INFANT DIET AND SALMONELLOSIS

Background: The Island of Guam has experienced a high incidence of infant salmonellosis for a number of years (age-specific incidence rate of approximately 3700 cases per 100 000 infants in 1984). Interviews of case parents suggested that the use of high-iron infant formula was associated with this problem. Methods: A case-control study was conducted to test this hypothesis. Information on feeding practices and a variety of medical and socio-economic factors was collected for 78 laboratory-confirmed Salmonella-case infants and 167 control infants. Results: Case infants were less likely to have been breast-fed (odds ratio [OR] = 9.15; 95% confidence interval [CI] = 2.71-30.9) and more likely to have been fed infant formula with an iron content of 10 mg/L or greater (OR = 2.96; 95% CI = 1.24-7.08) than were control infants. Conclusions: Although the precise means by which infants are most commonly exposed to Salmonella bacteria remain obscure, breast-feeding apparently protects against the development of physician-diagnosed clinical salmonellosis, while the feeding of high-iron infant formula has the opposite effect.