LAUSR

OBJECTIVE Deficiency in 25-hydroxyvitamin D (25OHD) has been associated with sleep disorders in adults. Only three cross-sectional studies were performed in children; which showed an association between 25OHD deficiency and both obstructive sleep apnea syndrome and primary snoring. No longitudinal study has been performed in children from the general population. We analyzed the association between cord-blood vitamin D levels at birth and night-sleep duration trajectories for children between 2 and 5-6 years old in a non-clinical cohort. METHOD We included 264 children from the French EDEN mother-child birth-cohort with cord-blood 25OHD level determined by radio-immunoassay at birth, and night-sleep trajectories for children between 2 and 5-6 years old obtained by the group-based trajectory modeling method. Associations between 25OHD and sleep trajectories were assessed by multinomial logistic regression adjusted for maternal and child characteristics. RESULTS The trajectories short sleep (<10h30/night), medium-low sleep (10h30-11h00/night), medium-high sleep (≈11h30/night), long sleep (≥11h30/night) and changing sleep (decreased from ≥11h30 to 10h30-11h00/night) represented 5%, 46%, 37%, 4% and 8% of the children, respectively. The mean 25OHD level was 19 ng/ml (SD = 11, range 3-63). It was 12 (SD = 7), 20 (SD = 11), 19 (SD = 10), 14 (SD = 7) and 16 (SD = 8) ng/ml for children with short, medium-low, medium-high, long and changing sleep trajectories, respectively. On adjusted analysis, for each 1-ng/ml decrease in 25OHD level, the odds of belonging to the short sleep versus medium-high sleep trajectory was increased (odds ratio = 1.12, 95% confidence interval [1.01-1.25]). We found no other significant association between 25OHD level and other trajectories. CONCLUSION A low 25OHD level at birth may be associated with an increased probability of being a persistent short sleeper in preschool years. These results need confirmation.