Proportional hazards analysis of diarrhea due to enterotoxigenic Escherichia coli and breast feeding in a cohort of urban Mexican children.

Ninety-eight women-infant pairs were followed for up to 50 weeks in the northern part of Guadalajara, Mexico, from August 1986 to July 1987 as part of a community-based, prospective study of the relation between infant feeding patterns and enterotoxigenic Escherichia coli producing heat-labile toxin (LT-ETEC) diarrheal disease. Strictly formula-fed children had an incidence of diarrhea over three times that of strictly breast-fed infants and twice that of breast-fed and supplementally fed children. Strictly formula-fed infants colonized by LT-ETEC were symptomatic for diarrhea nearly three times as often as strictly breast-fed infants and twice as often as infants receiving a mixed diet. The fitting of parametric hazard models to durations until LT-ETEC colonization revealed that the hazard for the first colonization was time invariant. The hazard of diarrhea increased by 400-500% during the rainy season or among children 3 months of age or older who received avena, a barley drink. The best-fitting hazard models to durations until symptomatic expression of LT-ETEC infection all increased through time. This hazard was inversely impacted by the overall amount of LT-ETEC-specific, immunoglobulin A antibodies the infant received via the mother's breast milk and by the provision of traditional medicinal teas.