Reactive agility and motor-cognitive assessments for distinguishing community-dwelling older adults with and without a history of falls: a dual-centre cross-sectional study

Although motor and cognitive functions are closely interconnected in daily life, they are typically assessed separately in fall assessments. This dual-center cross-sectional, observational study examined the discriminative performance of agility and motor-cognitive assessments to distinguish between community-dwelling older adults with and without a history of falls, compared with established motor and cognitive tests. A total of 231 participants (103 with a history of falls; age 71.9 ± 7.5 years) completed a SKILLCOURT test battery, including the Random Star Run (unplanned, reactive changes of direction; agility) and motor-cognitive stepping tasks involving reactive responses while standing to simple and choice reaction, task-switching, 2-back, and Stroop word-color stimuli. Motor tests included the Timed-Up-and-Go (TUG), Sit-to-Stand, and walking speed, each performed under both single- and dual-task conditions. Seated PC-based cognitive tests assessed reaction speed and executive functions corresponding to the SKILLCOURT tasks. Age- and study location-adjusted general linear models identified significant group effects after Holm adjustment for the Random Star Run (η 2 p = 0.04, p = .050) and the simple stepping reaction task (η 2 p = 0.03, p = .048), with faster completion and response times for participants without a history of falls, whereas established assessments showed no relevant group differences (η 2 p = 0.001–0.013, p > .05). Z-standardised adjusted logistic regression analyses revealed significant associations between Random Star Run performance and fall history (Odds Ratio; OR = 1.82 per 1 SD increase, p = .009), with longer completion times associated with higher odds of having a history of falls. The adjusted model yielded an AUC value of 0.73, with 70% sensitivity and 71% specificity, and showed significantly higher discriminative performance than the established functional mobility measure, the TUG (AUC = 0.63, p = .017). Similarly, the SKILLCOURT simple reaction task was associated with fall history (OR = 1.58, p = .009) and showed an AUC value of 0.72, with 63% sensitivity and 75% specificity, but did not demonstrate superior discriminative performance compared with the corresponding PC-based simple reaction test (AUC = 0.63, p > .05). These findings suggest that agility-based measures may provide additional value beyond conventional approaches for distinguishing older adults with and without a history of falls. Prospective studies are warranted to determine whether incorporating such measures can improve the prediction of future falls beyond established motor and cognitive assessments. Trial registration: DRKS00034524 (July 8, 2024).