What is the evidence for interactions between filaggrin null mutations and environmental exposures in the aetiology of atopic dermatitis? A systematic review

H. Blakeway, V. Van-de-Velde, V. B. Allen, G. Kravvas, L. Palla, M. J. Page, C. Flohr, R. B. Weller, A. D. Irvine, T. McPherson, A. Roberts, H. C. Williams, N. Reynolds, S. J. Brown, L. Paternoster, S. M. Langan, on behalf of UK TREND Eczema Network

Research output: Contribution to journalReview ArticleResearchpeer-review

2 Citations (Scopus)


Background: Epidemiological studies indicate that gene–environment interactions play a role in atopic dermatitis (AD). Objectives: To review the evidence for gene–environment interactions in AD aetiology, focusing on filaggrin (FLG) loss-of-function mutations. Methods: A systematic search from inception to September 2018 in Embase, MEDLINE and BIOSIS was performed. Search terms included all synonyms for AD and filaggrin/FLG; any genetic or epidemiological study design using any statistical methods were included. Quality assessment using criteria modified from guidance (ROBINS-I and Human Genome Epidemiology Network) for nonrandomized and genetic studies was completed, including consideration of power. Heterogeneity of study design and analyses precluded the use of meta-analysis. Results: Of 1817 papers identified, 12 studies fulfilled the inclusion criteria required and performed formal interaction testing. There was some evidence for FLG–environment interactions in six of the studies (P-value for interaction ≤ 0·05), including early-life cat ownership, older siblings, water hardness, phthalate exposure, higher urinary phthalate metabolite levels (which all increased AD risk additional to FLG null genotype) and prolonged breastfeeding (which decreased AD risk in the context of FLG null genotype). Major limitations of published studies were the low numbers of individuals (ranging from five to 94) with AD and FLG loss-of-function mutations and exposure to specific environmental factors, and variation in exposure definitions. Conclusions: Evidence on FLG–environment interactions in AD aetiology is limited. However, many of the studies lacked large enough sample sizes to assess these interactions fully. Further research is needed with larger sample sizes and clearly defined exposure assessment.

Original languageEnglish
Pages (from-to)443-451
Number of pages9
JournalBritish Journal of Dermatology
Issue number3
Publication statusAccepted/In press - 1 Jan 2019

Cite this