Associations of Voluntary and Required Group Physical Activity with Depressive Symptoms Among Chinese Adolescents: A Cross-sectional and Longitudinal Approach

Technical Summary: Associations of Voluntary and Required Group Physical Activity with Depressive Symptoms Among Chinese Adolescents

Problem Statement
While the general relationship between physical activity (PA) and mental health in adolescents is well-documented, existing literature often treats PA as a monolithic construct. This study addresses a specific gap in the Chinese context: the lack of distinction between voluntary PA and required group PA (e.g., mandatory group running and broadcast calisthenics organized by schools). Previous research has not sufficiently examined how these distinct forms of activity, driven by different motivations (autonomy vs. coercion), differentially impact depressive symptoms. Furthermore, there is a scarcity of longitudinal studies that simultaneously compare these associations while accounting for gender differences, particularly given self-determination theory's suggestion that autonomy is crucial for psychological well-being.

Methodology
The study employed a mixed-methods approach utilizing both cross-sectional and longitudinal designs with a large sample of Chinese adolescents.

• Participants:
  • Cross-sectional (Sample 1): 13,888 valid responses from students across six junior high schools and one senior high school (September–October 2025).
  • Longitudinal (Sample 2): A six-month follow-up of 1,380 students from one junior high school.
• Measures:
  • Physical Activity: Assessed using the Physical Activity Rating Scale-3 (PARS-3), modified to distinguish between "voluntary PA" and "required group PA" (including group running and broadcast calisthenics). Scores were calculated by multiplying intensity, frequency, and duration (0–100 scale).
  • Depressive Symptoms: Measured using the Patient Health Questionnaire-9 (PHQ-9) at both time points.
  • Covariates: Gender, age, residence (town/rural), and subjective socioeconomic status (SES) via the MacArthur Scale.
• Statistical Analysis:
  • Hierarchical Linear Regression: Used to examine associations between PA types and depressive symptoms, controlling for covariates and autoregressive effects (baseline depression) in the longitudinal model.
  • Specification Curve Analysis (SCA): Employed at the cross-sectional level to test the robustness of associations across 16 different model specifications involving various covariate combinations. This involved generating null distributions via 500 bootstrap samples to determine if observed effects deviated significantly from chance.
  • Moderation Analysis: Interaction terms were tested to examine whether gender moderated the longitudinal associations.

Key Results

1. Cross-Sectional Findings:

   • Voluntary PA: Showed a robust, negative association with depressive symptoms. The SCA confirmed this robustness (Median $\beta$ = −0.032, $p$ < 0.001; NSRPD = 16/16).
   • Required Group PA: The association with depressive symptoms was not robust across different model specifications (Median $\beta$ = −0.003, $p$ = 0.626; NSRPD = 2/16). While a standard regression showed a small positive association ($\beta$ = 0.025), the SCA indicated this was not consistent across analytical choices.

2. Longitudinal Findings (6-month follow-up):

   • Voluntary PA: After controlling for baseline depressive symptoms, voluntary PA was not significantly associated with follow-up depressive symptoms ($\beta$ = −0.001, $p$ = 0.913).
   • Required Group PA: Showed a significant, positive association with follow-up depressive symptoms ($\beta$ = 0.060, $p$ = 0.007). Higher levels of required group PA at baseline predicted higher depressive symptoms six months later.
   • Gender Moderation: Gender significantly moderated the longitudinal relationship between required group PA and depressive symptoms.
     • Boys: No significant association ($\beta$ = −0.004, $p$ = 0.889).
     • Girls: A strong, significant positive association was observed ($\beta$ = 0.129, $p$ < 0.001).

Key Contributions

• Differentiation of PA Types: The study empirically distinguishes between voluntary and required group PA, revealing that they have divergent relationships with mental health outcomes.
• Robustness Testing: By utilizing Specification Curve Analysis, the study moves beyond single-model reporting to demonstrate that the negative link between voluntary PA and depression is robust, whereas the cross-sectional link for required group PA is fragile.
• Longitudinal Insight: The research challenges the assumption that all PA is protective. It identifies that required group PA may act as a risk factor for future depressive symptoms, particularly for girls, when controlling for baseline mental health.
• Gender Specificity: The findings highlight a critical gender disparity, suggesting that mandatory group exercise may be detrimental to the mental health of adolescent girls, potentially due to concerns regarding body image, public performance, and lack of autonomy.

Significance and Claims
The authors claim that this study deepens the understanding of how the form and motivation behind physical activity influence adolescent mental health. The significance lies in providing evidence that school-level health promotion policies should not merely focus on increasing PA volume but must consider the organizational form of that activity.

Specifically, the paper suggests:

• Policy Implications: Schools should enrich PA experiences to foster autonomy, choice, and competence, rather than relying solely on mandatory, synchronized group exercises.
• Targeted Interventions: Interventions should be gender-sensitive, acknowledging that required group PA may have adverse effects on girls.
• Theoretical Alignment: The findings partially support Self-Determination Theory, suggesting that the lack of autonomy in required group PA may undermine psychological needs, leading to negative outcomes, particularly in the long term for girls.

Limitations Acknowledged
The authors modestly note several constraints:

• The longitudinal model controlled for autoregressive effects but did not separate within-person from between-person effects.
• The six-month follow-up is relatively short, limiting conclusions to short-term associations.
• The longitudinal sample was restricted to junior high students, potentially limiting generalizability to senior high students.
• The cross-sectional sample was from a region with lower economic development.
• The study relied on self-reported data and did not directly measure autonomy or motivation during the activities.