Breast-feeding, maternal IgE, and total serum IgE in childhood

Background: There is controversy regarding the relationship of the effect of breast-feeding on markers of allergy such as total serum IgE in childhood. Objective: This study, using longitudinal data, tested the hypothesis that the relation of breast-feeding to IgE in childhood differs depending on maternal total IgE level, Methods: Total serum IgE was assessed with the paper radioimmunosorbent test at 4 ages in nonselected children enrolled at birth into the prospective Tucson Children's Respiratory Study. Children were classified as never breast-fed, breast-fed less than 4 months, or breast-fed 4 months or longer, on the basis of physician report or questionnaires completed by parents by the time the child was 18 months old. A longitudinal random effects model was used to test for group differences and temporal trends in IgE for children classified,vith reference to maternal IgE (high tertile vs all others) and breast-feeding history, A total of 664 children with 1457 observations were included. Results: Among children whose mothers were in the 2 lower tertiles of IgE, breast-feeding was associated with lower total serum IgE at age 6 years (24.2 vs 44.3 IU/mL for never breastfed children; P < .02); similar trends existed at age 11 years. In contrast, for children whose mothers were in the highest tertile of IgE, breast feeding of 4 months or longer was associated with higher IgE levels in the child compared with those never breast-fed or breast-fed less than 4 months (97.0 vs 38.9 IU/mL; P < .005). These cross-sectional analyses were confirmed with the longitudinal random effects model, which also showed no effect of confounders. Paternal IgE showed no similar relation with child IgE. Conclusion: Breast feeding appears to have paradoxic relations with IgE in childhood, depending on maternal IgE level. These findings may help explain the contradictory results found in other investigations of the relation of breast-feeding to allergic symptoms and markers.