Increased psychological and attenuated cortisol and alpha-amylase responses to acute psychosocial stress in female patients with borderline personality disorder

Psychoneuroendocrinology (2010) 35, 1565—1572

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

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

Increased psychological and attenuated cortisol and alpha-amylase responses to acute psychosocial stress in female patients with borderline personality disorder Urs M. Nater a,1,*, Martin Bohus b,1, Elvira Abbruzzese a, Beate Ditzen a, Jens Gaab a, Nikolaus Kleindienst b, Ulrich Ebner-Priemer b, Jana Mauchnik b, Ulrike Ehlert a a b

University of Zurich, Institute of Psychology, Dept. of Clinical Psychology & Psychotherapy, Switzerland Central Institute of Mental Health, Dept. of Psychosomatic Medicine and Psychotherapy, Mannheim, Germany

Received 2 December 2008; received in revised form 2 June 2010; accepted 2 June 2010

KEYWORDS Borderline personality disorder; Stress; Cortisol; Alpha-amylase; ACTH; Catecholamines

Summary Objective: Borderline personality disorder (BPD) is characterized by increased self-reported stress and emotional responding. Knowledge about the psychological and physiological mechanisms that underlie these experiences in BPD patients is scarce. The objective was to assess both psychological and endocrinological responses to a standardized psychosocial stressor in female BPD patients and healthy controls. Methods: A total of 15 female BPD patients and 17 healthy control subjects were included in a case—control study. All subjects were free of any medication, had a regular menstrual cycle, and were investigated during the luteal phase of their menstrual cycle. Co-occurring current major depression, current substance abuse/dependence, and lifetime schizophrenia or bipolar I disorder were excluded. Psychological measures of stress, salivary cortisol, salivary alpha-amylase, plasma ACTH, plasma norepinephrine and epinephrine concentrations were measured before, during, and after exposure to a standardized psychosocial stress protocol. Results: BPD patients displayed maladaptive cognitive appraisal processes regarding the upcoming stressor as well as significantly higher subjective stress, coupled with a substantial cortisol and alpha-amylase hyporeactivity to the stressor in comparison to the controls. No significant differences for ACTH and catecholaminergic responses were observed, while the ACTH:cortisol ratio was higher in BPD patients than in controls.

* Corresponding author at: University of Zurich, Institute of Psychology, Clinical Psychology & Psychotherapy, Binzmuehlestrasse 14/Box 26, 8050 Zurich, Switzerland. Tel.: +41 44 635 7382; fax: +41 44 635 7359. E-mail address: [email protected] (U.M. Nater). 1 These authors contributed equally to this work. 0306-4530/$ — see front matter # 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.psyneuen.2010.06.002

1566

U.M. Nater et al. Conclusions: Attenuated cortisol responsiveness in BPD patients might in part be explained by decreased adrenal responsiveness to endogenous ACTH and altered central noradrenergic activation as reflected by alpha-amylase. # 2010 Elsevier Ltd. All rights reserved.

1. Introduction Borderline personality disorder (BPD) presents as a heterogeneous constellation of symptoms with dysfunctional affect regulation at its core (Lieb et al., 2004b). Several studies clearly indicate that individuals with BPD experience higher emotional reactivity, emotional intensity, and slower return to baseline of affective arousal, compared to healthy controls and individuals with other personality disorders (Koenigsberg et al., 2002; Stiglmayr et al., 2005; Ebner-Priemer et al., 2007). BPD patients are further characterized by displaying dysfunctional behavior such as non-suicidal selfinjury as a strategy to regulate these intense levels of distress (Kleindienst et al., 2008). Stressful life events have been reported to accumulate under daily life conditions in adult BPD patients (Zanarini et al., 2005). Interestingly, experimental studies clearly demonstrate that central stress coping mechanisms fail within this group of patients. For instance, functional neuroimaging studies in BPD patients revealed strong evidence for alterations of the limbic circuitry including insula, hippocampus, and amygdala hyperreactivity, and alterations in frontal and prefrontal regions (Mauchnik et al., 2005). The neurochemical underpinnings of these limbic dysregulations, however, are unclear. There is general agreement that the stress hormones cortisol and corticosterone target the limbic brain affecting cognitive processes and emotional arousal (De Kloet et al., 2005). Recent evidence shows that corticosteroid hormones exert both rapid and slow non-genomic effects on neurons in the hippocampus, the amygdale, and frontal and prefrontal regions. According to De Kloet et al. (2008), corticosteroids seem to modulate a biphasic stress response: during the initial phase corticosteroids promote hippocampal excitability and amplify the effect of other stress hormones. These fast responses are complemented by slower glucocorticoid receptor mediated effects which facilitate downregulation of temporarily raised excitability and recovery from stressful experience. From this perspective, both cortisol hyper- or hypo-secretion could play a role in promoting affective dysregulation in BPD patients. However, so far, studies examining alterations of the hypothalamic—pituitary—adrenal (HPA) axis in BPD have been scarce. It has been fairly consistently shown that dexamethasone suppression (DST) in BPD is neither particularly sensitive nor specific (Lahmeyer et al., 1988; Korzekwa et al., 1991; De la Fuente and Mendlewicz, 1996). The majority of these studies included BPD patients with comorbid major depressive disorder (MDD) or posttraumatic stress disorder (PTSD), which both are known to affect the HPA axis (Steckler et al., 1999; Yehuda, 2002). In a more naturalistic study, Lieb et al. (2004a) collected salivary cortisol during daily life conditions for 14 h after awakening. Patients with BPD displayed significantly higher salivary cortisol levels than healthy controls as demonstrated by higher total cortisol in response to awakening and higher total daily cortisol levels. However,

no correlation could be found between monitored subjective levels of distress and related salivary cortisol levels. We are aware of two published studies addressing physiological responses to experimentally induced acute stress in BPD, both showing unclear results. In one study (Simeon et al., 2007), 13 BPD patients and 11 healthy controls were exposed to the Trier Social Stress Test, a standardized psychosocial stressor (TSST; Kirschbaum et al., 1993). No difference in cortisol stress reactivity between BPD patients and healthy controls was found, but BPD patients with high levels of long-term dissociation showed elevated plasma cortisol peaks compared with low dissociation BPD patients and controls. In the second study (Walter et al., 2008), BPD patients (n = 9) did not show different cortisol stress reactivity levels to an interpersonal conflict task in comparison with 12 healthy controls, but displayed elevated cortisol levels during recovery. However, in both studies the samples consisted of male and female subjects and the menstrual cycle of the latter were not timed. Since this variable has a strong impact on cortisol secretion during the TSST (Kirschbaum et al., 1999), the interpretation of the data is difficult. Thus, it remains to be determined whether BPD patients display altered HPA axis stress reactivity. The same is true for information on the autonomic nervous system (ANS) in BPD. Emotional dysregulation has been shown to correlate with imbalances in noradrenergic neurotransmission, although the findings are mixed (Gurvits et al., 2000). Ebner-Priemer et al. (2008) assessed distress and heart rates in 50 BPD patients and 50 healthy controls under 24 h daily life conditions. BPD patients reported significantly more distress and displayed elevated heart rates, but the correlation between physiological arousal and psychological distress were equal in both groups. Enhanced ANS reactivity, however, cannot be excluded, since most of the patients were on medication. The lack of sufficient studies measuring ANS parameters might in part be explained by the insufficient availability of markers that reliably reflect sympathetic activity (Grassi and Esler, 1999). With the recent suggestion of the salivary enzyme alpha-amylase being a valid and reliable marker of central sympathetic activity (Ehlert et al., 2006; Nater and Rohleder, 2009; Rohleder and Nater, 2009) it is now possible to measure both HPA axis and ANS activity in a convenient and non-invasive manner (Nierop et al., 2006). Summarizing, there is evidence that BPD patients display insufficient cognitive and neural stress regulation, but the role of neither the HPA axis nor the ANS is sufficiently studied. The objective of the current study was therefore to assess both psychological and neuroendocrine responses to a standardized psychosocial stressor in a female population of BPD patients and healthy subjects, controlled for known covariates. We hypothesized (1) that BPD patients experience enhanced subjective stress in face of an acute psychosocial stressor, compared to healthy controls; (2) that secretion of cortisol and ACTH in BPD is disturbed, indicating dysregulated

Acute stress response in BPD stress reactivity of the HPA axis; and (3) that ANS reactivity would be increased in BPD patients.

2. Materials and methods 2.1. Subjects Fifteen subjects meeting DSM-IV criteria for BPD and 17 healthy control subjects participated in the study. Due to the female preponderance in treatment seeking subjects with BPD, only female participants were included. Subjects presenting features of borderline personality were recruited via Internet advertisement and screened by telephone interview. Subjects who fulfilled DSM-IV criteria for borderline personality disorder, and did not present current alcohol/ substance abuse, acute medication, and current major depression were eligible to participate. Diagnosis of BPD and axis II disorders was determined by the International Personality Disorder Examination (IPDE) (Loranger, 1999). The IPDE was assessed by experienced and trained psychologists. Interrater reliability was 0.85. Axis I diagnoses were assessed by a trained psychologist (JM) using the Structured Clinical Interview for DSM-IV axis I disorders (DSM-IV, SCID) (First et al., 1997). Comorbid current axis I disorders in the BPD sample were panic disorder (n = 5), agoraphobia without panic disorder (n = 4), specific phobia (n = 1), social phobia (n = 6), generalized anxiety disorder (n = 2), posttraumatic stress disorder (n = 5), and obsessive-compulsive disorder (n = 2). Additional axis II disorders were paranoid personality disorder (PD) (n = 3), avoidant PD (n = 8), dependent PD (n = 1), and obsessive-compulsive PD (n = 1). All participants were free of major medical illness according to medical history and physical examination, and did not show any substance or alcohol abuse or dependence within 6 months prior to the study. Participants had also been free of psychotropic medication for at least 8 weeks prior to the beginning of the study. Further exclusion criteria were hormonal contraceptives, irregular menstrual cycle, current and lifetime schizophrenia, current and lifetime bipolar disorder I, current major depressive disorder, current substance abuse, and current anorexia (BMI < 17.5). Healthy control participants were recruited via advertisement and were required to be free of any axis I or II disorder (SKID II). Otherwise, the same exclusion criteria were applied to the control subjects as to the BPD patients. Demographic information (age, education) was collected from all participants. The subjects were asked not to undergo excessive physical activity for the 48 h prior to the experiment and to refrain from any sporting activities at all for the 24 h before the study. Intake of alcohol was prohibited for the 48 h prior to the experiment. Caffeine, tea and smoking were not permitted within 2 h prior to the study. At least 60 min before the study, subjects had to refrain from brushing their teeth (to avoid gingival bleeding) or eating. To minimize variations of physiological variables due to time of day and circadian rhythm, all experiments were carried out in the afternoon (between 2 and 4 PM). All participants took part in the experiment during their late luteal phase (i.e., around the 25th day of the menstrual cycle) to minimize hormonal variations across the menstrual cycle. To control for a possible modulation of endocrine measures by changes of estrogen or other hormones during the menstrual cycle, only

1567 subjects with regular menstrual cycles were included. All subjects were residents of Germany or the German speaking part of Switzerland. The subjects were remunerated for participation in the study with 100 Swiss Francs or 60 Euro, respectively. After receiving a thorough explanation of the study, subjects provided written informed consent for participation. The study was approved by the ethics committees of the University of Zurich and the University of Heidelberg Medical School.

2.2. Stress challenge A standardized psychosocial stress protocol, the Trier Social Stress Test (TSST), was used (Kirschbaum et al., 1993). This test comprises a mock job interview (5 min) and a mental arithmetic task (5 min) in front of an audience. In the current study, participants were introduced to the audience 20 min prior to the actual start of the TSST. Briefly, the upcoming task of giving a job interview was explained to the subjects. Then, the subject was accompanied to an adjacent room in which she could prepare the task ahead. All laboratory assessments took place at the University of Zurich, Switzerland. Saliva samples were taken 35 and 20 min before the stressor, as well as immediately before the stressor, and 0, 10, 20, 30, 40, 50, 60, 70, and 80 min after the stressor. Blood samples (from the indwelling venous catheter) were taken immediately before and after the TSST, with further samples taken 0, 10, 20, 30, 40, 50, 60, 70 min after the stressor. The TSST was always performed between 2 and 4 PM. All procedures applied to both patients and controls. Three of initially 18 patients were reluctant to perform the stress test, although they were otherwise compliant with the rest of the protocol. Thus, in the following analyses, data from 15 patients and 17 controls are included. The three non-compliant participants did not differ from the compliant ones in any measure.

2.3. Measures 2.3.1. Psychological measures Stress-related anticipatory cognitive appraisal processes were assessed with the Primary Appraisal Secondary Appraisal (PASA) (Gaab et al., 2005) questionnaire. This questionnaire is suitable to assess each of the two main cognitive appraisal processes relevant in an acute stress situation: primary stress appraisal (‘‘Is the situation a threat or challenge?’’), and secondary appraisal: (‘‘What possibilities do I have to cope with the stress situation adequately?’’). The PASA was administered after the introduction to the social stressor. Internal consistency for both scales are moderate to high (Crohnbach’s alpha = 0.80 and 0.74, respectively). The State-Trait Anxiety Inventory (STAI) (Laux et al., 1981) was used to assess psychological stress responses to the TSST. The STAI comprises two scales: the trait and the state form, indicating the presence or absence of anxiety symptoms. In the current study, only results for STAI state are reported. STAI state was administered immediately before and after the social stressor. Internal consistency scores among a variety of samples are high (Crohnbach’s alpha > 0.90). Finally, the stressfulness of the TSST was assessed by a visual analogue scale (VAS) administered immediately after the social stressor.

1568 Table 1

U.M. Nater et al. Demographic characteristics. Patients (n = 15)

Controls (n = 17)

Statistics

Mean

SD

Mean

SD

Age BMI

32.6 24.9

7.8 6.5

27.2 21.4

6.2 2.3

Smoking (no. cig.) <10/day >10/day

11 4

17 0

x2 = 5.1; p = .038

Education (years)

11

12

Z=

2.3.2. Physiological measures Salivary samples for the assessment of both cortisol and alpha-amylase were collected using Salivette (Sarstedt) collection devices and stored at 20 8C after completion of the session until biochemical analysis took place. Salivary free cortisol was analyzed by using a commercial chemiluminescence immunoassay (LIA) (IBL Hamburg, Germany). Interand intraassay coefficients of variation were below 10%. To reduce error variance caused by imprecision of the intraassay, all samples of one subject were analyzed in the same run. Salivary alpha-amylase (sAA) activity was determined using the automatic analyser Cobas Mira and assay kits obtained from Roche, as previously described (Nater et al., 2006). The assay is a kinetic colorimetric test. Inter- and intraassay variance was below 1%. Fifty minutes prior to the TSST, a catheter was inserted into the antecubital vein. The catheter was kept patent with infusions of saline. Catheter insertion and subsequent blood collection failed in 4 patients due to inaccessibility of the veins. Thus, results for ACTH, NE, and EP are reported for 11 patients. Blood samples were taken with EDTA-coated monovettes (Sarstedt, Sevelen, Switzerland), and plasma was pipetted in pre-cooled aliquots. Samples were stored at 80 8C. Plasma ACTH from the blood samples collected during the experiment was analyzed by a two-site commercial chemiluminescence assay (Nichols, Bad Nauheim, Germany). Plasma norepinephrine and epinephrine were determined by means of HPLC and electrochemical detection after liquid—liquid extraction (as recently described Ehrenreich et al., 1997). The limit of detection was 10 pg/ml. Inter- and intraassay variance was lower than 5% for both epinephrine and norepinephrine. 2.3.3. Statistical analyses Two-way analyses of variance (ANOVAs) for repeated measures were computed to reveal possible time and group effects. Mean differences were calculated with Student’s t-tests. Data were tested for normal distribution and homogeneity of variance using a Kolmogorov—Smirnov and Levene’s test. All reported ANOVA results were corrected by the Greenhouse—Geisser procedure where appropriate (violation of sphericity assumption), reflected by the degrees of freedom with decimal values. Mann—Whitney U-test was used for variables that were not normally distributed. For all analyses, the significance level was a = 5%. Unless indicated, all results shown are means  standard error of means (SEM).

Z= Z=

1.9; p = .053 1.2; p = .25

1.8; p = .13

3. Results 3.1. Subject characteristics The two groups did not significantly differ with regard to age, BMI and education. There were slightly more heavy smokers (>10 cigarettes/day) in the patient group (n = 4) than in the control group (n = 0) (Table 1).

3.2. Psychological stress responses Patients and controls differed significantly in their primary appraisal. Patients interpreted the situation rather as a threat than as a challenge (3.55 for controls vs. 4.40 for patients; t23.31 = 2.95, p = 0.007). The two groups also differed regarding their secondary appraisal (patients saw less possibilities to cope with the stress situation: 4.41 for controls vs. 3.63 for patients; t30 = 2.39, p = 0.024). However, evaluation of the general stress experience measured immediately after the stressor using the VAS was similar in both groups (t30 = 1.33, p = 0.195). State anxiety was significantly higher in the BPD group before and after the stressor (group effect: F(1/30) = 19.43; p < 0.001).

[(Figure_1)TD$IG]

Figure 1 Mean salivary cortisol concentrations during the course of the study in 15 BPD patients and 17 healthy controls. The social stressor (TSST) is indicated by grey bars (the narrow grey bar indicates the introduction to the TSST). The data shown are the mean plus standard error of mean.

Acute stress response in BPD

[(Figure_3)TD$IG]

1569

3.3. Cortisol responses The stress test resulted in significant changes of cortisol levels in the total group (time effect: F(1.58/47.41) = 5.21; p = 0.014). Patients and controls differed significantly with regard to cortisol levels, with the patient group displaying lower cortisol levels (group effect: F(1/30) = 7.32; p = 0.011; see Fig. 1). Since patients and controls displayed different cortisol baseline levels (with higher levels in the control group; t25.48 = 2.08, p = 0.047), the first cortisol measurement was taken as a covariate, resulting in a significant time  group interaction effect in cortisol reaction to the stressor (F(2.12/ 61.47) = 3.30; p = 0.041). Cortisol increases elicited by the stressor (Dcortisol, computed by subtracting the last measurement time point from the peak value 10 min after the stressor) were significantly higher in the healthy controls than in the patients (11.23 nmol/l for controls vs. 4.44 nmol/l for patients; t25.11 = 2.75, p = 0.011), indicating an almost threefold increase in cortisol reactions in healthy controls compared to patients.

Figure 3 Mean salivary alpha-amylase concentrations during the course of the study in 15 BPD patients and 17 healthy controls. The social stressor (TSST) is indicated by grey bars (the narrow grey bar indicates the introduction to the TSST). The data shown are the mean plus standard error of mean.

3.4. ACTH responses The stress test resulted in significant ACTH changes in the total group (time effect: F(2.11/54.85) = 10.85; p < 0.001). Patients and controls did not differ in their ACTH levels (group effect: F(1/26) = 0.91; p = 0.349; no significant interaction effect; see Fig. 2). ACTH increases elicited by the stressor (DACTH, computed by subtracting the last measurement time point from the peak value 10 min after the stressor) were not significantly different between healthy controls and patients (24.58 pg/ml for controls vs. 13.34 pg/ ml for patients; t26 = 1.56, p = 0.131).

3.5. ACTH:cortisol ratio In addition, the ACTH:cortisol ratio was computed. The results show a significant different ACTH:cortisol ratio between BPD patients and healthy controls. BPD patients [(Figure_2)TD$IG]

revealed a significant increased ACTH:cortisol ratio (139.0 for controls vs. 168.2 for patients; t26 = 2.1, p = 0.044).

3.6. Alpha-amylase responses For sAA, the stress test resulted in significant changes in both groups (time effect: F(6.44/193.18) = 9.8; p < 0.001). Patients and controls did not differ with regard to sAA in their overall levels (group effect: F(1/30) = 1.78; p = 0.192), but there was a significant time  group interaction effect (F(6.44/193.18) = 2.58; p = 0.018) (see Fig. 3), with the patients displaying consistently lower sAA increases throughout the experiment.

3.7. Catecholamines responses For NE, the stress test resulted in significant changes in both groups (time effect: F(2.79/72.46) = 12.67; p < 0.001). Patients and controls did not differ with regard to NE levels (group effect: F(1/26) = 0.45; p = 0.509; no significant interaction effect; see Fig. 4A). For EP, the stress test also resulted in significant changes (time effect: F(2.58/64.57) = 6.12; p = 0.002). Patients and controls did not differ with regard to EP levels (group effect: F(1/25) = 0.91; p = 0.351; no significant interaction effect; see Fig. 4B). The influence of age, BMI, and smoking was examined for all analyses, revealing no main effects of these variables on endocrine outcomes.

4. Discussion

Figure 2 Mean ACTH concentrations during the course of the study in 11 BPD patients and 17 healthy controls. The social stressor (TSST) is indicated by grey bars (the narrow grey bar indicates the introduction to the TSST). The data shown are the mean plus standard error of mean.

We applied an established psychosocial stress paradigm (TSST) to a medication-free group of female individuals meeting BPD criteria. In the TSST, BPD patients reported significantly higher and more threatening acute stress compared to healthy controls. They also rated their stress coping resources as deficient. We found significantly reduced cortisol and alpha-amylase levels at baseline and in response to

1570 [(Figure_4)TD$IG]

U.M. Nater et al.

Figure 4 Mean norepinephrine (A) and epinephrine (B) concentrations during the course of the study in 11 BPD patients and 17 healthy controls. The social stressor (TSST) is indicated by grey bars (the narrow grey bar indicates the introduction to the TSST). The data shown are the mean plus standard error of mean.

stress in BPD patients compared to healthy controls. ACTH levels, however, did not differ between patients and healthy controls. Consistently, the data revealed a significantly elevated ACTH/cortisol ratio in BPD patients. The reported data are contradictory to the findings by Simeon et al. (2007) who also applied the TSST to a smaller group of BPD patients and did not find attenuated cortisol response. Beside the already discussed problems like mixed gender and not controlling for menstrual cycle, the patients in this study did not differ from the healthy control group regarding the level of self-reported distress. Our findings might be carefully interpreted as an indication of decreased adrenal responsiveness to endogenous ACTH in BPD patients. Adrenal glucocorticoid regulation is a complex and highly sensitive system and many potential reasons for disconcordance between measured plasma ACTH and cortisol levels can be discussed (for a review, see Bornstein et al., 2008). However mismatches between ACTH and corticosterone determinations are frequently observed in studies of HPA axis regulation following acute and chronic stress. In animal studies anxiety-producing stressors result in a diminished response (Bornstein et al., 2008). The majority of BPD patients report a long history of traumatic experience incl. physical abuse, sexual abuse and emotional neglect (Battle et al., 2004) and many adult BPD patients report persistent severe problems like ongoing abuse during adulthood (Zanarini et al., 2000; Yen et al., 2002; Zanarini et al., 2005). In our study, the BPD participants retrospectively reported significantly more stress during the previous weeks than healthy controls. It could be assumed (McEwen, 1998; Heim et al., 2000; Ehlert et al., 2001) that prolonged or excessive activation of the HPA axis would lead to the development of hyporesponsiveness and that, therefore, chronic psychological stress in BPD patients might lead to attenuated cortisol release due to adrenal hypofunction or to alterations in suprapituitary pathways. Regarding the underlying mechanisms of adrenal hypofunction, current research indicates that a large number of neuropeptides, neurotransmitters, opioids, growth factors, gonadal steroids, and cytokines, are capable of modulating adrenal glucocorticoid release independently of pituitary

ACTH (Bornstein and Chrousos, 1999). Any of these components, including centrally controlled aminergic transmitters could be involved in HPA dysregulation of BPD patients. And finally, genetic influence of ACTH receptor promoter polymorphism on adrenal secretion has also been shown recently, affecting ACTH receptor gene transcription and secretion, leading to attenuated adrenocortical response (Reisch et al., 2005). Thus, the potential role of long-term gene-environmental interaction should be considered as well. Stress not only activates the HPA axis, but also the ANS, involving brain noradrenergic mechanisms which control autonomic output (Pacak and Palkovits, 2001). Concomitant assessment of parameters of both systems has rarely been undertaken in a single study in BPD research. We did not find significant differences in peripheral norepinephrine and epinephrine levels between the two groups. These results are in line with the study by Simeon et al. (2007), who did not find differences in norepinephrine stress reactivity between BPD patients and healthy controls, either. However, the validity that peripheral catecholamines reflect central noradrenergic processes is low. In contrast, recent research indicates that alpha-amylase (sAA) might reflect central noradrenergic activity. Specifically, sAA secretion is determined by sympathetic nervous system activation, with beta-adrenergic mechanisms as the main contributing factor in sAA secretion, which has been shown both in animal and human studies (Ehlert et al., 2006; Nater and Rohleder, 2009; Rohleder and Nater, 2009). We found that the sympathetic stress response, as indicated by alpha-amylase, was substantially attenuated in BPD patients. We intend to investigate this intriguing finding in future studies in order to determine the relationship between altered sympathetic activity and BPD symptoms. Although it is a major strength of the current study that a well-defined patient sample was included and great effort was taken to exclude participants with conditions that might have confounded our results, some limitations should be noted. First, it should be acknowledged that the study sample was relatively small and our results should be replicated in a larger sample. Second, it would have been useful to conduct pharmacological challenge tests in our study.

Acute stress response in BPD Future studies should incorporate both basal and dynamic measurements in order to detect underlying abnormalities in HPA and ANS regulation in BPD patients. Third, both major depressive disorder (MDD) and posttraumatic stress disorder (PTSD) are known to affect the HPA axis (Steckler et al., 1999; Yehuda, 2002). While MDD was an exclusion condition in our study, 5 BPD patients suffered from comorbid PTSD. The presence of traumatic experiences might explain findings in pharmacological challenge tests (Lange et al., 2005) or baseline results (Flory et al., 2009) in BPD or personality disorders in general (Grossman et al., 2003). In our study however, additional analyses did not show an impact of a PTSD diagnosis on our results (data not shown), which might be due to the small sample size. Taken together, we were able to show attenuated cortisol responsiveness and a reduced ACTH/cortisol ratio in unmedicated female patients suffering from BPD compared to healthy control subjects. Our results indicate that this observation might in part be explained by altered central noradrenergic activation. Highly accelerated and prolonged stress response is a key feature of BPD patients (EbnerPriemer et al., 2007). Since glucocorticoid release plays an important role in downregulation of temporarily raised excitability (De Kloet et al., 2005, 2008), one could argue that an interplay between dysfunctional cognitive appraisal and attenuated cortisol response leads to the characteristic attenuated recovery from stressful experience in BPD patients. Certainly, more experimental work is needed to test this proposed working model. In addition, it would be most relevant to investigate stress-specific appraisal and psychobiological stress markers in patients’ everyday lives and relate these data to laboratory results.

Role of the funding source This work was supported by a Young Investigator Grant of the University of Zurich (UMN). The funding source had no role in the design of the study, data collection and analysis, or drafting the manuscript.

Conflicts of interest The authors have no conflicts of interest and declare no financial interests.

Acknowledgements We gratefully acknowledge the help of Sabrina Aisslinger, Barbara Gla ¨ser, Mirjam Thoma, Katrin Ku ¨nkele, and Petra Luda ¨scher in conducting the experiments.

References Battle, C.L., Shea, M.T., Johnson, D.M., Yen, S., Zlotnick, C., Zanarini, M.C., Sanislow, C.A., Skodol, A.E., Gunderson, J.G., Grilo, C.M., McGlashan, T.H., Morey, L.C., 2004. Childhood maltreatment associated with adult personality disorders: findings from the Collaborative Longitudinal Personality Disorders Study. J. Personal. Disord. 18, 193—211. Bornstein, S.R., Chrousos, G.P., 1999. Clinical review 104: adrenocorticotropin (ACTH)- and non-ACTH-mediated regulation of the

1571 adrenal cortex: neural and immune inputs. J. Clin. Endocrinol. Metab. 84, 1729—1736. Bornstein, S.R., Engeland, W.C., Ehrhart-Bornstein, M., Herman, J.P., 2008. Dissociation of ACTH and glucocorticoids. Trends Endocrinol. Metab. 19, 175—180. De Kloet, E.R., Joels, M., Holsboer, F., 2005. Stress and the brain: form adaptation to disease. Nat. Rev. Neurosci. 6, 463—475. De Kloet, E.R., Karst, H., Joels, M., 2008. Corticosteroid hormones in the central stress response: quick-and-slow. Front. Neuroendocrinol. 29, 268—272. De la Fuente, J.M., Mendlewicz, J., 1996. TRH stimulation and dexamethasone suppression in borderline personality disorder. Biol. Psychiatry 40, 412—418. Ebner-Priemer, U.W., Kuo, J., Schlotz, W., Kleindienst, N., Rosenthal, M.Z., Detterer, L., Linehan, M.M., Bohus, M., 2008. Distress and affective dysregulation in patients with borderline personality disorder: a psychophysiological ambulatory monitoring study. J. Nerv. Ment. Dis. 196, 314—320. Ebner-Priemer, U.W., Welch, S.S., Grossman, P., Reisch, T., Linehan, M.M., Bohus, M., 2007. Psychophysiological ambulatory assessment of affective dysregulation in borderline personality disorder. Psychiatry Res. 150, 265—275. Ehlert, U., Erni, K., Hebisch, G., Nater, U., 2006. Salivary alphaamylase levels after yohimbine challenge in healthy men. J. Clin. Endocrinol. Metab. 91, 5130—5133. Ehlert, U., Gaab, J., Heinrichs, M., 2001. Psychoneuroendocrinological contributions to the etiology of depression, posttraumatic stress disorder, and stress-related bodily disorders: the role of the hypothalamus—pituitary—adrenal axis. Biol. Psychol. 57, 141— 152. Ehrenreich, H., Schuck, J., Stender, N., Pilz, J., Gefeller, O., Schilling, L., Poser, W., Kaw, S., 1997. Endocrine and hemodynamic effects of stress versus systemic CRF in alcoholics during early and medium term abstinence. Alcohol Clin. Exp. Res. 21, 1285—1293. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B., 1997. Structured Clinical Interview for DSM-IV. American Psychiatric Press, Washington. Flory, J.D., Yehuda, R., Grossman, R., New, A.S., Mitropoulou, V., Siever, L.J., 2009. Childhood trauma and basal cortisol in people with personality disorders. Compr. Psychiatry 50, 34—37. Gaab, J., Rohleder, N., Nater, U.M., Ehlert, U., 2005. Psychological determinants of the cortisol stress response: the role of anticipatory cognitive appraisal. Psychoneuroendocrinology 30, 599— 610. Grassi, G., Esler, M., 1999. How to assess sympathetic activity in humans. J. Hypertens. 17, 719—734. Grossman, R., Yehuda, R., New, A., Schmeidler, J., Silverman, J., Mitropoulou, V., Sta Maria, N., Golier, J., Siever, L., 2003. Dexamethasone suppression test findings in subjects with personality disorders: associations with posttraumatic stress disorder and major depression. Am. J. Psychiatry 160, 1291—1298. Gurvits, I.G., Koenigsberg, H.W., Siever, L.J., 2000. Neurotransmitter dysfunction in patients with borderline personality disorder. Psychiatr. Clin. North Am. 23, 27—40 vi. Heim, C., Ehlert, U., Hellhammer, D.H., 2000. The potential role of hypocortisolism in the pathophysiology of stress-related bodily disorders. Psychoneuroendocrinology 25, 1—35. Kirschbaum, C., Kudielka, B.M., Gaab, J., Schommer, N.C., Hellhammer, D.H., 1999. Impact of gender, menstrual cycle phase, and oral contraceptives on the activity of the hypothalamus—pituitary—adrenal axis. Psychosom. Med. 61, 154—162. Kirschbaum, C., Pirke, K.M., Hellhammer, D.H., 1993. The ‘Trier Social Stress Test’–—a tool for investigating psychobiological stress responses in a laboratory setting. Neuropsychobiology 28, 76—81. 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 self-injury

1572 among women with borderline personality disorder. J. Nerv. Ment. Dis. 196, 230—236. Koenigsberg, H.W., Harvey, P.D., Mitropoulou, V., Schmeidler, J., New, A.S., Goodman, M., Silverman, J.M., Serby, M., Schopick, F., Siever, L.J., 2002. Characterizing affective instability in borderline personality disorder. Am. J. Psychiatry 159, 784—788. Korzekwa, M., Steiner, M., Links, P., Eppel, A., 1991. The dexamethasone suppression test in borderlines: is it useful? Can. J. Psychiatry 36, 26—28. Lahmeyer, H.W., Val, E., Gaviria, F.M., Prasad, R.B., Pandey, G.N., Rodgers, P., Weiler, M.A., Altman, E.G., 1988. EEG sleep, lithium transport, dexamethasone suppression, and monoamine oxidase activity in borderline personality disorder. Psychiatry Res. 25, 19—30. Lange, W., Wulff, H., Berea, C., Beblo, T., Saavedra, A.S., Mensebach, C., Wingenfeld, K., Driessen, M., 2005. Dexamethasone suppression test in borderline personality disorder–—effects of posttraumatic stress disorder. Psychoneuroendocrinology 30, 919—923. Laux, L., Glanzmann, P., Schaffner, P., Spielberger, C.D., 1981. Das State-Trait-Angstinventar. Theoretische Grundlagen und Handanweisungen, Weinheim, Beltz. Lieb, K., Rexhausen, J.E., Kahl, K.G., Schweiger, U., Philipsen, A., Hellhammer, D.H., Bohus, M., 2004a. Increased diurnal salivary cortisol in women with borderline personality disorder. J. Psychiatr Res. 38, 559—565. Lieb, K., Zanarini, M.C., Schmahl, C., Linehan, M.M., Bohus, M., 2004b. Borderline personality disorder. Lancet 364, 453—461. Loranger, A.W., 1999. International Personality Disorder Examination (IPDE): DSM-IV and ICD-10 modules. Psychological Assessment Resources, Odessa. Mauchnik, J., Schmahl, C., Bohus, M., 2005. New findings in the biology of borderline personality disorder. Dir. Psychiatry 25, 197—215. McEwen, B.S., 1998. Protective and damaging effects of stress mediators. N. Engl. J. Med. 338, 171—179. Nater, U.M., La Marca, R., Florin, L., Moses, A., Langhans, W., Koller, M.M., Ehlert, U., 2006. Stress-induced changes in human salivary alpha-amylase activity–—associations with adrenergic activity. Psychoneuroendocrinology 31, 49—58. Nater, U.M., Rohleder, N., 2009. Salivary alpha-amylase as a noninvasive biomarker for the sympathetic nervous system: current state of research. Psychoneuroendocrinology 34, 486—496. Nierop, A., Bratsikas, A., Klinkenberg, A., Nater, U.M., Zimmermann, R., Ehlert, U., 2006. Prolonged salivary cortisol recovery in

U.M. Nater et al. second-trimester pregnant women and attenuated salivary alpha-amylase responses to psychosocial stress in human pregnancy. J. Clin. Endocrinol. Metab. 91, 1329—1335. Pacak, K., Palkovits, M., 2001. Stressor specificity of central neuroendocrine responses: implications for stress-related disorders. Endocr. Rev. 22, 502—548. Reisch, N., Slawik, M., Zwermann, O., Beuschlein, F., Reincke, M., 2005. Genetic influence of an ACTH receptor promoter polymorphism on adrenal androgen secretion. Eur. J. Endocrinol. 153, 711—715. Rohleder, N., Nater, U.M., 2009. Determinants of salivary alphaamylase in humans and methodological considerations. Psychoneuroendocrinology 34, 469—485. Simeon, D., Knutelska, M., Smith, L., Baker, B.R., Hollander, E., 2007. A preliminary study of cortisol and norepinephrine reactivity to psychosocial stress in borderline personality disorder with high and low dissociation. Psychiatry Res. 149, 177—184. Steckler, T., Holsboer, F., Reul, J.M., 1999. Glucocorticoids and depression. Baillieres Best Pract. Res. Clin. Endocrinol. Metab. 13, 597—614. Stiglmayr, C.E., Grathwol, T., Linehan, M.M., Ihorst, G., Fahrenberg, J., Bohus, M., 2005. Aversive tension in patients with borderline personality disorder: a computer-based controlled field study. Acta Psychiatr. Scand. 111, 372—379. Walter, M., Bureau, J.F., Holmes, B.M., Bertha, E.A., Hollander, M., Wheelis, J., Brooks, N.H., Lyons-Ruth, K., 2008. Cortisol response to interpersonal stress in young adults with borderline personality disorder: a pilot study. Eur. Psychiatry 23, 201—204. Yehuda, R., 2002. Current status of cortisol findings in post-traumatic stress disorder. Psychiatr. Clin. North Am. 25, 341—368 vii. Yen, S., Shea, M.T., Battle, C.L., Johnson, D.M., Zlotnick, C., DolanSewell, R., Skodol, A.E., Grilo, C.M., Gunderson, J.G., Sanislow, C.A., Zanarini, M.C., Bender, D.S., Rettew, J.B., McGlashan, T.H., 2002. Traumatic exposure and posttraumatic stress disorder in borderline, schizotypal, avoidant, and obsessive-compulsive personality disorders: findings from the collaborative longitudinal personality disorders study. J. Nerv. Ment. Dis. 190, 510—518. Zanarini, M.C., Frankenburg, F.R., Reich, D.B., Hennen, J., Silk, K.R., 2005. Adult experiences of abuse reported by borderline patients and Axis II comparison subjects over six years of prospective follow-up. J. Nerv. Ment. Dis. 193, 412—416. Zanarini, M.C., Frankenburg, F.R., Reich, D.B., Marino, M.F., Lewis, R.E., Williams, A.A., Khera, G.S., 2000. Biparental failure in the childhood experiences of borderline patients. J. Personal. Disord. 14, 264—273.