Body Mass Index (BMI) and Race Affect Adenoma Formation: A Large Single Center Cohort of Patients Undergoing Screening Colonoscopy

Abstract

Introduction: Obesity is reported as a risk factor for colorectal adenomas, but results have been varied especially in relation to race. We examined the effect of BMI on colorectal adenomas in the context of race in a large single-center cohort of patients undergoing screening colonoscopy. Methods: Average risk screening colonoscopies between 10/1/06 and 10/31/15 were included. Following races were studied-White, Black, Asian, South Asian, Other (Native American, Hawaiian, Pacific Islander), and unknown. Chi square tested for effects of BMI, and the interaction of BMI and race on the incidence of tubular adenoma (TA), tubulovillous adenoma (TVA), villous adenoma (VA) and sessile serrated adenoma (SSA). Multivariate logistic regression models were fitted to control for possible confounding of age, gender, race, and diabetes. Results: Of 53,164 patients, 17.4% were obese (BMI > 30) and 47.7% were overweight (BMI 25-29). In univariate analysis, higher BMI was a significant risk factor for all adenomas except SSA. The interaction of BMI with race increased risk of all adenoma except SSA in Whites only, and in addition for TA and TVA in “other”race, and for TVA and VA in South Asians. In multivariate logistic regression analysis, only Whites had an increased risk of TA if overweight (OR 1.11; 1.04-1.15), or obese (OR 1.27; 1.19-1.36). With Whites as the reference group, Blacks had higher risk of TA with BMI < 25 (OR 1.28; 1.05-1.55), “other” race had lower risk when overweight (OR 0.88; 0.76-0.99), and South Asians had a lower risk when obese (OR 0.71; 0.51-0.99). In the case of SSA, higher BMI did not increase risk in any race in multivariate regression analysis. With Whites as reference group, Blacks had a lower risk across BMI groups (< 25, 25-29 and > 30), while Asians had lower risk in < 25 and 25-29 BMI group, and South Asians had lower risk in 25-29 BMI group. The interaction of BMI and race could not be studied in multivariate regression analysis for TVA and VA due to small sample sizes. BMI >30 was associated with increased risk of TVA (OR 1.2, 95% C.I. 1.00-1.43) and VA (OR1.17; 0.99-1.38). Conclusion: Race strongly modifies BMI as a risk factor for colorectal adenomas. Also, the interaction of BMI and race varies based on type of adenoma. A linear relationship with increasing BMI was noted only for TA in Whites. BMI did not increase risk for SSA in any race. Whites had a higher SSA risk regardless of BMI category.243_A Figure 1. Baseline Characteristics243_B Figure 2. BMI & Tubular Adenoma243_C Figure 3. BMI & Sessile Serrated Adenoma