Introduction Adolescent mental health in the United States has deteriorated markedly over the past decade, with diagnosed mental or behavioral health conditions among 12-to-17-year-olds increasing 35% between 2016 and 2023, driven by a 61% rise in anxiety diagnoses and a 45% rise in depression. Community-based interventions combining physical activity, creative expression, and social connection represent a promising response. Methods This prospective repeated-measures study examined the association between structured dance education and mood among 256 children and adolescents (ages 5–17) at four studio locations. Over 4 months (February–May 2025), students completed 4, 224 paired pre- and post-class mood ratings using a single-item emoji scale; the primary analytic sample comprised 4, 059 sessions from 251 students. PANAS-C administrations provided a concurrent measure of dispositional affect, with each student’s score anchored to the administration nearest the study close. Multilevel models accounted for the nested data structure (students contributing 1–109 sessions each). Results The intraclass correlation was 0. 432. Post-class mood was significantly higher than pre-class mood (session-level Cohen’s d = 0. 27 primary estimate and per-student dᵦ = 0. 43 p 0. 001) with 85. 8% of sessions showing maintained or improved mood. A Tobit sensitivity analysis confirmed these patterns after adjusting for the 28. 8% post-class ceiling rate. No covariate (including weekly frequency, class type, proficiency level, or instructor experience) explained significant additional variance. Individual differences dominated: random slopes improved model fit substantially (ΔAIC = 201). Convergent validity between the emoji scale and PANAS-C Positive Affect was significant (r = 0. 19, p = 0. 011, n = 175), with the strongest association among students attending 4–6 classes per week (r = 0. 51, n = 25). Conclusion Dance participation was associated with a consistent acute mood benefit generalizing across genres, schedules, and experience levels; individual characteristics, rather than program features, accounted for most meaningful variation. The observational design precludes causal attribution.
Henderson et al. (Thu,) studied this question.