Vitamin D supplementation in the first 2 years and autism spectrum traits at 6-8 years - a randomized clinical trial

Early life vitamin D levels may be associated with autism spectrum disorder (ASD) and related traits, but causality is unknown. We examine whether higher-than-standard vitamin D₃ supplementation during the first 2 years, as well as higher pregnancy and childhood 25-hydroxyvitamin D (25(OH)D) levels and their trajectories, are associated with lower ASD trait scores at ages 6-8 years in a non-clinical cohort.

This secondary analysis of the double-blind randomized clinical trial vitamin D intervention in infants (VIDI) comprised 366 Finnish children aged 6-8 years, 177 of whom were randomized to receive 400-IU and 189 to receive 1,200-IU daily oral vitamin D₃ supplementation between ages 2 weeks and 2 years. ASD-related traits were assessed at mean age 7.2 years (SD 0.4) using the parent-reported Autism Spectrum Screening Questionnaire (ASSQ). Predictor variables were supplementation group, 25(OH)D concentrations measured during pregnancy and at ages 1 and 2 years, as well as 25(OH)D trajectories (high vs. low) derived from these time points.

None of the predictor variables of interest were associated with the outcome in the full sample. After sex stratification, among boys, 25(OH)D concentrations at 1 and 2 years were inversely associated with ASSQ scores (mean difference -0.2 of normalized SD score (95% CI -0.3 to -0.1, p = .003) and -0.2 (95% CI -0.3 to -0.05, p = .01) per 10 ng/mL 25(OH)D) after adjustment for age, breastfeeding, parental education, maternal depressive symptoms, and season of 25(OH)D assessment as was belonging to the higher 25(OH)D trajectory, -0.45 SD (95% CI -0.79 to -0.10, p = .01).

We found no indication that higher-than-normal vitamin D₃ supplementation between ages 0 and 2 years decreases ASD-related trait scores at ages 6-8 years. Sex-stratified analysis suggested an inverse association, among boys, between early life 25(OH)D concentrations and ASD-related traits, warranting further studies on potential causal direction and sex specificity of associations.