Neonatal vitamin D status is not associated with later risk of type 1 diabetes: results from two large Danish population-based studies

Two nationwide Danish population-based observational studies: a case-cohort design (912 T1D cases and 2866 non-cases; births 1981–2002; follow-up through 2012) and a nested case-control design (527 matched pairs; births 1981–1999; follow-up through May 2004). Neonatal 25(OH)D levels were measured in dried blood spots by LC-MS, focusing on 25(OH)D3 (25(OH)D2 largely below the lower limit of quantification). Analyses used weighted Cox regression for the case-cohort study and conditional logistic regression for the case-control study. Data sources included the Danish Civil Registration System, DanDiabKids, and the Danish Newborn Screening Biobank; no dosing information or pregnancy-level interventions were involved.

No association between neonatal 25(OH)D3 levels at birth and risk of type 1 diabetes by age 18 in either study. Case-control: per-25 nmol/L increase in 25(OH)D3 gave OR 1.02 (0.80–1.28) without adjusting for HLA risk and OR 0.86 (0.60–1.22) with HLA risk; both non-significant. Case-cohort: no trend across 25(OH)D3 quintiles. Conclusion: birth-time 25(OH)D3 levels do not predict late-onset type 1 diabetes risk. Authors note that higher maternal 25(OH)D3 during pregnancy—potentially via higher-dose supplementation than currently recommended—could be protective, but this possibility cannot be ruled out and requires further study.

Limitations include 25(OH)D2 values largely below the lower limit of quantification, necessitating focus on 25(OH)D3; neonatal DBS measurements may not capture broader prenatal vitamin D exposure; DBS LC-MS calibration and lack of a universal reference standard introduce measurement variation (estimated total assay variation ~15%); not all DBS cards were available, and some analyses were excluded; in case-control, adjustment for non-HLA genetic susceptibility was not possible; observational design in a Danish population limits generalizability; there was overlap of some cases between the two study designs, which could affect independence of results.