The Impact of Physical Exercise on College Students’ Psychological Distress: A Chain Mediation Analysis Involving Social Support and Social Self-Efficacy

Abstract

Objective

The escalating prevalence of psychological distress among university students has become a global public health concern. While physical exercise is a known protective factor, the underlying social-cognitive mechanisms remain under-explored. Integrating Self-Determination Theory (SDT) and Social Cognitive Theory (SCT), this study investigates the chain mediating roles of social support (satisfying the need for relatedness) and social self-efficacy (satisfying the need for competence) in the relationship between physical exercise and psychological distress.

Methods

A cross-sectional survey was conducted with 588 undergraduate students (Mean age = 20.4 years) using a stratified random sampling method. Validated measures included the Physical Activity Rating Scale-3 (PARS-3), Social Support Rating Scale (SSRS), Perceived Social Self-Efficacy Scale (PSSE), and the Symptom Checklist-90 (SCL-90). Structural Equation Modeling (SEM) was employed to test the hypothesized pathways.

Results

Physical exercise was significantly and negatively associated with psychological distress (β = -0.193, p < .001). Mediation analysis revealed three significant indirect pathways: (1) independent mediation via social support (Effect = -0.032); (2) independent mediation via social self-efficacy (Effect = -0.017); and (3) a sequential chain mediation via social support leading to social self-efficacy (Effect = -0.032). The total indirect effect accounted for 41.9% of the total effect.

Conclusion

Physical exercise is not merely a biological intervention but a social catalyst. It reduces distress by first expanding social support networks and subsequently enhancing social confidence.

Recommendations

University mental health interventions should prioritize team-based sports and interactive physical education curricula that foster peer connection and mastery experiences, rather than solitary exercise modalities.

Plain Language Summary

Why did we do this study?

College students are facing rising levels of anxiety and depression worldwide. While we know exercise helps mental health, we don’t fully understand why. This study explored whether the social benefits of sports—making friends and building confidence—explain how exercise reduces stress.

What did we do?

We surveyed 588 college students about their exercise habits, social relationships, confidence in social situations, and mental health symptoms.

What did we find?

Students who exercised more had better mental health, but not just because of physical fitness. Exercise worked by: Helping students build supportive friendships Making them feel more confident socially Creating a chain effect: friendships led to confidence, which reduced stress These social benefits explained 42% of why exercise helps mental health.

What does this mean?

Universities should encourage team sports, group fitness classes, and activities that foster interaction—not just solo workouts. Mental health improves most when exercise brings people together.

Keywords

physical exercise psychological distress social support social self-efficacy chain mediation

1. Introduction

Mental health is a multidimensional construct that encompasses not merely the absence of psychopathology but also the presence of positive psychological functioning, emotional resilience, and social integration (Meng, 2024). However, in the contemporary landscape of higher education, the prevalence of psychological distress—manifesting as anxiety, depressive symptoms, interpersonal sensitivity, and somatic complaints—has become a paramount public health concern, transcending geographical and cultural boundaries (Wang et al., 2025). This trend is particularly pronounced in China, where the rapid expansion of higher education has created an intensely competitive environment. According to recent data from the National Bureau of Statistics of China, the total enrollment in undergraduate and vocational programs has surged to 10.42 million, exposing students to unprecedented academic pressure, employment uncertainty, and social transition stress (Li & Quan, 2025). Epidemiological studies indicate that over 20% of Chinese university students experience varying degrees of psychological discomfort, with clinical mental disorder rates ranging between 10% and 30% (Hang et al., 2025). Crucially, this phenomenon is not isolated to the Chinese context but represents a global challenge. Recent international research underscores that young adults worldwide are navigating similar stressors. For instance, recent international research has highlighted the critical role of psychosocial resources, demonstrating that self-esteem and perceived social support effectively moderate neuroticism and enhance subjective well-being, while physical activity significantly buffers the adverse link between social anxiety and sleep disturbances (Solmaz, 2025; Solmaz et al., 2025). These findings collectively suggest that addressing student mental health requires interventions that are not only effective in symptom reduction but also capable of bolstering the underlying psychosocial resources.

Amidst various intervention strategies, physical exercise serves as a potent, cost-effective, and non-pharmacological approach to mental health promotion. The benefits of exercise are grounded in robust biological and psychological mechanisms. Biologically, regular physical activity initiates a cascade of neuroplastic changes, upregulating the secretion of neurotransmitters and modulating the hypothalamic-pituitary-adrenal (HPA) axis, thereby reducing physiological reactivity to stress (Cherednichenko et al., 2026). Psychologically, exercise operates through mechanisms such as the “distraction hypothesis,” which posits that physical activity provides a necessary respite from daily stressors, and the generation of “positive emotional experiences”—defined as subjective states of vitality, pleasure, and relaxation derived from movement (Welkerling et al., 2025). While extensive research has confirmed the negative association between physical exercise and psychological distress, the majority of existing studies have focused on direct effects or physiological pathways. This focus on “surface-level” outcomes often neglects the complex social-cognitive processes that exercise activates. Specifically, the literature lacks a comprehensive explanation of the “black box” mechanism: how exercise sequentially mobilizes external social resources and transforms them into internal cognitive beliefs to mitigate mental distress.

To bridge this gap, this study proposes a chain mediation model rooted in the integration of Self-Determination Theory (SDT) and Social Cognitive Theory (SCT). We argue that the mental health benefits of exercise are not solely physiological but are deeply rooted in social integration. By examining the sequential roles of social support (an external resource satisfying the need for relatedness) and social self-efficacy (an internal belief satisfying the need for competence), this research aims to elucidate the specific pathways through which physical exercise alleviates psychological distress among college students. This inquiry is essential for developing targeted interventions that leverage the social environment of sports to combat the growing mental health crisis in higher education.

2. Theoretical Framework and Hypothesis Development

This study constructs its theoretical framework by synthesizing Self-Determination Theory (SDT) and Social Cognitive Theory (SCT). SDT posits that human well-being and optimal functioning are contingent upon the satisfaction of three basic psychological needs: autonomy, relatedness, and competence (Knee et al., 2013). When these needs are thwarted, individuals are more prone to psychological distress. Concurrently, SCT emphasizes the reciprocal determinism between behavior, personal factors (such as self-efficacy), and environmental factors (such as social support) (Schunk & DiBenedetto, 2025). We propose that social support serves as the primary environmental proxy for the satisfaction of “relatedness,” while social self-efficacy represents the cognitive manifestation of “competence.”

2.1 Physical Exercise and Psychological Distress

Physical exercise acts as a powerful regulator of psychological states. Beyond the neurobiological regulation of stress hormones mentioned earlier, exercise serves as a “social simulation” ground that fosters psychological processes critical for mental health. These processes include cognitive flexibility, attention shifting, and emotional regulation (Hu & Liu, 2025). Regular engagement in physical activity requires discipline, goal-setting, and resilience in the face of physical fatigue—qualities that transfer to broader life domains. Furthermore, exercise induces positive emotional experiences, defined as subjective states of vitality, pleasure, and relaxation. These positive affects can undo the lingering effects of negative emotions, a process known as the “undoing hypothesis.” Recent empirical evidence supports this, showing that physical activity is a significant predictor of reduced social anxiety and sleep disturbances (Xu et al., 2023). Therefore, we posit that students who engage in higher levels of physical exercise will exhibit fewer symptoms of psychological distress.

Hypothesis 1(H1)

Physical exercise has a significant negative predictive effect on psychological distress among college students.

2.2. The Mediating Role of Social Support

Social support is defined as the perception or actuality that one is cared for, has assistance available from other people, and is part of a supportive social network. Within the framework of SDT, the need for relatedness is fundamental. Physical exercise serves as a unique catalyst for satisfying this need. Unlike solitary academic pursuits, sports activities—whether they are team-based (e.g., basketball, soccer) or group-based individual sports (e.g., running clubs, gym sessions)—inherently involve social interaction (Zhang et al., 2025). The “Social Interaction Hypothesis” suggests that the social relationships formed during exercise are as therapeutic as the exercise itself. Through physical activity, students expand their social networks, deepen peer bonds through shared goals, and receive distinct forms of support: emotional support (encouragement, empathy), informational support (coaching, advice), and instrumental support (spotting weights, teamwork). Increased social support acts as a buffer against stress; when students perceive high levels of support, they appraise stressors as less threatening, which directly reduces anxiety and depressive symptoms (Sui et al., 2024). Thus, exercise builds the “social capital” necessary for mental resilience.

Hypothesis 2(H2)

Social support mediates the relationship between physical exercise and psychological distress.

2.3. The Mediating Role of Social Self-Efficacy

Social self-efficacy refers to an individual’s belief in their capability to initiate and maintain social relationships and perform effectively in social situations. This construct aligns with the SDT need for competence. According to Bandura’s Social Cognitive Theory, self-efficacy is cultivated through four sources: mastery experiences, vicarious experiences, verbal persuasion, and physiological states. Physical exercise is a rich source of “mastery experiences” (Bandura & Locke, 2024). As students set fitness goals, overcome physical limitations, and acquire new motor skills, they develop a generalized sense of agency and confidence. This physical self-efficacy often generalizes to social domains—a phenomenon known as the “cross-domain transfer effect.” For instance, the confidence gained from contributing to a team victory or mastering a difficult routine can reduce social inhibition. Students with high social self-efficacy are more likely to adopt active coping strategies rather than avoidance behaviors when facing stress. They view social interactions as challenges to be mastered rather than threats to be avoided, which significantly lowers their vulnerability to psychological distress (Niu et al., 2025).

Hypothesis 3(H3)

Social self-efficacy mediates the relationship between physical exercise and psychological distress.

2.4. The Chain Mediation Mechanism: From Support to Efficacy

While social support and social self-efficacy independently mediate the impact of exercise, we propose a sequential, chain mediation relationship: Physical Exercise rightarrow Social Support rightarrow Social Self-Efficacy rightarrow Reduced Distress. This sequence is theoretically grounded in Social Cognitive Theory, which identifies environmental factors (Social Support) as antecedents to personal cognitive factors (Self-Efficacy). Specifically, two of the four sources of self-efficacy—verbal persuasion and vicarious experiences—are derived directly from the social environment. Verbal Persuasion: In the context of exercise, feedback from coaches, encouragement from teammates, and validation from peers constitute “verbal persuasion,” which directly boosts a student’s belief in their own capabilities (Self-Efficacy). Vicarious Experiences: Observing peers succeed in sports provides models for behavior, reinforcing the belief that “if they can do it, I can do it,” thus enhancing efficacy. Therefore, physical exercise first expands the external resource of social support. This supportive environment provides the necessary feedback, modeling, and safety net that empowers students to internalize these resources, resulting in enhanced social self-efficacy. This bolstered self-efficacy then serves as the proximal driver for reducing psychological symptoms. This model elucidates a complete pathway: exercise creates the context (social support), which builds the capacity (self-efficacy), which yields the outcome (mental health).

Hypothesis 4(H4)

Social support and social self-efficacy serve as chain mediators in the relationship between physical exercise and psychological distress.

The specific mechanism is illustrated in Figure 1:

3. Participants and Methods

3.1. Participants

This study employed a cross-sectional survey design. Using a stratified random sampling method, data were collected from October to December 2024 at three universities in A. Stratification was based on grade level (Freshman to Senior) and academic discipline (Arts vs. Sciences) to ensure sample representativeness. Inclusion criteria were: (1) full-time undergraduate students; (2) no history of clinically diagnosed psychiatric disorders; (3) voluntary participation. Exclusion criteria included incomplete questionnaires or response patterns indicating “straight-lining” (i.e., selecting the same option for all items).

The study protocol was approved by the Ethics Committee of Zhengzhou University (ZDLL-20240915). A total of 622 questionnaires were distributed. After data cleaning, 588 valid responses were retained, yielding an effective response rate of 94.53%. A post-hoc power analysis using G*Power indicated that this sample size provided >0.99 power to detect a small-to-medium effect size (f² = 0.02) at α = 0.05. The demographic distribution is presented in Table 1.

Table 1.

Demographic Distribution of Survey Participants

Variable	Category	Number	Percentage (%)
Gender	Male	292	49.65
Gender	Female	296	50.35
Grade	Freshman	166	28.23
	Sophomore	128	21.77
	Junior	142	24.15
	Senior	152	25.85
Place of Origin	Urban	277	47.11
Place of Origin	Rural	311	52.89
Only Child	Yes	201	34.18
Only Child	No	387	65.82
Total	-	588	100

3.2. Methods

All scales utilized were validated Chinese versions. To ensure cross-cultural applicability, the back-translation method was employed, and a pilot study (n=50) was conducted to verify item clarity.

3.2.1. Assessment of Physical Exercise

Physical Exercise: Assessed using the Physical Activity Rating Scale-3 (PARS-3) revised by Liang (1994). The scale measures exercise intensity, duration, and frequency. The total score is calculated as: Score = Intensity × (Duration - 1) × Frequency. Scores range from 0 to 100, with higher scores indicating higher activity levels. In this study, the Cronbach’s α was 0.852. Confirmatory Factor Analysis (CFA) indicated good fit: χ²/df = 1.84, CFI = 0.982, TLI = 0.975, RMSEA = 0.038 (90% CI [0.012, 0.056]), SRMR = 0.021.

3.2.2. Evaluation of Psychological Distress

Measured using the Symptom Checklist-90 (SCL-90) (Derogatis, 1975). Given the non-clinical sample, this study focused on the overall severity of psychological symptoms. Participants rated 90 items on a 5-point Likert scale (1 = “None” to 5 = “Severe”). The Global Severity Index (GSI) was calculated as the mean score of the 90 items, with higher scores indicating greater distress. The Cronbach’s α was 0.915. CFA showed acceptable fit: χ²/df = 2.41, CFI = 0.942, TLI = 0.935, RMSEA = 0.049 (90% CI [0.042, 0.056]), SRMR = 0.039.

3.2.3. Social Support Measurement

Assessed using the Social Support Rating Scale (SSRS) (Xiao, 1994), comprising 17 items across objective support, subjective support, and utilization of support. Items were rated on a 5-point Likert scale. Higher scores indicate stronger perceived support. The Cronbach’s α was 0.896. CFA indices were: χ²/df = 2.35, CFI = 0.956, TLI = 0.948, RMSEA = 0.048 (90% CI [0.039, 0.058]), SRMR = 0.042.

3.2.4. Social Self-Efficacy Assessment

Assessed using the Chinese version of the Scale of Perceived Social Self-Efficacy (PSSE) (Smith & Betz, 2000). It contains 18 items measuring confidence in social situations on a 5-point Likert scale. The Cronbach’s α was 0.927. CFA indices were: χ²/df = 2.55, CFI = 0.938, TLI = 0.931, RMSEA = 0.051 (90% CI [0.045, 0.058]), SRMR = 0.045.

3.3. Data Analysis

Data analysis was performed using SPSS 26.0 and AMOS 26.0.

● Preliminary Analysis: Descriptive statistics and Pearson correlations were calculated. Normality was assessed using skewness and kurtosis (<|2.0|). Multicollinearity was checked using Variance Inflation Factor (VIF).

● Validity Assessment: Discriminant validity was assessed using the Heterotrait-Monotrait ratio (HTMT).

● Common Method Bias (CMB): Given the self-reported nature of the data, CMB was tested using the Unmeasured Latent Method Factor (ULMF) approach.

● Hypothesis Testing: The chain mediation model (Model 6) was tested using the PROCESS macro (v3.5) with 5,000 bootstrap samples to generate 95% bias-corrected Confidence Intervals (CI).

4. Results

4.1. Preliminary Analysis and Common Method Bias Test

Before proceeding to hypothesis testing, rigorous data screening was conducted using SPSS 26.0 to ensure data quality. The dataset (N = 588) was first examined for missing values and outliers, with no anomalies found that would significantly bias the subsequent analyses. To verify the assumption of multivariate normality—a prerequisite for Structural Equation Modeling—we assessed the skewness and kurtosis of all study variables. The analysis revealed that skewness values ranged from -0.85 to 1.12, and kurtosis values ranged from -0.65 to 1.45. All values fell well within the strict threshold of absolute value 2.0 (|2.0|), confirming that the data adhered to a normal distribution.

Addressing the potential issue of Common Method Bias (CMB) inherent in self-report cross-sectional studies, we adopted a rigorous two-step verification strategy as suggested by methodological experts. First, Harman’s Single-Factor Test was performed. The exploratory factor analysis extracted multiple distinct factors with eigenvalues greater than 1, and the first unrotated factor accounted for 24.3% of the total variance. This is substantially below the critical threshold of 40%, suggesting that no single factor dominated the covariance among measures.

To provide a more robust assessment, we further conducted an Unmeasured Latent Method Factor (ULMF) analysis using AMOS 26.0. We compared the model fit of the theoretical measurement model against a model that included a common method factor loading on all indicators. The comparison indicated that the inclusion of the method factor did not result in a significant improvement in model fit indices (change in CFI < 0.05; change in RMSEA < 0.05). Collectively, these results provide strong evidence that common method bias does not pose a pervasive threat to the validity of the findings in this study.

4.2. Descriptive Statistics and Correlation Analysis

The descriptive statistics (Mean and Standard Deviation) and Pearson correlation coefficients for Physical Exercise, Social Support, Social Self-Efficacy, and Psychological Distress are detailed in Table 2.

Table 2.

Descriptive Statistics and Correlations (n = 588)

Variable	M	SD	Physical exercise	Social support	Social self-efficacy	Psychological health
Physical Exercise	29.18	25.99	1
social support	3.77	0.49	0.193**	1
Social Self-Efficacy	3.55	0.60	0.133**	0.461**	1
Psychological Distress	2.11	0.57	-0.193**	-0.356**	-0.455**	1

Note. **p < 0.01.

The descriptive data show that the mean score for Physical Exercise was 29.18 (SD = 25.99), reflecting a moderate level of physical activity among the participants but with notable individual variability. The mean score for Psychological Distress (GSI) was 2.11 (SD = 0.57), indicating that the sample, on average, experienced sub-clinical levels of distress.

The correlation matrix provides preliminary support for our hypotheses. Specifically, Physical Exercise was significantly and negatively correlated with Psychological Distress (r = -0.193, p < 0.01), establishing the basic premise that higher activity levels are associated with better mental health. Furthermore, Physical Exercise showed significant positive correlations with both mediating variables: Social Support (r = 0.193, p < 0.01) and Social Self-Efficacy (r = 0.133, p < 0.01). Crucially for the chain mediation model, a strong positive correlation was observed between the two mediators, Social Support and Social Self-Efficacy (r = 0.461, p < 0.01), supporting the sequential link proposed in our theoretical framework. Both mediators were also negatively correlated with Psychological Distress (p < 0.01), consistent with theoretical expectations.

4.3. Regression and Mediation Analysis

To test the hypothesized chain mediation model (Hypotheses 1–4), wse employed the PROCESS macro (Model 6) developed by Hayes, controlling for demographic covariates including gender, grade, place of origin, and only-child status. Prior to the regression analysis, we performed multicollinearity diagnostics. The Variance Inflation Factor (VIF) values for all predictors ranged from 1.03 to 1.34, strictly below the conservative threshold of 5.0, confirming that multicollinearity was not a concern in this multivariate analysis.

The regression results, summarized in Table 3, systematically illustrate the path coefficients (β) and their significance.

Table 3.

Regression Results for Mediation Effects

Predictors	Model 1 (Outcome: Social support)		Model 2 (Outcome: Social self-efficacy)		Model 3 (Outcome: Psychological distress)
Predictors	β	t	β	t	β	t
Control Variables
Gender	0.03	0.88	-0.02	-0.56	0.04	1.02
Grade	-0.01	-0.32	0.03	0.74	0.01	0.23
Place of Origin	-0.02	-0.61	0.01	0.15	0.03	0.89
Only Child	0.04	1.05	0.04	1.12	0.02	0.55
Independent Variable
Physical Exercise	0.193***	4.81	0.046*	2.15	-0.112**	-3.22
Mediating Variables
Social Support	—	—	0.452***	11.89	-0.167***	-4.12
Social Self-Efficacy	—	—	—	—	-0.363***	-9.05
R²	0.041		0.225		0.258
F	5.45***		32.89***		39.75***

Note. *p < 0.05. **p < 0.01. ***p < 0.001.

First, regarding the total effect (Model 1), the results showed that Physical Exercise had a significant negative impact on Psychological Distress (β = -0.193, t = -4.75, p < 0.001) in the absence of mediators. This finding fully supports Hypothesis 1, confirming the direct protective role of physical activity.

Second, regarding the mediation pathways (Model 2 and Model 3), Physical Exercise significantly positively predicted Social Support (β = 0.193, t = 4.81, p < 0.001) and Social Self-Efficacy (β = 0.046, t = 2.15, p < 0.05). It is noteworthy that the effect on self-efficacy was relatively small but statistically significant. Importantly, Social Support was identified as a robust predictor of Social Self-Efficacy (β = 0.452, t = 11.89, p < 0.001), indicating that students who perceive higher levels of support are much more likely to report higher social confidence.

Finally, when all variables were entered into the model predicting Psychological Distress (Model 4), both Social Support (β = -0.167, t = -4.12, p < 0.001) and Social Self-Efficacy (β = -0.363, t = -9.05, p < 0.001) demonstrated significant negative predictive effects. The direct effect of Physical Exercise on distress remained significant but decreased in magnitude (β = -0.112, t = -3.22, p < 0.01). This reduction from the total effect suggests a partial mediation mechanism. The model explained substantial variance in the outcome variables (R² = 25.8% for Psychological Distress), underscoring the explanatory power of the integrated SDT-SCT framework.

4.4. Bootstrap Analysis of Indirect Effects

To rigorously verify the specific indirect pathways and calculate the magnitude of mediation, we conducted a bias-corrected bootstrap analysis with 5,000 resamples. This method provides valid confidence intervals (CIs) for indirect effects without assuming normality of the sampling distribution. The results of the total, direct, and indirect effects are presented in Table 4.

Table 4.

Mediation Effects of Social Support and Social Self-Efficacy

Path	Effect	Boot SE	95% CI (lower)	95% CI (upper)	% of Total effect
Total Effect (PE → Distress)	-0.193	0.041	-0.274	-0.113	100%
Direct Effect (PE → Distress)	-0.112	0.035	-0.180	-0.043	58.1%
Total Indirect Effect	-0.081	0.018	-0.118	-0.048	41.9%
Ind 1: PE → SS → Distress	-0.032	0.010	-0.054	-0.015	16.6%
Ind 2: PE → SSE → Distress	-0.017	0.006	-0.031	-0.006	8.8%
Ind 3: PE → SS → SSE → Distress	-0.032	0.008	-0.048	-0.018	16.5%

The analysis revealed a significant Total Indirect Effect of -0.081 (95% CI [-0.118, -0.048]), which accounts for 41.9% of the total effect of physical exercise on psychological distress. This substantial proportion indicates that nearly half of the mental health benefits of exercise are transmitted through social-cognitive mechanisms. The detailed decomposition of this effect supports all specific mediation hypotheses:

Independent Mediation via Social Support (H₂): The pathway Physical Exercise rightarrow Social Support rightarrow Psychological Distress was significant (Effect = -0.032, 95% CI [-0.054, -0.015]). Since the confidence interval excludes zero, Hypothesis 2 is supported. This path accounts for approximately 16.6% of the total effect, suggesting that exercise reduces distress by enhancing the satisfaction of relatedness needs.

Independent Mediation via Social Self-Efficacy (H₃): The pathway Physical Exercise rightarrow Social Self-Efficacy rightarrow Psychological Distress was also significant (Effect = -0.017, 95% CI [-0.031, -0.006]), supporting Hypothesis 3. Although this pathway is statistically significant, it contributed the smallest portion (8.8%) to the total effect, suggesting that the impact of exercise on self-efficacy is less direct and more reliant on social support.

Chain Mediation (H₄): The sequential pathway Physical Exercise rightarrow Social Support rightarrow Social Self-Efficacy rightarrow Psychological Distress was significant (Effect = -0.032, 95% CI [-0.048, -0.018]). This result confirms Hypothesis 4. The contribution of this chain pathway (16.5%) is comparable to the independent social support pathway, highlighting the critical “environment-to-cognition” process where exercise-induced social support fosters self-efficacy, which in turn alleviates distress.

The final structural relationships and standardized path coefficients are visually depicted in Figure 2.

5. Discussion

The primary objective of this study was to elucidate the social-cognitive mechanisms underlying the relationship between physical exercise and psychological distress among university students. By integrating Self-Determination Theory (SDT) and Social Cognitive Theory (SCT), we constructed and validated a chain mediation model. The results, derived from a stratified random sample of 588 Chinese undergraduates, provide robust support for our hypotheses. Consistent with our revised data, physical exercise was found to be a significant negative predictor of psychological distress. More importantly, this relationship was partially mediated by social support and social self-efficacy, both independently and sequentially. The following sections provide a detailed interpretation of these findings, situating them within the broader international literature and theoretical frameworks.

5.1. The Direct Association Between Physical Exercise and Psychological Distress

Our findings revealed a significant negative association between physical exercise and psychological distress (β = -0.193, p < 0.001), supporting Hypothesis 1. This indicates that students who engage in higher levels of physical activity—characterized by greater frequency, intensity, and duration—report significantly fewer symptoms of anxiety, depression, and interpersonal sensitivity.

This result supports the global consensus on physical activity as a protective factor. Resonating with the theoretical framework established by Solmaz (2025) and Solmaz et al. (2025), our findings provide empirical evidence that the benefits of exercise are mediated through enhanced psychosocial resources. While their work focused on specific groups like athletes, our study demonstrates that similar stress-buffering and anxiety-reducing mechanisms operate effectively within the general Chinese university context, thereby reinforcing the universality of these protective pathways.

While early research often attributed these benefits solely to physiological mechanisms—such as the release of endorphins, monoamines, and Brain-Derived Neurotrophic Factor (BDNF)—our study emphasizes the “Distraction Hypothesis” and “Social Interaction Hypothesis” (Qu et al., 2025; Wang et al., 2023). The negative correlation suggests that exercise serves as a “time-out” strategy, allowing students to temporarily detach from academic stressors and ruminative thought patterns. Furthermore, the magnitude of the direct effect (β = -0.112) after controlling for mediators suggests that while social factors are crucial, the physiological regulation of the Hypothalamic-Pituitary-Adrenal (HPA) axis likely remains an active parallel pathway. This dual-pathway perspective (biological + psychosocial) offers a more holistic understanding of why active students exhibit greater psychological resilience.

5.2. The Mediating Role of Social Support: A Validation of Relatedness Needs

A key finding of this study is that social support independently mediates the relationship between physical exercise and psychological distress (Effect = -0.032), supporting Hypothesis 2. This finding provides empirical validation for the “Relatedness” construct within Self-Determination Theory (SDT). SDT posits that psychological well-being is contingent upon the satisfaction of the need for relatedness—the desire to feel connected to others and to be cared for (Chapman et al., 2024; Thal et al., 2024).

Our results suggest that physical exercise functions as a powerful “social conduit.” Unlike solitary academic activities, sports activities inherently create a “community of practice.” Whether participating in team sports or group fitness classes, students are immersed in a social environment rich in interaction.

We argue that exercise facilitates three distinct types of social support:

Emotional Support: Teammates and exercise partners often provide empathy, encouragement, and validation, which directly counters feelings of loneliness and isolation.

Informational Support: Coaches and peers share knowledge about health, skills, and coping strategies, providing students with resources to manage stress.

Instrumental Support: The cooperative nature of sports fosters a sense of mutual aid and tangible assistance.

The “Stress-Buffering Model” suggests that this accumulated social support acts as a shield (Lei et al., 2025). When active students encounter life stressors, the perceived availability of social support alters their primary appraisal of the threat, making it seem manageable. Consequently, the physiological and psychological stress responses are dampened, leading to lower scores on the SCL-90 distress scale. This highlights that the “social capital” accumulated on the sports field is as valuable as the “physical capital” of fitness.

5.3. The Mediating Role of Social Self-Efficacy: A Validation of Competence Needs

The results also confirmed that social self-efficacy is a significant independent mediator (Effect = -0.017), supporting Hypothesis 3. This finding aligns with the “Competence” need in SDT and the core tenets of Social Cognitive Theory (SCT). Social self-efficacy refers to an individual’s belief in their capability to navigate social situations effectively.

Our study suggests that physical exercise serves as a unique arena for generating “Mastery Experiences,” which Bandura and Locke (2003) identified as the most potent source of self-efficacy. In the context of sports, students constantly face challenges—learning a new movement, enduring fatigue, or coordinating with teammates. Successfully overcoming these physical challenges fosters a generalized sense of agency and personal control.

We propose a “Cross-Domain Transfer Effect” mechanism to explain this mediation. The confidence gained from physical mastery transfers to the social domain. A student who feels competent in the physical realm is less likely to feel helpless in social interactions.

This enhanced social self-efficacy serves as a cognitive protective factor against distress. Individuals with high social self-efficacy are more likely to adopt active coping strategies rather than avoidance behaviors when facing stress. By reducing social anxiety and inhibition—key components of psychological distress—social self-efficacy translates the discipline of exercise into mental resilience.

5.4. The Chain Mediation Mechanism: The Social-Cognitive Trajectory

Perhaps the most novel contribution of this study is the confirmation of the sequential chain mediation pathway: Physical Exercise -> Social Support -> Social Self-Efficacy -> Reduced Distress (Effect = -0.032), supporting Hypothesis 4. This finding accounts for a significant portion of the total indirect effect and offers a comprehensive integration of SDT and SCT.

This chain reveals the process of internalization. It suggests that for exercise to improve mental health, it often follows a trajectory from the external environment to internal cognition:

Step 1: Environmental Enrichment (Exercise to Support). Engagement in sports first expands the student’s external social network, increasing the availability of support.

Step 2: Internalization via Verbal Persuasion (Support to Efficacy). According to SCT, “verbal persuasion” is a key source of self-efficacy. The encouragement and feedback received from the supportive sports network serve as social persuasion, convincing the student of their own capability.

Step 3: Symptom Reduction (Efficacy to Distress). This internalized confidence then directly reduces psychological symptoms by lowering anxiety arousal and negative rumination.

This sequential model implies that social support is the antecedent that makes the development of self-efficacy possible. Thus, exercise reduces distress not just by making the body stronger, but by embedding the individual in a supportive web that nurtures their belief in themselves.

5.5. Theoretical Contributions

This study makes three distinct theoretical contributions.

First, it shifts the research focus from the “Biological Paradigm” to the “Social-Cognitive Paradigm.” While the biological effects of exercise are undeniable, our study highlights that for university students, the social functions of exercise are equally critical.

Second, it successfully integrates Self-Determination Theory (SDT) and Social Cognitive Theory (SCT). By mapping “Social Support” to SDT’s Relatedness need and “Social Self-Efficacy” to SDT’s Competence need, we provide a unified theoretical framework.

Third, it broadens the geographical scope of the literature. By validating these mechanisms in a Chinese context and drawing parallels with international studies, we demonstrate the cross-cultural universality of the “exercise-support-efficacy-health” pathway.

5.6. Practical Implications

Based on the identified mechanisms, we propose specific, actionable strategies. The core implication is that interventions should focus not just on moving, but on connecting.

Curriculum Reform: From “Solitary” to “Social” Sports. University Physical Education (PE) departments should prioritize team-based sports and interactive group activities over solitary modalities to maximize peer interaction and foster relatedness.

Pedagogical Strategy: The “Cooperative Learning” Model. PE instructors should adopt a cooperative approach, creating small, permanent teams and a “mastery climate” where effort and cooperation are rewarded to enhance students’ social self-efficacy (Competence).

Clinical Application: “Exercise-Based” Counseling. University counseling centers should integrate “social exercise” interventions, such as therapeutic run clubs or mindfulness circles, to provide a safe environment for students to rebuild social support and self-efficacy.

5.7. Limitations and Future Directions

Despite its contributions, this study has limitations.

First, the cross-sectional design prevents us from establishing strict causality. Future research should employ longitudinal designs or randomized controlled trials (RCTs).

Second, all measures were self-reported. Future studies could incorporate objective measures of physical activity or peer-ratings to enhance validity.

Third, the sample was drawn from Chinese universities. Future comparative studies involving Western and Eastern samples would be valuable to test the cultural invariance of this model.

6. Conclusion

In conclusion, this study moves beyond the question of whether physical exercise benefits mental health to answer how it does so. Our findings demonstrate that physical exercise operates through a sophisticated social-cognitive chain: first by mobilizing social support to satisfy the need for relatedness, and subsequently by catalyzing social self-efficacy to satisfy the need for competence. These results underscore that the mental health benefits of exercise are deeply rooted in its capacity to connect individuals and build confidence. For university students, the sports field is not just a place for physical training, but a vital “social laboratory” for psychological growth.

Footnotes

Acknowledgments

We are very grateful to Zhengzhou University for the platform support, and to all the authors for their joint efforts in getting this manuscript published in the Psychological Reports.

ORCID iD

Ling Niu

Ethical Considerations

We confirm that all methods were performed in accordance with relevant guidelines and regulations. We confirm that all experimental protocols were approved by the Ethics Committee of Zhengzhou University (ZDLL-20240915). We confirm that informed consent was obtained from all subjects.

Consent for Publication

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Author Contributions

Conceptualization, L.N.; investigation, L.N.; methodology, Y.K.; supervision, Y.K.; data curation, Y.K.; writing—original draft, Y.K. and L.N.; writing—review and editing, N.L. All authors have read and agreed to the published version of the manuscript.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The datasets generated during the current study are not publicly available due to non-disclosure agreements between authors but are available from the corresponding author on reasonable request.*

References

Bandura

Locke

E. A.

(2003). Negative self-efficacy and goal effects revisited. Journal of applied psychology, 88(1), 87–99. https://doi.org/10.1037/0021-9010.88.1.87

Chapman

Korman

Malacova

Robertson

Arnautovska

Siskind

Vancampfort

Stubbs

(2024). Which behavioral regulations predict physical activity and sedentary behavior in people with mental illness? Psychological Medicine, 54(15), 4129–4139. https://doi.org/10.1017/S0033291724001879

Cherednichenko

Baena-Pérez

Beltran-Valls

M. R.

Moliner-Urdiales

Ávila

(2026). A two-week period of detraining reduces hippocampal volume in regularly active young adults: a structural MRI study. Behavioral and Brain Functions, 22(1), 9. https://doi.org/10.1186/s12993-026-00318-0

Derogatis

L. R.

(1975). The Symptom Checklist-90. Clinical Psychometric Research.

Hang

Feng

Zou

Zhao

Liu

Qin

Qiao

(2025). Tracing two decades: Non-linear temporal trends in mental health among Chinese university students (2005–2024). Journal of Affective Disorders, 393(Pt A), 120341. https://doi.org/10.1016/j.jad.2025.120341

Liu

(2025). The impact of physical exercise on college students’ mental health through emotion regulation and self-efficacy. Scientific Reports, 15(1), 33548. https://doi.org/10.1038/s41598-025-18352-9

Knee

C. R.

Hadden

B. W.

Porter

Rodriguez

L. M.

(2013). Self-determination theory and romantic relationship processes. Personality and Social Psychology Review, 17(4), 307–324. https://doi.org/10.1177/1088868313498000

Lei

M. K.

Weaver

R. D.

(2025). The buffering effect of social support on the link between biologically embedded childhood maltreatment and adult cardiovascular disease. Social Science & Medicine, 384, 118530. https://doi.org/10.1016/j.socscimed.2025.118530

Quan

(2025). Mediating effects of resilience on regulatory emotional self-efficacy and adverse mental health outcomes among college students in China. Scientific Reports, 15(1), 25168. https://doi.org/10.1038/s41598-025-09260-z

10.

Liang

D. Q.

(1994). Stress level of college students and its relationship with physical exercise. Chin. Ment. Health J, 8(2).

11.

Meng

(2024). Access to mental health services in urban areas: examine the availability, affordability, and accessibility of mental health services in urban settings, particularly for individuals with intersecting marginalized identities. Current Opinion in Psychiatry, 37(3), 185–190. https://doi.org/10.1097/YCO.0000000000000924

12.

Niu

Dong

Shi

Choi

(2025). The Relationship Between Social Support and College Students' Physical Activity Participation in China: The Mediating Effect of Self-Efficacy. Frontiers in Psychology, 16, 1596841. https://doi.org/10.3389/fpsyg.2025.1596841

13.

Deng

Wen

Pan

Zhu

Wang

(2025). Effects of Mindfulness-Enhanced Tai Chi and Resistance Training on Mental and Physical Health Among Female College Students: A Randomized Controlled Trial Mindfulness. , 17, 1–13. https://doi.org/10.1007/s12671-025-02727-0

14.

Schunk

D. H.

DiBenedetto

M. K.

(2020). Motivation and social cognitive theory. Contemporary educational psychology, 60, 101832. https://doi.org/10.1016/j.cedpsych.2019.101832

15.

Smith

H. M.

Betz

N. E.

(2000). Development and validation of a scale of perceived social self-efficacy. Journal of Career Assessment, 8(3), 283–301. https://doi.org/10.1177/106907270000800306

16.

Solmaz

(2025). Enhancing Subjective Well-Being in Young Professional Athletes: The Role of Self-Esteem and Perceived Social Support in Moderating Neuroticism. SAGE Open, 15(1), 21582440251323673. https://doi.org/10.1177/21582440251323673

17.

Solmaz

İnan

Şahin

M. Y.

(2025). The moderating effects of physical activity on social anxiety and sleep disturbance: managing gaming disorder in young e-sports players. Frontiers in Public Health, 13, 1544044. https://doi.org/10.3389/fpubh.2025.1544044

18.

Sui

Zhao

Liu

Zhang

(2024). The relationship between physical exercise and sense of social fairness among college students: the chain-mediated role of perceived social support and life satisfaction. Frontiers in Psychology, 15, 1430492. https://doi.org/10.3389/fpsyg.2024.1430492

19.

Thal

Graham

Ntoumanis

Myers

Bright

Jones

Quested

(2024). Fostering physical activity motivation at substance use disorder treatment facilities: A qualitative study grounded in self-determination theory. Mental Health and Physical Activity, 27, 100650. https://doi.org/10.1016/j.mhpa.2024.100650

20.

Wang

Deng

Wang

(2025). The heterogeneity in psychological distress among undergraduate nursing students: a latent profile analysis. BMC nursing, 24(1), 801. https://doi.org/10.1186/s12912-025-03373-6

21.

Wang

Jiang

Wang

Xue

Wang

(2023). Relationship between physical activity and individual mental health after traumatic events: a systematic review. European journal of psychotraumatology, 14(2), 2205667. https://doi.org/10.1080/20008066.2023.2205667

22.

Welkerling

Niess

Schneeweiss

Sudeck

Rohe

Wolf

(2025). Single bout of exercise reduces self-reported and decoded rumination in favor of distraction in patients with major depression. Journal of affective disorders, 397, 120829. https://doi.org/10.1016/j.jad.2025.120829

23.

Xiao

(1994). Theoretical basis and research application of social support rating scale. The Journal of Clinical Psychiatry, 4(2), 98.

24.

C. Y.

Zhu

K. T.

Ruan

X. Y.

Zhu

X. Y.

Zhang

Y. S.

Tong

W. X.

(2023). Effect of physical exercise on sleep quality in college students: Mediating role of smartphone use. PloS one, 18(11), e0288226. https://doi.org/10.1371/journal.pone.0288226

25.

Zhang

Liu

Yin

Kong

Gong

(2025). The effect of exercise motivation on adolescents’ subjective Well-being: chain mediation of physical activity and Social-emotional competency. BMC Public Health, 25(1), 1698. https://doi.org/10.1186/s12889-025-22965-0