Active without thinking? Distinguishing between automatic responses as predictors of physical activity

Background . Regular physical activity (PA) improves both physical and mental health, yet most people are insufficiently active. Identifying modifiable determinants of PA can aid development of effective PA promotion interventions. It is well recognised that PA is at least partly regulated by automatic processes, which capture simple associations that trigger behaviour without conscious thought. Yet, few attempts have been made to differentiate the roles of specific automatic processes on different forms of PA. This study aimed to model the unique effects of three automaticity variables – i.e., automatic valuations, approach-avoidance tendencies, and habit strength – on engagement in sport, leisure-based and total PA.  Methods . A cross-sectional survey design was used. A sample of 226 individuals (mean age 23.70 years, SD = 5.43, range = 17-61) self-reported engagement in sport-based PA, leisure-time PA, and active travel. They also reported habit strength for sport and for leisure-time activity, and completed response-time tasks capturing automatic valuations and approach tendencies towards PA. Correlation analyses were used to assess the magnitude of relationships between all variables, and regression analyses to model associations between automaticity variables and sport and leisure-time PA, respectively.  Results . Automatic valuations, approach tendencies, and sport habit strength all independently and positively predicted sport-based PA and total PA engagement. However, leisure-time PA was predicted only by leisure activity habit strength, not automatic valuations or approach-avoidance. Leisure-time habit strength was the only predictor of a composite PA engagement measure combining total time in sport, leisure-time activity and active travel.  Discussion . Study limitations, including over-representation of females, and people who were highly active, demand that caution is exercised when extrapolating from our findings, and our study warrants replication using more rigorous methods. Nonetheless, our results preliminarily suggest that promoting positive automatic valuations, approach tendencies, and habit strength could encourage engagement in sport-based PA, whereas interventions to increase PA more generally might more fruitfully emphasis PA habit formation.