Optimal cord blood vitamin D concentration does not protect against type 1 diabetes: a longitudinal cohort study

Abstract Calcidiol (25-hydroxy vitamin D [25(OH)D]) has a putative role in protection against auto-immune diseases, including type 1 diabetes (T1D). However, contradictory findings have been reported. We compared concentrations of the 25(OH)D in cord-blood in 83 children who later developed T1D and compared with 296 randomly selected healthy controls from the ABIS (All Babies in Southeast Sweden) general population birth cohort. We used the weighted Cox proportional hazard models to estimate the hazard ratios (HRs) for T1D. Models were adjusted for T1D heredity, smoking during pregnancy, fish-intake during pregnancy, parity, delivery-mode, vitamin supplement intake during pregnancy, sex, season of birth, birth weight, and preterm-status. No association between concentration of 25(OH)D and T1D was found [per 10 nmol/L increase HR (95% CI): 1.10 (0.96-1.26)]. The robustness of lack of association between 25(OH)D and T1D is confirmed using stratified models for Season [Autumn: 0.72 (0.30-1.75); Spring:0.74 (0.51-1.07); Summer: 1.27 (0.85-1.89); Winter: 1.44 (0.92-2.23)], Sex [Male:1.17 (0.94-1.44); Female: 0.86 (0.68-1.09)] and supplementary vitamin intake during pregnancy [Yes: 1.03 (0.83-1.26); No: 1.24 (0.96-1.61)]. The absence of significant association of 25(OH)D and T1D was consistent across age of disease onset. In conclusion, optimal cord blood 25(OH)D concentration does not seem to protect against T1D.