Purchase this article with an account.
Madeleine K. Adams, Julie A. Simpson, Luba Robman, Galina Makeyeva, Khin Zaw Aung, Andrea Richardson, Robyn Guymer, Paul N. Baird; Does Age Modify Associations with Complement Factor H in Age Related Macular Degeneration?. Invest. Ophthalmol. Vis. Sci. 2012;53(14):1324.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To examine whether age modifies single nucleotide polymorphism (SNP) associations between the complement factor H (CFH) gene and Age-Related Macular Degeneration (AMD).
4588 participants (2,294 cases of early or late AMD and 2,294 controls, individually matched on sex, age and country of birth), aged 48-86 years at AMD detection, were selected from the Melbourne Collaborative Cohort Study of 22287 participants. Fundus photographs were graded for AMD. Genotypes and allele frequencies for the CFH SNPs rs1061170 (Y402H); rs2274700; rs3766404; rs393955; rs800292 were determined and categorised as homozygous wild type, heterozygous and homozygous risk. Estimates of the odds ratios (OR) were determined using logistic regression. Effect modification by age group (75 years) were investigated by fitting interaction terms between the categorical variable and genotype, and tested by the likelihood ratio test.
Differences in genotype frequency were seen for all 5 CFH SNPs included in this study across the differing age groups in controls, where the wild type homozygote prevalence rose with increasing age group. Associations with early AMD were strongly modified by age group for 3 of the 4 SNPs (interaction P-value 0.01 - 0.00003). In addition, a reversal of direction of association for each SNP from a negative to a positive association was evident with increasing age. Odds ratios for risk homozygotes for each SNP ranged from 0.37 to 0.48 in younger age groups, and from 1.87 to 2.8 in older age groups. The inverse relationship reached significance in two of the SNPs (rs1061170 and rs393955) P<0.05; the positive relationship was significant across all SNPs (P<0.001).
These results show a reversal of genetic associations with increasing age and have important implications for predictive models for AMD that extrapolate risks from older cohorts, or do not stratify by age, as they assume a homogeneity of risk across all age groups, which in fact may not exist.
This PDF is available to Subscribers Only