Evaluation of Ancestral Admixture and Substructure in Latinos and Its Impact on Determining the Risk of Age-Related Macular Degeneration (AMD)

Abstract

Purpose: : The amount of ancestral structure and admixture differs substantially between Latinos and Caucasians. It has been suggested that their genetic landscape consists of diverse proportions of European, Native American, and African descent. Preliminary evidence suggests that there are differences in risk for AMD due to population background. In this study we sought to (1) evaluate the degree of population admixture and substructure in Latinos and (2) determine the impact different ancestries may have on the risk of having AMD among Latinos.

Methods: : A total of 224 prevalent AMD cases and 224 controls with no AMD were selected from the LALES cohort. Also, 5 Mexican Indians and 25 Mayan Amerindians were chosen from Coriell’s Human Population Collection to establish a Native American (NA) reference set. LALES and NA individuals were genotyped for 233 ancestry informative markers through Illumina’s platform. These markers exhibit large differences in allele frequencies between different ethnicities and confer an increased power for detecting levels of population stratification. Population structure was analyzed through the clustering algorithm implemented in Structure2.0. LALES admixture coefficients were compared with African American, European, and Asian descent individuals from the Multiethnic Cohort study of cancer. Estimated individual ancestries for the LALES subjects were then incorporated in allelic tests of association models to evaluate the effect of population admixture on risk of having AMD. We are currently analyzing the NA set with the purpose of comparing it with the LALES ancestry data.

Results: : Based on a 4-cluster model, the estimated ancestral background for the LALES population was distributed as 2.8% African, 7.2% Asian, 30% European, and 60% of a distinct ancestry. Immediate comparison of the NA set with the preponderant distinct substructure proportion of the LALES subjects is currently in progress. Allelic association tests for the LALES sample show a variation in genetic risk variants depending mainly on the amount of European or potentially Native American descent.

Conclusions: : Population stratification confounds genetic association studies particularly in Latinos. The LALES population has a decreased amount of European ancestry, an important factor that may contribute to the lower risk of having late AMD in Latinos compared to Caucasians.

Keywords: age-related macular degeneration • genetics • clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology