Brittany K. Wise-Oringer, Aviva B. Sopher, Patricia M Vuguin, Sharon E. Oberfield, Ilene Fennoy, Kristen M. Williams, Anne Claire Burghard, Abeer Hassoun, Renu Nandakumar, Patrick O’Day, and Richard J. Auchus
Premature adrenarche (PA), the early onset of pubic hair and/or axillary hair/odor in children, is associated with elevated adrenal androgens and precursors in the absence of gonadotropin-dependent puberty. Laboratory data in PA classically demonstrate increased DHEAS, T, and A4 levels that correlate with pubic hair development. In premature pubarche (PP), the clinical presentation occurs in the absence of elevated DHEAS, T, and A4. PA is associated with insulin resistance and progression to metabolic syndrome (MetS) and PCOS; it is unclear which of these children are at risk for metabolic abnormalities. Adrenally-derived 11-oxygenated C19 steroids (11oAs) have comparable androgenic potency to T and DHT and are elevated in disorders of androgen excess. We sought to characterize the 11oA profiles of children with PA/PP and controls and to correlate them with traditional androgens and metabolic markers, including criteria for childhood MetS. A prospective cross-sectional study was performed of subjects with PA or PP (5 M, 14 F) and controls (2 M, 6 F) ages 3 – 8 yrs (F) or 3 – 9 yrs (M). Children with precocious puberty, steroid use, or recent illness were excluded. Fasting early morning serum was collected, a complete physical exam was performed, and BP and waist circumference were measured; a bone age was obtained only in PA/PP subjects. 11oAs (11OHT, 11KT, 11OHA4, 11KA4) were analyzed by LC-MS. Subjects were divided into PA (DHEAS ≥ 50 µg/dL, n=10) or PP (DHEAS < 50 µg/dL, n=9) for sub-analysis. There were no significant differences in sex, race/ethnicity, BMI z-score, preterm gestation, birth weight, family history, or clinical criteria for childhood MetS. T, A4, DHT, DHEAS, and all 11oAs were significantly higher in PA/PP subjects. While lipids did not differ, insulin and HOMA-IR were higher in PA/PP vs. controls {insulin Mdn = 8.2 (IQR 3.5 – 10.0) vs. 2.0 (2.0 – 3.3) µIU/mL, p < 0.03; HOMA-IR Mdn = 1.8 (IQR 0.8 – 2.1) vs. 0.4 (0.4 – 0.8), p < 0.03}. In a sub-analysis of PA vs. PP, there were no differences in baseline characteristics or metabolic markers. DHEAS was elevated in PA vs. PP {Mdn = 95 (IQR 73 – 111) vs. 42 (36 – 46) µg/dL, p < 0.00003}, although no differences were noted in 11oA levels. Correlations of androgens and their precursors suggested best correlation of 11KT and 11OHA4 with T (ρ=0.87; ρ=0.87) and A4 (ρ=0.87; ρ=0.88). There was moderate correlation of 11KT and 11OHT with insulin (ρ=0.47; ρ=0.51) and HOMA-IR (ρ=0.43; ρ=0.47). We conclude that PA and PP differ only by DHEAS (by definition) and not by insulin sensitivity or 11oA, consistent with 11oA – rather than DHEAS – mediating the phenotypic changes of pubarche. These pilot data are the first to report the early morning steroid metabolite levels including 11oAs in a phenotypically and metabolically well-defined group of PA, PP, and age-matched male and female controls. The relationships between PA, PP, risk for MetS, and 11oA warrant further study.