LAUSR

BACKGROUND It is currently uncertain whether non-alcoholic fatty liver disease (NAFLD) is associated with an increased risk of colorectal tumours. We performed a meta-analysis of relevant observational studies to quantify the magnitude of the association between NAFLD and risk of colorectal adenomas and cancer. METHODS We searched PubMed, Scopus and Web of Science from January 2000 to November 2017 using pre-defined keywords to identify observational studies of asymptomatic adults undergoing screening colonoscopy, in which NAFLD was diagnosed by imaging or histology. Data from selected studies were extracted and meta-analysis was performed using random-effects modelling. RESULTS Eleven observational studies (8 cross-sectional and 3 longitudinal) with aggregate data on 91,124 asymptomatic adults (32.1% with NAFLD) of predominantly Asian descent accounting for a total of 14,911 colorectal adenomas and 1684 cancers were included in the final analysis. NAFLD was associated with an increased risk of prevalent colorectal adenomas (n = 7 studies using liver imaging techniques; random-effects odds ratio [OR] 1.28, 95% CI 1.11-1.48; I2 = 82.9% or n = 1 study using liver biopsy; random-effects OR 1.61, 95% CI 0.90-2.89) and cancer (n = 4 studies using liver imaging techniques; random-effects OR 1.56, 95% CI 1.25-1.94; I2 = 65.6% or n = 1 study using liver biopsy; random-effects OR 3.04, 95% CI 1.29-7.18). NAFLD was also associated with an increased risk of incident colorectal adenomas (n = 3 studies; random-effects hazard ratio [HR] 1.42, 95% CI 1.18-1.72; I2 = 0%) and cancer (n = 1 study; random-effects HR 3.08, 95% CI 1.02-9.03). These risks were independent of age, sex, smoking, body mass index and diabetes (or metabolic syndrome). Sensitivity analyses did not alter these findings. Funnel plot and Egger's test did not reveal significant publication bias. CONCLUSIONS This meta-analysis of observational studies (involving asymptomatic individuals of predominantly Asian descent undergoing screening colonoscopy) suggests that NAFLD (detected by imaging or biopsy) is independently associated with a moderately increased prevalence and incidence of colorectal adenomas and cancer. However, the observational design of the studies does not allow for proving causality, and the possibility of residual confounding by some unmeasured factors cannot be ruled out. More prospective studies, particularly in European and American individuals, and mechanistic studies are required to better understand the association between NAFLD and colonic carcinogenesis.