An Early Phase Trial Testing the Proof of Concept for a Gamified Smartphone App in Manipulating Automatic Evaluations of Exercise

People who are more physically active tend to have more favorable automatic evaluations of exercise (i.e., nonconscious evaluations based on mental associations between "exercise" and "pleasant" or "unpleasant" that manifest into approach tendencies). Although some interventions have been shown to modify automatic evaluations in lab-based settings, the training regimes may not translate into scalable real-world interventions. The aim of these studies were to (a) test how often people tend to engage with the app in a "real-world" setting, and (b) test whether an app with gamification features and evaluative conditioning strategies change automatic evaluations of exercise versus sedentary behavior. Participants (N = 289, 238 female, M age = 33) were randomly allocated to have access to either Flex Exercise-a game-based app which contained 70% exercise-related content or Flex Control-the same game-based app with no exercise content. Participants completed an Implicit Association Test (IAT) as assessments of automatic evaluations immediately after exposure to Flex and 24 hr later. No significant between-group difference was observed immediately after exposure to Flex for automatic evaluations; however, 1 day following exposure, those in the Flex Exercise condition had significantly more favorable automatic evaluations of exercise than those in the Flex Control condition (d = 0.24). This effect was driven by a change in automatic evaluations, as assessed through the IAT, in the control condition more favorable toward sedentary behavior relative to physical activity and was magnified by user engagement. This mHealth intervention may have inadvertently enhanced sedentary automatic evaluations