A Robust and Unified Framework for Estimating Heritability in Twin Studies using Generalized Estimating Equations

TL;DR

This paper introduces a flexible GEE2 framework for estimating heritability in twin studies that improves robustness, allows covariate adjustment, and outperforms traditional methods under non-normal data conditions.

Contribution

The paper develops a unified GEE2-based approach for fitting ACE models, integrating NACE and Falconer's methods, with enhanced robustness and covariate incorporation.

Findings

GEE2 models provide better coverage of heritability estimates under non-normal data.

The GEE2-Falconer method can incorporate covariates affecting heritability.

NACE model can produce biased estimates in certain scenarios.

Abstract

The development of a complex disease is an intricate interplay of genetic and environmental factors. "Heritability" is defined as the proportion of total trait variance due to genetic factors within a given population. Studies with monozygotic (MZ) and dizygotic (DZ) twins allow us to estimate heritability by fitting an "ACE" model which estimates the proportion of trait variance explained by additive genetic (A), common shared environment (C), and unique non-shared environmental (E) latent effects, thus helping us better understand disease risk and etiology. In this paper, we develop a flexible generalized estimating equations framework ("GEE2") for fitting twin ACE models that requires minimal distributional assumptions, rather only the first two moments need to be correctly specified. We prove that two commonly used methods for estimating heritability, the normal ACE model ("NACE") and Falconer's method, can both be fit within this unified GEE2 framework, which additionally provides robust standard errors. Although the traditional Falconer's method cannot directly adjust for covariates, we show that the corresponding GEE2 version ("GEE2-Falconer") can incorporate covariate effects for both mean and variance-level parameters (e.g. let heritability vary by sex or age). Given non-normal data, we show that the GEE2 models attain significantly better coverage of the true heritability compared to the traditional NACE and Falconer's methods. Finally, we demonstrate an important scenario where the NACE model produces biased estimates of heritability while Falconer's method remains unbiased. Overall, we recommend using the robust and flexible GEE2-Falconer model for estimating heritability in twin studies.