Genetic risk score Mendelian randomization shows obesity measured as body mass index, but not waist:hip ratio, is causal for endometrial cancer

Abstract

$\textbf{Background:}$ The strongest known risk factor for endometrial cancer (EC) is obesity. To determine whether single nucleotide polymorphisms (SNPs) associated with increased body mass index (BMI) or waist-hip ratio (WHR) are associated with EC risk, independent of measured BMI, we investigated relationships between 77 BMI and 47 WHR SNPs and EC in 6,609 cases and 37,926 country-matched controls.

$\textbf{Methods:}$ Logistic regression analysis and fixed-effects meta-analysis were used to test for associations between EC risk and (i) individual BMI or WHR SNPs, (ii) a combined weighted genetic risk score (wGRS) for BMI or WHR. Causality of BMI for EC was assessed using Mendelian randomization, with BMIwGRS as instrumental variable.

$\textbf{Results:}$ The BMIwGRS was significantly associated with EC risk ($P$=3.4x10$^{-17}$). Scaling the effect of the BMIwGRS on EC risk by its effect on BMI, the EC odds ratio (OR) per 5kg/m$^2$ of genetically predicted BMI was 2.06 (95% confidence interval (CI)=1.89-2.21), larger than the observed effect of BMI on EC risk (OR=1.55, 95% CI 1.44-1.68, per 5kg/m$^2$). The association attenuated but remained significant after adjusting for BMI (OR=1.22, 95% CI=1.10-1.39, $P$=5.3x10$^{-4}$). There was evidence of directional pleiotropy ($P$=1.5x10$^{-4}$). BMI SNP rs2075650 was associated with EC at study-wide significance ($P$<4.0x10$^{-4}$), independent of BMI. EC was not significantly associated with individual WHR SNPs or the WHRwGRS.

$\textbf{Conclusions:}$ BMI, but not WHR, is causally associated with EC risk, with evidence that some BMI-associated SNPs alter EC risk via mechanisms other than measurable BMI.

$\textbf{Impact:}$ The causal association between BMI SNPs and EC has possible implications for EC risk modeling.