The role of hyperinsulinemia vs. insulin resistance in obese girls with hyperandrogenemia

Program: Abstracts - Orals, Featured Poster Presentations, and Posters
Session: SUN 498-523-Female Reproductive Endocrinology & Case Reports
Sunday, June 16, 2013: 1:45 PM-3:45 PM
Expo Halls ABC (Moscone Center)

Poster Board SUN-505
Amy Denise Anderson*1, Jessicah S. P. Collins1, Ruchi Bhabhra2, Christine Michele Burt Solorzano2, John C Marshall1, Anne C Gabel3 and Christopher Rolland McCartney1
1University of Virginia, Charlottesville, VA, 2Division of Endocrinology, University of Virginia, Charlottesville, VA, 3University of Virginia
The role of hyperinsulinemia vs. insulin resistance in obese girls with hyperandrogenemia

A.D. Anderson, J. S. P. Collins, C. Burt Solorzano, R. Bhabhra, A.C. Gabel, J.C. Marshall, C.R. McCartney

University of Virginia, Charlottesville, VA

Polycystic ovary syndrome (PCOS) is associated with obesity and insulin resistance.  Compensatory hyperinsulinemia contributes to hyperandrogenemia (HA) by augmenting androgen production and inhibiting SHBG production.  Adolescent HA can be a forerunner of PCOS. A majority of peripubertal girls with obesity demonstrate HA, but specific causes remain unknown. While morning LH and fasting insulin predict free testosterone (T) in obese girls, these are imprecise measures of overall LH and insulin levels. We pursued a detailed study with (a) sampling for insulin from 1 h before to 2 h after a standardized meal (at 1900 h) and while fasting (0700-0900 h); (b) frequent sampling for LH and T (1800-0900 h); and (c) a hyperinsulinemic euglycemic clamp (0900-1100 h). An estimate of 24-h insulin assumed that 16 of 24 h were reflected by periprandial insulin and 8 of 24 h by fasting insulin. Based on preliminary data, we have reported that ambient LH concentration is a significant predictor of free T in obese girls; however, in contrast to our expectations, estimated 24-h insulin does not. We assessed whether free T is better predicted by either (a) an expanded insulin data set or (b) insulin sensitivity index (ISI). We have studied 11 obese subjects: age 12.3 ± 2.4 (mean ± SD); Tanner stage 3.7 ± 1.5; BMI-for-age-percentile (BMI%) 98.6 ± 1.0. In addition to the above times, insulin was measured in stored blood (2100-0700 h), and an alternate estimate for 24-h insulin assumed that 18 of 24 h was reflected by insulin values for 6 h post-meal(1900-0100 h) and 6 of 24 h by fasting levels (0200-0700 h). Alternate and original estimates for 24-h insulin were highly correlated (r = 0.94, p <0.0001). Estimated 24-h LH correlated with free T (r = 0.68, p = 0.02). Free T did not correlate with either estimated 24-h insulin exposure (by either method) or ISI (r = 0.24-0.33, p > 0.3 for all). After adjusting for differences in Tanner stage and estimated 24-h LH, ISI significantly correlated with free T (r = 0.766, P = 0.016), but estimates of 24-h insulin did not (r = 0.44-0.45, p > 0.2 for both). These data suggest that ISI may be a better predictor of free T than estimates of 24-h insulin in obese girls. This could reflect limitations of periodic insulin sampling to accurately quantify ambient hyperinsulinemia, but may also imply that other factors involved in insulin resistance are important in the genesis of HA.

Nothing to Disclose: ADA, JSPC, RB, CMB, JCM, ACG, CRM

*Please take note of The Endocrine Society's News Embargo Policy at