Low vitamin D levels have been reported in multiple immune disorders such as type 1 diabetes mellitus (T1DM). The purpose of our study was to determine vitamin D levels in children at the onset of T1DM compared with children with other diseases and to test the hypothesis that low vitamin D may increase the odds for developing diabetes. All the children (n = 58) that were consecutively admitted to our clinic at T1DM onset between May 2010 and July 2012 were compared with a control group of children (n = 166) hospitalized for other diseases, matched for sex, season of visit, and age. For each subject, we considered clinical and anthropometric data, the season at time of hospitalization, and serum 25-hydroxyvitamin D (25(OH)D), which were analyzed and compared using multivariable conditional logistic regression. Median 25(OH)D was significantly lower in the diabetic patients (36.2 nmol/l, range = 7.5-121.0 nmol/l) than in controls (48.7 nmol/l, range = 7.5-190.2 nmol/l), p = 0.010. Low 25(OH)D levels seem to increase the odds for developing T1DM (odds ratio (OR) = 3.45 for 25(OH)D 51-74 nmol/l, OR = 5.56 for 25(OH)D ≤ 50 nmol/l). There was no seasonal effect on the risk of developing T1DM. Median 25(OH)D level was significantly lower in patients admitted with diabetic ketoacidosis (30.2 nmol/l, range = 7.5-101.8 nmol/l) than in patients without ketoacidosis (40.7 nmol/l, range = 15.2-121.1 nmol/l), p = 0.019; but when adjusted for season, the p value was 0.116.
Conclusions: Children at onset of T1DM have lower vitamin D serum levels than those with other diseases. Further longitudinal studies on children before the onset of T1DM will allow clinicians to explore the causal relationship between vitamin D and T1DM.