PURPOSE: This study evaluated exercise-related predictors of successful long-term weight control in women by analyzing the extent to which sustained exercise participation and self-determination theory (SDT)-based exercise motivation variables mediated the impact of a behavioral weight control intervention on 3-yr weight change. METHODS: Longitudinal randomized controlled trial consisting of a 1-yr SDT-based intervention and a 2-yr follow-up with 221 female participants (means ± SD: age = 37.6 ± 7 yr, body mass index = 31.6 ± 4.1 kg·m(-2)). The tested model incorporated experimentally manipulated perceived need support, motivational regulations, and 2-yr exercise adherence as mediators of the intervention's impact on 3-yr weight change. Paths were tested using partial least squares analysis. Where there were significant intervening paths, tests of mediation were conducted. RESULTS: Treatment had significant effects on 1- and 2-yr autonomous regulations, 2-yr physical activity, and 3-yr weight change, fully mediated by the tested paths (effect ratio = 0.10-0.61). Moderate and vigorous exercise at 2 yr had a significant effect (P < 0.001) on weight loss success at 3 yr and partially mediated the effect of treatment on weight change. The 2-yr autonomous regulation effects on follow-up weight change were only partially mediated by physical activity (effect ratio = 0.42). CONCLUSIONS: This application of SDT to physical activity and weight management showed that not all types of motivation predict long-term behavioral outcomes and that sustained moderate and vigorous exercise mediated long-term weight change. It provides strong evidence for a link between experimentally increased autonomous motivation and exercise and long-term weight loss maintenance. Results highlight the importance of interventions targeting the internalization of exercise behavioral regulation and making exercise and physical activity positive and meaningful experiences rather than simply focusing on immediate behavior change in overweight/obese women.
Stored only when both papers are in the corpus — 297 citations exist upstream.