(PS6-A2) Relationships Between Exercise and Disinhibited Eating During Weight Loss and Maintenance

Preliminary work has suggested that when individuals with overweight and obesity enroll in behavioral weight loss programs (BWL), exercise and eating behavior may have important bi-directional effects on each other. Exercise engagement may maximize self-regulatory processes for controlling eating behavior. It is also possible that engaging in disinhibited eating may reduce willingness to exercise, perhaps because disinhibited eating reduces self-efficacy and a sense of control over weight-related behaviors. While the exercise-disinhibited eating relationship has been explored during the initial phase of BWL programs, the nature of this relationship during later stages of the behavior change process remains relatively understudied, which is a notable gap in the literature given the importance of exercise for long-term weight control. The aims of this study were: 1) examine the change in total weekly activity levels (TWA, measured in METs) and disinhibited eating (DE) over the course of an 18-month BWL program and 18-month follow-up, 2) determine if TWA predicts DE at each timepoint, and 3) determine how TWA and DE relate to each other throughout treatment and follow-up. Participants (N = 320) were adults enrolled in BWL treatment who completed self-report measures of DE (Three Factor Eating Questionnaire- Disinhibition) and weekly activity levels (Paffenbarger Physical Activity Questionnaire) at 0, 6 (mid-treatment), 18 (post-treatment), and 36 months (18-month follow-up). Weekly activity levels increased at 6 and 18 months (p’s < .001) relative to baseline but returned to pre-treatment levels by 36 months (p > .05). DE improved (i.e., decreased) from baseline to 6 and 18 months (p’s < .01) but rebounded to baseline levels by 36 months (p > .05). In regression analyses, lower levels of weekly activity predicted higher DE throughout treatment and at 18-month follow-up (p’s < .01). To further explore the pattern of these results, a median split was used to create groups of high vs. low weekly activity levels, and high vs. low levels of DE. At baseline, participants with low weekly activity reported marginally higher DE (M = 9.26, SD = 3.20), compared to those with high levels of weekly activity (M = 8.52, SD = 3.52, p = .052). Significant differences in DE between groups of low vs. high levels of activity were also found at 6, 18, and 36 months (p’s < .01). At baseline, participants with high DE reported lower weekly activity levels (M = 329.29, SD = 67.65) than those with low DE (M = 367.58, SD = 67.04, p < .001). This pattern was maintained at 6 and 18 months, but not at 36 months (p > .05). Results suggest that BWL participants who are less active are also more likely to exhibit disinhibited eating throughout BWL and during the no-intervention follow-up period, adding new insight into the sustained relationship between eating and exercise behaviors. Further research is needed to understand causal direction of effects. The results are consistent with the notion that to promote long-term weight loss, participants likely need to be using skills to maintain reductions in disinhibited eating (which might have a positive effect on exercise engagement) and maintain a high level of exercise (which might have a positive effect on disinhibited eating).