Higher-dose vitamin D3 supplementation during pregnancy was associated with modestly higher verbal and visual memory scores among children at age 10 years, according to a post hoc secondary analysis of a randomized clinical trial.
The findings, published in JAMA Network Open, did not show statistically significant differences in estimated intelligence or most other cognitive functions between children whose mothers received higher-dose vs standard-dose vitamin D3 supplementation.
Researchers analyzed data from the blinded, placebo-controlled Copenhagen Prospective Studies on Asthma in Childhood 2010 trial in Denmark. Pregnant patients were randomized at 24 weeks' gestation to receive either an additional 2,400 IU/d of vitamin D3 plus the recommended 400 IU/d dose, for a total of 2,800 IU/d, or standard-dose supplementation of 400 IU/d through 1 week postpartum.
The original 2 × 2 factorial trial was designed to evaluate persistent wheeze or asthma in offspring and also included randomization to fish oil or olive oil supplementation. The cognitive analysis was post hoc and was not prespecified as the trial's primary outcome.
The current analysis included 498 children who completed neurocognitive testing at a mean age of 10 years. The test battery assessed 8 cognitive domains and 11 cognitive functions, including estimated intelligence, processing speed, reaction time, sustained attention, motor speed, verbal and visual memory, verbal and spatial working memory, flexibility or set shift, and planning.
Following adjustment for child sex, age at testing, season of birth, maternal preintervention 25-hydroxyvitamin D level, and concurrent omega-3 fatty acid supplementation, higher-dose vitamin D3 was associated with higher verbal memory and visual memory scores compared with standard-dose supplementation. The effect sizes were modest, at 0.17 SD for verbal memory and 0.24 SD for visual memory; both associations remained statistically significant after false discovery rate correction.
Higher-dose supplementation was also associated with better flexibility or set shift before correction for multiple comparisons, but that finding did not remain statistically significant after false discovery rate correction. No statistically significant differences were found for estimated intelligence, processing speed, reaction time, sustained attention, motor speed, verbal working memory, spatial working memory, or planning.
The mean estimated intelligence score was similar between groups, at about 108 in both the higher-dose and standard-dose groups. Among the 11 cognitive functions assessed, only the two memory measures remained significantly associated with supplementation after correction for multiple comparisons.
Two prior analyses of the same COPSAC2010 cohort found no effect of prenatal higher-dose vitamin D3 supplementation on neurodevelopment from birth to age 6 years and no effect on neurodevelopmental disorders at age 10 years.
In secondary observational analyses, maternal 25-hydroxyvitamin D levels measured at 24 weeks' gestation were not associated with verbal or visual memory. The researchers noted that the discrepancy between the supplementation findings and the measured-level analyses may reflect differences in exposure timing or residual confounding.
Analyses based on postpartum 25-hydroxyvitamin D levels of 40 ng/mL or higher vs less than 40 ng/mL also did not identify a threshold associated with improved cognitive outcomes. About 58% of children in the higher-dose group had mothers who reached postpartum concentrations of 40 ng/mL or higher; the researchers noted that limited exposure contrast may have contributed to the lack of a clear threshold.
Sensitivity analyses found no statistically significant interactions by child sex, maternal preintervention vitamin D status, child vitamin D levels at 6 months or 6 years, or concurrent omega-3 fatty acid supplementation. In analyses restricted to children without attention-deficit/hyperactivity disorder, the associations with verbal and visual memory persisted; however, interaction testing was not statistically significant, limiting interpretation of this subgroup finding. Analyses by autism spectrum disorder status were limited by the small number of children with autism spectrum disorder.
The researchers noted several limitations, including the post hoc and non-prespecified nature of the analysis, multiple cognitive comparisons, and the fact that the trial was not originally powered for cognitive outcomes. The cohort was predominantly White and relatively vitamin D sufficient, with median maternal preintervention 25-hydroxyvitamin D levels above 30 ng/mL, which may limit generalizability to more diverse or vitamin D–deficient populations. Supplementation also began at 24 weeks' gestation, so the findings do not address earlier prenatal exposure.
In the Discussion, the researchers wrote that the findings, together with existing evidence, support recommendations for increasing routine antenatal vitamin D dosing, while noting that the cognitive analysis was post hoc, non-prespecified, and limited to modest associations in memory measures.
The authors noted that the cohort's relatively high baseline vitamin D levels may have constrained the observed effects, which "may lead to underestimation of associations in more deficient populations," wrote lead study author Olivia Frigast Frederiksen, BSc, of Copenhagen Prospective Studies on Asthma in Childhood, Herlev and Gentofte Hospital, University of Copenhagen, and colleagues.
Disclosures can be found in the study.
Source: JAMA Network Open
