Abstract
CONTEXT: The study of gonadal hormone effects on adolescent wellbeing has been limited by logistical challenges. Urine hormone profiling offers new opportunities to understand the health and behavioral implications of puberty hormones. OBJECTIVE: To characterize pubertal change in urinary testosterone and estradiol among male and female adolescents, respectively. DESIGN: Three-year prospective cohort study. SETTING: Australian regional community. PARTICIPANTS: 282 (163 male) normally developing adolescents aged 11.8 ± 1.0 years at baseline. MAIN OUTCOME MEASURE: Quarterly urine measurements of testosterone and estradiol (mass spectrometry); annual anthropometric assessment and Tanner stage (TS) self-report. RESULTS: Two-class sigmoidal and quadratic growth mixture models (centered on age at TS3) were identified as best-fit for describing testosterone (male) and estradiol (female) change. Classes 1 (male: 63%; female: 82%) and 2 (male: 37%; female: 18%) were respectively named the "stable" and "unstable" trajectories, characterized by different standard deviation of quarterly hormone change and magnitude of hormone peaks and troughs (all P < 0.001). Compared with class 1 (stable), class 2 males were taller at baseline (154 vs 151 cm), reported earlier and faster TS progression (P < 0.01), and showed higher serum testosterone levels at baseline and 3 years (P ≤ 0.01). Class 2 females exhibited smaller height and weight gains over the 3 years and had higher baseline serum estradiol (249 vs 98 pmol/L; P = 0.002) than class 1. CONCLUSIONS: Adolescents showed 2 distinct urinary gonadal hormone trajectories, characterized by stability of change over time, which were not associated with consistent anthropometric differences. Results provide a methodology for studying gonadal hormone impacts on other aspects of biopsychosocial wellbeing. Identification of potential "at-risk" hormone groups would be important for planning supportive interventions.