Quantile regression analysis reveals widespread evidence for gene-environment or gene-gene interactions in myopia development

Abstract

A genetic contribution to refractive error has been confirmed by the discovery of more than 150 associated variants in genome-wide association studies (GWAS). Environmental factors such as education and time outdoors also demonstrate strong associations. Currently however, the extent of gene-environment or gene-gene interactions in myopia is unknown. We tested the hypothesis that refractive error-associated variants exhibit effect size heterogeneity, a hallmark feature of genetic interactions. Of 146 variants tested, evidence of non-uniform, non-linear effects were observed for 66 (45%) at Bonferroni-corrected significance ( P < 1.1 × 10 −4 ) and 128 (88%) at nominal significance ( P < 0.05). LAMA2 variant rs12193446, for example, had an effect size varying from −0.20 diopters (95% CI −0.18 to −0.23) to −0.89 diopters (95% CI −0.71 to −1.07) in different individuals. SNP effects were strongest at the phenotype extremes and weaker in emmetropes. A parsimonious explanation for these findings is that gene-environment or gene-gene interactions in myopia are pervasive.

Effect of Education on Myopia: Evidence from the United Kingdom ROSLA 1972 Reform

“…If such heterogeneity is genuine, it may indicate that genetically-predisposed individuals are likely to develop myopia irrespective of their exposure to education. Providing indirect support for this result, Pozarickij et al 38 recently demonstrated that 88% of known GWAS variants associated with refractive error exhibited evidence of interaction effects, including gene × education interaction effects.…”

Association of 5p15.2 and 15q14 with high myopia in Tujia and Miao Chinese populations

“…Although the SNPs in these studies are different from those we studied, they do suggest that interactions between genetics and environment have impact on the risk of myopia. The study by Pozarickij et al does show an interaction for the 15q14 region, although not for the specific SNPs we used [74]. For now, no paper has reported on the interaction of genetic and environmental factors in Tujia and Miao populations, so it is worth collecting more data about Tujia and Miao populations for further analysis and research, but that is beyond the scope of this report.…”

“…Unfortunately, this number of responses is not enough to serve as covariates in our study. Thirdly, there are also some studies that have reported the interaction of genetic and environmental factors on the risk of myopia [71,72,[74][75][76]. For example, the study of Fan et al found that three genome-wide associated loci, AREG, GABRR1 and PDE10A, showed strong interaction with education in Asian populations, but this interaction was not significant in European populations [77].…”

Association of 5p15.2 and 15q14 with High Myopia in Tujia and Miao Chinese Populations

“…Although the SNPs in these studies are different from those we studied, they do suggest that interactions between genetics and environment have impact on the risk of myopia. The study by Pozarickij et al does show an interaction for the 15q14 region, although not for the speci c SNPs we used [74]. For now, no paper has reported on the interaction of genetic and environmental factors in Tujia and Miao populations, so it is worth collecting more data about Tujia and Miao populations for further analysis and research, but that is beyond the scope of this report.…”

“…Unfortunately, this number of responses is not enough to serve as covariates in our study. Thirdly, there are also some studies that have reported the interaction of genetic and environmental factors on the risk of myopia [71,72,[74][75][76]. For example, the study of Fan et al found that three genome-wide associated loci, AREG, GABRR1 and PDE10A, showed strong interaction with education in Asian populations, but this interaction was not signi cant in European populations [77].…”