Purpose: The purpose of this study was to determine if gait velocity could predict faller status in a group of independently functioning older adults.
Method: A cross-sectional design was used to establish the sensitivity and specificity of gait velocity to predict faller status as measured during the performance of the 30-tt walk test at a preferred and maximum speed. Faller status was based on a 12-month retrospective self-reported fall history. A total of 413 independently functioning older adults (M = 78 years; SD = 6.5) performed the 30-ft walk test at preferred and maximum speed in a single test session.
Results: Binary logistic regression analysis indicated that gait velocity could be used to predict faller status in both the preferred and maximum speed test conditions. Receiver operating characteristic (ROC) analysis determined that a cut-off value of 4.23 ft/s produced the highest sensitivity (79%) combined with specificity (33%) in predicting faller status for the preferred speed condition while a second ROC analysis conducted for the maximum speed condition determined that a cut-off value of 5.42 ft/s produced the highest sensitivity (75%) and specificity (31%) in predicting faller status.
Conclusion: Although the results indicate that gait velocity is a relatively good predictor of who will fall among higher functioning community-dwelling older adults, it is a relatively poor predictor of who will not fall based on the low specificity values obtained for both test conditions. This is perhaps not surprising given the multifactorial nature of falls. Tests that evaluate more dimensions of balance and mobility may be more likely to better predict who will and will not fall, particularly among higher functioning older adults. This conclusion is supported by the fact that the 8-tt up-and-go test (Rikli & Jones, 1999), which combines transfer, turning, and walking skills performed at maximum speed, exhibits both good sensitivity (78%) and specificity (86%) when administered to older adults with a comparable level of physical function (Rose, Jones, & Lucchese, 2002). While gait velocity may be a good predictor of faller status in the case of older adults at lower levels of function, tests that measure more dimensions of balance and mobility should be used to predict who will or will not fall when evaluating older adults at higher levels of function.