The role of glycaemic, lipid, obesity and blood pressure risk factors as mediators of the effect of height on coronary artery disease and type 2 diabetes mellitus: A Mendelian randomisation study: E. Marouli 1; F. Del Greco M. 2; C.M. Astley 3,6; J. Yang 7,8; R.J.F Loos 4; Z. Kutalik 5; J.N. Hirschhorn 3,6; P. Deloukas 1; Genetic Investigation of ANthropometric Traits (GIANT) Consortium

Evidence from observational studies suggests that shorter stature is associated with cardiometabolic diseases. However, it is unclear whether this is a causal relationship or whether it is confounded by other factors. Using a Genetic Score for height (defined as adult stature) and Mendelian randomisation (MR), we investigated the causal effect of height on coronary artery disease (CAD) and type 2 diabetes (T2D). We also assessed whether these effects are mediated by known cardiometabolic risk factors (e.g. BMI, glycaemic traits, lipids, blood pressure) based on genetic evidence.
We used UK Biobank (449,094 unrelated British individuals) to investigate by several MR analyses: (1) the causal effect of height on CAD (23,755 cases vs 425,339 controls) and T2D (29,427 cases vs 416,908 controls), (2) the causal effect of height on possible mediators (known cardiometabolic risk factors), and (3) the extent to which these factors mediate any effect that height may have on CAD and T2D. In parallel, we used publically available summary statistics data from large consortia (GIANT, CARDIOGRAMplusC4D, DIAGRAM, MAGIC, GLGC) to perform two-sample MR analyses.
Two-sample MR analyses showed that one SD higher genetically determined height (~6.5 cm) was associated with decreased risk of CAD (OR: 0.83, 95% CI: 0.80-0.87). This association remained significant (OR: 0.86, 95% CI: 0.79-0.94) after performing sensitivity analyses. The height-CAD effect was reduced by 1-2%, in terms of magnitude, after adjustment for glycaemic (glucose, insulin, HBA1c, 2hGlu), lipid (Total Cholesterol, Low Density Lipoprotein, High Density Lipoprotein, Triglycerides) or blood pressure (BP; Systolic BP, Diastolic BP, Pulse Pressure) traits in Multivariate MR analyses. We observed a causal effect of height on T2D (OR: 0.93, 95% CI: 0.90-0.96, p=0.01), but it was entirely driven by its indirect effect through BMI (OR: 0.95, 95% CI: 0.92-0.99, p=0.12). There was no evidence for a causal relation between adult height and T2D adjusted for BMI (OR: 0.99, 95% CI: 0.96-1.02, p=0.95).
In summary, one SD higher genetically determined height reduces CAD risk by 17% and the association does not seem to be mediated by glycaemic, lipid and blood pressure traits, suggesting the observational association is due to a causal effect. On the other hand, we observed no direct causal effect of height on T2D risk, once its effect on BMI is taken into account.