LAUSR

Carome et al. report that exclusive human milk diet was associated with lower incidence of severe intraventricular hemorrhage (IVH)/periventricular leukomalacia (PVL) in extremely-low-birth-weight infants [1]. The differences in rates of IVH or PVL alone were not significant in the two groups. Authors provide an excellent summary of various mechanisms by which a human milk diet could offer a neuroprotective effect in these infants. These mechanisms could certainly play a role in prevention of PVL and will need further investigation. However, it appears unlikely that short exposure to small dose of human milk can reduce the incidence of severe IVH. Several studies have demonstrated that nearly all IVH in preterm infants occur within first 72 h of birth and a significant proportion of these occur as early as within first 6 h of birth [2–5]. Authors state that first enteral feeds occurred on a median age of 2 days and since infants at risk of IVH are usually most immature and unstable in first few days, it is quite unlikely that these infants can receive any meaningful amount of human milk during the period of highest risk of IVH. My other concern and question is related to the study population. It appears that the majority of infants in EHM group were born in 2015–2017, while most infants in nonEHM groups were born prior to that. This difference can have significant effect on their findings as the observed decrease in the rate of IVH/PVL could be related to other practice changes over time besides increase in use of human milk. I would also like the authors to provide information on use of practices reported to have an impact on incidence of IVH such as prophylactic indomethacin, elevated supine midline head positioning for first 72 h of life, and delayed cord clamping. Statistical analyses to control for these variables is important to exclude the possibility of a type I error.