Violence exposure in daily life and adolescent internet gaming disorder in the Chinese context: peer victimization and negative emotion as mediators in a three-wave study Authors: Kenan Qin, Ning Zhang, Xiong Gan, Huiling Fan Published: 2026 ABSTRACT AbstractObjectivesInternet gaming disorder (IGD), a growing concern among adolescents globally, remains insufficiently understood regarding the role of violence exposure, particularly in China. This study examined the prospective associations among violence exposure, peer victimization, negative emotion, and IGD among Chinese adolescents.MethodsA total of 386 junior high school students (47.9% boys; Mage = 12.23, SD = 0.45) from Hubei Province completed three surveys at six-month intervals. Participants reported on exposure to violence, peer victimization, negative emotion, and IGD.ResultsStructural equation modeling showed that violence exposure at baseline (T1) was prospectively associated with IGD one year later (T3). In addition, violence exposure was indirectly associated with later IGD through sequential links involving peer victimization and negative emotion at T2.ConclusionThe present findings provide longitudinal evidence that violence exposure may be associated with adolescent IGD through peer victimization and negative emotion. These findings contribute to a better understanding of the environmental, interpersonal, and emotional processes associated with adolescent IGD and may offer directions for future prevention research.Keywords: Violence exposure, Internet gaming disorder, Peer victimization, Negative emotion, Chinese adolescents, Three-wave longitudinal study Introduction With the rising popularity of digital games, excessive use of internet games has become a public health concern worldwide [59]. Internet gaming disorder (IGD) refers to persistent and compulsive involvement in online games that leads to social and emotional problems [90]. Due to biological and psychological immaturity, adolescence is a pivotal period of increased risk for IGD, especially among boys [26, 77]. Prior studies have demonstrated that IGD is associated with various physical and mental health problems, including executive dysfunction, depression, anxiety, and stress [18, 43]. In addition, IGD has been linked to maladaptive outcomes such as impaired academic performance and diminished social skills [31, 54]. In China, adolescent IGD has become an issue of growing concern. Beyond differences in diagnostic criteria and measurement tools [61, 75], cultural and social contextual factors may also contribute to the relatively high prevalence of IGD among Chinese adolescents [40]. For example, collectivist cultural values place greater emphasis on interpersonal relationships and social belongingness [74], potentially increasing adolescents’ sensitivity to peer experiences and social evaluation. In addition, Chinese adolescents often face intense academic pressure and high parental expectations [6, 9], which may further increase psychological stress. From this perspective, online gaming may function as a compensatory coping strategy through which adolescents attempt to escape from stressful life experiences and emotional distress [44]. Grounded in General Strain Theory (GST; [2], the present study conceptualizes violence exposure as a stressful environmental experience that may contribute to adolescents’ interpersonal difficulties, emotional distress, and maladaptive coping behaviors. Rather than treating violence exposure, peer victimization, negative emotion, and IGD as isolated psychological phenomena, the present study conceptualizes them as a sequential developmental process linking unsafe environments, social threat, emotional distress, and compensatory online engagement. Social Compensation Theory further suggests that adolescents experiencing unmet emotional and interpersonal needs in offline contexts may become more likely to seek compensation through online activities such as gaming [78]. Together, these perspectives provide an integrated framework for understanding the potential pathways from violence exposure to adolescent IGD. Violence exposure and adolescent IGD Violence exposure refers to encounters with violent stimuli that individuals hear, witness, or experience in their daily lives, whether in real-world settings or through media [7, 69]. Such experiences are considered important environmental stressors and have been associated with a range of psychological and behavioral adjustment problems [60, 71]. From a developmental perspective, repeated exposure to unsafe or violent environments may heighten adolescents’ perceptions of threat and psychological vulnerability, thereby increasing the likelihood of maladaptive coping behaviors. Grounded in GST, stressful and adverse life experiences may contribute to maladaptive behaviors aimed at coping with distress [2]. Consistent with this perspective, prior empirical studies have found that violence exposure was associated with problematic internet use and non-substance addictions such as IGD [58, 92]. Liang et al. [47] and Quancai et al. [66] further reported that community and domestic violence exposure were associated with more severe IGD symptoms among Chinese adolescents. This perspective is also broadly consistent with cognitive-behavioral accounts of problematic internet use, which emphasize the role of distal environmental stressors in the development of maladaptive online behaviors [14]. Although violence exposure may take different forms (e.g., hearing about, witnessing, or directly experiencing violence), the present study conceptualizes these experiences as indicators of a broader threat-related environmental stressor. From the perspective of General Strain Theory, these forms of violence exposure share a common characteristic in that they signal an unsafe environment and generate chronic psychological strain, which may subsequently influence adolescents’ social adjustment, emotional regulation, and coping processes. Accordingly, the association between violence exposure and adolescent IGD may not operate directly but rather through intermediate social and emotional adjustment processes. Mediating role of peer victimization Peer victimization refers to repeated and intentional harm inflicted by one or more peers over time, accompanied by an imbalance of power between victims and perpetrators [57, 72]. Because peer interactions are central to adolescents’ social development [30], negative peer experiences such as victimization may undermine psychological well-being and social adjustment [64, 73]. Although peer victimization itself involves experiences of harm, it is conceptually distinct from violence exposure. Violence exposure reflects adolescents’ broader exposure to violence across multiple social contexts (e.g., family, school, and community), whereas peer victimization specifically refers to repeated harmful experiences occurring within peer relationships. Therefore, peer victimization is conceptualized in the present study as a relational experience through which broader environmental risks may influence adolescent adjustment. Within the present framework, peer victimization may represent an important form of social threat through which unsafe environmental experiences become linked to adolescents’ emotional and behavioral adjustment problems. GST suggests that stressful interpersonal experiences may contribute to adolescents’ maladaptive coping behaviors [2]. Consistent with this perspective, adolescents exposed to violence are more likely to experience peer victimization than their non-exposed counterparts [42]. Longitudinal studies have further found that violence exposure was prospectively associated with increased peer victimization [79, 80]. This pathway is also supported by social developmental perspectives emphasizing the role of peer experiences in adolescent behavioral adjustment [32]. Peer victimization may further contribute to adolescents’ problematic online behaviors. According to Social Compensation Theory, adolescents experiencing unmet interpersonal and emotional needs in offline peer relationships may become more likely to rely on online gaming for emotional compensation and social fulfillment [78]. Similarly, self-determination perspectives suggest that online gaming may temporarily compensate for unmet needs for belongingness and relatedness [68]. Empirical studies in both Western and Chinese contexts have consistently linked peer victimization to higher levels of adolescent IGD [22, 35, 48, 81]. Taken together, violence exposure may be associated with adolescent IGD indirectly through subsequent peer victimization experiences. Building on this perspective, we next examine whether negative emotion may further explain this longitudinal association. Mediating role of negative emotion Apart from interpersonal difficulties such as peer victimization, violence exposure may also be associated with adolescent IGD through emotional processes. Grounded in General Strain Theory (GST; [2], stressful and adverse experiences may elicit emotional distress, which in turn contributes to maladaptive coping behaviors. Within this framework, negative emotion may represent an important psychological pathway linking violence exposure to adolescent IGD. Negative emotion refers to individuals’ subjective experiences of distressing emotional states. In the present study, negative emotion is conceptualized as a broader state of emotional distress reflected by symptoms of depression, anxiety, and stress [82]. These emotional responses are among the most common psychological reactions to stressful and adverse life experiences and are highly relevant to the emotional processes emphasized in General Strain Theory. Therefore, the present study focuses on adolescents’ overall emotional distress in response to adverse experiences rather than on any specific emotional symptom individually. Prior studies have consistently shown that adolescents exposed to violence are more likely to experience emotional distress than their non-exposed peers [17]. Exposure to unsafe and stressful environments may be associated with difficulties in emotional regulation and psychological adjustment [27, 69], thereby increasing levels of negative emotion. Consistent with GST, empirical evidence has demonstrated that violence exposure is prospectively associated with elevated depression, anxiety, stress, and other internalizing symptoms [4, 15]. In the Chinese context, studies have similarly found that exposure to violence in family and school settings was associated with higher levels of depression and anxiety among adolescents [45, 85]. Taken together, these findings suggest that violence exposure may be prospectively associated with negative emotion among Chinese adolescents. Negative emotion may further contribute to adolescent IGD. According to Social Compensation Theory, adolescents experiencing emotional distress and unmet psychological needs in offline contexts may become more likely to seek emotional relief and compensation through online gaming [78]. This interpretation is also broadly consistent with psychological compensation perspectives suggesting that excessive online engagement may function as a maladaptive coping strategy for emotional distress [23]. Empirical studies have consistently shown that depression, anxiety, and stress are positively associated with adolescent IGD symptoms [50, 76, 89]. Adolescents with higher levels of emotional distress may therefore become more likely to rely on excessive gaming as a means of escaping or regulating undesirable emotional experiences [49]. Beyond evidence for these separate pathways, previous studies have also identified negative emotion as a potential mediator linking violence exposure to adolescent maladjustment [4, 15]. However, this mediating role has received limited attention in research on violence exposure and IGD among Chinese adolescents. Therefore, the present study proposes that violence exposure may be indirectly associated with adolescent IGD through negative emotion. Peer victimization and negative emotion Prior studies have consistently demonstrated a close association between peer victimization and adolescent emotional problems [53, 84]. Within the present framework, peer victimization may represent an important interpersonal stressor that contributes to adolescents’ emotional distress. According to the interpersonal risk model, negative peer experiences such as rejection, exclusion, and victimization may undermine adolescents’ sense of belonging and psychological security, thereby increasing vulnerability to negative emotions [41]. Because adolescence is a developmental period characterized by heightened emotional sensitivity and increasing interpersonal concerns, adolescents may be particularly vulnerable to the emotional consequences of peer victimization [88]. Empirical evidence from culturally diverse contexts has consistently supported the association between peer victimization and negative emotion. Different forms of victimization, including physical and relational victimization, have been prospectively associated with depression, anxiety, and stress among adolescents in the United States, Japan, and Australia [20, 36, 67]. In China, studies have similarly shown that bullying victimization and other adverse peer experiences were associated with higher levels of depression, anxiety, and loneliness among adolescents [12, 34, 84]. Together, these findings suggest that peer victimization may be prospectively associated with negative emotion among Chinese adolescents. A growing number of studies have further suggested that peer victimization and negative emotion may operate sequentially in the development of adolescent maladjustment. For example, Li et al. [46] and Niu et al. [56] found that negative emotions mediated the association between peer victimization and addictive behaviors, including internet addiction and video game addiction, among adolescents in China. Building on these findings, the present study further proposes that violence exposure may be associated with adolescent IGD through the sequential mediating roles of peer victimization and negative emotion. Specifically, violence exposure may first be associated with increased peer victimization, which may subsequently contribute to elevated negative emotion and, in turn, problematic online gaming behaviors among adolescents. Current study Drawing on General Strain Theory and Social Compensation Theory, as well as prior empirical findings, the present study aims to further examine the longitudinal association between violence exposure and adolescent IGD, along with the underlying psychological mechanisms. Specifically, we constructed a theoretically specified sequential mediation model in which peer victimization was modeled as preceding negative emotion based on prior theory and empirical evidence. Based on the current literature, the following hypotheses are proposed (see Fig. 1): Conceptual sequential mediating model across three time points Hypothesis 1: T1 violence exposure would be positively associated with adolescent T3 IGD. Hypothesis 1: T1 violence exposure would be positively associated with adolescent T3 IGD. Hypothesis 2: T2 peer victimization would mediate the association between T1 violence exposure and T3 IGD among adolescents. Hypothesis 2: T2 peer victimization would mediate the association between T1 violence exposure and T3 IGD among adolescents. Hypothesis 3: T2 negative emotion would mediates the association between T1 violence exposure and T3 IGD among adolescents. Hypothesis 3: T2 negative emotion would mediates the association between T1 violence exposure and T3 IGD among adolescents. Hypothesis 4: T2 peer victimization and T2 negative emotion would sequentially mediate the association between T1 violence exposure and T3 IGD among adolescents. Hypothesis 4: T2 peer victimization and T2 negative emotion would sequentially mediate the association between T1 violence exposure and T3 IGD among adolescents. Method Participants and procedures Using random cluster sampling, participants were recruited from public secondary schools in Hubei Province, China. The data collection process began in the first semester of the academic year 2021–2022, and then the survey was conducted every six months over three waves. At Time 1 (October 2021), a total of 412 students took part in our study. Of them, 399 (attrition rate = 3.155%) and 386 (attrition rate = 6.311%) adolescents continued to participate in the second and third rounds, respectively (see Table 1). To test attrition, we conducted independent samples t-tests comparing key variables between the dropout and retained groups. The results indicated no significant differences, supporting the assumption that attrition was random in this study (see Table S1 in supplementary materials). Thus, the attrition is unlikely to have caused systematic bias or affected the generalizability of the results. The final sample consisted of 386 adolescents ranging in age from 11 to 13 years (Mage = 12.225, SD = 0.454), of whom 185 (47.9%) were boys and 201 (52.1%) were girls. Table 1Participant attrition across time pointsTime PointParticipants RetainedParticipants Dropped OutAttrition Rate (%)Time 1412--Time 2399133.155Time 3386266.311Attrition rate is calculated based on the number of participants who dropped out at each time point divided by the initial sample size at Time 1 Participant attrition across time points Attrition rate is calculated based on the number of participants who dropped out at each time point divided by the initial sample size at Time 1 Furthermore, all procedures involving human participants were reviewed and approved by the Ethics Committee of the corresponding author’s institution (Protocol # csz2024‑001).At first, they explained the purpose of the survey and the principles of voluntariness, confidentiality. Then, paper-and-pencil questionnaires were distributed by class during school hours to measure subjects’ basic information and their development regarding the study variables. All procedures were similar at each time point and lasted about 15 min. At the end of each survey, we thank all students and teachers for their cooperation. Measures Violence exposure at time 1 We utilized the Chinese version of the Violence Exposure Questionnaire (VEQ), developed by Schwartz and Proctor [69] and amended by Wu [86], to investigate the status of violence that teens see, hear, or personally experience in their daily lives. The unidimensional questionnaire includes 25 items (e.g., “Threatened or injured by a knife, gun, or other sharp object”). Participants were asked to score the items on a 5-point scale (from “1 = never” to “5 = always”), with the higher total mean scores suggesting higher real-life exposure to violent information. The Cronbach’s α and McDonald’s ω coefficients of the VEQ were 0.948 and 0.949, respectively. Peer victimization at time 2 The present study used the victimization sub-questionnaire of the Chinese version of the Olweus Bully/Victim Questionnaire (OBVQ), revised by Zhang and Wu [93], to measure the extent to which adolescents were victimized by peers. The questionnaire consists of 6 items, which are classified as physical, relational, and verbal victimization. All items were rated on a 5-point scale (from “1 = never happened” to “5 = several times a week”), with the higher overall mean scores indicating more frequent victimization. According to past research, this measurement has demonstrated good reliability in the Chinese context [65]. The Cronbach’s α and McDonald’s ω coefficients were 0.725 and 0.778, respectively. Negative emotion at time 2 This study assessed teenagers’ negative emotional experiences by using the Chinese version of the Depression Anxiety Stress Scale (DASS), developed by Lovibond and Lovibond [52] and revised by Gong et al. [25]. The scale comprises 21 items divided into three dimensions, including depression, anxiety, and stress. Students were required to score on a 4-point scale (from “0 = not like me at all” to “3 = very much like me”), with the higher total mean scores showing higher levels of negative emotion. Also, the DASS was found to have good reliability and validity among the Chinese adolescent population [28]. The Cronbach’s α and McDonald’s ω coefficients were 0.930 and 0.932, respectively. Internet gaming disorder at time 3 The current study used the Chinese version of the Internet Gaming Disorder Questionnaire (IGDQ), adapted from Gentile [24] by Yu et al. [91], to evaluate IGD symptoms among juveniles. It is unidimensional and contains 11 items. These items were rated on a 3-point scale (from “1 = never” to “3 = often”), with the higher total mean scores representing a greater likelihood of IGD. The IGDQ showed good reliability among Chinese teens in the past [65]. The Cronbach’s α and McDonald’s ω coefficients were 0.794 and 0.828, respectively. Covariates at time 1 In past studies, gender and age have been identified as key influences on adolescent peer victimization, negative emotion, and IGD [4, 47, 79]. Accordingly, adolescents’ gender (1 = boys,2 = girls) and age were included as covariates in subsequent analyses. Analytic strategy SPSS 25.0 and Mplus 8.3 were used to test the hypotheses. Initially, SPSS 25.0 was utilized to carry out preliminary analyses. Second, we examined the sequential mediation model by constructing measurement and path models using latent variable SEM via Mplus 8.3 [55]. To assess the model fit, indices including chi-square (χ2), degrees of freedom (df), comparative fit index (CFI), Tucker-Lewis index (TLI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR) were reported. It has been pointed out that a model could be considered a good fit if it meets the criteria of χ2/df less than 3, CFI and TLI values greater than 0.95, and RMSEA and SRMR values less than 0.06 [39]. Indirect effects were tested with bias‑corrected bootstrapping, and 5,000 resamples were used as commonly recommended for stable estimation [62]. The mediating effects were statistically significant if the 95% confidence intervals (CIs) did not include zero. Result Common method bias test In order to determine whether the bias had an impact on the interpretation of the results, both exploratory and confirmatory procedures were used in our study. First, in terms of exploratory procedure, the results of the Harman single-factor analysis showed that there were 13 factors with characteristic root values greater than 1, and the first factor explained only 23.205% of the variance, which is much less than the threshold of 40% [29]. Second, the confirmatory procedure was conducted through model comparisons. This approach involves testing whether the difference between the confirmatory factor analysis (CFA) model (M1) and another model (M2) with method factors added to M1 is statistically significant [87]. The results of the model comparison (M2-M1) demonstrated that the change value of each fit index was less than 0.040 (ΔTLI = 0.038, ΔCFI = 0.030, ΔRMSEA = −0.004, and ΔSRMR = −0.009), indicating that the model with the addition of the method factor did not significantly improve the baseline model. As a result, our data did not have a serious problem of common method bias. Descriptive statistics and bivariate analyses Table 2 presents the results of descriptive statistics and bivariate analyses for the main variables. More specifically, there were significant and positive correlations (rs = 0.155 ~ 0.342, ps < 0.001) between two of the four variables, including T1 violence exposure, T2 peer victimization, T2 negative emotion, and T3 IGD. Moreover, regarding the relationships between control variables and study variables, only gender was found to be positively related to T2 negative emotion (r = 0.155, p = 0.002), but none of the other relationships were significant. Table 2Descriptive statistics and correlations among main variablesMSDVariables123451. Gender––12. T1Age12.2250.454–13. T1VE1.3030.460−0.0100.00914. T2PV1.1950.405−0.0720.0410.155**15. T2NE0.3550.4560.155**0.0790.312***0.319***16. T3IGD1.1390.211−0.0340.0110.194***0.342***0.338***N = 386. VE Violence Exposure, PV Peer Victimization, NE Negative Emotion, IGD Internet Gaming Disorder, SD Standard Deviation**p < 0.01***p < 0.001 Descriptive statistics and correlations among main variables N = 386. VE Violence Exposure, PV Peer Victimization, NE Negative Emotion, IGD Internet Gaming Disorder, SD Standard Deviation **p < 0.01 ***p < 0.001 Mediating effects of peer victimization and negative emotion The current study constructed a sequential mediation model through SEM. In the first step, we used CFA to test the measurement model, which is made up of four latent variables. Among them, peer victimization contained three observable variables, including physical, relational, and verbal victimization. Negative emotion comprised three observable variables, namely depression, anxiety, and stress. For violence exposure and IGD, which are measured in a unidimensional way, we packaged them into four and three observable variables, respectively, via the item-structure balance method that contributed to the stability and fit of the model estimation [51]. This method involves grouping items to create balanced parcels that reflect the underlying construct evenly, helping to reduce measurement error and improve model estimation reliability, especially in structural equation modeling. To conclude, the CFA results indicated good measurement model fit data (χ2/df = 2.000, CFI = 0.982, TLI = 0.977, RMSEA = 0.051, and SRMR = 0.034). In the second step, this study tested the path model. We first investigated the direct effect of T1 violence exposure on T3 IGD. The results revealed a direct effect model with good fit (χ2/df = 1.397, CFI = 0.995, TLI = 0.993, RMSEA = 0.032, and SRMR = 0.022). The direct pathway from T1 violence exposure to T3 IGD was significant and positive (β = 0.218, p = 0.001) after controlling for gender and age, thus supporting hypothesis 1. Next, we further inspected the mediating effects of T2 peer victimization and T2 negative emotion by incorporating control variables, identifying hypotheses 2 and 3. The results found the sequential mediation model also fitted the data well (χ2/df = 1.688, CFI = 0.984, TLI = 0.979, RMSEA = 0.042, and SRMR = 0.032). Figure 2 illustrates the path coefficients and their significance. In particular, T1 violence exposure significantly predicted peer victimization and negative emotion at the next time (β = 0.184, p = 0.007; β = 0.279, p < 0.001). At T2, peer victimization predicted an increase in negative emotion (β = 0.323, p < 0.001). Both T2 peer victimization and T2 negative emotion had positive effects on T3 IGD (β = 0.296, p = 0.004; β = 0.278, p < 0.001). However, the direct effect of T1 violence exposure on T3 IGD was nonsignificant (β = 0.067, p = 0.310), showing that the mediating variables of this model acted as complete mediators. Tests of the mediation model showed the indirect effects of T2 peer victimization and T2 negative emotion in the association between T1 violence exposure and adolescent T3 IGD. Note. N = 386. All estimated parameters are standardized. **p < 0.01, ***p < 0.001 As displayed in Table 3, the findings of the bootstrapping analysis of the sequential mediation model indicated that the total indirect effect was significant (β = 0.148, SE = 0.042, 95% CI = [0.081, 0.253]). Of these, both T2 peer victimization and T2 negative emotion separately mediated the relationship between T1 violence exposure and T3 IGD (ind1: β = 0.054, SE = 0.027, 95% CI = [0.017, 0.134]; ind2: β = 0.077, SE = 0.031, 95% CI = [0.030, 0.149]). More crucially, the sequential mediating effect of peer victimization and negative emotion at T2 was also significant (ind3: β = 0.016, SE = 0.008, 95% CI = [0.006, 0.044]), validating hypothesis 4. Besides, the effect values of three indirect pathways accounted for 25.116%, 35.814%, and 7.442% of the total effect value, respectively. Table 3The results of total, direct, and indirect effects of sequential mediation modelEffectsEstimateBoot SEBoot LLCIBoot ULCIInd1: T1VE → T2PV → T3IGD0.0540.0270.0170.134Ind2: T1VE → T2NE → T3IGD0.0770.0310.0300.149Ind3: T1VE → T2PV → T2NE → T3IGD0.0160.0080.0060.044Total indirect effect0.1480.0420.0810.253Direct effect0.0670.068−0.0600.206Total effect0.2150.0740.0840.376N = 386. VE Violence Exposure, PV Peer Victimization, NE Negative Emotion, IGD Internet Gaming Disorder, SE Standard Error, Bootstrap Sample Size = 5000, CI Confidence Interval, LL Lower Limit, UL Upper Limit The results of total, direct, and indirect effects of sequential mediation model N = 386. VE Violence Exposure, PV Peer Victimization, NE Negative Emotion, IGD Internet Gaming Disorder, SE Standard Error, Bootstrap Sample Size = 5000, CI Confidence Interval, LL Lower Limit, UL Upper Limit Discussion Considering the high harm of IGD in adolescents, our study explored its risk factors from both internal (i.e., negative emotion) and external (i.e., violence experiences and peer victimization) perspectives using a longitudinal design. Employing a sequential mediation model within the SEM framework, we found that T1 violence exposure was prospectively associated with T3 IGD primarily through through the separate and sequential mediating roles of T2 peer victimization and negative emotion, whereas the direct pathway was not statistically significant. Specific interpretations and discussions of the results are provided in the following paragraphs. Violence exposure and internet gaming disorder This study revealed that exposure to violent situations was prospectively associated with later IGD among adolescents, consistent with prior findings in China [47, 66]. The difference is that we used a study design spanning three time points to capture the dynamic interplay between them, rather than cross-sectional designs. Hence, the present findings are broadly consistent with the cognitive-behavioral model of internet use and the social compensation hypothesis [14, 78]. Witnessing and experiencing violence may elicit teenagers’ stress responses and behavioral changes, such as pathological online gaming use [14, 71]. Concretely, positive experiences arising from excessive online gaming play an important role in compensating for violent episodes, thus potentially increasing adolescents’ reliance on internet gaming as a coping strategy [78, 94]. The role of peer victimization The findings indicated that peer victimization played a mediating role in the association between early violence exposure and later IGD. This aligns with the social development model [32], which highlights peers as key links between environmental factors and developmental outcomes in Chinese adolescents [63]. On the one hand, T1 violence exposure positively predicted T2 peer victimization, in line with previous results [16]. This supports social learning theory [3, 5], which suggests that individuals exposed to violence internalize aggressive behaviors. As a result, they may normalize violence or aggression directed at them [19, 38]. These changes can impair behavioral functioning, increasing the likelihood of negative peer interactions [70, 79]. On the other hand, T2 peer victimization positively predicted T3 IGD, in line with past research [48]. Juveniles who experience frequent bullying may struggle to fulfill basic psychological needs in their peer groups. To compensate, they may turn to internet gaming [68, 78]. Online games provide an alternative identity and a temporary escape from real-life stress [83]. However, over time, this coping pattern may be associated with a greater likelihood of problematic gaming behaviors over time. The role of negative emotion Our results suggested that T2 negative emotion played a mediating role in the association between T1 violence exposure and T3 IGD. This finding is consistent with General Strain Theory (GST, [1, 2]. According to GST, violent experiences may be associated with elevated negative emotions. In turn, adolescents may use online games to cope, which may be associated with higher levels of IGD. Specifically, T1 violence exposure positively predicted T2 negative emotion, consistent with prior research [11, 33]. From the perspective of emotional security, violence from family, peers, or the community reduces adolescents’ sense of safety, increasing negative affect [13, 45]. In the second half of the pathway, T2 negative emotion positively predicted T3 IGD, in line with previous studies [49, 76]. Online gaming provides adolescents a way to relieve distress and temporarily divert attention from negative emotions [37]. However, chronic overplaying of online games creates maladaptive compensatory coping pattern, and further, this loss-compensation-satisfaction paradigm exacerbates online game use, ultimately may be linked to greater vulnerability to IGD [21, 23]. The roles of peer victimization and negative emotion We found that higher violence exposure was associated with greater adolescent peer victimization, which in turn was associated with higher levels of negative emotion and eventually was linked to IGD in Chinese contexts. Regarding the association between both mediators, T2 peer victimization was positively associated with T2 negative emotion. This finding is similar to prior studies [20, 67] and offers empirical support for the interpersonal risk model [41]. Teens who undergo more negative social experiences such as exclusion, rejection, conflict, and victimization are more likely to form negative cognitive schemas and self-evaluations [41, 63]. As poor interpersonal conditions intensify, individuals ultimately are more likely to suffer from mental health problems [10]. Overall, this study underscores the potential roles of peer and emotional processes in the association between violence exposure and IGD, offering further insight into the possible pathways underlying this relationship. Limitations and future prospects While the present study provides valuable insights, it has some limitations. First, baseline levels of IGD, peer victimization, and negative emotion were not controlled in the longitudinal mediation model. Therefore, the present findings mainly support temporally ordered associations rather than strong evidence of developmental change or causal processes. Future studies should incorporate baseline-adjusted or cross-lagged panel models to further examine longitudinal developmental dynamics. Second, self-reported measures may introduce biases due to social desirability and recall deviations. Future research should incorporate multi-informant data, such as reports from parents, teachers, and peers, to improve reliability. Third, the sample was limited to adolescents from central China, restricting generalizability. Future studies should include diverse regions or cross-cultural comparisons to validate the findings. Fourth, the six-month measurement interval may not fully capture emotional fluctuations from peer victimization. Longer intervals with more frequent follow-ups could better reveal developmental patterns. Fifth, the current study included a relatively limited set of covariates, and some important factors (e.g., socioeconomic status and prior gaming behavior) were not included in the analyses, the possibility of omitted variable bias cannot be fully ruled out. Therefore, the indirect path should be interpreted as consistent with the proposed sequential mechanism, but it is not definitive evidence. Sixth, our deficit-focused approach may overlook adolescents’ developmental potential [21]. Subsequent research could adopt a positive youth development (PYD) perspective to explore protective factors against IGD. In addition, future studies could further investigate subjective psychological processes associated with violence exposure, such as perceived vulnerability, insecurity, or social threat, to provide a more nuanced understanding of the mechanisms underlying adolescent IGD [8]. Conclusion The present study employed a three-wave longitudinal design to examine the associations among violence exposure, peer victimization, negative emotion, and adolescent IGD. The findings indicated that: (1) violence exposure prospectively predicted later IGD; (2) peer victimization served as an indirect pathway linking violence exposure to IGD; (3) negative emotion also served as an indirect pathway linking violence exposure to IGD; and (4) peer victimization and negative emotion jointly formed a theory-driven indirect pathway between violence exposure and IGD. These findings contribute to a better understanding of the environmental, interpersonal, and emotional processes associated with adolescent IGD. Acknowledgements The authors thank the participants who cooperated in this study. Authors’ contributions Q.K. was involved in conceptualization, drafting and revising the manuscript. Z.N. was involved in data processing and analysis, and result visualization. G.X. was responsible for project funding, project management, and participation in discussions. F.H. participated in data collection, manuscript review, and data verification. Q.K. and Z.N. are co-first authors of the article. Funding This research was supported by the General Project of Education under the National Social Science Fund: “Research on Identification and Intervention of Short Video Addiction among Left-Behind Children Based on Multimodal Data and Machine Learning” (BBA250048). Data availability The data that support the finding of this study was available from the corresponding author upon reasonable requests. Declarations Ethics approval and consent to participate This survey was approved by the ethics committee of psychological research at Yangtze University and adhered to the Declaration of Helsinki. Informed consent was obtained from both participants and their parents. Consent for publication All authors have approved the manuscript for publication. Consent for publication is not applicable as the study does not involve any identifying images or personal/clinical details of participants that compromise anonymity. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Kenan Qin and Ning Zhang are co-first authors of the article. Contributor Information Xiong Gan, Email: 307180052@qq.com. Huiling Fan, Email: 121528529@qq.com. References Associated Data Data Availability Statement The data that support the finding of this study was available from the corresponding author upon reasonable requests.