Genome-wide association studies (GWAS) exploit the correlation structure in the genome, due to linkage disequilibrium (LD), by testing a representative subset of genetic markers for association with disease. As a result, we expect GWAS to highlight markers correlated with causal loci rather than to discover the causal loci directly. Depending on the strength of LD between causal and marker single nucleotide polymorphisms (SNPs), the observed disease effect at the marker will be an imperfect representation of the true disease effect.
