Abstract
BACKGROUND: Vitamin D inadequacy is globally prevalent among pregnant women, but its impact on offspring cardiometabolic risks remains inconclusive. This study aimed to evaluate associations between in utero vitamin D status and childhood cardiometabolic risk factors. METHODS: In the original Hyperglycaemia and Adverse Pregnancy Outcome (HAPO) Study, all pregnant women booked for antenatal care, except teenage pregnancy, plan to delivery at another hospital, uncertain last menstrual period and no ultrasonographic estimated gestational age, inability to complete the oral glucose tolerance test (OGTT), multiple pregnancy, conception by assisted reproduction, glucose testing before recruitment, or diagnosis of diabetes during or before the current pregnancy and required medication. Archived maternal serum samples from 24 to 32 gestational weeks and umbilical cord serum samples at birth in the original HAPO Study at the Hong Kong Centre were assayed for total 25-hydroxyvitamin D [25(OH)D] levels by using liquid chromatography-tandem mass spectroscopy. Children's clinical and biochemistry parameters were collected at the HAPO Follow-Up Study at around age 7, including anthropometry, blood pressure (BP), carotid-femoral pulse wave velocity (cfPWV), fasting lipid profile, plasma glucose (PG), and insulin levels at OGTT. Linear regression analyses were used to evaluate the associations of in-utero 25(OH)D levels with children's cardiometabolic risk factors. FINDINGS: There was no association between maternal serum total 25(OH)D level in 24-32 weeks of gestation and childhood cardiometabolic risk factors. In contrast, umbilical cord serum total 25(OH)D level at birth was negatively associated with offspring diastolic BP, cfPWV, and fasting PG at around age 7, whereas no significant results were observed in other cardiometabolic risk factors. Every 1-SD (18.0 nmol/L) increase in the umbilical cord serum total 25(OH)D level was independently associated with a reduction in diastolic BP, cfPWV, and fasting PG by 0.73 mmHg (95% CI [-1.44, -0.02], p = 0.044), 0.08 m/s ([-0.14, -0.03], p = 0.004), and 0.04 mmol/L ([-0.07, -0.01], p = 0.017), respectively. However, only associations with cfPWV and fasting PG remained significant after correction for multiple testing. INTERPRETATION: A low umbilical cord serum 25(OH)D level at birth, but not maternal serum 25(OH)D in 24-32 weeks of gestation, was associated with higher childhood diastolic BP, arterial stiffness, and fasting PG levels. Although maternal total 25(OH)D levels between 24 and 32 gestational weeks were not associated with childhood cardiometabolic risks, our findings indicated that neonatal vitamin D status at birth may be relevant to childhood vascular and glucose metabolic health. Given the observational study design and modest effect size, these findings should be interpreted cautiously and warrant further investigations.