Gait quality and velocity influences activity tracker accuracy in individuals post-stroke

Objective: Fitbit Zip™ (FBZ) is a low-cost user-friendly device that could help motivate individuals post-stroke to increase their physical activity. However, inaccuracy in step counts from altered gait variables could cause dissatisfaction and reduce compliance. The aim of this study was to determine the influence of gait variables on the concurrent validity of the FBZ. Method: In a cross-sectional study, 19 community-dwelling stroke survivors (mean 66 (SD 8) years)  wore a FBZ at the non-paretic hip, and were videoed completing a six minute walk test (6MWT). Steps recorded by the FBZ were compared against the criterion standard of manually counted steps from video-recordings. Spatio-temporal gait parameters were calculated, and gait quality was assessed using the Wisconsin Gait Analysis (WGA) tool. Concurrent validity of FBZ was determined using Kendall's Tau correlation coefficient. Linear regression analysis determined the association between gait parameters, quality, and FBZ accuracy. Results: A very strong correlation was observed between the FBZ steps and manual counting (τ = 0.80, p < .001). Step difference demonstrated significant negative association with gait velocity (R² = 0.35, B = -59.94, p = .007), and positive association with WGA score (R² = 0.69, B = 4.59, p < .001), indicating poor FBZ accuracy in participants with lower speed (≤0.8m/s) and poor gait quality (WGA score>16). Conclusions: FBZ is an accurate measure of step activity in independent ambulators with stroke walking at speeds > 0.8m/s, but accuracy can be compromised with lower speed and poor gait quality. Clinicians should consider gait velocity and quality before advising FBZ as a motivational tool to increase physical activity in individuals post-stroke.