\usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\hat b_{xy} = - 0.21$$\end{document}b^xy=-0.21 and OR = 0.81), which is obvious because one of the diagnostic criteria for dyslipidemia is an abnormally low level of HDL-c. In addition, there was a highly significant risk effect (OR = 1.36) of HDL-c on age-related macular degeneration (AMD) in the case–control data, consistent with the result from a recent MR study42. The associations between lipids and AMD are controversial and results from different observational studies are inconsistent43. Our results support the observations that increased HDL-c is associated with increased risk of AMD43–45. It should be noted that LDL-c and TG also appeared to be associated with AMD before HEIDI-outlier filtering but the effects were not significant after HEIDI-outlier filtering (Supplementary Fig. 12), implying that the observed association between LDL-c (or TG) and AMD in epidemiological studies43 might be due to pleiotropy.
