OBJECTIVE To further investigate the association between the type of feeding in infancy and the development of IDDM.

RESEARCH DESIGN AND METHODS We have carried out a case-control study in the area of Sassari (northern Sardinia, Italy), which is characterized by an ethnically homogenous population at high risk of IDDM. The study subjects comprised 100 IDDM patients and 100 control subjects, matched for sex and age and selected from children admitted at the Department of Pediatrics of the University of Sassari. Diabetic children (53 boys, 47 girls) had been diagnosed between 1983 and 1994, and their age at diagnosis ranged between 1 and 15 years. Information on feeding patterns during the 1st year of life was collected through questionnaires administered to the mothers. The questionnaire was designed to evaluate the duration of complete or partial breast-feeding and the age at which dietary products containing cow's milk were introduced into the diet.

RESULTS A larger proportion of the diabetic children rather than the control children had been breast-fed, and the risk of IDDM among children who had not been breast-fed was below unity (odds ratio [OR] 0.41; 95% CI 0.19–0.91). No clear difference was observed between diabetic and control subjects in the duration of breast-feeding (medians: 3 and 2 months, respectively), even if, overall, the data suggested a slight increase in the risk of IDDM with longer duration of breast-feeding (OR 1.10; 95% CI 0.99–1.22 per month). Although a larger proportion of control children rather than diabetic children had been given cow's milk–derived formula and solid food before the age of 3 months, there was no time-risk relationship.

CONCLUSIONS Our data do not support the existence of a protective effect of breast-feeding on the risk of IDDM, nor do the data indicate that early exposure to cow's milk and dairy products has any influence on the development of IDDM in a high-risk population.