Higher DHEA-S (dehydroepiandrosterone sulfate) levels are associated with depressive symptoms during the menopausal transition: results from the PENN Ovarian Aging Study

Longitudinal cohort study with 11 assessments over 11 years of 436 population-based premenopausal women aged 35-47 at enrollment, including African American and Caucasian participants. Exclusion criteria included current psychotropic or hormonal medications. Mood outcomes were depressive symptoms (CES-D) and major depression diagnosed via PRIME-MMD (periods 7-11). DHEA-S levels were measured from blood samples and analyzed after natural-log transformation; repeated measures were modeled using generalized estimating equations.

DHEA-S levels were positively associated with depressive symptoms over time, both unadjusted (estimate 1.21, p=0.007) and adjusted for age, menopausal stage, race, smoking, and BMI (estimate 1.07, p=0.013). There was no association between DHEA-S and major depression (adjusted odds ratio 0.95, 95% CI 0.72–1.26, p=0.72). The DHEA-S–depressive symptoms relationship persisted across baseline age groups with no significant interactions by age or race. Adding testosterone or estradiol to the model did not alter the association. Menopausal status effects showed lower depressive symptoms during perimenopause and postmenopause compared with premenopause; late perimenopause showed substantially lower odds of major depression (OR 0.27). DHEA-S levels declined about 13% from baseline to the 11-year endpoint (106.4 to 92.3 μg/dL). Conclusion: DHEA-S is positively linked to depressive symptoms but not to major depression during the menopausal transition; findings suggest adrenal hormones relate to mood during this period but DHEA-S alone is not a predictor of major depression.

Limitations: only women in the menopausal transition; two racial groups (African American and Caucasian); excluded antidepressant/hormonal medication users, yielding a healthier sample than other midlife cohorts; DHEA-S measured during the day (diurnal fluctuations exist but are small); cortisol not measured; PENN-5 criteria differ from STRAW and require ongoing validation; potential underpower to detect small associations with major depression; substantial attrition from 436 baseline to 300 at year 11 (and data missing for some measures).