BACKGROUND: Shorter weeknight sleep duration has not been assessed as a mediating mechanism linking earlier pubertal timing to a greater burden of adolescent depression symptoms. METHODS: Among 1138 participants (48.6 % female) from Project Viva, a pre-birth longitudinal cohort, we examined relationships among pubertal timing measures, actigraphy-captured and self-reported weeknight sleep duration across mid-adolescence, and depression symptoms in late adolescence. We assessed pubertal timing using age at peak height velocity, self-reported adrenarche (Tanner pubic hair stage) and parent-reported Pubertal Development Scale scores in early adolescence. Relationships were adjusted for age, sex, mid-childhood BMI, and socioeconomic status
effect modification by sex was considered throughout. The mediational g-formula estimated the indirect effect of pubertal timing on depression symptoms via weeknight sleep duration. RESULTS: All measures of earlier pubertal timing predicted shorter actigraphy-measured sleep duration
self-reported adrenarche predicted self-reported sleep duration and depression symptoms. No effect modification by sex was found. In both sexes, a 1-stage advance in adrenarche was associated with less (-0.14 h (95 % CI -0.23,-0.05)) self-reported sleep on weeknights across the mid-adolescent years. The relationship between earlier adrenarche and depression symptoms was mediated by self-reported weeknight sleep duration - a decrease in sleep duration in mid-adolescent years attributable to a 1-stage advance in adrenarche increased depression symptoms by 0.22 points (95 % CI 0.08,0.40). LIMITATIONS: The geographic specificity and attrition in Project Viva limit the generalizability of our findings. CONCLUSIONS: Maintaining adequate weeknight sleep across adolescent years may mitigate the impact of earlier adrenarche on depressed mood in late adolescence.