Increased testosterone levels and cortisol awakening responses in patients with borderline personality disorder: Gender and trait aggressiveness matter

Psychoneuroendocrinology (2015) 55, 116—127

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/psyneuen

Increased testosterone levels and cortisol awakening responses in patients with borderline personality disorder: Gender and trait aggressiveness matter Juliane Rausch a, Andrea Gäbel a, Krisztina Nagy a, Nikolaus Kleindienst b, Sabine C. Herpertz a, Katja Bertsch a,∗ a

Department for General Psychiatry, Center of Psychosocial Medicine, University of Heidelberg, Germany Department of Psychosomatic Medicine, Central Institute of Mental Health Mannheim, Medical Faculty Mannheim, Heidelberg University, Heidelberg, Germany

b

Received 20 November 2014; received in revised form 4 February 2015; accepted 4 February 2015

KEYWORDS Endocrinology; Hypothalamus— pituitary—adrenal axis; Hypothalamus— pituitary—gonadal axis; Stress; Androgen metabolism; Reproductive system

Summary Background: Borderline personality disorder (BPD) is characterized by antagonism, negative affectivity, disinhibition, and impairments in interpersonal functioning, including enhanced impulsive aggression. Interpersonal dysfunctions may be related to alterations in endocrine systems. The current study investigated alterations in basal activity of the hypothalamus— pituitary—gonadal (HPG) reproductive and the hypothalamus—pituitary—adrenal (HPA) stress system in BPD patients and their association to anger-related aggression with a particular focus on effects of gender and comorbid conditions of depression and posttraumatic stress disorder (PTSD). Method: Saliva testosterone levels as well as cortisol awakening responses were assessed in 55 medication-free female and male patients with BPD and compared to 47 gender-, age, and intelligence-matched healthy volunteers. In addition, analyses controlling for current depression and PSTD and bivariate correlations between testosterone and cortisol levels on the one hand and anger and aggressiveness on the other hand were performed. Results: The results revealed increased saliva testosterone levels in female and male patients with BPD as well as elevated cortisol awakening responses in female, but not male patients with BPD compared to healthy volunteers. Cortisol awakening responses were positively related to anger and aggressiveness in female patients with BPD, but no associations were found with testosterone levels.

∗ Corresponding author at: Department of General Psychiatry, Center for Psychosocial Medicine, University of Heidelberg, Voßstraße 4, 69115 Heidelberg, Germany. Tel.: +49 6221 56 36502; fax: +49 6221 56 5374. E-mail address: [email protected] (K. Bertsch).

http://dx.doi.org/10.1016/j.psyneuen.2015.02.002 0306-4530/© 2015 Elsevier Ltd. All rights reserved.

Testosterone and cortisol in BPD

117 Conclusion: In line with previous reports, the present results suggest endocrine alterations in BPD which may be associated with interpersonal impairments, such as increased anger-related aggressive behavior and could have implications for the development of new (psychopharmaco-) therapeutic interventions that may help to restore the alterations in the HPA and HPG systems. © 2015 Elsevier Ltd. All rights reserved.

1. Introduction Borderline personality disorder (BPD) is a severe mental disorder characterized by antagonism, negative affectivity, disinhibition, and impairments in interpersonal functioning, including enhanced anger-related impulsive aggression (American Psychiatric Association, 2013). Interpersonal problems such as impulsive aggression may be related to endocrine alterations. In healthy volunteers and highly aggressive populations, there is some evidence for associations between the hypothalamus—pituitary—gonadal (HPG) and hypothalamus—pituitary—adrenal (HPA) stress axis functioning and aggression (for reviews, see Van Honk et al., 2010; Carre and Mehta, 2011; Haller, 2013, 2014). Although results are inconsistent (Archer et al., 2005), increased basal testosterone levels have been reported in more aggressive individuals, and (weak) positive correlations between testosterone and aggression were found — particularly in those with low basal cortisol secretion (Popma et al., 2007; Carre and Mehta, 2011). As for cortisol secretion, positive associations with aggression have been found with both decreased and increased levels in healthy volunteers and aggressive individuals (e.g., Gerra et al., 2007; van Goozen et al., 2007; Böhnke et al., 2010a, 2010b) suggesting a general association between HPA axis dysregulations and aggressiveness (Haller, 2013, 2014). In BPD, only few studies have investigated endocrine alterations and their association with the patients’ interpersonal impairments. This is surprising as structural and functional alterations in brain regions that are crucially involved in the processing of social threats, the regulation of the fight/flight response, and the endocrine system have been reported in BPD. Besides reduced hippocampal and amygdalar gray matter volumes, which are the most consistent structural alteration in BPD (for meta-analysis, see Nunes et al., 2009), abnormalities in gray matter volume of the anterior cingulate cortex (e.g., Hazlett et al., 2005; Minzenberg et al., 2008), the hypothalamus (Kuhlmann et al., 2013), and pituitary (Garner et al., 2007) have been observed. In addition, exaggerated and prolonged amygdala responses (e.g., Herpertz et al., 2001; Hazlett et al., 2012) seem to be a neural correlate for BPD patients’ hypersensitivity to social threats (Bertsch et al., 2013a). Patients with BPD also report more frequent and intense daily hassles (Jovev and Jackson, 2006) and elevated levels of stress-associated inner tension (Kuo and Linehan, 2009), which often precede self-injury and impulsive aggression (Kleindienst et al., 2008). As about 80% of patients diagnosed with BPD report traumatic childhood experiences, early stress and related HPA axis alterations have been discussed to play a prominent role in the etiology of BPD (Herman et al., 1989; Ogata et al., 1990). Although results of the few previous studies on HPA axis functioning are

heterogeneous and may be influenced by comorbid disorders, in particular current major depression and posttraumatic stress disorder (PTSD; for review, see Zimmerman and Choi-Kain, 2009; Wingenfeld et al., 2010), there is evidence for elevated cortisol secretion and thus a basal HPA hyperactivity in BPD (Lieb et al., 2004; Wingenfeld et al., 2007; Carvalho Fernando et al., 2012). Regarding HPG activity, so far only one study investigating the occurrence of polycystic ovary (PCO) syndrome in a small group of partly medicated female patients with BPD and healthy women has shown increased plasma testosterone levels that could not be explained by PCO status or weight (Roepke et al., 2010). Taken together, elevated basal cortisol and testosterone secretion have been reported in BPD who are characterized by more frequent and intense reactions to potential social threats and stressors and elevated stress-related inner tension which often triggers dysfunctional behaviors, such as anger-related impulsive aggression. However, little is known about effects of gender and comorbid disorders — current major depression and PTSD, in particular — as well as trait anger and aggressiveness with respect to HPG and HPA axes functioning in BPD. The current study therefore aimed to investigate alterations in HPG and HPA axes functioning in female and male patients with BPD compared to healthy women and men controlling for comorbid conditions of current major depression and PTSD and their relationship with self-reported anger and aggressiveness. Therefore, we recruited a relatively large clinical sample of medication-free female and male patients with a current DSM-IV diagnosis of BPD and healthy women and men who provided two saliva samples to assess mean saliva testosterone levels as well as eight saliva samples at four fixed time points after awakening on two consecutive weekdays to assess the cortisol awakening response, a reliable measure for the HPA axis functioning (Schmidt-Reinwald et al., 1999; Hellhammer et al., 2007). In addition, anger and aggressiveness — as well as depressiveness, borderline symptom severity, and history of childhood traumatization — were measured with self-report questionnaires. Based on previous results, we hypothesized increased saliva testosterone levels and cortisol awakening responses in patients with BPD compared to healthy volunteers. We also expected increased testosterone levels in male compared to female patients and healthy volunteers and explored group by gender interactions as well as associations between endocrine data and self-reported anger and aggressiveness in BPD.

2. Material and methods 2.1. Participants Participants were 55 patients with a current DSM-IV diagnosis of BPD (including self-injury and/or aggression; BPD;

118 N = 35 women; Mage = 27.4, SD = 7.2, range: 18—44 years) and 47 age-, gender-, and intelligence-matched healthy volunteers (CON; N = 26 women, Mage = 28.0, SD = 6.5, range: 19—49 years) who had never received a psychiatric diagnosis or undergone any psychological or psychiatric treatment. The study was part of the German Clinical Research Unit: Mechanisms of Disturbed Emotion Processing in Borderline Personality Disorder (Schmahl et al., 2014). As systematic investigations of gender differences in patients with BPD are missing (Mancke et al., in press), we recruited female and male patients and healthy volunteers through the resident’s registration office, advertisements and clinical referral from in- and out-patient units. General exclusion criteria comprised: current alcohol/drug abuse (urine toxicology screenings), or alcohol/drug abuse in the last two months (interview); severe medical illness; and regular medication (except for oral contraceptives). Additional exclusion criteria for patients were lifetime diagnosis of schizophrenia, schizoaffective or bipolar disorder, reported alcohol/drug dependence in the last 12 months. No testosterone data were available from one female BPD patient, one male BPD patient, and one female healthy volunteer. In addition, cortisol data of 9 female and 5 male BPD patients as well as 5 female and 2 male healthy volunteers had to be excluded due to a lack of compliance regarding the data collection (assessed with protocols and electronic devices). Thus, analyses of testosterone data included 54 BPD patients (34 female) as well as 46 healthy volunteers (25 female), while analyses of cortisol data were based on 41 BPD patients (26 female) and 40 healthy volunteers (21 female). The Ethics Committee of the Medical Faculty of the University of Heidelberg approved the study. All participants provided written informed consent, after the study procedures were fully explained and were paid for their participation.

2.2. Measures Qualified diagnosticians assessed borderline personality diagnosis as well as axis I and II comorbidities (axis I diagnoses: SCID-I; First et al., 1995; axis II diagnoses including BPD: International Disorder Examination; Loranger et al., 1994). State and trait anger was assessed with the StateTrait Anger Expression Inventory (STAXI; Schwenkmezger et al., 1992) and trait aggressiveness with the Buss and Perry Aggression Questionnaire (BPAQ; Buss and Perry, 1992). In addition, borderline symptom severity was assessed with the Borderline Symptom List (BSL-23; Bohus et al., 2009); depressiveness with the Beck Depression Inventory (BDI; Beck and Steer, 1995); history of childhood traumatization was assessed with the Childhood Trauma Questionnaire (CTQ; Bernstein and Fink, 1998). These instruments are well established and validated, and have been used in multiple previous studies assessing trait characteristics and their association with biological data in BPD patients and healthy volunteers. Demographic information including possible confounding variables such as age, smoking, and for female participants, contraceptive intake and current menstrual cycle phase were assessed in standardized questionnaires. Height and weight was measured at the day

J. Rausch et al. of testosterone assessment. Raven’s Progressive Matrices were used as an estimate for intelligence (Raven and Horn, 2009).

2.3. Testosterone and cortisol assessment and analysis 2.3.1. Mean testosterone levels Basal testosterone levels were assessed with two native saliva samples collected in the afternoon (first sample between 1.30 and 2.00 p.m. and second sample between 1530 h and 1600 h in 2-ml polypropylene tubes IBL, Hamburg, Germany) according to a procedure described by Schultheiss and Stanton (2009). Samples were immediately frozen at −20 ◦ C for biochemical analysis. Testosterone concentration was measured using a competitive chemiluminescence immunoassay (LIA) with a sensitivity of 0.0025 ng/ml (IBL) and intra- and interassay coefficients between 10% and 12%. To increase reliability, we used the average of the two saliva samples for statistical analyses (mean testosterone).

2.3.2. Cortisol awakening response To get a reliable trait measure of basal HPA axis activity, we assessed the cortisol awakening response (CAR) on two consecutive weekdays (Hellhammer et al., 2007). The participants collected saliva samples with salivette devices (Starstedt, Rommelsdorf, Germany) at home each day at awakening and 30, 45, and 60 min later. The participants had to gently chew on a cotton swab for about 1 min and then put it back in the devise without touching it with their hands. Sample collection was protocolled and locked with electronic monitoring systems (MEMS® , Medication Event Monitoring System), which provide a time signature for each removal of a salivette, and enhance compliance of participants (Kudielka et al., 2012). During the sampling period, participants drank nothing but water and refrained from brushing their teeth, eating, and exercising. The participants stored all samples in the refrigerator or freezer until returning them to our laboratory. The samples were then frozen at −20 ◦ C until biochemical analysis. Cortisol concentration was measured using a commercially available chemiluminescence immunoassay (CLIA) with high sensitivity of 0.16 ng/ml (IBL) and intra- and interassay coefficients of less than 6% and 8%, respectively. We computed the area under the curve with respect to ground (AUCG ) and the mean cortisol increase (MnInc) of the cortisol awakening response as trait measures of HPA axis activity (Hellhammer et al., 2007). AUCG , which represents the entire area under the cortisol awakening response with respect to ground, was calculated according to a formula described by Pruessner et al. (2003) and MnInc, which represents the cortisol increase after awakening, was calculated with the formula provided by Wust et al. (2000). AUCG and MnInc were calculated for each participant and day and then averaged over the 2 days to form reliable indicators of overall or total cortisol secretory activity (AUCG ; Pruessner et al., 2003; Clow et al., 2004) and cortisol dynamic following awakening (MnInc; Wust et al., 2000; Clow et al., 2004).

Testosterone and cortisol in BPD

2.4. Statistical analysis Statistical analyses were performed with IBM SPSS 20. Demographic and psychometric data as well as data regarding the BPD symptomatology and the history of traumatization were analyzed with two-factorial analyses of variance (ANOVAs) with the between-subject factors group (patients, volunteers) and gender (female, male) or chi2 -tests for nonparametric variables (e.g., comorbidities, menstrual cycle, or smoking). Group differences in mean testosterone levels as well as in the mean AUCG and MnInc of the CAR were also analyzed with separate two-factorial ANOVAs with the between-subject factors group (BPD, CON) and gender (female, male). In addition, we performed a 2 × 2 × 4 repeated-measures ANOVA with the between-subject factors group (BPD, CON) and gender (female, male) and the within-subject factor time (0, +30, +45, +60 min after awakening). Analyses of covariance (ANCOVAs) were calculated in order to analyze effects of potential confounding variables, i.e., comorbid major depression or PTSD, smoking, age, and body-mass index, and for female participants contraceptive intake and phase of menstrual cycle. Where appropriate, the Huynh—Feldt procedure (Huynh and Feldt, 1976) was applied to correct for potential violations of the sphericity assumption. All statistical analyses employed a two-tailed p< .05. Effect sizes of significant results are reported as proportion of explained variance (partial eta squared [2 ]). In cases of significant effects, we used Dunn’s multiple comparison as post hoc tests, which include a Bonferoni correction for multiple testing. We additionally performed Pearson correlations to investigate the relationship between testosterone levels and cortisol awakening responses and self-reported anger and aggressiveness in the four subgroups. Please note that the results of these correlational analyses may only be interpreted descriptively due to the limited power in the separate analyses for the four subgroups (female BPD, male BPD, female CON, male CON).

3. Results 3.1. Comorbidities, demographic and psychometric data 3.1.1. Comorbidities The comorbidity of current major depression was 23% (lifetime: 86%) in female BPD patients and 25% (lifetime: 75%) in male BPD patients and current PTSD was diagnosed in 31% (lifetime: 37%) of female and 35% (lifetime: 35%) of male BPD patients. Female and male patients did not differ significantly in the frequency of comorbid major depression or PTSD (p > .157; see Table 1 for details and statistical parameters). Further comorbid disorders comprised: anxiety disorders (female BPD: 49% current, 80% lifetime; male BPD: 50% current, 55% lifetime), eating disorders (female BPD: 40% current, 80% lifetime; male BPD: 50% current, 55% lifetime), remitted substance abuse (female BPD: 43%; male BPD: 45%), obsessive—compulsive disorder (female BPD: 11% current, 6% lifetime; male BPD: 0% current and lifetime), somatoform disorder (female BPD: 11% current; male BPD: 0% current), adjustment disorder (female BPD: 9% current) as well as antisocial (female BPD: 2.9% current, 2.9%

119 lifetime; male BPD: 15% current, 25% lifetime) and avoidant (females BPD 28.6% current, 31.4% lifetime; male BPD: 25% current, 35% lifetime) personality disorders. Please note that a history of any psychiatric disorder was an exclusion criterion for healthy volunteers leading to significant differences between BPD patients and healthy volunteers in the reported psychiatric diagnoses. 3.1.2. Demographic and psychometric data BPD patients did not differ from healthy volunteers in terms of the matching variables (gender distribution, age, and intelligence; for main effects of group as well as group by gender interactions; p > .313, see Table 1 for details) and there was no significant difference between female patients and healthy volunteers regarding menstrual cycle phase or oral contraceptive intake (p ≥ .10, see Table 1). However, body-mass-index was significantly higher in BPD patients than in healthy volunteers (p = .044) and the BPD patient sample tended to comprise more smokers (p = .064) compared to the healthy volunteer sample. In addition, there was a significant main effect of gender for age (p = .021), body-mass index (p =.012), and intelligence (p =.013) with older age, higher body-mass indices, and lower intelligence levels in male compared to female participants across the BPD patient and healthy volunteer groups (no significant group by gender interactions; p ≥ .179; see Table 1 for details and statistical parameters).

3.2. BPD symptomatology and history of childhood traumatization On average, BPD patients fulfilled 6.4 DSM-IV BPD criteria and thus differed significantly from healthy volunteers, who fulfilled 0.2 BPD criteria (main effect of group, p< .001). Importantly, female and male patients did not differ significantly in terms of fulfilled BPD criteria as shown by non-significant main effect of gender and a non-significant group by gender interaction (p ≥ .623). As expected, patients reported significantly higher levels of borderline symptom severity (BSL-23), depressiveness (BDI), aggressiveness (BPAQ), and anger (STAXI; main effect of group, p < .001; see Table 1 for details). Female and male BPD patients and male and female healthy volunteers did not differ regarding borderline symptom severity, depressiveness, or anger (no significant main effect of gender or gender by group interaction, p ≥ .151, see Table 1). However, self-reported aggressiveness in the BPAQ was generally increased in males compared to females (main effect of gender, p = .003), and in male compared to female BPD patients, in particular (group by gender interaction, p = .084). As expected, BPD patients reported significantly higher levels of childhood traumatization (CTQ) compared to healthy volunteers (main effect of group, p< .001; no significant main effect of gender or group by gender interaction, p ≥ .179; Table 1).

3.3. Basal testosterone and cortisol levels 3.3.1. Mean testosterone levels In line with our hypothesis, BPD patients had increased basal testosterone levels compared to healthy volunteers

120 Table 1 Demographic, and psychometric, and endocrine data as well as BPD symptomatology and history of childhood traumatization of female (F-BPD) and male (M-BPD) patients and female (F-CON) and male (M-CON) healthy volunteers. Values represent means (M) and standard errors (SE) as well as F- and 2 -values for parametric variables and Ns and % as well as 2 and -values for non-parametric variables. Current and lifetime diagnoses of major depression (depression) and posttraumatic stress disorder (PTSD) were based on Structured Clinical Interview for DSM-IV (SCID-I). Number of currently fulfilled BPD criteria was assessed with the International Examination for Personality Disorders (IPDE). BSL: Borderline Symptom List, BDI: Beck Depression Inventory; BPAQ: Buss and Perry Aggression Questionnaire; STAXI: State-Trait Anger Expression Inventory; CTQ: Childhood trauma Questionnaire; AUCG Cortisol: Mean area under the curve with respect to ground of the cortisol awakening response; MInc Cortisol: mean increase of the cortisol awakening response.

Smokers (N (%)) Age (years) Body-mass index Menstrual cycle phase Oral contraceptive users Raven-intelligence (IQ) Current depression (N (%)) Lifetime depression (N (%)) Current PTSD (N (%)) Lifetime PTSD (N (%)) Borderline criteria BSL-borderline symptoms BDI-depressiveness BPAQ-aggressiveness STAXI-state anger STAXI-trait anger CTQ-total score Mean testosterone (pg/ml) AUCG Cortisol MInc Cortisol

F-BPD

F-CON

M-BPD

M-CON

Group

M ± SE/N (%)

M ± SE/N (%)

M ± SE/N (%)

M ± SE/N (%)

F/2

p

2 /

F/2

14 (40%) 26.5 ± 1.1 24.9 ± 0.9 9/14/6/60% 3 (9%) 109.8 ± 2.0 8 (23%) 30 (86%) 11 (31%) 13 (37%) 6.4 ± 0.2 2.1 ± 0.1 27.1 ± 1.2 64.3 ± 2.1 13.6 ± 0.7 24.9 ± 0.9 59.2 ± 2.3 26.3 ± 4.9 1659.6 ± 117.5 16.5 ± 1.9

8 (31%) 26.3 ± 1.3 22.8 ± 1.2 8/12/8/35% 5 (19%) 112.0 ± 2.3 0 0 0 0 0.4 ± 0.2 0.1 ± 0.1 3.4 ± 1.4 43.1 ± 2.8 10.3 ± 0.9 13.7 ± 1.1 32.6 ± 2.7 11.5 ± 5.8 1321.0 ± 130.7 5.9 ± 2.2

11 (55%) 29.3 ± 1.5 28.0 ± 1.2 — — 103.7 ± 2.6 5 (25%) 15 (75%) 7 (35%) 7 (35%) 6.3 ± 0.2 2.1 ± 0.1 29.0 ± 1.6 76.4 ± 2.4 15.9 ± 0.8 25.2 ± 1.3 66.3 ± 3.1 83.7 ± 6.4 1174.3 ± 154.7 7.9 ± 2.6

5 (24%) 30.1 ± 1.5 25.5 ± 1.2 — — 106.2 ± 2.5 0 0 0 0 0.0 ± 0.2 0.2 ± 0.1 3.6 ± 1.6 46.5 ± 2.7 10.4 ± 0.9 14.3 ± 1.2 32.2 ± 3.1 67.7 ± 6.3 1315.4 ± 137.4 7.0 ± 2.3

3.43 0.95 4.20 0.79 1.02 1.03 13.26 69.70 18.70 22.17 824.89 362.20 278.94 101.37 27.90 93.47 119.44 6.84 0.53 6.58

.064 .758 .044 .851 .313 .313 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 <.001 .010 .469 .012

.18 <.01 .05 .13 .10

0.09 5.52 6.53 — — 6.40 0.03 1.99 0.11 1.74 0.24 0.07 0.50 9.35 2.09 0.15 1.46 93.24 3.27 2.80

Gender

.37 .84 .43 .47 .89 .80 .75 .52 .23 .50 .56 .07 .01 .08

p

2 /

.762 .021 .012

.03 .05 .08

— —

— —

.013 .853 .157 .915 .420 .623 .792 .482 .003 .151 .698 .230 <.001 .074 .098

.02 .14 .01 .13 <.01 <.01 .01 .09 .02 <.01 .02 .49 .04 .04

Group

x

Gender

F/2

p

2 /

4.90 .179 0.18 .677 0.038 .845 — — — — 0.01 .928 13.30 .004 70.59 <.001 18.94 <.001 25.43 <.001 0.10 .751 0.12 .731 0.35 .553 3.05 .084 1.61 .208 0.02 .898 1.82 .179 0.01 .924 3.12 .081 4.74 .033

.22 <.01 <.01 — — .37 .84 .44 .50 <.01 <.01 <.01 .03 .02 <.01 .02 <.01 .04 .06

J. Rausch et al.

Testosterone and cortisol in BPD

Figure 1 Mean saliva testosterone levels (+ one standard error) of female (FBPD) and male BPD (MBPD) patients and female (FCON) and male (MCON) healthy volunteers. **p < .01.

(main effect of group, F(1,96) = 6.84, p = .010, 2 = .07; Fig. 1). Additionally and as expected, testosterone levels were generally increased in male compared to female participants (main effect of gender, F(1,96) = 93.24, p <.001, 2 = .49) in both the patient and healthy control sample (no significant group by gender interaction, F(1,96) = 0.01, p = .924, 2 < .01). Group and gender differences in testosterone levels remained stable after controlling for possible confounding variables, i.e., comorbid diagnosis of major depression or PTSD, smoking, and body-mass index, as well as menstrual cycle phase and oral contraceptive intake. To further explore potential influences of comorbid major depression or PTSD, we compared testosterone levels between patients with and without a current diagnosis of major depression as well as between patients with and without a current diagnosis of PTSD using Student’s t-test. These analyses revealed no significant difference between patients with and without a current diagnosis of major depression (p = .691) or PTSD (p = .858). 3.3.2. Cortisol awakening response Female BPD patients showed increased and steeper cortisol awakening response compared to male BPD patients as well as female and male healthy volunteers (see Figs. 2 and 3). According the statistical analyses, AUCG tended to be increased in female compared to male participants (main effect of gender, F(1,77) = 3.27, p = .074, 2 = .41). This main effect of gender was, however, qualified by a trendlevel significant group by gender interaction (F(1,77) = 3.12, p = .081, 2 = .04) indicating increased AUCG in female BPD patients compared to male BPD patients (p < .05) and compared to healthy women (p < .10) and healthy men (p < .10; Fig. 3A). In addition, MnInc were significantly steeper in BPD patients compared to healthy volunteers (main effect of group, F(1,77) = 6.58, p = .012, 2 = .08). According to a significant group by gender interaction (F(1,77) = 4.74, p = .033, 2 = 0.6; Fig. 2B), this was, however, only the case for female

121 BPD patients compared to female volunteers (p < .01), while male patients did not significantly differ from healthy male volunteers (p > .05) or female volunteers (p > .05). Post hoc tests also revealed significantly increased MnInc in female BPD patients compared to male BPD patients (p < .05). The repeated-measure ANOVA including all four time points of saliva collection in the morning revealed a main effect of time (F(3,246) = 31.72, p < .001, ε = .57, 2 = .28) replicating the often reported inverted U-shape of the cortisol awakening response with rising cortisol levels after awakening (p < .01 from 0 min to 30, 45, and 60 min after awakening) followed by a drop (p < .05 from 30/45 min to 60 min after awakening). According to significant interactions of gender by time (F(3,246) = 3.57, p = .038, ε = .57, 2 = .04), group by time (F(3,246) = 4.15, p = .023, ε = .57, 2 = .05) as well as group by gender by time (F(3,246) = 3.26, p = .049, ε = .57, 2 = .04), female BPD patients had significantly higher cortisol levels 30, 45, and 60 min after awakening compared to both male BPD patients (30 min: p < .10, 45 min and 60 min: p < .01) as well as female and male healthy volunteers (all p < .01; Fig. 2A). There was no significant difference between the groups at the first time of measurement (0 min after awakening, p > .10). In all four groups, cortisol levels 30, 45, and 60 min after awakening were significantly higher than directly (0 min) after awakening (all p < .05). Importantly, effects remained stable even after controlling for possible confounding variables, i.e., comorbid diagnosis of major depression or PTSD, smoking, and bodymass index, as well as menstrual cycle phase and oral contraceptive intake. To further explore potential influences of comorbid major depression or PTSD, we compared cortisol awakening responses (AUCG and MnInc) between patients with and without a current diagnosis of major depression as well as between patients with and without a current diagnosis of PTSD using Student’s t-test. These analyses revealed no significant difference between patients with and without a current diagnosis of depression (p ≥ .257) or PTSD (p ≥ .142). In an additional, exploratory ANOVA, we compared mean testosterone levels of patients and healthy volunteers with high vs. low cortisol awakening responses (median split according to MnInc levels; the same analysis was repeated with a median split according to AUCG which however yielded the same results). A significant group by MnInc interaction (F(1,76) = 4.61, p = .035, 2 = .06; no significant main effects of group or MnInc, F(1,76) ≤ 1.58, p ≥ .213, 2 ≤ .02) indicated enhanced testosterone levels in those patients with lower cortisol awakening responses compared to patients with higher cortisol awakening responses (p < .10) as well as healthy volunteers (p < .10).

3.4. Association between endocrine data and self-reported anger and aggressiveness Contrary to our hypothesis, we did not find a significant association between testosterone levels and self-reported anger or aggressiveness in female BPD patients (r(32) ≥ .19, p ≥ .293) or male BPD patients (r(18) ≥ .18, p ≥ .473) and healthy women (r(23) ≥ .07, p ≥ .750) or healthy men (r(19)≤ ≥ .27, p ≥ .258). Additional correlational analyses however revealed positive associations between AUCG as

122

J. Rausch et al.

Figure 2 Mean cortisol awakening response (± one standard error) of female (FBPD) and male BPD (MBPD) patients and female (FCON) and male (MCON) healthy volunteers. (A) Saliva cortisol levels at 0, +30, +45, and +60 min after awakening averaged across two consecutive weekdays. (B) Mean Increase (MnInc) of the cortisol awakening response averaged across two consecutive weekdays. **p < .01.

well as MnInc of the cortisol awakening response and self-reported aggressiveness (AUCG: r(24) = .49, p = .013, MnInc: r(24) = .33, p = .104), and the STAXI subscales state anger (AUCG: r(24) = .52, p = .008, Fig. 3B; MnInc: r(24) = .23, p = .234) and trait anger (AUCG: r(24) = .64, p = .001, MnInc: r(24) = .42, p = .036; Fig. 3B) in female BPD patients. No significant correlations were found in male BPD patients (r(13) ≥ .14, p ≤ .628), healthy women (r(19) ≥ .18, p ≥ .461), or healthy men (r(17) ≤ .21, p ≥ .398). An additional hierarchical multiple regression analysis (for details, see Mehta and Josephs, 2010) with self-reported aggressiveness as outcome and the following variables as

predictors: gender, group (Step 1), mean testosterone (zstandardized within gender) and cortisol (z-standardized) (Step 2), testosterone × cortisol interaction (Step 3), and testosterone × cortisol × gender × group interaction (Step 4) did not reveal a significant main effect of cortisol or any significant testosterone × cortisol interactions (neither using AUCG , MnInc, baseline cortisol, nor mean cortisol levels, all p > .10). Additional correlational analyses revealed that neither testosterone nor cortisol levels were significantly correlated with borderline symptom severity (BSL-23; p ≥ .55), depressiveness (BDI; p ≥ .35), or childhood traumatization (p > .20),

Figure 3 (A) Area under the curve with respect to ground (AUCG ) of the cortisol awakening response (+ one standard error) averaged across two consecutive weekdays in female (FBPD) and male BPD (MBPD) patients and female (FCON) and male (MCON) healthy volunteers. (B) Relationship between AUCG and the STAXI Trait Anger scale for female BPD patients. *p < .05, # p < .10.

Testosterone and cortisol in BPD and there were no significant associations between testosterone and cortisol levels in female or male BPD patients or healthy volunteers (p > .10).

4. Discussion The major findings of the present study are increased basal testosterone levels in a relatively large clinical sample of medication-free female and male patients with BPD as well as elevated cortisol awakening responses in female patients with BPD compared to healthy volunteers. Furthermore, cortisol awakening responses were positively correlated with trait anger and aggressiveness in female patients with BPD, while data did not indicate an association between anger/aggressiveness and testosterone in male and female BPD patients or healthy volunteers. Together with previous studies, the present results may suggest endocrine alterations including the stress and reproductive systems in BPD that may at least partly be related to the patients’ interpersonal impairments. Our results in terms of increased basal testosterone levels confirm and extend previously reported findings by Roepke et al. (2010) who reported elevated serum testosterone levels in a small group of partly medicated female BPD patients. Although female BPD patients in this single previous study had a significantly higher prevalence of PCO syndrome, this did not explain the increased testosterone levels in this group. In line with Roepke et al. (2010), we found no evidence for menstrual cycle, body mass index, smoking, or comorbid major depression and PTSD explaining the elevated testosterone levels in BPD which may rule out physiological factors influencing androgen metabolism, such as increased abdominal fat (Kahl et al., 2005), as only explanation of these results. In accordance with the present findings of increased testosterone levels in male and female BPD patients, this may hence indicate an altered HPG axis in BPD. It should however be mentioned that the mean testosterone values of 26.3 pg/ml in the female BPD patients and of 83.7 pg/ml in the male BPD patients are in the normal range observed in salivary samples (see, for instance, Popma et al., 2007; Carre and McCormick, 2008; Mehta et al., 2008; Mehta and Josephs, 2010) suggesting alterations in the finetuning rather than an absolute malfunctioning of the HPG axis in BPD. The functioning of the complex HPG reproductive system has been shown to be influenced by a variety of factors, including genetic factors, early life experiences including pre-, peri-, and postnatal maternal behavior and chronic early life stress, as well as acute stressful events in adulthood (for review, see Toufexis et al., 2014). Particularly maternal care has been reported to have a major impact on the HPG reproductive system — and the HPA stress system — by disturbing sexual differentiation during vulnerable phases of infantile development that may dysregulate the conversion of testosterone to estrogen via aromatase activity (MacLusky et al., 1987; Kawata, 1995). Consistent to this and other previous studies (e.g., Herman et al., 1989; Bertsch et al., 2013b), male and female BPD patients of the current study retrospectively reported significantly higher levels of traumatic childhood experiences compared to healthy volunteers, although there was no significant association

123 between testosterone or cortisol levels and severity of childhood traumatization within the patient samples. In addition, the enhanced frequency and intensity of acute daily stressful events that are experienced by patients with BPD (Jovev and Jackson, 2006; Kuo and Linehan, 2009) might also mediate HPG activity via activations of the HPA and the sympathetic stress system (Toufexis et al., 2014). Apart from stress-related modulations, increased testosterone levels might reflect a genetic vulnerability for BPD symptomatology considering the aspect of a possible genetic determination of modified androgen metabolism (Franks et al., 2000). Together with traumatic childhood experiences this might lead to an increased sensitivity for potential social threats (Domes et al., 2008; Bertsch et al., 2013a), which could trigger impulsive aggressive reactions (Mancke et al., in press) and may, on a neural level, be associated with increased and prolonged amygdala responses (Herpertz et al., 2001; Hazlett et al., 2012; Bertsch et al., 2013a) and reduced prefrontal regulation of the amygdala (Schulze et al., 2011). Although meta-analyses only suggest a weak positive correlation between testosterone and aggression across studies (r = .08, see Archer et al., 2005; Carre and Mehta, 2011), enhanced testosterone levels have been implicated in reduced prefronto-amygdalar inhibition (e.g., Volman et al., 2011) as well as competitive and dominant behaviors (Dabbs et al., 1991; Popma et al., 2007; Mehta and Josephs, 2010). This seems to be particularly the case if dominance behaviors (such as leadership or physical aggression in sports) are assessed and may reveal an important short-coming of the inflated use of the term ‘‘aggression’’ that ranges from paper-pencil tests for trait anger, hostile thoughts, or aggressiveness to actual physical aggression or antisocial behaviors (Mazur and Booth, 1998; Chichinadze et al., 2012). The fact that we did not find any significant association between trait anger or aggressiveness and basal testosterone levels in the current study might hence indicate the previously reported necessity to provoke and measure impulsive aggression experimentally and assess testosterone increase during the experiment (Carre and Mehta, 2011). Our results are also consistent with a study of Coccaro et al. (2007) who did not find significant associations between aggressiveness and testosterone levels in cerebrospinal fluid in a sample of 31 male patients with various personality disorders, including four patients with BPD. In addition, it should be mentioned that the relationship between testosterone and aggression is known to be stronger in those participants with low basal cortisol levels (Carre and Mehta, 2011), while no or only weak associations between testosterone and aggression were found in participants with high cortisol levels (Popma et al., 2007; Mehta and Josephs, 2010) — similar to those found in our female BPD patients. In line with our additional findings of highest testosterone levels in those BPD patients with low cortisol awakening responses, Carre and Mehta (2011) suggested that the testosterone-behavior pathway only works efficiently when HPA output is low. Following this, they suppose that high levels of environmental stress and consequently high HPA output may inhibit effects of testosterone on reproductively relevant behaviors, such as aggression and dominance, which might be metabolically inefficient or even dangerous in the context of acute stress, which could at least partly explain the weak associations between testosterone

124 and anger or aggressiveness in the current as well as in previous studies (Mehta and Josephs, 2010; Carre and Mehta, 2011). Before drawing strong conclusions, further studies are needed to disentangle the relationship between different facets of experimentally provoked and self-reported aggression and salivary testosterone levels, especially in clinical samples. This is particularly relevant with regard to the correlational results reported in the present study, which are based on rather small subgroups with inadequate statistical power and could therefore only be interpreted descriptively. The increased cortisol awakening responses in our female BPD patients are consistent with previous reports of exaggerated salivary cortisol responses to awakening, higher salivary cortisol levels over the day (Lieb et al., 2004; Carvalho Fernando et al., 2012), and enhanced overnight urinary cortisol levels in (female) patients with BPD (Wingenfeld et al., 2007). In the current study, this effect remained significant after controlling for comorbid diagnoses of major depression and PTSD as well as other possible confounding variables such as menstrual cycle, body-mass index, or smoking, and might together with previous studies suggest a BPD-associated hyperactivity of the HPA axis. It has been hypothesized that HPA axis alterations in BPD — and other psychiatric disorders — might be a consequence of early life stress (for review, see Wingenfeld et al., 2010). In line with previous studies (e.g., Herman et al., 1989), our BPD patients reported significantly increased levels of early traumatization although childhood traumatization was not significantly associated with cortisol awakening responses in female or male patients with BPD. Similarly, Carvalho Fernando et al. (2012) also found only weak positive associations between self-reported childhood traumatization and basal saliva cortisol levels and moderate associations with cortisol responses after dexamethasone administration in BPD. This could indicate a weak association between basal HPA hyperactivity and early traumatization in BPD, or, alternatively be due to a restricted variance regarding selfreported early traumatization — almost 80% of patients with BPD have a history of childhood traumatization — and/or basal cortisol levels in the patient group. In addition, HPA hyperactivity may fit to the reports of more frequent and more intense daily hassles and inner tension in patients with BPD and differentiate on an endocrine level patient with BPD from those with PTSD (without BPD), in whom reduced basal cortisol levels have been reported in various studies (for review and meta-analysis, see Morris et al., 2012). In female BPD patients, we additionally found indicators for positive associations between cortisol awakening responses and trait anger and aggressiveness. This is consistent with previous reports of anger and anger-related impulsive (auto)aggression being often triggered by social stressors and stress-associated inner tension (Kleindienst et al., 2008) suggesting interpersonal implications of HPA alterations in BPD. Investigating a relatively large clinical sample of female and male patients with BPD with similar patterns of comorbid disorders for the first time allowed an investigation of gender-specific differences in basal cortisol levels. Interestingly, we only found significantly elevated cortisol awakening responses in female patients with BPD, while male patients did not differ from healthy volunteers.

J. Rausch et al. Although stronger HPA hyperactivity has also been reported previously in female compared to male patients with major depression (Young et al., 1991; Matsuzaka et al., 2013) and associations between HPA axis activity and depression or work stress seem to be stronger in female than in male participants (e.g., Ruttle et al., 2014; Sjörs et al., 2014), the reasons for these gender differences remain mostly unclear. Gender-specific influences of early emotional maltreatment and/or of enhanced testosterone secretion on the functioning of the HPA axis and brain regions crucially involved in the regulation of threat processing and stress regulation (e.g., Arai et al., 1996; Russo et al., 2014) might be partial explanations. However, further investigations are needed to replicate the present results and to disentangle gender-specific alterations of basal HPA axis functioning in stress-associated disorders, such as BPD. The current study has several strengths, revealing elevated basal saliva testosterone levels in a relatively large clinical sample of medication-free female and male as well as enhanced cortisol awakening responses in female patients with BPD. However, several limitations should be noted. First, diagnostic interviews revealed a number of comorbidities in the patient sample, which reflects a typical pattern of comorbid psychiatric disorders in BPD patients. Importantly, female and male patients did not differ in the pattern of comorbid disorders and the statistical control for major depression and PTSD did not change the effects. However, due to the lack of a clinical control group, borderline-specific conclusions should be drawn with caution and only with reference to previous studies which either excluded BPD patients with comorbid disorders (e.g., depression in Lieb et al., 2004), included clinical control groups (e.g., with depression; Carvalho Fernando et al., 2012), or subgroups of BPD patients with and without PTSD symptomatology (Wingenfeld et al., 2010). Second, despite matching the samples with regard to a number of variables, we could not avoid significant differences between patients and volunteers in smoking and between female and male participants with regard to intelligence and bodymass index. We therefore statistically controlled for these as well as other possible confounding variables, which however did not change the pattern of results. Third, the present results indicate alterations in basal functioning of two major endocrine systems, the HPG and the HPA axis, however, leaving out other associated and interacting hormones, such as estradiol or oxytocin, the latter of which also seems to be altered in BPD (Bertsch et al., 2013b) and in women with early childhood traumatization (Heim et al., 2009). Further studies assessing the basal functioning, reactivity, and interaction of several endocrine systems are urgently needed. Fourth, because of the mean group differences in endocrine and self-report data, we had to perform separate correlational analyses for the four groups. Thus, the results of these analyses should only be interpreted descriptively and need to be replicated in larger samples before strong conclusions can be drawn. Finally, aggressiveness was only assessed with a self-report questionnaire, which might be problematic due to enhanced alexithymia (New et al., 2012) in BPD. In several previous studies correlations between endocrine parameters and aggression were strongest, if aggression was experimentally provoked and measured.

Testosterone and cortisol in BPD

5. Summary and conclusions Taken together, the results of the present study indicate increased basal saliva testosterone levels in a relatively large clinical sample of medication-free female and male patients with BPD as well as elevated cortisol awakening responses in female patients with BPD compared to a well-matched group of healthy women and men. In addition, cortisol awakening responses were positively related to trait anger and aggressiveness in female patients with BPD, but no associations were found with basal testosterone levels. In line with previous reports, the present results suggest endocrine alterations including the HPG reproductive and the HPA stress system in BPD, which may be associated with interpersonal impairments, such as increased anger-related aggressive behavior. The examination of wellcharacterized female and male BPD patients and healthy volunteers allowed the investigation of gender differences while controlling for influences of stress-associated comorbid disorders such as major depression and PTSD. Endocrine alterations may not only indicate long-lasting consequences of traumatic childhood experiences, but might also have implications for the development of new (psychopharmaco) therapeutic interventions that could help to restore the alterations in the HPA and HPG systems of patients with BPD.

Role of funding sources The study was funded by the German Research Foundation (DFG; He 2660/12-1; He 2660/7-2).

Conflict of interest The authors declare no conflict of interest.

Acknowledgement The study was part of the clinical research unit 256 (www.kfo256.de; Schmahl et al., 2014), funded by the German Research Foundation (DFG; He 2660/12-1; He 2660/7-2).

References American Psychiatric Association, 2013. Diagnostic and Statistical Manual of Mental Disorders, 5th ed. American Psychiatric Publishing (APP), Arlington, VA. Arai, Y., Sekine, Y., Murakami, S., 1996. Estrogen and apoptosis in the developing sexually dimorphic preoptic area in female rats. Neurosci. Res. 25, 403—407. Archer, J., Graham-Kevan, N., Davies, M., 2005. Testosterone and aggression: a reanalysis of Book, Starzyk, and Quinsey’s (2001) study. Aggress. Violent Behav. 10, 241—261. Beck, A.T., Steer, R.A., 1995. Beck-Depressions-Inventar: BDI, 2., überarb. Aufl. Huber/Bern, Göttingen. Bernstein, D.P., Fink, L., 1998. Childhood Traume Questionnaire. A Retrospective Self-report: Maunal. The Psychological Corporation, Orlando. Bertsch, K., Gamer, M., Schmidt, B., Schmidinger, I., Walther, S., Kastel, T., Schnell, K., Buchel, C., Domes, G., Herpertz, S.C., 2013a. Oxytocin and reduction of social threat hypersensitivity

125 in women with borderline personality disorder. Am. J. Psych. 170, 1169—1177. Bertsch, K., Schmidinger, I., Neumann, I., Herpertz, S.C., 2013b. Reduced plasma oxytocin levels in female patients with borderline personality disorder. Horm. Behav. 63, 424—429. Böhnke, R., Bertsch, K., Kruk, M., Naumann, E., 2010a. The relationship between basal and acute HPA axis activity and aggressive behavior in adults. J. Neural Transm. 117, 629—637. Böhnke, R., Bertsch, K., Kruk, M.R., Richter, S., Naumann, E., 2010b. Exogenous cortisol enhances aggressive behavior in females, but not in males. Psychoneuroendocrinology 35, 1034—1044. Bohus, M., Kleindienst, N., Limberger, M.F., Stieglitz, R.D., Domsalla, M., Chapman, A.L., Steil, R., Philipsen, A., Wolf, M., 2009. The short version of the Borderline Symptom List (BSL23): development and initial data on psychometric properties. Psychopathology 42, 32—39. Buss, A.H., Perry, M., 1992. The aggression questionnaire. J. Pers. Soc. Psychol. 63, 452—459. Carre, J.M., McCormick, C.M., 2008. Aggressive behavior and change in salivary testosterone concentrations predict willingness to engage in a competitive task. Horm. Behav. 54 (3), 403—409. Carre, J.M., Mehta, P.H., 2011. Importance of considering testosterone—cortisol interactions in predicting human aggression and dominance. Aggress. Behav. 37, 489—491. Carvalho Fernando, S., Beblo, T., Schlosser, N., Terfehr, K., Otte, C., Lowe, B., Wolf, O.T., Spitzer, C., Driessen, M., Wingenfeld, K., 2012. Associations of childhood trauma with hypothalamic—pituitary—adrenal function in borderline personality disorder and major depression. Psychoneuroendocrinology 37, 1659—1668. Chichinadze, K., Lazarashvili, A., Chichinadze, N., Gachechiladze, L., 2012. Testosterone dynamics during encounter: role of emotional factors. J. Comp. Physiol. A: Neuroethol. Sens. Neural Behav. Physiol. 198, 485—494. Clow, A., Thorn, L., Evans, P., Hucklebridge, F., 2004. The awakening cortisol response: methodological issues and significance. Stress 7, 29—37. Coccaro, E.F., Beresford, B., Minar, P., Kaskow, J., Geracioti, T., 2007. CSF testosterone: relationship to aggression, impulsivity, and venutresomeness in adult males with personality disorder. J. Psychiatr. Res. 41, 488—492. Dabbs Jr., J.M., Jurkovic, G.J., Frady, R.L., 1991. Salivary testosterone and cortisol among late adolescent male offenders. J. Abnorm. Child Psychol. 19, 469—478. Domes, G., Czieschnek, D., Weidler, F., Berger, C., Fast, K., Herpertz, S.C., 2008. Recognition of facial affect in Borderline Personality Disorder. J. Pers. Disord. 22, 135—147. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., 1995. Structured Clinical Interview for DSM-IV (SCID-I). Biometrics Research Department, New York State Psychiatric Institute, New York. Franks, S., Gilling-Smith, C., Gharani, N., McCarthy, M., 2000. Pathogenesis of polycystic ovary syndrome: evidence for a genetically determined disorder of ovarian androgen production. Hum. Fertil. (Camb.) 3, 77—79. Garner, B., Chanen, A.M., Phillips, L., Velakoulis, D., Wood, S.J., Jackson, H.J., Pantelis, C., McGorry, P.D., 2007. Pituitary volume in teenagers with first-presentation borderline personality disorder. Psychiatry Res. 156, 257—261. Gerra, G., Zaimovic, A., Raggi, M.A., Moi, G., Branchi, B., Moroni, M., et al., 2007. Experimentally induced aggressiveness in heroin-dependent patients treated with buprenophine: comparison of patients receiving metahdone and healthy subjects. Psychiatry Res. 149, 201—213. Haller, J., 2013. The neurobiology of abnormal manifestations of aggression — a review of hypothalamic mechanisms in cats, rodents, and humans. Brain Res. Bull. 93, 97—109.

126 Haller, J., 2014. The gucocorticoid/aggression relationship in animals and humans: an analysis sensitive to behavioral characteristics, glucocorticoid secretion patterns, and neural mechanisms. Curr. Top. Behav. Neurosci. 17, 73—109. Hazlett, E.A., New, A.S., Newmark, R., Haznedar, M.M., Lo, J.N., Speiser, L.J., Chen, A.D., Mitropoulou, V., Minzenberg, M., Siever, L.J., Buchsbaum, M.S., 2005. Reduced anterior and posterior cingulate gray matter in borderline personality disorder. Biol. Psychiatry 58, 614—623. Hazlett, E.A., Zhang, J., New, A.S., Zelmanova, Y., Goldstein, K.E., Haznedar, M.M., Meyerson, D., Goodman, M., Siever, L.J., Chu, K.-W., 2012. Potentiated amygdala response to repeated emotional pictures in borderline personality disorder. Biol. Psychiatry 72, 448—456. Heim, C., Young, L.J., Newport, D.J., Mletzko, T., Miller, A.H., Nemeroff, C.B., 2009. Lower CSF oxytocin concentrations in women with a history of childhood abuse. Mol. Psychiatry 14, 954—958. Hellhammer, J., Fries, E., Schweisthal, O.W., Schlotz, W., Stone, A.A., Hagemann, D., 2007. Several daily measurements are necessary to reliably assess the cortisol rise after awakening: stateand trait components. Psychoneuroendocrinology 32, 80—86. Herman, J.L., Perry, J.C., van der Kolk, B.A., 1989. Childhood trauma in borderline personality disorder. Am. J. Psych. 146, 490—495. Herpertz, S.C., Dietrich, T.M., Wenning, B., Krings, T., Erberich, S.G., Willmes, K., Thron, A., Sass, H., 2001. Evidence of abnormal amygdala functioning in borderline personality disorder: a functional MRI study. Biol. Psychiatry 50, 292—298. Huynh, H., Feldt, L.S., 1976. Estimation of the box correction for degrees of freedom from sample data in randomized block and split-plot designs. J. Educ. Stat. 1, 69—82. Jovev, M., Jackson, H.J., 2006. The relationship of borderline personality disorder, life events and functioning in an Australian psychiatric sample. J. Pers. Disord. 20, 205—217. Kahl, K.G., Bester, M., Greggersen, W., Rudolf, S., Dibbelt, L., Stoeckelhuber, B.M., Gehl, H.B., Sipos, V., Hohagen, F., Schweiger, U., 2005. Visceral fat deposition and insulin sensitivity in depressed women with and without comorbid borderline personality disorder. Psychosom. Med. 67, 407—412. Kawata, M., 1995. Roles of steroid hormones and their receptors in structural organization in the nervous system. Neurosci. Res. 24, 1—46. Kleindienst, N., Bohus, M., Ludascher, P., Limberger, M.F., Kuenkele, K., Ebner-Priemer, U.W., Chapman, A.L., Reicherzer, M., Stieglitz, R.D., Schmahl, C., 2008. Motives for nonsuicidal selfinjury among women with borderline personality disorder. J. Nerv. Ment. Dis. 196, 230—236. Kudielka, B.M., Gierens, A., Hellhammer, D.H., Wüst, S., Schlotz, W., 2012. Salivary cortisol in ambulatory assessment—–some dos, some don’ts, and some open questions. Psychosom. Med. 74, 418—431. Kuhlmann, A., Bertsch, K., Schmidinger, I., Thomann, P.A., Herpertz, S.C., 2013. Morphometric differences in central stress-regulating structures between women with and without borderline personality disorder. J. Psychiatry Neurosci. 38, 129—137. Kuo, J.R., Linehan, M.M., 2009. Disentangling emotion processes in borderline personality disorder: physiological and self-reported assessment of biological vulnerability, baseline intensity, and reactivity to emotionally evocative stimuli. J. Abnorm. Psychol. 118, 531—544. Lieb, K., Rexhausen, J.E., Kahl, K.G., Schweiger, U., Philipsen, A., Hellhammer, D.H., Bohus, M., 2004. Increased diurnal salivary cortisol in women with borderline personality disorder. J. Psychiatr. Res. 38, 559—565. Loranger, A.W., Sartorius, N., Andreoli, A., Berger, P., Buchheim, P., Channabasavanna, S.M., Coid, B., Dahl, A., Diekstra, R.F.,

J. Rausch et al. Ferguson, B., et al., 1994. The International Personality Disorder Examination. The World Health Organization/Alcohol, Drug Abuse, and Mental Health Administration international pilot study of personality disorders. Arch. Gen. Psychiatry 51, 215—224. MacLusky, N.J., Clark, A.S., Naftolin, F., Goldman-Rakic, P.S., 1987. Estrogen formation in the mammalian brain: possible role of aromatase in sexual differentiation of the hippocampus and neocortex. Steroids 50, 459—474. Mancke, F., Herpertz, S.C., Bertsch, K., 2015. Aggression in borderline personality disorder — a multidimensional model. Pers. Disord.: Theory Res. Treat., http://dx.doi.org/10.1037/ per0000098 (in press). Matsuzaka, H., Maeshima, H., Kida, S., Kurita, H., Shimano, T., Nakano, Y., Baba, H., Suzuki, T., Arai, H., 2013. Gender differences in serum testosterone and cortisol in patients with major depressive disorder compared with controls. Int. J. Psychiatry Med. 46, 203—321. Mazur, A., Booth, A., 1998. Testosterone and dominance in men. Behav. Brain Sci. 21, 353—363, discussion 363-397. Mehta, P.H., Jones, A.C., Josephs, R.A., 2008. The social endocrinology of dominance: basal testosterone predicts cortisol changes and behavior following victory and defeat. J. Pers. Soc. Psychol. 94 (6), 1078—1093. Mehta, P.H., Josephs, R.A., 2010. Testosterone and cortisol jointly regulate dominance: evidence for a dual-hormone hypothesis. Horm. Behav. 58, 898—906. Minzenberg, M.J., Fan, J., New, A.S., Tang, C.Y., Siever, L.J., 2008. Frontolimbic structural changes in borderline personality disorder. J. Psychiatr. Res. 42, 727—733. Morris, M.C., Compas, B.E., Garber, J., 2012. Relations among posttraumatic stress disorder, comorbid depression, and HPA function: a systematic review and meta-analysis. Clin. Psychol. Rev. 32, 301—315. New, A.S., aan het Rot, M., Ripoll, L.H., Perez-Rodriguez, M.M., Lazarus, S., Zipursky, E., Weinstein, S.R., Koenigsberg, H.W., Hazlett, E.A., Goodman, M., Siever, L.J., 2012. Empathy and alexithymia in borderline personality disorder: clinical and laboratory measures. J. Pers. Disord. 26, 660— 675. Nunes, P.M., Wenzel, A., Borges, K.T., Porto, C.R., Caminha, R.M., de Oliveira, I.R., 2009. Volumes of the hippocampus and amygdala in patients with borderline personality disorder: a meta-analysis. J. Pers. Disord. 23, 333—345. Ogata, S.N., Silk, K.R., Goodrich, S., Lohr, N.E., Westen, D., Hill, E.M., 1990. Childhood sexual and physical abuse in adult patients with borderline personality disorder. Am. J. Psych. 147, 1008—1013. Popma, A., Vermeiren, R., Geluk, C.A., Rinne, T., van den Brink, W., Knol, D.L., Jansen, L.M., van Engeland, H., Doreleijers, T.A., 2007. Cortisol moderates the relationship between testosterone and aggression in delinquent male adolescents. Biol. Psychiatry 61, 405—411. Pruessner, J.C., Kirschbaum, C., Meinlschmid, G., Hellhammer, D.H., 2003. Two formulas for computation of the area under the curve represent measures of total hormone concentration versus time-dependent change. Psychoneuroendocrinology 28, 916—931. Raven, J.C., Horn, R., 2009. Standard progressive matrices: SPM (SPM-C/SPM-P/SPM Plus)Raven’s progressive matrices und vocabulary scales [2. Aufl.]. Pearson, Frankfurt, p. Test. Roepke, S., Ziegenhorn, A., Kronsbein, J., Merkl, A., Bahri, S., Lange, J., Lubbert, H., Schweiger, U., Heuser, I., Lammers, C.H., 2010. Incidence of polycystic ovaries and androgen serum levels in women with borderline personality disorder. J. Psychiatr. Res. 44, 847—852. Russo, M., Proujansky, R., Gilbert, A., Braga, R.J., Burdick, K.E., 2014. Initial evidence for sex-specific effects of early emotional

Testosterone and cortisol in BPD abuse on affective processing in bipolar disorder. Eur. Psychiatry 29, 52—57. Ruttle, P.L., Klein, M.H., Slattery, M.J., Kalin, N.H., Armstrong, J.M., Essex, M.J., 2014. Adolescent adrenocortical activity and adiposity: differences by sex and exposure to early maternal depression. Psychoneuroendocrinology 47, 68—77. Schmahl, C., Herpertz, S.C., Bertsch, K., Ende, G., Flor, H., Kirsch, P., Meyer-Lindenberg, A., Rietschel, M., Schneider, M., Spanagel, R., Treede, R.-D., Bohus, M., 2014. Mechanisms of disturbed emotion processing and social interaction in borderline personality disorder: state of knowledge and research agenda of the German Clinical Research Unit. Borderline Pers. Disord. Emot. Dysregul. 1 (12), http://dx.doi.org/10.1186/2051-6673-1-12. Schmidt-Reinwald, A., Pruessner, J.C., Hellhammer, D.H., Federenko, I., Rohleder, N., Schurmeyer, T.H., Kirschbaum, C., 1999. The cortisol response to awakening in relation to different challenge tests and a 12-hour cortisol rhythm. Life Sci. 64, 1653—1660. Schultheiss, O.C., Stanton, S.J., 2009. Assessment of salivary hormones. In: Harmon-Jones, E., Beer, J.S. (Eds.), Methods in the Neurobiology of Social and Personality Psychology. Guilford, New York, NY, pp. 17—44. Schulze, L., Domes, G., Krüger, A., Berger, C., Fleischer, M., Prehn, K., Schmahl, C., Grossmann, A., Hauenstein, K., Herpertz, S.C., 2011. Neuronal correlates of cognitive reappraisal in borderline patients with affective instability. Biol. Psychiatry 69, 564—573. Schwenkmezger, P., Hodapp, V., Spielberger, C.D., 1992. Das State-Trait-Ärgerausdrucks-Inventar STAXI: Handbuch, 1. Aufl. Huber/Bern, Göttingen/Toronto. Sjörs, A., Ljung, T., Jonsdottir, I.H., 2014. Diurnal salivary cortisol in relation to perceived stress at home and at work in healthy men and women. Biol. Psychol. 99, 193—197.

127 Toufexis, D., Rivarola, M.A., Lara, H., Viau, V., 2014. Stress and the reproductive axis. J. Neuroendocrinol. 26, 573—586. van Goozen, S.H.M., Fairchild, G., Snoek, H., Harold, G.T., 2007. The evidence for a neurobiological model of childhood antisocial behavior. Psychol. Bull. 133, 149—182. Van Honk, J., Harmon-Jones, E., Morgan, B.E., Schutter, D.J., 2010. Socially explosive minds: the triple imbalance hypothesis of reactive aggression. J. Pers. 78, 67—94. Volman, I., Toni, I., Verhagen, L., Roelofs, K., 2011. Endogenous testosterone modulates prefrontal-amygdala connectivity during social emotional behavior. Cereb. Cortex 21, 2282— 2290. Wingenfeld, K., Hill, A., Adam, B., Driessen, M., 2007. Dexamethasone suppression test in borderline personality disorder: impact of PTSD symptoms. Psychiatry Clin. Neurosci. 61, 681— 683. Wingenfeld, K., Spitzer, C., Rullkotter, N., Lowe, B., 2010. Borderline personality disorder: hypothalamus pituitary adrenal axis and findings from neuroimaging studies. Psychoneuroendocrinology 35, 154—170. Wust, S., Wolf, J., Hellhammer, D.H., Federenko, I., Schommer, N., Kirschbaum, C., 2000. The cortisol awakening response — normal values and confounds. Noise Health 2, 79—88. Young, E.A., Haskett, R.F., Murphy-Weinberg, V., Watson, S.J., Akil, H., 1991. Loss of glucocorticoid fast feedback in depression. Arch. Gen. Psychiatry 48, 693—699. Zimmerman, D.J., Choi-Kain, L.W., 2009. The hypothalamic—pituitary—adrenal axis in borderline personality disorder: a review. Harv. Rev. Psychiatry 17 (3), 167—183.