Psychometric properties of 3-meter backward walk test (3MBWT) in people with Parkinson disease.

Background and purpose: The purpose of this study was to investigate the 3-meter backward walk test (3MBWT) in individuals with Parkinson Disease (PD) to determine the following: (1) concurrent validity with other gait velocity measures and (2) interrater and intrarater reliability of in-person and video assessment.

Methods: A convenience sample of 25 people with PD participated. Forward gait velocity was measured using a computerized walkway (Zeno Walkway System), the 10-meter walk test (10MWT), and 4-meter walk test (4MWT). Backward gait velocity was measured using the 3MBWT. Concurrent validity was assessed using Pearson's correlations. Reliability was assessed using intra-class correlation coefficients (ICC_(2,1)).

Results: All relationships between the 3MBWT and gait outcome measures were significant. The 3MBWT demonstrated strong correlations with the 4MWT dual task (r = .795, p=<.001) and moderate correlations with 4MWT comfortable walking speed (r = .658, p < .001), 4MWT fast walking speed (r = .601,p=.002), 10MWT comfortable walking speed (r = .512, p = .009), and 10MWT dual task (r = .535, p = .006). A low yet significant correlation was noted with the 10MWT fast walking speed (r = .398, p = .049). Association between the 3MBWT and the Zeno Walkway System revealed moderate correlations. All reliability tests were significant at p < .001. Interrater reliability ICC_(2,1) values were very high for 3MBWT (ICC_(2,1) = 0.93, [0.83-0.91]). Intrarater reliability was also very high (ICC_(2,1) = 0.96 [0.90-0.98]).

Conclusion: The 3MBWT demonstrates validity and reliability as a tool for assessing gait speed in the posterior direction in people with PD.