Revisiting heritability accounting for shared environmental effects and maternal inheritance

Area of Science:

• Quantitative genetics
• Human genetics
• Biostatistics

Background:

• Heritability quantifies genetic contribution to phenotypic variation.
• Complex traits are influenced by genetic and various environmental/familial factors.
• Confounding effects of non-genetic variance components on heritability are understudied in humans.

Purpose of the Study:

• To obtain unbiased heritability estimates for complex traits considering multiple variance components.
• To estimate the contribution of non-genetic variance components to phenotypic variation.
• To compare regression and variance component methods for heritability estimation under complex variance structures.

Main Methods:

• Simulations were used to compare heritability estimation methods with and without additional variance components.
• Variance component models were applied to Framingham Heart Study (FHS) data.
• Models accounting for spousal, maternal, and sibling effects were compared to traditional familial models.

Main Results:

• Failure to account for variance components leads to biased heritability estimates.
• Heritability for FHS traits was overestimated by 4-47% when non-genetic factors were ignored.
• Spousal correlation explained 14-36% of variation in anthropometric/lifestyle traits.
• Maternal and sibling effects contributed 3-7% to variation in anthropometric/metabolic traits.

Conclusions:

• Accurate heritability estimation necessitates incorporating multiple variance components.
• Non-genetic factors like spousal, maternal, and sibling effects significantly influence complex trait variation.
• These findings may help resolve the 'missing heritability' problem in genetic studies.