Discrimination of standing postures between young and elderly people based on center of pressure

Abstract

Posturography is utilized to assess the influence of aging on postural control. Although this measurement is advantageous for finding group-level differences between the young and the elderly, it is unclear whether it has the potential to differentiate elderly individuals who are affected by various impacts of aging. The purpose of this study was to determine the utility of posturography to discriminate elderly individuals from young adults. We investigated the performances of the random forest classifiers constructed from center of pressure (COP) indices for discriminating standing postures between healthy elderly and young people. Postural sways in 19 young and 31 community-dwelling elderly participants were measured using force plates in 4 standing conditions: bipedal standing, standing on a narrow base, standing on foam rubber, and standing with eyes closed. We further verified the informative predictors that contributed to the prediction model. As the results, the classifier based on the COP indices for standing on foam rubber showed the best performance (accuracy: 93.4%, sensitivity: 94.4%, specificity: 93.6%, area under the curve of receiving operator characteristics: 0.95), followed by the classifier for standing with eyes closed. The informative predictors varied depending on the postural conditions. Our findings demonstrated the potential of posturography for identifying elderly postures. The evaluation of sensory re-weighting using the appropriate COP indices would be a useful clinical tool for detecting the progress of aging on postural control.