Duration of breast feeding and bovine serum albumin antibody levels in type 1 diabetes: a case-control study.

To compare the levels of bovine serum albumin (BSA) antibodies and their relationship with duration of breast feeding, age of exposure to cow's milk, and human leukocyte antigen (HLA-DQ) genotype in children with and without type 1 diabetes. Serum samples from 143 (0.3-14.7 yr) newly diagnosed children with type 1 diabetes and 107 unrelated control children (0.8-13.5 yr) were evaluated for BSA antibodies. Duration of breast feeding and exposure to cow's milk were recorded on questionnaires. HLA-DQ typing was determined by polymerase chain reaction. One hundred percent of the diabetic children were positive for BSA antibodies compared to 1.9% for healthy controls (p < 0.001). Diabetic children also had higher levels of immunoglobulin G antibodies than unrelated controls (55.1 vs. 17.8 ng/mL, p < 0.0001). Duration of breast feeding (5.4 vs. 7.6 months, p < 0.02), but not age of exposure to cow's milk (8.3 vs. 9.2 months, p = 0.11), differed between cases and controls. There was no difference in antibody titer by duration of breast feeding or age of exposure to cow's milk in the cases or controls. Higher levels of antibodies to BSA were found in children recently diagnosed with type 1 diabetes compared to the controls, particularly those with high or moderate HLA-DQ genotypes. The BSA profile, however, does not seem to depend on duration of breast feeding or age of exposure to cow's milk in this population.