Linking Individual Motives to the Type of Exercise and Sport Activity: Toward Recommendations for Optimal Activity Matching Through a Machine Learning Approach.

Abstract

Background: To effectively promote exercise and sport behavior, it is often emphasized that individual motives should be more strongly considered. The assumption is that people are more likely to maintain an activity if their motives are satisfied. This study investigates how individual motives relate to different types of exercise and sport activities, aiming to improve the empirical basis for tailored recommendations.

Methods: 20,613 adults (Mage = 36.37 y, 67.74% women) completed a 1-time survey. Using a machine learning approach, associations between 7 motives (eg, social contact, stress reduction), sociodemographic variables (eg, sex), and weekly exercise volume were analyzed as predictors of 10 categories of exercise and sport activities (eg, team sports, group-oriented fitness activities).

Results: Overall, the motives of social contact, aesthetics, and fitness/health, along with age, weekly volume of exercise and sport, and sex, emerged as the strongest predictors. However, a closer look reveals distinct combinations of variables associated with participation in each category of activities. For example, team sports were mainly chosen by younger, highly active men who score high in social contact and competition/performance and low on aesthetics.

Conclusions: The findings pave the way for empirically grounded, tailored recommendations that align activity types with individuals' motives and sociodemographic characteristics. When integrated into counseling, such recommendations may enhance long-term adherence by focusing more on personal motivation.