Dose-Response Relationships between Physical Exercises and Mental Health during Early Adolescence: an Investigation of the Underlying Neural and Genetic Mechanisms from the ABCD Study

Gechang Yu, Xinran Wu, Zhaowen Liu, Mai Shi, Huaxin Fan, Yu Liu, Nanyu Kuang, Songjun Peng, Zhengxu Lian, Jingyang Chen, Senyou Yang, Chuiguo Huang, Hongjiang Wu, Baoqi Fan, Jianfeng Feng, Wei Cheng, Barbara J. Sahakian, Trevor W. Robbins, Benjamin Becker, Jie ZhangmedRxiv 2023 · DOI

TL;DR

This study of over 11,000 adolescents reveals that while specific team sports (like soccer and basketball) significantly improve mental health, "more" is not always better; instead, starting at an earlier age provides the most consistent benefits. The research identifies specific neural pathways—particularly connectivity between attention and default-mode networks—and demonstrates that exercise can actually buffer the detrimental effects of high genetic risk for disorders like ADHD.

What Problem Does This Paper Address?

While the link between physical exercise (PE) and adult mental health is well-documented, the specifics for adolescents remain "elusive." We know exercise is good, but we don't know the "dose-response" relationship: Does more exercise always lead to better outcomes? Does the type of sport matter?

Furthermore, the biological "why" is often a black box. This paper seeks to uncover the underlying neural (brain structure and function) and genetic mechanisms that explain how exercise protects adolescent mental health during a period of high developmental vulnerability.

Methods

The researchers employed a multi-modal approach, integrating behavioral, neuroimaging, and genomic data:

  • Linear Mixed Models (LMM): Used to associate 7 dosage measures (e.g., frequency, duration, age of initiation) across 15 types of exercise with mental health scores, adjusting for socioeconomic and biological covariates.
  • Neuroimaging Analysis: Analyzed structural MRI (cortical thickness, volume) and resting-state functional MRI (connectivity between 12 predefined networks) to find neural correlates of exercise.
  • Mediation Analysis: Used 10,000 bootstraps to determine if changes in brain connectivity actually explained the relationship between exercise and improved mental health.
  • Polygenic Risk Scores (PRS): Calculated genetic predispositions for 10 psychiatric disorders to see if exercise could "buffer" or offset high-risk genotypes.

Datasets

The study utilized data from the Adolescent Brain Cognitive Development (ABCD) Study, a massive, multi-site longitudinal cohort in the US.

  • Sample Size: 11,878 adolescents (aged 9–10 years at baseline).
  • Metrics: Included the Child Behavior Checklist (CBCL) for mental health, parental reports for exercise habits, and high-resolution 3T MRI scans.
  • Genetics: Focused on a subset of 4,706 unrelated adolescents of European ancestry to ensure genomic consistency for the PRS analysis.

Key Insights and Ideas

  • The "Diminishing Returns" of Dosage: Unlike pharmacological interventions where more often equals more effect, the researchers found that for many sports, higher frequency or duration didn't necessarily yield extra mental health benefits. A "lower dose" was often sufficient.
  • The "Early Start" Advantage: The most consistent predictor of better mental health wasn't how many hours a kid played per week, but how early they started. Starting in preschool (3–5 years) or middle childhood (6–8 years) was consistently advantageous.
  • The Martial Arts Outlier: Interestingly, while most sports were beneficial or neutral, participation in martial arts was associated with increased psychopathology scores. The authors suggest this might be due to "pre-selection bias," where children with existing behavioral challenges might be more likely to be enrolled in martial arts.
  • Neural "Cross-Talk": The study found that playing football specifically strengthens the communication between the Dorsal Attention Network (DAN) and the Default Mode Network (DMN). This suggests that team sports may help "normalize" attentional biases that contribute to depression and anxiety.

Key Results

  • Beneficial Sports: Five specific exercises—basketball, baseball/softball, soccer, football, and skiing—were significantly associated with reduced mental health problems.
  • Soccer as a Powerhouse: Soccer showed the strongest dose-response relationships. For example, earlier participation in soccer was significantly associated with lower Total Problems ($\beta = 0.77, p < 1.8 \times 10^{-8}$).
  • Genetic Buffering: Participation in any sport significantly buffered the genetic risk of ADHD. Youths with high genetic risk for ADHD who exercised showed significantly lower mental health burdens than those who did not exercise ($\beta = -0.90, z = -3.39$).
  • Brain Structure: Earlier initiation of baseball, basketball, and soccer was linked to increased cortical thickness in the visual and somatosensory cortex, potentially attenuating the rapid cortical thinning typically seen in early adolescence.

Limitations and Future Work

  • Cross-Sectional Nature: The current analysis is primarily cross-sectional. While it shows strong associations, longitudinal data (which the ABCD study will eventually provide) is needed to confirm causality.
  • Ancestry Bias: The genetic analysis was limited to individuals of European ancestry to avoid population stratification issues, which limits the generalizability of the genetic findings to other ethnic groups.
  • Subjective Reporting: Exercise data relied on parental reports, which can be subject to recall bias. Future work could incorporate objective data from wearable sensors (e.g., Fitbits).
  • Unorganized Activity: The study focused on "organized" PE. The impact of unorganized physical activity (like walking or playing in a park) remains an open question for future research.

Summary updated

Abstract

Abstract Adolescence is a critical developmental with increased vulnerability to mental disorders. While the positive impact of physical exercise on adult mental health is well-established, dose-response relationships and the underlying neural and genetic mechanisms in adolescents remain elusive. Leveraging data from &gt;11,000 pre-adolescents (9-10 years, ABCD Study) we examined associations between seven different measures of exercise dosage across 15 exercises and psychopathology, and the roles of brain function and structure and psychiatric genetic risks. Five specific exercises (basketball, baseball/softball, soccer, football, and skiing) were associated with better mental health while the beneficial effects varied with exercise types, dosage measures and dimensions of psychopathology. Interestingly, more exercise does not always translate to better mental health whilst earlier initiation was consistently advantageous. Communication between attention and default-mode brain networks mediated the beneficial effect of playing football. Crucially, exercise mitigates the detrimental effects of psychiatric genetic risks on mental health. We offer a nuanced understanding of exercise effects on adolescent mental health to promote personalized exercise-based interventions in youth.
BibTeX
@misc{yu2023dose,
  author = {Gechang Yu and Xinran Wu and Zhaowen Liu and Mai Shi and Huaxin Fan and Yu Liu and Nanyu Kuang and Songjun Peng and Zhengxu Lian and Jingyang Chen and Senyou Yang and Chuiguo Huang and Hongjiang Wu and Baoqi Fan and Jianfeng Feng and Wei Cheng and Barbara J. Sahakian and Trevor W. Robbins and Benjamin Becker and Jie Zhang},
  title = {{Dose-Response Relationships between Physical Exercises and Mental Health during Early Adolescence: an Investigation of the Underlying Neural and Genetic Mechanisms from the ABCD Study}},
  year = {2023},
  doi = {10.1101/2023.09.20.23295830},
  howpublished = {medRxiv},
}
Figure