Objective mobility tests may streamline fall-risk screening in older adults
image: ROC curves of four machine learning classifiers (logistic regression, support vector machine, random forest, and artificial neural network) across six feature sets (DGI, CSQ, MTC, DGI + CSQ, DGI + MTC, and All), showing that objective mobility tests achieved the strongest single-domain discrimination for fall history classification (AUC up to 0.89).
Credit: HIGHER EDUCATION PRESS
Falls are a major health threat to older adults, often leading to injury, loss of independence, and rising healthcare burden. Because a previous fall is one of the strongest predictors of future falls, identifying older adults with a history of falls is an important first step in practical risk screening. Yet clinicians still face a core question: which assessment information is most useful, and which combination is efficient enough for real-world use?
In a new study published in Healthcare and Rehabilitation, researchers compared three commonly used domains in fall-risk assessment: demographic information, clinical scales and questionnaires, and mobility tests with contextual factors. Using data from 163 older adults in the public G-STRIDE dataset, the team evaluated multiple machine learning models to determine which assessment strategy best distinguished fallers from non-fallers.
The results showed that mobility tests and contextual variables provided the strongest single-domain performance. Objective mobility assessment achieved an area under the receiver operating characteristic curve of 0.89, clearly outperforming demographic information alone. When mobility measures were combined with basic demographic data, performance remained strong, while adding more questionnaire-based assessments provided only limited incremental benefit.
The researchers also identified a compact seven-variable subset that preserved nearly all of the predictive power of the full clinical feature pool. Most of these selected variables were mobility-related, reinforcing the idea that practical screening may be possible without relying on extensive questionnaire-based assessments.
These findings support the development of efficient fall-risk assessment tools centered on objective mobility evaluation, with basic demographic information used as a complement. Such streamlined strategies may help reduce assessment burden while maintaining clinically meaningful discrimination in rehabilitation and geriatric care settings. The work titled “Comparative study of demographic information, clinical scales and questionnaires, and mobility tests for fall risk assessment in older adults” was published online in Healthcare and Rehabilitation (available online on Mar. 27, 2026).