Search

[Purpose] We investigated the usefulness of assessment tools used in fall prevention classroom. [Subjects] The subjects were 33 community-dwelling elderly women. [Method] Assessment tools for measurement were selected based on motor function, stride time variability, gait parameters, such as root mean square (RMS), and level of daily activities. In univariate analysis, stride time variability was found to be significantly greater for fallers than for non-fallers. The cut-off value and the area under the curve (AUC) were determined from the receiver-operating-characteristic (ROC). [Results] Significant differences were observed between the fallers group and no-fallers group in stride time variability and level of daily activities. Logistic regression analysis showed stride time variability and level of daily activities were significantly associated with falls. AUC of stride time variability was 0.79, and that of the level of daily activities was 0.20. The degree of sensitivity of stride time variability for falls was 69% (specificity: 85%, accuracy: 79%) when the cut-off value was estimated to be 2.26%. [Conclusion] Our study suggests that stride time variability is a useful assessment tool for predicting falls. [ABSTRACT FROM AUTHOR]