Enteral High-Dose Docosahexaenoic Acid and Neurodevelopment in Extremely Preterm Infants: A Systematic Review and Meta-analysis

Background: Enteral high-dose docosahexaenoic acid (DHA) may be required for neurodevelopment, including cognition, of extremely preterm infants. High-level summative evidence is lacking. Objectives: This study aims to examine associations between enteral high-dose DHA during the neonatal period and neurodevelopment in infants born ≤29 wk of gestation. Methods: The following databases were searched (from inception to 11 April, 2024): CINAHL, Cochrane Library, Embase, Medline, Scopus, and Web of Science. Eligible randomized controlled trials (RCTs) in infants born ≤29 wk, assessing direct enteral administration ≥ 40 mg/kg/d DHA, or breast milk/formula with DHA ≥ 0.60% total fatty acids, reporting neurodevelopmental outcomes. Two reviewers independently screened articles, extracted data, and assessed quality using the Cochrane Handbook guidance. Data were pooled using fixed or random-effect meta-analyses. The primary outcome was global cognitive scores from a standardized test. Results: We screened 1978 articles and included 3 high-quality RCTs (2028 infants born ≤29 wk). Enteral high-dose DHA was not associated with overall differences in global cognition scores at a corrected age (CA) of 18–36 mo [3 RCTs, 638 children, mean difference (MD) 0.67; 95% confidence interval (CI): –1.80, 3.15; P = 0.59; I² = 0%] or CA of 5–7 y (2 RCTs, 852 children; MD: 2.22; 95% CI: –0.14, 4.57; P = 0.06; I² = 33%); however, benefit was observed in the largest RCT with a direct enteral emulsion (656 children, CA of 5 y, MD 3.45; 95% CI: 0.38, 6.52; P = 0.03). Associations with most secondary outcomes were not seen; however, high-dose DHA was associated with reduced mild motor (3 RCTs, CA of 18–36 mo) and cognitive (2 RCTs, CA of 5–7 y) impairment. No negative impacts were observed. Conclusions: Enteral high-dose DHA in extremely preterm infants was not associated with differences in global cognition scores on meta-analysis; however, higher scores were observed with the use of a direct emulsion. Results support contemporary recommendations. This trial was registered at PROSPERO as CRD42022382744 (https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022382744).