JMIR Publications

ABSTRACT

Background:

The success of a smartphone application (app) in changing health behaviour is thought to be contingent on engagement, commonly operationalised as frequency of use.

Objective:

The current study examined user engagement with a 100-day physical activity (PA) intervention delivered via an app to examine rates of engagement, associations between user characteristics and engagement, and whether engagement was related to intervention efficacy.

Methods:

Engagement was captured in a real-time log of user interactions with either a gamified app (n=141) or a non-gamified version of the same app (n=160). PA was assessed via accelerometry and self-report at baseline and three-month follow-up. Survival analysis was used to assess time to churn (non-use attrition). Mixed models examined associations between user characteristics and engagement. Characteristics of super-users (top quartile of users) and regular-users (lowest three quartiles) were compared using t-tests and chi-square analysis. Linear mixed models were used to assess whether being a super-user was related to change in physical activity over time.

Results:

Engagement was generally high. Churn occurred in 32% and 39% of the gamified and basic groups, respectively with no significant between group differences in time to churn (p = .42). Users with a body mass index (BMI) in the healthy range had higher engagement (M= 230.5, 95% CI 190.6-270.5, F2 = 8.67, p < 0.001), compared to users whose BMI was overweight or obese (M= 170.6, 95% CI 139.5-201.6; M= 132.9, 95% CI 104.8-161.0). Older users engaged more (M= 200.4, 95% CI 171.9-228.9, F1 = 6.385, p = .01) than younger users (M= 155.6, 95% CI 128.5-182.6). Super-users were 4.6 years older (t(297) = 3.6, p = <.001) and less likely to have a BMI in the obese range (X2(2) = 15.1, p = <.001). At the three-month follow-up, super-users were completing 28.2 (95% CI 9.4-46.9) more minutes of objectively measured PA than-regular-users (F = 4.76, p = .03).

Conclusions:

Engagement was high across the 100-day intervention period and the inclusion of gamified features enhanced engagement. Participants who engaged the most saw significantly greater increases to their objectively measured PA over time, supporting the theory that intervention exposure is linked to efficacy. Further research is needed to determine whether these findings are replicated in other app-based interventions, including those conducted in real-world settings. Clinical Trial: Australian and New Zealand Clinical Trial Registry, protocol: 12617000113358.