Abstract
IMPORTANCE Gamification, the application of game design elements such as points and levels in nongame contexts, is often used in digital health interventions, but evidence on its effectiveness is limited. OBJECTIVE To test the effectiveness of a gamification intervention designed using insights from behavioral economics to enhance social incentives within families to increase physical activity. DESIGN, SETTING, AND PARTICIPANTS The Behavioral Economics Framingham Incentive Trial (BE FIT) was a randomized clinical trial with a 12-week intervention period and a 12-week follow-up period. The investigation was a community-based study between December 7, 2015, and August 14, 2016. Participants in the modified intent-to-treat analysis were adults enrolled in the Framingham Heart Study, a long-standing cohort of families. INTERVENTIONS All participants tracked daily step counts using a wearable device or a smartphone, established a baseline, selected a step goal increase, and received daily individual feedback on goal performance by text message or email for 24 weeks. Families in the gamification arm could earn points and progress through levels based on physical activity goal achievement during the 12-week intervention. The game design was meant to enhance collaboration, accountability, and peer support. MAIN OUTCOMES AND MEASURES The primary outcome was the proportion of participant-days that step goals were achieved during the intervention period. Secondary outcomes included the proportion of participant-days that step goals were achieved during the follow-up period and the change in the mean daily steps during the intervention and follow-up periods. RESULTS Among 200 adults comprising 94 families, the mean age was 55.4 years, and 56.0% (n = 112) were female. During the intervention period, participants in the gamification arm achieved step goals on a significantly greater proportion of participant-days (0.53 vs 0.32; adjusted difference, 0.27; 95% CI, 0.20-0.33; P < .001) and had a significantly greater increase in the mean daily steps compared with baseline (1661 vs 636; adjusted difference, 953; 95% CI, 505-1401; P < .001) than the control arm. During the follow-up period, physical activity in the gamification arm declined but remained significantly greater than that in the control arm for the proportion of participant-days achieving step goals (0.44 vs 0.33; adjusted difference, 0.12; 95% CI, 0.05-0.19; P < .001) and the mean daily steps compared with baseline (1385 vs 798; adjusted difference, 494; 95% CI, 170-818; P < .01). CONCLUSIONS AND RELEVANCE Gamification designed to leverage insights from behavioral economics to enhance social incentives significantly increased physical activity among families in the community.
Generated Summary
The study is a randomized clinical trial named the Behavioral Economics Framingham Incentive Trial (BE FIT). It was conducted between December 7, 2015, and August 14, 2016, with a 12-week intervention and a 12-week follow-up. The investigation was community-based study among families in the Framingham Heart Cohort. The study used insights from behavioral economics to enhance social incentives, such as collaboration, accountability, and peer support, to increase physical activity. The primary outcome was the proportion of participant-days that step goals were achieved during the intervention period. Secondary outcomes included the proportion of participant-days that step goals were achieved during the follow-up period and the change in the mean daily steps during the intervention and follow-up periods. Participants tracked daily step counts using a wearable device or a smartphone, established a baseline, selected a step goal increase, and received daily individual feedback on goal performance by text message or email for 24 weeks. Families in the gamification arm could earn points and progress through levels based on physical activity goal achievement during the 12-week intervention.
Key Findings & Statistics
- Among 200 adults comprising 94 families, the mean age was 55.4 years, and 56.0% (n = 112) were female.
- During the intervention period, participants in the gamification arm achieved step goals on a significantly greater proportion of participant-days (0.53 vs 0.32; adjusted difference, 0.27; 95% CI, 0.20-0.33; P < .001) and had a significantly greater increase in the mean daily steps compared with baseline (1661 vs 636; adjusted difference, 953; 95% CI, 505-1401; P < .001) than the control arm.
- During the follow-up period, physical activity in the gamification arm declined but remained significantly greater than that in the control arm for the proportion of participant-days achieving step goals (0.44 vs 0.33; adjusted difference, 0.12; 95% CI, 0.05-0.19; P < .001) and the mean daily steps compared with baseline (1385 vs 798; adjusted difference, 494; 95% CI, 170-818; P < .01).
- The mean (SD) numbers of participant baseline daily steps were 7662 (3378) in the control arm and 7244 (3368) in the gamification arm, which were not significantly different (P = .41).
- The percentage increase in participant step goals from baseline was not significantly different between study arms (P = .74), with a mean (SD) of 2722 (1740) steps in the control arm and 2340 (1153) steps in the gamification arm.
- Almost 40% (79 of 200) of participants selected a custom step goal, and on average these goals were a 49.5% increase from baseline.
- During the intervention period, step data that were missing or had values less than 1000 steps per day represented 12.7% (1090 of 8568 participant-days) of observations in the control arm and 10.1% (835 of 8232 participant-days) of observations in the gamification arm.
- The mean (SD) unadjusted proportions of participant-days that step goals were achieved during the intervention period were 0.32 (0.24) in the control arm and 0.53 (0.29) in the gamification arm.
- The unadjusted mean (SD) steps per day during the intervention period were 8298 (3836) in the control arm and 8905 (3382) in the gamification arm.
- In the main adjusted model, the gamification arm achieved step goals on a significantly greater proportion of participant-days than the control arm during the intervention period (adjusted difference, 0.27; 95% CI, 0.20-0.33; P < .001) and during the follow-up period (adjusted difference, 0.12; 95% CI, 0.05-0.19; P < .001).
- The gamification arm also had a significantly greater change in the mean daily steps than the control arm during the intervention period (adjusted difference, 953; 95% CI, 505-1401; P < .001) and during the follow-up period (adjusted difference, 494; 95% CI, 170-818; P = .003).
Other Important Findings
- The gamification arm received daily feedback for 24 weeks.
- The unadjusted proportions of participant-days that step goals were achieved during the follow-up period were 0.33 (0.20) in the control arm and 0.44 (0.22) in the gamification arm.
- Ninety-seven percent (194 of 200) of participants completed the entire 24-week study.
- Most participants in both study arms had positive perceptions about their experiences in the study, and many stated that they would continue to use their activity tracking devices after the study concluded.
Limitations Noted in the Document
- The study’s participants were members of the Framingham Heart Study, had European ancestry, and needed either a smartphone or a computer, which may limit generalizability.
- The study evaluated physical activity using step counts and did not have data on other measures of physical activity, device wear time, or other health outcomes.
- Future studies could evaluate both changes in step counts and other clinical outcomes over longer periods.
- The study did not test the effect of the intervention in nonfamily networks.
Conclusion
The findings indicate that gamification designed using insights from behavioral economics to enhance social incentives within families significantly increased physical activity during the 12-week intervention. The mean increase from baseline among participants in the gamification arm was approximately 1700 steps, which is almost an additional 1 mile per day. The study suggests that gamification may offer a promising approach to change health behaviors if designed using insights from behavioral economics to enhance social incentives. The results of this trial expand our understanding of using gamification and social incentives to change health behaviors and reveal important implications for the design of future interventions. First, an important element of our study was the use of principles from behavioral economics within the gamification design to address predictable barriers to behavior change. While gamification is used widely across various industries, evidence on its effectiveness is inconsistent, and most study designs have not appropriately incorporated theories from health behavior. Second, although several retrospective studies have identified associations between social networks and health behaviors, there have been few studies evaluating effects prospectively.