Psychoneuroendocrinology (2009) 34, 939—946

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / p s y n e u e n

Passionate love and relationship thinkers:

Experimental evidence for acute cortisol
elevations in women
Timothy J. Loving *, Erin E. Crockett, Aubri A. Paxson

Department of Human Development and Family Sciences, The University of Texas at Austin, 1 University Station, Box A2700, Austin,
TX 78712, United States

Received 22 September 2008; received in revised form 2 January 2009; accepted 22 January 2009

KEYWORDS Summary We assessed the impact of an individual difference variable, relationship-focused

Passionate love; thinking, on women’s acute salivary cortisol responses during and after a guided imagery task.
Cortisol; Specifically, 29 healthy women, all of whom were experiencing high levels of passionate love, but
Relationship thinking; varied on levels of relationship-focused thinking, were assigned to one of two experimental
Guided imagery; conditions: a partner reflection condition or a cross-sex friend reflection condition. Results
Falling in love; indicated that women experiencing passionate love evidenced increased cortisol levels when
Women asked to reflect on their romantic partner and relationship relative to women asked to reflect on a
cross-sex friendship, but this difference was particularly pronounced and relatively long-lasting
for those women characterized by a high amount of relationship-focused thinking. Our study
significantly expands extant work on the passionate love—cortisol link by isolating the impact of a
specific psychological variable, relationship-focused thinking, on the physiological experience of
falling in love. We believe our work highlights the advances that can be made when established
work in the close relationships and neuroendocrine fields are integrated.
# 2009 Elsevier Ltd. All rights reserved.

The early stages of romantic attraction and romantic rela-
tionship development are defined by an array of physiologi-
cal, psychological, and behavioral consequences (Carter,
1998; Fisher, 2000). This experience, commonly referred
to as early-stage intense romantic love (Aron et al., 2005;
Brand et al., 2007), falling in love, or passionate love, is
hypothesized to have biological substrates that evolved as a
core feature of the human mating system. Several lines of
research support this hypothesis. For example, results from
* Corresponding author. Tel.: +1 512.471.0973;
fax: +1 512 475 8662.
E-mail address: tjloving@mail.utexas.edu (T.J. Loving).
fMRI studies indicate that the experience of passionate love
activates portions of the brain responsible for motivational
drive-states (Aron et al., 2005; Bartels and Zeki, 2000).
Further, passionate love is associated with changes in circu-
lating levels of neurotrophins and stress hormones. Indivi-
duals who indicate that they think about their partner at
least four hours a day and score high on a self-report measure
of passionate love show increased levels of the neurotrophin
nerve growth factor (NGF) relative to individuals neither
romantically involved nor in long-term established romantic
relationships (Emanuele et al., 2005). Importantly, upregula-
tion of NGF can induce activation of the hypothalamic—
pituitary—adrenocortical (HPA) axis of the endocrine system

0306-4530/$ — see front matter # 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.psyneuen.2009.01.010
940
(Angelucci, 1994), which ultimately stimulates the adrenal
cortex to secrete cortisol, the human body’s primary gluco-
corticoid (Rabin, 1999). A more recent study suggests that
this biological cascade may indeed occur. Individuals experi-
encing passionate love display significantly greater circulat-
ing levels of cortisol relative to control groups (Marazziti and
Canale, 2004).
Although these studies, and others, have expanded knowl-
edge considerably of passionate love’s physiological conse-
quences, extant work has ignored potentially key moderating
variables that could impact passionate love’s physical
effects. The experience of passionate love is a significant
interpersonal transition that requires individuals to adapt
affectively (Nieder and Seiffge-Krenke, 2001), behaviorally
(Kelley and Thibaut, 1978), and cognitively (Aron et al.,
1995). Thus, any variable that affects individuals’ psycholo-
gical experience of falling in love may alter the physiological
outcomes associated with this life transition.
Past work on this topic is built on the assumption that the
number of hours individuals spend thinking about their part-
ners is an important component of the passionate love
experience (Emanuele et al., 2005; Marazziti and Canale,
2004). We chose to utilize a validated measure of relation-
ship-focused cognition to determine if, indeed, individual
differences in how much people think about their partners
and romances alter HPA-axis activity. Broadly speaking, a
host of individual dispositions moderate physiological reac-
tivity and health outcomes (Contrada and Guyll, 2001; Ouell-
ette and DiPlacido, 2001); however, the extent to which
individual psychological differences may actually moderate
the biological impact of passionate love has not received
empirical attention. Given passionate love’s effects on
neural processes and attentiveness to environmental stimuli
(Aron et al., 2005; Bianchi-Demicheli et al., 2006), we
reasoned that individuals who demonstrate a tendency to
think more about their relationships would be particularly
susceptible to increased cortisol levels when in the throes of
a passionate relationship. In other words, people who are
cognitively more relationship-focused should show signs of
being more affected by the experience of passionate love
than individuals who are less relationship-focused.
To test this hypothesis, we employed an experimental
design, which enabled us to rule out the effect of any
unmeasured, potentially confounding variables. Specifically,
we assessed acute shifts in salivary cortisol levels after
participants, all of whom were experiencing passionate love,
were asked to either reflect on their romantic partner or an
opposite-sex friend. The use of an opposite-sex friend served
as a control condition and follows the method employed in
recent fMRI research on passionate love (Aron et al., 2005).
Further, we focused our investigation on women because of
the relatively profound role one-on-one relationships play in
women’s lives relative to men’s (Gabriel and Gardner, 1999;
Kiecolt-Glaser and Newton, 2001).
1. Methods
1.1. Participants
Twenty-nine women participated in a 2-h laboratory study.
Participants had to be involved in a nonmarital romantic
relationship of no more than 1-year duration and be in
T.J. Loving et al.
generally good health. The mean age of participants was
19.79 years (SD = 1.72; range = 18—25). The majority (52%) of
participants self-identified as ‘White’ (24% Asian; 14% Latino
or Hispanic; 10% other). Median relationship length was 17
weeks (M = 20.48, SD = 13.77).
1.2. Measures
1.2.1. Screening
Interested participants completed an online screening inter-
view. Participants were excluded if they reported that (a)
they were involved in a nonmarital relationship of more than
1-year duration, (b) they were not able to visit the laboratory
during the week between the hours of 2 p.m. and 6 p.m., or
(c) they would not be able to provide a picture of their
partner and a friend. We also excluded participants reporting
behaviors or conditions with known HPA-axis implications,
including smoking, taking certain medications, having a his-
tory of hormone problems, having depression or anxiety,
being currently pregnant, or working night shifts. Although
the use of hormonal contraceptives are known to influence
cortisol responses (Kirschbaum et al., 1999), we chose to not
exclude women if they indicated taking a daily-dose or multi-
phasic form of birth control; rather, we included birth control
use as a covariate in analyses (n = 11). This approach was
motivated by our desire to assess cortisol reactivity in a
sample of healthy, ‘normal’ individuals involved in a romantic
relationship. Given the prevalence of birth control use
amongst nonmarried women (Mosher et al., 2004), we
believe excluding these women would have significantly
reduced the generalizability of our findings.
Interested participants were also asked to complete a
15-item version of Hatfield and Sprecher’s (1986) 30-item
Passionate Love Scale (PLS). All women were presented
with the last 15 items of the published measure (e.g.,
‘‘Sometimes my body trembles with excitement at the sight
of my partner’’; 1 (untrue) to 6 (true); a = 0.76; possible
range = 15—90). Consistent with past work (Emanuele
et al., 2005), only women scoring in the upper one third
of the scale range were invited to participate in the study
(M = 78.76, SD = 6.56; range = 67—90).
1.2.2. Passionate love–—Second assessment
We believed it prudent to control for a more proximal
measure of passionate love in our analysis to ensure that
any observed effects for relationship-focused thinking
(described below) were independent of individuals’ passio-
nate love experiences. Thus, the evening prior to their
participation in the laboratory component of the study, we
asked participants to complete the remaining 15 items of the
PLS (a = 0.82; M = 76.52, SD = 7.39).
1.2.3. Relationship-focused thinking
To assess individuals’ tendencies to think about romantic
relationships, participants completed Cate and colleagues’
(1995) 20-item relationship thinking scale (e.g., ‘‘I find
myself at times drifting off and thinking about my relation-
ship with my partner’’; 1 (extremely unlike me) to 5 (extre-
mely like me); a = 0.81; M = 78.76, SD = 9.75). Relationship-
focused thinking was not significantly associated with the
night before measure of passionate love, r(29) = 0.30,
p > 0.10.
Passionate love and cortisol reactivity
1.2.4. Salivary cortisol
To assess circulating levels of free cortisol, all participants
were asked to provide a series of saliva samples via the
Salivette (Sarstedt, Germany). Altogether, participants pro-
vided five saliva samples at standard intervals (approxi-
mately every 15 min; see below). To reduce participant
reactivity to the sampling procedure, the first sample served
as a practice sample and was discarded. Cortisol concentra-
tions, reported in mg/dL, were determined via SalimetricsLLC
expanded range high sensitivity salivary cortisol enzyme
immunoassay kit. As per kit instructions, all samples were
frozen at 20 8C until assayed. Each participant’s samples
were assayed in the same batch and high and low control
samples provided by SalimetricsLLC were included in each
batch to ensure reliability. As is standard practice, obtained
cortisol values were subjected to a log10 transformation prior
to analyses to normalize their distribution (e.g., Bower et al.,
2005; Diamond et al., 2008; Smyth et al., 1997).
1.3. Procedure
1.3.1. Recruitment
Participants responded to either an online ad (i.e., face-
book.com) or flyers posted around the campus of a large
university in the southwestern United States soliciting parti-
cipants who were ‘‘madly, deeply in love.’’ Prospective
participants were directed to a secure website where they
were presented with detailed information about the study
including the study consent form. Interested participants
completed an online screening form (described above); eli-
gible participants were then scheduled to visit the first
author’s laboratory. All participants were asked to bring
(or e-mail us ahead of time) a photograph of their romantic
partner and a ‘‘friend who is the same sex and about the same
age as your dating partner, but someone with whom you are
not romantically involved.’’ Study sessions were conducted in
the afternoon between the hours of 2 p.m. and 6 p.m. to
control for diurnal patterns in cortisol secretion (Kirschbaum
and Hellhammer, 1994); participants were instructed to not
eat or drink anything at least 1 h prior to participation.
Participants were paid US$ 20 for completion of the 2-h
study. All procedures were approved by the university’s
Institutional Review Board.
1.3.2. Night before survey
The evening prior to their scheduled laboratory session,
participants were asked to complete a brief online survey
that assessed the second portion of the PLS, relationship-
focused thinking, and additional measures unrelated to the
current analysis. We did not assess these relationship-focused
measures during the laboratory session so as to prevent
contaminating the thoughts of participants assigned to the
‘friend’ condition (see below).
1.3.3. Laboratory session and experimental
manipulation
Upon arrival to the laboratory, consent was collected and
participants were provided with basic instructions regarding
the collection of salivary samples. At this point, the ‘prac-
tice’ saliva sample was administered. Participants were then
asked to complete a brief survey and sit quietly for 15 min
until the collection of the second saliva sample (i.e., base-
941
line). Participants were given the option of looking through
an emotionally neutral picture book while they waited for the
15 min to pass.
Next, participants were guided through a brief relaxation
and guided imagery exercise. The experimenter first induced
a state of relaxation in participants, following standard
relaxation exercise protocols (e.g., deep breathing, eyes
closed). Depending upon condition assignment, the experi-
menter then instructed participants to picture the face of
either their partner or friend ‘‘and try to visualize all the
details about him/her.’’ (Prior to this point in the procedure,
participants were not given any indication of why they were
asked to provide a picture of their partner and a friend.) This
began the guided imagery portion of the protocol, which
continued for approximately 5 min (contact the first author
for a detailed script of the procedure). Participants were
asked to recall how they met their partner (or friend), when
they first realized they were in love with their partner (or first
realized they wanted to be friends with their friend), things
they enjoy doing together, things they like about their part-
ner (or friend), and their feelings for their partner (or friend).
At no point did the verbal prompts in either condition refer-
ence a physical relationship (or lack thereof). We henceforth
refer to the randomly assigned experimental conditions as
the partner condition (n = 14) and friend condition (n = 15).
Importantly, as expected based on the random assignment to
conditions, there were no differences between conditions on
average participant levels of passionate love, relationship
length, or relationship-focused thinking (all ps > 0.20).
Upon conclusion of the guided imagery protocol, partici-
pants were asked to talk into a tape recorder for 3 min ‘‘about
all the things going through your head.’’ Participants were then
left alone with a digital recorder and the picture of their
partner/friend (used to keep the image of the partner or
friend clear in their minds). After 3 min, the experimenter
returned, collected the recorder, and then instructed parti-
cipants to write about their partner/friend for 10 min. Alto-
gether, participants underwent approximately 5 min of guided
imagery and an additional 13 min of partner/friend reflection.
Participants then provided a third saliva sample (timed to tap
individual reactions to the guided imagery manipulation).
Next, participants completed an additional questionnaire
and sat quietly for 15 min. The fourth saliva sample was then
collected (i.e., recovery), followed by an additional ques-
tionnaire and ‘quiet time’ for 15 min. The fifth saliva sample
was collected after an additional 15 min. If participants
completed any of the questionnaires early during the
15 min periods, they were provided with a book of landscapes
without any text. Further, participants were asked to refrain
from any other activities during the recovery period.
Although the content of the remaining questionnaires is
unrelated to the current analyses, it is important to note
that none of the laboratory questionnaires contained ques-
tions about participants’ friendships or romances (i.e., friend
and partner information was collected the night prior to the
laboratory session). Following collection of the final saliva
sample, participants were debriefed and paid.
2. Results
We tested our hypothesis with multilevel statistical models
created in HLM (version 6.06) for Windows, allowing us to
942 T.J. Loving et al.

account for cortisol measurements nested within individuals.
We also assessed the significance of a number of potential
covariates, including health behaviors known to affect cor-
tisol (i.e., age, birth control use, waking up time that day,
and alcohol consumption), as well as relationship length and
night before PLS. Only age, relationship length, and night
before PLS were significant control variables in analyses;
these variables were retained in the final model. No other
potential covariates were significant and were subsequently
dropped from analyses.
Mean cortisol values prior to transformation, and irrespec-
tive of study condition, were as follows (SD in parentheses):
Baseline = 0.145 (0.080); Guided Imagery = 0.156 (0.079);
Recovery = 0.143 (0.072); Recovery + 15 = 0.130 (0.081). Des-
criptive plots of average cortisol levels at each measurement
indicated that cortisol levels changed nonlinearly over time.
Thus, in addition to a time component, we included a quadratic
time variable (time2) in the model. Although time was not a
significant predictor of within person change (b = 0.040,
p = 0.12), time2 was a significant predictor (b = 0.020, p =
0.02). As such, both terms were maintained in the final model.
To test our hypothesis, we also included an interaction
term between relationship-focused thinking and condition
(i.e., partner vs. friend), as well as their main effects.
The within-individual level of the analysis allowed each
participant’s cortisol level to be modeled as a function of
time and the quadratic of time. The equation was as follows:
Ci j ¼ b0i þ b1i T i j þ b2i Ti2j þ ei j
where Cij is the cortisol level for individual i at time j; Tij is
the time of the cortisol measure; Ti2j is the quadratic com-
ponent (time squared) and eij is a residual component specific
to individual i at time j. The coefficient b0i is the regression
intercept for individual i and represents cortisol level at
baseline.
The between-individual level of the analysis estimated
the random effects for the within-individual variables
and allowed us to test whether individuals’ cortisol
levels differed as a function of relationship-focused
thinking and condition while controlling for covariates.
The between-individual equation for the intercept was
as follows:
b0i ¼ g 00 þ g 01 RLi þ g 02 RTi þ g 03 Ai þ g 04 PLSi þ u0i
This model specifies that the intercept for individual i
was a function of a grand mean intercept that is common
across individuals (g00), adjusted to reflect the individual’s
relationship length (g01RLi), levels of relationship-focused
thinking (g02RTi), age (g03Ai), night before levels of pas-
sionate love (g04PLSi) and a residual component specific
to individual i. The between-individual equations for
the effects of time and time2 were similar, but included
two additional parameters: condition (g05Ci) and the inter-
action between relationship-focused thinking and condi-
tion (g06Ii). Condition and the interaction between
condition and relationship-focused thinking were excluded
from the intercept because baseline measures of cortisol
were taken prior to experimental manipulation. The

Table 1 HLM results: effect of control variables, relationship-focused thinking, and experimental condition on salivary cortisol
levels.

b SE t df p-Value
Intercept
Intercept 1.745 0.468 3.73 23 0.001
Relationship length 0.009 0.004 2.22 23 0.037
age 0.051 0.032 1.58 23 0.128
Night before PLS 0.006 0.007 0.87 23 0.394
RT 0.006 0.005 1.38 23 0.182
Time
Intercept 0.157 0.376 0.42 21 0.680
Relationship length 0.003 0.001 1.87 21 0.076
Age 0.018 0.012 1.44 21 0.165
Night before PLS 0.008 0.004 2.12 21 0.046
RT 0.013 0.003 4.52 21 0.000
Condition a 0.912 0.301 3.03 21 0.007
RT  condition a 0.011 0.004 3.03 21 0.007
Time squared
Intercept 0.032 0.102 0.32 21 0.755
Relationship length 0.001 0.000 1.31 21 0.206
Age 0.008 0.004 1.80 21 0.086
Night before PLS 0.003 0.001 2.98 21 0.008
RT 0.005 0.001 4.74 21 0.000
Condition a 0.377 0.100 3.79 21 0.001
RT  condition a 0.005 0.001 3.88 21 0.001
PLS, Passionate Love Scale; RT, Relationship thinking.
a
1 = Partner, 0 = Friend
Passionate love and cortisol reactivity
between-individual equations for time and time2 were as
follows:
b0i ¼ g 00 þ g 01 RLi þ g 02 RTi þ g 03 Ai þ g 04 PLSi þ g 05 Ci
þ g 06 Ii þ u0i
Results are presented in Table 1. Consistent with extant
work (Marazziti and Canale, 2004), relationship length was a
significant predictor of baseline cortisol (i.e., the intercept),
such that women in relationships of a shorter duration had
higher baseline cortisol levels relative to women in longer
relationships.
In addition to the significant controls noted above, there
was a main effect of condition and relationship-focused
thinking on changes in cortisol over time. As hypothesized,
these main effects were qualified by a significant condi-
tion  relationship-focused thinking interaction. Predicted
levels of cortisol at each time point as a function of relation-
ship-focused thinking (1 SD) and condition are displayed in
Fig. 1. As can be seen, cortisol levels of high relationship-
focused thinking women in the partner condition increased
following the guided imagery exercise and continued to
increase through the last saliva sample collected 45 min after
the start of the manipulation. In contrast, cortisol levels of
high relationship-focused thinking women in the friend con-
dition and low relationship-focused thinking women in the
friend condition generally declined throughout the proce-
dure. Finally, low relationship-focused thinking women in the
partner condition displayed an initial rise in cortisol during
the guided imagery, but then steadily declined following the
conclusion of the guided imagery exercise. Thus, women
currently experiencing high levels of passionate love evi-
dence increased cortisol levels when asked to reflect on their
romantic partner and relationship relative to women cur-
rently experiencing high levels of passionate love but asked
to reflect on a cross-sex friendship, but this difference in
cortisol responses is particularly pronounced and relatively
Figure 1 Women’s cortisol over time as a function of experimental condition (partner vs. friend) and relationship-focused thinking
(high vs. low; 1 SD).
943
long-lasting for those women who tend to engage in a high
amount of relationship-focused thinking.
3. Discussion
The present study extends recent work on the hormonal
consequences of falling in love (i.e., passionate love) by
demonstrating that women who engage in more relation-
ship-focused thinking experience acute increases in cortisol
after simply being asked to reflect on their romantic relation-
ship and partner. Interestingly, women with a lower tendency
to engage in relationship-focused thinking also showed an
initial rise in cortisol after reflecting on their partners, but
the effect diminished shortly after our experimental manip-
ulation concluded. These findings suggest that the HPA-axis
activation seen in those experiencing passionate love is
pronounced for some types of individuals relative to others.
Our work extends Marazziti and Canale’s (2004) recent
study in which they found increased circulating levels of
cortisol when individuals were experiencing early-stage pas-
sionate love relative to controls. Marazziti and Canale relied
on a single blood sample for assessment of plasma cortisol;
however, their study essentially involved a ‘partner condi-
tion’ for all participants. Thus, it is difficult to ascertain
whether individuals experiencing passionate love demon-
strate generalized hyperactivation of the HPA axis or simply
showed cortisol elevations in response to the study protocol.
In our study, although visual inspection of the predicted
means (Fig. 1) could suggest that high relationship-focused
thinking women may experience chronic increases in cortisol
during early-stage passionate love, analyses indicate that
relationship-focused thinking was not a significant predictor
of baseline levels of salivary cortisol. Thus, we are inclined to
believe that the observed elevated cortisol levels in both
studies are apparent when individuals are focused on their
romances, and that some individuals focus more on these
relationships than others. Indeed, when we asked more
944
relationship-focused women to reflect on their relationship
with a friend, their cortisol levels dropped throughout
the study procedure, suggesting it is possible to ‘deactivate’
the physiological consequences associated with the passio-
nate love experience.
It is important to consider the implications of these
findings for women, who on average tend to be highly
relationship-focused, independent of their passionate love
state (Gabriel and Gardner, 1999). Do the increased cortisol
responses we observed for high relationship-focused women
imply that these individuals are more stressed when falling
in love?
There are several reasons to think that this is not the case.
All available evidence suggests that transitions into romantic
relationships are associated with positive outcomes (Lee and
Gramotnev, 2007), and that falling in love may procure
particularly protective benefits for individuals (Esch and
Stefano, 2005). Events such as ‘‘starting a love relationship’’
and ‘‘begin to date’’ are considered positive life events in the
stress and health literature (Reich and Zautra, 1981).
Although the transition into a new romance does require
adaptation on the part of the individual, any potential
negative effects of this process are likely offset by the
positive aspects of enjoyment that accompany positive life
events (Davidson et al., 2006). Indeed, individuals in passio-
nate love show increased positive emotions and favorable
sleep patterns (Brand et al., 2007) and more quickly recog-
nize positive feeling words when primed with their partners’
names (Bianchi-Demicheli et al., 2006). Additionally, life
events individuals perceive as being ‘good’ (versus ‘bad’)
have a beneficial effect on individual outcomes, including
physical health (e.g., lymphocyte proliferation; Snyder
et al., 1993). In other words, although both positive and
negative life events may produce physiological arousal typi-
cally construed as ‘‘stress’’ (Rietveld and van Beest, 2007),
whether that physiological arousal results in positive or
negative health outcomes depends on the subjective inter-
pretation of the event.
From a physiological perspective, mild to moderate phy-
siological arousal is associated with, if not required for,
social bonding (Diamond, 2001), and other outcomes that
are concomitant with bonding actually reduce stress
responses (e.g., production of oxytocin; Diamond, 2001).
Passionate love also increases the prevalence of positive
emotions (Kim and Hatfield, 2004); positive emotions pro-
tect individuals from negative physical and mental health
outcomes by facilitating adaptation to stressful situations
(Cohen and Hoberman, 1983; Folkman et al., 2007). Finally,
as noted in the introduction, NGF levels also rise during the
early developmental phase of romantic relationships. NGF
upregulates the HPA axis (Angelucci, 1994), suggesting the
increased cortisol levels may reflect a byproduct of
increased NGF (Emanuele et al., 2005). Future work should
explore this meditational path.
Although we suspect our pattern of responses do not
wholly reflect a traditional stress response, per se, we cannot
fully rule out such an interpretation. Recently, researchers
have elucidated several mechanisms responsible for the HPA-
axis reactivity observed during the Trier Social Stress Test
(TSST), a widely used, standardized acute stress task (Kirsch-
baum et al., 1993); the identified mechanisms may provide
additional insight into our findings. Specifically, the HPA axis
T.J. Loving et al.
is particularly responsive when individuals experience or
suspect social rejection, such as is the case during the TSST,
a task characterized by a high degree of ego involvement
(Dickerson and Kemeny, 2004; Kajantie and Phillips, 2006). It
is possible that priming women, who appear to be especially
reactive to stressors containing a social rejection compo-
nent, to focus on their passionate love experience renders
more salient concerns of relationship dissolution. Put another
way, perceptions of uncontrollability are key predictors of
HPA-axis responses, especially when individual goals are
threatened and attainment of these goals is not exclusively
under the volitional control of the individual (Dickerson and
Kemeny, 2004). The subjective passionate love experience
may occur at the individual level, but the progression of any
newly initiated romance to a more stable, committed form is
dependent on the thoughts, behaviors, and actions of both
individuals involved in the romance (Loving et al., in press).
In other words, by having participants think about falling in
love, we may have inadvertently caused them to think about
the future of their budding romances, a future that is char-
acterized by a high degree of uncontrollability given the
interdependence that defines romantic relationships (Kelley
and Thibaut, 1978). Altogether, it is clear that multiple
mechanisms may underlie the hormonal consequences of
falling in love; gaining a more complete understanding of
how and why passionate love results in physiological out-
comes requires that researchers consider the individual- and
couple-level characteristics that define the experience.
The implications of our results need to be balanced
against some study limitations. Our study sample was rela-
tively small, although in line with past work on this topic, and
the observed difference between salivary cortisol responses
for women in the partner versus friend conditions suggests
strong effect sizes. Additionally, because our analyses were
limited to women, we cannot speak to whether relationship-
focused thinking functions in a similar manner across sexes,
but it is important to note that no sex differences have been
found in past work (e.g., Aron et al., 2005; Marazziti and
Canale, 2004). Further, although our experimental design
helps to rule out potential confounding variables that may
affect results between conditions, the design does not rule
out potentially confounding variables within conditions, or, in
our case, constructs that may covary with our relationship-
focused thinking moderator. For example, we excluded par-
ticipants reporting a current or past history of anxiety or
depression, but we did not assess obsessive-compulsive ten-
dencies (e.g., the Yale—Brown Obsessive Compulsive Scale;
Goodman et al., 1989), as has been done in prior work
(Marazziti and Canale, 2004). It is possible that our measure
of relationship-focused thinking may share some conceptual
overlap with obsessionality. Although we believe the fact
that women scoring higher on relationship-focused thinking
demonstrated declines in cortisol when asked to focus on a
nonromantic, opposite-sex friend minimizes this concern,
future work would benefit from being able to definitively
rule out this possibility. Perhaps more plausibly, the relation-
ship-focused thinking measure may reflect underlying
genetic factors linked to the experience of passionate love
and formation of social attachments (Emanuele et al., 2007;
Gillath et al., 2008). Given the relevance of the attachment
system to the initiation and early development of romantic
relationships (e.g., Eastwick and Finkel, 2008), it is possible
Passionate love and cortisol reactivity
that genetic factors at least partly account for individual
differences in relationship-focused thinking.
These limitations notwithstanding, we believe a notable
strength of our study is the employment of an imaginary task
within an experimental design setting. Having participants
recall specific social relationships and/or features of those
relationships provides researchers with an opportunity to
study how ‘real’ (vs. hypothetical or laboratory induced)
relationships may affect biological outcomes (e.g., Bloor
et al., 2004). Importantly, the administration of such tasks
is relatively simple. By manipulating what individuals think
about during the study session, we were able to develop a
more nuanced understanding of how individual differences
(i.e., relationship-focused thinking) influences the physiolo-
gical outcomes associated with passionate love. This same
design has the potential to provide researchers the ability to
study an unlimited range of individual difference variables
that may affect the physiological experiences associated
with major life transitions. For example, the impact of past
relationship experiences (e.g., prior exposure to falling in
love), orientations towards relationships (e.g., adult attach-
ment, implicit relationship theories; Diamond et al., 2008;
Knee et al., 2003), or specific characteristics of the romances
(e.g., self-expansion; Aron and Aron, 1997; Lewandowski
et al., 2006) could all be studied with this design.
In conclusion, our study significantly expands extant work
on the passionate love—cortisol link by isolating the impact of
a specific psychological variable, relationship-focused think-
ing, on the physiological experience of falling in love. We
believe our work highlights the advances that can be made
when established work in the close relationships and neu-
roendocrine fields are integrated (Loving et al., 2006). As the
two fields continue to merge, our understanding of how
relationship initiation affects romantic partners physiologi-
cal and health outcomes will benefit tremendously.
Role of funding source
Funding for this study was provided by a research grant from
the University of Texas at Austin to the first author. The
University had no further role in study design; in the collection,
analysis and interpretation of data; in the writing of the
report; and in the decision to submit the paper for publication.
Conflict of interest statement
There are no conflicts of interest.
Acknowledgements
We are thankful to Alex Hsieh, Shena Sandle, Ashley Senegal,
and Sung Sim for their assistance in data collection. We also
thank Eshkol Rafaeli for his early input on the guided imagery
procedure.
References
Angelucci, L., 1994. The hypothalamus—pituitary—adrenocortical
axis: epigenetic determinants, changes with aging, involvement
of NGF. Neurochemistry International 25 (1), 53—59.
945
Aron, A., Aron, E.N., 1997. Self-expansion motivation and including
other in the self. In: Duck, S. (Ed.), Handbook of personal
relationships: Theory, research and interventions. 2nd ed. John
Wiley & Sons, Inc., pp. 251—270.
Aron, A., Fisher, H., Mashek, D.J., Strong, G., Li, H., Brown, L.L.,
2005. Reward, motivation, and emotion systems associated with
early-stage intense romantic love. Journal of Neurophysiology 94
(1), 327—337.
Aron, A., Paris, M., Aron, E.N., 1995. Falling in love: prospective
studies of self-concept change. Journal of Personality and Social
Psychology 69 (6), 1102—1112.
Bartels, A., Zeki, S., 2000. The neural basis of romantic love.
Neuroreport: For Rapid Communication of Neuroscience Research
11 (17), 3829—3834.
Bianchi-Demicheli, F., Grafton, S.T., Ortigue, S., 2006. The power of
love on the human brain. Social Neuroscience 1 (2), 90—103.
Bloor, L.E., Uchino, B.N., Hicks, A., Smith, T.W., 2004. Social rela-
tionships and physiological function: the effects of recalling social
relationships on cardiovascular reactivity. Annals of Behavioral
Medicine 28 (1), 29—38.
Bower, J.E., Ganz, P.A., Dickerson, S.S., Petersen, L., Aziz, N., Fahey,
J.L., 2005. Diurnal cortisol rhythm and fatigue in breast cancer
survivors. Psychoneuroendocrinology 30 (1), 92—100.
Brand, S., Luethi, M., von Planta, A., Hatzinger, M., Holsboer-
Trachsler, E., 2007. Romantic love, hypomania, and sleep pattern
in adolescents. Journal of Adolescent Health 41 (1), 69—76.
Carter, S.C., 1998. Neuroendocrine perspectives on social attach-
ment and love. Psychoneuroendocrinology 23 (8), 779—818.
Cate, R.M., Koval, J., Lloyd, S.A., Wilson, G., 1995. Assessment of
relationship thinking in dating relationships. Personal Relation-
ships 2 (2), 77—95.
Cohen, S., Hoberman, H.M., 1983. Positive events and social supports
as buffers of life change stress. Journal of Applied Social Psychol-
ogy 13 (2), 99—125.
Contrada, R.J., Guyll, M., 2001. On who gets sick and why: the role of
personality and stress. In: Baum, A., Revenson, T.A., Singer, J.E.
(Eds.), Handbook of Health Psychology. Lawrence Erlbaum Associ-
ates, Inc., Mahwah, NJ, pp. 59—84.
Davidson, L., Shahar, G., Lawless, M.S., Sells, D., Tondora, J., 2006.
Play, pleasure, and other positive life events: ‘‘non-specific’’
factors in recovery from mental illness? Psychiatry: Interpersonal
and Biological Processes 69 (2), 151—163.
Diamond, L.M., 2001. Contributions of psychophysiology to research
on adult attachment: review and recommendations. Personality
and Social Psychology Review 5 (4), 276—295.
Diamond, L.M., Hicks, A.M., Otter-Henderson, K.D., 2008. Every time
you go away: changes in affect, behavior, and physiology asso-
ciated with travel-related separations from romantic partners.
Journal of Personality and Social Psychology 95 (2), 385—403.
Dickerson, S.S., Kemeny, M.E., 2004. Acute stressors and cortisol
responses: a theoretical integration and synthesis of laboratory
research. Psychological Bulletin 130 (3), 355—391.
Eastwick, P.W., Finkel, E.J., 2008. The attachment system in fledg-
ling relationships: an activating role for attachment anxiety.
Journal of Personality and Social Psychology 95 (3), 628—647.
Emanuele, E., Brondino, N., Pesenti, S., Re, S., Geroldi, D., 2007.
Genetic loading on human loving styles. Neuroendocrinology
Letters 28 (6), 815—821.
Emanuele, E., Politi, P., Bianchi, M., Minoretti, P., Bertona, M.,
Geroldi, D., 2005. Raised plasma nerve growth factor levels
associated with early-stage romantic love. Psychoneuroendocri-
nology 20, 1—7.
Esch, T., Stefano, G.B., 2005. Love promotes health. Neuroendocri-
nology Letters 26 (3), 264—267.
Fisher, H., 2000. Lust, attraction, attachment: biology and evolu-
tion of the three primary emotion systems for mating, reproduc-
tion, and parenting. Journal of Sex Education & Therapy 25 (1),
96—104.
946
Folkman, S., Moskowitz, J.T., Hewstone, M., Schut, H.A.W., De Wit,
J.B.F., Van Den Bos, K., et al., 2007. Positive affect and meaning-
focused coping during significant psychological stress. In: The
Scope of Social Psychology: Theory and Applications, Psychology
Press, New York, pp. 193—208.
Gabriel, S., Gardner, W.L., 1999. Are there ‘his’ and ‘hers’ types of
interdependence? The implications of gender differences in col-
lective versus relational interdependence for affect, behavior,
and cognition. Journal of Personality and Social Psychology 77
(3), 642—655.
Gillath, O., Shaver, P.R., Baek, J.-M., Chun, D.S., 2008. Genetic
correlates of adult attachment style. Personality and Social
Psychology Bulletin 34 (10), 1396—1405.
Goodman, W.K., Price, L.H., Rasmussen, S.A., Mazure, C., 1989. The
Yale—Brown Obsessive Compulsive Scale: I. Development, use, and
reliability. Archives of General Psychiatry 46 (11), 1006—1011.
Hatfield, E., Sprecher, S., 1986. Measuring passionate love in inti-
mate relationships. Journal of Adolescence 9 (4), 383—410.
Kajantie, E., Phillips, D.I.W., 2006. The effects of sex and hormonal
status on the physiological response to acute psychosocial stress.
Psychoneuroendocrinology 31 (2), 151—178.
Kelley, H.H., Thibaut, J.W., 1978. Interpersonal Relations: A Theory
of Interdependence. Wiley-Interscience, New York.
Kiecolt-Glaser, J.K., Newton, T.L., 2001. Marriage and health: his and
hers. Psychological Bulletin 127 (4), 472—503.
Kim, J., Hatfield, E., 2004. Love types and subjective well-being: a cross
cultural study. Social Behavior and Personality 32 (2), 173—182.
Kirschbaum, C., Hellhammer, D.H., 1994. Salivary cortisol in psy-
choneuroendocrine research: recent developments and applica-
tions. Psychoneuroendocrinology 19 (4), 313—333.
Kirschbaum, C., Kudielka, B.M., Gaab, J., Schommer, N.C., Hellham-
mer, D.H., 1999. Impact of gender, menstrual cycle phase, and
oral contraceptives on the activity of the hypothalamic—pitui-
tary—adrenal axis. Psychosomatic Medicine 61 (2), 154—162.
Kirschbaum, C., Pirke, K.M., Hellhammer, D.H., 1993. The ‘Trier
Social Stress Test’–—a tool for investigating psychobiological stress
responses in a laboratory setting. Neuropsychobiology 28, 76—81.
Knee, C.R., Patrick, H., Lonsbary, C., 2003. Implicit theories of
relationships: orientations toward evaluation and cultivation.
Personality and Social Psychology Review 7 (1), 41—55.
T.J. Loving et al.
Lee, C., Gramotnev, H., 2007. Life transitions and mental health in a
national cohort of young Australian women. Developmental Psy-
chology 43 (4), 877—888.
Lewandowski, G.W.J., Aron, A., Bassis, S., Kunak, J., 2006. Losing a
self-expanding relationship: implications for the self-concept.
Personal Relationships 13 (3), 317—331.
Loving, T.J., Gleason, M.E.J., Pope, M.T., in press. Transition novelty
moderates daters’ cortisol responses when talking about mar-
riage. Personal Relationships.
Loving, T.J., Heffner, K.L., Kiecolt-Glaser, J.K., 2006. Physiology
and interpersonal relationships. In: Vangelisti, A., Perlman, D.
(Eds.), Cambridge Handbook of Personal Relationships. Cam-
bridge University Press, New York, pp. 385—405.
Marazziti, D., Canale, D., 2004. Hormonal changes when falling in
love. Psychoneuroendocrinology 29, 931—936.
Mosher, W.D., Martinez, G.M., Chandra, A., Abma, J.C., Willson, S.J.,
2004. Use of contraception and use of family planning services in
the United States: 1982—2002. Advance Data 350, 1—36.
Nieder, T., Seiffge-Krenke, I., 2001. Coping with stress in different
phases of romantic development. Journal of Adolescence 24 (3),
297—311.
Ouellette, S.C., DiPlacido, J., 2001. Personality’s role in the protec-
tion and enhancement of health: where the research has been,
where it is stuck, how it might move. In: Baum, A., Revenson,
T.A., Singer, J.E. (Eds.), Handbook of Health Psychology. Lawr-
ence Erlbaum Associates, Inc, Mahwah, NJ, pp. 175—193.
Rabin, B.S., 1999. Stress, Immune Function and Health: The Con-
nection. Wiley-Liss & Sons, New York.
Reich, J.W., Zautra, A., 1981. Life events and personal causation:
some relationships with satisfaction and distress. Journal of
Personality and Social Psychology 41 (5), 1002—1012.
Rietveld, S., van Beest, I., 2007. Rollercoaster asthma: when positive
emotional stress interferes with dyspnea perception. Behaviour
Research and Therapy 45 (5), 977—987.
Smyth, J.M., Ockenfels, M.C., Gorin, A.A., Catley, D., Porter, L.S.,
Kirschbaum, C., et al., 1997. Individual differences in the diurnal
cycle of cortisol. Psychoneuroendocrinology 22 (2), 89—105.
Snyder, B.K., Roghmann, K.J., Sigal, L.H., 1993. Stress and psycho-
social factors: effects on primary cellular immune response.
Journal of Behavioral Medicine 16 (2), 143—161.