Mendelian randomisation study of age at menarche and age at menopause and the risk of colorectal cancer

Sonja Neumeyer,Peter T Campbell,Michael O Woods,Mingyang Song,Robert E Schoen,John L Hopper,Li Hsu,Graham Casey,Steven Gallinger,Clemens Schafmayer,Katja Butterbach,Bette J Caan,William M Grady,Alicja Wolk,Kenneth Offit,John D Potter,Stephen B Gruber,Emily White,Andrew T Chan,Volker Arndt,Anna H Wu,Graham G Giles,Susanna C Larsson,Stéphane Bézieau ,Amit D Joshi ,Aung Ko Win ,Loı̈c Le Marchand ,Edward Giovannucci ,Paul D.p Pharaoh ,Vı́ctor Moreno ,Daniel D Buchanan ,Jiyan Dai ,Yi Lin,Hermann Brenner,Mark A Jenkins,Gad Rennert,Polly A Newcomb,Mathieu Lemire,Lihong Qi,Michael Hoffmeister,Barbara L Banbury,Annika Lindblom,Stephen J Chanock,Li Li,Martha L Slattery,Stephanie A Bien,Jochen Hampe,Ulrike Peters,Cornelia M Ulrich,Sonja I Berndt,Jenny Chang‐Claude

doi:10.1038/s41416-018-0108-8

Abstract

BackgroundSubstantial evidence supports an association between use of menopausal hormone therapy and decreased colorectal cancer (CRC) risk, indicating a role of exogenous sex hormones in CRC development. However, findings on endogenous oestrogen exposure and CRC are inconsistent.MethodsWe used a Mendelian randomisation approach to test for a causal effect of age at menarche and age at menopause as surrogates for endogenous oestrogen exposure on CRC risk. Weighted genetic risk scores based on 358 single-nucleotide polymorphisms associated with age at menarche and 51 single-nucleotide polymorphisms associated with age at menopause were used to estimate the association with CRC risk using logistic regression in 12,944 women diagnosed with CRC and 10,741 women without CRC from three consortia. Sensitivity analyses were conducted to address pleiotropy and possible confounding by body mass index.ResultsGenetic risk scores for age at menarche (odds ratio per year 0.98, 95% confidence interval: 0.95–1.02) and age at menopause (odds ratio 0.98, 95% confidence interval: 0.94–1.01) were not significantly associated with CRC risk. The sensitivity analyses yielded similar results.ConclusionsOur study does not support a causal relationship between genetic risk scores for age at menarche and age at menopause and CRC risk.

Highlights

Colorectal cancer (CRC) is the third most common cancer worldwide and incidence rates are higher in men than in women.[1]
We examined the association between genetic risk scores (GRS) and colorectal cancer (CRC) risk using logistic regression models adjusted for study, age as well as principal components (PCs) of genetic ancestry to account for potential population stratification
Sensitivity analyses adjusting for BMI using a GRS yielded similar results for age at menarche

Summary

Introduction

Colorectal cancer (CRC) is the third most common cancer worldwide and incidence rates are higher in men than in women.[1]. METHODS: We used a Mendelian randomisation approach to test for a causal effect of age at menarche and age at menopause as surrogates for endogenous oestrogen exposure on CRC risk. RESULTS: Genetic risk scores for age at menarche (odds ratio per year 0.98, 95% confidence interval: 0.95–1.02) and age at menopause (odds ratio 0.98, 95% confidence interval: 0.94–1.01) were not significantly associated with CRC risk. CONCLUSIONS: Our study does not support a causal relationship between genetic risk scores for age at menarche and age at menopause and CRC risk

Objectives

Methods

Results

Conclusion