Abstract
Population stratification represents the principle that genetic variation differs across populations, and these differences may lead to problematic interpretations of epidemiologic studies when the composition of the study population could lead to unmeasured confounding. Advances in genotyping technology greatly facilitate genetic association studies, yet it is critical to understand the relationship between genotype, haplotype, and functional risk alleles. Insights from extended studies of a chromosomal region can provide perspective into the ancestral heritage of founder populations and risk alleles. Examples from studies of colorectal cancer in Ashkenazi Jewish populations, and studies of smoking behavior in Caucasians of European origin, Mexican-Americans, Japanese and Han Chinese illustrate these principles.