Purchase this article with an account.
Christopher Hammond, Pirro Hysi, Virginie Verhoeven, Caroline Klaver, CREAM Consortium on Myopia and Refractive Errors; Heterogeneity of effects in refractive error: lessons from the CREAM Consortium genome-wide association meta-analysis. Invest. Ophthalmol. Vis. Sci. 2013;54(15):5958.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Refractive error (RE) is a complex disease which is highly heritable, but world-wide variations in prevalence and distribution of RE suggest environmental factors are also important. The Consortium on Refractive Error and Myopia (CREAM) has recently identified genetic variants associated with RE in a meta-analysis of 27 genome-wide association studies (GWAS) including 38,593 Caucasian subjects. The purpose of this study was to explore the determinants of heterogeneity in RE-associated genetic variants across these multiple population-based studies around the world.
The 20 most significant loci associated with RE from the CREAM meta-analysis were taken forward and random effect meta-analyses were carried out regressing for putative sources of genetic effect heterogeneity. We attempted to investigate some of the known risk factors, including economic factors, education, sunlight and variations in RE allelic architecture. We used national statistics on Gross Domestic Product, tertiary education enrolment, geographic latitude, and specific allele frequencies at CREAM participating centers as national proxies for the above factors.
Heterogeneity of effects was generally low for most loci which were associated at genome-wide significance in the CREAM meta-analysis (maximum i2=0.197 for rs12193446 within the LAMA2 gene). The differing allele frequency at each center was the major source of heterogeneity for almost all the loci (p=0.002 for rs6495367 near RASGRF1 gene, for example). Other factors were associated with heterogeneity at a minority of loci such as geographic latitude with a locus in proximity to the GJD2 gene (p=0.008 for rs634990), and education and affluence (GDP) for a locus within the LAMA2 gene(both p=0.01 for rs12205363).
Our results show that differences in allele frequencies, which represent variations in RE allelic architecture between populations, are the main source of heterogeneity of genetic effect. However, environmental factors may have a significant role in modifying the effect at some specific genetic loci, namely latitude (and possible light exposure) influencing the GJD2 gene locus, and education/affluence at the LAMA2 gene. More specific and precise measures of environmental risk factors may allow identification of gene-environmental interactions at other loci.
This PDF is available to Subscribers Only