See

discussions, stats, and author profiles for this publication at: http://www.researchgate.net/publication/45694427

Changes in body weight during

pharmacological treatment of depression
ARTICLE in THE INTERNATIONAL JOURNAL OF NEUROPSYCHOPHARMACOLOGY APRIL 2011
Impact Factor: 5.26 DOI: 10.1017/S1461145710000933 Source: PubMed

CITATIONS

DOWNLOADS

VIEWS

10

26

281

21 AUTHORS, INCLUDING:
Robert Keers

Jana Strohmaier

King's College London

Universitt Heidelberg

27 PUBLICATIONS 307 CITATIONS

72 PUBLICATIONS 1,835 CITATIONS

SEE PROFILE

SEE PROFILE

Katherine J Aitchison

Peter Mcguffin

University of Alberta

King's College London

161 PUBLICATIONS 2,884 CITATIONS

653 PUBLICATIONS 27,137 CITATIONS

SEE PROFILE

SEE PROFILE

Available from: Erik Roj Larsen

Retrieved on: 06 July 2015

International Journal of Neuropsychopharmacology (2011), 14, 367375. f CINP 2010

doi:10.1017/S1461145710000933

ARTICLE

Changes in body weight during

pharmacological treatment of depression
Rudolf Uher1, Ole Mors2, Joanna Hauser3, Marcella Rietschel4, Wolfgang Maier5,
Dejan Kozel6, Neven Henigsberg7, Daniel Souery8, Anna Placentino9, Robert Keers1,
Joanna M. Gray1, Mojca Zvezdana Dernovsek10, Jana Strohmaier4, Erik Roj Larsen2,
Astrid Zobel5, Aleksandra Szczepankiewicz3, Petra Kalember7, Julien Mendlewicz8,
Katherine J. Aitchison1, Peter McGun1 and Anne Farmer1
1
Institute of Psychiatry, Kings College London, UK ; 2 Centre for Psychiatric Research, Aarhus University Hospital, Risskov,
Denmark ; 3 Laboratory of Psychiatric Genetics, Department of Psychiatry, Poznan University of Medical Sciences, Poland ;
4
Central Institute of Mental Health, Division of Genetic Epidemiology in Psychiatry, Mannheim, Germany ; 5 Department of
Psychiatry, University of Bonn, Germany ; 6 Institute of Public Health, Ljubljana, Slovenia ; 7 Croatian Institute for Brain Research,
Medical School, University of Zagreb, Croatia ; 8 Universite Libre de Bruxelles, Erasme Academic Hospital, Department of
Psychiatry, Brussels, Belgium ; 9 Biological Psychiatry Unit, San Giovanni di Dio IRCCS-FBF, and UOP 23-Department of Mental
Health, Spedali Civili Hospital, Brescia, Italy ; 10 Educational and Research Institute Ozara, Ljubljana, Slovenia

Abstract
The risk of weight gain is an important determinant of the acceptability and tolerability of antidepressant
medication. To compare changes in body weight during treatment with dierent antidepressants, body
weight and height were measured at baseline and after 6, 8, 12 and 26 wk treatment with escitalopram or
nortriptyline in 630 adults with moderate-to-severe unipolar depression participating in GENDEP, a partrandomized open-label study. Weight increased signicantly more during treatment with nortriptyline
compared to escitalopram. The weight gain commenced during the rst 6 wk of nortriptyline treatment,
reached on average 1.2 kg at 12 wk (0.44-point BMI increase), and continued throughout the 6-month
follow-up period. Participants who were underweight at baseline gained most weight. Participants who
were obese at baseline did not gain more weight during treatment. Weight gain occurred irrespective of
whether weight loss was a symptom of current depressive episode and was identied as an undesired
eect of the antidepressant by most participants who gained weight. There was little weight change
during treatment with escitalopram, with an average increase of 0.14 kg (0.05-point BMI increase) over
12 wk of treatment. In conclusion, treatment with the tricyclic antidepressant nortriptyline was associated
with moderate weight gain, which cannot be explained as a reversal of symptomatic weight loss and is
usually perceived as an undesired adverse eect. While treatment with nortriptyline may be recommended in underweight subjects with typical neurovegetative symptoms, escitalopram is a suitable
alternative for subjects at risk of weight gain.
Received 9 April 2010 ; Reviewed 25 May 2010 ; Revised 4 June 2010 ; Accepted 15 July 2010 ;
First published online 18 August 2010
Key words : Antidepressants, adverse eects, body weight, major depressive disorder.

Introduction
A complex relationship exists between depression,
antidepressants and body weight. Although depressive episodes are typically associated with weight loss,
there is a positive association between depression
and obesity in the population (Luppino et al. 2010).
Address for correspondence : Dr R. Uher, P080 SGDP, Institute of
Psychiatry, 16 De Crespigny Park, SE5 8AF, London, UK.
Tel. : +44 207 848 0891 Fax : +44 207 848 0866
Email : rudolf.uher@kcl.ac.uk

Eective antidepressant medication can be expected

to reverse the weight loss caused by depression, but
weight gain is frequently reported as an undesired
eect of antidepressant drugs and a reason for discontinuing pharmacotherapy (Berken et al. 1984 ;
Cassano & Fava, 2004 ; Fava, 2000 ; Goethe et al. 2007 ;
Uher et al. 2009a). The links between depression, antidepressants and weight gain raise several questions
relevant to the clinical treatment of depression.
The rst question is whether certain antidepressants
are less prone to cause weight gain and are more

368

R. Uher et al.

suited for individuals who are overweight or at risk of

gaining weight. Weight increase has been reported as
an adverse eect of some tricyclic antidepressants and
mirtazapine (Berken et al. 1984 ; Fava, 2000 ; Paykel
et al. 1973). However, data are inconsistent on weight
changes during treatment with newer antidepressants,
such as selective serotonin reuptake inhibitors (SSRIs)
(Fava, 2000 ; Fava et al. 2000 ; Patten et al. 2009). It
remains to be established if SSRI antidepressants are
associated with signicantly less weight gain in direct
comparison with tricyclic antidepressants. The Genome
Based Therapeutic Drugs for Depression (GENDEP)
study presents an opportunity to answer this question
as it is the largest comparison of a SSRI (escitalopram)
and a tricyclic antidepressant (nortriptyline) and
weight and height measurements at baseline and several follow-ups are available for the majority of participants. In addition to testing the hypothesis that
nortriptyline is associated with more weight gain than
escitalopram, we explore the relationship between
changes in body weight and changes in depression
severity, and the role of sex and smoking status as
moderators of weight gain during treatment with
antidepressants.
A second question concerns the timing and persistence of weight changes during antidepressant treatment. While it has been reported that early reduction
in weight during the rst weeks of treatment with
uoxetine may be oset by later weight increase
(Michelson et al. 1999), the time-course of weight
changes for other antidepressants is unclear. Specically, it remains to be established whether weight gain
is limited to the initial weeks of treatment with tricyclic antidepressants or if it continues with long-term
treatment. An answer to this question may help to
inform a patients decision to continue or discontinue
an antidepressant following weight gain in the initial
weeks of treatment (Goethe et al. 2007).
A third question is whether weight gain during
antidepressant treatment should be considered as a
desirable reversal of depressive weight loss or an undesirable adverse eect (Benazzi, 1998 ; Fava, 2000).
Although this is a crucial distinction for decisions
about antidepressant treatment, to our knowledge, it
has not yet been directly examined. In the GENDEP
dataset, we address this question in three ways.
First, we investigate how much antidepressant-related
weight gain occurs in subjects who are underweight,
overweight or obese at baseline compared to those
whose weight is within the normal range. While
weight gain in individuals who are underweight can
be considered desirable, weight gain in individuals
who are already overweight or obese is likely to be

deleterious. Second, we compare weight changes

during antidepressant treatment in subjects who had
experienced weight loss and weight gain as symptoms
of depression. Preferential weight gain in subjects who
had experienced weight loss during the current depressive episode can be considered a desirable restitution of normality. On the other hand, further weight
gain in those who experienced weight gain during the
present depressive episode is clearly undesirable.
Third, we compared weight changes during antidepressant treatment in subjects who did and did not
report weight gain as side-eect related to the antidepressant. This subjective perception of weight gain
as undesirable and antidepressant-induced has arguably the strongest impact on the individuals decision
to continue or discontinue an antidepressant drug.
Method
Study design
GENDEP is a 12-wk open-label, part-randomized
multi-centre study with two active pharmacological
treatment arms (Uher et al. 2009 b). It was designed
to establish clinical and genetic determinants of
therapeutic response and adverse reactions to two
antidepressants with contrasting modes of action :
nortriptyline (a tricyclic antidepressant with strong
anity to noradrenaline transporter) and escitalopram
(a SSRI). A total of 811 adults diagnosed with ICD-10/
DSM-IV unipolar major depression of at least moderate severity established in the SCAN interview (Wing
et al. 1998) were recruited in eight European countries :
Belgium, Croatia, Denmark, Germany, Italy, Poland,
Slovenia, and UK, between July 2004 and December
2007. Recruitment was restricted to individuals of
white European parentage. Exclusion criteria were
personal or family history of bipolar disorder or
schizophrenia and current substance dependence. The
study was approved by ethics boards in all participating centres. All participants provided written consent after the procedures were explained. GENDEP
is registered at EudraCT (No.2004-001723-38, http://
eudract.emea.europa.eu) and ISRCTN (No. 03693000,
http://www.controlled-trials.com). A detailed description of the GENDEP sample and design is available
elsewhere (Uher et al. 2009b).
Interventions
Participants for whom the two antidepressants were
clinically considered to be at equipoise were randomly
allocated to receive escitalopram or nortriptyline using
a random number generator, stratied by centre and

Body weight and antidepressants

performed independently of the assessing clinician.
Patients with contraindications for one of the drugs
were allocated non-randomly to the other antidepressant. Escitalopram was initiated at 10 mg/d and increased to a target dose of 15 mg/d within the rst
2 wk, and could be further increased to 20 mg/d (and
up to 30 mg/d in cases where there was clinical
agreement that a higher dose was needed). Nortriptyline was initiated at 50 mg/d and titrated to a target
dose of 100 mg/d within the rst 2 wk, and could be
further increased to 150 mg/d (and up to 200 mg/d in
cases where there was clinical agreement that a higher
dose was needed). Other psychotropic medication
was prohibited with the exception of occasional use
of hypnotics. Compliance was monitored weekly by
self-reported pill count and plasma levels of antidepressants were measured at week 8. Of the 811
participants, 628 (77 %) completed 8 wk and 527 (65 %)
completed 12 wk on the originally allocated antidepressant. Individuals randomly allocated to nortriptyline were more likely to discontinue treatment
than individuals allocated non-randomly or those
allocated to escitalopram. Individuals treated with
escitalopram and nortriptyline improved to a similar
degree on measures of depressive symptoms (Uher
et al. 2009b).
Measures
Body weight and height were measured by calibrated
medical scales and measures at baseline assessment
prior to the initiation of study medication and after 6,
8, 12 and 26 wk of treatment. Body mass index (BMI), a
continuous measure of body weight relative to height,
was calculated as body weight in kilograms divided
by squared height in metres (kg mx2). Categorical
denitions of underweight (BMI <19), overweight
(BMI >25 to 30) and obesity (BMI >30) follow the
recommendations of the World Health Organization
(WHO, 1998) and the US National Institute of Health
Clinical Guidelines on the Identication, Evaluation
and Treatment of Overweight and Obesity in Adults
(NIH, 1998) and are consistent with previous reports
(Khan et al. 2007 ; Kloiber et al. 2007 ; Papakostas et al.
2005 ; Uher et al. 2009c).
Depression severity was measured weekly using
several rating scales, including the clinician-rated
10-item MontgomeryAsberg Depression Rating Scale
(MADRS ; Montgomery & Asberg, 1979), with high
inter-rater reliability (Uher et al. 2008). To explore the
relationship between changes in body weight and
changes in depression severity, we considered the
content overlap between weight change and MADRS,
due to the inclusion of appetite in the MADRS scale

369

(item 5). To avoid confounding due to content overlap,

we used the previously described observed mood
dimension, which contains core symptoms of depression but excludes appetite and other neurovegetative symptoms, as an index of depression severity
(Uher et al. 2008).
Adverse eects of antidepressants, including weight
gain and weight loss, were measured using the
clinician-rated UKU Side-eect Rating Scale (Lingjaerde
et al. 1987) and the self-rated Antidepressant SideEect Check list (ASEC ; Uher et al. 2009a). Both scales
record whether any adverse experience is present and
whether it is considered likely to be linked to the
medication in question. There was an excellent agreement between the two scales (Uher et al. 2009a). For
the purpose of the present investigation, the adverse
eect of weight gain/loss was recorded as positive if it
was reported on either UKU or ASEC and considered
as linked to the antidepressant. Smoking status was
assessed by two questions in the initial diagnostic
interview : Do you smoke cigarettes or other tobacco
products? (yes/no) and How many cigarettes a day
on average do you smoke?
Statistical analysis
The outcome variables were BMI at weeks 6, 8, 12 and
26. To make use of repeated weight measurements,
predictors of changes in BMI were tested using linear
mixed regression models tted with maximum likelihood as previously reported (Uher et al. 2009b). The
mixed models account for the clustering of individuals
within centres, allow use of all repeated measurements, and provide unbiased estimates in the presence
of missing data under less restrictive assumptions
than methods used for single time-point analyses
(Gueorguieva & Krystal, 2004 ; Lane, 2008). All regression models included xed eects of time (linear
and quadratic), BMI at baseline, drug and sex, and
random eects of individual and recruitment centre.
To explore the relationship between changes in depression severity and changes in body weight, body
weight was used as a time-varying covariate. Since the
treatment groups diered on baseline BMI and not all
subjects were randomized, baseline BMI was always
included as a covariate and positive results including
drug comparison were conrmed in sensitivity analyses restricted to participants who were randomly allocated to treatment. The signicance of predictors
of interest was tested using likelihood ratio tests.
Statistical signicance threshold for all analyses was
set at a=0.05, uncorrected. Statistical analyses were
performed in Stata 10 (StataCorp, 2007).

370

R. Uher et al.

Table 1. Body mass index (BMI) at baseline and on

follow-up visits by study week and treatment arm in
observed cases. The table gives the average BMI (mean),
the standard deviation (S.D.) and the number of cases with
available BMI data at each time-point
BMI
Escitalopramtreated subjects

Nortriptylinetreated subjects

Week

Mean

S.D.

Mean

S.D.

0
6
8
12
26

25.28
25.05
25.33
25.33
24.72

4.65
4.76
4.57
4.60
4.01

384
286
334
302
134

26.15
26.23
26.26
26.47
26.83

5.27
4.68
4.83
5.02
5.02

246
179
211
168
61

Results
Sample characteristics
Of the 811 GENDEP participants, 630 (78 %) had BMI
measured both at baseline and at least one follow-up
visit and were included in the analyses. BMI
measurement was available on 25 occasions per individual (mean 3.6, median 4 occasions ; Table 1). The
included subjects did not dier from the remaining
181 participants on sex, age, depression severity or
BMI at baseline (all p>0.05). The sample included
235 men and 395 women with a mean age of 42.8
(s.d.=11.6, range 1972) yr, and a mean MADRS score
of 28.4 (s.d.=6.6, range 1447). A total of 384 participants were treated with escitalopram [192 (50 %) of
them randomly allocated] and 246 with nortriptyline
[157 (64 %) of them randomly allocated].
Body-weight changes during treatment
Table 1 shows observed values of BMI at each timepoint. However, since there are a number of missing
values at each time-point, the numbers do not represent the same groups of individuals. To enable
comparison across time-points, Fig. 1 shows the timecourse across the 26 wk with missing values replaced
by the best linear unbiased estimate from drug-specic
mixed regression models which included both linear
and quadratic eects of time.
Over the rst 12 wk treatment, participants treated
with nortriptyline gained on average 1.22 kg, increasing their BMI score by 0.44 (95 % CI 0.370.51,
p<0.0001). After 6 months of treatment with nortriptyline, the average weight gained reached 1.82 kg,

increasing the BMI score by 0.64. The weight gain affected a signicant proportion of nortriptyline-treated
individuals : after 6 months treatment 95 (38.6 %)
were estimated to gain o2 kg compared to only six
(2.4 %) estimated to lose o2 kg. A small proportion of
nortriptyline-treated individuals gained a substantial
amount of weight [15 (6.1 %) gained o5 kg ; only 2
(0.8 %) lost o5 kg].
The weight of escitalopram-treated participants
remained relatively stable, with a non-signicant
average increase of 0.14 kg, or 0.05 on the BMI (95 % CI
x0.001 to 0.10, p=0.0541). After 6 months treatment
with escitalopram, an average participant gained
0.34 kg, increasing their BMI score by 0.12. After 6
months treatment, 51 (13.3 %) escitalopram-treated
participants gained o2 kg and 29 (7.6 %) lost o2 kg.
Only a small number of escitalopram-treated individuals experienced major weight changes [eight
(2.1 %) gained and 10 (2.6 %) lost o5 kg].
The dierence in weight gain between the two
antidepressants after controlling for baseline BMI
was signicant [baseline adjusted dierential weight
increase of 1.16 kg, or BMI change of 0.40 (95 %
CI 0.270.53, p<0.0001) ; Fig. 1] and was further
strengthened in a sensitivity analysis restricted to
randomly allocated participants [baseline adjusted
dierential weight increase of 1.48 kg or BMI change
0.51 (95 % CI 0.330.69, p<0.0001)]. The weight gain
did not dier signicantly by sex or smoking status
and two-way interactions of sex and smoking with
drug were non-signicant (all p>0.05).
The relationship between changes in depression
severity and changes in BMI diered by drug, reected by a signicant interaction between drug and
BMI in their eects on observed mood (p<0.0001),
which remained signicant in a sensitivity analysis
restricted to randomly allocated participants (p<
0.0001). This was due to a strong negative correlation
between the change in BMI and change in depression
severity among those treated with nortriptyline (r=
0.22, p=0.0007) but no signicant relationship between change in BMI and change in depression
severity among those treated with escitalopram
(r=0.06, p=0.2275).
Body-weight changes by weight category
The weight gain during antidepressant treatment was
unequally distributed across weight categories and
treatment arms, with a signicant interaction between
baseline BMI and drug (p=0.048). The interaction was
conrmed in a sensitivity analysis restricted to randomly allocated participants (p=0.10 due to smaller

Body weight and antidepressants

371

0.7
Nortriptyline

0.6

BMI change

0.5
0.4
0.3
0.2
0.1
Escitalopram

0
0.1
0

12
Treatment week

26

Fig. 1. Time-course of body mass index (BMI) change over 26 wk treatment by antidepressant. The y-axis represents mean
BMI change with missing values at each time-point substituted with the best linear unbiased estimate adjusted for baseline
dierences in BMI. Error bars represent 1 standard error.

sample size). Among escitalopram-treated participants, a small weight gain was limited to those whose
weight was in the normal range at baseline (BMI
change 0.14, 95 % CI 0.070.21, p<0.001 ; Fig. 2 a).
Among nortriptyline-treated participants, those who
were underweight gained the most weight, followed
by normal weight and overweight individuals. Those
who were obese at baseline tended to lose weight with
both antidepressants (Fig 2 a).
Weight change as a symptom of depression
In order to dierentiate between resolution of depressive symptoms and unintended eects of antidepressants, we explored the weight changes during
treatment in relation to symptoms of weight loss and
weight gain reported at baseline. In total, 290 (46.0 %)
participants reported weight loss and 80 (12.7 %)
reported weight gain as a symptom of depression at
the baseline assessment. Weight change during treatment with either escitalopram or nortriptyline was
unrelated to these baseline symptoms and there was
no signicant drug x symptom interaction (all p>0.1 ;
Fig. 2b).
Weight changes as adverse eects of antidepressants
Figure 2 c shows that subjective reports of weight gain
and weight loss as adverse eects of antidepressants
accurately reected actual measured weight changes.
Among nortriptyline-treated participants, 127 (51.6 %)
reported undesired weight gain linked to the use of
nortriptyline and these individuals gained on average

1.78 kg (average 0.62-point BMI increase, 95 % CI

0.510.72, p<0.0001). Of the 95 individuals who
gained o2 kg during nortriptyline treatment, 78 (82 %)
reported weight gain as an adverse eect. Among
escitalopram-treated participants, 135 (35.2%) reported
undesired weight gain linked to the use of escitalopram and these individuals gained on average 0.92 kg
(average 0.32-point BMI increase, 95 % CI 0.220.42,
p<0.0001). Of the 51 individuals who gained o2 kg
during treatment with escitalopram, 37 (73 %) reported weight gain as an adverse eect. Weight gain
was signicantly more often reported as an adverse
eect of nortriptyline than of escitalopram [x2(1)=
16.91, p<0.0001].
Reports of undesired weight loss related to the
use of antidepressants were less frequent. Among
nortriptyline-treated participants, 27 (11.0 %) reported
undesired weight loss linked to the use of nortriptyline and these individuals lost on average 1.29 kg
(average BMI change x0.44, 95 % CI x0.66 to x0.23,
p<0.0001). Of the six individuals who lost o2 kg
during nortriptyline treatment, four (67 %) reported
weight loss as an adverse eect. Among escitalopramtreated participants, 50 (13.0 %) reported undesired
weight loss linked to the use of escitalopram and these
individuals lost on average 1.30 kg (average BMI
change x0.45, 95 %CI x0.58 to x0.32, p<0.0001). Of
the 29 individuals who lost o2 kg during treatment
with escitalopram, 10 (34 %) reported weight loss as
an adverse eect. Undesired weight loss was not
related to the choice of antidepressant [x2(1)=0.57,
p=0.448].

372

R. Uher et al.

BMI change

(a) BMI change by weight category

1.4
1.2
1.0
0.8
0.6
0.4
0.2
0
0.2
0.4
Under Norm

Over Obese

Under Norm

Escitalopram

Over Obese

Nortriptyline

BMI change

(b) BMI change by depression symptoms of weight loss/gain

0.8
0.6
0.4
0.2
0
0.2
0.4
0.6
Loss

None

Gain

Loss

Escitalopram

None

Gain

Nortriptyline

BMI change

(c) BMI change by adverse effects of weight loss/gain

0.8
0.6
0.4
0.2
0
0.2
0.4
0.6
Loss

None
Escitalopram

Gain

Loss

None

Gain

Nortriptyline

Fig. 2. Body mass index (BMI) change over 12 wk of treatment with escitalopram or nortriptyline by : (a) baseline weight
category (x-axis : Under, underweight ; Norm, weight within normal range ; Over, overweight). (b) Depression symptoms of
weight-loss/gain reported at baseline (x-axis : Loss, weight loss reported as a symptom of depression at baseline ; None,
neither weight loss nor weight gain reported as symptom of depression at baseline ; Gain, weight gain reported as symptom
of depression at baseline). (c) Reports of weight gain or weight loss as adverse eects linked to antidepressants (x-axis :
Loss, weight loss reported as an adverse eect linked to the antidepressant ; None, neither weight loss nor weight gain
reported as an adverse eect linked to the antidepressant ; Gain, an adverse eect linked to the antidepressant). The y-axes
represent the best linear unbiased estimate of BMI change over 12 wk, adjusted for baseline dierences in BMI. Error bars
represent 1 standard error.

Discussion
In GENDEP, nortriptyline was associated with a signicant increase in weight in a large proportion of
men and women treated with this antidepressant.
The weight gain continued throughout the 6-month
observation period. Nortriptyline-related weight gain
was disproportionately larger in individuals with low
initial body weight, but could not be explained as a
restoration of premorbid body weight as it occurred
irrespective of whether weight loss was a feature of
the current depressive episode. Less weight change

occurred in escitalopram-treated individuals, but there

was a tendency towards slight weight increase.
The majority of individuals who experienced signicant weight change during antidepressant treatment
identied it as an undesirable eect of the antidepressant.
Although nortriptyline is considered to be a less
potent inducer of weight gain than other tricyclic
antidepressants (Fernstrom & Kupfer, 1988 ; Paradis
et al. 1992), it was associated with substantial weight
gain in a signicant proportion of treated individuals
and most of them identied the weight gain as

Body weight and antidepressants

undesirable. Not only was nortriptyline treatment associated with a higher proportion of individuals reporting drug-linked undesired weight gain, but these
individuals gained twice as much weight as those who
reported undesired weight gain linked to the use of
escitalopram. The impact of nortriptyline-induced
weight gain is moderated by the concentration of
weight increase in individuals with lower weight at
baseline and absence of further drug-induced weight
gain among individuals who were already obese.
However, since nortriptyline induced weight gain irrespective of whether weight loss had been a symptom
of depression, it cannot be considered as a curative
eect. The deleterious character of drug-induced
weight changes is underlined by subjects identifying
the changes as drug-related and undesirable. On the
other hand, nortriptyline-treated individuals who
gained more weight also experienced more improvement in their mood, measured by a scale excluding
appetite and sleep symptoms, suggesting an intricate
relationship between weight change and the therapeutic eect of tricyclic antidepressants.
Drug-induced weight gain makes treatment with
nortriptyline and other tricyclic antidepressants less
acceptable to patients and clinicians, especially for
long-term use. However, since it is more eective
against neurovegetative symptoms of depression (Uher
et al. 2009b) and its ecacy is less aected by being
underweight (Uher et al. 2009c), nortriptyline may be a
good option in depressed individuals suering from
severe insomnia, loss of appetite and weight loss
and in those who are underweight. In individuals
who require nortriptyline for treatment-resistant depression (Nierenberg et al. 2003), weight gain should
be prevented or managed using adjunct pharmacological or non-pharmacological interventions as advocated during treatment with antipsychotics (Baptista
lvarez-Jimenez et al. 2008). Through
et al. 2008 ; A
establishing predictors of individual dierences in
antidepressant-related weight gain, pharmacogenetic
studies may help avoid prescribing nortriptyline
and other weight-inducing antidepressants to individuals at highest risk (Keers et al. in press ; Secher et al.
2009).
Escitalopram was associated with little weight
change in both men and women over up to 6 months
of treatment. The non-signicant result is consistent
with escitalopram either being weight-neutral or
causing a very small weight increase. Consistent with
a previous report on uoxetine (Michelson et al. 1999),
there was a non-signicant tendency towards initial
weight loss in the rst 68 wk of escitalopram treatment, followed by a small weight increase at weeks 12

373

and 26. Although there was a longer-term tendency

towards small weight gain even among escitalopramtreated individuals, the lack of weight increase in
subjects who were obese or reported weight gain as a
symptom of depression means that the small weight
changes can be considered as restorative rather than
pathological. Still, 13 % of escitalopram-treated individuals reported weight gain as an undesired eect
of the medication. Weight gain is by no means rare
among SSRI-treated individuals with depression. In
the STAR*D study, 71 % of subjects who achieved
remission of depression with citalopram treatment
reported undesired weight increase (Nierenberg et al.
2010). The lack of weight measurement in STAR*D
precluded substantiation of the reported weight gain.
The present GENDEP data suggest that weight gain
during escitalopram treatment is on average small.
This makes escitalopram a more acceptable antidepressant, especially for long-term use. However,
when making decisions about long-term pharmacotherapy, the relatively low risk of weight gain needs
to be balanced against higher risks of other adverse
eects, such as sexual disturbance (Demyttenaere et al.
2005) and osteoporosis (Haney & Warden, 2008).
As the largest comparison of a tricyclic antidepressant and a SSRI to date, and measured weight
and height available for the majority of participants,
GENDEP presents a unique opportunity for exploring
antidepressant-induced weight changes. The random
allocation of the majority of participants to escitalopram or nortriptyline allowed a drug comparison
which is free from selection bias and overcomes
the shortcomings of naturalistic studies. Although a
proportion of subjects were non-randomly allocated
to treatment, drug comparisons proved robust in
sensitivity analyses restricted to randomly allocated
subjects. The interpretation of the present ndings is
limited by the absence of a placebo group. For this
reason, it is impossible to establish whether the small
weight gain during treatment with escitalopram reects the natural course of recovery from depression
or is a pharmacological eect of the drug.
In conjunction with reports that body weight signicantly moderated the therapeutic eect of antidepressants (Khan et al. 2007 ; Oskooilar et al. 2009 ;
Papakostas et al. 2005 ; Uher et al. 2009c), the present
results show that body weight and its changes cannot
be ignored when making clinical decisions about depression treatment. By quantifying the risk of weight
gain related to the use of nortriptyline and escitalopram, the present study can inform decisions about
initiation and continuation of these antidepressants.
Future studies are needed in order to establish if

374

R. Uher et al.

these ndings generalize to other antidepressants and

populations.

Acknowledgements
The GENDEP project was funded by the European
Commission Framework 6 grant, EC Contract Ref. :
LSHB-CT-2003-503428. Lundbeck provided nortriptyline and escitalopram for the GENDEP study.
GlaxoSmithKline and the UK National Institute for
Health Research of the Department of Health contributed to the funding of the sample collection at the
Institute of Psychiatry, London. The sponsors had no
role in the design and conduct of the study, or in data
collection, analysis, interpretation or writing of the
report.
We especially acknowledge the contribution of
Andrej Marusic and Jorge Perez, who were the principal investigators at Ljubljana, Slovenia, and at
Brescia, Italy, and who both passed away during the
conduct of the study. We also acknowledge the contribution of the following collaborators : Amanda
Elkin, Bhanu Gupta, Cerisse Gunasinghe, Desmond
Campbell, Richard J Williamson, Helen Dean, Maja
Bajs, Mara Barreto, Thomas Schulze, Christine Schmal,
Susanne Hofels, Anna Schuhmacher, Ute Pfeier,
Sandra Weber, Anne Schinkel Stamp, Alenka Tancic,
Jerneja Sveticic, Zrnka Kovacic, Pawe Kapelski, Maria
Skibinska, Aleksandra Rajewska, Anna LeszczynskaRodziewicz, Elzbieta Cegielska, Caterina Giovannini,
Cristian Bonvicini, and Luciana Rillosi.
Statement of Interest
Henigsberg and Kalember have participated in clinical
trials sponsored by pharmaceutical companies including GlaxoSmithKline and Lundbeck. Henigsberg
has received honoraria for participating on expert
panels from pharmaceutical companies including
Lundbeck. Souery is a member of a national advisory
boards for AstraZeneca, BristolMyers Squibb, Eli
Lilly and Lundbeck. Aitchison, Farmer and McGun
and have received consultancy fees and honoraria for
participating on expert panels from pharmaceutical
companies including Lundbeck and GlaxoSmithKline.
References
lvarez-Jimenez M, Hetrick SE, Gonzalez-Blanch C,
A
Gleeson JF, et al. (2008). Non-pharmacological
management of antipsychotic-induced weight gain :
systematic review and meta-analysis of randomised
controlled trials. British Journal of Psychiatry 193, 101107.

Baptista T, ElFakih Y, Uzcategui E, Sandia I, et al. (2008).

Pharmacological management of atypical antipsychoticinduced weight gain. CNS Drugs 22, 477495.
Benazzi F (1998). Weight gain in depression remitted with
antidepressants : pharmacological or recovery eect?
Psychotherapy Psychosomatics 67, 271274.
Berken GH, Weinstein DO, Stern WC (1984). Weight gain.
A side-eect of tricyclic antidepressants. Journal of
Aective Disorders 7, 133138.
Cassano P, Fava M (2004). Tolerability issues during
long-term treatment with antidepressants. Annals of
Clinical Psychiatry 16, 1525.
Demyttenaere K, Albert A, Mesters P, Dewe W, et al. (2005).
What happens with adverse events during 6 months of
treatment with selective serotonin reuptake inhibitors?
Journal of Clinical Psychiatry 66, 859863.
Fava M (2000). Weight gain and antidepressants. Journal of
Clinical Psychiatry 61 (Suppl. 11), 3741.
Fava M, Judge R, Hoog SL, Nilsson ME, et al. (2000).
Fluoxetine vs. sertraline and paroxetine in major
depressive disorder : changes in weight with long-term
treatment. Journal of Clinical Psychiatry 61, 863867.
Fernstrom MH, Kupfer DJ (1988). Antidepressant-induced
weight gain : a comparison study of four medications.
Psychiatry Research 26, 265271.
Goethe JW, Woolley SB, Cardoni AA, Woznicki BA, et al.
(2007). Selective serotonin reuptake inhibitor
discontinuation : side eects and other factors that
inuence medication adherence. Journal of Clinical
Psychopharmacology 27, 451458.
Gueorguieva R, Krystal JH (2004). Move over ANOVA :
progress in analyzing repeated-measures data and its
reection in papers published in the Archives of
General Psychiatry. Archives of General Psychiatry 61,
310317.
Haney EM, Warden SJ (2008). Skeletal eects of serotonin
(5-hydroxytryptamine) transporter inhibition : evidence
from clinical studies. Journal of Musculoskeletal and
Neuronal Interactions 8, 133145.
Keers R, Bonvicini C, Scassellati C, Uher R, et al. (in press).
Variation in GNB3 predicts response and adverse reactions
to antidepressants. Journal of Psychopharmacology.
Khan A, Schwartz KA, Kolts RL, Brown WA (2007). BMI,
sex, and antidepressant response. Journal of Aective
Disorders 99, 101106.
Kloiber S, Ising M, Reppermund S, Horstmann S, et al.
(2007). Overweight and obesity aect treatment response
in major depression. Biological Psychiatry 62, 321326.
Lane P (2008). Handling drop-out in longitudinal clinical
trials : a comparison of the LOCF and MMRM approaches.
Pharmaceutical Statistics 7, 93106.
Lingjaerde O, Ahlfors UG, Bech P, Dencker SJ, et al. (1987).
The UKU side eect rating scale. A new comprehensive
rating scale for psychotropic drugs and a cross-sectional
study of side eects in neuroleptic-treated patients. Acta
Psychiatrica Scandinavica 334 (Suppl.), 1100.
Luppino FS, de Wit LM, Bouvy PF, Stijnen T, et al. (2010).
Overweight, obesity, and depression : a systematic review

Body weight and antidepressants

and meta-analysis of longitudinal studies. Archives of
General Psychiatry 67, 220229.
Michelson D, Amsterdam JD, Quitkin FM, Reimherr FW,
et al. (1999). Changes in weight during a 1-year trial of
uoxetine. American Journal of Psychiatry 156, 11701176.
Montgomery SA, Asberg M (1979). A new depression scale
designed to be sensitive to change. British Journal of
Psychiatry 134, 382389.
Nierenberg AA, Husain MM, Trivedi MH, Fava M, et al.
(2010). Residual symptoms after remission of major
depressive disorder with citalopram and risk of relapse :
a STAR*D report. Psychological Medicine 40, 4150.
Nierenberg AA, Papakostas GI, Petersen T, Kelly KE, et al.
(2003). Nortriptyline for treatment-resistant depression.
Journal of Clinical Psychiatry 64, 3539.
NIH (1998). Clinical guidelines on the identication,
evaluation, and treatment of overweight and obesity in
adults the evidence report. National Institutes of Health.
Obesity Research 6 (Suppl. 2), 51S209S.
Oskooilar N, Wilcox CS, Tong ML, Grosz DE (2009). Body
mass index and response to antidepressants in depressed
research subjects. Journal of Clinical Psychiatry 70,
16091610.
Papakostas GI, Petersen T, Iosifescu DV, Burns AM, et al.
(2005). Obesity among outpatients with major depressive
disorder. International Journal of Neuropsychopharmacology 8,
5963.
Paradis CF, Stack JA, George CJ, Miller MD, et al.
(1992). Nortriptyline and weight change in depressed
patients over 60. Journal of Clinical Psychopharmacology 12,
246250.
Patten SB, Williams JV, Lavorato DH, Brown L, et al.
(2009). Major depression, antidepressant medication and

375

the risk of obesity. Psychotherapy Psychosomatics 78,

182186.
Paykel ES, Mueller PS, De l V (1973). Amitriptyline, weight
gain and carbohydrate craving : a side eect. British Journal
of Psychiatry 123, 501507.
Secher A, Bukh J, Bock C, Koefoed P, et al. (2009).
Antidepressive-drug-induced bodyweight gain is
associated with polymorphisms in genes coding for COMT
and TPH1. International Clinical Psychopharmacology 24,
199203.
StataCorp (2007). Stata Statistical Software : Release 10.
Stata Corp LP : College Station, TX.
Uher R, Farmer A, Henigsberg N, Rietschel M, et al. (2009a).
Adverse reactions to antidepressants. British Journal of
Psychiatry 195, 202210.
Uher R, Farmer A, Maier W, Rietschel M, et al. (2008).
Measuring depression : comparison and integration of
three scales in the GENDEP study. Psychological Medicine
38, 289300.
Uher R, Maier W, Hauser J, Marusic A, et al. (2009b).
Dierential ecacy of escitalopram and nortriptyline on
dimensional measures of depression. British Journal of
Psychiatry 194, 252259.
Uher R, Mors O, Hauser J, Rietschel M, et al. (2009c). Body
weight as a predictor of antidepressant ecacy in the
GENDEP project. Journal of Aective Disorders 118, 147154.
WHO (1998). Obesity : preventing and managing the global
epidemic. In Report of a World Health Organisation
Consultation on Obesity, Geneva, 35 June 1997. World
Health Organisation : Geneva.
Wing JK, Sartorius N, Ustin TB (1998). Diagnosis and Clinical
Measurement in Psychiatry. A Reference Manual for SCAN.
World Health Organization : Geneva.