Understanding internet gaming disorder in Mexican university students: prevalence and associated factors Authors: Júlia Gisbert-Pérez, Manuel Martí-Vilar, Guillermo M. Chans, Laura Badenes-Ribera Published: 2026 ABSTRACT This study estimated the prevalence of Internet Gaming Disorder (IGD) among Mexican university students and identified key risk and protective factors. Using a quantitative cross-sectional design, data were collected from a convenience sample of 712 students (Mage = 19.76 years, SD = 1.68; 42.7% female). Pearson chi-squared tests, correlation analyses, and binomial logistic regression were employed to examine associations and predictors of IGD. The estimated prevalence of probable IGD was 2.1% in the overall sample and 5.4% among active gamers. Gaming-related factors, such as weekly gaming hours, exposure to toxic gaming communities, and motivations based on violent gratification or coping, were positively associated with IGD, as were specific impulsivity traits. Conversely, gaming for fun, perceived social support, and emotional intelligence were negatively associated with IGD. Logistic regression analysis revealed that lack of perseverance (OR = 1.324, 95% CI [1.056 to 1.660]) and lower perceived peer social support (OR = 0.795, 95% CI [0.647 to 0.977]) were statistically significantly associated with increased likelihood of IGD among Mexican gamers. These findings provide evidence for developing culturally tailored interventions in university settings to mitigate the risks associated with problematic gaming behaviors.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1038/s41598-026-58721-6. Introduction Video games have become one of the most popular forms of entertainment and social interaction worldwide. Recent global survey data indicate that approximately 79% of young adults aged 18–29 engage in gaming at least weekly, reflecting the widespread integration of gaming into daily life across regions such as North America, Europe, Asia-Pacific, and Latin America1. Given this high level of engagement, concerns have emerged about the potential negative effects of excessive gaming, particularly the development of Internet Gaming Disorder (IGD)2. IGD is characterized by a pattern of gaming behavior that causes significant impairment in personal, academic, or social functioning3. While the gaming landscape provides new opportunities for learning, recreation, and socialization, it also poses challenges in managing the risks associated with problematic gaming behavior. According to the Diagnostic and Statistical Manual of Mental Disorders, 5th ed., text Rev. (DSM-5-TR)4, IGD is defined as a persistent and recurrent pattern of gaming that leads to clinically significant impairment or distress over 12 months. Diagnosis requires meeting at least five of nine criteria: (1) preoccupation with gaming; (2) withdrawal symptoms, such as irritability, anxiety, or sadness when gaming is unavailable; (3) tolerance, or the need to spend increasing amounts of time gaming; (4) unsuccessful attempts to control play; (5) loss of interest in other hobbies and activities; (6) continued excessive gaming despite awareness of psychosocial problems; (7) deception about the amount of time spent gaming; (8) using gaming to escape or relieve negative moods; and (9) jeopardizing or losing significant relationships, academic performance, or career opportunities due to gaming. IGD is included in Section III of the DSM-5-TR as a “Condition for Further Study.” Similarly, the International Classification of Diseases, 11th Revision (ICD-11)3 recognizes Gaming Disorder as a behavioral pattern marked by impaired control over gaming, prioritization of gaming over other activities, and continuation despite adverse consequences. This condition may occur in both online and offline contexts and must result in substantial functional impairment. A recent meta-analysis across 17 countries estimated the global prevalence of IGD at 3.05%5. Prevalence varies across regions, with higher rates reported in Asia (5.08%) and the United States (2.74%)5, 6. In Latin America, Mexico stands out as one of the fastest-growing gaming markets, underscoring the need to understand how IGD manifests in emerging digital contexts7, 8. IGD also shows age-related variability. It commonly emerges during adolescence9, with approximately 4.6% affected10, and more recent studies indicate rising prevalence among both adolescents and adults (8.8%–10.4%)11. University students represent a particularly vulnerable group12, as early adulthood is marked by socioemotional instability13 and intensive engagement with internet-connected technologies for academic, social, and leisure purposes14, 15. IGD in this population has been linked to poorer academic performance and reduced social functioning16. Following the COVID-19 pandemic, increased screen time contributed to a rise in gaming-related problems17. Before the pandemic, IGD prevalence among Mexican university students was estimated at 5.2% (95% CI: 4.7%–5.8%)18, situated within the broader range reported internationally (3.8%–11%)12, 19. Despite this growth, post-pandemic data on IGD in Mexico remain limited, highlighting the need for updated evidence in this population. The pandemic also reinforced the use of video games as social tools, leading to a surge in online gaming20. Although these spaces can foster social connection, popular titles such as League of Legends, Counter-Strike, and Call of Duty are often associated with toxic community dynamics. “Toxicity” encompasses antisocial and disruptive behavior directed at other players, including verbal abuse, trolling, and hateful communication in-game chats and social media21. The causes of such behavior remain uncertain but may involve increasing game difficulty and frustration22. While early research primarily examined the role of violent content in fostering aggression, emerging evidence indicates that the social context of play may exert a stronger influence on players’ behavior than the content itself23. Because IGD criteria often involve manifestations of impulsivity, loss of control, and problematic social conduct24, it is crucial to investigate how gaming community toxicity and individual gaming motivations may contribute to the development and maintenance of IGD. Understanding how these psychosocial and motivational factors contribute to IGD can be further clarified through the Interaction of Person–Affect–Cognition–Execution (I-PACE) model25, 26. This model proposes that IGD results from the interplay between individual predispositions (e.g., impulsivity), affective responses (e.g., difficulties in emotion regulation, maladaptive beliefs, or coping motives), and deficits in executive control. Recent updates to the model highlight that low emotional intelligence, stress-reactive coping, and poor self-regulation increase susceptibility to problematic gaming. At the same time, reward sensitivity and habitual responding may reinforce excessive use27. Recent systematic reviews and meta-analyses have identified gaming motivations as significant predictors of IGD, alongside gaming time11, 28, 29. In particular, coping and escapism motivations—using gaming to avoid negative emotions or to disengage from real-life stressors—are strongly associated with IGD symptoms30. Motivations related to satisfying basic psychosocial needs, such as social interaction, achievement, and the flow experience, have also shown significant correlations with IGD31. Although the motivation to compete is positively linked to IGD, its association tends to be weaker than that of escapism motives29. While the relationship between gaming motivations and IGD is well established32, psychological needs and life contexts differ considerably across developmental stages. For instance, escapism and competition motives exhibit stronger associations with IGD among adolescents than adults33. Because adolescents and university-aged individuals face distinct developmental tasks, social roles, and stressors, it is essential to identify which gaming motivations increase IGD risk within each age group. Beyond motivational factors, several psychosocial characteristics, such as emotional intelligence, perceived social support, and impulsivity, play a key role in the onset and maintenance of behavioral addictions, including IGD. Emotional intelligence, defined as the ability to perceive, understand, and regulate one’s emotions and those of others, appears to serve a protective function34. Individuals with lower emotional intelligence may experience greater psychological distress, heightening their vulnerability to maladaptive gaming behaviors33. Similarly, perceived social support is a buffer against IGD, as individuals with strong social networks are better equipped to manage daily challenges and satisfy their interpersonal needs29, 31. Although gaming can promote new social connections27, individuals with lower social competence or limited offline support tend to be more susceptible to IGD35. Impulsivity has consistently shown a positive association with IGD, reflecting difficulties in inhibitory control and a preference for immediate rewards36. Many video games provide instant feedback and gratification, which can reinforce maladaptive patterns among individuals with poor self-regulation. Moreover, deficits in executive functioning can impair impulse control and behavioral regulation, contributing to excessive gaming37. Because executive functions and emotional regulation continue to mature across development, age likely moderates the relationship between psychosocial factors and IGD33. Cultural context should also be considered: cross-national studies have revealed meaningful variations in the predictors of internet-related addictions38. This highlights the need to examine how psychosocial and motivational correlates of IGD operate across different age groups and sociocultural settings. Given the established association between psychosocial and gaming-related factors in the development of IGD and the particular vulnerability of university students in rapidly expanding gaming markets such as Mexico, it is crucial to examine how these variables interact in the post-pandemic context. Understanding these interactions within current sociocultural and temporal frameworks is essential for designing educational and mental health interventions tailored to local needs. Accordingly, this study was guided by the following research question: What is the prevalence of IGD among Mexican university students, and how are psychosocial and gaming-related factors associated with its presence? To address this question, the study pursued two main objectives: (1) to estimate the prevalence of IGD among Mexican university students and (2) to examine the associations between psychosocial and gaming-related factors and IGD, and to evaluate their relative contribution to the likelihood of meeting IGD criteria. Based on prior evidence and the post-pandemic context of gaming in Mexico, we formulated the following hypotheses: (H0) the prevalence of IGD among Mexican university students would be higher than pre-pandemic estimates reported for this population; (H1) higher weekly gaming hours, greater exposure to toxic gaming communities, and playing primarily with acquaintances would be positively associated with IGD severity; (H2) coping/escapism and violent gratification motives would be positively associated with IGD, whereas (H3) gaming for fun would be negatively associated with IGD severity; (H4) higher levels of impulsivity would be related to greater IGD severity; (H5) higher perceived social would be negatively associated with IGD; (H6) higher emotional intelligence across its components would be negatively associated with IGD; and (H7) these psychosocial and gaming-related factors would significantly predict the likelihood of meeting IGD criteria in this population. Materials and methods Sample and procedure A cross-sectional design with non-probabilistic sampling was employed. Data were collected through an online survey administered in university classrooms in Tecnologico de Monterrey (Mexico) between January and April 2025. The recruitment process began with on-site invitations during in-person classes for various undergraduate courses. A trained, standardized team researcher briefly presented the project to the students and invited them to participate voluntarily and individually. To avoid self-selection bias, no incentives were offered. To reduce social desirability bias, participants were assured of anonymity in their responses. Participants provided informed consent before data collection. The study followed the ethical principles of the Declaration of Helsinki (1964), and subsequent amendments, and was approved by the Institutional Committee on Research Ethics of Tecnologico de Monterrey (protocol no. EHE-2023-08). The target population consisted of Spanish-speaking university students aged 18 or older residing in Mexico. An a priori sample size calculation was estimated using G*Power, which indicated a minimum sample size of n = 596 (α = 0.05, power = 0.80, event probability = 0.06). The final sample comprised 712 participants (Mage = 19.76, SDage = 1.68; range: 18–30 years, 42.7% female). Regarding households’ monthly income (expressed in Mexican pesos, MXN), half of the sample (49.2%) reported earnings within the $10,001–$50,000 MXN range, 33.3% reported incomes above $80,000 MXN, and 10.1% reported incomes below $10,000 MXN. No missing data were identified. To ensure data quality, cases exhibiting invariant response patterns, defined as providing the same response to all items, were removed from the dataset, as such patterns are considered indicators of low response engagement. Measures Sociodemographic data Participants reported their age, gender, and current activity (studying, working, or both). Gaming data Participants provided information on their average weekly gaming time (in hours), preferred gaming mode (alone, with family or friends, acquaintances, or strangers), and the perceived level of toxicity in their gaming communities. Toxicity was rated using a single item created by the authors and scored on a scale of 1–5 (how toxic do you consider the community or players to be in the games you usually play?). Gaming motivations The Motives in Video Game Use scale (e-MUV39 assesses eight dimensions of gaming motivation through 32 items: Immersion (items 1–4), Customization (items 5–8), Violent gratification (items 9–12), Coping (items 13–16), Fun (items 17–20), Cognitive challenge (items 21–24), Competition (items 25–28), and Social interaction (items 29–32). Items are rated on a 5-point Likert-type scale (1 = strongly disagree; 5 = strongly agree), with higher scores reflecting greater importance of each motivational domain. In the present study, internal consistency was satisfactory for all subscales, with Cronbach’s α and McDonald’s ω as follows: Immersion (α = 0.87, ω = 0.88), Customization (0.83, 0.84), Violent gratification (0.88, 0.88), Coping (0.82, 0.81), Fun (0.92, 0.92), Cognitive challenge (0.82, 0.82), Competition (0.74, 0.75), and Social interaction (0.79, 0.79). Internet gaming disorder The Internet Gaming Disorder Scale (IGDS40 assesses levels of Internet Gaming Disorder through nine items reflecting the DSM-5 diagnostic criteria. Responses are rated on a six-point Likert-type scale (0 = never, 1 = one to four times in the past year, 2 = five to 11 times in the past year, 3 = once to three times per month, 4 = once or more per week, and 5 = every day or almost every day). A criterion is considered met when a participant responds with a 4 or 5. Participants who meet at least five of the nine criteria are classified as disordered40. An alternative scoring method proposed by Evren et al.41, in which responses from 3 to 5 are recognized as meeting the criterion, was also examined. The IGDS was translated into Spanish following the backward-forward translation procedure42. A confirmatory factor analysis (CFA) using Diagonal Weighted Least Squares (DWLS) was conducted to assess the unidimensional structure of the IGDS, which exhibited adequate fit. The present study’s internal consistency was excellent (Cronbach’s α = 0.92; McDonald’s ω = 0.92). Impulsive behavior The short version of the UPPS-P Impulsive Behavior Scale43(Spanish short version of Cándido et al.44) evaluates five impulsivity traits using 20 items: negative urgency (NU; items 4, 7, 12 and 17), lack of premeditation (LPM; items 1, 6, 13 and 19), lack of perseverance (LPS; items 5, 8, 11 and 16), sensation seeking (SS; items 3, 9, 14 and 18) and positive urgency (PU; items 2, 10, 15 and 20). Responses are rated on a 4-point Likert-type scale (1 = strongly agree; 4 = strongly disagree), with higher scores indicating greater impulsive behavior (α = 0.70, ω = 0.70), and subscale reliability values were satisfactory: NU (0.76, 0.76), LPM (0.78, 0.78), LPS (0.85, 0.85), SS (0.76, 0.76), and PU (0.77, 0.77). Perceived social support The Multidimensional Scale of Perceived Social Support (MSPSS45; Spanish version of Santos-Vega et al.46) measures perceived support from three sources across 12 items: perceived family support (PFS; items 1–4), perceived peer support (PPS; items 5–8), and support from other significants (POS; items 9–12). Responses are rated on a 4-point Likert scale (1 = never or almost never; 4 = always or almost always). Higher scores indicate stronger perceived social support. In this study, the total scale showed excellent reliability (α = 0.93, ω = 0.93). Subscale reliability coefficients were also high: PFS (0.88, 0.88), PPS (0.90, 0.90), and POS (0.93, 0.93). Emotional intelligence The Wong and Law Emotional Intelligence Scale (WLEIS-S47; Spanish version of Pacheco et al.48) assesses emotional intelligence across 16 items representing four dimensions: self-emotion appraisal (SEA, items 1, 5, 9, 13), others’ emotions appraisal (OEA, items 2, 6, 10, 14), emotion awareness (EA, items 3, 7, 11, 15), and emotion regulation (ER, items 4, 8, 12, 16). Responses are rated on a 7-point Likert scale (1 = totally disagree, 7 = totally agree). Higher scores indicate greater emotional intelligence. The total scale demonstrated excellent reliability (α = 0.95, ω = 0.95), with subscale reliability coefficients as follows: (0.84, 0.84), OEA (0.83, 0.83), EA (0.86, 0.86), and ER (0.87, 0.87). Statistical analysis Categorical variables were summarized using frequencies and percentages, while continuous variables were described using means and standard deviations. The prevalence of IGD among gamers was calculated by dichotomizing IGDS scores according to DSM-5 criteria. Univariate skewness and kurtosis values for all variables fell within acceptable limits for normality (± 2 for skewness, ± 7 for kurtosis49, 50. Two approaches were identified for transforming the ordinal IGDS scores into dichotomous scores to establish criteria fulfillment and IGD cutoff: one conservative40 and one lenient41. Thus, the proportions of IGD were calculated based on both methods. Correlations between IGD scores and continuous variables were analyzed using Pearson’s correlation coefficients, while chi-square (χ2) tests were used to assess relationships between IGD and categorical variables. Effect sizes for chi-square tests were estimated using Cramer’s V. A binary logistic regression with forward (LR) selection was conducted to identify significant predictors of IGD51. Because IGD was a low-frequency outcome and multiple predictors were considered, forward (LR) selection was used to derive a parsimonious exploratory logistic model, noting that stepwise procedures may yield sample-dependent solutions. Given the small number of IGD cases and the relatively large set of candidate predictors, we applied an initial bivariate prescreening to reduce the number of variables entered into the forward logistic regression and to obtain a more parsimonious multivariable model, thereby minimizing the risk of overfitting52. All variables assessed in the preliminary analyses were initially considered for candidate selection. These included sociodemographic variables, gaming-related variables, gaming motivation dimensions, impulsivity dimensions, perceived social support dimensions, and emotional intelligence dimensions. Variables showing a statistically significant bivariate association with IGD at p <.05 were retained as candidate predictors for the multivariable logistic regression. A forward stepwise logistic regression procedure was then applied, with variables entered sequentially based on statistical significance (entry criterion: p <.05; removal criterion: p >.10). Model fit was evaluated using several indices53: (1) the χ2 test for overall model significance (change in −2LL); (2) the Hosmer-Lemeshow test for goodness of fit and Nagelkerke’s R2 for explained pseudo-variance; and (3) classification accuracy to assess predictive performance. The Wald statistic was used to test the significance of each regression coefficient. Odds ratios (ORs) and 95% confidence intervals were calculated as effect size estimates, testing the null hypothesis that OR = 1. Before conducting the regression analyses, multicollinearity among the predictor variables was examined. Tolerance values lower than 0.10 and Variance Inflation Factor (VIF) values exceeding 10 were considered indicative of serious multicollinearity54. In the present data, all predictors exhibited tolerance values well above 0.20 and VIF values below 5. Internal validation was conducted using stratified bootstrap resampling to evaluate the stability of logistic regression coefficients under sampling variability55. This approach enables a more robust assessment of parameter uncertainty in exploratory models with limited data and class imbalance56 (see Supplementary Materials). All analyses were performed using IBM SPSS Statistics (Version 28.0 for Windows). Results Sociodemographic characteristics Table 1 presents the sociodemographic characteristics of the sample, categorized according to two IGD cutoff criteria. Most participants (73.5%, n = 523) reported studying exclusively, while 39.3% (n = 280) indicated that they currently play video games. Table 1Characteristics of the Mexican gamer subgroups based on IGD criteria (n = 280).Answers 4/5 as criterion metAnswers 3/4/5 as criterion metNon-IGD(n = 265)IGD (n = 15)χ2dfp-valueVNon-IGD(n = 236)IGD (n = 44)χ2dfp-valueVMage (SD)19.55(1.70)19.60(1.18)19.53(1.73)19.73(1.31)Gender (% female)23.033.30.7210.390.0522.529.52.0810.150.08Current act. (% study and work)19.626.70.9330.810.0518.229.53.6030.300.11Mode GamingAlone70.253.31.8910.160.0871.259.12.5510.110.09Family/Friends63.466.70.0610.790.0164.856.81.0210.310.06Acquaintances14.333.33.9410.040.1214.022.72.1810.140.08Strangers10.926.73.3710.060.1111.015.90.8510.360.05M(SD) H/w gaming5.31(4.01)6.67(4.45)5.22(4.09)6.33(3.64)M Mean, SD Standard deviation, V Cramer’s V, Current act. Current activity, df degrees of freedom, H/w hours per week. Table 1Characteristics of the Mexican gamer subgroups based on IGD criteria (n = 280).Answers 4/5 as criterion metAnswers 3/4/5 as criterion metNon-IGD(n = 265)IGD (n = 15)χ2dfp-valueVNon-IGD(n = 236)IGD (n = 44)χ2dfp-valueVMage (SD)19.55(1.70)19.60(1.18)19.53(1.73)19.73(1.31)Gender (% female)23.033.30.7210.390.0522.529.52.0810.150.08Current act. (% study and work)19.626.70.9330.810.0518.229.53.6030.300.11Mode GamingAlone70.253.31.8910.160.0871.259.12.5510.110.09Family/Friends63.466.70.0610.790.0164.856.81.0210.310.06Acquaintances14.333.33.9410.040.1214.022.72.1810.140.08Strangers10.926.73.3710.060.1111.015.90.8510.360.05M(SD) H/w gaming5.31(4.01)6.67(4.45)5.22(4.09)6.33(3.64)M Mean, SD Standard deviation, V Cramer’s V, Current act. Current activity, df degrees of freedom, H/w hours per week. Characteristics of the Mexican gamer subgroups based on IGD criteria (n = 280). 19.55 (1.70) 19.60 (1.18) 19.53 (1.73) 19.73 (1.31) 5.31 (4.01) 6.67 (4.45) 5.22 (4.09) 6.33 (3.64) M Mean, SD Standard deviation, V Cramer’s V, Current act. Current activity, df degrees of freedom, H/w hours per week. Estimated prevalence of internet gaming disorder Using the criterion of endorsing five or more items with response options 4 or 540, the estimated prevalence of IGD in the total sample was 2.1% (n = 15). When considering only current gamers (n = 280), the prevalence increased to 5.4%. Applying the less restrictive cutoff, in which endorsement of response options 3, 4, or 5 on at least five items was considered positive41, the estimated prevalence rose to 6.2% (n = 44) in the total sample and 15.7% among current gamers. Regarding estimated prevalence by sex, the prevalence was 1.6% for females and 2.5% for males. With the less restrictive criterion, the estimated prevalence was 4.3% and 6.5%, respectively. No significant differences in age, gender, or current activity were found between non-IGD and probable IGD gamers under either criterion (Table 1). However, a significant difference emerged for the variable playing with acquaintances under the more restrictive cutoff. Potential risk factors of internet gaming disorder Gaming-related factors Weekly gaming hours were positively associated with IGD severity (r =.168, p =.005), as were exposure to toxic gaming communities (r =.177, p =.003) and playing with acquaintances (r =.148, p =.013). Regarding gaming motivations, higher IGD scores were significantly associated with Violent Gratification (r =.134, p =.025) and Coping (r =.146, p =.014) motives, whereas Fun motives were negatively correlated with IGD (r = −.155, p =.009; Table 2). Table 2Correlations between IGD and gaming-related factors in Mexican gamers.HGTOXIMMCUSVIOCOPFUNCHALCOMPSOCIGDS r 0.1680.1770.086−0.0470.1340.146−0.1550.0670.0390.030 p-value 0.0050.0030.1490.4340.0250.0140.0090.2610.5170.614IGDS Internet Gaming Disorder Scale, HG Gaming hours/week, TOX Toxicity, IMM Immersion, CUST Customization, VIO Violent Gratification, COP Coping, FUN Fun, CHAL Cognitive Challenge, COMP Competition, SOC Social Interaction. Table 2Correlations between IGD and gaming-related factors in Mexican gamers.HGTOXIMMCUSVIOCOPFUNCHALCOMPSOCIGDS r 0.1680.1770.086−0.0470.1340.146−0.1550.0670.0390.030 p-value 0.0050.0030.1490.4340.0250.0140.0090.2610.5170.614IGDS Internet Gaming Disorder Scale, HG Gaming hours/week, TOX Toxicity, IMM Immersion, CUST Customization, VIO Violent Gratification, COP Coping, FUN Fun, CHAL Cognitive Challenge, COMP Competition, SOC Social Interaction. Correlations between IGD and gaming-related factors in Mexican gamers. IGDS Internet Gaming Disorder Scale, HG Gaming hours/week, TOX Toxicity, IMM Immersion, CUST Customization, VIO Violent Gratification, COP Coping, FUN Fun, CHAL Cognitive Challenge, COMP Competition, SOC Social Interaction. Psychological factors IGD severity was positively correlated with overall impulsivity (r =.256, p <.001). Among impulsivity dimensions, Negative urgency (r =.133, p =.028), Lack of premeditation (r =.231, p <.001), and Lack of perseverance (r =.333, p <.001) were statistically significantly correlated with higher IGD scores (Table 2). Potential protective factors of internet gaming disorder Psychological factors Lower IGD levels were significantly associated with higher perceived social support (r = −.188, p =.002) and higher emotional intelligence (r = −.239, p <.001). Specifically, perceived social support from family (r = −.183, p =.002), peers (r = −.171, p =.004), and significant others (r = −.139, p =.020) was inversely related to IGD. All four components of emotional intelligence were negatively correlated with IGD (SEA: r = −.194, p =.001; OEA: r = −.158, p =.008; EA: r = −.258, p <.001; ER: r = −.252, p <.001)(Table 3). Table 3Variable selection process: Measures of global adjustment of the models in Mexican gamers (n = 280).−2LLχ2dfR2 Cox y SnellR2 NagelkerkeAccuracyModel 0132.406––––94.9Model 1104.0156.531*10.0110.07194.9Model 299.0611.486*20.0030.12494.9df degrees of freedom of χ2. *p <.05.Model 0 =\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:logit\left(Non-IGD=1\right)={\beta\:}_{0}+\epsilon\:$$\end{document}.Model 1 =\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:logit\left(Non-IGD=1\right)={\beta\:}_{0}+{\beta\:}_{1}Lack\:of\:perseverance+\epsilon\:$$\end{document}.Model 2 =\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:logit\left(Non-IGD=1\right)={\beta\:}_{0}+{\beta\:}_{1}Lack\:of\:perseverance+{\beta\:}_{2}Social\:Support\:of\:Peers+\epsilon\:$$\end{document}. Table 3Variable selection process: Measures of global adjustment of the models in Mexican gamers (n = 280).−2LLχ2dfR2 Cox y SnellR2 NagelkerkeAccuracyModel 0132.406––––94.9Model 1104.0156.531*10.0110.07194.9Model 299.0611.486*20.0030.12494.9df degrees of freedom of χ2. *p <.05.Model 0 =\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:logit\left(Non-IGD=1\right)={\beta\:}_{0}+\epsilon\:$$\end{document}.Model 1 =\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:logit\left(Non-IGD=1\right)={\beta\:}_{0}+{\beta\:}_{1}Lack\:of\:perseverance+\epsilon\:$$\end{document}.Model 2 =\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:logit\left(Non-IGD=1\right)={\beta\:}_{0}+{\beta\:}_{1}Lack\:of\:perseverance+{\beta\:}_{2}Social\:Support\:of\:Peers+\epsilon\:$$\end{document}. Variable selection process: Measures of global adjustment of the models in Mexican gamers (n = 280). df degrees of freedom of χ2. *p <.05. Model 0 =\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:logit\left(Non-IGD=1\right)={\beta\:}_{0}+\epsilon\:$$\end{document}. Model 1 =\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:logit\left(Non-IGD=1\right)={\beta\:}_{0}+{\beta\:}_{1}Lack\:of\:perseverance+\epsilon\:$$\end{document}. Model 2 =\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:logit\left(Non-IGD=1\right)={\beta\:}_{0}+{\beta\:}_{1}Lack\:of\:perseverance+{\beta\:}_{2}Social\:Support\:of\:Peers+\epsilon\:$$\end{document}. Predictors of IGD Given the significant differences observed across IGD levels, a binomial logistic regression analysis was conducted using Forward Selection: LR was performed to examine the relationship between potential predictor variables and the likelihood of developing IGD. In the first step, Lack of perseverance (Model 1) was included, resulting in a statistically significant improvement in model fit compared with the null model (χ2(1) = 6.531, p =.011). In the second step, Perceived social support of peers (Model 2) was added, further improving the model fit relative to Model 1 (χ2(2) = 11.486, p =.003). This improvement was confirmed by the non-significant Hosmer–Lemeshow’s test, (Model 2: χ2(8) = 9.863, p =.275), indicating an adequate model fit. The Nagelkerke R2 coefficient showed that the predictors accounted for approximately 12.4% of the variance in IGD probability (Table 4). However, compared with Model 1, Model 2 did not increase classification accuracy, likely due to the imbalance between IGD and non-IGD cases. Although the model exhibited high specificity, its sensitivity was limited in identifying individuals meeting IGD criteria. Table 4Results of the final binary logistic regression model for predicting IGD in Mexican gamers (n = 280).βSEWalddfp-valueOR95% CI for ORLLULLack of perseverance0.2810.1155.93210.0151.3241.0561.660Peers social support−0.2290.1054.76110.0290.7950.6470.977Intercept−2.7301.5203.22310.0730.065β regression coefficient, SE Standard error, df degrees of freedom of the Wald test, p probability value associated with the Wald test, OR Odds Ratio, 95% CI 95% Confidence Interval for OR, LL Lower limit, UL Upper limit. Table 4Results of the final binary logistic regression model for predicting IGD in Mexican gamers (n = 280).βSEWalddfp-valueOR95% CI for ORLLULLack of perseverance0.2810.1155.93210.0151.3241.0561.660Peers social support−0.2290.1054.76110.0290.7950.6470.977Intercept−2.7301.5203.22310.0730.065β regression coefficient, SE Standard error, df degrees of freedom of the Wald test, p probability value associated with the Wald test, OR Odds Ratio, 95% CI 95% Confidence Interval for OR, LL Lower limit, UL Upper limit. Results of the final binary logistic regression model for predicting IGD in Mexican gamers (n = 280). β regression coefficient, SE Standard error, df degrees of freedom of the Wald test, p probability value associated with the Wald test, OR Odds Ratio, 95% CI 95% Confidence Interval for OR, LL Lower limit, UL Upper limit. For Lack of perseverance, the regression coefficient (β = 0.281) and odds ratio (OR = 1.324, 95% CI [1.056, 1.660]) indicate that each one-unit increase in this variable raised the odds of IGD by 32.4%. Conversely, for Perceived social support from peers, after controlling for Lack of perseverance, the regression coefficient (β = −0.229) and odds ratio (OR = 0.795, 95% CI [0.647, 0.977]) suggest that each one-unit increase in perceived social support from friends reduced the likelihood of IGD by 20.5%. As a robustness analysis, stratified bootstrap resampling was conducted to assess the stability of the logistic regression coefficients. The bootstrap estimates were consistent with the original model for Lack of perseverance, supporting the stability of its positive association with IGD. For perceived peer social support, the bootstrap estimate maintained the negative direction observed in the original model, although the confidence interval slightly included zero (see Supplementary Materials). Discussion This study estimated the prevalence of IGD among Mexican university students, examined its association with psychosocial and gaming-related factors, and identified variables that contribute to explaining IGD. Under the strict diagnostic criterion, the estimated prevalence of IGD was 2.1% among university students and 5.4% among current gamers. Both estimates are consistent with previously reported meta-analytic ranges for young adults (3.4%, 95% CI [1.9–5.9])53, as our results fall within the confidence interval. Previous Mexican research using the same DSM-5 cut-off of five or more symptoms reported a prevalence estimate of 5.2% (95% CI [4.7–5.8])18. Therefore, no statistically significant change in IGD prevalence can be inferred. While increases in IGD have been documented in other countries during the pandemic57, the present study cannot reflect temporal changes, as comparisons must rely on equivalent diagnostic criteria and populations. Meta-analytic estimates from Asia (10.3%, 95% CI [8.6–11.9]), Europe (9.2%, 95% CI [6.4–12.1]), and the Americas (11.5%, 95% CI [4.6–18.3])11 show broader variability across regions. Notably, the present estimate among current gamers (5.4%) also falls within the confidence interval reported for the Americas, indicating compatibility with the range observed in that region. Consistent with previous studies, our results showed that greater gaming time, exposure to toxic interactions, and playing mainly with acquaintances were all associated with higher IGD levels. Previous meta-analyses have consistently shown that greater gaming time is associated with elevated IGD symptoms23, 58. However, not only the quantity but also the quality of the gaming environment appears to play a critical role in IGD development. Gandolfi et al.59 demonstrated that exposure to toxic behaviors within gaming communities predicts higher IGD scores. Such toxic environments are often linked to poor social skills and frustration, traits that—when combined with competitiveness and intense social interaction—may reinforce excessive gaming behavior and increase IGD risk21, 59. Similarly, playing primarily with acquaintances may create greater social pressure or a stronger sense of obligation to perform well, compared with playing with friends (where trust mitigates competitiveness) or strangers (where expectations are minimal60. As a hypothesis-generating idea, these social dynamics could heighten players’ engagement and desire to continue gaming, potentially contributing to excessive play61. Motivational factors also play a critical role. Coping and escapism motives were positively associated with IGD, whereas gaming for fun or recreation showed a negative relationship. However, these findings should be interpreted with caution, as only three motives showed statistically significant correlations with IGD, but with small effect sizes. Prior research has identified coping and competition motives as key risk factors for IGD29, 33, 62. Network analyses suggest that escapism and competition motives are particularly central among players with IGD63. While recreational motives generally provide protection, they may lead to problematic gaming when combined with coping or avoidance strategies. Nonetheless, the current results suggest that such interpretations are still tentative and require further empirical research. Impulsivity emerged as another significant risk factor for IGD in this study. Among the impulsivity dimensions, those related to acting without forethought and difficulties sustaining effort were the most strongly associated with IGD in our research. Specifically, negative urgency, lack of premeditation, and lack of perseverance showed the clearest patterns, suggesting that students who struggle to manage emotional impulses, think ahead before acting, or maintain focus on effortful tasks may be more vulnerable to excessive gaming36. These results align with theoretical accounts proposing that impaired self-regulation and urgency-driven behaviors heighten vulnerability to problematic gaming. Elevated impulsivity reflects deficits in executive control and behavioral inhibition, both associated with addictive behaviors58. However, this relationship may also be bidirectional: individuals with high IGD scores often display greater impulsive decision-making64, suggesting that prolonged problematic gaming can further impair self-regulation. Consequently, individuals with heightened impulsivity may fail to control their gaming habits, which can be exacerbated by the reinforcing nature of IGD itself28. The results suggest that emotional intelligence may protect against IGD among Mexican university students. This pattern is reflected in the specific EI components examined in the study. All four subdimensions—self-emotion appraisal (SEA), others’ emotion appraisal (OEA), use of emotion (EA), and emotion regulation (ER)—showed small but statistically significant negative correlations with IGD. Previous studies have consistently identified an inverse association between emotional intelligence and IGD in the general adult population33, 34. Consistent with these findings, research has shown that young adults with IGD tend to score lower on measures of emotional intelligence compared to regular gamers and non-gamers65. Within IGD groups, individuals who experience emotional comorbidities such as anxiety, depression, or social anxiety often exhibit even lower levels of emotional intelligence. Difficulties in recognizing and managing emotions may lead to increased reliance on gaming as an unhealthy coping strategy, which aligns with the associations observed in this sample. Although the effect sizes were modest, these results reinforce the notion that emotional intelligence is negatively associated with IGD and may play a crucial role in understanding the emotional mechanisms underlying IGD. All dimensions of perceived social support were inversely associated with IGD. This pattern aligns with previous literature identifying perceived social support as a protective factor against IGD66, 67. Studies involving adolescents, emerging adults, and college students consistently report that individuals with IGD tend to perceive lower levels of social support68, 69 and often demonstrate weaker social networks and interpersonal skills70. Teng et al.69 argued that the interpersonal deterioration associated with IGD can further undermine psychosocial well-being. Moreover, Durak et al.71 highlighted that the relationship between perceived social support and IGD may vary depending on the source of support (online vs. offline) and the individual’s gaming experience. In sum, these findings reinforce the view that social support can buffer against IGD, though its protective efficacy likely depends on personal, social, or contextual factors. Finally, the logistic regression analysis identified lack of perseverance and perceived peer support as the most influential predictors of IGD in this sample of Mexican university students. Specifically, lower perseverance—reflecting difficulty maintaining sustained attention on extended tasks—was statistically significantly associated with a greater likelihood of developing IGD. This finding aligns with a systematic review by Dullur et al.72, which emphasized that attentional difficulties are strongly associated with IGD. Conversely, greater perceived social support from friends or peers was associated with a reduced risk of IGD. In the Mexican context, where peer relationships often constitute a key source of social integration and coping, strengthening peer-support networks may be an effective strategy for preventing problematic gaming behaviors. It should be noted that certain dimensions of emotional intelligence, such as emotional awareness and regulation, correlated strongly with IGD but were not included in the model because they shared variance with other predictors. Instead, peers’ social support was retained for its unique contribution, despite a lower correlation. Stepwise procedures may exclude variables whose predictive variance substantially overlaps with that of other predictors73. Moreover, multicollinearity diagnostics indicated that VIF and tolerance values were within acceptable ranges, maybe suggesting redundancy among predictors, but no severe multicollinearity that would compromise coefficient stability. Several context-specific patterns emerged that distinguish Mexican university gamers from those in European samples74. Mexican students consistently exhibited negative associations across all EI dimensions. These findings contrast with those observed among Italian students, where emotional intelligence showed no correlation with IGD, and differ from those among Spanish students, for whom only certain EI components were correlated with IGD. This suggests that emotional competencies may play a more central protective role in the Mexican context. Peer support also appeared to be a stronger protective factor in Mexico than in Spain or Italy, underscoring the importance of peer-based social integration in Mexican university life. Regarding impulsivity, Mexican students exhibited difficulties with perseverance and planning, a pattern partly similar to Spain but differing from Italy, where urgency was the primary correlate of IGD. In terms of social gaming, playing with acquaintances was relevant in Mexico, aligning with Spanish findings, but was less evident in Italy. In summary, although IGD prevalence levels were similar across Mexico, Spain, and Italy, the Mexican pattern appeared more strongly influenced by socioemotional vulnerability. These findings indicate several specific needs among Mexican university students: strengthening emotional regulation skills, reinforcing peer-support networks, and providing targeted strategies for managing stress and coping without relying on gaming. Additionally, difficulties in perseverance and planning underscore the need for interventions that support academic self-regulation. Together, these needs point to culturally informed priorities for prevention and early intervention in this population. Limitations of the study This study has several limitations. First, the sample consisted exclusively of Mexican university students recruited through convenience sampling at Tecnologico de Monterrey with middle- to upper-middle socioeconomic backgrounds, which limits representativeness. Although this approach facilitated data collection, the absence of random selection introduces potential selection bias, reducing the generalizability of the findings to the broader population of young adults in Mexico. Likewise, no data was collected on whether any of the participants were Spanish-speaking international students. Second, the cross-sectional design precludes causal inferences regarding the relationships among variables. Moreover, although some correlations were statistically significant, their effect sizes were small, so interpretations should be made cautiously, as statistical significance does not guarantee practical importance. Third, reliance on self-reported measures may have introduced social desirability bias, potentially affecting the accuracy of the results. Fourth, as toxicity was measured using a single-item self-report, the reliability and construct coverage are unknown, and it is not possible to rule out measurement error. Therefore, findings relating to toxicity should be considered exploratory. Fifth, the use of a forward stepwise approach introduces certain methodological limitations. This procedure may yield models sensitive to sample fluctuations, which can affect the stability and generalizability of the results. In addition, the initial bivariate prescreening used to reduce the set of candidate predictors may have excluded theoretically relevant variables that did not show significant zero-order associations but could become relevant after multivariable adjustment; therefore, the multivariable findings should be interpreted as exploratory and warrant replication/validation in independent samples. Although perceived peer social support was a statistically significant predictor in the original logistic regression model, the stratified bootstrap sensitivity analysis indicated that its confidence interval slightly included zero, suggesting sensitivity to sampling variability. Therefore, this finding should be interpreted with caution and requires replication in larger samples with a greater number of IGD cases. Sixth, the study focused solely on IGD and did not assess other technology-related behavioral addictions. Finally, although the instruments were translated and reviewed by native Spanish speakers using standard forward-backward translation procedures42, full psychometric validation within the Mexican university population remains needed. Future research should address several directions derived from the study’s limitations. Longitudinal designs are needed to clarify causal relationships and examine the development of IGD over time. Studies should also expand the scope beyond IGD to include broader technology-related behavioral addictions and continue validating the instruments within Mexican university populations to establish culturally appropriate cut-off scores. In addition, improving sampling procedures to enhance representativeness and reduce potential bias remains a critical methodological priority. Finally, future studies could build on these findings by using dimensional constructs within multiple linear regression models, which may provide more parsimonious results than data-driven stepwise methods. Beyond these directions, future research should examine psychosocial predictors of regular (non-pathological) gaming, explore potential gender differences in the associations identified, and test the moderating role of perceived social support—an approach highlighted in previous work39—to provide a more comprehensive understanding of gaming behavior and IGD. Practical implications of the study This study offers several practical implications for educational and psychological contexts in Mexican universities. When assessing IGD, cultural, psychological, and gaming-related variables must be considered to understand their relationship with IGD symptoms and their influence on problematic gaming behaviors. Such insights can guide the design of targeted intervention strategies that address the specific needs of Mexican university students. Insights from the findings suggest that prevention strategies should focus on online social environments and peer norms, especially given the links between increased gaming time, toxic gaming experiences, and higher IGD scores among students. Preventative initiatives, such as workshops or seminars, could promote healthy gaming habits and strengthen personal and social competencies. For example, gaming-focused modules might include strategies for managing gaming time, reinforcing fun and cooperative motives for play, and fostering positive peer relationships. Since coping and escapism were linked to IGD, interventions should focus on teaching adaptive emotion regulation and stress management strategies. Modules focused on personal development should prioritize self-awareness, emotional regulation, goal-setting, and impulsivity management. Given the links between IGD and impulsivity, programs enhancing planning and attentional control may benefit at-risk students. Social skills programs can promote empathy and collaboration by using recreational gaming to build connections. Furthermore, improving students’ emotional intelligence is crucial, as it serves as a protective measure against IGD. Finally, since peer support is associated with IGD risk, universities should consider implementing peer mentoring or social integration initiatives to strengthen supportive networks. Although the overall prevalence of IGD appears relatively low, estimates vary depending on the diagnostic threshold applied. Therefore, it is crucial to develop culturally and contextually tailored prevention and intervention programs that address both protective and risk factors associated with IGD, ensuring the psychological well-being of Mexican university students. These programs should address both protective and risk factors to support psychological well-being, based on empirical patterns observed in this study among Mexican university students. Conclusions This study highlights the need to address IGD among Mexican university students, with an estimated prevalence of 2.1% in this sample. Significant risk factors included the amount of time spent gaming, interactions with acquaintances, competitive and coping motivations, exposure to toxic gaming communities, and higher impulsivity. In contrast, gaming for fun, greater emotional intelligence, and stronger perceived social support emerged as protective factors. To prevent IGD within this population, fostering peer relationships and developing interventions that strengthen perseverance and emotional regulation may be particularly effective. These findings underscore the importance of implementing emotional education initiatives and campaigns to promote responsible, balanced gaming practices in university settings. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1