Psychological Correlates of Problematic Gaming Behaviour in Romanian Medical Students: a Cross-Sectional Analysis of Structural and Contextual Factors

1. Introduction

Videogaming, defined as “the experiential and interactive engagement facilitated by electronic games” [1], has become one of the most prominent forms of leisure and digital entertainment among younger populations worldwide [2]. Beyond its recreational dimension, videogaming represents a complex sociocultural and technological phenomenon that integrates cognitive stimulation, emotional involvement, and interpersonal interaction within digitally mediated environments. Contemporary video games encompass a broad spectrum of formats and genres, including single-player, cooperative, and competitive multiplayer modalities, each characterized by distinct patterns of participation, communication, and social engagement. These gaming environments often promote immersive experiences through reward systems, narrative structures, real-time interaction, and opportunities for social connectedness, thereby increasing their appeal among adolescents and young adults. As a result, the global videogame industry has experienced substantial growth in recent decades, further accelerated by the COVID-19 pandemic—during which social isolation and restricted mobility increased engagement with digital entertainment—, with video gaming also serving as a means of facilitating social connectedness, managing stress, and reducing experiences of loneliness [3]. Current projections suggest further expansion of the industry, driven by rapid technological innovation and the integration of videogaming into broader educational, professional, and social domains. In particular, advances in virtual and augmented reality technologies, the increasing use of simulation-based training in medical and healthcare education, and the expanding global influence of electronic sports (e-sports) are expected to further consolidate the role of videogaming in contemporary society [4].

In terms of public perception, videogaming is increasingly conceptualized not only as entertainment, but also as a multifaceted digital activity, with significant psychological, social, educational, and economic implications, a view supported by a growing body of evidence documenting its potential benefits. Reported positive outcomes include enhanced creativity, neuroplasticity, and specific memory functions, as well as improvements in problem-solving capacities and spatial cognition [5]. Additional research has identified beneficial effects on mood and emotional regulation [6], increased levels of physical activity [7], and greater self-efficacy and social competence. Furthermore, videogaming has been associated with the facilitation of social networking and the development of interpersonal relationships [8,9].

Nevertheless, despite these undeniable advantages, concerns persist regarding the potential adverse consequences of excessive video game engagement. In particular, children, adolescents, and young adults are considered especially vulnerable to the development of problematic gaming behaviour (PGB) and video game addiction. In particular, online gaming can lead to a pattern of problematic or compulsive behaviour. Recognizing the potential risks associated with excessive gaming, Diagnostic and Statistical Manual of Mental Disorders (5th Edition) included Internet Gaming Disorder (IGD) in Section III as a condition warranting further study [10,11]. Subsequently, IGD was formally recognized as a diagnosable condition in the International Classification of Diseases (11^th Revision) [12]. Both diagnostic frameworks emphasize the importance of distinguishing normative gaming behaviour from maladaptive or pathological patterns of use. They also highlight the relevance of therapeutic interventions for excessive gaming behaviours, while underscoring the necessity of differentiating recreational engagement from problematic use.

Recent estimates suggest that more than two billion individuals worldwide engage in video gaming [13,14]. Among these, approximately 60 million individuals are thought to meet criteria associated with video game addiction [15,16], the majority of whom are adolescents and young adults. Current evidence indicates that the prevalence of IGD among young people ranges between 6.7% and 9.9% [17,18]. However, substantial regional variability has been reported, with estimates ranging from 1.2% to 15%, generally lower in Western countries and higher in Asian populations [19,20,21,22]. These differences may be partially explained by variations in diagnostic criteria and assessment methods, as well as by the increasing global accessibility of the internet and the broad adoption of smartphones and tablet devices [18].

In the academic environment, a previous meta-analysis reported a pooled prevalence of internet addiction of 30.1%, representing nearly a fivefold increase relative to the general population, while the pooled prevalence of IGD specifically was estimated at 6.2%, approximately twice that observed in the broader population [23]. This latter finding is particularly noteworthy, given that IGD diagnosis requires the endorsement of at least five DSM-5 criteria, suggesting that subclinical gaming-related difficulties may be considerably more prevalent in this particular category.

1.1. Factors Associated with Videogaming

1.1.1. Structural Factors

Structural factors associated with IGD refer to relatively stable personal and / or biographical characteristics that shape vulnerability to problematic gaming.

(a) Demographic variables

Young adult males appear to demonstrate a greater propensity toward excessive video game use compared with their female counterparts [24]. This pattern has been consistently observed across diverse populations of young adults in different academic settings. For instance, a European study involving students aged 18–21 years identified male gender as a significant structural risk factor, with this association being mediated by a predisposition toward dissociative experiences [25]. Similar findings have been reported among male medical students in India, the Arab region, and Egypt [26,27,28]. One proposed explanation is that the gaming industry’s historical emphasis on targeting male audiences may contribute to men’s increased vulnerability to developing IGD [29].

Individuals who engage most extensively in gaming tend to be younger, possess lower levels of educational attainment, and are less likely to be employed full-time. Students, in particular, represent a population at elevated risk due to the relative flexibility of their schedules [30]. Among medical students specifically, a weekly gaming duration exceeding 20 hours has emerged as a prominent predictor of PGB, with gender and psychopathology acting as moderating variables [31]. Furthermore, studies have identified a declining prevalence gradient across academic years in both general student populations [32] and medical student cohorts [23], with first-year students exhibiting the highest prevalence rates.

(b) Personality characteristics

They have been consistently associated with an increased risk of problematic gaming. These include introversion, hostility, impulsivity, higher neuroticism and low conscientiousness, whereas traits such as extraversion, openness, agreeableness, conscientiousness, self-control, self-esteem, life satisfaction, and higher academic achievement appear to exert protective effects [33,34,35,36]. Within the framework of the Big Five personality model, distinct gamer profiles have been identified, ranging from socially oriented “seekers” to achievement-driven “achievers,” each characterized by specific personality trait configurations [37,38]. Lower levels of conscientiousness were identified in students with a higher number of hours dedicated to video games (7 or more hours per week). Regarding openness, higher scores were found in students who played a greater variety of games, and positive relationships were found between openness and game genre preferences (e.g., role-playing games, puzzle games, strategy games, platformers, action-adventure games, and simulation games) [39].

In medical students, studies examining personality profiles indicate that higher levels of neuroticism positively correlate with video game addiction, whereas agreeableness, extraversion, and conscientiousness demonstrate a negative correlation [33,40]. Notably, conscientiousness is the sole personality trait that exerts a statistically significant protective function against the development of gaming pathology [33,41].

1.1.2. Contextual Factors Associated with Videogaming

Contextual risk factors associated with problematic gaming among university students and young adults have been extensively documented across diverse cultural settings. Longitudinal findings identified both risk and protective factors, thereby supporting the conceptualization of IGD as a dynamic and potentially modifiable condition across the lifespan.

(a) Psychiatric comorbidity has important connections to gaming behaviour, with evidence suggesting complex and frequently bidirectional relationships (Wu et al., 2018). Early investigations of internet addiction among college students identified depression, attention-deficit/hyperactivity disorder (ADHD), social phobia, and hostility as significant correlates [23,42].

Although problematic gaming has been associated with anxiety symptoms [43], evidence regarding anxiety as a direct etiological factor in gaming addiction remains limited. In fact, moderate levels of gaming have been shown to alleviate anxiety, facilitate social connectedness, and improve mood [43], particularly among adolescents and socially withdrawn individuals [6]. Similarly, depressive symptomatology has been associated with both increased gaming engagement and adverse psychosocial outcomes [35,44]. Nevertheless, certain forms of gaming, particularly physically interactive “exergames,” have demonstrated beneficial effects in reducing depressive symptoms [6,7,45].

These relationships appear especially interesting in the context of medical education. Compared with students enrolled in other academic programs, medical students experience longer periods of study and substantially greater academic and emotional demands. Such pressures may precipitate or exacerbate psychopathological conditions, including depression, anxiety, and burnout, potentially contributing to the elevated prevalence of addictive behaviours observed in this population. A previous meta-analysis reported a pooled prevalence of internet addiction among medical students of approximately 30.1%, representing nearly a fivefold increase relative to the general population, while the pooled prevalence of IGD specifically was estimated at 6.2%, approximately twice that observed in the broader population [23]. This latter finding is particularly noteworthy, given that IGD diagnosis requires the endorsement of at least five DSM-5 criteria, suggesting that subclinical gaming-related difficulties may be considerably more prevalent. Supporting this interpretation, a recent cross-sectional study employing the IGDS9-SF among Indian undergraduate medical students found that IGD was significantly associated with insomnia, poor academic performance, and aggressive behaviour, thereby underscoring the clinical importance of screening for PGB [46]. Furthermore, a large-scale Indonesian study identified weekly gaming duration, male gender, and membership in gaming communities as primary contextual predictors of IGD among medical students, with psychopathological variables functioning as moderators of these relationships [31].

(b) Social and interpersonal factors

The association between social and interpersonal factors and video gaming is multifaceted, given that video gaming may contribute to both beneficial outcomes [47] and adverse consequences for social relationships [48,49]. The global proliferation of video gaming has changed contemporary social engagement patterns across diverse demographics: excessive gaming behaviour may precipitate a decline in interpersonal connectivity, manifest as diminished participation in familial, peer-based, and community activities [49].

Much less studied is how social variables (family structure and group dynamics, including social support and prosocial behaviour within them) can influence video gaming behaviour.

(b1) Family structure

Family represents a fundamental contextual determinant of gaming behaviour and psychological well-being, exerting a substantial influence on both vulnerability to and protection from problematic gaming. Family systems characterized by inconsistent parenting practices, limited supervision, and low emotional responsiveness are associated with a higher risk of video game addiction among adolescents and young adults [50]. More broadly, dysfunctional family structures often show reduced adaptability in reorganizing relational roles, boundaries, and interactional patterns in response to developmental changes, a rigidity that has been linked to a range of addictive behaviours [51]. Key structural disruptions, including parental divorce, socioeconomic disadvantage, and parental mental health difficulties, may further impair developmental adjustment and increase reliance on gaming as a maladaptive coping strategy, while excessive gaming can in turn intensify familial strain and reinforce a cycle of relational disengagement. Consistent with this, long-term interpersonal conflicts within the family have been associated with greater self-isolation and gaming involvement [52], while the family’s low socioeconomic status has been linked to increase in gaming time [53].

Oppositely, well-functioning family structures provide an important protective buffer against problematic gaming. High-quality parent–young adult relationships, characterized by emotional support, effective communication, and appropriate monitoring are particularly protective during transitions to university, when parental oversight typically decreases [35]. Population-based findings similarly indicate that perceived social support, parent–child relationship quality, and family connectedness are inversely associated with IGD severity [54]. Parental monitoring has been identified as a key protective factor across genders [55], with reduced oversight during entry into higher education proposed as a contributor to increased IGD prevalence in early university years. In contrast, overly restrictive or punitive parenting strategies—such as rigid rule enforcement and strict limitation of access—do not appear effective in reducing problematic gaming and may, in some cases, exacerbate symptoms, particularly among male students [56].

(b2) Peer-group dynamics have also been shown to influence PGB. Membership in peer groups characterized by maladaptive social influences has been associated with increased gaming time, loneliness, social media addiction, aggressive behaviour, anxiety, ADHD symptoms, depression, sensation seeking, social vulnerability, and impulsivity [35]. Although in-game communication may facilitate the formation of meaningful social connections through processes, such as self-disclosure and reciprocity [57], these online relationships often lack the emotional depth and stability associated with offline interpersonal relationships and may displace real-world social interaction [58,59]. Moreover, lower levels of perceived social support have consistently been identified as predictors of IGD severity [58,60].

(b3) Finally, social support may serve as a mediator that mitigates the impact of stress and anxiety on IGD [61]. An important aspect revealed by previous research is that offline social support, including interactions with family and peers, has a stronger protective effect than online social interactions, and has the effect of reducing anxiety, stress, and depressive symptoms in students with IGD [62].

In academic populations, contextual factors (satisfaction with relationships with family and colleagues, involvement in social activities) are variables that affect mental health [63,64] and may increase vulnerability to IGD [65,66].

Taken together, the aforementioned findings delineate a consistent profile of risk factors associated with PGB; however, while these patterns provide a useful framework for understanding the phenomenon, evidence from Eastern European populations—and particularly Romania—remains limited, underscoring the need for further investigation. Recent Romanian data indicate substantial engagement in video gaming, with approximately 7.9 million individuals aged 15–64 participating and an estimated 11.4% prevalence of video game addiction among adolescents, particularly in specific educational subgroups [67,68]. However, existing evidence remains largely confined to prevalence estimates, with limited insight into the psychological determinants, risk mechanisms, and broader contextual correlates of IGD. In academic contexts, despite the growing scholarly interest in preserving students’ health and well-being, empirical research specifically examining PGB remains scarce. This gap in the literature constitutes a primary rationale for the present study.Top of Form

1.2. Objective and Hypotheses

The present study aimed to investigate the relationships between IGD, structural factors (demographic characteristics, personality traits) and contextual psychosocial factors (psychiatric comorbidity - depression and anxiety symptoms, social and interpersonal factors - family structure and social support, and prosocial behaviour) in Romanian medical students. The study was conceptually guided by the IGD framework, but the outcome is referred to PGB, because participants were assessed using a self-report screening instrument rather than a structured professional diagnostic interview.

The investigated hypotheses were:

H1. PGB is significantly predicted by a combination of demographic characteristics, gaming-related variables, personality traits, psychological distress, perceived social support, and prosocial behaviour:

H1.a. Structural factors

Demographic characteristics are expected to show associations with PGB, although their independent contribution may be limited after accounting for personality and gaming-related variables.

H1.b. Gaming-related contextual factors

Higher daily gaming time and preferred game type, particularly competitive gaming, are expected to be significant predictors of PGB.

H1.c. Personality traits

Personality traits are expected to contribute to PGB, with lower conscientiousness predicting higher PGB.

H1.d. Perceived social support and prosocial behaviour

Higher perceived social support and higher prosocial behaviour are expected to be negatively associated with PGB.

H2. Psychological distress and PGB

Depressive and anxiety symptoms are expected to be positively associated with PGB and to contribute to gaming-associated symptoms severity.

2. Materials and Methods

2.1. Design

The design of the study was cross-sectional, with a single administration of a series of standardized psychometric instruments.

2.2. Participants

Participants were undergraduate students, both Romanian and international, undergoing their studies and training at the “Carol Davila” University of Medicine and Pharmacy Bucharest (CDUMP), the largest and one of the most prestigious medical schools of Romania. Inclusion criteria were: being at least 18 years old, having the status of current undergraduate students in the abovementioned institution, having agreed to the informed consent form, having completed all study instruments. The exclusion criteria were incomplete or inappropriate answers to the questionnaires, being underage, not having given their consent and participants who specified that they do not play video games.

2.3. Procedure

Data collection was carried out between December 2023 and December 2025 by using a convenience sampling strategy and administering an online questionnaire (Google Form) distributed across students’ social media platforms, specifically Facebook groups and institutional email addresses.

Before taking part in this research, all participants received a brief explanatory statement about the study and completed informed consent forms. The study was run in accordance with the World Medical Association Declaration of Helskinki and was approved by the CDUMP Institutional Review Board (no. 20738 / 21.07.2023). The anonymity was guaranteed and no email addresses or other identifying data were collected. A researcher (M.D.) was available by email, in case there were questions or any type of disagreement regarding the process of filling in the questionnaires and for those participants intending to withdraw from the study. Final results were included in an IBM-SPSS 26^® database.

2.4. Instruments

After providing informed consent, all participants completed a set of questions collecting demographic information, including age, year of study, gender, area of origin, family structure, marital status, engagement in professional activities in addition to studying, and, where applicable, the type of employment contract associated with their job. The next section referred to their gaming habits, specifically to the age of starting gaming activity, favourite devices (smartphone, personal computer, laptop, tablet, video game console), type of internet connection (Integrated Services Digital Network, satellite, Digital Subscriber Line, cable, dial-up, broadband, mobile, wireless, hotspot), daily average time spent in gaming (less than one hour, 2-3 hours, 4-5 hours, more than 6 hours), preference regarding the time of week to play (during the weekend, during the week or the same), preferred type of game (competitive, cooperative, individual-casual), average amount of money (RON) spent monthly on games in the last year, perceived social support from family and friends (on a scale from 1 to 10), and a subjective evaluation about the existence of a person they can rely on, regardless of the moment or problem (expressed as a dichotomic answer yes/no).

Following the collection of demographic information, participants completed a battery of standardized psychological instruments selected on the basis of previous empirical evidence supporting their psychometric adequacy and relevance to the study objectives.

2.4.1. Big Five Inventory–2 Extra-Short Form (BFI-2-XS)

The Big Five Inventory–2 Extra-Short Form (BFI-2-XS) [69] was used to assess the five major personality dimensions: extraversion, agreeableness, conscientiousness, negative emotionality, and openness to experience. The instrument consists of 15 items rated on a 5-point Likert scale ranging from 1 (“strongly disagree”) to 5 (“strongly agree”). Previous studies have reported satisfactory internal consistency, with Cronbach’s alpha coefficients ranging from 0.74 to 0.92 [70].

2.4.2. Patient Health Questionnaire-9 (PHQ-9)

Depressive symptoms were assessed using the Patient Health Questionnaire-9 (PHQ-9) [71], a self-report measure corresponding to the nine diagnostic criteria for major depressive disorder outlined in the DSM-5 [10]. Participants rated the frequency of each symptom on a 4-point Likert scale ranging from 0 (“not at all”) to 3 (“almost every day”). The PHQ-9 represents the self-administered version of the PRIME-MD diagnostic instrument for common mental disorders and has demonstrated excellent internal reliability, with a reported Cronbach’s alpha of .89 [72].

2.4.3. Generalized Anxiety Disorder-7 Scale (GAD-7)

Anxiety symptoms were measured using the Generalized Anxiety Disorder-7 Scale (GAD-7) [73], a self-report instrument designed to screen for probable cases of generalized anxiety disorder and evaluate symptom severity. The scale comprises seven items rated on a 4-point Likert scale ranging from 0 (“not at all”) to 3 (“almost every day”). The GAD-7 has shown good internal consistency, with Cronbach’s alpha coefficients exceeding 0.82 [74].

2.4.4. Multidimensional Scale of Perceived Social Support (MSPSS)

Perceived social support was assessed using the Multidimensional Scale of Perceived Social Support (MSPSS) [75]. The MSPSS consists of 12 items measuring perceived support from three sources: family, friends, and a significant other. Responses are provided on a 7-point Likert scale ranging from 1 (“strongly disagree”) to 7 (“strongly agree”). The instrument has demonstrated strong psychometric properties, including good validity and internal consistency, with Cronbach’s alpha coefficients ranging from 0.88 to 0.97 [76,77].

2.4.5. Prosocialness Scale for Adults

Prosocial behaviour was evaluated using the Prosocialness Scale for Adults [78]. This 16-item instrument assesses prosocial tendencies across domains such as helping, sharing, caring, and empathic responsiveness to others’ needs and requests. Items are rated on a 5-point Likert scale ranging from 1 (“never/almost never true”) to 5 (“almost always/always true”). Previous research has reported excellent internal consistency, with Cronbach’s alpha values ranging from 0.903 [79] to 0.932 [80].

2.4.6. Internet Gaming Disorder Scale–Short Form (IGDS9-SF)

PGB was assessed using the Internet Gaming Disorder Scale–Short Form (IGDS9-SF) [81]. The scale is based on the nine diagnostic criteria for IGD included in the DSM-5 [10] and is designed to evaluate the severity of gaming-related symptoms and their associated negative consequences. The instrument comprises nine items rated on a 5-point Likert scale ranging from 1 (“never”) to 5 (“very often”), referring to gaming activities occurring over the preceding 12 months. The IGDS9-SF has demonstrated good reliability, with reported Cronbach’s alpha values of approximately 0.87 [82].

Although higher IGDS9-SF cut-off scores have been proposed for the identification of probable clinical cases, lower thresholds are commonly employed in non-clinical and exploratory research to identify individuals exhibiting elevated levels of gaming-related symptoms or PGB. This approach is consistent with the broader literature, in which problematic gaming is frequently operationalized through questionnaire-based thresholds, although the specific cut-off values vary across instruments and populations [83,84,85].

In the present study, a cut-off score of 24 was adopted to classify participants as exhibiting PGB. This threshold was selected based on both conceptual and empirical considerations. Conceptually, a score of 24 reflects moderate-to-frequent endorsement of IGD criteria, without necessarily indicating the presence of severe or clinically significant impairment. Empirically, this score corresponded to values above the 90^th percentile of the sample distribution, whereas a score of 18 represented the 75^th percentile. The selected threshold ensured an adequate subgroup size for statistical analyses, while preserving sensitivity and statistical power. Consequently, findings derived from this classification should be interpreted as reflecting PGB or elevated IGD symptomatology rather than a formal clinical diagnosis of IGD.

2.5. Data Analysis

The analysis included firstly a descriptive level, with calculation of minimum, maximum, mean, standard error and standard deviation of numerical variables (e.g., age, age of starting gaming activity). Frequencies were also determined for the nominal variables (e.g., year of study, gender, country of origin).

Subsequently, multiple variable association tests (Pearson chi-square, t-tests for independent samples, Sperman correlations and ANOVA) were carried out between demographic data, gaming habits and the psychometric instruments results, while also comparing Romanian and international students.

Hierarchical multiple regression analyses were conducted to examine the contribution of structural factors (demographic variables and personality traits) and contextual factors (gaming-related behaviours) to psychological outcomes. In the main model variables were entered in blocks as follows: (1) demographic variables, (2) personality traits, (3) gaming-related variables, and (4) perceived social support. An additional regression model was run to assess contribution of depressive symptoms (PHQ-9), anxiety symptoms (GAD-7), perceived social support and prosocial behaviour to PGB. Standardized regression coefficients (β) and R² were reported to evaluate the incremental contribution of each block.

For all calculations, the threshold of statistical significance was p < 0.05.

3. Results

A total of 261 students consented to participate in the study. Of these, six participants were excluded because they were under the age of 18, and one participant was excluded due to missing age information. Additionally, 11 respondents were not included in the final sample, because they reported not engaging in video gaming. Consequently, the final analytical sample comprised 243 participants.

3.1. Demographic Characteristics

The mean age of the participants was 20.53 years (SD = 2.23; range = 18–36). Participants reported initiating video gaming at a mean age of 10.61 years (SD = 3.40; range = 3–24), suggesting early exposure to gaming activities. Among the 204 participants who reported spending money on video games, the average monthly expenditure during the previous year was 65.52 RON (SD = 283.99; range = 0–3000). The substantial standard deviation indicates considerable variability in spending patterns across participants.

With regard to sociodemographic characteristics, the sample was predominantly female (58.0%), primarily composed of individuals residing in urban areas (86.1%), and largely consisted of Romanian students (72.0%), with international students accounting for 28.0% of the sample. Most participants reported having siblings (69.1%) and were not engaged in paid employment at the time of data collection (94.6%).

Regarding relationship status, the majority of participants identified as single (61.7%). Smaller proportions reported being in a non-cohabiting romantic relationship (27.2%), cohabiting with a partner (8.6%), or being married (2.5%).

The descriptive characteristics of all continuous variables are reported in Table 1.

3.2. Gaming Behaviour

Gaming behaviour was generally characterized by low-to-moderate daily engagement. Most participants reported gaming for less than 1 hour/day (59.3%), followed by 2–3 hours/day (34.2%). Only a minority reported gaming for 4–5 hours/day (4.9%) or more than 6 hours/day (1.6%). The majority of participants reported gaming during weekends (77.4%), whereas weekday-dominant gaming was rare (4.9%) (Table 2).

With regard to gaming preferences, individual/casual games (43.2%) and competitive games (41.6%) were most commonly reported, whereas cooperative games (15.2%) were least preferred. A descriptive gender pattern was also observed, with male participants showing a greater preference for competitive gaming, whereas female participants more frequently were engaged in individual/casual games.

In terms of devices, gaming was predominantly mobile-based. The smartphone was the most frequently used gaming device (72.4%), followed by the PC (52.7%), laptop (51.9%), console (42.0%), and tablet (35.0%). The most frequently reported connection type was wireless internet (77.4%), followed by mobile data (38.3%) and cable (37.9%), whereas other types of connection were rare. Connectivity relied mainly on wireless internet, reflecting high digital accessibility.

The mean IGDS9-SF score was 14.99 (SD = 6.03; range: 9–38). Using a cut-off score of 24 on the IGDS9-SF, 25 participants (10.28%) were classified as displaying PGB.

3.3. Psychological and Social Variables

Participants reported relatively high perceived support from both family and friends. On a 1–10 rating scale, the mean perceived family support was 8.40, while the mean perceived support from friends was 8.02. In addition, 96.7% of participants reported that they had at least one person they could rely on regardless of the situation.

Most participants reported high levels of perceived social support across different areas. Specifically, 70.8% indicated high partner support, 70.8% reported high family support, 62.6% reported high friend support, and 71.2% reported high overall support. Mean scores on the MSPSS dimensions were 5.63 (SD=1.66) for partner support, 5.59 (SD=1.5) for family support, 5.35 (SD=1.43) for friend support, and 5.52 (SD=1.32) for total support.

Regarding personality traits, the highest average score was for openness to experience (M = 10.87, SD = 2.25), followed by conscientiousness (M = 10.18, SD = 2.07) and agreeableness (M = 10.10, SD = 2.14). In contrast, lower average values were found for emotional stability (M = 9.45, SD = 2.60) and extraversion (M = 9.27, SD = 2.44). The mean total BFI-2-XS score was 48.96 (SD = 5.70).

Regarding depressive symptoms, the mean PHQ-9 score was 9.13 (SD = 5.70). Although 24.3% of participants reported no depressive symptoms and 33.3% were in the subthreshold range, 25.5% presented moderate to moderately severe symptoms and 16.9% met the threshold for major depression. With respect to functional impairment associated with depressive symptoms, 56.0% reported that these symptoms made work, household tasks, or social interactions “somewhat difficult”, while 15.7% reported severe or extreme impairment.

The mean GAD-7 score was 8.19 (SD = 5.56). Mild anxiety was the most frequent category (35.8%), followed by moderate (30.0%), moderately severe (22.6%), and severe anxiety (11.5%).

The mean score for prosocial behaviour was 60.79 (SD = 13.64), suggesting an overall moderate-to-high level of self-reported prosocial tendencies in the sample.

3.4. Factors Associated with PGB

Chi-square analyses were performed to examine associations between PGB and selected categorical variables. Significant associations were identified for family structure (χ² = 4.647, p = 0.031), daily gaming time (χ² = 15.283, p = 0.020), and preferred game type (χ² = 10.897, p = 0.040). Participants with siblings, those reporting longer daily gaming time, and those preferring competitive games showed higher frequencies of problematic gaming (Table 3).

A marginal association was also observed for student origin (χ² = 7.976, p = 0.050). Specifically, international students exhibited a greater prevalence of problematic gaming compared to their Romanian counterparts. No statistically significant relationships were identified concerning year of study, gender, area of residence, marital status, or preferred gaming time. Further chi-square analyses, which explored the psychological and psychosocial correlates of problematic gaming, revealed significant associations with monthly gaming expenditure (χ² = 46.368, p = 0.016), conscientiousness (χ² = 25.857, p = 0.040), depressive symptoms (PHQ-9) (χ² = 39.170, p = 0.047), partner support (χ² = 37.855, p = 0.036), total perceived social support (χ² = 118.659, p = 0.020), and prosocial behaviour (χ² = 70.006, p = 0.032).

Associations with family support and friend support approached, but did not reach conventional statistical significance. No significant association was observed between PGB and extraversion, agreeableness, emotional stability, openness to experience, total BFI-2-XS score, or anxiety symptoms.

In summary, the descriptive and inferential results show that while gaming was mostly moderate in this population, a clinically significant subset had problematic gaming activity. These behaviours were linked to a mix of behavioural, structural, and psychological factors.

3.5. Hierarchical Regression Analysis

3.5.1. Predictors of PGB

A hierarchical multiple regression was conducted to examine the contribution of structural factors, gaming-related variables, and perceived social support to PGB, operationalized as IGDS9-SF total score. Demographic variables were entered in Block 1, personality traits in Block 2, gaming-related variables in Block 3, perceived social support (total score) in Block 4 and prosocial behaviour in Block 5.

The final model was statistically significant, explaining 28.9% of the variance in IGDS9-SF scores (R² = 0.289, F(20,131) = 2.657, p < 0.001). Demographic variables entered in Block 1 did not significantly predict PGB, ΔR² = 0.062, p = 0.227. The addition of personality traits in Block 2 resulted in a modest but non-significant increase in explained variance, ΔR² = 0.054, p = 0.139. However, the inclusion of gaming-related variables in Block 3 significantly improved the model, ΔR² = 0.155, p < 0.001, representing the largest incremental contribution to IGDS9-SF scores.

The addition of total perceived social support in Block 4 produced a small, marginal increase in explained variance, ΔR² = 0.017, p = 0.078, but did not reach conventional statistical significance. Finally, the addition of prosocial behaviour in Block 5 did not improve the model, ΔR² < 0.001, p = 0.792.

In the final model, higher daily gaming time remained a significant positive predictor of PGB, β = 0.292, p = 0.001. Preferred game type was also a significant predictor, β = −0.232, p = 0.006. Given the coding of this variable, this result indicates that lower scores on the game-type variable, corresponding to competitive gaming, were associated with higher IGDS9-SF scores. Conscientiousness was a significant negative predictor, β = −0.201, p = 0.017, suggesting that lower levels of self-regulation, organization, and persistence were associated with higher PGB. Total perceived social support showed a negative but non-significant association with IGDS9-SF scores, β = −0.158, p = 0.142, and prosocial behaviour was not a significant predictor, β = −0.025, p = 0.792 (Table 4).

Overall, PGB was primarily associated with behavioural engagement, with a smaller contribution from personality traits and no independent effect of perceived social support or prosocial behaviour.

3.5.2. Additional Psychological–Social Model

Because depressive and anxiety symptoms were theoretically relevant but were not included in the main gaming-behaviour model, an additional sensitivity analysis was conducted to examine their contribution to IGDS9-SF scores alongside personality, social support, and prosocial behaviour. The results of the sensitivity hierarchical regression model are summarized in Supplementary Table S1.

In this regression model, IGDS9-SF total score remained the dependent variable. Demographic variables were entered in Block 1, Big Five personality traits in Block 2, depressive and anxiety symptoms in Block 3, perceived social support in Block 4, and prosocial behaviour in Block 5. The country variable was excluded from the analysis because it was constant in the analysed subsample.

The final model was statistically significant, F(18, 146) = 2.097, p = 0.009, explaining 20.5% of the variance in IGDS9-SF scores, although the adjusted R² was modest, at 0.107. Demographic variables did not significantly contribute to the model, ΔR² = 0.063, p = 0.165, and the addition of personality traits showed only a marginal contribution, ΔR² = 0.060, p = 0.069. The inclusion of depressive and anxiety symptoms in Block 3 significantly improved the model, ΔR² = 0.069, p = 0.002. However, perceived social support, ΔR² = 0.012, p = 0.545, and prosocial behaviour, ΔR² = 0.001, p = 0.616, did not add significant incremental explanatory value. In the final model, Conscientiousness was the only significant independent predictor, β = −0.207, p = 0.013, indicating that lower conscientiousness was associated with higher IGDS9-SF scores.

These findings support the relevance of psychological distress and conscientiousness in relation to problematic gaming symptom severity, while suggesting that perceived social support and prosocial behaviour did not independently explain additional variance after controlling for demographic, personality, and distress-related variables.

4. Discussion

The current study offers an in-depth investigation of video gaming activity and its psychological associations among young people, encompassing both native Romanian and international students. The findings partially supported the proposed hypotheses, suggesting a differentiated and multidimensional framework integrating behavioural, personality, and social factors.

Overall, gaming behaviour in this sample was generally characterized by low-to-moderate engagement, with most participants reporting less than one hour of daily gaming. Notwithstanding this comparatively restricted exposure at the group level, a clinically significant subset (10.28%) demonstrated PGB, reinforcing the idea that maladaptive gaming behaviours might arise even in groups with generally moderate usage.

These findings align closely with prior study conducted among Romanian student populations. Rusu et al. indicated that personality factors are variably linked to gaming motivations and preferences [86]. Individuals exhibiting lower levels of openness and conscientiousness were more inclined to engage in gaming as a means of alleviating boredom, while those with strong extraversion indicated social motivations for gaming. Moreover, personality factors were associated with certain gaming preferences, with males favouring competitive genres and females opting for cognitively engaging or narrative-driven games.

The present study contributes to the existing body of research by demonstrating that, while personality characteristics are associated with gaming preferences and motivations, behavioural engagement, including time spent gaming and a preference for competitive games, is a more significant factor in the development of PGB. This suggests that personality may influence how individuals interact with games, whereas behavioural patterns determine the likelihood of maladaptive usage.

At the descriptive level, individuals indicated elevated levels of perceived social support, especially from family and close ties. Nonetheless, this did not completely protect against psychological distress, since a significant portion of subjects reported moderate to severe depression and anxiety symptoms. This indicates that simply having access of social support may be inadequate to mitigate emotional vulnerability. The correlation between gaming duration and problematic usage aligns with current literature, which recognizes excessive involvement as a key behavioural indicator of problematic gaming [82,87,88,89]. In the present study, even moderate increases in daily gaming time (e.g., 2–3 hours) were associated with higher rates of problematic gaming, suggesting that risk may emerge at relatively low thresholds in vulnerable individuals.

The type of the game also played a significant effect. Individuals favouring competitive games had elevated levels of problematic gaming in contrast to those who preferred cooperative or relaxed games. This discovery aligns with prior studies emphasizing the reinforcing characteristics of competitive gaming settings, encompassing reward mechanisms, ranking dynamics, and social comparison processes [82,90].

Interestingly, family structure was associated with problematic gaming at the bivariate level, with higher rates observed among participants with siblings; however, this association did not remain significant in multivariate analyses. This finding may reflect shared gaming environments or peer-like social modelling within the family context [91], although current evidence on sibling influences in gaming behaviour remains limited and inconclusive.

At the inferential level, problematic gaming activity was correlated with multiple behavioural, structural, and psychosocial aspects. Bivariate analysis revealed correlations with gaming length, game type, family structure, financial investment, and psychiatric symptoms. Nevertheless, multivariate studies uncovered a more nuanced pattern, indicating that gaming-related characteristics served as the primary predictors, whilst personality traits assumed a secondary, still significant role.

Taken together, these findings support a three-layer model of gaming and psychological functioning.

First, problematic gaming appears to be primarily behaviour-driven, with engagement patterns, such as time spent gaming and preference for competitive games playing a central role. This is consistent with behavioural addiction models emphasizing reinforcement mechanisms, reward systems, and sustained engagement.

Second, personality factors appear to be more strongly associated with problematic gaming than prosocial functioning. Across the regression models, lower conscientiousness consistently emerged as an independent predictor of higher IGDS9-SF score. This data supports the idea that self-regulation, planning, persistence, and impulse control may be important protective variables in gaming-related outcomes. In the context of medical education, where students face high academic demands and increasing autonomy, lower conscientiousness may represent a vulnerability factor for difficulties in regulating leisure activities, including gaming.

Third, psychological distress was relevant in the supplementary psychological–social model. The inclusion of depressive and anxiety symptoms significantly improved the prediction of IGDS9-SF scores, suggesting that emotional distress is associated with problematic gaming symptom severity. However, in the final supplementary model, conscientiousness remained the only significant independent predictor. Therefore, this trend may reflect the significant overlap between depressive and anxious symptoms, as well as their mutual relationship with internal vulnerability. This shows that emotional discomfort may be part of a larger sensitivity profile linked with problematic gaming, rather than acting as a distinct predictor in this cohort. Given the cross-sectional methodology, these results should be viewed as association rather than directionality.

Fourth, in the context of problematic gaming, perceived social support appears to play a limited influence. Although lower levels of perceived support had small bivariate relationships with IGDS9-SF scores, social support did not provide substantial additive explanatory value in multivariate models. In the extra psychological-social model, the block containing family, friends, and significant-other support was not important, and the following addition of prosocial behaviour did not improve the model. Similarly, prosocial behaviour did not explain additional variance after demographic variables, personality traits, gaming-related characteristics, and perceived social support were controlled. Thus, the findings do not support the hypothesis that reduced prosocial functioning is a major independent predictor of problematic gaming in this cohort. This supports a differentiated interpretation: problematic gaming appears to be mainly behaviour-driven and self-regulation-related, whereas prosocial behaviour appears to be more strongly embedded in personality and interpersonal support.

At the same time, the absence of an independent effect of perceived social support should not be interpreted as evidence that social factors are irrelevant. Rather, the findings may indicate that broad perceived social support is not sufficiently specific to capture the social mechanisms involved in problematic gaming. More specific variables, such as online versus offline support, gaming motivation, peer-group gaming norms, family monitoring, loneliness, academic stress, and the social function of gaming may better explain the relationship between social functioning and PGB.

These results have practical implications for medical students’ well-being. Screening for problematic gaming in medical students should not focus only on the number of hours spent gaming, but also on the type of gaming involvement, especially competitive gaming, and on students’ broader self-regulation capacities. Preventive interventions may benefit from addressing time management, coping strategies, sleep routines, academic stress, and awareness of highly reinforcing gaming environments. Interventions targeting problematic gaming should focus on behavioural regulation, including reducing excessive engagement and addressing reinforcement mechanisms associated with competitive gaming. In contrast, interventions aimed at reducing depression and anxiety should prioritize emotion regulation and vulnerability-related traits, particularly negative emotionality. Finally, interventions focused on strengthening supportive interpersonal environments may be more relevant for improving general psychosocial functioning than for directly reducing problematic gaming symptoms.

Future research should adopt longitudinal designs to clarify the temporal relationships between personality, gaming behaviour, and psychological outcomes. In addition, investigating potential mediating and moderating mechanisms would provide a more refined understanding of how these factors interact over time. Particular attention should be given to the qualitative aspects of social support, including its functionality and perceived adequacy, as these may be more relevant than its simple presence.

Overall, the present findings support a multidimensional and domain-specific perspective on gaming and mental health, emphasizing the need to move beyond one-dimensional models that uniformly link gaming behaviour to psychological outcomes.

5. Limitations

The study has a number of methodological limitations. The first limitation identified is that the sample used for the study was relatively homogeneous. The study sample was drawn from a single institution, with all participants enrolled in the same medical faculty. There was also a disproportion regarding the number of students in the preclinical years (I, II – 173 students) and the clinical years (III, IV, V, VI – 70 students). Moreover, participation in the study was determined by the voluntary decision of individuals through a self-selection process, generating a possible bias (the answers were provided by those with higher motivation or greater interest in the topic of the research) and a limitation of the generalizability of the results. Subsequently, the inclusion criteria eliminated the number of responses (some of the registered students were either minors or specified that they did not usually play) and the final sample size was average (243 participants). A second limitation is represented by the measurement of time spent in video gaming through self-reporting by the student and the lack of objective assessment of it. Estimation errors can also be found within other variables regarding gaming habits, such as the age at which the game was started or the average monthly amount spent for this purpose.

Another possible source of error is represented by the psychological variables assessed using standardized question tests. For example, anxiety or depression reported by students can be influenced by various external social or academic factors, such as personal life, financial situation or stress associated with the exam session. Therefore, interpretation should be made with caution, as there are possible associations between these variables themselves and gaming behaviour.

While the IGDS9-SF scores were derived from a self-report screening questionnaire and no professional diagnostic interview was conducted, the results must be considered rather than indicative of PGB or IGD symptom severity and not a formal diagnosis of Internet Gaming Disorder.

6. Conclusions

The current study found that PGB among medical students is mostly linked to gaming involvement and personality-related self-regulation. Higher daily gaming time, preferred game type, and poorer conscientiousness were the strongest predictors of IGDS9-SF results. These findings are especially important for medical students, who confront intense academic pressure, ongoing cognitive demands, and increased autonomy. In this setting, problematic gaming may indicate challenges with self-management and coping, which could have an impact on sleep, study routines, academic functioning, and psychological well-being.

Psychological distress increased IGDS9-SF scores at the block level, but depression and anxiety did not remain independent predictors in the final model. Perceived social support and prosocial behaviour had no significant explanatory value, implying that problematic gaming in this population was not largely caused by poor social or prosocial functioning. The data should be taken as indicating PGB or elevated IGDS9-SF symptom severity, rather than clinically diagnosed Internet Gaming Disorder. Future longitudinal studies should examine factors such as gaming motivation, sleep, academic stress, burnout, and coping strategies to further elucidate the complex relationships between gaming behaviour, psychological well-being, and adaptation among medical students.