A nomogram for predicting disability-free survival in older adults over 15 years: A population-based cohort study

This article aimed to develop and validate a simple-to-use nomogram to predict 15-year disability-free survival among older adults.

A cohort of 1878 disability-free participants aged ≥60 was followed for 15 years. Participants were randomly divided into a training cohort for nomogram development (n = 1314 [70 %]) and validation cohort to confirm the model's performance (n = 564 [30 %]). Information on socio-demographic, lifestyle factors, the Life Satisfaction Index A (LSI-A), chronic diseases, and Mini-Mental State Examination (MMSE) score, and biomarkers were collected through interviews, clinical and neuropsychological examinations, and medical records. Disability-free survival was defined as survival in the absence of dementia and physical disability, and the composite endpoint is first occurrence of events of death, dementia and physical disability. We developed a nomogram summing the number of risk points corresponding to weighted covariates to predict disability-free survival. Validation of the nomogram using C statistic, calibration plots, and Kaplan-Meier curves.

In the multivariate-adjusted model, factors associated with composite end point were younger age, high MMSE (hazard ratio [HR], 0.93; [95 % CI, 0.87-0.99]), high LSI-A (0.78, [0.64-0.97]), non-smoking (0.74, [0.59-0.94]), engagement in physical leisure activity (0.62, [0.48-0.78]), and absence of chronic diseases (0.78, [0.66-0.91]). Incorporating these 6 factors, the nomogram achieved C-statistics of 0.78 (95 % CI, 0.75-0.81) and 0.77 (95 % CI, 0.74-0.80) in predicting disability-free survival in the training and validation cohorts, respectively, and had good calibration curves.

The nomogram was able to predict long-term of disability-free survival and performed well on internal validation, and may be considered for use in effective surveillance, promote, management of clinical and public health ageing.