Incremental predictive value of intrinsic capacity and environmental characteristics in the risk prediction of incident disability.

Objective: To examine the incremental value of intrinsic capacity (IC) and environmental characteristics in the risk prediction of disability. Method: Secondary analysis was performed on a longitudinal sample of individuals aged 50 years or above. The selected subsample was ambulant and cognitively intact, and did not have any disabilities in instrumental activities of daily living (IADL) at baseline. A set of 18 indicators were first used to assess conditions associated with declines in IC and environmental characteristics. Participants were then followed up for approximately one year, and the IADL status (i.e., disabled or not) was treated as the outcome variable in the logistic regression models. The incremental predictive value of IC was examined by comparing the baseline model that only included traditional risk factors (e.g., health conditions and lifestyle factors), against the full model that also included the aforementioned 18 indicators. The comparison was performed using the change in area under the receiver operating characteristic curve (ROCAUC) and the continuous net reclassification index (NRI). Results: Among 10,993 participants (mean age = 73.3, 82.1 % women), 680 (6.2 %) developed disability during the concerned period. The full model significantly outperformed the baseline model, with the ROCAUC improving from 0.707 to 0.729 (change = 0.021; 95 % CI: 0.013-0.030). The continuous NRI was 0.361 (95 % bootstrap CI: 0.280-0.450). Conclusions: Measurements of IC and environmental characteristics have a significant incremental value in predicting disability. In practice, the full model can be implemented as a calculator for identifying older populations at risk of disability in the community settings.