Background: The last decade has seen a substantial increase in the use of mobile health apps and research into the effects of those apps on health and health behaviour. In parallel, research has aimed at identifying population subgroups that are more likely to use those health apps. There are two major limitations to the present evidence base. First, research into mobile health apps has focused on vary broad health apps or health apps for behavioral categories. No research has investigated which population subgroups are more likely to use apps for a specific health behaviour. In this study, we focused on walking apps. Second, research has tended to focus on health app users versus non-users: little research effort has been directed at subgroups of non-users, including populations who previously used apps but decided not to use them anymore. Objective: We aimed to provide profile distributions of current users, previous users, and informed non-users, and to identify predictor variables relevant for profile classification. Methods: Data were available from 1683 participants who were participants of a national walking event in September 2017. They provided information on demographics, walking behaviour and walking app usage, and items from User Acceptance of Information Technology, in an online survey. Data were analyzed using discriminant function analyses and multinominal logistic regression analysis. Results: The majority was a current walking app user (n = 899, 53.4%) and only a small part was a previous walking app user (n = 121, 7.2%). Current walking app users were more likely to walk on at least five days per week and for at least 30 minutes for bout (OR = 1.4, 95% CI [1.1; 1.9], P = .005), but also to be more likely to be overweight, OR = 1.7, 95% CI [1.2 - 2.4], P = .001, or obese, OR = 1.4, 95% CI [1.1; 1.9], P = .005, as compared to informed non-users. Further, current walking app users perceived the app to be less boring, to be easier to ease and to retrieve information, and to be more helpful to obtain their goal. Effect sizes ranged from 0.10, 95% CI [0.08 – 0.30] to 1.58, 95% CI [1.47 – 1.70]. Conclusions: Results demonstrated the usefulness to focus on behaviour-specific apps and on subgroups of non-users. The findings provide indications for health practitioners to stimulate walking app usage. For app-developers, findings indicate suggestions to prevent people to stop using (walking) apps.
