Heritability estimation of sex-specific effects on human quantitative traits

Recent studies have suggested that sex-specific genetic architecture could be because of the effects of autosomal genes that are differentially expressed in males and females. Yet, few studies have explored the effects of X-linked genes on sex-specific genetic architecture. In this study, we extended the variance component, maximum likelihood method to evaluate the relative contributions of sex-specific effects on both autosomes and the X chromosome to estimates of heritability of 20 quantitative human phenotypes in the Hutterites. Seventeen of these traits were previously analyzed in this population under a model that did not include X chromosomal effects; three traits are analyzed for the first time (age at menarche, percent fat and fat-free mass [FFM]). Seven traits (systolic blood pressure (SBP), adult height, fasting insulin, triglycerides, lipoprotein (a) [Lp(a)], serotonin, and age at menarche) showed significant X-linked effects; three of these (SBP, adult height, and triglycerides) showed X-linked effects only in males. Four traits (Lp(a), low-density lipoprotein cholesterol, ratio of percent predicted forced expiratory volume at 1 s/forced vital capacity, and FFM) showed significant sex-environment interactions, and two traits (high-density lipoprotein cholesterol and FFM) showed significant sex-specific autosomal effects. Our analyses demonstrate that sex-specific genetic effects may not only be common in human quantitative traits, but also that the X chromosome both plays a large role in these effects and has a variable influence between the sexes.