Validation of Self-Reported Gait Speed and Grip Strength Measures With Objective Measures in Older Adults With Mild Frailty in England.

Background: Clinical research often relies on in-person assessments, which can limit participation among frailer older adults who have difficulty attending clinical settings. Developing valid self-reported alternatives could improve inclusivity. We aimed to validate self-reported gait speed and grip strength against objective measures in adults aged ≥ 65 with mild frailty participating in the HomeHealth trial. Methods: Participants with paired subjective and objective gait speed and/or grip strength measures were included. Internal consistency was assessed using Cronbach's alpha for gait speed (4 items) and grip strength (2 items), stratified by sex. Criterion validity was evaluated using sensitivity and specificity of self-reported measures to identify frailty (Fried phenotype thresholds). Construct validity was examined through regression analyses with instrumental activities of daily living and quality of life outcomes. Results: A total of 254 participants were included (161 females, 93 males). Sensitivity and specificity for detecting slow gait speed were 65.5% and 65.8% in men and 79.5% and 59.0% in women. For weak grip strength, sensitivity and specificity were 23.0% and 66.7% in men and 49.3% and 93.3% in women. Internal consistency was low ( Conclusions: Self-reported gait speed and grip strength measures demonstrated limited validity compared with objective assessments in older adults with mild frailty and should be used cautiously as substitutes for measuring weakness and slowness in this population.