Effects of physical activity on anxiety levels in college students: mediating role of emotion regulation

Relationship between physical activity and anxiety

Physical activity is an activity with a certain intensity, frequency, and duration for the purpose of improving physical health, with physical activity as content and means (Li, 2017). The Guidelines on Physical Activity and Sedentary Behavior (2020) issued by WHO clearly suggests the recommendation to enhance physical activity (Bull et al., 2020). Empirical studies have also demonstrated that anxiety levels can be effectively lowered through regular and moderate physical activity (Carek, Laibstain & Carek, 2011; Dong et al., 2022; Li et al., 2023). Facilitating physical activity is a health promotion approach to preventing or treating anxiety disorders (Kandola & Stubbs, 2020). Additionally, studies have found that physical activity with different levels and frequency have different effects on anxiety. High-intensity exercise relieves anxiety more than low-intensity exercise does, and those who exercise more frequently having fewer anxiety symptoms compared to those who exercise less frequently (Ji et al., 2022). Therefore, the existing studies suggest that physical activity negatively predicts anxiety levels.

Mediating role of emotion regulation

Emotion regulation is the process by which individuals regulate their emotions based on the environment they are in, and the way people change their emotional responses (Niu et al., 2023). It consists of extrinsic and intrinsic processes that are responsible for detecting, assessing, and changing emotional responses (Thompson, 1994). Emotion regulation is a multidimensional structure that includes different strategies of emotion regulation, such as reappraisal, expressive inhibition, distraction, and avoidance (Cisler & Olatunji, 2012). Emotion regulation is related to general cognitive abilities and involves brain regions in the prefrontal cortex associated with executive functions. It downregulates the responses of emotion-related areas, including the amygdala (Mohammed, Kosonogov & Lyusin, 2022; Steward et al., 2021). Inadequate emotion regulation has been proved to be an important cause of anxiety and depression (Wirtz et al., 2014). College students who regularly do physical activity show stronger emotion regulation skills (Xue & Zhao, 2018), and emotion regulation is an important antecedent variable that college students commit to physical activity, and it has a significant effect. Therefore, improving the emotion regulation ability of college students can effectively improve the physical activity engagement status (Zhu, Lu & Dong, 2016) Emotion regulation plays a mediating role in the relationship between physical activity and mental health (Jiang, Zhang & Mao, 2018). Moreover, physical activity helps to cultivate positive emotions and deepen the effect of emotion regulation while relieving college students’ anxiety (Xu & Du, 2021). Psychological problems can also be alleviated indirectly by influencing the use of emotion regulation strategies (Xu, Zhang & Yang, 2022).

Objectives and hypotheses of the study

To conclude, although there have been studies that have elaborated the relationship between physical activity and anxiety, physical activity and emotion regulation, and emotion regulation and anxiety from their respective viewpoints at present, very few of them have explored the relationship between physical activity, anxiety, and emotion regulation. There neither have been studies that have explained the roles between the three factors mentioned and the overall effects. Physical activity improves emotion regulation and anxiety from a physiological mechanism perspective by regulating the hypothalamic-pituitary-adrenal axis. Psychological mechanisms suggest that physical activity can enhance cognitive function, and self-efficacy, and reduce anxiety sensitivity (Szuhany, Bugatti & Otto, 2015), which is consistent with the self-efficacy hypothesis (Artino Jr, 2012). Additionally, based on Bandura’s self-regulation theory (Bandura, 1991), emotion self-regulation is reflected in the use of automatic or intentional strategies to generate, maintain, adjust, or express one’s emotions. Based on these hypotheses and theories, an integrated model can be proposed to explain the relationships between physical activity, anxiety, and emotion regulation. By investigating the anxiety levels of college students and exploring the effects of physical activity and emotion regulation on the anxiety levels of college students as well as the mediating role mechanism, this study can provide a basis for alleviating the anxiety of the college students from a multidimensional perspective. On this basis, this study takes college students as the research object and adopts a process-oriented approach to test the direct relationship of physical activity and anxiety and how emotion regulation affects the relationship indirectly. Two main research questions will be clarified in this study: (1) Is there a positive relationship between physical activity and anxiety among college students in China? (2) Does emotion regulation play a mediating role in the relationship between physical activity and anxiety? The following hypotheses were proposed to conduct the study by using the method outlined by Hayes & Matthes (2009). In order to clearly identify the effect differences of emotion regulation strategies playing a role in the physical activity-anxiety relationship, model hypotheses were constructed as shown in Fig. 1.

Hypothesis 1 (H1): Physical activity is a negative predictor of anxiety.

Hypothesis 2 (H2): Physical activity is a positive predictor of emotion regulation.

Hypothesis 3 (H3): Emotion regulation is a negative predictor of anxiety.

Hypothesis 4 (H4): Emotion regulation plays a mediating role in the indirect relational effects of physical activity on anxiety.

Hypothesis 5 (H5): Both cognitive reappraisal and expressive inhibition play a mediating role in the indirect relational effects of physical activity on anxiety.

Participants and survey procedures

Based on the sample size calculation for complex mediation models proposed by Schoemann, Boulton & Short (2017), and referring to the correlation coefficients and mean standard deviations of variables from previous studies, the path coefficients α, β, and c’ were set at 0.125, −0.118, and −0.087, respectively. The results indicated that to achieve a statistical test capacity of 0.8, the minimum sample size required is 874, which this study meets.

Convenience sampling is a commonly used non-probability sampling method in educational and social science research, with advantages such as low workload, low cost, minimal time investment, and simple operation. However, this method has certain limitations. Distributing questionnaires at different locations to achieve diversity can help mitigate these drawbacks, thereby capturing the typical characteristics of the target population (Golzar, Tajik & Noor, 2022). Convenience sampling method was used, and East China, South China, and Central China were selected as the survey areas. The questionnaire survey was conducted in October 2021 and undergraduates of six universities in Shanghai, Jiangsu, Shandong, Guangxi, and Hunan were selected as respondents. The inclusion criteria of the subjects were no mental illness, no use of psychotropic drugs, no contraindications to exercise, no history of head trauma, and no consumption of caffeinated and alcoholic beverages within 24 h of filling in the questionnaire. A total of 1,721 questionnaires were distributed in this study, 84 invalid questionnaires were excluded after verification, 1,637 valid questionnaires were taken back, and the validity rate of the questionnaires was 95.12%.

A questionnaire was used in this study. The questionnaires were distributed by trained principals and collected online and with the assistance of course lecturers. Guided by uniform prompts, students are supervised in the classroom to use cell phones or computers to complete the questionnaires. The recovery rate of the questionnaires was 100%. After the questionnaires were collected, further preliminary screening of the quality of the questionnaires was conducted and the following exclusion criteria were set to obtain a valid questionnaire: (1) Patterned responses: questionnaires where responses exhibit artificial patterns, such as repetitive or consistent answers like “wave-like” or “parallel”; (2) missing data: questionnaires with more than one-third of the responses missing; (3) identical answers: questionnaires where all responses are completely identical. These criteria are set to ensure the validity of the collected questionnaires.

This study was approved by the Ethics Committee of Shanghai University of Sports (102772022RT114). All college students participated voluntarily, and all researchers involved in data collection, data analysis, and manuscript writing were familiarized with the survey procedures and received thorough ethics training prior to data collection.

Measuring tools

Data analysis

Data analysis were conducted by adopting SPSS 24.0. K-S non parametric test was used to test the normality of the data, one-way ANOVA test was used to verify the homoscedasticity of the data, and Pearson Correlation Analysis was performed. The data were tested to be in conformity with normal distribution and satisfied the requirement of homoscedasticity and linearity. Furthermore, as this study adopted a questionnaire self-report format, this may lead to common method biases. Harman’s single-factor test was used to verify the possible common method bias, and the results showed that the first factor explained 19.99% of the total variance, which is less than the critical value of 40%. Therefore, there was no significant common method bias in this study.

After the questionnaire, SPSS 24.0 was used for statistical analysis. Firstly, descriptive statistics and independent samples t-tests were performed on the scores of the two dimensions of anxiety level and emotion regulation of college students respectively. Secondly, one-way multiple analysis of covariance was used to control for BMI, place of student source and other relevant demographic variables, to discuss the relationship between grade, physical activity and the two dimensions of anxiety level and emotion regulation of college students respectively. Lastly, mediating effect regression model Process Model 4 was used (M is said to be a mediating variable between X and Y or M plays a mediating role between X and Y if the effect of the independent variable X on the dependent variable Y is studied through a variable M) to control demographic variables (gender and grade) and to analyze the relationship between physical activity, anxiety level and emotion regulation (Demming, Jahn & Boztug, 2017). The statistical significance criterion was set based on whether the Bootstrap confidence interval (CI) of 5,000 repeat samples included zero. If the CI did not include zero, the effect was considered significant.

Demographic characteristics

The survey included a total of 1,637 university students, comprising 749 males (45.8%) and 888 females (54.2%), with an age range of 18 to 22 years old. Among them, 1,562 students were from the Han nationality (93.2%), and 111 students were from ethnic minorities (6.8%). A total of 822 students (50.2%) were from urban areas, while 815 students (49.8%) were from rural areas.

Basic characteristics of anxiety levels and emotion regulation in college students

The results in Table 1 show that the overall anxiety level of college students is close to the dividing value for anxiety symptoms, with a total score of (44.72 ± 10.37); and emotion regulation is at a moderately low level, with a cognitive reappraisal score of (30.16 ±6.51) and an expression inhibition score of (16.96 ± 4.99).

The results of one-way ANOVA showed that from the perspective of gender, the score of male students’ anxiety level were slightly higher than that of female students’ and male students’ cognitive reappraisal level was slightly lower than female students’, neither of which was significantly different (P > 0.05). Male students’ expression inhibition level was significantly higher than female students’, and there was a significant difference (P < 0.05). From the perspective of grades, the anxiety level was the highest in junior grade and the lowest in the freshman grade, and there was a significant difference between grades (F = 7.649, P < 0.01). The cognitive reappraisal level was the highest in the freshman grade and the lowest in the junior grade, and there was a significant difference between grades (F = 5.092, P < 0.01). However, there was no significant differences in the level of expression inhibition (P > 0.05). Regarding the place of student source, there were no significant differences in anxiety levels, cognitive reappraisal and expression inhibition between urban and rural college students (P > 0.05), indicating that family conditions are not the main factor influencing college students’ anxiety levels and emotion regulation.

Correlation analysis of physical activity, anxiety level and emotion regulation of college students

To test the direct effects of physical activity and emotion regulation on college students’ anxiety levels, bivariate Pearson correlation analysis were conducted for each of the three variables. The correlation coefficients are shown in Table 2.

As can be seen from Table 2, there is a significant correlation between the level of physical activity and anxiety level and the two dimensions of emotion regulation (r = 0.128 ∼0.078, P < 0.01). There is also a significant correlation between the two dimensions of emotion regulation and anxiety level (r = 0.112 ∼0.263, P < 0.01). There was a high degree of correlation between physical activity, anxiety level, and emotion regulation among college students, and they interacted with each other. It indicates that the data are suitable for mediation model test. In addition, the correlation between the three variables were consistent with theoretical expectations, which provides preliminary evidence for subsequent test of the hypotheses.

The effect of physical activity on anxiety level and emotion regulation of college students

As can be seen in Table 3, the differences between anxiety level and emotion regulation and multiple back testing were statistically significant in terms of physical activity. Significant differences were observed in anxiety levels and physical activity levels (F = 6.102, P < 0.01), with the highest level of anxiety in light exercise and the lowest in hard exercise. The effect of light exercise on anxiety levels is greater than that of hard exercise When it comes to the two dimensions of emotion regulation, significant differences were observed in the level of cognitive reappraisal (F = 9.376, P < 0.001), with the lowest level of cognitive reappraisal in light exercise and the highest in hard exercise. Significant differences were also observed in the level of expression inhibition (F = 4.089, P < 0.05), with the lowest expression inhibition levels for medium exercise and the highest for hard exercise. Based on back testing of significant effects, the impact of light exercise on the cognitive reappraisal dimension is smaller than that of medium exercise. Medium exercise has a smaller effect on the cognitive reappraisal dimension compared to hard exercise. Moreover, medium exercise has a greater effect on the expressive suppression dimension compared to hard exercise. Hard exercise consistently shows greater effects on the overall dimension of emotion regulation compared to both light and medium exercise. The results suggest that hard exercise is more favorable for college students to ease anxiety and improve the level of emotion regulation.

Mediation model test of physical activity affecting college students’ anxiety level through emotion regulation

In this study, the mediating effect regression model was constructed with the level of physical activity as the independent variable, the level of anxiety as the dependent variable, and emotion regulation as the mediating variable. In the mediation model of physical activity, anxiety level, and emotion regulation, the test results from the model constructed in Fig. 2 and Table 4 showed that physical activity had a direct negative effect on anxiety level, which means the greater the amount of physical activity, the lower the anxiety level (β = −0.042 P < 0.001). Hypothesis 1 thus held. Physical activity had a significant positive effect on the overall level of emotion regulation (β = 0.05, P < 0.001), and hypothesis 2 thus held. Emotion regulation has a negative predictive effect on anxiety level, i.e., the higher the level of emotion regulation, the lower the anxiety level (β = −0.12, P < 0.001). Hypothesis 3 thus held. Based on the combination of hypotheses 2 and 3, the above results indicate that emotion regulation has a mediating role between physical activity and the anxiety level of college students, so hypothesis 4 held. Emotion regulation plays a mediating role in the relationship between physical activity and anxiety level.

In the mediation model analysis of physical activity, anxiety level, cognitive reappraisal and expression inhibition (as shown in Fig. 3 and Table 5), the effect of physical activity on emotion regulation can be concluded as follows: physical activity had a significant positive effect on cognitive reappraisal (R² = 0.14, β = 0.04, P < 0.001), while physical activity did not have a significant effect on expression inhibition (R ² = 0.17, β = 0.01, P = 0.27). The effects of physical activity and emotion regulation on anxiety are as follows: physical activity (β = −0.0.03, P = 0.012), cognitive reappraisal (β = −0.59, P < 0.001), and expressive inhibition (β = 0.57, P < 0.001) had a significant effect (R² = 0.37) on anxiety levels. The model effect relationship is that the direct and indirect effects of physical activity on anxiety level were −0.745 and −0.354, respectively.

As can be seen in Table 6, physical activity has an indirect negative effect on anxiety levels by affecting cognitive reappraisal, i.e., the greater the amount of physical activity, the higher the level of cognitive reappraisal (β = 0.4, P < 0.001), and the lower the level of anxiety (β = −0.003, P < 0.001). However, physical activity does not have an effect on expression inhibition (β = 0.01, P >0.05); expression inhibition had a direct effect on anxiety levels, and the lower the level of expression inhibition, the lower the anxiety level (β = 0.57, P < 0.001). therefore, hypothesis 5 could not hold.

Anxiety levels and emotion regulation of college students

The results of the study show that the anxiety level among college students differs in terms of grade and gender, which is similar to what previous studies have indicated (Naceanceno et al., 2022). In terms of gender, previous studies have shown that female students have higher anxiety levels than male students do (Gao, Ping & Liu, 2020), while the results of this study shows that male students score slightly higher than female students do in anxiety levels. The potential reason for this may be due to the fact that male students face more pressure from finances, family burdens, employment, etc. when they are about to graduate. In terms of grade level, the anxiety level is the lowest in freshman year, the highest in junior year, and the second highest in senior year. The reason for this phenomenon may lie in the following: freshmen feel freshness and joy about the new environment, and they are in a state of relaxation, so it is normal that the anxiety level is the lowest in this period. The highest anxiety level appears in the third year of college, indicating that the anxiety state of college students has already been ahead of schedule. With the rapid development of the society, the expansion of colleges and universities, and the increase in the pressure of talent competition, the prospect of employment is not optimistic, and have to prepare for the graduation of college, making the anxiety level increase with grades (Fu & Zhai, 2023). The result suggests that schools should pay more attention to the mental health of juniors in advance. In terms of place of student source, the difference in anxiety levels between urban and rural college students is not significant, which means that place of student source is not the main factor affecting anxiety levels.

The two dimensions of emotion regulation show that there is no significant difference in the level of cognitive reappraisal between male and female students, which is consistent with the findings of foreign scholars (Caprara et al., 2008) However, there is a significant difference between male and female students in expression inhibition. Male students’ expression inhibition level is significantly higher than female students’, which is due to the fact that male students tend not to express, while female students express more emotions than male students do (Wang et al., 2013; Gross & John, 1995; Gross & John, 1997; Wang et al., 2019; Zhang & Bian, 2020). This may be related to the fact that family and social environments have different role expectations for male and female students (Jiang, Zhang & Mao, 2018). There is a significant difference in cognitive reappraisal between grades, which may be related to the level of anxiety faced by themselves. The anxiety level is the lowest in freshman year. Freshmen are enthusiastic, resulting in high level of cognition reappraisal. The anxiety level is the highest in junior year. Juniors are depressed, resulting in the lowest level of cognitive reappraisal. There is no significant difference in the expression inhibition between grades, which suggests that college students do not express their emotions easily with the increase in grades. easily express their emotions. The place of student source is not a major factor affecting emotion regulation, indicating that there is no difference in emotion regulation among college students regardless of whether they come from rural or urban areas.

Relationship between physical activity and anxiety

Studies have found that physical activity have a direct negative effect on anxiety levels. The effect was even more significant as the amount of exercise increased and anxiety levels among college students decreased dramatically, which is in line with the results of the study by Ji et al. (2022) on the effect of physical activity on the anxiety of college students. The relief of anxiety in the high-intensity exercise group was significantly higher than that in the low-intensity exercise group (Ji et al., 2022). The potential reasons why physical activity is effective in relieving anxiety levels are as follows. Firstly, exercise promotes an increase in brain-derived neurotrophic factor (BDNF), which promotes neuroplasticity, neuronal growth and differentiation (Ji et al., 2022). Secondly, exercise stimulates the secretion of endorphins, a chemical in the brain. It is a naturally occurring substance that affects mood and makes people feel exciting, which makes people feel relaxing and joyful, and effectively reduces negative emotions such as anxiety (Duan et al., 2014). The larger amount of exercise can facilitate the brain and the whole body’s blood circulate, make people feel exciting, which is beneficial to college students to relieve psychological pressure, distract their attention, and thus effectively reduce the anxiety level of college students.

The results of this study are slightly different from other studies. Medium amount of exercise improves and reduces anxiety levels of college students (Duan et al., 2014; Ma, 2017; Yu, 2008). However, medium and small amount of exercise may be more helpful to reduce the anxiety level (Li, 2005). This may be related to the way, program, and time of participating in physical activity. It is also related to having exercise habits or not, because different forms of physical activity, such as aerobic exercises, strength exercises, etc.; different kinds of exercise programs, such as running, playing ball games, Tai Chi, yoga, etc.; and different exercise times and exercise intensities may contribute to different degrees of psychological benefits. There is also an interaction between these influencing factors (Li, 2005; Yu, 2008). It may also be related to the physical activity rating scale used. The scale used in this study (PARS-3) was revised by Liang Deqing et al. from Wuhan Sports University, and the amount of exercise is divided into grades by intensity, time, and frequency, and the amount of exercise =intensity × time × frequency. The amount of exercise in the existing studies is divided into light, medium, and high intensity based on the Rating of Perceived Exertion (RPE) scores(Li, 2005), or in the form of a self-report scale (Xie, Zhao & Wang, 2002). As a result, the degree of quantification of the amount of physical activity may vary. In addition, the physical activity scenario is also an important factor influencing the amount of exercise. A period of physical activity at a medium amount of exercise is effective in relieving anxiety in both the autonomous and cooperative scenarios, but physical activity performed in the cooperative scenario relieves anxiety better than physical activity in the autonomous scenario (Li, 2013; Johnston et al., 2021).

Mediating role of emotion regulation

In the relationship between physical activity and the effect of emotion regulation among college students, the amount of the exercise has a significant effect on emotion regulation. With the increase of the amount of exercise, the level of cognitive reappraisal of college students is enhanced. The lowest level of expression inhibition is found in the medium amount of exercise, and the highest level of expression inhibition is found in the large amount of exercise. This suggests that increasing the amount of physical activity moderately can effectively enhance the effect of emotion regulation among college students, and that the medium amount of physical activity is a key factor influencing the effect of emotion regulation among college students, which is consistent with the results of Xue & Zhao (2018). Research in cognitive neuroscience has found that the brain regions activated by aerobic exercise and those of emotion regulation overlap in the ventral and dorsal prefrontal and anterior cingulate cortex regions (Etkin, Buechel & Gross, 2015), revealing the brain-neural mechanisms by which aerobic exercise affects emotion regulation. Thus, physical activity enhances emotion regulation, such as the generation of emotion regulation strategies (Brand et al., 2018). In addition, we found that college students with high levels of physical activity frequently use cognitive reappraisal of emotion regulation strategies, which enable them to increase the effectiveness of cognitive reappraisal, and blunt emotional responses to negative emotional cues. It is typically associated with a thicker anterior cingulate cortex in the right hemisphere of the brain (Rothermund & Koole, 2011; Wu et al., 2022). However, our findings showed that physical activity did not have a significant effect on expression inhibition and previous findings did not show that the thickness of thicker anterior cingulate cortex in the right brain was associated with expression inhibition in college students, which may be due to the fact that the participants were in early adulthood and there was less direct correspondence between their anterior cingulate cortex of right brain and expression inhibition. Based on this, physical activity mainly improves college students’ emotion regulation by influencing cognitive reappraisal of emotion regulation strategies and thus improving their emotion regulation.

The present study found that emotion regulation is a negative predictor of anxiety level. In other words, anxious patients engaged in less reappraisal, which is similar to what previous studies have shown. The close connection between the prefrontal cortex, striatum, and amygdala gets mature when emotion regulation is enhanced, which enhances prefrontal control over emotionally reactive subcortical areas, and in turn improves the ability of negative emotion regulation (Ahmed, Bittencourt-Hewitt & Sebastian, 2015; Ernst, 2014).

The mediating regression model constructed in this study confirms the mediating effect mechanism of emotion regulation between physical activity and college students’ anxiety level, i.e., physical activity can not only directly lower the anxiety level of college students, but also indirectly lower it by improving their cognitive reappraisal level, which is similar to the results of the existing research conducted by Jiang, Zhang & Mao (2018). A number of studies have shown that physical activity can play a role through emotion regulation, which can ease anxiety and release repressed emotions, and can effectively reduce or eliminate college students’ anxiety (Hu et al., 2004; Lin & Xu, 2020; Xu & Xu, 2010). The results of this study further indicate that physical activity has a significant positive effect on the cognitive reappraisal of college students. Besides, physical activity and cognitive reappraisal have a significant negative effect on the anxiety level of college students, which suggests that the effect of physical activity on the anxiety level of college students is mediated by the cognitive reappraisal of emotion regulation. Therefore, enhancing the emotion regulation ability of college students is an important way to leverage the strength of physical activity on college students’ anxiety level, which can not only directly lower the anxiety level by promoting the secretion of endorphins and other hormones through physical activity, but also lower the anxiety level of college students by enhancing their sense of satisfaction and joyful emotional experience in physical activity (Zhu, Lu & Dong, 2016).

Limitations and influences

There are some limitations to this study. Firstly, the study utilized convenience sampling, which may introduce sample biases and systematic errors. Future research should expand to larger-scale surveys to validate the findings of this study. Secondly, subjective self-perception assessments have inherent limitations. Future research suggests that measurements should be made with the help of scientific instruments, such as accelerometers. Lastly, the cross-sectional study design inherently limits establishing causal relationships. Future research should include longitudinal intervention experiments to further investigate causal relationships between variables.

Overall, this study theoretically supports the existing research on the relationship between physical activity and college students’ anxiety and enriches the relevant theoretical support on anxiety research. In addition, the constructed mediating regression model of physical activity, anxiety, and emotion regulation reveals the interaction between the three and provides an empirical framework. These findings offer theoretical support for understanding the potential mechanisms underlying the relationship between physical activity and anxiety among college students.