Game usage refers to the amount of time spent playing video games, including both clinical and commercial types. Video-game elements promote intrinsic motivation and enjoyment, which can sometimes contribute to excessive gaming behaviours. Reference Arbeau, Thorpe, Stinson, Budlong and Wolff1 However, the mental health consequences of video-game engagement are complex and depend on a range of factors such as the duration and content of the games. Reference Hartanto, Lua, Quek, Yong and Ng2 While playing video games can provide short-term pleasure and psychological benefits, excessive gaming is partially linked to deficits in executive functions, which can impair an individual’s ability to regulate participation in reward-seeking behaviours such as gaming. Reference Dong and Potenza3 This reduction in cognitive-control abilities may also contribute to the development of mental health problems, such as depression. Reference Dotson, McClintock, Verhaeghen, Kim, Draheim and Syzmkowicz4
Apart from shared cognitive risk factors, the connection between excessive video gaming and depression may operate through social deficits. The social displacement theory suggests that excessive engagement with video games may diminish opportunities for physical activity and real-life social interactions. Over time, this can lead to a social deficit pathway, whereby depressive symptoms are exacerbated through a reduction in both the quantity and quality of offline social connections. Reference Kowert, Domahidi, Festl and Quandt5 As proposed by the social support and buffering theory, Reference Cohen and Wills6 there are two models – the buffering model and the main-effect model. The buffering model suggests that social relationships can buffer the negative effects of stressful events by providing affected individuals with support and care. On the other hand, the main-effect model proposes that having a large social network, such as a group of friends, benefits well-being by providing individuals with stable and socially rewarding roles within their social community. Different studies have shown the benefits of social support across different support sources in the prevention and treatment of depression. Reference Rueger, Malecki, Pyun, Aycock and Coyle7
Social support refers to actions and behaviours that convey to an individual, either directly or indirectly, that they are appreciated and looked after by others. A classification system for social support outlines five key categories of support: sharing information, emotional support, feeling respected, having a social network and practical help. Reference Ko, Wang and Xu8 Of special note, online interactions are not as effective as offline interactions in providing practical help and showing respect. Reference Liu, Wright and Hu9 Furthermore, authentic support from real-life friends is essential for receiving meaningful emotional assistance. Reference Meshi and Ellithorpe10 Individuals who are dissatisfied with the support they receive from offline contacts or lack satisfactory support from offline social contacts may become involved in online social interactions to seek alternative sources of support. Reference Chung11 However, a meta-analysis of studies on the relationship between online social support and adolescent mental health found that the support received from internet acquaintances may not yield the same positive impact as the support received from real-life acquaintances. Reference Zhou and Cheng12 When the need for relatedness is not met, individuals may experience social isolation, rejection and feelings of loneliness. It is noteworthy that real-life relationships include not just friends but also parents, who are important sources of social support. Research shows that optimal parenting involves guiding children’s screen use with reasoning and by considering their perspective. Reference Detnakarintra, Trairatvorakul, Pruksananonda and Chonchaiya13 This approach can help children develop social skills, make friends and balance gaming with other activities. Reference Goering and Mrug14,Reference Goagoses, Bolz, Eilts, Schipper, Schütz and Rademacher15
Given the potential effect of social factors, this study aimed to examine three research questions on the link between video-gaming duration and depressive symptoms in Chinese junior high school students: (a) whether frequent video gaming is significantly associated with adolescent depressive symptoms; (b) whether this association can be mediated by perceived support from friends and (c) whether this association can be moderated by childhood experiences of optimal parenting. The hypothesised model was shown in Fig. 1. Based on the theories and empirical evidence reviewed above, we expected that longer durations of video gaming were associated with higher levels of depressive symptoms (Hypothesis 1). Additionally, longer durations of video gaming were associated with lower levels of perceived support from friends, which in turn were linked to higher levels of depressive symptoms (Hypothesis 2). However, the role of friend support was moderated by childhood experiences of optimal parenting, with more pronounced peer influences observed in adolescents who lacked optimal parenting (Hypothesis 3).
Fig. 1 Hypothesised model.
Method
Participants
We recruited eligible adolescents aged 11–14 who can read Chinese or English and had no physical disabilities through local high schools, social media and adolescent community centres in Hong Kong and invited them to complete our survey either in paper format or electronically. We chose this age group because early adolescence represents a transition period marked by a gradual shift from complete reliance on parents to greater independence and peer influences. Studying this population can capture the influences of both parental support and peer support. By utilising this multi-recruitment sites approach, we were able to gather valid data from 1071 Chinese adolescents.
Procedure
This study was approved by the Institutional Review Board of The University of Hong Kong and the Hospital Authority Hong Kong West Cluster (UW 19-722). It involved two sets of questionnaires, one for adolescents to self-report and another for parents to report on their family demographics. The online survey platform provided instructions for completing each survey and ensured anonymity. All participants provided their consent and completed the survey voluntarily, with the option to withdraw at any time. On average, the set of questionnaires took approximately 30 min to complete.
Measures
Video-game usage
Participants completed a questionnaire about the amount of time they spent playing video games on handheld and/or other types of consoles on both weekdays and weekends over the past two months. The questionnaire is a reliable tool for assessing screen usage and has been used in previous studies involving children and adolescents in Hong Kong. Reference Wong, Tung, Rao, Leung, Hui and Tso16–Reference Wong, Tung, Rao, Ho, Chan and Fu18 Total gaming time was calculated using a weighted average formula ((2 × weekend + 5 × weekday) ÷ 7) and reported in hours.
Support from friends
The friend support subscale of the Multidimensional Scale of Perceived Social Support (MSPSS) Reference Zimet, Dahlem, Zimet and Farley19 was used to assess participants’ perceptions of support received from friends in the past month. The MSPSS has been widely used in studies involving Chinese adolescents and has demonstrated good reliability and validity. Reference Wong, Tung, Fu, Bacon-Shone, Molasiotis and Li20,Reference Wan, Yang, Liu, Zhang, Liu and Jia21 The friend support subscale comprised four items (e.g. ‘My friends really try to help me.’) rated on a 7-point Likert scale, ranging from 1 (strongly disagree) to 7 (strongly agree). Higher scores indicated higher levels of support from friends. In our study, Cronbach’s α for this scale was 0.93.
Childhood experiences of optimal parenting
The Parental Bonding Instrument (PBI) Reference Parker, Tupling and Brown22 was used to assess participants’ perceptions of parental care and support during their childhood. The Chinese version of the PBI has been validated in Hong Kong and shown to be valid and reliable. Reference Ngai, Cheung, Xie, Ng, Ngai and Liu23,Reference Wong, Tung, Chan, Wong, Tsang and Chow24 The PBI consists of two subscales: one evaluates care (12 items) and the other evaluates overprotection (13 items). Participants responded to these items separately for their mothers and fathers using a 4-point Likert scale, ranging from 0 (very unlike) to 3 (very like). Higher scores indicated a greater inclination towards the measured parenting practice. Furthermore, we categorised their mothers and fathers into ‘high’ and ‘low’ groups based on the specific subscale cut-off scores (care score of 27 for mothers and 24 for fathers, and overprotection score of 13.5 for mothers and 12.5 for fathers). Reference Parker, Tupling and Brown22 Optimal parenting was defined as having high care and low overprotection. Participants whose mothers and fathers did not meet the criteria were considered to have suboptimal parenting in subsequent analyses. In our study, Cronbach’s α for the care scale was 0.85 for the mother and 0.88 for the father. Additionally, Cronbach’s α for the overprotection scale was 0.78 for the mother and 0.80 for the father.
Depressive symptoms
The depression subscale of the Depression Anxiety Stress Scale (DASS-21) Reference Lovibond and Lovibond25 was used to measure levels of depressive symptoms in participants at the time of completing the scale. DASS-21 is a suitable assessment tool for Chinese adolescents as it has been proven to be both reliable and valid. Reference Mellor, Vinet, Xu, Mamat, Richardson and Román26,Reference Chan, Chan, Chen, Hui, Lee and Suen27 The depression subscale included 7 items (e.g. ‘I found it difficult to work up the initiative to do things’) rated on a 4-point Likert scale, ranging from 0 (strongly disagree) to 3 (strongly agree). The total score was generated by adding the item scores and then multiplying the sum by 2. Higher scores indicated more depressive symptoms. In our study, Cronbach’s α for this scale was 0.81.
Covariates
Since previous studies have demonstrated significant associations of attention-deficit hyperactivity disorder (ADHD) with screen time and depression, Reference Eirich, McArthur, Anhorn, McGuinness, Christakis and Madigan28–Reference Dardani, Davey Smith, Leppert, O’Donovan, Rice and Riglin30 adjusting for ADHD symptoms can minimise the likelihood of a false association between video-gaming duration and depressive symptoms in participants. Specifically, we asked their primary caregiver to complete the Chinese version of the Strengths and Weaknesses of ADHD Symptoms and Normal Behaviour Scale (SWAN), which has been validated in Hong Kong. Reference Lai, Leung, Luk, Wong, Law and Ho31 This scale consists of 18 items rated on a 7-point Likert scale, with values ranging from −3 (far above normal) to +3 (far below normal). The total score, obtained by summing individual item scores, indicates the severity of ADHD symptoms in participants. Additionally, the primary caregiver provided their educational and employment details, as well as their spouse’s, along with their monthly household income. These demographic details were combined to generate a socioeconomic status (SES) index. This index was calculated using principal component analysis with varimax rotation, with greater values indicating higher social classes. Reference Vyas and Kumaranayake32 Participants’ age and gender were collected through self-report.
Data analysis
Data analysis was performed using the software package SPSS 22.0. To summarise participants’ sociodemographic characteristics, we computed descriptive statistics, including the mean and s.d. for continuous variables, and frequencies and percentages for categorical variables. Pearson’s correlation analyses were conducted to examine the relationships between the variables of interest. Additionally, we employed model 4 of the PROCESS macro to examine the mediating role of perceived friend support between video-gaming duration and depressive symptoms. To explore the role of childhood experiences of optimal parenting in these associations, we utilised model 7 of the PROCESS macro, which is commonly used for testing moderated mediation models. To assess the significance of conditional direct and indirect effects, we estimated bias-corrected bootstrap confidence intervals using 5000 bootstrap resamples. A significant effect was indicated by a confidence interval that did not include zero.
Results
Descriptive statistics and correlations among the variables
Table 1 provides descriptive statistics. The mean age of the 1071 participants (471 females and 600 males) was 13.62 years old, with an s.d. of 0.95. Around 30% of their fathers and mothers had completed tertiary education. Additionally, a majority of fathers (87.7%) were engaged in full-time employment, while a substantial proportion of mothers (40.7%) were housewives. Moreover, 559 (52.2%) families had a monthly household income that fell below the median income in Hong Kong (HKD 30 000). Correlation analyses revealed that depressive symptoms were associated negatively with support from friends (r = −0.26, p < 0.001) and positively with ADHD symptoms (r = 0.19, p < 0.001), whereas support from friends was negatively associated with ADHD symptoms (r = −0.14, p < 0.001).
Table 1 Participant characteristics
ADHD, attention-deficit hyperactivity disorder.
a. Missing data not shown.
Based on the recommended cut-off score of 21 on the DASS depression scale, Reference Mellor, Vinet, Xu, Mamat, Richardson and Román26,Reference Chan, Chan, Chen, Hui, Lee and Suen27 79 participants (7.4%) were identified as having significant depressive symptoms. Compared to their counterparts, these participants reported significantly lower levels of friend support (16.86 v. 21.24, p < 0.001), longer durations of video gaming (4.71 v. 3.97, p = 0.016), reduced perceived maternal care (19.89 v. 25.73, p < 0.001) and lower paternal care (17.91 v. 23.95, <0.001). Additionally, they perceived higher maternal overprotection (16.39 v. 13.34, p < 0.001) and paternal overprotection (14.32 v. 11.34, p < 0.001).
Testing for mediation
As shown in Table 2, after adjusting for age, gender, family SES and ADHD symptoms, video-game usage was positively associated with depressive symptoms (β = 1.31, p = 0.006) and depressive symptoms were negatively associated with support from friends (β = −0.34, p < 0.001). The association of video-game usage with depressive symptoms remained significant after further adjusting for support from friends (β = 1.14, p = 0.014). However, video-game usage was not associated with support from friends (β = −0.53, p = 0.127). Therefore, support from friends did not mediate the relationship between video-game usage and depressive symptoms before taking parenting influences into account.
Table 2 Testing the effect of video-game usage on depressive symptoms via support from friends
N = 1071, Dep, depressive symptoms; FSES, family socioeconomic status; ADHD, attention-deficit hyperactivity disorder; VGU, video-game usage. Bootstrap sample size = 5000. The beta coefficient was unstandardised. *p < 0.05, **p < 0.01, ***p < 0.001.
Testing for moderated mediation
Hypothesis 3 predicted that perceived suboptimal care and overprotection from parents during childhood would strengthen the association between frequent video gaming and depressive symptoms via reduced support from friends. As shown in Table 3, after adjusting for age, gender, family SES and ADHD symptoms, a significant interaction effect of video-game usage and childhood experiences of optimal parenting on support from friends was observed (β = −1.66, p = 0.014). Simple slope tests showed that the association between video-game usage and support from friends was significant only in adolescents who experienced suboptimal parenting during childhood (β = −1.12, p = 0.011), but not in those who perceived optimal care and support from at least one parent (β = 0.55, p = 0.297). That is, the association between frequent video gaming and depressive symptoms through reduced support from friends was significant only when adolescents lacked optimal care and support from both parents during childhood (indirect effect 0.38, s.e. = 0.16, 95% CI = 0.08–0.72). Therefore, Hypothesis 3 was supported.
Table 3 Testing the moderated mediation effect of video-game usage on depressive symptoms
N = 1071, Dep, depressive symptoms; FSES, family socioeconomic status; ADHD, attention-deficit hyperactivity disorder; VGU, video game usage; LOP, lack of optimal parenting. Bootstrap sample size = 5000. The beta coefficient was unstandardised. *p < 0.05, **p < 0.01, ***p < 0.001.
Discussion
The findings of this study indicate the amount of time spent playing video games as a significant lifestyle factor associated with depressive symptoms in adolescents. We found that adolescents who spent more time playing video games were more likely to experience depressive symptoms. Additionally, perceived support from friends served as a conditional mediator between frequent video gaming and depressive symptoms. The significance of peer support was moderated by childhood experiences of optimal parenting, with more pronounced peer influences observed in adolescents who lacked optimal care and support from both parents during childhood. Nonetheless, perceived support from friends, when not stratified by childhood experiences of optimal parenting, did not mediate the association between frequent video gaming and adolescent depressive symptoms. These findings highlight the significance of parenting influences, which can counteract the negative effect of reduced peer support on adolescent mental health. However, the development of depression is complex, involving various biological, cognitive, emotional and interpersonal risk factors. Reference Reivich, Gillham, Chaplin, Seligman, Goldstein and Brooks33 Future research should consider exploring other pathways through which video-game usage could potentially influence lifestyle habits and how these lifestyle changes relate to depressive symptoms. For example, the social displacement theory Reference Kowert, Domahidi, Festl and Quandt5 suggests that increased gaming time may encroach on opportunities for healthy activities, such as physical exercise. Furthermore, excessive gaming can lead to cognitive overstimulation, which may impair sleep quality and potentially exacerbate depression. Reference Jiang, Zhao, Wang, Hua, Chen and Yao34,Reference Swann, Kilpatrick, Breslin and Oddy35
We found evidence for the association between frequent video gaming and depressive symptoms through reduced peer support but only for adolescents who lacked optimal care and support from both parents during childhood. In other words, adolescents who grow up in a neglectful or overprotective family environment may be more susceptible to depression when they concurrently experience insufficient peer support due to excessive video gaming. Our results align with the principles of basic psychological need theory, Reference Vansteenkiste, Ryan and Soenens36 which emphasise the importance of relatedness for maintaining health and well-being. Previous research has shown that real-life social support has a greater positive impact on mental health than support received through social media. Reference Meshi and Ellithorpe10 Excessive gaming may reduce opportunities to foster feelings of relatedness and belonging in real-world relationships. Our findings also demonstrate that increased time spent playing video games may adversely influence adolescents’ perceptions of peer support, particularly when parental support is lacking, potentially undermining their mental health.
According to Bronfenbrenner’s bioecological theory, Reference Bronfenbrenner37 an individual’s growth and development occur in distinct patterns that are shaped by the interplay of protective and risk exposures within and across different ecological levels. These exposures, either alone or in combination, can have an impact on health outcomes and developmental trajectories. Consistent with this perspective, our findings show that suboptimal interpersonal factors, such as a lack of support from both parents and friends, combined with suboptimal lifestyle choices such as frequent video gaming, can increase negative emotions. Previous research has identified certain individual factors, such as neuroticism, that may elevate the risk of developing depressive symptoms. Reference Ono, Takaesu, Nakai, Ichiki, Masuya and Kusumi38 It is plausible that positive environmental factors, such as supportive surroundings, can mitigate the risks associated with suboptimal behavioural choices or innate characteristics to adolescent mental health. For example, optimal parenting was found to be a protective factor in this study, which aligns with previous research showing that caring and understanding parenting approaches can promote children’s social skills, such as children’s ability to accurately interpret emotions and navigate social interactions in a positive direction. Reference Zimmer-Gembeck, Rudolph, Kerin and Bohadana-Brown39 There is also evidence suggesting that adolescents with higher levels of parental support are less likely to experience emotional problems due to a lack of support from another source. Reference Lyell, Coyle, Malecki and Santuzzi40
It is worth noting that the mediating role of peer support was significant only for participants who experienced suboptimal parenting during childhood. Our study did not examine whether and how frequent video gaming may influence adolescent mental health in a supportive family context. Notably, some studies showed that individuals who heavily engage in video games may struggle with self-control issues, which can be partly attributed to changes in brain structure associated with depression. Reference Dotson, McClintock, Verhaeghen, Kim, Draheim and Syzmkowicz4 These cognitive problems can be inborn or caused by factors outside of the family context and warrant further investigation. The results of this study have several practical implications. First, interventions should prioritise reducing problematic video gaming. Due to the widespread availability of video gaming, it can be challenging to completely abstain from it. Therefore, interventions can focus on moderating gaming use and increasing participation in non-technology or offline activities. Second, given that perceived peer support is a conditional mediator linking video gaming to depressive symptoms, we should promote offline social activities for high game players to enhance their sense of peer support. For instance, mentorship programmes can be established where experienced high game players can offer guidance and support to newer or struggling players. This can help to improve relationships between players and provide opportunities for peer support that extend beyond the virtual gaming world. Third, the moderating effect of parenting quality indicates that proper parenting strategies, such as showing higher levels of care and avoiding excessive overprotection, provide crucial social resources to mitigate the social deficit pathway that connects video gaming to depressive symptoms.
Several limitations should be considered. First, this study relied on a cross-sectional survey design, making it impossible to establish causal associations. Future research could benefit from longitudinal data or experimental studies to better ascertain the findings. It is also important to investigate whether the association between gaming and depression is bidirectional and driven by the same underlying mechanisms. Second, while the self-report method is commonly used in literature to measure depressive symptoms, future studies should consider employing multiple methods to gather data from various sources including parents and peers. Third, caution should be exercised when generalising our findings from early adolescents to other age groups. Furthermore, while our results were adjusted for ADHD symptoms, other unmeasured neurodevelopmental problems, such as autism, may also have an impact on the findings. Last, this study examined perceived support from friends and parents among participants who are predominantly below the clinical depression threshold. Other factors such as school connectedness and loneliness may also influence the association. Future research should investigate how the severity of gaming behaviour, such as addiction, is linked to clinical depression, whether in a manner similar to or different from the findings of this study.
In conclusion, this study sheds light on a social deficit pathway through which frequent video gaming is linked to depressive symptoms in early adolescents. Support from friends and parents plays a role in the association between frequent video gaming and depressive symptoms in adolescents. Parental support can mitigate the negative effect of frequent video gaming on perceived peer support and depressive symptoms in adolescents. Promoting peer support is conducive to mitigating depressive symptoms particularly for adolescent gamers who lack optimal parenting.
Data availability
The data that support the findings of this study are available on request from the corresponding author.
Acknowledgement
We thank participants for their participation in this study.
Author contributions
R.S.W. conceptualised the paper, analysed the data and wrote the initial draft. K.T.S.T. coordinated data collection and critically revised the manuscript. P.I. supervised data collection and critically revised the manuscript. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
Funding
All authors have indicated they have no financial relationships relevant to this article to disclose.
Declaration of interest
None.
Ethical standards
This research was approved by the Institutional Review Board of The University of Hong Kong and the Hospital Authority Hong Kong West Cluster (UW 19-722). Written informed consent and informed assent were obtained from participants and their parents.
Open access
eLetters
No eLetters have been published for this article.