Infant formulas with partially or extensively hydrolyzed milk proteins for the prevention of allergic diseases: A systematic review and meta-analysis of clinical trials

Despite the widely recommended usage of partially or extensively hydrolyzed milk protein formula (PHF or EHF) for preventing allergic diseases (ADs), clinical studies have been inconclusive regarding their efficacy compared with cow's milk formula (CMF) or breast milk (BM). We aimed to systematically evaluate the effects of PHF or EHF versus CMF or BM on the risk of ADs (cow's milk allergy, allergic rhinitis, eczema, asthma, wheeze, food allergy, and sensitization) in children. We searched PubMed, Embase, Cochrane Library, and Web of Science for clinical trials published from inception to October 21st, 2022. We utilized the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach to grade the strength of evidence. Overall, 24 trials (10950 infants) were included, 17 of which specifically included high-risk infants. GRADE was low for the evidence that compared with CMF, infants early-fed with EHF had a lower risk of cow’s milk allergy at age 0-2 years (Relative Risk [RR], 95%CI: 0.62,0.39-0.99). Moderate evidence supported that PHF and EHF reduced the risk of eczema in children under or over 2 years of age, respectively (RR: 0.71, 0.52-0.96 and 0.79, 0.67-0.94). We also identified moderate systematic evidence indicating that PHF reduced the risk of wheeze at age 0-2 years compared with CMF (RR: 0.50, 0.29-0.85), but PHF and EHF increased the risk compared with BM (RR: 1.61, 1.11-2.31 and 1.64, 1.26-2.14). Neither PHF nor EHF had significant effects on other ADs in children of any age. In conclusion, compared with CMF, PHF or EHF had different preventive effect on cow’s milk allergy, eczema, and wheeze. Compared with BM, both PHF and EHF may increase the risk of wheeze, but not other ADs. Given that most trials included only high-risk infants, more research on non-high-risk infants is warranted before any generalization is attempted.This protocol was registered in PROSPERO (CRD42022320787).