SIgA, secretory IgA EID, electroimmunodiffusion ELISA, enzyme-linked immunosorbent assayThe decision to feed human milk to VLBW infants often is based on the potential enhancement of host defense. Two approaches have been used in the few studies d~signed to e.valuate the protection afforded by feeding human milk to these mfants: 1) assessment of morbidity and 2) the J?easuremen~of the concentrations of specific immune factors m serum, saliva, and feces of infants fed either human milk or cow's milk-based formulas (1-4). Most epidemiologic studies of morbidity in fullterm infants usually have concluded that breast-feeding is protective but interpretations of these reports often have been confo~nded by uncontrolled demographic variables (5). ",:ssessments of morbidity in VLBW infants have focused on sepSIS and necrotizing enterocolitis. No consistent benefits have been reported (6-8). Concentrations of immuno~obulinsin. the serum were similar in both groups or elevated m VLBW mfants fed cow's milk preparations (9-11).The effects of human milk feeding on fecal excretion of lactoferrin, lysozyme, and SIgA have been reporte~(~-3). Fecal levels of these selected immune factors are greater m mfants fed human milk than in those fed a formula (1-3). In that regard, these selected factors also resist proteolytic digestion (12, 13). These observations suggest that the factors remain intact throughout the infant's gastrointestinal tract. The few studies of immune factors in the feces of VLBW infants, however, have been of short duration and have not been designed to evaluate the effects of the quantity of immune factors fed, variations in nutrient intakes that would influence growth and maturation (1, 2), or intrinsic differences in the characte~stics ofth~population studied such as gastrointestinal absorptIve capaCIty. Furthermore the impact of augmented levels of lactoferrin, lysozyme, and SIgA in milk obtained from women delivering. prema~u~ely has not been considered in the design of most published climcal evaluations (14, 15). Ideally, a study of the effects of human milk on premature infants should be carried out wit~breast-fed infants. This study condition is precluded in VLBW mfants because of their difficulties in suckling and their nutritional needs that may not be met by the nutrient concentrations of human milk (16). " Our investigations of the effects of feedmg human mIlk to VLBW infants began with detailed assessments of the levels and stability of selected nutritional and immune factors during the collection, storage, heat processing, and fractionation of human milk (17-20). A protocol to fortify human milk was developed 711 Abbreviations ABSTRACT. The amounts of lactoferrin, lysozyme, total IgA, secretory IgA (SIgA), and specific SIgA antibodies to a pool of Escherichia coli 0 antigens were measured lD 96-h collections of feces obtained from 28 very low birth weight infants, 28-30 wk of gestation, studied at 2.5 and 6 wk of age. Eighteen of these infants were fed their mot~ers' milk fortified with fractions ...