Independent and joint associations of aerobic and muscle-strengthening exercise with mental health in adolescents: A cross-sectional analysis before and during COVID-19 using the 2015–2021 National Youth Risk Behavior Survey

Abstract

<p>Purpose: COVID-19 led to social isolation that may have compromised adolescent mental health. This study examined the independent and joint associations of aerobic physical activity (PA) and muscle-strengthening exercise (MSE) with mental health problems in adolescents. Methods: Participants were US adolescents who completed the 2015–2021 National Youth Risk Behavior Survey (N = 61,298; 45.7% female). Data were collected between 2015 and 2021 and analyzed in 2023. Outcomes were binary response items asking about feeling sad/hopeless, having difficulty concentrating, remembering, or making decisions, and having a suicidal ideation. Preventive exposure variables were items asking about frequencies of aerobic PA and MSE with responses dichotomized to align with recommendations. Independent and joint associations were examined using robust Poisson regression with covariates selected using double selection lasso. Structural equation models examined the associations treating PA and MSE as continuous predictors and poor mental health as a latent dependent variable. Results: Meeting either recommendation alone associated with a 4–10% lower prevalence of mental health problems (APR = 0.90–0.96, p < 0.05), and meeting both recommendations associated with a 15%–20% lower prevalence of mental health problems (APR = 0.80–0.85, p < 0.001). Although categorical joint associations were stronger in males (p < 0.05), multiplicative interactions were observed in females using continuous variables for PA and MSE (β = −0.09, p < 0.001). Conclusion: Meeting aerobic PA and MSE recommendations associated with lower prevalence of mental health problems. Participation in MSE below recommended levels may be beneficial for females when combined with aerobic PA. Future research should examine these associations by acquiring contextual information and device-based assessments.</p>