Early infant diet and risk of IDDM in blacks and whites. A matched case-control study

To investigate the role of early infant feeding in the development of insulin-dependent diabetes mellitus (IDDM) and to determine whether an association exists in both blacks and whites.

Black and white diabetic subjects were recruited from the Allegheny County and Children's Hospital of Pittsburgh IDDM Registries. Extensive infant diet histories were obtained from the diabetic subjects and their nondiabetic siblings, who were used as nondiabetic control subjects. Each diabetic subject was matched outside his/her family to an unrelated nondiabetic control subject on birth order, birth year (+/- 2 yr), and race, which resulted in 211 case-control pairs with a mean birth year of 1967.

In whites, diabetic subjects were less likely to have been breast-fed than control subjects (odds ratio [OR] 0.5, 95% confidence interval [CI] 0.3, 0.9). Breast-feeding prevalence did not differ between black diabetic subjects and control subjects. Duration of overall and exclusive breast-feeding did not differ between diabetic and control subjects in the black and white cohorts. The following analyses, which examined whether the timing of the first breast milk substitute to which the infant was exposed differed between diabetic and control subjects, were conducted for exposure to any breast milk substitute and to breast milk substitutes that were cow's milk based. In whites, age at exposure to any breast milk substitutes and cow's milk-based substitutes were similar between diabetic and control subjects. In blacks, the first exposure to breast milk substitutes occurred significantly earlier for any substitute (5.1 vs. 11.9 wk, P = 0.02) and marginally earlier for cow's milk-based substitutes (3.9 vs. 8.5 wk, P = 0.07) in diabetic subjects compared with control subjects. The first exposure to breast milk substitutes was more likely to occur by 3 mo of age in black diabetic subjects compared with black control subjects (OR 3.3, 95% CI 1.1-10.0) after adjusting for maternal age at birth. The addition of breast-feeding status to the model only slightly weakened this association in blacks.

The analyses of this study cohort suggest that the observed protective effect of breast-feeding on the risk of IDDM may be related to differences in the age at exposure to breast milk substitutes in blacks but not in whites.