Dataset and Codebook for: Long-term efficacy of exercise across diagnostically heterogenous mental disorders and the mediating role of affect regulation skills

Background: Exercise interventions are efficacious in reducing disorder-specific symptoms in various mental disorders. However, little is known about long-term transdiagnostic efficacy of exercise across heterogenous mental disorders and the potential mechanisms underlying treatment effects. Methods: Physically inactive outpatients, with depressive disorders, anxiety disorders, insomnia or attention deficit hyperactivity disorder were randomized to a standardized 12-week exercise intervention, combining moderate exercise with behaviour change techniques (BCTs) (n = 38), or a passive control group (n = 36). Primary outcome was global symptom severity (Symptom Checklist-90, SCL-90-R) and secondary outcomes were self-reported exercise (Physical Activity, Exercise, and Sport Questionnaire), exercise-specific affect regulation (Physical Activity-related Health Competence Questionnaire) and depression (SCL-90-R) assessed at baseline (T1), post-treatment (T2) and one year after post-treatment (T3). Intention-to-treat analyses were conducted using linear mixed models and structural equations modeling. Results: From T1 to T3, the intervention group significantly improved on global symptom severity (d = -0.43, p = .031), depression among a depressed subsample (d = -0.62, p = .014), exercise (d = 0.45, p= .011) and exercise-specific affect regulation (d = 0.44, p = .028) relative to the control group. The intervention group was more likely to reveal clinically significant changes from T1 to T3 (p = .033). Increases in exercise-specific affect regulation mediated intervention effects on global symptom severity (ß = -0.28, p = .037) and clinically significant changes (ß = -0.24, p = .042). Conclusions: The exercise intervention showed long-term efficacy among a diagnostically heterogeneous outpatient sample and led to long-lasting exercise behaviour change. The long-term increases in exercise-specific affect regulation within exercise interventions seems to be essential for long-lasting symptom reduction beyond an intervention period.