Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Aug 1.
Published in final edited form as: Int J Eat Disord. 2022 Jul 2;55(8):1100–1107. doi: 10.1002/eat.23767

Dehydroepiandrosterone sulfate levels predict weight gain in women with anorexia nervosa

Allison Kimball 1,2, Caitlin Colling 1,2, Melanie S Haines 1,2, Erinne Meenaghan 1, Kamryn T Eddy 2,3, Madhusmita Misra 1,2, Karen K Miller 1,2
PMCID: PMC9357210  NIHMSID: NIHMS1817890  PMID: 35779065

Abstract

Objective:

Anorexia nervosa (AN) is a serious condition characterized by undernutrition, complicated by endocrine dysregulation, and with few predictors of recovery. Urinary free cortisol (UFC) is a predictor of weight gain, but 24-hour urine samples are challenging to collect. We hypothesized that serum dehydroepiandrosterone sulfate (DHEAS), which like cortisol is regulated by adrenocorticotropic hormone (ACTH), would predict weight gain and increases in fat mass in women with AN.

Methods:

We prospectively studied 34 women with AN and atypical AN, mean age 27.4±7.7 years (mean±SD), who received placebo in a 6-month randomized trial. Baseline DHEAS and 24-hour UFC were measured by liquid chromatography with tandem mass spectrometry. Body composition was assessed at baseline and 6 months by DXA and cross-sectional abdominal CT at L4.

Results:

Mean baseline DHEAS level was 173±70 μg/dL (0.7±0.3 times the mean normal range for age) and mean baseline UFC (n=15) was 20±18 μg/24h (normal: 0–50 μg/24h). Higher DHEAS levels predicted weight gain over 6 months (r=0.61, p<0.001). DHEAS levels also predicted increases in fat mass (r=0.40, p=0.03), appendicular lean mass (r=0.38, p=0.04), and abdominal adipose tissue (r=0.60, p<0.001). All associations remained significant after controlling for age, baseline BMI, OCP use, duration of AN, and SSRI/SNRI use. DHEAS levels correlated with UFC (r=0.61, p=0.02).

Discussion:

In women with AN, higher serum DHEAS predicts weight gain and increases in fat and muscle mass. Further studies are needed to confirm these findings and further elucidate the association between DHEAS and weight gain.

Keywords: DHEAS, cortisol, anorexia nervosa, eating disorder

Introduction

Anorexia nervosa (AN) is an eating disorder defined by restriction of food intake resulting in low body weight, a fear of gaining weight, and distorted body image. Atypical AN shares the diagnostic criteria of AN except that body mass index (BMI) is >18.5 kg/m2. AN is complicated by endocrine system dysregulation, including hypothalamic amenorrhea, growth hormone resistance, and activation of the hypothalamic-pituitary-adrenal (HPA) axis (Miller, 2011; Misra & Klibanski, 2014; Schorr & Miller, 2017). The endocrine complications are related to the degree of undernutrition and are inversely correlated with BMI (Schorr & Miller, 2017). Studies have demonstrated elevated 24-hour urine free cortisol (UFC) levels in about a third of patients with low-weight AN (Biller et al., 1989; Putignano et al., 2001). Elevations in 24-hour mean serum cortisol levels (Boyar et al., 1977; Doerr, Fichter, Pirke, & Lund, 1980), in overnight mean serum cortisol levels by frequent sampling (Lawson, Donoho, et al., 2009; Misra et al., 2004), in the cortisol response to adrenocorticotropic hormone (ACTH) (Lawson, Misra, et al., 2009), and in midnight salivary cortisol levels (Miller, 2011; Putignano et al., 2001) have also been demonstrated in patients with AN. Normalization of cortisol levels is observed in many patients with AN who have recovered (Doerr et al., 1980; Lawson et al., 2013).

AN is associated with high morbidity and mortality. However, treatment with behavioral, psychiatric and medical therapies results in recovery in only about 50% of adults (Lowe et al., 2001). There are no validated predictors of weight recovery, and many patients’ courses are marked by periods of relapse and remission. Cushing’s disease, a condition of hypercortisolism, is associated with increased food cravings and weight gain (Geer et al., 2016). Prior studies have shown that higher levels of serum cortisol area under the curve (AUC), measured with frequent sampling, predict recovery of menses and greater increases in fat mass and BMI in girls with AN (Misra et al., 2006). Serum cortisol levels measured with frequent sampling and 24-hour UFC have been shown to correlate with each other (r=0.52) (Misra et al., 2006).

Like cortisol, the adrenal androgen dehydroepiandrosterone (DHEA) and its sulfated form DHEAS, which has a longer half-life, are secreted by the adrenal glands and are under the control of the corticotropin-releasing hormone-ACTH axis. In vivo, intra-adrenal cortisol has been shown to stimulate DHEA secretion by inhibiting 3β-hydroxysteroid dehydrogenase-2 in adrenal cells (Topor, Asai, Dunn, & Majzoub, 2011). Cortisol measures can be challenging to interpret and use clinically because of diurnal variation, which limits the interpretability of a single serum measurement. 24-hour urine collections for UFC provide a more integrated assessment of cortisol levels but are onerous to collect. In contrast to cortisol, DHEAS levels are stable throughout the day and a single serum measure can be clinically meaningful. However, studies examining DHEAS levels in women with AN report variable results, demonstrating either low levels compared to a laboratory reference range (Devesa et al., 1988; Gordon et al., 1999), or normal (Lawson, Misra, et al., 2009; Miller et al., 2007) or elevated levels compared to controls (Oskis, Loveday, Hucklebridge, Thorn, & Clow, 2012). It is unknown whether the variability in DHEAS levels represent the spectrum of HPA axis activation in AN and whether DHEAS levels are markers of recovery. We hypothesized that DHEAS levels would correlate with UFC and be a predictor of weight gain in women with AN.

Methods

Study participants

The study group included 34 community-dwelling women with AN, aged 18–45 years, who received placebo during a 6-month trial; the parent trial was a randomized placebo-controlled study of low-dose testosterone therapy in women with AN (Kimball et al., 2019). In this parent study, 35 participants completed six months of placebo treatment; one participant was excluded from this post-hoc analysis due to a baseline DHEAS level that was an outlier (defined as more than three times the interquartile range above the third quartile). Participants fulfilled criteria for AN (American Psychiatric Association, 2013) or atypical AN (defined as weight loss and all psychological features of AN but body mass index (BMI) >18.5 kg/m2) by the Structured Clinical Interview for DSM (SCID) (First, Karg, Spitzer, & Williams, 2015). All participants had outpatient treatment teams. Exclusion criteria for the trial included unstable medical illness or psychiatric illness, including severe current depressive symptoms (defined as a Hamilton Depression Rating Scale (Hamilton, 1960) score >20, excluding two eating/weight loss items related to the symptoms of AN) or serious suicide risk; bipolar I disorder; and psychotic disorder. An inclusion criterion for the parent trial was a free testosterone level less than the median of the reference range for premenopausal women; no participants were excluded based on this criterion. No participants had received androgens or androgen precursors within 3 months of study enrollment. Pre-treatment clinical characteristics, change in DHEAS levels, and change in weight were previously published (Kimball et al., 2019), but UFC, body composition, and the relationship between these variables have not been reported.

Protocol

The study was approved by the Mass General Brigham Institutional Review Board, and all participants gave informed written consent before study participation. Study participants were recruited through collaborating physicians and through advertisements, and eligibility was determined at a screening visit. Nutritional evaluation, including weight in a gown and height, was performed, and BMI was calculated. Psychiatric disorders, including major depressive disorder (MDD) and generalized anxiety disorder (GAD), were diagnosed by SCID (Spitzer, First, Gibbon, & Williams, 2002). At baseline, blood was drawn for DHEAS, and fifteen participants completed 24-hour urine collections for measurement of UFC. Body composition was assessed by dual energy x-ray absorptiometry (DXA) and cross-sectional abdominal computed tomography (CT) at baseline and following 6 months of testosterone or placebo treatment; this substudy examined predictors of weight gain in those participants who received placebo only. Depression symptom severity, anxiety symptom severity, and eating disorder psychopathology were assessed at baseline and 6 months by the Hamilton Depression Rating Scale (HAM-D) (Hamilton, 1960), Hamilton Anxiety Rating Scale (HAM-A) (Hamilton, 1959), Eating Disorder Examination-Questionnaire (EDE-Q) (Fairburn & Beglin, 2008), and Eating Disorder Inventory-2 (EDI-2) (Fairburn & Cooper, 1991).

Laboratory methods

DHEAS and 24-hour UFC were measured by LC-MS/MS (Endocrine Sciences, Calabasas Hills, CA). As the normal range for DHEAS varied by age, a DHEAS index was also calculated for each participant, defined as the DHEAS level divided by the mean of the normal range for age. The normal range for UFC was 0–50 μg/24h.

Imaging methods

Participants underwent DXA (Discovery A; Hologic Inc., Bedford, MA, USA) for measurement of fat mass and appendicular lean mass (precision of 0.01 g/cm2 at the spine and 3% for fat mass). Participants underwent single slice CT (LightSpeed Pro, GE Healthcare, Waukesha, WI, USA) of the abdomen through the mid-portion of the L4 level for measurement of total abdominal adipose tissue (TAT), visceral adipose tissue (VAT), and subcutaneous adipose tissue (SAT). Scan parameters were standardized: 144 mm table height, 80 kV and 70 mA for the abdomen, scanning time of 2 seconds, 1 cm section thickness, and 48 cm field of view. Abdominal adipose tissue was identified using a threshold set for −50 to −250 Hounsfield units (HU) (Borkan et al., 1982). Manual delineation was used to separate VAT, and cross-sectional area (mm2) of abdominal VAT and SAT were reported. Analyses were performed using Osirix software version 3.2.1 (www.osirix-viewer.com/index.html). At our institution, the coefficient of variation for repeated measurements of the same scan on consecutive days by the same analyst is 1.7% for SAT and 2.3% for VAT.

Statistical analysis

JMP Statistical Database Software (version Pro 14.0; SAS Institute, Cary, NC) was used for statistical analyses. Variables were assessed for normality using the Shapiro-Wilk test, and if non-normal, were log-transformed. ANOVA was used to compare baseline characteristics between participants who gained weight and those who did not. Linear regression analyses were performed to investigate associations between baseline hormone levels and changes in weight and body composition, and Pearson’s correlation coefficients are reported. With n=34, a correlation coefficient ≥0.46 can be detected with 80% power at a two-sided 0.05 significance level. Multivariate standard least squares were constructed to control for age, baseline BMI, duration of AN, oral contraceptive use, and selective serotonin reuptake inhibitor (SSRI)/serotonin-norepinephrine reuptake inhibitor (SNRI) use.

Results

Participant characteristics

At baseline, mean weight was 51.3 ± 4.9 (SD) kg and mean BMI was 18.6 ± 1.4 kg/m2. Forty-four percent of participants met criteria for atypical AN with BMI>18.5 kg/m2, and 86% of participants with atypical AN had a history of low weight (BMI<18.5 kg/m2). One participant identified as Asian, and one participant identified as Hispanic; the remainder of participants identified as White and non-Hispanic. Thirty-five percent of participants were eumenorrheic, 29% were amenorrheic (no menstrual period within the preceding 3 months), and 35% were taking combination oral contraceptives. Mean baseline DHEAS level did not differ between participants taking vs not taking combination oral contraceptives (178 ± 54 vs 170 ± 78 μg/dL, p=0.73) or between participants with typical AN vs atypical AN (165 ± 69 vs 183 ± 71, p=0.48).

Eighteen participants (53%) gained weight from baseline to 6 months with a mean weight gain of 3.6 ± 2.8 kg (range 0.1–10.3 kg). Sixteen participants lost weight from baseline to 6 months; mean weight loss was 1.1 ± 1.1 kg with a range of 0–3.5 kg. Baseline clinical characteristics and hormone levels are presented in Table 1 by group (participants who gained weight vs participants who did not gain weight). Participants who gained weight had a shorter duration of AN (7.8 ± 6.0 vs 14.4 ± 8.6 years, p=0.01). Otherwise, there were no significant differences in baseline clinical characteristics between groups, including age, BMI, proportion of typical vs atypical AN, reproductive status, and SSRI/SNRI use. Mean baseline DHEAS level (199 ± 72 vs 144 ± 55 μg/dL, p=0.02) and DHEAS index (0.8 ± 0.3 vs 0.6 ± 0.2, p=0.03) were higher in participants who gained weight compared to participants who did not. Mean baseline UFC for the fifteen participants who performed 24-hour collections (of whom eight gained weight and seven did not gain weight) was 20 ± 18 μg/24h (normal range 0–50 μg/24h) and did not differ between groups (p=0.59).

Table 1.

Baseline characteristics.

Participants who gained weight (n=18) Participants who did not gain weight (n=16) p-value
Clinical characteristics
 Age (years) 26.2 ± 7.1 28.9 ± 8.3 0.30
 Weight (kg) 51.2 ± 3.4 51.3 ± 6.3 0.96
 BMI (kg/m2) 18.3 ± 0.3 18.8 ± 0.3 0.33
 Duration of AN (years) 7.8 ± 6.0 14.4 ± 8.6 0.01
 Typical vs atypical AN (%) 61 / 39 50 / 50 0.51
 AN subtype (%): restricting vs binge-purge 78 / 22 69 / 31 0.55
 Reproductive status:
  Amenorrhea (%) 39 19 0.27
  Eumenorrhea (%) 28 44 0.48
  Combination oral contraceptive pill use (%) 33 38 1.00
 Major depressive disorder (%) 56 56 1.00
 Generalized anxiety disorder (%) 78 69 0.70
 SSRI/SNRI use (%) 56 56 1.00
 HAM-D score 15.8 ± 6.0 14.6 ± 3.5 0.48
 HAM-A score 16.3 ± 6.0 14.3 ± 5.1 0.29
Hormone measurements
 DHEAS level (μg/dL) 199 ± 72 144 ± 55 0.02
  DHEAS index 0.8 ± 0.3 0.6 ± 0.2 0.03
 Urinary free cortisol* (0–50) (μg/24h) 21 ± 18 19 ± 20 0.59
Open in a new tab
*

Includes values from 15 subjects who completed 24-hour urine collections for free cortisol.

Values reported as mean ± SD. Abbreviations: BMI, body mass index; AN, anorexia nervosa; SSRI, selective serotonin reuptake inhibitor; SNRI, serotonin and norepinephrine reuptake inhibitor; HAM-D, Hamilton Depression Rating Scale; HAM-A, Hamilton Anxiety Rating Scale. DHEAS index defined as DHEAS level divided by mean normal range for age.

At 6 months, 22 participants (65%) had a BMI >18.5 kg/m2. There was a trend toward a higher baseline DHEAS level (189 ± 72 vs 145 ± 58 μg/dL, p=0.08) and DHEAS index (0.77 ± 0.28 vs 0.60 ± 0.20, p=0.07) in participants with a BMI >18.5 kg/m2 at 6 months compared to participants with a BMI ≤18.5 kg/m2 at 6 months.

DHEAS levels predict weight gain and changes in body composition

In the entire cohort, higher DHEAS levels at baseline predicted weight gain at 3 and 6 months (r=0.49, p=0.004 and r=0.61, p<0.001, respectively; Figure 1A and Figure 1B). This association remained significant after controlling for age, baseline BMI, oral contraceptive use, SSRI/SNRI use, and duration of AN (p<0.001). Duration of AN was negatively associated with weight gain (r=−0.40, p=0.02), but none of the other covariates were predictors of weight gain. When analyzing participants with AN vs atypical AN separately, the association between baseline DHEAS levels and weight gain remained significant within each group. In participants with baseline DHEAS levels less than 100 μg/dL, eight of nine participants lost weight at 6 months. All five participants with baseline DHEAS levels greater than 250 μg/dL gained weight at 6 months. Baseline DHEAS levels also predicted weight at 6 months (r=0.45, p=0.007) and BMI at 6 months (r=0.38, p=0.03).

Figure 1.

Figure 1.

Open in a new tab

Higher baseline DHEAS levels predicted weight gain over (A) 3 months (r=0.49, p=0.004) and (B) 6 months (r=0.61, p<0.001).

Baseline DHEAS levels predicted an increase in fat mass (r=0.40, p=0.03; Figure 2A) and appendicular lean mass (r=0.38, p=0.04; Figure 2B). Baseline DHEAS levels also predicted an increase in TAT (r=0.60, p<0.001; Figure 2C) and SAT (r=0.60, p<0.001), and there was a trend towards an association between baseline DHEAS levels and increase in VAT (r=0.34, p=0.06). All associations remained significant after controlling for age, baseline BMI, oral contraceptive use, SSRI/SNRI use, and duration of AN. Every 1 μg/dL increase in baseline DHEAS was associated with a 0.014 kg increase in total fat mass (95% CI: 0.002, 0.026; p=0.03), 0.006 kg increase in trunk fat (95% CI: 0.000, 0.013; p=0.04), and 0.006 kg increase in lean mass (95% CI: 0.000, 0.011; p=0.04). Baseline DHEAS levels were not associated with a change in extremity fat (p=0.18). After controlling for change in weight, there were no significant associations between baseline DHEAS levels and change in HAM-D score, HAM-A score, EDE-Q scores, or EDI-2 scores.

Figure 2.

Figure 2.

Open in a new tab

Baseline DHEAS levels predicted an increase in (A) fat mass (r=0.40, p=0.03), (B) appendicular lean mass (r=0.38, p=0.04), and (C) total abdominal adipose tissue (r=0.60, p<0.001) over 6 months.

24-hour UFC

Baseline DHEAS levels were positively associated with 24-hour UFC (r=0.61, p=0.02; Figure 3). Baseline 24-hour UFC did not predict change in weight at 6 months (r=0.37, p=0.17) or changes in fat mass, appendicular lean mass, TAT, VAT, or SAT.

Figure 3.

Figure 3.

Open in a new tab

Baseline DHEAS levels were positively associated with baseline 24-hour UFC (r=0.61, p=0.02). UFC values are from the 15 subjects who completed 24-hour urine collections.

Discussion

AN is a severe psychiatric condition marked by treatment resistance, periods of recovery and relapse, and a significant risk of death. Currently, there are no known biomarkers of weight recovery (Malcolm & Phillipou, 2021; Mitchell & Peterson, 2020). DHEAS has been identified as a neuroactive steroid with neuroprotective, neuroexcitatory, anti-depressive and memory enhancing effects (Stárka, Dušková, & Hill, 2015). We report that higher DHEAS levels predict weight gain and increases in fat mass, skeletal muscle mass, and abdominal fat in women with AN over six months. These data suggest that DHEAS has promise as a biomarker in AN.

Outcomes for women with AN are heterogenous; 50% achieve full remission, 20% have chronic illness, and the remainder have residual symptoms (Eddy et al., 2017; Khalsa, Portnoff, McCurdy-McKinnon, & Feusner, 2017). Biomarkers are objective, measurable, and reproducible indicators of a specific medical state that can be used to predict the incidence or outcome of disease and the effects of treatments (Strimbu & Tavel, 2010). Identifying a biomarker for weight gain in women with AN would improve prognostication and may inform timing of intensification of treatment. Leptin, which is secreted by adipose tissue, has generated interest as a biomarker as serum levels are low in underweight patients with AN and increase to normal or supra-normal levels with weight gain. Preliminary evidence suggests that low leptin levels may predict early weight relapse after acute inpatient treatment of AN (Dardennes et al., 2021). However, leptin has not been found to predict increases in BMI during inpatient treatment for AN (Hebebrand et al., 1997; Hebebrand, Muller, Holtkamp, & Herpertz-Dahlmann, 2007; Mantzoros, Flier, Lesem, Brewerton, & Jimerson, 1997; Wabitsch et al., 2001).

DHEA, DHEAS and cortisol are hormones secreted by the adrenal cortex in response to ACTH. DHEAS levels have been shown to correlate with cortisol levels and are abnormal in conditions with dysregulation of the HPA axis, both in conditions of low and elevated cortisol. For example, DHEAS levels are low in adrenal insufficiency (a condition of low cortisol) and elevated in Cushing’s disease (a condition of elevated cortisol) (Li et al., 2020; Nasrallah & Arafah, 2003; Yamaji, Ishibashi, Sekihara, Itabashi, & Yanaihara, 1984). An exception is that during some acute physiologic traumas, such as burns and critical illness, although cortisol levels are high as expected, DHEAS levels are suppressed (Lephart, Baxter, & Parker, 1987; Parker, Levin, & Lifrak, 1985). A subset of patients with AN have activation of the HPA axis; prior studies have demonstrated elevated levels of urinary, serum (using frequent sampling) and midnight salivary cortisol (Biller et al., 1989; Boyar et al., 1977; Doerr et al., 1980; Lawson, Donoho, et al., 2009; Miller, 2011; Misra et al., 2004; Putignano et al., 2001). Examination of DHEAS levels in AN have been inconsistent, with studies reporting low, normal, and elevated levels (Gordon et al., 1999; Lawson, Misra, et al., 2009; Miller et al., 2007; Oskis et al., 2012). In a prior study of women with AN examining cortisol and DHEAS response to ACTH after dexamethasone suppression, cortisol levels were increased and DHEAS levels were unchanged compared to controls (Lawson, Misra, et al., 2009). BMI was not associated with DHEAS but was negatively associated with baseline morning cortisol, cortisol AUC and peak cortisol after ACTH stimulation. However, to our knowledge, the relationship between endogenous DHEAS levels and subsequent weight gain in women with AN has not been previously reported.

We found that in women with AN and atypical AN, DHEAS levels predicted weight gain and increases in fat mass, appendicular lean mass, TAT, and SAT but were not associated with changes in extremity fat. In addition to predicting change in weight, baseline DHEAS levels predicted absolute weight and BMI at 6 months, which suggests that DHEAS may be serve as a marker of weight restoration. As DHEAS levels are highest in the brain (Kroboth, Salek, Pittenger, Fabian, & Frye, 1999), the weight gain and body composition changes observed in participants with higher DHEAS levels may reflect central activity of DHEAS. In rats, DHEAS treatment resulted in higher calorie intake in lean rats and a reduction of calorie intake in obese rates, suggesting a divergent effect depending on baseline weight (Wright, Browne, Svec, & Porter, 1993). Alternatively, elevated DHEAS levels may be a proxy for higher cortisol levels as cortisol is known to stimulate appetite and increase consumption of high-fat sweet foods (Epel, Lapidus, McEwen, & Brownell, 2001; Torres & Nowson, 2007). Some studies have shown that DHEA replacement may improve mood and quality of life in women in adrenal insufficiency (Alkatib et al., 2009), though the preponderance of evidence does not support a clinically meaningful effect. In a study of young women with AN comparing DHEA treatment with conventional hormone replacement therapy (HRT), DHEA and HRT treatment did not increase bone mineral density after accounting for weight gain, but DHEA resulted in improvement in some psychological endpoints (Gordon et al., 2002). We did not find any associations between DHEAS levels and change in depression symptom severity, anxiety symptom severity, or eating disorder psychopathology after controlling for change in weight.

We found that UFC correlated with DHEAS levels in women with AN, consistent with published literature in other populations such as Cushing’s disease and adrenal insufficiency (Li et al., 2020; Nasrallah & Arafah, 2003; Yamaji et al., 1984). UFC did not predict change in weight, fat mass, or appendicular lean mass, which may have been due to the limited number of participants with UFC data. It has previously been shown that high cortisol AUC and UFC predicted recovery of menses and greater increases in fat mass and BMI in girls with AN (Misra et al., 2006). In our study, only 15 participants completed the 24-hour urine collection for cortisol, which speaks to the limited practicality of this test.

In our analysis, we did not find a cutoff for DHEAS that predicts weight gain. We observed that eight of nine women with baseline DHEAS levels less than 100 μg/dL lost weight at 6 months and all five women with baseline DHEAS levels greater than 250 μg/dL gained weight at 6 months. As such, DHEAS levels less than 100 μg/dL and greater than 250 μg/dL may be useful markers for future weight loss or weight gain, respectively, but further studies are needed to validate this finding.

Limitations of this study include the missing UFC values, which prevents us from drawing conclusions regarding UFC as a predictor of clinical endpoints in women with AN. This highlights the difficulty of obtaining 24-hour urine collections and the need for identifying markers of HPA activation and predictors of recovery that are easier to obtain. Longitudinal measures of DHEAS would further elucidate the relationship between DHEAS and weight gain but were not available in these participants. In addition, follow-up beyond 6 months would be necessary to examine the durability of the observed relationship between DHEAS and weight gain. Finally, as the majority of participants identified as White and non-Hispanic, these data may not be generalizable to individuals of other races and ethnicities.

In summary, AN is a severe condition with significant morbidity and mortality, and predictors of weight recovery are needed to improve prognostication and guide therapeutic decision making. DHEAS, a readily available blood test, holds promise as a more practical biomarker for weight gain in women in AN, although confirmatory studies are needed. Further studies are also required to examine the mechanism by which higher DHEAS levels may lead to weight gain in women with AN.

Public Significance Statement.

Anorexia nervosa is a severe psychiatric condition, and predictors of weight recovery are needed to improve prognostication and guide therapeutic decision making. While urinary cortisol is a predictor of weight gain, 24-hour urine collections are challenging to obtain. Like cortisol, dehydroepiandrosterone sulfate (DHEAS) is a hormone produced by the adrenal glands. As a readily available blood test, DHEAS holds promise as more practical biomarker of weight gain in anorexia nervosa.

Funding:

NIH Grants R01 MH083657, T32 DK007028, K24 HL092902, K23 DK115903-01, and K23 MH092560. This work was also conducted with support from the Harvard Catalyst/The Harvard Clinical and Translational Science Center (Grants 1UL1TR001102, 8 UL1 TR000170 from the National Center for Advancing Translational Science, and 1 UL1 RR025758 from the National Center for Research Resources).

Footnotes

Conflicts of interest:

Procter & Gamble provided the study medication (placebo) for the placebo-controlled study at no cost; only data from the placebo arm was used for the analysis reported in this manuscript. Dr. Miller has received study medication from Pfizer and investigator-initiated research grants and study medication from Amgen, and she has had equity in Bristol-Myers Squibb, General Electric, Boston Scientific, Amgen, and Becton Dickinson. The other authors declare no relevant conflicts of interest.

Ethics approval:

Approval was obtained from the Massachusetts General Hospital Institutional Review Board. The procedures used in this study adhere to the tenets of the Declaration of Helsinki.

Data availability:

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

  1. Alkatib AA, Cosma M, Elamin MB, Erickson D, Swiglo BA, Erwin PJ, & Montori VM (2009). A systematic review and meta-analysis of randomized placebo-controlled trials of DHEA treatment effects on quality of life in women with adrenal insufficiency. J Clin Endocrinol Metab, 94(10), 3676–3681. doi: 10.1210/jc.2009-0672 [DOI] [PubMed] [Google Scholar]
  2. American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Disorders (A. P. Association Ed. 5th ed.). Washington, DC. [Google Scholar]
  3. Biller BM, Saxe V, Herzog DB, Rosenthal DI, Holzman S, & Klibanski A (1989). Mechanisms of osteoporosis in adult and adolescent women with anorexia nervosa. J Clin Endocrinol Metab, 68(3), 548–554. doi: 10.1210/jcem-68-3-548 [DOI] [PubMed] [Google Scholar]
  4. Borkan GA, Gerzof SG, Robbins AH, Hults DE, Silbert CK, & Silbert JE (1982). Assessment of abdominal fat content by computed tomography. Am J Clin Nutr, 36(1), 172–177. doi: 10.1093/ajcn/36.1.172 [DOI] [PubMed] [Google Scholar]
  5. Boyar RM, Hellman LD, Roffwarg H, Katz J, Zumoff B, O’Connor J, … Fukushima DK (1977). Cortisol secretion and metabolism in anorexia nervosa. N Engl J Med, 296(4), 190–193. doi: 10.1056/NEJM197701272960403 [DOI] [PubMed] [Google Scholar]
  6. Dardennes R, Tolle V, Lavoisy G, Grouselle D, Alanbar N, Duriez P, … Epelbaum J (2021). Lower leptin level at discharge in acute anorexia nervosa is associated with early weight-loss. Eur Eat Disord Rev, 29(4), 634–644. doi: 10.1002/erv.2830 [DOI] [PubMed] [Google Scholar]
  7. Devesa J, Pérez-Fernández R, Bokser L, Gaudiero GJ, Lima L, & Casanueva FF (1988). Adrenal androgen secretion and dopaminergic activity in anorexia nervosa. Horm Metab Res, 20(1), 57–60. doi: 10.1055/s-2007-1010748 [DOI] [PubMed] [Google Scholar]
  8. Doerr P, Fichter M, Pirke KM, & Lund R (1980). Relationship between weight gain and hypothalamic pituitary adrenal function in patients with anorexia nervosa. J Steroid Biochem, 13(5), 529–537. doi: 10.1016/0022-4731(80)90209-5 [DOI] [PubMed] [Google Scholar]
  9. Eddy KT, Tabri N, Thomas JJ, Murray HB, Keshaviah A, Hastings E, … Franko DL (2017). Recovery From Anorexia Nervosa and Bulimia Nervosa at 22-Year Follow-Up. J Clin Psychiatry, 78(2), 184–189. doi: 10.4088/JCP.15m10393 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Epel E, Lapidus R, McEwen B, & Brownell K (2001). Stress may add bite to appetite in women: a laboratory study of stress-induced cortisol and eating behavior. Psychoneuroendocrinology, 26(1), 37–49. doi: 10.1016/s0306-4530(00)00035-4 [DOI] [PubMed] [Google Scholar]
  11. Fairburn CG, Beglin S (2008). Appendix II Eating Disorder Examination Questionnaire. In CG F (Ed.), Cognitive Behavior Therapy and Eating Disorders. New York, NY: Guildford Press. [Google Scholar]
  12. Fairburn CG, Cooper Z (1991). Eating Disorder Inventory-2. Professional Manual. Odessa: Psychological Assessment Research, Inc. [Google Scholar]
  13. First MB, Spitzer RL, Gibbon M, Williams JB (2002). Structured clinical interview for DSM-IV axis I disorders, research version, non-patient edition. (SCID-I/NP). New York: Biometrics Research. [Google Scholar]
  14. First MB, Williams JB, Karg RS, Spitzer RL (2015). Structured Clinical Interview for DSM-5—Research Version (SCID-5 for DSM-5, Research Version; SCID-5-RV). Arlington, VA: American Psychiatric Association. [Google Scholar]
  15. Geer EB, Lalazar Y, Couto LM, Cohen V, Lipton LR, Shi W, … Freda PU (2016). A prospective study of appetite and food craving in 30 patients with Cushing’s disease. Pituitary, 19(2), 117–126. doi: 10.1007/s11102-015-0690-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gordon CM, Grace E, Emans SJ, Feldman HA, Goodman E, Becker KA, … LeBoff MS (2002). Effects of oral dehydroepiandrosterone on bone density in young women with anorexia nervosa: a randomized trial. J Clin Endocrinol Metab, 87(11), 4935–4941. doi: 10.1210/jc.2002-020545 [DOI] [PubMed] [Google Scholar]
  17. Gordon CM, Grace E, Emans SJ, Goodman E, Crawford MH, & Leboff MS (1999). Changes in bone turnover markers and menstrual function after short-term oral DHEA in young women with anorexia nervosa. J Bone Miner Res, 14(1), 136–145. doi: 10.1359/jbmr.1999.14.1.136 [DOI] [PubMed] [Google Scholar]
  18. Hamilton M (1959). The assessment of anxiety states by rating. Br J Med Psychol, 32(1), 50–55. [DOI] [PubMed] [Google Scholar]
  19. Hamilton M (1960). A rating scale for depression. J Neurol Neurosurg Psychiatry, 23, 56–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hebebrand J, Blum WF, Barth N, Coners H, Englaro P, Juul A, … Remschmidt H (1997). Leptin levels in patients with anorexia nervosa are reduced in the acute stage and elevated upon short-term weight restoration. Mol Psychiatry, 2(4), 330–334. doi: 10.1038/sj.mp.4000282 [DOI] [PubMed] [Google Scholar]
  21. Hebebrand J, Muller TD, Holtkamp K, & Herpertz-Dahlmann B (2007). The role of leptin in anorexia nervosa: clinical implications. Mol Psychiatry, 12(1), 23–35. doi: 10.1038/sj.mp.4001909 [DOI] [PubMed] [Google Scholar]
  22. Khalsa SS, Portnoff LC, McCurdy-McKinnon D, & Feusner JD (2017). What happens after treatment? A systematic review of relapse, remission, and recovery in anorexia nervosa. J Eat Disord, 5, 20. doi: 10.1186/s40337-017-0145-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kimball A, Schorr M, Meenaghan E, Bachmann KN, Eddy KT, Misra M, … Miller KK (2019). A Randomized Placebo-Controlled Trial of Low-Dose Testosterone Therapy in Women With Anorexia Nervosa. J Clin Endocrinol Metab, 104(10), 4347–4355. doi: 10.1210/jc.2019-00828 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kroboth PD, Salek FS, Pittenger AL, Fabian TJ, & Frye RF (1999). DHEA and DHEA-S: a review. J Clin Pharmacol, 39(4), 327–348. doi: 10.1177/00912709922007903 [DOI] [PubMed] [Google Scholar]
  25. Lawson EA, Donoho D, Miller KK, Misra M, Meenaghan E, Lydecker J, … Klibanski A (2009). Hypercortisolemia is associated with severity of bone loss and depression in hypothalamic amenorrhea and anorexia nervosa. J Clin Endocrinol Metab, 94(12), 4710–4716. doi: 10.1210/jc.2009-1046 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lawson EA, Holsen LM, Desanti R, Santin M, Meenaghan E, Herzog DB, … Klibanski A (2013). Increased hypothalamic-pituitary-adrenal drive is associated with decreased appetite and hypoactivation of food-motivation neurocircuitry in anorexia nervosa. Eur J Endocrinol, 169(5), 639–647. doi: 10.1530/EJE-13-0433 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lawson EA, Misra M, Meenaghan E, Rosenblum L, Donoho DA, Herzog D, … Miller KK (2009). Adrenal glucocorticoid and androgen precursor dissociation in anorexia nervosa. J Clin Endocrinol Metab, 94(4), 1367–1371. doi: 10.1210/jc.2008-2558 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lephart ED, Baxter CR, & Parker CR (1987). Effect of burn trauma on adrenal and testicular steroid hormone production. J Clin Endocrinol Metab, 64(4), 842–848. doi: 10.1210/jcem-64-4-842 [DOI] [PubMed] [Google Scholar]
  29. Li T, Cunningham JL, Gilliam WP, Loukianova L, Donegan DM, & Bancos I (2020). Prevalence of opioid induced adrenal insufficiency in patients taking chronic opioids. J Clin Endocrinol Metab. doi: 10.1210/clinem/dgaa499 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lowe B, Zipfel S, Buchholz C, Dupont Y, Reas DL, & Herzog W (2001). Long-term outcome of anorexia nervosa in a prospective 21-year follow-up study. Psychol Med, 31(5), 881–890. [DOI] [PubMed] [Google Scholar]
  31. Malcolm A, & Phillipou A (2021). Current directions in biomarkers and endophenotypes for anorexia nervosa: A scoping review. J Psychiatr Res, 137, 303–310. doi: 10.1016/j.jpsychires.2021.02.063 [DOI] [PubMed] [Google Scholar]
  32. Mantzoros C, Flier JS, Lesem MD, Brewerton TD, & Jimerson DC (1997). Cerebrospinal fluid leptin in anorexia nervosa: correlation with nutritional status and potential role in resistance to weight gain. J Clin Endocrinol Metab, 82(6), 1845–1851. doi: 10.1210/jcem.82.6.4006 [DOI] [PubMed] [Google Scholar]
  33. Miller KK (2011). Endocrine dysregulation in anorexia nervosa update. J Clin Endocrinol Metab, 96(10), 2939–2949. doi: 10.1210/jc.2011-1222 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Miller KK, Lawson EA, Mathur V, Wexler TL, Meenaghan E, Misra M, … Klibanski A (2007). Androgens in women with anorexia nervosa and normal-weight women with hypothalamic amenorrhea. J Clin Endocrinol Metab, 92(4), 1334–1339. doi: 10.1210/jc.2006-2501 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Misra M, & Klibanski A (2014). Endocrine consequences of anorexia nervosa. Lancet Diabetes Endocrinol, 2(7), 581–592. doi: 10.1016/S2213-8587(13)70180-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Misra M, Miller KK, Almazan C, Ramaswamy K, Lapcharoensap W, Worley M, … Klibanski A (2004). Alterations in cortisol secretory dynamics in adolescent girls with anorexia nervosa and effects on bone metabolism. J Clin Endocrinol Metab, 89(10), 4972–4980. doi: 10.1210/jc.2004-0723 [DOI] [PubMed] [Google Scholar]
  37. Misra M, Prabhakaran R, Miller KK, Tsai P, Lin A, Lee N, … Klibanski A (2006). Role of cortisol in menstrual recovery in adolescent girls with anorexia nervosa. Pediatr Res, 59(4 Pt 1), 598–603. doi: 10.1203/01.pdr.0000203097.64918.63 [DOI] [PubMed] [Google Scholar]
  38. Mitchell JE, & Peterson CB (2020). Anorexia Nervosa. N Engl J Med, 382(14), 1343–1351. doi: 10.1056/NEJMcp1803175 [DOI] [PubMed] [Google Scholar]
  39. Nasrallah MP, & Arafah BM (2003). The value of dehydroepiandrosterone sulfate measurements in the assessment of adrenal function. J Clin Endocrinol Metab, 88(11), 5293–5298. doi: 10.1210/jc.2003-030449 [DOI] [PubMed] [Google Scholar]
  40. Oskis A, Loveday C, Hucklebridge F, Thorn L, & Clow A (2012). Diurnal patterns of salivary cortisol and DHEA in adolescent anorexia nervosa. Stress, 15(6), 601–607. doi: 10.3109/10253890.2012.661493 [DOI] [PubMed] [Google Scholar]
  41. Parker LN, Levin ER, & Lifrak ET (1985). Evidence for adrenocortical adaptation to severe illness. J Clin Endocrinol Metab, 60(5), 947–952. doi: 10.1210/jcem-60-5-947 [DOI] [PubMed] [Google Scholar]
  42. Putignano P, Dubini A, Toja P, Invitti C, Bonfanti S, Redaelli G, … Cavagnini F (2001). Salivary cortisol measurement in normal-weight, obese and anorexic women: comparison with plasma cortisol. Eur J Endocrinol, 145(2), 165–171. doi: 10.1530/eje.0.1450165 [DOI] [PubMed] [Google Scholar]
  43. Schorr M, & Miller KK (2017). The endocrine manifestations of anorexia nervosa: mechanisms and management. Nat Rev Endocrinol, 13(3), 174–186. doi: 10.1038/nrendo.2016.175 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Stárka L, Dušková M, & Hill M (2015). Dehydroepiandrosterone: a neuroactive steroid. J Steroid Biochem Mol Biol, 145, 254–260. doi: 10.1016/j.jsbmb.2014.03.008 [DOI] [PubMed] [Google Scholar]
  45. Strimbu K, & Tavel JA (2010). What are biomarkers? Curr Opin HIV AIDS, 5(6), 463–466. doi: 10.1097/COH.0b013e32833ed177 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Topor LS, Asai M, Dunn J, & Majzoub JA (2011). Cortisol stimulates secretion of dehydroepiandrosterone in human adrenocortical cells through inhibition of 3betaHSD2. J Clin Endocrinol Metab, 96(1), E31–39. doi: 10.1210/jc.2010-0692 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Torres SJ, & Nowson CA (2007). Relationship between stress, eating behavior, and obesity. Nutrition, 23(11–12), 887–894. doi: 10.1016/j.nut.2007.08.008 [DOI] [PubMed] [Google Scholar]
  48. Wabitsch M, Ballauff A, Holl R, Blum WF, Heinze E, Remschmidt H, & Hebebrand J (2001). Serum leptin, gonadotropin, and testosterone concentrations in male patients with anorexia nervosa during weight gain. J Clin Endocrinol Metab, 86(7), 2982–2988. doi: 10.1210/jcem.86.7.7685 [DOI] [PubMed] [Google Scholar]
  49. Wright BE, Browne ES, Svec F, & Porter JR (1993). Divergent effect of dehydroepiandrosterone on energy intakes of Zucker rats. Physiol Behav, 53(1), 39–43. doi: 10.1016/0031-9384(93)90008-4 [DOI] [PubMed] [Google Scholar]
  50. Yamaji T, Ishibashi M, Sekihara H, Itabashi A, & Yanaihara T (1984). Serum dehydroepiandrosterone sulfate in Cushing’s syndrome. J Clin Endocrinol Metab, 59(6), 1164–1168. doi: 10.1210/jcem-59-6-1164 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

RESOURCES