OR25-6 Higher Ovarian Androgen Responses to Ambient Luteinizing Hormone and to Recombinant Human Chorionic Gonadotropin Stimulation in Late Pubertal Girls with Obesity

Abstract

Our studies suggest that 60% of obese peripubertal girls have androgen excess, a precursor to polycystic ovary syndrome. Since the ovary is a major source of androgens in women, we studied steroid levels after dexamethasone (DEX) and after recombinant human chorionic gonadotropin (r-hCG) in girls to assess how obesity affects ovarian androgens during puberty. We studied 41 peripubertal girls in the mid-follicular phase (if relevant, 11 were pre-menarcheal): age 13.2 ± 3.0 (7.3-18.8) y (mean ± SD [range]); BMI Z-score 1.2 ± 1.1 (-1.08 to +2.64); Tanner breast 4.0 ± 1.3 (1-5); bone age 14.2 ± 2.9 (7.8-18) y. In early pubertal girls (Tanner 1-3), 5 were normal weight (NW, BMI%-for-age [BMI%] 5-84) and 6 overweight (OW, BMI% ≥ 85). In late pubertal girls (Tanner 4-5), 12 were NW and 18 OW. DEX (1 mg po) was given at 10 pm. At 7 am the next day, progesterone (P4), 17-OH progesterone (OHP), 17-OH pregnenolone, androstenedione (A4), dehydroepiandrosterone (DHEA), testosterone (T), estradiol (E2), and anti-mullerian hormone were measured immediately before r-hCG (25 mcg IV) was given. DEX was again given at 10 pm; hormone levels were repeated 24h after r-hCG. Free T was calculated using SHBG. Forty girls also had frequent (q 10 min) LH sampling from 7 pm-7 am. Comparisons were made using Student’s t-tests. In early pubertal girls, NW girls had higher DHEA levels after r-hCG compared to OW girls (2.46 ± 1.07 vs. 1.23 ± 0.68 ng/dL, p=0.04). In late puberty, OW girls had higher free T after DEX (19.3 ± 15.6 vs. 8.6 ± 4.1 pmol/L, p=0.03) and after r-hCG (27.2 ± 22.3 vs. 10.7 ±4.5 pmol/L, p=0.02) versus NW girls. OW girls also had higher total T levels after DEX (21.7 ± 15.2 vs. 16.7 ± 7.4 ng/dL, p=0.3) and after r-hCG (30.0 ± 20.4 vs. 21.1 ± 10.2 ng/dL, p=0.2), but did not achieve significance. SHBG was lower in OW versus NW girls in late puberty only (20.7 ± 13.4 vs. 48.0 ± 21.8 nmol/L, p=0.0002). No other hormone levels were different. Overnight mean LH was not different for NW and OW groups, for either early or late puberty. However, in late pubertal girls, 33% in OW group had mean LH values ≥ 5 (5.3-9.3) mIU/mL compared to 17% in NW group. Compared to OW girls with mean LH < 5, OW girls with mean LH ≥ 5 had higher OHP, A4, free T, and total T after DEX and after r-hCG (p ≤ 0.05 for all, except A4 after r-hCG [p=0.06]). Results suggest that OW late pubertal girls had higher free T levels after DEX, which may reflect steady-state ovarian androgen production related to ambient LH milieu. OW late pubertal girls had higher free T levels after r-hCG, which may suggest increased capacity to produce ovarian androgens in response to stimulation. In a subgroup of OW girls, obesity during late puberty was associated with higher mean LH values and higher androgen levels after DEX and r-hCG, in keeping with our earlier reports that LH levels predict free T levels in girls with obesity.