Enteral Feeding During Therapeutic Hypothermia for Hypoxic-Ischemic Encephalopathy in Neonates: A Multicentre Study from Italy and Belgium.

Background

Enteral feeding during therapeutic hypothermia (TH) for hypoxic-ischemic encephalopathy (HIE) remains heterogeneous across centres because of concerns regarding feeding intolerance and the risk of necrotizing enterocolitis (NEC). Evidence regarding optimal feeding initiation and advancement remains limited.

Methods

We performed a secondary analysis of a multicentre retrospective cohort of infants with HIE treated with TH at two tertiary neonatal intensive care units in Italy and Belgium between 2011 and 2024. We included survivors with complete nutritional data, aware that this restriction, adopted because feeding-related endpoints cannot be assessed in non-survivors, may introduce survivor bias. Feeding practices, enteral feeding progression, feeding tolerance, and gastrointestinal outcomes were evaluated.

Results

Of 302 infants in the original cohort, 276 were included. Enteral feeding during TH was administered to 166 infants (60.1%). Enteral feeding volumes remained low during the first three days of life but progressively increased thereafter. Despite inter-center differences in feeding strategies (initiation and advancement), time to full enteral feeding (median 9 days; IQR 7-12) was comparable overall (Italy: median 9 days, IQR 7-12; Belgium: median 10 days, IQR 7-14; p = 0.323). Infants receiving enteral nutrition during TH were observed to achieve full enteral feeding earlier than those not fed during cooling (median 8.0 vs. 12.0 days, p < 0.001) and to have a shorter length of hospital stay (median 11.0 vs. 15.0 days, p < 0.001). Breastfeeding at discharge was also observed more frequently among infants who received enteral nutrition during TH (79.5% vs. 54.5%, p < 0.001). Feeding intolerance occurred in 47 infants (17.0%) and temporary feeding withdrawal in 31 (11.2%). Only three infants (1.1%) developed Bell stage IIA NEC, with no progression to more advanced disease or surgery. NEC was more frequent among infants with Sarnat stage III HIE than among those with stage I-II disease (8.3% vs. 0.4%, p = 0.022).

Conclusions

Enteral feeding during TH was feasible, and gastrointestinal complications were uncommon. These findings support the feasibility of cautious enteral feeding during TH. The concentration of NEC events among infants with severe HIE, although based on only three cases, is hypothesis-generating and suggests that future studies should evaluate feeding strategies according to both neurological severity and physiological stability. Whether earlier initiation or faster advancement is beneficial in clinically stable infants cannot be determined from these observational data and should be addressed in prospective studies.