To help shed some light on the 50-year-old controversy concerning the possible protective effect of breast feeding on subsequent atopic disease, I developed 12 standards pertaining to both biologic and methodologic aspects of exposure (infant feeding), outcome (atopic conditions), and statistical analysis for studies of atopic eczema, asthma, allergic rhinitis, cow milk allergy, and other food allergy. Among the published studies on atopic eczema, the nine claiming a protective benefit of breast feeding performed less well than the 12 not making such a claim on "methodologic" standards relating to strict diagnostic criteria and blind ascertainment of outcome. The positive studies were somewhat stronger, however, on the "biologic" standards bearing on sufficient duration and exclusivity of breast feeding and on separate analysis of children at high risk. For the other atopic conditions, there were no important differences between positive and negative studies. In few negative or positive studies was there adequate control for confounding variables or examination of potential benefits relating to the severity or age at onset of atopic disease. To avoid another 50 years of unresolved controversy, future studies should improve both the biologic and methodologic aspects of their design and analysis.
PIP: In the last 50 years many studies have been published that purportedly prove or refute the hypothesis that breast feeding protects infants against 6 immediate hypersensitivity-mediated disease states -- atopic eczema, asthma, allergic rhinitis, cow milk allergy, other food allergy, and combinations of these. A MEDLINE search for articles published between 1983 and 1986 and an examination of end references back to 1936 provided a supply of articles on both sides of the debate. These articles were rated against 12 biological and methodological standards to determine whether their conclusions were based on rigorous examination of the facts. The 12 standards were: nonreliance on recall, blind ascertainment of infant feeding history, breast feeding duration of at least 2 months, specifically stated exclusivity of breast feeding, strict diagnostic criteria for the atopic diseases, blind ascertainment of outcome, measured severity of disease outcome, age at onset of disease, control for confounding variables, assessment of dose-response effect, assessment of genetic of risk factors for atopic disease, and rigorous statistical methods. The total retrieval included 22 studies of atopic eczema, 13 studies of asthma, 7 studies of rhinitis, 3 studies of cow milk allergy, 4 studies of other food allergy, and 8 studies of mixtures of 2 or more atopic diseases. These studies were rated against the 12 standards. Serious flaws reduced the value of all studies in greater or lesser degree. Nonblind ascertainment of outcome, failure to control for confounding variables, failure to ascertain disease severity, and failure to record age at onset of disease were the most common drawbacks. It is not possible to determine from these studies whether or not breast feeding has a prophylactic effect on atopic disease. More research is obviously necessary, but a rigorous study design must come first.