| Abstract: | BACKGROUND AND AIM: This meta-analysis aims to explore the risk of colorectal polyps among non-alcoholic fatty liver disease (NAFLD) patients. METHODS: We searched PubMed, EMBASE, and Cochrane library databases using predefined search term to identify eligible studies (published up to 7 November 2019). Data from selected studies were extracted by using a standardized information collection form, and meta-analyses were performed using random-effects model. The statistical heterogeneity among studies (I2 ), subgroup analyses, meta-regression analyses, and the possibility of publication bias were assessed. RESULTS: Twenty observational (12 cross-sectional, two case-control, and six cohort) studies met the eligibility criteria, involving 142 387 asymptomatic adults. In cross-sectional/case-control studies, NAFLD was found to be associated with an increased risk of colorectal polyps (odds ratio [OR] = 1.34; 95% confidence interval [CI] = 1.23-1.47) (including unclassified colorectal polyps, hyperplastic polyps, adenomas, and cancers) with statistically significant heterogeneity (I2  = 67.8%; P < 0.001). NAFLD was also associated with a higher risk of incident colorectal polyps (hazard ratio = 1.60; 95% CI = 1.36-1.87) with low heterogeneity (I2  = 21.8%; P = 0.263) in longitudinal studies. The severity of NAFLD was associated with a higher risk of colorectal adenomas (OR = 1.57; 95% CI = 1.30-1.88), but not colorectal cancer (OR = 1.37; 95% CI = 0.92-2.03). The subgroup analysis according to gender showed that NAFLD was significantly associated with a higher risk of colorectal polyps in the male population without significant heterogeneity (OR = 1.47; 95% CI = 1.29-1.67, I2  = 0%), but not in the female population (OR = 0.88; 95% CI = 0.60-1.29, I2  = 34.2%). CONCLUSIONS: NAFLD was associated with an increased risk of colorectal polyps. There was a significant difference of the relationship between genders, which suggested more precise screening colonoscopy recommendation in NAFLD patients according to gender. |
