[Competing risks of death and Hardy-Weinberg equilibrium in case-control studies of gene-disease association]

Abstract

Objective: To study the impact of competing risks on Hardy-Weinberg equilibrium and their consequences in case-control studies of gene-late onset disease association.

Methods: Based on a population born in Hardy-Weinberg equilibrium for a particular gene, the genetic composition when the gene is associated with a lethal early-onset disease and its consequences on a late-onset disease can be deduced. Odds ratios estimates are unbiased in case-control studies when controls are sampled by density, even if the controls are in Hardy-Weinberg disequilibrium.

Results: An example in which a mutant gene is associated with early mortality is presented, producing a departure from Hardy-Weinberg equilibrium; as a result, controls in later ages are in disequilibrium, producing an odds ratio equal to 1.61.

Conclusion: Although the main causes of Hardy-Weinberg disequilibrium in controls are selection bias or genotyping error, a competing risk of death associated with the mutant gene would also result in Hardy-Weinberg disequilibrium among controls.