PPLR

European Psychiatry 22 (2007) 309e312 http://france.elsevier.com/direct/EURPSY/

Short communication

Overnight urinary cortisol release in women with borderline personality disorder depends on comorbid PTSD and depressive psychopathology

Katja Wingenfeld a,*, Martin Driessen

a,b,d , Bettina Adam

b,c , Andreas Hill

b,e

a Department of Psychiatry and Psychotherapy Bethel, Ev. Hospital Bielefeld, Remterweg 69-71, D-33617 Bielefeld, Germany b University of Schleswig-Holstein, Campus Luebeck, Germany c August-Bier-Hospital, Bad Malente-Gremsmühlen, Germany

d University of Bielefeld, Bielefeld, Germany e Institute of Sexology and Forensic Psychiatry, Center for Psychosocial Medicine, University Hospital Hamburg-Eppendorf, Hamburg, Germany

Received 13 July 2006; received in revised form 5 September 2006; accepted 5 September 2006

Available online 1 December 2006

Abstract

Free cortisol was investigated in BPD patients and healthy controls. A positive association was found between cortisol and depression scores, while the number of PTSD symptoms was negatively correlated with cortisol release. These findings suggest that alterations in cortisol release in BPD are strongly associated with the severity of psychopathology. � 2006 Elsevier Masson SAS. All rights reserved.

Keywords: Borderline personality disorder; PTSD; Cortisol; HPA axis; Childhood trauma

1. Introduction

Alterations of the hypothalamusepituitaryeadrenal (HPA) axis have been found in different psychiatric disorders and there is evidence for an association with early adverse life ex- periences [1,26]. Early life stress results in an increased risk for the development of mood and anxiety disorders [9e11,17]. Posttraumatic stress disorder (PTSD), and major depression disorder (MDD) occur frequently in patients with borderline personality disorder (BPD) [7,18,28]. While there is evidence for an enhanced central release of corticotropin releasing factor in patients suffering from PTSD or MDD [2,20], basal cortisol levels as well as feedback regulation differ between these disorders [19,26].

Only a few studies of BPD have evaluated HPA axis func- tion yielding varying results [3,12,13,15,23]. These diverging results may be due to different prevalence rates of comorbid psychiatric disorders like PTSD and/or MDD in the BPD

* Corresponding author. Tel.: þ49 521 772 78522; fax: þ49 521 772 78511. E-mail address: [email protected] (K. Wingenfeld).

0924-9338/$ - see front matter � 2006 Elsevier Masson SAS. All rights reserve doi:10.1016/j.eurpsy.2006.09.002

samples [14,21,26]. Surprisingly, only a few studies have in- vestigated basal cortisol release in BPD. One study reported increased free cortisol on awakening and higher total daily cortisol [15]. Another investigated 24 h urine cortisol release in PTSD patients with and without comorbid BPD [24]. The authors found higher cortisol in patients with just PTSD in comparison to the BPD group, which may partially explain di- vergent findings in PTSD literature.

The aim of this study was to further evaluate unstimulated free cortisol excretion in patients with BPD and its association with other clinical features.

2. Method

2.1. Subjects

We recruited 21 female patients with BPD and 24 healthy female controls. There were no differences in age (mean [SD]: patients 28.1 [5.4], controls 27.7 [6.9]) or body mass in- dex (patients 24.4 [5.8], controls 24.1 [5.1]). No participants were pregnant or suffered from significant physical illness.

d.

310 K. Wingenfeld et al. / European Psychiatry 22 (2007) 309e312

Additional exclusion criteria were anorexia, schizophrenia, schizoaffective disorder, major depressive disorder with psy- chotic symptoms, alcohol and/or drug dependence. All sub- jects were drug free for at least 7 days before endocrine assessments. All participants gave their written informed consent. The study was approved by the institutional review board (Ethics Commission of the Medical University of Luebeck).

2.2. Clinical assessment

All participants were evaluated for BPD using the Struc- tured Clinical Interview for DSM-IV Personality Disorders (SCID-II) and for PTSD using the SCID-I [25]. Severity of PTSD was estimated by the number of symptoms reported. Depressive mood state was measured using Beck Depression Inventory (BDI) [4]. Trauma history was assessed using Child- hood Trauma Questionnaire (CTQ) [5].

2.3. Cortisol assessment

Urine was collected over three consecutive nights (7 p.m. to 7 a.m.). Urinary free cortisol concentrations (mg/12 h) were as- sayed by radio-immunoassay (RIA, Immunotech). Inter-assay coefficient of variation was 5.4e9.2% and intra-assay coeffi- cient of variation was 3.5e5.1%.

2.4. Statistical analyses

Demographic and clinical data were analyzed using t-test or one-way analysis of variances. Mean cortisol was computed from measurements on three consecutive nights because of potential intra-individual variation of hormonal release and analysis of covariances was also applied. Partial correlations were performed to test associations between clinical and endo- crine data.

3. Results

Patients reported a significantly more severe childhood trauma history (CTQ) and reached higher mean depression (BDI) compared to controls (Table 1). Sixteen (76.2%) pa- tients suffered concurrently from PTSD.

In the patient group partial correlations between cortisol re- lease and clinical variables were performed. When controlling for depression (BDI score), the number of PTSD symptoms was significantly negatively associated with mean 12 h cortisol (r ¼�0.637, p ¼ 0.003). There were significant partial correla- tions between cortisol and the symptom clusters re-experience (r ¼ 0.688, p ¼ 0.001) and avoidance (r ¼�0.520, p ¼ 0.019), but only a trend for an association with arousal (r ¼�0.429, p ¼ 0.059). BDI scores showed a significant positive correlation with cortisol (r ¼ 0.352, p ¼ 0.019), when controlling for the number of PTSD symptoms. Number of PTSD symptoms did not correlate with BDI scores (r ¼ 0.272, p ¼ 0.233).

When statistically controlling for CTQ and BDI scores (one-way ANCOVA) patients with BPD showed higher mean cortisol release in comparison with controls (Table 1). CTQ scores were used as covariate, as PTSD symptoms were only assessed in patients because the controls had suf- fered no trauma experiences.

To evaluate the impact of PTSD symptoms on endocrine alterations the BPD group was divided by median-split (me- dian ¼ 9) into subgroups with low and high numbers of PTSD symptoms. These subgroups significantly differed with respect to the number of reported PTSD symptoms (tdf:19 ¼�5.303, p ¼ 0.001) but not with respect to depres- sive symptoms. When compared with the controls, only the subgroup with a low number of PTSD symptoms showed a higher level of cortisol release (Fdf:2,42 ¼ 4.62, p ¼ 0.021, Fig. 1).

Depressive symptoms were controlled because of their im- pact on HPA axis function. The patients were divided into subjects with high and low BDI scores by median-split (me- dian ¼ 25). No difference was found between these subgroups with respect to their cortisol levels (tdf:19 ¼�0.18, p ¼ 0.863).

Table 1

Sample characteristics and mean 12 h urinary cortisol in BPD patients and controls a

BPD Group (N ¼ 21) Control group (N ¼ 24) Statistics BDI 29.7 (12.5) 3.0 (3.6) tdf:43 ¼ 9.98, p < 0.001 Number of PTSD symptoms 9.14 (3.5)

(min: 0, max: 15)

e e

CTQ total b

11.9 (3.2) 5.6 (0.7) tdf:43 ¼ 9.48, p ¼ 0.001 Physical abuse 3.3 (0.8) 1.4 (0.3) tdf:43 ¼ 9.42, p ¼ 0.001 Emotional neglect 3.5 (0.6) 2.0 (0.4) tdf:43 ¼ 9.01, p ¼ 0.001 Physical neglect 2.3 (0.7) 1.1 (0.2) tdf:43 ¼ 7.69, p ¼ 0.001 Sexual abuse 2.9 (1.4) 1.0 (0.1) tdf:43 ¼ 6.60, p ¼ 0.001

Mean urinary cortisol (mg/12 h)

Observed means 14.39 (14.1) 10.29 (4.3)

Estimated means corrected

for CTQ and BDI

19.79 (3.8) 5.44 (3.4) Fdf:1,412 ¼ 4.55, p ¼ 0.039

a Data are given as mean (SD). b

Weighted scores are given.

311K. Wingenfeld et al. / European Psychiatry 22 (2007) 309e312

4. Discussion

This first investigation of overnight urinary free cortisol in patients with BPD compared to controls showed higher corti- sol levels in patients with BPD. There was a negative associ- ation between cortisol and PTSD symptoms. When dividing the BDP group in patients with a high or low number of PTSD symptoms, exaggerated cortisol levels were only found in BPD patients with a low number of PTSD symptoms.

Consistent with these results, Lieb and colleagues reported enhanced free cortisol in saliva collected over the day in BPD [15]. They also found reduced feedback sensitivity in response to 0.5 mg dexamethasone. Comorbid PTSD was not assessed. Non-suppression in the DST was previously only found in BPD patients without comorbid PTSD, while BPD patients with PTSD were comparable to controls [14].

For PTSD there are several studies on 24 h urinary cortisol yielding mixed results (see Refs. [26,27]), while only few studies evaluated overnight cortisol release [6,16,22]. In one study 12 h urinary cortisol was measured in mothers of child cancer survivors with and without PTSD [8]. Exaggerated cor- tisol was only found in the non-PTSD group compared to con- trols. Another study found no differences in over night urinary cortisol in PTSD patients and controls [22]. Both results agree with those of this study. Higher 24 h urinary cortisol levels have been found in patients with PTSD compared to patients with PTSD and comorbid BPD [24]. It was suggested that these differences might reflect differences in the severity of PTSD symptoms rather than factors related to BPD per se [24]. Unfortunately, this study did not use controls, making it difficult to compare with the present data. Furthermore, 24 h and 12 h urinary cortisol measurements are not compara- ble, due to the circadian rhythm of cortisol release.

In sum, the data presented here further support the hypoth- esis of the relevance of comorbid PTSD symptoms in BPD for HPA axis alterations.

m ea

n ur

in ar

y co

rti so

l ( µg

/1 2h

)

0

5

10

15

20

25

30

Low number of PTSD

Symptoms

High number of PTSD

Symptoms

Controls

Fig. 1. Mean (SE) over night urinary cortisol release in BPD patients with high

(N ¼ 9) and low (N ¼ 12) number of PTSD symptoms compared to healthy controls (N ¼ 24) (Fdf:2,42 ¼ 4.62, p ¼ 0.021).

Despite the significant positive partial correlation between depression score and cortisol release there were no group dif- ferences between patients with high and low BDI scores. The PTSD effect may be the stronger one, which is indicated by the high negative correlation between re-experience symptoms and cortisol. Another point is the lack of an association be- tween PTSD symptoms and BDI scores, possibly caused by an equal distribution of patients with low or high PTSD symp- toms in the two ‘‘depression subgroups’’. Noteworthy, Rinne et al. showed comparable cortisol levels after DEX in BPD patients with PTSD and patients with PTSD þ MDD, while the BPD þ MDD group differed significantly from the BPD þ PTSD group [21].

4.1. Limitations

Depressive symptoms were measured with a self-rating scale and MDD was not evaluated by a structured interview. Furthermore, no standardized PTSD rating scale was used. Another limitation was the small sample size, especially for the subgroup analyses. Furthermore, only women were stud- ied, making comparison with other studies difficult [24]. Un- fortunately, there were no controls for menstrual cycle phase or oral contraceptive intake although they are known to influ- ence HPA axis outcomes. Time of awakening could not be evaluated and therefore possible interfering effects could not be excluded. A more general problem is that measurement of childhood trauma with the CTQ is only retrospective. How- ever, most studies in this field use such retrospective questionnaires.

In sum, there is evidence for HPA axis hyperactivity in BPD, e.g. enhanced cortisol excretion and reduced feedback sensitiv- ity [14,15]. These alterations seem to be mediated by trauma- related symptoms and depressive psychopathology. Further studies are needed, evaluating comorbidity-related subgroups.

Acknowledgements

This study was funded by the Deutsche Forschungsgemein- schaft (Dr 358/2-1).

References

[1] Antonijevic IA. Depressive disorders e is it time to endorse different

pathophysiologies? Psychoneuroendocrinology 2006;31:1e15.

[2] Baker DG, Ekhator NN, Kasckow JW, Dashevsky B, Horn PS,

Bednarik L, et al. Higher levels of basal serial CSF cortisol in combat

veterans with post-traumatic stress disorder. Am J Psychiatry

2005;162:992e4.

[3] Baxter L, Edell W, Gerner R, Fairbanks L, Gwirtsman H. Dexametha-

sone suppression test and Axis I diagnoses of inpatients with DSM-III

borderline personality disorder. J Clin Psychiatry 1984;45:150e3.

[4] Beck AT, Steer RA. Beck-Depressions-Inventar: BDI. In: Testhandbuch.

Bern: Huber; 1994.

[5] Bernstein D, Fink L. CTQ childhood trauma questionnaire: a retrospec-

tive self-report. San Antonio: The Psychological Corporation; 1998.

[6] Delahanty DL, Nugent NR, Christopher NC, Walsh M. Initial urinary

epinephrine and cortisol levels predict acute PTSD symptoms in child

trauma victims. Psychoneuroendocrinology 2005;30:121e8.

312 K. Wingenfeld et al. / European Psychiatry 22 (2007) 309e312

[7] Driessen M, Beblo T, Reddemann L, Rau H, Lange W, Silva A, et al. Ist

die Borderline-Persönlichkeitsstörung eine komplexe posttraumatischen

Belastungsstörung? Nervenarzt 2002;73:820e9.

[8] Glover DA, Poland RE. Urinary cortisol and catecholamines in mothers

of child cancer survivors with and without PTSD. Psychoneuroendocri-

nology 2002;27:805e19.

[9] Heim C, Nemeroff CB. Neurobiology of early life stress: clinical studies.

Semin Clin Neuropsychiatry 2002;7:147e59. [10] Heim C, Newport DJ, Wagner D, Wilcox MM, Miller AH, Nemeroff CB.

The role of early adverse experience and adulthood stress in the predic-

tion of neuroendocrine stress reactivity in women: a multiple regression

analysis. Depress Anxiety 2002;15:117e25. [11] Johnson JG, Cohen P, Brown J, Smailes EM, Bernstein DP. Childhood

maltreatment increases risk for personality disorders during early adult-

hood. Arch Gen Psychiatry 1999;56:600e6.

[12] Krishnan KR, Davidson JR, Rayasam K, Shope F. The dexamethasone

suppression test in borderline personality disorder. Biol Psychiatry

1984;19:1149e53.

[13] Lahmeyer HW, Val E, Gaviria FM, Prasad RB, Pandey GN, Rodgers P,

et al. EEG sleep, lithium transport, dexamethasone suppression, and

monoamine oxidase activity in borderline personality disorder. Psychia-

try Res 1988;25:19e30.

[14] Lange W, Wulff H, Berea C, Beblo T, Saavedra AS, Mensebach C, et al.

Dexamethasone suppression test in borderline personality disorder-

effects of posttraumatic stress disorder. Psychoneuroendocrinology

2005;30:919e23.

[15] Lieb K, Rexhausen JE, Kahl KG, Schweiger U, Philipsen A,

Hellhammer DH, et al. Increased diurnal salivary cortisol in women

with borderline personality disorder. J Psychiatr Res 2004;38:559e65.

[16] Mason JW, Wang S, Yehuda R, Lubin H, Johnson D, Bremner JD, et al.

Marked lability in urinary cortisol levels in subgroups of combat veterans

with posttraumatic stress disorder during an intensive exposure treatment

program. Psychosom Med 2002;64:238e46.

[17] Newport DJ, Heim C, Bonsall R, Miller AH, Nemeroff CB. Pituitarye adrenal responses to standard and low-dose dexamethasone suppression

tests in adult survivors of child abuse. Biol Psychiatry 2004;55:10e20.

[18] Ogata SN, Silk KR, Goodrich S, Lohr NE, Westen D, Hill EM. Child-

hood sexual and physical abuse in adult patients with borderline person-

ality disorder. Am J Psychiatry 1990;147:1008e13.

[19] Pariante CM, Miller AH. Glucocorticoid receptors in major depression:

relevance to pathophysiology and treatment. Biol Psychiatry

2001;49:391e404.

[20] Raadsheer FC, van Heerikhuize JJ, Lucassen PJ, Hoogendijk WJ,

Tilders FJ, Swaab DF. Corticotropin-releasing hormone mRNA levels

in the paraventricular nucleus of patients with Alzheimer’s disease and

depression. Am J Psychiatry 1995;152:1372e6.

[21] Rinne T, de Kloet ER, Wouters L, Goekoop JG, DeRijk RH, van den

Brink W. Hyperresponsiveness of hypothalamicepituitaryeadrenal axis to combined dexamethasone/corticotropin-releasing hormone challenge

in female borderline personality disorder subjects with a history of sus-

tained childhood abuse. Biol Psychiatry 2002;52:1102e12.

[22] Rohleder N, Joksimovic L, Wolf JM, Kirschbaum C. Hypocortisolism

and increased glucocorticoid sensitivity of pro-inflammatory cytokine

production in Bosnian war refugees with posttraumatic stress disorder.

Biol Psychiatry 2004;55:745e51. [23] Soloff PH, George A, Nathan RS. The dexamethasone suppression test in

patients with borderline personality disorders. Am J Psychiatry

1982;139:1621e3.

[24] Southwick SM, Axelrod SR, Wang S, Yehuda R, Morgan 3rd CA,

Charney D, et al. Twenty-four-hour urine cortisol in combat veterans

with PTSD and comorbid borderline personality disorder. J Nerv Ment

Dis 2003;191:261e2.

[25] Wittchen HU, Zaudig M, Fydrich T. Strukturiertes Klinisches Interview

für DSM-IV. Achse I: Psychische Störungen. SKID-I. Hogrefe; 1997.

[26] Yehuda R. Current status of cortisol findings in post-traumatic stress dis-

order. Psychiatr Clin North Am 2002;25:341e68. vii. [27] Young EA, Breslau N. Cortisol and catecholamines in posttraumatic

stress disorder: an epidemiologic community study. Arch Gen Psychiatry

2004;61:394e401.

[28] Zanarini MC, Frankenburg FR, Dubo ED, Sickel AE, Trikha A,

Levin A, et al. Axis I comorbidity of borderline personality disorder.

Am J Psychiatry 1998;155:1733e9.