Preliminary concurrent validity of the Fitbit-Zip and ActiGraph activity monitors for measuring steps in people with polymyalgia rheumatica

Background
Activity monitors provide objective measurements of physical activity, however, the accuracy of these devices in people with polymyalgia rheumatica (PMR) is unknown. Therefore, this study aimed to obtain preliminary evidence of the accuracy of two activity monitors and explore if clinical and gait-related factors altered device accuracy in people with PMR.

Methods
The ActiGraph with low frequency extension (+LFE) and standard (-LFE) algorithms, Fitbit-Zip (waist) and Fitbit-Zip (shirt) were concurrently tested using a two-minute walk test (2MWT) and stairs test in 27 people with PMR currently treated with prednisolone. To determine accuracy, activity monitor step-count was compared to a gold-standard step-count (GSSC; calculated from video recording) using Bland-Altman plots.

Results
The Fitbit-Zip (waist) achieved closest agreement to the GSSC for the 2MWT (mean bias (95%CI): 10 (-3, 23); 95%LOA: −55, 74). The ActiGraph (+LFE) achieved closest agreement to the GSSC for the stairs test (mean bias (95%CI): 0 (-1, 1); 95%LOA: −5, 5). The ActiGraph (-LFE) performed poorly in both tests. All devices demonstrated reduced accuracy in participants with lower gait velocity, reduced stride length, longer double-limb support phase and greater self-reported functional impairment.

Conclusion
Our preliminary results suggest that in controlled conditions, the Fitbit-Zip fairly accurately measures step-count during walking in people with PMR receiving treatment. However, device error was greater than data published in healthy people. The ActiGraph may not be recommended without activation of the LFE. We identified clinical and gait-related factors associated with higher levels of functional impairment that reduced device accuracy. Further work is required to evaluate the validity of the activity monitors in field conditions.