The estimated effect of season and vitamin D in the first trimester on pubertal timing in girls and boys : A cohort study and an instrumental variable analysis

Background: Season of birth has been associated with age at menarche. Maternal vitamin D levels in pregnancy may explain this effect. We investigated whether the season of first trimester or maternal 25-hydroxyvitamin D3 [25(OH)D3] levels were associated with pubertal timing in children. Methods: We conducted a follow-up study of 15 819 children born in 2000-03 from the Puberty Cohort, nested in the Danish National Birth Cohort (DNBC). Mean differences in attaining numerous pubertal markers, including a combined estimate for the average age at attaining all pubertal markers, were estimated for low (November-April) relative to high (May-October) sunshine exposure season in the first trimester using multivariable interval-censored regression models. Moreover, we conducted a two-sample instrumental variable analysis using season as an instrument for maternal first-Trimester 25(OH)D3 plasma levels obtained from a non-overlapping subset (n = 827) in the DNBC. Results: For the combined estimate, girls and boys of mothers who had their first trimester during November-April had earlier pubertal timing than girls and boys of mothers whose first trimester occurred during May-October:-1.0 months (95% CI:-1.7 to-0.3) and-0.7 months (95% CI:-1.4 to-0.1), respectively. In the instrumental variable analysis, girls and boys also had earlier pubertal timing: respectively,-1.3 months (95% CI:-2.1 to-0.4) and-1.0 months (95% CI:-1.8 to-0.2) per SD (22 nmol/L) decrease in 25(OH)D3. Conclusions: Both first pregnancy trimester during November-April and lower 25(OH)D3 were associated with earlier pubertal timing in girls and boys.