医学
优势比
多项式logistic回归
内科学
逻辑回归
人口学
可能性
潜在类模型
物理疗法
儿科
统计
数学
机器学习
社会学
计算机科学
作者
Jiayi Zhou,Melissa Y Wei,Jingyi Zhang,Hua Liu,Chenkai Wu
出处
期刊:Age and Ageing
[Oxford University Press]
日期:2022-08-01
卷期号:51 (8)
标识
DOI:10.1093/ageing/afac177
摘要
Abstract Objective to identify multimorbidity patterns among middle-aged and older adults in China and examine how these patterns are associated with incident disability and recovery of independence. Methods data were from The China Health and Retirement Longitudinal Study. We included 14,613 persons aged ≥45 years. Latent class analysis (LCA) was conducted to identify multimorbidity patterns with clinical meaningfulness. Multinomial logistic models were used to determine the adjusted association between multimorbidity patterns and incident disability and recovery of independence. Results we identified four multimorbidity patterns: ‘low morbidity’ (67.91% of the sample), ‘pulmonary–digestive–rheumatic’ (17.28%), ‘cardiovascular–metabolic–neuro’ (10.77%) and ‘high morbidity’ (4.04%). Compared to the ‘low morbidity’ group, ‘high morbidity’ (OR = 2.63, 95% CI = 1.97–3.51), ‘pulmonary–digestive–rheumatic’ (OR = 1.89, 95% CI = 1.63–2.21) and ‘cardiovascular–metabolic–neuro’ pattern (OR = 1.61, 95% CI = 1.31–1.97) had higher odds of incident disability in adjusted multinomial logistic models. The ‘cardiovascular–metabolic–neuro’ (OR = 0.60, 95% CI = 0.44–0.81), ‘high morbidity’ (OR = 0.68, 95% CI = 0.47–0.98) and ‘pulmonary–digestive–rheumatic’ group (OR = 0.75, 95% CI = 0.60–0.95) had lower odds of recovery from disability than the ‘low morbidity’ group. Among people without disability, the ‘cardiovascular–endocrine–neuro’ pattern was associated with the highest 2-year mortality (OR = 2.42, 95% CI = 1.56–3.72). Conclusions multimorbidity is complex and heterogeneous, but our study demonstrates that clinically meaningful patterns can be obtained using LCA. We highlight four multimorbidity patterns with differential effects on incident disability and recovery from disability. These studies suggest that targeted prevention and treatment approaches are needed for people with multimorbidity.
科研通智能强力驱动
Strongly Powered by AbleSci AI