Objectives: Impact of the proportion of human milk (HM) in mixed feeding on necrotizing enterocolitis (NEC) remains unknown. This study explores the influence of different proportions of HM on the risk of NEC. Materials and Methods: A retrospective cohort study was performed in infants with very low birth weight (VLBW). A spline smoothing curve was used to evaluate the dose-dependent association between HM and the risk of NEC. Univariate and multivariate analyses were performed to detect the association between the proportion of HM and NEC. Results: Twenty-four infants developed NEC, with 4 (1.9%) in the high HM group, 18 (28.1%) in the low HM group, and 2 (8.0%) in the exclusive formula group (p < 0.001). After adjusting for the relevant confounders, low HM (proportion of HM ≤54%) (OR 33.526, 95% confidential interval [CI] 7.183-156.475, p < 0.001) and exclusive formula feeding (OR 8.493, 95% CI 1.107-65.187, p = 0.040) significantly increased the incidence of NEC, compared with the high HM feeding (proportion of HM >54%). Similarly, low HM was independently associated with an increased risk of feeding intolerance compared with high HM feeding (OR 4.383, 95% CI 2.243-8.564, p < 0.001). Conclusion: A low ratio of HM (≤54%) significantly increased the risk of intestinal complications in VLBW infants. Mixed feeding should relate to the proportion of HM in premature infants.
Keywords: feeding intolerance; human milk; necrotizing enterocolitis; proportion.