Changing the Physical Activity Behavior of Adults With Fitness Trackers: A Systematic Review and Meta-Analysis.

Abstract

To examine whether a fitness tracker (FT) intervention changes physical activity (PA) behavior compared to a control condition or compared to an alternative intervention. Searches between January 01, 2010, and January 01, 2019, were conducted in PubMed, CINAHL, Cochrane CENTRAL, EMBASE, and PsycINFO. Randomized clinical trials of adults using an FT to change PA behavior were included. Nonclinical trials, studies that included the delivery of structured exercise, and/or studies that only used the FT to assess PA were excluded. Extracted features included characteristics of the study population, intervention components, PA outcomes, and results. Papers were pooled in a statistical meta-analysis using a fixed effects model. Where statistical pooling was not possible, standardized mean difference (SMD) and 95% confidence intervals (CI) were calculated. Findings were presented in a narrative form and tables. Of 2076 articles found, 21 were included in the review. A small yet significant positive effect (SMD = 0.25, 95% CI = 0.17-0.32; P < .01; I2 = 56.9%; P = .03) was found in step count for interventions compared to control. A small yet significant negative effect (SMD = -0.11, 95% CI = -0.20 to -0.02; P = .02; I2 = 58.2%; P = 0.03) was found in moderate-to-vigorous PA for interventions compared to an alternative intervention. Trackers may enhance PA interventions, as a general positive effect is found in step count compared to a control. However, there is no evidence of a positive effect when interventions are compared to an alternative intervention. It is unknown whether results are due to other intervention components and/or clinical heterogeneity.