The effect of physical activity on resilience of Chinese children: the chain mediating effect of executive function and emotional regulation

Abstract

Background

Children’s mental health has become a major public health challenge, with approximately 19.3% of Chinese children and adolescents facing mental health problems. Resilience, a positive adaptive trait, serves as a key protective factor in children’s mental health. While studies indicate that physical activity (PA) enhances resilience, its underlying mechanisms remain unclear. This study aims to examine the relationship between PA and resilience, with a focus on the mediated roles of executive function (EF) and emotion regulation (ER).

Methods

A cross-sectional study was conducted with a sample of 971 primary school students from grades 2 to 6 (mean age = 9.8 years) in Yangzhou City, Jiangsu Province. The Physical Activity Rating Scale (PARS-3), the Chinese Adolescent Resilience Scale (RSCA), the Emotion Regulation Strategies Questionnaire for Children and Adolescents (ERQ-CA), and the Computerized Neuropsychological Assessment System were used to evaluate the relevant variables. SPSS 29.0 and AMOS 29.0 were used for data analysis and structural equation model testing.

Results

The results revealed significant positive correlations between PA and resilience (r = 0.545, p < 0.01) and ER (r = 0.413, p < 0.01), as well as a significant negative correlation with EF (r = -0.341, p < 0.01), where shorter reaction times indicate better EF. PA was found to have a significant direct effect on resilience (β = 0.294, p < 0.001). Bootstrap mediation analysis revealed that EF and ER jointly formed a significant chain-mediated pathway between PA and resilience, accounting for 17.39% of the total indirect effect. Among them, emotion regulation played the most prominent independent mediating role, contributing 68.38% of the indirect effect.

Conclusions

This study elucidates the mechanisms through which PA influences children’s resilience via executive function and emotion regulation, highlighting the crucial role of cognitive and ER skills in this process. These findings not only enhance the understanding of the mechanisms underlying children’s resilience development but also offer new perspectives for designing interventions to promote children’s mental health. Future educational and public health policies should integrate PA with cognitive-emotional training to effectively enhance children’s ability to cope with stress and challenges.

Introduction

According to the World Health Organization (WHO), about one-third of global mental health problems occur in childhood, posing a major challenge for public health worldwide [1]. In China, rapid social transformation, increasing educational competition, the nuclearization of the family structure, and the widespread use of digital technology have exacerbated children’s mental health issues [2]. Epidemiological studies indicate that the prevalence of mental illness among Chinese children and adolescents has reached 19.3% [3], making them a leading cause of non-fatal disability in this population [4]. More concerningly, the problem is showing a trend towards younger age groups [5]. Primary school children are in a critical transitional phase from childhood to adolescence, a period of heightened risk for psychological, emotional, and behavioral problems. Physical, psychological, and interpersonal changes during this stage increase children’s susceptibility to negative emotions, including depression and anxiety [6, 7]. Mental health problems in children not only impair their learning and social functioning but also have long-term consequences for psychological and social adjustment, as well as health behaviors in adulthood. Additionally, these issues may elevate the risk of extreme behaviors [8], including self-harm and suicide, imposing a substantial burden on individuals, families, and society. Given the growing challenges to children’s mental health, reducing the incidence of mental illness and identifying and strengthening protective factors have become key research priorities.

Resilience, a positive adaptive capacity, enables children to maintain or restore their mental health when facing adversity and serves as a key protective factor in coping with stress and change [9]. This study is grounded in Richardson’s Metatheory of Resilience and Resiliency [10], which conceptualizes resilience not merely as a trait but as a dynamic adaptive process. According to this theoretical framework, resilience develops through cycles of disruption and reintegration, where individuals encounter stressors, experience some degree of disorganization, and then reintegrate with varying degrees of resilience depending on available protective factors. Richardson’s model identifies three levels of protective factors: biological (such as neurobiological systems), psychological (such as cognitive and emotional processes), and environmental (such as physical activity, social support).

Consistent with this theoretical framework, individuals with high resilience demonstrate notable psychological strengths and can flexibly mobilize internal and external resources to convert negative experiences into motivation for personal growth [11], thereby reducing the risk of negative psychological symptoms such as depression and anxiety [12–14]. Therefore, resilience is particularly important for children and adolescents, not only helping them better cope with challenges during their growth but also promoting the development of their psychological and social adaptability. Cultivating resilience enables individuals to develop more effective coping mechanisms, enhance their adaptability to challenges, and ultimately achieve better psychological well-being [15, 16]. Given the critical protective role of resilience in children’s mental health, exploring effective strategies to enhance resilience and its underlying mechanisms is of great significance. This study aims to investigate the key factors influencing children’s resilience and their mechanisms of action, with particular focus on the interrelationships among behavioral, cognitive, and emotional factors, aligning with Richardson’s emphasis on understanding the interplay between different protective systems [17]. The findings will provide a scientific foundation and practical insights for interventions aimed at promoting children’s mental health.

Physical activity and resilience

Resilience is a relatively stable positive psychological trait that enables individuals to cope with stress or adversity, and to maintain and promote their healthy growth and well-being [18]. In the student population, resilience not only affects individuals’ academic achievement and social adaptation, but also has a long-term impact on psychological health and quality of life in adulthood [19]. According to resilience theory, psychological resilience is a dynamic adaptive process, and individuals can continuously develop and strengthen their psychological resilience through a variety of coping strategies and protective factors in the face of pressure and challenges [20, 21]. In recent years, increasing research has explored the impact of environmental factors on psychological resilience, including social support, educational interventions, and lifestyle adjustments [10]. Among these highly malleable external factors, physical activity(PA) is widely recognized as a key contributor to promoting individual resilience and has been shown to have positive effects across different age groups [22, 23].

PA refers to any bodily movement caused by skeletal muscle contraction that increases energy expenditure [24], encompassing exercise, leisure activities, and daily physical tasks. From the perspective of resilience theory, PA acts as a key external protective factor that fosters resilience by enhancing individuals’ coping abilities and stress resistance. Extensive empirical research has demonstrated that regular moderate-to-high-intensity PA significantly enhances resilience levels [25]. For example, a longitudinal study on adolescents found that those who maintained a regular PA routine exhibited higher resilience levels than their sedentary counterparts [26]. Furthermore, systematic reviews have identified a significant positive correlation between PA and resilience, with higher exercise intensity yielding greater resilience improvements [27]. Additionally, different types of PA may influence resilience through distinct mechanisms. Research has shown that aerobic exercise reduces depressive symptoms and improves emotional stability [28]. Flexibility training reduces anxiety and improves stress coping skills [29]. These findings suggest that different PA modalities may collectively enhance resilience through distinct mechanisms, including mood regulation, cognitive resource enhancement, and improved stress coping.

Beyond these behavioral pathways, emerging neuroscience research has begun to elucidate the neurobiological foundations underlying PA’s effects on resilience. Recent neuroscience research suggests that PA may influence resilience through neurobiological pathways, including enhanced neuroplasticity and improved prefrontal-limbic connectivity [30–32].Meta-analytic evidence further suggests that exercise-induced neuroplasticity may be particularly pronounced during childhood and adolescence, given the heightened malleability characteristic of developing neural systems [33]. However, existing research has focused on adolescents and adults, with limited studies on the effects of PA on children’s resilience. Given that childhood is a critical period for resilience development, this study proposes Hypothesis H1: PA positively predicts children’s resilience levels.

The mediating role of executive function

Executive function (EF), as a higher-order cognitive function, encompasses the initiation, adaptation, regulation, monitoring, and control of information processing and behavior [34]. These abilities develop rapidly throughout childhood and adolescence, playing a crucial role in academic achievement, social adaptation, and mental health [35]. In recent years, a growing body of research has examined the positive impact of PA on executive function, as evidenced by randomized controlled trials and meta-analyses. For example, a 12-month resistance training intervention led to significant improvements in EF [36]. Additionally, acute aerobic exercise has been shown to rapidly improve cognitive and EF in individuals [37]. Meta-analyses have further confirmed that various types and intensities of PA exert positive effects on EF across different age groups [38].

EF enables flexible cognitive processing and rapid adaptation to environmental changes, which are essential components in resilience development [39]. Studies have found that individuals with stronger EF demonstrate enhanced abilities such as maintaining focus, resisting impulsive reactions, and navigating new and unexpected challenges [35, 40] - skills that are conceptually aligned with resilience development. Consistent with this perspective, Shariati’s research documented significant associations between EF abilities and resilience, finding that individuals with higher EF scores demonstrated superior abilities to flexibly adjust their strategies in response to challenges, more effectively manage stressful events, and exhibit higher resilience levels [41].

From a theoretical perspective, this relationship can be understood through self-regulation theory, which positions EF as a central component of an individual’s capacity to regulate thoughts, emotions, and behaviors in response to environmental demands [42]. This theoretical framework suggests that EF may relate to resilience through enhancing cognitive control processes that are valuable when responding to stressors and challenges [35].Importantly, research has begun to explore how physical activity might influence resilience through cognitive pathways. A recent study by Sibbick found that the beneficial effects of PA on resilience appear to be partially mediated by improvements in cognitive function [43], suggesting that cognitive enhancement might be one mechanism through which PA contributes to resilience development. Based on this theoretical foundation and empirical evidence, this study proposes hypothesis H2: EF mediates the relationship between PA and resilience in children.

The mediating role of emotion regulation

Emotion regulation (ER) refers to the process through which individuals modulate their emotional responses using various cognitive and behavioral strategies [44], playing a crucial role in mental health and adaptive development. Research has shown that ER ability not only affects an individual’s emotional stability, but is also closely related to the way he or she copes with stress [45]. Effective ER is considered a protective factor for psychological resilience [46], helping individuals to cope more effectively with negative emotions and to adapt positively in the face of adversity. This capacity to regulate emotions people expands cognitive perspectives, cultivating greater receptiveness to new information while enhancing positive emotional experiences [47]. In addition, emotion regulation not only influences individuals responses to short-term stress, but also plays a role in long-term adaptation. Studies suggest that the emotional control and positive cognition dimensions of resilience largely depend on effective ER strategies, with individuals possessing strong ER skills typically exhibiting higher resilience levels [48].

On the other hand, PA is regarded as an important external factor for promoting ER. PA stimulates the release of dopamine and serotonin, helping to reduce anxiety and depression while enhancing emotional stability [49]. Moreover, sport serves as a positive ER strategy, fostering well-being and vitality while mitigating negative emotions [50]. While studies have examined the positive effects of PA on ER and resilience, there remains a lack of systematic investigation into whether ER mediates the relationship between PA and resilience. In summary, this study proposes hypothesis H3: ER mediates the relationship between PA and resilience in children.

The chain mediating role of executive function and emotion regulation

EF refers to the psychological process through which individuals exert conscious control over their thoughts, emotions, and behaviors [51], while ER involves a series of internal and external mechanisms for monitoring, evaluating, and adjusting the intensity and duration of emotional responses [52]. As essential components of an individuals cognitive and emotional self-regulation, these two elements develop rapidly during childhood and early adolescence, and are closely linked to subsequent mental health [53] and academic achievement [54].

According to the cognitive theory of emotion, emotional generation is influenced by three key factors: environmental events, physiological conditions, and cognitive processes. Among these, cognition plays a central role in determining emotional experiences [55]. As a higher-order cognitive function that modulates other cognitive processes, EF enables children and adolescents to regulate their thought processes effectively, guiding them in making rational emotional appraisals across various situations and events. This cognitive regulatory ability is essential for children and adolescents in managing daily challenges, including academic pressure and social interactions. Furthermore, it serves as a critical foundation for their socio-emotional development. A longitudinal study on primary school student demonstrated that EF at T1 and T2 significantly predicted ER at a later stage [56].

The current study

Although previous research has examined the relationship between PA and resilience, the underlying mechanisms remain insufficiently understood. Specifically, how PA influences EF and ER, and whether these factors mediate its relationship with resilience, remains underexplored in children. More importantly, identifying the pathways through which PA enhances resilience is crucial, as childhood represents a critical period for its development—a stage during which psychological difficulties may persist into adulthood, leading to substantial personal and societal costs. These findings will substantially enhance our understanding of resilience development mechanisms and provide a foundation for evidence-based interventions aimed at promoting children’s mental health. This study aims to construct a research model of how PA, EF, and ER potentially influence resilience in children (Fig. 1) and to test the hypotheses proposed.

Fig. 1.

The hypothetical structure model

Methods

Participants

This study used a stratified cluster convenience sampling method to select students from grades 2 to 6 at a primary school in Yangzhou City, Jiangsu Province. The test was conducted collectively, with each class as a unit. The trained postgraduate student administered the test, explained the requirements, and guided students in completing the questionnaire, which was collected immediately after completion. A total of 1,025 questionnaires were collected, with 971 valid responses retained after screening. Participants ranged in age from 7 to 12 years, with a mean age of 9.8 years. The sample included 499 boys (51.4%) and 472 girls (48.6%). The study was approved by the Ethics Committee of the Medical College of Yangzhou University (Approval No.: YXYLL-2023-129). Written informed consent was obtained from the parents or legal guardians of all participants. Age-appropriate instructions were provided to participants, who participated voluntarily. The study was conducted in accordance with the Declaration of Helsinki (1964) and related guidelines and regulations.

Measures

Physical activity scale

This study assessed physical activity levels using the previously published Physical Activity Rating Scale (PARS-3) developed by D.C. Liang [57]. This scale has been widely applied in Chinese children and adolescents [58, 59]. The scale evaluates three dimensions: exercise frequency, duration, and intensity. A 5-point Likert scale (1 to 5) was used. The total score was calculated using the formula: physical activity score = exercise frequency score × (exercise time score − 1) × exercise intensity score.The total score ranges from 0 to 100, with higher scores reflecting greater physical activity levels. The original PARS-3 developed by Liang demonstrated good reliability and validity among Chinese populations, with reported Cronbach’s alpha ranging from 0.82 to 0.89 and acceptable criterion validity. The scale demonstrated good reliability in this study, with a Cronbach’s alpha of 0.815.

Resilience scale

The previously published Resilience Scale for Chinese Adolescents (RSCA), developed by Hu Yueqin et al., was used to assess resilience [60]. This scale has been extensively used in previous studies with Chinese populations [61, 62]. The scale consists of 27 items across five dimensions: focus on goal, emotional control, positive cognition, family support, and interpersonal assistance. Responses were rated on a 5-point Likert scale (1–5). Reverse-scored items were adjusted before computing the total score, with higher scores reflecting greater resilience. In the original validation study by Hu, the RSCA showed excellent psychometric properties with Cronbach’s alpha of 0.85 for the total scale. The scale demonstrated good reliability in this study, with a Cronbach’s alpha of 0.908.

Executive function test tasks

Executive function was assessed using a computerized neuropsychological assessment system specifically developed for Chinese children by Professor Aiguo Chen [63]. It has been widely applied in previous studies involving Chinese children and adolescents, demonstrating stable psychometric properties and suitability for small-group administration [64, 65]. During the test, participants were instructed to press a key as quickly and accurately as possible. EF was primarily evaluated based on reaction time. Shorter reaction times indicated better EF.

Inhibition was assessed using the Flanker task, which included two trial types: consistent (e.g., LLLLL, FFFFFF) and inconsistent (e.g., LLFLL, FFLFF). Participants were instructed to quickly and accurately identify the middle letter.The response time difference was calculated by subtracting the consistent trial reaction time from the inconsistent trial reaction time. Smaller response time differences and higher accuracy reflect better inhibition performance.

Refresh was assessed using the 1-back task. Participants viewed a letter on the screen (B, D, L, Y, or O) and quickly determined whether it matched the previously shown letter. Shorter reaction times indicated better updating performance.

Shifting was assessed using the More-Odd shifting task, which included two trial types: homogeneous (e.g., large/small or odd/even judgments) and heterogeneous (e.g., switching between large/small and odd/even judgments). The test score was the mean reaction time in the switching condition compared to the no-switching condition. A smaller difference indicated better switching performance.

Emotion regulation scale

This study used the previously published Emotion Regulation Strategies Questionnaire for Children and Adolescents (ERSQ-CA). The English version developed by Gullone and Taffe based on the ERQ [66], was adapted into Chinese by Chen Liang [67]. This scale has been validated and applied in previous research studies [68, 69]. The questionnaire includes two subscales: cognitive reappraisal and expressive suppression, comprising 10 items. Responses were rated on a 5-point Likert scale (1–5), ranging from ‘totally disagree’ to ‘totally agree’. The Chinese version adapted by Chen Liang showed reliability with alpha coefficients of 0.87 and 0.71 for the respective subscales. The scale demonstrated good reliability in this study, with a Cronbach’s alpha of 0.893.

Reliability test

The validity of the measurement model was assessed using standardized evaluation methods. Reliability was evaluated using three key metrics: the Cronbach’s alpha coefficient, the combined reliability (CR) value, and the average variance extracted (AVE). Specifically, Cronbach’s alpha measures internal consistency, CR reflects inter-item correlation, and AVE assesses convergent validity. Empirical tests indicated that all scales met reliability and validity criteria: Cronbach’s α > 0.8, CR > 0.7, and AVE > 0.5.These results indicate that the measurement model has robust reliability and convergent validity, thus confirming the reliability of the findings. Table 1 presents the detailed reliability analysis results.

Table 1.

Reliability and validity testing of the scale

Variable	Factor loading	AVE	CR
focus on goal	0.831,0.841,0.817,0.777,0.803	0.663	0.908
emotion control	0.834,0.828,0.821,0.854,0.858,0.869	0.713	0.937
positive cognition	0.831,0.794,0.803,0.830	0.664	0.888
family support	0.836,0.824,0.848,0.831,0.855,0.844	0.705	0.935
interpersonal assistance	0.797,0.827,0.821,0.849,0.867,0.854	0.699	0.933
cognitive reappraisal	0.813,0.811,0.788,0.796,0.811,0.826	0.652	0.918
expressive suppression	0.846,0.824,0.837,0.811	0.688	0.898
physical activity	0.735,0.826,0.766	0.603	0.820
executive function	0.931,0.957,0.909	0.869	0.952

Validation factor analysis

Confirmatory factor analysis (CFA) was performed in AMOS 29.0, and model fit results are presented in Table 2. The χ²/df = 2.70, below the 5.000 threshold, suggesting an acceptable model fit. The RMSEA = 0.042, below 0.080, indicating good model fit. Other fit indices exceed the 0.900 threshold. These results confirm a good fit between the sample data and the theoretical model, supporting further empirical analysis.

Table 2.

Model fit Indics

Evaluation indicators	Model Fit value	Judgment standard
χ2/df	2.70	< 5.000, acceptable;<3.000,good fit
RMSEA	0.042	< 0.080, good fit
CFI	0.975	> 0.900, good fit
GFI	0.975	> 0.800, acceptable;0.900, good fit
AGFI	0.961	> 0.800, acceptable;0.900, good fit
NFI	0.973	> 0.900, good fit
IFI	0.983	> 0.900, good fit

Data analytic strategy

Amos 29.0 software was utilized to conduct structural equation modeling (SEM). The model’s suitability was assessed, ensuring that the sample size was at least ten times the number of measured variables. With 43 variables and an effective sample size of 971 participants, the data met the necessary analytical requirements. Statistical significance was determined at a threshold of p < 0.05. Additionally, the Bootstrap method was applied to examine mediation effects, and results were reported with a 95% confidence interval.

Results

Common method bias test

This study may have common method bias due to self-reported data collection. Therefore, this study used both procedural control and statistical control to reduce the occurrence of common method bias. Procedural control ensured adherence to psychometric testing standards, including reliable scales, anonymous administration, and other safeguards. For statistical control, Harman’s single-factor test was conducted on the collected data. The results identified seven factors with eigenvalues greater than 1. The first factor explained 26.95% of the variance, below the 40% threshold [70], indicating no serious common method bias.

Analysis of group differences in physical activity, resilience, executive function, emotional regulation

The results of the Mann-Whitney U-test for gender and the Kruskal-Wallis H-test for grade level are shown in Table 3, where PA differed between grades (H(4) = 12.206, p = 0.016). Post-hoc analyses using Dunn’s test with Bonferroni correction revealed that 4th grade students had significantly higher physical activity levels compared to 3rd grade students (p = 0.018, effect size r = 0.28) [71]. No other significant pairwise differences were observed between grade levels (p > 0.05). Resilience, ER, and EF did not differ significantly between grades. Gender differences in PA, resilience ER, and EF were not significant (p > 0.05).

Table 3.

Gender and grade test(M ± SD)

Variables	Physical activity	Resilience	Emotional regulation	Executive function
Mann-Whitney U	111921.500	113635.500	116920.500	115180.500
Wilcoxon W	223549.500	225263.500	228548.500	239930.500
Z	−1.343	−0.945	−0.193	−0.591
Asymptotic saliency (two-tailed)	0.179	0.345	0.847	0.554
Kruskal-Wallis H	12.206	2.360	2.869	0.646
df	4	4	4	4
p-value	0.016*	0.670	0.580	0.958
Male	25.41 ± 21.54	3.01 ± 0.50	2.98 ± 0.62	1112.89 ± 370.75
Female	24.82 ± 20.37	2.98 ± 0.49	2.97 ± 0.61	1098.52 ± 365.92
2nd grade	26.93 ± 22.77	2.97 ± 0.53	2.98 ± 0.64	1109.27 ± 406.46
3nd grade	19.14 ± 16.15	2.97 ± 0.48	2.93 ± 0.59	1109.13 ± 380.53
4nd grade	27.54 ± 21.66	3.04 ± 0.49	3.05 ± 0.65	1117.74 ± 371.50
5nd grade	25.24 ± 21.20	3.00 ± 0.50	2.98 ± 0.62	1098.76 ± 340.56
6nd grade	25.81 ± 23.03	3.02 ± 0.52	2.96 ± 0.62	1131.40 ± 390.29

*means p< 0.05

Descriptive statistics and correlation analysis

The results of descriptive statistics and correlation analyses for each variable are shown in Table 4. After controlling for grade and gender, PA was positively correlated with resilience and ER (r = 0.544, p < 0.01; r = 0.413, p < 0.01). EF was negatively correlated with PA, resilience, and ER (r = −0.341, p < 0.01; r = −0.444, p < 0.01; r = −0.407, p < 0.01), where the negative correlations reflect the reaction time measurement approach (shorter times indicate better EF performance). Resilience was positively correlated with ER (r = 0.460, p < 0.01). These findings align with theoretical expectations and provide a foundation for hypothesis testing.

Table 4.

Mean and correlation between variables

	M ± SD	1	2	3	4	5	6	7	8	9
1. Physical activity	25.41 ± 21.54	1
2. Resilience	3.01 ± 0.0.50	0.544***	1
3. Focus on goal	2.90 ± 0.81	0.403***	0.703***	1
4. Emotional control	2.99 ± 0.78	0.348***	0.653***	0.369***	1
5. Positive cognition	3.10 ± 0.80	0.378***	0.736***	0.420***	0.410***	1
6. Family support	3.04 ± 0.76	0.275***	0.520***	0.215***	0.118***	0.233***	1
7. Interpersonal assistance	3.00 ± 0.80	0.323***	0.562***	0.213***	0.177***	0.265***	0.118***	1
8. Emotion rugalation	2.98 ± 0.62	0.413***	0.460***	0.383***	0.295***	0.314***	0.231***	0.237***	1
9. Executive function	1112.89 ± 370.75	−0.341***	−0.444***	−0.345***	−0.237***	−0.313***	−0.244***	−0.270***	−0.407***	1

The data in the table are correlation factors. ***means p< 0.001

Construction of the model

The structural equation model was constructed using AMOS 26.0 and the model fit indices and structural relationships were checked as shown in Fig. 2. The standardized path coefficients used were all significant (p < 0.05). The results indicated that PA was positively associated with resilience (β = 0.29, p < 0.001) and ER (β = 0.55, p < 0.001). ER was also positively associated with resilience (β = 0.48, p < 0.001). PA was negatively associated with EF (β = −0.46, p < 0.001). EF was negatively associated with resilience (β = −0.12, p < 0.05) and negatively associated with ER (β = −0.30, p < 0.001).

Fig. 2.

Path analysis results of PA, resilience, EF, ER. Note: The coefficients in the graphs are standardized; ***p < 0.001,**p < 0.01,*p < 0.05

Examination of the mediating effects of executive function and emotion regulation

The mediating effect was verified using the bias-corrected non-parametric percentile Bootstrap method, where 95% confidence intervals (CIs) were calculated with 5,000 Bootstrap samples, and a significant mediating effect was indicated if the 95% CI for the standardised path coefficient did not contain zero. Table 5 presents the results, showing a significant mediating effect of EF between PA and resilience (CI = [0.002, 0.066]), confirming hypothesis H2. The mediating effect of ER between PA and resilience was significant (CI = [0.110, 0.268]), confirming hypothesis H3. Additionally, the chain mediation effect of EF and ER was significant (CI = [0.023, 0.074]), confirming hypothesis H4. Further analysis showed that the total effect of PA on resilience was 0.444, the direct effect was 0.191, and the total indirect effect was 0.253. The mediating effects were distributed as follows: PA → EF → resilience (CI = 0.036, 14.23%), PA → ER → resilience (CI = 0.173, 68.38%), and PA → EF → ER → resilience (indirect effect = 0.044, 17.39%).

Table 5.

Chain mediated path effect test for executive function and emotion regulation

path	Effect value	Bootstrap SE	Bootstrap 95%CI	Relative intermediary effect %
			Lower upper
physical activity→executive function→resilience	0.036	0.016	[0.002,0.066]	14.23%
physical activity→emotion regulation→resilience	0.173	0.040	[0.110,0.268]	68.38%
Physical activity→executive function→emotion regulation→resilience	0.044	0.013	[0.023,0.074]	17.39%
Total indirect effect	0.253	0.045	[0.177, 0.354]	56.88%
Direct effect	0.191	0.054	[0.078,0.292]	43.02%
Total effect	0.444	0.033	[0.380,0.509]	100%

Discussion

The relationship between physical activity and resilience

The results indicate that PA positively predicts resilience, validating our hypothesis H1 that physical activity would positively influence children’s resilience levels. Our findings are consistent with previous research demonstrating positive correlations between PA and resilience across different age groups [25, 72, 73]. For instance, longitudinal studies have shown that adolescents maintaining regular PA routines exhibit higher resilience levels than sedentary counterparts [26], while our cross-sectional findings extend this relationship to younger elementary school populations.

According to Richardson’s Metatheory of Resilience and Resiliency, resilience develops through cycles of disruption and reintegration, where individuals encounter stressors and then reintegrate with varying degrees of resilience depending on available protective factors [10]. Our findings provide empirical support for this theoretical framework by demonstrating that PA serves as a key environmental protective factor within Richardson’s three-level system (biological, psychological, and environmental), acting as an external factor that fosters resilience by enhancing individuals’ coping abilities and stress resistance. This finding further supports Self-Determination Theory (SDT) [74], which suggests that PA enhances resilience by fostering autonomy, competence, and relatedness while promoting more positive responses to stress and challenges. From a developmental psychology perspective, PA not only supports children’s physical and mental health but also serves as a key driver of psychological capacity development. Its long-term cumulative effects may foster more stable ER abilities and greater stress resilience over time.

At the neurophysiological level, The significant PA-resilience relationship provides evidence for the theoretical framework wherein PA stimulates BDNF secretion and enhances neuroplasticity, improving cognitive flexibility and ER [32]. Furthermore, our results corroborate neuroimaging evidence that regular PA enhances prefrontal cortex structure and function, optimizing integration with large-scale networks such as the FPN, DMN and corticolimbic system [75]. These neurobiological adaptations provide the physiological foundation for PA’s beneficial effects on resilience development.

In addition, evidence at the behavioural level suggests that regular physical activity participation provides beneficial environments and conditions for children’s resilience development. Physical activity experiences can improve physiological and psychological states while enhancing emotional management skills, which ultimately contributes to the development of children’s resilience [76, 77]. Given PA’s role in enhancing resilience, schools, families, and society should take active measures to promote children’s PA. Schools should optimise the physical education curriculum, families should actively participate in parent-child sports, and society should improve sports facilities and enhance publicity to jointly promote children’s physical and mental development.

Grade-level differences in physical activity patterns

The post-hoc analysis revealed that 4th grade students demonstrated significantly higher physical activity levels compared to 3rd grade students, which aligns with motor development theory suggesting that children’s PA engagement increases with age during elementary school years [78]. This finding is consistent with previous research indicating that older elementary school children tend to exhibit greater physical competence, motor skill proficiency, and autonomous motivation for physical activity participation [79, 80]. Several factors may contribute to this grade-level difference. First, developmental improvements in fundamental motor skills typically occur between ages 8–10, enabling 4th graders to engage more confidently in various physical activities [81]. Second, 4th grade students may experience greater autonomy and self-efficacy in initiating and sustaining physical activities compared to younger peers [82]. Third, school-based physical education curricula often become more diverse and challenging in higher grades, potentially fostering greater interest and engagement [83]. This grade-specific pattern suggests that interventions aimed at enhancing resilience through physical activity should consider developmental appropriateness and may need to be tailored differently for younger versus older elementary school students.

The mediating effect of executive function

This study reveals the mediating role of EF in the relationship between PA and resilience, with Hypothesis 2 being confirmed. This indicates that PA enhances resilience through the improvement of EF, which is consistent with previous research [84, 85]. Although the effect size for the mediating role of EF in this study is small, as Walters pointed out in his systematic analysis of mediation effects, mediation effects are typically small, reflecting the inherent nature of indirect effects as products of two individual effects. In developmental research involving complex psychological processes, small mediation effects can still represent meaningful mechanisms as long as they achieve statistical significance and theoretical support [86].

Self-regulation theory suggests that EF is the core of an individual’s self-regulation ability [42] and plays an important role in the individual’s ability to cope with stress, regulate emotions and adapt to environmental changes. EF enables individuals to manage emotions, adjust cognitive strategies, and regulate behavior when facing stress or challenges, thereby enhancing adaptability and resilience. At the neurophysiological level, multiple studies have confirmed the mechanisms through which PA influences EF. PA enhances EF by optimizing cognitive resource allocation through increased physiological activation. PA activates the sympathetic nervous system, increasing heart rate and blood flow. These physiological changes enhance attention and information processing, improving cognitive performance [87]. In addition, from a psychophysiological perspective, acute PA promotes the release of a variety of neurotransmitters (e.g., adrenaline, dopamine, and the brain-derived neurotrophic factor BDNF), and these changes contribute to enhanced cognitive processes [88].

Neuroimaging studies further demonstrate the direct effects of PA on EF-related brain regions [89]. Acute low-intensity exercise activates the prefrontal cortex, positively impacting EF [90]. The prefrontal cortex (PFC) serves as the neural hub for EF, governing impulse inhibition, cognitive flexibility, and working memory. Regular PA enhances cognitive control by stimulating PFC activity.It has been found that higher levels of cardiorespiratory fitness better induce haemodynamic responses in frontal, parietal and temporal lobe regions to achieve higher levels of cognitive function [91]. These neural mechanisms provide a physiological basis for PA’s influence on resilience via EF, further highlighting PA’s role in cognitive regulation and stress adaptation.

Based on these findings, future PA interventions for elementary school children should leverage the systematic nature of school-based physical education (PE) programs. PE represents a structured component of PA that, while accounting for approximately 15–20% of children’s total weekly PA, provides universal access through mandatory curriculum implementation across primary schools. For children in Grades 2–6 (the age range of our study participants, mean age 9.8 years), PA interventions should be developmentally tailored to enhance both EF and resilience. Specifically, elementary PE curricula should integrate cognitive challenges with physical activities, incorporating strategy-based games, rule-following sports, and coordination exercises that require planning and problem-solving capabilities appropriate for 8–11 year-old children. Families should be encouraged to support executive-function-enhancing physical activities appropriate for children’s developmental stage, including structured games requiring planning and behavioral regulation. Communities should provide safe, accessible opportunities for elementary school-aged children to engage in cognitively demanding physical activities that promote strategic thinking and sustained attention. These multi-environmental approaches can maximize the EF-mediated benefits for resilience development in this age group.

The mediating effect of emotion regulation

This study confirmed the mediating role of ER between PA and resilience, suggesting that PA indirectly enhances resilience by improving ER. These findings support Hypothesis H3 and align with previous research [92]. According to Broaden-and-Build Theory, positive emotions broaden cognitive and behavioral styles and help build emotional, psychological, and social resources [93]. Therefore, individuals are able to call upon their emotional and cognitive resources more effectively in the face of challenges and adversity in order to develop stable psychological adaptive strategies, thereby enhancing resilience. ER theory posits that individuals manage emotions through cognitive reappraisal and expressive suppression [94]. Cognitive reappraisal involves altering one’s perception of a situation to regulate emotions, while expressive suppression involves controlling emotional responses by inhibiting outward expression. Both strategies are crucial for resilience in coping with stress and challenges.At the neurophysiological level, PA enhances emotion regulation by modulating emotion-related neural networks. It strengthens PFC function and cognitive control, promoting the use of cognitive reappraisal over expressive suppression during ER [95]. Additionally, PA reduces amygdala hyperactivity, aiding in the regulation of emotional responses to negative stimuli, which helps reduce mood swings and enhance psychological stability [76]. This process involves the top-down regulation of the amygdala by the PFC. By enhancing the PFC’s inhibitory control over the amygdala, PA improves impulse control and optimizes ER strategies [96].

Beyond neural mechanisms, PA optimises an individual’s pattern of emotion regulation through behavioural pathways. PA is thought to reduce reliance on expressive inhibitory strategies. By providing a healthy pathway for emotional release, PA can help individuals express and manage emotions in a more positive way when faced with negative emotions, rather than simply repressing them [97]. Studies show that regular PA promotes structural and functional improvements in emotion-related brain regions (e.g., prefrontal cortex and amygdala). These changes provide a solid neural basis for effective ER under stress [98]. Therefore, combining PA with psychological interventions during development can strengthen children’s ER strategies and enhance resilience. Relying solely on PA may be insufficient, so integrating psychological strategies ensures more comprehensive and lasting improvements in resilience.

The chain mediating role of executive function and emotion regulation

This study confirmed the chain-mediated role of EF and ER between PA and resilience, supporting Hypothesis H4 and aligning with previous research [99]. This chain-mediated effect reveals the complete cognitive-emotional pathway through which PA affects resilience, and further enriches research on the mechanisms by which PA promotes resilience. It should be noted that the chain mediation effect size (0.044) is small according to conventional standards. While modest in magnitude, this finding was statistically significant (95% bootstrap CI: [0.027, 0.076]) and aligns with previous studies reporting similarly small chain mediation effects in developmental research [100, 101].

According to process modeling theory [102], emotion regulation depends on executive functions, including working memory (maintaining regulatory goals), inhibitory control (suppressing inappropriate responses), and cognitive flexibility (adapting regulatory strategies).n particular, the effective implementation of cognitive reappraisal, an emotion regulation strategy, is highly dependent on the functioning of the prefrontal executive control network [103]. Thus, PA provides a cognitive foundation for effective emotion regulation by enhancing EF, thereby strengthening resilience.

At the neurophysiological level, this chain-mediated pathway reflects a cascade effect of PA on key brain region functions.PA promotes neuroplasticity and functional integration in the PFC, enhancing executive network efficiency [89]. Regular exercisers show optimized PFC-amygdala connectivity during executive control and emotion regulation tasks. These neural changes support the PA-EF-ER-resilience pathway [104]. Behavioral studies show that regular exercisers better focus attention, inhibit impulsive responses, and flexibly use cognitive reappraisal strategies during negative emotional events [105]. From a developmental perspective, the PA-EF-ER-resilience chain may follow a certain temporal sequence during childhood development. Research suggests that childhood PA first promotes the development of EF, which in turn matures in a process that provides the cognitive basis for the acquisition of more complex ER strategies [75]. This developmental sequence allows individuals to demonstrate greater mental toughness in the face of stress. Therefore, age-specific PA interventions are particularly important and should be designed according to the developmental stage.

Conclusion

This study explored how physical activity (PA) influences resilience and confirmed the mediating roles of executive function (EF) and emotion regulation (ER).The findings suggest that: (1) PA positively predicts resilience; (2) EF partially mediates the relationship between PA and resilience; (3) ER partially mediates the relationship between PA and resilience; and (4) EF and ER form a chain mediating pathway between PA and resilience. Moreover, our findings reveal age-related differences in physical activity, suggesting that future interventions should be tailored to students at different elementary school stages to optimize their effectiveness. These findings reveal that PA influences resilience through a dual cognitive-emotional pathway, offering new insights into the mechanisms of PA in promoting mental health.

Limitations of the study

This study has several limitations.First, the cross-sectional design limited the study to exploring relationships at a single time point, without determining causality. While our findings demonstrate significant associations between PA and resilience, the direction of causality remains unclear, and bidirectional relationships are possible (e.g., resilient children may be more inclined to engage in physical activity). Future studies should adopt longitudinal or experimental designs to establish temporal precedence and validate the long-term effects of PA on EF, ER, and resilience.

Second, while this study confirmed the key role of EF and ER in the PA-resilience pathway, resilience may also be influenced by factors like personality traits, social support, and coping strategies. Our model focused on cognitive and emotional pathways, but resilience development is multifactorial and may involve genetic, social, and environmental influences that were not examined in this study. Future studies should integrate other theoretical models to explore multiple pathways through which PA promotes resilience, including the roles of family support systems, peer relationships, genetic predispositions, and broader social-ecological factors.

Third, future research would benefit from incorporating diverse measurement approaches and assessment tools to better capture the complex, dynamic relationships between these variables, ultimately enriching understanding of these mechanisms and informing the development of more comprehensive, evidence-based intervention approaches.

Abbreviations

PA

Physical Activity

EF

Executive Function

ER

Emotion Regulation

Authors’ contributions

Yifan Xu, Shimeng Wang and Dandan Chen: Conceptualized and designed the study, data analysis, manuscript writing. Kai Qi, Shuqiao Meng, Xiaoxiao Dong, Aiguo Chen: Refinement of the study and critical revisions of the manuscript. All authors reviewed the manuscript.

Funding

This research was supported by grants from the National Natural Science Foundation of China (31771243) and the National Social Science Foundation of China (23ATY008).

Data availability

The data presented in this study are available on request from the corresponding author.

Declarations

Ethics approval and consent to participate

The study was approved by the Ethics Committee of the Medical College of Yangzhou University (Approval No.: YXYLL-2023-129). Written informed consent was obtained from the parents or legal guardians of all participants. Age-appropriate instructions were provided to participants, who participated voluntarily. The study was conducted in accordance with the Declaration of Helsinki (1964) and related guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.