The present study, involving 23 elderly in need of nursing care, examined the reproducibility and validity of walking at the optimum and maximum speeds. The elderly were asked to walk at their optimum and maximum speeds twice each, and the reproducibility was calculated using inter-correlation coefficients. To evaluate the validity, correlation coefficients between walking and the FRT and Frail CS-10 (10-Second Chair Stand) tests were calculated. When the walking speed was optimum and maximum, the reproducibilities for walking-related factors such as the walking speed and rate were high, whereas that for the LNG (Length of Total COP) at the optimum speed was low, at 0.67. Both at the optimum and maximum speeds, marked correlations were noted between a number of walking-related factors and the FRT and Frail CS-10 scores. However, when the walking speed was optimum, there was no significant correlation between the stride width or LNG and Frail CS-10 scores. The results suggest that although the reproducibility and validity of both cases: walking at the optimum and maximum speeds, have been confirmed, the latter is superior as a test to evaluate the elderly's ability to maintain balance.
