(P27-017-25) Predictors of Energy Density in Preterm Human Milk: A Systematic Review and Meta-Analysis

Objectives: The purpose of this analysis was to evaluate the extent to which gestational age (GA), postpartum day (PD), and analytical methods predicted energy density in preterm human milk (HM).

Methods: These analyses are a part of a larger systematic review and meta-analysis to evaluate preterm HM nutrient composition. A research librarian reviewed Ebsco, PubMed, and Scopus through July 2023 and two independent reviewers evaluated titles, abstracts, and full text articles for inclusion criteria: studies published in English and that reported HM composition from mothers who delivered preterm. Samples from countries with a Human Development Index < 0.8 and studies with a sample size of one were excluded from analysis.  Meta-regressions tested the effects of GA at delivery (wk), PD at HM collection, and the interaction between the effects. A separate model was fit to explore differences across analytical methods (bomb calorimetry, mid-infrared spectroscopy, calculations from macronutrients, and other/not specified). Studies that reported multiple time points were aggregated before fitting the random effects meta-analyses for the analytical method analyses only.

Results: Of the 48 studies that reported energy density, 30 were included in this meta-analysis. Among all HM samples, the effects of GA and PD were dependently associated with energy density. This association between PD and energy density was attenuated as GA increased (GA = 0.25(0.006), p < 0.001; PD= -0.82(0.003), p < 0.001; and GA×PD= -0.026(0.001) p < 0.001). The magnitudes of the associations varied by milk type. The energy density estimates of all HM types combined differed across methodology. Bomb calorimetry was 71.62 kcal/100mL (95% CI: 68.33, 74.92)[Q=38.00, df=9, p< 0.001; I2=70.5%] mid-infrared spectroscopy was 65.11 kcal/100mL (95% CI: 61.62, 68.61) [Q=75.86, df=8, p< 0.001; I2=89.8%], calculated was 68.48 kcal/100mL (95% CI: 61.23, 75.74) [Q=62.96, df=7, p< 0.001; I2=88.9%], and other/not specified was 68.55 kcal/100 mL (95%CI: 64.20, 72.89) [Q=4.05, df=2, p=0.132; I2=54.3%).

Conclusions: These data reflect a positive association between PD and energy density of preterm HM for GA < 30 weeks but null to inverse associations at GA ≥ 30. These results underscore the need for better standardization across methodologies that measure HM energy density.

Funding Sources: Mead Johnson Nutrition