In Reply : Breast-Feeding and Infectious Illness, Power and Confidence Intervals

The issues raised by Hopkinson in her letter in this issue are important questions. Hopkinson doubted that our findings are applicable to the United States, because the overall rate of initiating breast-feeding is different from Denmark. We suggested that our findings could be applied to middle class populations in the US because breast-feeding rates in Denmark are similar for this group, but we stated ". . . nor can they be extrapolated to rural or poor urban populations in developed nations." By not making a distinction between different groups within the US or other industrialized nations, Hopkinson unnecessarily limits the potential implications of our study.