Associations of serum vitamin D concentrations with obstetric glucose metabolism in a subset of the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study cohort

Abstract

To assess associations between maternal serum vitamin D concentration and glucose metabolism in a cohort of pregnant women living in an Australian subtropical environment. Cross-sectional assessment of 25-hydroxy vitamin D concentrations in 399 Hyperglycemia and Adverse Pregnancy Outcome ancillary study participants, treated at an obstetric teaching hospital in Brisbane, Australia. All patients underwent a blinded 75-g oral glucose tolerance test at 24-32 (target 28) weeks' gestation. The mean (± standard deviation) fasting plasma glucose was 4.5 ± 0.4 mmol/l. Mean (± standard deviation) serum 25-hydroxy vitamin D was 132.5 ± 44.0 nmol/l. A difference of one standard deviation in maternal 25-hydroxy vitamin D was inversely related to fasting glucose (fasting glucose lower by 0.047 mmol/l, P=0.012) when assessed with multiple linear regression after adjusting for confounders. Maternal 25-hydroxy vitamin D correlated with β-cell function as estimated by the log-transformed homeostasis model assessment-β-cell function equation (r=0.131, P=0.009), but not with the homeostasis model assessment of insulin resistance. An association between mid-gestational 25-hydroxy vitamin D and fasting glucose was confirmed in a largely normoglycaemic and vitamin D-replete pregnant population. The correlation between 25-hydroxy vitamin D and β-cell function suggests that vitamin D may influence glucose metabolism through this mechanism. Intervention studies are required to determine causality and the role of vitamin D replacement in deficient individuals.