Establishment of classification and regression tree model for assessing of a risk for future occurence of life-style related disease using a community-based cohort data.

2016 Fiscal Year Final Research Report

• Project/Area Number: 25460768
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Epidemiology and preventive medicine
• Research Institution: Sapporo Medical University
• Principal Investigator: Ohnishi Hirofumi 札幌医科大学, 医学部, 准教授 (20359996)
• Project Period (FY): 2013-04-01 – 2017-03-31
• Keywords: 特定健診・特定保健指導 / 生活習慣病罹患ハイリスク者 / 決定木分析 / CART / 地域一般住民コホート

Outline of Final Research Achievements

We tried to establish a useful Classification and Regression Tree （CART）model for assessing a risk for future occurrence of life-style related disease, which includes hypertension, type 2 diabetes and cardiovascular disease (CVD), using a community-based cohort data. In the predicting model for hypertension, sensitivity and specificity were 68.2% and 71.1%, respectively. After exclusion of individuals with metabolic syndrome (MetS) or pre-MetS, sensitivity and specificity changed into 65.3% and 84.2%, respectively. In the predicting models for type 2 diabetes, sensitivity and specificity were 54.2% and 96.3%, respectively. In the predicting model for CVD, sensitivity and specificity were 7.4% and 99.3%, respectively. To find individuals at high risk for future occurrence of life-style related disease, using a CART model may be useful in the situation of health checkup and health guidance.

Free Research Field

疫学