Post-mini-puberty testosterone and infant alpha-amylase during caregiver separation: A secondary, age-restricted analysis

Human males exhibit a testosterone surge during the first months of postnatal life, an early HPG-axis activation known as mini-puberty. Rodent work has established that early HPG-axis activity organizes the calibration of later HPA reactivity, with a smaller literature also implicating the sympathetic branch. Whether individual variation in human mini-puberty androgen exposure predicts later stress-system function has not been directly tested. We addressed this question in a secondary analysis of an existing male infant cohort. The primary analytic sample was restricted to visits within a peri-mini-puberty age window anchored to the cohort's empirical cross-sectional testosterone inflection, and results were compared against the full cohort and against a tighter age-restriction cutoff as robustness checks. This study examined whether salivary testosterone, measured by enzyme immunoassay, at 3 and 10 months predicted cortisol and salivary alpha-amylase (sAA) responses to a Strange Situation caretaker separation episode. Bayesian path models tested baseline, reactivity, and recovery for both hormones, adjusting for infant age at each visit, time of day at each collection, and illness at each visit. Contrary to predictions from rodent models, testosterone sampled within the mini-puberty window did not credibly predict any outcome. However, testosterone at 10 months-after this window had closed-was concurrently associated with stronger sAA reactivity, robust across priors, the tighter age-restriction cutoff, and adjustment for behavioral distress, perinatal and 10-month body size, and feeding time, and less robustly with higher tonic sAA. A cortisol reactivity association present in the full cohort attenuated below the credibility threshold in the restricted window. Our results suggest that the developmental signature of infant testosterone on later stress responses may sit outside the mini-puberty window and involve the sympathetic branch, though these conclusions rest on this restricted secondary analysis of a modest male infant sample. Our discussion includes a call for replication, specifically using LC-MS/MS-measured testosterone in independent cohorts.