Abstract
Purpose: The purpose of this study was to ascertain whether we could predict falls by the examination of vibration sense using a tuning fork.
Subject: The subjects were 62 community-dwelling elderly outpatients and day-care clients who could walk by themselves, and who had no disorders of cognitive function on the central nervous system.
Methods: We assessed them with an examination of vibration sense using a tuning fork and three performance tests (Timed “Up & Go” test, Modified functional reach test, Timed 10m walk test). In addition, subjects were monitored for six months after those assessments to learn whether they experienced a fall. They were then divided into a “fallers” group and a “non-fallers” group, and the measured values were compared. We calculated the cutoff value with Receiver Operating Characteristic (ROC) curve analysis on each performance test and vibration sense.
Results: There were 22 fallers (35.5%) during the six months, and the results for vibration sense of the non-fallers group were superior to those of the fallers group. The area under the ROC curve of vibration sense was equal to that of performance tests (0.83). The degree of sensitivity of vibration sense to falls was 77% (specificity: 68%, accuracy: 71%) when the cutoff value was estimated as 5.5 seconds.
Conclusion: The results suggested that the examination of vibration sense using a tuning fork is useful for fall prediction.
