Milk Cortisol and Childhood Obesity: Clarifications and Future Directions

We are thankful for the interest with which Finken and colleagues read our paper and appreciate the opportunity to discuss the important ideas they present. Finken et al. express concern that we assessed milk cortisol at one time-point to predict changes in infant BMIP because cortisol concentrations in milk vary across the day, reaching their apex shortly after waking and their nadir at night.1,2 While we agree that collecting 24-hour milk samples will constitute an important methodological advance in future research, we regard our current findings to be reliable for several reasons. First, we included the time-of-day of sample collection in our multivariate growth curve models. By statistically adjusting for temporal variation, we isolated the contribution of milk cortisol on infant BMIP. Second, although it is true that the morning apex and nightly nadir of milk cortisol differ by ~500%,1 cortisol variation in the timeframe relevant to our study (1100 – 1700) only differs by ~30%.1 Consistent with this limited variation, the correlation between time-of-day and milk cortisol in our study was relatively modest and only marginally significant (β = .27, p = .050). Given these complementary considerations, our study most likely captured between-mother variation in milk cortisol, as intended, rather than within-mother variation in diurnal milk cortisol. Furthermore, the vast majority of prior research on early glucocorticoid exposure and later obesity risk derives from samples taken at one time-point.3