2023 Volume 2 Issue 1 Pages 10-17
Objective: This longitudinal observational study aimed to investigate the predictive accuracy of falls based on walking ability and fall-related self-efficacy in community-dwelling older people.
Methods: The participants were 280 people aged ≥65 years, who could perform their activities of daily living indepen dently. The maximum and comfortable walking speeds and Short Falls Efficacy Scale International (Short FES-I) scores were assessed in the baseline survey, and the walking reserve was estimated using the walking speed recorded at baseline. Furthermore, the number of comorbidities, number of prescribed medicines, instrumental activities of daily living, and history of falls were investigated as covariates. The occurrence of falls was investigated in a follow-up survey after 1 year. The factors associated with the incidence of falls were analyzed using decision tree analysis (Classification and Regression Trees algorithm).
Results: The results of the decision tree analysis revealed that along with history of falls, number of comorbidities, and age, both walking reserve and Short FES-I scores were significantly associated with the occurrence of falls. The discrimination accuracy for the occurrence of falls based on these associated factors showed 62.5% sensitivity and 92.0% specificity.
Conclusions: Assessment of both walking ability and fall-related self-efficacy was suggested to contribute to the accurate discrimination of the risk of falls.
