Autonomous motivation, self-efficacy and identity predict MVPA maintenance in a weight loss trial

Background 
Physical activity (PA) is a key predictor of successful weight loss maintenance. However, psychosocial predictors of PA initiation and maintenance in adults with overweight or obesity (OW/OB) remain unclear. This study used a data-driven approach to define device-measured PA initiation and maintenance and examined theory-based predictors of PA, including three promising predictors of maintenance within a behavioral weight loss program.

Methods 
This secondary analysis included adults with OW/OB (age 18-60 years, BMI 27-46 kg/m2) without diabetes or cardiovascular disease enrolled in a 12-month weight loss trial. Participants were randomized to 4:3 intermittent fasting or daily caloric restriction (weekly target energy deficit: 34%). Both groups received the same PA prescription (300 min/week of moderate-to-vigorous physical activity [MVPA]), behavioral support, and fitness center access. MVPA was estimated over 7 days with the activPALTM micro at months 0, 3, 6, and 12, along with validated measures of self-efficacy, autonomous motivation, outcome expectancies and exercise identity. At month 12, three yes/no items assessed weight loss satisfaction, whether PA became habitual, and whether it became part of participants’ identity. Random intercept mixed models were used to determine if MVPA and predictors changed during the trial. Three changepoint models tested linear MVPA change: (1) 0–12 months, (2) slope change at 3 months, and (3) slope change at 6 months. The best-fitting model (lowest -2LL with significant changepoint) defined initiation and maintenance phases. Predictors of initiation and maintenance were tested via four linear regressions: (1) baseline factors predicting initiation (changepoint MVPA) controlling for baseline MVPA, (2) baseline predictors of maintenance (12-month MVPA controlling for changepoint MVPA), (3) 3-month predictors of maintenance, and (4) 6-month predictors of maintenance. Two-sample t-tests compared 0–12 month MVPA change by yes/no responses to satisfaction, habit, and identity items.

Results 
The analytic sample included n=162 participants who had at least one valid measure of PA (mean±SD: age 42±9, BMI 34±4, 74.1% female, 86% White). MVPA, autonomous motivation, exercise identity (p’s<0.001), and outcome expectancies (p=0.047) increased from 0-12 months, while self-efficacy declined (p<0.001). Mean MVPA increase was 29% (95%CI: 15, 46%). The model with slope change at 3 months was selected, defining MVPA initiation as 0-3 months, and maintenance as 3-12 months. None of the psychosocial factors were associated with MVPA initiation. Baseline autonomous motivation was associated with MVPA maintenance (p=0.017). Self-efficacy at 3 months (p=0.001) and 6-months (p=0.047) and exercise identity at 3 months (p=0.037) and 6 months (p<0.001) were associated with MVPA maintenance. Significant differences in MVPA change were only observed between those who said PA had become part of their identity vs. those who did not (11±15 vs 4±15 mins/day of MVPA, p=0.011).

Conclusion 
Using changepoint modeling, PA initiation was defined as 0-3 months and maintenance as months 3-12 of a 12-month behavioral weight loss program. While psychosocial variables did not predict PA initiation, PA maintenance is associated with baseline autonomous motivation, and exercise self-efficacy and identity after intervention exposure.