Given the estimates from conditional GSMR analyses (Fig. 5; Supplementary Table 11), we could use an approximate approach to calculate the aggregate effect of multiple risk factors on a disease, i.e., \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\log \left( {{\mathrm{OR}}} \right) = {\sum} {[x_i\log \left( {{\mathrm{OR}}_i} \right)]}$$\end{document} logOR=∑[xi logORi]. Here is a hypothetical example. If all the risk factors increase by 1 SD (i.e., ~4 kg m−2 for BMI, ~1 mmol L−1 for LDL-c, ~1 mmol L−1 for TG and ~19 mm Hg for SBP), we would have an increased risk of ~2.3-fold to T2D (e1.01−0.17), and 4.5-fold to CAD (e0.41+0.47+0.14+0.48).
