Are insulin resistance and central adiposity associated with lower DHEA-S levels in women?

The relationship between adrenal androgens, fat distribution, and metabolic health in women is complex and changes significantly across the lifespan. While cellular mechanisms suggest that insulin can suppress adrenal androgen production, clinical evidence reveals a highly stage-dependent relationship.

Clinical and epidemiological evidence

• Premenopausal women: In younger, premenopausal cohorts, observational studies demonstrate an inverse correlation where higher endogenous DHEA levels are associated with lower visceral and central adiposity, suggesting a protective, anti-adipogenic role at the tissue level.
• Postmenopausal women: This association reverses in late menopause. In postmenopausal women, higher DHEA-S levels are positively correlated with increased abdominal fat accumulation, indicating that the loss of ovarian estrogen alters how adrenal androgens affect adipose tissue distribution.
• Insulin resistance associations: Epidemiological data indicate that DHEA-S does not independently predict insulin resistance or type 2 diabetes risk in women once adjustments are made for age, overall adiposity, and other circulating sex hormones.

Mechanistic explanations

• SULT2A1 downregulation: In vitro studies in adrenocortical cells show that insulin exposure downregulates the transcription of the SULT2A1 gene. Because the SULT2A1 enzyme is responsible for sulfating DHEA into DHEA-S, this downregulation directly reduces DHEA-S synthesis and secretion.
• Androgen interaction: Cellular models indicate that insulin acts synergistically to enhance testosterone's inhibitory effects on both DHEA-S and cortisol secretion in adrenal cells.
• Adipose-insulin axis: Visceral adiposity functions as a primary driver of insulin resistance in women, which in turn can modulate adrenal steroidogenesis via the insulin-SULT2A1 pathway.

Bottom line

• The association between central adiposity, insulin resistance, and lower DHEA-S levels is plausible but highly modified by menopausal status; a clear negative correlation exists in premenopausal women, whereas postmenopausal women show the opposite pattern.