Fitbit wear-time and patterns of activity in cancer survivors throughout a physical activity intervention and follow-up: Exploratory analysis from a randomised controlled trial.

Objective: There has been growing interest in the use of smart wearable technology to promote physical activity (PA) behaviour change. However, little is known concerning PA patterns throughout an intervention or engagement with trackers. The objective of the study was to explore patterns of Fitbit-measured PA and wear-time over 24-weeks and their relationship to changes in Actigraph-derived moderate-to-vigorous PA (MVPA).

Methods: Twenty-nine intervention participants (88%) from the wearable activity technology and action-planning (WATAAP) trial in colorectal and endometrial cancer survivors accepted a Fitbit friend request from the research team to permit monitoring of Fitbit activity. Daily steps and active minutes were recorded for each participant over the 12-week intervention and throughout the follow-up period to 24-weeks. Accelerometer (GT9X) derived MVPA was assessed at end of intervention (12-weeks) and end of follow-up (24-weeks).

Results: Fitbit wear-time over the 24-weeks of data was remarkably consistent, with median adherence score of 100% for all weeks. During the intervention, participants recorded a median 8006 steps/day. Daily step count was slightly increased through week-13 to week-24 with a median of 8191 steps/day (p = 0.039). Actigraph and Fitbit derived measures were highly correlated but demonstrated poor agreement overall. Fitbit measured activity was closest to MVPA measured using Freedson cut-points as no bias was observed.

Conclusions: Step count was maintained throughout the trial displaying promise for the effectiveness of smart-wearable interventions to reduce sedentary behaviour beyond the intervention period. Further worthwhile work should compare more advanced smart-wearable technology with accelerometers in order to improve agreement and explore less resource-intensive methods to assess PA that could be scalable.