Enhancing cognitive–emotional self-regulation in girls with conduct disorder: a controlled intervention study comparing virtual reality and club-based martial arts training

Conduct disorder (CD) is characterized by persistent rule-violating and aggressive behaviors that are often associated with difficulties in emotional and behavioral self-regulation. Interventions that strengthen cognitive-emotional self-regulation may therefore help improve behavioral functioning in affected youth. The present study compared the effects of immersive VR-based martial arts training and traditional club-based martial arts training on cognitive-emotional self-regulation in girls with conduct disorder. A three-arm, quasi-experimental, non-randomized controlled intervention study was conducted in Jiangxi, China, involving 143 girls aged 8–11 years with DSM-based diagnoses of conduct disorder. Participants were allocated, based on their availability and scheduling considerations, to VR-based martial arts training (VR-MA; n = 48), club-based martial arts training (Club-MA; n = 48), or a no-exercise control group ( n = 47), while approximately balanced group sizes were maintained. The intervention groups completed 21 sessions over seven weeks (three sessions/week, 30–50 min/session). Cognitive-emotional self-regulation was assessed at baseline (T0), post-intervention (T1), and 30-day follow-up (T2) using a 24-item study-specific adapted composite measure assessing emotional response modulation and impulse control. The measure was reviewed by three independent specialists for content relevance, item clarity, and developmental appropriateness and demonstrated satisfactory internal consistency in the present sample (subscale Cronbach’s α = 0.82–0.84; total-scale α = 0.88). Data were analyzed using mixed repeated-measures analyses and baseline-adjusted ANCOVA. At post-intervention, a significant baseline-adjusted group effect was observed for total cognitive-emotional self-regulation, F(2, 139) = 19.42, p < .001, ηp² = 0.22. Compared with the control group, the VR-MA group showed a moderate improvement (Δ = 5.2, p = .013), whereas the Club-MA group showed a larger improvement (Δ = 11.3, p < .001). At follow-up, group differences remained significant, F(2, 139) = 14.87, p < .001, ηp² = 0.18, although the difference between the VR-MA and control groups was no longer statistically significant. The Club-MA group had higher adjusted cognitive-emotional self-regulation scores than the VR-MA and control groups at post-intervention, and these differences remained evident at the 30-day follow-up. VR-based martial arts training was associated with modest short-term benefits.