Early Nutrition – First, Do No Harm

In modern practice, risk of adverse effects is an expected consequence of necessary clinical interventions – so the dictum of Hippocrates could be changed in a modern context to reflect the need to understand, quantify, and minimize risk (harm) that inevitably accompanies interventions. However, in very sensitive periods of development, as we encounter in neonatology, adverse effects of interventions may be complex, multiple, and difficult to discern as causal against a background of diverse unrelated pathologies seen in sick neonates; hence the particular importance of evidence-based medicine in this area.

The example dissected in this lecture is the use of cow’s milk–based products in neonatal care for the smallest preterm infants. Since human milk alone does not meet nutritional needs for the rapidly growing preterm infant, cow’s milk–based formulas and fortifiers were introduced in the 1970s and 1980s. The objective in using these products is to improve growth and neurodevelopment; however, my focus is on the characterization and quantification of risk of these products as a basis for optimizing future nutritional care and achieving quality improvement.

I consider key methodological issues, since the value of individual publications and meta-analyses has been by affected by flawed dietary definition. In all, 15 significant adverse neonatal or long-term programmed clinical outcomes are identified in randomized trials in relation to the use of cow’s milk products in neonatal care. A number of these outcome effects are further established in our new meta-analyses of clinical trials, presented in my lecture. Further adverse outcomes have been identified in cohort studies and meta-analyses of these studies. This suggests a broad impact of diet on the biology of the organism. However, my talk focuses principally on the neonatal period, notably on necrotizing enterocolitis and proven sepsis (noting the secondary impact of these two latter conditions on cerebral palsy risk and later IQ), mortality, severe retinopathy of prematurity (ROP), chronic lung disease, and patent ductus arteriosus, where there is evidence for major and significant adverse effects with important implications for practice.

In current practice there is now less use of preterm formulas, with more units electing to achieve a 100% human milk–based diet with mother’s own plus donor breast milk. The question is whether there is evidence that a cow’s milk–based fortifier used in this circumstance would have adverse effects. Studies were identified where the base diet was all human milk and where it was possible to compare a cow’s milk–based fortifier with a human milk–based fortifier used as the control. Two new analyses are included. Our finding from the totality of evidence available to date is that cow’s milk–based fortifier had significant adverse effects for a number of neonatal outcomes. One possible reason for this is the fact that even when the base diet is 100% human milk, fortifiers provide an unexpectedly high proportion of total protein intake.

In summary, our data indicate that cow’s milk–based products, when used in the smallest preterm infants, are associated with a major increase in some of the most serious outcomes in neonatal care. These findings have important implications for explorations aimed at underpinning future quality improvement.