Physical exercise and problematic smartphone use among adolescents in Western China: a multi-center cross-sectional study on the serial mediation of self-control and meaning in life Authors: Xingyu He, Min Liu, Ping Mo, Wenxiu Luo, Jiaming Luo Published: 2026 ABSTRACT ObjectiveProblematic smartphone use has become a prominent public health issue in the information era. Although research has shown that exercise is beneficial in alleviating problematic smartphone use among adolescents, the specific underlying mechanisms have not been fully elucidated. This study aims to explore the internal mechanisms between self-control and the meaning in life in adolescents’ problematic smartphone use (PSU) through physical exercise.MethodsA multi-center, large-sample cross-sectional study was conducted among students from 34 primary and secondary schools in western China. Assessments were conducted using the Physical Activity Rating Scale (PARS-3), the Brief Self-Control Scale (BSCS), Meaning in Life Questionnaire (MLQ), and the Smartphone Addiction Scale-Short Version (SAS-SV). Data were analyzed using the Mann–Whitney U test, correlation analysis, and mediation analysis based on structural equation modeling.ResultsA total of 8,245 participants were included. The rates of PSU in males and females were found to be 23.97% and 19.47%, respectively. After adjusting for relevant demographic covariates, structural equation modeling analysis indicated that exercise was negatively associated with PSU (β = -0.056, p < 0.001). Both self-control (β = -0.034, p < 0.001) and meaning in life (β = -0.019, p < 0.001) played potential partial mediating roles in the relationship between physical exercise and PSU. Furthermore, self-control and meaning in life served as potential partial chain mediators between physical exercise and adolescent PSU (β = -0.011, p < 0.001), accounting for 9.17% of the total association.ConclusionMaintaining regular physical exercise can enhance adolescents’ self-control ability and meaning in life thereby helping to prevent or alleviate PSU behaviors. Introduction With the widespread adoption of smartphones, problematic smartphone use (PSU) has emerged as a global public-health concern. PSU is defined as a behavioral addiction where individuals experience impairment of social functions and psychological and behavioral problems due to excessive, uncontrolled smartphone use [1]. Studies indicate that the global incidence rate of PSU among individuals aged 13–18 is 34.5% [2]. China is among the countries with higher rates of PSU [3]. The incidence rate of PSU among Chinese adolescents is approximately 23.5%, and it has been steadily increasing in recent years, becoming a prominent public health concern [4]. PSU harms adolescents’ physical, mental, and behavioral health and is closely linked to anxiety, depression, and loneliness. In severe cases, it significantly increases the risk of suicidal ideation and behaviors [5–7]. Affected adolescents commonly report poorer sleep quality and insomnia, visual fatigue, cervical spine complaints, and hand problems, all of which pose health risks [7, 8]. Reliance on smartphone-mediated communication can weaken real-world social skills, heighten family and interpersonal conflicts, and hinder social adaptation [8, 9]. From a theoretical perspective, PSU can be understood as a form of behavioral dysregulation. According to self-regulation theory, individuals’ capacity to regulate their thoughts, emotions, and behaviors plays a critical role in preventing maladaptive behaviors. When self-regulation fails, individuals are more likely to seek immediate gratification, thereby increasing the risk of problematic smartphone use [10, 11]. Against this backdrop, exploring the protective factors for PSU and the relationships between related variables holds urgent theoretical and practical significance. The association between physical exercise and PSU Accumulating evidence indicates that physical exercise alleviates symptoms of PSU in adolescents. Prior studies report a significant inverse association between physical exercise and PSU [12]. On one hand, physical exercise can directly reduce smartphone use time, thereby preventing overuse and helping adolescents re-establish healthier routines [13, 14]. On the other hand, physical exercise promotes the release of neurotransmitters such as endorphins, improving mood and alleviating stress, thereby reducing smartphone dependence behavior triggered by negative emotions and stress. After exercise, adolescents’ craving for smartphones significantly decreases [15]. Accordingly, we propose hypothesis H1: Physical exercise is negatively associated with adolescents’ PSU. The mediating role of self-control between physical exercise and PSU Self-control is the capacity to regulate thoughts, emotions, and behaviors in the face of desires and impulses so that actions align with long-term goals and values [16, 17]. Previous studies have shown that physical exercise can significantly enhance an individual’s self-control ability [18]. In a meta-analysis of 83 studies (n = 80,681), Li et al. [19] examined the link between self-control and PSU in student samples. The results showed a moderate negative association between self-control and PSU. Although previous studies have preliminarily explored the mediating role of self-control in the relationship between physical exercise and PSU, these investigations often rely on smaller, localized samples. Therefore, the epidemiological robustness and generalizability of this mediating mechanism among adolescents—particularly in underrepresented regions—require rigorous validation through large-scale epidemiological samples. Theoretically, this relationship can be elucidated through the Strength Model of Self-Control within Self-Regulation Theory. Self-Regulation Theory posits that self-control relies on a limited resource, and its depletion underlies addictive behaviors like PSU [11]. Drawing on the ‘muscle metaphor,’ regular physical exercise acts as a structured practice that strengthens this ‘self-regulatory muscle’ over time, thereby replenishing self-regulatory resources [20]. Based on this theoretical premise, hypothesis H2 is proposed: Self-control serves as a potential mediator in the relationship between physical exercise and adolescents’ PSU. The mediating role of meaning in life between physical exercise and PSU Meaning in life typically refers to an individual’s subjective experience of understanding, goals, and values [21]. Research indicates a significant positive correlation between physical exercise and adolescents’ meaning in life [22]. Individuals who regularly engage in physical exercise tend to experience enhanced well-being and pleasure, which in turn may foster a richer perception of meaning in life [23]. Besides, meaning in life plays a crucial role in preventing and alleviating PSU. Higher levels of meaning in life are associated with lower risk and severity of PSU [24]. Among adolescents, meaning in life functions as a psychological protective factor for PSU and may buffer the effects of impulsivity and emotional distress on PSU [25]. Previous intervention models for PSU have predominantly focused on mitigating negative emotional states, empirical research exploring positive psychological resources remains scarce, particularly regarding the mediating role of meaning in life between physical exercise and adolescent PSU. From the perspective of Self-Regulation Theory, when physical exercise replenishes self-regulatory resources and enhances individual well-being, it provides adolescents with the cognitive capacity to construct a strong cognitive anchor—meaning in life [11, 26]. This cognitive anchor effectively buffers the negative consequences of impulsivity and emotional distress, steering adolescents away from the pathological dopamine pleasure of smartphones. Therefore, we propose hypothesis H3: Meaning in life serves as a potential mediator in the relationship between physical exercise and PSU. The chain mediating role of self-control and meaning in life between physical exercise and PSU Previous studies suggest that physical exercise, self-control, and meaning in life are antecedent variables of PSU. Additionally, research shows a close relationship between self-control and meaning in life, with self-control being positively associated with meaning in life [27]. Individuals with higher self-control are more likely to choose meaningful activities over immediate pleasure, which may foster a stronger meaning in life [28]. People with stronger self-control tend to identify goals and values and to experience greater coherence, thereby developing and consolidating meaning in life [26]. However, although empirical work links physical exercise to lower PSU, existing chain mediation models have predominantly focused on the alleviation of negative emotional states. The role of positive psychological resources, particularly the meaning in life, in interventions for behavioral addictions like PSU remains insufficiently explored. Furthermore, the relationship between self-control and meaning in life requires further establishment within the specific context of adolescent development. To address these gaps, this study examines the association between physical exercise and adolescent PSU and tests the potential mediating roles of self-control and meaning in life. Theoretically, rather than examining these variables in isolation, Self-Regulation Theory unifies them: exercise replenishes self-regulatory resources, which manifests behaviorally as improved self-control. This enhanced self-control, in turn, fosters a stronger cognitive anchor and goal orientation, leading to an elevated meaning in life. Therefore, hypothesis H4 is proposed: Self-control and meaning in life play a potential chain mediating role in the association between physical exercise and PSU. Current study Guided by the integrative theoretical framework discussed above, the present study seeks to empirically validate these proposed pathways. While previous research has confirmed isolated associations among these four variables, no study to date has tested a comprehensive model that jointly evaluates physical exercise, self-control, and meaning in life as potential interactive pathways underlying PSU. Moreover, generalizations from smaller, single-center adult cohorts may not fully capture the epidemiological reality of adolescents in developing regions. Therefore, utilizing a large-scale, multi-center sample of adolescents from Western China, this study aims to further provide theoretical and empirical evidence for how physical exercise alleviates PSU. By shifting the focus from simple behavioral correction to an integrated psychological resource model, this study reveals the sequential pathway relationships of these relevant psychological factors. On the basis of previous theories and hypotheses, a hypothetical model is constructed, as shown in Fig. 1. Fig. 1Hypothetical model of physical exercise, self-control, meaning in life, and problematic smartphone use Hypothetical model of physical exercise, self-control, meaning in life, and problematic smartphone use Methods Participants This study is a cross-sectional research using a multistage cluster random sampling method, surveying students from 34 elementary and secondary schools in Sichuan Province, China. The sampling process was conducted as follows: First, two counties were selected from the six county-level administrative districts in Nanchong City using a simple random sampling approach. Second, within each of the two selected counties, approximately 30% of the primary and secondary schools were randomly chosen using a lottery method, resulting in a total of 34 schools being included in this study. In this study, the ‘clusters’ were defined at the school level. All eligible students (ranging from grades 4 to 12) within these selected clusters were invited to participate. The inclusion of both urban and rural schools in the sample may help improve the geographic diversity of the sample. To ensure data quality, data collection was conducted by psychiatrists, graduate students, and undergraduates specializing in mental health. The survey was answered anonymously, with emphasis on voluntary participation. The content was strictly confidential and the results were used only for scientific research. All subjects were informed of the study’s purpose and characteristics and signed informed consent forms. The inclusion criteria for the study were: students currently enrolled between the ages of 10 and 19, without intellectual disabilities or severe physical illnesses, and capable of independently completing the survey questions.Exclusion criteria included: students or their parents who refused to participate in the survey, and students or parents who refused to sign the informed consent form. This study was approved by the Ethics Committee of North Sichuan Medical College (Project Number: NSMC [2024]014). In May 2024, a total of 10,046 questionnaires were collected. To ensure high data quality, 1,801 invalid questionnaires were carefully excluded based on the following strict criteria: (1) participants who do not meet the age of 10–19 years; (2) extreme completion times falling outside three median absolute deviations (± 3 MAD) from the median response time, indicating potential random or inattentive responding; and (3) incomplete or invalid response patterns, including excessive missing data on key variables (for which listwise deletion was employed) and uniform responding (e.g., selecting the same option for all scale items). Ultimately, 8,245 valid questionnaires were obtained, with an effective return rate of 82.07%. Measurement Physical activity scale (PARS-3) This study used the Physical Activity Rating Scale (PARS-3) developed by Japanese scholar Masao Hashimoto and revised by Chinese scholar Deqing Liang and others [29]. The scale includes three items: exercise intensity, exercise frequency, and exercise duration. Using a Likert 5-point scale, exercise intensity and frequency are rated from 1 to 5, while exercise duration is rated from 0 to 4. The total amount of physical exercise is quantified using the formula “exercise intensity × exercise duration × exercise frequency.” The score ranges from 0 to 100, with higher scores indicating greater physical exercise. The Cronbach’s α for this study is 0.76. Smartphone Addiction scale (SAS-SV) The validated SAS-SV questionnaire developed by M Kwon, DJ Kim, H Cho and S Yang [30] was used in this study. It comprises 10 items rated on a 6-point Likert scale, yielding a total score of 10–60. The instrument assesses the severity of problematic smartphone use; higher scores indicate greater addictive propensity. According to Kwon et al., the cutoff values for problematic smartphone use are scores higher than 33 for females and higher than 31 for males. The Chinese version of this scale has demonstrated good reliability and validity [31]. The Cronbach’s α for this study is 0.91. Brief self-control scale (BSCS) The Brief self-control scale (BSCS) was revised by ME Morean, KS DeMartini, RF Leeman, GD Pearlson, A Anticevic, S Krishnan-Sarin, JH Krystal and SS O’Malley [32]. The scale consists of two dimensions: Self-Discipline (3 items) and Impulse Control (4 items), totaling 7 items. The Brief Self-Control Scale (BSCS) assesses participants’ perceptions, responses, and levels of agreement as evaluation indicators, using a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The Cronbach’s α for this study is 0.70. Meaning in life questionnaire (MLQ) The Meaning in Life Questionnaire (MLQ) consists of two dimensions: Presence of Meaning and Search for Meaning.A 7-point scale is used, ranging from “Strongly Disagree” to “Strongly Agree,” scored from 1 to 7. Higher scores indicate a stronger meaning in life for the individual. This study used the Chinese version of the MLQ revised by Wang Mengcheng (2008) for the survey. This scale demonstrated good reliability and validity [33]. The Cronbach’s α coefficient of the scale in this study is 0.89. Statistical processing We used SPSS Statistics 25.0 for descriptive statistical analysis. Harman’s single-factor test was used to examine common method bias. Continuous PSU scores were categorized to define adolescents with and without PSU. As detailed in the measurement instruments section, this categorization was based on established gender-specific cutoffs (scores > 31 for males and > 33 for females) [30]. Prior to inferential testing, the distributional characteristics of the variables were examined. Because the variables exhibited non-normal distributions, the non-parametric Mann–Whitney U test was appropriately selected to compare adolescents with and without PSU in terms of physical exercise, self-control, and meaning in life. For the correlation analysis, Pearson correlation analysis was retained over Spearman’s rank correlation for two methodological reasons. First, given our large sample size, parametric tests are robust to non-normality according to the Central Limit Theorem. Second, all study variables were analyzed as continuous measures, and the primary objective was to assess linear associations among variables. Therefore, Pearson correlation was considered appropriate for the present analyses and consistent with the subsequent SEM framework. Following the mediation testing process proposed by Z Wen and B Ye [34], AMOS 29.0 was utilized to construct a structural equation model. The model was estimated after adjusting for several demographic and contextual covariates, including age, gender, grade, residence, and poverty. We examined the direct association between physical exercise and PSU and the potential mediating role of self-control and meaning in life in the relationship between physical exercise and adolescent PSU. Specifically, to examine the serial mediation model, the following three potential indirect pathways were tested: (1) physical exercise → self-control → PSU; (2) physical exercise → meaning in life → PSU; and (3) physical exercise → self-control → meaning in life → PSU. The significance of these indirect associations was tested using the bias-corrected percentile Bootstrap method with 5,000 resamples. An indirect association was considered statistically significant if its 95% confidence interval (CI) did not include 0. The significance level for all indicators was set at p < 0.05. Results Common method deviation test As this study utilized a survey method, with data obtained from participants’ subjective ratings and questionnaire responses, corresponding control methods such as anonymous questionnaire measurement, positive and negative scoring, and standardized testing were employed during the testing process. Harman’s single-factor test was used to investigate common method bias [35]. The results showed that there were six factors with eigenvalues greater than 1, with the first factor explaining 25.03% of the variance, which is less than the critical standard of 40%. This indicates that there is no common method bias in this study. Demographic characteristics of participants A total of 8,245 valid questionnaires were obtained.The age range of participants was 10–19 years, with an average age of 13.61 ± 2.33. There were 4,301 males (52.16%) and 3,944 females (47.84%).The basic information of the study subjects selected for this research is shown in Table 1. Table 1Basic Information of survey subjects (N = 8245)Demographic VariablesCategorizationNumber of PeoplePercentage (%)Gendermale430152.16%female394447.84%GradeElementary school323039.17%Junior high school244029.60%High school257531.23%Location of Their HomeUrban530764.37%Countryside293835.63%Left-behind childrenYes467656.71%No356943.29%PovertyYes181922.06%No642677.94%Parents’ marital statusNot divorced640577.68%Divorced110313.38%remarry7378.94% Table 1Basic Information of survey subjects (N = 8245)Demographic VariablesCategorizationNumber of PeoplePercentage (%)Gendermale430152.16%female394447.84%GradeElementary school323039.17%Junior high school244029.60%High school257531.23%Location of Their HomeUrban530764.37%Countryside293835.63%Left-behind childrenYes467656.71%No356943.29%PovertyYes181922.06%No642677.94%Parents’ marital statusNot divorced640577.68%Divorced110313.38%remarry7378.94% Basic Information of survey subjects (N = 8245) Results of difference tests for variables related to PSU Applying the gender-specific cutoff scores of the SAS-SV (males > 31, females > 33), the detection rate of PSU among male participants was 23.97%, while that among female participants was 19.47%. According to Table 2, non-parametric Mann-Whitney U tests were conducted between the PSU and non-PSU groups for males, yielding results for physical exercise (Z = -3.78, p < 0.001), self-control (Z = -22.45, p < 0.001), and meaning in life (Z = -0.22, p = 0.824). For females, the same non-parametric tests between the PSU and non-PSU groups showed results for physical exercise (Z = -6.78, p < 0.001), self-control (Z = -21.20, p < 0.001), and meaning in life (Z = -3.46, p < 0.001). Table 2Comparison of physical exercise, self-control, and meaning in lifemalesfemalesPSU(median score)Not-PSU(median score) p PSU(median score)Not-PSU(median score) p physical exercise12.0015.00< 0.0016.009.00< 0.001self-control21.0024.00< 0.00120.0024.00< 0.001meaning in life47.0047.000.82446.0047.00< 0.001Participants were categorized into the PSU group based on the gender-specific cutoff scores of the SAS-SV (males > 31, females > 33) Table 2Comparison of physical exercise, self-control, and meaning in lifemalesfemalesPSU(median score)Not-PSU(median score) p PSU(median score)Not-PSU(median score) p physical exercise12.0015.00< 0.0016.009.00< 0.001self-control21.0024.00< 0.00120.0024.00< 0.001meaning in life47.0047.000.82446.0047.00< 0.001Participants were categorized into the PSU group based on the gender-specific cutoff scores of the SAS-SV (males > 31, females > 33) Comparison of physical exercise, self-control, and meaning in life Participants were categorized into the PSU group based on the gender-specific cutoff scores of the SAS-SV (males > 31, females > 33) Correlation analysis of physical exercise, self-control, meaning in life, and PSU among adolescents This study used SPSS 25.0 for descriptive statistical analysis of the variables, as shown in Table 3. problematic smartphone use was negatively correlated with physical exercise (r = − 0.09, p < 0.001), self-control (r = − 0.43, p < 0.001), and meaning in life (r = − 0.20, p < 0.001). Physical exercise was positively correlated with self-control (r = 0.14, p < 0.001) and meaning in life (r = 0.14, p < 0.001). There was a significant positive correlation between self-control and meaning in life (r = 0.32, p < 0.001). The correlations between the main variables reached a significant level, providing a solid basis for subsequent mediation analysis. Table 3Correlation analysis among major variables *** p < 0.001PSUM ± SDPSUphysical exerciseself-controlmeaning in life23.61 ± 10.931physical exercise16.59 ± 8.91−0.09***1self-control23.67 ± 5.26−0.43***0.14***1meaning in life42.37 ± 14.78−0.20***0.14***0.32***1 Table 3Correlation analysis among major variables *** p < 0.001PSUM ± SDPSUphysical exerciseself-controlmeaning in life23.61 ± 10.931physical exercise16.59 ± 8.91−0.09***1self-control23.67 ± 5.26−0.43***0.14***1meaning in life42.37 ± 14.78−0.20***0.14***0.32***1 Correlation analysis among major variables *** p < 0.001 Mediating role of self-control and meaning in life between physical exercise and PSU This study employed Amos 29.0 to conduct structural equation modeling analysis after adjusting for demographic covariates (i.e., age, gender, grade, residence, and poverty) in order to examine the mediating roles of self-control and meaning in life between physical exercise and PSU. The model fit indices were: χ2/df = 12.60, RMSEA = 0.04, GFI = 0.96, CFI = 0.96, AGFI = 0.95, TLI = 0.95, IFI = 0.96. Although the χ²/df ratio exceeds the conventional threshold, the chi-square statistic is highly sensitive to sample size; therefore, overall model fit should be evaluated comprehensively by incorporating multiple fit indices that are less sensitive to sample size (e.g., CFI, RMSEA) [36]. Given that the RMSEA is less than 0.08, and the CFI, GFI, and TLI all exceed 0.90, the overall model fit is considered excellent and suitable for mediation analysis. The indirect associations of each path were examined using the Bootstrap method (5,000 resampling iterations). It should be noted that the reported path coefficients (β) are standardized estimates, whereas the 95% bootstrap confidence intervals (CIs) are based on unstandardized estimates. Results are shown in Fig. 2; Table 4. The results showed: (1) Physical exercise showed a significant negative association with PSU (β = -0.056, p < 0.001), indicating a direct relationship between physical exercise and adolescent PSU. (2) Physical exercise was significantly positively associated with self-control (β = 0.233, p < 0.001), and self-control was significantly negatively associated with PSU (β = -0.144, p < 0.001). The results indicated a significant indirect association in the “physical exercise → self-control → PSU” path (95% CI = -0.049,-0.027), accounting for 28.33% of the total association. (3) Physical exercise was significantly positively associated with the meaning in life (β = 0.121, p < 0.001), and the meaning in life was significantly negatively associated with PSU (β = -0.157, p < 0.001). The results indicated a significant indirect association in the “physical exercise → meaning in life → PSU” path (95% CI = -0.028,-0.015), accounting for 15.83% of the total association. (4) Self-control was significantly positively associated with the meaning in life (β = 0.303, p < 0.001), indicating a significant serial indirect association in the “physical exercise → self-control → meaning in life → PSU” chain (95% CI =-0.016,-0.009), with an indirect coefficient of 0.233 × (0.303) ×-0.157=-0.011, accounting for 9.17% of the total association. The 95% confidence intervals for the indirect associations through self-control, meaning in life, and the path from self-control to the meaning in life did not include 0, indicating that the indirect associations of these pathways are significant. Because the direct association between physical exercise and PSU remains significant after including the mediators, these results collectively indicate a partial mediation model. Fig. 2The serial mediation model between self-control and meaning in life. Standardized path coefficients (β) are presented. The model adjusts for demographic covariates (i.e., age, gender, grade, residence, and poverty). *** p<0.001 The serial mediation model between self-control and meaning in life. Standardized path coefficients (β) are presented. The model adjusts for demographic covariates (i.e., age, gender, grade, residence, and poverty). *** p<0.001 Table 4Potential mediation pathways between physical exercise and problematic smartphone useAssociationsPathBSEBoot 95% CIβProportionDirect associationPE→PSU−0.0620.018(-0.100,-0.024)−0.05646.67%Mediating associationPE→SC→PSU−0.0370.006(-0.049,-0.027)−0.03428.33%PE→MIL→PSU−0.0210.003(-0.028,-0.015)−0.01915.83%PE→SC→MIL→PSU−0.0120.002(-0.016,-0.009)−0.0119.17%Total indirect association−0.0700.007(-0.085,-0.058)−0.06453.33%Total association−0.1320.019(-0.172,-0.096)−0.120Boot 95% CI is based on unstandardized indirect effects with 5,000 bootstrap resamplesPE Physical exercise, SC Self-control, MIL Meaning in life, PSU Problematic smartphone use. B denotes the unstandardized coefficient, whereas β denotes the standardized coefficient Table 4Potential mediation pathways between physical exercise and problematic smartphone useAssociationsPathBSEBoot 95% CIβProportionDirect associationPE→PSU−0.0620.018(-0.100,-0.024)−0.05646.67%Mediating associationPE→SC→PSU−0.0370.006(-0.049,-0.027)−0.03428.33%PE→MIL→PSU−0.0210.003(-0.028,-0.015)−0.01915.83%PE→SC→MIL→PSU−0.0120.002(-0.016,-0.009)−0.0119.17%Total indirect association−0.0700.007(-0.085,-0.058)−0.06453.33%Total association−0.1320.019(-0.172,-0.096)−0.120Boot 95% CI is based on unstandardized indirect effects with 5,000 bootstrap resamplesPE Physical exercise, SC Self-control, MIL Meaning in life, PSU Problematic smartphone use. B denotes the unstandardized coefficient, whereas β denotes the standardized coefficient Potential mediation pathways between physical exercise and problematic smartphone use Boot 95% CI is based on unstandardized indirect effects with 5,000 bootstrap resamples PE Physical exercise, SC Self-control, MIL Meaning in life, PSU Problematic smartphone use. B denotes the unstandardized coefficient, whereas β denotes the standardized coefficient Discussion The relationship between physical exercise and PSU This study confirmed that physical exercise is significantly negatively associated with adolescent PSU, supporting hypothesis H1. This finding is highly consistent with conclusions from several global studies: J Kim and K Lee [13] found in their survey of Korean adolescents that participating in moderate physical exercise more than 5 days a week, vigorous physical exercise more than 3 days a week, or strength training more than 3 days a week can reduce the risk of PSU. Z Ren, J Tan, B Huang, J Cheng, Y Huang, P Xu, X Fang, H Li, D Zhang and Y Gao [14] confirmed through component data analysis in Foshan, China, that the increase in daily exercise time can improve the incidence of PSU. The mechanisms of action may be explained through two pathways: Time Displacement Theory: Physical exercise directly reduces the duration of smartphone use. Due to the finite nature of a 24-hour day, an increase in active behaviors naturally occupies and displaces the time available for screen-based sedentary behaviors. This temporal displacement mechanism has been empirically supported by recent experimental interventions, which demonstrate that actively engaging in physical exercise instead of using smartphones effectively reduces overall screen time and prevents problematic use [37]. Neuro-Reward Substitution Mechanism: At the neurobiological level, physical exercise promotes the release of neurotransmitters and normalizes dopaminergic signaling in the brain’s reward circuit. This effectively substitutes the pathological dopamine pleasure typically derived from excessive smartphone use [38]. Consistent with this neurobiological premise, empirical evidence at the behavioral level supports this substitution effect; for instance, a randomized controlled trial by Yang et al. [15] demonstrated that engaging in physical exercise significantly reduces immediate post-exercise cravings for smartphones, acting as a healthy behavioral alternative. Therefore, regular physical exercise can reduce pathological smartphone use among adolescents, alleviate adolescent PSU, and serves as an effective intervention strategy for preventing and addressing this issue. Mediating role of self-control This study confirmed that self-control serves as a potential partial mediator in the relationship between physical exercise and adolescent PSU, supporting previous research [39, 40]. Crucially, this finding provides empirical support for the Strength Model of Self-Control within Self-Regulation Theory. Studies have found that individuals who engage in long-term physical exercise exhibit stronger self-control when faced with tasks requiring persistence and self-discipline, particularly in terms of enduring self-control [41]. Consistent with the “muscle metaphor” of this theory, during physical exercise, individuals need to set goals, overcome fatigue, resist temptations, and persist in completing training. This rigorous process acts as a structured practice that continuously reinforces and replenishes their limited self-regulatory resources [42, 43]. Equipped with replenished cognitive resources, self-control can help individuals suppress the impulse to go online, rationally schedule their time, and reduce excessive internet use driven by immediate gratification [43, 44]. Secondly, self-control acts as a “protective umbrella” by regulating negative emotions (such as anxiety and stress), reducing the risk of individuals relying on the internet due to emotional distress [45]. In summary, self-control effectively alleviates PSU through multiple mechanisms, including impulse inhibition and emotion regulation. Mediating role of meaning in life This study confirmed the potential partial mediating role of meaning in life between physical exercise and PSU. Previous studies have shown that physical exercise is positively associated with the meaning in life [22], while meaning in life is negatively associated with PSU [24]. Physical exercise strengthens individuals’ meaning in life, closely related to the positive experiences, achievement, and self-worth derived from exercise itself. From the perspective of Self-Regulation Theory, as exercise replenishes self-regulatory resources, adolescents gain the necessary psychological capacity to establish this strong cognitive anchor [27, 46]. Furthermore, improvements in physical fitness and health status during exercise imbue individuals with hope and motivation for life, leading to a more positive view of their existence [46]. Physical exercise often involves social interaction and teamwork, which enhance feelings of belonging and self-identification, further strengthening the perception of life’s meaning [47]. The relationship between of meaning in life on PSU can primarily be explained through behavioral choices. Individuals with a high meaning in life often possess clearer life goals and value pursuits, and thus are more likely to consciously reduce meaningless leisure activities. meaning in life inclines individuals to seek achievement and belonging through positive real-world activities rather than relying on the virtual world for comfort [48, 49]. In summary, meaning in life helps individuals effectively resist PSU risk through enhanced goal orientation, sense of value, and real-life engagement [48, 49]. Chain mediation role of self-control and meaning in life Another finding of this study is that self-control and meaning in life act as potential partial chain mediators in the association between physical exercise and adolescent PSU, which aligns with the integrated pathways proposed by Self-Regulation Theory. Self-control can increase the meaning in life, mainly through the sense of life structure and goal achievement it brings. Self-control helps individuals resist temptation and reduce impulsive behavior, making life more orderly and controllable, and this sense of mastery further reinforces the meaning of life [50]. When confronted with stress and challenges, individuals with strong self-control are better able to maintain higher-order goals and respond to difficulties proactively, facilitating the experience of meaning in life even in adversity [51]. Physical exercise can increase self-control and meaning in life. Physical exercise is an activity that requires sustained willpower and self-discipline. Long-term adherence to exercise can effectively enhance individuals’ self-control abilities, allowing them to better self-regulate and manage behavior in the face of temptation and stress [52, 53], thereby reducing dependence on immediate gratification and excessive smartphone use. Additionally, meaning in life derives from goal pursuit and self-realization. Physical exercise provides individuals with clear goals, challenges, and growth opportunities, allowing them to experience life’s value and meaning through overcoming difficulties and achieving self-transcendence [27, 46], enhancing engagement and satisfaction in real life, thereby reducing PSU behaviors. Therefore, we confirm that physical exercise is indirectly associated with adolescent PSU through the potential mediation of self-control and meaning in life. Furthermore, we reveal that physical exercise can enhance self-control, thereby increasing meaning in life, ultimately achieving the prevention or improvement of PSU behaviors. Importantly, while the serial indirect association accounts for a relatively modest proportion (9.17%) of the total association from a statistical perspective, its practical significance in public health should not be overlooked. In behavioral interventions, even small statistical effects can yield substantial population-level benefits, particularly when the intervention—such as physical exercise—is universally accessible and low-cost. The identification of this “self-control → meaning in life” pathway demonstrates that physical exercise does not merely act as a physical distraction from smartphones; rather, it triggers a positive psychological cascade. For educators and clinicians, this implies that integrating physical exercise with targeted cognitive-behavioral training can produce cumulative and synergistic benefits in preventing and alleviating adolescent PSU. Summary This multi-center, large-sample cross-sectional study explored the underlying mechanisms linking physical exercise and PSU among adolescents. The findings demonstrated that self-control serves as a significant independent potential mediator between physical exercise and PSU, thereby supporting existing theoretical frameworks. Specifically, our findings empirically validate the application of Self-Regulation Theory in the context of behavioral addiction. The novelty of this study lies in, to our knowledge, the very limited research identifying meaning in life as a distinct independent mediator in this relationship among adolescents. Based on this mediation, the study identifies and verifies a chain mediation path: “self-control → meaning in life.” Unlike previous studies restricted to smaller college cohorts, our findings are fortified by a large multi-center epidemiological sample from Western China, demonstrating robust statistical power. Guided by this theoretical framework, this study promotes a paradigm shift in PSU prevention and intervention from simple behavioral correction toward an integrated psychological resource model. It underscores the importance for clinicians, parents, and educators to attend to adolescents’ characteristics of self-control and meaning in life. Developing regular exercise habits can serve as a key entry point to enhance adolescents’ self-control by reinforcing goal persistence and training delayed gratification. Designing meaningful physical activities can help adolescents discover self-worth through bodily challenges, thereby fostering their meaning in life. During this process, dynamic monitoring of adolescents’ self-control and meaning in life can facilitate systematic assessment of their capacity to regulate PSU-related behaviors. These insights open new avenues for strategies aimed at preventing PSU among adolescents. Limitations This study, based on cross-sectional surveys and study design, constructed a structural equation model of physical exercise and adolescent PSU. While it can reveal path relationships between variables to a certain extent, it cannot establish precise causal relationships in the model. In future research, employing longitudinal designs could help to deeper understand the relationships between variables. Second, the sample of this study was drawn exclusively from 34 schools in western Sichuan Province, which limits the generalizability of our findings to the entire population of Chinese adolescents. Regional differences in economic development, educational resources, and physical education facilities may influence adolescents’ exercise habits and smartphone use. Future studies should include more diverse and nationally representative samples to validate our findings. Additionally, a notable limitation of this study is the use of self-reported questionnaires from adolescents. However, it should be noted that common methodological biases likely inherent in such an approach may limit the results. While we employed Harman’s single-factor test to assess common method bias, this approach has relatively low sensitivity. The current study did not implement a time-lagged measurement design or latent method factor control, which means potential method bias cannot be completely ruled out. Future research should adopt these more rigorous methodological and statistical controls to further validate our findings. Furthermore, due to the original data anonymization procedures, school identifiers were not retained in the final analytic dataset; therefore, we were unable to conduct multilevel modeling or apply cluster-robust standard error estimation methods. Consequently, the potential non-independence of observations within schools was not accounted for. This may have resulted in underestimated standard errors and overly narrow bootstrap confidence intervals, thereby inflating the apparent statistical significance of certain associations. Accordingly, the reported findings should be interpreted in light of this limitation. Future research is encouraged to balance participant confidentiality with the retention of anonymized cluster identifiers to enable appropriate multilevel analyses. Finally, this study primarily examines the overall relationship between physical exercise, self-control, meaning in life, and PSU, without categorizing and comparing participants’ types and intensity of physical exercise. Subsequent research could categorize and compare composite variables of physical exercise to explore whether the type of sports may affect study outcomes. Conclusions This study validated that physical exercise, as a key antecedent variable, has a significant direct negative association with on adolescent PSU. Additionally, structural equation model analysis further confirmed the potential partial mediating roles of self-control and meaning in life between the two factors. Importantly, the study revealed that physical exercise is also indirectly associated with PSU through a sequential mediating pathway of “self-control → meaning in life”. This finding deepens the understanding of the underlying mechanisms through which physical exercise influences PSU and expands the theoretical framework within this field. The results indicate that regular physical exercise can effectively enhance adolescents’ self-control and meaning in life, serving as an important means to promote psychological health and foster healthy internet use behaviors among adolescents.