Gender differences in hypothalamic–pituitary–adrenal (HPA) axis reactivity

Psychoneuroendocrinology (2006) 31, 642–652

www.elsevier.com/locate/psyneuen

Gender differences in hypothalamic–pituitary– adrenal (HPA) axis reactivity Magdalena Uharta, Rachel Y. Chonga, Lynn Oswaldb, Ping-I Linc, Gary S. Wanda,b,* a

Department of Medicine, The Johns Hopkins University School of Medicine, 720 Rutland Avenue, Room 863, Baltimore, MD 21205, USA b Department of Psychiatry, The Johns Hopkins University School of Medicine, 550 North Broadway, Suite 506, Baltimore, MD 21205, USA c Section of Medical Genetics, Department of Medicine, Center for Human Genetics, Duke University, 595 LaSalle Street, Durham, NC 27710, USA Received 5 May 2005; received in revised form 12 December 2005; accepted 8 February 2006

KEYWORDS Gender; Cortisol; Adrenocorticotropin (ACTH); Hypothalamic–pituitary–adrenal (HPA) axis; Trier social stress test; Naloxone challenge

Summary The present study was designed to determine whether there are gender differences in hormonal response patterns to HPA axis activation. To this end, two methods of activating the HPA axis were employed: a standardized psychological stress test and a pharmacological challenge. Healthy subjects (mean age 23.4 years, SD 7.0 years) completed a naloxone challenge and/or the modified Trier Social Stress Test (TSST). For the naloxone challenge, two baseline blood samples were obtained. Placebo was then administered (0 min), followed by increasing doses of intravenous naloxone (50, 100, 200 and 400 mg/kg) at 30-min intervals. Post-placebo blood samples were collected at 15-min intervals for 180 min. The TSST consisted of 5 min of public speaking followed by 5 min of mental arithmetic exercises. Three baseline and five post-TSST blood samples were drawn. Eighty subjects (53 male, 27 female) underwent the TSST. Following the psychological stressor, adrenocorticotropin (ACTH) and cortisol responses were significantly greater in male subjects compared to female subjects (zZK2.34, pZ 0.019 and zZK2.12, pZ0.034, respectively). Seventy-two subjects (52 male, 20 female) underwent HPA axis activation induced by naloxone. In contrast to the TSST, cortisol responses to the naloxone challenge were significantly greater in female subjects compared to male subjects (zZ4.11, p!0.001). Forty-one subjects (25 male, 16 female) completed both the TSST and naloxone challenge. In this subset, ACTH and cortisol responses to the TSST did not differ significantly by gender, although the effect size was moderate to large. Adrenocorticotropin and cortisol

* Corresponding author. Address: Division of Endocrinology, The Johns Hopkins University School of Medicine, Ross Research Building, Room 863, 720 Rutland Avenue, Baltimore, MD 21205, USA. Tel.: C1 410 955 7225; fax: C1 410 955 0841. E-mail address: [email protected] (G.S. Wand).

0306-4530/$ - see front matter Q 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.psyneuen.2006.02.003

Gender and HPA axis hormonal response

643

responses to the naloxone challenge were significantly greater in female subjects compared to male subjects (zZ2.29, pZ0.022 and zZ4.34, p!0.001, respectively). In summary, male subjects had greater HPA axis responses to a psychological stressor than female subjects, and females had greater hormonal reactivity than males to pharmacological stimulation with naloxone. Such differences are of interest as potential contributors to gender differences in health risks. Q 2006 Elsevier Ltd. All rights reserved.

1. Introduction Activation of the hypothalamic–pituitary–adrenal (HPA) axis is an essential adaptive mechanism that enables the human body to maintain physiological stability in response to stressful stimuli (Herman and Cullinan, 1997; Chrousos 1998; Tsigos and Chrousos, 2002). Following the perception of stress, corticotropin-releasing hormone (CRH) neurons in the hypothalamus receive regulatory impulses from several major neurotransmitter systems, including direct and indirect inhibitory signals from b-endorphin-producing neurons (Jackson et al., 1990; Calogero, 1995; Jessop, 1999). CRH release stimulates the synthesis and release of adrenocorticotropin (ACTH) by the anterior pituitary, which in turn stimulates the synthesis and release of cortisol by the adrenal cortex. There is evidence that healthy individuals react differently to stressful stimuli (Berger et al., 1987; Kirschbaum et al., 1995a), and that both enhanced and attenuated hormonal responses to stress are maladaptive. Chronic HPA axis dysregulation is associated with the development of mood and anxiety disorders, such as depression (Sapolsky, 2000; Gold and Chrousos, 2002; Sherwood Brown et al., 2004). Moreover, excess cortisol exposure is related to a variety of medical conditions including hypertension, atherosclerosis, obesity, insulin resistance, dyslipidemia, bone demineralization, and impaired immunity (McEwen, 1998; Tsigos and Chrousos, 2002). Likewise, it contributes to the development and maintenance of substance use disorders, such as alcohol dependence (Gianoulakis, 1998). Interestingly, pronounced differences in the prevalence of several of these disorders have been shown between men and women (Boyd and Weissman, 1981; Grant et al., 2004). Previous studies suggest that gender influences the HPA axis hormonal responses to stress. In preclinical models, ACTH and corticosterone levels in response to stress have been shown to be consistently greater in females compared with males (Kitay, 1961; Handa et al., 1994; Armario et al., 1995). However, in human studies, no such clear-cut gender differences have been established.

In response to psychological stressors in young subjects, certain studies have shown higher cortisol and ACTH responses in male subjects compared to females (Kirschbaum et al., 1992, 1995a,b). Nevertheless, other studies have suggested that there are no significant gender differences in young subjects in response to stress (Collins and Frankenhaeuser, 1978; Frankenhaeuser et al., 1978). In a study of healthy young adults, Kirschbaum et al. (1999) showed that ACTH responses were elevated in men compared to women and that free cortisol responses were similar between men and women in the luteal phase of their menstrual cycle whereas women in the follicular phase or taking oral contraceptives showed lower free cortisol responses compared to males. The same gender effect was demonstrated in elderly subjects, with higher ACTH and cortisol responses in male subjects compared to females (Kudielka et al., 1998; Traustadottir et al., 2003). Conversely, higher HPA axis hormonal responses to stress in elderly females compared to males have also been reported (Seeman et al., 1995, 2001). Thus, the impact of gender on the HPA axis hormonal response to stressful events remains inconclusive. Activation of the HPA axis can be evoked by numerous methods that act at different levels of the HPA system. Removing the endogenous inhibitory opioid tone on CRH neurons using naloxone, a non-selective opioid receptor antagonist, induces a rise in ACTH and cortisol (Volavka et al., 1979; Morley et al., 1980; Wand et al., 1998) and thus provides an assessment strategy for the functional evaluation of the hypothalamic opioid tone. Although several studies have evaluated effects of opioid blockade on the HPA axis (Cohen et al., 1983; Kreek, 1996; Wand et al., 1999, 2001), little research has addressed gender differences in the opioid–HPA axis interactions in healthy subjects. The aim of the present study was to determine whether there are gender differences in HPA axis hormonal response patterns to two different methods of HPA axis activation: a standardized psychological stress test and a pharmacological challenge with naloxone.

644

M. Uhart et al.

2. Methods 2.1. Subjects One hundred and eleven healthy subjects between the ages of 18 and 50 (mean age 23.4 years, SD 7.0 years) from the Baltimore area were recruited by newspaper advertisements and posted fliers. Persons who appeared to qualify for research participation based on a telephone screen were invited to the laboratory for an interview. After being given complete description of the study, volunteers provided written informed consent for the protocol approved by the Johns Hopkins Medicine Institutional Review Board. Subject assessment included a medical history and physical exam performed by a physician, a complete blood count, comprehensive metabolic panel (including renal and hepatic function tests), electrocardiogram, urinalysis, alcohol breathalyzer test and urine toxicology screen. A urine pregnancy test was obtained on female subjects. The semistructured assessment for the genetics of alcoholism (SSAGA) (Bucholz et al., 1994) was administered by a master’s degree-level interviewer to identify DSM-IV axis I psychiatric diagnoses. Exclusion criteria were as follows: (a) presence of a serious medical condition, (b) presence of a DSM-IV axis I disorder, including alcohol/drug abuse or dependence, (c) use of any psychoactive medications within the past 30 days, (d) treatment in the last 6 months with any medication that may affect opioid or HPA axis function, including antidepressants, neuroleptics, sedative hypnotics, glucocorticoids, appetite suppressants, estrogens (including anti-contraceptive pills), opiates, or dopamine medications, (e) presence of a seizure disorder or history of closed head trauma, (f) consumption of more than 30 alcoholic drinks per Table 1

month, (g) positive urine toxicology screen, (h) or, for females, pregnancy or lack of effective nonhormonal methods of birth control. Menstrual cycle phase was determined by measurements of estradiol and progesterone levels on the day of the challenge (Table 1). Subjects with progesterone levels greater than or equal to 3 ng/mL were categorized as being in the luteal phase of their menstrual cycle (Yen et al., 1999). Eighty subjects underwent the TSST and 72 subjects completed the naloxone challenge. Of these, 41 subjects completed both sessions. Thus, there were 111 subjects in all. In the TSST group, there were 53 male subjects and 27 female subjects. Based on progesterone levels, all female subjects completed the TSST during the follicular phase of their menstrual cycle. In the naloxone group, there were 52 male subjects and 20 female subjects. Based on progesterone values, all female subjects underwent the naloxone challenge during the follicular phase of their menstrual cycle.

2.2. General procedure Following the initial assessment interview, subjects reported to the Johns Hopkins Hospital Outpatient General Clinical Research Center (GCRC) to complete either the naloxone challenge and/or the Trier Social Stress Test (TSST). For the subset of subjects who completed both challenges, one challenge was completed on one day, and the other challenge on a separate day. The time between the administrations of both challenges was 7–14 days, and the order in which the challenges were completed was randomized. Subjects were instructed to get adequate sleep the night prior to the challenges and to report any stressful situations that may have occurred during the previous week. They were also instructed to

Demographic characteristics by gender. TSST, nZ80a

Sample size Age, mean (SD), (years) Race, no. (%) Caucasian African–American Asian BMI, mean (SD), (kg/m2) Estradiol, mean (SE), (pg/mL) Progesterone, mean (SE), (ng/mL)

Naloxone challenge, nZ72a

Male

Female

Male

Female

53 26.7 (10.1)

27 22.6 (5.3)

52 21.4 (2.5)

20 20.9 (2.6)

40 (75.5) 8 (15.1) 5 (9.4) 25.2 (3.5)

17 (63.0) 10 (37.0) 0 (0.0) 23.9 (3.9) 46.3 (8.0) 0.5 (0.1)

43 (82.7) 5 (9.6) 4 (7.7) 24.3 (3.0)

12 (60.0) 7 (35.0) 1 (5.0) 24.9 (3.5) 49.7 (14.1) 0.7 (0.1)

a Eighty subjects underwent the TSST and 72 subjects completed the naloxone challenge. Of these, 41 subjects completed both sessions. Thus, there were 111 subjects in all.

Gender and HPA axis hormonal response refrain from any alcohol, illicit drugs, or over-thecounter medications for 48 h prior to participating in the study protocol. Urine toxicology screens were completed before each session. On the day of each challenge, subjects fasted from 1000 h until testing was completed.

2.3. Trier Social Stress Test (TSST) The TSST consists of 5 min of public speaking followed by 5 min of a mental arithmetic task and was completed as previously described in detail (Kirschbaum et al., 1993; Uhart et al., 2004). Upon arrival at the GCRC, an intravenous catheter was inserted into a forearm vein at 1200 h. Baseline blood samples for cortisol and ACTH were obtained at 1300, 1315, and 1330 h. Subjects then participated in the TSST. Immediately following completion of the TSST and at 15-min intervals, five additional blood specimens were drawn for ACTH and cortisol.

2.4. Naloxone challenge For the naloxone challenge, subjects received five doses of naloxone according to the 1 day, singlesession protocol reported by Mangold et al. (2000). Upon arrival at the GCRC, subjects had an intravenous catheter inserted into a forearm vein at 1200 h. One hour later, a bolus of 0.9% saline was administered as a placebo; this was designated as time 0 min. Subsequently, incremental doses of naloxone dissolved in 0.9% saline were administered at 30, 60, 90, and 120 min. In all, a total of five increasing doses of naloxone were administered (0, 50, 100, 200 and 400 mg/ kg). Baseline blood samples for cortisol and ACTH were obtained 15 min before and immediately prior to placebo administration. Post-placebo blood samples for cortisol and ACTH were drawn at 15-min intervals for 180 min.

2.5. Hormone assays Hormones were assayed as previously described (Blevins et al., 1994). Plasma concentrations of ACTH were assayed by a two-site IRMA (Nichols immunoradiometric assay). Plasma concentrations of cortisol were measured by radioimmunoassay (Diagnostic Products Corporation, Inc., Los Angeles, CA, USA). Plasma concentrations of estradiol and progesterone were measured in female subjects (Diagnostic Products Corporation, Inc., Los Angeles, CA, USA). Intra-assay and inter-

645 assay coefficients of variance for all assays was less than 10%.

2.6. Statistical analysis Preliminary analyses included evaluation of gender group differences in demographic characteristics with t-tests for continuous variables (age and body mass index) and c 2 analyses for categorical variables (race). The major outcomes of interest were hormonal measurements of ACTH and cortisol. All hormonal measurements were transformed to the logarithmic scale due to non-normality. The mean baseline for each variable was calculated by taking the average of the baseline measurements (K30, K15, and 0 min time points for the pre-TSST measurements and K15 and 0 min time points for the pre-naloxone challenge measurements). The mean baseline differences in hormonal levels between gender groups were also analyzed with t-tests. We then carried out four different sets of analyses to assess the effect of gender on hormone responses in each of the two challenge groups, and in the subset of subjects who completed both challenges. For each analysis, age and race were incorporated into the model for the TSST data whereas age, race, baseline hormones and BMI were incorporated into the model for the naloxone data set. Baseline hormone values were added because baseline hormone values differed by gender for the naloxone challenge group only; BMI was added because a pharmacological agent was administered. First, we performed longitudinal data analysis. Each hormone measurement at each time point was treated as the outcome in the generalized linear model using generalized estimating equations (GEE) to take into account the within-individual correlation residuals arising from repeated measurements for each individual (Zeger and Liang, 1986). The model included a contrast for gender difference as a major covariate of interest, time, and time squared to adjust for non-linear time trend. Second, we conducted post-hoc regression analyses to evaluate gender effect on hormone differences at each time point. Third, we compared gender differences in peak hormone responses to the TSST and naloxone challenge. Peak hormone level was defined as the highest level reached during each of the challenges. We treated the peak hormone response as the outcome variable and gender as the major covariate of interest in the generalized linear model. Fourth, we carried out area under the

646

M. Uhart et al. generalized linear model. Lastly, we calculated the effect size of the hormonal responses in each of the two challenge groups, and in the subset of subjects who completed both challenges. Effect size was calculated as the difference between the mean peak hormone response values of the male group and female group divided by the standard deviation of the overall sample (Cohen, 1988). The analyses were two sided with a 0.05 significance level and were performed by using the software STATA 8.0. Finally, we plotted the unadjusted means of ACTH and cortisol response concentrations to the two challenges against time by gender.

3. Results 3.1. Demographics

Figure 1 (a) Plasma ACTH response to TSST by gender. Values reflect unadjusted means (SE). *Plasma ACTH levels differed significantly by gender at the following time points: 25 min, pZ0.007; 40 min, pZ0.021. (b) Plasma cortisol response to TSST by gender. Values reflect unadjusted means (SE). *Plasma cortisol levels differed significantly by gender at the following time points: 25 min, pZ0.027; 85 min, pZ0.006.

curve (AUC) analysis. ACTH and cortisol AUC values by gender were computed by using the trapezoid algorithm and the effect of gender on differences in the AUC was assessed by using the

Table 2

Demographic characteristics for the subjects undergoing the TSST (nZ80) and the naloxone challenge (nZ72) are in Table 1. Subjects were healthy and predominantly Caucasian, and all subjects were non-smokers. The groups differed in racial composition (TSST, c2Z6.76, pZ0.034; naloxone, c2Z6.70, pZ0.035), and race was adjusted for in our statistical models. There were no statistically significant differences between gender groups in terms of age and BMI. In the TSST group, males were 18–50 years (mean 26.7 years, SD 10.1 years) and females were 18–42 years (mean 22.6 years, SD 5.3 years). In the naloxone group, males were 18–32 years (mean 21.4 years, SD 2.5 years) and females were 18–27 years (mean 20.9 years, SD 2.6 years).

3.2. Trier social stress test (TSST) There were no mean baseline differences in plasma ACTH and cortisol levels by gender (tZ1.49, pZ

Adjusted mean hormonal values for the TSST and naloxone challenge by gender.

Hormone/measure

Sample size ACTH AUC, mean (SE) Peak, mean (SE) Cortisol AUC, mean (SE) Peak, mean (SE)

TSST, nZ80

Naloxone challenge, nZ72

Male

Female

53

27

232.2 (7.5) 29.1 (1.1)

209.2 (8.7) 19. 5 (1.1)

201.2 (5.4) 17.9 (1.0)

185.7 (7.9) 13.7 (1.1)

p Value

Male

Female

52

20

zZK1.93, pZ0.054 zZK2.42, pZ0.016

517.1(8.5) 32.1(1.0)

546.4 (15.0) 44.7(1.1)

zZ1.63, pZ0.103 zZ2.68, pZ0.007

zZK1.59, pZ0.111 zZK2.15, pZ0.031

458.8 (5.5) 20.4(1.0)

500.5 (8.7) 27.9(1.0)

zZ3.99, p!0.001 zZ5.06, p!0.001

ACTH, adrenocorticotropic hormone; AUC, area under the curve.

p Value

Gender and HPA axis hormonal response 0.139 and tZ1.70, pZ0.092, respectively). Following the psychological stressor, ACTH and cortisol responses to the TSST were significantly greater in male subjects compared to female subjects (zZK2.34, pZ0.019 and zZK2.12, pZ 0.034, respectively) (Fig. 1a and b). Male subjects also had greater peak ACTH and cortisol responses compared to female subjects (zZK2.42, pZ0.016 and zZK2.15, pZ0.031, respectively). In addition, ACTH area under the curve responses were marginally greater in males as compared to females (zZK1.93, pZ0.054) (Table 2).

647

3.3. Naloxone challenge There were no mean baseline differences in plasma cortisol levels by gender (tZ1.09, pZ0.281). However, ACTH baseline levels differed by gender (tZ2.57, pZ0.012) and baseline hormonal levels were used as a covariate in the analysis. Cortisol responses to the naloxone challenge were significantly greater in female subjects compared to male subjects (zZ4.11, p!0.001). In addition, ACTH responses were marginally greater in female subjects compared to male subjects (zZ1.70, pZ0.089) (Fig. 2a and b). Female subjects also had statistically significantly greater peak ACTH and cortisol responses compared to male subjects (zZ 2.68, pZ0.007 and zZ5.06, p!0.001, respectively). In addition, female subjects had greater cortisol area under the curve response compared to males (zZ3.99, p!0.001) (Table 2).

3.4. Trier Social Stress Test (TSST) and naloxone challenge Forty-one subjects completed both the TSST and the naloxone challenge. Demographic characteristics are presented in Table 3. There were no statistically significant differences between gender groups in terms of age, race and BMI. Adrenocorticotropin and cortisol responses to the TSST did not differ significantly by gender (Fig. 3a and b, and Table 4). Given the small sample size and lack of significance, effect sizes were calculated showing Cohen’s dZ0.58 and 0.77 for ACTH and cortisol, respectively. However, ACTH and cortisol responses to the naloxone challenge were significantly greater in

Table 3 Demographic characteristics by gender in subjects who completed both the TSST and naloxone challenge. TSST and naloxone challenge, nZ41

Figure 2 (a) Plasma ACTH response to naloxone by gender. Values reflect unadjusted means (SE). Pl denotes time of placebo (saline) administration. N denotes times of incremental naloxone administration. (b) Plasma cortisol response to naloxone by gender. Values reflect unadjusted means (SE). Pl denotes time of placebo (saline) administration. N denotes times of incremental naloxone administration. *Plasma cortisol levels differed significantly by gender at the following time points: 60 min, pZ0.050; 75 min, pZ0.009; 90 min, p!0.001; 105 min, p!0.001; 120 min, p!0.001; 135 min, p! 0.001; 150 min, p!0.001; 165 min, p!0.001; 180 min, p!0.001.

Sample size Age, mean (SD), (y) Race, no. (%) Caucasian African–American Asian BMI, mean (SD), (kg/m2) Estradiol, mean (SE), (pg/mL) Progesterone, mean (SE), (ng/mL)

Male

Female

25 20.8 (2.8)

16 20.5 (2.4)

20 (80) 2 (8) 3 (12) 24.8 (3.3)

11 (68.7) 5 (31.3) 0 (0) 25.1 (3.6) TSST 48.8 (11.1)

Naloxone 46.1 (16.6)

0.5 (0.1)

0.7 (0.1)

648

M. Uhart et al. cortisol responses compared to male subjects (zZ 3.24, p!0.001 and zZ5.0, p!0.001, respectively). In addition, female subjects had greater ACTH and cortisol area under the curve response compared to males (zZ2.30, pZ0.022 and zZ4.25, p!0.001) (Table 4).

4. Discussion

Figure 3 (a) Plasma ACTH response to TSST by gender in subjects who completed both the TSST and naloxone challenge. Values reflect unadjusted means (SE). (b) Plasma cortisol response to TSST by gender in subjects who completed both the TSST and naloxone challenge. Values reflect unadjusted means (SE).

female subjects compared to male subjects (zZ 2.29, pZ0.022 and zZ4.34, p!0.001, respectively) (Fig. 4a and b). Female subjects also had statistically significantly greater peak ACTH and

In the present study, we observed that healthy male subjects demonstrate a more robust ACTH and cortisol response to a psychological stress compared to females. In contrast, healthy females had a higher cortisol and marginally higher ACTH response to naloxone compared to males. Overall, our findings were consistent whether our analysis was conducted using a between-subject or withinsubject design. Our findings in response to the TSST lend further support to a number of studies that have shown greater cortisol and ACTH responses in young and elderly male subjects to a psychological challenge, including public speaking and mental arithmetic exercises, compared to females (Kirschbaum et al., 1992, 1995a,b; Kudielka et al., 1998). The evidence in human studies of the effect of gender on HPA axis stress response, however, has been conflicting. Certain studies have suggested higher HPA axis responses in elderly females compared to elderly males employing a driving simulation and a cognitive paradigm challenge (Seeman et al., 1995, 2001), and following lumbar puncture (Petrie et al., 1999). Still, other studies have reported no significant gender difference in cortisol response to examination stress and a cognitive-conflict task in young subjects (Collins and Frankenhaeuser, 1978; Frankenhaeuser et al., 1978). Interestingly, in contrast to the findings after exposure to psychological stress, we observed

Table 4 Adjusted mean hormonal values for the TSST and naloxone challenge by gender in subjects who completed both the TSST and naloxone challenge. Hormone/ measure Sample size ACTH AUC, mean (SE) Peak, mean (SE) Cortisol AUC, mean (SE) Peak, mean (SE)

TSST, nZ41

Naloxone challenge, nZ41

Male

Female

25

16

221.9 (9.7) 27.9 (0.1)

205.0 (11.3) 19.8 (1.1)

197.3 (7.7) 17.6 (1.0)

181.6 (11.5) 13.4 (1.1)

p Value

Male

Female

25

16

zZK1.07, pZ0.284 zZK1.53, pZ0.125

512.3 (14.2) 31.1 (1.0)

563.9 (17.3) 47.9 (3.0)

zZ2.30, pZ0.022 zZ3.24, p!0.001

zZK1.08, pZ0.281 zZK1.54, pZ0.124

466.5 (9.3) 22.2 (1.0)

522.2 (9.1) 30.8 (1.0)

zZ4.25, p!0.001 zZ5.0, p!0.001

ACTH, adrenocorticotropic hormone; AUC, area under the curve.

p Value

Gender and HPA axis hormonal response

Figure 4 (a) Plasma ACTH response to naloxone by gender in subjects who completed both the TSST and naloxone challenge. Values reflect unadjusted means (SE). Pl denotes time of placebo (saline) administration. N denotes times of incremental naloxone administration. *Plasma ACTH levels differed significantly by gender at the following time points: 45 min, pZ0.013; 75 min, p! 0.001; 90 min, pZ0.008; 150 min, pZ0.021; 180 min, pZ 0.040. (b) Plasma cortisol response to naloxone by gender in subjects who completed both the TSST and naloxone challenge. Values reflect unadjusted means (SE). Pl denotes time of placebo (saline) administration. N denotes times of incremental naloxone administration. *Plasma cortisol levels differed significantly by gender at the following time points: 45 min, pZ0.015; 60 min, pZ 0.007; 75 min, pZ0.003; 90 min, p!0.001; 105 min, p! 0.001; 120 min, p!0.001; 135 min, pZ0.001; 150 min, pZ0.001; 165 min, pZ0.001; 180 min, p!0.001.

higher cortisol response in female subjects to naloxone compared to males. These results are in agreement with HPA axis hormonal response to other pharmacological challenges, including administration of CRH alone or combined with dexamethasone or arginine-vasopressin, which have shown elevated HPA axis responsivity and decreased feedback sensitivity in female subjects compared to males (Gallucci et al., 1993; Heuser

649 et al., 1994; Born et al., 1995). However, following a CRH challenge, other studies have reported no significant gender difference in hormonal response (Hermus et al., 1984). The possible role of sex steroids such as estrogen on the gender differences in response to the TSST observed in our study is suggested by several lines of evidence. In animal studies, estrogens enhance HPA axis activity (Viau and Meaney, 1991; Burgess and Handa, 1992) and, in response to stress females have consistently shown greater increases in ACTH and corticosterone compared with males (Kitay, 1961; Handa et al., 1994; Armario et al., 1995). In humans, sex steroids seem to modulate the HPA axis stress response as suggested by the observation that cortisol responses to stress were similar in males compared to females in the luteal phase of the menstrual cycle whereas in the follicular phase females had blunted response compared to males (Kirschbaum et al., 1999). However, the direct impact of estrogens on HPA axis regulation in humans remains contradictory. Whereas in one study a short-term estradiol application enhanced cortisol responsivity to stress in males (Kirschbaum et al., 1996), other studies show that the effect of estrogen on stress reactivity is, if anything reversed. For example, a 2-week estradiol treatment in postmenopausal women did not modify stress-induced HPA axis responses and feedback sensitivity seemed to be increased resulting in a blunted cortisol response to dexamethasone-CRH test (Kudielka et al., 1999). It is plausible that sex steroids may also influence opioid regulation of HPA axis activation and contribute to explain the gender differences in response to naloxone observed in our study. The hypothalamic CRH neurons receive direct inhibitory input from b-endorphin-producing neurons located in the arcuate nucleus via the m-opioid receptor (Tsagarakis et al., 1990). In addition, b-endorphin-producing neurons inhibit norepinephrine neurons, which provide direct stimulatory input to hypothalamic CRH neurons (Jackson et al., 1990). There is prior evidence that expression of m-opioid receptors is modulated by gonadal steroid hormones in rat brain (Hammer et al., 1994), as shown by increased receptor binding to naloxone in hypothalamic regions following exposure to estradiol (Brown et al., 1996). Furthermore, the regional brain content of opioid peptides is modulated by gonadal steroid hormones; for example, in rodents estrogen induces the expression and seems to stimulate the release of endogenous opioid peptides that activate opioid receptors in the limbic system and hypothalamus (Hammer et al., 1994; Priest et al., 1995; Eckersell et al., 1998). It is plausible that if opioid

650 activity differs among genders, then HPA axis activity would also differ as a function of gender following naloxone administration. The observation that acute psychological stressors on one hand and pharmacological stimulation tests on the other hand seem to result in different gender-specific patterns of HPA axis responsivity is intriguing. Reported gender differences could possibly be attributed to differences in the applied HPA axis stimulation procedures. In this regard Stroud et al. (2002) observed that men and women show different adrenocortical response to different stressors; in their study, men showed greater cortisol responses to achievement challenges, but women showed greater cortisol responses to a social rejection challenge. Thus, it appears that different mechanisms of HPA axis activation stimulate the HPA axis in a singular way and are under different neurochemical regulation. However, in our study, the HPA axis stimulation procedures also differed in their duration. For example, the psychological stressor was administered over a 10min period and, differently, the naloxone challenge was administered over a 2-h period. Thus, it is possible that the gender differences observed in response to naloxone administration could be related to other mechanisms such as variation in feedback regulation. In this regard, there is evidence in preclinical models that estradiol modulates mineralocorticoid (MR) and glucocorticoid (GR) receptors and thus modifies cortisol negative feedback within the HPA axis (Peiffer et al., 1991; Burgess and Handa, 1992; Handa et al., 1994). Specifically, in estrogen treated rats, an increase in the magnitude and the duration of the corticosterone response to stress has been shown, suggesting an impairment of the GR-mediated negative feedback (Burgess and Handa, 1992). Thus, the possibility that decreased cortisol negative feedback in females as compared to males could contribute to the increased hormonal response observed following naloxone administration cannot be ruled out. This observation may reflect that, in addition to the nature of HPA axis activation, the duration of the activation should also be considered in future studies. There is interest in investigating gender differences in HPA axis activation in response to challenges as it may contribute to explain gender differences in the prevalence of diseases associated with HPA axis dysregulation. For example, gender differences in HPA axis responses to stress may be one mechanism underlying gender differences in prevalence of depression (Boyd and Weissman, 1981; Stroud et al., 2002). Similarly, gender differences in endogenous opioid activity could

M. Uhart et al. play a role in the observed gender differences in the development and maintenance of substance use disorders such as alcohol dependence (Grant et al., 2004; Oswald and Wand, 2004). The present study has several weaknesses. Although the sample size for the naloxone challenge and TSST was ample, it would have been ideal if the sample size was larger for the subgroup of subjects that completed both the naloxone and TSST. The moderate to large effect size suggests that this would have resulted in finding significant gender differences to the TSST. Second, it also needs to be acknowledged that there are potential confounding differences for participants in the between-subject versus the within-subject portions of the study. For the within-subject design, participants underwent two challenges which activated the HPA axis. It is plausible that a certain degree of desensitization occurred between the first and second challenge as subjects become accustomed to the stress of a novel environment and intravenous line placement. However, the wash out period between studies should have minimized any significant desensitization. Third, although race could have been a confounder as previously described (Yanovski et al., 2000), adjusting for race and other demographic characteristics in our statistical models likely mitigated bias induced by baseline group differences. In addition, one should be cautious about generalizing the pattern of responses seen in artificial settings to real-life situations. Furthermore, whether gender would predict hormonal responses to other types of HPA axis activators needs to be investigated. Further research is needed to replicate the findings reported here and to extend our understanding of the underlying mechanisms for any gender differences in HPA axis hormonal response to challenge. In summary, male subjects demonstrate a more robust HPA axis hormonal response to a psychological stressor compared to females and, in contrast, females had greater hormonal reactivity to pharmacological stimulation with naloxone compared to males. Such gender differences are of interest as potential contributors to gender differences in health risks.

Acknowledgements This work was supported by NIH grants AA 10158 (GSW), AA 12303 (GSW) and AA 12837 (MEM), and a gift from the Kenneth A. Lattman Foundation (GSW).

Gender and HPA axis hormonal response

References Armario, A., Gavalda, A., Marti, J., 1995. Comparison of the behavioural and endocrine response to forced swimming stress in five inbred strains of rats. Psychoneuroendocrinology 20, 879–890. Berger, M., Bossert, S., Krieg, J.C., Dirlich, G., Ettmeier, W., Schreiber, W., von Zerssen, D., 1987. Interindividual differences in the susceptibility of the cortisol system: an important factor for the degree of hypercortisolism in stress situations? Biol Psychiatry 22, 1327–1339. Blevins Jr., L.S., Dobs, A.S., Wand, G.S., 1994. Naloxone-induced activation of the hypothalamic–pituitary–adrenal axis in suspected central adrenal insufficiency. Am. J. Med. Sci. 308, 167–170. Born, J., Ditschuneit, I., Schreiber, M., Dodt, C., Fehm, H.L., 1995. Effects of age and gender on pituitary–adrenocortical responsiveness in humans. Eur. J. Endocrinol. 132, 705–711. Boyd, J.H., Weissman, M.M., 1981. Epidemiology of affective disorders. A reexamination and future directions. Arch. Gen. Psychiatry 38, 1039–1046. Brown, L.L., Pasi, S., Etgen, A.M., 1996. Estrogen regulation of mu opioid receptor density in hypothalamic premammillary nuclei. Brain Res. 742, 347–351. Bucholz, K.K., Cadoret, R., Cloninger, C.R., Dinwiddie, S.H., Hesselbrock, V.M., Nurnberger Jr., J.I., Reich, T., Schmidt, I., Schuckit, M.A., 1994. A new, semi-structured psychiatric interview for use in genetic linkage studies: a report on the reliability of the SSAGA. J. Stud. Alcohol 55, 149–158. Burgess, L.H., Handa, R.J., 1992. Chronic estrogen-induced alterations in adrenocorticotropin and corticosterone secretion, and glucocorticoid receptor-mediated functions in female rats. Endocrinology 131, 1261–1269. Calogero, A.E., 1995. Neurotransmitter regulation of the hypothalamic corticotropin-releasing hormone neuron. Ann. NY Acad. Sci. 771, 31–40. Chrousos, G.P., 1997. Stressors, stress, and neuroendocrine integration of the adaptive response. The 1997 hans selye memorial lecture. Ann. NY Acad. Sci. 851, 311–335. Cohen, J. 1988. Statistical power analysis for the behavioral sciences, 2nd ed. Hillsdale, NJ. Cohen, M.R., Pickar, D., Dubois, M., 1983. The role of the endogenous opioid system in the human stress response. Psychiatr. Clin. North Am. 6, 457–471. Collins, A., Frankenhaeuser, M., 1978. Stress responses in male and female engineering students. J. Human Stress 4, 43–48. Eckersell, C.B., Popper, P., Micevych, P.E., 1998. Estrogeninduced alteration of m-opioid receptor immunoreactivity in the medial preoptic nucleus and medial amygdala. J. Neurosci. 18, 3967–3976. Frankenhaeuser, M., von Wright, M.R., Collins, A., von Wright, J., Sedvall, G., Swahn, C.G., 1978. Sex differences in psychoneuroendocrine reactions to examination stress. Psychosom. Med. 40, 334–343. Gallucci, W.T., Baum, A., Laue, L., Rabin, D.S., Chrousos, G.P., Gold, P.W., Kling, M.A., 1993. Sex differences in sensitivity of the hypothalamic–pituitary–adrenal axis. Health Psychol. 12, 420–425. Gianoulakis, C., 1998. Alcohol-seeking behavior: the roles of the hypothalamic–pituitary–adrenal axis and the endogenous opioid system. Alcohol Health Res. World 22, 202–210. Gold, P.W., Chrousos, G.P., 2002. Organization of the stress system and its dysregulation in melancholic and atypical depression: high vs low CRH/NE states. Mol. Psychiatry 7, 254–275.

651 Grant, B.F., Dawson, D.A., Stinson, F.S., Chou, S.P., Dufour, M.C., Pickering, R.P., 2002. The 12-month prevalence and trends in DSM-IV alcohol abuse and dependence: United States, 1991–1992 and 2001–2002. Drug. Alcohol Depend. 74, 223–234. Hammer Jr., R.P., Zhou, L., Cheung, S., 1994. Gonadal steroid hormones and hypothalamic opioid circuitry. Horm. Behav. 28, 431–437. Handa, R.J., Burgess, L.H., Kerr, J.E., O’Keefe, J.A., 1994. Gonadal steroid hormone receptors and sex differences in the hypothalamo–pituitary–adrenal axis. Horm. Behav. 28, 464–476. Herman, J.P., Cullinan, W.E., 1997. Neurocircuitry of stress: central control of the hypothalamo–pituitary–adrenocortical axis. Trends Neurosci. 20, 78–84. Hermus, A.R., Pieters, G.F., Smals, A.G., Benraad, T.J., Kloppenborg, P.W., 1984. Plasma adrenocorticotropin, cortisol, and aldosterone responses to corticotropin-releasing factor: modulatory effect of basal cortisol levels. J. Clin. Endocrinol. Metab. 58, 187–191. Heuser, I.J., Gotthardt, U., Schweiger, U., Schmider, J., Lammers, C.H., Dettling, M., Holsboer, F., 1994. Ageassociated changes of pituitary–adrenocortical hormone regulation in humans: importance of gender. Neurobiol. Aging 15, 227–231. Jackson, R.V., Grice, J.E., Jackson, A.J., Hockings, G.I., 1990. Naloxone-induced ACTH release in man is inhibited by clonidine. Clin. Exp. Pharmacol. Physiol. 17, 179–184. Jessop, D.S., 1999. Review: central non-glucocorticoid inhibitors of the hypothalamo–pituitary–adrenal axis. J. Endocrinol. 160, 169–180. Kirschbaum, C., Wust, S., Hellhammer, D., 1992. Consistent sex differences in cortisol responses to psychological stress. Psychosom. Med. 54, 648–657. 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. Kirschbaum, C., Klauer, T., Filipp, S.H., Hellhammer, D.H., 1995a. Sex-specific effects of social support on cortisol and subjective responses to acute psychological stress. Psychosom. Med. 57, 23–31. Kirschbaum, C., Pirke, K.M., Hellhammer, D.H., 1995b. Preliminary evidence for reduced cortisol responsivity to psychological stress in women using oral contraceptive medication. Psychoneuroendocrinology 20, 509–514. Kirschbaum, C., Schommer, N., Federenko, I., Gaab, J., Neumann, O., Oellers, M., Rohleder, N., Untiedt, A., Hanker, J., Pirke, K.M., Hellhammer, D.H., 1996. Shortterm estradiol treatment enhances pituitary–adrenal axis and sympathetic responses to psychosocial stress in healthy young men. J. Clin. Endocrinol. Metab. 81, 3639–3643. 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. Kitay, J.I., 1961. Sex differences in adrenal cortical secretion in the rat. Endocrinology 68, 818–824. Kreek, M.J., 1996. Opioid receptors: some perspectives from early studies of their role in normal physiology, stress responsivity, and in specific addictive diseases. Neurochem. Res. 21, 1469–1488. Kudielka, B.M., Hellhammer, J., Hellhammer, D.H., Wolf, O.T., Pirke, K.M., Varadi, E., Pilz, J., Kirschbaum, C., 1998. Sex differences in endocrine and psychological responses to

652 psychosocial stress in healthy elderly subjects and the impact of a 2-week dehydroepiandrosterone treatment. J. Clin. Endocrinol. Metab. 83, 1756–1761. Kudielka, B.M., Schmidt-Reinwald, A.K., Hellhammer, D.H., Kirschbaum, C., 1999. Psychological and endocrine responses to psychosocial stress and dexamethasone/corticotropinreleasing hormone in healthy postmenopausal women and young controls: the impact of age and a two-week estradiol treatment. Neuroendocrinology 70, 422–430. Mangold, D., McCaul, M.E., Ali, M., Wand, G.S., 2000. Plasma adrenocorticotropin responses to opioid blockade with naloxone: generating a dose–response curve in a single session. Biol. Psychiatry 48, 310–314. McEwen, B.S., 1998. Protective and damaging effects of stress mediators. N. Engl. J. Med. 338, 171–179. Morley, J.E., Baranetsky, N.G., Wingert, T.D., Carlson, H.E., Hershman, J.M., Melmed, S., Levin, S.R., Jamison, K.R., Weitzman, R., Chang, R.J., Varner, A.A., 1980. Endocrine effects of naloxone-induced opiate receptor blockade. J. Clin. Endocrinol. Metab. 50, 251–257. Oswald, L.M., Wand, G.S., 2004. Opioids and alcoholism. Physiol. Behav. 81, 339–358. Peiffer, A., Lapointe, B., Barden, N., 1991. Hormonal regulation of type II glucocorticoid receptor messenger ribonucleic acid in rat brain. Endocrinology 129, 2166–2174. Petrie, E.C., Wilkinson, C.W., Murray, S., Jensen, C., Peskind, E.R., Raskind, M.A., 1999. Effects of Alzheimer’s disease and gender on the hypothalamic–pituitary–adrenal axis response to lumbar puncture stress. Psychoneuroendocrinology 24, 385–395. Priest, C.A., Eckersell, C.B., Micevych, P.E., 1995. Estrogen regulates preproenkephalin-A mRNA levels in the rat ventromedial nucleus: temporal and cellular aspects. Brain Res. Mol. Brain Res. 28, 251–262. Sapolsky, R.M., 2000. Glucocorticoids and hippocampal atrophy in neuropsychiatric disorders. Arch. Gen. Psychiatry 57, 925–935. Seeman, T.E., Singer, B., Charpentier, P., 1995. Gender differences in patterns of HPA axis response to challenge: macarthur studies of successful aging. Psychoneuroendocrinology 20, 711–725. Seeman, T.E., Singer, B., Wilkinson, C.W., McEwen, B., 2001. Gender differences in age-related changes in HPA axis reactivity. Psychoneuroendocrinology 26, 225–240. Sherwood Brown, E., Varghese, F.P., McEwen, B.S., 2004. Association of depression with medical illness: does cortisol play a role? Biol. Psychiatry 55, 1–9.

M. Uhart et al. Stroud, L.R., Salovey, P., Epel, E.S., 2002. Sex differences in stress responses: social rejection versus achievement stress. Biol. Psychiatry 52, 318–327. Traustadottir, T., Bosch, P.R., Matt, K.S., 2003. Gender differences in cardiovascular and hypothalamic–pituitary– adrenal axis responses to psychological stress in healthy older adult men and women. Stress 6, 133–140. Tsagarakis, S., Rees, L.H., Besser, M., Grossman, A., 1990. Opiate receptor subtype regulation of CRF-41 release from rat hypothalamus in vitro. Neuroendocrinology 51, 599–605. Tsigos, C., Chrousos, G.P., 2002. Hypothalamic–pituitary–adrenal axis, neuroendocrine factors and stress. J. Psychosom. Res. 53, 865–871. Uhart, M., McCaul, M.E., Oswald, L.M., Choi, L., Wand, G.S., 2004. GABRA6 gene polymorphism and an attenuated stress response. Mol. Psychiatry 9, 998–1006. Viau, V., Meaney, M.J., 1991. Variations in the hypothalamic– pituitary–adrenal response to stress during the estrous cycle in the rat. Endocrinology 129, 2503–2511. Volavka, J., Cho, D., Mallya, A., Bauman, J., 1979. Naloxone increases ACTH and cortisol levels in man. N. Engl. J. Med. 300, 1056–1057. Wand, G.S., Mangold, D., El Deiry, S., McCaul, M.E., Hoover, D., 1998. Family history of alcoholism and hypothalamic opioidergic activity. Arch. Gen. Psychiatry 55, 1114–1119. Wand, G.S., Mangold, D., Ali, M., 1999. Adrenocorticotropin responses to naloxone in sons of alcohol-dependent men. J. Clin. Endocrinol. Metab. 84, 64–68. Wand, G.S., McCaul, M.E., Gotjen, D., Reynolds, J., Lee, S., 2001. Confirmation that offspring from families with alcohol-dependent individuals have greater hypothalamic– pituitary–adrenal axis activation induced by naloxone compared with offspring without a family history of alcohol dependence. Alcohol Clin. Exp. Res. 25, 1134–1139. Yanovski, J.A., Yanovski, S.Z., Boyle, A.J., Gold, P.W., Sovik, K.N., Sebring, N.G., Drinkard, B., 2000. Hypothalamic–pituitary–adrenal axis activity during exercise in African American and Caucasian women. J. Clin. Endocrinol. Metab. 85, 2660–2663. Yen, S., Jaffe, R., Barbieri, R., 1999. Reproductive endocrinology: physiology, pathophysiology, and clinical management. Saunders, Philadelphia, PA. Zeger, S.L., Liang, K.Y., 1986. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 42, 121– 130.