A comprehensive prognostic tool for older adults: Predicting death, ADL disability, and walking disability...

Area of Science:

• Gerontology
• Public Health
• Biostatistics

Background:

• Clinical and financial decisions for older adults rely on predicting future disability and mortality risks.
• Existing prognostic tools lack comprehensive long-term disability risk assessment for the general older population.

Purpose of the Study:

• To develop a unified prognostic tool predicting mortality, activities of daily living (ADL) disability, and walking disability simultaneously.
• To identify a shared set of variables for predicting these multiple adverse outcomes in older adults.

Main Methods:

• Longitudinal analysis of the Health and Retirement Study (HRS) including 6646 community-dwelling adults aged ≥70 years.
• Development of three optimized prediction models (mortality, ADL disability, walking disability) using Cox and competing-risk regression.
• Evaluation of 40 predictors and validation of model performance using bootstrapping for optimism-corrected statistics.

Main Results:

• The final prognostic tool incorporates 16 variables, demonstrating good discrimination (iAUC: 0.799 for mortality, 0.685 for walking disability, 0.703 for ADL disability).
• Significant variability in disability risk was observed even among individuals with similar mortality risk predictions.
• Model performance remained robust even with one or two missing predictors.

Conclusions:

• The developed prognostic tool provides individualized risks for mortality and disabilities in older adults.
• This tool can inform clinical and financial decision-making by accounting for diverse risk profiles.
• The model is accessible on ePrognosis for practical application.