The Association between Reported Dietary Intake of Vitamin D and Cow’s Milk Proteins in the Development of Islet Autoimmunity (IA): The Diabetes Autoimmunity Study in the Young (DAISY)
Abstract Number: 1006-P
Authors: MELISSA SIMPSON, HEATHER BRADY, XIANG YIN, KATHY BARRIGA, MARIAN REWERS, JILL NORRIS, Aurora, CO
Institutions: Aurora, CO
Results: Vitamin D is hypothesized to be protective in type 1 diabetes (T1D) whereas cow’s milk protein is hypothesized to be diabetogenic. In the United States, dairy manufacturers are required to fortify fluid milk products with vitamin D, thus milk contains two factors with potentially opposite effects on diabetes risk. We investigated the association between vitamin D, protein from cow’s milk sources and the time to development of islet autoimmunity (IA) in children in DAISY.
DAISY follows children who either possess a higher risk HLA genotype or are a first degree relative of someone with type 1 diabetes for development of IA, which is defined as being positive for insulin, glutamic acid decarboxylase or IA-2 autoantibodies on at least two consecutive occasions and still positive or diabetic at last follow-up. Data on vitamin D and cow’s milk protein intake were collected annually starting at the age of 1, using a food frequency questionnaire. The reported data on vitamin D intake was associated with plasma 25,OH vitamin D (p=0.04) in a subcohort of 257 DAISY children, adjusting for season of blood draw and ethnicity. Of the 1785 children with complete dietary data, 57 children developed IA, after a mean follow-up of 4.6 years. Hazard ratios (HR) and 95% confidence intervals (CI) were calculated based on a standard deviation difference in nutrient intake. Nutrients were treated as time-varying to reflect changing intake during childhood. The vitamin D variable includes that from foods, fortification and supplements. The cow’s milk protein variable represents the sum of the protein (grams) from cow’s milk in foods and milk. We did not have a quantitative measure of intake during infancy; therefore we did not assess risk due to infant exposures.
After adjusting for caloric intake, intake of polyunsaturated fatty acids, having a first degree relative with T1D, and HLA-DR3/4 status, we found that higher cow’s milk protein intake (HR:1.37 (CI: 1.07 - 1.74) and lower vitamin D intake (HR: 0.71, CI: 0.50 - 0.99) are significantly associated with the risk for IA. Neither cow’s milk protein nor vitamin D was associated with IA without the other in the model, suggesting that accounting for one is necessary to see the effect of the other.
These results suggest that cow’s milk protein and vitamin D are associated with IA, albeit in opposite directions, alluding to the complexity of the role that cow’s milk may have in IA risk.