Association results of PPARGC1A SNPs for NV AMD in three cohorts and in meta-analysis using multivariable models.

Abbreviations: 3′UTR, 3′ untranslated region; SNP, single-nucleotide polymorphism. a, SNPs in nearly complete linkage disequilibrium (r2 = 0.96) – no other SNPs were in linkage disequilibrium; ADD, additive model (minor allele count –2|1|0); DOM, dominant model (grouping minor allele homozygotes with heterozygotes). SNPs were tested from the panel of the ILLUMINA HumanCNV370v1 chip (SNP batch IDs at http://www.ncbi.nlm.nih.gov/SNP/snp_viewBatch.cgi?sbid=1047132). People in the reference groups (controls) were AMD-free and at least 65-years-of-age at the time of phenotype classification. We computed odds ratios (ORs) and 95% confidence intervals (95% CI) from age-, sex, and smoking-adjusted logistic regression analyses on 506 cases and 512 controls in the Discovery Cohort (University of Michigan), 123 cases and 198 controls in Replication Cohort 1 (University of Pennsylvania), and 205 cases and 314 controls in Replication Cohort 2 (Mayo Clinic, Rochester). Combined estimates (ORmeta) were computed with age-, sex, and smoking-adjusted meta-regression – random effects models were applied in instances indicated by Cochrane’s Q statistic. All P values are 2-sided, with the exception of those for the replication cohorts. Exact (empirical) P values are from max(T) permutation with 10000 iterations on the full sample.