Abstract MP59: Revisiting Heritability Accounting For Common Environmental Effects And Maternal Inheritance
Background: Heritability measures the proportion of phenotypic variation attributable to genetic factors. Reliable heritability estimates are used to determine the necessary sample size and power for subsequent genetic studies of susceptibility genes. In addition to a shared nuclear genetic component, common environmental factors and maternal or mitochondrial inheritance may have strong effects on some phenotypes. Failure to account for necessary variance components may give rise to biased heritability estimates.
Aims: Our primary goal was to investigate how heritability estimates are biased in the presence of common environmental or/and maternal effects. The second goal was to estimate mitochondrial inheritance for a number of common phenotypes.
Methods: We employed variance component methods to account for additional variance components using both simulated and Framingham Heart Study (FHS) pedigree data to revisit heritability estimates in the presence of common environmental and maternal effects.
Results: Using both simulated families and actual FHS extended pedigrees, we demonstrated that heritability is greatly overestimated when key variance components are not properly accounted for. The inflation in heritability ranged from 9% to 214% across several anthropometric, metabolic and life-style phenotypes when we compared models that consider correct variance components and simple models that only consider familial relationships (Table 1). Maternal inheritance was observed in most phenotypes investigated. The estimated maternal inheritance ranged from ~1 to 5%, which is considerable because the mitochondrial genome is about the size of an average gene.
Conclusion/Discussion: We systematically investigated the influence of common environmental effects and maternal inheritance in heritability estimates using extended pedigrees. Our findings may explain, in part, the missing heritability for some traits and may facilitate further collaborations in the genetic study of mitochondrial DNA in disease susceptibility for common phenotypes.
Author Disclosures: C. Liu: None. J. Dupuis: None. M.G. Larson: None. L. Cupples: None. J.M. Ordovas: None. R.S. Vasan: None. J. Meigs: None. P.F. Jacques: None. D. Levy: None.
- © 2014 by American Heart Association, Inc.