Suppression of Serum Dehydroepiandrosterone Sulfate Levels by Insulin: An Evaluation of Possible Mechanisms *

Abstract

We previously demonstrated a progressive decline in serum dehydroepiandrosterone sulfate (DHEA-S) levels in women during a hyperinsulinemic-euglycemic clamp. To determine whether this fall in serum DHEA-S levels might have been due to insulin-stimulated 1) hydrolysis of DHEA-S to dehydroepiandrosterone (DHEA), 2) conversion of DHEA-S/DHEA to androstenedione, and/or 3) urinary excretion of these steroids, 10 additional men were studied by the hyperinsulinemic-euglycemic clamp technique. Each man received a 0.1 U/kg (0.72 nmol/kg) insulin bolus dose, followed by a 10 mU/kg-min (72 pmol/kg-min) insulin infusion for 4 h. An average insulin level of 12,390 ± 259 (± SE) pmol/L (1,726.8 / 36 μU/mL) was achieved; serum glucose was maintained at 5.0 / 0.1 mmol/L (90.5 / 2.3 mg/dL). During the hyperinsulinemia, serum DHEAS levels fell progressively and were significantly lower than baseline at 4 and 6 h of study (85.5 ± 5.9% and 79.1 ± 3.2% of baseline values, respectively; P < 0.05). Serum DHEA levels fell concurrently and were significantly lower than baseline at 2, 4, and 6 h of study (66.2 ± 12.3%, 61.6 ± 11.2%, and 52.9 ± 10.2% of baseline values, respectively; P < 0.05). The percent fall in serum DHEA levels correlated positively with the percent fall in serum DHEA-S levels (r = 0.44; P < 0.02). Serum androstenedione levels also fell progressively during hyperinsulinemia and were significantly lower than baseline at 2, 4, and 6 h of study 71.5 ± 4.1%, 71.0 ± 7.2%, and 48.1 ± 3.3% of baseline values, respectively; P < 0.05). No change in serum DHEA-S, DHEA, or androstenedione levels occurred in paired control studies, during which 0.45% saline was infused at rates matched exactly to the rates of the dextrose and insulin infusions during the hyperinsulinemic clamp studies. Despite decreasing serum DHEA-S and DHEA levels during hyperinsulinemia, urinary DHEA-S and DHEA glucuronide excretions were increased by 50% (P < 0.05) and 86% (P=0.05), respectively, compared to urinary excretion of these steroids during control studies. In contrast, urinary excretion of unconjugated DHEA was unchanged. Quantitatively, however, increased urinary excretion of conjugated DHEA during hyperinsulinemia accounted for only about 5% of the concomitant fall in serum DHEA-S concentrations. These observations indicate that hyperinsulinemia reduces serum DHEA-S, DHEA, and androstenedione levels in men and suggest that neither insulin-stimulated hydrolysis of DHEA-S nor increased conversion of DHEA-S/DHEA to androstenedione is a likely mechanism for the fall in serum DHEA-S levels. Moreover, although hyperinsulinemia increases urinary excretion of conjugated DHEA, the contribution of this mechanism to the insulin-induced fall in serum DHEA-S levels is negligible.