[Predictive effects of pubertal timing and tempo on adolescent depressive symptoms: a 6-year prospective cohort study].

To examine the prospective association of pubertal timing and tempo with depressive symptoms in adolescents. Since 2013, 2 084 students in grade 1-3 were selected from two primary schools in Bengbu, Anhui Province were selected by using convenience sampling method to establish the adolescence pubertal development cohort. Followed up for 6 years, physical examination, secondary sexual development evaluation (testicular volume for boys and breast development for girls) and depressive symptoms were evaluated biennially. Non-linear growth model was used to estimate pubertal timing and tempo for boys and girls respectively. Depressive symptoms were interviewed by using the Short Mood & Feeling Questionnaire (SMFQ) at baseline and Mood & Feeling Questionnaire (MFQ) during follow-up for students in grade 1-2. Children Depression Inventory (CDI) was used for students in grade 3 at baseline and during follow-up. Depressive symptom scores were standardized by using the -score method. Multivariate linear regression model was used to analyze the predictive effects of modeling pubertal timing and tempo on depressive symptoms of adolescence boys and girls. There were 1 909 students with complete questionnaire and puberty development information, including 1 052 boys (59.19%) and 857 girls (43.81%), with average age about (13.94±0.87) years and 91.60 percent follow-up rate. The average modeling pubertal timing of girls (11.25 years) was earlier than that of boys (12.70 years), and the average pubertal tempo of girls about 1.47 Tanner stage/year was faster than that of boys about 1.28 Tanner stage/year. After controlling for depressive symptoms, maternal education and adverse childhood experiences at baseline and age, body mass index (BMI) classification and sleep time during follow-up, this predictive effect of pubertal timing and tempo on depressive symptoms was only significant among girls. Compared with girls with on time pubertal timing, girls in the delay timing group had a lower level of depressive symptoms (β=-0.19, 95% :-0.34,-0.01). Compared with girls in average pubertal tempo group, the fast tempo group associated with an increasing risk of depressive symptoms (β0.23, 95%: 0.05, 0.40), while the slow tempo group associated with an decreasing risk of depressive symptoms (β-0.21, 95％:-0.39,-0.03). Insignificant effects were found in puberty timing and tempo on depressive symptoms of boys (>0.05). Fast pubertal tempo increases the risk of development of depressive symptoms of adolescent girls. There is no predictive effect of pubertal timing and tempo on depression symptoms of adolescent boys.