Catherine Rochefort: Risk Factors Predicting Exercise Avoidance: Comparisons across a Two-Part Exercise Outcome

Nearly 25% of US adults report engaging in no regular exercise at all. To date, there are no data identifying risk factors for exercise avoidance. Using cross-sectional data from Amazon’s MTurk and a student participant pool (N=1277), we examined potential risk factors for exercise avoidance. We modeled physical activity as a two-part outcome: exercise avoidance (engagement in 0 minutes of exercise) and exercise amount (non-zero weekly exercise minutes). We conducted bivariate logistic regressions to identify predictors of exercise avoidance and then a multivariate model to identify predictors contributing unique variance. To examine whether predictors are unique to exercise avoidance, we examined associations with exercise amount in a multivariate gamma regression. In bivariate models, age (p=.02), enjoyment (p<.001), self-efficacy (p<.001), mindfulness (p=.01), conscientiousness (p<.001), and neuroticism (p=.02) predicted exercise avoidance. In the multivariate model, age (p=.02), enjoyment (p<.001), self-efficacy (p<.001), and conscientiousness (p=.03) predicted unique variance in exercise avoidance. All predictors except conscientiousness were associated with exercise amount (ps<.001) but with more modest effects. We identified several risk factors for exercise avoidance. In most cases, the risk factors had a more meaningful effect on exercise avoidance than exercise amount