1. Introduction
Internet gaming disorder was included in Section III of the current version of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) as a new condition warranting more clinical research and experience [Reference American Psychiatric Association1]. Internet gaming disorder refers to a problematic (online or offline) use of video games and is characterized by 9 different diagnostic criteria (preoccupation with playing, withdrawal symptoms when not playing, tolerance, unsuccessful attempts to reduce or stop playing, giving up other activities, continuation of playing despite problems, deceiving or covering up playing, playing to escape adverse moods, and risking or losing relationships or career opportunities because of excessive playing) [Reference Petry, Rehbein, Ko and O’Brien2]. To confirm a diagnosis of Internet gaming disorder, at least 5 out of these 9 criteria have to be met for the past 12 months [Reference American Psychiatric Association1]. A cut-off point of 5 criteria was empirically verified in several studies [Reference Ko, Yen, Chen, Wang, Chen and Yen3,Reference Lemmens, Valkenburg and Gentile4]. Petry and colleagues [Reference Petry, Rehbein, Gentile, Lemmens, Rumpf and Mößle5] explored the fit of 8 established instruments measuring “problematic gaming” or “Internet addiction” that were published before the release of the DSM-5 [Reference American Psychiatric Association1], to these new criteria, and found none of them to cover all 9 diagnostic criteria of Internet gaming disorder. Meanwhile, several new questionnaires covering all aspects of Internet gaming disorder were developed and published.
Comprehensive theoretical models for the etiology of the disorder are still lacking, but Kiraly et al. [Reference Király, Griffiths and Demetrovics6] underlined the interplay of 3 key factors for the development of Internet gaming disorder: Structural characteristics of the video games, psychological characteristics of the gamer (the present study focused on this aspect) and motivations for playing games.
1.1. Prevalence of internet gaming disorder
According to Petry et al. [Reference Petry, Rehbein, Ko and O’Brien2], only 2 prevalence estimates based on appropriate instruments have been published for Internet gaming disorder. Rehbein et al. [Reference Rehbein, Kliem, Baier, Mößle and Petry7] surveyed a large sample of adolescents (aged between 13 and 18 years) in Lower Saxony (1 of the 16 federal states in Germany) and observed a prevalence rate of 1.2% [95% confidence interval (95%CI): 1.0–1.4%] for Internet gaming disorder. Lemmens et al. [Reference Lemmens, Valkenburg and Gentile4] investigated a representative sample of Dutch adolescents and adults (aged between 13 and 40 years) and reported a prevalence estimate for Internet gaming disorder of 3.8% (based on a latent class analysis, 95%CI was not reported in the publication) or 5.4% (based on the cut-off point ≥5). For the youngest age group (13–20 years) a prevalence of 5.5% (cut-off point ≥5) was observed [Reference Lemmens, Valkenburg and Gentile4]. Quite recently, Pontes et al. [Reference Pontes, Macur and Griffiths8] stated a prevalence rate of 2.5% (95%CI: 1.7–3.7%) for Internet gaming disorder in a sample of Slovenian adolescents (N = 1071, age ranged from 12 to 16 years).
1.2. Associations between Internet gaming disorder and mental health
Kiraly et al. [Reference Király, Griffiths and Demetrovics6] emphasized the psychological characteristics of the gamer as 1 of the 3 key factors for the development of Internet gaming disorder, but currently, associations between Internet gaming disorder and mental health are largely unexplained. In addition, only very few studies used entirely DSM-5-based instruments [Reference Petry, Rehbein, Ko and O’Brien2] to investigate associations between Internet gaming disorder and mental health. Lemmens et al. [Reference Lemmens, Valkenburg and Gentile4] reported statistically significant bivariate associations between Internet gaming disorder and lower self-esteem, less prosocial behavior, a lower satisfaction with life, as well as higher levels of loneliness and aggression in a population sample (2444 adolescents and adults). King and Delfabbro [Reference King and Delfabbro9] examined 844 adolescents (average age: 14.1 years, SD = 1.5 years) with regard to Internet gaming disorder, depression, anxiety, stress and gaming-specific cognitions. They observed statistically significant bivariate correlations between Internet gaming disorder and depression, anxiety and stress [Reference King and Delfabbro9]. But in a multivariate model (adjusted for age, gender, gaming frequency and gaming specific cognitions) none of the 3 psychopathological aspects was statistically significantly related to Internet gaming disorder. Thus, first findings indicate dimensional interrelations between Internet gaming disorder and adolescent mental health problems such as aggression and low self-esteem [Reference Lemmens, Valkenburg and Gentile4], but further research is needed.
For a healthy development of children and adolescents, the mental health status of their parents is also of high importance. Recently, Lam [Reference Lam10] reported an empirical association between parental depression and adolescent Internet addiction. But to our knowledge no published study examined, if Internet gaming disorder is also related to parental mental health. Therefore, the combination of adolescent self-reports and parental ratings is a new approach for the investigation of Internet gaming disorder.
1.3. Aim and research questions
The objective of the present study was thus to examine with a dimensional and a categorical approach the association of Internet gaming disorder with both adolescent and parental mental health in early adolescence. We explored the following research questions:
Which aspects of adolescent and parental mental health are associated with Internet gaming disorder in early adolescence?
How much variance can be explained by adolescent and parental mental health?
2. Methods
2.1. Participants
The sample consisted of 1095 family dyads (an adolescent and a related caregiver each). Overall, we surveyed 539 girls (49.2%) and 556 boys (50.8%). Moreover, 98.8% of caregivers were biological parents (931 mothers, 85.0% and 151 fathers, 13.8%), 0.6% were new partners of the mother (5 cases) or of the father (1 case), and 0.7% grandparents (5 grandfathers, 2 grandmothers). In the following, all caregivers are labeled as parents. The mean age of adolescents was 12.99 (SD = 0.82, range: 12–14) and of the mean age of the parents was 41.63 (SD = 5.76, range: 26–80) years. Based on the current school performance of the adolescent, the parent was requested to predict the prospective level of graduation of her or his child (forecast). Overall, the parents predicted a graduation on a high educational level for 40.8% of the adolescents, on a medium educational level for 48.2% and on a low level for 11.0% of the sample.
2.2. Measures
Internet gaming disorder was measured for the last 12 months using the Internet Gaming Disorder Scale (IGDS) [Reference Lemmens, Valkenburg and Gentile4]. The IGDS consists of 9 self-report items on the presence of Internet gaming disorder symptoms (binary response format: 0 = “no”, 1 = “yes”). By summing up the responses, an IGDS sum score was calculated with a higher sum indicating a higher risk of Internet gaming disorder. Following Lemmens et al. [Reference Lemmens, Valkenburg and Gentile4], adolescents who answered yes to 5 or more of the criteria (positive screening result) were classified as having Internet gaming disorder.
To assess self-reported adolescent mental health within the last 6 months, we used the Screening psychischer Störungen im Jugendalter-II (SPS-J-II) [Reference Hampel and Petermann11], which is the German adaption of the Reynolds Adolescent Adjustment Screening Inventory [Reference Reynolds12]. The SPS-J-II consists of 32 items with a 3-level response format (0 = “never or almost never”, 1 = “sometimes”, 2 = “nearly all the time”) which is divided into 4 subscales evaluating the frequency of antisocial behavior, anger control problems, emotional distress (combined measure of anxiety and depressiveness), and self-esteem problems. Higher scores indicate a greater degree of psychopathological problems.
To assess a parental rating of adolescent hyperactivity/inattention within the last 6 months, we applied 1 subscale of the Strengths and Difficulties Questionnaire (SDQ) [Reference Goodman13]. The SDQ-scale hyperactivity/inattention consists of 5 items with a 3-level response format (0 = “not true”, 1 = “somewhat true”, 2 = “certainly true). A higher score indicates a higher level of hyperactivity/inattention.
Parental depression was assessed with the Patient Health Questionnaire-2 (PHQ-2) [Reference Kroenke, Spitzer and Williams14]. The PHQ-2 consists of 2 items rating the frequency of depressed mood and anhedonia over the last 2 weeks (4-level response format: 0 = “not at all”, 1 = “several days”, 2 = “more than half the days”, 3 = “nearly everyday”). A higher score indicates a higher severity of depression.
Parental anxiety within the last 2 weeks was measured using the Generalized Anxiety Disorder Scale-2 (GAD-2) [Reference Kroenke, Spitzer, Williams, Monahan and Löwe15]. The GAD-2 also consists of only 2 questions rating nervousness and worrying with a 4-level response format (0 = “not at all”, 1 = “several days”, 2 = “more than half the days”, 3 = “nearly everyday”). A higher sum indicates a higher severity of anxiety.
2.3. Procedure
According to published results, the prevalence estimates for Internet gaming disorder in Europe are in a single-digit percentage range [Reference Lemmens, Valkenburg and Gentile4,Reference Rehbein, Kliem, Baier, Mößle and Petry7,Reference Pontes, Macur and Griffiths8]. Therefore, in population samples only a small number of persons are expected to be affected by Internet gaming disorder. A very small number of cases lead to restrictions concerning suitable statistical analysis methods. To counteract these problems, adolescents with subjectively perceived problems in digital media use were oversampled in the present survey. Based on parental and adolescent responses to 2 screening items (assessing the self-perceived frequency of adolescent excessive digital media use and the severity of problematic digital media use), the adolescents were assigned to a group with more subjectively perceived problems in digital media use (higher risk group) or to another group with less subjectively perceived problems (lower risk group). Data was collected from January to March 2016 by an experienced German market research institute in 757 higher risk group families (target value: 70% of the whole sample, target achievement: 69.1% of the sample) and 338 lower risk group (target value: 30%, achievement: 30.9%) families. The data collection was carried out in all 16 German federal states by trained interviewers in the family's home. Face-to-face interviews were conducted separately with a parent and the adolescent.
2.4. Statistical analyses
Statistical analyses were performed on all 1095 family dyads (regardless of whether an adolescent was assigned to the higher or lower risk group). Overall, 76 adolescents (6.9% of the whole sample) stated that they had never played video games (we asked for all types of online or offline games played on computer, tablet computer, games console and smartphone) and therefore, did not have to answer the 9 IGDS questions. For these cases, we determined the IGDS sum value on 0 and assigned them to the group without Internet gaming disorder. We used SPSS version 22.0 (IBM, 2013, New York, USA) to calculate linear regression analysis (dimensional approach) and logistic regression analysis (categorical approach). The response variable for the linear regression analysis was the IGDS sum score. For the logistic regression analysis, it was a diagnosis of Internet gaming disorder (“no” or “yes”, based on the cut-off point ≥ 5) [Reference Lemmens, Valkenburg and Gentile4]. As explanatory variables in all models, we used gender of the adolescent, gender of the parent, antisocial behavior, anger control problems, emotional distress, self-esteem problems, hyperactivity/inattention of the adolescent, as well as depression and anxiety of the parent. We tested multicollinearity with the condition index. The goodness of fit for the multivariate logistic regression model was verified by the Hosmer–Lemeshow test. In a sensitivity analysis, we weighted the sample based on the Internet gaming disorder diagnosis using inverse probability weighting. We aimed to obtain a prevalence of Internet gaming disorder that corresponds to the figure (5.5%) reported by Lemmens et al. [Reference Lemmens, Valkenburg and Gentile4] for a population sample and repeated the statistical analyses for this weighted sample.
3. Results
3.1. Dimensional approach
In the linear regression analysis we found statistically significant associations between Internet gaming disorder and (male) gender of the adolescent, adolescent antisocial behavior, anger control problems, emotional distress, self-esteem problems, hyperactivity/inattention and parental anxiety (see Table 1). In the conducted multivariable analyses a higher level of psychopathological problems was always related to a higher degree of Internet gaming disorder. Parental depression and gender of the parent were not statistically significantly associated with Internet gaming disorder. For the multivariate model, the corrected R 2 was 0.41, i.e., two fifths of the variance in the outcome could be explained by the predictors. The condition index for the model was 10.57. Accordingly, we did not see any indications for multicollinearity. Remarkably, gender of the adolescent is the strongest related factor, but the combined aspects of mental health explain a clearly greater proportion of variance in the multivariate model (corrected R 2 for an alternative model including only the different aspects of adolescent and parental mental health: 0.30).
Table 1 Associations of different aspects of adolescent and parental mental health with Internet gaming disorder.
a P < 0.05.
b P < 0.01.
c P < 0.001.
d Ratings by adolescents.
e Ratings by parents.
SPS-J-II: screening psychischer Störungen im Jugendalter-II; SDQ: strengths and difficulties questionnaire; PHQ-2: patient health questionnaire-2; GAD-2: generalized anxiety disorder scale-2.
3.2. Categorical approach
In the logistic regression analysis, male gender, more antisocial behavior, anger control problems, emotional distress, self-esteem problems, hyperactivity/inattention, as well as less parental depression and more parental anxiety were associated with higher odds of Internet gaming disorder diagnosis (see right column in Table 1 for adjusted odds ratios of the multivariable analyses). Gender of the parent was not related to Internet gaming disorder. The result of the Hosmer–Lemeshow Test (P = 0.742) indicates a good model fit. Nagelkerke's R 2 showed a value of 0.41, indicating that two fifths of the variation between the two groups (adolescents with and without Internet gaming disorder) was explained by these factors (Nagelkerke's R 2 for an alternative model including only the aspects of adolescent and parental mental health: 0.30).
3.3. Sensitivity analyses with a weighted sample
Due to the oversampling of adolescents with subjectively perceived problems in digital media use, we observed a high percentage of persons with a positive screening result for Internet gaming disorder in our sample (260 adolescents). In a sensitivity analysis, we thus repeated the same statistical analyses as reported above, using a weighting scheme that adjusted the prevalence rate for Internet gaming disorder (5.5%, which is equivalent to a proportion previously observed in a representative population sample) [Reference Lemmens, Valkenburg and Gentile4]. For adolescents reporting an IGDS sum score below 5 (n = 835, 76.3% of the whole sample) we used a weight of 1.239 and if the IGDS sum score was 5 or higher (n = 260, 23.7% of the sample), we used a weight of 0.232 (adjustment to a level of 5.5%). In the linear regression analysis with the weighted sample, we observed a corrected R 2 of 0.26 and standardized Beta coefficients that were similar to those in the unweighted sample (gender of adolescent: −0.31, gender of parent: 0.03, antisocial behavior: 0.16, anger control problems: 0.06, emotional distress: 0.06, self-esteem problems: 0.13, hyperactivity/inattention: 0.15, parental depression: −0.06, parental anxiety: 0.06). In the logistic regression analysis with the weighted sample, Nagelkerke's R 2 was 0.27 and the odds ratios were very similar to those in the unweighted sample (gender of adolescent: 0.15, gender of parent: 0.94, antisocial behavior: 1.11, anger control problems: 1.09, emotional distress: 1.09, self-esteem problems: 1.15, hyperactivity/inattention: 1.27, parental depression: 0.75, parental anxiety: 1.30). In the weighted sample, 3 explanatory variables (male gender, more self-esteem problems and hyperactivity/inattention) were statistically significantly associated with Internet gaming disorder, both in a dimensional and in a categorical statistical approach (in the linear regression analysis only, antisocial behavior was additionally related to Internet gaming disorder).
4. Discussion
In the present study, self-reports and external ratings were combined to examine associations between Internet gaming disorder and different aspects of adolescent and parental mental health for the first time. We used a dyadic approach, surveying an adolescent and a related parent. Internet gaming disorder appears to be associated with internalizing and externalizing problems in adolescents. Moreover, the findings of the present study provide first evidence that not only adolescent but also parental mental health is relevant to Internet gaming disorder in early adolescence.
Concerning externalizing problems, we found Internet gaming disorder to be related to adolescent antisocial behavior, anger control problems, and hyperactivity/inattention. This is in line with the result of Lemmens et al. [Reference Lemmens, Valkenburg and Gentile4] reporting an association between Internet gaming disorder and aggressive behavior. To our knowledge, there are no published findings regarding Internet gaming disorder and hyperactivity in adolescents so far, but according to Yen et al. [Reference Yen, Liu, Wang, Chen, Yen and Ko16] attention deficit and hyperactivity disorder (ADHD) is a common comorbidity of Internet gaming disorder among young adults. In summary, the results of the present study regarding externalizing problems fit into the pattern of findings from previously published cross-sectional surveys.
Concerning internalizing problems, we observed associations between Internet gaming disorder and both adolescent emotional distress and self-esteem problems. Lemmens et al. [Reference Lemmens, Valkenburg and Gentile4] also reported a relation between Internet gaming disorder and a lower self-esteem. Furthermore, Kim et al. [Reference Kim, Hwang, Choi, Kim, Demetrovics and Király17] observed in an adult sample (aged 20–49 years) statistically significantly higher depression and anxiety scores for the Internet gaming disorder risk group compared to healthy controls. Interestingly, not only internalizing problems of the adolescent seem to be related to Internet gaming disorder but also higher parental anxiety. Thus, adolescent Internet gaming disorder could be interpreted as an attempt to cope with own mental health problems or to compensate problematic interactions with her or his parent (e.g. promoted by parental anxiety). Surprisingly, a lower parental depression was associated with Internet gaming disorder in the categorical statistical analysis, but this finding needs replication. To sum up, it can be stated that not only aspects of adolescent but also of parental mental health seem to be related to Internet gaming disorder and have explanatory value even in multivariate models (in dimensional as well as in categorical statistical approaches).
The present study has several limitations. We have used an approach to oversample adolescents with different degrees of severity of problematic media use. Therefore, we did not survey a representative population sample and it remains unclear, if the results are valid for all German adolescents. To examine the robustness of our findings, we conducted a sensitivity analysis with a weighted sample that is likely to be more similar to a population sample. In this weighted sample with a lower prevalence of Internet gaming disorder, the standardized beta coefficients and the odds ratios were very similar to the results for the unweighted sample. However, in the weighted sample the explained variance or variation between the two groups was lower. Furthermore, the conducted sensitivity analysis is only as reliable, as the prevalence estimate reported by Lemmens et al. [Reference Lemmens, Valkenburg and Gentile4] fits for German adolescents.
As another limitation, the parents were only asked to rate their own anxiety and depression, but it is conceivable that other aspects of mental health (e.g., parental aggression) could be also associated with adolescent Internet gaming disorder. In future studies, it would be interesting to investigate the relation between adolescent and parental mental health and Internet gaming disorder in early adolescence in comprehensive models (e.g. including mediator or moderator variables or interaction effects).
Due to the cross-sectional design, it remains unclear if the related factors such as parental anxiety promote adolescent Internet gaming disorder or vice versa, Internet gaming disorder leads to a higher level of anxiety for the parent. Moreover, the gender ratio in the sample was only balanced for the adolescents but not for parents. We investigated much more mothers than fathers, and it cannot ruled out, that the results would differ, if the ratio was more balanced. Further on, meanwhile the new DSM-5 condition Internet Gaming Disorder was criticized by several researchers (e.g. [Reference Kuss, Griffiths and Pontes18]), and future modifications of the name or the concept are conceivable.
In sum, however, and despite these limitations, this study makes an important contribution to the existing research on Internet gaming disorder. We added some interesting aspects to the currently limited knowledge about the new DSM-5 condition Internet gaming disorder in early adolescence. Generally, longitudinal studies are urgently needed in the future research of Internet gaming disorder. Specifically, the importance of different aspects of adolescent and parental mental health as predictors for the development of Internet gaming disorder should be clarified.
5. Conclusions
We observed relations between several different internalizing and externalizing problems in adolescents and Internet gaming disorder. Additionally, in the present study the importance of parental mental health for Internet gaming disorder was revealed for the first time, even though the relationships were weaker than for certain aspects of adolescent mental health (e.g. hyperactivity/inattention). Based on the findings, adolescent and parental mental health should be taken into consideration when preventing or treating Internet gaming disorder in adolescence.
Disclosure of interest
The authors declare that they have no competing interest.
Acknowledgements
The VEIF study was supported by the German Research Foundation [DFG, grant to the project leader Prof. Kammerl, grant number KA 1611/6-1].
Comments
No Comments have been published for this article.