Intrinsic capacity impairment and modifiable associated factors among community-dwelling older adults with chronic...

To examine the prevalence and domain-specific patterns of intrinsic capacity impairment among community-dwelling older adults with chronic diseases from selected urban communities in Nanjing, China; to identify factors associated with any-domain intrinsic capacity impairment; and to explore theory-driven, hypothesis-generating associations among eHealth literacy, sleep disorder, social frailty and intrinsic capacity.

Community-based cross-sectional study.

13 selected urban communities in Nanjing, Jiangsu province, China, from January 2025 to December 2025.

A total of 1063 community-dwelling adults aged 60 years or older with at least one physician-diagnosed chronic disease were included in the final analysis.

The primary outcome was any-domain intrinsic capacity impairment, defined as impairment in at least one of five domains: locomotion, cognition, vitality, psychological capacity and sensory capacity. Secondary outcomes included multidomain intrinsic capacity impairment, domain-specific impairment patterns and sociodemographic, health-related, behavioural and psychosocial factors associated with intrinsic capacity impairment.

The mean intrinsic capacity score was 3.37±1.38. Overall, 746 participants (70.18%) had any-domain intrinsic capacity impairment and 561 (52.78%) had multidomain impairment. Sensory impairment was the most common domain deficit (56.16%), followed by impaired vitality or nutritional risk (32.17%), cognitive impairment (28.69%), impaired psychological capacity (27.19%) and locomotor impairment (18.34%). In the fully adjusted logistic regression model, age older than 70 years (OR 2.15, 95% CI 1.59 to 2.92), living alone (OR 1.53, 95% CI 1.11 to 2.11), having two or more chronic diseases (OR 1.85, 95% CI 1.37 to 2.49), regular exercise (OR 0.56, 95% CI 0.41 to 0.76), low eHealth literacy (OR 1.73, 95% CI 1.28 to 2.35), sleep disorder (OR 2.94, 95% CI 2.12 to 4.07) and social frailty (OR 2.01, 95% CI 1.35 to 3.00) were associated with any-domain intrinsic capacity impairment. The model showed acceptable discrimination, with an area under the receiver operating characteristic curve of 0.774, and the Hosmer-Lemeshow test did not indicate poor calibration (χ²=9.890, p=0.273). In the trimmed non-saturated path model, social frailty showed the strongest direct association with intrinsic capacity, and an indirect association between eHealth literacy and intrinsic capacity through social frailty was observed.

Any-domain intrinsic capacity impairment was common among community-dwelling older adults with chronic diseases from selected urban communities in Nanjing, and more than half of the participants had multidomain impairment. Older age, living alone, multimorbidity, lack of regular exercise, low eHealth literacy, sleep disorder and social frailty were associated with intrinsic capacity impairment. These findings may inform multidomain screening and hypothesis generation for future longitudinal and intervention studies. Causal or temporal relationships should not be inferred from these cross-sectional data.