Vitamin D deficiency is widespread globally in pregnant women and is suggested to contribute to offspring risk of symptoms of autism spectrum disorders (ASD), but findings are inconsistent.

To investigate whether low prenatal 25OHD exposure and 25OHD status in early life increase risk of early symptoms of ASD.

To investigate early signs of ASD.

From the prospective birth cohort, Odense child cohort (n = 2.549), Denmark, we included singletons with their mothers with available serum concentrations of 25OHD at early pregnancy (n = 1.231), late pregnancy (n = 1.361), and birth (n = 2.082). ASD symptoms were rated by parents on the pervasive developmental problem (PDP) scale of the Child behavior checklist for ages 1½-5 (CBCL 1½-5). Associations between 25OHD and PDP score were analyzed by multiple linear and logistic regression models. A priori included covariates were gestational age, child gender, birth weight, season of birth, parity, parental psychiatric diagnoses, maternal age, smoking habits, alcohol consumption, education level, vitamin D supplementation, and pre-gestational BMI.

Means (SD) of 25OHD for early pregnancy, late pregnancy, and cord were 65.5 (21.5), 78.7 (27.0), and 47.0 (21.7) nmol/L, respectively. PDP data were obtained at a mean (SD) of 2.7 (0.6) years of age. Data from the regression analyses are pending.

In this well-off cohort with relatively high levels of cord 25OHD, power calculations allows us to detect a true change of 0.02 PDP scores per 10 nmol/L change in 25OHD. Further studies will clarify whether early PDP scores track later ASD diagnosis.

The authors have not supplied their declaration of competing interest.