Lack of accuracy of force platform in predicting falls in older adults

Abstract

Background: Postural instability is related to falls in older adults. One tool for assessing postural instability is the force platform, but there is little information on its accuracy in discriminating fall risk in older adults. Objective: To determine the accuracy of the force platform in predicting falls and discriminating fallers from non-fallers older adults. Method: Was a single-arm cross-sectional study of 96 older adults aged ≥60 years, divided into fallers (single and recurrent) and non-fallers. Center of pressure (COP) sway was evaluated on a force platform under the conditions of eyes open/closed and stable/unstable surface in the variables: anteroposterior(Xavg) and mediolateral (Yavg) COP displacement; total COP displacement (path length); mediolateral(Ysd) and anteroposterior(Xsd) standard deviations of COP; mean COP velocity(Vavg); anteroposterior (Vx) and mediolateral (Vy) COP velocity; 95 % ellipse area. Comparisons between groups were made using one-way ANOVA or the Kuskal-Wallis test and the association between falls and the variables was made using logistic regression. The Receiver Operating Characteristic Curve (ROC curve) was used to determine the accuracy of the force platform. Results: Significant difference was found for path length, Vavg, Vx and Xsd and Vavg, showing that recurrent fallers had greater COP oscillation. There was an association between area, Xsd and Ysd (single fallers) and Xavg (recurrent fallers). The ROC curve showed no accuracy in predicting falls. Conclusions: The force platform was able to distinguish non-fallers from recurrent fallers in some situations and balance variables. Despite this, the study did not find reasonable accuracy of the instrument in predicting falls.