Dr Donnelly suggests that breast-feeding might contribute a protective effect against development of childhood cancer. The scope of the Editorial1 precluded an exhaustive critical discussion of a number of practices, including breast-feeding, that might prevent pediatric neoplastic diseases.
That breast-feeding per se decreases the risk for all childhood cancers is not firmly established in the literature. Dr Donnelly cites two epidemiologic studies that support this beneficial effect, but these retrospective reports were weakened by small sample size, limited geographical area, and questions about maternal recall bias regarding duration and exclusiveness of breast-feeding.2,3 In a study to evaluate a general-purpose epidemiologic questionnaire of possible environmental factors in development of childhood cancer, Schwartzbaum et al4 found an association (P=.04) between duration of breast-feeding and risk for development of childhood Hodgkin's disease; however, this study also pointed out the limited value of these types of survey instruments for identification of