Abstract
Purpose: :
To compare distribution and clinical significance between the 3 single nucleotide polymorphisms (SNPs) of the CFH and ARMS2 (major)genes and the 9 SNPs of other reported age-related macular degeneration (AMD) related (non-major) genes in an older Australian sample.
Methods: :
Of 4828 Blue Mountains Eye Study participants, 2534 (52.5%) had good quality genome-wide association data. The 3 major SNPs include rs1061170, rs10737680 (CFH) and rs10490924 (ARMS2), and 9 SNPs of non-major genes include rs13081855 (COL8A1), rs4698775 (CFI), rs429608 (C2.CFB), re943068 (VEGFA), rs3812111 (CIL10A1), rs13278062 (TNFRSF10A), rs920915 (LIPC), rs1864163 (CETP) and rs5749482 (SYN3). The presence of at least one risk allele (versus none) was used to define the risk status of individuals for each SNP, and these binary risk indicators were group for major and non-major genes. Genetic associations of cumulative (prevalent and incident) late AMD prevalence, assessed using logistic regression models, were compared between major and non-major SNP groups. Age-onset of incident late AMD over 15 years was compared between subjects with and without any risk allele for each SNP.
Results: :
Of 2534 subjects, 1974 (78%) had data on the 12 SNPs, and 28.8%, 46.4% and 24.8% of these had risk status for 0-1, 2 and 3 major SNPs, and 9.4%, 77.5% and 13.2 % had risk status for 0-3, 4-6 and 7-9 non-major SNPs, respectively. All SNPs of major and only 2 of 9 SNPs (COL8A1, CFI) of non-major genes were significantly associated with late AMD. Compared to subjects with risk status for 0-1 major SNPs, those with risk status for 2 (adjusted odds ratio, OR 4.0, 95% confidence interval, CI, 1.9-8.5) or 3 (OR 6.1, 95% CI 2.7-13.6) were significantly more likely to have late AMD. Compared to subjects with risk status for 0-3 non-major SNPs, having risk status for 4-6 (OR 0.7, 95% CI 0.3-1.6) or 7-9 (OR 1.2, 95% CI 0.5-3.1) was not significantly associated with late AMD, after adjusting for age, sex, smoking and genetic risk groupings of major and non-major SNPs. Adding any or all non-major SNPs to an AMD prediction model containing the major SNPs did not improve the model fit. A younger age-onset of late AMD was observed only in cases with risk status for rs10490924 (ARMS2, 77.8 vs 81.6 years, p=0.02) and rs1864163 (CETP, 78.2 vs 81.1 years, p=0.08).
Conclusions: :
Over 70% of this general population had 2+ risk alleles in CFH and ARMS2,and over 90% had 4+ risk alleles in the known non-major genes. While the association of late AMD with the major genes was robust, the clinical significance of these non-major genes was negligible, questioning the huge effort spent on finding more non-major genes for AMD.