Background: Participation in regular physical activity is important for improving health, but sedentary behavior is difficult to change. One option is to provide feedback on physical activity with wearable motion sensing technologies (activity devices). This review sets out to synthesize the literature on the effectiveness of these devices for physical activity, weight, and patient satisfaction outcomes, and to describe moderating factors that may impact effectiveness (ie, population characteristics, location where device is worn on body, or device role in overall intervention approach).
Methods: We searched MEDLINE, Embase, CINAHL, SPORTDiscus, and Cochrane CENTRAL from January 1, 2000, to January 6, 2015, for peer-reviewed, English-language randomized controlled trials among adults (≥18 years of age). Article inclusion, data abstraction, and quality assessment were conducted through a duplicate process, with discussion to resolve discrepancies. Trial quality was evaluated as low, unclear, or high risk of bias (ROB). Strength of evidence (SOE) was summarized as high, moderate, or low. Random-effects models were used to produce standardized mean differences (SMDs) for physical activity outcomes and mean differences (MDs) for weight outcomes. Heterogeneity was measured with I2. Qualitative synthesis was conducted for outcomes with <3 studies.
Results: We identified 4787 unique citations; 14 trials met eligibility criteria. Women comprised 62.5% of the population. Median age was 49.7 years (range 28.7 to 79.8 years). Study sizes ranged from 20 to 544 participants (median 62), with the majority of studies (n=8) randomizing <70 participants. Although all of the interventions had multiple components, in the majority of studies (n=8), the wearable device was used in a major role (ie, central motivational enhancement). The device was an accelerometer in all 14 studies.
Twelve trials (2 at low ROB, 2 at unclear ROB, 8 at high ROB) examined accelerometer interventions for increasing physical activity; the majority (n=9) used an inactive comparator. Overall, a small significant effect was found for increasing physical activity (SMD 0.26; 95% CI 0.04 to 0.49) with high heterogeneity (I2=64.7%). Moderate SOE was found for small increases in physical activity when compared with an inactive comparator (SMD 0.29; 95% CI 0.03 to 0.55) with high heterogeneity (I2=70.3%). Low SOE and no statistically significant effect (SMD 0.17; 95% CI -1.09 to 1.43) were found when compared with an active comparator.
Eleven trials (2 at low ROB, 3 at unclear ROB, 6 at high ROB) examined the effect of accelerometer interventions on weight loss or maintenance. The overall pooled estimate showed a small significant effect for weight loss (MD -1.65 kg; 95% CI -3.03 to -0.28) with high heterogeneity (I2=81%). Moderate SOE and no significant effect were found for accelerometers versus inactive comparators (MD -1.44 kg; 95% CI -3.08 to 0.19). A positive trend with low SOE for accelerometers was found in 2 trials on weight loss, but only one was statistically significant.
No studies reported the outcome of patient satisfaction with healthcare. Also, no moderating factors were found to significantly impact effectiveness or explain heterogeneity.
Conclusions: The small positive effects produced by interventions that include accelerometers may not result in a clinically significant impact on physical activity or weight loss; however, the small sample sizes with moderate to high heterogeneity in the current studies limit the conclusions that may be drawn. Larger, well-designed randomized controlled trials are needed. Clinicians and policymakers should consider these findings and the existing gaps in the literature before widespread use of these technologies.