Longitudinal Investigation of Pubertal Milestones and Hormones as a Function of Body Fat in Girls

Background: Studies comparing the timing and pace of puberty in overweight/obese girls (OW/OB) vs normal weight girls (NW) have produced conflicting results; some suggest earlier activation of the central components of the reproductive axis in OB while others are more consistent with a peripheral source of estrogen (e.g. adipose tissue) driving puberty in OB. Importantly, there have been no longitudinal assessments of both clinical and biochemical pubertal markers in OB vs. NW. Methods: 90 healthy pre-menarchal girls (26 OW/OB, 54 NW) from the community, aged 8.2-14.7 years, completed 2.8 ± 1.7 (mean, SD) study visits over the course of 4 years. Visits included dual-energy x-ray absorptiometry to calculate percent total body fat (TBF), Tanner staging, breast ultrasound for morphological staging (BMORPH; stages A-E), pelvic ultrasound, hand x-ray (bone age, BA), blood tests for reproductive hormones, and urine collection to determine a vaginal maturation index (VMI), an index of estrogen exposure in urogenital epithelial cells. Menarchal status was determined at each visit and via follow-up questionnaires. The effect of TBF on hormones and markers of estrogen action, the pace of breast maturation, and age at menarche were determined using a mixed, multi-state, or Cox proportional hazards model, respectively. Mixed and Cox models controlled for BMORPH at visit 1 (V1) and race. Results: NW girls were older than OW/OB (11.3 vs. 10.2 yrs, p<0.01) at V1, more likely to be non-Hispanic White (66 vs. 40%, p=0.03), and had more advanced breast morphology BMORPH (p<0.01). LH, E2, VMI, BA, and ovarian and uterine volume increased with time with no effect of TBF. There was an interaction between time and TBF for FSH, INHB, E1, Total T, Free T, and A’dione (p<0.05): levels were initially similar in all TBF groups, but after 1 yr, levels increased in girls with higher TBF, plateaued in girls with mid-range TBF, and decreased in girls with lower TBF. Girls with higher TBF progressed through BMORPH stage D (corresponding to growth/arborization of the breast ductal system), more slowly than girls with lower TBF but achieved menarche at a younger age (risk 1.04x higher per 1 unit increase in TBF). Conclusions: Intensive reproductive phenotyping of girls during the pubertal transition reveals that both the neuroendocrine and ovarian components of the axis are generally preserved in girls with higher TBF but that the axis appears to be activated earlier than in girls with lower TBF. In late puberty, however, girls with higher TBF demonstrate subtle differences in standard hormonal (e.g. serum FSH, INHB, and androgen) and clinical (e.g. delayed growth of breast bud) markers of puberty. Investigation of the mechanistic basis for these differences and their potential clinical consequences for girls with higher TBF deserves further study.