Adherence And Affect In Exercise Interventions For Adolescents With Obesity: A Randomized Clinical-trial

Adherence and sustainability to physical activity interventions for the treatment of obesity in adolescents challenging. PURPOSE: To compare the effects of self-selected exercise intensity (SEI) with predetermined exercise intensity (PEI) in training sessions on exercise adherence and affective valence in adolescents with obesity. METHODS: Thirty-seven adolescents, with obesity, were randomly allocated into SEI (n = 18; 6 girls) or PEI (n = 19; 6 girls). Exercise sessions consisted of 35 minutes of running/walking (treadmill), 3x/week, for 12 weeks. SEI group chose the speed/intensity at the beginning of each session and were allowed to adjust it themselves every 5 minutes. Individuals in the PEI group exercised at an intensity of 60-70% of heart rate reserve. Affective valences in the adolescents during exercise were measured using the Feeling Scale. A chi-square test was used to investigate the distribution of participants from the two in the different adherence group clusters, one independent samples t-test was calculated to test whether the absolute mean frequency of participation in the two treatment groups. The linear mixed effects model was also used to predict the affective valence with the treatment. RESULTS: Mean adherence to training sessions in the participants was 22.3 times (SD = 4.87) for the PEI group, and 23.4 times (SD = 3.15) for SEI, out of a total of 34 offered. This difference is not statistically different, t (18.60) = 0.64, p = .266 (one-tailed), d = 1.05. We found 4 adherers in the PEI condition and 7 in the SEI condition, and 8 non-adherers in the PEI and 6 in the SEI condition. This observed distribution is not different from a random distribution, χ2(1) = 0.40, p = .529, V = 0.13. Additionally predicting affect in the unconditional means model with study participants as a random factor revealed that the average affect across all participants and sessions was 2.90 (95% CI [2.55-3.25], p < .001). Adding treatment as a fixed factor to the model did not significantly contribute to goodness of fit; χ2(1) = 1.16, p = .28. This means that there was no significant difference in mean affective valence between the two treatment groups. CONCLUSIONS: The present results show no differences between SEI and PEI on affective responses during the sessions and adherence to the training program, in adolescents with obesity.