Multimorbidity patterns and domain-specific intrinsic capacity impairments among rural older adults: a multiple correspondence analysis and k-means clustering study

Abstract

Abstract Background Multimorbidity is prevalent among older adults, but its specific patterns and differential associations with impairments across intrinsic capacity (IC) domains remain under-explored. This study aimed to identify stable multimorbidity patterns in rural older adults using multiple correspondence analysis (MCA) combined with k-means clustering and to examine their independent associations with domain-specific IC impairments. Methods This cross-sectional study included 3,641 community-dwelling older adults with multimorbidity (≥ 2 of 15 chronic conditions) from rural China. MCA was used to reduce dimensionality, followed by k-means clustering to identify multimorbidity patterns. Multivariable logistic regression, adjusted for age, sex, education, income, and smoking status, was used to estimate the associations between the identified patterns and impairments in each of the six IC domains (cognition, physical function, vitality, psychology, hearing, and vision), with the high metabolic burden pattern used as the reference. Results Five stable multimorbidity patterns were identified: high metabolic burden (36.4%), musculoskeletal degenerative (28.4%), gastrointestinal-tumor risk pattern (13.8%), cardiovascular-heart failure pattern (13.0%) and respiratory pattern (8.4%). Compared with the high metabolic burden pattern, the gastrointestinal-tumor risk pattern was associated with higher odds of vitality impairment (OR = 1.70, 95% CI: 1.30–2.21); the cardiovascular-heart failure pattern was associated with higher odds of physical function impairment (OR = 1.40, 95% CI: 1.11–1.77); and the gastrointestinal-tumor risk pattern, cardiovascular-heart failure pattern, and respiratory pattern were all associated with higher odds of psychological impairment (ORs from 1.61 to 2.27). The respiratory and gastrointestinal-tumor risk patterns were also associated with higher odds of hearing and visual impairments. Notably, the high metabolic burden pattern demonstrated the strongest risk for cognitive impairment. Conclusion Distinct multimorbidity patterns are differentially associated with specific IC domain impairments. These findings support pattern-oriented risk stratification and targeted functional management in older adults, offering a practical framework for precision geriatric care in resource-limited rural settings.