A systematic review of randomized controlled trials (RCTs) of exercise interventions using digital activity trackers (E-DAT) in cancer patients

108 Background: Exercise ameliorates cancer and treatment-related toxicities but adherence to exercise is a barrier. Few studies in the general population suggest that E-DAT may improve exercise adherence. We conducted a systematic review to examine the effects of E-DAT on physical function, health-related quality of life (HRQOL), and serum inflammatory markers during and after cancer treatment. Methods: We used PubMed, Embase, and the Cochrane Library to identify RCTs of E-DAT in cancer patients aged ≥18 years published in English between 1/1/2008 and 7/27/2017. Two authors independently reviewed the titles of articles from the search (n = 160), removed duplicates (n = 49), and reviewed the remaining 111 articles for eligibility and substantive results. We excluded RCTs that used digital activity trackers solely for data collection. Results: Twelve RCTs met eligibility criteria, including 1,450 patients (mean age: 50-70 years) with the following cancers: breast (n = 5 RCTs), colon or breast (n = 2), prostate (n = 1), acute leukemia (n = 1) and various type (n = 3). Duration of E-DAT ranged from 4-24 weeks. Follow-up period was 4-24 weeks with retention of 54-95%. Half of the RCTs had E-DAT with in-person exercise training and the rest had self-directed training. The technology component of E-DAT included pedometers (n = 8); pedometers with smart phone application (n = 1), Wii-Fit (n = 1), or heart rate monitor (n = 1); and a wireless sensor with accelerometer, gyroscope, and magnetometer (n = 1). Adherence to E-DAT was > 70% in 5 of 7 RCTs. Compared to controls (exercise interventions without digital activity trackers or standard of care), E-DAT significantly improved the step count in 60% of 5 RCTs, activity level in 56% of 9 RCTs, and HRQOL in 56% of 9 RCTs (all p < 0.05), with no significant changes of inflammatory markers (i.e. TNF, CRP, c-peptide) in 2 RCTs. No significant correlations were found between duration of E-DAT with adherence (Spearman’s r = 0.16; p = 0.75) or study retention (Spearman’s r = -0.42; p = 0.17). Conclusions: This systematic review shows that E-DAT is feasible to implement in cancer patients. Future research should examine the optimal type, dose and schedule of E-DAT.