Emollients in infancy to prevent atopic dermatitis: A systematic review and meta-analysis

Background Several studies have evaluated prophylactic emollients as a preventive strategy against atopic dermatitis (AD) and food allergy (FA). We aimed to synthesize the evidence on efficacy and safety of prophylactic emollients started during the first 6 weeks of infancy for prevention of AD and FA. Methods MEDLINE, Embase, CINAHL, BIOSIS, and the Cochrane Library databases were searched systematically for randomized controlled trials published between January 2000 and July 2020, which assessed the effects of prophylactic emollients initiated within the first 6 weeks of life on the development of AD within 24 months of age, compared to no treatment. Risk of bias and certainty of evidence were assessed using the Cochrane Collaboration's tool and GRADE process, respectively. Results Of the 1486 articles identified, 10 studies fulfilled inclusion criteria. In infants given emollients, there was no significant reduction on the development of AD (risk ratio (RR) 0.84, 95% confidence interval (CI) 0.64, 1.10) compared to the control group. However, there was significant benefit of prophylactic emollients (RR 0.75, 95% CI 0.62-1.11) in the high-risk population (n = 8 studies). There was also significant benefit (RR 0.59, 95% CI 0.43, 0.81) in studies (n = 6) where emollients were used continuously to the point of AD assessment; but not when treatment was ceased for an interval before AD assessment. There were no protective effects on FA found. Conclusion The prophylactic application of emollients initiated in early infancy may prevent AD, especially in high-risk populations and when used continuously. We hypothesize that emollients may delay rather than prevent AD.

Automated summary

Prophylactic use of emollients initiated in early infancy may prevent AD, especially in high-risk populations and when used continuously. However, emollients may delay rather than prevent the development of AD. No protective effects on FA were found.