lable at ScienceDirect

Journal of Psychiatric Research xxx (2014) 1e8

Contents lists avai

Journal of Psychiatric Research

journal homepage: www.elsevier .com/locate/psychires

Effects of social exclusion on emotions and oxytocin and cortisol levelsin patients with chronic depression

Andrea Jobst a, Lena Sabass a, Anja Palagyi a, Christine Bauriedl-Schmidt a,Maria Christine Mauer a, Nina Sarubin a, Anna Buchheim c, Babette Renneberg b,Peter Falkai a, Peter Zill a, Frank Padberg a, *

a Department of Psychiatry und Psychotherapy, Ludwig Maximilian University, Munich, Germanyb Department of Psychology, Free University of Berlin, Berlin, Germanyc Department of Psychology, Clinical Psychology, University of Innsbruck, Austria

a r t i c l e i n f o

Article history:Received 7 August 2014Received in revised form24 October 2014Accepted 1 November 2014

Keywords:OxytocinChronic depressionSocial exclusionCyberballOstracismSocial painSocial stress

* Corresponding author. Tel.: þ49 89 440053358; fE-mail address: padberg@med.uni-muenchen.de (

http://dx.doi.org/10.1016/j.jpsychires.2014.11.0010022-3956/© 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Jobst A, etdepression, Journal of Psychiatric Research

a b s t r a c t

Objective: Patients with chronic depression (CD) experience a high burden of disease, severe co-morbidity, and increased mortality. Although interpersonal dysfunction is a hallmark of CD, the under-lying mechanisms are largely unexplored. Oxytocin (OT) has been proposed to play a crucial role in thesocial deficits of mental disorders and has been found to be dysregulated after social exclusion (ostra-cism) in patients with borderline personality disorder. This study investigated how social exclusion af-fects emotions, OT levels, and cortisol (CT) levels in CD patients.Method: Twenty-one patients diagnosed with CD and 21 healthy controls (HC) matched for gender, age,and education underwent repeated neuroendocrine measurements in a standardized laboratory settingwhile playing Cyberball, a virtual ball-tossing game that mimics a social exclusion situation. Emotionalreactions, plasma OT and cortisol levels were assessed at baseline and 5, 15, and 40 min after Cyberball.Results: At baseline, there were no group differences in OT levels. Immediately after playing Cyberball,plasma OT levels showed divergent changes in CD patients and HC; the difference in direction of changewas significant with a reduction in CD patients compared to HC (p ¼ .035*); CT levels did not differbetween groups at any time point, but decreased over time. Patients showed more threatened emotionalneeds and increased negative emotions, especially anger and resentment, and showed higher sensitivityto ambiguous threat of social exclusion than healthy controls.Conclusions: CD patients react to ostracism with pronounced negative emotions. The reduction in OTlevels in CD patients after social exclusion may contribute to their interpersonal dysfunction and theirdifficulty in coping adequately with aversive social cues.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Chronic depression (CD) is a severe mental disorder that de-velops in about 20%e30% of patients with major depressive dis-order (MDD) (Angst et al., 2009; Kessler et al., 1994). CD can becharacterized by interpersonal problems that may result fromdisturbed attachment and parenting as well as interpersonaltrauma during childhood (Wiersma et al., 2009). This interper-sonal dysfunction plays a major role in sustaining a depressivestate and has been the focus of novel psychotherapeutic ap-proaches, e.g. the cognitive behavioral system of psychotherapy

ax: þ49 89 440053930.F. Padberg).

al., Effects of social exclusion(2014), http://dx.doi.org/10.10

(CBASP) (Keller et al., 2000). However, little is known about theneurobiological background of impaired interpersonal function inCD.

Theoretically, changes in interpersonal function, i.e. disturbedaffiliative behaviors, impaired reading of social cues, impairedestablishment of trust, reduced capacity for attachment, andinability to cope with social stressors, may be related to alteredregulation of the oxytocin (OT) system (Eckstein and Hurlemann,2013; Herpertz and Bertsch, 2014; Meyer-Lindenberg et al., 2011;Stanley and Siever, 2010; Striepens et al., 2011). OT, a hypothalamicneuropeptide, stimulates milk ejection and uterine contraction atparturition in mammals and has been linked to prosocial behaviorand social approach in humans (Carter, 1998; Kosfeld et al., 2005).High peripheral levels of plasma OT are associated with trust and

on emotions and oxytocin and cortisol levels in patients with chronic16/j.jpsychires.2014.11.001

A. Jobst et al. / Journal of Psychiatric Research xxx (2014) 1e82

trustworthiness (Zak et al., 2005), and plasma OT levels can beincreased by the warm touch of others (Holt-Lunstad et al., 2008).OT modulates the formation of social memories (Domes et al.,2012) and is involved in reading mental states of other in-dividuals (mentalizing) (Domes et al., 2007). Moreover, it plays amajor role in maternal care as well as parental and pair bonding,e.g. Ditzen and coworkers demonstrated an increase of positivecommunication behavior in relation to negative behavior togetherwith a reduction of salivary cortisol levels after intranasal OTadministration during a standard conflict discussion in heterosex-ual couples (Ditzen et al., 2009). Another study demonstrated apositive association between peripheral OT levels and positivecommunication in couples using a structured social support inter-action task (Gouin et al., 2010).

OT has been shown to exhibit stress-attenuating functions in theTrier Social Stress Test (TSST) (Heinrichs et al., 2003), possiblymediated by a suppressive effect on the HPA axis, as has beenshown in male rats (Neumann et al., 2000). Central OT releaseunder stress has been observed in animals, accompanied bysecretion of OT into the peripheral circulation (Nishioka et al., 1998;Wotjak et al., 1998). In humans, however, findings regarding theinteraction of stress and OT release are inconsistent. Heinrichs andcolleagues did not observe OT release in response to the TSST inbreastfeeding women (Heinrichs et al., 2001), whereas Pierre-humbert and coworkers found a clear OT response during the TSSTin people who had been exposed to traumatizing events such asabuse or potentially fatal illnesses during childhood or adolescence(Pierrehumbert et al., 2010). In addition, this response was relatedto the respective attachment representations (Pierrehumbert et al.,2012).

Dysregulation of the OT system has been suggested as a puta-tive mechanism that contributes to interpersonal dysfunctions inconditions with social deficits, such as borderline personality dis-order (BPD), autistic disorder, and schizophrenia (Goldman,Marlow-O'Connor, Torres and Carter, 2008; Green et al., 2001;Stanley and Siever, 2010). Women with BPD had lower plasmaOT levels than healthy controls, and their plasma OT levels werenegatively correlated with a history of childhood trauma (Bertschet al., 2012). Moreover, BPD patients showed an altered regula-tion of peripheral OT levels after social exclusion (Jobst et al., 2014),which may underlie the deficits in repairing broken cooperation(King-Casas et al., 2008). There is recent evidence of an altered OTregulation in MDD: Salivary OT levels were found to be reduced inmothers with CD and their children compared to non-depressedmothers (Apter-Levy et al., 2013). Similarly, plasma OT levelswere lower in women with MDD than in healthy controls (Yuenet al., 2014). However, to date OT has not been investigated inCD, even though interpersonal deficits are particularly pronouncedin this disorder.

Rejection and feeling isolated from others can be experimentallymimicked with social exclusion (ostracism) paradigms, e.g. theCyberball paradigm. In the Cyberball paradigm, participants areexcluded from a computer-based virtual ball-tossing game(Williams and Jarvis, 2006). Using this paradigm, Renneberg andcolleagues observed higher scores of negative emotions (anger)after exclusion in BPD patients than in healthy controls. BPD pa-tients felt excluded even when they were not, and the exclusionresulted in dysfunctional behavioral intentions (Renneberg et al.,2012; Staebler et al., 2011).

To our knowledge, so far no studies have addressed the effects ofsocial exclusion on the regulation of emotions, OT and CT levels inpatients with CD. Consequently, this pilot study used the Cyberballsocial exclusion paradigm to investigate the effects of socialexclusion on these variables in a group of CD patients and healthycontrols.

Please cite this article in press as: Jobst A, et al., Effects of social exclusiondepression, Journal of Psychiatric Research (2014), http://dx.doi.org/10.10

2. Methods and materials

2.1. Participants

Twenty-one inpatients (15 men, 6 women, mean age45.79 ± 15.55, range: 20e72 yrs) with a diagnosis of CD and 21healthy controls (15 men, 6 women, mean age 46.43 ± 14.37),matched for gender, age and education, were included in the study.The patients were recruited at the Department of Psychiatry andPsychotherapy, Ludwig Maximilian University of Munich. CD di-agnoses and comorbid axis I and II diagnoses were assessed withthe German version of the Structural Clinical Interview for DSM-IV(First et al., 1995) (SCID-I-Screening, SCIDeIIeInterview); all pa-tients met the diagnostic criteria for dysthymia and depressiveepisode lasting longer than 2 years. Individuals with a diagnosis ofsubstance dependence, schizophrenia, schizoaffective disorder, orbipolar disorder were excluded. The mean frequency of comorbiddiagnoses on SCID-II was 1.42 (SD 1.61); patients had the followingco-morbid personality disorders: Depressive (N ¼ 5), avoidant(N¼ 4), obsessive-compulsive (N¼ 4), paranoid (N¼ 2), dependent(N ¼ 2), antisocial (N ¼ 2), negativistic (N ¼ 1), schizoid (N ¼ 1),narcissistic (N ¼ 1), borderline (N ¼ 1), and histrionic (N ¼ 1). TheHC were recruited by advertisements in newspapers and on noti-ceboards; they did not have a current psychiatric disorder, hadnever received psychiatric treatment and had not received anypsychological treatment in the past 10 years, as confirmed with theSCID-I and eII screening instruments (First et al., 1995).

Most of the CD patients were receiving pharmacological treat-ment, i.e. antidepressants (N ¼ 17), second generation antipsy-chotics (SGA, N ¼ 10), or mood stabilizers (N ¼ 9). At the time ofassessment, 8 patients were taking benzodiazepines.

The studywas approved by the Institutional Review Board of theFaculty of Medicine, Ludwig Maximilian University, Munich. Allparticipants provided written informed consent. The healthy con-trols received V70 financial compensation.

2.2. Procedure and measures

Plasma OT and CT levels were measured in all participantsbefore and after a social exclusion paradigm that used the stan-dardized virtual ball-tossing game Cyberball to evoke social stress(Williams and Jarvis, 2006). Participants had to visit the laboratorythree times. At the first visit, we checked inclusion and exclusioncriteria and obtained informed consent. At the second visit, par-ticipants completed the following psychometric questionnaires:Beck Depression Inventory II (Beck et al., 1996), 24-item HamiltonDepression Rating Scale (Hamilton, 1960), the German version ofthe short form of the Childhood Trauma Questionnaire (CTQ-SF)(Bernstein et al., 2003; Klinitzke et al., 2012), and the RejectionSensitivity Questionnaire (RSQ) (Downey and Feldman, 1996). Atthe third session, participants played Cyberball in a standardizedexperimental setting, as previously reported (Jobst et al., 2014).Before starting the game, an intravenous catheter was inserted intoa forearm vein to allow through-the-wall blood drawings via a longcatheter that ran through a soundproof lock to an adjacent labo-ratory. Ten to 15 min after inserting the catheter (during whichtime the participants completed questionnaires on their currentemotional state), a blood sample was collected for baseline mea-surements, and the Cyberball game was started on a computerscreen positioned in front of the participants. During the experi-ment, participants sat in a comfortable chair and had no visualcontact with the investigators. The Cyberball paradigm includestwo virtual players who are controlled by the software (Williamsand Jarvis, 2006). An explanation on the computer screen tellsthe participants that they will take part in a virtual game with two

on emotions and oxytocin and cortisol levels in patients with chronic16/j.jpsychires.2014.11.001

A. Jobst et al. / Journal of Psychiatric Research xxx (2014) 1e8 3

other real players to simulate their mental visualization skills. Inour study, participants played one round of the game, which lastedabout three minutes. They got the ball 3 times (10% of the throws)at the beginning, but afterwards theywere excluded from the gamewithout being told why and were only able to watch the otherplayers playing. After Cyberball, blood samples were collected forpost exclusion measurements (see Fig. 1).

2.3. Oxytocin and cortisol measurements

OT and CT plasma levels were measured as previously reported(Jobst et al., 2014). In the group of women, special attention waspaid to themenstrual cycle, because of neuroendocrine interactionswith OT. Participants taking hormonal contraception were evalu-ated between the 3rd and 18th day of the intake period, andwomen with a menstrual cycle who were not taking hormonalcontraception were evaluated during the follicular phase, i.e. be-tween the 5th and 12th day of themenstrual cycle, because gonadalhormones are more stable during this period (Salonia et al., 2005).Five female participants were in menopause (CD group: n ¼ 2; HCgroup: n ¼ 3). All assessments took place between 8 and 11 a.m. inorder to control for the circadian change in hormone levels. Dataindicate that OT levels are more stable in the morning: A diurnalfluctuation of peripheral OT levels has been found, with concen-trations reaching a nadir in the late afternoon and a peak at 2 p.m.(Forsling et al., 1998). Participants were told not to eat or drink onehour before the time of their appointment.

Blood samples for OT and CT were collected at four time points:1 min before (t0) the Cyberball game started and 5 min (t1), 15 min(t2) and 40min (t3) after the game (see Fig.1). Tubeswere preparedwith aprotinin (500 IU/ml) to prevent OT degradation. Afterwards,the tubes were centrifuged at 1.600� g for 15 min and storedat�80 �C until the biochemical analyses. Analyses of plasma OTandplasma CT were performed by the Neurochemical Laboratory at theDepartment of Psychiatry, University of Munich. Serum CT levelswere determined with an immunoassay analyzer (Elecsys Cortisol

Fig. 1. The Cyberball game and experimental design. Cyberball is an experimental ball-tossfrom the game partway through without being told why. Before and after Cyberball, particlevels were measured at four time points.

Please cite this article in press as: Jobst A, et al., Effects of social exclusiondepression, Journal of Psychiatric Research (2014), http://dx.doi.org/10.10

Test; Roche Diagnostics, Mannheim, Germany) according to themanufacturer's instruction. OT was assessed with a commerciallyavailable ELISA Kit (Enzo Life Sciences, Germany). Measurementswere performed in duplicates. The interassay coefficient of varia-tion was below 17%, and the intra-assay coefficient of variation was15%.

2.4. Emotional reactions

Emotional reactions were assessed by self-rating questionnairesbefore and after Cyberball (see flowchart in Fig. 1). We used a 14-item emotion scale (Gross and Levenson, 1995), modified for theCyberball paradigm (Staebler et al., 2009), consisting of thefollowing items: Amusement, affection, contentment, surprise,loneliness, hurt, despair, sadness, fear, contempt, anger, pride,boredom, and resentment. The emotions were divided into threesub-scales: Positive emotions (affection, contentment, amusement,surprise, pride), self-focused negative emotions (fear, sadness, hurt,loneliness, despair), and other-focused negative emotions (anger,resentment, contempt, boredom). In addition, we asked the par-ticipants to assess aversive inner tension, expressed as a percentageof maximal tension. After the game, we used the NeedseThreatQuestionnaire (NTQ) (Williams et al., 2000) to assess the extent towhich the participants felt that their needs (belonging, self-esteem,meaningful existence, control) were threatened by the game.Moreover, participants had to rate estimated percentage of tossesreceived and the extent to which they felt disregarded andexcluded. The aversive impact experienced during the game wascalculated from items of the NTQ.

2.5. Statistics

We used SPSS software version 17 for statistical analyses. De-mographic and psychometric analyses were performed with chi-square tests (CTQ) and independent samples t-tests (age, educa-tion, BDI, HDRS). We used repeated measure analysis of variance

ing game played on a computer screen with virtual partners. Participants are excludedipants completed a battery of self-rating questionnaires. Oxytocin and cortisol plasma

on emotions and oxytocin and cortisol levels in patients with chronic16/j.jpsychires.2014.11.001

A. Jobst et al. / Journal of Psychiatric Research xxx (2014) 1e84

(rmANOVA) and independent sample t-tests to analyze outcomeswithin the Cyberball game and rmANOVA for comparisons ofplasma OT and plasma CT levels between both groups. We thenapplied paired and independent sample post hoc t-tests. OT, CT, andpsychometric values were compared by Pearson and Spearmancorrelation. Baseline OT plasma measurements showed a largeinter-individual range of 162.4 pg/ml to 1809.4 pg/ml. Two patients(both men) were excluded from the statistical analyses, becausetheir OT levels were greater than three standard deviations fromthe mean of the whole sample (outliers between 1809.4 and1691.1 ng/ml); the matched HC were excluded as well. Thisapproach was chosen instead of log transforming our data as suchextremely elevated OT levels may indicate pathological conditionsnot necessarily related to CD and the two outliers can thus not beconsidered as representative scores drawn from the population weintended to sample. Data were normally distributed (Kolmogor-oveSmirnov test); data samples for OT ranged from 162.4 pg/ml to1183.3 pg/ml (mean ¼ 577.98, SD ¼ 341.39). CT plasma levelsranged from 180.1 nmol/l to 688.2 nmol/l (mean ¼ 439.66,SD ¼ 148.79) and were normally distributed (Kolmogor-oveSmirnov test).

3. Results

3.1. Sample characteristics

CD patients and HC did not differ significantly with regard to ageand education. As expected, depression scores were significantlyhigher in patients than in HC (Table 1). The CD group had higherCTQ scores for emotional abuse and emotional neglect (moderatetrauma as defined by Bernstein and Fink (Bernstein & Fink, 1998)than the HC, indicating early emotional childhood trauma, but notfor physical abuse, physical neglect, or sexual abuse. The controlgroup scored below cut-off values on all CTQ subscales (Table 1).Patients also had significantly higher scores of rejection sensitivitythan controls, as assessed by the RSQ (Table 1).

3.2. Psychological measures

Analysis of inner tension showed a significant effect for group(rmANOVA: F1,35 ¼ 5.33; p ¼ .027*), but not for time (F1,35 ¼ 0.26;

Table 1Sample characteristics of patients with chronic depression (CD) and healthy controls(HC). No significant differences were found between patient and control group inage, gender and education, but significant differences were found in psychometricdata.

Characteristics Group; mean (SD) Test statistic p value

CD (n ¼ 19) HC (n ¼ 19)

Age, yrs 45.79 (15.56) 46.58 (14.40) t36 ¼ 0.162 .872Gender, n 13 male

6 female13 male6 female

X21 ¼ 0.000 1.000

Education, yrs 11.58 (1.54) 11.63 (1.67) t36 ¼ 0.101 .920BDI-II 25.95 (9.72) 2.47 (2.99) t21.37 ¼ 10.06 >.001**

HAMD-24 20.74 (8.38) 1.37 (2.03) t20.11 ¼ 9.79 >.001**

CTQEmotional abuse 11.05 (4.80) 6.0 (1.45) Z ¼ 3.737 >.001**

Physical abuse 6.74 (3.16) 5.42 (0.77) Z ¼ 1.479 .201Sexual abuse 6.05 (4.40) 4.95 (0.52) Z ¼ 0.874 .583Emotional neglect 12.95 (4.47) 8.63 (3.53) Z ¼ 2.974 .001*

Physical neglect 7.16 (1.77) 6.53 (2.25) Z ¼ 1.373 .191RSQ 12.69 (5.52) 6.42 (3.62) t33 ¼ 4.033 >.001**

BDI-II¼ Beck Depression Inventory; HAMD-24¼Hamilton Depression Rating Scale;RSQ¼ Rejection Sensitivity Questionnaire; SD¼ standard deviation. * ¼ significanceaccording to level of significance set at a ¼ 0.05, ** ¼ significance according to levelof significance set at a ¼ 0.01.

Please cite this article in press as: Jobst A, et al., Effects of social exclusiondepression, Journal of Psychiatric Research (2014), http://dx.doi.org/10.10

p ¼ .612) or time � group interaction (F1,35 ¼ 0.80; p ¼ .378). Atbaseline, patients tended to score higher than HC in the assessmentof inner tension (t35¼1.94; p¼ .060); this differencewas significantafter playing Cyberball (t35 ¼ 2.46; p¼ .019*), but playing Cyberballdid not significantly change the amount of inner tension in eithergroup (Table 2).

All participants reported disregard and exclusion whenanswering the NTQ index questions (mean ratings > 3). Both pa-tients and controls judged ball ownership realistically and feltdisregarded and excluded to a similar extent. However, patientshad a significantly higher rejection sensitivity (RSQ: t33 ¼ 4.63;p � .001*), and the exclusion paradigm had a significantly higheraversive impact on patients than on HC (NTQ: Aversive impactfactor: t31 ¼ 2.77; p ¼ .009*). CD patients experienced more nega-tive emotions during the game (NTQ: negative mood: t34 ¼ 2.93;p ¼ .006*), and their needs tended to be more threatened byexclusion than were those of the HC, as need fulfillment was lowerin the CD group (NTQ: needs: t34 ¼ 1.84; p ¼ .075). All subscales ofthe NTQ are represented within Table 2.

At baseline, levels of all negative emotions were significantlyhigher in patients than in HC (t36 ¼ 5.37; p � .001*) and levels ofpositive emotions were significantly lower (t36 ¼ 5.21; p � .001*).We conducted rmANOVAs on the factors self-focused negativeemotions, other-focused negative emotions, and positive emotionsto analyze change of emotions by Cyberball: At both time points(before and after the game) patients showed significantly highernegative emotions focused on others (significant group effect:F1,35 ¼ 8.30; p ¼ .007*) and significantly higher self-focused nega-tive emotions (significant group effect: F1,35 ¼ 24.93; p � .001*) aswell as lower positive emotions (significant group effect:F1,35 ¼ 10.78; p ¼ .002*) compared to HC. Other-focused negativeemotions, self-focused negative emotions and positive emotionswere not significantly influenced by Cyberball (no significant timeor time � group interactions). However, among other-focusednegative emotions, anger and resentment increased in patients,but not in HC (significant group � time interaction effects:F1,35 ¼ 5.20; p ¼ .029*).

The psychological reaction to social exclusion (NTQ) correlatedsignificantly with clinical characteristics in the CD group: Positivecorrelations were found between the severity of depression (BDI)and the extent of subjective exclusion (r ¼ .52; p ¼ .039*), theaversive impact of feeling excluded (r¼ .64; p¼ .015*), and negativemood (r ¼ .69; p ¼ .002*), and a negative correlation was foundbetween the BDI and the fulfillment of needs (r ¼ �.49; p ¼ .046*)experienced during Cyberball.

3.3. Plasma levels of oxytocin and cortisol before and afterCyberball

OT plasma levels showed a marked variation between subjects,but intra-individual levels were rather stable. Mean OT levels areshown in Fig. 2. To analyze OT and CT measurements over time, weperformed rmANOVAs for the four time points (t0et3). For OTlevels (Fig. 2A), a significant time � group interactionwas observed(F2.64,94.96 ¼ 3.135; p ¼ .035*, rmANOVA), but no significant effectsfor time (F2.64,94.96¼ .806; p¼ .480) or group (F1,36¼ .590; p¼ .447).An independent sample t-test revealed divergent changes in OTlevels from t0 to t2 between patients and HC (t36 ¼ 2.757;p ¼ .009*). Baseline OT levels did not correlate with age, severity ofclinical symptoms (HAMD-24, BDI-II, CTQ, RSQ), measures ofemotional reactions, or the day of the menstrual cycle and did notdiffer significantly between patients and HC in either women ormen. Within the CD group, women showed significantly lower OTlevels than men (t15.5 ¼ 2.224; p¼ .041*), which was not the case incontrol group.

on emotions and oxytocin and cortisol levels in patients with chronic16/j.jpsychires.2014.11.001

Table 2Emotional reactions to Cyberball in patients with chronic depression (CD) and healthy controls (HC).

CD (n ¼ 19) HC (n ¼ 19) Between-groupcomparison

Before After p value Before After p value p value p valueD pre-post

mean (SD) mean (SD)

Inner tension 45.74 (27.23) 46.67 (23.83) .700 29.47 (23.39) 26.05 (26.85) .423 <.059 pre<.029*post

.378

Emotion ScalePositive emotions 2.60 (1.21) 2.43 (1.10) .381 3.95 (1.28) 3.67 (1.45) .128 <.002*pre

.005*post.668

Self-focused negative emotions 3.21 (11.61) 3.18 (1.57) .809 1.39 (0.75) 1.17 (0.46) .032* <.001*pre<.001*post

.261

Other- focused negative emotions 2.10 (1.21) 2.29 (1.33) .226 1.38 (0.52) 1.28 (0.43) .306 .038*pre<.006*post

.121

Anger/resentment 2.06 (1.43) 2.42 (1.55) .154 1.47 (0.86) 1.16 (0.34) .090 .141pre.001*post

.029*

Needs-threat scale CD HC p value

“I felt disregarded” 3.69 (0.87) 3.21 (1.36) .219“I felt excluded” 3.81 (1.122) 3.11 (1.45) .132“I owned the ball” 10.84% (7.17) 9.47% (5.31) .521Aversive impact factor 28.21 (56.01) 21.45 (7.48) .009*Negative mood 24.71 (6.95) 18.05 (6.66) .006*Needs overall fulfillment 9.39 (3.10) 11.42 (3.49) .075Belonging 2.76 (1.28) 3.12 (1.21) .380Self-esteem 2.44 (0.99) 3.07 (0.93) .057Meaningful existence 3.04 (1.18) 3.47 (1.33) .380Control 1.53 (0.54) 1.75 (0.79) .333

CD¼ chronic depression; SD¼ standard deviation. * ¼ significance according to level of significance set at a¼ 0.05, **¼ significance according to level of significance set at a¼0.01.

A. Jobst et al. / Journal of Psychiatric Research xxx (2014) 1e8 5

CT levels (Fig. 2B) were analyzed over all four time points byrmANOVAs. There was a significant effect for time(F1.44,51.88 ¼ 19.506; p � .001*), with a decrease of CT levels in bothpatient and HC group. No significant time � group effect(F1.44,51.88 ¼ 0.126; p ¼ .814) and no significant group effect(F1,36 ¼ 0.110; p ¼ .742) could be observed. Correlation analysesshowed that cortisol levels were not correlated with age, BDI-II,HAMD-24, RSQ, or CTQ subscales.

OT plasma levels did not correlate with CT levels in either groupor at any time point.

4. Discussion

The main finding of this study was a differential regulation ofemotions and OT plasma levels in CD patients and HC after social

A

0 5 10 15 20 25 30 35 40 45

350

400

450

500

550

600

Time [min]

Plas

ma

leve

ls [p

g/m

l]

Oxytocin

Cyb

erba

ll

Fig. 2. A) Peripheral oxytocin and B) cortisol levels over time. A repeated measurements ANbetween patients with chronic depression (CD) and healthy controls (HC) (F2.64,94.96 ¼ 3.135;p ¼ <.001*).

Please cite this article in press as: Jobst A, et al., Effects of social exclusiondepression, Journal of Psychiatric Research (2014), http://dx.doi.org/10.10

exclusion (ostracism). Immediately after playing Cyberball, plasmaOT levels were lower in patients than at baseline, but higher inhealthy controls; the difference in the direction of the change wassignificant (p ¼ .035*). CT levels did not differ between the groupsat any timepoint. Patients' emotional needs were more threatenedby the exclusion paradigm (NTQ), and anger and resentmentincreased (emotion scale). To our knowledge, this is the first studyto investigate emotional reaction and OT plasma levels in CD duringa social exclusion paradigm.

The findings of the present study resemble our previous resultsin BPD (Jobst et al., 2014), i.e. changes in OT plasma levels in CDpatients are similar to those in patients with BPD. There is growingevidence of altered OT regulation in BPD patients (Bertsch et al.,2013, 2012), which potentially contributes to their deficit inrepairing social conflict (King-Casas et al., 2008). CD patients often

B

0 5 10 15 20 25 30 35 40 45

Time [min]

350

400

450

500

CD

HC

Plas

ma

leve

ls [n

mol

/l]

Cortisol

Cyb

erba

ll

OVA over all time points showed a significant difference for the development over timep ¼ .035*). Cortisol decreased significantly over time in both groups (F1.44,51.88 ¼ 19.506;

on emotions and oxytocin and cortisol levels in patients with chronic16/j.jpsychires.2014.11.001

A. Jobst et al. / Journal of Psychiatric Research xxx (2014) 1e86

show severe interpersonal dysfunction (Zobel et al., 2010), a sig-nificant number of early adverse life events (Nemeroff et al., 2003),and a disease course that results in chronic impairment in psy-chosocial dimensions (Judd et al., 2000).

In humans, one function of OT is to maintain and strengthensocial bonding (Francis et al., 2000). Previous studies clearly sup-port this crucial role in affiliation and prosocial behavior (Feldman,2012; Feldman et al., 2007). Intranasally administered OT signifi-cantly increases the experience of attachment security in insecurelyattached adults (Buchheim et al., 2009). In contrast, the Cyberballexclusion condition can be regarded as a negative social stimulusthat may evoke an attempt to counter-regulate and repair thenegative social exclusion situation. Our data support the notion thatthe neurobiological mechanisms underlying this attempt maydiffer between patients and HC and may involve an altered OTresponse.

We did not observe any differences in CT plasma levels betweenpatients and HC, suggesting that the different OT reactions in thetwo groups were not due to differences in their global stress re-actions. These results converge with previous findings: i.e. Zollerand colleagues found no salivary CT release after a Cyberballexclusion situation in healthy subjects (Zoller et al., 2010;Zwolinski, 2012) and may reflect the presence of a defensiveresponse (emotional analgesia) to social exclusion in which centralstress pathways controlling cortisol release are inhibited, as sug-gested by Bass and colleagues (Bass et al., 2014). Moreover, OTplasma levels were found to be independent from CT levels in MDD(Yuen et al., 2014). The decrease of CT levels observed in bothgroups over time convergewith the findings by Bass and colleagues(Bass et al., 2014) and may also reflect a habituation of the CTresponse to the novelty/stress of undergoing the experimentalprotocol and a possibly elevated baseline.

Neither patients nor HC showed a significant increase in innertension during social exclusion. After social exclusion, however,patients remained in more pronounced negative emotional states(NTQ) than HC. Moreover, after social exclusion patients but not HCshowed an increase in negative emotions, in particular anger andresentment (emotion scale). In patients, negative mood experi-enced during social exclusion (NTQ) was higher the more depres-sive symptoms were reported. To our knowledge, so far no otherstudies have reported about the experimentally induced emotionalreaction following a social exclusion challenge (Cyberball or others)in CD patients. In BPD patients, however, negative emotional effectsof social exclusion have been observed with the Cyberball para-digm: The social exclusion situation induced especially anger andcontempt (Jobst et al., 2014; Renneberg et al., 2012), BPD patientsfelt excluded to a greater extent than healthy controls even thoughboth groups estimated similar ball ownership (Jobst et al., 2014),and patients felt excluded even if they were not (Renneberg et al.,2012). CD patients also reacted to social exclusion with an in-crease in negative emotions focused on others, namely anger andresentment, and the exclusion had a similar aversive impact onthem.

Taken together, the negative emotional state observed in CDafter social exclusion (Cyberball) and the accompanying reductionin OT levels compared to HC extends previous findings in BPDpatients (Jobst et al., 2014; King-Casas et al., 2008; Renneberget al., 2012). To our knowledge, no other studies have examinedthis issue with a social exclusion challenge in clinical CD pop-ulations. A study on the effects of cocaine use in pregnancy foundthat women who used cocaine during pregnancy show lowerplasma OT concentrations, experience greater hostility and

Please cite this article in press as: Jobst A, et al., Effects of social exclusiondepression, Journal of Psychiatric Research (2014), http://dx.doi.org/10.10

depressed mood and have less social support and impairedadaptive coping strategies than healthy non-cocaineeusingmothers (Light et al., 2004). These findings may point in a similardirection to our results and suggest altered OT levels in womenafter cocaine use in pregnancy.

In contrast to the findings of Bertsch et al. in BPD (Bertsch et al.,2012) and Yuen et al. in CD (Yuen et al., 2014), we did not observe asignificant difference in baseline OT levels between patients andcontrols in either women or men and we did not find a significantcorrelation between OT baseline and CTQ scores. However, this lackof a significant difference may be due to the small sample size inour study, because mean OT levels were numerically but notsignificantly lower in patients than HC at all four time points andour findings do not prove that there is no correlation between OTand childhood trauma. We were able to confirm the gender dif-ferences recently reported by Yuen et al. for MDD (Yuen et al.,2014), because we also observed significantly lower baseline OTlevels in women than in men. Of course, this observation raises thequestion whether our findings are specific for CD or may begeneralizable to MDD patients; we cannot answer this question atthe present time. Regarding OT plasma levels, findings inMDD vary,and both elevated and reduced OT plasma levels have been re-ported (Cyranowski et al., 2008; Frasch et al., 1995; Parker et al.,2010; Yuen et al., 2014), although these studies do not reportclinical information regarding MDD chronicity.

We propose two hypotheses for the difference in OT levelsbetween patients and HC after the social exclusion paradigm: First,the relative reduction in OT levels after social exclusion mightrepresent an endogenous deficit leading to isolation and inter-personal failure; second, the reduction might be an endogenousprotection mechanism in CD that enables patients to better copewith social stress, i.e. the lower OT levels might regulate aversivesocial cues so that the social stress stimulus is experienced as lessaversive. The second hypothesis of an endogenous protectionmechanism might be supported by recent findings by Ecksteinet al., who demonstrated that OT facilitated the sensation of socialstress in the absence of altered endocrine responses in healthymale participants (Eckstein et al., 2014). If the first hypothesis ofan endogenous deficit is true, the therapeutic application of OTcould theoretically improve interpersonal problems by compen-sating for the deficit in endogenous OT. If it is not true, however,OT substitution might even worsen depressive symptoms, as hasbeen observed in single studies (MacDonald et al., 2013; Mahet al., 2013).

The methodological strengths of our study are the highly stan-dardized experimental setting in which we conducted Cyberball,the fact that we controlled for the menstrual cycle and contracep-tive intake, and the comparisonwith awell-selected control samplematched for sex, age, and education. However, our study also haslimitations. These include a small sample size, pharmacologicaltreatment in the CD group, and no measurement of estrogen andprogesterone levels. Moreover, we examined OT only during theCyberball exclusion condition and did not compare it to an inclu-sion situation. The lacking Cyberball induced distress may appearas another limitation of our study putatively questioning its validityas social exclusion paradigm. Psychological reactions immediatelyfollowing the social exclusion ranged from more thwarted psy-chological need states (Zwolinski, 2012) to more negative moodand fewer positive ruminative thoughts and affect, relative toincluded counterpart (Iffland et al., 2014). Thus, we suggest tofurther optimize the Cyberball paradigm to more reliably inducesocial exclusion related distress in HC, e.g. by extending the

on emotions and oxytocin and cortisol levels in patients with chronic16/j.jpsychires.2014.11.001

A. Jobst et al. / Journal of Psychiatric Research xxx (2014) 1e8 7

duration of the inclusion phase or using a more naturalistic balltossing animation. Finally, the relationship between peripheral andcentral OT levels with respect to social behavior needs to be furtherelucidated (Meyer-Lindenberg et al., 2011).

5. Conclusions

In summary, our results suggest a higher sensitivity of CDpatients to ambiguous threat of social exclusion as we observedan increase of negative emotions and subjective threat toemotional needs and a reduction in OT plasma levels after asocial exclusion paradigm (Cyberball) compared to HC. Thesefindings resemble our previous observations in BPD (Jobst et al.,2014). The reduced OT reactivity to a negative bonding stimulusmight indicate a disrupted attachment system in patients withemotional childhood traumatization, as is common in CD andBPD.

Financial disclosures

F.P. reports having received grants and research support fromBrainsway Inc., Jerusalem, Israel, and neuroConn GmbH, Ilmenau,Germany. Until 2010, P.F. received grants and research support fromAstra Zeneca; speaker honoraria from Astra Zeneca, BMS, Essex,GSK, Janssen-Cilag, and Lundbeck; and was a speaker or member ofthe advisory boards of Janssen-Cilag, Astra Zeneca, Lilly, andLundbeck. The remaining authors reported no biomedical financialinterests or potential conflicts of interest.

Role of the funding source

The study was supported by the F€oFoLe Program of the Facultyof Medicine of the Ludwig Maximilian University, Munich (grant toA. Jobst, grant number 732).

Author contribution

Andrea Jobst and Frank Padberg have done (1) substantialcontributions to conception and design of the work and theacquisition, analysis, and interpretation of data for the work; and(2) drafting of the work or revising it critically for important in-tellectual content; and (3) final approval of the version to bepublished' and (4) agreement to be accountable for all aspects ofthe work in ensuring that questions related to the accuracy orintegrity of any part of the work are appropriately investigatedand resolved.

Lena Sabaß, Nina Sarubin, Anna Buchheim, Babette Renne-berg, Christine Mauer, Peter Falkai have done 1) substantialcontributions to conception of the work, and (2) revising of thework critically for important intellectual content; and (3) finalapproval of the version to be published' and (4) agreement to beaccountable for all aspects of the work in ensuring that questionsrelated to the accuracy or integrity of any part of the work areappropriately investigated and resolved.

Anja Palagyi, Christine Bauriedl-Schmidt, and Peter Zill havedone (1) substantial contributions to design of the work, to theacquisition and analysis of data for the work; and (2) revising itcritically for important intellectual content; and (3) final approvalof the version to be published' and (4) agreement to be accountablefor all aspects of the work in ensuring that questions related to theaccuracy or integrity of any part of the work are appropriatelyinvestigated and resolved.

Please cite this article in press as: Jobst A, et al., Effects of social exclusiondepression, Journal of Psychiatric Research (2014), http://dx.doi.org/10.10

Acknowledgments

This study was part of A. Palagyi's MD thesis at the Faculty ofMedicine of the Ludwig Maximilian University, Munich (inpreparation).

We thank the following clinicians, psychologists, and studentswho provided support for the study: Anna Albert, Amalie Boldi-schar, Tanja Daltrozzo, Jonathan Hall, Eva Holzamer, ConstanceJacob, Ariane Klughardt, Aischa Letters, Annika Pl€oger, RichardMusil, Anja Palagyi, Uli Palm, Elias Seidl, Christina Selberdinger, andCharlotte Wink. We thank also Jacquie Klesing, Board-certifiedEditor in the Life Sciences (ELS), for editing assistance with themanuscript.

References

Angst J, Gamma A, Rossler W, Ajdacic V, Klein DN. Long-term depression versusepisodic major depression: results from the prospective Zurich study of acommunity sample. J Affect Disord 2009;115:112e21.

Apter-Levy Y, Feldman M, Vakart A, Ebstein RP, Feldman R. Impact of maternaldepression across the first 6 years of life on the child's mental health, socialengagement, and empathy: the moderating role of oxytocin. Am J Psychiatry2013;170:1161e8.

Bass EC, Stednitz SJ, Simonson K, Shen T, Gahtan E. Physiological stress reactivityand empathy following social exclusion: a test of the defensive emotionalanalgesia hypothesis. Soc Neurosci 2014;9:504e13.

Beck AT, Steer RA, Ball R, Ranieri W. Comparison of Beck Depression Inventories -IAand -II in psychiatric outpatients. J Pers Assess 1996;67:588e97.

Bernstein DP, Fink L. Childhood trauma questionnaire. A retrospective selfreportmanual. San Antonio (TX): The Psychological Corporation; 1998.

Bernstein DP, Stein JA, Newcomb MD, Walker E, Pogge D, Ahluvalia T, et al.Development and validation of a brief screening version of the ChildhoodTrauma Questionnaire. Child Abuse Negl 2003;27:169e90.

Bertsch K, Gamer M, Schmidt B, Schmidinger I, Walther S, Kastel T, et al. Oxytocinand reduction of social threat hypersensitivity in women with borderline per-sonality disorder. Am J Psychiatry 2013;170:1169e77.

Bertsch K, Schmidinger I, Neumann ID, Herpertz SC. Reduced plasma oxytocin levelsin female patients with borderline personality disorder. Horm Behav 2012;63:424e9.

Buchheim A, Heinrichs M, George C, Pokorny D, Koops E, Henningsen P, et al.Oxytocin enhances the experience of attachment security. Psychoneur-oendocrinology 2009;34:1417e22.

Carter CS. Neuroendocrine perspectives on social attachment and love. Psycho-neuroendocrinology 1998;23:779e818.

Cyranowski JM, Hofkens TL, Frank E, Seltman H, Cai HM, Amico JA. Evidence ofdysregulated peripheral oxytocin release among depressed women. PsychosomMed 2008;70:967e75.

Ditzen B, Schaer M, Gabriel B, Bodenmann G, Ehlert U, Heinrichs M. Intranasaloxytocin increases positive communication and reduces cortisol levels duringcouple conflict. Biol Psychiatry 2009;65:728e31.

Domes G, Heinrichs M, Michel A, Berger C, Herpertz SC. Oxytocin improves “mind-reading” in humans. Biol Psychiatry 2007;61:731e3.

Domes G, Sibold M, Schulze L, Lischke A, Herpertz SC, Heinrichs M. Intranasaloxytocin increases covert attention to positive social cues. Psychol Med 2012:1e7.

Downey G, Feldman SI. Implications of rejection sensitivity for intimate relation-ships. J Pers Soc Psychol 1996;70:1327e43.

Eckstein M, Hurlemann R. Oxytocin: evidence for a therapeutic potential of thesocial neuromodulator. Nervenarzt 2013;84:1321e8.

Eckstein M, Scheele D, Weber K, Stoffel-Wagner B, Maier W, Hurlemann R.Oxytocin facilitates the sensation of social stress. Hum Brain Mapp 2014;35:4741e50.

Feldman R. Oxytocin and social affiliation in humans. Horm Behav 2012;61:380e91.

Feldman R, Weller A, Zagoory-Sharon O, Levine A. Evidence for a neuroendocri-nological foundation of human affiliation: plasma oxytocin levels across preg-nancy and the postpartum period predict mother-infant bonding. Psychol Sci2007;18:965e70.

First MB, Spitzer RL, Gibbon M, Williams JBW. Structured clinical interview forDSM-IV (SCID-I). New York: Biometrics Research Department. New York StatePsychiatric Insitute; 1995.

Forsling ML, Montgomery H, Halpin D, Windle RJ, Treacher DF. Daily patterns ofsecretion of neurohypophysial hormones in man: effect of age. Exp Physiol1998;83:409e18.

Francis DD, Champagne FC, Meaney MJ. Variations in maternal behaviour areassociated with differences in oxytocin receptor levels in the rat.J Neuroendocrinol 2000;12:1145e8.

on emotions and oxytocin and cortisol levels in patients with chronic16/j.jpsychires.2014.11.001

A. Jobst et al. / Journal of Psychiatric Research xxx (2014) 1e88

Frasch A, Zetzsche T, Steiger A, Jirikowski GF. Reduction of plasma oxytocin levelsin patients suffering from major depression. Adv Exp Med Biol 1995;395:257e8.

Goldman M, Marlow-O'Connor M, Torres I, Carter CS. Diminished plasma oxytocinin schizophrenic patients with neuroendocrine dysfunction and emotionaldeficits. Schizophr Res 2008;98:247e55.

Gouin JP, Carter CS, Pournajafi-Nazarloo H, Glaser R, Malarkey WB, Loving TJ, et al.Marital behavior, oxytocin, vasopressin, and wound healing. Psychoneur-oendocrinology 2010;35:1082e90.

Green L, Fein D, Modahl C, Feinstein C, Waterhouse L, Morris M. Oxytocin andautistic disorder: alterations in peptide forms. Biol Psychiatry 2001;50:609e13.

Gross JJ, Levenson RW. Emotion elicitation using films. Cogn Emot 1995;9(1):87e108.

Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry 1960;23:56e62.

Heinrichs M, Baumgartner T, Kirschbaum C, Ehlert U. Social support and oxytocininteract to suppress cortisol and subjective responses to psychosocial stress.Biol Psychiatry 2003;54:1389e98.

Heinrichs M, Meinlschmidt G, Neumann I, Wagner S, Kirschbaum C, Ehlert U, et al.Effects of suckling on hypothalamic-pituitary-adrenal axis responses to psy-chosocial stress in postpartum lactating women. J Clin Endocrinol Metab2001;86:4798e804.

Herpertz SC, Bertsch K. The social-cognitive basis of personality disorders. CurrOpin Psychiatry 2014;27:73e7.

Holt-Lunstad J, Birmingham WA, Light KC. Influence of a “warm touch” supportenhancement intervention among married couples on ambulatory blood pres-sure, oxytocin, alpha amylase, and cortisol. Psychosom Med 2008;70:976e85.

Iffland B, Sansen LM, Catani C, Neuner F. Rapid heartbeat, but dry palms: reactionsof heart rate and skin conductance levels to social rejection. Front Psychol2014;5:956.

Jobst A, Albert A, Bauriedl-Schmidt B, Mauer C, Renneberg B, Buchheim A, et al.Leads to divergent changes of oxytocin levels in borderline patients comparedto healthy subjects. Psychother Psychosom 2014;83:252e4.

Judd LL, Akiskal HS, Zeller PJ, Paulus M, Leon AC, Maser JD, et al. Psychosocialdisability during the long-term course of unipolar major depressive disorder.Arch Gen Psychiatry 2000;57:375e80.

Keller MB, McCullough JP, Klein DN, Arnow B, Dunner DL, Gelenberg AJ, et al.A comparison of nefazodone, the cognitive behavioral-analysis system of psy-chotherapy, and their combination for the treatment of chronic depression.N Engl J Med 2000;342:1462e70.

Kessler RC, McGonagle KA, Zhao S, Nelson CB, Hughes M, Eshleman S, et al. Lifetimeand 12-month prevalence of DSM-III-R psychiatric disorders in the UnitedStates. Results from the National Comorbidity Survey. Arch Gen Psychiatry1994;51:8e19.

King-Casas B, Sharp C, Lomax-Bream L, Lohrenz T, Fonagy P, Montague PR. Therupture and repair of cooperation in borderline personality disorder. Science2008;321:806e10.

Klinitzke G, Romppel M, Hauser W, Brahler E, Glaesmer H. The German Version ofthe Childhood Trauma Questionnaire (CTQ): psychometric characteristics in arepresentative sample of the general population. Psychother Psychosom MedPsychol 2012;62:47e51.

Kosfeld M, Heinrichs M, Zak PJ, Fischbacher U, Fehr E. Oxytocin increases trust inhumans. Nature 2005;435:673e6.

Light KC, Grewen KM, Amico JA, Boccia M, Brownley KA, Johns JM. Deficits inplasma oxytocin responses and increased negative affect, stress, and bloodpressure in mothers with cocaine exposure during pregnancy. Addict Behav2004;29:1541e64.

MacDonald K, MacDonald TM, Brune M, Lamb K, Wilson MP, Golshan S, et al.Oxytocin and psychotherapy: a pilot study of its physiological, behavioral andsubjective effects in males with depression. Psychoneuroendocrinology2013;38:2831e43.

Mah BL, Van Ijzendoorn MH, Smith R, Bakermans-Kranenburg MJ. Oxytocin inpostnatally depressed mothers: its influence on mood and expressed emotion.Prog Neuropsychopharmacol Biol Psychiatry 2013;40:267e72.

Please cite this article in press as: Jobst A, et al., Effects of social exclusiondepression, Journal of Psychiatric Research (2014), http://dx.doi.org/10.10

Meyer-Lindenberg A, Domes G, Kirsch P, Heinrichs M. Oxytocin and vasopressin inthe human brain: social neuropeptides for translational medicine. Nat RevNeurosci 2011;12:524e38.

Nemeroff CB, Heim CM, Thase ME, Klein DN, Rush AJ, Schatzberg AF, et al. Differ-ential responses to psychotherapy versus pharmacotherapy in patients withchronic forms of major depression and childhood trauma. Proc Natl Acad Sci U SA 2003;100:14293e6.

Neumann ID, Kromer SA, Toschi N, Ebner K. Brain oxytocin inhibits the (re)activityof the hypothalamo-pituitary-adrenal axis in male rats: involvement of hypo-thalamic and limbic brain regions. Regul Pept 2000;96:31e8.

Nishioka T, Anselmo-Franci JA, Li P, Callahan MF, Morris M. Stress increases oxytocinrelease within the hypothalamic paraventricular nucleus. Brain Res 1998;781:56e60.

Parker KJ, Kenna HA, Zeitzer JM, Keller J, Blasey CM, Amico JA, et al. Preliminaryevidence that plasma oxytocin levels are elevated in major depression. Psy-chiatry Res 2010;178:359e62.

Pierrehumbert B, Torrisi R, Ansermet F, Borghini A, Halfon O. Adult attachmentrepresentations predict cortisol and oxytocin responses to stress. Attach HumDev 2012;14:453e76.

Pierrehumbert B, Torrisi R, Laufer D, Halfon O, Ansermet F, Beck Popovic M.Oxytocin response to an experimental psychosocial challenge in adults exposedto traumatic experiences during childhood or adolescence. Neuroscience2010;166:168e77.

Renneberg B, Herm K, Hahn A, Staebler K, Lammers CH, Roepke S. Perception ofsocial participation in borderline personality disorder. Clin Psychol Psychother2012;19:473e80.

Salonia A, Nappi RE, Pontillo M, Daverio R, Smeraldi A, Briganti A, et al. Menstrualcycle-related changes in plasma oxytocin are relevant to normal sexual functionin healthy women. Horm Behav 2005;47:164e9.

Staebler K, Gebhard R, Barnett W, Renneberg B. Emotional responses in borderlinepersonality disorder and depression: assessment during an acute crisis and 8months later. J Behav Ther Exp Psychiatry 2009;40:85e97.

Staebler K, Renneberg B, Stopsack M, Fiedler P, Weiler M, Roepke S. Facial emotionalexpression in reaction to social exclusion in borderline personality disorder.Psychol Med 2011;41:1929e38.

Stanley B, Siever LJ. The interpersonal dimension of borderline personality disorder:toward a neuropeptide model. Am J Psychiatry 2010;167:24e39.

Striepens N, Kendrick KM, Maier W, Hurlemann R. Prosocial effects of oxytocin andclinical evidence for its therapeutic potential. Front Neuroendocrinol 2011;32:426e50.

Wiersma JE, Hovens JG, van Oppen P, Giltay EJ, van Schaik DJ, Beekman AT, et al. Theimportance of childhood trauma and childhood life events for chronicity ofdepression in adults. J Clin Psychiatry 2009;70:983e9.

Williams KD, Cheung CK, Choi W. Cyberostracism: effects of being ignored over theInternet. J Pers Soc Psychol 2000;79:748e62.

Williams KD, Jarvis B. Cyberball: a program for use in research on interpersonalostracism and acceptance. Behav Res Methods 2006;38:174e80.

Wotjak CT, Ganster J, Kohl G, Holsboer F, Landgraf R, Engelmann M. Dissociatedcentral and peripheral release of vasopressin, but not oxytocin, in response torepeated swim stress: new insights into the secretory capacities of peptidergicneurons. Neuroscience 1998;85:1209e22.

Yuen KW, Garner JP, Carson DS, Keller J, Lembke A, Hyde SA, et al. Plasma oxytocinconcentrations are lower in depressed vs. healthy control women and are in-dependent of cortisol. J Psychiatr Res 2014;51:30e6.

Zak PJ, Kurzban R, Matzner WT. Oxytocin is associated with human trustworthiness.Horm Behav 2005;48:522e7.

Zobel I, Werden D, Linster H, Dykierek P, Drieling T, Berger M, et al. Theory of minddeficits in chronically depressed patients. Depress Anxiety 2010;27:821e8.

Zoller C, Maroof P, Weik U, Deinzer R. No effect of social exclusion on salivarycortisol secretion in women in a randomized controlled study. Psychoneur-oendocrinology 2010;35:1294e8.

Zwolinski J. Psychological and neuroendocrine reactivity to ostracism. AggressBehav 2012;38:108e25.

on emotions and oxytocin and cortisol levels in patients with chronic16/j.jpsychires.2014.11.001