Risk factors and predictive model for comorbidity in the elderly population:Taking Guangzhou City as an example
Objective To explore the risk factors of comorbidity in the elderly population and to construct a risk prediction model.Methods Multi-stage sampling method was used to investigate the demographic charac-teristics,educational background,marital status,and comorbidity of 1,100 elderly people aged 60 and above in Guangzhou City.Univariate tests,logistic regression analysis,and decision tree CHAID model were used to ex-plore the Influencing factors of comorbidity in the elderly population.Results Logistic regression analysis results showed that gender,urban or rural area,waist to hip ratio,sleep duration,daily slow walking time,and daily low-intensity household chores were related factors for comorbidity in the elderly population,among which central obe-sity(OR=1.502,95％CI:1.368-1.685),urban area(OR=1.298,95％CI:1.228-1.392),and sleep du-ration less than 7 hours(OR=1.367,95％CI:1.232-1.471)were risk factors for comorbidity in the elderly population,and male(OR=0.870,95％CI:0.820-0.961),average daily slow walking time more than 30 mi-nutes(OR=0.623,95％CI:0.470-0.824),low-intensity household chores more than 30 minutes(OR=0.638,95％CI:0.485-0.839),and sleep duration more than 8 hours were protective factors for comorbidity in the elderly population.The CHAID decision tree model showed that central obesity,smoking habit,and short sleep duration were important risk factors for comorbidity in the elderly population.Conclusions There is a close relationship between comorbidity in the elderly population and central obesity,smoking habit,and short sleep time.Decision trees and multiple logistic regression models can complement each other in analyzing the influencing factors of comorbidity in the elderly population.
国家科技期刊平台
NSTL
万方数据
