基于聚类分析的社区衰弱老年人多维健康类型及其特征识别OA

Objective:To identify the multidimensional health profiles and related characteristics of community-dwelling(pre)frail older adults,and to provide a scientific basis for stratified and precise health management.Methods:A total of 824 community-dwelling(pre)frail older adults from 17 communities in Shanxi province were recruited between May and September 2025 using a multistage stratified sampling method.Data were collected using a general information questionnaire,the FRAIL Scale,the Katz Activity of Daily Living Scale(ADL),the Lawton Instrumental Activity of Daily Living Scale(IADL),the Short Portable Mental Status Questionnaire(SPMSQ),the 5-item Geriatric Depression Scale(GDS-5),the University of California Los Angeles Loneliness Scale-8(ULS-8),and the Lubben Social Network Scale-6(LSNS-6).K-medoids cluster analysis was performed to identify multidimensional health profiles,and the demographic and health-related characteristics among different profile groups were compared.Results:Four distinct multidimensional health profiles were identified among community-dwelling(pre)frail older adults:the comprehensive health profile(n=207,25.1%),the cognitive-loneliness profile(n=222,26.9%),the general stability profile(n=328,39.8%),and the multiple impairment profile(n=67,8.1%).Significant differences were observed among the four profiles in age,gender,educational level,marital status,living arrangement,subjective economic status and number of chronic diseases(P<0.05).Conclusions:The multidimensional health profiles identified through cluster analysis demonstrated substantial heterogeneity among community-dwelling(pre)frail older adults.Profile-based management tailored to the characteristics of each subgroup may facilitate the transformation of health management from a homogeneous approach to a more refined and individualized model,thereby improving management efficiency and promoting healthy aging.