Social support is a key variable affecting young people's physical exercise activities (Courneya et al., 2000). Adequate social support is an effective way to cultivate college students to participate in physical exercise (Alshehri et al., 2021). Dong et al. found that social support can buffer the negative impact of physical exercise discomfort on the body and mind, and enhance the sense of pleasure and satisfaction brought by exercise (Dong et al., 2018). Family support and significant others support as important factors for exercise adherence behavior (Courneya and Mcauley, 1995). College students who perceive social support can gain intrinsic motivation to actively participate in physical exercise. Based on this, this study proposes the following hypotheses:

Social support and psychological capital are not independent of each other. The theory of positive psychological capital proposes that supportive environments have a positive impact on psychological capital (Luthans et al., 2008). Psychological capital can regulate the impact of social support on physical exercise, and people with higher hope, self-efficacy, and respect benefit more from support, making them more likely to convert support into physical exercise behavior (Kang et al., 2019). Research has shown that a satisfactory social environment and interpersonal support are beneficial for increasing individuals' confidence in participating in sports activities (Ren et al., 2020). Therefore, the following hypotheses are proposed:

Psychological capital has a certain correlation with behavioral variables. By increasing the content of individual psychological capital, it can effectively stimulate people's subjective initiative and provide a continuous source of motivation for the sustainability of individual physical exercise behavior (Teixeira et al., 2012). The higher the psychological capital content of college students, the more positive their attitude toward physical exercise is. The easier it is for them to persist in physical exercise, which helps to shape a healthy personality (Yang et al., 2013). Li et al. also believes that the higher the level of psychological capital among college students, the higher their attitude toward sports, which in turn promotes good physical exercise behavior among individuals (Li, 2019). Therefore, the following hypotheses are proposed:

In summary, in order to explore the relationship between social support, psychological capital, and physical exercise behavior, this study aims to construct a chain mediation model (Figure 1).

A questionnaire survey was conducted among college students in Chongqing using both on-site and online methods. The survey was carried out in two stages. In the first stage, 145 students from two classes at Southwest University were selected for the survey. After removing 10 invalid questionnaires, 135 valid questionnaires were obtained and used for reliability and validity testing. Upon successful completion of the testing, the second stage of the formal investigation commenced. In the second stage, a random sampling method was employed, with a total of 385 questionnaires distributed and 359 valid questionnaires collected. The responses to the questions were measured using a 5-point Likert scale, where 1–5 points corresponded to “strongly agree,” “agree,” “neutral,” “disagree,” and “strongly disagree,” respectively. When analyzing the physical exercise behavior of college students, the duration, intensity, and frequency of independent physical exercise were considered, and scores ranging from 1 to 5 were assigned.

All the scales employed have corresponding Chinese versions, and the following offers a detailed elaboration.

Data analysis was conducted using IBM SPSS 26.0 and STATA 17.0. The following steps were taken to ensure the robustness and validity of the results:

Among the 359 participants, those aged 18–22 accounted for 71.87% of the total, with a higher proportion of males (57.10%) than females. In addition, students majoring in psychology, education, agriculture, pharmacy, and other fields account for 65.18%, while students majoring in sports account for 34.82%.

Confirmatory factor analysis (CFA) was used to verify the construct validity. When AVE > 0.5 and CR > 0.7, it indicated convergent validity and composite reliability (Sum et al., 2018). Independent confirmatory factor analysis was conducted on peer support, teacher support, family support, psychological capital and physical exercise behavior respectively. After eliminating the indexes with factor load < 0.5, the results showed that peer support (AVE = 0.72, CR = 0.89), teacher support (AVE = 0.75, CR = 0.93), parent support (AVE = 0.73, CR = 0.92), psychological capital (AVE = 0.50, CR = 0.98), physical exercise behavior (AVE = 0.76, CR = 0.92) show that the model converges effectively. In addition, we examined the degree of fit between the measurement model and the actual collected data, and the results showed that χ2/df = 3.638, RMESA = 0.052, GFI = 0.913, AGFI = 0.901, NFI = 0.923, CFI = 0.942, IFI = 0.943. All the fitting indexes are higher than the standard of 0.9, indicating a good degree of fitting.

Table 6 examines the mediating role of psychological capital. The results show that peer support, teacher support, and family support are significantly positively correlated with psychological capital (P < 0.01). The indirect effects of social support on college students' physical exercise behavior, mediated by psychological capital, account for 87.34%, 36.95%, and 30.69% of the total effects, respectively. In other words:

These findings confirm that psychological capital plays a significant mediating role in the relationship between social support and college students' physical exercise behavior.

The results indicate that peer support shows the strongest positive correlation with college students' physical exercise behavior. Peer relationships play a crucial role in motivating college students to engage in physical exercise (Bull et al., 2020). which can be attributed to the following reasons: First, the similarity in health behaviors among peers and the exemplary role of peers effectively encourage individual participation in sports activities (Ji et al., 2022). Specifically, when college students join various activities or sports clubs, the sports atmosphere created by their peers often inspires them to participate in physical exercise collectively. Additionally, peer support is positively associated with college students' happiness and satisfaction during physical exercise, fostering positive emotional experiences, boosting confidence, and facilitating the acquisition of new sports skills (Chen et al., 2017). Haidar et al., focusing on college students as their research subjects, explored the relationship between peer support and physical exercise. Their findings indicate that every unit of peer support increases the likelihood of college students participating in physical activities by at least 1.15 times (Haidar et al., 2019). The findings of this paper align with the conclusions of the aforementioned scholars, confirming that peer support has a significant positive impact on college students' physical exercise behavior.

The association between teacher support on college students' physical exercise behavior is second only to peer support. For students, school is the most significant environment in their daily lives and the primary setting where they are likely to engage in physical exercise. Consequently, sports activities exhibit strong correlations with multiple facets of school life, including the promotion of students' values and the shaping of their ideological consciousness. Teachers, as direct transmitters of knowledge and values, play a pivotal role in students' development. Perceived teacher support is a key variable positively correlated with personal psychological factors—including goal orientation and achievement motivation (Ghorbani et al., 2020). I Moreover, teacher support helps students better understand motor skills and reduces psychological barriers to participation. For instance, teachers design exercises and competitive activities tailored to students' ability levels, enabling them to experience moderate challenges and a sense of achievement, thereby enhancing their interest in sports (Liu et al., 2025). Chatzisarantis et al. conducted a large-scale intervention study, confirming the positive impact of teacher support on students' participation in extracurricular sports activities. Compared to students under neutral teacher support conditions, those receiving active teacher support demonstrated a stronger willingness to exercise during leisure time and participated in sports activities more frequently (Chatzisarantis and Hagger, 2009). Therefore, teacher support not only provides students with a sense of security but also significantly promotes their engagement in physical exercise.

Compared with teacher support and peer support, family support has a relatively weaker impact on college students' physical exercise behavior. While the family is a primary environment for children's lives alongside school, it remains an important source of support for college students' physical activity behavior (Belanger and Patrick, 2018). Research indicates that family support positively promotes adolescents‘ participation in sports activities, adherence to physical exercise routines, and acquisition of sports skills (Jiang and Xiao, 2024). Dong et al. found that parental support enhances adolescents' control beliefs, not only encouraging their active participation in exercise but also influencing their level of engagement and strengthening their persistence in exercise (Dong and Mao, 2018). However, during college, factors like physical distance and emotional independence weaken the association between family support and students‘ physical exercise behavior. This shift in dynamics explains why family support plays a less prominent role in shaping college students' exercise behavior compared to peer and teacher support.

Further research has revealed that social support can indirectly predict college students' physical exercise behavior through the mediating effect of psychological capital. The dual engines driving college students' physical exercise behavior are internal psychological appeals and external perceptual support. The interaction between these internal and external factors effectively enhances college students' health awareness and behavior (Dong et al., 2018). The more support students receive from significant figures during fitness activities, the stronger their self-esteem protection becomes. This confidence helps them eliminate negative emotions during sports activities, thereby increasing their psychological capital and fostering self-awareness and autonomy in exercise (Zhou et al., 2014). In this study, teacher support, peer support, and family support were found to promote physical exercise behavior among college students by enhancing their psychological capital, including self-efficacy, resilience, hope, and optimism. Specifically, teacher support is positively associated with students' self-efficacy, which in turn correlates with greater enthusiasm for sports participation (Dutrisac et al., 2023). Peer support strengthens students' optimism through emotional, informational, and instrumental support, which further promotes individual physical exercise behavior (Luo et al., 2022). Family support, particularly through a positive sports atmosphere, significantly increases students' confidence in physical exercise and encourages their participation and persistence in such activities (Jiang and Xiao, 2024). These findings confirm that psychological capital plays a mediating role in promoting college students' physical exercise behavior. The mediation model constructed in this study is feasible and provides valuable insights into the internal mechanisms through which social support improves college students' physical exercise behavior. Additionally, it offers practical guidance for initiatives aimed at enhancing college students' engagement in physical exercise.

The results of this study are limited to the mediating role of psychological capital in the relationship between social support and physical exercise behavior among college students. Other potential mediating variables, such as family environment, self-control, and health concepts, remain unexplored and warrant further investigation in subsequent research. Additionally, the study focuses exclusively on college students over 19 years old, resulting in a relatively homogeneous sample. Caution is necessary when extending variables such as social support and physical exercise behavior to other age groups. Future research should aim to expand the sample range to include diverse populations. Furthermore, this study relies on self-reported data from college students, which may introduce method bias and affect the validity of the findings. To address this limitation, future research could consider incorporating perspectives from teachers, peers, and parents as additional research subjects. This multi-source approach would provide a more comprehensive understanding of the relationships between the variables.