Measuring the Burden of Modifiable Risk Factors for Colorectal Cancer in an Inner City Cohort

Abstract

Introduction: The rising prevalence of obesity projects it to exceed cigarette smoking as the leading preventable cause of death in the US. Increasing evidence demonstrates a link between obesity and colorectal cancer. With minority populations carrying a disproportionate degree of incidence and mortality secondary to colorectal cancer, this study sought to examine the burden of modifiable risk factors within a predominantly African American (AA) cohort at a major inner city teaching hospital. Methods: All adults who had undergone a screening or diagnostic colonoscopy from 2013 to 2015 were included in this study. A total of 4359 patients were analysed after eliminating the duplicate data.The following variables, BMI categories, cigarette smoking status, ethnicity, Diabetes Mellitus and alcohol intake were analysed to examine if there were differences in risk factors (DM, smoking, BMI and alcohol) between these groups. Results: Our population consisted of 77.3% AA. The prevalence of obesity, BMI ≥ 30, in this cohort, was 44.9%. 44.8% with DM, 19.7% current smokers and 31.1% reporting alcohol consumption. 40 subjects (0.9%) were diagnosed with CRC over the 3 year period. Subjects were more likely to be obese or overweight (77.1%). AA in this population were more likely to be overweight and obese (OR-1.25, 95% CI- 1.06 - 1.47) and to consume alcohol products (OR- 1.45; 95% CI- 1.23 - 1.70) There was no statistical differences in smoking status and DM in AA. In the analysis of predictors of CRC in the study population, gender was independently associated with the development of CRC, with females being less likely to be diagnosed by almost half (OR- 0.45: 95% CI- 0.23- 0.86). When adjusted for all the other predictors including increasing. There was still a statistically significant association between gender, females vs males (OR- 0.46; 95% CI; 0.23- 0.93) and increasing age (OR- 1.06, 95% CI; 1.02- 1.10). Conclusion: Overall incidence of CRC was higher in this population compared to US averages. AA were more likely to have a greater burden of modifiable risk factors (BMI>25 and alcohol intake), but these factors did not significantly predict risk of CRC. It is possible that in low socioeconomic cohorts, other unmeasured confounders like diet and environmental factors may play a greater role as modifiable risk factors than obesity and DM. This study highlights the striking prevalence and additional burden of modifiable risk factors in an underserved, predominantly AA cohort.