\usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\sqrt {h_i^2h_j^2}$$\end{document} where ρg is the genetic covariance between traits, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$h_i^2$$\end{document} is the heritability of trait i and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$h_j^2$$\end{document} is the heritability of trait j36. Near-zero h2 estimates can therefore cause the rG estimate to become out of bounds (rG > 1), as observed in three out of five nominally significant traits, with large standard errors72, as observed in all nominally associated traits. Fig. 3 shows the genetic correlation of meta-analyzed depression subtypes with significantly correlated health traits.
