“…In multivariate analysis, we included factors such as the use of NSAIDs, low-dose aspirin, and non-aspirin antiplatelet agents, which showed a significant association with bleeding in the univariate analysis. In addition, we also included age, gender, alcohol, smoking, and metabolic factors, such as diabetes mellitus, dyslipidemia, and hypertension, and chronic kidney disease, which showed association with obesity in univariate analysis and which affects the risk of diverticular bleeding, as previously reported [7,8,22]. Thus, the multivariate model was developed using age, gender, alcohol, smoking, NSAIDs, low-dose aspirin, non-aspirin antiplatelet agents, metabolic factors, and chronic kidney disease (Model 1).…”