Do breast-feeding and delayed introduction of solid foods protect against subsequent atopic eczema?

Previous studies relating infant feeding to subsequent atopic eczema have had methodologic flaws that include insensitive study designs, nonblind observation, and failure to control for confounding variables. To avoid these flaws, we conducted a case-control study of 636 patients attending a dermatology clinic. The dermatologic problem in each patient was classified as atopic eczema (case), dermatologic condition unrelated to atopy (control), or dermatologic condition with questionable relation to atopy (uncertain). The feeding history was ascertained later, along with family history and demographic data, by an assistant blind both to the question under study and to the case vs control status of each subject. Breast-feeding was not associated with any reduction in the estimated relative risk of developing atopic eczema. No significant relationship was found among the cases between severity of disease and breast-feeding nor between age of onset of disease and duration of breast-feeding or age at introduction of solid foods. Even when breast-feeding was redefined as "pure" and exclusive (no nonhuman milk or solids) for greater than or equal to 2 months, no protective effect was uncovered. We conclude that breast-feeding and delayed introduction of solids do not protect against atopic eczema, and that previous claims of protective effects were based on data probably biased by nonblinding and important confounders.