG, axonal labeling extended
medial (a), dorsolateral (b), lateral (c) or ventrolateral mostly to the caudal PAG, but avoided the DLPAG. (d)
(d) columns of PAG. Solid dots represent PHA-L Following an injection in VLPAG, labeled axons were
labeled cell bodies. (a) Following an injection in much more concentrated in the CePAG. Additional
dmPAG, labeled axons extended throughout the PAG, abbreviations: 3 oculomotor nucleus, Dk nucleus of
involving all subnuclei. At rostral levels, the labeling Darkschewitsch, EW Edinger-Westphal nucleus, fr fas-
was more concentrated in DPAG and precommissural ciculus retroflexus, Su3 supraoculomotor periaqueduc-
nucleus (PrC). (b) Following an injections into the tal grey, Su3C supraoculomotor cap. Other labels as in
DLPAG, axons labeling extended mostly throughout the the abbreviation list (from [164]
42 L.C. Schenberg

DLPAG [52]. Therefore, although the respiratory rons intrinsically sensitive to hypoxia and/or
symptoms are more prominent in respiratory- hypercapnia. Because the diagram activations
type panics, non-respiratory panics may present (arrows) were based mostly on c-fos immunohis-
similar symptomatology owing to the overlap- tochemistry, it remains unclear whether they rep-
ping output pathways to LPBA, CnF and sympa- resent the depolarization of inhibitory or
thoexcitatory centers of the medulla. Among the excitatory neurons. Neither is it clear whether
possible differences of respiratory and non- these neurons were activated by local (pH, PCO2,
respiratory panic systems, PAG projections to PO2) or extrinsic inputs (peripheral and/or central
LPBA are presumptively more active in respira- chemoreceptors). In any event, boutons were rep-
tory panic attacks. Indeed, studies with PHAL resented either as excitatory (pathway color) or
[284] and c-fos immunohistochemistry [285] as inhibitory (black) based on immunohisto-
showed that whereas the LPAG and VLPAG proj- chemical, electrophysiological and pharmaco-
ect to all subnuclei of LPBA, DMPAG and logical studies.
DLPAG project almost exclusively to the supe- Most notably, whereas in vitro neurons of neo-
rior subnucleus of LPBA. Most notably, however, cortex and hippocampus are depressed by both
Krout et al. [284] showed that the CePAG (therein hypoxia and hypercapnia [238], those of NTS,
named aqueductal PAG or central PAG) sends VLM, VMS, LC, PAG, PH and PMD are instead
a unique projection to the inner division of excited (red arrows). In particular, while the PH
eLPBA. Remarkably, as well, Hayward et al. neurons are intrinsically sensitive to both hypoxia
[270] showed that the microinjection of gluta- and hypercapnia [85, 232, 238], PAG and LC neu-
mate antagonists into the eLPBA blocked the rons respond predominantly to hypoxia or hyper-
respiratory responses to electrical stimulations of capnia, respectively. On the other hand, although
DPAG. The eLPBA appears also involved in the the neurons of NTS, VLM and VMS are also
entrainment of respiration and afferent muscle intrinsically sensitive to hypoxia and/or hyper-
activity [286]. The latter mechanism could capnia [184, 225, 237, 288, 289], there are no
explain why exercise aborts panic attacks [146, reports that the stimulation of these nuclei pro-
287]. Indeed, Klein [24] had long proposed that duce defensive behaviors. Conversely, the PMD,
exertion aborts panic by providing countervailing PH, PAG and LC are both intrinsically sensitive to
information to the suffocation monitor. It is nev- hypoxia and/or hypercapnia and traditionally
ertheless unclear whether the PBA is a monitor or involved in arousal and/or defensive behaviors.
an effector of the suffocation alarm system. The latter structures are thus the best candidates
to translate incoming signals of asphyxia to alert-
ing and/or defensive responses.
2.8.7 Overview of Suffocation Existing evidence also suggests that respiratory
Alarm System of Mammals: responses to hypercapnia are facilitated by both the
An Evidence-Based PH [290] and the PAG [282, 283]. On the other
Hypothesis hand, data from lesion studies suggest that DPAG
(but not the VLPAG or the PH) inhibits the ventila-
The presumptive structure of the suffocation tory response to hypoxia [283, 290]. While the
alarm system of mammals is outlined in Fig. 2.16. former mechanism is already expected for a suffo-
The diagram depicts the probable pathways cation alarm system, the latter would reduce oxy-
involved respiratory (cyan) and non-respiratory gen consumption upon lack of useful air. Because
(green) panics, as well as the main inputs (white) DPAG lesions attenuated predominantly the tidal
and outputs (yellow) of these systems. The 5-HT volume [283], the DPAG appears to inhibit the
neurons (pink) are emphasized as well. In turn, pBC. It is proposed that this effect is mediated by
the arrows depict nuclei and pathways that are LPBA activation of an inhibitory interneuron
activated by peripheral and/or central signals of (Fig. 2.16a). In particular, Subramanian and
asphyxia. In particular, the red arrows depict neu- Holstege [278] reported that chemical stimulations
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 43

Fig. 2.15 Diagram of

connections modulating the
sympathetic function as
revealed by double-labeling
with anterograde
(leucoagglutinin of Phaseolus
vulgaris, PHA-L) and
retrograde (pseudorabies virus,
PRV) markers. PHA-L was
injected into the VLPAG
(Stage 1) and, 6 days later, the
PRV was injected into the
stellate ganglion (Stage 2)
(from Jansen et al. [164], with

Fig. 2.16 The suffocation alarm system. Arrows repre- and respiratory (suffocation-like) panic attacks. Black
sent structures activated by hypoxia and/or hypercapnia. boutons represent inhibitory inputs. White and yellow
Red arrows represent structures intrinsically sensitive to neurons represent modulatory inputs and effector mech-
hypoxia, hypercapnia or both conditions, as assessed in anisms, respectively. See text for explanation. G gabaer-
in vitro slices. Green and blue neurons and boutons rep- gic neuron, Op opioidergic neuron. Other labels as in
resent the core nuclei of non-respiratory (predation-like) abbreviation list
44 L.C. Schenberg

of the ventral part of LPAG produce tachypnea VLPAG. Subramanian and Holstege [248]
while inhibiting the pBC. The latter paradoxical showed in addition that whereas the chemical
effects suggest that the PAG takes over the control stimulation of the ventral part of the LPAG
of respiration during panic-like reactions. (8 mm from bregma) produced tachypnea, that
Although the PAG appears to be the fulcrum of of VLPAG led to expiratory apneusis. Moreover,
suffocation signals from both forebrain and hind- chemical stimulations of rather close regions of
brain, the CePAG is the only PAG region targeted LPAG and VLPAG elicit the opposite responses
by afferents from commissural, medial and ventro- of quiescence and escape [158, 293]. Taken
lateral subnuclei of NTS [171, 172] that are the together, data from chemical stimulations and
terminal fields of sinus nerve afferents [291] c-fos immunohistochemistry suggest that the
(Fig. 2.16b). Interestingly, as well, whereas the LPAGcv and the VLPAG harbor independent
repeated injection of high doses of KCN (>60 g) populations of neurons sensitive to peripheral
produced widespread activations of the DPAG of and central signals of hypoxia and hypercapnia,
the anesthetized rat [253], the single injection of a respectively (Fig. 2.16b and e). Data also suggest
low dose of KCN (40 g) that elicited a short- that VLPAG-CePAG and CePAG-LPAGcv might
lasting escape response activated only the rostrolat- be crucial in early dyspnoea and late tachypnea
eral (LPAGr) and caudoventrolateral (VLPAGc) of respiratory-type panic attacks, respectively. In
regions of PAG (C.J.T. Mller, unpublished this model, the CePAG appears as a comparator
results). In contrast, rats that developed escape of hypercapnia signals from VMS and hypercap-
upon severe ambient hypoxia (8 % O2) showed sig- nia plus hypoxia signals from NTS. The preva-
nificant activations in rostral sectors (7.08 mm lence of the latter signals would result in a
from bregma) of both DLPAG and LPAG, but not respiratory-type panic attack.
in DMPAG or VLPAG [200]. Importantly, as well, In turn, the hypoxia appears to activate the
the NTS was labeled by both the ambient hypoxia DLPAG both directly, via local actions (red
and KCN [200, 253, C.J.T. Mller, unpublished arrow), and indirectly, via PH excitatory projec-
results]. While the differences in PAG activations tions and NTS disinhibitory projections [225,
by KCN (LPAGr and VLPAGc) and hypoxia 232] (Fig. 2.16f and g). Although highly specula-
(DLPAG and LPAGr) may be due to hypoxia direct tive, afferents from hypothalamus are likely to
excitation of DLPAG [238], the only PAG region convey complex signals of suffocation/asphyxia,
reliably activated by hypoxia was the LPAGr. In such as stuffy or stale air, drowning, predators
turn, whereas the LPAGr makes most connections suffocation bite of nose or throat of preys, or even
with forebrain and hypothalamus, it is also targeted strangling in humans. The latter possibilities are
by afferents from both caudal LPAG and LPBA supported by data of Lopes et al. [283] showing
[167]. Remarkably, as well, the CePAG was mark- that DMPAG/DLPAG (but not LPAG/VLPAG)
edly labeled following repeated injections of higher inhibits ventilatory responses to hypoxia that
doses of KCN (see Fig. 6 of Hayward and Von could be counterproductive upon hypoxic condi-
Reizenstein [253]). Moreover, Sandkhler and tions of preys caught by predators or victims of
Herdegen [292] showed that restricted chemical avalanches, landslides or drowning.
stimulations of CePAG activate the LPAGr (see Overall, above data suggest that respiratory-
Fig. 5b, c in Sandkhler and Herdegen [292]). type panics to both environmental hypoxia and
Taken together, these data support the involvement KCN [200, 246] are mediated by NTS projec-
of both CePAG and LPAGr in respiratory-type tions either direct or indirect to CePAG-LPAGcv
panic attacks. In turn, the VLPAGc appear to medi- neurons that project in turn to the LPBA
ate the behavioral inhibitory effects of normoxic (Fig. 2.16h) and, less markedly, to CnF midbrain
hypercapnia (Fig. 2.16c). locomotor region (Fig. 2.16i). By contrast, non-
The NTS could also project, either directly or respiratory type panic attacks appear to be medi-
indirectly, to neurons of the ventral part of LPAG ated by DLPAG afferents from VMH, PMD and
(Fig. 2.16d) mistakenly regarded as part of DLSC that convey pre-processed exteroceptive
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 45

signals (olfactory, auditory or visual) of an immi- affect the rat behavior. The latter results are con-
nent threat [52]. Output neurons of DLPAG sonant with the NTS modulation of VLPAGc.
would in turn project to CnF and LC neurons that Lastly, although the lesions of caudal VLPAG
mediate escape and increased in attention to envi- block conditioned freezing [96, 173], chemical
ronmental cues, respectively (Fig. 2.16i and j). stimulations of VLPAG produce only quiescence
The apparently spontaneous clinical panic would [293], being pleasurable in both rats [180] and
thus be the result of the activation of either the humans [101]. While the latter findings make
CePAG-LPAGcv or the VMHdm-PMd-DLPAG unlikely the VLPAG mediation of panic attacks,
efferent systems by as-yet-unknown molecular they suggest that the activation of VLPAG upon
mechanism. normoxic hypercapnia might be responsible for
Krout et al. [284] showed, on the other hand, the lack of defensive responses of both rats and
that the PAG sends profuse projections to most cats exposed to CO2 concentrations as high as
nuclei of LPBA (Fig. 2.16h). Most notably, how- 20 % [86, 221, 223, 246]. The latter mechanism
ever, they showed that the CePAG is the only might also explain the attenuation of DPAG-evoked
region that projects to the inner division of defensive responses by 8 % and 13 % CO2 [246].
eLPBA. Interestingly enough, whereas the phar- As a corollary, the hypoxia-induced escape may
macological blockade of eLPBA produces robust be explained by the direct excitatory actions of
attenuations of DPAG-evoked respiratory responses hypoxia on neurons of DPAG. Similarly, the lack
[270], chemical stimulations of the CePAG pro- of activation of both DLPAG and LPAG in rats
duces inspiratory apneusis [248]. Existing evidence exposed to 1 % CO [235] explains why the vic-
suggests, on the other hand, that the inner division tims of CO intoxication die quietly in bed.
of eLPBA is specifically involved in the entrain-
ment of respiratory and locomotor patterns [286]. It
is then proposed that only CePAG activations in the 2.8.8 Insights into a Neurochemical
absence of exercise led to the dyspneic symptoms Puzzle
of panic attacks.
Neurons of in vitro slices of PAG are hardly The molecular mechanisms underlying PAG
excited by hypercapnia [238]. Conversely, the intrinsic sensitivity to hypoxia remain completely
VLPAG is activated in rats exposed to a wide unknown. Neurotransmission within the suffoca-
range of hypercarbic mixtures (520 %) [86, 223, tion alarm pathways is likewise unsolved.
225] (Table 2.3). These data suggest that VLPAG Difficulties are further aggravated by the apalling
activations by CO2 are mostly of central origin. complexity of neurotransmission within the PAG,
Remarkably, as well, hyperoxia (60 % O2) did not involving aminoacids both excitatory and inhibi-
affect VLPAG activations by hypercapnia [86] tory, NE, 5-HT, CCK, nitric oxide (NO), sub-
(Table 2.3). Because glossopharyngeal fibers stance P (SP), enkephalins (ENK), endomorphins
convey signals of both hypoxia and hypercapnia, (EDM), prepro-thyrotropin releasing hormone,
the latter data makes unlikely the VLPAG modu- orexin, galanin, somatostatin and vasoactive
lation by peripheral chemoreceptor inputs. The intestinal polypeptide, among other chemicals
VLPAG modulation by pH/CO2 central inputs is [162, 170, 172, 295302].
nevertheless supported by the profuse projections Notwithstanding, it is a quite remarkable fact
of VMS neurons to the caudal sectors of both that whereas the traditional benzodiazepines were
LPAG and VLPAG [294] (Fig. 2.16e). However, ineffective in DPAG-evoked panic responses [60],
recent results of our group showed that the low-doses (0.11.0 mg kg1) of clinically-effective
VLPAGc atop the LDTg nucleus is activated by benzodiazepine panicolytics (alprazolam, clonaze-
both 13 % CO2 and, quite unexpectedly, 40 g pam) produced significant attenuations of latter
KCN (C.J.T. Mller, unpublished results). The responses [303]. DPAG-mediated panic-like
latter author also showed that KCN microinjec- responses to chemoreceptor stimulations were
tions (20200 pg/100 mL) into the PAG did not even more sensitive than those evoked by electrical
46 L.C. Schenberg

stimulation, being markedly attenuated or virtually seems also sensitive to PO2 and/or peripheral che-
suppressed by clonazepam doses within the clini- moreceptors [86]. If so, the latter data suggest that
cal range (0.010.3 mg kg1) [254, 255]. Conversely, the NRD exerts a specific modulation of CePAG-
microinjections of GABA-A receptor antagonists LPAGcv suffocation alarm system. Moreover,
into the DPAG elicit all behavioral, cardiovascular although prior studies found that the DPAG is the
and respiratory defensive responses of the rat [266, recipient of 5-HT afferents from NRD, NRO,
304306]. In particular, Behbehani et al. [306] NRPo and NRMn [215], Jansen et al. [164]
reported that the CePAG (i.e., the medial and reported that the CePAG (i.e., the LPAG close to
medioventral parts of the PAG) is the region most the wall of the aqueduct) is the main target of
sensitive to the microinjection of the GABA-A NRD projections. Indeed, Ruiz-Torner et al. [169]
antagonist bicuculline. These data suggest that found that the CePAG is further delimited by
DPAG and CePAG are tonically inhibited by gab- 5-HT immunostaining. Not surprisingly, recent
aergic synapses as would be expected for a system studies showed that escape reactions to either
dedicated to cope with emergencies only. Indeed, hypoxia or KCN are attenuated by chronic treat-
Lovick and Stezhka [307] showed that although the ments with fluoxetine and DPAG microinjections
DPAG harbors two populations of output neurons, of both 5-HT1A and 5-HT2A receptor agonists
only 18 % and 37 % cells of each population fired [254, 255]. Overall, these data suggest that SSRIs
spontaneously, both at low-firing rate (<4 Hz). attenuation of panic attacks might be mediated by
Since the CePAG possesses the highest density of the facilitation of 5-HT inhibitory actions both
GABA-A receptors [306], it is worth examining direct and indirect of DLPAG and CePAG-
whether the CO2 inhibition of DPAG-evoked panic LPAGcv circuit, respectively (Fig. 2.16).
responses is mediated by these receptors. Remarkably, as well, the CePAG is the main
On the other hand, plenty of evidence suggests target of DMH glutamatergic neurons that are
that 5-HT exerts opposite roles in DPAG and presumptively activated by interoceptive cues
VLPAG. In particular, whereas the 5-HT is pre- [311, 312]. In turn, whereas the CePAG is fairly
dominantly excitatory in DPAG (72 % of respon- delimited by the expression of EDM [172] and
sive neurons) [308], it is virtually inhibitory in CCK [298], the SP is widely expressed across the
VLPAG (94 % of responsive neurons) [309]. full extension of DMPAG and LPAG and also
Stezhka and Lovick [177, 310] showed, on the caudal sectors of DLPAG [298]. Similarly, SP
other hand, that the dorsal NRD sends excitatory receptors (SPr) are particularly dense in cell bod-
projections to DLPAG. Brando et al. [308] also ies and dendrites of approximately 6 % of neu-
showed that the DPAG harbors both low-firing rons of broad areas of DPAG, being observed in
cells excited by 5-HT2 agonists and high-firing postsynaptic and nonsynaptic regions as well. A
cells inhibited by 5-HT1A agonists. Because the small proportion of axons (4.2 %) and axon ter-
DPAG is tonically inhibited by gabaergic syn- minals (5.3 %) also showed SPr immunoreactiv-
apses, it is tempting to speculate that high- and ity in axo-dendritic synapses predominantly
low-firing cells correspond to inhibitory interneu- asymmetric (70 %) [313]. Whole-cell patch-
rons and excitatory output neurons, respectively. clamp studies showed, on the other hand, that SP
Studies with c-fos immunocytochemistry produces dose-dependent depolarizations of
showed, on the other hand, that whereas the NRD 60 % of neurons of PAG [314], a greater propor-
activity is markedly increased (200450 %) by tion of which (70 %) was inhibited by met-
prolonged exposures (2 h) to moderate hypercap- ENK. Taken together, these data suggest that SP
nia (10 % and 15 % CO2) [86], it is unchanged by may act in a diffuse, nonsynaptic manner, modu-
both prolonged exposures (3 h) to light hypercap- lating excitatory neurotransmission both presyn-
nia (5 % CO2) [222] and short exposures (5-min) aptically and postsynaptically.
to severe hypercapnia (20 % CO2) [223] The balance between CCK, SP and opioid
(Table 2.3). Because NRD activations by CO2 transmission into the PAG appears thus to be of
were blocked by hyperoxia (60 % O2), the NRD major importance in subjects vulnerability to
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 47

PD. Most notably, while CCK is a recognized pani- 2.9 Modeling Lactate
cogen, Brodin et al. [315] showed that 1- or 7-day Vulnerability in Rats
isolations of adult rats upregulate SP expression in
DPAG. In turn, recent studies of Bassi et al. [316] Because LAC is a major respiratory metabolite
showed that 1-day isolations significantly increased under hypoxic conditions, Pitts and McClure [7]
the number and duration of 22 kHz ultrasonic seminal observations that infusions of 0.5 M LAC
vocalizations (USVs), which were reversed by precipitates panic attacks in predisposed individu-
resocialization. Conversely, 14-day isolations pro- als but not in healthy patients had a major impact
duced reductions in USVs that could not be in panic research and in the development of SFA
reversed by resocialization. The USVs were also theory [24, 27, 42]. Surprisingly, however, most
reduced by the microinjection of SP into the dPAG pre-clinical research suggests that LAC infusions
(35 pmol/0.2 L), an effect blocked by pretreat- activate hypothalamic circuits (Fig. 2.17) that
ment with the SP antagonist spantide (100 pmol/ show little overlap with pathways presumptively
0.2 L). Bassi et al. [316] concluded that 1- and involved in suffocation [65, 66, 231].
14-day isolations recruit distinct brain defensive The hypothesis of the hypothalamic mediation of
systems. Most importantly, however, these studies panic attacks was based on early observations that
suggest that SP might be implicated in separation inhibitions of gabaergic transmission at PH facilitate
anxiety and social loss predisposing effects on the freezing in experimental conflict, escape in Sidmans
development of PD. non-signaled avoidance and anxiety in the elevated
Although the neurotransmission of both plus-maze (EPM) [317322]. Conversely, injections
DLPAG and CePAG-LPAGcv is largely unsolved, of a GABA-A receptor agonist (muscimol) into the
data herein discussed suggests that the DPAG is PH released punished behavior [319]. Although the
endowed with the capability to process concomi- DMH involvement in defensive behaviors was dis-
tant signals of hypoxia and hypercapnia that make carded in earlier studies, further research identified
up real-life asphyxia. It is then proposed that suf- the effective sites in DMH properly [320] and, more
focation sensations arise when the VMS-mediated recently, in PeF [231]. As well, Shekhar [320]
hypercapnia inhibition of CePAG-LPAGcv is sur- showed that baseline anxiety in EPM is either
passed by NTS-mediated excitation of suffocation increased or decreased by blocking or enhancing the
signals of hypoxic hypercapnia. Indeed, although gabaergic transmission in DMH, respectively.
the high-altitude illness is devoid of panic symp- Shekhar [323] also showed that DMH-evoked
toms [148], Nepalese travelers eventually report defensive behaviors were blocked by subchronic
nocturnal panic attacks when hypoxia is further treatments with the clinically effective panicolytics
aggravated by hypercapnia [149]. imipramine (5 and 15 mg kg1, 7 days) and clonaz-
From a molecular point of view, spontaneous epam (5 mg kg1, 3 days).
panic attacks might be triggered by multiple Most notably, however, the DMH depletion of
mechanisms, including (1) phasic failure of 5-HT GABA by chronic microinfusions of the inhibitor
inhibition of CePAG-LPAGcv circuit, (2) CePAG- of glutamate decarboxilase l-allylglycine (L-AG)
LPAGcv activations by DMH-mediated interocep- (Fig. 2.17a) rendered rats sensitive to LAC infu-
tive signals, (3) epigenetic deficiency of EDM/ sions that precipitate panic in patients (i.e., 0.5 M,
ENK transmission in CePAG and/or LPAGcv, (4) 10 mL kg1 per 15 min) [64]. In particular,
epigenetic upregulation of CCK transmission in whereas the LAC infusions increased heart rate,
CePAG (5) epigenetic upregulation of CCK in blood pressure and anxiety in rats treated with
CePAG, (6) epigenetic upregulation of substance P L-AG, they were ineffective in controls treated
transmission in DPAG. Although the occurrence with the inactive isomer d-allylglycine. Shekhar
of panic attacks in room-air conditions makes et al. [324] also showed that LAC microinjec-
unlikely the involvement of NTS-CePAG path- tions into the organum vasculosum lamina termi-
way, panics might also be predisposed by a genet- nalis (OVLT) produce anxiety-like behaviors that
ically-determined increase in DPAG neurons were blocked by local microinjections of both
intrinsic sensitivity to hypoxia. tetrodotoxin and glutamate antagonists in OVLT
48 L.C. Schenberg

Fig. 2.17 Evidence-based presumptive pathways of lac- represent excitatory and inhibitory inputs, respectively.
tate induced panic attacks. Red arrows represent c-fos See text for explanation. G gabaergic neuron, Op opioi-
labeled structures following LAC infusions in l- dergic neuron. Other labels as in abbreviation list
allylglycine (L-AG) treated rats. White and black boutons

and DMH, respectively. In contrast, the microin- angiotensin and orexin efferents to amygdala-
jections of tetrodotoxin in the subfornical organ projecting neurons of DMH [322, 330] (Fig. 2.17).
(SFO) and medial preoptic area were without Although the microinjections of LAC in the
effects, making unlikely the involvement of the subfornical organ (SFO) and medial preoptic
angiotensinergic pathways that mediate thirst area were without effects [324], the latter finding
[325]. do not rule out the question whether the LAC-
Further studies showed that LAC-evoked anxi- induced anxiety of rats is a by-product of either
ety responses were sensitive to both panicogens thirst [331, 332] or hunger [333] states supposed
and panicolitics. For instance, whereas the LAC- to be mediated by angiotensin and orexin, respec-
induced anxiety was facilitated by the putative tively. Orexin appears also involved in behavioral
panicogen yohimbine [321, 326], it was blocked arousal responses [334, 335] that could have
by the panicolytic benzodiazepine alprazolam affected the rat performance in both SI and
[327]. Moreover, chemical kindling of the amyg- EPM. As a matter of fact, further studies reported
dala by repe ated microinjections of GABA-A the puzzling finding that L-AG treated panic-
and urocortin antagonists rendered rats sensitive prone rats showed increased anxiety to intrave-
to LAC infusions as shown by the increase of nous infusions of 0.5 M sodium chloride [336].
anxiety in social interaction test (SI) [328, 329]. Because isosmolar infusions of D-mannitol were
Next, Shekhar and collaborators showed that ineffective, the authors concluded that LAC-
whereas the LAC-induced anxiety was facilitated evoked panic is due to the activation of osmosen-
by the microinjection of angiotensin-II into the sitive pathways by increased concentration of
DMH, it was attenuated by both the blocking of sodium rather than of lactate. Conversely, how-
angiotensin-II receptor [322] and the silencing of ever, Kellner et al. [337] showed that neither PD
orexin-1 receptor gene [330]. Taken together, patients, nor healthy volunteers developed panic
these studies provided support to the argument to intravenous infusions of a 2.7 % NaCl solution
that LAC-evoked anxiety is mediated by OVLT isosmolar to 0.5 M LAC.
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 49

Another drawback of above studies is the eval- following the intravenous infusion of either LAC
uation of panic through animal models of GAD or 0.9 % NaCl. Compared to saline treated con-
(SI, EPM). Accordingly, we cannot rule out the trols, LAC infusions of panic-prone rats produced
involvement of both angiotensin and orexin in significant increases in c-fos expression in OVLT
anxieties accompanying thirst and hunger. The (171 %) but not in SFO or AP regions which are
hypothesized participation of the amygdala in also devoid of blood-brain barrier. There were
panic attacks was in turn severely compromised also significant activations of DMH (84 %), PVN
by the demonstration that fear-unresponsive (150360 %), SON (243 %), CeA (241 %), PBA
Urbach-Wiethe disease patients lacking the amyg- (382 %) and NTS (700 %) (red arrows in
dala develop panic attacks both spontaneously Fig. 2.17). Remarkably, the authors stressed that
[76] and in response to 35 % CO2 [77]. Conversely, the PBA was the only major respiratory center
pre-clinical studies in rats showed that kindling of that clearly responded to LAC infusions.
the amygdala facilitates fear-like resistance to Moreover, LAC-induced activations of PBA were
capture while inhibiting the panic-like responses significantly correlated with increases in respira-
to electrical stimulation of DPAG [338]. tory rate, but not with heart rate, blood pressure or
There are at least two more reasons that make anxiety. Although the NTS neurons were also
unlikely the DMH mediation of panic attacks markedly labeled in panic-prone rats, neurons
properly. First, although the LAC-induced panic were very often immunoreactive for both c-fos
attacks of humans are not accompanied by activa- and GAD, thereby suggesting the depolarization
tions of HPA axis, chemical stimulations of DMH of gabaergic interneurons rather than output neu-
in awake rats produced over 650 % increases in rons [341]. In the same vein, LAC infusions failed
ACTH plasma levels [339, 340]. Not surprisingly, in activating the ventral medullary areas which
the PVN is the main recipient of DMH projec- are established targets of NTS, including rostral
tions [311]. Second, data from our laboratory and caudal VLM, pBC and nucleus ambiguous.
showed that high-resolution frequency-varying Neither did the LAC activated the VMS. The only
low-intensity stimulation (<50 A, a.c.) of DMH exception was a light though significant activation
pars diffusa (DMHd) produced only exophthal- of C1 adrenergic neurons involved in both tonic
mus, whereas that of DMH pars compacta and phasic control of blood pressure.
(DMHc) produced exophthalmus, immobility, Lactate brain activations are consonant with
defecation and micturition (A.C. Alves, unpub- respiratory effects of 0.5 M and 2 M LAC infu-
lished results). The limited repertoire of DMHd sions in freely-moving Wistar rats [342]. Indeed,
was confirmed by chemical stimulations with 0.5 M LAC increased respiratory rate while
N-methyl-D-aspartic acid (NMDA). The latter decreasing the tidal volume, thereby suggesting
study showed in addition that while the electrical that LAC activate predominantly the PBA [265].
and chemical stimulations of DMHd elicited Most notably, Johnson et al. [341] showed that
robust ingestive behaviors in sated rats, uninten- LAC infusions activate the same subnuclei of
tional stimulations of dorsomedial VMH LPBA targeted by the LPAG (i.e., superior lat-
(VMHdm) elicited all defensive responses of the eral, rostral lateral, dorsal lateral, medial and lat-
rat, including trotting, galloping and jumping, at eral crescent areas) and the CePAG (inner
thresholds similar to those of DPAG. The latter division of eLPBA) (note, however, that the inner
data suggest that the eventual elicitation of escape division of eLPBA was mistakenly named ven-
during DMH stimulations may be due to the trolateral PBN, personal communication of
spreading of either current or drug to VMHdm. P.L. Johnson). These areas show partial overlap
As a matter of fact, a recent study in humans with terminal fields of the NTS efferents to dor-
reported panic attack provocation following elec- sal, central and eLPBA subnuclei of LPBA [343
trical stimulation of VMH [78]. 345]. Conversely, LAC marked activations of
Most importantly, Johnson et al. [341] exam- PBA are hardly explained by DMH occasional
ined c-fos protein expression in L-AG treated rats farthest projections to this nucleus [311].
50 L.C. Schenberg

Recent data suggest, on the other hand, that mate antagonists into the DLPAG/LPAG sup-
PBA activations may be independent from NTS presses the cardiovascular responses to chemical
afferents. In particular, Kaur et al. [346] showed stimulations of DMH [351, 352]. In turn, the
that 2-h exposures of mice to 10 % CO2 activates NRDv projects mostly to VLPAG and CePAG
VLM glutamatergic projections to both the exter- [164]. Because 5-HT actions are predominantly
nal lateral and the lateral crescent subnuclei of inhibitory in VLPAG/NRDlw [309], these data
PBA which project in turn to respiratory motor suggest that LAC behavioral effects may be due
neurons. These authors also showed that selective to the NRDv-mediated inhibition of NRDlw
delection of the gene of vesicular glutamate inhibitory inputs to CePAG-LPAGcv (Fig. 2.17).
transporter-2 blocked respiratory arousal to It is also proposed that 5-HT excites CePAG via
hypercapnia [347]. Taken together, these studies 5-HT2 receptors [309] (Fig. 2.17). The disruption
suggest that VLM glutamatergic projections to of gabaergic tone in the DMH could thus facili-
LPBA are implicated in respiratory activations to tate the respiratory-type panics both directly, via
CO2. However, it is worth remembering that facilitation of DMH glutamatergic projections to
0.5 M LAC infusions did not activate any medul- NRDd and CePAG, and indirectly, via NRDv dis-
lary respiratory structure. inhibition of CePAG-LPAGcv panic circuit
Importantly, as well, PBA neurons are (Fig. 2.17). These data give support to the argu-
depressed by opioids [348, 349]. Accordingly, it is ment that subjects vulnerable to LAC infusions
tempting to speculate that respiratory activations present deficiencies in both GABA and 5-HT
to naloxone-preceding LAC infusions in healthy inhibitory inputs to DMH and CePAG-LPAGcv
volunteers were due to blockade of opioid recep- panic circuit, respectively.
tors of PBA [120]. This mechanism could also On the other hand, it is a quite remarkable fact
explain why these infusions produced hyperventi- that 0.5 M LAC infusions neither activated the
lation but not panic [120]. Consonant with the lat- PAG [341], nor facilitated the KCN-evoked panic
ter finding, LAC infusions of panic-prone rats attacks of rats (E.A. Moraes, unpublished results),
did not increase c-fos expression in any region of or produced panics in naloxone-treated healthy
PAG [341]. The latter data are also in agreement volunteers [120]. LAC infusions also did not
with the lack of effects of 0.5 M LAC infusions in facilitate KCN-evoked panics of rats that were
KCN-evoked experimental panic presumptively either pre-treated with naloxone or neonatally-
mediated at DPAG (E.A. Moraes, unpublished isolated throughout lactation (E.A. Moraes,
results). It remains to be examined whether the unpublished results). Nonetheless, LAC infusions
KCN-evoked panic is facilitated by higher concen- produced marked activations of the LPBA and,
trations of LAC. Indeed, Olsson et al. [342] sug- mostly, of the inner division of eLPBA (see Fig.
gested that human infusions with 0.5 M LAC are S3b, d of Johnson et al. [341] which is the specific
better modeled by rat infusions with 2 M LAC. target of CePAG [284]. Remarkably, as well, the
On the other hand, double-labeling of both LPBA sends profuse efferents to primary intero-
c-fos protein and tryptophan hydroxylase showed ceptive areas of PIC [353] that project in turn to
that whereas the LAC infusions activate 5-HT the VLPAG [354, 355] (Fig. 2.17). The LPBA-
neurons of NRDlw in D-AG treated controls PIC-VLPAG circuit could thus be the basis of
(yellow arrow in Fig. 2.17), they are ineffective in panic attack cognitive theories of the catastro-
L-AG treated panic-prone rats [231, 260]. fization of bodily symptoms [2123]. Indeed, the
Although the DMH does not project significantly insula is one of few structures activated by both
to DPAG or NRDlw, it does project to CePAG LAC infusions and CO2 exposures in patients and
[310, 312] (Fig. 2.17) and, indirectly, to NRDv volunteers, respectively [205, 206].
via massive projections to PeF/lateral hypothala- Concluding, although the evidence is mixed,
mus [52]. There is also evidence that nearly 65 % pre-clinical data suggest that LAC-evoked behav-
of DMH projections to PAG are glutamatergic ioral effects might be provoked by (1) GABA
[297, 350]. Indeed, the microinjection of gluta- deficient inhibition of DMH excitatory inputs to
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 51

CePAG-LPAGv, (2) DMH/NRDv inhibition of suffocation HPA axis activation would counter-
NRDlw inhibitory inputs to CePAG-LPAGv, (3) productively increase catabolic activity beyond
NTS-mediated activation of PBA-CePAG-LPAG, the adaptive capacity of the organism.
and, (4) LAC direct activations of LPBA-PAG or Cortisol concentrations were nevertheless
LPBA-PIC-VLPAG circuits. Admittedly, the increased in the saliva of patients having severe
mechanisms underlying LAC-evoked panic panic attacks [367]. The HPA axis is also acti-
attacks remain largely obscure almost 50 years vated in human, fear-like, panics marked by pal-
from its discovery by Pitts and McClure [7]. pitations, tremor and sweating, but devoid of
suffocation symptoms. These fear-like panics can
be provoked by drugs that either produce anxiety,
2.10 Modeling Neuroendocrine such as -carboline, yohimbine, pentylenetetra-
Unresponsiveness of Panic zole and fenfluramine [368, 369] or stimulate in
Attacks vitro neurons of the PVN [368]. Therefore,
although the CCK-related peptides (CCK-4,
The concept of stress was coined by Hans Selye CCK-8S, and pentagastrin) produce panic attacks
[356, 357] to denote a complex syndrome caused and HPA activations [370372], neuroendocrine
by various noxious agents thereafter named effects of these peptides appear to be due to phar-
stressors. Selyes stress hypothesis had a macological stimulation of PVN neurons by
widespread influence on medical world and, mechanisms unrelated to panic. Indeed, both in
mostly, immunology and psychiatry. In particu- vivo and in vitro studies showed that CCK stimu-
lar, clinical and experimental studies implicate lates DMH and PVN neurons either directly or
stress in depression, panic, and post-traumatic indirectly via the activation of peripheral recep-
stress disorder [358362]. Selyes hallmark was, tors [368, 373]. Moreover, the chronic treatment
however, the doctrine of the non-specificity of the with a clinically effective panicolytic (citalo-
alarm reaction according to which stress pram) decreased the intensity of the CCK-
became almost a synonym of the HPA axis acti- induced panic attacks; however, it did not change
vation that follow the exposure of mammals to a the HPA axis response to this neuropeptide [374].
perplexing number of stressors. Further studies Lastly, CCK-induced panic attacks were blocked
showed that PRL secretion is also increased in by the 5-HT3 antagonist odasentron [375], vagot-
response to a great number of physical and psy- omy and CCK-A receptor antagonists [368, 376],
chological stressors [363365]. thereby suggesting the contribution of a peripher-
Notably, however, plenty of evidence showed ally mediated component.
that ACTH, COR and PRL are unaltered during Because humans are susceptible to both sug-
panic attacks. Indeed, neither the situationally gestion and procedural anxiety, the demonstration
provoked panic attacks of agoraphobics [16] nor that DPAG-evoked panic responses do not acti-
the experimentally provoked panics to LAC and vate the HPA axis would be invaluable evidence
CO2 [13, 15, 1719, 366] increased stress hor- of DPAG mediation of panic attacks. Indeed, a
mones significantly relative to healthy controls. preliminary study by our group showed that nei-
COR plasma levels actually decreased in ten pan- ther the ACTH nor the PRL secretion increased 5
icking subjects exposed to 7 % CO2 [366]. and 15 min after the panic-like behaviors pro-
Despite a small sample size, Cameron et al. [14] duced by 1-min electrical stimulation of the
also did not find any change in stress hormones DPAG [377]. Nonetheless, further analyses
either at peak or 10 and 60 min after nine sponta- showed that corticosterone (CORT) plasma levels
neous panic attacks in four hospitalized patients. were significantly increased (285.2 8 ng/mL)
The neuroendocrine unresponsiveness of clinical one week after the electrode implantation (unpub-
panic is both intriguing and an important clue for lished results). Accordingly, the lack of ACTH
understanding the neurobiology of PD. In partic- responses might have been due to the CORT
ular, Preter and Klein [27] suggested that upon increased level inhibition of HPA axis in rats
52 L.C. Schenberg

recently implanted with both an electrode (7 days that PAG glutamatergic efferents to the PVN
prior to testing) and an indwelling catheter (2 days arise only from the VLPAG and the yet undiffer-
prior to testing). Additionally, Lim et al. [378] entiated commissural PAG. These regions are
reported a conspicuous increase (160 %) in CORT either much ventral or much rostral to the stimu-
plasma level 30 min after a DPAG-evoked 1-min lated sites of our studies, respectively.
explosive flight bout (running with aimless direc- The above data might suggest that HPA axis
tion) in 4-week surgery-recovered rats presenting unresponsiveness in clinical panic is solely due to
reduced baseline levels of CORT (70 ng/mL). the lack of DPAG projections to PVN. Remarkably,
Such conflicting data might be explained by the however, the HPA axis is likewise unresponsive
different degrees of physical effort during the to 20-kHz artificial alarm calls that cause full-
DPAG-evoked flight behaviors of rats stimulated blown flight behaviors, tachycardia and wide-
in arenas of rather different areas (0.3 m2 versus spread activations of DPAG, DMH, BLA, CeA
1 m2). That physical effort is relevant in HPA axis and paraventricular nucleus of the thalamus
responses was already shown by Schenberg et al. (PVT) but not of PVN [381, 385]. Because the
[377] report of a moderate (83 %) though non- HPA axis is robustly activated by stimulations of
significant increase in ACTH plasma levels fol- both DMH [340] and CeA [386], the HPA axis is
lowing the exhausting effort of DPAG-evoked most likely inhibited during both the DPAG stim-
repetitive flight bouts. ulation and the 20-kHz artificial alarm call. In
Most importantly, however, a recent study from particular, Bhatnagar et al. [387] presented evi-
our group showed that stress hormones remain dence of a PVT-mediated inhibition of HPA axis
unchanged when DPAG-evoked escape is pre- by CCK afferents from VLPAG (note, however,
vented by stimulating the rat in a 20-cm diameter that the CCK neurons of PAG are predominantly
roofed cylinder (0.03 m2) with stimuli that elicited localized along the CePAG [298]).
full-blown flight behaviors in a 55-cm diameter Because the HPA axis is markedly activated in
open-field (1 m2) [379] (Fig. 2.18). Neither did the both stressful and non-stressful conditions, a
ACTH increase when physical exertion was statis- recent multinational authoritative consensus
tically adjusted to the average effort of non-stimu- review on stress proposed that the use of terms
lated controls as measured by LAC plasma levels. stress and stressor should be restricted to
In contrast, foot-shocks of the same duration as the conditions and stimuli where predictability and
intracranial stimulus produced marked neuroendo- controllability are at stake [388]. Yet, panic
crine responses that were not correlated with mus- attacks are neither controllable nor predictable
cle activity (Fig. 2.19). and do not activate the HPA axis. Koolhaas et al.
The importance of physical exertion in DPAG- [388] added that stress should be reconceptual-
evoked neuroendocrine responses is further evi- ized as a stimulus or environmental condition in
denced by studies showing that rat escape which the response demands exceed the adaptive
responses to both DPAG stimulations and 20-kHz capacity of the organism. As a result, the home-
artificial alarm calls attain maximum average ostatic responses should be modified (allostatic
speed of 1.9 m/s (6.8 km/h) [380, 381]. This changes) to achieve stability in anticipation of
speed is over four times the rat treadmill speed physiological requirements of stress [388]. The
(1.5 km/h) required to produce significant latter argument is consonant with Preter and
increases in both PVN c-fos expression and LAC Kleins [27] contention that under suffocation,
and ACTH plasma levels [382]. the acute activation of the HPA axis would coun-
The HPA axis unresponsiveness to DPAG terproductively increase the oxygen demand
stimulations is further supported by the lack of beyond the adaptive capacity of the organism.
the DPAG excitatory projections to PVN. Indeed, Accordingly, Preter and Klein [27] proposed that
Pittman et al. [383] showed that PAG stimula- fear neuroendocrine response should be modified
tions activated only 2 of 188 neurons tested in the during asphyxia to allow energy conservation
PVN. In turn, Ziegler et al. [384] recently showed until a possible escape.
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 53

Fig. 2.18 Hormone and metabolite plasma levels follow- Fig. 2.19 Hormone and metabolite plasma levels 3 or
ing electrical stimulations of DPAG of rats placed in a small 30 min after the application of 1-min foot-shocks either
compartment (20-cm diameter roofed-cylinder) that pre- effective (1 mA) or fictive. Note the marked increase in
vented flight behavior. Rats were stimulated either fictively ACTH, PRL and LAC 3 min after the end of foot-shock.
(SHAM) or at the flight threshold intensity (DPAG) deter- Symbols indicate (#) significant differences from sham-
mined 2 weeks before in a 60-cm diameter open-field. shocked group for the same blood sampling interval, or
Columns represent means (SEM) from blood samples col- (*) significant differences from 3-min plasma level for the
lected 3 or 30 min after the end of 1-min intracranial stimu- same shock condition. Other details as in Fig. 2.18 (from
lus. Groups which were further exposed to a 60-cm Armini et al. [379], with permission)
diameter brightly-lit open-field (30-min groups) showed
marked increases in both CORT and ACTH, confirming the
HPA responsiveness to a novel environment. ACTH corti-
cotropin, CORT corticosterone, GLU glucose, LAC lactate, tion of the carotid body [389]. HPA axis activa-
PRL prolactin. Symbols indicate (*) significant differences tions of anesthetized animals were even more
between 3- and 30-min plasma levels for the same treat- conspicuous with higher levels of both hypoxia
ment and (#) significant differences between SHAM and and hypercapnia [390, 391]. Most importantly,
DPAG groups for the same sample interval (from Armini
et al. [379], with permission) however, acute exposures (20 min) of conscious
rats to normocapnic hypoxia (7 % O2), hypercap-
nia (8 % CO2) or hypercapnic hypoxia (7 % O2,
Nevertheless, whereas the 20-min exposure of 8 % CO2) produced ACTH increases of approxi-
anesthetised dogs to low levels of hypercapnia mately 100 %, 200 % and 300 %, respectively
(6 % CO2) did not alter the secretion of both [227]. While these findings contradict Preter and
ACTH and CORT, 20-min exposures to moderate Kleins [27] argument, the lack of HPA responses
hypoxia (10 % O2) produced a 175 % increase in appears nevertheless to hold in prolonged
the ACTH plasma level that was attenuated hypoxia. For instance, neither the secretion of
(albeit not abolished) by chronic chemodenerva- ACTH, nor that of CORT showed any change fol-
54 L.C. Schenberg

lowing the 42-h exposure of conscious rats to in adulthood. To rule out the influences of genetic
hypocapnic hypoxia [392]. Consequently, data background and caregiving behaviors, the authors
suggest that the PAG inhibition of HPA axis employed a split-litter design in which half of
might be masked during acute asphyxia by neu- male pups were subjected to 3-h daily isolations
roendocrine activations brought about by NTS throughout lactation (PN2-PN21) while siblings
excitatory projections to PVN, DMH, CeA and and dam were moved to another box. Because
bed nucleus of stria terminalis [187]. Finally, siblings were both handled and kept with a
although the HPA axis is reportedly activated separation-anxious mother, they were considered
during panics caused by two tidal-volume inhala- a sham-isolated group. Experimental groups
tion of 35 % CO2, NTS projections to both PAG were further compared with controls that
and PVN are believed to be inactive during room- remained undisturbed in home-cages until wean-
air conditions of spontaneous panic attacks. ing. At 60 days of age, the rats were implanted
Irrespective of the mechanism involved, the with electrodes into the DPAG and, 1 week later,
lack of stress hormone responses in the DPAG- subjected to sessions of DPAG stimulation, EPM
evoked panic-like behaviours stands out as a and forced-swimming (FST) for the assessment
compelling evidence of the DPAG mediation of of panic vulnerability, baseline anxiety and
clinical panic. depressive mood, respectively.
Results showed that DPAG-evoked panic-like
responses of immobility, exophthalmus, trotting,
2.11 Modeling the Comorbidity galloping and jumping were significantly facili-
of Panic Disorder tated in NSI rats relative to both sham-isolated
with Childhood Separation and control groups (Fig. 2.20). In contrast, NSI
Anxiety rats did not show any change in scores of either
anxiety or depression relative to sham-isolated
Although clinical and epidemiological evidence rats. Groups also did not differ with respect to
suggests that CSA predisposes the subject to defecation or micturition responses to DPAG
adult-onset PD (for review, see Aschebrand et al. stimulations. The latter observations agree with
[37], other studies reported that neither CSA previous studies suggesting that DPAG-evoked
[393] nor early-life adversity [35, 41] had any defensive behaviors (freezing and flight) and pel-
effect on later incidence of panic attacks. For vic viscera responses (micturition and defeca-
instance, Roberson-Nay et al. [41] showed that tion) are processed by functionally distinct
although PD and CSA share a common genetic systems [60, 165, 246, 394, 395]. As it regards, it
diathesis, childhood adversities account for only is pertinent that urges of defecation and micturi-
1.2 % of adult-onset panic attacks. Conversely, tion are neither experienced by patients during
Spatola et al. [153] reported that early-life adver- panic attacks [53, 396] nor recognised as symp-
sities do correlate with CO2 hypersensitivity pro- toms typical of clinical panic [397, 398]. In any
vided that the stressful events took place before event, the NSI facilitation of DPAG-evoked
18 years of age. Accordingly, the influence of the defensive behaviors are the first behavioral evi-
early-life adversities in the later development of dence in animals that early-life separation stress
PD remains unclear. Whether genetic or epige- selectively facilitates panic-like behaviors in
netic, the brain mechanisms whereby CSA pre- adulthood. Importantly, as well, these data impli-
disposes subjects to PD are completely unknown. cate the DPAG not only in panic attacks but also
Although the comorbidity of psychiatric dis- in the predispositions of separation-anxious chil-
orders is rarely studied in animals, a recent study dren to the later development of PD.
of Quintino-dos-Santos et al. [137] examined The study of Quintino-dos-Santos et al. [137]
whether the neonatal social isolation (NSI), a is consonant with previous data showing that
model of CSA, facilitates panic-like behaviors neonatally-isolated adult rats present sex-
produced by electrical stimulations of the DPAG dependent facilitations of panic-like respiratory
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 55

responses to either hypoxia (males) or hypercap- [128, 406] or to 3-h daily periods of maternal
nia (females) [37, 138, 399402]. As a corollary, separations [400, 407] during the stress hypore-
the PAG is unresponsive to hypoxia during sponsive period.
PN2-PN12 [225] stress hyporesponsive period The study of Quintino-dos-Santos et al. [137]
in which the mother shields the pup against early- is reminiscent of Rachel Kleins [36] 15-year
life adversities [256258]. Although Dumont double-blind interview-based follow-up study of
et al. [138] suggested that NSI facilitates the cen- children presenting manifest symptoms of CSA
tral processing of chemoreceptor afferent inputs, (school refusal). The latter study showed that
Quintino-dos-Santos et al. [137] presented evi- although the diagnoses of major depression dis-
dence that neonatal separation may sensitise order (MDD) did not differ from controls,
DPAG regions that mediate defensive responses. separation-anxious subjects showed significant
As a matter of fact, predators are the major threat increases in the frequency of both panic attacks
to the separated pup. Otherwise, NSI facilitations and hospitalizations due to depressive episodes.
of DPAG-evoked panic could be the result of Unexpectedly, however, recent studies from our
enduring plastic changes of both ascending and laboratory showed that rats subjected to 3-h daily
descending projections of DPAG. Lastly, panic maternal separations along the lactation period
facilitations could be the outcome of early-life (PN2-PN21) were more resilient than both controls
programming of HPA axis [403]. Indeed, the and rats separated during stress hyporesponsive
HPA axis is known to be hyperactive in both period only (PN2-PN12) (A.C.B. Aguiar, unpub-
panic patients [404, 405] and adult rats subjected lished results). In particular, 21-day separated rats
to a 24-h single period of mother deprivation showed longer swimming duration in FST,

Fig. 2.20 Median threshold intensities (I50 SE) of and sham-isolated rats (+) for Bonferronis 5 % criterion
DPAG-evoked behaviors of neonatally-isolated rats, (likelihood ratio chi-square tests for curve location) (from
sham-isolated rats and non-handled controls. Symbols Quintino-dos-Santos et al. [137], with permission)
represent values significantly different from controls (*)
56 L.C. Schenberg

increased appetite for sugar in sucrose preference differ in controls and depressed outpatients with-
test (SPT), higher thresholds of DPAG-evoked out a history of panic attacks.
panic responses and higher weight gains. Procedures Nonetheless, because spontaneous panic attacks
of latter study were nevertheless rather distinct from are conspicuously uncontrollable stress, a recent
Quintino-dos-Santos et al. [137]. For instance, study of our laboratory assessed the late effects of
whereas the treatments of latter study were applied uncontrollable shocks, a presumptive model of
to siblings from the same litter, those of recent study depression and/or trauma, on DPAG-evoked panic
were applied to pups of independent litters. behaviors [416]. Briefly, rats with electrodes in the
Moreover, whereas the previous study employed DPAG were subjected to a 7-day shuttle-box one-
group-reared adult rats, the recent study examined way escape yoked training with escapable (ES) or
rats reared individually. As well, Quintino-dos- inescapable (IS) foot-shocks. Controls were sub-
Santos et al. [137] surgery (bone removal for sinus jected to fictive shock (FS) sessions. The day after
exposure) and recovery period (7-day in glass- the termination of one-way escape training, the rats
walled brightly-lit individual cages) were more were trained in a two-way escape novel task to
stressful than the surgery (burr hole only) and ascertain the effectiveness of uncontrollable stress.
recovery (28-day period in polypropylene individ- Data showed that the IS group performed signifi-
ual cages) of the recent study. Yet, because sham- cantly poorer than the ES group in two-way escape
isolated rats of former study did not differ from task. Unexpectedly, however, IS rats showed a
controls [137], the NSI facilitation of DPAG-evoked marked attenuation of DPAG-evoked freezing and
behaviors was most probably due to the interaction flight behaviors relative to both ES and FS groups,
of early-life and adult-life stress. Indeed, Roberson- 2 and 7 days after one-way escape training (Fig. 2.21).
Nay et al. [41] presented epidemiological evidence Moreover, whereas the threshold of DPAG-evoked
that 60 % of variation of adult-onset panic attacks is freezing and flight behaviors of IS rats remained
due to environmental factors of adulthood. high or were further increased 7 days after escape
training, thresholds of DPAG-evoked defecation
and micturition did not change or were even
2.12 Modeling the Comorbidity reduced. The latter data support previous studies
of Panic and Depression [246, 395] suggesting the separate processing of
defensive behaviors (freezing and flight) and pelvic
Panic disorder is highly comorbid with depres- viscera responses (micturition and defecation) pro-
sion [154, 408, 409]. For instance, a 10-year duced by electrical stimulation of DPAG. Most
interview-based prospective study showed that importantly, IS inhibited DPAG-evoked defensive
13.6 % and 27.3 % PD patients at age 30 had behaviors in spite of the striking differences in the
either MDD or recurrent brief depression (RBD), aversive stimulus (foot-shock vs intracranial stimu-
respectively [154]. Notably, as well, the latter lus) and context (shuttle-box vs open-field) [416].
authors found that 10-year prevalence of depres- Accordingly, effects cannot be attributed to either a
sion and suicide attempts in PD patients is 6.8 context conditioning or a stimulus sensitisation.
and 4.2 times higher relative to the general popu- In the same vein, experiments carried out in the
lation. Clinical data also suggest that panic is elevated T-maze (ETM) detected panicolytic
facilitated by both acute and post-traumatic stress effects the day after the rat exposure to either IS,
disorders [410414]. In contrast to the existing FST or restraint stress [417]. Thus, whereas the
evidence of a common genetic diathesis of PD ETM anxiety-like behavior (avoidance from open-
and CSA [41], mechanisms underlying the arms) was enhanced, the ETM panic-like behavior
comorbidity of panic and depression disorders (escape from open-arms) was attenuated. The latter
remain nevertheless completely obscure. It is effect closely resembles the attenuation of DPAG-
also unclear whether PD is enhanced by any kind evoked escape in inescapably-shocked rats [416].
of depressive disorder. In particular, McGrath Evidence amassed in last decades suggests
et al. [415] reported that LAC sensitivity did not that subjects exposed to uncontrollable stress
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 57

develop a depression-like syndrome character- Although most researchers associate the out-
ised by decreased motivation to respond to the comes of uncontrollable stress with putative
same or other aversive stimuli, by cognitive defi- changes in the amygdala [423425], hippocam-
cit (learned helplessness), and by emotion and pus [424, 426431], or both structures [432, 433],
mood effects including an early increase in anxi- Lino-de-Oliveira et al. [434, 435] showed that
ety and the later development of depression. Data whereas the microinjections of glutamate into the
from yoked experiments further showed that DPAG reduce floating behavior, microinjections
these effects result from the subject learning that of lidocaine had the opposite effect. Moreover,
stress is beyond his control and not from the they showed that sub-chronic administrations of
stressor aversiveness [418421]. Similarly, the antidepressants reduce FST-induced increases in
FST is a widespread procedure for the screening FLI in most columns of the PAG [434, 435]. Most
of potential antidepressants [422] based on the notably, however, evidence from positron-emis-
assumption that floating correlates with depres- sion tomography in rats (microPET) showed that
sive mood. Regardless of whether or not uncon- whereas the PAG is markedly activated during the
trollable stress produces true depressed mood, IS FST training session, it remains inactive during
inhibition of escape to both foot-shock and intra- next day test session [436]. Taken together, these
cranial stimulus implicates the DPAG as the data implicate the DPAG in behavioral effects of
motivational substrate of both escape responses. uncontrollable stress.

Fig. 2.21 Percent changes of median threshold intensi- training (days 7 to 0). Symbols indicate significant dif-
ties (I50 SE) of DPAG-evoked defensive behaviors in ferences relative to (*) baseline value, () DPAG second
the 2nd (2) and 7th (7) days after the end of one-way stimulation session, or to (+) ES and (#) FS groups
escape training with foot shocks either fictive (FS), escap- (Bonferronis 5 % criterion of likelihood ratio chi-square
able (ES) or inescapable (IS). Screening sessions were tests for location of parallel-fitted threshold curves) (from
carried out the day before the onset of one-way escape Quintino-dos-Santos et al. [416], with permission)
58 L.C. Schenberg

Remarkably, as well, freezing was also attenuated the consequences of uncontrollable stress have
one week after the exposure to IS. Accordingly, the been used as a model of both depression [418421,
IS attenuation of DPAG-evoked escape behaviors 443] and trauma [444], the DPAG-evoked panic-
cannot be ascribed to an enhancement of freezing at like behaviors were expected to be facilitated in
the expense of trotting and galloping. The impair- inescapably-shocked rats. Accordingly, the IS
ment of both passive (freezing) and active (flight) unexpected inhibition of DPAG-evoked panic
defensive behaviors is best explained by a deactiva- behaviors may be a idiosyncratic feature of the
tion of a DPAG in-built motivational system. If so, present model of depression. Indeed, whereas PD
the presumptive panicolytic effects of uncontrolla- is most often associated with MDD and RBD
ble stress on ETM [417] and DPAG-evoked defen- [154], the exposure to uncontrollable stress is rem-
sive behaviors [416] are better explained by a iniscent of the reactive depression nowadays
decrease in resilience (escape failure) to stress. The termed adjustment disorder with depressed mood
latter argument is further supported by the higher [397]. Indeed, DPAG-evoked panic behaviors
degree of attenuation of trotting and galloping which were markedly facilitated in olfactory bulbectomy
are the common responses to both the DPAG stimu- model of depression (J.W. Quintino-dos-Santos,
lations and the inescapable foot-shocks. unpublished results).
Conversely, however, Strong et al. [437] pre- In conclusion, the conjoint inhibition of pas-
sented evidence that %-HT transmission in the sive (freezing) and active (flight) defensive
dorsal striatum (DS) plays a crucial role in learn- behaviors suggests that IS deactivates a DPAG
ing deficits of inescapably-shocked rats. Indeed, in-built motivational system that may be impli-
whereas the microinjections of a 5-HT2C antago- cated in depressed patients difficulties to cope
nist into the DS prevented escape failures of rats with the stressors of daily life.
previously exposed to IS, microinjections of a
5-HT2C agonist impaired learning even in the
absence of the prior exposure to IS. Therefore, it 2.13 Modeling Female
is tempting to speculate that DPAG and DS medi- Vulnerability to Panic
ate motivational and cognitive effects of helpless- Disorder
ness, respectively. On the other hand, Tannure
et al. [338] presented evidence of a two-process Despite the high variability of epidemiological
flight system. Indeed, whereas the kindling of the survey data, the prevalence of panic attacks is con-
amygdala facilitated the flight responses from the sistently higher in women. In particular, at age 30,
experimenter (capture resistance), it inhibited the the Zurich study found panic attacks in 10 % of the
DPAG-evoked flight behaviors. Be this as it may, population, with a sex ratio of 3:1 in favor of
these systems might be bridged by DPAG profuse females [154]. Although panic attacks are rarely
efferents to intralaminar and midline thalamic observed before puberty or after menopause, they
nuclei [438] that exert diffuse excitatory actions are more frequent and severe during the premen-
on both cortex and striatum [439]. strual phase [445]. Patients with LLPDD) also
In turn, clinical and epidemiological evidence show higher sensitivity to panicogenic challenges
suggest that the first episode of MDD is very often in the late luteal phase [446448].
precipitated by uncontrollable stress, including These and other data prompted Lovick and col-
social loss, bond breakdown, disease and unem- laborators to carry out several studies that exam-
ployment. There is additional evidence that PD ined the neural excitability of the DPAG during the
predisposes to the later development of both estrous cycle [155, 449451]. These authors
depression [154, 408410, 440] and trauma [410 showed that the neurons of the DLPAG increase
412, 414, 441]. Furthermore, patients with post- the expression of the alpha4/beta1/delta (4/1/)
traumatic stress disorder (PTSD) not only subunits of the GABA-A receptor in the late dies-
experience the physiological symptoms of panic trus relative to the other phases of the estrous cycle
but also fear these symptoms [414, 442]. Because or to their expression in male rats [452]. Moreover,
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 59

they showed that most GABA4/1/ receptors of Anxiety is nevertheless reinstated following the
the DLPAG are localised in gabaergic neurons sudden withdrawal of progesterone, which mim-
(autoreceptors), as revealed by the double label- ics the rapid decay of this hormone preceding
ling of these subunits and glutamic acid decarbox- menses [457, 458].
ylase [453, 454]. Because the GABA4/1/ According to Smith and collaborators, the
receptors are highly sensitive to GABA [452], the increases in anxiety in the early and late luteal
upregulation of GABA4/1/ autoreceptors is phases appear to be due to a allopregnanolone-
expected to reduce the GABA inhibitory tonus mediated upregulation of the 4 subunit of
within the DLPAG. Accordingly, Lovick and col- GABA-insensitive receptors [459461]. Increases
laborators suggested that the increased expression in progesterone withdrawal anxiety in rats can be
of GABA4/1/ autoreceptors may contribute prevented by indomethacin blockage of the break-
both to the development of LLPDD and to the high down of progesterone to allopregnanolone.
comorbidity of LLPDD and PD. Therefore, allopregnanolone rather than proges-
Lovick and collaborators also showed that the terone is the active compound [462]. Indeed, the
pressor response, tachycardia, and tachypnea post-treatment allopregnanolone levels in LLPDD
produced by the systemic injection of a CCK- patients are significantly lower in improved sub-
related panicogen (pentagastrin) were enhanced jects relative to unimproved ones. Improvement
in the late diestrus. For rats in estrus, the pressor was also significantly associated with lower allo-
response and tachycardia, but not the tachypnea, pregnanolone levels for premenstrual depression
were significantly larger than the response evoked and appetite changes [463].
in the early diestrus [455]. Overall, the above studies suggest that the
Lastly, extracellular recordings from output increased incidence of anxiety and panic in the late
neurons in the DPAG showed that the increased luteal phase appears to be mediated by the distinct
firing rate produced either by the intravenous effects of allopregnanolone in GABA transmis-
administration of pentagastrin or by the ionto- sion, i.e., the upregulation of GABA-insensitive
phoretic injection of the GABA-A receptor receptors in amygdala areas controlling anxiety
antagonist bicuculline, is significantly increased and the upregulation of GABA-hypersensitive
during the estrus and late diestrus relative to pro- autoreceptors in periaqueductal areas controlling
estrus and early diestrus [456]. panic. These mechanisms may be complementary
On the other hand, the bimodal incidence of in determining the high incidence of panic in
catamenial epilepsy suggests that brain excitabil- women.
ity increases in the middle of the cycle, between
the peak of estradiol and the early surge of pro-
gesterone, and during the sharp fall of proges- 2.14 Conclusion
terone prior to menses [457]. Although the
neuroexcitant properties of estradiol are respon- Existing evidence suggests that DLPAG is an
sible for the increased excitability around the exteroceptive column that integrates both the
midcycle peak, the higher excitability in early and amygdala-processed odor inputs from VMH and
late luteal phases are most likely associated with PMD and visuo-acoustic inputs from DLSC [249,
changes in brain levels of the neuroactive metabo- 464468]. These circuits appear to mediate the non-
lite of progesterone, allopregnanolone. Indeed, respiratory type of panic attacks [52]. By contrast,
whereas the acute injections of both progesterone evidence herein discussed suggests that the CePAG
and allopregnanolone have manifest anxiolytic is an interoceptive column that receives afferents
effects in rats, the 72-h exposure to progesterone from NTS and projects to both the LPAGcv and to
increases anxiety, akin to that observed in the the inner division of eLPBA. As shown in the pres-
early surge of progesterone in women. Thereafter, ent study, the CePAG-LPAGcv circuit may be the
anxiety decreases to baseline levels in spite of the very core of respiratory-type panic attacks. This cir-
sustained exposure of rats to progesterone. cuit appears to be modulated by NRDlw inhibitory
60 L.C. Schenberg

projections that gate asphyxia signals. In turn, both There are important misses as well. Indeed,
DLPAG and LPAGcv project to the CnF (the mid- although the DPAG appears to be activated by both
brain locomotor region) [162]. Although electrical LAC infusions and CO2 inhalations in humans, nei-
and chemical stimulations of CnF produce behav- ther the LAC nor the CO2 produce panic behaviors
ioral and cardiorespiratory responses quite similar in rats. Moreover, DPAG-evoked behaviors are not
to those of the stimulation of the DPAG [280, attenuated by chronic administration of the stan-
L.C. Schenberg, unpublished results], it is unknown dard panicolytic imipramine. The negative result
whether these responses are attenuated by clini- may be due to the high dose of imipramine
cally-effective panicolytics, as shown for DPAG- (10 mg kg1 day1, 21 days). Indeed, the 21-day
mediated panic-like responses to both electrical treatment with 5 mg kg1 fluoxetine was less effec-
stimulations and cyanide injections. tive in attenuating DPAG-evoked panic-like behav-
The PAG-evoked panic responses satisfy most iors than the 21-day treatments with 1 or 2 mg kg1
criteria of a translational model of clinical panic, fluoxetine [60, C.S. Bernab, unpublished results].
i.e., face validity (homology of symptoms and Although frequently presented as evidence of
physiological responses), predictive validity (drug DPAG mediation of panic attacks, the attenuation
sensitivity) and construct validity (facilitation by of DPAG-evoked shuttle-box escape responses by
hypoxia, female hormones and early-life stress). acute treatments with SSRIs is another inconsis-
Moreover, PAG stimulations in humans produce tency of this model [303, 469, 470]. These effects
emotional, neurological and autonomic responses may be due to specific features of the shuttle-box.
strikingly similar to those of panic attacks either Indeed, acute injections of low doses of either clo-
spontaneous or provoked by intravenous infu- mipramine or fluoxetine failed in attenuating the
sions of LAC. The PAG model of panic is also DPAG-evoked panic-like responses of rats stimu-
endorsed by the attenuation of DPAG-evoked lated in an open-field [60, 471]. Although the
panic-like behaviors by clinically effective pani- inhibitory effect of hypercapnia on DPAG-evoked
colytics given at doses and regimens similar to panic responses appears incongruent, pre-expo-
those of panic therapy. In particular, DPAG- sures to hypercapnia produce a reliable facilitation
evoked panic was markedly attenuated by a of KCN-evoked panic responses. Accordingly, the
21-day administration of daily doses of fluoxetine former effect may be due to the non-selective fea-
as low as 1 mg kg1 and, conversely, facilitated by tures of electrical stimulation.
systemic injection of the putative panicogen pen- Most importantly, however, recent results from
tylenetetrazol. Remarkably, as well, DPAG stimu- our laboratory showed that 0.5 M LAC infusions
lations are not accompanied by stress hormone failed in facilitating the KCN-evoked escape
responses in the absence of muscular effort. responses presumptively mediated at the PAG
DPAG-evoked responses are also facilitated in (E.A. Moraes, unpublished results). Although the
both female rats and neonatally-isolated adult latter results may be due to the smaller effects of
rats. Lastly, recent evidence showed that the 0.5 M LAC infusions in rats [342], they also sug-
DPAG mediates respiratory-type panic attacks to gest that LAC vulnerability to panic attacks is
KCN selective stimulation of chemoreceptors and mediated by mechanisms other than those of
that these responses are potentiated by hypercap- KCN-evoked panic attacks. This possibility is
nia and attenuated by acute and chronic treat- supported both by the provocation of panic attacks
ments with clinical doses of the established by d-LAC [141] and by the lack of activation of
panicolytics clonazepam and fluoxetine, respec- PAG in rats infused with LAC [341].
tively [254]. Similarly, escape responses to ambi- In conclusion, although the animal models are
ent hypoxia were attenuated by acute and chronic not expected to reproduce clinical disorders
administrations of alprazolam and fluoxetine, exactly, a translational model of PD should (1)
respectively, and also by intra-periaqueductal present face validity, (2) be sensitive to clinically-
injections of 5-HT and its agonists [255]. effective panicolytics in doses and regimens
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 61

Table 2.4 Truth table of translational features of panic attacks evoked by the stimulation of dorsal periaqueductal gray
matter (DPAG) of the midbrain of animals and humans
Clinical panic
DPAG-evoked panic in YES Symptomatologya Attenuation by acute SSRIsj
animals or humans Lack of HPA axis responsesb
Lack of PRL axis responsesb
Attenuation by:
Chronic fluoxetine
Chronic clomipramine
Acute alprazolam
Acute clonazepam
Facilitation by:
Hypercapnic hypoxiad
Late luteal phase
Early-life stresse
Major depressionf
NO Attenuation by chronic imipramineg Attenuation by:
Facilitation by hypercapnia Acute maprotiline
Facilitation by LAC i.v. infusionsh Chronic maprotiline
Facilitation by reactive depressioni Acute diazepam
Acute midazolam
Acute buspirone
Chronic buspirone
Predisposed by specific
Notes refer to DPAG-evoked panic attacks in animals
See Table 2.2
Stimulation of rats in a 20-cm diameter arena that prevented flight behavior [379]
KCN-evoked selective cytotoxic hypoxia of chemoreceptors
CO2 potentiation of DPAG-mediated panic attacks evoked by intravenous injections of KCN
Neonatal social isolation
Olfactory bulbectomy model of depression (unpublished results)
Dose above clinical range
0.5 M only (unpublished results)
Exposure to inescapable shocks
Doses above clinical range
See Tannure et al. [338]

similar to those of panic therapy, (3) be sensitive References

to clinically-specific panicogens (57 % CO2 and
0.5 M LAC), (4) be devoid of neuroendocrine 1. Freud S. On the right to separate from neurasthenia a
responses, (5) be facilitated in female subjects definite symptom-complex as anxiety neurosis. In:
and (6) reproduce the clinical comorbidity of Brill JAA, editor. Selected papers on hysteria and
other psychoneuroses. New York: The Journal of
panic disorder. Together, DPAG and KCN mod- Nervous and Mental Disease Publishing Company;
els of panic attacks meet most of these criteria 1912. p. 13354. Chapter VI.
(Table 2.4). Preliminary modeling of the comor- 2. Freud S. The psychotherapy of hysteria. In: Brill
bidity of PD and CSA further implicates the PAG JAA, editor. Selected papers on hysteria and other
psychoneuroses. New York: The Journal of Nervous
in the predisposition of separation-anxious chil- and Mental Disease Publishing Company; 1912.
dren to the later development of PD. p. 75120. Chapter IV.
62 L.C. Schenberg

3. Freud S. Obsessions and phobias: their psychical attacks: do patients and controls respond differently?
mechanism and their aetiology. Complete psycho- Psychiatry Res. 1986;17:295308.
logical works, standard edition. London: Hogarth 22. Margraf J, Ehlers A, Roth WT. Sodium lactate infu-
Press; 1962. p. 81. sions and panic attacks: a review and critique.
4. APA. Diagnostic and statistical manual of mental Psychosom Med. 1986;48:2351.
disorders. 3rd ed. Washington, DC: American 23. Clark DM, Salkovskis PM, Ost LG, Breitholtz E,
Psychiatry Association; 1980. Koehler KA, Westling BE, et al. Misinterpretation of
5. Drury A. The percentage of carbon dioxide in the body sensations in panic disorder. J Consult Clin
alveolar air, and the tolerance to accumulating car- Psychol. 1997;65:20313.
bon dioxide in cases of so-called irritable heart of 24. Klein DF. False suffocation alarms, spontaneous
soldiers. Heart. 1920;7:16573. panics, and related conditions. An integrative
6. Cohen ME, White PD. Life situations, emotions hypothesis. Arch Gen Psychiatry. 1993;50:30618.
and neurocirculatory asthenia (anxiety neurosis, 25. Klein DF. Panic may be a misfiring suffocation alarm.
neurasthenia, effort syndrome). Res Publ Assoc In: Montgomery SA, editor. Psychopharmacology of
Res Nerv Ment Dis. 1949;29:83269. panic. 1st ed. New York: Oxford University Press;
7. Pitts FN, McClure JN. Lactate metabolism in anxi- 1993. p. 6773.
ety neurosis. N Engl J Med. 1967;277:132936. 26. Shavitt RG, Gentil V, Mandetta R. The association
8. Cowley DS, Roy-Byrne PP. Hyperventilation and of panic/agoraphobia and asthma. Contributing fac-
panic disorder. Am J Med. 1987;83:92937. tors and clinical implications. Gen Hosp Psychiatry.
9. Klein DF, Fink M. Psychiatric reaction patterns to 1992;14:4203.
imipramine. Am J Psychiatry. 1962;119:4328. 27. Preter M, Klein DF. Panic, suffocation false alarms,
10. Klein DF. Delineation of two drug-responsive anxiety separation anxiety and endogenous opioids. Prog
syndromes. Psychopharmacologia. 1964;5:397408. Neuropsychopharmacol Biol Psychiatry. 2008;32:
11. Bowlby J. Attachment, separation anxiety, loss. 2nd 60312.
ed. NY: Basic Books; 1969. 28. Griez E, Schruers K. Experimental pathophysiology
12. Gorman JM, Askanazi J, Liebowitz MR, Fyer AJ, of panic. J Psychosom Res. 1998;45:493503.
Stein J, Kinney JM, et al. Response to hyperventila- 29. Liebowitz MR, Fyer AJ, Gorman JM, Dillon D,
tion in a group of patients with panic disorder. Am Davies S, Stein JM, et al. Specificity of lactate infu-
J Psychiatry. 1984;141:85761. sions in social phobia versus panic disorders. Am
13. Liebowitz MR, Gorman JM, Fyer AJ, Levitt M, J Psychiatry. 1985;142:94750.
Dillon D, Levy G, et al. Lactate provocation of panic 30. Rifkin A, Klein DF, Dillon D, Levitt M. Blockade by
attacks: II. Biochemical and physiological findings. imipramine or desipramine of panic induced by
Arch Gen Psychiatry. 1985;42:70919. sodium lactate. Am J Psychiatry. 1981;138:6767.
14. Cameron OG, Lee MA, Curtis GC, McCann 31. Liebowitz MR, Fyer AJ, Gorman JM, Dillon D,
DS. Endocrine and physiological changes during spon- Appleby IL, Levy G, et al. Lactate provocation of
taneous panic attacks. Psychoneuroendocrinology. panic attacks: I. Clinical and behavioral findings.
1987;12:32131. Arch Gen Psychiatry. 1984;41:76470.
15. Levin AP, Doran AR, Liebowitz MR, Fyer AJ, 32. Woods SW, Charney DS, Delgado PL, Heninger
Gorman JM, Klein DF, et al. Pituitary adrenocortical GR. The effect of long-term imipramine treatment
unresponsiveness in lactate-induced panic. on carbon dioxide-induced anxiety in panic disorder
Psychiatry Res. 1987;21:2332. patients. J Clin Psychiatry. 1990;51:5057.
16. Woods SW, Charney DS, McPherson CA, Gradman 33. Yeragani VK, Pohl R, Balon R, Rainey JM, Berchou
AH, Heninger GR. Situational panic attacks. R, Ortiz A. Sodium lactate infusion after treatment
Behavioral, physiologic, and biochemical character- with tricyclic antidepressants: behavioral and physi-
ization. Arch Gen Psychiatry. 1987;44:36575. ological findings. Biol Psychiatry. 1988;24:76774.
17. Woods SW, Charney DS, Goodman WK, Heninger 34. Gittelman-Klein R, Klein DF. School phobia: diag-
GR. Carbon dioxide-induced anxiety. Behavioral, nostic considerations in the light of imipramine
physiologic, and biochemical effects of carbon diox- effects. J Nerv Ment Dis. 1973;156:199215.
ide in patients with panic disorders and healthy sub- 35. Battaglia M, Bertella S, Politi E, Bernardeschi L,
jects. Arch Gen Psychiatry. 1988;45:4352. Perna G, Gabriele A, et al. Age at onset of panic dis-
18. Hollander E, Liebowitz MR, Cohen B, Gorman JM, order: influence of familial liability to the disease
Fyer AJ, Papp LA, et al. Prolactin and sodium lactate- and of childhood separation anxiety disorder. Am
induced panic. Psychiatry Res. 1989;28:18191. J Psychiatry. 1995;152:13624.
19. Hollander E, Liebowitz MR, Gorman JM, Cohen B, 36. Klein RG. Is panic disorder associated with childhood
Fyer A, Klein DF. Cortisol and sodium lactate-induced separation anxiety disorder? Clin Neuropharmacol.
panic. Arch Gen Psychiatry. 1989;46:13540. 1995;18 Suppl 2:S714.
20. Hollander E, Liebowitz MR, DeCaria C, Klein 37. Battaglia M, Ogliari A, D'Amato F, Kinkead
DF. Fenfluramine, cortisol, and anxiety. Psychiatry R. Early-life risk factors for panic and separation
Res. 1989;31:2113. anxiety disorder: insights and outstanding questions
21. Ehlers A, Margraf J, Roth WT, Taylor CB, Maddock arising from human and animal studies of CO sensi-
RJ, Sheikh J, et al. Lactate infusions and panic tivity. Neurosci Biobehav Rev. 2014;46:45564.
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 63

38. Bernstein GA, Borchardt CM, Perwien AR, Crosby factor and cluster analysis. Acta Psychiatr Scand.
RD, Kushner MG, Thuras PD, et al. Imipramine plus 1996;93:806.
cognitive-behavioral therapy in the treatment of 56. Perna G, Caldirola D, Namia C, Cucchi M, Vanni G,
school refusal. J Am Acad Child Adolesc Psychiatry. Bellodi L. Language of dyspnea in panic disorder.
2000;39:27683. Depress Anxiety. 2004;20:328.
39. Roberson-Nay R, Klein DF, Klein RG, Mannuzza S, 57. Briggs AC, Stretch DD, Brandon S. Subtyping of
Moulton III JL, Guardino M, et al. Carbon dioxide panic disorder by symptom profile. Br J Psychiatry.
hypersensitivity in separation-anxious offspring of 1993;163:2019.
parents with panic disorder. Biol Psychiatry. 2010;67: 58. Nardi AE, Nascimento I, Valenca AM, Lopes FL,
11717. Mezzasalma MA, Zin WA, et al. Respiratory panic
40. Battaglia M, Pesenti-Gritti P, Spatola CA, Ogliari A, disorder subtype: acute and long-term response to
Tambs K. A twin study of the common vulnerability nortriptyline, a noradrenergic tricyclic antidepres-
between heightened sensitivity to hypercapnia and sant. Psychiatry Res. 2003;120:28393.
panic disorder. Am J Med Genet B Neuropsychiatr 59. Roberson-Nay R, Latendresse SJ, Kendler KS. A
Genet. 2008;147B:58693. latent class approach to the external validation of
41. Roberson-Nay R, Eaves LJ, Hettema JM, Kendler respiratory and non-respiratory panic subtypes.
KS, Silberg JL. Childhood separation anxiety disor- Psychol Med. 2012;42:46174.
der and adult onset panic attacks share a common 60. Schenberg LC, Bittencourt AS, Sudr ECM, Vargas
genetic diathesis. Depress Anxiety. 2012;29:3207. LC. Modeling panic attacks. Neurosci Biobehav
42. Preter M, Klein DF. Lifelong opioidergic vulnerabil- Rev. 2001;25:64759.
ity through early life separation: a recent extension 61. Schenberg LC, Schimitel FG, Armini RS, Bernabe
of the false suffocation alarm theory of panic disor- CS, Rosa CA, Tufik S, et al. Translational approach
der. Neurosci Biobehav Rev. 2014;46:10351. to studying panic disorder in rats: hits and misses.
43. Lawson EE, Waldrop TG, Eldridge FL. Naloxone Neurosci Biobehav Rev. 2014;46:47296.
enhances respiratory output in cats. J Appl Physiol. 62. Mobbs D, Petrovic P, Marchant JL, Hassabis D,
1979;47:110511. Weiskopf N, Seymour B, et al. When fear is near:
44. Bonham AC. Neurotransmitters in the CNS control threat imminence elicits prefrontal-periaqueductal
of breathing. Respir Physiol. 1995;101:21930. gray shifts in humans. Science. 2007;317:107983.
45. van der Schier R, Roozekrans M, van VM, Dahan A, 63. Gorman JM, Kent JM, Sullivan GM, Coplan
Niesters M. Opioid-induced respiratory depression: JD. Neuroanatomical hypothesis of panic disorder,
reversal by non-opioid drugs. F1000Prime Rep revised. Am J Psychiatry. 2000;157:493505.
2014;6:79. 64. Smoller JW, Gallagher PJ, Duncan LE, McGrath
46. Kalin NH, Shelton SE, Barksdale CM. Opiate modu- LM, Haddad SA, Holmes AJ, et al. The human
lation of separation-induced distress in non-human ortholog of acid-sensing ion channel gene ASIC1a is
primates. Brain Res. 1988;440:28592. associated with panic disorder and amygdala struc-
47. Kalin NH, Shelton SE, Lynn DE. Opiate systems in ture and function. Biol Psychiatry. 2014;76:90210.
mother and infant primates coordinate intimate con- 65. Shekhar A, Keim SR. The circum ventricular organs
tact during reunion. Psychoneuroendocrinology. form a potential neural pathway for lactate sensitiv-
1995;20:73542. ity: implications for panic disorder. J Neurosci.
48. Panksepp J, Herman B, Conner R, Bishop P, Scott 1997;17:972635.
JP. The biology of social attachments: opiates alleviate 66. Johnson PL, Federici LM, Shekhar A. Etiology, trig-
separation distress. Biol Psychiatry. 1978;13:60718. gers and neurochemical circuits associated with
49. Panksepp J, Herman BH, Vilberg T, Bishop P, unexpected, expected, and laboratory-induced panic
DeEskinazi FG. Endogenous opioids and social attacks. Neurosci Biobehav Rev. 2014;46:42954.
behavior. Neurosci Biobehav Rev. 1980;4:47387. 67. Gorman JM, Liebowitz MR, Fyer AJ, Stein J. A neu-
50. Deakin JFW, Graeff FG. 5-HT and mechanisms of roanatomical hypothesis for panic disorder. Am
defence. J Psychopharmacol. 1991;5:30515. J Psychiatry. 1989;146:14861.
51. Graeff FG. Serotonin, the periaqueductal gray and 68. Blanchard DC, Griebel G, Blanchard RJ. The mouse
panic. Neurosci Biobehav Rev. 2004;28:23959. defense test battery: pharmacological and behavioral
52. Canteras NS, Graeff FG. Executive and modulatory neu- assays for anxiety and panic. Eur J Pharmacol.
ral circuits of defensive reactions: implications for panic 2003;463:97116.
disorder. Neurosci Biobehav Rev. 2014;46:35264. 69. Beitman BD, Basha I, Flaker G, DeRosear L,
53. Goetz RR, Klein DF, Gorman JM. Symptoms essen- Mukerji V, Lamberti J. Non-fearful panic disorder:
tial to the experience of sodium lactate-induced panic. panic attacks without fear. Behav Res Ther. 1987;25:
Neuropsychopharmacology. 1996;14:35566. 48792.
54. Dillon DJ, Gorman JM, Liebowitz MR, Fyer AJ, 70. Fleet RP, Martel JP, Lavoie KL, Dupuis G, Beitman
Klein DF. Measurement of lactate-induced panic and BD. Non-fearful panic disorder: a variant of panic in
anxiety. Psychiatry Res. 1987;20:97105. medical patients? Psychosomatics. 2000;41(4):31120.
55. Shioiri T, Someya T, Murashita J, Takahashi S. The 71. Fleet RP, Lavoie KL, Martel JP, Dupuis G, Marchand
symptom structure of panic disorder: a trial using A, Beitman BD. Two-year follow-up status of emer-
64 L.C. Schenberg

gency department patients with chest pain: was it normoxic and hyperoxic hypercapnia. J Comp
panic disorder? CJEM. 2003;5:24754. Neurol. 1997;388:16990.
72. Figueiredo HF, Bodie BL, Tauchi M, Dolgas CM, 87. Nutt DJ. Altered central alpha 2-adrenoceptor sensi-
Herman JP. Stress integration after acute and chronic tivity in panic disorder. Arch Gen Psychiatry.
predator stress: differential activation of central 1989;46:1659.
stress circuitry and sensitization of the hypothalamo- 88. Glue P, Nutt DJ. Benzodiazepine receptor sensitivity
pituitary-adrenocortical axis. Endocrinology. 2003; in panic disorder. Lancet. 1991;337:563.
144(12):524958. 89. Muskin PR, Fyer AJ. Treatment of panic disorder.
73. Hauger RL, Millan MA, Lorang M, Harwood JP, J Clin Psychopharmacol. 1981;1:8190.
Aguilera G. Corticotropin-releasing factor receptors 90. Reiman EM, Raichle ME, Butler FK, Herscovitch P,
and pituitary adrenal responses during immobiliza- Robins E. A focal brain abnormality in panic disorder,
tion stress. Endocrinology. 1988;123:396405. a severe form of anxiety. Nature. 1984;310:6835.
74. Fredrikson M, Sundin O, Frankenhaeuser M. Cortisol 91. Reiman EM, Raichle ME, Robins E, Butler FK,
excretion during the defense reaction in humans. Herscovitch P, Fox P, et al. The application of posi-
Psychosom Med. 1985;47:3139. tron emission tomography to the study of panic dis-
75. Furlan PM, DeMartinis N, Schweizer E, Rickels K, order. Am J Psychiatry. 1986;143:46977.
Lucki I. Abnormal salivary cortisol levels in social 92. Gray JA, McNaughton N. The neuropsychology of
phobic patients in response to acute psychological but anxiety. 2nd ed. Oxford: Oxford Medical
not physical stress. Biol Psychiatry. 2001;50:2549. Publications; 2000.
76. Wiest G, Lehner-Baumgartner E, Baumgartner 93. Kaplan JS, Arnkoff DB, Glass CR, Tinsley R, Geraci
C. Panic attacks in an individual with bilateral selec- M, Hernandez E, et al. Avoidant coping in panic dis-
tive lesions of the amygdala. Arch Neurol. 2006;63: order: a yohimbine biological challenge study.
1798801. Anxiety Stress Coping. 2012;25:42542.
77. Feinstein JS, Buzza C, Hurlemann R, Follmer RL, 94. Kaitin KI, Bliwise DL, Gleason C, Nino-Murcia G,
Dahdaleh NS, Coryell WH, et al. Fear and panic in Dement WC, Libet B. Sleep disturbance produced
humans with bilateral amygdala damage. Nat by electrical stimulation of the locus coeruleus in a
Neurosci. 2013;16:2702. human subject. Biol Psychiatry. 1986;21:7106.
78. Wilent WB, Oh MY, Buetefisch CM, Bailes JE, 95. Libet B, Gleason CA. The human locus coeruleus
Cantella D, Angle C, et al. Induction of panic attack and anxiogenesis. Brain Res. 1994;634:17880.
by stimulation of the ventromedial hypothalamus. 96. Ledoux JE, Iwata J, Cicchetti P, Reis DJ. Different
J Neurosurg. 2010;112:12958. projections of the central amygdaloid nucleus medi-
79. Charney DS, Redmond DE. Neurobiological mecha- ate autonomic and behavioral correlates of condi-
nism in human anxiety: evidence supporting central tioned fear. J Neurosci. 1988;8:251729.
noradrenergic hyperactivity. Neuropharmacology. 97. Ledoux JE, Cicchetti P, Xagoraris A, Romanski
1983;22:15316. LM. The lateral amygdaloid nucleus: sensory inter-
80. Charney DS, Heninger GR, Redmond Jr face of the amygdala in fear conditioning. J Neurosci.
DE. Yohimbine induced anxiety and increased nor- 1990;10:10629.
adrenergic function in humans: effects of diazepam 98. Hitchcock J, Davis M. Lesions of the amygdala, but
and clonidine. Life Sci. 1983;33:1929. not of the cerebellum or red nucleus, block condi-
81. Charney DS, Woods SW, Goodman WK, Heninger tioned fear as measured with the potentiated startle
GR. Neurobiological mechanisms of panic anxiety: bio- paradigm. Behav Neurosci. 1986;100:1122.
chemical and behavioral correlates of yohimbine-induced 99. Nashold Jr BS, Wilson WP, Slaughter DG. Sensations
panic attacks. Am J Psychiatry. 1987;144:10306. evoked by stimulation in the midbrain of man.
82. Elam M, Yao T, Thoren P, Svensson TH. Hypercapnia J Neurosurg. 1969;30:1424.
and hypoxia: chemoreceptor-mediated control of 100. Amano K, Tanikawa T, Iseki H, Kawabatake H,
locus coeruleus neurons and splanchnic, sympa- Notani M, Kawamura H, et al. Single neuron analy-
thetic nerves. Brain Res. 1981;222:37381. sis of the human midbrain tegmentum. Appl
83. Liebowitz MR, Fyer AJ, McGrath PA, Klein Neurophysiol. 1978;41:6678.
DF. Clonidine treatment of panic disorder. 101. Young RF. Brain and spinal stimulation: how and to
Psychopharmacol Bull. 1981;17:1223. whom! Clin Neurosurg. 1989;35:42947.
84. Hoehn-Saric R, Merchant AF, Keyser ML, Smith 102. Fernandez de Molina A, Hunsperger RW. Organization
VK. Effects of clonidine on anxiety disorders. Arch of the subcortical system governing defence and flight
Gen Psychiatry. 1981;38:127882. reactions in the cat. J Physiol. 1962;160:20013.
85. Pineda J, Aghajanian GK. Carbon dioxide regulates the 103. Adams DB. Brain mechanisms for offense, defense
tonic activity of locus coeruleus neurons by modulating and submission. Behav Brain Sci. 1979;2:20141.
a proton- and polyamine-sensitive inward rectifier 104. Hilton SM, Redfern WS. A search for brain stem cell
potassium current. Neuroscience. 1997;77:72343. groups integrating the defence reaction in the rat.
86. Teppema LJ, Veening JG, Kranenburg A, Dahan A, J Physiol. 1986;378:21328.
Berkenbosch A, Olievier C. Expression of c-fos in 105. Graeff FG. The anti-aversive action of drugs. In:
the rat brainstem after exposure to hypoxia and to Thompson T, Dews PB, Barret JE, editors. 6th ed.
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 65

Advances in Behavioral Pharmacology: Volume 6: 119. Schlegel S, Steinert H, Bockisch A, Hahn K,

Neurobehavioral Pharmacology (v. 6). Hillsdale: Schloesser R, Benkert O. Decreased benzodiazepine
Lawrence Erlbaum Associates Inc.; 1987 p. 12956. receptor binding in panic disorder measured by
106. Graeff FG. Animal models of aversion. In: Simon P, IOMAZENIL-SPECT. A preliminary report. Eur
Soubrie P, Wildlocher D, editors. Selected models of Arch Psychiatry Clin Neurosci. 1994;244:4951.
anxiety, depression and psychosis. 1st ed. Basel: 120. Preter M, Lee SH, Petkova E, Vannucci M, Kim S,
Karger AG; 1988. p. 11541. Klein DF. Controlled cross-over study in normal
107. Paul ED, Lowry CA. Functional topography of seroto- subjects of naloxone-preceding-lactate infusions;
nergic systems supports the Deakin/Graeff hypothesis respiratory and subjective responses: relationship to
of anxiety and affective disorders. J Psychopharmacol. endogenous opioid system, suffocation false alarm
2013;27:1090106. theory and childhood parental loss. Psychol Med.
108. Paul ED, Johnson PL, Shekhar A, Lowry CA. The 2011;41:38593.
Deakin/Graeff hypothesis: focus on serotonergic inhibi- 121. Roncon CM, Biesdorf C, Santana RG, Zangrossi Jr
tion of panic. Neurosci Biobehav Rev. 2014;46:37996. H, Graeff FG, Audi EA. The panicolytic-like effect
109. Tanaka KF, Samuels BA, Hen R. Serotonin receptor of fluoxetine in the elevated T-maze is mediated by
expression along the dorsal-ventral axis of mouse serotonin-induced activation of endogenous opioids
hippocampus. Philos Trans R Soc Lond B Biol Sci. in the dorsal periaqueductal grey. J Psychopharmacol.
2012;367:2395401. 2012;26:52531.
110. Pitknen A, Pikkarainen M, Nurminen N, Ylinen 122. Roncon CM, Biesdorf C, Coimbra NC, Audi EA,
A. Reciprocal connections between the amygdala Zangrossi Jr H, Graeff FG. Cooperative regulation
and the hippocampal formation, perirhinal cortex, of anxiety and panic-related defensive behaviors
and postrhinal cortex in rat. A review. Ann N Y Acad in the rat periaqueductal grey matter by 5-HT1A
Sci. 2000;911:36991. and mu-receptors. J Psychopharmacol. 2013;27:
111. Nutt DJ, Glue P, Lawson C, Wilson S. Flumazenil 11418.
provocation of panic attacks. Evidence for altered 123. Breuer J, Freud S. Studien ber Hysterie. Vienna:
benzodiazepine receptor sensitivity in panic disor- Franz Deudicke; 1895.
der. Arch Gen Psychiatry. 1990;47:91725. 124. Bowlby J. A secure base: parent-child attachment
112. Barnard EA, Skolnick P, Olsen RW, Mohler H, and healthy human development. New York: Basic
Sieghart W, Biggio G, et al. International union of Books; 1988.
pharmacology: XV. Subtypes of gamma-aminobutyric 125. Heim C, Shugart M, Craighead WE, Nemeroff
acid-A receptors: classification on the basis of subunit CB. Neurobiological and psychiatric consequences
structure and receptor function. Pharmacol Rev. of child abuse and neglect. Dev Psychobiol. 2010;52:
1998;50:291313. 67190.
113. Olsen RW, Sieghart W. International union of phar- 126. Ladd CO, Owens MJ, Nemeroff CB. Persistent
macology: LXX. Subtypes of gamma-aminobutyric changes in corticotropin-releasing factor neuronal sys-
acid(A) receptors: classification on the basis of sub- tems induced by maternal deprivation. Endocrinology.
unit composition, pharmacology, and function. 1996;137:12128.
Update. Pharmacol Rev. 2008;60:24360. 127. Ladd CO, Huot RL, Thrivikraman KV, Nemeroff
114. Strhle A, Kellner M, Yassouridis A, Holsboer F, CB, Plotsky PM. Long-term adaptations in gluco-
Wiedemann K. Effect of flumazenil in lactate- corticoid receptor and mineralocorticoid receptor
sensitive patients with panic disorder. Am J Psychiatry. mRNA and negative feedback on the hypothalamo-
1998;155:6102. pituitary-adrenal axis following neonatal maternal
115. Strhle A, Kellner M, Holsboer F, Wiedemann separation. Biol Psychiatry. 2004;55:36775.
K. Behavioral, neuroendocrine, and cardiovascular 128. Rots NY, de Jong J, Workel JO, Levine S, Cools
response to flumazenil: no evidence for an altered AR, de Kloet ER. Neonatal maternally deprived rats
benzodiazepine receptor sensitivity in panic disor- have as adults elevated basal pituitary-adrenal activ-
der. Biol Psychiatry. 1999;45:3216. ity and enhanced susceptibility to apomorphine.
116. Potokar J, Lawson C, Wilson S, Nutt D. Behavioral, J Neuroendocrinol. 1996;8:5016.
neuroendocrine, and cardiovascular response to flu- 129. Anisman H, Zaharia MD, Meaney MJ, Merali Z. Do
mazenil: no evidence for an altered benzodiazepine early-life events permanently alter behavioral and
receptor sensitivity in panic disorder (Comment on). hormonal responses to stressors? Int J Dev Neurosci.
Biol Psychiatry. 1999;46:170911. 1998;16:14964.
117. Kaschka W, Feistel H, Ebert D. Reduced benzodiaz- 130. Biagini G, Pich EM, Carani C, Marrama P, Agnati
epine receptor binding in panic disorders measured by LF. Postnatal maternal separation during the stress
iomazenil SPECT. J Psychiatr Res. 1995;29:42734. hyporesponsive period enhances the adrenocortical
118. Malizia AL, Cunningham VJ, Bell CJ, Liddle PF, response to novelty in adult rats by affecting feed-
Jones T, Nutt DJ. Decreased brain GABA(A)- back regulation in the CA1 hippocampal field. Int
benzodiazepine receptor binding in panic disorder: J Dev Neurosci. 1998;16:18797.
preliminary results from a quantitative PET study. 131. van Oers HJ, de Kloet ER, Levine S. Early vs. late
Arch Gen Psychiatry. 1998;55:71520. maternal deprivation differentially alters the endo-
66 L.C. Schenberg

crine and hypothalamic responses to stress. Brain 146. Strhle A, Feller C, Onken M, Godemann F, Heinz
Res Dev Brain Res. 1998;111:24552. A, Dimeo F. The acute antipanic activity of aerobic
132. Kinkead R, Gulemetova R. Neonatal maternal sepa- exercise. Am J Psychiatry. 2005;162:23768.
ration and neuroendocrine programming of the res- 147. Beck JG, Ohtake PJ, Shipherd JC. Exaggerated anxi-
piratory control system in rats. Biol Psychol. 2010; ety is not unique to CO2 in panic disorder: a com-
84:2638. parison of hypercapnic and hypoxic challenges.
133. Meaney MJ, Diorio J, Francis D, Weaver S, Yau J, J Abnorm Psychol. 1999;108:47382.
Chapman K, et al. Postnatal handling increases the 148. Hackett PH, Roach RC. High-altitude illness. N
expression of cAMP-inducible transcription factors Engl J Med. 2001;345:10714.
in the rat hippocampus: the effects of thyroid hor- 149. Fagenholz PJ, Murray AF, Gutman JA, Findley JK,
mones and serotonin. J Neurosci. 2000;20:392635. Harris NS. New-onset anxiety disorders at high alti-
134. Gardner KL, Thrivikraman KV, Lightman SL, tude. Wilderness Environ Med. 2007;18:3126.
Plotsky PM, Lowry CA. Early life experience alters 150. Tweed JL, Schoenbach VJ, George LK, Blazer
behavior during social defeat: focus on serotonergic DG. The effects of childhood parental death and
systems. Neuroscience. 2005;136:18191. divorce on six-month history of anxiety disorders. Br
135. Gardner KL, Hale MW, Lightman SL, Plotsky PM, J Psychiatry. 1989;154:8238.
Lowry CA. Adverse early life experience and social stress 151. Stein MB, Walker JR, Anderson G, Hazen AL, Ross
during adulthood interact to increase serotonin trans- CA, Eldridge G, et al. Childhood physical and sexual
porter mRNA expression. Brain Res. 2009;1305:4763. abuse in patients with anxiety disorders and in a com-
136. Gardner KL, Hale MW, Oldfield S, Lightman SL, munity sample. Am J Psychiatry. 1996;153:2757.
Plotsky PM, Lowry CA. Adverse experience during 152. Battaglia M, Pesenti-Gritti P, Medland SE, Ogliari
early life and adulthood interact to elevate tph2 mRNA A, Tambs K, Spatola CA. A genetically informed
expression in serotonergic neurons within the dorsal study of the association between childhood separa-
raphe nucleus. Neuroscience. 2009;163:9911001. tion anxiety, sensitivity to CO(2), panic disorder, and
137. Quintino-Dos-Santos JW, Muller CJ, Bernabe CS, the effect of childhood parental loss. Arch Gen
Rosa CA, Tufik S, Schenberg LC. Evidence that the Psychiatry. 2009;66:6471.
periaqueductal gray matter mediates the facilitation 153. Spatola CA, Scaini S, Pesenti-Gritti P, Medland SE,
of panic-like reactions in neonatally-isolated adult Moruzzi S, Ogliari A, et al. Gene-environment inter-
rats. PLoS One. 2014;9:e90726. actions in panic disorder and CO(2) sensitivity:
138. Dumont FS, Biancardi V, Kinkead R. Hypercapnic effects of events occurring early in life. Am J Med
ventilatory response of anesthetized female rats sub- Genet B Neuropsychiatr Genet. 2011;156B:7988.
jected to neonatal maternal separation: insight into the 154. Angst J, Wicki W. The epidemiology of frequent and
origins of panic attacks? Respir Physiol Neurobiol. less frequent panic attacks. In: Montgomery SA, edi-
2011;175:28895. tor. Psychopharmacology of panic. New York:
139. Grosz HJ, Farmer BB. Pitts and McClures lactate- Oxford University Press; 1993. p. 724.
anxiety study revisited. Br J Psychiatry. 1972;120: 155. Lovick TA. Sex determinants of experimental panic
4158. attacks. Neurosci Biobehav Rev. 2014;46:46571.
140. Gorman JM, Battista D, Goetz RR, Dillon DJ, 156. Nashold Jr BS, Wilson WP, Slaughter GS. The mid-
Liebowitz MR, Fyer AJ, et al. A comparison of brain and pain. In: Bonica JJ, editor. International sym-
sodium bicarbonate and sodium lactate infusion in posium on pain. New York: Raven; 1974. p. 1916.
the induction of panic attacks [published erratum 157. Kumar K, Toth C, Nath RK. Deep brain stimulation
appears in Arch Gen Psychiatry 1991;48:772]. Arch for intractable pain: a 15-year experience.
Gen Psychiatry. 1989;46:14550. Neurosurgery. 1997;40:73646.
141. Gorman JM, Goetz RR, Dillon D, Liebowitz MR, 158. Bandler R, Depaulis A. Midbrain periaqueductal
Fyer AJ, Davies S, et al. Sodium D-lactate infusion gray control of defensive behavior in the cat and rat.
of panic disorder patients. Neuropsychopharmacology. In: Depaulis A, Bandler R, editors. The midbrain
1990;3:1819. periaqueductal gray matter. New York: Plenum
142. Ewaschuk JB, Naylor JM, Zello GA. D-Lactate in Press; 1991. p. 17598.
human and ruminant metabolism. J Nutr. 2005;135: 159. Bandler R, Keay KA. Columnar organization in the
161925. midbrain periaqueductal gray and the integration of
143. Dager SR, Rainey JM, Kenny MA, Artru AA, emotional expression. Prog Brain Res. 1996;107:
Metzger GD, Bowden DM. Central nervous system 285300.
effects of lactate infusion in primates. Biol 160. Carrive P. Functional organization of PAG neurons
Psychiatry. 1990;27:193204. controlling regional vascular beds. In: Depaulis A,
144. Dager SR, Marro KI, Richards TL, Metzger Bandler R, editors. The midbrain periaqueductal
GD. Localized magnetic resonance spectroscopy mea- gray matter: functional, anatomical, and neurochem-
surement of brain lactate during intravenous lactate infu- ical organization. New York: Plenum Press; 1991.
sion in healthy volunteers. Life Sci. 1992;51:97385. p. 67100.
145. Stein JM, Papp LA, Klein DF, Cohen S, Simon J, 161. Carrive P. The periaqueductal gray and defensive
Ross D, et al. Exercise tolerance in panic disorder behavior: functional representation and neuronal
patients. Biol Psychiatry. 1992;32:2817. organization. Behav Brain Res. 1993;58:2747.
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 67

162. Keay KA, Bandler R. Periaqueductal gray. In: 175. Lovick TA. Stimulation in the ventral periaqueductal
Paxinos G, editor. The rat nervous system. 3rd ed. grey matter modulates the cardiovascular response
San Diego: Elsevier; 2004. p. 24357. evoked from the midbrain defence area in anaesthe-
163. Kingsbury MA, Kelly AM, Schrock SE, Goodson tized rats. J Physiol. 1990;91P.
JL. Mammal-like organization of the avian midbrain 176. Lovick TA. Inhibitory modulation of the cardiovascu-
central gray and a reappraisal of the intercollicular lar defence response by ventrolateral periaqueductal
nucleus. PLoS One. 2011;6:e20720. grey matter in rats. Exp Brain Res. 1992;89:1339.
164. Jansen AS, Farkas E, Mac SJ, Loewy AD. Local 177. Lovick TA. Influence of the dorsal and median raphe
connections between the columns of the periaque- nuclei on neurons in the periaqueductal gray matter:
ductal gray matter: a case for intrinsic neuromodula- role of 5-hydroxytryptamine. Neuroscience. 1994;59:
tion. Brain Res. 1998;784:32936. 9931000.
165. Bittencourt AS, Carobrez AP, Zamprogno LP, Tufik 178. Lovick TA, Parry DM, Stezhka VV, Lumb
S, Schenberg LC. Organization of single compo- BM. Serotonergic transmission in the periaqueductal
nents of defensive behaviors within distinct columns gray matter in relation to aversive behaviour: mor-
of periaqueductal gray matter of the rat: role of phological evidence for direct modulatory effects on
N-methyl-D-aspartic acid glutamate receptors. identified output neurons. Neuroscience. 2000;95:
Neuroscience. 2004;125:7189. 76372.
166. Teixeira KV, Carobrez AP. Effects of glycine or 179. Pobbe RL, Zangrossi Jr H. 5-HT(1A) and 5-HT(2A)
(+/)-3-amino-1-hydroxy-2-pyrrolidone microinjec- receptors in the rat dorsal periaqueductal gray mediate
tions along the rostrocaudal axis of the dorsal periaque- the antipanic-like effect induced by the stimulation of
ductal gray matter on rats performance in the elevated serotonergic neurons in the dorsal raphe nucleus.
plus-maze task. Behav Neurosci. 1999;113:196203. Psychopharmacology (Berl). 2005;183:31421.
167. Mota-Ortiz SR, Sukikara MH, Felicio LF, Canteras 180. Schenberg LC, Graeff FG. Role of the periaqueductal
NS. Afferent connections to the rostrolateral part of gray substance in the antianxiety action of benzodiaz-
the periaqueductal gray: a critical region influencing epines. Pharmacol Biochem Behav. 1978;9:28795.
the motivation drive to hunt and forage. Neural Plast. 181. Spyer KM, Gourine AV. Chemosensory pathways in
2009; ID612698. the brainstem controlling cardiorespiratory activity.
168. Holstege G, Kerstens L, Moes MC, Vanderhorst Philos Trans R Soc Lond B Biol Sci. 2009;364:
VG. Evidence for a periaqueductal gray-nucleus 260310.
retroambiguus-spinal cord pathway in the rat. 182. Finley JC, Katz DM. The central organization of
Neuroscience. 1997;80:58798. carotid body afferent projections to the brainstem of
169. Ruiz-Torner A, Olucha-Bordonau F, Valverde- the rat. Brain Res. 1992;572(12):10816.
Navarro AA, Martinez-Soriano F. The chemical 183. Loeschcke HH, Koepchen HP, Gertz KH. ber den
architecture of the rat's periaqueductal gray based on Einflub von Wasserstoffionenkonzentration und
acetylcholinesterase histochemistry: a quantitative CO2-Druck im Liquor cerebrospinalis auf die
and qualitative study. J Chem Neuroanat. 2001;21: Atmung. Pflugers Arch. 1958;266:56985.
295312. 184. Loeschcke HH. Central chemosensitivity and the
170. Onstott D, Mayer B, Beitz AJ. Nitric oxide syn- reaction theory. J Physiol. 1982;332:124.
thase immunoreactive neurons anatomically define 185. Schlaefke ME, See WR, Loeschcke HH. Ventilatory
a longitudinal dorsolateral column within the mid- response to alterations of H+ ion concentration in
brain periaqueductal gray of the rat: analysis using small areas of the ventral medullary surface. Respir
laser confocal microscopy. Brain Res. 1993;610: Physiol. 1970;10:198212.
31724. 186. Schlaefke ME. Central chemosensitivity: a respira-
171. Bandler R, Tork I. Midbrain periaqueductal grey region tory drive. Rev Physiol Biochem Pharmacol. 1981;90:
in the cat has afferent and efferent connection with soli- 171244.
tary tract nuclei. Neurosci Lett. 1987;74:16. 187. Ricardo JA, Koh ET. Anatomical evidence of direct
172. Lv BC, Ji GL, Huo FQ, Chen T, Li H, Li projections from the nucleus of the solitary tract to
YQ. Topographical distributions of endomorphinergic the hypothalamus, amygdala, and other forebrain
pathways from nucleus tractus solitarii to periaqueduc- structures in the rat. Brain Res. 1978;153:126.
tal gray in the rat. J Chem Neuroanat. 2010;39:16674. 188. Luft U. Aviation physiology-the effects of altitude.
173. Iwata J, Ledoux JE, Reis DJ. Destruction of intrinsic In: Handbook of physiology. Washington, DC: Am.
neurons in the lateral hypothalamus disrupts the Physiol. Soc.; 1965. p. 1099145.
classical conditioning of autonomic but not behav- 189. Moosavi SH, Golestanian E, Binks AP, Lansing RW,
ioral emotional responses in the rat. Brain Res. Brown R, Banzett RB. Hypoxic and hypercapnic
1986;368:1616. drives to breathe generate equivalent levels of air
174. Kiser RS, Brown CA, Sanghera MK, German hunger in humans. J Appl Physiol. 2003;94:14154.
DC. Dorsal raphe nucleus stimulation reduces cen- 190. Banzett RB, Lansing RW, Evans KC, Shea SA. Stimulus-
trally elicited fearlike behavior. Brain Res. 1980;191: response characteristics of CO2-induced air hunger in
26572. normal subjects. Respir Physiol. 1996;103:1931.
68 L.C. Schenberg

191. Banzett RB, Lansing RW, Reid MB, Adams L, 207. Drevets WC, Videen TQ, MacLeod AK, Haller JW,
Brown R. Air hunger arising from increased PCO2 Raichle ME. PET images of blood flow changes dur-
in mechanically ventilated quadriplegics. Respir ing anxiety: correction. Science. 1992;256:1696.
Physiol. 1989;76:5367. 208. Benkelfat C, Bradwejn J, Meyer E, Ellenbogen M,
192. Gandevia SC, Killian K, McKenzie DK, Crawford Milot S, Gjedde A, et al. Functional neuroanatomy
M, Allen GM, Gorman RB, et al. Respiratory sensa- of CCK4-induced anxiety in normal healthy volun-
tions, cardiovascular control, kinaesthesia and tran- teers. Am J Psychiatry. 1995;152:11804.
scranial stimulation during paralysis in humans. 209. Javanmard M, Shlik J, Kennedy SH, Vaccarino FJ,
J Physiol. 1993;470:85107. Houle S, Bradwejn J. Neuroanatomic correlates of
193. van den Hout MA, Boek C, van der Molen GM, CCK-4-induced panic attacks in healthy humans: a
Jansen A, Griez E. Rebreathing to cope with hyper- comparison of two time points. Biol Psychiatry.
ventilation: experimental tests of the paper bag 1999;45:87282.
method. J Behav Med. 1988;11:30310. 210. Reiman EM, Raichle ME, Robins E, Mintun MA,
194. Moosavi SH, Banzett RB, Butler JP. Time course of Fusselman MJ, Fox PT, et al. Neuroanatomical cor-
air hunger mirrors the biphasic ventilatory response relates of a lactate-induced anxiety attack. Arch Gen
to hypoxia. J Appl Physiol. 2004;97:2098103. Psychiatry. 1989;46:493500.
195. Spengler CM, Banzett RB, Systrom DM, Shannon 211. Goossens L, Leibold N, Peeters R, Esquivel G, Knuts
DC, Shea SA. Respiratory sensations during heavy I, Backes W, et al. Brainstem response to hypercap-
exercise in subjects without respiratory chemosensi- nia: a symptom provocation study into the patho-
tivity. Respir Physiol. 1998;114:6574. physiology of panic disorder. J Psychopharmacol.
196. Shea SA, Andres LP, Shannon DC, Guz A, Banzett 2014;28:44956.
RB. Respiratory sensations in subjects who lack a 212. Brust RD, Corcoran AE, Richerson GB, Nattie E,
ventilatory response to CO2. Respir Physiol. Dymecki SM. Functional and developmental identi-
1993;93:20319. fication of a molecular subtype of brain serotonergic
197. Banzett RB, Mulnier HE, Murphy K, Rosen SD, neuron specialized to regulate breathing dynamics.
Wise RJ, Adams L. Breathlessness in humans acti- Cell Rep. 2014;9:215265.
vates insular cortex. Neuroreport. 2000;11:211720. 213. Bett K, Sandkuhler J. Map of spinal neurons acti-
198. Craig AD. How do you feel? Interoception: the sense vated by chemical stimulation in the nucleus raphe
of the physiological condition of the body. Nat Rev magnus of the unanesthetized rat. Neuroscience.
Neurosci. 2002;3:65566. 1995;67:497504.
199. Craig AD. Human feelings: why are some more aware 214. Schenberg LC, Lovick TA. Attenuation of the
than others? Trends Cogn Sci. 2004;8:23941. midbrain-evoked defense reaction by selective stim-
200. Casanova JP, Contreras M, Moya EA, Torrealba F, ulation of medullary raphe neurons in rats. Am
Iturriaga R. Effect of insular cortex inactivation on J Physiol. 1995;269:R137889.
autonomic and behavioral responses to acute 215. Beitz AJ, Clements JR, Mullett MA, Ecklund
hypoxia in conscious rats. Behav Brain Res. 2013; LJ. Differential origin of brainstem serotoninergic
253:607. projections to the midbrain periaqueductal gray and
201. Prabhakar NR. Sensing hypoxia: physiology, genet- superior colliculus of the rat. J Comp Neurol.
ics and epigenetics. J Physiol. 2013;591:224557. 1986;250:498509.
202. Katzman MA, Struzik L, Vijay N, Coonerty- 216. Fischer H, Andersson JL, Furmark T, Fredrikson
Femiano A, Mahamed S, Duffin J. Central and M. Brain correlates of an unexpected panic attack: a
peripheral chemoreflexes in panic disorder. human positron emission tomographic study.
Psychiatry Res. 2002;113:18192. Neurosci Lett. 1998;251:13740.
203. Corfield DR, Fink GR, Ramsay SC, Murphy K, 217. Sato M, Severinghaus JW, Basbaum AI. Medullary
Harty HR, Watson JD, et al. Evidence for limbic sys- CO2 chemoreceptor neuron identification by c-fos
tem activation during CO2-stimulated breathing in immunocytochemistry. J Appl Physiol. 1992;73:
man. J Physiol. 1995;488:7784. 96100.
204. Brannan S, Liotti M, Egan G, Shade R, Madden L, 218. Erickson JT, Millhorn DE. Hypoxia and electrical
Robillard R, et al. Neuroimaging of cerebral activa- stimulation of the carotid sinus nerve induce Fos-
tions and deactivations associated with hypercapnia like immunoreactivity within catecholaminergic and
and hunger for air. Proc Natl Acad Sci U S A. serotoninergic neurons of the rat brainstem. J Comp
2001;98:202934. Neurol. 1994;348:16182.
205. Liotti M, Brannan S, Egan G, Shade R, Madden L, 219. Haxhiu MA, Yung K, Erokwu B, Cherniack NS. CO2-
Abplanalp B, et al. Brain responses associated with induced c-fos expression in the CNS catecholaminer-
consciousness of breathlessness (air hunger). Proc gic neurons. Respir Physiol. 1996;105:3545.
Natl Acad Sci U S A. 2001;98:203540. 220. Hirooka Y, Polson JW, Potts PD, Dampney
206. Reiman EM, Fusselman MJ, Fox PT, Raichle RA. Hypoxia-induced Fos expression in neurons
ME. Neuroanatomical correlates of anticipatory projecting to the pressor region in the rostral ventro-
anxiety. Science. 1989;243:10714. lateral medulla. Neuroscience. 1997;80:120924.
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 69

221. Larnicol N, Wallois F, Berquin P, Gros F, Rose D. 237. Coates EL, Li A, Nattie EE. Widespread sites of
c-fos-like immunoreactivity in the cats neuraxis fol- brain stem ventilatory chemoreceptors. J Appl
lowing moderate hypoxia or hypercapnia. J Physiol Physiol. 1993;75:514.
Paris. 1994;88:818. 238. Kramer JM, Nolan PC, Waldrop TG. In vitro responses
222. Berquin P, Bodineau L, Gros F, Larnicol of neurons in the periaqueductal gray to hypoxia and
N. Brainstem and hypothalamic areas involved in hypercapnia. Brain Res. 1999;835:197203.
respiratory chemoreflexes: a Fos study in adult rats. 239. Kim Y, Bang H, Kim D. TASK-3, a new member of
Brain Res. 2000;857:3040. the tandem pore K(+) channel family. J Biol Chem.
223. Johnson PL, Fitz SD, Hollis JH, Moratalla R, 2000;275:93407.
Lightman SL, Shekhar A, et al. Induction of c-Fos in 240. Rajan S, Wischmeyer E, Xin LG, Preisig-Muller R,
panic/defence-related brain circuits following brief Daut J, Karschin A, et al. TASK-3, a novel tandem
hypercarbic gas exposure. J Psychopharmacol. pore domain acid-sensitive K+ channel. An extracel-
2011;25:2636. lular histiding as pH sensor. J Biol Chem.
224. Mitchell RA, Herbert DA. The effect of carbon diox- 2000;275:166507.
ide on the membrane potential of medullary respira- 241. Buckler KJ. TASK-like potassium channels and
tory neurons. Brain Res. 1974;75:3459. oxygen sensing in the carotid body. Respir Physiol
225. Horn EM, Kramer JM, Waldrop TG. Development of Neurobiol. 2007;157:5564.
hypoxia-induced Fos expression in rat caudal hypo- 242. Talley EM, Solorzano G, Lei Q, Kim D, Bayliss
thalamic neurons. Neuroscience. 2000;99:71120. DA. CNS distribution of members of the two-pore-
226. Feldman JL, Mitchell GS, Nattie EE. Breathing: domain (KCNK) potassium channel family.
rhythmicity, plasticity, chemosensitivity. Annu Rev J Neurosci. 2001;21:7491505.
Neurosci. 2003;26:23966. 243. Bayliss DA, Talley EM, Sirois JE, Lei Q. TASK-1 is
227. Raff H, Roarty TP. Renin, ACTH, and aldosterone a highly modulated pH-sensitive leak K(+) channel
during acute hypercapnia and hypoxia in conscious expressed in brainstem respiratory neurons. Respir
rats. Am J Physiol. 1988;254:R4315. Physiol. 2001;129:15974.
228. Argyropoulos SV, Bailey JE, Hood SD, Kendrick AH, 244. Washburn CP, Sirois JE, Talley EM, Guyenet PG,
Rich AS, Laszlo G, et al. Inhalation of 35 % CO(2) Bayliss DA. Serotonergic raphe neurons express
results in activation of the HPA axis in healthy volun- TASK channel transcripts and a TASK-like pH- and
teers. Psychoneuroendocrinology. 2002;27:71529. halothane-sensitive K+ conductance. J Neurosci.
229. Kaye J, Buchanan F, Kendrick A, Johnson P, Lowry 2002;22:125665.
C, Bailey J, et al. Acute carbon dioxide exposure in 245. Washburn CP, Bayliss DA, Guyenet
healthy adults: evaluation of a novel means of inves- PG. Cardiorespiratory neurons of the rat ventrolat-
tigating the stress response. J Neuroendocrinol. eral medulla contain TASK-1 and TASK-3 channel
2004;16:25664. mRNA. Respir Physiol Neurobiol. 2003;138:
230. van Duinen MA, Schruers KR, Maes M, Griez EJ. CO2 1935.
challenge induced HPA axis activation in panic. Int 246. Schimitel FG, de Almeida GM, Pitol DN, Armini
J Neuropsychopharmacol. 2007;10:797804. RS, Tufik S, Schenberg LC. Evidence of a suffoca-
231. Johnson P, Lowry C, Truitt W, Shekhar A. Disruption tion alarm system within the periaqueductal gray
of GABAergic tone in the dorsomedial hypothalamus matter of the rat. Neuroscience. 2012;200:5973.
attenuates responses in a subset of serotonergic neu- 247. Vianna DM, Landeira-Fernandez J, Brando
rons in the dorsal raphe nucleus following lactate- ML. Dorsolateral and ventral regions of the periaq-
induced panic. J Psychopharmacol. 2008;22:64252. ueductal gray matter are involved in distinct types of
232. Ryan JW, Waldrop TG. Hypoxia sensitive neurons in fear. Neurosci Biobehav Rev. 2001;25:7119.
the caudal hypothalamus project to the periaqueduc- 248. Subramanian HH, Balnave RJ, Holstege G. The mid-
tal gray. Respir Physiol. 1995;100:18594. brain periaqueductal gray control of respiration.
233. Shams I, Avivi A, Nevo E. Oxygen and carbon diox- J Neurosci. 2008;28:1227483.
ide fluctuations in burrows of subterranean blind 249. Schenberg LC, Pvoa RMF, Costa AL, Caldellas
mole rats indicate tolerance to hypoxic-hypercapnic AV, Tufik S, Bittencourt AS. Functional specializa-
stresses. Comp Biochem Physiol A Mol Integr tions within the tectum defense systems of the rat.
Physiol. 2005;142:37682. Neurosci Biobehav Rev. 2005;29:127998.
234. Teppema LJ, Veening JG, Berkenbosch A. Expression 250. Steeves JD, Jordan LM. Autoradiographic demon-
of c-fos in the brain stem of rats during hypercapnia. stration of the projections from the mesencephalic
Adv Exp Med Biol. 1995;393:4751. locomotor region. Brain Res. 1984;307:26376.
235. Bodineau L, Larnicol N. Brainstem and hypothalamic 251. Franchini KG, Krieger EM. Cardiovascular
areas activated by tissue hypoxia: Fos-like immunore- responses of conscious rats to carotid body chemore-
activity induced by carbon monoxide inhalation in the ceptor stimulation by intravenous KCN. J Auton
rat. Neuroscience. 2001;108:64353. Nerv Syst. 1993;42:639.
236. Dillon GH, Waldrop TG. In vitro responses of cau- 252. Franchini KG, Oliveira VL, Krieger
dal hypothalamic neurons to hypoxia and hypercap- EM. Hemodynamics of chemoreflex activation in
nia. Neuroscience. 1992;51:94150. unanesthetized rats. Hypertension. 1997;30:699703.
70 L.C. Schenberg

253. Hayward LF, Von Reitzentstein M. c-Fos expression 269. Guyenet PG, Koshiya N. Working model of the sym-
in the midbrain periaqueductal gray after chemore- pathetic chemoreflex in rats. Clin Exp Hypertens.
ceptor and baroreceptor activation. Am J Physiol 1995;17:16779.
Heart Circ Physiol. 2002;283:H197584. 270. Hayward LF, Castellanos M, Davenport
254. Schimitel FG, Muller CJ, Tufik S, Schenberg PW. Parabrachial neurons mediate dorsal periaque-
LC. Evidence of a suffocation alarm system sensitive ductal gray evoked respiratory responses in the rat.
to clinically-effective treatments with the panicolytics J Appl Physiol. 2004;96:114654.
clonazepam and fluoxetine. J Psychopharmacol. 271. Zhang W, Hayward LF, Davenport PW. Respiratory
2014;28:11848. responses elicited by rostral versus caudal dorsal
255. Spiacci Jr A, de Oliveira ST, da Silva GS, Glass ML, periaqueductal gray stimulation in rats. Auton
Schenberg LC, Garcia-Cairasco N, et al. Serotonin Neurosci. 2007;134:4554.
in the dorsal periaqueductal gray inhibits panic-like 272. Haibara AS, Tamashiro E, Olivan MV, Bonagamba
defensive behaviors in rats exposed to acute hypoxia. LG, Machado BH. Involvement of the parabrachial
Neuroscience. 2015;307:1918. nucleus in the pressor response to chemoreflex activa-
256. Suchecki D, Mozaffarian D, Gross G, Rosenfeld P, tion in awake rats. Auton Neurosci. 2002;101:607.
Levine S. Effects of maternal deprivation on the ACTH 273. Adams DB, Baccelli G, Mancia G, Zanchetti
stress response in the infant rat. Neuroendocrinology. A. Relation of cardiovascular changes in fighting to
1993;57:20412. emotion and exercise. J Physiol. 1971;212:32135.
257. Suchecki D, Rosenfeld P, Levine S. Maternal regula- 274. Hilton SM. The defence-arousal system and its rele-
tion of the hypothalamic-pituitary-adrenal axis in the vance for circulatory and respiratory control. J Exp
infant rat: the roles of feeding and stroking. Brain Biol. 1982;100:15974.
Res Dev Brain Res. 1993;75:18592. 275. Lovick TA. Interactions between descending path-
258. Suchecki D, Nelson DY, Van OH, Levine ways from the dorsal and ventrolateral periaqueduc-
S. Activation and inhibition of the hypothalamic- tal gray matter in the rat. In: Depaulis A, Bandler R,
pituitary-adrenal axis of the neonatal rat: effects of editors. The midbrain periaqueductal gray matter.
maternal deprivation. Psychoneuroendocrinology. New York: Plenum Press; 1991. p. 10120.
1995;20:16982. 276. Lovick TA. Midbrain and medullary regulation of
259. Schenberg LC. Towards a translational model of defensive cardiovascular functions. Prog Brain Res.
panic attack. Psychol Neurosci. 2010;3:937. 1996;107:30113.
260. Johnson PL, Lightman SL, Lowry CA. A functional 277. Carobrez AP, Schenberg LC, Graeff
subset of serotonergic neurons in the rat ventrolat- FG. Neuroeffector mechanisms of the defense reac-
eral periaqueductal gray implicated in the inhibition tion in the rat. Physiol Behav. 1983;31:43944.
of sympathoexcitation and panic. Ann N Y Acad 278. Subramanian HH, Holstege G. Stimulation of the
Sci. 2004;1018:5864. midbrain periaqueductal gray modulates pre-
261. Sachs E. On the relation of the optic thalamus to res- inspiratory neurons in the ventrolateral medulla in
piration, circulation, temperature, and the spleen. the in vivo rat. J Comp Neurol. 2013;521:308398.
J Exp Med. 1911;14:40840. 279. Hilton SM. Inhibition of baroreceptor reflexes on
262. Tan ES. Brain-stem regions for stimulus-bound and hypothalamic stimulation. J Physiol.
stimulus-related respiration. Exp Neurol. 1967;17: 1963;165:56P7.
51728. 280. Coote JH, Hilton SM, Zbrozyna AW. The ponto-
263. Paydarfar D, Eldridge FL. Phase resetting and dys- medullary area integrating the defence reaction in
rhythmic responses of the respiratory oscillator. Am the cat and its influence on muscle blood flow.
J Physiol. 1987;252:R5562. J Physiol. 1973;229:25774.
264. Hayward LF, Swartz CL, Davenport PW. Respiratory 281. Schenberg LC, Vasquez EC, da Costa MB. Cardiac
response to activation or disinhibition of the dorsal baroreflex dynamics during the defence reaction in
periaqueductal gray in rats. J Appl Physiol. 2003;94: freely moving rats. Brain Res. 1993;621:508.
91322. 282. Lopes LT, Patrone LG, Bicego KC, Coimbra NC,
265. Cohen MI. Neurogenesis of respiratory rhythm in Gargaglioni LH. Periaqueductal gray matter modu-
the mammal. Physiol Rev. 1979;59:110573. lates the hypercapnic ventilatory response. Pflugers
266. Schenberg LC, de Aguiar JC, Graeff FG. GABA mod- Arch. 2012;464:15566.
ulation of the defense reaction induced by brain electri- 283. Lopes LT, Biancardi V, Vieira EB, Leite-Panissi C,
cal stimulation. Physiol Behav. 1983;31:42937. Bicego KC, Gargaglioni LH. Participation of the
267. Bizzi E, Libretti A, Malliani A, Zanchetti A. Reflex dorsal periaqueductal grey matter in the hypoxic
chemoceptive excitation of diencephalic sham rage ventilatory response in unanaesthetized rats. Acta
behavior. Am J Physiol. 1961;200:9236. Physiol (Oxf). 2014;211:52837.
268. Hilton SM, Joels N. Facilitation of chemoreceptor 284. Krout KE, Jansen AS, Loewy AD. Periaqueductal
reflexes during the defence reaction. J Physiol. gray matter projection to the parabrachial nucleus in
1965;176:202. rat. J Comp Neurol. 1998;401:43754.
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 71

285. Hayward LF, Castellanos M. Increased c-Fos expres- 300. Mennicken F, Hoffert C, Pelletier M, Ahmad S,
sion in select lateral parabrachial subnuclei following O'Donnell D. Restricted distribution of galanin recep-
chemical versus electrical stimulation of the dorsal peri- tor 3 (GalR3) mRNA in the adult rat central nervous
aqueductal gray in rats. Brain Res. 2003;974:15366. system. J Chem Neuroanat. 2002;24:25768.
286. Potts JT, Rybak IA, Paton JF. Respiratory rhythm 301. Harding A, Paxinos G, Halliday G. The serotonin
entrainment by somatic afferent stimulation. and tachykinin systems. In: Paxinos G, editor. The
J Neurosci. 2005;25:196578. rat nervous system. 3rd ed. San Diego: Elsevier;
287. Orwin A. The running treatment: a preliminary 2004. p. 120556.
communication on a new use for an old therapy 302. Darwinkel A, Stanic D, Booth LC, May CN,
(physical activity) in the agoraphobic syndrome. Br Lawrence AJ, Yao ST. Distribution of orexin-1 recep-
J Psychiatry. 1973;122:1759. tor-green fluorescent protein- (OX1-GFP) expressing
288. Sun MK, Reis DJ. Hypoxia-activated Ca2+ currents neurons in the mouse brain stem and pons: co-local-
in pacemaker neurones of rat rostral ventrolateral ization with tyrosine hydroxylase and neuronal nitric
medulla in vitro. J Physiol. 1994;476:10116. oxide synthase. Neuroscience. 2014;278:25364.
289. Pascual O, Morin-Surun MP, Barna B, Deavit-Saubie 303. Jenck F, Moreau JL, Martin JR. Dorsal periaqueduc-
M, Pequignot JM, Champagnat J. Progesterone tal gray-induced aversion as a simulation of panic
reverses the neuronal responses to hypoxia in rat anxiety: elements of face and predictive validity.
nucleus tractus solitarius in vitro. J Physiol. 2002;544: Psychiatry Res. 1995;57:18191.
51120. 304. Brando ML, de Aguiar JC, Graeff FG. GABA medi-
290. Peano CA, Shonis CA, Dillon GH, Waldrop ation of the anti-aversive action of minor tranquiliz-
TG. Hypothalamic GABAergic mechanism involved ers. Pharmacol Biochem Behav. 1982;16:397402.
in respiratory response to hypercapnia. Brain Res 305. Graeff FG, Brando ML, Audi EA, Milani H. Role
Bull. 1992;28:10713. of GABA in the anti-aversive action of anxiolytics.
291. Panneton WM, Loewy AD. Projections of the carotid Adv Biochem Psychopharmacol. 1986;42:7986.
sinus nerve to the nucleus of the solitary tract in the 306. Behbehani MM, Jiang MR, Chandler SD, Ennis
cat. Brain Res. 1980;191:23944. M. The effect of GABA and its antagonists on mid-
292. Sandkuhler J, Herdegen T. Distinct patterns of acti- brain periaqueductal gray neurons in the rat. Pain.
vated neurons throughout the rat midbrain periaque- 1990;40:195204.
ductal gray induced by chemical stimulation within 307. Lovick TA, Stezhka VV. Neurones in the dorsolat-
its subdivisions. J Comp Neurol. 1995;357:54653. eral periaqueductal grey matter in coronal slices of
293. Morgan MM, Carrive P. Activation of the ventrolat- rat midbrain: electrophysiological and morphologi-
eral periaqueductal gray reduces locomotion but not cal characteristics. Exp Brain Res. 1999;124:538.
mean arterial pressure in awake, freely moving rats. 308. Brandao ML, Lopez-Garcia JA, Graeff FG, Roberts
Neuroscience. 2001;102:90510. MH. Electrophysiological evidence for excitatory
294. Loewy AD, Wallach JH, Kellar S. Efferent connec- 5-HT2 and depressant 5-HT1A receptors on neu-
tions of the ventral medulla oblongata in the rat. rones of the rat midbrain tectum. Brain Res.
Brain Res Rev. 1981;3:6380. 1991;556:25966.
295. Sandner G, Dessort D, Schmitt P, Karli P. Distribution 309. Jeong HJ, Lam K, Mitchell VA, Vaughan
of GABA in the periaqueductal gray matter. Effects CW. Serotonergic modulation of neuronal activity in
of medial hypothalamic lesions. Brain Res. 1981; rat midbrain periaqueductal gray. J Neurophysiol.
224:27990. 2013;109:27129.
296. Van den Bergh P, Wu P, Jackson IM, Lechan 310. Stezhka VV, Lovick TA. Inhibitory and excitatory
RM. Neurons containing a N-terminal sequence of projections from the dorsal raphe nucleus to neurons
the TRH-prohormone (preproTRH53-74) are pre- in the dorsolateral periaqueductal gray matter in
sent in a unique location of the midbrain periaque- slices of midbrain maintained in vitro. Neuroscience.
ductal gray of the rat. Brain Res. 1988;461:5363. 1994;62:17787.
297. Beitz AJ, Williams FG. Localization of putative 311. Thompson RH, Canteras NS, Swanson LW. Organization
amino acid transmitters in the PAG and their rela- of projections from the dorsomedial nucleus of the hypo-
tionship to the PAG-raphe magnus pathway. The thalamus: a PHA-L study in the rat. J Comp Neurol.
midbrain periaqueductal gray matter. New York: 1996;376:14373.
Plenum Press; 1991. p. 30527. 312. Veening J, Buma P, Ter Horst GJ, Roeling TAP,
298. Smith GST, Savery D, Marden C, Costa JJL, Averill Luiten PGM, Nieuwenhuys R. Hypothalamic pro-
S, Priestley JV, et al. Distribution of messenger jections to the PAG in the rat: topographical,
RNAs encoding enkephalin, substance P, somato- immuno-electronmicroscopical and functional
statin, galanin, vasoactive intestinal polypeptide, aspects. In: Depaulis A, Bandler R, editors. The mid-
neuropeptide Y, and calcitonin gene related peptide brain periaqueductal gray matter. New York: Plenum
in the midbrain periaqueductal grey in the rat. Press; 1991. p. 387415.
J Comp Neurol. 1994;350:2340. 313. Barbaresi P. Immunocytochemical localization of
299. Mihaly E, Legradi G, Fekete C, Lechan RM. Efferent substance P receptor in rat periaqueductal gray mat-
projections of ProTRH neurons in the ventrolateral ter: a light and electron microscopic study. J Comp
periaqueductal gray. Brain Res. 2001;919:18597. Neurol. 1998;398:47390.
72 L.C. Schenberg

314. Drew GM, Mitchell VA, Vaughan CW. Postsynaptic 329. Sajdyk TJ, Shekhar A. Sodium lactate elicits anxiety
actions of substance P on rat periaqueductal grey neu- in rats after repeated GABA receptor blockade in the
rons in vitro. Neuropharmacology. 2005;49:58795. basolateral amygdala. Eur J Pharmacol. 2000;
315. Brodin E, Rosen A, Schott E, Brodin K. Effects of 394:26573.
sequential removal of rats from a group cage, and of 330. Johnson PL, Truitt W, Fitz SD, Minick PE, Dietrich
individual housing of rats, on substance P, cholecysto- A, Sanghani S, et al. A key role for orexin in panic
kinin and somatostatin levels in the periaqueductal grey anxiety. Nat Med. 2010;16:1115.
and limbic regions. Neuropeptides. 1994;26:25360. 331. Antunes-Rodrigues J, Castro M, Elias LL, Valenca
316. Bassi GS, Nobre MJ, Carvalho MC, Brando MM, McCann SM. Neuroendocrine control of body
ML. Substance P injected into the dorsal periaqueduc- fluid metabolism. Physiol Rev. 2004;84:169208.
tal gray causes anxiogenic effects similar to the long- 332. Fitzsimons JT. Angiotensin, thirst, and sodium appe-
term isolation as assessed by ultrasound vocalizations tite. Physiol Rev. 1998;78:583686.
measurements. Behav Brain Res. 2007;182:3017. 333. Rodgers RJ, Ishii Y, Halford JC, Blundell JE. Orexins and
317. Shekhar A, DiMicco JA. Defense reaction elicited appetite regulation. Neuropeptides. 2002;36:30325.
by injection of GABA antagonists and synthesis 334. Siegel JM. Narcolepsy: a key role for hypocretins
inhibitors into the posterior hypothalamus in rats. (orexins). Comment. Cell. 1999;98:40912.
Neuropharmacology. 1987;26:40717. 335. Taheri S, Zeitzer JM, Mignot E. The role of hypocre-
318. Shekhar A, Hingtgen JN, DiMicco JA. Selective tins (orexins) in sleep regulation and narcolepsy.
enhancement of shock avoidance responding elicited Ann Rev Neurosci. 2002;25:283313.
by GABA blockade in the posterior hypothalamus of 336. Molosh AI, Johnson PL, Fitz SD, DiMicco JA,
rats. Brain Res. 1987;420:11828. Herman JP, Shekhar A. Changes in central sodium and
319. Shekhar A, Hingtgen JN, DiMicco JA. GABA recep- not osmolarity or lactate induce panic-like responses
tors in the posterior hypothalamus regulate experi- inamodelofpanicdisorder.Neuropsychopharmacology.
mental anxiety in rats. Brain Res. 1990;512:818. 2010;35:133347.
320. Shekhar A. GABA receptors in the region of the dor- 337. Kellner M, Wiedemann K, Holsboer F. Atrial natri-
somedial hypothalamus of rats regulate anxiety in uretic factor inhibits the CRH-stimulated secretion
the elevated plus-maze test. I Behav Meas Brain Res. of ACTH and cortisol in man. Life Sci. 1992;50:
1993;627:916. 183542.
321. Shekhar A, Johnson PL, Sajdyk TJ, Fitz SD, Keim 338. Tannure RM, Bittencourt AS, Schenberg LC. Short-
SR, Kelley PE, et al. Angiotensin-II is a putative term full kindling of the amygdala dissociates natu-
neurotransmitter in lactate-induced panic-like ral and periaqueductal gray-evoked flight behaviors
responses in rats with disruption of GABAergic inhi- of the rat. Behav Brain Res. 2009;199:24756.
bition in the dorsomedial hypothalamus. J Neurosci. 339. Bailey TW, DiMicco JA. Chemical stimulation of
2006;26:920515. the dorsomedial hypothalamus elevates plasma
322. Johnson PL, Shekhar A. Panic-prone state induced ACTH in conscious rats. Am J Physiol Regul Integr
in rats with GABA dysfunction in the dorsomedial Comp Physiol. 2001;280:R815.
hypothalamus is mediated by NMDA receptors. 340. Zaretskaia MV, Zaretsky DV, Shekhar A, DiMicco
J Neurosci. 2006;26:7093104. JA. Chemical stimulation of the dorsomedial hypo-
323. Shekhar A. Effects of treatment with imipramine thalamus evokes non-shivering thermogenesis in
and clonazepam on an animal model of panic disor- anesthetized rats. Brain Res. 2002;928:11325.
der. Biol Psychiatry. 1994;36:74858. 341. Johnson PL, Truitt WA, Fitz SD, Lowry CA, Shekhar
324. Shekhar A, Keim SR, Simon JR, McBride A. Neural pathways underlying lactate-induced panic.
WJ. Dorsomedial hypothalamic GABA dysfunction Neuropsychopharmacology. 2008;33:2093107.
produces physiological arousal following sodium 342. Olsson M, Ho HP, Annerbrink K, Thylefors J,
lactate infusions. Pharmacol Biochem Behav. Eriksson E. Respiratory responses to intravenous
1996;55:24956. infusion of sodium lactate in male and female Wistar
325. Swanson LW, Lind RW. Neural projections subserv- rats. Neuropsychopharmacology. 2002;27:8591.
ing the initiation of a specific motivated behavior in 343. Herbert H, Moga MM, Saper CB. Connections of
the rat: new projections from the subfornical organ. the parabrachial nucleus with the nucleus of the sol-
Brain Res. 1986;379:399403. itary tract and the medullary reticular formation in
326. Shekhar A, Sajdyk TS, Keim SR, Yoder KK, Sanders the rat. J Comp Neurol. 1990;293:54080.
SK. Role of the basolateral amygdala in panic disor- 344. Chamberlin NL, Saper CB. Topographic organiza-
der. Ann N Y Acad Sci. 1999;877:74750. tion of respiratory responses to glutamate micro-
327. Shekhar A, Keim SR. LY354740, a potent group II stimulation of the parabrachial nucleus in the rat.
metabotropic glutamate receptor agonist prevents J Neurosci. 1994;14:650010.
lactate-induced panic-like response in panic-prone 345. Chamberlin NL. Functional organization of the para-
rats. Neuropharmacology. 2000;39:113946. brachial complex and intertrigeminal region in the
328. Sajdyk TJ, Schober DA, Gehlert DR, Shekhar A. Role control of breathing. Respir Physiol Neurobiol.
of corticotropin-releasing factor and urocortin within 2004;143:11525.
the basolateral amygdala of rats in anxiety and panic 346. Kaur S, Pedersen NP, Yokota S, Hur EE, Fuller PM,
responses. Behav Brain Res. 1999;100:20715. Lazarus M, et al. Glutamatergic signaling from the
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 73

parabrachial nucleus plays a critical role in hyper- 361. Jovanovic T, Norrholm SD, Blanding NQ, Davis M,
capnic arousal. J Neurosci. 2013;33:762740. Duncan E, Bradley B, et al. Impaired fear inhibition
347. Yokota S, Kaur S, Vanderhorst VG, Saper CB, is a biomarker of PTSD but not depression. Depress
Chamberlin NL. Respiratory-related outputs of glu- Anxiety. 2010;27:24451.
tamatergic, hypercapnia-responsive parabrachial 362. Keen-Rhinehart E, Michopoulos V, Toufexis DJ,
neurons in mice. J Comp Neurol. 2015;523:90720. Martin EI, Nair H, Ressler KJ, et al. Continuous
348. Huang GF, Besson JM, Bernard JF. Intravenous expression of corticotropin-releasing factor in the
morphine depresses the transmission of noxious central nucleus of the amygdala emulates the dys-
messages to the nucleus centralis of the amygdala. regulation of the stress and reproductive axes. Mol
Eur J Pharmacol. 1993;236:44956. Psychiatry. 2009;14:3750.
349. Huang GF, Besson JM, Bernard JF. Morphine 363. Neill JD. Effects of stress on serum prolactin and
depresses the transmission of noxious messages in luteinizing hormone levels during the estrous cycle
the spino(trigemino)-ponto-amygdaloid pathway. of the rat. Endocrinology. 1970;87:11927.
Eur J Pharmacol. 1993;230:27984. 364. Siegel RA, Conforti N, Chowers I. Neural pathways
350. Beitz AJ. Possible origin of glutamatergic projec- mediating the prolactin secretory response to acute
tions to the midbrain periaqueductal grey and deep neurogenic stress in the male rat. Brain Res. 1980;
layer of the superior colliculus of the rat. Brain Res 198:4353.
Bull. 1989;23:2535. 365. Dijkstra H, Tilders FJH, Hiehle MA, Smelik
351. da Silva LG, de Menezes RC, dos Santos RA, PG. Hormonal reactions to fighting in rat colonies
Campagnole-Santos MJ, Fontes MA. Role of periaq- prolactin rises during defence, not during offence.
ueductal gray on the cardiovascular response evoked Physiol Behav. 1992;51:9618.
by disinhibition of the dorsomedial hypothalamus. 366. Sinha SS, Coplan JD, Pine DS, Martinez JA, Klein
Brain Res. 2003;984:20614. DF, Gorman JM. Panic induced by carbon dioxide
352. da Silva LG, Menezes RC, Villela DC, Fontes inhalation and lack of hypothalamic-pituitary-adrenal
MA. Excitatory amino acid receptors in the periaq- axis activation. Psychiatry Res. 1999;86:938.
ueductal gray mediate the cardiovascular response 367. Bandelow B, Wedekind D, Pauls J, Broocks A,
evoked by activation of dorsomedial hypothalamic Hajak G, Ruther E. Salivary cortisol in panic attacks.
neurons. Neuroscience. 2006;139:112939. Am J Psychiatry. 2000;157:4546.
353. Allen GV, Saper CB, Hurley KM, Cechetto 368. Kamilaris TC, Johnson EO, Calogero AE, Kalogeras
DF. Organization of visceral and limbic connections KT, Bernardini R, Chrousos GP, et al. Cholecystokinin-
in the insular cortex of the rat. J Comp Neurol. octapeptide stimulates hypothalamic-pituitary-adrenal
1991;311:116. function in rats: role of corticotropin-releasing hor-
354. Shipley MT, Ennis M, Rizvi TA, Behbehani mone. Endocrinology. 1992;130:176474.
MM. Topographical specificity of forebrain inputs to 369. Graeff FG, Garcia-Leal C, Del-Ben CM, Guimaraes
the midbrain periaqueductal gray: evidence for dis- FS. Does the panic attack activate the hypothalamic-
crete longitudinally organized input columns. In: pituitary-adrenal axis? An Acad Bras Cienc. 2005;77:
Depaulis A, Bandler R, editors. The midbrain peri- 47791.
aqueductal gray matter. New York: Plenum Press; 370. De Montigny C. Cholecystokinin tetrapeptide induces
1991. p. 41748. panic-like attacks in healthy volunteers. Preliminary
355. Floyd NS, Price JL, Ferry AT, Keay KA, Bandler findings. Arch Gen Psychiatry. 1989;46:5117.
R. Orbitomedial prefrontal cortical projections to 371. Bradwejn J, Koszycki D, Annable L, Couetoux DT,
distinct longitudinal columns of the periaqueductal Reines S, Karkanias C. A dose-ranging study of the
gray in the rat. J Comp Neurol. 2000;422:55678. behavioral and cardiovascular effects of CCK-
356. Selye H. A syndrome produced by diverse nocuous tetrapeptide in panic disorder. Biol Psychiatry.
agents. Nature. 1936;138:31. 1992;32:90312.
357. Selye H. Forty years of stress research: principal 372. Abelson JL, Liberzon I. Dose response of adrenocorti-
remaining problems and misconceptions. Can Med cotropin and cortisol to the CCK-B agonist pentagas-
Assoc J. 1976;115:536. trin. Neuropsychopharmacology. 1999;21:48594.
358. Strohle A, Holsboer F. Stress responsive neurohor- 373. Kobelt P, Paulitsch S, Goebel M, Stengel A,
mones in depression and anxiety. Pharmacopsychiatry. Schmidtmann M, van der Voort I, et al. Peripheral
2003;36 Suppl 3:S20714. injection of CCK-8S induces Fos expression in the
359. Jovanovic T, Norrholm SD, Fennell JE, Keyes M, dorsomedial hypothalamic nucleus in rats. Brain
Fiallos AM, Myers KM, et al. Posttraumatic stress Res. 2006;1117:10917.
disorder may be associated with impaired fear 374. Shlik J, Aluoja A, Vasar V, Vasar E, Podar T, Bradwejn
inhibition: relation to symptom severity. Psychiatry J. Effects of citalopram treatment on behavioural, car-
Res. 2009;167:15160. diovascular and neuroendocrine response to cholecys-
360. Jovanovic T, Norrholm SD, Blanding NQ, Phifer JE, tokinin tetrapeptide challenge in patients with panic
Weiss T, Davis M, et al. Fear potentiation is associated disorder. J Psychiatry Neurosci. 1997;22:33240.
with hypothalamic-pituitary-adrenal axis function in 375. Depot M, Caille G, Mukherjee J, Katzman MA,
PTSD. Psychoneuroendocrinology. 2010;35:84657. Cadieux A, Bradwejn J. Acute and chronic role of
74 L.C. Schenberg

5-HT3 neuronal system on behavioral and neuroen- 388. Koolhaas JM, Bartolomucci A, Buwalda B, de Boer
docrine changes induced by intravenous cholecysto- SF, Flugge G, Korte SM, et al. Stress revisited: a crit-
kinin tetrapeptide administration in humans. ical evaluation of the stress concept. Neurosci
Neuropsychopharmacology. 1999;20:17787. Biobehav Rev. 2011;35:1291301.
376. Chen DY, Deutsch JA, Gonzalez MF, Gu Y. The 389. Raff H, Shinsako J, Dallman MF. Renin and ACTH
induction and suppression of c-fos expression in the responses to hypercapnia and hypoxia after chronic
rat brain by cholecystokinin and its antagonist carotid chemodenervation. Am J Physiol. 1984;247:
L364,718. Neurosci Lett. 1993;149:914. R4127.
377. Schenberg LC, Dos Reis AM, Ferreira Povoa RM, 390. Raff H, Shinsako J, Keil LC, Dallman
Tufik S, Silva SR. A panic attack-like unusual stress MF. Vasopressin, ACTH, and corticosteroids during
reaction. Horm Behav. 2008;54:58491. hypercapnia and graded hypoxia in dogs. Am
378. Lim LW, Blokland A, van DM, Visser-Vandewalle J Physiol. 1983;244:E4538.
V, Tan S, Vlamings R, et al. Increased plasma corti- 391. Raff H, Shinsako J, Keil LC, Dallman
costerone levels after periaqueductal gray MF. Vasopressin, ACTH, and blood pressure during
stimulation-induced escape reaction or panic attacks hypoxia induced at different rates. Am J Physiol.
in rats. Behav Brain Res. 2011;218:3017. 1983;245:E48993.
379. Armini RS, Bernabe CS, Rosa CA, Siller CA, 392. Raff H, Sandri RB, Segerson TP. Renin, ACTH, and
Schimitel FG, Tufik S, et al. In a rat model of panic, adrenocortical function during hypoxia and hemorrhage
corticotropin responses to dorsal periaqueductal in conscious rats. Am J Physiol. 1986;250:R2404.
gray stimulation depend on physical exertion. 393. Kossowsky J, Wilhelm FH, Schneider S. Responses
Psychoneuroendocrinology. 2015;53:13647. to voluntary hyperventilation in children with sepa-
380. Beckett S, Marsden CA. Computer analysis and ration anxiety disorder: implications for the link to
quantification of periaqueductal grey- induced panic disorder. J Anxiety Disord. 2013;27:62734.
defence behaviour. J Neurosci Methods. 1995;58: 394. Vargas LC, Marques TA, Schenberg LC. Micturition
15761. and defensive behaviors are controlled by distinct
381. Neophytou SI, Graham M, Williams J, Aspley S, neural networks within the dorsal periaqueductal
Marsden CA, Beckett SR. Strain differences to the gray and deep gray layer of the superior colliculus of
effects of aversive frequency ultrasound on behav- the rat. Neurosci Lett. 2000;280:458.
iour and brain topography of c-fos expression in the 395. Schenberg LC, Marcal LPA, Seeberger F, Barros
rat. Brain Res. 2000;854:15864. MR, Sudr ECM. L-type calcium channels selec-
382. Soya H, Mukai A, Deocaris CC, Ohiwa N, Chang H, tively control the defensive behaviors induced by
Nishijima T, et al. Threshold-like pattern of neuronal electrical stimulation of dorsal periaqueductal gray
activation in the hypothalamus during treadmill run- and overlying collicular layers. Behav Brain Res.
ning: establishment of a minimum running stress 2000;111:17585.
(MRS) rat model. Neurosci Res. 2007;58:3418. 396. Goetz RR, Klein DF, Gorman JM. Consistencies
383. Pittman QJ, Blume HW, Renaud LP. Connections of between recalled panic and lactate-induced panic.
the hypothalamic paraventricular nucleus with the Anxiety. 1994;1:316.
neurohypophysis, median eminence, amygdala, lat- 397. APA. Diagnostic and statistical manual of mental
eral septum and midbrain periaqueductal gray: an disorders. 4th-R ed. Washington, DC: American
electrophysiological study in the rat. Brain Res. Psychiatry Association; 2000.
1981;215:1528. 398. WHO. The ICD-10 classification of mental and
384. Ziegler DR, Edwards MR, Ulrich-Lai YM, Herman behavioral disorders. Diagnostic criteria for research.
JP, Cullinan WE. Brainstem origins of glutamatergic Geneva: 1993.
innervation of the rat hypothalamic paraventricular 399. D'Amato FR, Zanettini C, Lampis V, Coccurello R,
nucleus. J Comp Neurol. 2012;520:236994. Pascucci T, Ventura R, et al. Unstable maternal envi-
385. Klein S, Nicolas LB, Lopez-Lopez C, Jacobson LH, ronment, separation anxiety, and heightened CO2
McArthur SG, Grundschober C, et al. Examining face sensitivity induced by gene-by-environment inter-
and construct validity of a noninvasive model of panic play. PLoS One. 2011;6:e18637.
disorder in Lister-hooded rats. Psychopharmacology 400. Kinkead R, Genest SE, Gulemetova R, Lajeunesse Y,
(Berl). 2010;211:197208. Laforest S, Drolet G, et al. Neonatal maternal separa-
386. Feldman S, Weidenfeld J. The excitatory effects of the tion and early life programming of the hypoxic venti-
amygdala on hypothalamo-pituitary-adrenocortical latory response in rats. Respir Physiol Neurobiol.
responses are mediated by hypothalamic norepineph- 2005;149:31324.
rine, serotonin, and CRF-41. Brain Res Bull. 1998;45: 401. Genest SE, Gulemetova R, Laforest S, Drolet G,
38993. Kinkead R. Neonatal maternal separation and sex-
387. Bhatnagar S, Viau V, Chu A, Soriano L, Meijer OC, specific plasticity of the hypoxic ventilatory response
Dallman MF. A cholecystokinin-mediated pathway to in awake rat. J Physiol. 2004;554:54357.
the paraventricular thalamus is recruited in chronically 402. Genest SE, Gulemetova R, Laforest S, Drolet G,
stressed rats and regulates hypothalamic-pituitary- Kinkead R. Neonatal maternal separation induces
adrenal function. J Neurosci. 2000;20:556473. sex-specific augmentation of the hypercapnic venti-
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 75

latory response in awake rat. J Appl Physiol. iors generated by the elevated T-maze. Brain Res
2007;102:141621. Bull. 2004;64:1818.
403. Francis DD, Meaney MJ. Maternal care and the 418. Maier SF, Seligman ME. Learned helplessness: the-
development of stress responses. Curr Opin ory and evidence. J Exp Psychol. 1975;105:346.
Neurobiol. 1999;9:12834. 419. Maier SF. Learned helplessness and animal models
404. Abelson JL, Curtis GC. Hypothalamic-pituitary- of depression. Prog Neuropsychopharmacol Biol
adrenal axis activity in panic disorder. 24-Hour Psychiatry. 1984;8:43546.
secretion of corticotropin and cortisol. Arch Gen 420. Maier SF, Watkins LR. Stressor controllability, anxi-
Psychiatry. 1996;53:32331. ety, and serotonin. Cogn Ther Res. 1998;22:595613.
405. Schreiber W, Lauer CJ, Krumrey K, Holsboer F, 421. Maier SF, Watkins LR. Stressor controllability and
Krieg JC. Dysregulation of the hypothalamic- learned helplessness: the roles of the dorsal raphe
pituitary-adrenocortical system in panic disorder. nucleus, serotonin, and corticotropin-releasing fac-
Neuropsychopharmacology. 1996;15:715. tor. Neurosci Biobehav Rev. 2005;29:82941.
406. Rentesi G, Antoniou K, Marselos M, Fotopoulos A, 422. Porsolt RD, Lenegre A, McArthur RA. Pharmacological
Alboycharali J, Konstandi M. Long-term conse- models of depression. In: Olivier B, Slangen JL, Mos J,
quences of early maternal deprivation in serotoner- editors. Animal models in psychopharmacology. Basle:
gic activity and HPA function in adult rat. Neurosci Birkhaeuser-Verlag; 1991. p. 13761.
Lett. 2010;480:711. 423. Maier SF, Grahn RE, Kalman BA, Sutton LC,
407. Fournier S, Allard M, Gulemetova R, Joseph V, Kinkead Wiertelak EP, Watkins LR. The role of the amygdala
R. Chronic corticosterone elevation and sex-specific and dorsal raphe nucleus in mediating the behavioral
augmentation of the hypoxic ventilatory response in consequences of inescapable shock. Behav Neurosci.
awake rats. J Physiol. 2007;584(Pt 3):95162. 1993;107:37788.
408. Gorman JM, Coplan JD. Comorbidity of depression 424. Amat J, Matus-Amat P, Watkins LR, Maier
and panic disorder. J Clin Psychiatry. 1996;57 Suppl SF. Escapable and inescapable stress differentially
10:3441. and selectively alter extracellular levels of 5-HT in
409. Kaufman J, Charney D. Comorbidity of mood and the ventral hippocampus and dorsal periaqueductal
anxiety disorders. Depress Anxiety. 2000;12 Suppl gray of the rat. Brain Res. 1998;797:1222.
1:6976. 425. Hammack SE, Richey KJ, Watkins LR, Maier
410. Safadi G, Bradwejn J. Relationship of panic disorder SF. Chemical lesion of the bed nucleus of the stria ter-
to posttraumatic stress disorder. Arch Gen Psychiatry. minalis blocks the behavioral consequences of uncon-
1995;52:768. trollable stress. Behav Neurosci. 2004;118:4438.
411. Koenen KC, Lyons MJ, Goldberg J, Simpson J, 426. Leshner AI, Segal M. Fornix transection blocks
Williams WM, Toomey R, et al. A high risk twin learned helplessness in rats. Behav Neural Biol.
study of combat-related PTSD comorbidity. Twin 1979;26:497501.
Res. 2003;6:21826. 427. Petty F, Sherman AD. Learned helplessness induc-
412. Nixon RD, Resick PA, Griffin MG. Panic following tion decreases in vivo cortical serotonin release.
trauma: the etiology of acute posttraumatic arousal. Pharmacol Biochem Behav. 1983;18:64950.
J Anxiety Disord. 2004;18:193210. 428. Joca SR, Padovan CM, Guimaraes FS. Activation of
413. Hinton D, Hsia C, Um K, Otto MW. Anger- post-synaptic 5-HT(1A) receptors in the dorsal hip-
associated panic attacks in Cambodian refugees with pocampus prevents learned helplessness develop-
PTSD; a multiple baseline examination of clinical ment. Brain Res. 2003;978:17784.
data. Behav Res Ther. 2003;41:64754. 429. Joca SR, Zanelati T, Guimaraes FS. Post-stress facil-
414. Cougle JR, Feldner MT, Keough ME, Hawkins KA, itation of serotonergic, but not noradrenergic, neuro-
Fitch KE. Comorbid panic attacks among individu- transmission in the dorsal hippocampus prevents
als with posttraumatic stress disorder: associations learned helplessness development in rats. Brain Res.
with traumatic event exposure history, symptoms, 2006;1087:6774.
and impairment. J Anxiety Disord. 2010;24:1838. 430. Malberg JE, Duman RS. Cell proliferation in adult
415. McGrath PJ, Stewart JW, Liebowitz MR, Markowitz hippocampus is decreased by inescapable stress: rever-
JM, Quitkin FM, Klein DF, et al. Lactate provoca- sal by fluoxetine treatment. Neuropsychopharmacology.
tion of panic attacks in depressed outpatients. 2003;28:156271.
Psychiatry Res. 1988;25:417. 431. Zhou J, Li L, Tang S, Cao X, Li Z, Li W, et al. Effects
416. Quintino-Dos-Santos JW, Muller CJ, Santos AMC, of serotonin depletion on the hippocampal GR/MR
Tufik S, Rosa CA, Schenberg LC. Long-lasting and BDNF expression during the stress adaptation.
marked inhibition of periaqueductal gray-evoked Behav Brain Res. 2008;195:12938.
defensive behaviors in inescapably-shocked rats. 432. Wu J, Kramer GL, Kram M, Steciuk M, Crawford
Eur J Neurosci. 2014;39:27586. IL, Petty F. Serotonin and learned helplessness: a
417. de Paula Soares V, Zangrossi Jr H. Involvement of regional study of 5-HT1A, 5-HT2A receptors and
5-HT1A and 5-HT2 receptors of the dorsal periaque- the serotonin transport site in rat brain. J Psychiatr
ductal gray in the regulation of the defensive behav- Res. 1999;33:1722.
76 L.C. Schenberg

433. Yang LM, Hu B, Xia YH, Zhang BL, Zhao H. Lateral CO2 inhalation in panic disorder: a comparison with
habenula lesions improve the behavioral response in major depression and premenstrual dysphoric disor-
depressed rats via increasing the serotonin level in dor- der. Am J Psychiatry. 2001;158:5867.
sal raphe nucleus. Behav Brain Res. 2008;188:8490. 449. Lovick TA. Plasticity of GABA-A receptor subunit
434. Lino-de-Oliveira C, De Lima TC, Carobrez expression during the oestrous cycle of the rat:
AP. Dorsal periaqueductal gray matter inhibits pas- implications for premenstrual syndrome in women.
sive coping strategy elicited by forced swimming Exp Physiol. 2006;91:65560.
stress in rats. Neurosci Lett. 2002;335:8790. 450. Lovick TA. GABA in the female brainoestrous
435. Lino-de-Oliveira C, De Oliveira RM, Padua CA, De cycle-related changes in GABAergic function in the
Lima TC, Del Bel EA, Guimaraes FS. Antidepressant periaqueductal grey matter. Pharmacol Biochem
treatment reduces Fos-like immunoreactivity induced Behav. 2008;90:4350.
by swim stress in different columns of the periaque- 451. Lovick TA, Devall AJ. Progesterone withdrawal-
ductal gray matter. Brain Res Bull. 2006;70:41421. evoked plasticity of neural function in the female
436. Jang DP, Lee SH, Lee SY, Park CW, Cho ZH, Kim periaqueductal grey matter. Neural Plast. 2009;2009:
YB. Neural responses of rats in the forced swim- 730902.
ming test: [F-18]FDG micro PET study. Behav Brain 452. Lovick TA, Griffiths JL, Dunn SM, Martin
Res. 2009;203:437. IL. Changes in GABA-A receptor subunit expres-
437. Strong PV, Christianson JP, Loughridge AB, Amat J, sion in the midbrain during the oestrous cycle in
Maier SF, Fleshner M, et al. 5-Hydroxytryptamine 2C Wistar rats. Neuroscience. 2005;131:397405.
receptors in the dorsal striatum mediate stress-induced 453. Griffiths J, Lovick T. Withdrawal from progesterone
interference with negatively reinforced instrumental increases expression of alpha4, beta1, and delta
escape behavior. Neuroscience. 2011;197:13244. GABA(A) receptor subunits in neurons in the peri-
438. Krout KE, Loewy AD. Periaqueductal gray matter aqueductal gray matter in female Wistar rats. J Comp
projections to midline and intralaminar thalamic Neurol. 2005;486:8997.
nuclei of the rat. J Comp Neurol. 2000;424:11141. 454. Griffiths JL, Lovick TA. GABAergic neurones in the
439. Macchi G, Bentivoglio M, Molinari M, Minciacchi rat periaqueductal grey matter express alpha4, beta1
D. The thalamo-caudate versus thalamo-cortical projec- and delta GABAA receptor subunits: plasticity of
tions as studied in the cat with fluorescent retrograde expression during the estrous cycle. Neuroscience.
double labeling. Exp Brain Res. 1984;54:22539. 2005;136:45766.
440. Kaufman J, Plotsky PM, Nemeroff CB, Charney 455. Brack KE, Jeffery SM, Lovick TA. Cardiovascular
DS. Effects of early adverse experiences on brain and respiratory responses to a panicogenic agent in
structure and function: clinical implications. Biol anaesthetised female Wistar rats at different stages of
Psychiatry. 2000;48:77890. the oestrous cycle. Eur J Neurosci. 2006;23:330918.
441. Faravelli C, Pallanti S. Recent life events and panic 456. Brack KE, Lovick TA. Neuronal excitability in the
disorder. Am J Psychiatry. 1989;146:6226. periaqueductal grey matter during the estrous cycle in
442. Falsetti SA, Resnick HS. Frequency and severity of female Wistar rats. Neuroscience. 2007;144:32535.
panic attack symptoms in a treatment seeking sample 457. Smith SS, Woolley CS. Cellular and molecular
of trauma victims. J Trauma Stress. 1997;10:6839. effects of steroid hormones on CNS excitability.
443. Sherman AD, Sacquitne JL, Petty F. Specificity of Cleve Clin J Med. 2004;71 Suppl 2:S410.
the learned helplessness model of depression. 458. Gallo MA, Smith SS. Progesterone withdrawal
Pharmacol Biochem Behav. 1982;16:44954. decreases latency to and increases duration of elec-
444. Maier SF. Exposure to the stressor environment pre- trified prod burial: a possible rat model of PMS anxi-
vents the temporal dissipation of behavioral depres- ety. Pharmacol Biochem Behav. 1993;46:897904.
sion/learned helplessness. Biol Psychiatry. 2001;49: 459. Gulinello M, Gong QH, Li X, Smith SS. Short-term
76373. exposure to a neuroactive steroid increases alpha4
445. Yonkers KA, Pearlstein T, Rosenheck GABA-A receptor subunit levels in association with
RA. Premenstrual disorders: bridging research and increased anxiety in the female rat. Brain Res.
clinical reality. Arch Womens Ment Health. 2003;6: 2001;910:5566.
28792. 460. Gulinello M, Gong QH, Smith SS. Progesterone with-
446. Le Melldo JM, Van DM, Coupland NJ, Lott P, drawal increases the alpha4 subunit of the GABA-A
Jhangri GS. Response to flumazenil in women with receptor in male rats in association with anxiety and
premenstrual dysphoric disorder. Am J Psychiatry. altered pharmacologya comparison with female
2000;157:8213. rats. Neuropharmacology. 2002;43:70114.
447. Gorman JM, Kent J, Martinez J, Browne S, Coplan J, 461. Smith SS. Withdrawal properties of a neuroactive
Papp LA. Physiological changes during carbon diox- steroid: implications for GABA(A) receptor gene
ide inhalation in patients with panic disorder, major regulation in the brain and anxiety behavior. Steroids.
depression, and premenstrual dysphoric disorder: 2002;67:51928.
evidence for a central fear mechanism. Arch Gen 462. Smith SS, Gong QH, Hsu FC, Markowitz RS,
Psychiatry. 2001;58:12531. Ffrench-Mullen JM, Li X. GABA-A receptor alpha4
448. Kent JM, Papp LA, Martinez JM, Browne ST, Coplan subunit suppression prevents withdrawal properties of
JD, Klein DF, et al. Specificity of panic response to an endogenous steroid. Nature. 1998;392:92630.
2 A Neural Systems Approach to the Study of the Respiratory-Type Panic Disorder 77

463. Freeman EW, Frye CA, Rickels K, Martin PA, Smith activity in rat brain following exposure to a preda-
SS. Allopregnanolone levels and symptom improve- tory odor. Neuroscience. 2001;104:108597.
ment in severe premenstrual syndrome. J Clin 468. Bittencourt AS, Nakamura-Palacios EM, Mauad H,
Psychopharmacol. 2002;22:51620. Tufik S, Schenberg LC. Organization of electrically
464. Redgrave P, Dean P. Does the PAG learn about and chemically evoked defensive behaviors within
emergencies from the superior colliculus? In: the deeper collicular layers as compared to the peri-
Depaulis A, Bandler R, editors. The midbrain peri- aqueductal gray matter of the rat. Neuroscience.
aqueductal gray matter. New York: Plenum Press; 2005;133:87392.
1991. p. 199209. 469. Jenck F, Broekkamp CL, Van Delft AML. The effect of
465. King SM, Shehab S, Dean P, Redgrave P. Differential antidepressants on aversive periaqueductal gray stim-
expression of fos-like immunoreactivity in the ulation in rats. Eur J Pharmacol. 1990;177:2014.
descending projections of superior colliculus after 470. Hogg S, Michan L, Jessa M. Prediction of anti-panic
electrical stimulation in the rat. Behav Brain Res. properties of escitalopram in the dorsal periaqueduc-
1996;78:13145. tal grey model of panic anxiety. Neuropharmacology.
466. Canteras NS, Chiavegatto S, Valle LE, Swanson 2006;51:1415.
LW. Severe reduction of rat defensive behavior to a 471. Schenberg LC, Capucho LB, Vatanabe RO, Vargas
predator by discrete hypothalamic chemical lesions. LC. Acute effects of clomipramine and fluoxetine on
Brain Res Bull. 1997;44:297305. dorsal periaqueductal grey-evoked unconditioned
467. Dielenberg RA, Hunt GE, McGregor IS. When a defensive behaviours of the rat. Psychopharmacology
rat smells a cat: the distribution of Fos immunore- (Berl). 2002;159:13844.
The Hippocampus and Panic
Disorder: Evidence from Animal
and Human Studies

Gisele Pereira Dias and Sandrine Thuret

Contents Abstract
3.1 Anatomy of the Hippocampus 80 Panic disorder (PD) is a highly incapacitating
psychiatric disorder. Its wide range of somatic
3.2 The Role of the Hippocampus in Emotional
Regulation 80 and psychological symptoms makes it plausi-
ble that a number of different brain structures
3.3 Involvement of the Hippocampus
in PD: Evidence from Rodent Studies 83
and circuits are likely to mediate this condition.
In this chapter we highlight the possible contri-
3.4 Involvement of the Hippocampus
butions of the hippocampus, a key brain region
in PD: Evidence from Human Studies 84
involved in the regulation of cognition (learn-
3.5 The Role of the Hippocampus ing/memory), mood and defensive responses
in PD: Future Directions 85
(fear/anxiety), for the pathophysiology of
3.6 Conclusion 87 PD. This chapter will present the anatomy of
References 88 the hippocampus and highlight its role in emo-
tional regulation, so that an understanding of
the involvement of the hippocampus in PD can
be drawn. Evidence from both animal and
human findings on this topic will be approached.
Particularly, the capacity of the hippocampus to
continually generate newly functional neurons
throughout life, a phenomenon called adult hip-
G.P. Dias
Laboratory of Panic and Respiration, Institute of
pocampal neurogenesis, will be pointed as part
Psychiatry, Federal University of Rio de Janeiro, of the key future directions for the study of the
Rio de Janeiro, Brazil neurobiological basis of PD.
e-mail: giseledias@ufrj.br
S. Thuret (*)
Department of Basic and Clinical Neuroscience,
Laboratory of Adult Neurogenesis and Mental
Health, Institute of Psychiatry, Psychology and
Neuroscience, Kings College London, London, UK Hippocampus Panic disorder Fear circuitry
e-mail: sandrine.1.thuret@kcl.ac.uk Anxiety Stem cells Animal models

Springer International Publishing Switzerland 2016 79

A.E. Nardi, R.C.R. Freire (eds.), Panic Disorder, DOI 10.1007/978-3-319-12538-1_3
80 G.P. Dias and S. Thuret

3.1 Anatomy the processed information is sent to other parts of

of the Hippocampus the brain through projections to the subiculum
and EC [5, 6].
The hippocampus is one of the brain structures The hippocampal formation extends dorso-
implicated in the regulation of many of the fea- ventrally within the temporal lobes, an anatomic
tures involved in panic disorder (PD), such as feature that appears to render functional conse-
those related to defensive responses to threat. quences. In this sense, evidence indicates that the
Moreover, in the clinical practice with PD information processed in the dorsal hippocampus
patients, it is very common that this condition is is more likely to integrate into different circuits
accompanied by another highly impairing disor- than that processed by the ventral portion of the
der, agoraphobia. Agoraphobic patients present hippocampus. Thus, due to its projections to
intense fear of being in contexts where they can- regions such as the dorsal septum, the mammil-
not escape or find immediate help. Thus, although lary complex and the lateral entorhinal cortex,
normally the first panic attack occurs spontane- the dorsal hippocampus appears to be more
ously (i.e., without the presence of a specific and related to the regulation of cognitive abilities,
identifiable triggering stimulus), there is often an whereas mood and anxiety are believed to be
association of the attack with the surrounding more fundamentally regulated by the circuitry of
environmental (exteroceptive) and/or internal the ventral hippocampus [79], considering its
(interoceptive) cues. This latter is known to be outputs to the amygdala, hypothalamus, nucleus
mediated primarily by the insula but growing evi- accumbens and the prefrontal cortex [10]
dence from both human and rodent studies show (Fig. 3.1).
that the processing of exteroceptive cues is fun-
damentally a function of the hippocampus, along
with the processing of self-location [1], which in 3.2 The Role
itself is an essential requirement for agoraphobic of the Hippocampus
associations. in Emotional Regulation
Before exploring the evidence on the func-
tional contribution of the hippocampus to the It is well established that the hippocampus plays
development of PD-related symptomatology, it is a fundamental role for cognitive abilities, such
useful to understand how the hippocampal for- as spatial recognition and declarative memory
mation is organized and integrates the emotional [1114]. The classic case of patient H.M., who
circuits of the limbic system. had a bilateral temporal lobe resection as an
The hippocampus is a C-shaped structure intervention to reduce chronic epileptic seizures
located deeply within the subcortical region of and, as a consequence of the removal of the hip-
the temporal lobes, extending longitudinally pocampus, started to suffer from a severe inca-
along the brain. It is rolled-up in 2 laminae, the pacity to form new factual memories [15],
cornu Ammonis (CA; for its resemblance with established the hippocampal formation as the
Jupiter Ammons horn [2]) and the gyrus denta- house of memory in the brain. Considering the
tus (dentate gyrus, DG) [3]. It is well established large amount of evidence that came later with
that the information processing in the hippocam- hundreds of animal studies pointing for the hip-
pus occurs through its tri-synaptic circuit, start- pocampus as one of the ultimate structures
ing with the glutamatergic granule cells of the implicated in cognitive regulation, it might seem
DG receiving input from the entorhinal cortex counterintuitive that this brain region also sig-
(EC), via the perforant path. From the DG projec- nificantly participates in the regulation of mood
tions constituting the so-called Mossy fibers and anxiety. However, as we shall see in Sect. 3.4,
extend their axons to CA3, which then estab- the hippocampus integrates a defensive neural
lishes connection with the pyramidal neurons at network, proposed in the end of the 1980s by
CA1, via the Schaffer collaterals [4]. From CA1, Gorman et al. [16] and known as the fear circuitry,
3 The Hippocampus and Panic Disorder: Evidence from Animal and Human Studies 81

Fig. 3.1 Schematic representation of the dorsoventral tum, the mammillary complex and the lateral entorhinal
division of the hippocampus. The hippocampus is a cortex, is believed to be involved in cognitive abilities; the
C-shaped structure that extends longitudinally along the ventral hippocampus, in turn, projects to the amygdala,
septo-temporal axis of the brain. The dorsal hippocampus, hypothalamus, nucleus accumbens and prefrontal cortex,
due to its projections to structures such as the dorsal sep- thus being more related with emotional regulation [10]

a well-established circuit implicated in processing The hippocampal formation is, therefore, an

external and internal cues to prepare the indi- essential component of the emotional system of
vidual for a defensive response. This network the brain, playing an important functional role in
includes the amygdala, nucleus accumbens, hip- the modulation of complex behavioral patterns,
pocampus, ventromedial hypothalamus, periaq- along with cortical and subcortical areas.
ueductal gray, several brain stem and thalamic But which could be the putative roles of the
nuclei, the insular cortex, and some prefrontal hippocampus in the fear circuitry? One potential
regions (Box 3.1). candidate is the processing of risk assessment.
When it comes to defensiveness behaviors,
risk assessment emerges as a fundamental aspect
Box 3.1. Fear Circuitry of emotional regulation. This behavioral pattern
When a stimulus is perceived as potentially is believed to aim at evaluating the odds of threat/
harmful for the integrity of the individual, a potential danger in contrast with those of reward
series of neurochemical, neuroendocrine [17]. This defense strategy whereby the so-
and behavioral responses arises which, called non-defensive behaviors (self-grooming,
altogether, potentiate the likelihood of sur- locomotion, feeding, etc.) are fundamentally
vival. In neurobiological terms, threatening inhibited [18] has been pointed to be likely a
stimuli activate the so-called fear circuitry. function of the hippocampus [19] and its circuits
Fear responses are triggered through the with the septum and the amygdala [20]. The input
activation of a subcortical structure, the to the hippocampus from the medial entorhinal
amygdala, which receives afferents from a cortex (MEC) on the spatial context of an experi-
number of structures involved in cognitive ence, and from the lateral entorhinal cortex
processing, such as the sensorial cortex, the (LEC) on the content of an experience [21] would
thalamus and, of our special interest, the make it possible the delivery of information
hippocampus [16]. needed for risk assessment processing by the hip-
pocampal formation, thus closely relating a
82 G.P. Dias and S. Thuret

Fig. 3.2 Schematic representation of one of the potential the BLA, which in turn rapidly connects with the
roles of the hippocampus in the fear circuitry. The hippo- CEA. Through its connections to the dPAG and brain stem
campus encodes information about contexts for exam- nuclei, the CEA triggers the fight/flight fear response.
ple, information on the surrounding environmental stimuli BLA basolateral amygdala, CEA central amygdala, dPAG
associated with a previous panic attack and projects it to dorsal periaqueductal gray

possible hyperactivity of this function to the campal lesions interfere in contextual fear condi-
threat-biased information processing observed in tioning. Anagnostaras et al. [25] reinforced this
anxious patients. Nevertheless, some authors and showed further that the participation of the
relate risk assessment to be a function primarily hippocampus in contextual fear learning is time-
related to the mediation of danger in generalized framed, since lesions in this structure one day
anxiety disorder (GAD), and not in PD [20] after conditioning prevent fear learning, whereas
which would be a disorder more closely related hippocampal lesions 28 days after the condition-
to fear- rather than anxiety- (and therefore risk ing session render no effect over the storage of
assessment) related responses. Further research contextual fear memories. In this way, animals
on this topic should help unravel if and which the with hippocampal lesions do not present contex-
contributions of the hippocampus for risk assess- tual fear responses, given that they are not able to
ment in the context of panic responses are, since form a contextual representation necessary for
evidence from clinical practice largely point for the association of the context with the uncondi-
the presence of such behavioral patterns among tioned stimulus and send it to the amygdala so
PD patients as well. that the expected fear responses are triggered
Another putative role of the hippocampus in [26]. A schematic representation of this putative
PD particularly in cases involving agoraphobic role of the hippocampus within the fear circuitry
associations is through its well-established is illustrated in Fig. 3.2.
mediation of contextual fear learning. With Further and stronger evidence for the potential
effect, the hippocampus is one of the main neural role of the hippocampus on emotional regulation,
structures contributing to the fear/anxiety mani- expanding its functional roles beyond those
festations triggered by learned fear, most likely involved in cognitive processing, come from stud-
via its descending projections to the amygdala ies investigating the effects of anxiolytic drugs
[22]. Maren and Fanselow [23] showed that elec- injected specifically to the ventral hippocampus.
trolytic lesions in regions of the hippocampal for- Indeed, growing evidence point for the notion that
mation that send projections to the amygdala the functions regulated by the hippocampus are
eliminate pavlovian fear conditioning in situa- topographically distributed along its septo-tempo-
tions of contextual conditioned stimuli. Phillips ral axis, with the dorsal hippocampus being more
and Ledoux [24] also demonstrated that hippo- related to the regulation of cognition [79] and
3 The Hippocampus and Panic Disorder: Evidence from Animal and Human Studies 83

the ventral hippocampus more closely regulating showed this region to influence anxiety modula-
emotional processing [10], considering its projec- tion [2931]; the dorsoventral dichotomy [32],
tions to more limbic areas. According to Behrendt thus, seems to be less simple than thought, and
[27], information processed by the dorsal hippo- deserves special attention from the next genera-
campus would be translated into orienting and tion of studies aiming to unravel the particulari-
locomotor actions, whilst that processed by the ties of the contributions of the hippocampus to
ventral hippocampal formation would lead to the emotional regulation.
mapping of motivationally salient environmental
information, via its projections to the ventral
subiculum -ventromedial prefrontal cortex and 3.3 Involvement
ventral striatum. In addition, Behrendt [27] of the Hippocampus in PD:
emphasizes that not only external cues would be Evidence from Rodent
processed by the ventral hippocampus, but also Studies
emotional information extracted from the sur-
rounding internal physiological milieu, inducing Despite the growing evidence pointing for the
the individual to enter a behavioral mode of emo- involvement of the hippocampal circuitry in the
tion-guided arousal. Although the author high- mediation of anxiety-related behaviors, one of
lights the putative links of the ventral hippocampus the most established rodent models for PD has
with appetitive behavior, considering other the hypothalamus and not the hippocampus at
sources of evidence pointing for a role of this por- its center. The hypothalamic model is very effi-
tion of the hippocampus in defensiveness, this cient in inducing panic vulnerability in rats, by
becomes of special interest when thinking about chronically disrupting the inhibition promoted by
the possible contributions of the hippocampus to GABAergic neurons in the hypothalamus fol-
PD. Indeed, if it is true that at least the ventral por- lowed by intravenous injections of panicogenic
tion of the hippocampus is fundamental for the stimuli, such as sodium lactate infusion [33]. The
processing of exteroceptive and interoceptive model presents robust face, predictive and con-
cues, then the hippocampus could achieve a struct validities and its well-conceived rationale
higher status among the neural substrates of should also provide the basis for the development
PD. This is a reasonable hypothesis, considering of future rodent models aiming to unravel the
that PD/agoraphobia mostly emerge as a conse- contributions of other key brain areas involved in
quence of aversive associations with contexts fear, anxiety and PD, such as the hippocampus.
where previous panic attacks took place, as well At least two points should be highlighted in
as with bridges, tunnels and other agoraphobic our pursuit to understand the contributions of the
stimuli (external cues), in addition with higher hippocampus and other brain structures in animal
levels of CO2 and other panicogenic stimuli in the models of PD: (1) although the hippocampus has
blood and cerebrospinal fluid (internal cues). become a very popular structure to study due to
Supporting this idea, a recent study has identi- its highly neuroplastic nature (please see
fied that microinjections of neuropeptide S a Sect. 3.5), it is far from being the only brain
potential novel treatment for anxiety-related ill- region involved in such complex cognitive and
nesses such as PD into the ventral CA1 are suf- emotional processing, such as that present in anx-
ficient to reduce the anxious behavior of iety disorders, thus making it fundamental the
C57BL/6N mice [28]. The participation of the existence of models investigating other brain
hippocampus in emotional regulation appears to areas; (2) part of the scientific literature on rodent
be well established, however future studies are studies more clearly distinguish the panic attack
still needed for further investigation of the spe- from the anxiety concept. It is worth noting here
cific contributions of the dorsal and ventral por- that the panic attack is a fear-related response
tions of the hippocampus in this process. Indeed, characterized by the flight/freeze strategy medi-
some studies of the dorsal hippocampus also ated by the hypothalamus [33] and the dorsal
84 G.P. Dias and S. Thuret

periaqueductal gray to a real, present danger parvalbumin-positive interneurons in the hippo-

[20], not to mention the amygdala which is the campus 12 h after the panic-like reaction observed
hallmark structure underlying fear [19]; the anxi- in the open field [39]. Interestingly, these authors
ety concept, on the other hand, relates more to the had previously shown that electrical stimulation
anticipatory emotion to a potential threat. In the of the dorsal periaqueductal gray induced an
clinical practice, both emotions appear to be increase in the activation of cells in the CA1 and
present in PD patients: the fear response that DG 2 h after the panic reaction [40]. This rein-
characterizes the panic attack itself and the anxi- forces the notion that not only hippocampal sub-
ety presented in anticipation to the advent of a populations of neurons mediate the regulation of
new attack or to the absence of immediate help, long-lasting anxiety- and fear-related behavioral
in the case of PD patients with agoraphobia. patterns but also that they are necessary at differ-
But the hypothalamic rat model of PD is not the ent time frames of the fear learning processing.
only rodent model aiming at investigating the neu-
robiological basis of this disorder. In this context,
it has been shown, for instance, that intraperitoneal 3.4 Involvement
injections of lactate which produced panic-like of the Hippocampus in PD:
behavior, as seen by the induction of tachycardia Evidence from Human
and freezing response increased the neuronal fir- Studies
ing of neurons in the rat hippocampus [34].
Important evidence for the contribution of the Neuroimaging studies have immensely contrib-
hippocampus in PD have also come from studies uted to our current understanding of PD [41, 42]
using transgenic mice that overexpress the neuro- and other anxiety disorders, especially with
trophin tyrosine kinase receptor type 3 (NTRK3) regard the to fear circuitry [43]. According to the
a protein whose expression has been found to be original proposition by Gorman et al. [16], PD
altered in PD and other anxious patients [35]. patients would present a more sensitive fear net-
TgNTRK3 mice present good: face validity, as work, whose (hyper)activation would then result
seen by their heightened anxiety and panic-related in the arousal of panic attacks. Some authors,
responses; construct validity, observed by their however, suggest that this circuitry should be
increased density of noradrenergic neurons in the revised as to include other areas, such as the ante-
locus coeruleus, and finally, predictive validity, rior cingulate [44].
given that they respond well to diazepam in the Specifically with regard to the putative role of
elevated T maze [36], a paradigm for assessing the hippocampus in mediating both trait and state
panic-like behavioral patterns. These animals also anxiety in PD, data is conflicting despite a trend
present higher susceptibility to stress, as shown towards pointing the hippocampal circuits as
by their altered circadian corticosterone rhythm being fundamental for our comprehension of the
and more passive behaviors under certain chronic neurobiology of PD. In this sense, a quantitative
stress conditions [37], reinforcing this mouse line volumetric magnetic resonance imaging (MRI)
as an appropriate model of this disorder. Of spe- study showed that, although a bilateral reduction
cial note, it has been recently found that these of the temporal lobe volume was revealed in PD
mice present hyperexcitability in the hippocampal patients, the amygdala-hippocampus complex
subcircuit CA3CA1 and that this unbalanced (AHC) was found to be normal [45]. Similar
excitation-to-inhibition ratio in the hippocampus findings had been reported by Vythilingam et al.
underlies their also increased fear memories [38]. [46] who measured the volume of the temporal
Finally, additional evidence from animal stud- lobe, hippocampus and whole brain in 13 patients
ies for the contribution of the hippocampus in PD with PD in comparison with 14 healthy controls,
come from a study showing that the fear-like revealing that the mean volume of both left and
behavior induced by deep brain stimulation of right temporal lobes was significantly decreased
the periaqueductal gray leads to deactivation of in the patients group, without changes in the
3 The Hippocampus and Panic Disorder: Evidence from Animal and Human Studies 85

hippocampal volume. The assumption that defi- ing the contribution of the hippocampal formation
cits in hippocampal volume appear to be a minor in PD goes beyond the mapping of the neural
issue for the development of PD has also received substrates of this disorder: it is becoming increas-
recent support [47]. ingly relevant to comprehend how distinguished
Nevertheless, other studies point for the oppo- activation patterns of certain brain areas to a
site, i.e., for volume reductions of the hippocam- symptom-eliciting task can predict pharmaco-
pus and the left parahippocampal gyrus in PD logical and psychotherapeutic treatment out-
patients ([48]; reviewed in [49]). Future studies come. In this context, a recent study has shown,
with larger samples and more advanced imaging among others, that activation of the hippocampus
techniques are, thus, needed so that a more com- during maintenance of emotional responses to
prehensive view of the hippocampal volume as a negative images during fMRI scanning of PD
possible biomarker of PD can be pinpointed. and GAD patients was greater in responders than
But macroscopic measures, such as the vol- non-responders to cognitive-behavioral therapy
ume of a given structure, are not the only param- (CBT) [57]. Another study reinforced the idea of
eters implicating a certain region or network in the hippocampus (particularly, of increased right
the mediation of anxiety or other traits/states. hippocampal gray matter volume), along with a
Metabolic and functional aspects can add impor- differential activation of other structures, as a
tant evidence for understanding the contribution predictor of improved CBT outcome for PD [58].
of the hippocampus (or of any other brain struc- Such studies are strongly encouraged as they
ture) for the emotional processing characteristic open important avenues for the development of
of PD. In this particular, changes in metabolism more personalized and effective treatment strate-
in hippocampal [50] and parahippocampal [51] gies to psychiatric patients.
areas have been found in PD studies using single
photon emission computed tomography
(SPECT). The hippocampus, along with the 3.5 The Role
amygdala and insula, has also been shown to of the Hippocampus in PD:
present increased activation among PD patients Future Directions
in a functional MRI (fMRI) study [52]. Enhanced
hippocampal activation in response to a safety Due to their remarkable plasticity that is, to the
signal has also been revealed at baseline for ability to change morphologically and function-
medication-free PD patients with agoraphobia ally upon environmental demands and internal
[53]. Furthermore, reduced binding properties of signals hippocampal circuits are the subject of
the serotonergic receptor 5-HT1A has been found innumerous studies aiming at identifying effec-
in the hippocampus, amygdala, as well as in fron- tive interventions capable to positively regulate
tal and temporal cortical areas, of patients with cognition and emotion. In this context of neural
PD [54]. Reduced binding potential of the seroto- plasticity, one of the most noticeable features of
nergic transporter (5-HTT) has also been reported the hippocampus is its ability to continuously
in males with PD when compared with healthy generate functional neurons throughout the lifes-
males [55]. These data strongly support the idea pan of the individual.
that the hippocampus, especially at the metabolic Indeed, for many years, neuroscientists
and neurotransmission levels, present important believed that once the organism was born no fur-
components that contribute for the abnormal ther neuronal cells could be added to the brain. In
information processing typical of PD. Other stud- the 1960s, nevertheless, the technical possibility
ies, using magnetic resonance spectroscopy, also of identifying and tracking newly born neurons in
pointed for abnormalities of different neurotrans- the postnatal brain through autoradiography [59],
mitter and metabolites in PD, highlighting the and later, by the injection of the thymidine analog
involvement of the hippocampus in this disorder bromodeoxyuridine (BrdU) [60] opened a new
(reviewed in [56]). The relevance of understand- avenue in the study of how the brain works and
86 G.P. Dias and S. Thuret

how it can be modified. This represented the role in the ability of the DG to distinguish similar
breakthrough for a new mentality in neurobiologi- stimuli. This process is known as pattern separa-
cal sciences, based on the fundamental assump- tion, a mechanism without which the individual
tion that the adult brain could be shaped by the loses the ability to convert similar experiences
generation of new neurons capable to integrate into discrete, nonoverlapping representations.
into specific circuits. This ability, in turn, is thought to be an important
Adult neurogenesis occurs in the so-called factor underlying the development of anxiety dis-
neurogenic niches brain regions characterized orders [72].
by the presence of neural progenitor cells at con- Mostly important, it has been recently and
stant self-renewal activity and holding the poten- elegantly demonstrated that AHN is a process
tiality to differentiate into neurons in response to happening not only in rodents and monkeys, but
specific molecular signaling. In the mammalian also in humans. Previous postmortem analysis of
brain, these niches are the subventricular zone the human brain had already indicated the occur-
(SVZ) [61, 62], adjacent to the lateral ventricles, rence of AHN in our species [73] but not only
and the subgranular zone (SGZ) of the DG in the replication was needed; functional inferences
hippocampus [63, 64]. The generation of neurons were necessary as well, so that AHN could more
in this latter niche during postnatal life is, thus, definitely emerge as a form of relevant neural
known as adult hippocampal neurogenesis plasticity in humans. This came with the study by
(AHN) and is of special interest for our topic on Jonas Frisns group in 2013 [74]: in an inge-
mental health. nious way, these researchers took advantage of
AHN is a highly complex and intricately regu- the increased levels of C14 in the atmosphere after
lated process [65], counting on the influence of a the nuclear bomb tests of the 1950s and 60s to
number of different regulatory molecular signal- hypothesize that, if adult neurogenesis was true
ing pathways, many of which are still largely for humans, then the postmortem analysis of C14
unknown. These molecular pathways specifically in the brain should show neurons that were born
target the regulation of each of AHNs stages: (1) in years posterior to the individuals year of birth.
maintenance of the pool of progenitors; (2) cell Curiously, not only this was demonstrated to be
proliferation; (3) fate commitment to a neuronal the case, but also that: (1) this was shown primar-
phenotype; (4) acquisition of structural and func- ily to happen indeed in the hippocampus; (2)
tional characteristics of a mature granule neuron through mathematical modelling, the group could
(maturation); (5) survival; (6) integration into show that AHN takes place in the human brain at
pre-existing circuits. These, in turn, are believed similar rates to those found in the rodent brain.
to be the circuits upon regulation by the hippo- This is especially encouraging to behavioral neu-
campal formation, that is, circuits involved in roscientists, in that it gives support to the idea
cognitive functions (especially spatial and refer- that understanding the mechanisms underlying
ence working memory) and mood/anxiety. AHN and behavioral change in animals might
With effect, a number of papers have demon- generate knowledge that is also, to great extent,
strated that the lack of AHN induces cognitive applicable to humans.
impairment [66, 67], as well as depressive- [68] One of the greatest challenges to modern neu-
and anxiety-like [69] behavior in rodents. Despite roscience is, thus, to identify specific interven-
the consistent acknowledgement in the scientific tions and their respective mechanisms that are
community implicating neurogenesis deficits capable of upregulating AHN in health and dis-
with cognitive and mood dysfunction, not every ease. With effect, it has been demonstrated that
study have succeeded in showing such associa- AHN can be altered by: (1) different classes of
tion ([70]; reviewed in [71]). Nevertheless and drugs, such as hypnotics [75] and, classically, by
of special relevance for our topic on anxiety antidepressants [76, 77]; (2) environmental
there is more consensus in recognizing that the factors, like the exposure to an enriched environ-
newly generated neurons appear to have a crucial ment [78, 79] and physical exercise [80, 81]; (3)
3 The Hippocampus and Panic Disorder: Evidence from Animal and Human Studies 87

dietary factors (reviewed in [82]), such as with genetic component, which is the case of a
regard to meal intake (caloric restriction; [83]), multitude of mental illnesses, including PD [92].
meal frequency (intermittent fasting; [84]), meal The use of these cells offers a variety of relevant
texture (reviewed in [85]) and meal content advantages, such as: (1) the fact that it is not only
(reviewed in [86]; for example, polyphenol- or human, but patient-specific tissue; (2) collecting
omega-3 fatty acids-enriched diets high fat samples for the generation of iPSCs is a non-
diets). All these three levels of interventions have invasive procedure; (3) it allows for in vivo cor-
been shown, in animal studies, to be able to mod- relation of the cellular/molecular findings with
ify mental health-related behaviors likely via the those from the same patients in neuropsychologi-
induction of AHN. This opens a new avenue for cal assessment and neuroimaging evaluation. Of
studies using PD models. In this regard, to our special interest for our topic, it has been recently
knowledge, only one studied aimed at associating demonstrated that DG granule neurons can be
AHN to a panic-related response [87]. generated from human pluripotent stem cells
Specifically, these authors showed that chronic [93]. Subsequently, a next frontier in the study of
treatment with corticosterone induced anxiety the neurobiology of PD, particularly with regard
[88] and decreased AHN in rats [87], an effect to the contributions of the hippocampus, could be
that could be counteracted by the antidepressant the generation of DG neurons from iPSCs of PD
imipramine. However, the effects of chronic patients.
treatment with corticosterone on anxiety were
related to avoidance behavior, which has been
associated with generalized anxiety, and not to 3.6 Conclusion
escape behavior, a panic-related response.
Besides, chronic treatment with corticosterone The lifetime prevalence of PD is estimated to
has failed to induce specific panic-like behaviors; reach 3.7 %, with panic attacks reaching alarming
it can induce generalized anxiety- [88] and 22.7 % [94], thus posing the study of the biopsy-
depressive- [89] behavioral phenotypes, both of chosocial features related to this impairing disor-
which can be clinically comorbid with PD but are der as an important challenge for contemporary
not necessarily and specifically a feature of panic. science.
On the other hand, higher levels of awakening This chapter discussed empirical evidence on
cortisol have been found in PD patients [90] and the putative roles of the hippocampal formation
higher levels of corticosterone are largely known for the threat-biased emotional processing char-
to be associated with downregulation of AHN acteristic of PD. Although, as described, the lit-
[89]. Thus, the study of hippocampal neurogene- erature comprises growing evidence for a
sis in more specific animal models of PD, which contribution of the hippocampus in PD and other
might have high levels of corticosterone as one of anxious states, it is noteworthy that the complex
their biomarkers of stress, may render invaluable cognitive and emotional processing of any neuro-
responses on the mechanisms by which the hip- psychiatric illness results from abnormal func-
pocampus may contribute to anxiety in PD, as tioning of neural circuits encompassing a number
well as serve as a platform for the screening of of brain regions, each contributing for different
potentially effective interventions modifying aspects of the disease. The future of our under-
AHN and panic-related behaviors. standing of the neurobiological basis of PD, and
But animal models are not the only way to of other psychiatric disorders, lies thus on the
study the putative roles of AHN in PD. The active integration of knowledge not only about
recent advent of patient-specific induced pluripo- the brain regions involved, but also on their func-
tent stem cells (iPSCs) (reviewed in [91]) has tional connectivity. This is a dynamic result of
immensely added to the toolkit of scientists aim- intricate genetic and epigenetic factors regulating
ing to unravel the cellular and molecular sub- neurochemical, endocrine and behavioral sys-
strates of potentially any disease with a strong tems. Interestingly, if on the one hand these
88 G.P. Dias and S. Thuret

factors are at the very basis of psychiatric disorder, 17. Blanchard D, Blanchard RJ. Ethoexperimental
approaches to the biology of emotion. Annu Rev
on the other they also hold promise to be modifi-
Psychol. 1988;39:4368.
able targets for effective pharmacological, 18. Blanchard R, Nikulina JN, Sakai RR, McKittrick C,
psychotherapeutic/environmental and dietary McEwen B, Blanchard DC. Behavioral and endocrine
interventions. change following chronic predatory stress. Physiol
Behav. 1998;63(4):5619.
19. Morris R. Stress and the hippocampus. In: Andersen
P, Morris R, Amaral D, Bliss T, OKeefe J, editors.
References The hippocampus book. New York: Oxford University
Press; 2007. p. 75168.
1. Witter M, Canto CB, Couey JJ, Koganezawa N, 20. Shuhama R, Del-Ben CM, Loureiro SR, Graeff
O'Reilly KC. Architecture of spatial circuits in the FG. Animal defense strategies and anxiety disorders.
hippocampal region. Philos Trans R Soc Lond B Biol An Acad Bras Cinc. 2007;79(1):97109.
Sci. 2013;369(1635):20120515. 21. Knierim J, Neunuebel JP, Deshmukh SS. Functional
2. Pearce J. Ammons horn and the hippocampus. correlates of the lateral and medial entorhinal cortex:
J Neurol Neurosurg Psychiatry. 2001;71(3):351. objects, path integration and local-global reference
3. Destrieux C, Bourry D, Velut S. Surgical anatomy of frames. Philos Trans R Soc Lond B Biol Sci.
the hippocampus. Neurochirurgie. 2013;59(45): 2013;369(1635):20130369.
14958. 22. Anagnostaras S, Craske MG, Fanselow MS. Anxiety:
4. Szirmai I, Buzski G, Kamondi A. 120 years of hip- at the intersection of genes and experience. Nat
pocampal Schaffer collaterals. Hippocampus. 2012; Neurosci. 1999;2(9):7802.
22(7):150816. 23. Maren S, Fanselow MS. Synaptic plasticity in the
5. Amaral D, Lavenex P. Hippocampal neuroanatomy. basolateral amygdala induced by hippocampal forma-
In: Andersen P, Morris R, Amaral D, Bliss T, OKeefe tion stimulation in vivo. J Neurosci.
J, editors. The hippocampus book. New York: Oxford 1995;15(11):754864.
University Press; 2007. p. 37114. 24. Phillips R, Ledoux J. Differential contribution of
6. Doetsch F, Hen R. Young and excitable: the function amygdala and hippocampus to cued and contextual fear
of new neurons in the adult mammalian brain. Curr conditioning. Behav Neurosci. 1992;106(2):27485.
Opin Neurobiol. 2005;15(1):1218. 25. Anagnostaras S, Maren S, Fanselow MS. Temporally
7. Fanselow M, Dong H. Are the dorsal and ventral hip- graded retrograde amnesia of contextual fear after
pocampus functionally distinct structures? Neuron. hippocampal damage in rats: within-subjects exami-
2010;65:719. nation. J Neurosci. 1999;19(3):110614.
8. Kheirbek M, Hen R. Dorsal vs ventral hippocampal 26. Ledoux J. The emotional brain: the mysterious under-
neurogenesis: implications for cognition and mood. pinnings of emotional life. New York: Phoenix; 1999.
Neuropsychopharmacology. 2011;36(1):3734. 27. Behrendt R. Situationally appropriate behavior: trans-
9. Tanti A, Rainer Q, Minier F, Surget A, Belzung lating situations into appetitive behavior modes. Rev
C. Differential environmental regulation of neurogen- Neurosci. 2013;24(6):577606.
esis along the septo-temporal axis of the hippocam- 28. Dine J, Ionescu IA, Stepan J, Yen YC, Holsboer F,
pus. Neuropharmacology. 2012;63(3):37484. Landgraf R, et al. Identification of a role for the ven-
10. Sahay A, Hen R. Adult hippocampal neurogenesis in tral hippocampus in neuropeptide S-elicited anxioly-
depression. Nat Neurosci. 2007;10(9):11105. sis. PLoS One. 2013;8(3):e60219.
11. Squire L. Memory and the hippocampus: a synthesis 29. Gonzalez L, Ouagazzal AM, File SE. Stimulation of
from findings with rats, monkeys, and humans. benzodiazepine receptors in the dorsal hippocampus
Psychol Rev. 1992;99:195231. and median raphe reveals differential GABAergic
12. Squire L, Ojemann JG, Miezin FM, Petersen SE, control in two animal test of anxiety. Eur J Neurosci.
Videen TO, Raichle ME. Activation of the hippocam- 1998;10(12):367380.
pus in normal humans: a functional anatomical study of 30. Nazar M, Jessa M, Panik A. Benzodiazepine-
memory. Proc Natl Acad Sci USA. 1992;89:183741. GABAA receptor complex ligands in two models of
13. Squire L. The hippocampus and spatial memory. anxiety. J Neural Transm. 1997;104(67):73346.
Trends Neurosci. 1993;16:567. 31. Rezayat M, Roohbakhsh A, Zarrindast MR, Massoudi
14. Kosaki Y, Lin TC, Horne MR, Pearce JM, Gilroy R, Djahanguiri B. Cholecystokinin and GABA inter-
KE. The role of the hippocampus in passive and action in the dorsal hippocampus of rats in the ele-
active spatial learning. Hippocampus. 2014;24(12): vated plus-maze test of anxiety. Physiol Behav.
163352. 2005;84(5):77582.
15. Squire L. The legacy of patient H.M. for neurosci- 32. Strange B, Witter MP, Lein ES, Moser EI. Functional
ence. Neuron. 2009;61(1):69. organization of the hippocampal longitudinal axis.
16. Gorman J, Liebowitz MR, Fyer AJ, Stein J. A neuro- Nat Rev Neurosci. 2014;15(10):65569.
anatomical hypothesis for panic disorder. Am 33. Johnson P, Shekhar A. An animal model of panic vul-
J Psychiatry. 1989;146(2):14861. nerability with chronic disinhibition of the dorsomedial/
3 The Hippocampus and Panic Disorder: Evidence from Animal and Human Studies 89

perifornical hypothalamus. Physiol Behav. 2012; and voxel-based morphometric MRI study. Psychol
107(5):68698. Med. 2010;40(11):187986.
34. Bergold P, Pinkhasova V, Syed M, Kao HY, Jozwicka 46. Vythilingam M, Anderson ER, Goddard A, Woods SW,
A, Zhao N, et al. Production of panic-like symptoms Staib LH, Charney DS, et al. Temporal lobe volume in
by lactate is associated with increased neural firing panic disordera quantitative magnetic resonance
and oxidation of brain redox in the rat hippocampus. imaging study. Psychiatry Res. 2000;99(2):7582.
Neurosci Lett. 2009;453(3):21924. 47. Lai C. Hippocampal and subcortical alterations of
35. Muios-Gimeno MGM, Kagerbauer B, Martn- first-episode, medication-nave major depressive dis-
Santos R, Navins R, Alonso P, et al. Allele variants order with panic disorder patients. J Neuropsychiatry
in functional MicroRNA target sites of the neuro- Clin Neurosci. 2014;26(2):1429.
trophin-3 receptor gene (NTRK3) as susceptibility 48. Massana G, Serra-Grabulosa JM, Salgado-Pineda P,
factors for anxiety disorders. Hum Mutat. Gast C, Junqu C, Massana J, et al. Parahippocampal
2009;30(7):106271. gray matter density in panic disorder: a voxel-based
36. Dierssen M, Gratacs M, Sahn I, Martn M, Gallego morphometric study. Am J Psychiatry.
X, Amador-Arjona A, et al. Transgenic mice overex- 2003;160(3):5668.
pressing the full-length neurotrophin receptor TrkC 49. Del Casale A, Serata D, Rapinesi C, Kotzalidis GD,
exhibit increased catecholaminergic neuron density in Angeletti G, Tatarelli R, et al. Structural neuroimag-
specific brain areas and increased anxiety-like behav- ing in patients with panic disorder: findings and limi-
ior and panic reaction. Neurobiol Dis. 2006;24(2): tations of recent studies. Psychiatr Danub. 2013;
40318. 25(2):10814.
37. Amador-Arjona A, Delgado-Morales R, Belda X, 50. Sakai Y, Kumano H, Nishikawa M, Sakano Y, Kaiya
Gagliano H, Gallego X, Keck ME, et al. Susceptibility H, Imabayashi E, et al. Cerebral glucose metabolism
to stress in transgenic mice overexpressing TrkC, a associated with a fear network in panic disorder.
model of panic disorder. J Psychiatr Res. 2010;44(3): Neuroreport. 2005;16(9):92731.
15767. 51. Koh K, Kang JI, Lee JD, Lee YJ. Shared neural activ-
38. Santos M, D'Amico D, Spadoni O, Amador-Arjona A, ity in panic disorder and undifferentiated somatoform
Stork O, Dierssen M. Hippocampal hyperexcitability disorder compared with healthy controls. J Clin
underlies enhanced fear memories in TgNTRK3, a Psychiatry. 2010;71(12):157681.
panic disorder mouse model. J Neurosci. 2013;33(38): 52. Wittmann A, Schlagenhauf F, John T, Guhn A,
1525971. Rehbein H, Siegmund A, et al. A new paradigm
39. Temel Y, Blokland A, Lim LW. Deactivation of the (Westphal-Paradigm) to study the neural correlates of
parvalbumin-positive interneurons in the hippocam- panic disorder with agoraphobia. Eur Arch Psychiatry
pus after fear-like behaviour following electrical Clin Neurosci. 2011;261(3):18594.
stimulation of the dorsolateral periaqueductal gray of 53. Lueken U, Straube B, Konrad C, Wittchen HU,
rats. Behav Brain Res. 2012;233(2):3225. Strhle A, Wittmann A, et al. Neural substrates of
40. Lim L, Temel Y, Visser-Vandewalle V, Blokland A, treatment response to cognitive-behavioral therapy in
Steinbusch H. Fos immunoreactivity in the rat fore- panic disorder with agoraphobia. Am J Psychiatry.
brain induced by electrical stimulation of the dorsolat- 2013;170(11):134555.
eral periaqueductal gray matter. J Chem Neuroanat. 54. Nash J, Sargent PA, Rabiner EA, Hood SD,
2009;38(2):8396. Argyropoulos SV, Potokar JP, et al. Serotonin
41. de Carvalho M, Dias GP, Cosci F. de-Melo-Neto VL, 5-HT1A receptor binding in people with panic disor-
Bevilaqua MC, Gardino PF, et al. Current findings of der: positron emission tomography study. Br
fMRI in panic disorder: contributions for the fear neu- J Psychiatry. 2008;193(3):22934.
rocircuitry and CBT effects. Expert Rev Neurother. 55. Maron E, Tru I, Hirvonen J, Tuominen L, Lumme V,
2010;10(2):291303. Vasar V, et al. Gender differences in brain serotonin
42. de Carvalho M, Rozenthal M, Nardi AE. The fear cir- transporter availability in panic disorder.
cuitry in panic disorder and its modulation by J Psychopharmacol. 2011;25(7):9529.
cognitive-behaviour therapy interventions. World 56. Pannekoek J, van der Werff SJ, Stein DJ, van der Wee
J Biol Psychiatry. 2010;11(2 Pt 2):18898. NJ. Advances in the neuroimaging of panic disorder.
43. Shin L, Liberzon I. The neurocircuitry of fear, stress, Hum Psychopharmacol. 2013;28(6):60811.
and anxiety disorders. Neuropsychopharmacology. 57. Ball T, Stein MB, Ramsawh HJ, Campbell-Sills L,
2010;35(1):16991. Paulus MP. Single-subject anxiety treatment outcome
44. Dresler T, Guhn A, Tupak SV, Ehlis AC, Herrmann prediction using functional neuroimaging.
MJ, Fallgatter AJ, et al. Revise the revised? New Neuropsychopharmacology. 2014;39(5):125461.
dimensions of the neuroanatomical hypothesis of 58. Reinecke A, Thilo K, Filippini N, Croft A, Harmer
panic disorder. J Neural Transm. 2013;120(1):329. CJ. Predicting rapid response to cognitive-behavioural
45. Sobanski T, Wagner G, Peikert G, Gruhn U, Schluttig treatment for panic disorder: the role of hippocampus,
K, Sauer H, et al. Temporal and right frontal lobe insula, and dorsolateral prefrontal cortex. Behav Res
alterations in panic disorder: a quantitative volumetric Ther. 2014;S0005-7967(14):00120-X.
90 G.P. Dias and S. Thuret

59. Altman J. Autoradiographic investigation of cell pro- campal neurogenesis in adult humans. Cell.
liferation in the brains of rats and cats. Anat Rec. 2013;153(6):121927.
1963;145(4):57391. 75. Methippara M, Bashir T, Suntsova N, Szymusiak R,
60. Nowakowski R, Lewin SB, Miller MW. McGinty D. Hippocampal adult neurogenesis is
Bromodeoxyuridine immunohistochemical determi- enhanced by chronic eszopiclone treatment in rats.
nation of the lengths of the cell cycle and the DNA- J Sleep Res. 2010;19(3):38493.
synthetic phase for an anatomically defined 76. Su X, Li XY, Banasr M, Duman RS. Eszopiclone
population. J Neurocytol. 1989;18(3):3118. and fluoxetine enhance the survival of newborn
61. Petreanu L, Alvarez-Buylla A. Maturation and death neurons in the adult rat hippocampus. Int
of adult-born olfactory bulb granule neurons: role of J Neuropsychopharmacol. 2009;12(10):14218.
olfaction. J Neurosci. 2002;22(14):610613. 77. Mateus-Pinheiro A, Pinto L, Bessa JM, Morais M,
62. Ernst A, Alkass K, Bernard S, Salehpour M, Perl S, Alves ND, Monteiro S, et al. Sustained remission
Tisdale J, et al. Neurogenesis in the striatum of the from depressive-like behavior depends on hippocam-
adult human brain. Cell. 2014;156(5):107283. pal neurogenesis. Transl Psychiatry. 2013;3:e210.
63. Alvarez-Buylla A, Lim DA. For the long run: main- 78. Monteiro B, Moreira FA, Massensini AR, Moraes
taining germinal niches in the adult brain. Neuron. MF, Pereira GS. Enriched environment increases neu-
2004;41(5):6836. rogenesis and improves social memory persistence in
64. van Praag H, Schinder AF, Christie BR, Toni N, socially isolated adult mice. Hippocampus.
Palmer TD, Gage FH. Functional neurogenesis in the 2014;24(2):23948.
adult hippocampus. Nature. 2002;415(6875):10304. 79. Hattori S, Hashimoto R, Miyakawa T, Yamanaka H,
65. Mu Y, Lee SW, Gage FH. Signalling in adult neuro- Maeno H, Wada K, et al. Enriched environments
genesis. Curr Opin Neurobiol. 2010;20:41623. influence depression-related behavior in adult mice
66. Rola R, Raber J, Rizk A, Otsuka S, VandenBerg SR, and the survival of newborn cells in their hippocampi.
Morhardt DR, et al. Radiation-induced impairment of Behav Brain Res. 2007;180(1):6976.
hippocampal neurogenesis is associated with cogni- 80. Farioli-Vecchioli S, Mattera A, Micheli L, Ceccarelli
tive deficits in young mice. Exp Neurol. 2004; M, Leonardi L, Saraulli D, et al. Running rescues
188(2):31630. defective adult neurogenesis by shortening the length
67. Villeda S, Luo J, Mosher KI, Zou B, Britschgi M, of the cell cycle of neural stem and progenitor cells.
Bieri G, et al. The ageing systemic milieu negatively Stem Cells. 2014;32(7):196882.
regulates neurogenesis and cognitive function. Nature. 81. Winocur G, Wojtowicz JM, Huang J, Tannock
2011;477(7362):904. IF. Physical exercise prevents suppression of
68. Snyder J, Soumier A, Brewer M, Pickel J, Cameron hippocampal neurogenesis and reduces cognitive
HA. Adult hippocampal neurogenesis buffers stress impairment in chemotherapy-treated rats.
responses and depressive behaviour. Nature. 2011; Psychopharmacology (Berl). 2014;231(11):231120.
476(7361):45861. 82. Murphy T, Dias GP, Thuret S. Effects of diet on brain
69. Revest J, Dupret D, Koehl M, Funk-Reiter C, Grosjean plasticity in animal and human studies: mind the gap.
N, Piazza PV, et al. Adult hippocampal neurogenesis Neural Plast. 2014;2014:563160.
is involved in anxiety-related behaviors. Mol 83. Park J, Glass Z, Sayed K, Michurina TV, Lazutkin A,
Psychiatry. 2009;14(10):95967. Mineyeva O, et al. Calorie restriction alleviates the age-
70. Groves J, Leslie I, Huang GJ, McHugh SB, Taylor A, related decrease in neural progenitor cell division in the
Mott R, et al. Ablating adult neurogenesis in the rat aging brain. Eur J Neurosci. 2013;37(12):198793.
has no effect on spatial processing: evidence from a 84. Mattson M, Duan W, Guo Z. Meal size and frequency
novel pharmacogenetic model. PLoS Genet. affect neuronal plasticity and vulnerability to disease:
2013;9(9):e1003718. cellular and molecular mechanisms. J Neurochem.
71. Tanti A, Belzung C. Hippocampal neurogenesis: a 2003;84(3):41731.
biomarker for depression or antidepressant effects? 85. Zainuddin M, Thuret S. Nutrition, adult hippocampal
Methodological considerations and perspectives for neurogenesis and mental health. Br Med Bull.
future research. Cell Tissue Res. 2013;354(1): 2012;103(1):89114.
20319. 86. Dias G, Cavegn N, Nix A, do Nascimento Bevilaqua
72. Sahay A, Scobie KN, Hill AS, O'Carroll CM, MC, Stangl D, Zainuddin MS, et al. The role of
Kheirbek MA, Burghardt NS, et al. Increasing adult dietary polyphenols on adult hippocampal neurogen-
hippocampal neurogenesis is sufficient to improve esis: molecular mechanisms and behavioural effects
pattern separation. Nature. 2011;472(7344):46670. on depression and anxiety. Oxid Med Cell Longev.
73. Eriksson P, Perfilieva E, Bjrk-Eriksson T, Alborn 2012;2012:541971.
AM, Nordborg C, Peterson DA, et al. Neurogenesis in 87. Diniz L, dos Santos TB, Britto LR, Cspedes IC,
the adult human hippocampus. Nat Med. Garcia MC, Spadari-Bratfisch RC, et al. Effects of
1998;4(11):13137. chronic treatment with corticosterone and imipramine
74. Spalding K, Bergmann O, Alkass K, Bernard S, on fos immunoreactivity and adult hippocampal neu-
Salehpour M, Huttner HB, et al. Dynamics of hippo- rogenesis. Behav Brain Res. 2013;238:1707.
3 The Hippocampus and Panic Disorder: Evidence from Animal and Human Studies 91

88. Diniz L, Dos Reis BB, de Castro GM, Medalha CC, induced pluripotent stem cells for the modelling of
Viana MB. Effects of chronic corticosterone and Autistic Spectrum Disorders. Psychopharmacology
imipramine administration on panic and anxiety- (Berl). 2013;231(6):107988.
related responses. Braz J Med Biol Res. 2011; 92. Gregersen N, Dahl HA, Buttenschn HN, Nyegaard
44(10):104853. M, Hedemand A, Als TD, et al. A genome-wide study
89. Murray F, Smith DW, Hutson PH. Chronic low dose of panic disorder suggests the amiloride-sensitive cat-
corticosterone exposure decreased hippocampal cell ion channel 1 as a candidate gene. Eur J Hum Genet.
proliferation, volume and induced anxiety and depres- 2012;20(1):8490.
sion like behaviours in mice. Eur J Pharmacol. 93. Yu D, Marchetto MC, Gage FH. How to make a hip-
2008;583(1):11527. pocampal dentate gyrus granule neuron. Development.
90. Vreeburg S, Zitman FG, van Pelt J, Derijk RH, 2014;141(12):236675.
Verhagen JC, van Dyck R, et al. Salivary cortisol lev- 94. Kessler R, Chiu WT, Jin R, Ruscio AM, Shear K,
els in persons with and without different anxiety dis- Walters EE. The epidemiology of panic attacks, panic
orders. Psychosom Med. 2010;72(4):3407. disorder, and agoraphobia in the National Comorbidity
91. Cocks G, Curran S, Gami P, Uwanogho D, Jeffries Survey Replication. Arch Gen Psychiatry. 2006;63(4):
AR, Kathuria A, et al. The utility of patient specific 41524.
Panic Disorder, Is It Really a Mental
Disorder? From Body Functions 4
to the Homeostatic Brain

Giampaolo Perna, Giuseppe Iannone,

Tatiana Torti, and Daniela Caldirola

Contents Abstract
4.1 Introduction 94 Panic disorder (PD), is characterized by
repeated PAs (i.e. abrupt surges of anxiety and
4.2 Respiration in Panic Disorder 95
fear accompanied by physical e.g. pounding
4.3 Cardiovascular System and PD 99 heart, sweating, trembling, etc. and cogni-
4.4 Panic Disorder tive e.g. fear of dying, fear of losing control,
and the Balance System 101 etc. symptoms that usually reach their peak
4.4.1 Balance and the Visual System within 10 min), and major changes in behavior
in PD 103
or persistent anxiety over having further attacks
4.5 Photosensitivity and PD 105 for at least 1 month. Since PD can be treated
4.6 Conclusions and Clinical Implications 105 with psychotropic drugs and/or psychotherapy
References 107
it has been commonly considered a mental dis-
order. However, recent evidence indicates that
patients with PD exhibit subclinical anomalies
in the respiratory, cardiac, and balance systems.
In addition, apart from reducing panic symp-
toms, many antipanic pharmacotherapies
(e.g. SSRIs) improve the functioning of the

G. Iannone D. Caldirola
G. Perna (*)
Department of Clinical Neurosciences, Villa San
Department of Clinical Neurosciences, Villa San
Benedetto Menni, Hermanas Hospitalarias, FoRiPsi,
Benedetto Menni, Hermanas Hospitalarias, FoRiPsi,
Albese con Cassano, Italy
Albese con Cassano, Italy
e-mail: g.iannone@alumni.maastrichtuniversity.nl;
Department of Psychiatry and Neuropsychology, caldiroladaniela@gmail.com
Maastricht University, Maastricht, The Netherlands
T. Torti
Department of Psychiatry and Behavioral Sciences, Department of Clinical Neurosciences, Villa San
Leonard Miller School of Medicine, University of Benedetto Menni, Hermanas Hospitalarias, FoRiPsi,
Miami, Miami, FL, USA Albese con Cassano, Italy
AIAMC (Italian Association for Behavioural AIAMC (Italian Association for Behavioural
Analysis, Modification and Behavioural Analysis, Modification and Behavioural and
and Cognitive Therapies), Milan, Italy Cognitive Therapies), Milan, Italy
e-mail: pernagp@gmail.com e-mail: tatiana.torti@gmail.com

Springer International Publishing Switzerland 2016 93

A.E. Nardi, R.C.R. Freire (eds.), Panic Disorder, DOI 10.1007/978-3-319-12538-1_4
94 G. Perna et al.

abovementioned systems. Therefore, some existence of hypersensitive alarm systems in

authors believe PAs may be real alarms arising PD. According to Klein PAs occur when the suf-
from transient instability of homeostatic body focation alarm system is erroneously triggered [5];
functions. The idea PD is a mere psychiatric Gormans neuroanatomical model postulates PAs
disease may be challenged by acknowledging a are conditioned fear responses mediated by an
paramount role also to aberrant homeostatic overly sensitive fear network [6, 7]; the three-
functioning. This might pave the way to a more alarms (true, false, learned) theory deems PAs as
integrated approach of treating PD. the results of both spontaneous firing of the fear
system and conditioning processes to internal or
external cues [8]. Clark et al. [9] consider PAs
Keywords catastrophic misinterpretations of harmless bodily
Panic disorder Homeostasis Respiration sensations. These theories share the assumption
Cardiac system Balance Vision that alarms are false because patients with PD are
Photosensitivity physically healthy. (2) PD is treated with psycho-
tropic drugs and/or psychotherapies. Withal there
has been some debate whether to consider PD just
a mental disorder.
4.1 Introduction Beside cognitive symptoms, patients with PD
often complain of several somatic symptoms
According to the latest edition of the Diagnostic including respiratory difficulties, irregular heart-
and Statistical Manual of Mental Disorders (DSM- beat, dizziness, and photophobia. Usually physi-
5) panic disorder (PD) is an anxiety disorder char- cians and psychiatrists after conducting standard
acterized by recurrent unexpected panic attacks procedures (such as physical and/or clinical tests)
(PAs), consisting of physical and cognitive symp- reassure patients that their bodies function per-
toms such as palpitations, dyspnea, dizziness, fectly, and they ascribe the somatic symptoms
derealisation, fear of losing control, and fear of entirely to anxiety. Conversely numerous scien-
dying, that surge abruptly and that reach a peak tific findings suggest that patients with PD may
within minutes, provoking intense fear or discom- suffer from subclinical abnormal organic systems
fort. Beyond the PAs themselves, which are the functioning (e.g. in the cardio-respiratory and the
hallmark of the disorder, another key feature of PD balance systems), which may be associated with
is fear (1 month at least) of having future PAs, hyperreactivity to hypercapnic and hypoxic inha-
which can lead to important maladaptive changes lations, subclinical autonomic hyperreactivity,
in behaviour (i.e. anticipatory anxiety and phobic and space and motion discomfort.
avoidance of places and situations where an attack In the next paragraphs we examine evidence
has occurred or where patients believe it may that patients with PD may manifest subtle physi-
occur). PD frequently occurs in comorbidity with ological functions abnormalities as well as
other mental disorders, such as agoraphobia, major reduced adaptability to changes, and that their
depression disorder, and bipolar disorder, it can be brains are more akin to react when these systems
chronic and disabling, can cause distress and are stimulated. In particular, we will focus on
impair quality of life [1]. PD has a lifetime preva- three main systems: the cardiac, the respiratory,
lence of approximately 3.5 % in the general popu- and the balance systems. We hypothesize that
lation and 58 % in primary care settings [24]. PAs are true alarms signaling aberrant function-
Etiology of PD has not been fully unfolded; how- ing of one or more of these body systems. Finally
ever research suggests interaction of genetic pre- we speculate that patients with PD are physiolog-
disposition and specific environmental factors. PD ically different from healthy subjects and there-
is considered a mental disorder because: (1) PAs fore PD may be reconsidered from two non
are deemed false alarms. It has been more than mutually-exclusive perspectives (a somatic one
20 years since germinal theories postulated the and a psychological one) and that patients may
4 Panic Disorder, Is It Really a Mental Disorder? From Body Functions to the Homeostatic Brain 95

benefit from therapeutic approaches that simulta-

neously take into account both facets of the Box 4.1. Panic Disorder
disorder. Respiration:
Respiratory symptoms and respiratory
diseases are very frequent.
4.2 Respiration in Panic Disorder Hypersensitivity to hypercapnic gas-
mixture inhalation.
Breathing is involved in the phenomenology and Baseline respiratory irregularity (e.g.
the biological mechanisms of PD. Both experi- chronic hyperventilation).
mental and clinical observations provide strong Respiratory PD subtype.
evidence that a connection between panic and
respiration does exist [10, 11]. Respiratory symp- Cardiac system:
toms are very frequent among patients with PD Comorbidity with cardiac disorders.
[12], both during spontaneous PAs and during Comorbidity with coronary artery
daily-life [13], and constitute a hallmark of PAs disease.
and PD. Further information corroborates this Decreased cardiac vagal function.
correlation. First, an association between PD and Higher exercise avoidance and worse
hyperventilation is known. Indeed, several stud- cardiopulmonary performance.
ies reported low resting partial pressure of carbon
dioxide (CO2) in patients with PD [14, 15]. Also Balance and vision:
up to 40 % of patients with PD suffers from High prevalence of vestibular symptoms.
hyperventilation and both patients with PAs and Balance control mainly relies on non-
patients with hyperventilation syndrome exhibit vestibular cues, especially visual cues.
similar respiratory symptoms, such as dyspnea
[16]. Hence it might be that hyperventilation Photosensitivity:
causes PAs in patients suffering from Lowered threshold of tolerance to light.
PD. However chronic hyperventilation (and Photophobic behavior.
hypocapnia) has been reported in less than 50 %
of the patients with PD and it does not seem to be
unique of PD. In fact it is also prevalent in PAs. Given that during the PA, tidal volume (and
patients suffering from other anxiety disorders heart rate) increased and pCO2 dropped, suggest-
and it may just reflect background anxiety [17]. ing hyperventilation, it might plausible to assume
Hyperventilation seems to induce anxiety, but not that hyperventilation is merely a consequence
PAs, in patients with PD [18]. Panic precedes (and probably a compensatory mechanism) rather
hyperventilation, as it is the case during hyper- than a cause of PAs. Skin conductance levels
capnic challenges [19], and a review claimed that raised in the hour preceding PAs as well as during
a causal role of hyperventilation in the etiology the attacks. These changes were largely absent
of PD seems unlikely [11]. Recently Meuret et al. when an attack did not occur. These findings sug-
[20] examined changes in respiration, heart rate, gest that unexpected PAs are preceded by signifi-
and skin conductance level 60 min before and cant autonomic irregularities and invite to rethink
10 min after PAs in individuals suffering from the classic nosotaxy between situational and
PD. The authors observed important patterns of spontaneous PAs.
autonomic and respiratory irregularity that were Second, an association between PD and respi-
not detected by the patients, as early as 47 min ratory diseases was found. Goodwin, Pine [21]
before panic onset. Respiratory changes, such as documented an association between self-
decreased tidal volume and pCO2 increase, char- reported respiratory diseases and increased like-
acterized the final minutes preceding a PA. These lihood of PAs among adults in the general
findings suggest that hypoventilation precedes population, also when controlling for differences
96 G. Perna et al.

in sociodemographic characteristics, physical the top of that, absence of hypothalamic pituitary

illnesses, and comorbid mental disorders adrenal activation during PAs contrasts the
(adjusted OR = 1,795 % CI) and concluded that assumption that panic is a manifestation of a
this association is specific to PAs. Up to 40 % of hypersensitive fear system [28]. However both
patients with PD have a history of respiratory normal and increased hormonal activity has been
disease, in particular asthma and bronchitis [11]. reported in PD [29]. It might be that HPA-axis
PD prevalence in patients suffering from chronic abnormal function may render some individuals
obstructive pulmonary disease is higher both more vulnerable to PD (and to psychopathology
than in people suffering from other mental disor- in general) over time, perhaps by increasing vul-
ders (i.e. obsessive-compulsive, depressive, and nerability to future stressors.
eating disorders) and than in healthy controls Children suffering from anxiety disorders
[22, 23]. The nature of this association remains exhibit greater changes in somatic symptoms
largely unknown. It may be that either PAs pre- after CO2 inhalation and those who developed
cede the onset of respiratory disease or that panic symptoms manifest respiratory rate
respiratory or lung disease lead to the develop- increases in response to CO2 breathing and ele-
ment of PAs. Alternatively a third factor (e.g. vated mean tidal volume levels, and higher respi-
cigarette smoking) may increase co-occurrence ratory rate variability during room-air breathing
of respiratory disease and PAs. Future studies [30]. Behavioral hyperreactivity to CO2 has been
that identify genetic and/or environmental routes found in healthy first-degree relatives of patients
of transmission are warranted to determine the with PD. Finally, even mentally healthy individu-
specific mechanisms of this association, albeit als with first-degree relatives suffering from PD
mounting evidence indicates that respiratory ill- experience behavioral and respiratory hyperac-
ness precedes PD. tivity in response to CO2 administration when
Third, patients with PD showed behavioural compared to controls without such a family his-
and respiratory hypersensitivity to hypercapnic tory [31, 32].
gas-mixture inhalation [24, 25] and subclinical Our group investigated the breath-by-breath
abnormalities in respiratory patterns. Klein [5] complexity of respiration dynamics in patients
speculated that PAs are false suffocations with PD and in healthy controls. We found that
alarms resulting from carbon dioxide hypersen- patients exhibited greater baseline respiratory
sitivity that induces the brains suffocation irregularity, which may be a vulnerability factor
monitor to erroneously signal lack of air and to PAs [33]. We also found higher respiratory
breathlessness. This might lead to respiratory dis- irregularity in children of patients with PD when
tress, hyperventilation and, eventually, to a compared to children of psychiatrically healthy
PA. Preter and Klein [26] further developed the parents, even when children with anxiety disor-
suffocation false alarm theory and proposed that ders were excluded. Hence irregular breathing
endogenous opioidergic regulation dysfunctions may represent a risk marker of familial vulner-
increase suffocation sensitivity, separation anxi- ability to PD that is independent of state
ety, and PAs. Just recently, the same authors dem- effects and it may anticipate PD [34]. In addi-
onstrated an association between endogenous tion, subjects with PD show increased respira-
opioid system deficiency, panic-like suffocation tory variability during mild physical activity
sensitivity, and childhood parental loss. Finally, even in the absence of full blown respiratory
this theory reshaped the role of the amygdala diseases [24, 35].
(and in general of the fear system) in PD. This is In a recent meta-analytic review our research
consistent with findings of patients with focal group compared baseline respiratory and
bilateral amygdale lesions who also exhibited hematic parameters related to the respiratory
PAs in response to CO2 inhalation [27]. Hence, function in subjects with PD and in healthy con-
an alternative alarm system beyond the amygdala trols. We found higher baseline mean minute
is likely to underlie hypersensitivity to CO2. On ventilation (MV), and lower end-tidal partial
4 Panic Disorder, Is It Really a Mental Disorder? From Body Functions to the Homeostatic Brain 97

pressure (et-pCO2) and venous pCO2 in subjects that state anxiety during respiratory assessment
with PD, which indicated a condition of base- does not fully explain these respiratory abnor-
line hyperventilation. In addition, reduced malities. Yet, patients with PD exhibit specific
HCO3 (bicarbonate ion) and PO4 (phosphate emotional, cognitive and behavioral characteris-
ion) venous concentrations suggest that hyper- tics, such as fear of somatic sensations, panic-
ventilation may be chronic and not just related specific beliefs, and phobic/protective behaviors.
to higher anxiety states during the respiratory Given that these features are not completely
assessment. Finally preliminary indication of captured by trait/state anxiety measurements,
higher MV, respiration rate (RR) and tidal vol- they may specifically influence respiration in
ume (TV) variability, higher RR and TV irregu- this population.
larity, and higher rate of sighs and apneas in the Several hypotheses have been advanced to
respiratory patterns of these subjects was found elucidate the origin of respiratory irregularity in
[14]. Whether these respiratory abnormalities PD. It is known that complex regulatory systems
are peculiar to PD or whether they are common modulate respiration. Many brainstem regions
to other anxiety disorders is currently debated. containing CO2-sensitive neurons are implicated
On the one hand it is known that anxiety influ- in regulating both ventilation and panic [40, 41].
ences respiration [36], therefore respiratory Hence it is plausible to hypothesize that some
abnormalities may be not specific to PD but they overlap exist between neurons that serve respira-
might be common also in subjects suffering tion and those that elicit panic. Protopopescu
from other anxiety disorders. Previous research et al. [42] found increased brainstem gray matter
on baseline respiratory parameters (especially volume in the ventral and dorsal midbrain, and in
mean RR, TV, MV, and et-pCO2) in subjects the rostral pons of patients with PD compared
with PD and in subjects suffering from other with healthy controls. In a recent review, Perna
anxiety disorders yielded conflicting results et al. [43] also suggested that the brainstem vol-
[3739]. Just recently in a meta-analysis our ume is larger in patients with PD and that aber-
research team observed significant differences rant functioning of the brainstem serotonergic
between the baseline respiratory parameters in system (i.e. altered serotonergic receptors and
subjects with PD and in subjects with social 5-HT-transporter bindings) is likely to be
phobia (SP) or generalized anxiety disorder involved in panic modulation as well as in (car-
(GAD). We found significantly lower mean end- dio)respiratory activity. The pre-Btzinger com-
tidal partial pressure of CO2 (et-pCO2) in sub- plex is a cluster of brainstem neurons that
jects with PD than in those with SP or GAD, and contains the basic circuits for respiratory rhythm
higher mean respiratory rate, lower venous et- and pattern generation [44]. Breathing irregulari-
pCO2 and HCO3 concentration in subjects with ties may arise from an intrinsic deranged activity
PD than in those with SP. These findings sug- of the pre-Btzinger complex neurons, that fail to
gested that subjects with PD have a condition of adequately cope with external stimuli and that
baseline hyperventilation when compared to would underlie respiratory phenomena, such as
subjects with SP or GAD. Hematic variables sighs and gasps, and/or from compensatory-like
suggested that the hyperventilation may be responses to abnormal central and/or peripheral
chronic. Conversely, we did not find significant signals (the brainstem communicates with both
differences in respiratory abnormalities between sensory afferents from central and peripherical
subjects with SP or GAD and healthy controls chemoreceptors and from pulmonary/chest wall
[15]. These results support the idea that baseline receptors). In addition, more rostral areas such as
respiratory abnormalities are specific to PD. It is the hypothalamus, cerebellum and cortex, modu-
still not clear whether hyperventilation may late respiration across different physiological
arise from an intrinsic malfunction of the respi- states [45, 46].
ratory system or rather from panic-related anxi- The limbic circuit regulates respiration dur-
ety. Chronic hyperventilation in PD indicates ing arousal and emotional states. The dorsal
98 G. Perna et al.

periaqueductal gray seems to regulate uncondi- a trait marker vulnerability of PD is a question

tioned defensive responses to proximal threats, that deserves a conclusive and definite answer.
including physical stimuli, therefore it might be So far, preliminary findings supported the latter
implicated in PD [47]. Respiratory irregularity in hypothesis.
patients with PD may derive from abnormalities Smoking has been seen as a risk factor for the
in these brain centers. Finally, a wider and more first occurrence of PAs and the onset of PD [50,
general dysfunction across the homeostatic sys- 51]. Withal the biological mechanisms underly-
tems, including the respiratory system, the cardiac ing the link between PAs and smoking are largely
system, and the balance system may underlie such unknown. Our team investigated the effects of
respiratory irregularity. For instance the parabra- smoking on respiratory irregularity in patients
chial nucleus in the brainstem filters and orga- with PD. When compared with healthy controls,
nizes interoceptive stimuli from our basic both smoker and non-smoker patients exhibited
homeostatic functions, and maintains a represen- greater respiratory irregularity but smoker
tation of internal stability and bodily well-being patients showed higher irregularities than non-
[48]. Given that basic physiologic systems work smoker patients. On the contrary, smoking did
in concert with mutual modulation, we speculated not influence the regularity of respiratory pat-
that breathing irregularities may be related to per- terns in healthy subjects. Overall, smokers had
turbations of the cardiovascular system and the more severe PAs than nonsmokers [52]. It might
balance system and that the respiratory irregular- be that smoking impairs respiration in patients
ity in patients with PD might originate from a with PD, and influences the onset and/or mainte-
more general dysfunction of the brainstem cir- nance of the disorder. Abelson et al. [53] claimed
cuits that modulate homeostatic functions. Within that respiration irregularity in PD is influenced by
this framework, PAs may represent a primal emo- neither doxapram-induced hyperventilation
tion arising from those specific brain circuits that (doxapram is a respiratory stimulant) nor cogni-
process bodily sensations and perceptions that are tive manipulation but it appears to have intrinsic
related to homeostatic functions. Functional fail- and stable features. However replication studies
ure in these centers might account for the emer- with a greater number of participants necessitate
gence of PAs [25]. In conclusion homeostatic to yield more robust and convincing evidence and
dysfunctions (in particular respiratory dysfunc- to confirm behavioral and/or respiratory hyper-
tions) may be crucial in the pathophysiology of PD reactivity both at baseline and following CO2
and trigger neuroanatomical networks involved administration in patients with PD.
in PAs. Maddock [49] claimed that excessive Fourth, there has been a long scientific debate
response to lactate in the brain may underlie these on whether PD might be categorized in different
abnormal respiratory findings. Similarly, Esquivel subtypes according to specific clusters of symp-
et al. [45] maintained that much of the connection toms. Several investigators speculated about the
between panic and respiration might be explained existence of a respiratory and a non-respiratory
by altered acid-base levels in the brain: acute PD. Klein [5] hypothesized a connection between
brain acidosis may be linked to PD and PAs might the respiratory system and PD. Ley [54] also pos-
represent a defensive response to subtle potentially tulated the existence of a subcategory of PD with
threatening acid-base alteration. Administration dyspnea, heart palpitations, terror, and a strong
of panicogenic substances (e.g. lactate or CO2) desire to flee, as predominant symptoms.
would activate the centers that govern brain pH Similarly Briggs et al. [13] contemplated the
and evoke spontaneous PAs. It is plausible to existence of a subgroup of PD characterized by
assume that the degree of brain acidosis is rele- more respiratory symptoms (e.g. shortness of
vant to the panic symptoms induced by CO2 inha- breath, feelings of choking, etc.), higher occur-
lation and it may reflect an underlying metabolic rence of spontaneous PAs, and better response to
disturbance (see next paragraphs). Whether higher imipramine, whereas the non-respiratory group
respiratory irregularity is a consequence or rather seems to suffer more from situational PAs and to
4 Panic Disorder, Is It Really a Mental Disorder? From Body Functions to the Homeostatic Brain 99

respond better to alprazolam. More recently Song studies are encouraged to provide a definite
et al. [55] found that patients with the respiratory answer on whether it is appropriate to separate
subtype exhibited earlier onset of PD, more PD into distinct subcategories.
severe clinical symptoms, higher fear of respira- In conclusion, the respiratory system of indi-
tory symptoms, higher co-occurrence of agora- viduals suffering from PD appears to be more
phobia, and better response to SSRIs, when unstable and sensitive than subjects without
compared to patients with non-respiratory sub- PD. It is pivotal to verify to which extent the
type. Those with predominant respiratory symp- respiratory abnormalities arise from intrinsic
toms were also more sensitive to CO2 challenges, malfunctions of the respiratory system.
exhibited higher familial prevalence of PD and
earlier onset of the disorder, as well as more pre-
vious depressive episodes [56]. 4.3 Cardiovascular System
35 % CO2 inhalation induced dyspnea in and PD
patients with PD [22]. We found that patients
with PD who reacted to 35 % CO2 were charac- Heart palpitations or a racing heart, and chest
terized by greater baseline pattern of tidal volume pain or discomfort are among the core somatic
and inspiratory drive compared to patients who symptoms that characterize PAs. Subjects who
did not react to 35 % CO2 [34]. Meuret et al. [57] experience PAs are often concerned they may
identified three dimensions of panic: the cardio- suffer from a cardiac disease or that they are
respiratory (with palpitations, shortness of breath, dying of a heart attack and repeatedly refer emer-
choking, chest pain, numbness, fear of dying) the gency room. Of these individuals, more than
autonomic/somatic (with sweating, trembling, 20 % is not diagnosed with a cardiac problem but
nausea, chills/hot flashes, and dizziness) and the rather with PD [59]. Although routinely cardiac
cognitive (with derealisation, fear of going crazy examination yields negative result, the symptoms
or losing control). A systematic review by these subjects experience feel so real that reassur-
Niccolai et al. [24] indicated that patients with ance from the doctors is not enough to convince
PD exhibit an abnormal breathing pattern when them their cardiovascular system performs nor-
compared to controls, during rest/baseline, chal- mally. Most of these individuals believe there
lenge, and recovery conditions, and that respira- might be some undetected abnormality in their
tory variability may be a possible endophenotype cardiovascular system [60]. As a results they tend
of PD. Roberson-Nay et al. [58] compared the to worry about future attacks and develop antici-
respiratory and non-respiratory panic to investi- patory anxiety, which negatively impacts on their
gate whether these subtypes represent a single quality of life and daily functioning. Also, fre-
disorder, whether they differed on a quantitative quent special consultation results in increased
(i.e. severity) or rather qualitative (i.e. distinct health care consumption and costs [61]. A link
patterns) dimension from one another. Their between PD and cardiac disorders (CDs)
results suggested that the two panic subtypes had emerged. Some historical studies suggested an
distinct symptom profiles that differed in severity association between PD, arrhythmias, sudden
and that respiratory panic represented the most cardiac death or idiopathic cardiomyopathy [62
severe form of the disorder. In sum, sufficient sci- 64]. Several reports found a relationship between
entific evidence supports that patients with PD PD and coronary artery disease (CAD). PD prev-
with prominent respiratory symptoms constitute alence ranged from 11 % to 53 % in patients with
a discrete subgroup with more severe symptoms. documented CAD who visited emergency rooms
Finally, from a review by it emerged that hyper- (ERs) or outpatient cardiology clinics [6567],
sensitivity to CO2 might be a valid marker of the especially in those with atypical chest pain [65].
respiratory PD subtype [2]. We believe it is plau- Chance of suffering from CAD was 26 % in
sible to distinguish between respiratory and non- patients with PD who referred to ERs for chest
respiratory PD. Genetic and neuroimaging pain [68]. Conversely, other studies failed to find
100 G. Perna et al.

an association between PD and CAD in patients sequent CAD (AMI, unstable angina or angina
presenting with chest pain to ERs or cardiology pectoris) than subjects without PD (about 40,000
settings [6971]. Such discrepancy may arise subjects), even after adjusting for age at entry in
from methodological limitations of the studies, the cohort, smoking, obesity, use of angiotensin
such as low sample sizes, lack of standardized converting enzyme inhibitors, beta blockers,
tests for cardiac diagnoses and/or clinician- diuretics, and statins. The average time span
reported questionnaires to evaluate PD, and/or between diagnosis of PD and incident diagnosis
comorbid psychiatric disorders associated with of a CDs event was about 1.5 years. Both direct
cardiac risk. (e.g. physiological alterations) and indirect (e.g.
Although many studies did not find an associ- unhealthy behaviors) mechanisms may underlie
ation between PD and CAD [72, 73], others that the association between panic and cardiac risk.
used more robust methodological criteria (e.g. Many studies described imbalanced autonomic
inclusion of subjects with primary diagnosis of regulation and reduced heart rate variability
PD performed by a clinical interview and/or (lower parasympathetic activity and higher sym-
structured clinician-administered interview, diag- pathetic/parasympathetic ratio) in patients with
nosis of CDs obtained by medical examination, PD. Defective neuronal noradrenaline reuptake
sensitive standardized tests, and standardized cri- in the heart may contribute to adverse cardiac
teria) support a cross-sectional association events in subjects with PD by augmenting the
between current PD and CAD (prevalence across sympathetic cardiac firing [80]. Important asso-
studies ranged from 4.721 %) [7476]. This ciations between PD and increased arterial stiff-
suggests that past history of PD may also be rel- ness (that predicts cardiovascular mortality),
evant for the occurrence of CAD, even in the poor cardiovascular fitness and several factors
absence of current panic symptoms at the time of negatively affecting the endothelial function, and
CAD diagnosis. Thus, lack of investigating life- the development of atherosclerosis, (such as
time PD in some studies may have hampered the increased homocysteine levels, platelet aggrega-
identification of a relationship between PD and tion, lower levels of nitric oxide, lipid pattern
CDs. In addition, in older subjects with cardio- abnormalities, and higher inflammatory indexes)
vascular diseases (CVDs) higher rates of sub- emerged [8183]. In subjects with PD, inhalation
threshold panic-phobic symptoms emerged and of 35 % CO265 % O2 gas mixture (which is
this suggests a possible association between known to induce PAs), provoked transient myo-
CVDs and panic even in the absence of a full- cardial ischemia at least in high risk CAD
blown PD [77]. Finally, a nationwide population- patients [84]. Niccolai et al. [85] found that heart
based study by examined prospectively the rate in response to the 35 % CO2 challenge was
relationship between PD and acute myocardial higher in subjects with PD than controls.
infarction risk within 1 year of follow-up and Moreover patients needed more time to recover,
found that almost 5 % of patients with PD experi- and showed increased respiratory parameters
enced an acute myocardial infarction episode variability. Taken together these findings suggest
within a year, compared with less than 3 % in the that PAs are disturbing events for the cardio-
comparison cohort [78]. The association per- respiratory system that may contribute to increase
sisted also when controlling for hypertension, cardiac risk in this population over time. As we
coronary heart diseases, and age and the authors discussed in the previous chapter, subjects with
concluded that PD may represent an independent PD have also irregular respiratory patterns [33,
risk factor for developing acute myocardial 39] and baseline hyperventilation [14]. Since res-
infarction. These results confirm previous find- piration influences the autonomic regulation of
ings of a large cohort study based on the US cardiac activity [86] and coronary vasospasm can
National Managed Cara Database [79]. They be precipitated by hyperventilation [87, 88], such
showed that patients with PD (about 40,000 sub- peculiar respiratory features may increase vul-
jects) had a nearly twofold increased risk for sub- nerability to CAD.
4 Panic Disorder, Is It Really a Mental Disorder? From Body Functions to the Homeostatic Brain 101

The QT interval indicates the time the ven- monary exercise performance when compared to
tricular myocardium needs to depolarize and healthy controls. We investigated cardiorespira-
repolarize. A lengthened QT interval is a marker tory fitness levels in subjects with PD and in a
for ventricular tachyarrhythmia and a risk factor group of age-, gender-, weight-, and physical
for sudden death. QT variability is elevated in activity levels-matched healthy controls. We also
patients with PD than in healthy controls [89]. investigated whether psychological variables (i.e.
QT interval variability is closely related to state and trait anxiety, fear of physical sensations,
HRV. Decreased heart rate and heart period vari- and fear of autonomic arousal) influenced cardio-
ability in patients with PD compared to normal respiratory responses and perceived exertion of
control subjects, suggests decreased cardiac patients PD during a submaximal exercise test.
vagal function [90, 91]. Increased QT variability We found that although patients had poorer car-
in combination with decreased HRV may signifi- diorespiratory fitness and spent more effort dur-
cantly increase the risk for cardiovascular mor- ing physical exercise, this was not related to the
bidity and sudden death [92, 93]. Both HRV and psychological variables examined but it might be
QT interval can be influenced by sympathetic related to a diminished ability of the cardiac sys-
mechanisms. Considering that PAs are associated tem to efficiently respond to physical efforts
with several autonomic symptoms, including [104]. This might contribute to a sedentary life-
chest pain, heart pounding, tachycardia, and style and to increased cardiovascular risk in the
shortness of breath, such enhanced autonomic long run [103, 105].
activity can result in a significant increase in QT Taking into account the available data and
variability [94]. Increased QT in patients with PD given the involvement of the cardio-respiratory
may be due to increased sympathetic activity system in the pathophysiology of PD, a more in-
which may put them at a greater risk for signifi- depth investigation of the association between
cant cardiovascular events and sudden death. panic and CDs is strongly recommended.
Finally, some studies performing surface electro- Therefore research, preferably longitudinal, is
cardiogram in subjects with PD found that the encouraged to confirm the association between
QT interval augmented even more during hyper- both full-blown and subthreshold panic and
ventilation challenges [95]. Patients with PD also CAD, to investigate the association between
manifest increased dispersion of the QT and panic and other CDs, such as arrhythmias, and
P-wave [96, 97], which indicates higher regional hypertension and to investigate whether specific
heterogeneity of ventricular repolarization and subgroups of subjects with PD (e.g. those with
atrial depolarization, respectively, and are con- biological subclinical risk factors and/or higher
sidered indicators of arrhythmia and sudden cardio-respiratory instability and symptoms)
death risk [98100]. might be at higher risk of cardiovascular disease.
The physiologic alterations found in PD Unfortunately PD is often under-recognized in
patients are consistent with the idea that abnor- cardiological care settings and diagnosing PD
mal regulation of the body homeostatic functions may result in failure to recognize cardiac diseases
may be involved in the pathophysiology of PD [70]. Hence a better understanding of this rela-
[25, 43] and confer a peculiar vulnerability to tionship between may contribute to improve
medical diseases, including CDs [25, 101]. treatment and prevention of both PD and CDs.
Lastly, behavioral risk factors, such as smoking
[102, 103] (which is known to be associated with
increased cardiovascular risk) and physical exer- 4.4 Panic Disorder
cise avoidance [103] may contribute to cardiac and the Balance System
morbidity in PD. Cigarette smoking prevalence is
high in patients with PD when compared with Prevalence of vestibular symptoms, such as ver-
both healthy controls and subjects with other tigo, instability, and lightheadedness is higher in
anxiety disorders [50]. Patients with PD exhibit individuals suffering from PD, when compared
higher exercise avoidance and worse cardiopul- to control populations [106]. Jacob [107] noticed
102 G. Perna et al.

that 75 % of patients with PD manifest postural the sole cause of the disorder. The second one is
instability. Up to one third of patients with PD an otogenic pattern, in which a neurotologic con-
and agoraphobia with chronic dizziness suffer dition triggers the development of anxiety [112].
from peripheral vestibular alterations [108]. The mismatch theory of Furman and Jacob [113]
Stambolieva et al. assessed postural instability by also suggested that vestibular dysfunctions may
static posturographic tests of standing on stable provoke space and motion discomfort and elicit
and foam surfaces with open and closed eyes in anxiety in patients with PD. The third one is an
30 patients with PD and in 30 sex- and age- interactive pattern in which a neurotologic condi-
matched healthy controls. They noticed that tion is responsible for the onset of dizziness but
almost 84 % of the patients experienced dizziness also exacerbates preexisting or prodromal anxi-
and imbalance, both during and between PAs. No ety or panic symptoms. Therefore the connection
differences of sway velocity (an indicator of pos- between vestibular manifestations and anxiety
tural stability) emerged between the groups while disorders seems to be bidirectional: on the one
standing on both surfaces with open eyes. hand vestibular disorders can trigger anxiety and
However the sway velocity of the patients with on the other hand anxiety symptoms can trigger
PD was higher when compared to controls while vestibular symptoms. This implies that at least
standing with closed eyes both on the stable and some anomalies of the balance system can be
the foam surfaces. The authors concluded that ascribed to psychological factors rather than to
visual information plays a more important role in vestibular dysfunctions.
maintaining postural stability when sensory con- Neuroanatomical/neurophysiological circuit-
flict exists [109]. ries that might account for the relationships
Finally, peripheral vestibular disorders seem between the vestibular system and PD encom-
to be more prevalent in patients with PD and ago- pass: (1) connections between the locus coeru-
raphobia than in patients with PD alone [110]. In leus and the lateral vestibular nucleus [114], (2)
a double-blind, random, cross-over study Perna vestibular inputs to the raphae nuclei [115], (3)
et al. [111] compared a group of patients with serotonergic influence on the vestibular system
PD, with and without agoraphobia, and a group [116]. Finally vestibularrespiratory connections
of sex- and age-matched healthy controls who have been proposed (4). Vestibular nuclei project
underwent static posturography in three condi- to and receive from the caudal parabrachial
tions (eyes open, eyes closed and neck extension) nucleus and the subparabrachial nucleus (i.e. the
and the 35 % CO2 challenge. Symptomatological Klliker-Fuse) [117, 118]. A review of anatomic
reactivity to CO2 correlated with balance system and physiologic studies demonstrated direct con-
dysfunction in patients only in the eyes-closed nections between the vestibular nuclei and the
condition. Up to 42 % of patients with PD (com- brainstem regions that influence both sympa-
pared to 05 % in controls) exhibited aberrant thetic and parasympathetic activity [119]. A
balance system functioning, which correlated combination of autonomic, vestibular, and limbic
with agoraphobic avoidance. The authors con- information both in the brainstem and forebrain
cluded that the balance system seems to influence areas seems to regulate balance control. In par-
the psychobiological mechanisms underlying ticular, the parabrachial nucleus (PBN), which is
agoraphobic avoidance and therefore plays a role located in the pons, influences cardiovascular,
in the behavioural features of PD. respiratory, and autonomic responses and trans-
Staab et al. suggested three patterns of illness mits interoceptive information, as suggested by
in patients with clinical syndrome of what they chemical stimulation and lesion studies [120
labeled psychogenic dizziness, which is charac- 122]. Respiration and the balance system are
terized by vague and elusive physical symptoms intertwined and such connections may partly
of vertigo and lightheadedness, in the absence of explain the association between hyperventilation
objective clinical tests abnormalities. The first and postural instability in patients with PD (who
one is a psychogenic pattern, in which anxiety is are in fact in a chronic state of hyperventilation),
4 Panic Disorder, Is It Really a Mental Disorder? From Body Functions to the Homeostatic Brain 103

in patients with vestibular diseases, and in healthy for changing conditions and to begin fast
controls. These neuroanatomical connections responses. It is generally considered to be
indicate that respiration and the balance system involved in motion processing, posture, move-
are intertwined. In subjects with dizziness, hyper- ment and balance and in orienting and defensive
ventilation may induce nystagmus and reveal reactions to visual motion, involving short latency
vestibular dysfunction [123]. Accordingly, we postural adjustments as well as head and eye
believe chronic hyperventilation contributes to movements. Finally peripheral vision also recog-
postural instability and might aggravate dizziness nizes harmful threats coming up alongside [130].
in subjects with PD. In patients with PD balance control seems to
The PBN mediates both awareness of, and rely mainly on non-vestibular cues, such as pro-
affective and emotional responses to intrinsic and prioceptive and especially visual cues. In these
extrinsic stimuli that alter the sense of physiolog- individuals vision is perhaps the sensory systems
ical well-being (i.e. interoception) [124127] most strongly associated with postural balance in
and, in concert with the central amygdala, it also particular when sensory conflict exists or when
plays a role in recognizing innate danger stim- the other sensory systems are compromised or
uli [128]. Because interconnections between the damaged (e.g. while standing on an unstable sur-
PBN, the limbic system and the prefrontal cortex face, such as a foam platform, in broad spaces,
seem to be involved in the development and heights, crowded places, or in eyes closed condi-
expression of PD, it has been proposed that these tions). Impaired visual inputs enhance sensory
structures may also be a substrate for the co- conflict and may lead to adulterated propriocep-
occurrence of balance disorders and PD [48]. tive information and balance impairment which
In conclusion, patients with PD seem to pres- in turn may elicit panic symptoms. There is evi-
ent subclinical abnormalities in their balance sys- dence that in patients with PD, anxiety and dis-
tem. According to Balaban [129] such vestibular comfort may arise when visual information is
instability may be linked to changes in homeo- inaccurate as a result of hyperexcitability of the
static autonomic responses which in turn would locus coeruleus and the vestibular brain nuclei,
trigger affective and emotional responses, includ- which would result in higher postural instability
ing panic. [109, 111]. Our group found that patients with
PD and agoraphobia manifest postural instability
during peripheral visual stimulation whereas
4.4.1 Balance and the Visual controls did not. Conversely, the two groups
System in PD showed similar patterns of postural instability
during central visual stimulation [104]. Hence we
Maintaining proper balance and posture partly concluded that patients with PD, especially those
depends on visual information. Scientific litera- with comorbid agoraphobia, might be hypersen-
ture identified that the visual system has two sitive to the influence of peripheral visual system
visual pathways referred to as central (or focal/ on balance. Such higher sensitivity is perhaps
parvocellular) and peripheral (or ambient/mag- linked to a more active visual alarm system that
nocellular). Visual information from central and scans the environment for possible threats.
peripheral visual fields have complementary Connections between visual, vestibular and lim-
roles and might differently affects postural con- bic areas may increase postural sway when the
trol. The central vision enables object identifica- visual environment is changing in an uncertain
tion and recognition, it works largely consciously way, such as during motion in the peripheral
and in isolation and enables direct visual impact visual field [131]. Indeed, the relationship
on objects that appear straight in front of the between state anxiety and postural instability
visual field. The peripheral vision underlies the during peripheral stimulation supports the idea
perception of self-motion and body stance in the that these situations are emotionally relevant for
environment, and the ability to scan surroundings patients with PD. Such hypersensitivity might be
104 G. Perna et al.

specific of PD and arise from multiple sources. It ties in patients suffering from PD is possible.
might be an idiosyncratic perceptual habit that, in Jacob et al. [134] claimed that 812 weeks of
concert with PAs, might lead to consequent ago- vestibular rehabilitation alone was sufficient to
raphobia. Alternatively, it might follow PD and reduce anxiety and avoidance in patients suffer-
agoraphobia, which would act as disrupting ing from PD and from vestibular abnormalities
factors on balance control systems by vestibular and that 4 weeks of cognitive-behavioral therapy
brainstem-limbic connections. Also, panic- produced little benefit prior to vestibular rehabili-
phobic conditions might involve activation of tation. Teggi et al. [110] found that early vestibu-
complex alarm systems including interoceptive lar rehabilitation had even greater beneficial
conditioning processes linked to destabilizing effects on anxiety than on the balance function.
visual stimuli and operant learning processes Pharmacotherapy is another viable treatment
related to the avoidance of visual experiences and option. Mezzasalma et al. [108] evaluated the
provoking discomfort in everyday life [8]. efficacy and effectiveness of imipramine on the
In conclusion, even though patients with PD treatment of comorbid chronic dizziness and PD
report normal baseline sway, they seem unable to in nine patients with a diagnosis of PD with ago-
effectively counterbalance somatosensory and/or raphobia. The authors found peripheral vestibu-
vestibular information during irrelevant visual lar alteration in about one third of patients with
stimuli and increase postural sway more than PD and agoraphobia with chronic dizziness.
control subjects when exposed to visual sensory After a 3-month treatment with imipramine,
conflict. Such increase is unlikely to result from patients exhibited a significant decrease in anxi-
general destabilization, given that they tend to ety, dizziness, quality of life, and PD severity.
sway synchronously with the optic stimulus Preliminary results indicate reduced anxious
[132]. Patients may sway excessively also when symptoms and impairment due to dizziness after
standing on a fixed platform with a static visual 12 weeks of fluoxetine treatment in patients with
scene, and this suggests generalized balance vestibular dysfunction (e.g. dizziness that
abnormality rather than a problem limited to sen- included vertigo, motion sickness, nausea, and
sory integration in situations involving sensory anxiety) without anxiety disorders [135]. Our
conflict. Alternatively, it might be that these indi- team assessed posturography in a small group of
viduals are less capable of ignoring the misleading patients with PD and agoraphobia who were par-
visual information (i.e., they depend more on tial responders to pharmacotherapy and
visual cues). This increased reliance on vision cognitive-behavioral therapy, before and after 10
may be reflected in complaints of disequilibrium weeks of vestibular rehabilitation treatment con-
in complex moving visual environments. Whether sisting of stimulation of the peripheral visual
such visual dependence hinges on trait- or state- field in concomitance with a series of head/body
like enhanced vigilance needs to be determined. movement patients had to perform. Following
Many complications emerge when examining the vestibular rehabilitation patients exhibited
patients with vestibular dysfunctions and comor- improved balance performance as well as dimin-
bid psychiatric disorders [133]. Multidisciplinary ished agoraphobic symptoms (unpublished
evaluation of these patients is warranted, unfortu- study). In light of this, it is central that both
nately psychiatrists usually do not refer patients patients and clinicians acknowledge that PD and
with PD for otoneurologic evaluations as they vestibular anomalies can occur separately or in
refer them to cardiologic evaluations for symp- comorbidity. Clinicians should be advised to
toms like palpitations. On the other hand, it is conduct exhaustive medical investigation in
also rare that vestibular specialists refer patients order to convey the most accurate diagnosis to
to a psychiatrist to exclude the diagnosis of PD, their patients, to inform them about the causes of
which may be present in addition to the organic their symptoms, and about the treatment options,
vestibular disease. Exploratory research indicates which encompasses specific rehabilitation proto-
that adequate treatment of vestibular abnormali- cols addressed to re-establish proper functioning
4 Panic Disorder, Is It Really a Mental Disorder? From Body Functions to the Homeostatic Brain 105

of the vestibular system and decrease anxiety- activated or deactivated by light. Finally, whether
and dizziness-related symptomatology, as well photosensitivity is a state characteristic second-
as SSRIs, TCAs, and CBT. ary to the active disease or a trait predisposing to
the full-blown disorder is currently debated
although research tends to favor the former
4.5 Photosensitivity and PD hypothesis on the basis that photophobia in PD
renormalized after cognitive behavior therapy
Photosensitivity (i.e. an abnormally high sensi- treatment [141].
tivity to light exposure) seems to contribute both
to the etiopathogenesis of and response to ther-
apy in PD [136]. Indeed patients with PD have a 4.6 Conclusions and Clinical
lowered threshold of tolerance to light and Implications
develop the tendency to adopt photophobic
behavior when compared to healthy controls (e.g. Patients with PD often complain of somatic
to protect themselves from light by wearing sun- symptoms such as abnormalities in the cardiore-
glasses and/or by avoiding to go out during day- spiratory and vestibular systems. They exhibit
time) [137]. Indeed they score significantly poorer physical fitness [103, 104], higher respira-
higher on photophobia (light avoidance) and sig- tory variability during mild physical activity,
nificantly lower on photophilia (light pursuit) higher respiratory dysfunctions/breathing pat-
symptoms, as measured by the Photosensitivity terns irregularities, when compared with subjects
Assessment Questionnaire (PAQ). Interestingly, without PD [14, 24, 33]. Even unexpected PAs
photosensitivity seems not to be linked to the have been associated with significant autonomic
mere presence of a diagnosed PD, in fact it cor- and respiratory instability that largely preceded
related with the panic-agoraphobic spectrum, panic onset [20]. Patients often manifest a less
regardless of diagnosis, within both clinical and efficient cardiovascular system [101] and higher
healthy populations. These results suggest that cardiovascular risk [142]. Finally subclinical
photosensitivity may belong to the core of PD, abnormalities in the balance system, as well as
regardless of the presence of current active higher photosensitivity, are also common [33,
symptomatology. More precisely, photophobia 111, 138].
may be integrated in the panic-agoraphobic spec- Taken together these findings partly contradict
trum, as it seems to run in parallel to other panic the assumption that PAs are just false alarms. PAs
spectrum dimensions [138]. Photophobia has may be real alarms and reflect reduced adaptabil-
been associated with agoraphobia as well. Light ity to changes, and true homeostatic instability,
frequently elicits anxious/panic symptoms also in which may sustain the experience of anticipatory
these patients [139]. anxiety and phobic avoidance and increase vul-
Light sensitivity may be correlated with sub- nerability to panic. For these reasons we feel con-
clinical autonomic system dysfunctions in fident to state that patients with PD are
patients with PD. Research indicates anatomical physiologically different from healthy subjects
links between the amygdala, which exerts a key and that cardiac, respiratory, and balance symp-
role in anxiety, and the Edinger-Westphal toms may be the outcome of the inability of these
nucleus, a midbrain center that controls pupil systems to relate with the environment.
movement, lens accommodation, and eyes con- In addition mounting evidence suggests that
vergence [140]. The specific associations antipanic medications are not merely psychotro-
between light exposure and the neurotransmitters pic medications. For instance, the selective sero-
involved in PD are largely uninvestigated. For tonin re-uptake inhibitors (SSRIs), which are
instance, it might be interesting to explore to considered the first choice drug treatment of PD,
which extent serotonin or melatonin (both of also act on the respiratory, cardiovascular and
which have been involved in PD) synthesis is balance systems. Asymptomatic patients with PD
106 G. Perna et al.

(with and without agoraphobia) with no pulmo- We believe SSRIs might exert their anti-panic
nary diseases exhibited improved lung function effect also by reducing homeostatic dysfunctions
following administration of antipanic drugs (i.e. in patients with PD. Besides cognitive-behavior
paroxetine, imipramine, and clonazepam) when psychotherapy, other non-pharmacologic treat-
compared to those in the washout period [143]. ments, such as breathing therapies, and physical
These results indicate that anti-panic drugs ame- exercise, can help to normalize homeostatic dys-
liorate pulmonary function in this population. functions in PD. Indeed breathing therapies [147]
In line with these findings, Perna et al. [31] and aerobic physical exercise reduce panic symp-
found that patients with PD manifested abnormal toms, therefore they might represent valid adjunc-
values for many dynamic lung volumes (i.e. peak tive treatment options for PD [148]. In addition,
expiratory flow rate, expiratory flow at 75 % of preliminary data suggest that vestibular rehabili-
vital capacity, and maximum mid-expiratory flow tation might also be beneficial for patients with
rate) when compared to a group of healthy con- PD (especially in those with comorbid agorapho-
trols. The authors believe that such functional bia). Although we cannot exclude that these
abnormalities indicate subclinical obstruction of somatic treatments act by increasing the per-
lung airways, which are perhaps relevant to the ceived sense of control in patients with PD, they
mechanisms related to PD. might improve panic symptoms also via their
In rats a significant increase in baseline respi- positive effect on the homeostatic body func-
ratory rate was found after 5 and 15 weeks of tions. A recent meta-analysis indicated that com-
treatment with paroxetine. In particular following bining exposure, relaxation training, and
15 weeks of treatment the rats exhibited reduced breathing retraining is more efficacious than cog-
respiratory rate in response to CO2 exposure. nitive therapy alone, suggesting that somatic
These results indicate that the regulation of respi- interventions are more effective than the mental
ration may be an important factor for the parox- ones [149]. Taken together, these findings rein-
etine antipanic effect [144]. Lungs are the main force the assumption that PAs may represent real
reservoir of the serotonin transporter, which is the alarms. Homeostatic dysfunctions may represent
main target of the SSRIs, and this may explain the a candidate endophenotype of panic vulnerability
favourable effect of the SSRIs on respiratory and underlie maintenance of defensive active
irregularities as well as in pulmonary arterial mechanisms such as anticipatory anxiety and
hypertension. Paroxetine treatment at 20 mg/day phobic avoidance. Many patients complain of
for 4 weeks increased heart rate variability and somatic and cognitive symptoms even when not
total parasympathetic activity and decreased total experiencing a PA. Therefore, treatment should
sympathetic activity in patients with PD [145]. focus not only on reducing PAs occurrence but
These results indicate that paroxetine (and argu- also aim at reaching a thorough state of physical
ably SSRIs in general) may protect against or well being. Anti-panic drugs posology might be
even decrease cardiovascular morbidity and mor- adjusted until a full sensation of physical wellbe-
tality in patients with PD. Finally, we investigated ing is reached. Physical exercise could also be
the effects of a 6-week treatment with citalopram recommended as an additional intervention to
on the balance system function in 15 patients with potentiate cardiovascular fitness. We believe
PD, with or without agoraphobia, who underwent diagnosis and treatment of patients with PD must
static posturography on days 0 and 42. After 6 be approached with a multidisciplinary evalua-
weeks of treatment with citalopram there was a tion, and therefore all treatment options must be
significant decrease of four out of six posturogra- considered. Since cardiorespiratory and vestibu-
phy measures in eyes-closed and neck extension lar symptoms can aggravate psychiatric symp-
conditions [146]. This study suggests that seroto- toms and psychiatric disorders can complicate
nergic modulation can improve the balance sys- even further the evaluation of patients with
tem function in patients with PD, especially when somatic complaints, every therapeutic option
visual information is lacking. available must be carefully considered. Hopefully
4 Panic Disorder, Is It Really a Mental Disorder? From Body Functions to the Homeostatic Brain 107

in the future PD treatment will involve interven- Psychiatr Res. 2012;21(3):16984. doi:10.1002/
tions that address not only the core symptoms but
4. Goodwin RD, Faravelli C, Rosi S, Cosci F, Truglia
also focus on harmonizing respiration and bal- E, de Graaf R, et al. The epidemiology of panic dis-
ance, and increase heart rate variability. Regaining order and agoraphobia in Europe. Eur
physical wellbeing will help patients overcoming Neuropsychopharmacol. 2005;15(4):43543.
anticipatory anxiety and agoraphobia more rap-
5. Klein DF. False suffocation alarms, spontaneous
idly and effectively. In conclusion, the assump- panics, and related conditions. An integrative
tion that PD is merely a mental disorder seems hypothesis. Arch Gen Psychiatry. 1993;50(4):
unsatisfactory. We believe PD should be inte- 30617.
6. Gorman JM, Liebowitz MR, Fyer AJ, Stein J. A neu-
grated in a broader vision that acknowledges a
roanatomical hypothesis for panic disorder. Am
central role to the homeostatic systems in the J Psychiatry. 1989;146(2):14861.
etiopathogenesis of the disorder. Actually the 7. Gorman JM, Kent JM, Sullivan GM, Coplan
somatic complains of our patients might reflect JD. Neuroanatomical hypothesis of panic disorder,
revised. Am J Psychiatry. 2000;157(4):493505.
real bodily dysfunctions. In the future we are
8. Bouton ME, Mineka S, Barlow DH. A modern learn-
convinced that patients with PD may benefit from ing theory perspective on the etiology of panic disor-
multidisciplinary evaluation, therefore it is cru- der. Psychol Rev. 2001;108(1):432.
cial that mental health professionals, vestibular 9. Clark DM, Salkovskis PM, Ost LG, Breitholtz E,
Koehler KA, Westling BE, et al. Misinterpretation of
specialists, cardiac specialists, and lung special-
body sensations in panic disorder. J Consult Clin
ists mutually exchange information in order to Psychol. 1997;65(2):20313.
assess the contingent co-occurrence of PD and its 10. Bellodi L, Perna G. The panic respiration connec-
related somatic diseases. This will allow a better tion. Milan, Italy: Medical Media; 1998.
11. Davidson Jr, Nutt D. Anxiety and respiration. In:
diagnosis and lay the foundation for more ade-
Disorders eA, editor. Anxiety disorders. Oxford:
quate treatment options that may encompass Blackwell Science Ltd.; 2003.
pharmacological, psychological and somatic 12. Schruers K, Griez E. The effects of tianeptine or
treatments, and physical exercise, in a joint effort paroxetine on 35 % CO2 provoked panic in panic dis-
order. J Psychopharmacol. 2004;18(4):5538.
to decrease panic symptomatology on the one
hand and to re-establish proper functioning of the 13. Briggs AC, Stretch DD, Brandon S. Subtyping of
homeostatic systems, on the other hand. We trust panic disorder by symptom profile. Br J Psychiatry.
this approach will favor not only remission but 1993;163:2019.
14. Grassi M, Caldirola D, Vanni G, Guerriero G,
recovery (which encompasses both symptoms
Piccinni M, Valchera A, et al. Baseline respiratory
remission and more functional aspects of the parameters in panic disorder: a meta-analysis.
patients wellbeing, such as social functionality J Affect Disord. 2013;146(2):15873. doi:10.1016/j.
and quality of life), and avoid relapse in the jad.2012.08.034.
15. Grassi M, Caldirola D, Di Chiaro NV, Riva A, Dacco
S, Pompili M, et al. Are respiratory abnormalities
specific for panic disorder? A meta-analysis.
Neuropsychobiology. 2014;70(1):5260.
References doi:10.1159/000364830.
16. Cowley DS, Roy-Byrne PP. Hyperventilation and
1. Association AP. Diagnostic and statistical manual of panic disorder. Am J Med. 1987;83(5):92937.
mental disorders. 5th ed. Arlington, VA: American 17. van den Hout MA, Hoekstra R, Arntz A, Christiaanse
Psychiatric Publishing; 2013. M, Ranschaert W, Schouten E. Hyperventilation is
2. Amaral JM, Spadaro PT, Pereira VM, Silva AC, not diagnostically specific to panic patients.
Nardi AE. The carbon dioxide challenge test in panic Psychosom Med. 1992;54(2):18291.
disorder: a systematic review of preclinical and clini- 18. Griez E, Zandbergen J, Lousberg H, van den Hout
cal research. Rev Bras Psiquiatr. 2013;35(3):31831. M. Effects of low pulmonary CO2 on panic anxiety.
doi:10.1590/1516-4446-2012-1045. Compr Psychiatry. 1988;29(5):4907.
3. Kessler RC, Petukhova M, Sampson NA, Zaslavsky 19. Gorman JM, Papp LA, Martinez J, Goetz RR,
AM, Wittchen HU. Twelve-month and lifetime prev- Hollander E, Liebowitz MR, et al. High-dose carbon
alence and lifetime morbid risk of anxiety and mood dioxide challenge test in anxiety disorder patients.
disorders in the United States. Int J Methods Biol Psychiatry. 1990;28(9):74357.
108 G. Perna et al.

20. Meuret AE, Rosenfield D, Wilhelm FH, Zhou E, 35. Nardi AE. Panic disorder is closely associated with
Conrad A, Ritz T, et al. Do unexpected panic attacks respiratory obstructive illnesses. Am J Respir Crit
occur spontaneously? Biol Psychiatry. 2011;70(10): Care Med. 2009;179(3):2567. doi:10.1164/
98591. doi:10.1016/j.biopsych.2011.05.027. ajrccm.179.3.256.
21. Goodwin RD, Pine DS. Respiratory disease and 36. Homma I, Masaoka Y. Breathing rhythms and
panic attacks among adults in the United States. emotions. Exp Physiol. 2008;93(9):101121.
Chest. 2002;122(2):64550. doi:10.1113/expphysiol.2008.042424.
22. Perna G, Battaglia M, Garberi A, Arancio C, Bertani 37. Munjack DJ, Brown RA, Cabe DD, McDowell DE,
A, Bellodi L. Carbon dioxide/oxygen challenge test Baltazar PL. A naturalistic follow-up of panic
in panic disorder. Psychiatry Res. patients after short-term pharmacologic treatment.
1994;52(2):15971. J Clin Psychopharmacol. 1993;13(2):1568.
23. Spinhoven P, Ros M, Westgeest A, Van der Does 38. Wilhelm FH, Gerlach AL, Roth WT. Slow recovery
AJ. The prevalence of respiratory disorders in panic from voluntary hyperventilation in panic disorder.
disorder, major depressive disorder and V-code Psychosom Med. 2001;63(4):63849.
patients. Behav Res Ther. 1994;32(6):6479. 39. Wilhelm FH, Gevirtz R, Roth WT. Respiratory dys-
24. Niccolai V, van Duinen MA, Griez EJ. Respiratory regulation in anxiety, functional cardiac, and pain
patterns in panic disorder reviewed: a focus on biologi- disorders. Assessment, phenomenology, and treat-
cal challenge tests. Acta Psychiatr Scand. 2009;120(3): ment. Behav Modif. 2001;25(4):51345.
16777. doi:10.1111/j.1600-0447.2009.01408.x. 40. Putnam RW, Filosa JA, Ritucci NA. Cellular
25. Perna G, Caldirola D, Bellodi L. Panic disorder: mechanisms involved in CO(2) and acid signaling in
from respiration to the homeostatic brain. Acta chemosensitive neurons. Am J Physiol Cell
Neuropsychiatr. 2004;16:5767. Physiol. 2004;287(6):C1493526. doi:10.1152/
26. Preter M, Klein DF. Panic, suffocation false alarms, ajpcell.00282.2004.
separation anxiety and endogenous opioids. Prog 41. Bailey JE, Argyropoulos SV, Lightman SL, Nutt
Neuropsychopharmacol Biol Psychiatry. 2008;32(3): DJ. Does the brain noradrenaline network mediate
60312. doi:10.1016/j.pnpbp.2007.07.029. the effects of the CO2 challenge? J Psychopharmacol.
27. Feinstein JS, Buzza C, Hurlemann R, Follmer RL, 2003;17(3):2529.
Dahdaleh NS, Coryell WH, et al. Fear and panic in 42. Protopopescu X, Pan H, Tuescher O, Cloitre M,
humans with bilateral amygdala damage. Nat Goldstein M, Engelien A, et al. Increased brainstem
Neurosci. 2013;16(3):2702. doi:10.1038/nn.3323. volume in panic disorder: a voxel-based morphomet-
28. Preter M, Klein DF. Lifelong opioidergic vulnerabil- ric study. Neuroreport. 2006;17(4):3613.
ity through early life separation: A recent extension doi:10.1097/01.wnr.0000203354.80438.1.
of the false suffocation alarm theory of panic disor- 43. Perna G, Guerriero G, Brambilla P, Caldirola
der. Neurosci Biobehav Rev. 2014;46:34551. D. Panic and the brainstem: clues from neuroimag-
doi:10.1016/j.neubiorev.2014.03.025. ing studies. CNS Neurol Disord Drug Targets.
29. Faravelli C, Lo Sauro C, Godini L, Lelli L, Benni L, 2014;13(6):104956.
Pietrini F, et al. Childhood stressful events, HPA axis 44. Remmers JE. Central neuron control of breathing.
and anxiety disorders. World J Psychiatry. In: Altose MD, Kawakami Y, editors. Control of
2012;2(1):1325. doi:10.5498/wjp.v2.i1.13. breathing in health and disease (lung biology in
30. Pine DS, Klein RG, Coplan JD, Papp LA, Hoven health and disease). New York: Marcel Dekker;
CW, Martinez J, et al. Differential carbon dioxide 1999.
sensitivity in childhood anxiety disorders and nonill 45. Esquivel G, Schruers KR, Maddock RJ, Colasanti A,
comparison group. Arch Gen Psychiatry. 2000; Griez EJ. Acids in the brain: a factor in panic?
57(10):9607. J Psychopharmacol. 2010;24(5):63947.
31. Perna G, Cocchi S, Bertani A, Arancio C, Bellodi doi:10.1177/0269881109104847.
L. Sensitivity to 35 % CO2 in healthy first-degree 46. Horn EM, Waldrop TG. Suprapontine control of res-
relatives of patients with panic disorder. Am piration. Respir Physiol. 1998;114(3):20111.
J Psychiatry. 1995;152(4):6235. 47. Roncon CM, Biesdorf C, Santana RG, Zangrossi Jr
32. Coryell W, Pine D, Fyer A, Klein D. Anxiety H, Graeff FG, Audi EA. The panicolytic-like
responses to CO2 inhalation in subjects at high-risk effect of fluoxetine in the elevated T-maze is
for panic disorder. J Affect Disord. 2006;92(1):63 mediated by serotonin-induced activation of endog-
70. doi:10.1016/j.jad.2005.12.045. enous opioids in the dorsal periaqueductal grey.
33. Caldirola D, Bellodi L, Caumo A, Migliarese G, J Psychopharmacol. 2012;26(4):52531.
Perna G. Approximate entropy of respiratory pat- doi:10.1177/0269881111434619.
terns in panic disorder. Am J Psychiatry. 48. Balaban CD, Thayer JF. Neurological bases for
2004;161(1):7987. balance-anxiety links. J Anxiety Disord.
34. Perna G, Bertani A, Caldirola D, Gabriele A, Cocchi 2001;15(12):5379.
S, Bellodi L. Antipanic drug modulation of 35 % 49. Maddock RJ. The lactic acid response to alkalosis in
CO2 hyperreactivity and short-term treatment out- panic disorder: an integrative review.
come. J Clin Psychopharmacol. 2002;22(3):3008. J Neuropsychiatry Clin Neurosci. 2001;13(1):2234.
4 Panic Disorder, Is It Really a Mental Disorder? From Body Functions to the Homeostatic Brain 109

50. Knuts IJ, Cosci F, Esquivel G, Goossens L, van 65. Fleet R, Lavoie K, Beitman BD. Is panic disorder
Duinen M, Bareman M, et al. Cigarette smoking and associated with coronary artery disease? A critical
35 % CO(2) induced panic in panic disorder patients. review of the literature. J Psychosom Res. 2000;
J Affect Disord. 2010;124(1-2):2158. doi:10.1016/j. 48(45):34756.
jad.2009.10.012. 66. Jeejeebhoy FM, Dorian P, Newman DM. Panic dis-
51. Breslau N, Klein DF. Smoking and panic attacks: an order and the heart: a cardiology perspective.
epidemiologic investigation. Arch Gen Psychiatry. J Psychosom Res. 2000;48(45):393403.
1999;56(12):11417. 67. Huffman JC, Pollack MH. Predicting panic disorder
52. Caldirola D, Bellodi L, Cammino S, Perna among patients with chest pain: an analysis of the
G. Smoking and respiratory irregularity in panic dis- literature. Psychosomatics. 2003;44(3):22236.
order. Biol Psychiatry. 2004;56(6):3938. doi:10.1176/appi.psy.44.3.222.
doi:10.1016/j.biopsych.2004.06.013. 68. Lynch P, Galbraith KM. Panic in the emergency
53. Abelson JL, Weg JG, Nesse RM, Curtis GC. room. Can J Psychiatry. 2003;48(6):3616.
Persistent respiratory irregularity in patients with 69. Katerndahl D. Panic plaques: panic disorder & coro-
panic disorder. Biol Psychiatry. 2001;49(7):58895. nary artery disease in patients with chest pain. J Am
54. Ley R. The many faces of Pan: psychological and Board Fam Pract. 2004;17(2):11426.
physiological differences among three types of panic 70. Katerndahl DA. The association between panic dis-
attacks. Behav Res Ther. 1992;30(4):34757. order and coronary artery disease among primary
55. Song HM, Kim JH, Heo JY, Yu BH. Clinical charac- care patients presenting with chest pain: an updated
teristics of the respiratory subtype in panic disorder literature review. Prim Care Companion J Clin
patients. Psychiatry Investig. 2014;11(4):4128. Psychiatry. 2008;10(4):27685.
doi:10.4306/pi.2014.11.4.412. 71. Katerndahl DA. Chest pain and its importance in
56. Nardi AE, Valenca AM, Mezzasalma MA, Lopes patients with panic disorder: an updated literature
FL, Nascimento I, Veras AB, et al. 35 % carbon review. Prim Care Companion J Clin Psychiatry.
dioxide and breath-holding challenge tests in panic 2008;10(5):37683.
disorder: a comparison with spontaneous panic 72. Bunevicius A, Staniute M, Brozaitiene J, Pop VJ,
attacks. Depress Anxiety. 2006;23(4):23644. Neverauskas J, Bunevicius R. Screening for anxiety
doi:10.1002/da.20165. disorders in patients with coronary artery disease.
57. Meuret AE, White KS, Ritz T, Roth WT, Hofmann Health Qual Life Outcomes. 2013;11:37.
SG, Brown TA. Panic attack symptom dimensions doi:10.1186/1477-7525-11-37.
and their relationship to illness characteristics in 73. Frasure-Smith N, Lesperance F. Depression and
panic disorder. J Psychiatr Res. 2006;40(6):5207. anxiety as predictors of 2-year cardiac events in
doi:10.1016/j.jpsychires.2005.09.006. patients with stable coronary artery disease. Arch
58. Roberson-Nay R, Latendresse SJ, Kendler KS. A Gen Psychiatry. 2008;65(1):6271. doi:10.1001/
latent class approach to the external validation of archgenpsychiatry.2007.4.
respiratory and non-respiratory panic subtypes. 74. Parker GB, Owen CA, Brotchie HL, Hyett MP. The
Psychol Med. 2012;42(3):46174. doi:10.1017/ impact of differing anxiety disorders on outcome
S0033291711001425. following an acute coronary syndrome: time to start
59. Soares-Filho GL, Machado S, Arias-Carrion O, worrying? Depress Anxiety. 2010;27(3):3029.
Santulli G, Mesquita CT, Cosci F, et al. Myocardial doi:10.1002/da.20602.
perfusion imaging study of CO(2)-induced panic 75. Todaro JF, Shen BJ, Raffa SD, Tilkemeier PL,
attack. Am J Cardiol. 2014;113(2):3848. Niaura R. Prevalence of anxiety disorders in men
doi:10.1016/j.amjcard.2013.09.035. and women with established coronary heart disease.
60. Starcevic V, Berle D. Cognitive specificity of anxiety J Cardiopulm Rehabil Prev. 2007;27(2):8691.
disorders: a review of selected key constructs. Depress doi:10.1097/01.HCR.0000265036.24157.e7.
Anxiety. 2006;23(2):5161. doi:10.1002/da.20145. 76. Dammen T, Arnesen H, Ekeberg O, Friis
61. Batelaan N, De Graaf R, Van Balkom A, Vollebergh S. Psychological factors, pain attribution and medi-
W, Beekman A. Thresholds for health and thresholds cal morbidity in chest-pain patients with and without
for illness: panic disorder versus subthreshold panic coronary artery disease. Gen Hosp Psychiatry.
disorder. Psychol Med. 2007;37(2):24756. 2004;26(6):4639. doi:10.1016/j.genhosppsych.
doi:10.1017/S0033291706009007. 2004.08.004.
62. Katon WJ. Chest pain, cardiac disease, and panic 77. Grenier S, Potvin O, Hudon C, Boyer R, Preville M,
disorder. J Clin Psychiatry. 1990;51(Suppl):2730. Desjardins L, et al. Twelve-month prevalence and
discussion 503. correlates of subthreshold and threshold anxiety in
63. Coryell W, Noyes Jr R, House JD. Mortality among community-dwelling older adults with cardiovascu-
outpatients with anxiety disorders. Am J Psychiatry. lar diseases. J Affect Disord. 2012;136(3):72432.
1986;143(4):50810. doi:10.1016/j.jad.2011.09.052.
64. Chignon JM, Lepine JP, Ades J. Panic disorder in 78. Chen YH, Tsai SY, Lee HC, Lin HC. Increased risk
cardiac outpatients. Am J Psychiatry. 1993;150(5): of acute myocardial infarction for patients with
7805. panic disorder: a nationwide population-based study.
110 G. Perna et al.

Psychosom Med. 2009;71(7):798804. doi:10.1097/ variability in patients with panic disorder: a study of
PSY.0b013e3181ad55e3. Holter ECG records. Psychiatry Res.
79. Gomez-Caminero A, Blumentals WA, Russo LJ, 1998;78(12):8999.
Brown RR, Castilla-Puentes R. Does panic disorder 91. Yeragani VK, Pohl R, Berger R, Balon R, Ramesh C,
increase the risk of coronary heart disease? A cohort Glitz D, et al. Decreased heart rate variability in
study of a national managed care database. panic disorder patients: a study of power-spectral
Psychosom Med. 2005;67(5):68891. analysis of heart rate. Psychiatry Res. 1993;
80. Alvarenga ME, Richards JC, Lambert G, Esler 46(1):89103.
MD. Psychophysiological mechanisms in panic dis- 92. Molgaard H, Sorensen KE, Bjerregaard P. Attenuated
order: a correlative analysis of noradrenaline spill- 24-h heart rate variability in apparently healthy sub-
over, neuronal noradrenaline reuptake, power jects, subsequently suffering sudden cardiac death.
spectral analysis of heart rate variability, and psy- Clin Auton Res. 1991;1(3):2337.
chological variables. Psychosom Med. 2006;68(1):8 93. Bigger Jr JT, Fleiss JL, Rolnitzky LM, Steinman
16. doi:10.1097/01.psy.0000195872.00987.db. RC. Frequency domain measures of heart period
81. Yapislar H, Aydogan S, Ozum U. Biological under- variability to assess risk late after myocardial infarc-
standing of the cardiovascular risk associated with tion. J Am Coll Cardiol. 1993;21(3):72936.
major depression and panic disorder is important. Int 94. Yeragani VK, Pohl R, Jampala VC, Balon R, Kay J,
J Psychiatry Clin Pract. 2012;16(1):2732. Igel G. Effect of posture and isoproterenol on beat-to-
doi:10.3109/13651501.2011.620127. beat heart rate and QT variability. Neuropsychobiology.
82. Jakovljevic M, Reiner Z, Milicic D. Mental disor- 2000;41(3):11323. doi:26642.
ders, treatment response, mortality and serum cho- 95. Sullivan GM, Kent JM, Kleber M, Martinez JM,
lesterol: a new holistic look at old data. Psychiatr Yeragani VK, Gorman JM. Effects of hyperventilation
Danub. 2007;19(4):27081. on heart rate and QT variability in panic disorder pre-
83. Hoge EA, Brandstetter K, Moshier S, Pollack MH, and post-treatment. Psychiatry Res. 2004;125(1):
Wong KK, Simon NM. Broad spectrum of cytokine 2939. doi:10.1016/j.psychres.2003.10.002.
abnormalities in panic disorder and posttraumatic 96. Atmaca M, Yavuzkir M, Izci F, Gurok MG,
stress disorder. Depress Anxiety. 2009;26(5):447 Adiyaman S. QT wave dispersion in patients with
55. doi:10.1002/da.20564. panic disorder. Neurosci Bull. 2012;28(3):24752.
84. Fleet R, Lesperance F, Arsenault A, Gregoire J, 97. Yavuzkir M, Atmaca M, Dagli N, Balin M, Karaca I,
Lavoie K, Laurin C, et al. Myocardial perfusion Mermi O, et al. P-wave dispersion in panic disorder.
study of panic attacks in patients with coronary Psychosom Med. 2007;69(4):3447. doi:10.1097/
artery disease. Am J Cardiol. 2005;96(8):10648. PSY.0b013e3180616900.
doi:10.1016/j.amjcard.2005.06.035. 98. Tomaselli GF, Beuckelmann DJ, Calkins HG, Berger
85. Niccolai V, van Duinen MA, Griez EJ. Objective and RD, Kessler PD, Lawrence JH, et al. Sudden cardiac
subjective measures in recovery from a 35 % carbon death in heart failure. The role of abnormal repolar-
dioxide challenge. Can J Psychiatry. 2008;53(11): ization. Circulation. 1994;90(5):25349.
73744. 99. Manttari M, Oikarinen L, Manninen V, Viitasalo
86. Guyenet PG, Stornetta RL, Abbott SB, Depuy SD, M. QT dispersion as a risk factor for sudden cardiac
Fortuna MG, Kanbar R. Central CO2 chemorecep- death and fatal myocardial infarction in a coronary
tion and integrated neural mechanisms of cardiovas- risk population. Heart. 1997;78(3):26872.
cular and respiratory control. J Appl Physiol. 100. Dilaveris PE, Gialafos EJ, Andrikopoulos GK,
2010;108(4):9951002. doi:10.1152/ Richter DJ, Papanikolaou V, Poralis K, et al. Clinical
japplphysiol.00712.2009. and electrocardiographic predictors of recurrent
87. Girotti LA, Crosatto JR, Messuti H, Kaski JC, atrial fibrillation. Pacing Clin Electrophysiol.
Dyszel E, Rivas CA, et al. The hyperventilation test 2000;23(3):3528.
as a method for developing successful therapy in 101. Fisher AJ, Woodward SH. Cardiac stability at differ-
Prinzmetals angina. Am J Cardiol. 1982;49(4): ing levels of temporal analysis in panic disorder,
83441. post-traumatic stress disorder, and healthy controls.
88. Fujii H, Yasue H, Okumura K, Matsuyama K, Psychophysiology. 2014;51(1):807. doi:10.1111/
Morikami Y, Miyagi H, et al. Hyperventilation- psyp.12148.
induced simultaneous multivessel coronary spasm in 102. Cosci F, Knuts IJ, Abrams K, Griez EJ, Schruers
patients with variant angina: an echocardiographic KR. Cigarette smoking and panic: a critical review
and arteriographic study. J Am Coll Cardiol. of the literature. J Clin Psychiatry. 2010;71(5):606
1988;12(5):118492. 15. doi:10.4088/JCP.08r04523blu.
89. Yeragani VK, Pohl R, Jampala VC, Balon R, Ramesh 103. Muotri RW, Bernik MA. Panic disorder and exercise
C, Srinivasan K. Increased QT variability in patients avoidance. Rev Bras Psiquiatr. 2014;36(1):6875.
with panic disorder and depression. Psychiatry Res. doi:10.1590/1516-4446-2012-1012.
2000;93(3):22535. 104. Caldirola D, Namia C, Micieli W, Carminati C,
90. Yeragani VK, Sobolewski E, Igel G, Johnson C, Bellodi L, Perna G. Cardiorespiratory response to
Jampala VC, Kay J, et al. Decreased heart-period physical exercise and psychological variables in
4 Panic Disorder, Is It Really a Mental Disorder? From Body Functions to the Homeostatic Brain 111

panic disorder. Rev Bras Psiquiatr. 2011;33(4): 120. Reilly S, Trifunovic R. Lateral parabrachial nucleus
3859. lesions in the rat: neophobia and conditioned taste
105. Sardinha A, Araujo CG, Soares-Filho GL, Nardi aversion. Brain Res Bull. 2001;55(3):35966.
AE. Anxiety, panic disorder and coronary artery dis- 121. Hayward LF, Felder RB. Lateral parabrachial
ease: issues concerning physical exercise and cogni- nucleus modulates baroreflex regulation of sympa-
tive behavioral therapy. Expert Rev Cardiovasc Ther. thetic nerve activity. Am J Physiol. 1998;274(5 Pt
2011;9(2):16575. doi:10.1586/erc.10.170. 2):R127482.
106. Furman JM, Redfern MS, Jacob RG. Vestibulo- 122. Chamberlin NL, Saper CB. Topographic organiza-
ocular function in anxiety disorders. J Vestib Res. tion of respiratory responses to glutamate micro-
2006;16(45):20915. stimulation of the parabrachial nucleus in the rat.
107. Jacob RG. Panic disorder and the vestibular system. J Neurosci. 1994;14(11 Pt 1):650010.
Psychiatr Clin North Am. 1988;11(2):36174. 123. Huh YE, Kim JS. Bedside evaluation of dizzy
108. Mezzasalma MA, Mathias Kde V, Nascimento I, patients. J Clin Neurol. 2013;9(4):20313.
Valenca AM, Nardi AE. Imipramine for vestibular doi:10.3988/jcn.2013.9.4.203.
dysfunction in panic disorder: a prospective case 124. Craig AD. How do you feel? Interoception: the sense
series. Arq Neuropsiquiatr. 2011;69(2A):196201. of the physiological condition of the body. Nat Rev
109. Stambolieva K, Angov G. Balance control in quiet Neurosci. 2002;3(8):65566. doi:10.1038/nrn894.
upright standing in patients with panic disorder. Eur 125. Craig AD. How do you feelnow? The anterior
Arch Otorhinolaryngol. 2010;267(11):16959. insula and human awareness. Nat Rev Neurosci.
doi:10.1007/s00405-010-1303-2. 2009;10(1):5970. doi:10.1038/nrn2555.
110. Teggi R, Caldirola D, Fabiano B, Recanati P, Bussi 126. Gauriau C, Bernard JF. Pain pathways and parabra-
M. Rehabilitation after acute vestibular disorders. chial circuits in the rat. Exp Physiol. 2002;87(2):
J Laryngol Otol. 2009;123(4):397402. doi:10.1017/ 2518.
S0022215108002983. 127. Zylka MJ. Nonpeptidergic circuits feel your pain.
111. Perna G, Dario A, Caldirola D, Stefania B, Cesarani Neuron. 2005;47(6):7712. doi:10.1016/j.neuron.
A, Bellodi L. Panic disorder: the role of the balance 2005.09.003.
system. J Psychiatr Res. 2001;35(5):27986. 128. Fanselow MS. Neural organization of the defensive
112. Staab JP, Ruckenstein MJ. Which comes first? behavior system responsible for fear. Psychon Bull
Psychogenic dizziness versus otogenic anxiety. Rev. 1994;1(4):42938. doi:10.3758/BF03210947.
Laryngoscope. 2003;113(10):17148. 129. Balaban CD. Vestibular autonomic regulation
113. Furman JM, Jacob RG. A clinical taxonomy of diz- (including motion sickness and the mechanism of
ziness and anxiety in the otoneurological setting. vomiting). Curr Opin Neurol. 1999;12(1):2933.
J Anxiety Disord. 2001;15(12):926. 130. Berencsi A, Ishihara M, Imanaka K. The functional
114. Schuerger RJ, Balaban CD. Immunohistochemical role of central and peripheral vision in the control of
demonstration of regionally selective projections posture. Hum Mov Sci. 2005;24(56):689709.
from locus coeruleus to the vestibular nuclei in rats. doi:10.1016/j.humov.2005.10.014.
Exp Brain Res. 1993;92(3):3519. 131. Balaban CD. Neural substrates linking balance con-
115. Yates BJ, Goto T, Kerman I, Bolton PS. Responses trol and anxiety. Physiol Behav. 2002;77(45):
of caudal medullary raphe neurons to natural vestib- 46975.
ular stimulation. J Neurophysiol. 1993;70(3): 132. Redfern MS, Yardley L, Bronstein AM. Visual influ-
93846. ences on balance. J Anxiety Disord. 2001;15(12):
116. Licata F, Li Volsi G, Maugeri G, Santangelo F. 8194.
Excitatory and inhibitory effects of 133. Szirmai A, Kisely M, Nagy G, Nedeczky Z,
5-hydroxytryptamine on the firing rate of medial Szabados EM, Toth A. Panic disorder in otoneuro-
vestibular nucleus neurons in the rat. Neurosci Lett. logical experience. Int Tinnitus J. 2005;11(1):
1993;154(12):1958. 7780.
117. Balaban CD. Projections from the parabrachial 134. Jacob RG, Whitney SL, Detweiler-Shostak G,
nucleus to the vestibular nuclei: potential substrates Furman JM. Vestibular rehabilitation for patients
for autonomic and limbic influences on vestibular with agoraphobia and vestibular dysfunction: a pilot
responses. Brain Res. 2004;996(1):12637. study. J Anxiety Disord. 2001;15(12):13146.
118. Balaban CD. Vestibular nucleus projections to the 135. Simon NM, Parker SW, Wernick-Robinson M,
parabrachial nucleus in rabbits: implications for ves- Oppenheimer JE, Hoge EA, Worthington JJ, et al.
tibular influences on the autonomic nervous system. Fluoxetine for vestibular dysfunction and anxiety: a
Exp Brain Res. 1996;108(3):36781. prospective pilot study. Psychosomatics. 2005;
119. Balaban CD, Beryozkin G. Vestibular nucleus pro- 46(4):3349. doi:10.1176/appi.psy.46.4.334.
jections to nucleus tractus solitarius and the dorsal 136. Palazzo L, Clemente R, Bersani G. Disturbo di
motor nucleus of the vagus nerve: potential sub- panico e stagionalit: ipotesi di una sottotipizzazione
strates for vestibulo-autonomic interactions. Exp del disturbo sulla base del suo andamento stagionale.
Brain Res. 1994;98(2):20012. Giorn Ital Psicopat. 2005;11:2356.
112 G. Perna et al.

137. Lelliott P, Marks I, McNamee G, Tobena A. Onset of and pulmonary function in panic disorder patients.
panic disorder with agoraphobia. Toward an inte- Rev Psiq Cln. 2009;36:1239.
grated model. Arch Gen Psychiatry. 1989;46(11): 144. Olsson M, Annerbrink K, Bengtsson F, Hedner J,
10004. Eriksson E. Paroxetine influences respiration in rats:
138. Bossini L, Frank E, Campinoti G, Valdagno M, implications for the treatment of panic disorder. Eur
Caterini C, Castrogiovanni P, et al. Neuropsychopharmacol. 2004;14(1):2937.
Photosensitivity and panic-agoraphobic spectrum: a 145. Tucker P, Adamson P, Miranda Jr R, Scarborough A,
pilot study. Riv Psichiatr. 2013;48(2):10812. Williams D, Groff J, et al. Paroxetine increases heart
doi:10.1708/1272.14034. rate variability in panic disorder. J Clin
139. Kellner M, Wiedemann K, Zihl J. Illumination per- Psychopharmacol. 1997;17(5):3706.
ception in photophobic patients suffering from panic 146. Perna G, Alpini D, Caldirola D, Raponi G, Cesarani
disorder with agoraphobia. Acta Psychiatr Scand. A, Bellodi L. Serotonergic modulation of the balance
1997;96(1):724. system in panic disorder: an open study. Depress
140. Bitsios P, Szabadi E, Bradshaw CM. The inhibition Anxiety. 2003;17(2):1016. doi:10.1002/da.10092.
of the pupillary light reflex by the threat of an elec- 147. Meuret AE, Wilhelm FH, Ritz T, Roth WT. Feedback
tric shock: a potential laboratory model of human of end-tidal pCO2 as a therapeutic approach for
anxiety. J Psychopharmacol. 1996;10(4):27987. panic disorder. J Psychiatr Res. 2008;42(7):5608.
doi:10.1177/026988119601000404. doi:10.1016/j.jpsychires.2007.06.005.
141. Bossini L, Fagiolini A, Valdagno M, Padula L, 148. Broocks A, Bandelow B, Pekrun G, George A,
Hofkens T, Castrogiovanni P. Photosensitivity in Meyer T, Bartmann U, et al. Comparison of aerobic
panic disorder. Depress Anxiety. 2009;26(1):E346. exercise, clomipramine, and placebo in the treatment
doi:10.1002/da.20477. of panic disorder. Am J Psychiatry. 1998;155(5):
142. Alici H, Ercan S, Bulbul F, Alici D, Alpak G, 6039.
Davutoglu V. Circadian blood pressure variation in 149. Sanchez-Meca J, Rosa-Alcazar AI, Marin-Martinez
normotensive patients with panic disorder. Angiology. F, Gomez-Conesa A. Psychological treatment of
2014;65(8):7479. doi:10.1177/0003319713512172. panic disorder with or without agoraphobia: a meta-
143. Nascimento I, de Melo-Neto VL, Valenca AM, Lops analysis. Clin Psychol Rev. 2010;30(1):3750.
FL, Freire RC, Cassabian LA, et al. Antipanic drugs doi:10.1016/j.cpr.2009.08.011.
Staging of Panic Disorder:
Implications for Neurobiology
and Treatment

Fiammetta Cosci

Contents Abstract
5.1 Introduction 114 The staging model of panic includes the fol-
lowing stages. Stage 1 in which subclinical
5.2 The Staging Model of Panic Disorder 114
symptoms of agoraphobia or social phobia or
5.3 Subclinical Symptoms 116 generalized anxiety disorder or hypochondria-
5.4 Psychological Development of Panic and sis are present. Stage 2 characterized by the
Neurobiology 118 acute manifestations of agoraphobia or social
5.5 Psychological Development of Panic and phobia or generalized anxiety disorder or
Treatment 120 hypochondriasis. Panic Disorder (PD) with
5.6 Conclusions 122 worsening of anxiety and hypochondriacal
symptoms characterizes stage 3 together with
References 123
demoralization or major depression. Chronic
PD and agoraphobia or social phobia or gen-
eralized anxiety disorder or hypochondriasis
together with increased liability to major
depression may occur at stage 4. This staging
model is applicable in clinical practice. In a
substantial proportion of patients with PD a
prodromal phase and, despite successful treat-
ment, residual symptoms can be identified.
Both prodromes and residual symptoms allow
to monitor the evolution of the disorder during
recovery via the rollback phenomenon. The
different stages of PD and the steps of the roll-
back have a correspondence in its neurobiol-
ogy and in its treatment. The translation of
staging in the neurobiology of panic identifies
different phases in the development of PD
which involve the amygdala, the hippocam-
pus, and the medial/orbital prefrontal cortex.
F. Cosci (*) The treatment implications, although still too
Department of Health Sciences,
University of Florence, Florence, Italy disregarded, emphasize the importance to
e-mail: fiammetta.cosci@unifi.it consider residual symptoms as the final target

Springer International Publishing Switzerland 2016 113

A.E. Nardi, R.C.R. Freire (eds.), Panic Disorder, DOI 10.1007/978-3-319-12538-1_5
114 F. Cosci

of the therapy. In addition, psychotherapy room in customary clinical taxonomy, such as

(mainly cognitive-behavioral therapy) has types, severity, and sequence of symptoms; rate
shown good effectiveness while sequential or of progression in illness (staging); severity of
stage-oriented treatments have shown promis- comorbidity; problems of functional capacity;
ing results. reasons for medical decision; and other aspects
of daily life, such as well-being, distress, and
Keywords mental pain [1, 46]. In recent years, there have
Staging Stage Panic Panic disorder been several exemplifications of this approach in
Neurobiology Treatment research on mood and anxiety disorders, reviewed
by Fava et al. [7].
Although current diagnostic entities (e.g., the
5.1 Introduction Diagnostic and Statistical Manual of Mental
Disorders DSM) are based on clinimetric prin-
Current emphasis in psychiatry is on cross- ciples, their use is still strongly influenced by
sectional assessment of symptoms resulting in psychometric models [3, 8]. This means that the
diagnostic criteria and on comorbidity. The use of severity of each disorder is determined by the
diagnostic criteria is derived from the traditional number of symptoms and not by their intensity or
method of clinical medicine, in which they pro- quality, to the same extent that a score in a self-
vide operating specifications for making a clinical rating scale depends on the number of symptoms
decision about the existence of a particular disease that are scored as positive [8]. As a consequence,
[1]. However, clinicians usually evaluate, in their the preferential target of therapy tends to become
daily practice, also issues that do not simply apply the syndrome resulting from a certain number of
to the severity of the disorder, such as its longitu- symptoms (which may be of mild intensity and
dinal development; social support and adaptation; of doubtful impact on quality of life), instead of
the resilience and reaction to previous conflicts, individual symptoms that may be incapacitating
threats, or losses; the motivation and compliance for the patient [3]. Moreover, clinicians may find
with the treatment; the pre-morbid personality and some difficulties in formulating a treatment plan
potential abnormal personality traits. for people who, for instance, do not reach the
Over the time, a clinical reasoning which known diagnostic threshold. It might be some-
goes through a series of transfer stations [2], times difficult to disentangle whether the symp-
where potential connections between presenting toms the patient is complaining about should be
symptoms and pathophysiological processes are addressed for treatment or whether the conditions
drawn and which are amenable to longitudinal need to be addressed in an integrated way, and if
verification and modification as long as thera- so, by which professional figures.
peutic goals are achieved [3], has been proposed.
Notwithstanding this, a lack of attention to the
longitudinal development of the disorders has 5.2 The Staging Model of Panic
been maintained and has apparently deprived the Disorder
clinical process of a number of important trans-
fer stations. In 1993, Fava and Kellner [9] proposed to use in
The creator of this innovative and emerging psychiatry a staging method that allows charac-
staging approach for assessment was Feinstein terizing a disorder according to seriousness,
who introduced in 1987 the term clinimetrics [1]. extension, features, and longitudinal develop-
Clinimetrics indicates a domain concerned with ment following standardized and well-defined
indices, rating scales, and other expressions that models. They published a seminal paper suggest-
are used to describe or measure symptoms, phys- ing a staging model for schizophrenia, mood, and
ical signs, and other distinctly clinical phenom- panic disorder (PD). The core concept was that
ena in medicine. Further, it provides a home for a these psychiatric disorders develop according to
number of clinical phenomena which do not find main stages. The first stage usually involves the
5 Staging of Panic Disorder: Implications for Neurobiology and Treatment 115

Table 5.1 Stages of panic disorder with agoraphobia Table 5.2 Staging of panic disorder according to Cosci
according to Fava and Kellner [9] and Fava [12]
Stages Stages
1 Prodromal or predisposing: anxiety 1 Prodromal phase: subclinical symptoms of
sensitivity, health anxiety, harm avoidance agoraphobia and/or social phobia and/or
and dependence generalized anxiety disorder and/or
2 Agoraphobia of mild or moderate severity hypochondriasis
(DSM IIIR) (APA, 1987) 2 Acute manifestations of agoraphobia and/or
3 Panic disorder, acute phase (DSM IIIR) social phobia and/or generalized anxiety
(APA, 1987). Worsening of agoraphobia, disorder and/or hypochondriasis
anxiety and hypochondriacal fears and beliefs 3 Panic disorder with worsening of anxiety and
4 Panic disorder, chronic phase (longer than hypochondriacal symptoms. Demoralization
6 months). Increased liability to major and/or major depression may occur
depression 4 Chronic panic disorder and agoraphobia and/
or social phobia and/or generalized anxiety
disorder and/or hypochondriasis (in attenuated
presence of predisposing factors (e.g., genetic or persistent form). Increased liability to
vulnerabilities, pre-morbid personality, lack of major depression
psychological well-being); the second stage is
characterized by the acute symptoms; the third
includes the residual symptoms; the fourth Over the time, the staging model of panic has
implies sub-chronic symptoms; and the fifth been updated [11, 12]. According to the most
stage, when present, is characterized by the recent revision [12], prodromal phase (stage 1)
chronic illness [9]. may include subclinical symptoms of agorapho-
Regarding PD, Fava and Kellner [9] described bia and/or social phobia and/or generalized anx-
a staging model, suggesting that, in a substantial iety disorder and/or hypochondriasis; stage 2 is
proportion of patients, anxiety sensitivity, health characterized by the acute manifestations of
anxiety, harm avoidance and dependence are the agoraphobia and/or social phobia and/or gener-
prodromes of the disorder (stage 1). In the same alized anxiety disorder and/or hypochondriasis;
staging model agoraphobia precedes the first PD with worsening of anxiety and hypochon-
panic attack (stage 2), and acute manifestations driacal symptoms characterizes stage 3 together
of panic disorder, worsened by agoraphobia, with demoralization and/or major depression;
hypochondriacal fears and beliefs (stage 3), can chronic PD and agoraphobia and/or social pho-
be followed by a stage 4 characterized by the bia and/or generalized anxiety disorder and/or
chronic manifestations of PD and increased lia- hypochondriasis together with increased liabil-
bility to major depression (see Table 5.1). ity to major depression may occur at stage 4 (see
This four-stage model was consistent with Table 5.2).
symptomatic patterns of improvement upon This staging model includes agoraphobia,
behavioral treatment of panic disorder as well as social anxiety, generalized anxiety, hypochon-
with the rollback phenomenon upon drug treat- driasis in stages 1, 2, and 4 suggesting to consider
ment. However, since, at least in some patients, the agoraphobia, social anxiety, generalized anxiety,
first panic attack apparently occurred without con- or hypochondriasis as a stage of development of
spicuous prodromal symptoms, while anticipatory PD. Thus, PD seems to become an aspecific clini-
anxiety, phobic avoidance and hypochondriasis cal manifestation in the frame of other anxiety
may develop subsequently, Sheehan and Sheehan disorders. A confirmation of such view comes
[10] outlined a different staging process: stage 1 from Kessler et al. [13] who found that isolated
(subpanic) characterized by panic attacks with lim- panic attacks are quite common and significantly
ited symptoms; stage 2 (panic); stage 3 (hypochon- comorbid with other DSM-IV disorders [14].
driasis); stage 4 (single phobia, that is the setting in A growing literature has supported the exis-
which panic occurs); stage 5 (social phobia); stage tence of the above mentioned clinical stages in
6 (agoraphobia); stage 7 (depression). the longitudinal development of panic.
116 F. Cosci

Roy-Byrne and Cowley [15] observed that, patients under antidepressant treatment before
despite the availability of effective anti-panic starting the behavioral therapy had a worse out-
treatments, PD remains a chronic illness and the come than those antidepressant-free [20].
presence of agoraphobia, major depression, and
personality disorder predict a poor outcome. In
the general population of the Epidemiologic 5.3 Subclinical Symptoms
Catchment Area study, the prodromal period was
about 1015 years long. Panic attacks occurring Prodromes can be identified with the early symp-
in the year before the first interview and the toms and signs of a disease. The prodromal phase
perception that one is a nervous person were connotes a time interval between the onset of
strong predictors of the onset of PD [16]. prodromal symptoms and the onset of the charac-
According to Keller and colleagues [17], the teristic manifestations of the fully developed ill-
probability of having a panic-free interval by ness. Residual symptoms have been identified
12 months was 0.68 for subjects with PD and with the persistent symptoms and signs despite
0.55 for those with PD with agoraphobia [17]. apparent remission or recovery. Of course, in any
ORourke et al. [18] almost confirmed these chronic and recurring medical illness, subclinical
results observing that at 12 month follow-up, 23 fluctuations, either in terms of symptomatology
(33.8 %) PD patients had recovered and remained and laboratory markers, may occur [21].
well, 12 reported relapses and remissions, 19 Detre and Jarecki [22] provided a model for
were still improved short of full recovery, and 14 relating prodromal and residual symptomatology
had persistent PD. The strongest predictor of sus- in psychiatric illness: the rollback phenome-
tained recovery was good clinical status between non. According to this phenomenon, as the ill-
6 and 12 months from baseline, while personality ness remits, it progressively recapitulates even
dysfunction was the most important characteris- though in a reverse order many of the stages and
tic of patients with persistent PD. Over a 1560 symptoms that were seen during the time it devel-
month period of follow up, Cowley et al. [19] oped [21]. According to the rollback model, there
found that only 10 % of the PD patients were is also a temporal relationship between the time
asymptomatic; the strongest predictors of overall of development of a disorder and the duration of
improvement were avoidance coping for out- the phase of recovery. Moreover, there appears to
come at 12 months and Axis I comorbidity for be a relationship between residual and prodromal
outcome at the time of the follow-up evaluation. symptomatology since certain prodromal symp-
Finally, 23 % of 132 PD with agoraphobia toms may persist and progress to become pro-
patients treated with behavioral methods based dromes of relapse.
on exposure homework and followed for a Fava et al. reviewed the literature on several
median period of 8 years had a relapse. The esti- psychiatric disorders and, among them PD [21,
mated cumulative percentage of patients remain- 23, 24], paving the way for the characterization
ing in remission was: 93.1 for at least 2 years, of the phenomenological development of these
82.4 after 5 years, 78.8 after 7 years, and 62.1 illnesses [25] and the application of sequential
after 10 years. The presence of a personality dis- treatment [3].
order and the pretreatment level of depressed Regarding PD with agoraphobia, Fava and
mood indicated a worse prognosis while patients Mangelli [24] illustrated prodromes starting from
who completely overcame agoraphobic avoid- the Kleins model [26] who observed that when
ance had a better outcome. Two additional risk patients are suddenly struck by the first panic
factors involved the use of psychotropic drugs: attack, they develop persistent anticipatory anxi-
those who were still taking benzodiazepines at ety and hypochondriacal fears, leading to avoid-
the end of exposure therapy had a less favorable ant behaviour and agoraphobia. This view has
outcome than those who were drug free; and been supported over time [27, 28]. However, also
5 Staging of Panic Disorder: Implications for Neurobiology and Treatment 117

different sequences of events have been observed. of 64 PD outpatients comparing those who
Wittchen et al. [28] pointed out that 45.1 % of developed residual symptoms at 1-year of fol-
individuals at risk with PD and 76.5 % of those at low-up with those who did not. Surprisingly, the
risk for panic attacks did not develop agorapho- two groups did not differ in terms of achieved
bia (AG). Faravelli et al. [29] re-interviewed 41 improvement measured via the frequency of
subjects with a lifetime history of AG 4 years panic attacks, the degree of avoidant behavior,
after the first evaluation and found that 12 cases the depressive symptoms, or the anxious symp-
had the original diagnosis of agoraphobia with- toms. Those who developed residual symptoms
out a history of panic attacks, 29 had PD with had higher rates in history of anxious disorders in
agoraphobia, 2 no longer met the criterion for childhood, pretreatment panic attacks, presence
agoraphobia turning into social phobia or into of depersonalization and derealization symptoms
specific phobia. In 2010, Wittchen et al. [30] con- during the panic attacks, comorbidity (particu-
firmed that there is no empirical evidence which larly with simple phobia) than those who did not
unequivocally demonstrates that agoraphobia is develop residual symptoms. In addition, the like-
temporally primarily and exclusively a function lihood of developing residual symptoms was
of panic attack (PA) or panic-like features. higher in patients who reported dyspnea as the
According to this body of scientific evidences, main symptom during panic attacks.
the DSM 5 [31] introduced agoraphobia as a In a study conducted by Marchesi et al. [39],
clinically significant disorder that exists indepen- 65 PD patients were followed over 12 months
dently from panic. and randomly treated with paroxetine or citalo-
Some authors also found generalized anxiety pram. A complete remission was achieved
to be prodromal of panic [27]; others observed by only 47.6 % of the subjects, whereas in the
that agoraphobic avoidance, generalized anxiety, remaining patients, that is those who did not
hypochondriacal fears and beliefs occur before reach a complete remission, limited symptom of
the first panic attack [3234]; further works panic attacks, anticipatory anxiety, phobic avoid-
showed that some patients have prodromal ance, and depression were present at the end of
depression, anxiety, or avoidance [35, 36]. the study.
Since 1990s, PD has been recognized as a Several studies have addressed the issue of
chronic illness with little spontaneous improve- the sequence of improvement of symptoms in
ment, high rates of relapse after remission, and patients with PD upon behavioral [4043] or
longer episodes when agoraphobia is a part of the pharmacological [4447] treatment. A seminal
constellation of symptoms [17]. Thus, residual paper was proposed by Fava et al. in 1991 [40] in
symptoms have been found extremely common which they specifically investigated the rollback
and encompassing phobic and anxiety distur- phenomenon in 25 PD patients who received 12
bances, social impairment, and dependence. sessions of in vivo exposure. After the first six
An interesting brief report by Fava et al. [37] sessions, agoraphobia was significantly improved
assessed psychological well-being and residual while 12 sessions of treatment yielded the disap-
symptoms in a sample of 30 patients who had pearance of panic attacks and a further reduction
recovered from PD with agoraphobia and 30 con- of agoraphobic avoidance. The phenomenologi-
trols. Remitted patients displayed significantly cal sequence observed retrospectively for the
more psychological distress (i.e., anxiety, depres- prodromal symptoms of PD was: phobic avoid-
sion, somatic symptoms) than controls; the most ance and hypochondriasis which leaded to panic
common residual symptoms were generalized and, thereafter, to more phobic avoidance and
anxiety, somatic anxiety, low self-esteem, agora- hypochondriasis. On the other hand, the rollback
phobia, and hypochondriasis. Patients with PD phenomenon observed was: a decrease in avoid-
also showed less psychological and physical ance by exposure which seemed to improve
well-being than controls. On the other hand, agoraphobia and panic, with eventual disappear-
Corominas et al. [38] evaluated a clinical sample ance of panic in majority of patients, whereas
118 F. Cosci

agoraphobia persisted at least to a lesser degree. and having high sensitive instruments in detecting
Prodromal symptoms of PD with agoraphobia sub-threshold psychological distress as well as
thus tend to become residual symptoms, which, sub-threshold psychiatric comorbidity [54].
in turn, may progress to prodromal symptoms of
Other authors also found interesting results. 5.4 Psychological Development
Bouchard et al. [48] observed that cognitive of Panic and Neurobiology
changes precede changes in the level of panic
apprehension both when treated with cognitive A longitudinal view of PD, encompassing pro-
restructuring or exposure. Hoffart et al. [43] dromal and residual symptoms, can find interest-
administered an integrated cognitive and behav- ing links to the neurobiology of panic.
ioral model of agoraphobic avoidance to patients In 1989, Gorman et al. [55] articulated a neu-
with PD and agoraphobia and found a feedback roanatomical hypothesis of PD positing that a
loop of effects during treatment: the anxiety elic- panic attack stems from loci in the brainstem that
ited by bodily sensations influenced catastrophic involve serotonergic and noradrenergic transmis-
beliefs and such beliefs influenced avoidant sion and respiratory control, that anticipatory
behavior. A reduction of avoidance, in turn, anxiety arises after the kindling of limbic area
decreased a fear of bodily sensations. Thus, structures, and that phobic avoidance is a func-
avoidant behavior seems to be maintained by tion of pre-cortical activation. The hypothesis
cognitive appraisal, while avoidance maintained asserted that medication exerts its therapeutic
anxiety conditioned to bodily sensations. effect by normalizing brainstem activity in
The role of subclinical symptoms has been patients with panic disorder, whereas cognitive
further increased by the development of the behavioural therapy works at the cortical level.
Diagnostic Criteria for Psychosomatic Research However, this original idea has been surpassed
(DCPR) [49, 50]. Fava and Wise [51] suggested because it was almost completely divorced from
to modify the DSM-IV category [14] concerned research that has mapped out the neuroanatomi-
with psychological factors affecting medical con- cal basis for fear. For this reason, Gorman et al. in
ditions into an expanded category of psychologi- 2000 [56] proposed a revised neuroanatomical
cal factors affecting either identified or feared hypothesis of PD. According to this model, the
medical conditions. They proposed a new section sensory input for the conditioned stimulus runs
with 6 most frequent DCPR syndromes. Among through the anterior thalamus to the lateral
them, the DCPR health anxiety, disease phobia, nucleus of the amygdala and is then transferred to
persistent somatization, demoralization, and the central nucleus of the amygdala which stands
irritable mood offer interesting specifiers for as the central point for dissemination of informa-
subclinical symptoms. For instance, health anxi- tion that coordinates autonomic and behavioural
ety, that encompasses nonspecific dimensions of responses. Efferents of the central nucleus of the
abnormal illness and somatic amplification that amygdala have many targets: the parabrachial
readily respond to appropriate reassurance, may nucleus, producing an increase in respiratory
be the prodromal or the residual symptom of rate; the lateral nucleus of the hypothalamus,
PD. Fava et al. [52] found a significant associa- activating the sympathetic nervous system and
tion between PD and DCPR syndromes of health causing autonomic arousal and sympathetic
anxiety, disease phobia, patterns of somatization, discharge; the locus coeruleus, resulting in an
and irritable mood. increase in norepinephrine release and contribut-
Unfortunately, the DSM 5 [31] did not agree to ing to increase blood pressure, heart rate, and the
this proposal although DCPR are extremely behavioral fear response; and the paraventricular
important, being good predictors of impaired psy- nucleus of the hypothalamus, causing an increase
chosocial functioning in medically ill people [53], in the release of adrenocorticoids. A projection
5 Staging of Panic Disorder: Implications for Neurobiology and Treatment 119

from the central nucleus of the amygdala to the dorsal raphe nuclei play a role in modifying
periaqueductal gray region would be responsible defense/escape responses by means of their
for additional behavioural responses, including inhibitory influence on the periaqueductal gray
phobic avoidance. There are also important region. Third, long-term treatment with SSRIs
reciprocal connections between the amygdala may reduce hypothalamic release of corticotropin-
and the sensory thalamus, prefrontal cortex, releasing factor (CRF). CRF, which initiates the
insula, and primary somatosensory cortex. cascade of events that leads to adrenal cortical
Although the amygdala receives direct sensory production of cortisol, is also a neurotransmitter
input from brainstem structures and the sensory in the central nervous system and has been shown
thalamus enabling a rapid response to potentially to increase fear. According to Gorman et al. [56]
threatening stimuli, it also receives afferents equally intriguing was the possibility that SSRIs,
from cortical regions involved in the processing by increasing serotonergic activity, have an effect
and evaluation of sensory information. Poten- on the central nucleus of the amygdala itself, this
tially, a neurocognitive deficit in these cortical may be a prime site for the anxiolytic action of
processing pathways could result in the misinter- the SSRIs, whereby an increase in 5-HT inhibits
pretation of sensory information known to be a excitatory cortical and thalamic inputs from acti-
hallmark of panic disorder, leading to an inap- vating the amygdala. In addition to their psychic
propriate activation of the fear network via effects, drugs such as SSRIs may eliminate most
misguided excitatory input to the amygdala. It is of the troubling physical effects that may occur
thus conceivable that the misinterpretation of during panic by affecting heart rate, blood pres-
sensory information, which in clinical practice sure, breathing rate, and glucocorticoid release.
can be represented by prodromal health anxiety/ This would lead to a secondary decrease in antic-
hypochondriacal beliefs and fear, may kindle the ipatory anxiety as a patient recognizes that the
fear network inducing a panic attack. seemingly life-threatening physical manifesta-
In this framework, Gorman et al. [56] also tions of panic have been blocked. It is not uncom-
hypothesised the possible neurobiological mech- mon, particularly in the early stages of medication
anisms of Selective Serotonin Reuptake Inhi- treatment of a patient with PD, to hear I some-
bitors (SSRIs) as leading pharmacological times feel as if the attack is coming on but then
treatment of PD. They observed that serotonergic nothing happens. My thoughts dont seem to be
neurons originate in the brainstem raphe region able to cause a panic attack anymore. It is thus
and project widely throughout the entire central conceivable that SSRIs induce a rollback phe-
nervous system. Three of these projections are of nomenon in which pharmacological treatments
particular relevance to an understanding of the first reduce the severity and frequency of panic
SSRI antipanic effect. First, the projection of attacks and then, indirectly, ameliorate anticipa-
serotonin (5-HT) neurons to the locus coeruleus tory anxiety and avoidance.
is generally inhibitory, such that the greater the In 2005, Ninan et al. [57] interestingly
activity of the serotonergic neurons in the raphe, matched the sequence of events in a prototypical
the smaller the activity of the noradrenergic neu- panic attack and what follows with potential neu-
rons in the locus coeruleus. This suggests that robiological alterations. They made the example
SSRIs, by increasing serotonergic activity in the of a young lady experiencing an unexpected
brain, have a secondary effect of decreasing nor- panic attack while driving on the highway. Some
adrenergic activity leading to a decrease in many confluence of events triggers the amygdala, the
of the cardiovascular symptoms associated with central command switch and activates a fixed
panic attacks, including tachycardia and increased action patterns of responses in her brain and
diastolic blood pressure. Second, the projection body. Thus, she pulls over the side of the road
of the raphe neurons to the periaqueductal gray paralyzed with fear and, after a few minutes
region appears to modify defense/escape behav- when the worst is over, she gathers up her cour-
iors. Thus, the serotonergic projections from the age and slowly drives to the safety of her home.
120 F. Cosci

The terrifying experience leaves her an emotion patients as compared to healthy controls. Since
memory (i.e., a strengthening of synapses in the the insula and the ACC are thought to translate
lateral nucleus of the amygdala that represents interoceptive stimulation into feeling, and panic
the experience). Subsequent experiences, either patients overestimate bodily signals, they are a
anticipated or actual, that match components of likely neural substrate of interoceptive supersen-
that emotional memory, now trigger the anxiety sitivity, and a possible site of action of both drug
response. The conventional memory system also and cognitive behavior therapy. As a comple-
remembers the panic attack. Explicit memory ment, antidepressants seem also to prevent panic
involves the hippocampus which is crucial for attacks by enhancing 5-HT inhibition in the PAG.
the autobiographical memory of the attack.
Moreover, the involvement of hippocampus has
also a role in recording the context in which the 5.5 Psychological Development
attack occurred. Thus, for instance, the highway of Panic and Treatment
is now associated with panic attack and driving
may elevate the risk of a further panic attack. The staging model of psychiatric disorders has
Although not previously connected with fear, been strongly related to treatment in unipolar
driving is now associated with vigilance, anxiety, depression, bipolar disorder, and schizophrenia.
arousal. The anticipation of highway drive may On one hand, it has been observed that it allows
become dysphoric and the situation may be there- to recognize a disorder early enough to treat it
fore avoided. The functional anatomy of such precociously, that is making early intervention
avoidance is the medial/orbital prefrontal cortex and prevention. On the other hand, it has been
and its reciprocal connections with the amygdala. observed that it may help in identifying possible
Excessive activation of the amygdala decreases therapeutic strategies for treatment resistant
prefrontal activity, which, in turn, reduces the cases. Unfortunately, current therapeutic models
inhibitory control of amygdala. Thus, the learn- for PD [59, 60] disregard staging as well as sub-
ing of new information that may counter the clinical symptomatology. Yet, there should be
initial association is impaired and avoidance more emphasis on treatment outcome, especially
becomes lasting. Once again, this matched on long-term outcome because of the chronic
sequence of events of a panic attack and the nature of PD. Disappearance of residual symp-
corresponding neurobiological alterations might toms, upon abatement of panic attacks, should be
identify different stages of the development of the final target of therapy since they constitute a
PD which are the mirror of what happens during substantial risk of relapse. Adequate treatments
the rollback. of enduring effects should become of paramount
In 2008, Graeff and Del-Ben [58] further clar- importance including, at least in some patients,
ified the neuroanatomical model of panic focus- long-lasting treatment, sequential, or stage-
ing on the role of 5-HT as enhancer of inhibitory oriented combination of different therapeutic
avoidance in the forebrain and inhibitor of the modalities.
one-way escape in the midbrain periaqueductal In this framework, Shear et al. [61] proposed
gray (PAG). Indeed, experimental studies led to to use a better system to monitor the patients
the association of escape with PD; functional progress because, if residual symptoms are iden-
neuroimaging show activation of the insula and tified clearly and the progress of treatment can be
upper brain stem (including the PAG), as well as mapped, clinicians may appropriately increase or
deactivation of the Anterior Cingulated Cortex possibly decrease medication dosage.
(ACC) during experimental panic attacks; and For a long time, the literature has suggested to
voxel-based morphometric analyses of brain increase the rates of remission combining phar-
magnetic resonance images suggest an increase macological and psychological treatments. More
of grey matter volume in the insula and upper recently, evidence seems not to strongly agree
brain stem, and a decrease in the ACC of panic with it. While in the 1990s antidepressants plus
5 Staging of Panic Disorder: Implications for Neurobiology and Treatment 121

exposure in vivo were proposed to treat PD [62]; of treatment with imipramine combined with
in 2000s, the combination of psychotherapy and behavior therapy to 38 PD with agoraphobia
antidepressant showed higher effectiveness than patients. Of them, 63 % responded markedly to
antidepressant alone but no differences over psy- the sequential treatment; most of the improvement
chotherapy alone [63]. Similarly, psychotherapy in panic occurred during the first 8 weeks,
plus benzodiazepines did not provide a signifi- whereas improvement in severity, anxiety,
cant advantage over psychotherapy alone [64]. depression, and phobias, continued to be signifi-
Thus, for those PD patients who have access to cant between mid-treatment and end of study.
appropriate behavior therapy services, psycho- Further analyses revealed that improvement in
therapy alone might be the most favorable inter- phobic anxiety and avoidance in the first 8 weeks
vention. In terms of differential effectiveness of of treatment, rather than improvement in panic,
various forms of psychotherapy, only behavior predicted the final outcome. De Beurs et al. [72]
therapy and cognitive-behavior therapy were compared fluvoxamine plus exposure, psycho-
homogeneously effective while brief psychody- logical panic management plus exposure, and
namic and cognitive-interpersonal therapy pro- exposure alone and found that the combination of
duced mixed results [63]. fluvoxamine and exposure demonstrated efficacy
When appropriate behavior therapy services superior than that of other treatments at the end
are not available or when the patient does not of the trial. However, these advantages faded at a
want to engage in a psychological path, pharma- 2-year naturalistic follow up [73].
cological treatment remains the only option. In Finally, 63 patients with a primary diagnosis
this case, the choice of the treatment should be of PD who had residual symptoms (i.e., panic
done on the basis of a flourishing literature which attacks, anticipatory anxiety, phobic avoidance)
strongly suggests that there is no adequate evi- despite being on a stable dose of medications for
dence of SSRIs higher effectiveness if compared at least 4 months, were treated with CBT using a
to tricyclic antidepressants or benzodiazepines group format by means of 12 sessions over 4
[65, 66]. Benzodiazepine, in particular clonaze- months. Significant reductions in symptoms were
pam, diazepam, alprazolam, lorazepam [67], are evident for all outcome measures (i.e., frequency
a possible first-line treatment in a PD patient who of panic attacks, agoraphobia, anticipatory anxi-
do not have a comorbid depressive disorder and ety) across treatment, with maintenance of these
are not prone to addiction [67, 68]. However, it gains at 1-year follow-up. At least a 50 % reduc-
should be noted that alprazolam has been related tion in symptoms was achieved by 78 % of the
to continuous and high-dose use; thus, it should sample for ratings of agoraphobia, 62 % for antic-
be carefully used or simply avoided [69] ipatory anxiety, and 49 % for the Hamilton
Different treatment approaches have been also Anxiety Scale score. Fully 81 % of the sample
proposed. Some authors evaluated the effects of a achieved a panic-free status, and 64 % met crite-
sequential treatment, that is a planned sequential ria for remission. The presence of dysthymia,
administration of different therapies based on generalized anxiety disorder, or social phobia at
specific effects induced by each therapy that pro- pre-treatment was associated with a lower like-
vide additional benefits in the course of time. lihood of remission [74]. This study provides
Goldstein [70] administered a 8-week treat- further evidence for the efficacy of CBT as a
ment program, starting with alprazolam and next-step strategy for patients who fail to respond
switching gradually to imipramine, to 6 PD adequately to pharmacotherapy for PD. The
patients. Five completed the treatment program improvement was also maintained despite overall
and, among them, 4 had no panic attacks by the reductions in medication use, indicating that
end of the first week of treatment and maintained CBT can be used as a strategy for medication
the improvement throughout the shift to imipra- discontinuation, with longer-term maintenance
mine. Mavissakalian [71] proposed 8 weeks of of treatment gains. Indeed, persistent post-
treatment with imipramine followed by 8 weeks withdrawal disorders induced, for instance, by
122 F. Cosci

Table 5.3 Staging of levels of treatment resistance in Finally, considering that treatment resistance
panic disorder according to Cosci and Fava [12] in PD is a growing and emerging issue [79], a
Stages staging levels of treatment resistance was pro-
0 No history of failure to respond to therapeutic posed by Cosci and Fava [12] (see Table 5.3).
1 Failure of at least one adequate therapeutic
trial (either pharmacological or psychological)
2 Failure of at least two adequate therapeutic
5.6 Conclusions
trials, including at least one psychological
3 Failure of three or more adequate therapeutic Despite the relative paucity of research on
trials, including at least one concerned with psychological development of PD, the reports
psychotherapy summarized in this chapter address important
4 Failure of three or more adequate therapeutic clinical issues that deserve further study.
(either pharmacological or psychological)
trials, including at least one concerned with The prodromal period seems to be about
psychotherapy/pharmacotherapy combination 1015 years long, the perception that one is a
nervous person [16] or the presence of specific
personality characteristics are strong predictors
paroxetine can be successfully treated with a of the onset of panic disorder. The most common
specific cognitive behavioral therapy which prodromes are depressed mood, illness phobia,
includes explanatory therapy, monitoring of distress and avoidance of closed spaces, exces-
emergent symptoms, homework exposure for sive worries, negative affectivity, anxiety sensi-
avoidance patterns, lifestyle modifications, tech- tivity, health anxiety or fear of disease, separation
niques of decreasing abnormal reactivity to the anxiety. The phenomenological clinical sequence
social environment, and well-being therapy [75]. of PD with agoraphobia is: phobic avoidance
An interesting proposal of stage-oriented ther- and hypochondriasis leading to panic, which, in
apy comes from Fava et al. [76]. They adminis- turn, leads to more phobic avoidance and
tered well-being therapy or CBT of residual hypochondriasis. On the other hand, the rollback
symptoms to three PDA outpatients, one major phenomenon is: a decrease in avoidance by expo-
depressive disorder, four social phobia, one gen- sure, which improves agoraphobia and panic,
eralized anxiety disorder, and one obsessive com- with eventual disappearance of panic, whereas
pulsive disorder patient. Both treatments were agoraphobia persists although to a less degree.
associated with decrease in residual symptoms Prodromal symptoms of PD with agoraphobia
but a significant advantage of well-being therapy thus tend to become residual symptoms which, in
over cognitive behavioral strategies was observed turn, may progress to prodromal symptoms of
when the residual symptoms at the second assess- relapse [40]. Alternatively, the rollback might be
ment (after treatment) of the two groups were characterized by anxiety elicited by bodily sensa-
compared with the initial measurements as tions which influences catastrophic beliefs which,
covariates [76]. This study has obvious limita- in turn, influence avoidant behavior [43].
tions, nonetheless, it provides important clinical The translation of staging in the neurobiology
insights since a novel, and specific psychothera- of panic identifies different phases in the develop-
peutic technique addressed to increasing well- ment of PD. They are the mirror of the steps
being [77] was found to be significantly associated observed during the rollback. The experience of a
with a decrease in residual symptoms in patients panic attack triggers the amygdala which acti-
with mood and anxiety disorders. More recently, vates patterns of responses in the brain and body.
a treatment-resistant panic disorder patient hav- The terrifying experience leaves an emotion
ing difficulties in automatic thought identifica- memory because of which subsequent anticipated
tion with the distress oriented approach of or actual experience, that match components of
cognitive therapy, showed to improve when the that emotional memory, will trigger the anxiety
wellbeing therapy was administered [78]. response. The hippocampus is also involved and
5 Staging of Panic Disorder: Implications for Neurobiology and Treatment 123

records the context in which the attack occurred 11. Fava GA, Rafanelli C, Tossani E, Grandi S. Agora-
so that some details of the threatening experience phobia is a disease: a tribute to Sir Martin Roth.
Psychother Psychosom. 2008;77:1338. doi:10.1159/
are now associated with panic attack. The antici- 000116606.
pation of such details may become dysphoric, and 12. Cosci F, Fava GA. Staging of mental disorders: sys-
the situation may be, therefore, avoided. The tematic review. Psychother Psychosom. 2013;82:20
functional anatomy of such avoidance is the 34. doi:10.1159/000342243.
13. Kessler RC, Chiu WT, Jin R, Ruscio AM, Shear K,
medial/orbital prefrontal cortex and its reciprocal Walters EE. The epidemiology of panic attacks, panic
connections with the amygdala [57]. disorder, and agoraphobia in the National Comorbidity
The treatment implications of the longitudinal Survey Replication. Arch Gen Psychiatry. 2006;
model of PD, although still too disregarded, 63:41524. doi:10.1001/archpsyc.63.4.415.
14. American Psychiatric Association. Diagnostic and
emphasize the importance to consider residual statistical manual of mental disorders (DSM IV). 4th
symptoms as the final target of the therapy and ed. Washington: APA; 1994.
offer adequate treatments of enduring effects. In 15. Roy-Byrne PP, Cowley DS. Course and outcome in
this regard, psychotherapy (mainly cognitive- panic disorder: a review of recent follow-up
studies. Anxiety. 19941995;1:15160. doi:10.1002/
behavioral therapy) [63] has shown the highest anxi.3070010402.
effectiveness and sequential or stage-oriented 16. Eaton WW, Badawi M, Melton B. Prodromes and pre-
treatments have shown promising results. cursors: epidemiologic data for primary prevention of
disorders with slow onset. Am J Psychiatry.
17. Keller MB, Yonkers KA, Warshaw MG, Pratt LA,
References Gollan JK, Massion AO, et al. Remission and relapse
in subjects with panic disorder and panic with agora-
1. Feinstein AR. Clinimetrics. New Haven: Yale phobia: a prospective short-interval naturalistic
University Press; 1987. follow-up. J Nerv Ment Dis. 1994;182:2906.
2. Feinstein AR. Basic biomedical science and the doi:10.1097/00005053-199405000-00007.
destruction of the pathophysiological bridge from 18. O'Rourke D, Fahy TJ, Brophy J, Prescott P. The
bench to bedside. Am J Med. 1999;107:4617. Galway Study of Panic Disorder: III. Outcome at
doi:10.1016/S0002-9343(99)00264-8. 5 to 6 years. Br J Psychiatry. 1996;168:4629.
3. Fava GA, Tomba E. New modalities of assessment doi:10.1192/bjp.168.4.462.
and treatment planning in depression: the sequential 19. Cowley DS, Flick SN, Roy-Byrne PP. Long-term
approach. CNS Drugs. 2010;24:45365. doi:10.2165/ course and outcome in panic disorder: a naturalistic
11531580-000000000-00000. follow-up study. Anxiety. 1996;2:1321. doi:10.1002/
4. Cosci F, Fava GA. New clinical strategies of assess- (sici)1522-7154(1996)2:1<13::aid-anxi2>3.0.
ment of comorbidity associated with substance use co;2-e.
disorders. Clin Psychol Rev. 2011;31:41827. 20. Fava GA, Rafanelli C, Grandi S, Conti S, Ruini C,
doi:10.1016/j.cpr.2010.11.004. Mangelli L, et al. Long-term outcome of panic disor-
5. Fava GA, Tomba E, Sonino N. Clinimetrics: the sci- der with agoraphobia treated by exposure. Psychol
ence of clinical measurements. Int J Clin Pract. 2012; Med. 2001;31:8918.
66:115. doi:10.1111/j.1742-1241.2011.02825.x. 21. Fava GA. Subclinical symptoms in mood disorders:
6. Tossani E. The concept of mental pain. Psychother pathophysiological and therapeutic implications.
Psychosom. 2013;82:6773. doi:10.1159/000343003. Psychol Med. 1999;29:4761.
7. Fava GA, Rafanelli C, Tomba E. The clinical process 22. Detre TP, Jarecki HG. Modern psychiatric treatment.
in psychiatry: a clinimetric approach. J Clin Psychiatry. Philadelphia: Lippincott; 1971.
2012;73(2):17784. doi:10.4088/JCP.10r06444. 23. Fava GA, Kellner R. Prodromal symptoms in affec-
8. Bech P. Fifty years with the Hamilton scales for anxi- tive disorder. Am J Psychiatry. 1991;148:82330.
ety and depression. A tribute to Max Hamilton. 24. Fava GA, Mangelli L. Subclinical symptoms of
Psychother Psychosom. 2009;78:20211. doi:10.1159/ panic disorder: new insights into pathophysiology and
000214441. treatment. Psychother Psychosom. 1999;68:2819.
9. Fava GA, Kellner R. Staging: a neglected dimension doi:10.1159/000012345.
in psychiatric classification. Acta Psychiatr Scand. 25. Fava GA, Tomba E, Grandi S. The road to recovery
1993;87:22530. doi:10.1111/j.1600-0447.1993. from depressiondon't drive today with yesterday's
tb03362.x. map. Psychother Psychosom. 2007;76:2605.
10. Sheehan DV, Sheehan KH. The classification of anxi- doi:10.1159/000104701.
ety and hysterical states. Part II. Toward a more heu- 26. Klein DF. Anxiety reconceptualized. In: Klein DF,
ristic classification. J Clin Psychopharmacol. 1982; Rabkin J, editors. Anxiety: new research and chang-
2:38693. ing concepts. New York: Raven; 1981. p. 23563.
124 F. Cosci

27. Garvey MJ, Tuason VB. The relationship of panic dis- 42. Noda Y, Nakano Y, Lee K, Ogawa S, Kinoshita Y,
order to agoraphobia. Compr Psychiatry. 1984; Funayama T, et al. Sensitization of catastrophic cog-
25(5):52931. doi:10.1016/0010-440X(84)90052-X. nition in cognitive-behavioral therapy for panic disor-
28. Wittchen HU, Nocon A, Beesdo K, Pine DS, Hofler der. BMC Psychiatry. 2007;7:70. doi:10.1186/
M, Lieb R, et al. Agoraphobia and panic. Prospective- 1471-244X-7-70.
longitudinal relations suggest a rethinking of diagnos- 43. Hoffart A, Sexton H, Hedley LM, Martinsen EW.
tic concepts. Psychother Psychosom. 2008;77:14757. Mechanisms of change in cognitive therapy for panic
doi:10.1159/000116608. disorder with agoraphobia. J Behav Ther Exp Psychiatry.
29. Faravelli C, Cosci F, Rotella F, Faravelli L, Catena 2008;39:26275. doi:10.1016/j.jbtep.2007.07.006.
Dell'osso M. Agoraphobia between panic and pho- 44. Rifkin A. The sequence of improvement of the symp-
bias: clinical epidemiology from the Sesto Fiorentino toms encountered in patients with panic disorder.
Study. Compr Psychiatry. 2008;49:2837. doi: Compr Psychiatry. 1991;32:55960.
10.1016/j.comppsych.2007.12.001. 45. Deltito JA, Argyle N, Buller R, Nutzinger D, Ottosson
30. Wittchen HU, Gloster AT, Beesdo-Baum K, Fava GA, JO, Brandon S, et al. The sequence of improvement of
Craske MG. Agoraphobia: a review of the diagnostic the symptoms encountered in patients with panic
classificatory position and criteria. Depress Anxiety. disorder. Compr Psychiatry. 1991;32:1209.
2010;27:11333. doi:10.1002/da.20646. doi:10.1016/0010-440X(91)90003-U.
31. American Psychiatric Association. Diagnostic and 46. Mavissakalian MR. Phenomenology of panic attacks:
statistical manual of mental disorders. 5th ed. responsiveness of individual symptoms to imipra-
Arlington: American Psychiatric Publishing; 2013. mine. J Clin Psychopharmacol. 1996;16:2337.
32. Fava GA, Grandi S, Canestrari R. Prodromal symp- 47. Liebowitz MR, Asnis G, Mangano R, Tzanis E. A
toms in panic disorder with agoraphobia. Am double-blind, placebo-controlled, parallel-group,
J Psychiatry. 1988;145:15647. flexible-dose study of venlafaxine extended release
33. Fava GA, Grandi S, Rafanelli C, Canestrari R. capsules in adult outpatients with panic disorder.
Prodromal symptoms in panic disorder with agora- J Clin Psychiatry. 2009;70:55061. doi:10.4088/
phobia: a replication study. J Affect Disord. 1992; JCP.08m04238.
26:858. 48. Bouchard S, Gauthier J, Nouwen A, Ivers H, Vallires
34. Rudaz M, Craske MG, Becker ES, Ledermann T, A, Simard S, et al. Temporal relationship between
Margraf J. Health anxiety and fear of fear in panic dis- dysfunctional beliefs, self-efficacy and panic appre-
order and agoraphobia vs. social phobia: a prospec- hension in the treatment of panic disorder with agora-
tive longitudinal study. Depress Anxiety. 2010;27: phobia. J Behav Ther Exp Psychiatry. 2007;38:27592.
40411. doi:10.1002/da.20645. doi:10.1016/j.jbtep.2006.08.002.
35. Hayward C, Killen JD, Kraemer HC, Taylor 49. Fava GA, Freyberger HJ, Bech P, Christodoulou G,
CB. Predictors of panic attacks in adolescents. J Am Sensky T, Theorell T, et al. Diagnostic criteria for use
Acad Child Adolesc Psychiatry. 2000;39:20714. in psychosomatic research. Psychother Psychosom.
doi:10.1097/00004583-200002000-00021. 1995;63:18. doi:10.1159/000288931.
36. Lelliott P, Marks I, McNamee G, Tobea A. Onset of 50. Wise TH. Diagnostic criteria for psychosomatic
panic disorder with agoraphobia. Toward an inte- research are necessary for DSM V. Psychother Psy-
grated model. Arch Gen Psychiatry. 1989;46:10004. chosom. 2009;78:3302. doi:10.1159/000235735.
doi:10.1001/archpsyc.1989.01810110042006. 51. Fava GA, Wise TN. Psychological factors affecting
37. Fava GA, Rafanelli C, Ottolini F, Ruini C, Cazzaro M, either identified or feared medical conditions: a solu-
Grandi S. Psychological well-being and residual tion for somatoform disorders. Am J Psychiatry.
symptoms in remitted patients with panic disorder and 2007;164:10023.
agoraphobia. J Affect Disord. 2001;65:18590. 52. Fava GA, Porcelli P, Rafanelli C, Mangelli L, Grandi
38. Corominas A, Guerrero T, Vallejo J. Residual symp- S. The spectrum of anxiety disorders in the medically
toms and comorbidity in panic disorder. Eur Psy- ill. J Clin Psychiatry. 2010;71:9104. doi:10.4088/
chiatry. 2002;17:399406. doi:10.1016/S0924-9338 JCP.10m06000blu.
(02)00693-4. 53. Porcelli P, Bellomo A, Quartesan R, Altamura M,
39. Marchesi C, De Panfilis C, Cantoni A, Giannelli MR, Iuso S, Ciannameo I, et al. Psychosocial functioning
Maggini C. Effect of pharmacological treatment on in consultation-liaison psychiatry patients: influence
temperament and character in panic disorder. Psy- of psychosomatic syndromes, psychopathology and
chiatry Res. 2008;158:14754. doi:10.1016/j. somatization. Psychother Psychosom. 2009;78:352
psychres.2006.08.009. 8. doi:10.1159/000235739.
40. Fava GA, Grandi S, Canestrari R, Grasso P, Pesarin 54. Ferrari S, Galeazzi GM, Mackinnon A, Rigatelli
F. Mechanisms of change of panic attacks with expo- M. Frequent attenders in primary care: impact of med-
sure treatment of agoraphobia. J Affect Disord. ical, psychiatric and psychosomatic diagnoses.
1991;22:6571. Psychother Psychosom. 2008;77:30614. doi:10.1159/
41. Stanley MA, Beck JG, Averill PM, Baldwin LE, 000142523.
Deagle 3rd EA, Stadler JG. Patterns of change during 55. Gorman JM, Liebowitz MR, Fyer AJ, Stein J. A neu-
cognitive behavioral treatment for panic disorder. roanatomical hypothesis for panic disorder. Am
J Nerv Ment Dis. 1996;184:56772. J Psychiatry. 1989;146:14861.
5 Staging of Panic Disorder: Implications for Neurobiology and Treatment 125

56. Gorman JM, Kent JM, Sullivan GM, Coplan Neurother. 2014;14:127586. doi:10.1586/14737175.
JD. Neuroanatomical hypothesis of panic disorder, 2014.963057.
revised. Am J Psychiatry. 2000;157:493505. 69. Cosci F, Guidi J, Balon R, Fava GA. Clinical
57. Ninan PT, Dunlop BW. Neurobiology and etiology of Methodology Matters in Epidemiology: Not All
panic disorder. J Clin Psychiatry. 2005;66:37. Benzodiazepines Are the Same. Psychother Psy-
58. Graeff FG, Del-Ben CM. Neurobiology of panic dis- chosom. 2015;84(5):2624. doi:10.1159/000437201.
order: from animal models to brain neuroimaging. 70. Goldstein S. Sequential treatment of panic disorder
Neurosci Biobehav Rev. 2008;32:132635. with alprazolam and imipramine. Am J Psychiatry.
doi:10.1016/j.neubiorev.2008.05.017. 1986;143:1634.
59. American Psychiatric Association. Practice guideline 71. Mavissakalian M. Sequential combination of imipra-
for the treatment of patients with panic disorder. 2nd mine and self-directed exposure in the treatment of
ed. Washington: APA; 2009. panic disorder with agoraphobia. J Clin Psychiatry.
60. Royal Australian and New Zealand College of 1990;51:1848.
Psychiatrists Clinical Practice Guidelines Team for 72. de Beurs E, van Balkom AJ, Lange A, Koele P, van
Panic Disorder and Agoraphobia. Australian and New Dyck R. Treatment of panic disorder with agorapho-
Zealand clinical practice guidelines for the treatment bia: comparison of fluvoxamine, placebo, and psycho-
of panic disorder and agoraphobia. Aust N Z logical panic management combined with exposure
J Psychiatry. 2003;37:64156. doi:10.1111/j.1440- and of exposure in vivo alone. Am J Psychiatry.
1614.2003.01254.x. 1995;152:68391.
61. Shear MK, Clark D, Feske U. The road to recovery in 73. de Beurs E, van Balkom AJ, Van Dyck R, Lange
panic disorder: response, remission, and relapse. A. Long-term outcome of pharmacological and psy-
J Clin Psychiatry. 1998;8:48. chological treatment for panic disorder with agora-
62. van Balkom AJ, Bakker A, Spinhoven P, Blaauw BM, phobia: a 2-year naturalistic follow-up. Acta Psychiatr
Smeenk S, Ruesink B. A meta-analysis of the treat- Scand. 1999;99:5967. doi:10.1111/j.1600-0447.1999.
ment of panic disorder with or without agoraphobia: a tb05385.x.
comparison of psychopharmacological, cognitive- 74. Heldt E, Manfro GG, Kipper L, Blaya C, Isolan L,
behavioral, and combination treatments. J Nerv Ment Otto MW. One-year follow-up of pharmacotherapy-
Dis. 1997;185:5106. resistant patients with panic disorder treated with
63. Furukawa TA, Watanabe N, Churchill R. Combined cognitive-behavior therapy: outcome and predictors
psychotherapy plus antidepressants for panic disorder of remission. Behav Res Ther. 2006;44:65765.
with or without agoraphobia. Cochrane Database doi:10.1016/j.brat.2005.05.003.
Syst Rev. 2007;1:CD004364. doi:10.1002/14651858. 75. Belaise C, Gatti A, Chouinard VA, Chouinard
CD004364.pub2. G. Persistent postwithdrawal disorders induced by
64. Watanabe N, Churchill R, Furukawa TA. Combined paroxetine, a selective serotonin reuptake inhibitor,
psychotherapy plus benzodiazepines for panic disor- and treated with specific cognitive behavioral therapy.
der. Cochrane Database Syst Rev. 2009;1:CD005335. Psychother Psychosom. 2014;83:2478. doi:10.1159/
doi:10.1002/14651858.CD005335.pub2. 000362317.
65. Berney P, Halperin D, Tango R, Daeniker-Dayer I, 76. Fava GA, Rafanelli C, Cazzaro M, Conti S, Grandi
Schulz P. A major change of prescribing pattern in S. Well-being therapy. A novel psychotherapeutic
absence of adequate evidence: benzodiazepines ver- approach for residual symptoms of affective disor-
sus newer antidepressants in anxiety disorders. ders. Psychol Med. 1998;28:47580.
Psychopharmacol Bull. 2008;41:3947. 77. Ryff CD, Singer B. Psychological well-being: mean-
66. Offidani E, Guidi J, Tomba E, Fava GA. Efficacy and ing, measurement, and implications for psycho-
tolerability of benzodiazepines versus antidepressants therapy research. Psychother Psychosom. 1996;65:
in anxiety disorders: a systematic review and meta- 1423.
analysis. Psychother Psychosom. 2013;82:35562. 78. Cosci F. Well-Being Therapy in a Patient with Panic
doi:10.1159/000353198. Disorder Who Failed to Respond to Paroxetine and
67. Freire RC, Machado S, Arias-Carrin O, Nardi AE. Cognitive Behavior Therapy. Psychother Psychosom.
Current pharmacological interventions in panic disor- 2015;84(5):3189. doi:10.1159/000430789.
der. CNS Neurol Disord Drug Targets. 2014;13: 79. Sanderson WC, Bruce TJ. Causes and management of
105765. doi:10.2174/1871527313666140612125028. treatment-resistant panic disorder and agoraphobia: a
68. Starcevic V. The reappraisal of benzodiazepines in the survey of expert therapists. Cogn Behav Pract. 2007;
treatment of anxiety and related disorders. Expert Rev 14:2635. doi:10.1016/j.cbpra.2006.04.020.
Panic Disorder Respiratory
Morena Mourao Zugliani,
Rafael Christophe R. Freire,
and Antonio Egidio Nardi

Contents Abstract
6.1 Introduction 128 Research into panic disorder (PD) has long
indicated the possibility of distinct PD sub-
6.2 Psychopathology 128
types, such as respiratory, cardiovascular or
6.3 Demographic and Clinical Features 130 gastrointestinal, based on the symptoms
6.4 Psychological Factors 131 experienced during a panic attack (PA). In
6.5 Biological Factors 132
this chapter, we will elaborate on the new
6.5.1 Oxidative Stress and Inflammation 132 developments concerning the respiratory
6.5.2 Genetics 132 subtype (RS) of panic disorder (PD) since its
6.5.3 Respiratory Dysfunction 132 first description, presenting psychopatholog-
6.5.4 Diagnostic Challenge Tests 133
ical features, diagnostic criteria, genetic and
6.6 Treatment 135 physiopathological hypotheses, as well as
6.7 Discussion 135 therapeutic and prognostic characteristics.
6.8 Conclusion 136
Evidence drawn from the available literature
indicates a greater incidence of family his-
References 136 tory of PD, as well as higher comorbidity
rates for disorders of anxiety and depression,
among RS patients in comparison with
patients of the non-respiratory subtype
(NRS). RS patients were also more sensitive
to CO2, hyperventilation and caffeine. These
patients were clearly distinguished from the
NRS patients by certain characteristics, such
as the heightened sensitivity to CO2 and the
higher incidence of a family history of PD.
M.M. Zugliani R.C.R. Freire (*) A.E. Nardi Nonetheless, it was not possible to demon-
Laboratory of Panic and Respiration, Institute of strate differential responses to pharmacolog-
Psychiatry, Federal University of Rio de Janeiro, ical treatment and cognitive behavioral
Rio de Janeiro, Brazil
therapy across the subtypes. RS patients
e-mail: morezugli@gmail.com;
rafaelcrfreire@gmail.com; seem to respond faster than NRS to pharma-
antonioenardi@gmail.com cological treatment with antidepressants and

Springer International Publishing Switzerland 2016 127

A.E. Nardi, R.C.R. Freire (eds.), Panic Disorder, DOI 10.1007/978-3-319-12538-1_6
128 M.M. Zugliani et al.

benzodiazepines, but more studies are needed respiratory diseases are prone to developing
to confirm this finding. panic disorder and agoraphobia [1517]. There
are also findings which point to a link between
Keywords abnormalities in the respiratory control centers
Caffeine Hypercapnia Comorbidity and the physiopathology of PD [4, 18]. Klein [4]
Hyperventilation Nocturnal panic Dyspnea proposed that spontaneous PA happen when a
Respiratory diseases Respiration lack of useful air is signaled by the brain suffoca-
tion monitor, thereby acting as a hyper-sensitive
alarm system against suffocation. However, such
a dysfunction would cause an individual to be
6.1 Introduction susceptible to PA as episodes of false suffo-
cation alarms [4]. Several studies indicate that
Panic Disorder (PD) can be defined as experienc- patients suffering distinct respiratory symptoms
ing unexpected recurring episodes of panic attacks differ in their responses to respiratory and non-
(PA) along with concerns about having other PA, respiratory challenges, when compared to PD
whereby the subject often worries about the con- patients without such prominent symptoms
sequences of PA or suffers significant behavioral [1922].
changes related to the PA [1]. According to the In the studies of Briggs et al. [6], Roberson-
DSM-5, a panic attack may be expected or unex- Nay et al. [2, 3] and other authors [7, 19] there is
pected, and is defined as being a sudden episode of a great deal of evidence to indicate that patients
intense anxiety and fear with a number of symp- with prominent respiratory symptoms during PA
toms [1]. However, the heterogeneity among panic may represent a distinct PD subtype which is
symptoms indicates that there may be distinct PD stable through time. Briggs and colleagues stud-
subtypes [2, 3] such as respiratory, cardiovascular ied 1034 PD patients accounts of their most
or gastrointestinal; as proposed by Klein [4] and recent severe PA [6] and separated the patients
others [5], based on the most prominent symptoms into two groups, based on the presence of promi-
occurring during a typical PA. Distinct groups of nent respiratory symptoms. Patients considered
symptoms suffered by PD patients may have an as being RS showed at least four of the following
association with specific clinical courses, sensitiv- symptoms: Fear of death, pain/discomfort in the
ity to respiratory tests, and efficacy of pharmaco- chest, breathlessness, paresthesias, and the sensa-
logical treatment [6]. Moreover, the diverse tion of choking. Those considered as suffering
clinical presentations of PD may reflect the dis- more natural PA, who also seemed to respond
tinct pathways producing PD [7]. better to antidepressants, belonged to the RS
Although there are a number of similarities group. Those having more situational PA, who
between PA and common fear reactions, there are showed a better response to benzodiazepines,
also distinct psychopathological and neurobio- were part of the NRS group [6].
logical differences [8]. For instance, in PA there In this chapter, the psychopathology, demo-
is a marked feeling of air hunger, which is not graphic features, clinical features, psychological
normally associated with fear reactions that result factors, neurobiological factors and treatment of
of external danger. Another important difference, the respiratory subtype were reviewed. The most
demonstrated clinically and in PD challenge relevant findings regarding the RS since its first
studies [8], is that PA fail to cause activation of description in 1993 have been summarized [6].
the HPA (hypothalamic-pituitary-adrenal) axis.
Furthermore, a great deal of research has been
done on the connection between panic disorder 6.2 Psychopathology
and the respiratory system [4, 911]. Some of
these studies have indicated that PD patients may In 1993, Briggs et al. [6] studied the psychopathol-
suffer from subclinical respiratory abnormalities ogy of PD and discovered a frequent incidence
[1114], as well as the fact that patients with of respiratory symptoms. A principal component
6 Panic Disorder Respiratory Subtype 129

analysis separated two symptom groups: the et al. [26] trembling, palpitations and paresthesias
non-respiratory symptom group, which included were added to the cardio-respiratory cluster along
8 PD symptoms; and the respiratory symptom with fear of dying, chest pain and dyspnea.
group, which included fear of death, breathless- Three factors were recently identified in a later
ness, paresthesias, pain/discomfort in the chest study [28] by factor analysis: the cognitive, with
and a sensation of choking. The NRS group incor- fear of going mad or suffering loss of control and
porated most symptoms of PD and an overall derealization; the autonomic/somatic, including
severity factor was taken into consideration. A dif- dizziness, perspiration, tremors, nausea, chills/
ferent PD subtype was determined by the symp- hot flushes; and cardio-respiratory, involving pal-
toms from the second group, and another subtype pitations, pain in the chest, breathlessness, chok-
was thus defined through the absence of such ing, sensations of numbness and fear of death.
symptoms. When at least four of the five symp- The cardio-respiratory dimension was found to
toms from the respiratory symptom group were be associated with agoraphobic avoidance and
present, the RS was defined; otherwise the RS was severity of panic in multiple regression analy-
absent [6]. Bandelow et al. [23] performed an ses, while those suffering from interference in
oblique principal component cluster analysis on a daily life, mainly with panic preoccupation, were
sample of 330 PD patients from 14 centers in six linked to autonomic/somatic and cognitive sub-
different countries and his findings were almost scales [28].
identical to those of Briggs et al. [6]. It was shown A multi-centric study [29] conducted in Europe
that there was a latent tendency towards symp- indicated than the lifetime prevalence of PA with
toms such as fear of death, pain/discomfort in the sensation of shortness of breath (PASB) was
chest, tingling or numbness, dyspnea and choking/ 6.77 %, while the lifetime prevalence of PA
suffocation. Patients showing at least four of without this symptom was 3.14 %, the 12-month
these symptoms suffered fewer situational PA and prevalence of these PA were 2.26 % and 1 %
appeared to have episodes of a more spontaneous respectively. The prevalence of PASB was signifi-
nature [23]. cantly higher in Spain and Italy, compared to
Factor and group analyses were performed by France, Belgium, Germany and the Netherlands.
Shioiri et al. [24] among a Japanese sample of The PASB was also associated to any chronic
207 PD patents and a distinct RS was not found. physical condition and to high use of health ser-
The respiratory symptoms were not together, vices due to mental health problems [29].
they were divided into two different groups [24]. Roberson-Nay and Kendler [2] examined
Also, in the study from Rees et al. [25] the five panic symptoms across four samples from epide-
respiratory symptoms were not grouped together. miologic studies and one sample from a clinical
The authors made a principal component analysis study with the aim of determining if patients with
of a sample of 153 PD patients and identified five PD have a tendency to co-vary within distinct
components of somatic symptoms. Component subgroups as a function of symptomatic similar-
one included choking and shortness of breath, ity. Examination of panic symptoms from the
component five consisted of chest pain and Epidemiological Catchment Area (ECA), Virginia
numbness. The other PD somatic symptoms were Adult Twin Study of Psychiatric and Substance
distributed across the other three components [25]. Use Disorders (VATSPSUD) and the Cross-
Two studies [26, 27] have shown that cardio- National Collaborative Panic Study (CNCPS)
respiratory symptoms were frequently present revealed two very distinct groups. The first group
among Spanish PD patients. In both studies the was differentiated by its prominent respiratory
factor analysis grouped together the symptoms symptoms as well as higher incidence across a
breathlessness/dyspnea, fear of death and pain/ number of other symptoms of PD, while the other
discomfort in the chest, although in the study of group was characterized by its low endorsement
Marquez et al. [27] the choking sensation was rates in regard to respiratory symptoms of
also included in this cluster. In the study by Segui panic but high endorsement of non-respiratory.
130 M.M. Zugliani et al.

The analysis of samples from the National compared to NRS patients, RS patients showed
Comorbidity Study (NCS) and the National greater comorbidity with agoraphobia, specific
Epidemiologic Survey on Alcohol and Related phobias, social or generalized anxiety disorder
Conditions (NESARC) were also in agreement and major depression [3]. In the same study [3],
with the findings from Briggs et al. [6], with only an analysis of a clinical sample (CNCPS) identi-
slight differences [2]. fied only specific phobias and major depression
With the exceptions of the studies from as comorbidities associated with the RS
Bandelow et al. [23] and Roberson-Nay and (Table 6.1).
Kendler [2], other studies involving factor analy- Higher rates of alcohol consumption [21] and
sis [2428] failed to reproduce the precise respi- cigarette smoking [19] for RS patients were
ratory symptom distinction described by Briggs found by some authors, when compared to the
et al. [6], although there were still many similari- NRS, although other studies failed to show these
ties. These disparities may occur as a result of disparities [3, 3436]. Another study [37] recently
uneven sampling or reduced numbers within the discovered a negative connection between the RS
sample. A further possibility in relation to eluci- and neuroticism, while other PD subtypes were
dation for this phenomenon is that certain factors not correlated to this personality trait.
such as cross-cultural samples or hereditary It is not clear if the RS is associated only with
influences could have had an effect on the preva- spontaneous PA [6, 23], both spontaneous and situ-
lence of panic symptoms in different samples, ational PA [20] or if there are no differences between
affecting the PD subtyping. the subtypes regarding the types of PA [3].
There is also controversy as to whether or
not the RS correlates with nocturnal panic
6.3 Demographic and Clinical attacks (NPA). These occur, usually after 23 h
Features of sleep, when the subject suddenly wakes up
with anxiety, a sensation of fear and physical
It has also been shown in studies of Demographic symptoms [38]. During these attacks, marked
and clinical features that RS PD patients, accord- respiratory symptoms such as breathlessness or
ing to the criteria of Briggs et al. [6] have specific dyspnea [38], chest pains, a sensation of chok-
demographics and clinical aspects. ing, paresthesias and fear of death arise [39].
Some research indicated that in PD patients in 4969 % of PD sufferers experience this kind of
the RS there is a higher age of onset [19, 30], episode [4042]. It is therefore common for PD
although other studies indicated the opposite [21, patients to develop anticipatory anxiety [39],
31], while another recent study found no differ- sleep-onset insomnia and phobic avoidance of
ence between the subtypes regarding age of onset sleep secondary to the NPA [38]. Sarisoy et al.
[3]. Studies comparing the RS with the NRS have [40] also found patients with NPA showing
not found any significant difference in regard to marked respiratory symptoms. Said patients
gender, occupation, marital status or education presented substantially higher levels of chest
[3, 21]. pain or distress, the sensation of choking, par-
Several studies indicated a higher degree of esthesias, dizziness and fear of loss of control
family history of PD for RS patients, compared or going mad than those patients without NPA
to NRS patients [21, 32, 33] (Table 6.1). [40]. Despite evidence indicating a correlation
Previously, some studies indicated that the between NPA and RS, two more recent research
NRS had greater levels of comorbidity with papers [41, 42] failed to demonstrate this
major depression [32, 33] and found no differ- correlation.
ences between the subtypes regarding comorbid- Compared to the NRS, patients in the RS exhib-
ity with agoraphobia and other anxiety disorders ited higher scores on the Clinical Global Impression
[31, 34, 35]. A recent analysis of a large epide- (CGI) scale [31], a lengthier period of illness,
miologic sample (NESARC) demonstrated that, more severe panic and phobic symptoms [19].
6 Panic Disorder Respiratory Subtype 131

Table 6.1 Clinical variations from the respiratory subtype to the non-respiratory subtype
Evidence type RS NRS Reference
Family history of PD (%) 62.175.7 28.635.3 Freire et al. [21], Nardi
et al. [32], Nardi et al. [33]
With agoraphobia (OR/95 % CI) 2.16 (1.513.11)** Roberson-Nay et al. [3]a
With GAD (OR/95 % CI) 3.31 (2.165.08)** Roberson-Nay et al. [3]a
With SAD (OR/95 % CI) 1.78 (1.292.49)** Roberson-Nay et al. [3]a
With specific phobia (OR/95 % CI) 1.88 (1.402.54)** Roberson-Nay et al. [3]a
With major depression (OR/95 % CI) 2.00 (1.163.45)* Roberson-Nay et al. [3]a
Scores on scales of PD severity
Clinical Global Impression (median) 5** 4** Valenca et al. [31]
Anxiety Sensitivity Index (mean/SD) 35.1 (13.2)* 29.5 (13.2)* Onur et al. [34]
Panic-Agoraphobia Spectrum Scale 65.2 (14.3)** 55.6 (17.2)** Onur et al. [34]
WHOQOLb (mean/SD) 64.3 (15.2)* 48.1 (19.9)* De-Melo-Neto [43]
PD panic disorder, SAD social anxiety disorder, GAD generalized anxiety disorder, RS respiratory subtype, NRS non-
respiratory subtype, OR odd ratio, 95 % CI 95 % confidence interval, SD standard deviation, WHOQOL World Health
Organization Quality of Life scale
P 0.05
** P 0.01
Only the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) sample
The difference between the two subtypes was only in the psychological domain

Onur et al.[34] detected no subtype distinctions in

scores on the Panic Disorder Severity Scale 6.4 Psychological Factors
(PDSS) and the Panic and Agoraphobia Scale
(PAS). However, on the Anxiety Sensitivity Index Behavioral sensitization following near-
(ASI) and the Panic-Agoraphobia Spectrum Scale drowning or suffocation incidents may have a
(PAS-SR), patients in the RS still recorded higher significant role in the psychobiology of PD.
scores. This dissimilarity between the subtypes lay Research indicates that 19.333 % of PD patients
among a number of domains of the PAS-SR, such have a history of traumatic suffocation (TSH),
as panic-like symptoms, agoraphobia, reas- while only 6.7 % of normal controls had TSH
surance orientation and separation sensitivity. [44]. Near-drowning, torture or rape involving
These domains also appear to be statistically dis- suffocation or choking were some of the many
criminative of RS and the NRS [34]. A study [43] types of traumatic suffocation episode that were
using the World Health Organization Quality of reported [44]. In comparing PD patients with and
Life Scale (WHOQOL) showed higher scores in without this background, the authors found a
the psychological domain for the RS, compared to higher prevalence of respiratory symptoms and
NRS patients (Table 6.1). nocturnal PA in those with TSH, and more symp-
Two epidemiological studies [3, 29] indicated toms of agoraphobia or cardiovascular symptoms
that RS patients have a tendency to seek more in those patients without such a background [44].
psychosocial treatment and are often given pre- In this comparison no contrasts were presented
scription medicine for symptoms of panic. These concerning childhood separation anxiety or
findings may explain why there are so many dif- familial history of PD [44]. It was discovered in
ferences between clinical sample and epidemio- another study that PD patients who had suffered
logical studies regarding the RS. torture suffocation in the past, showed a higher
132 M.M. Zugliani et al.

incidence of depression and respiratory symptoms and NRS. High ADA activity indicates that
than other patients [45]. TSH patients also pre- inflammatory processes are present in PD patients
sented more posttraumatic stress symptoms on and may increase oxidative stress [46].
CAPS, the Clinician-Administered Posttraumatic
Stress Disorder Scale, than those without THS,
although this was not statistically significant [45]. 6.5.2 Genetics
Bouwer and Stein [44] put forward the hypoth-
esis that an actual suffocation experience could Some studies [21, 32, 33] indicated that RS
augment the sensitivity of the suffocation patients have a higher frequency of familial his-
alarm, which would subsequently be more easily tory of PD compared to NRS, suggesting that
activated. genetic factors may play a role in the physiopa-
thology of respiratory PD.
It is also important to study the role of steroid
6.5 Biological Factors hormones such as progesterone in PD patients
with respiratory symptoms because they may
6.5.1 Oxidative Stress have an effect on breathing [35]. Hormonal
and Inflammation changes directly influence the regulation of
breathing due to changes in the state of excitabil-
There is mounting evidence to indicate that oxi- ity of the respiratory center [35]. Progesterone
dative stress is involved in the development of influences ventilatory control while its impact on
neuropsychiatric disorders, including schizoph- breath stimulation among healthy male subjects
renia, bipolar disorder, depression and PD [46]. has also been demonstrated [35]. Two progester-
The free radicals involved in oxidative stress one receptor isoforms (A and B) mediated all the
have short half-lives but may be indirectly evalu- principal actions of progesterone [35]. The pro-
ated through measurement of the activity of gesterone receptor gene is located on chromo-
certain antioxidant enzymes like superoxide dis- some 11q22-23, and consists of a number of
mutase (SOD) and catalase or glutathione peroxi- polymorphic regions, including an ALU insertion
dase (GSH-Px). The study from Ozdemir et al. polymorphism in intron 7 (PROGINS) and a sin-
[46] indicated that SOD and GSH-Px activity gle nucleotide polymorphism at position +331
was significantly lower in PD patients compared (G331A) in the promoter region [35]. Pirildar
to healthy subjects, but there were no significant et al. [35] compared RS patients with healthy
differences in the comparison between the RS controls and discovered a trend towards a statis-
and NRS. The authors concluded that there is a tically significant difference (P = 0.06) in the
high level of oxidative stress in PD, regardless of PROGINS. The healthy subjects were signifi-
the subtype [46]. cantly different from the RS and NRS patients
There is also evidence of comorbidity between regarding the G331A, but no differences were
PD and immunological diseases such as asthma, found in the comparison between the RS and
in which there is an increase of T lymphocytes NRS [35].
[46]. Adenosine deaminase (ADA) is an enzyme Currently there is insufficient data to deter-
which is important to the maturation and function mine what role genetics may play in the RS and
of T lymphocytes, as well as being an indicator of further study is required.
cellular immunity. It has been suggested that
increased plasma activity of this enzyme occurs
in diseases of an inflammatory nature involving a 6.5.3 Respiratory Dysfunction
cell-mediated immune response. Ozdemir et al.
[46] found higher ADA activity in PD patients in Asmundson and Stein [47] carried out research
comparison with controls. The ADA activity was on pulmonary function in patients with PD and
not significantly different when comparing RS did not find any respiratory damage, although
6 Panic Disorder Respiratory Subtype 133

patients with a low forced expiratory flow, at among RS patients. A study with a challenge of
50 % of a forced expired vital capacity (FEF 5 % carbon dioxide rebreathing [50], had higher
50 %), were different to patients with high FEF numbers of RS patients terminating the proce-
50 %, in regard to panic symptoms. Compared to dure voluntarily, compared to the NRS patients.
patients who presented a high FEF 50 %, during a There was also higher respiratory frequency and
PA, those with low FEF 50 % had a higher num- a greater incidence of suffocation sensations in
ber of strong respiratory and cognitive symp- RS patients, than the other subtype [50]. The
toms, such as breathlessness, a sensation of respiratory ratio, which is a dimensional con-
smothering, giddiness or feeling unsteady, leth- struct based on the respiratory subtype, was also
argy or tingling sensations, fear of death, loss of positively correlated to carbon dioxide sensitivity
control or going mad [47]. It was hypothesized [51]. Freire et al. [51] found that the respiratory
that PD patients with marked respiratory and ratio could predict CO2-induced PA with a sensi-
cognitive symptoms may characterize a different tivity of 67.7 % and a specificity of 65.5 %, using
group of PD patients with early manifestation of a cutoff of 0.437 in the respiratory ratio score.
obstructive pulmonary disease [47]. Pfaltz et al. The idea of two marked PD subtypes was rein-
[48] found no differences between PD patients forced by the differential responses to CO2 and
and healthy subjects regarding respiratory fre- these findings are also compatible with the theory
quency, volumes and irregularities in ambulatory of false suffocation alarm [19]. Patients in the RS
monitoring. However, the severity of respiratory could be oversensitive to CO2, and any trivial
symptoms in PD patients was positively corre- increase in levels of carbon dioxide may be read-
lated with the breath time and the variability of ily misinterpreted as a lack of useful air, causing
the breath time [48]. There was also a negative severe respiratory symptoms as well as other
correlation between the intensity of respiratory symptoms of PA (Table 6.2).
symptoms and an index of rapid shallow breath- In a study by Nardi et al. [52] 62.0 % of the
ing [48]. patients who responded with PA to 35 % CO2
Moynihan and Gevirtz [49] discovered that inhalation and breath holding test were in the RS
RS patients showed lower partial pressure of end- group. Only 30.8 % of the non-responders were
tidal CO2 (PETCO2) (35.14, SD = 3.89 mmHg) RS patients [52]. In another study from the same
compared to NRS (39.27, SD = 3.33 mmHg) and group [53], patients who responded to the double-
healthy subjects (39.43, SD = 2.72 mmHg). The breath 35 % CO2 inhalation presented more
PETCO2 is a measure of the quantity of CO2 in respiratory symptoms chest pain/discomfort,
exhaled air as an indication of the partial pressure shortness of breath, paresthesias and feelings of
of CO2 in the arterial blood. The low PETCO2 in choking in a typical spontaneous PA compared
RS patients indicates that these patients hyper- to non-responders (Table 6.2).
ventilate and eliminate more CO2 than the sub- In another study by Nardi et al. [20], PD
jects in the other two groups [49]. patients were submitted to two panic provoking
challenges with a 1-week interval: a double-
breath 35 % CO2 inhalation and the ingestion of
6.5.4 Diagnostic Challenge Tests 480 mg of caffeine. In the CO2 test, 61.4 % of the
patients suffered PA, while in the caffeine test
Biber and Alkin [19] found that single-breath only 45.8 % had PA. Those who had PA in the
35 % CO2 inhalation induced PA in 79 % of caffeine challenge were also sensitive to CO2,
patients in the RS, while just 48 % of those in the and 76.3 % of these patients were in the RS
NRS suffered PA in this respiratory challenge. group. Among patients who did not respond to
Valenca et al. [31] also found that 93.7 % of the either of the two tests only 37.5 % were in the RS
RS patients had PA during a double-breath 35 % group [20] (Table 6.2).
CO2 inhalation, while only 43.4 % of the NRS It was demonstrated in two studies [22, 54]
had PA, indicating increased sensitivity to CO2 that of those who had PA in the hyperventilation
134 M.M. Zugliani et al.

Table 6.2 Sensitivity to respiratory and non-respiratory challenge tests

N non-
N respiratory respiratory 95 % CI 95 % CI
Author Year N subtype subtype Method OR lower upper P
Abrams 2006 33 10 23 Standardized 2.267 0.452 11.349 0.319
5 % CO2
Biber 1999 51 28 23 Single breath 4.000 1.182 13.525 0.025
35 % CO2/65 %
O2 mixture
Freire 2008 117 66 51 35 % CO2 15.500 4.367 55.011 <0.001
Hyperventilation 0.648 0.273 1.537 0.325
Valena 2002 27 16 11 35 % CO2 18.0 1.722 188.090 0.015
Nardi 2006a 76 39 37 35 % CO2 3.671 1.337 10.077 0.011
Breath holding 12.056 4.090 35.528 <0.001
Nardi 2004a 85 52 33 Hyperventilation 0.847 0.323 2.218 0.736
Breath holding 1.238 0.435 3.517 0.688
Nardi 2007 83 41 42 Caffeine 480 mg 8.861 3.266 24.037 <0.001
35 % CO2 2.197 0.888 5.431 0.088
double breath
Nardi 2006b 91 31 60 35 % CO2 1.222 0.475 3.138 0.676
Nardi 2004b 88 51 37 Hyperventilation 4.727 1.894 11.794 <0.001
95 % CI 95 % confidence interval, N number of subjects, OR odd ratio

challenge, 75.075.6 % were RS patients. Also carbon dioxide [30]. RS patients did not differ
70.0 % of the patients sensitive to the breath from NRS patients regarding the level of anxiety
holding test were in the RS group [22]. In patients and panic symptoms produced by the challenge.
sensitive to both tests 72.0 % were in the RS However, the RS group showed more ventilatory
group, in those not sensitive to any of these tests irregularities and lower PETCO2 than the other
only 37.5 % were in the RS group [55] (Table 6.2). group [30].
Several other methods are used to provoke PA Several studies demonstrated that RS patients
in PD patients; one of these methods is lactate are more prone to panic attacks in challenge tests.
infusion. Massana et al. [56] found that during Regarding sensitivity to CO2, these patients were
the infusion, those with the cardiorespiratory more responsive to the 5 % rebreathing challenge
subtype had tachycardia and localized sweating, [50], the 35 % single-breath challenge [19], the
while the pseudoneurological PD patients had 35 % double-breath challenge [21, 31, 52] and
bradycardia and generalized sweating. the breath holding challenge [22, 52], compared
Studies have shown that hypoxic challenge to NRS patients. The RS patients were also more
tests also provoke PA in PD patients, similarly to sensitive to the hyperventilation challenge test
6 Panic Disorder Respiratory Subtype 135

[21] and to the caffeine challenge test [20]. The the previously obtained improvement was
differential responses to diagnostic challenge maintained [57]. A promising new treatment [58]
tests seem to be the most relevant feature of the is capnometry-assisted breathing therapy, which
RS (Table 6.2). uses a feedback system based on the PETCO2.
This treatment consists of: educating patients on
the role of breathing in PD; correcting problem-
6.6 Treatment atic respiratory patterns; having them perform
different breathing maneuvers with capnometer
Briggs et al. [6] conducted a multi-centric, ran- feedback to experience how changes in breathing
domized, controlled and double-blind medication affect physiology, symptoms, and mood; teach-
trial in 1034 PD patients. At the beginning of the ing them to control PETCO2 level and respiratory
trial patients had been free of all medication for rate; and having them practice breathing exercises
at least 1 week, and they were randomized to daily [58]. PD patients, regardless of the subtype,
receive placebo, imipramine or alprazolam. The showed significant improvement with a decrease
treatment lasted for 8 weeks and the clinicians in PDSS scores after five weekly treatment
were allowed to adjust the dosage according to sessions [58].
response or adverse effects. Alprazolam acted These studies indicate that imipramine, alpra-
faster than imipramine, although at the endpoint zolam, nortriptyline and clonazepam are effec-
both active drugs were significantly superior tive medications in the treatment of all PD
to the placebo. Among RS patients, those who patients; nevertheless tricyclic antidepressants
received imipramine improved more than those may be more effective than benzodiazepines in
who received alprazolam. Among NRS patients, RS patients. Trials with newer drugs are needed
those who received alprazolam showed greater to ascertain if these medications produce differ-
improvement [6]. The higher efficacy of seroto- ent improvements in RS and NRS patients. The
nergic medications may indicate that the physio- CBT has proved itself equally effective in RS and
pathology of the RS may be linked to a serotonin NRS patients.
imbalance. Nardi et al. [32] treated 118 PD
patients with nortriptyline, with a dosage from
50 mg to 150 mg per day, for 52 weeks. The RS 6.7 Discussion
patients improved faster than the NRS patients,
and in week 8 there was a statistically significant In the last 20 years the PD RS subtype has been
difference in the outcome measures, nevertheless extensively studied, and although some findings
by week 52 both groups had improved equally were replicated in more than one study, there
[32]. In a 3-year follow-up study [33] there was were also some controversial findings regarding
also a significantly faster response in the RS the PD subtypes.
patients compared to the NRS patients, both There were significant differences in the prev-
being treated with clonazepam in doses from alence of respiratory symptoms in samples from
1 mg to 4 mg per day. At the endpoint, 3 years different studies, indicating that the prevalence of
later, there were no differences between these two the RS may vary from one population to the
groups regarding the efficacy of the treatment [33]. other. There is evidence of higher prevalence of
Taylor et al. [57] conducted a study on the choking in Caucasian Hispanic compared to
treatment of PD with cognitive behavioral ther- Caucasian non-Hispanic PD patients [59], also,
apy (CBT) with 22 un-medicated patients who African Americans have greater fear of dying and
were submitted to 10 weekly CBT sessions, with fear of insanity, as well as a higher incidence of
all patients improving. There were no significant tingling sensations than European Americans
differences between those with prominent respi- [60], indicating that ethnicity may play a role in
ratory symptoms and those without prominent the unbalanced distribution of PD symptoms
respiratory symptoms. At a 3-month follow-up, across populations. Cultural factors cannot be
136 M.M. Zugliani et al.

ruled out, but the authors believe that the distinct 6.8 Conclusion
gene pools from the studied populations are the
main reason for differences in the prevalence of The definition of the RS made by Briggs et al. [6]
PD symptoms. The increased risk of PD in RS has been confirmed by two other high quality
family members also indicates that genetics play studies [2, 23], but there continues to be no con-
an important role in the physiopathology of this sensus concerning the definition of the respira-
PD subtype. Unfortunately only one study [35] tory subtype of panic disorder. This problem may
addressed the genetic differences between the RS be due to methodological variability between
and NRS and the results were inconclusive. The studies and small sample sizes. The authors
genes implicated in anxiety and respiratory func- believe that the respiratory subtype should not be
tion should both be investigated in future studies defined exclusively based on a symptomatologi-
to identify which of these genes may be respon- cal profile, but also based on other features such
sible for the RS. as respiratory challenge profiles. Incorporating
The RS patients were also more sensitive new criteria would increase the validity of the
than NRS patients to challenge tests with CO2 respiratory subtype, and it would become a use-
inhalation, hyperventilation, breath holding and ful tool for research on panic disorder.
caffeine. The differential responses to diagnostic
challenges were consistent across several studies,
indicating that respiratory system dysfunctions References
may contribute to the physiopathology of the RS
subtype. Recent studies indicated that twin sib- 1. APA. Diagnostic and statistical manual of mental dis-
lings and other family members share carbon orders. 5th ed. Arlington, VA: American Psychiatric
Publishing; 2013.
dioxide sensitivity [6163], once again indicating 2. Roberson-Nay R, Kendler KS. Panic disorder and its
the importance of genetic factors as mediators of subtypes: a comprehensive analysis of panic symptom
this phenomenon. Given that RS is associated with heterogeneity using epidemiological and treatment
CO2 sensitivity and both features are influenced by seeking samples. Psychol Med. 2011;41(11):241121.
3. Roberson-Nay R, Latendresse SJ, Kendler KS. A
genetic factors, it is reasonable to assume that they latent class approach to the external validation of
share at least some genetic characteristics. respiratory and non-respiratory panic subtypes.
RS patients had high comorbidity with ago- Psychol Med. 2012;42(3):46174.
raphobia, social anxiety disorder, generalized 4. Klein DF. False suffocation alarms, spontaneous pan-
ics, and related conditions. An integrative hypothesis.
anxiety disorder, specific phobias and major Arch Gen Psychiatry. 1993;50(4):30617.
depression. These patients also had higher scores 5. Aronson TA, Logue CM. Phenomenology of panic
on clinical scales and lower neuroticism scores, attacks: a descriptive study of panic disorder patients
compared to NRS. These differences in the clini- self-reports. J Clin Psychiatry. 1988;49(1):813.
6. Briggs AC, Stretch DD, Brandon S. Subtyping of
cal presentations of PD, with greater comorbidity panic disorder by symptom profile. Br J Psychiatry.
and more intense PD symptoms, raise the ques- 1993;163:2019.
tion of whether the RS could predict a worse 7. Freire RC, Perna G, Nardi AE. Panic disorder respira-
response to pharmacological and psychological tory subtype: psychopathology, laboratory challenge
tests, and response to treatment. Harv Rev Psychiatry.
treatments compared to the NRS. With the excep- 2010;18(4):2209.
tion of the study from Briggs et al. [6] in which 8. Klein DF. Panic developments. Rev Bras Psiquiatr.
RS responded better to treatment with imipra- 2012;34 Suppl 1:S12.
mine and NRS responded better to alprazolam, 9. Freire RC, Nardi AE. Panic disorder and the respira-
tory system: clinical subtype and challenge tests. Rev
the other studies failed to demonstrate differen- Bras Psiquiatr. 2012;34 Suppl 1:S3241.
tial responses to pharmacological treatment and 10. Nardi AE, Freire RC, Zin WA. Panic disorder and
CBT across the subtypes. RS patients seem to control of breathing. Respir Physiol Neurobiol.
respond faster than NRS to pharmacological 2009;167(1):13343.
11. Gorman JM, Fyer MR, Goetz R, Askanazi J,
treatment with antidepressants and benzodiaze- Liebowitz MR, Fyer AJ, et al. Ventilatory physiology
pines, although further studies are required to of patients with panic disorder. Arch Gen Psychiatry.
confirm this finding. 1988;45(1):319.
6 Panic Disorder Respiratory Subtype 137

12. Stein MB, Millar TW, Larsen DK, Kryger MH. 28. Meuret AE, White KS, Ritz T, Roth WT, Hofmann
Irregular breathing during sleep in patients with panic SG, Brown TA. Panic attack symptom dimensions
disorder. Am J Psychiatry. 1995;152(8):116873. and their relationship to illness characteristics in panic
13. Martinez JM, Papp LA, Coplan JD, Anderson DE, disorder. J Psychiatr Res. 2006;40(6):5207.
Mueller CM, Klein DF, et al. Ambulatory monitoring 29. Fullana MA, Vilagut G, Ortega N, Bruffaerts R, de
of respiration in anxiety. Anxiety. 1996;2(6): Girolamo G, de Graaf R, et al. Prevalence and corre-
296302. lates of respiratory and non-respiratory panic attacks
14. Abelson JL, Weg JG, Nesse RM, Curtis GC. Persistent in the general population. J Affect Disord.
respiratory irregularity in patients with panic disorder. 2011;131(13):3308.
Biol Psychiatry. 2001;49(7):58895. 30. Beck JG, Shipherd JC, Ohtake P. Do panic symptom
15. Hasler G, Gergen PJ, Kleinbaum DG, Ajdacic V, profiles influence response to a hypoxic challenge in
Gamma A, Eich D, et al. Asthma and panic in young patients with panic disorder? A preliminary report.
adultsa 20-year prospective community study. Am Psychosom Med. 2000;62(5):67883.
J Respir Crit Care Med. 2005;171(11):122430. 31. Valenca AM, Nardi AE, Nascimento I, Zin WA,
16. Nascimento I, Nardi AE, Valenca AM, Lopes FL, Versiani M. Respiratory panic disorder subtype and
Mezzasalma MA, Nascentes R, et al. Psychiatric dis- sensitivity to the carbon dioxide challenge test. Braz
orders in asthmatic outpatients. Psychiatry Res. J Med Biol Res. 2002;35(7):7838.
2002;110(1):7380. 32. Nardi AE, Nascimento I, Valenca AM, Lopes FL,
17. Scott KM, Von Korff M, Ormel J, Zhang MY, Mezzasalma MA, Zin WA, et al. Respiratory panic
Bruffaerts R, Alonso J, et al. Mental disorders among disorder subtype: acute and long-term response to
adults with asthma: results from the World Mental nortriptyline, a noradrenergic tricyclic antidepressant.
Health Survey. Gen Hosp Psychiatry. 2007;29(2): Psychiatry Res. 2003;120(3):28393.
12333. 33. Nardi AE, Valenca AM, Nascimento I, Lopes FL,
18. Gorman JM, Kent JM, Sullivan GM, Coplan Mezzasalma MA, Freire RC, et al. A three-year fol-
JD. Neuroanatomical hypothesis of panic disorder, low-up study of patients with the respiratory subtype
revised. Am J Psychiatry. 2000;157(4):493505. of panic disorder after treatment with clonazepam.
19. Biber B, Alkin T. Panic disorder subtypes: differential Psychiatry Res. 2005;137(12):6170.
responses to CO2 challenge. Am J Psychiatry. 34. Onur E, Alkin T, Tural U. Panic disorder subtypes:
1999;156(5):73944. further clinical differences. Depress Anxiety. 2007;
20. Nardi AE, Valenca AM, Lopes FL, De-Melo-Neto 24(7):47986.
VL, Freire RC, Veras AB, et al. Caffeine and 35 % 35. Pirildar S, Bayraktar E, Berdeli A, Kucuk O, Alkin T,
carbon dioxide challenge tests in panic disorder. Hum Kose T. Progesterone receptor gene polymorphism in
Psychopharmacol Clin Exp. 2007;22(4):23140. panic disorder: associations with agoraphobia and
21. Freire RC, Lopes FL, Valenca AM, Nascimento I, respiratory subtype of panic disorder. Klin
Veras AB, Mezzasalma MA, et al. Panic disorder Psikofarmakol B. 2010;20(2):1539.
respiratory subtype: a comparison between responses 36. Freire RC, Nardi AE. Are patients with panic disorder
to hyperventilation and CO2 challenge tests. respiratory subtype more vulnerable to tobacco, alco-
Psychiatry Res. 2008;157(13):30710. hol or illicit drug use? Rev Psiquiatr Clin. 2013;
22. Nardi AE, Valenca AM, Lopes FL, Nascimento I, 40(4):1358.
Mezzasalma MA, Zin WA. Clinical features of panic 37. Kristensen AS, Mortensen EL, Mors O. The associa-
patients sensitive to hyperventilation or breath- tion between bodily anxiety symptom dimensions and
holding methods for inducing panic attacks. Braz the scales of the Revised NEO Personality Inventory
J Med Biol Res. 2004;37(2):2517. and the Temperament and Character Inventory.
23. Bandelow B, Amering M, Benkert O, Marks I, Nardi Compr Psychiatry. 2009;50(1):3847.
AE, Osterheider M, et al. Cardio-respiratory and other 38. Lepola U, Koponen H, Leinonen E. Sleep in panic
symptom clusters in panic disorder. Anxiety. 1996; disorders. J Psychosom Res. 1994;38:10511.
2(2):99101. 39. Lopes FL, Nardi AE, Nascimento I, Valenca AM, Zin
24. Shioiri T, Someya T, Murashita J, Takahashi S. The WA. Nocturnal panic attacks. Arq Neuropsiquiatr.
symptom structure of panic disorder: a trial using fac- 2002;60(3B):71720.
tor and cluster analysis. Acta Psychiatr Scand. 40. Sarisoy G, Boke O, Arik AC, Sahin AR. Panic
1996;93(2):806. disorder with nocturnal panic attacks: symptoms
25. Rees CS, Richards JC, Smith LM. Symptom clusters and comorbidities. Eur Psychiatry. 2008;23(3):
in panic disorder. Aust J Psychol. 1998;50(1):1924. 195200.
26. Segui J, Salvador-Carulla L, Garcia L, Canet J, Ortiz 41. Lopes FL, Nardi AE, Nascimento I, Valenca AM,
M, Farre JM. Semiology and subtyping of panic dis- Mezzasalma MA, Freire RC, et al. Diurnal panic
orders. Acta Psychiatr Scand. 1998;97(4):2727. attacks with and without nocturnal panic attacks: are
27. Marquez M, Segui J, Garcia L, Canet J, Ortiz M. Is there some phenomenological differences? Rev Bras
panic disorder with psychosensorial symptoms Psiquiatr. 2005;27(3):21621.
(depersonalization-derealization) a more severe clini- 42. Freire RC, Valenca AM, Nascimento I, Lopes FL,
cal subtype? J Nerv Ment Dis. 2001;189(5):3325. Mezzasalma MA, Zin WA, et al. Clinical features of
138 M.M. Zugliani et al.

respiratory and nocturnal panic disorder subtypes. CO2 challenge test-induced panic attacks: a compari-
Psychiatry Res. 2007;152(23):28791. son with spontaneous panic attacks. Compr Psychiatry.
43. De-Melo-Neto VL, King AL, Valenca AM, da Rocha 2006;47(3):20914.
Freire RC, Nardi AE. Respiratory and non-respiratory 54. Nardi AE, Lopes FL, Valenca AM, Nascimento I,
panic disorder subtypes: clinical and quality of life Mezzasalma MA, Zin WA. Psychopathological
comparisons. Rev Port Pneumol. 2009;15(5):85974. description of hyperventilation-induced panic attacks:
44. Bouwer C, Stein DJ. Association of panic disorder a comparison with spontaneous panic attacks.
with a history of traumatic suffocation. Am Psychopathology. 2004;37(1):2935.
J Psychiatry. 1997;154(11):156670. 55. Nardi A, Valenca A, Mezzasalma M, Levy S, Lopes F,
45. Bouwer C, Stein D. Panic disorder following torture Nascimento I, et al. Comparison between hyperventi-
by suffocation is associated with predominantly respi- lation and breath-holding in panic disorder: patients
ratory symptoms. Psychol Med. 1999;29(1):2336. responsive and non-responsive to both tests.
46. Ozdemir O, Selvi Y, Ozkol H, Tuluce Y, Besiroglu L, Psychiatry Res. 2006;142(23):2018.
Aydin A. Comparison of superoxide dismutase, gluta- 56. Massana J, Risueno JAL, Masana G, Marcos T,
thione peroxidase and adenosine deaminase activities Gonzalez L, Otero A. Subtyping of panic disorder
between respiratory and nocturnal subtypes of patients patients with bradycardia. Eur Psychiatry. 2001;
with panic disorder. Neuropsychobiology. 2012;66(4): 16(2):10914.
24451. 57. Taylor S, Woody S, Koch WJ, Mclean PD, Anderson
47. Asmundson GJG, Stein MB. A preliminary-analysis KW. Suffocation false alarms and efficacy of cognitive
of pulmonary-function in panic disorderimplica- behavioral therapy for panic disorder. Behav Ther.
tions for the dyspnea-fear theory. J Anxiety Disord. 1996;27(1):11526.
1994;8(1):639. 58. Meuret AE, Wilhelm FH, Ritz T, Roth WT. Feedback
48. Pfaltz MC, Michael T, Grossman P, Blechert J, of end-tidal pCO2 as a therapeutic approach for panic
Wilhelm FH. Respiratory pathophysiology of panic disorder. J Psychiatr Res. 2008;42(7):5608.
disorder: an ambulatory monitoring study. Psychosom 59. Hollifield M, Finley MR, Skipper B. Panic disorder
Med. 2009;71(8):86976. phenomenology in urban self-identified Caucasian-
49. Moynihan JE, Gevirtz RN. Respiratory and cognitive Non-Hispanics and Caucasian-Hispanics. Depress
subtypes of panicpreliminary validation of Leys Anxiety. 2003;18(1):717.
model. Behav Modif. 2001;25(4):55583. 60. Smith LC, Friedman S, Nevid J. Clinical and socio-
50. Abrams K, Rassovsky Y, Kushner MG. Evidence for cultural differences in African American and European
respiratory and nonrespiratory subtypes in panic dis- American patients with panic disorder and agorapho-
order. Depress Anxiety. 2006;23(8):47481. bia. J Nerv Ment Dis. 1999;187(9):54960.
51. Freire RC, Nascimento I, Valenca AM, Lopes FL, 61. Bellodi L, Perna G, Caldirola D, Arancio C, Bertani
Mezzasalma MA, de Melo Neto VL, et al. The panic A, Di BD. CO2-induced panic attacks: a twin study.
disorder respiratory ratio: a dimensional approach to Am J Psychiatry. 1998;155(9):11848.
the respiratory subtype. Rev Bras Psiquiatr. 2013; 62. Perna G, Cocchi S, Bertani A, Arancio C, Bellodi
35(1):5762. L. Sensitivity to 35 % CO2 in healthy first-degree rela-
52. Nardi AE, Valenca AM, Mezzasalma MA, Lopes FL, tives of patients with panic disorder. Am J Psychiatry.
Nascimento I, Veras AB, et al. 35 % carbon dioxide 1995;152(4):6235.
and breath-holding challenge tests in panic disorder: a 63. Battaglia M, Pesenti-Gritti P, Spatola CA, Ogliari A,
comparison with spontaneous panic attacks. Depress Tambs K. A twin study of the common vulnerability
Anxiety. 2006;23(4):23644. between heightened sensitivity to hypercapnia and
53. Nardi A, Valenca A, Lopes F, Nascimento I, Veras A, panic disorder. Am J Med Genet B Neuropsychiatr
Freire R, et al. Psychopathological profile of 35 % Genet. 2008;147B(5):58693.
Lifelong Opioidergic Vulnerability
Through Early Life Separation:
A Recent Extension of the False
Suffocation Alarm Theory
of Panic Disorder

Maurice Preter

Contents Abstract
7.1 Introduction 140 Suffocation-False Alarm Theory (Klein, Arch
Gen Psychiatry 50:306317, 1993) postulates
7.2 Panic and Comorbid Conditions 140
the existence of an evolved physiologic suffo-
7.3 Testing the Panic-Suffocation-False cation alarm system that monitors information
Alarm-Endogenous Opioid Connection 143
about potential suffocation. Panic attacks mal-
References 146 adaptively occur when the alarm is erroneously
triggered. The expanded Suffocation-False
Alarm Theory (Preter and Klein, Biol
Psychiatry 32(3):603612, 2008) hypothesizes
that endogenous opioidergic dysregulation
may underlie the respiratory pathophysiology
and suffocation sensitivity in panic disorder.
Opioidergic dysregulation increases sensitivity
to CO2, separation distress and panic attacks.
That sudden loss, bereavement and childhood
separation anxiety are also antecedents of
spontaneous panic requires an integrative
explanation. Our work unveiling the lifelong
endogenous opioid system impairing effects of
childhood parental loss (CPL) and parental
separation in non-ill, normal adults opens a
new experimental, investigatory area.

M. Preter (*) Keywords

Department of Psychiatry, College of Physicians
Affective neuroscience Childhood parental
and Surgeons, Columbia University,
New York, NY, USA loss (CPL) Endogenous opioids Panic
disorder pathophysiology Expanded
Department of Neurology, Mount Sinai School
of Medicine, New York, NY, USA Suffocation-False Alarm Theory Panic
e-mail: mp2285@cumc.columbia.edu disorder comorbidity

Springer International Publishing Switzerland 2016 139

A.E. Nardi, R.C.R. Freire (eds.), Panic Disorder, DOI 10.1007/978-3-319-12538-1_7
140 M. Preter

7.1 Introduction reported severe early separation anxiety that often

prevented school attendance. Further, panic, in this
I briefly reference previous material on Kleins group, was frequently precipitated by bereave-
suffocation false alarm theory (SFA) of panic dis- ment, or separation [] [2]. This was also noted
order [1] and its amplification in 2008 [2]. I then by others, e.g., Faravelli and Pallanti [4], Kaunonen
discuss a recent finding showing a fundamental et al. [5], Milrod et al. [6].
difference in endogenous opioid reactivity to a We noted that [p]atients highly comorbid for
naloxone challenge in psychiatrically and medi- multiple anxiety disorders are particularly likely
cally healthy adults, depending on whether or not to recall childhood SAD [7], and that [c]laims
they had experienced childhood separation and that separation anxiety equivalently antecedes
parental loss. other anxious states [8] may be due to diagnosti-
In 1993, Klein published the original SFA cally ambiguous limited symptom attacks and the
theory of panic disorder [1], attempting to inte- unreliability of the questionnaire method. We
grate the multiplicity of apparently unrelated concluded that in the only controlled, long-term,
clinical and laboratory observations. We posited direct, blind, clinical interview follow-up of
that a physiologic misinterpretation by a suffo- separation-anxious, school-phobic children, the
cation monitor misfires an evolved suffocation only significant finding was an increased PD
alarm system. This produces sudden respiratory rate [2].
distress followed swiftly by a brief hyperventila- Adding further support, Battaglia et al. [9]
tion, panic, and the urge to flee. Carbon dioxide showed that [s]eparation anxiety correlates with
hypersensitivity is seen as due to the deranged increased familial loading and early onset of
suffocation alarm monitor. If other indicators of PD. More recently (2012), as part of their series
potential suffocation provoke panic, this theoreti- of brilliant twin studies, Robertson-Nay et al.
cal extension is supported. In the original paper, demonstrated that, [childhood] separation anxi-
we tested the theory by examining Ondines ety disorder and adult onset panic attacks share a
curse as the physiologic and pharmacologic con- common genetic diathesis.
verse of panic disorder, splitting panic in terms of
symptomatology and challenge studies, reevalu-
ating the role of hyperventilation, and reinterpret- 7.2 Panic and Comorbid
ing the contagiousness of sighing and yawning, Conditions
as well as mass hysteria [1]. Original SFA
focused on relating the observed lactate and car- There has been a regained awareness of the rele-
bon dioxide hypersensitivity in panic disorder to vance of panic states to other clinical contexts, as
a putative dysfunction in a hypothesized suffoca- Freud suggested in his pioneering statement [10].
tion alarm. At the time, the underlying patho- In the US, nearly half of panic patients are ini-
physiology that might connect the apparent tially seen in the medical emergency room of a
disparate phenomena of panic during relaxation hospital. They may undergo extensive diagnostic
and sleep, late luteal phase dysphoric disorder, medical procedures, such as MRI scans of the
pregnancy, childbirth, pulmonary disease, sepa- brain for headaches, or coronary angiograms for
ration anxiety, and treatment, was unknown. chest pain. Cardiovascular symptoms, particu-
Over the intervening decades, much data larly pseudo-anginal chest pain resembling a
evolved linking Separation Anxiety, Panic, and heart attack are the most common symptoms in
respiratory dysfunction. This suggested a poten- these PD patients experience. Accordingly, 25 %
tial missing link: or more of outpatients seen by a cardiologist have
Klein and Fink [3] posited a developmental a current diagnosis of PD [11]. Since primary
pathophysiological link between [clinical levels complaints are of distress, rather than anxiety,
of] separation anxiety and PD and subsequent ago- they have been termed non-fearful panics.
raphobia, since 50 % of hospitalized agoraphobics This seemingly oxymoronic term nevertheless
7 Lifelong Opioidergic Vulnerability Through Early Life Separation 141

indicates that fearfulness is not essential to panic Air hunger and chronic sighing outside of the
disorder [12]. acute attack are hallmarks of panic that rarely
Migraine and other chronic headaches are occurs under acute, external-threat initiated fear
highly comorbid with panic disorder [13, 14]. [1, 25]. Increasing hypercapnia is a more salient
Having PD increases the risk of migraine four- indicator of potential suffocation than hypoxia,
fold, and vice versa. This bidirectionality suggests but hypoxia also serves this alarm function.
that the migraine-panic association is unlikely to Beck et al. [26, 27] showed that panic patients
be merely coincidental and that shared environ- respond with increased panic symptoms not only
mental and familial factors are involved [13]. In to CO2 inhalation, but also to normocapnic
a longitudinal study, separation and parental loss hypoxia, as predicted by SFA. [2]. Unsurpri-
early in life increased the risk of both headaches singly, numerous studies found that panic disor-
and psychiatric morbidity, mainly anxiety and der and lung disease commonly occur together
depression, in adulthood [15]. [2837]. More specifically, we wrote, early lung
Panic disorder is also comorbid with other disease, including asthma and COPD may predis-
somatic pain syndromes [16]. In a cross-sectional pose to PD [28, 3842], or present solely with
survey of 1219 female veterans studying the panic symptoms [43, 44]. Asthma and PD are
prevalence and frequency of mastalgia, women both characterized by acute episodes, salient
reporting frequent mastalgia were much more respiratory symptoms and anxiety with avoid-
likely to have comorbid panic disorder [OR 7.1], ance of situations related to acute attacks [1, 37].
but also post-traumatic stress disorder, mood dis- There is a significantly higher (6.524 %) preva-
orders, and other somatic pain syndromes, such lence of PD in asthmatics [36, 45, 46] than the
as fibromyalgia, chronic pelvic pain or irritable 13 % reported in the general population [47,
bowel syndrome. 48]. Perna et al. [41] found a significantly higher
Although panic attacks as they occur in panic prevalence of PD, sporadic panic attacks, and
disorder often prominently feature air hunger, social phobia in asthmatics than the general pop-
other panic subtypes have prominent vestibular ulation. In 90 % of asthmatics with PD, asthma
symptoms and unspecific dizziness/lightheaded- appeared first. Panic symptomatology during the
ness. True vertigo was historically recognized as asthmatic attack predicted longer hospitaliza-
a common presentation of panic disorder [17, 18]. tions in asthmatic patients [4951]. [2].
The current rigid distinction between psychiatry The recent amplification of SFA centers on the
and neurology [19] interferes with proper assess- observation that both separation anxiety and suf-
ment [20]. focation sensitivity are under endogenous opioi-
In Mandarin Chinese, tou yun refers to the dis- dergic control. We amplified the SFA theory by
abling sensation of a constant state of movement suggesting that PD may be due to an episodic
of oneself or one's surroundings. This dizziness functional endogenous opioid deficit [25]. The
(note that tou yun also describes vertigo), is prob- following is a necessarily brief explanation.
ably the most common expression of panic disor- The endogenous opioid system was discov-
der in Chinese patients [21], so by sheer numbers, ered in the early 1970s. Electrical stimulation the
this may well be the most prevalent panic sub- periaqueductal gray [52] produced analgesia that
type worldwide. was reversed by naloxone, suggesting an endo-
Taken together, the various comorbidities of genous opioid system. Opioid molecules are
panic disorder and untreated sequelae massively among the oldest evolved signaling substances.,
impact peoples quality of life [22]. functioning in many physiological processes e.g.,
One prominent characteristic of the panic pain perception, respiration [53]. Dyspnea is
attack and subthreshold panic-related anxiety is modulated by central and peripheral opioid levels
respiratory dysregulation and chaotic breathing. in both rodents and humans [54].
This can be experimentally reproduced in adult In mice, exposure to intermittent, severe
panic sufferers, but also in children with separa- hypoxia prolonged survival during subsequent
tion anxiety disorder [2, 23, 24]. lethal suffocation [55]. This effect was blocked
142 M. Preter

by naloxone, implying that endogenous opioids The developmental phase of separation

increase adaptability to hypoxic environments. anxiety serves as a biologic leash for the increas-
Opioid receptors, including non-conventional ingly mobile, but helpless infant who continually
ones, can be found throughout the respiratory checks for the mothers presence, becomes
tract. Nebulized morphine is an outstanding acutely distressed on discovering her absence,
treatment for chronic dyspnea ([5659]. and immediately attempts to elicit retrieval by
In our 2008 paper, we summarized data from crying. In humans, separation anxiety usually
developmental psychobiology and neuroanatomy wanes around age four when the now verbally
that point to a possible link between separation skilled child can successfully elicit care even
and the endogenous opioid system, as follows: from non-relatives.
Following birth, mammalian infants cannot Using electrical brain stimulation (ESB), DVs
survive independently. Survival requires reli- have been elicited in many species from homolo-
able distress signaling mechanisms to elicit gous areas, including the midbrain, dorsomedial
parental care and retrieval. Distress vocaliza- thalamus, ventral septum, preoptic area, and the
tions (DVs) are a primitive form of audio-vocal bed nucleus striae terminalis (BNST). In some
communication [60]. A common neuroanatomy higher species, one can obtain separation calls by
subserving DVs may be shared by all mammals, stimulating the central amygdala and dorsomedial
although substantial functional variations hypothalamus. All these sites have high opioid
depend on the ontogenetic niche. The latter [61] receptor densities and figure heavily in sexual
signifies the ecological and social legacies (the and maternal behaviors [60]. Cortically, electrical
inherited environment) in which a given set of stimulation of the rostral cingulate gyrus in mon-
genes develops. For instance, isolated altricial keys consistently elicits distress calls [65, 66]. The
(developmentally immature) infants do not emit cingulate cortex, found exclusively in mammals,
DVs compared to other species, since it is not is particularly well developed in humans and con-
likely they will stray from the nest [62]. tains high densities of opioid receptors [67].
Immature human infants practically never get Naloxone-blockable opioid agonists reduce
lost for their first 6 months. Despite frequent isolation-induced distress vocalizations (DVs)
maternal absence, separation anxiety in humans across mammalian species [6871]. In beagles,
develops only after their motor system matures. imipramine, the classic anti-panic agent, and
Young rats are not specifically attached to their morphine were the only psychotropic drug that
mother, i.e. any mother will do as heater or yielded specific DV reduction at nonsedating
feeder. Only once mobile do they socially bond, doses [64, 71].
but their responses do not compare with the Naloxone given to guinea pigs and young
vigor seen in other species. Rats also differ from chicks [71] increased baseline vocalizations (by
other species, including primates, dogs and 600 %), but only when the animals were in a group,
chicks in their greater DV suppression by benzo- since isolates already emitted maximum DVs.
diazepines [6264]. Since benzodiazepines dif- Kalin et al. [69] studied opioid modulation of
ferentially alleviate anticipatory anxiety, social separation distress in primates, showing mor-
isolation in young rodents, as compared to many phine (0.1 mg/kg) significantly decreased separa-
other mammals, may activate anxiety mecha- tion distress vocalizations without changes in
nisms other than separation distress. Thus, autonomic and hormonal activation. Naloxone
Panksepp emphasizes that when using cross-spe- (0.1 mg/kg) blocked this effect. Sympathetic
cies analogies, it is important to keep in mind blockade using the (2) agonist, clonidine, and
that the type and degree of social separation dis- the adrenergic antagonist, propranolol, had no
tress depends on ecological and developmental specific effect on separation-induced coos in
parameters [62]. infant rhesus monkeys [2, 72].
7 Lifelong Opioidergic Vulnerability Through Early Life Separation 143

7.3 Testing the Panic- Based on these initial findings, and cognizant
Suffocation-False Alarm- that previous experiments using smaller doses of
Endogenous Opioid both intravenous and oral opioid blocking agents
Connection had shown little results [83, 84] we decided to
conduct a controlled, randomized experimental
Panic Disorder is unique among psychiatric dis- study to investigate whether high-dose naloxone,
orders in that its salient component, the panic an intravenous opioid receptor antagonist, could
attack, can be reliably incited in laboratory set- change the regularly resistant normal controls to
tings by specific chemical challenges as well as become more sensitive to intravenous lactate as a
having challenges specifically blocked by anti- respiratory stimulus to tidal volume increment.
panic agents, e.g. imipramine. We can experi- Study design and statistical analysis are detailed
mentally turn panic on and off, producing elsewhere [85], but in addition to the usual stan-
trenchant causally related data rather than infer- dard recruitment procedure for healthy research
ences from naturalistic data. [2]. Specifically, subjects, eligible volunteers were further inter-
sodium lactate infusions and CO2 inhalation reg- viewed about potentially significant individual
ularly produce panic attacks in patients with and family antecedents and comorbidities of
panic disorder [7375]. However, while normal panic, such as near-suffocation, pulmonary dis-
controls or patients with other anxiety disorders ease, and migraine headache. Recent and child-
rarely show such reactivity (i.e., progress to a hood loss and separation events (parental divorce
full-blown panic attack) [1], higher concentra- or death, childhood abuse) were specifically
tions of inhaled CO2 are highly aversive and can reviewed. [85].
produce respiratory panic symptomatology in a Results showed that [n]ormal subjects, usu-
dose-dependent fashion [7678]. ally relatively insensitive to the TV effects of lac-
Both spontaneous and lactate induced panic tate infusion, in this study, given opioid antagonist
attacks in panic patients produce air hunger and pretreatment, developed TV and RR increments
marked, objective increases in tidal volume (Vt) resembling those occurring in both spontaneous
[79, 80]. Since sodium lactate infusion causes a clinical panic attacks and in panic patients who
metabolic alkalosis, a compensatory decrease in panic during lactate infusions [7375]. The
ventilation would be expected. This would hypothesis that a functioning endogenous opioid
homeostatically buffer blood pH, by increasing system buffers normal subjects from the behav-
CO2 retention. However, the converse actually ioral and physiological effects of lactate is conso-
occurs indicating a specific lactate stimulating nant with these results.
effect on respiration. The most interesting aspect of this study is
The usual response of healthy control subjects that for the first time the prolonged physiologi-
to a sodium lactate infusion is a minor, but defi- cal effects of actual separations and losses during
nite increase in Vt [81]. The lesser tidal volume childhood, i.e. parental death, parental separation
response in lactate challenged normal subjects or divorce, on the endogenous opioid system of
may be due to buffering by their intact endoge- healthy adults have been objectively, experimen-
nous opioid system. tally shown. Presence or absence of childhood
An open pilot study showed that naloxone parental loss (CPL) antecedents determined the
infusion (ranging from an initial 0.5 mg/kg to response to the naloxone-lactate probe. [85].
a maximum of 2 mg/kg) followed by lactate In these carefully screened medically and psy-
(N + L), caused significant tidal volume incre- chiatrically healthy subjects, a history of child-
ments similar to those observed during clinical hood parental loss or separation decreased the
and lactate induced panic attacks in 8 of 12 nor- naloxone + lactate effect, implying that there was
mal subjects, supporting the hypothesis that opi- an antecedent decrement in opioidergic activity,
oidergic deficiency might be necessary for lactate so that the naloxone had nothing to block. It was
to produce a marked increase in tidal volume in the subjects that had not suffered such separation
normal subjects [82]. events that showed the expectable tidal volume
144 M. Preter

increment (hyperventilation) used as outcome CPL as related to childhood separation anxi-

measure. ety, adult panic disorder (PD) and suffocation
The import of these findings is that analyses hypersensitivity was studied by Battaglia et al. [97].
attempting to relate CPL to other baseline vari- In a large sample of twins from Norway, CPL
ables may well fail since CPL impact may be spe- accounted in no small part for the covariation
cific to challenges to the endogenous opioid between separation anxiety in childhood, hyper-
system. [85]. Also it implies that separation- sensitivity to CO2 (as indexed by the anxiety
induced, baseline opioidergic deficiency, while it response to a 35 % CO2/65 % O2 mixture), and
may not be sufficient to induce overt disease, PD in adulthood. Note that in Battaglias study,
confers lifetime vulnerability even in healthy CPL increased reactivity to the 35 % CO2 probe.
adults. Whether it is longitudinally, or cross- [85] []
sectionally relevant to somatic pain syndromes Testing the specificity of the naloxone-lactate
such as migraine, and to opiate abuse ought to be model of clinical panic requires double-blind
determined. investigation whether specific anti-panic drugs,
Again, we emphasize that these CPL effects but not panic irrelevant drugs, block this effect.
were apparent in a normal sample. In a society If found, this has practical and heuristic
where divorce rates approach 50 %, the results implications.
raise the question whether current psychiatric First, there is currently no specific, screening
classification and diagnostic scales are sensitive method for testing putative anti-panic drugs
enough to detect the effects of childhood parental except by the experimental treatment of panic
loss. This applies as well to developing societies disorder patients. The naloxone + lactate effect in
like China, where the massive migration of normal humans may afford such a screening
mostly young individuals from the countryside to method, and may be extended to preclinical stud-
urban areas has left approximately 30 million ies. Second, these data offer heuristic support for
small children behind [85, 86]. the theory that an opioidergic dysfunction is the
It is known that early maternal separation pathophysiological mechanism underlying panic
is a risk factor for adult anxious-depressive disorder. If so, the appropriateness of opioidergic
and borderline psychopathology [8789]. therapeutic agents comes into question. The use
However, its detrimental long-term effect is of morphine or other simple agonists would prob-
not limited to psychiatric illness [90]. Using ably be rejected for fear of inducing addiction,
criteria similar to ours, the Adverse Childhood although the evidence for addiction during indi-
Experiences (ACE) Study, a CDC supported cated medical treatment is slim. However, recent
prospective cohort study of 16,908 adults found work with opioidergic mixed agonist-antagonists
a significant relationship between CPL and pre- [98, 99], e.g. buprenorphine, may be relevant.
mature death in adulthood [91]. Retrospective The concern about addiction would be mitigated
(e.g., [92, 93]) and prospective longitudi- by the fact that higher doses become receptor
nal data [15, 9496], link family disruption, blockers rather than agonists. Positive results
physical abuse, separation and maternal loss would foster investigations into basic molecular
in early life to chronic physical pain in adult- mechanisms. For instance, we note that the dose
hood. It should be explored whether the NL of naloxone used in our study (2 mg/kg) substan-
vs. SL probe has a differential effect on pain tially exceeds that needed for opioid receptor
perception and physiological pain measures, (MOR) blockade [100], suggesting a role for the
and whether the presence of childhood paren- opioid receptor (DOR). This could spark inter-
tal loss antecedents modifies this interaction. est in the development of specific DOR agonists
Unfortunately, our exploratory pain measure suitable for human use. Currently, such agents
was limited to a single item, and in retrospect, have not been developed, although agents suit-
was clearly inadequate. able for animal use are available. [85].
7 Lifelong Opioidergic Vulnerability Through Early Life Separation 145

Since the publication of our paper, exciting fear and anxiety-related behaviors. In humans,
new work in panic disorder has emerged, notably focal bilateral amygdala lesions are extraordi-
from Brazil. Appropriately, the First World narily rare, and such cases have been crucial for
Symposium On Translational Models Of Panic understanding the role of the human amygdala in
Disorder, was held in Vitoria, E.S., in November fear. [] The most intensively studied case is
of 2012. Moreira et al. [101] present a thoughtful patient SM, whose amygdala damage stems from
review of the use of rodents in panic disorder Urbach-Wiethe disease [] Previous studies
research. Graeff [102], studying an animal model have shown that patient SM does not condition to
of panic disorder found that the inhibitory action aversive stimuli [], fails to recognize fearful
of serotonin is connected with activation of faces [] and demonstrates a marked absence
endogenous opioids in the periaqueductal gray of fear during exposure to a variety of fear-
(PAG). Schenberg and colleagues [103, 104] sug- provoking stimuli, including life-threatening
gest [the PAG] harbors an anoxia-sensitive suf- traumatic events []. Patients with similar
focation alarm system. Activation precipitates lesions have largely yielded similar results [].
panic attacks and potentiates the subjects One stimulus not previously tested in humans
responses to hypercapnia. Notably, the resem- with amygdala damage is CO2 inhalation.
blance of these effects to panic disorder was sup- Inhaling CO2 stimulates breathing and can pro-
ported by their pharmacological parallel to panic voke both air hunger and fear [] Furthermore,
disorder treatment. This model was also sup- CO2 can trigger panic attacks, especially in
ported by demonstrating a lack of stress hormone patients with panic disorder []. Recent work in
release during DPAG stimulation thus paralleling mice found that the amygdala directly detects
panic disorder [105, 106]. The utility of opioider- CO2 and acidosis to produce fear behaviors [].
gic mixed agonist-antagonists in animal models Thus, we hypothesized that bilateral amygdala
of panic disorder and in treatment refractory lesions would reduce CO2-evoked fear in humans.
patients would seem promising. In contrast with our prediction, patient SM
The lack of hypothalamic pituitary adrenal reported fear in response to a 35 % CO2 inhala-
(HPA) activation during the panic attack, as it tion challenge. To the best of our knowledge, this
occurs in panic disorder, is a striking peculiarity was the first time patient SM experienced fear in
since it contradicts the belief that the panic is an any setting, laboratory or otherwise, since child-
expression of a hypersensitive fear mechanism hood []. To further explore this issue, we tested
that stimulates the HPA anti-stress response, sup- two additional patients (AM and BG), monozy-
posedly reactive to all dangers. This is usually gotic twin sisters with focal bilateral amygdala
understood as dependent on hyper-responsiveness lesions resulting from Urbach-Wiethe disease
of the amygdala. For instance, both Stein [107], [] As with patient SM, both patients also
and Gorman et al. [108] neglect or dismiss the reported experiencing fear during the CO2 chal-
incongruity of the lack of HPA response, claim- lenge. [109].
ing a supposed amygdala-based hypersensitive These startling observations affirm that the
fear system as central to panic. reaction to carbon dioxide must be due to an
Further damage to the amygdalocentric fear alternative alarm system, such as has been pro-
system theory is provided by Feinstein et al. posed for possible suffocation. We have sug-
[109] who studied three patients with amygdala gested that under conditions of threatened
damage produced by Urbach-Wiethe syndrome. asphyxia the activation of the HPA system would
It is worth extensive citation: produce a counterproductive hyperoxidative
A substantial body of evidence has empha- state and is therefore inactivated. This is in keep-
sized the importance of the amygdala in fear ing with the observation that the tachycardia dur-
[]. In animals, amygdala-restricted manipula- ing panic is produced by vagal withdrawal rather
tions interfere with the acquisition, expression than a counterproductive sympathetic oxidative
and recall of conditioned fear and other forms of surge [2].
146 M. Preter

In conclusion, we objectively, experimentally 5. Kaunonen M, Paivi AK, Paunonen M, Erjanti

showed a physiological link between endogenous H. Death in the Finnish family: experiences of
spousal bereavement. Int J Nurs Pract. 2000;6:
opioid system deficiency and panic-like suffoca- 12734.
tion sensitivity in healthy adults. This is conso- 6. Milrod B, Leon AC, Shear MK. Can interpersonal
nant with the expanded Suffocation-False Alarm loss precipitate panic disorder? Am J Psychiatry.
theory of panic suggesting an episodic functional 2004;161:7589.
7. Lipsitz JD et al. Childhood separation anxiety disor-
endogenous opioid deficit [25, 110]. The speci- der in patients with adult anxiety disorders. Am
ficity of the naloxone + lactate model of clinical J Psychiatry. 1994;151:9279.
panic should be tested using specific anti-panic 8. van der Molen GM, van den Hout MA, van Dieren
components, possibly including opioidergic mixed AC, Griez E. Childhood separation anxiety and
adult-onset panic disorders. J Anxiety Disord. 1989;
agonist-antagonists such as buprenorphine. If 3:97106.
specific, the naloxone + lactate effect in normal 9. Battaglia M et al. Age at onset of panic disorder:
humans affords a screening method for testing influence of familial liability to the disease and of
putative anti-panic drugs which is currently not childhood separation anxiety disorder. Am J Psy-
chiatry. 1995;152:13624.
available. This could obviate the experimental 10. Freud, S. [1895].ber die Berechtigung, von der
treatment of panic disorder patients in drug Neurasthenie einen bestimmten Symptomenkomplex
development. als Angst-Neurose abzutrennen. Neurol. Zbl.,
Our data also show for the first time that actual XIV, p. 50-66; G.W., I, p. 315342; On the grounds
for detaching a particular syndrome from neurasthe-
separations and losses during childhood, such nia under the description anxiety neurosis. SE, 3:
parental death, parental separation or divorce 1895;85115.
(CPL), effect lifelong alterations in the physio- 11. Ballenger JC. Treatment of panic disorder in the
logical reactivity of the endogenous opioid sys- general medical setting. J Psychosom Res. 1998;
tem of healthy adults. 12. Beitman BD, Kushner M, Lamberti JW, Mukerji
This result encourages epigenetic inquiry into V. Panic disorder without fear in patients with angi-
the effects of CPL on endogenous opioid sys- ographically normal coronary arteries. J Nerv Ment
tems, and the role of these systems in resilience, Dis. 1990;178:30712.
13. Breslau N, Schultz LR, Stewart WF, Lipton R,
but also in the chronification of symptoms. A sec- Welch KM. Headache types and panic disorder:
ond step would be to explore similarities and directionality and specificity. Neurology. 2001;
differences between CPL effects and those of 56:3504.
subthreshold events (such as maternal rejec- 14. Hamelsky SW, Lipton RB. Psychiatric comorbidity
of migraine. Headache. 2006;46(9):132733.
tion; e.g. [111, 112]). Finally, a redefinition of 15. Fearon P, Hotopf M. Relation between headache in
what constitutes a (truly) healthy control (with- childhood and physical and psychiatric symptoms in
out antecedents of CPL) in clinical research adulthood: national birth cohort study. BMJ. 2001;
protocols may be called for. 322(7295):1145.
16. Johnson KM, Bradley KA, Bush K, Gardella C,
Dobie DJ, Laya MB. Frequency of mastalgia among
women veterans. Association with psychiatric con-
References ditions and unexplained pain syndromes. J Gen
Intern Med. 2006;21 Suppl 3:S705.
1. Klein DF. False suffocation alarms, spontaneous 17. Benedikt M. Ueber Platzschwindel. Allg Wien Med
panics, and related conditions. An integrative Zig. 1870;15:48890.
hypothesis. Arch Gen Psychiatry. 1993;50:30617. 18. Frommberger UH, Tettenborn B, Buller R, Benkert
2. Preter M, Klein DF. Panic, suffocation false alarms, O. Panic disorder in patients with dizziness. Arch
separation anxiety and endogenous opioids. Prog Intern Med Mar. 1994;154(5):5901.
Neuropsychopharmacol Biol Psychiatry. 2008;32(3): 19. Staab JP. Chronic dizziness: the interface between
60312. psychiatry and neuro-otology. Curr Opin Neurol.
3. Klein DF, Fink M. Psychiatric reaction patterns to 2006;19(1):418.
imipramine. Am J Psychiatry. 1962;119:4328. 20. Preter M, Bursztajn HJ. Crisis and opportunitythe
4. Faravelli C, Pallanti S. Recent life events and panic DSM-V and its neurology quandary. Asian J Psy-
disorder. Am J Psychiatry. 1989;146:6226. chiatry. 2009;2(4):143.
7 Lifelong Opioidergic Vulnerability Through Early Life Separation 147

21. Park L, Hinton D. Dizziness and panic in China: 38. Craske MG, Poulton R, Tsao JC, Plotkin D. Paths to
associated sensations of zang fu organ disequilib- panic disorder/agoraphobia: an exploratory analysis
rium. Cult Med Psychiatry. 2002;26(2):22557. from age 3 to 21 in an unselected birth cohort.
22. Sareen J, Jacobi F, Cox BJ, Belik SL, Clara I, Stein J Am Acad Child Adolesc Psychiatry. 2001;40:
MB. Disability and poor quality of life associated 55663.
with comorbid anxiety disorders and physical 39. Hasler G, Gergen PJ, Kleinbaum DG, Ajdacic V,
conditions. Arch Intern Med. 2006;166(19): Gamma A, Eich D, et al. Asthma and panic in young
210916. adults: a 20-year prospective community study. Am
23. Pine DS et al. Differential carbon dioxide sensitivity J Respir Crit Care Med. 2005;171:122430.
in childhood anxiety disorders and nonill compari- 40. Karajgi B, Rifkin A, Doddi S, Kolli R. The preva-
son group. Arch Gen Psychiatry. 2000;57:9607. lence of anxiety disorders in patients with chronic
24. Pine DS et al. Response to 5 % carbon dioxide in obstructive pulmonary disease. Am J Psychiatry.
children and adolescents: relationship to panic disor- 1990;147:2001.
der in parents and anxiety disorders in subjects. Arch 41. Perna G, Bertani A, Politi E, Colombo G, Bellodi L.
Gen Psychiatry. 2005;62:7380. Asthma and panic attacks. Biol Psychiatry.
25. Preter M, Klein DF. Panic disorder and the suffoca- 1997;42:62530.
tion false alarm theory: current state of knowledge 42. Verburg K, Griez E, Meijer J, Pols H. Respiratory
and further implications for neurobiologic theory disorders as a possible predisposing factor for panic
testing. In: Bellodi L, Perna G, editors. The panic disorder. J Affect Disord. 1995;33:12934.
respiration connection. Milan: MDM Medical 43. Edlund MJ, McNamara ME, Millman RP. Sleep
Media; 1998. apnea and panic attacks. Compr Psychiatry. 1991;32:
26. Beck JG, Ohtake PJ, Shipherd JC. Exaggerated anxi- 1302.
ety is not unique to CO2 in panic disorder: a com- 44. Sietsema KE, Simon JI, Wasserman K. Pulmonary
parison of hypercapnic and hypoxic challenges. hypertension presenting as a panic disorder. Chest.
J Abnorm Psychol. 1999;108:47382. 1987;91:9102.
27. Beck JG, Shipherd JC, Ohtake P. Do panic symptom 45. Goodwin RD, Messineo K, Bregante A, Hoven CW,
profiles influence response to a hypoxic challenge in Kairam R. Prevalence of probable mental disorders
patients with panic disorder? A preliminary report. among pediatric asthma patients in an inner-city
Psychosom Med. 2000;62:67883. clinic. J Asthma. 2005;42:6437.
28. Goodwin RD, Eaton WW. Asthma and the risk of 46. Shavitt RG, Gentil V, Mandetta R. The association
panic attacks among adults in the community. of panic/agoraphobia and asthma. Contributing fac-
Psychol Med. 2003;33:87985. tors and clinical implications. Gen Hosp Psychiatry.
29. Goodwin RD, Fergusson DM, Horwood LJ. Asthma 1992;14:4203.
and depressive and anxiety disorders among young 47. Kessler RC, Chiu WT, Jin R, Ruscio AM, Shear K,
persons in the community. Psychol Med. 2004;34: Walters EE. The epidemiology of panic attacks,
146574. panic disorder, and agoraphobia in the National
30. Katon WJ, Richardson L, Lozano P, McCauley E. Comorbidity Survey Replication. Arch Gen
The relationship of asthma and anxiety disorders. Psychiatry. 2006;63:41524.
Psychosom Med. 2004;66:34955. 48. Weissman MM. The epidemiology of anxiety disor-
31. Klein DF. Asthma and psychiatric illness. JAMA. ders: rates, risks and familial patterns. J Psychiatr
2001;285:8812. Res. 1988;22 Suppl 1:99114.
32. Nascimento I et al. Psychiatric disorders in asth- 49. Baron C, Lamarre A, Veilleux P, Ducharme G, Spier
matic outpatients. Psychiatry Res. 2002;110:7380. S, Lapierre JG. Psychomaintenance of childhood
33. Roy-Byrne PP, Craske MG, Stein MB. Panic disor- asthma: a study of 34 children. J Asthma. 1986;
der. Lancet. 2006;368:102332. 23:6979.
34. Valena AM, Falco R, Freire RC, Nascimento I, 50. Brooks CM, Richards Jr JM, Bailey WC, Martin B,
Nascentes R, Zin WA, et al. The relationship Windsor RA, Soong SJ. Subjective symptomatology
between the severity of asthma and comorbidities of asthma in an outpatient population. Psychosom
with anxiety and depressive disorders. Rev Bras Med. 1989;51:1028.
Psiquiatr. 2006;28:2068. 51. Jurenec GS. Identification of subgroups of childhood
35. Wingate BJ, Hansen-Flaschen J. Anxiety and asthmatics: a review. J Asthma. 1988;25:1525.
depression in advanced lung disease. Clin Chest 52. Mayer DJ, Wolfle TL, Akil H, Carder B, Liebeskind
Med. 1997;18:495505. JC. Analgesia from electrical stimulation in the
36. Yellowlees PM, Haynes S, Potts N, Ruffin RE. brainstem of the rat. Science. 1971;174:13514.
Psychiatric morbidity in patients with life- 53. Stefano GB, Scharrer B, Smith EM, Hughes Jr TK,
threatening asthma: initial report of a controlled Magazine HI, Bilfinger TV, et al. Opioid and opiate
study. Med J Aust. 1988;149:2469. immunoregulatory processes. Crit Rev Immunol.
37. Yellowlees PM, Kalucy RS. Psychobiological 1996;16:10944.
aspects of asthma and the consequent research 54. Santiago TV, Edelman NH. Opioids and breathing.
implications. Chest. 1990;97:62834. J Appl Physiol. 1985;59:167585.
148 M. Preter

55. Mayfield KP, D'Alecy LG. Role of endogenous opi- tion in a group of patients with panic disorder. Am
oid peptides in the acute adaptation to hypoxia. J Psychiatry. 1984;141:85761.
Brain Res. 1992;582:22631. 74. Liebowitz MR, Fyer AJ, Gorman JM, Dillon D,
56. Baydur A. Nebulized morphine: a convenient and Appleby IL, Levy G, et al. Lactate provocation of
safe alternative to dyspnea relief? Chest. 2004; panic attacks: I. Clinical and behavioral findings.
125:3635. Arch Gen Psychiatry. 1984;41:76470.
57. Mahler DA. Understanding mechanisms and docu- 75. Papp LA, Klein DF, Gorman JM. Carbon dioxide
menting plausibility of palliative interventions for hypersensitivity, hyperventilation, and panic disor-
dyspnea. Curr Opin Support Palliat Care. 2011; der. Am J Psychiatry. 1993;150:114957.
5(2):716. 76. Esquivel G, Schruers KR, Maddock RJ, Colasanti A,
58. Zebraski SE, Kochenash SM, Raffa RB. Lung opioid Griez EJ. Acids in the brain: a factor in panic?
receptors: pharmacology and possible target for neb- J Psychopharmacol. 2010;24(5):63947.
ulized morphine in dyspnea. Life Sci. 2000; 77. Griez EJ, Colasanti A, van Diest R, Salamon E,
66:222131. Schruers K. Carbon dioxide inhalation induces dose-
59. Bruera E, Sala R, Spruyt O, Palmer JL, Zhang T, dependent and age-related negative affectivity.
Willey J. Nebulized versus subcutaneous morphine PLoS One. 2007;2(10):e987.
for patients with cancer dyspnea: a preliminary 78. Leibold NK, Viechtbauer W, Goossens L, De Cort
study. J Pain Symptom Manage 2005;29:6138. K, Griez EJ, Myin-Germeys I, et al. Carbon dioxide
60. Panksepp J. Affective neuroscience: the foundations inhalation as a human experimental model of panic:
of human and animal emotions. New York: Oxford the relationship between emotions and cardiovascu-
University Press; 1998. lar physiology. Biol Psychol. 2013;94(2):33140.
61. West MJ, King AP. Settling nature and nurture into an 79. Goetz RR, Klein DF, Gully D, Kahn J, Liebowitz M,
ontogenetic niche. Dev Psychobiol 1987;20:54962. Fyer A, et al. Panic attacks during placebo proce-
62. Panksepp J, Newman JD, Insel TR. Critical concep- dures in the laboratory: physiology and symptom-
tual issues in the analysis of separation-distress sys- atology. Arch Gen Psychiatry. 1993;50:2805.
tems of the brain. In: Strongman KT, editor. 80. Martinez JM, Papp LA, Coplan JD, Anderson DE,
International review of studies on emotion, vol. 2. Mueller CM, Klein DF, et al. Ambulatory monitor-
New York: Wiley; 1992. ing of respiration in anxiety. Anxiety. 1996;2:
63. Kalin NH, Shelton SE, Barksdale CM. Separation 296302.
distress in infant rhesus monkeys: effects of diaze- 81. Liebowitz MR, Gorman JM, Fyer AJ, Levitt M,
pam and Ro 15-1788. Brain Res. 1987;408:1928. Dillon D, Levy G, et al. Lactate provocation of panic
64. Scott JP. Effects of psychotropic drugs in separation attacks. II. Biochemical and physiological findings.
distress in dogs. Amsterdam: Proc IX Congress Arch Gen Psychiatry. 1985;42:70919.
ECNP Exc Med; 1974. 82. Sinha SS, Goetz RR, Klein DF. Physiological and
65. Jurgens U, Ploog D. Cerebral representation of behavioral effects of naloxone and lactate in normal
vocalization in the squirrel monkey. Exp Brain Res. volunteers with relevance to the pathophysiology of
1970;10:53254. panic disorder. Psychiatry Res. 2007;149:30914.
66. Ploog D. Neurobiology of primate audio-vocal 83. Esquivel G, Fernndez-Torre O, Schruers KR,
behavior. Brain Res. 1981;228:3561. Wijnhoven LL, Griez EJ. The effects of opioid
67. Wise SP, Herkenham M. Opiate receptor distribution receptor blockade on experimental panic provoca-
in the cerebral cortex of the Rhesus monkey. Science. tion with CO2. J Psychopharmacol. 2009;23(8):
1982;218:3879. 9758.
68. Hofer MA, Shair H. Ultrasonic vocalization during 84. Liebowitz MR, Gorman JM, Fyer AJ, Dillon DJ,
social interaction and isolation in 2-weeek-old rats. Klein DF. Effects of naloxone on patients with panic
Dev Psychobiol. 1978;11:495504. attacks. Am J Psychiatry. 1984;141(8):9957.
69. Kalin NH, Shelton SE, Barksdale CM. Opiate modu- 85. Preter M, Lee SH, Petkova E, Vannucci M, Kim S,
lation of separation-induced distress in non-human Klein DF. Controlled cross-over study in normal
primates. Brain Res. 1988;440:28592. subjects of naloxone-preceding-lactate infusions;
70. Kehoe P, Blass EM. Opioid-mediation of separation respiratory and subjective responses: relationship to
distress in 10-day-old rats: reversal of stress with endogenous opioid system, suffocation false alarm
maternal stimuli. Dev Psychobiol. 1986;19: theory and childhood parental loss. Psychol Med.
38598. 2011;41(2):38593.
71. Panksepp J, Herman B, Conner R, Bishop P, Scott 86. Liu Z, Li X, Ge X. Left too early: the effects of age
JP. The biology of social attachments: opiates alle- at separation from parents on Chinese rural chil-
viate separation distress. Biol Psychiatry. 1978;13: drens symptoms of anxiety and depression. Am
60718. J Public Health. 2009;99:204954.
72. Kalin NH, Shelton SE. Effects of clonidine and pro- 87. Bandelow B, Spath C, Tichauer GA, Broocks A,
pranolol on separation-induced distress in infant rhe- Hajak G, Ruther E. Early traumatic life events,
sus monkeys. Brain Res. 1988;470:28995. parental attitudes, family history, and birth risk fac-
73. Gorman JM, Askanazi J, Liebowitz MR, Fyer AJ, tors in patients with panic disorder. Compr
Stein J, Kinney JM, et al. Response to hyperventila- Psychiatry. 2002;43:26978.
7 Lifelong Opioidergic Vulnerability Through Early Life Separation 149

88. Crawford TN, Cohen PR, Chen H, Anglin DM, 100. Sluka KA, Deacon M, Stibal A, Strissel S, Terpstra
Ehrensaft M. Early maternal separation and the tra- A. Spinal blockade of opioid receptors prevents the
jectory of borderline personality disorder symptoms. analgesia produced by TENS in arthritic rats.
Dev Psychopathol. 2009;21(3):101330. J Pharmacol Exp Ther. 1999;289:8406.
89. Kendler KS, Neale MC, Kessler RC, Heath AC, 101. Moreira FA, Gobira PH, Viana TG, Vicente MA,
Eaves LJ. Childhood parental loss and adult psycho- Zangrossi H, Graeff FG. Modeling panic disorder in
pathology in women. A twin study perspective. Arch rodents. Cell Tissue Res. 2013;354:11925.
Gen Psychiatry. 1992;49:10916. 102. Graeff FG. New perspective on the pathophysiology
90. Shonkoff JP, Boyce WT, McEwen BS. Neuroscience, of panic: merging serotonin and opioids in the peri-
molecular biology, and the childhood roots of health aqueductal gray. Braz J Med Biol Res. 2012;
disparities: building a new framework for health 45(4):36675.
promotion and disease prevention. JAMA. 2009; 103. Schenberg LC, Schimitel FG, Armini Rde S,
301:22529. Bernabe CS, Rosa CA, Tufik S, et al. Translational
91. Brown DW, Anda RF, Tiemeier H, Felitti VJ, approach to studying panic disorder in rats: hits and
Edwards VJ, Croft JB, et al. Adverse childhood misses. Neurosci Biobehav Rev. 2014;46(Pt 3):
experiences and the risk of premature mortality. Am 47296.
J Prev Med. 2009;37:38996. 104. Schimitel FG, De Almeida GM, Pitol DN, Armini
92. Juang KD, Wang SJ, Fuh JL, Lu SR, Chen RS, Tufik S, Schenberg LC. Evidence of a suffoca-
YS. Association between adolescent chronic daily tion alarm system within the periaqueductal gray
headache and childhood adversity: a community- matter of the rat. Neuroscience. 2012;200:5973.
based study. Cephalalgia. 2004;24:549. 105. Armini R, Bernabe CS, Rosa CA, Siller CA,
93. Kopec JA, Sayre EC. Stressful experiences in child- Schimitel FG, Tufik S, et al. In a rat model of panic,
hood and chronic back pain in the general popula- corticotropin responses to dorsal periaqueductal
tion. Clin J Pain. 2005;21:47883. gray stimulation depend on physical exertion.
94. Harter MC, Conway KP, Merikangas KR. Asso- Psychoneuroendocrinology. 2015;53:13647.
ciations between anxiety disorders and physical ill- 106. Schenberg LC, Dos Reis AM, Ferreira Pvoa RM,
ness. Eur Arch Psychiatry Clin Neurosci. 2003; Tufik S, Silva SR. A panic attack-like unusual stress
253:31320. reaction. Horm Behav. 2008;54(5):58491.
95. Jones GT, Power C, Macfarlane GJ. Adverse events 107. Stein G. Panic disorder: the psychobiology of
in childhood and chronic widespread pain in adult external treat and introceptive distress. CNS
life: results from the 1958 British Birth Cohort Spectrums. 2008;13(1):2630.
Study. Pain. 2009;143:926. 108. Gorman JM, Kent JM, Sullivan GM, Coplan
96. Katerndahl DA. Chest pain and its importance in JD. Neuroanatomical hypothesis of panic disorder,
patients with panic disorder: an updated literature revised. Am J Psychiatry. 2000;157(4):493505.
review. Prim Care Companion J Clin Psychiatry. 109. Feinstein JS, Buzza C, Hurlemann R, Follmer RL,
2008;10:37683. Dahdaleh NS, Coryell WH, et al. Fear and panic in
97. Battaglia M, Pesenti-Gritti P, Medland SE, Ogliari humans with bilateral amygdala damage. Nat
A, Tambs K, Spatola CA. A genetically informed Neurosci. 2013;16(3):2702.
study of the association between childhood separa- 110. Preter M, Klein DF. Lifelong opioidergic vulnera-
tion anxiety, sensitivity to CO(2), panic disorder, and bility through early life separation: a recent exten-
the effect of childhood parental loss. Arch Gen sion of the false suffocation alarm theory of panic
Psychiatry. 2009;66:6471. disorder. Neurosci Biobehav Rev. 2014;46(Pt 3):
98. Gerra G, Leonardi C, DAmore A, Strepparola G, 34551.
Fagetti R, Assi C, et al. Buprenorphine treatment 111. Eisenberger NI, Lieberman MD. Why rejection hurts:
outcome in dually diagnosed heroin dependent a common neural alarm system for physical and
patients: a retrospective study. Prog Neuro- social pain. Trends Cogn Sci. 2004;8(7):294300.
psychopharmacol Biol Psychiatry. 2006;30:26572. 112. Hsu DT, Sanford BJ, Meyers KK, Love TM, Hazlett
99. Wallen MC, Lorman WJ, Gosciniak JL. Combined KE, Walker SJ, et al. It still hurts: altered endoge-
buprenorphine and chlonidine for short-term opiate nous opioid activity in the brain during social rejec-
detoxification: patient perspectives. J Addict Dis. tion and acceptance in major depressive disorder.
2006;25:2331. Mol Psychiatry. 2015;20(2):193200.
Circadian Rhythm in Panic
Michelle Levitan and Marcelo Papelbaum

Contents Abstract
8.1 Introduction 151 The relationship between panic disorder and
sleep problems has been studied, and possible
8.2 Sleep Cycles 152
explanations for this association are discussed
8.3 Sleep Quality Studies 153 in this chapter. So far, the results of polysom-
8.4 Mechanism of Association Between Panic nographic studies in PD patients are inconclu-
and Sleep Disorders 153 sive, but seem to suggest that patients with PD
8.4.1 Obstructive Sleep Apnea Syndrome 153 have impaired initiation and maintenance of
8.4.2 Nocturnal Panic Attacks 153
8.4.3 Depression 154
sleep. The presence of nocturnal panic attacks
8.4.4 The Cortisol and Panic Disorder 155 induce an intense fear of sleep, leading to
8.4.5 Anxiety Sensitivity 155 anticipatory anxiety and sleep avoidance,
8.5 Treatment 155 resulting in secondary insomnia and facilitat-
8.5.1 Pharmacological Treatment 156 ing the development of new panic attacks.
8.5.2 Psychological Treatment 157 Other hypotheses, as the co-occurrence of
8.6 Conclusions 157 depression, cortisol levels and anxiety sensi-
References 158
tivity are also raised as mechanisms related to
sleep problems in PD patients. Treatments are
available, as the cognitive behavioral treat-
ment and some novel treatments that may
improve panic attacks as well as insomnia.

Panic disorder Sleep Nocturnal panic
attacks Sleep treatment
M. Levitan (*)
Laboratory of Panic and Respiration, Institute of
Psychiatry, Federal University of Rio de Janeiro,
Rio de Janeiro, Brazil
e-mail: milevitan@gmail.com 8.1 Introduction
M. Papelbaum
State Institute of Diabetes and Endocrinology The relationship between psychiatric disorders
of Rio de Janeiro, Rio de Janeiro, Brazil and sleep problems has been widely studied.
e-mail: marcelo@papelbaum.com Indeed, sleep complains integrate such an

Springer International Publishing Switzerland 2016 151

A.E. Nardi, R.C.R. Freire (eds.), Panic Disorder, DOI 10.1007/978-3-319-12538-1_8
152 M. Levitan and M. Papelbaum

important feature of psychiatric disorders, being, 8.2 Sleep Cycles

for instance, part of the diagnostic criteria for
depression and generalized anxiety disorder. For a better understanding, sleep architecture
Specifically for panic disorder (PD), the presence stages are exposed according to the Committee
of sleep disturbance is very common, with on Sleep Medicine and Research [5] and differ-
6877 % of the patients complaining about trou- ential diagnosis are made between sleep prob-
ble with sleep [1]. However, there is a paucity of lems (Table 8.1). Overall, sleep is divided in two
research regarding the comorbidity between this main types: rapid eye movement (REM) and
disorder and sleep disturbance. non-REM (NREM), the latter divided in four
To evaluate sleep quality, subjective reports stages. A sleep episode begins in the first stage of
and polysomnographic data are often used. The NREM sleep through the four stage, and ends up
clinical interview and scales are important to at REM sleep. Individuals do not remain in REM
identify sleep complaints and accordingly, refer sleep the rest of the night but, rather, cycle
to a more accurate exam, as a polysomnography, between stages of NREM and REM throughout
that provides sleep cycles and help identifying the night.
any alteration. However, because PD is not Stage 1 is easily interrupted by a disruptive
directly associated to sleep problems, health pro- noise. Brain activity on the electroencephalo-
fessionals tend to spend less time evaluating its gram in stage 1 transitions from wakefulness
occurrence and impact. (marked by rhythmic alpha waves) to low-
Possible explanations for the association voltage, mixed-frequency waves, in which alpha
between PD and sleep difficulties are discussed waves are associated with a wakeful relaxation
in this chapter, focusing mostly in insomnia, the state. It constitutes 25 % of the total sleep.
most common sleep disorder in PD. A potential An individual in stage 2 sleep requires more
link between both disorders relies on the occur- intense stimuli than in stage 1 to awaken. In this
rence of nocturnal panic attacks (NPA), that hap- stage, brain activity on an EEG shows relatively
pen repeatedly in 2045 % of PD patients [2, 3] low-voltage, mixed-frequency activity character-
and appear to predispose patients to be fearful ized by the presence of sleep spindles and
and to stay awake to avoid their recurrence. Other K-complexes. It lasts 4555 % of total sleep.
association hypotheses are discussed, as the Sleep stages 3 and 4 are collectively referred
physiological arousal, increased base levels of to as slow-wave sleep, in which the EEG shows
cortisol [4] or high anxiety sensitivity. At last, increased high-voltage, slow-wave activity. It
the comorbid depression [3] as an intermediate lasts 1015 % of sleep. The REM stage is defined
between sleep complaints and PD is addressed. by the presence of desynchronized (low-voltage,

Table 8.1 Differential diagnosis between nocturnal panic and sleep complaints Institute of Medicine (US) Committee
on Sleep Medicine and Research
Disorder Definition Cycle of sleep
Nocturnal panic attack Waking up in a state of horror Late stage II or early
stage III sleep
Insomnia Trouble in falling or staying asleep Before sleep begins
Nightmares Distressing dream that forces Occur largely in REM
awakenings; the person usually sleep, after hours of sleep
remembers the episode
Night terrors Cause feelings of terror; for the Occur mostly in stage IV
sleeper is hard to awaken
Obstructive sleep apnea Upper airway obstruction occur Occurs in stage I and II
repeatedly during sleep and REM sleep
Sleep Disorders and Sleep Deprivation: An Unmet Public Health Problem. Washington (DC): National Academies
Press (US); 2006. 3, Extent and Health Consequences of Chronic Sleep Loss and Sleep Disorders [5]
8 Circadian Rhythm in Panic Disorder 153

mixed-frequency) brain wave activity and bursts 8.4 Mechanism of Association

of rapid eye movements, in which dreaming is Between Panic and Sleep
most often associated. Disorders

8.4.1 Obstructive Sleep Apnea

8.3 Sleep Quality Studies Syndrome

Irregular cycling sleep is associated with sleep Studies investigating the prevalence of both dis-
disorders. Subjective reports of PD patients orders in clinical samples are scarce, however
evidenced that 68 % of the sample described dif- attention must be drawn to this comorbidity
ficulties in falling asleep and 77 % reported because both conditions are relative common,
disturbed sleep [6]. Higher percentages of sleep with sleep apnea prevalence ranging between
complaints in PD patients compared to healthy 24 % on population-based studies [16]. In addi-
subjects were found in other study, especially tion, authors suggest that the presence of sleep
middle night insomnia (67 % vs 23 %) and late apnea could be associated to higher prevalence of
night insomnia (67 % vs 31 %) [7]. psychiatric comorbid conditions [17].
So far, the results of polysomnographic stud- Patients with PD have greater respiratory
ies in PD patients are inconclusive, but seem to variability than comparison subjects [18]. This
suggest that patients with PD have impaired ini- trait marker may predispose patients to CO2-
tiation and maintenance of sleep, characterized induced panic attack [19]. In addition, respira-
by increased sleep latency and increased time tory events during sleep, in the presence o
awake after sleep onset [3, 8]. Some authors also comorbid OSA, may trigger panic attacks. In this
observed a decrease in sleep efficiency, total way, treatment results indicate that the vast
sleep time and amount of non-REM sleep in majority of PD with OSA patients experienced a
stage 4 compared to healthy controls, whereas diminution or disappearance of panic attacks
others identified an increased percentage of non- with CPAP, besides a decrease in the use of
REM sleep stage 1 in PD patients [3, 9, 10]. alprazolam, what might suggest that CPAP works
Regarding patients with repeated NPAs, more as a treatment for diurnal and NPAs alone too
severe subjective sleep complaints were reported [17]. These data seem to be in accordance with
when compared to those with only diurnal panic an influential etiological hypothesis for PD, the
attacks, but no differences were found on electro- Kleins suffocation alarm theory [20], in which
encephalogram indices or polysomnographic panic attacks would be a result of inappropriate
parameters [11, 12]. activation of an alarm system by the brain that
Commentary must be made on the possible signals lack of air [18, 20].
co-occurrence of PD with another sleep disorder:
the obstructive sleep apnea (OSA). The OSA
may cause NPAs symptoms [13], and treating 8.4.2 Nocturnal Panic Attacks
one condition with continuous positive airway
pressure (CPAP), could impact on the prognosis NPAs refer to waking up in a state of panic,
of the other one [14, 15]. Thus, in addition defined as an abrupt and discrete period of intense
to electroencephalogram alterations, respiratory fear and discomfort accompanied by cognitive
parameters might be able to guide the accurate and physical symptoms of arousal [21]. They are
diagnosis and treatment of sleep problems, and NREM-related events, usually emerging from
impact the clinical course of panic symptoms. late stage 2 or early stage 3 of sleep [3] and differ
154 M. Levitan and M. Papelbaum

Fig. 8.1 Nocturnal

panic attacks cycle Nocturnal Panic Attacks Fear of Sleep

Sleep Avoidance

from sleep problems regarding clinical features 8.4.3 Depression

and stage of occurrence.
The association between PD and sleep com- Survey data on sleep and psychiatric disorders in
plaints is often attributed to NPAs, that may occur 14,915 subjects using questionnaires evidenced
in 3369 % of PD patients [7]. Due to the noctur- that insomnia is the most frequent sleep disorder in
nal occurrence, they induce an intense fear of psychiatric disorders [27]. However, the natural
sleep, leading to anticipatory anxiety and sleep course of it incidence tend to differ according to the
avoidance, resulting in secondary insomnia and type of psychopathology. Specifically, insomnia
facilitating the development of new NPAs [22] tended to precede the onset of mood disorders in
(Fig. 8.1). This perspective has its foundation in 40 % of cases, whereas in the case of anxiety disor-
the cognitive behavioral model, in which learned ders, in 80 % of the time it appeared either at the
fearfulness of bodily sensations leads to reactiv- same time or following the onset of anxious symp-
ity changes in bodily state. Indeed, slight changes toms [27]. Therefore, sleep disorders and psychiat-
in heartbeat and sweat have been observed pre- ric disorders seem to interact in multiple ways [27].
ceding NPAs [23]. Previous research has evidenced association
In a polysomnographic study, Stein et al. [3] between sleep complaints and comorbid depres-
found that non-depressed PD patients presented sive symptomatology in patients with PD. There
regular sleep stages indistinguishable from the are some possible explanations for this relation-
healthy group. An interesting finding of this study ship. First, the occurrence of sleep problems
was that healthy individuals presented brief might be associated directly to the comorbid
arousals ten times every hour. The authors point depression, commonly coexisted with PD. In fact,
out that for PD patients, each of these arousals depression occurs in 5065 % of PD patients.
represents an opportunity to become aware of his A second possibility is that comorbid depression
sensations and end up interpreting them as dan- may be more common in the presence of a severe
gerous, possibly leading to a NPA [24]. form of PD that features with sleep problems [28].
Some authors suggest that NPAs would be a Lastly, insomnia itself, as a part of PD symptoms
more severe subtype of PD. This hypothesis is might be a risk factor for depression comorbidity.
based in studies that found that patients who had Indeed, longitudinal studies with individuals with
NPAs reported more fears and a more severe his- insomnia demonstrated an increase in the likeli-
tory of psychopathology than those who only had hood of developing major depression, especially
daytime panic attacks [25]. Other studies also when the sleep disorder is persistent [29, 30].
found and a greater severity of symptoms, com- Moreover, insomnia might be an independent risk
pared to diurnal panic attacks, especially chest factor for the development of PD. In this way,
pain, nausea and discomfort [26]. Regarding results of the epidemiological catchment area
sleep quality, patients with NPAs did not differ revealed that individuals with insomnia were five
significantly from those with diurnal panic times more likely to have a panic attack and being
attacks, except a tendency to report a more severe at risk of developing PD than individuals without
self-rating of past and current sleep difficulties insomnia. For these subjects, insomnia persisted
with sleep onset [24]. after the remission of the mental disorder [31].
8 Circadian Rhythm in Panic Disorder 155

8.4.4 The Cortisol and Panic explanation for sleep complaints and PD would
Disorder rely on a selective attention to fear of anxiety and
physical sensations that would occur at night
The hypothalamic-pituitary-adrenal (HPA) axis while trying to fall asleep [39]. Studies show an
is frequently referred as disturbed in PD [32]. association between AS and sleep disturbance
HPA regulates cortisol secretion and is under the and suggest that AS may contribute to initial
influence of various factors such as sleep, that insomnia in PD [39].
exerts an inhibitory effect on cortisol secretion [34]. This hypothesis lead authors the suggestion to
During nocturnal awakenings there is a transient include insomnia-targeted components in the
elevation of cortisol levels, followed by tempo- cognitive behavioral therapy (CBT) interven-
rary inhibition of cortisol secretion [33]. tions, which are effective to reduce AS in PD
Studies have pointed out the importance of patients [40] and may interfere with a possible
stress hormone regulation in anxiety disorders development of sleep problems. Besides CBT
[32, 33]. Patients with PD and controls partici- techniques as interoceptive exposure, pharmaco-
pated in a study in which salivary and urinary therapy seems to decrease AS levels well [41].
cortisol levels were determined in 2-h spans dur-
ing 3 consecutive days. The more severe group of
clinical patients presented salivary cortisol, noc- 8.5 Treatment
turnal and urinary cortisol levels significantly
higher than the controls in the whole group of PD Mellman et al. points out that there is overlap
patients. Other studies found elevated baseline between interventions that target sleep distur-
plasma levels of cortisol in PD patients [34, 35]. bances and those that are used in treating anxiety
The results suggest that cortisol elevations disorders [2]. Overlapping approaches include
are more pronounced during the night, mainly in medications and cognitive behavioral strategies
severe ill PD patients [32]. that target worry, tension, and maladaptive cogni-
Despite the results, the authors still cannot tions. Optimal sequencing or integration of treat-
affirm whether increased cortisol levels in PD ments targeting anxiety and sleep disturbance
patients only reflect the chronic stress due to were not fully investigated.
panic attacks and anticipatory anxiety or are the Some authors suggest that once treatment is
expression of a neurobiological defect involved initiated at the earliest phase, sleep problems in
in the pathogenesis of panic attacks [32]. anxious patients are considered secondary symp-
toms that improve with the other symptomatol-
ogy. Clinical experience shows that most patients
8.4.5 Anxiety Sensitivity with comorbid anxiety and insomnia seek treat-
ment months to years after the initial presentation
Anxiety sensitivity (AS) is a dispositional feature of symptoms. At this point, it is necessary to treat
associated to a excessive fear of anxiety-related both symptoms [42].
sensations based on beliefs that these sensations However, others state that in this comorbid
are harmful [36]. The AS is considered a main treatment, insomnia should be treated concur-
factor in the development and maintenance of rently with, but independently of the anxiety dis-
PD. Studies indicate that the association between order per se. Many authors criticize the idea that
AS and PD is largely attributable to the somatic the clinician should wait to evaluate whether the
aspects of AS [37], considered not the fear of the insomnia improves as a consequence of the anxi-
general anxiety, but the fear of physical sensa- ety disorder treatment. Clinical experience has
tions that are associated to the panic related- shown that without targeted insomnia treatment,
psychology [38]. Based on these data, a possible insomnia frequently persists [43].
156 M. Levitan and M. Papelbaum

8.5.1 Pharmacological Treatment side effects might be helpful in patients with

anxiety who also report significant comorbid
Specific treatments for sleep disturbance are insomnia [46].
available; however studies with PD and these Mirtazapine is a noradrenergic and specific
conditions are scarce and hardly ever address this serotoninergic antidepressant with known seda-
complaint. In fact, most studies related to treat- tive effect. It has been studied in some uncon-
ment of anxiety disorders with comorbid insom- trolled and head-to-head studies (against active
nia were conducted with generalized anxiety treatments), evidencing efficacy in the remission
disorder and post-traumatic stress disorder. of panic symptoms [47]. However, specific sleep
Despite well-established pharmacological parameters were not evaluated. Nevertheless,
treatments for panic symptoms, persistence of two case reports showed remission of insomnia
sleep complaints is common and, therefore, dis- with mirtazapine added onto selective serotonin
continuation of medication for comorbid chronic reuptake inhibitors treatment [47, 48]. In this
insomnia proves to be a difficult task. Benzo- way, it is worth mentioning the use of antidepres-
diazepines (BDZ) represent an advance by offer- sant trazodone (serotonin antagonist and reup-
ing safer alternatives that demonstrated long-term take inhibitor) as a sleep-inducing medication,
benefit, but they also showed problems of occa- especially at lower doses [49]. Although, only
sional dependence and residual side effects. two past studies showed limited or negative ben-
Nevertheless, attempts to slow tapering chronic efit on PD, the use of trazodone as an add-on
use of BDZ in asymptomatic PD patients should therapy to mitigate sleep problems must be kept
be made. Nardi et al. [44] showed successful dis- in mind, especially when BDZ use should be
continuation of 3-years of more use of clonaze- avoided [50, 51].
pam in patients with PD over the course of Another possibility is agomelatine, an antide-
4 months with mild and transient withdrawal pressant with the novel mechanism of being a
symptoms. However, although panic attacks selective melatonergic MT1/MT2 receptor ago-
might be remitted, the re-occurrence of sleep nist with serotonin 5-HT2c and 5-HT2b receptor
problems on the follow-up is common. In a study antagonist activities [52]. Although still not fully
that intervened in the long-term use of benzodi- elucidated, it seems that its melatonergic ago-
azepines for the treatment of chronic insomnia, nism and its 5-HT2c antagonism could act syner-
efforts to discontinue medication resulted in a gistically in the restoration of disrupted circadian
high rate of return to the hypnotic over the rhythms [52]. So far, few studies evaluated the
2 years of follow-up [43]. efficacy of agomelatine in treating anxiety disor-
In the case of antidepressants, the improve- ders and sleep symptoms, mostly in generalized
ments of PD symptoms, commonly, do not reflect anxiety disorder. Overall, the use of agomelatine
into benefits of sleep complaints. Small but sig- was associated with a larger improvement in sub-
nificant change of the subjective quality of sleep jective sleep symptoms, including getting of to
might be noted, objective sleep parameters such sleep, quality of sleep and sleep awakening [54].
as total sleep time and sleep onset latency remain Independently of the reduction of the anxiety
unchanged [45]. Indeed, Todder et al. [46] found symptoms. Fornaro et al. [55] discussed a possi-
improvement of subjectively quality of sleep and ble benefit of agomelatine in diminishing the
persistence of objective parameter alterations in need for benzodiazepines, which are highly pre-
patients with PD treated with escitalopram. Also, scribed in patients with PD [53]. This effect of
the study did not found a correlation between agomelatine could be a protective factor in reduc-
clinical improvement and the changes in the ing the risk of benzodiazepines abuse or depen-
sleep quality, suggesting that the mechanism dence, with the benefit of improving sleep, being
associated to improvement of the sleep might be an off-label option in the treatment of PD, espe-
non-specific related to the reduction of panic cially when the usual serotonin reuptake inhibi-
attacks. In this way, antidepressants with sedative tors are not well-tolerated [56].
8 Circadian Rhythm in Panic Disorder 157

Finally, regarding the comorbidity between schedule and eliminate sleep incompatible
PD and OSA, consideration must be made regard- behaviors in an effort to force the development
ing the pharmacologic treatment of panic symp- of an efficient consolidated sleep pattern. Some
toms. In a PD patient with OSA symptoms such sleep hygiene orientations (sleep rules) are
as sleep apnea episodes, if the clinician does not given: (1) To choose a wake up time; (2) not to
recognize the OSA, he could prescribe benzodi- entertain with activities while in bed; (3) not to
azepines. In this case the benzodiazepine could stay in bed if not to sleep; (4) avoid daytime
suppress the tonus of the upper airway respiratory napping; (5) try not to worry with problems in
muscles, making sleep apnea worse [15, 17]. On bed and (6) go to bed when sleepy.
the other hand, when specific treatment for OSA Step 4: Cognition restructuring: Targets mis-
is initiated, PD patients could have their symp- appraisals of bodily sensations as threatening
toms diminished or disappeared with CPAP, a or dangerous, being the technique most used
treatment that can suppress upper airway collapse in CBT. The therapist helps the patient to
during sleep with a pneumatic splint [17, 57]. identify and correct negative thoughts by
evaluating evidence for and against them.
Patients with insomnia often develop errone-
8.5.2 Psychological Treatment ous thoughts that worsen their difficulties,
such as Ill never get to sleep tonight; Ill
CBT is a short-term, multi-component psycho- be a wreck tomorrow; Ill get sick unless I
therapy, currently considered the treatment of sleep eight hours a night. By the same way
choice for insomnia [58] and PD [59]. When NPAs patients tend to anticipate the conse-
these conditions co-occur, a CBT protocol that quences of their sensations, ultimately believ-
addresses both difficulties: panic attacks and ing that that they will have a heart attack or die
insomnia, is highly recommended [22]. Based on while sleeping.
the most intense difficulty for the patient (panic
attack or insomnia), the treatment may begin
with one target at a time or manage both at the 8.6 Conclusions
same time. This decision will depend on the
patient impairment and the therapist manage- The importance of sleep in PD should not be
ment. The sessions are divided into goals and underestimated. Nearly 80 % of PD patients com-
stages, adjusted to each patient; bellow are some plain about disturbed sleep. In fact, this relation-
concise outlines [5860]. ship is so intimate that a subtype of PD with
nocturnal symptoms is recognized. Indeed, due to
Step 1: Sleep and PD education: Provides the sudden arousal without an obvious trigger,
information about anxiety and sleep. The ther- NPAs are considered a severe form of PD that
apists help the patients to identify dysfunc- leads to anticipatory anxiety and sleep avoidance.
tional beliefs about panic attacks and sleep Several hypotheses attempt to explain the
and correct them. mechanisms related to the association between
Step 2: Breathing and relaxation techniques: PD and sleep difficulties. Regardless, in clinical
Helps the patients to drift off to sleep as well practice, investigation of sleep problems in
as deactivate the hyper stimulation of the patients with PD is of extreme importance.
autonomic nervous system. It is possible that Firstly, it can provide symptom relief to the
NPAs patients will be too sensitive to a relax- patient and, secondly, the alleviation of sleep dis-
ation state similar to the one felt during sleep, turbance can have a positive impact on panic
avoiding this exercise. This should be accom- symptoms. Additional, although not apparently
plished when the patient feels ready. related to the panic symptoms, clinical problems
Step 3: Behavioral changes: This behavioral such should also be investigated, especially
phase use stimulus control and sleep restriction because treatment of one condition could affect
strategies to regularize the patients sleep/wake the other one. Therefore, asking family members
158 M. Levitan and M. Papelbaum

about patients sleep or submitting patient to a parison with normal controls. Acta Psychiatr Scand.
polysomnography could help in the exclusion of 1996;93(3):1914.
11. Uhde TW, Cortese BM, Vedeniapin A. Anxiety and
differential diagnosis. sleep problems: emerging concepts and theoretical
Although pharmacological treatment for PD treatment implications. Curr Psychiatry Rep. 2009;
is effective, remission of panic symptoms is 11(4):26976.
commonly associated with persistence of sleep 12. Landry P, Marchand L, Mainguy N, Marchand A,
Montplaisir J. Electroencephalography during sleep
complaints, which makes it difficult to discon- of patients with nocturnal panic disorder. J Nerv Ment
tinue medication. In this way, the review of dif- Dis. 2002;190(8):55962.
ferent pharmacological strategies seen in this 13. Edlund MJ, McNamara ME, Millman RP. Sleep apnea
chapter, also showed the need of other non- and panic attacks. Compr Psychiatry. 1991;
pharmacological tools, such as the ones pre- 14. Hanly P, Powles P. Hypnotics should never be used in
sented based on CBT, to address issues related to patients with sleep apnea. J Psychosom Res. 1993;37
irrational thoughts, avoidance and sleep hygiene, Suppl 1:5965.
considered main components for the patients 15. Takaesu Y, Inoue Y, Komada Y, Kagimura T, Iimori
M. Effects of nasal continuous positive airway pres-
improvement. sure on panic disorder comorbid with obstructive
sleep apnea syndrome. Sleep Med. 2012;13(2):
References 16. Lee W, Nagubadi S, Kryger MH, Mokhlesi B.
Epidemiology of obstructive sleep apnea: a
population-based perspective. Expert Rev Respir
1. Papadimitriou GN, Linkowski P. Sleep disturbance in Med. 2008;2(3):34964.
anxiety disorders. Int Rev Psychiatry. 2005;17(4): 17. Sharafkhaneh A, Giray N, Richardson P, Young T,
22936. Hirshkowitz M. Sleep. 2005;28(11):140511.
2. Mellman TA. Sleep and anxiety disorders. Psychiatr 18. Martinez JM, Kent JM, Coplan JD, Browne ST, Papp
Clin North Am. 2006;29(4):104758. LA, Sullivan GM, et al. Respiratory variability
3. Stein MB, Enns MW, Kryger MH. Sleep in nonde- in panic disorder. Depress Anxiety. 2001;14(4):
pressed patients with panic disorder: II. Polysom- 232327.
nographic assessment of sleep architecture and sleep 19. Freire RC, Nardi AE. Panic disorder and the respira-
continuity. J Affect Disord. 1993;28(1):16. tory system: clinical subtype and challenge tests. Rev
4. Bonnet MH, Arand DL. Hyperarousal and insomnia: Bras Psiquiatr. 2012;34 Suppl 1:S3241.
state of the science. Sleep Med Rev. 2010;14(1):915. 20. Klein DF. False suffocation alarms, spontaneous pan-
5. Institute of Medicine (US) Committee on Sleep ics, and related conditions: an integrative hypothesis.
Medicine and Research; Colten HR, Altevogt BM, Arch Gen Psychiatry. 1993;50(4):30617.
editors. Sleep disorders and sleep deprivation: an 21. American Psychiatric Association. Diagnostic and
unmet public health problem. Washington, DC: statistical manual of mental disorders, 4th ed. Text
National Academies Press (US); 2006. 3, Extent and Revision (DSM-IV-TR). Arlington, VA: American
Health Consequences of Chronic Sleep Loss and Psychiatric Publishing; 2000.
Sleep Disorders. Available from: http://www.ncbi. 22. Craske MG, Tsao JC. Assessment and treatment of
nlm.nih.gov/books/NBK19961. nocturnal panic attacks. Sleep Med Rev. 2005;9(3):
6. Sheehan DV, Ballenger J, Jacobsen G. Treatment of 17384.
endogenous anxiety with phobic, hysterical and hypo- 23. Roy-Byrne P, Mellman T, Uhde T. Biological findings
chondriacal symptoms. Arch Gen Psychiatry. 1980; in panic disorder: neuroendocrine and sleep- related
37:519. abnormalities. J Anxiety Dis. 1988;2:1729.
7. Mellman TA, Uhde TW. Sleep in panic and gener- 24. Craske MG, Lang AJ, Mystkowski JL, Zucker BG,
alised anxiety disorder. In: Ballanger J, editor. Bystritsky A, Yan-Go F. Does nocturnal panic repre-
Neurobiology of panic disorder. New York, NY: Alan sent a more severe form of panic disorder? J Nerv
R. Liss; 1990. p. 3656. Ment Dis. 2002;190(9):6118.
8. Staner L. Sleep and anxiety disorders. Dialogues Clin 25. Labbate LA, Pollack MH, Otto MW, Langenauer S,
Neurosci. 2003;5(3):24958. Rosenbaum JF. Sleep panic attacks: an association
9. Sloan EP, Natarajan M, Baker B, Dorian P, Mironov with childhood anxiety and adult psychopathology.
D, Barr A, et al. Nocturnal and daytime panic Biol Psychiatry. 1994;36:5760.
attackscomparison of sleep architecture, heart rate 26. Craske MG, Barlow DH. Nocturnal panic. J Nerv
variability, and response to sodium lactate challenge. Ment Dis. 1989;177:1607.
Biol Psychiatry. 1999;45:131320. 27. Ohayon MM, Roth T. Place of chronic insomnia in the
10. Arriaga F, Paiva T, Matos-Pires A, et al. The sleep of course of depressive and anxiety disorders. J Psychiatr
non-depressed patients with panic disorder: a com- Res. 2003;37(1):915.
8 Circadian Rhythm in Panic Disorder 159

28. Mellman TA. Sleep and anxiety disorders. Sleep Med 46. Todder D, Baune BT. Quality of sleep in escitalopram-
Clin. 2008; 2618. treated female patients with panic disorder. Hum
29. Roberts RE, Shema SJ, Kaplan GA, Strawbridge WJ. Psychopharmacol. 2010;25(2):16773.
Sleep complaints and depression in an aging cohort: a 47. Milan Pavlovic Z. Remission of panic disorder with
prospective perspective. Am J Psychiatry. 2000;157: mirtazapine augmentation of paroxetine: a case
818. report. Prim Care Companion J Clin Psychiatry.
30. Ford DE, Kamerow DB. Epidemiologic study of sleep 2007;9(5):396.
disturbances and psychiatric disorders. An opportu- 48. Uguz F. Low dose mirtazapine added to selective
nity for prevention? JAMA. 1989;262:147984. serotonin reuptake inhibitors in pregnant women with
31. Breslau N, Roth T, Rosenthal L, Andreski P. Sleep major depression or panic disorder including symp-
disturbance and psychiatric disorders: a longitudinal toms of severe nausea, insomnia and decreased appe-
epidemiological study of young adults. Biol Psy- tite: three cases. J Matern Fetal Neonatal Med.
chiatry. 1996;39:4118. 2013;26(11):10668.
32. Bandelow B, Wedekind D, Sandvoss V, Broocks A, 49. Bossini L, Casolaro I, Koukouna D, Cecchini F,
Hajak G, Pauls J, et al. Diurnal variation of cortisol in Fagiolini A. Off label uses of trazodone: a review.
panic disorder. Psychiatry Res. 2000;95:24550. Expert Opin Pharmacother. 2012;13(12):170717.
33. Balbo M, Leproult R, Van Cauter E. Impact of sleep 50. Charney DS, Woods SW, Goodman WK, Rifkin B,
and its disturbances on hypothalamo-pituitary-adre- Kinch M, Aiken B, et al. Drug treatment of panic dis-
nal axis activity. Int J Endocrinol. 2010;vol. 2010, order: the comparative efficacy of imipramine, alpra-
Article ID 759234, 16 pages, 2010. doi:10.1155/ zolam, and trazodone. J Clin Psychiatry. 1986;47(12):
2010/759234. 5806.
34. Nesse RM, Cameron OG, Curtis GC, McCann DS, 51. Serretti A, Chiesa A, Calati R, Perna G, Bellodi L, De
Huber Smith MJ. Adrenergic function in patients Ronchi D. Novel antidepressants and panic disorder:
with panic anxiety. Arch Gen Psychiatry. 1984; evidence beyond current guidelines. Neuropsycho-
41:7716. biology. 2011;63(1):17.
35. Goldstein S, Halbreich U, Asnis G, Endicott J, Alvir 52. McAllister-Williams RH, Baldwin DS, Haddad PM,
J. The hypothalamic-pituitary-adrenal system in panic Bazire S. The use of antidepressants in clinical prac-
disorder. Am J Psychiatry. 1987;144:13203. tice: focus on agomelatine. Hum Psychopharmacol.
36. Reiss S, Peterson RA, Gursky DM, McNally RJ. 2010;25(2):95102.
Anxiety sensitivity, anxiety frequency and the predic- 53. Levitan MN, Papelbaum M, Nardi AE. A review of
tions of fearfulness. Behav Res Ther. 1986;24:18. preliminary observations on agomelatine in the treat-
37. McNally RJ. Anxiety sensitivity and panic disorder. ment of anxiety disorders. J Clin Psychol. 2012;
Biol Psychiatry. 2002;52(10):93846. 68(4):397402.
38. Deacon B, Abramowitz J. Anxiety sensitivity and its 54. Stein DJ, Ahokas AA, de Bodinat C. Efficacy of
dimensions across the anxiety disorders. J Anxiety agomelatine in generalized anxiety disorder: a ran-
Disord. 2006;20(7):83757. domized, double-blind, placebo-controlled study.
39. Hoge EA, Marques L, Wechsler RS, Lasky AK, J Clin Psychopharmacol. 2008;28(5):5616.
Delong HR, Jacoby RJ, et al. The role of anxiety sen- 55. Fornaro M. Agomelatine in the treatment of panic dis-
sitivity in sleep disturbance in panic disorder. order. Prog Neuropsychopharmacol Biol Psychiatry.
J Anxiety Disord. 2011;25(4):5368. 2011;35(1):2867.
40. Smits JA, Berry AC, Tart CD, Powers MB. The effi- 56. Millan MJ, Brocco M, Gobert A, Dekeyne A.
cacy of cognitive-behavioral interventions for reduc- Anxiolytic properties of agomelatine, an antidepres-
ing anxiety sensitivity: a meta-analytic review. Behav sant with melatoninergic and serotonergic properties:
Res Ther. 2008;46:104754. role of 5-HT2C receptor blockade. Psychophar-
41. Simon NM, Otto MW, Smits JA, Nicolaou DC, Reese macology (Berl). 2005;177:44858.
HE, Pollack MH. Changes in anxiety sensitivity with 57. American Thoracic Society. Indications and standards
pharmacotherapy for panic disorder. J Psychiatr Res. for use of nasalcontinuous positive airway pressure
2004;38:4915. (CPAP) in sleep apnea syndromes. American Thoracic
42. Marcks BA, Weisberg RB. Co-occurrence of insom- Society. Official statement adopted March 1944. Am
nia and anxiety disorders: a review of the literature. J Respir Crit Care Med. 1994;150:173845.
Am J Lifestyle Med. 2009;3:3009. 58. Mitchell MD, Gehrman P, Perlis M, Umscheid
43. Morin CM. Insomnia: psychological assessment and CA. Comparative effectiveness of cognitive behav-
management. New York: Guillford Press; 1993. ioral therapy for insomnia: a systematic review. BMC
44. Nardi AE, Freire RC, Valena AM, Amrein R, de Fam Pract. 2012;13:40.
Cerqueira AC, Lopes FL, et al. Tapering clonazepam 59. Barlow DH, Craske MG. Mastery of your anxiety
in patients with panic disorder after at least 3 years of and panic II. San Antonio, TX: The Psychological
treatment. J Clin Psychopharmacol. 2010;30(3): Corporation; 1994.
2903. 60. Edinger JD, Hoelscher TJ, Marsh GR, Ionescu-
45. Cervena K, Matousek M, Prasko J, Brunovsky M, Pioggia M, Lipper S. A cognitive-behavioral therapy
Paskova B. Sleep disturbances in patients treated for for sleep maintenance insomnia in older adults.
panic disorder. Sleep Med. 2005;6(2):14953. Psychol Aging. 1992;7:2829.
Some Genetic Aspects of Panic
Fabiana Leo Lopes

Contents Abstract
9.1 Introduction 162 Panic disorder (PD) is a multifactorial disease
and despite being the anxiety disorder with
9.2 Linkage Studies 162
higher heritability, the underlying genetic
9.3 Candidate Gene Association Studies 163 basis of this disorder remains poorly eluci-
9.4 Genome Wide Association Studies 164 dated. Several candidate genes have been
described so far, but they generally are charac-
9.5 Other Genetics Studies 165
terized by small sample sizes, have small
9.6 Conclusion 165 effects, lack replication and translational
References 166 models. Initial attempts to perform genome-
wide association studies (GWAS) in PD did
not lead to significant results nor were con-
firmed in subsequent studies. These facts
serve to call attention to the PD, as other psy-
chiatric disorders, is likely to be a multigenic
and heterogeneous disease, with small-effect
alleles that do not reach genome-wide signifi-
cant results. Moreover, even presenting a
polygenic basis, future genetic studies for PD
should comprise large-scale multicenter stud-
ies under international collaboration in order
to obtain representative samples. Still, the
techniques of next generation sequencing
which are already dominating the field in
other psychiatric disorders aim to reveal the
common and rare genetic variants associated
with the PD. With a lifetime prevalence of
approximately 4 % and outlined endopheno-
F.L. Lopes (*) types (i.e. carbon dioxide sensitivity), a better
Laboratory of Panic and Respiration, Institute of
understanding of the genetic basis and bio-
Psychiatry, Federal University of Rio de Janeiro,
Rio de Janeiro, Brazil logical mechanisms underlying PD is very
e-mail: lopes.fabiana@gmail.com important.

Springer International Publishing Switzerland 2016 161

A.E. Nardi, R.C.R. Freire (eds.), Panic Disorder, DOI 10.1007/978-3-319-12538-1_9
162 F.L. Lopes

Keywords and internalizing disorders in the general popula-

Panic disorder Single nucleotide polymorphism tion differ significantly in relation to gender. PD
Genome-wide association study Genetic asso- accounted for the largest sex difference in their
ciation studies Inheritance patterns review, demonstrating a much better index for
genetic risk in women than in men (heritability in
males: 21 %; heritability in females: 96 %). See
Fig. 9.1.
9.1 Introduction Segregation studies also consistently confirm
a genetic component to transmission of PD, but
Panic disorder (PD) is a multifactorial disease they had not supported any specific mode of
likely involving biological, psychological, inheritance. While some studies have suggested a
survival-related evolutionary factors and the dominant transmission pattern, others have found
interaction of all of the above factors. To under- support for both dominant and recessive modes.
stand the contribution of genetic factors in the In this way, segregation studies in PD have not
etiology of a condition of interest, some resources found a pattern of inheritance according to
from clinical genetics are frequently used, like Mendelian rules, indicating that PD, as like the
family studies, twin studies, adoption studies and whole psychiatric disorders, are complex genetic
segregation studies. For instance, family studies disorders raised by an interaction of many factors
of PD support a pattern of familial aggregation such as polygenic and environmental factors.
indicating that the disease runs in families. In The great majority of molecular studies con-
a comprehensive meta-analysis [1] the results ducted on PD still consist of linkage and candi-
showed a significant association between PD in date gene association studies. A summary of the
the probands and in their first-degree relatives main findings, advantages and pitfalls are detailed
(summary odds ratio of the studies taken into elsewhere. Unlike other psychiatric diseases
account: 5.0; 95% CI: 3.08.2). The estimated schizophrenia, bipolar disorder, major depression
relative risk to siblings of PD probands is five- to disorder, autism and attention deficit hyperactive
tenfold higher than the population risk, and the disorder genetics research in PD is still at the
heritability of PD is 0.48 [1]. beginning. One of the barriers leading to this sce-
Twin studies comprise monozygotic and dizy- nario could be the complex nature of PD pheno-
gotic ones. One could assume that the correlation type. First of all, we have to remember that PD is
(C) should be 1 or 100 % for monozygotics (MZ) characterized by increased pattern of comorbidi-
and 0.5 or 50 % for dizygotics (DZ). But what is ties with medical and psychiatric conditions.
generally found is a different condition (p.ex., Second, a decade of research trying to highlight
height CMZ: 0.94 and CDZ: 0.44; bipolar disor- the commonalities and differences between panic
der CMZ: 0.79 and CDZ: 0.24; measles CMZ: and fear has not reached a consensus yet. The
0.95 and CDZ: 0.87) [2] . The difference of con- main repercussion that comes out is a fuzzy phe-
cordance between monozygotic and dizygotic notype definition. And last, but not least, we cur-
twins is used to estimate the heritability. Indeed, rently have disparities in funding and advocacy
when the difference is large means that exists a leading the field of panic disorder to be behind
great role of genes in determining the condition other mental health diseases [5].
(as we can observe in both examples of height
and bipolar disorder). Hettema et al. [1] in their
meta-analysis pointed out the estimate heritabil- 9.2 Linkage Studies
ity for PD to be 43 %. For illustration, Fig. 9.1
depicts the estimate heritability of psychiatric Many linkage methods dating from the mid-1900
diseases [3]. In addition, Kendler and Myers [4] were performed to analyze genetic data from
have recently highlighted that disorders that opti- populations with rare traits. However, rare traits
mally index the genetic liability to externalizing are usually influenced by a major gene or high
9 Some Genetic Aspects of Panic Disorder 163






40 h2





Fig. 9.1 Heritability estimates (h2) for mood and anxiety Agoraphobia, Blood-Ph Blood Injury phobia, SAD Social
disorders. Adapted from Bienvenu et al. [3], Stein et al. Anxiety Disorder, PD Panic Disorder, MDD Major
[25] and Kendler et al. [26]. BD Bipolar Disorder, Ago Depression Disorder, GAD Generalized Anxiety Disorder

penetrant condition what does not seem to occur intervals and unweighted genome scan meta-
in complex disorders like PD. Therefore, the analysis (GSMA) approach. The analysis for
detection sensitivity of linkage studies conducted anxiety identified nominal significance (p < 0.05)
in PD is low, due to the small individual effect of for regions on chromosomes 1, 2, 5, 9, 11, 15, 16
single genes. Moreover, these studies give an and 22. Chromosome 1 was correlated to both
approximate chromosomal location of the gene phenotypes, being nominally significant. The
related to an inherited phenotype. Using affected authors hypothesized that this region harbors
pedigrees, the rationale is if a marker signifi- genes broadly underlying anxiety susceptibility.
cantly co-segregates with the disease then the Linkage evidence for a broad phenotype panic
region around this marker contains genes confer- disorder + bipolar disorder has been obtained on
ring a disease risk. These studies use the loga- chromosomes 2q (lodscore = 4.6) and 12
rithm of the odds (LOD), which is based on the (lodscore = 3.6), under a dominant model of
number of estimates of recombination frequency. inheritance [27]. Descriptions of linkage on
A LOD score greater than 3.0 is considered sig- chromosome 2p, 2q, 9p, 12 and 15q (near
nificant (indicating that a thousand to one in GABA-A receptor subunitor genes) among fam-
favor of genetic linkage) and a LOD score greater ilies with PD have been consistently found on
than 1.9 is considered suggestive [6]. the scientific literature [8].
Webb et al. [7] performed a meta-analysis of
Genome Wide Linkage Scan on independent
samples of Neuroticism and Anxiety (two stud- 9.3 Candidate Gene Association
ies of panic disorder and one using a broad anx- Studies
iety definition). The data comprised 5341
families encompassing 15,529 individuals (the This approach has been the major source of
anxiety sample consisted of 718 subjects from investigations and the available published data in
162 families). Rank based genome scan was PD. These studies are based on association para-
used to analyze each trait separately and com- digm using case-control and/or family based
bined. They presented the results for 10 CM analysis. Although they have stronger power than
164 F.L. Lopes

linkage studies to identify loci associated to between PD and the region of the GABRB3 and
complex diseases, there are important pitfalls that GABRA5 genes in both United States and
should be taken into account when facing those Sardinian families, and a number of novel
studies. sequence variants in the region of these two
So far, candidate genes have been identified genes among PD probands. They also found sup-
on the basis of our current knowledge regarding port for the genetic contribution to GABRB3
the pathophysiology of panic disorder. There are expression, which may suggest a regulatory
more than 360 candidate genes currently investi- mechanism for the hypothesized GABAergic
gated to be associated with PD [9, 10]. Despite dysregulation in PD [14].
this huge number, only a small number has been
replicated and only a handful has dealt with sam-
ple sizes higher than 200 subjects. In addition, 9.4 Genome Wide Association
the results are nominally significant or even pres- Studies
ent a trend but are often uncorrected for multiple
comparisons. Out of 364 genes, only about 56 The first genome-wide association study (GWAS)
presented with positive associations [10, 11]. in PD was a Japanese study conducted among
Due to all of these conditions, the results of can- 200 subjects with PD and 200 controls [15]. All
didate genes association studies in PD seem not of the included subjects had a Japanese ascen-
to be robust and indicate that most of the candi- dant. None of the initial nominal association find-
date studies are not likely to have a main role in ings were replicated in the subsequent study, with
the susceptibility to this disorder. a sample size comprising 558 cases and 566 con-
The genes that have been more consistently trols [16]. A meta-analysis of these two Japanese
studied are cholecystokinin (CCK), cholecysto- studies (718 cases/1717 controls) followed by
kinin B receptor (CCKBR), 5-hydroxytryptamine a replication analysis (329 cases/861 controls)
receptor 2A (HTR2A), solute carrier family found no significant genome-wide results [17].
6 (neurotransmitter transporter), member 4 Evidence emerging from this study support the
(SLC6A4), adenosine A2a receptor (ADORA2A), involvement of the previous candidate gene
catechol-O-methyltransferase (COMT) and mono- NPYSR (4q31.3-32; p = 6.4 104). In a genome-
amine oxidase A (MAO). The most recent data is wide scan using microsatellite markers among
for transmembrane protein 132D (TMEM132D) PD patients and controls, from the isolated popu-
[12], acid-sensing (proton gated) ion channel 1 lation of the Faroe Islands, Gregerson et al. [18]
(ASIC1) [13], gamma-aminobutyric acid found evidence for the amiloride-sensitive cation
(GABA) A receptor, alpha 5 (GABRA5) and channel 1 (ACCN1) located on chromosome
gamma-aminobutyric acid (GABA) A receptor, 17q11.2-q12 as a potential candidate gene for
beta3 (GABRB3) [14]. The function of PD. Further analyses of the ACCN1 gene using
TMEM132D still remains to be confirmed but single-nucleotide polymorphisms (SNPs) reve-
there is suggestion of its involvement in neuronal aled significant association with PD in an
sprouting and brain connectivity [12]. The extended Faroese case-control sample. However,
amiloride-sensitive cation channel 2 gene the authors were not able to replicate the findings
(ACCN2) is the human ortholog of Asic1a and is in a larger and independent Danish case-control
highly expressed in the amygdala [13]. In a case- sample and concluded that possible risk alleles
control study, the authors observed an increased associated with PD in the isolated population are
effect of ACCN2 alleles on early-onset PD and not those ones involved in the development of
on the respiratory subtype of PD. In this way, PD in a larger outbred population. A Japanese
they speculate that ACCN2 variants may lead to study evaluated genome-wide copy number vari-
PD risk by lowering the threshold for amygdala ation association in PD [19] and detected positive
sensing of acidosis [13]. The study of Hodges results for duplications in the peri-centromeric
et al. [14] found additional support for association region in the chromosome 16p11.2. Nevertheless,
9 Some Genetic Aspects of Panic Disorder 165

the association level was borderline, the results to further studies of mi-RNAs and/or the involve-
were not replicated, translational validation still ment of regulatory regions in the etiopathological
lacks and though, additional confirmation studies basis of PD.
are needed [20]. Epigenetic studies are also scarce in the field
Among a sample consisting of 1001 European of PD. Domschke et al. [23] investigated DNA
American bipolar cases and 1034 controls, a methylation patterns in the regions of glutamate
GWAS was conducted throughout bipolar related decarboxylases GAD1 and GAD2 promoter
phenotypes. Thus, in the panic attack-adjusted and GAD1 intron 2 to be associated with PD. To
analysis, top-ranking SNPs included rs599845 in this end, 65 subjects with PD and 65 controls
a protein-coding region of SGOL1. Although this were analyzed. PD patients exhibited signifi-
result has never been associated with PD before, cantly lower average GAD1 methylation than
it had already been linked to temperament traits healthy controls, particularly at three CpG sites
in bipolar disorder. The results in this study did in the promoter as well as in intron 2. The authors
not reach genome-wide significance [21]. point to a potentially compensatory role of GAD1
The most robust results are for the transmem- gene hypomethylation in PD probably mediating
brane protein gene 1342D located on chromo- the influence of negative life events.
some 12. This GWAS was initially conducted in Recently, Sasakis team [24] performed a
MaxPlanck and expanded with the Consortium pathway analysis from the SNPs data genotyped
of Panic, totaling 2678 PD cases and 3262 con- for GWAS and converted to genes associations.
trols. A correlation with European ancestry was They performed three different types of pathway
detected and the results were not replicated in the analysis and each one showed that those path-
Japanese population. The results involving the ways related to immunity had the strongest asso-
transmembrane protein gene 1342D were further ciation with PD. The authors focused on and
validated in animal models with greater anxiety investigated HLA-B and HLA-DRB1 as candi-
associated with increased expression of mRNA date susceptibility genes for PD. Therefore, 744
TMEM in the anterior cingulate cortex. In PD subjects and 1418 controls were typed for
humans, associations with increased expression HLA-B and HLA-DRB1. It came out that patients
in the frontal cortex in postmortem brains have with PD were significantly more likely to carry
been reported [12]. HLA-DRB1*13:02 (p = 2.50 104, odds ratio =
1.49). This study provided initial evidence that
genes involved in immune related pathways are
9.5 Other Genetics Studies associated with PD.

A study conducted by the Spanish team analyzed

the involvement and regulation of micro-RNAs 9.6 Conclusion
(mi-RNA) in candidate genes for PD. They used
712 SNPs covering 325 regions of mi-RNAs in a Studies of PD molecular genetics are at a prelimi-
sample of 203 subjects with PD and 341 controls. nary stage compared to other pathologies with
The strongest associations occurred for two SNP: greater investment in the field of research and
rs6502892 tagging miR-22 (p < .0002) and development, such as schizophrenia, depression
rs11763020 tagging miR-339 (p < .00008), and bipolar disorder. Despite being the anxiety
although such associations have not survived disorder with higher heritability, the underlying
after multiple corrections. Replications in Finnish genetic basis of PD remains poorly elucidated.
and Estonian samples did not support these asso- Several candidate genes have been reported pre-
ciations. Functional studies have shown that senting, in some way, a disappointing result.
miR-22 regulates the four candidate genes of Small sample size (N < 200, largely) with only
PD BDNF, HTR2C, MAOA and RGS2 [22]. small effects that fail replication have been
Though preliminary, such data may call attention almost the rule than exception, dominating thus
166 F.L. Lopes

the scientific literature of such studies. Initial 10. McGrath LM, Weill S, Robinson EB, MacRae R,
attempts to perform GWAS in PD did not lead to Smoller JW. Bringing a developmental perspective to
anxiety genetics. Dev Psychopathol. 2012;24(4):
significant results nor were confirmed in subse- 117993.
quent studies. These facts serve to call attention 11. Maron E, Hettema JM, Shlik J. Advances in molecu-
to the PD, as in other psychiatric disorders, is lar genetics of panic disorder. Mol Psychiatry.
likely to be a multigenic and heterogeneous dis- 2010;15(7):681701.
12. Erhardt A, Akula N, Schumacher J, Czamara D,
ease, with small-effect alleles that does not reach Karbalai N, Mller-Myhsok B, et al. Replication and
genome-wide significant results. Moreover, even meta-analysis of TMEM132D gene variants in panic
presenting likely a polygenic basis future delin- disorder. Transl Psychiatry. 2012;2:e156.
eations for panic disorder comprise larger-scale 13. Smoller JW, Gallagher PJ, Duncan LE, McGrath LM,
Haddad SA, Holmes AJ, et al. The human ortholog of
multicenter studies under international collabora- acid-sensing ion channel gene ASIC1a is associated
tion in order to obtain a representative sample. with panic disorder and amygdala structure and func-
Still, the techniques of next generation sequenc- tion. Biol Psychiatry. 2014;76(11):90210.
ing which are already dominating the field in 14. Hodges LM, Fyer AJ, Weissman MM, Logue MW,
Haghighi F, Evgrafov O, et al. Evidence for linkage
other psychiatric disorders aim to reveal the and association of GABRB3 and GABRA5 to panic
common and rare genetic variants associated disorder. Neuropsychopharmacology. 2014;39(10):
with the PD. With a lifetime frequency of approx- 242331.
imately 4 % and outlined endophenotypes (i.e. 15. Otowa T, Yoshida E, Sugaya N, Yasuda S, Nishimura
Y, Inoue K, et al. Genome-wide association study of
carbon dioxide sensitivity), a better understand- panic disorder in the Japanese population. J Hum
ing of the genetic pathways interactions and bio- Genet. 2009;54:1226.
logical mechanisms underlying PD became very 16. Otowa T, Tanii H, Sugaya N, Yoshida E, Inoue K,
important. Yasuda S, et al. Replication of a genome-wide asso-
ciation study of panic disorder in a Japanese popula-
tion. J Hum Genet. 2010;55(2):916.
17. Otowa T, Kawamura Y, Nishida N, Sugaya N, Koike
References A, Yoshida E, et al. Meta-analysis of genome-wide
association studies for panic disorder in the Japanese
1. Hettema JM, Neale MC, Kendler KS. A review and population. Transl Psychiatry. 2012;2:e186.
meta-analysis of the genetic epidemiology of anxiety 18. Gregersen N, Dahl HA, Buttenschn HN, Nyegaard
disorders. Am J Psychiatry. 2001;158:156878. M, Hedemand A, Als TD, et al. A genome-wide study
2. Thomas A. Thrive in genetics. Oxford: Oxford of panic disorder suggests the amiloride-sensitive
University Press; 2013. Quantitative genetics, p. 97. ction channel 1 as a candidate gene. Eur J Hum
3. Bienvenu OJ, Davydow DS, Kendler KS. Psychiatric Genet. 2012;20(1):8490.
diseases versus behavioral disorders and degree of 19. Kawamura Y, Otowa T, Koike A, Sugaya N, Yoshida
genetic influence. Psychol Med. 2011;41(1):3340. E, Yasuda S, et al. A genome-wide CNV association
4. Kendler KS, Myers J. The boundaries of internalizing study on panic disorder in a Japanese population.
and externalizing genetic spectra in men and women. J Hum Genet. 2011;56(12):8526.
Psychol Med. 2014;44:64755. 20. Erhardt A, Spoormaker VI. Translational approaches
5. Smoller JW. Whos afraid of anxiety genetics? Biol to anxiety: focus on genetics, fear extinction and brain
Psychiatry. 2011;69:5067. imaging. Curr Psychiatry Rep. 2013;15(12):417.
6. Lander ES, Schork NJ. Genetic dissection of complex 21. Cuellar-Barboza AB, Winham SJ, Colby C, Prieto M,
traits. Science. 1994;265(5181):203748. Chauhan M, McElroy SL, et al. Genome-wide asso-
7. Webb BT, Guo AY, Maher BS, Zhao Z, van den Oord ciation study of bipolar disorder related phenotypes:
EJ, Kendler KS, et al. Meta-analyses of genome-wide rapid cycling, alcohol use disorders and panic attacks.
linkage scans of anxiety-related phenotypes. Eur Biol Psychiatry. 2014;75:1S401.
J Hum Genet. 2012;20(10):107884. 22. Muios-Gimeno M, Espinosa-Parrilla Y, Guidi M,
8. Fyer AJ, Hamilton SP, Durner M, Haghighi F, Heiman Kagerbauer B, Sipil T, Maron E, et al. Human micro
GA, Costa R, et al. A third-pass genome scan in panic RNAs miR-22, miR-138-2, miR-148a, and miR-488
disorder: evidence for multiple susceptibility loci. are associated with panic disorder and regulate sev-
Biol Psychiatry. 2006;60(4):388401. eral anxiety candidate genes and related-pathways.
9. Maron E, Lang A, Tasa G, Liivlaid L, Tru I, Must A, Biol Psychiatry. 2011;69:52633.
et al. Associations between serotonin-related gene 23. Domschke K, Tidow N, Schrempf M, Schwarte K,
polymorphisms and panic disorder. Int J Neuro- Klauke B, Reif A, et al. Epigenetic signature of panic
psychopharmacol. 2005;8(2):2616. disorder: a role of glutamate decarboxylase 1 (GAD1)
9 Some Genetic Aspects of Panic Disorder 167

DNA hypomethylation? Prog Neuropsychopharmacol 26. Kendler KS, Karkowski LM, Prescott CA. Fears and
Biol Psychiatry. 2013;46:18996. phobias: reliability and heritability. Psychol Med.
24. Shimada-Sugimoto M, Otowa T, Miyagawa T, Khor 1999;29(3):53953.
SS, Kashiwase K, Sugaya N, et al. Immune-related 27. Logue SF, Grauer SM, Paulsen J, Graf R,
pathways including HLA-DRB1*13:02 are associated Taylor N, Sung MA, Zhang L, Hughes Z, Pulito VL,
with panic disorder. Brain Behav Immun. 2015; Liu F, Rosenzweig-Lipson S, Brandon NJ,
46:96103. Marquis KL, Bates B, Pausch M. The orphan
25. Stein MB, Jang KL, Livesley WJ. Heritability of GPCR, GPR88, modulates function of the striatal
social anxiety-related concerns and personality char- dopamine system: a possible therapeutic target for
acteristics: a twin study. J Nerv Ment Dis. 2002; psychiatric disorders? Mol Cell Neurosci. 2009;42(4):
190(4):21924. 43847.
Panic Disorder and Personality
Disorder Comorbidity
Ricard Navins, El Egmond,
and Roco Martn-Santos

Contents Abstract
10.1 Introduction 170 The present chapter systematically reviews
10.1.1 Epidemiological Data of Panic the relationship between panic disorder, with
Disorder 170 and without co-occurring anxiety or depres-
10.1.2 Panic Disorder and Comorbidity 170 sion, and current personality disorder. Data
10.1.3 What We Know About the
Association Between Panic Disorder were collected with an advanced document
and Personality Disorders 171 protocol according to MOOSE (Meta-analysis
10.2 Methodology of the Review 171
of Observational Studies in Epidemiology)
10.2.1 Studies 171 guidelines for observational studies. A com-
10.2.2 Search Strategy 171 prehensive, computerized literature search
10.2.3 Data Extraction 172 was conducted in Medline, PsycINFO, and
10.2.4 Inclusion and Exclusion Criteria 172
10.2.5 Statistical Analysis 172
LILACS. Cohort, case-control and cross-sec-
tional surveys studies evaluating the comor-
10.3 Results 172
bidity between DSM panic disorder and
10.4 Discussion 173 personality disorders were included. Overall
10.4.1 Evidence-Based on the Review 173 prevalence, comorbidity rates, and 95 % CI
10.4.2 Nature of the Relationship Between
Panic Disorder and Personality were calculated with a random effects model.
Disorder 179 From 97 initial selected papers, 24 entered in
10.4.3 Impact of Personality on Panic the review. Among patients with a current
Disorder 179 DSM-III/R/IV panic disorder, 44.3 % (34.6
10.5 Conclusions 180 54.2 %) had any personality disorder; 6.3 %
References 180 (3.110.4 %) had cluster A; 17.9 % (12.2
24.2 %) cluster B, and 34.9 % (25.644.7 %)
had cluster C. Among patients with a current
panic disorder and co-occurring anxiety or

E. Egmond
R. Navins R. Martn-Santos (*)
Department of Psychiatry and Psychology, Hospital
Department of Psychiatry and Psychology, Hospital
Clinic, Barcelona, Spain
Clinic, Institut dInvestigaci Biomdica August Pi I
Sunyer (IDIBAPS), Centro de Investigacin Department of Clinical and Health Psychology,
Biomdica en Red en Salud Mental (CIBERSAM), Faculty of Psychology, Universidad Autnoma de
G25, Universidad de Barcelona, Barcelona, Spain Barcelona, Cerdanyola del Valls, Barcelona, Spain
e-mail: rnavines@clinic.ub.es; rmsantos@clinic.ub.es e-mail: egmond@clinic.ub.es

Springer International Publishing Switzerland 2016 169

A.E. Nardi, R.C.R. Freire (eds.), Panic Disorder, DOI 10.1007/978-3-319-12538-1_10
170 R. Navins et al.

depression, 61.8 % (44.677.7 %) had any per- Female gender, low socioeconomic status, and
sonality disorder, 7.2 % (4.410.5 %) had anxious childhood temperament are common
cluster A; 24.0 % (17.630.9 %) cluster B, and risk factors for panic disorder. Panic disorder can
38.6 % (25.752.2 %) had cluster C. In con- produce marked distress and impairment, and is
clusion, comorbidity between panic disorder associated with significant suicide risk. Panic
and personality disorders is common. Cluster disorder appears to increase risk for all-cause
C was the most frequent personality disorder mortality because it may increase risk for cardio-
subtype related to panic disorder. Personality vascular disease [2, 4].
disorders were more prevalent among indi-
viduals with panic disorder and co-occurring
anxiety or depression. 10.1.2 Panic Disorder
and Comorbidity

The diagnosis is frequently associated with other

Keywords comorbid axis-I psychiatric disorders, especially
Panic disorder Personality disorder with depressive and other anxiety disorders [5].
Systematic review Depression comorbidity Moreover, panic disorder can also co-occur with
Cluster A Cluster B Cluster C Treatment comorbid axis-II psychiatric disorder, resulting in
non-responders the diagnosis of personality disorder [6].
A personality disorder is a persistent and mal-
adaptive pattern of internal experience and behav-
ior, that have their beginning in the adolescence or
10.1 Introduction first adult age, and that causes significant malaise
or deterioration in the activity of the individual
10.1.1 Epidemiological Data of Panic [DSM-IV]. Currently, personality disorders con-
Disorder stitute an important medical and social pattern,
shown by a high prevalence (1015 % in general
Panic disorder is an anxiety disorder character- population and until 50 % in psychiatric patients),
ized by unexpected and repeated episodes of as well as the personal repercussions and partner-
intense fear accompanied by physical symptoms relatives who tolerate [7]. Although personality
that may include chest pain, heart palpitations, disorders have been defined categorically through-
and shortness of breath, dizziness, or abdominal out the history of psychiatric nomenclatures, the
distress. Panic attacks usually produce a sense of DSM-5 Personality and Personality disorders
unreality, a fear of impending doom, or a fear of Work Group proposed a substantial shift to a
losing control [1]. dimensional conceptualization and diagnosis of
The estimated current prevalence rate for personality pathology [8]. The DSM-5 gives a
panic disorder is about 15 % of the adult popula- categorical classification of personality disorders,
tion [2]. In the National Epidemiologic Survey grouped into three clusters (2013) [9].
on Alcohol and Related Conditions, the overall The repercussion of personality disorders in
12-month and lifetime prevalence rates for panic panic disorder has been evaluated in different stud-
disorder (with or without agoraphobia) were ies, suggesting that this comorbidity is associated
2.1 % and 5.1 %. The 12-month and lifetime with a greater severity of the symptoms of panic
prevalence rates for panic disorder with agora- disorder [10], and also with a greater prevalence of
phobia were 0.6 % and 1.1 %, while the corre- comorbid agoraphobia [11]. Thus, the degree of
sponding rates for panic disorder without fear and phobia is usually more pronounced in
agoraphobia were 1.6 % and 4.0 %. Agoraphobia patients with axis II comorbidity, as is the level of
without panic disorder was uncommon (12- general psychopathology [10, 12]. This is also
month prevalence 0.05 %; lifetime prevalence reflected in health care cost analyses, showing that
0.17 %) [3]. axis II personality disorder patients represent a
10 Panic Disorder and Personality Disorder Comorbidity 171

markedly higher economic burden to the health phobia showing a higher prevalence of personal-
care system than for example patients with depres- ity disorders, while generalized anxiety disorder
sion and anxiety [13]. Moreover, having a person- and post-traumatic stress disorder showed no
ality disorder represents a strong vulnerability relationship between both. The study was also
factor for developing other axis I disorders [14]. criticized [26] for using a personality disorder
Moreover, treatment of comorbid personality rather than an anxiety disorders as the main
disorder is normally more complex and has less inclusion criteria. More recently, Friborg et al.
favorable outcomes for panic disorder patients [27] performed a systematic review and meta-
[1517], higher drop-out rates [18], less positive analysis to identify the proportions of comorbid
patient expectations [19], and more challenges personality disorder across the major subtypes of
establishing a durable and flexible therapeutic anxiety disorders. The rate of any comorbidity
alliance [19, 20]. Patients with cluster A or clus- in Axis II was high across all anxiety disorders,
ter B also appear to have a poorer treatment ranging from 35 % for post-traumatic stress dis-
response than cluster C patients [21, 22]. Also, order to 52 % for obsessive-compulsive disorder.
the degree of comorbid psychosocial impairment Globally, cluster C occurred more than twice
depends on the type of personality disorder. as often as cluster A or B. Within cluster C, the
A higher degree of impairment seems to exist avoidant personality disorder occurred most fre-
among schizotypal and borderline patients than quently, followed by the obsessive-compulsive
among obsessive-compulsive or avoidant per- and the dependent subtype. Gender or duration
sonality disorder patients [23]. of an anxiety disorder was not related to varia-
tion in personality disorder comorbidity.
In this chapter we undertook a systematic
10.1.3 What We Know review, and meta-analysis when possible, of
About the Association clinical observational studies, to summarize the
Between Panic Disorder relationship between personality disorders, and
and Personality Disorders prevalence of panic disorder with and without
anxiety or depression. The lack of previous sys-
Several studies of general associations between tematic reviews or meta-analyses using panic
personality disorder and panic disorder have disorder as the primary inclusion criteria was the
been published [6, 11, 14, 24]. The general con- main reason for the present study.
clusions from these reviews point out that the
comorbid personality disorder among patients
with a panic disorder diagnosis vary consider- 10.2 Methodology of the Review
ably, but the proportions of avoidant, dependent
and compulsive (cluster C) personality disorders 10.2.1 Studies
are the highest. These proportions are smaller in
the schizoid, schizotypal and paranoid (cluster A) For this review, we considered all relevant cohort,
as well as the dramatic, borderline and anti-social case-control and cross-sectional survey studies
personality disorders (cluster B). that evaluate the comorbidity between panic dis-
Very few systematic reviews and meta- order and personality disorders.
analyses on comorbidity between Axis II and
anxiety disorders in general have been pub-
lished. There is one previous systematic review 10.2.2 Search Strategy
and meta-analysis by Borenstein [25] that spe-
cifically focuses on examining comorbidity with Data were collected with an advanced document
dependent personality disorder. The ratio var- protocol according to MOOSE (Meta-analysis of
ied considerably between diagnostic groups of Observational Studies in Epidemiology) guide-
anxiety disorder, with panic disorder, obsessive- lines for observational studies [28]. A com-
compulsive disorder, agoraphobia, and social prehensive, computerized literature search was
172 R. Navins et al.

conducted in Medline (1984Jan 2012); participants also had to be diagnosed with

EMBASE (1984Dec 2014); PsychLIT (1984 categorical personality disorder using the
Dec 2014); CINAHL (1984Dec 2014); and Structured Clinical Interview for DSM-IV
LILACS (1984Dec 2014), for studies in humans (SCID-II) [32].
of the association between panic disorder and The exclusion criteria were studies with
personality disorders. Our search terms included patients with another axis I comorbidity, except
any combination of the key words panic, other anxiety disorder or depressive disorders,
panic disorder, personality, personality and studies where panic disorder diagnose was
disorder, depression, and comorbidity. We secondary to medical or substance use pathology.
also reviewed reference lists of the identified Studies were also excluded if they were not
studies and review articles to search for addi- published as full reports, such as conference
tional studies. abstracts and letters to editors; or if N/% was not
used in measuring the prevalence of personality
disorders. If multiple published reports were
10.2.3 Data Extraction available from the same study, we included only
the one with the most detailed information on the
The titles and abstracts were examined, and full- relationship in question. Studies which evaluated
text articles of potentially relevant studies were only a subtype of personality disorder were also
obtained. Subsequently, inclusion and exclusion excluded.
criteria were applied, and the selected articles
were included in this systematic review.
Data was extracted from each study using a 10.2.5 Statistical Analysis
standardized spreadsheet. Information extracted
included the following: title, author, year of Cross-tabulations were used to calculate overall
publication, study design, sample size, age, sex, prevalence and comorbidity rates with a random
methods of interview, panic disorder diagnose, effects model. Standard errors and 95 % confi-
comorbid DSM (III, IV, IV-R) diagnoses, person- dence limits were estimated.
ality disorder diagnose, duration of panic
disorder, and onset of panic disorder. We also
extracted the N, % of any cluster personality dis- 10.3 Results
order, and we calculated the 95 % confidence
interval. Using keywords, 448 articles were identified
Two clinical researchers (RN, and EE), a psy- and titles and abstracts were examined. At this
chiatrist and psychologist, performed each step in stage, 351 articles were eliminated because they
this literature research, study identification, study did not meet the selection criteria a priori. We
selection, and data extraction. Disagreements obtained 99 potentially relevant papers, which
were resolved by discussion, and consensus was were thoroughly examined. Twenty-four articles
achieved in the selection of articles for analysis. were rejected because they failed to meet inclu-
sion criteria [11, 22, 3354], 45 met exclusion
criteria for comorbidity [22, 45, 46, 50, 5568],
10.2.4 Inclusion and Exclusion no having a structured clinical diagnosis [33, 41,
Criteria 44, 6976], no measuring in N/% [45, 7780],
multiple published reports [56, 8186], partial
The inclusion criteria were clinical studies of inclusion of personality disorders [45, 57, 61,
subjects diagnosed with panic disorder using the 64, 82, 87], no specification of axis-I comorbid-
Structured Clinical Interview for DSM-III, DSM- ity [88], or because of language restriction [89];
III-R or DSM-IV (SCID-I) [2931] irrespective were letters or reviews [43, 75, 9095] (Fig. 10.1
of gender, race, age, or nationality. All study shows the flow chart). We finally selected 24
10 Panic Disorder and Personality Disorder Comorbidity 173

Fig. 10.1 Flow-chart of

Potential relevant studies
studies selection

Failed to meet
inclusion criteria*

Met exclusion


Final selected

Panic disorder & Panic disorder & depression

personality disorders & personality disorders
N=16 N=8

published studies, 16 evaluating the comorbid-

ity between panic disorder and personality dis- 10.4 Discussion
orders [16, 17, 24, 86, 96107], or eight between
panic disorder and co-occurring major depres- 10.4.1 Evidence-Based on the Review
sion and personality disorders [10, 68, 97, 102,
103, 107109], using the Structured Clinical Panic Disorder
Interview for DSM-III, DSM-III-R or DSM-IV and Comorbid Personality
(SCID-II). Disorder
Table 10.1 presents the characteristics of the This systematic review confirms that the comor-
selected studies of panic disorder patients and bidity between panic disorder with or without
comorbid personality disorder [16, 17, 24, 86, depression and personality disorder was com-
96107], and Table 10.1 of panic disorder patients mon. Personality disorders were more prevalent
with co-occurring depression and comorbid among individuals with panic disorder and co-
personality disorder [10, 68, 97, 102, 103, occurring depression. Furthermore, the asso-
107109]. ciations between panic disorder alone or with
Table 10.2 shows the overall prevalence and co-occurring anxiety or depression and Axis II
its 95 % confidence interval of the selected stud- disorder were all high. The study confirms that
ies of panic disorder patients and comorbid per- between panic patients with a personality disor-
sonality disorder, and Table 10.3 of panic disorder der, cluster C personality disorder was more fre-
patients with co-occurring depression and comor- quently associated with panic disorder than with
bid personality disorder. other personality disorders.

Table 10.1 Characteristics of the studies of panic disorder and comorbid personality disorders
Duration of Onset panic
Personality illness disorder
Women Age Panic disorder disorder (years) (years)
Author Year N (N, %) (Mean SD) Design diagnose diagnose (Mean SD) (Mean SD)
Mendoza et al. 2011 104 75 (71.1) 37.5 (8.8) Cross-sectional SCID-I (DSM-IV) SCID-II
Telch et al. 2011 173 128 (73) 35.2 () Cohort SCID-I (DSM-IV) SCID-II 9.1 ()
Gutirrez et al. 2008 157 108 (68.8) 34.9 (9.1) Cross-sectional SCID-I (DSM-IV) SCID-II
Iketani et al. 2004 105 58 (55.2) 36.9 (12.2) Cross-sectional SCID-P (DSM-III-R) SCID-II 33.3 (12.6)
Massion et al. 2002 386 260 (67.3) 39.8 (10.8) Cohort SCALUP(DSM-III-R) IPDE
Barzega et al. 2001 184 112 (60.9) 31.8 (9.8) Cross-sectional SCID I (DSM-IV) SCID-II 30.2 (10.3)
Dyck et al. 2001 230 153 (66.6) 41.0 (12.6) Cohort SCID I (DSM-III-R) IPDE
Latas et al. 2000 60 45 (75) 33.6 (7.7) Cross-sectional SCID I (DSM-IV) SCID-II 38.78 (44.05)
Ampollini et al. 1999 42 27 (64.3) 31.4 (9.6) Case-Control SCID I (DSM-III-R) SIDP-R
Langs et al. 1998 49 27 (55.1) 34.8 (9.7) Cross-sectional SCID I (DSM-III-R) SCID-II 4.4 (6.8) 30.4 (9.7)
Mauri et al. 1992 40 24 (60) 33.4 (9) Cross-sectional SCID I (DSM-III-R) PDE 6.4 (7) 26.7 (9)
Mellman et al. 1992 23 17 (78) 37.7 (11.9) Cohort SADS SID-P 12.7 (7.6) 25 (11.4)
Brooks et al. 1991 30 11 (37) 35 (12) Cross-sectional SCID-P (DSM-III-R) SCID-II
Sciuto et al. 1991 48 32 (66.7) 34.9 (10.2) Cross-sectional SCID-I-DSM-III-R SIDP-R 5.4 (5.7) 29.1 (10.3)
Alnaes et al. 1990 39 30 (1951) Cross-sectional SCID-I (DSM-III) SID-P
Reich et al. 1987 88 52 (59) 37.3 (1.9) Cross-sectional SCID I (DSM-III-R) SIDP
IPDE (International Personality Disorder Examination)-DSM-III-R criteria. SCID-DSM-III (Structured Clinical Interview Diagnoses for DSM-IIIv.patient). SCID-DSM-III-R
(Structured Clinical Interview Diagnoses for DSM-III-R v.patient). SCID-DSM-IV (Structured Clinical Interview Diagnoses for DSM-IV v.patient). SID-P (Structured Interview
for DSM-III Personality Disorders). SIDP-R (Revised Structured Interview for DSM-III-R Personality Disorders). PDE (Personality Disorder Examination-DSM-III-R and
R. Navins et al.
Table 10.2 Characteristic of studies of panic disorder and co-occurring depression and comorbid personality disorders
Panic Personality of illness Onset PD