Prevalence and Correlates of Internet Gaming Disorder Among an Advanced Level Student Population

. 1 Background: Internet Gaming Disorder (IGD) is on the rise in many low and middle-income countries owing 3 to the increasing popularity of electronic gaming and technology availability among adolescents. Despite this, 4 the epidemiology of IGD in South Asia remains largely unknown. Hence, we aimed to determine its 5 prevalence and associations including motivations for gaming, among a Sri Lankan school-going population. Methods: A cross-sectional study was conducted among Advanced Level students aged 16-18 years, 8 attending four schools in the Colombo Educational Zone. Of the 412 randomly sampled recruits, 395 9 consenting participants filled a pretested questionnaire exploring gaming habits, motivations, and 10 psychosocial factors. English and Sinhala versions of the Internet Gaming Disorder Scale Short-Form 11 (IGDS9-SF) were used and IGD was identified if ≥5 items in the scale were endorsed. The Sinhala IGDS9-SF 12 demonstrated acceptable internal consistency reliability and its factorial validity was affirmed via Confirmatory 13 Factor Analysis. Results: 81.5% (n=322) of the sample consisted of gamers, a majority of whom preferred mobile gaming 16 (64.0%) and Multiplayer Battle Royale games (27.0%). The prevalence of IGD was 5.06% (95% CI = 2.90– 17 7.22) and was significantly higher (p<0.05) among males who constituted 52.4% of the sample. Daily weekday 18 gaming for ≥6h, low involvement in student societies, poor relationship with parents, Escape motive and 19 Fantasy motive were positively associated, and the Competition motive was negatively associated with IGD in 20 multivariable analyses done using binary logistic regression. Conclusion: The prevalence of IGD was considerably high in our population and is associated with specific motives, poor parent-child relationship, and low extra-curricular involvement.


2
Electronic gaming is a healthy recreational activity which confers several cognitive, motivational, emotional, 3 and social benefits and is gaining popularity as a source of entertainment among adolescents. 1 Pathological 4 gaming however, is a type of maladaptive behavior and is described as "Internet Gaming Disorder" (IGD) 5 under Section III of the Diagnostic and Statistical Manual of Mental Disorders, 5 th Edition (DSM-5), where it is 6 defined as a pattern of persistent and recurrent participation in electronic gaming leading to significant 7 impairment or distress as indicated by five or more of nine core symptoms exhibited over a year. The 8 specified symptoms include pre-occupation with games, withdrawal, tolerance, lack of control, loss of interest 9 in previous hobbies, continued excessive use despite knowledge of psychosocial problems, deception of 10 family members and/or others, use of games to gain relief from negative emotions and jeopardizing important 11 relationships or opportunities due to gaming behavior. The condition was also recognized by the World Health 12 Organization and described as "Gaming Disorder" in the latest International Classification of Diseases (ICD-13 11). 2 Despite its name, IGD applies to both online and offline gaming activity. 3 It has to be noted that Internet  The global pooled prevalence of IGD is 3.05% 27 and ranges from 0.21-57.5%. 28 IGD is predominantly seen in 31 males and is higher among adults than younger adolescents, although the prevalence decreases with 32 increasing age among adults. 29 The prevalence is lower among European populations compared to South-

33
East Asian populations, with rates as low as 1.6% reported from representative European samples 17 and as 34 high as 10.1% in South-East Asia. 30 The prevalence of IGD was 3.5% in a study done among a sample of 15-

35
19-year-old school-going adolescents 31 and 3.6% among a sample of medical students from India. 32

36
The literature pertaining to the psychological aspects of electronic gaming is scarce in the South Asian region.

37
Evidence from Sri Lanka has hitherto been limited to case reports such as that done by Chandradasa and 38 Rodrigo in 2017, where IGD was described in four adolescents aged 13-16 years presenting to a Sri Lankan 39 tertiary care hospital. 33 In this light, our aim is to address this regional evidence gap by investigating the 40 prevalence of IGD in a Sri Lankan school-going adolescent population, using a simple validated screening tool 6 IJMS which can be used in native settings. We also aim to describe the role of interindividual and student-specific 1 factors as well as the role of gaming motivations which have not been adequately explored in the extant 2 literature.

3
A descriptive cross-sectional study with an analytical component was conducted among four conveniently 4 selected national schools located in Colombo, which is the commercial capital of Sri Lanka. Being national 5 schools, the students hail from around the country and therefore the population is diverse and representative 6 of the multi-ethnic demographic of Sri Lanka. Ethics approval was obtained from the Ethics Review

10
Sample Size 11 A sample size of 355 was deemed necessary at a confidence level of 95% and 5% margin of error for a finite 12 population of 4500 students. An estimated proportion of 50% was considered in the calculation to arrive at a 13 maximum sample size. 34

15
Participants 16 Sri Lankan school education is offered in four stages which are namely: Primary (Grades 1-5), Junior 17 Secondary (Grades 6-9), Senior Secondary (Grades 10-11) and Collegiate or Advanced Level (A/L) (Grades  incompletely/incoherently. Based on these criteria, 17 participants were excluded due to absence (n=4), 26 failure to return the consent and assent forms (n=2) and incompletely/incoherently filled questionnaires 27 (n=11), with 395 students finally included in the study, resulting in a response rate of 95.87%.

29
Study Procedure 30 Data collection was done throughout September 2020. Permission was initially obtained from the Zonal 31 Education Director and Principals of the selected schools, prior to approaching the selected classes for data 32 collection. The investigators approached the students during the free timeslots in their academic timetables.

33
The study's objectives and procedure were described before distributing the consent forms which were to be 34 signed by the parents. Informed assent was additionally obtained by the students. Students who returned the 35 completed consent and assent forms were given the questionnaire in their preferred language 36 (English/Sinhala). Personal details such as name and residential address were not requested to ensure 37 anonymity. The investigators were present at the time of data collection to answer any queries raised by the 38 participants, although the participants were not allowed to discuss the contents of the questionnaire amongst 39 themselves in order to prevent information bias.

1
The study employed a self-administered questionnaire, which was pre-tested among 32 A/L students from the Involvement in online learning activities, positive attitude regarding parent-child and teacher-student 7 relationships, lack of satisfaction with one's physical appearance and the presence of few/no friends were 8 assessed by single items graded on a five-point Likert scale, with higher values indicating greater levels of 9 agreement. Self-esteem was assessed with a Single-Item Self-esteem Scale which was graded on a similar 10 five-point scale. 35 Gaming habits and preferences including the preferred genre and device were also 11 explored in this section. The game that was played for the longest duration over the past year was used to 12 identify the preferred gaming genre.

24
Statistical Analysis

25
The internal consistency reliability of all scales was assessed by the Cronbach's alpha test and α ≥0.7 was 26 considered acceptable. Factorial validity of the Sinhala IGDS9-SF scale was assessed via Factor Analyses.

27
Mean, Standard deviation (SD), Counts and Proportions were used to describe the general characteristics of 28 the sample.

29
Bivariate analyses were conducted using chi-square and fisher's exact tests to identify associations of IGD 30 status. A p value of <0.05 was considered statistically significant at a confidence level of 95%. The Shapiro-

31
Wilk and Kolmogorov-Smirnov tests revealed a non-normal distribution of the IGDS9-SF scores and so, 32 correlational analyses of the IGD score and MOGQ subscale scores were done using Spearman's Rank(rs) 33 coefficients.

34
Multivariable analyses were then conducted using binary logistic regression with the IGD status as the 35 dependent variable. All significant bivariate associations were initially entered into the initial model and a 36 forward method of entry was used to obtain the final model which retained only significant variables. The 37 continuous variables were tested via the Box-Tidwell test a priori to verify the linearity assumption.

8
The Factorial Validity of the Sinhala IGDS9-SF scale was assessed using factor analyses. The adequacy of 9 the sample size was initially verified with the Keiser-Meyer-Olkin measurement which was acceptable at 10 0.886. The Bartlett's Test of Sphericity was significant (χ 2 =412, df=36, p<0.001). The Exploratory Factor 11 Analysis done with a Varimax rotation on the nine items revealed a one-factor solution with an Eigenvalue 12 threshold of one. This was further affirmed by examination of the Scree plot. The unidimensional structure 13 was further assessed by Confirmatory Factor Analysis. The test for exact fit was significant (χ 2 =44.9, df=27, unidimensional construct, like its validated English counterpart.

19
Descriptive Analysis of Sample Characteristics 20 The general and gaming-related characteristics of the sample are described in

26
Of the 395 participants, 322 identified themselves as "gamers" (81.5%). A gamer was defined as an individual 27 who had intentionally played electronic games for any duration of time, offline/online, using any device during 28 the past 12 months. A majority played games for ≤1h every day during the weekdays (53.4%, n=172) and 29 weekends (40.7%, n=131). Most of the students started to play games when they were 8-10 years old (24.8%,

34
Prevalence of IGD and Bivariate Analyses

8
All correlations between the IGDS9-SF score and the MOGQ subscales were significant (p<0.001). The

18
Gaming daily for six hours or more during the weekdays was associated with almost seven-fold higher odds of

2
To our knowledge, this is the first study which describes the prevalence of IGD in a Sri Lankan adolescent 3 population and is one of few conducted in the South Asian region. In addition to the disease burden, we have 4 discussed IGD from the perspective of school-going students by focusing on interindividual factors such as 5 the parent-child, teacher-child relationships, as well as the role of academic and extra-curricular activities 6 which are unique to this population and hitherto overlooked by researchers. Understanding the interplay 7 between motivation and addiction enables early identification of at-risk gamers at a community-level.

8
However, previous studies have described these motivations in relation to adult populations, and so, our study 9 furthers the field by describing the specific motivations that drive gaming addiction among an adolescent 10 population. Furthermore, the Sinhala IGDS9-SF which was translated from the original English version, was 11 found to be a potentially valid and reliable tool which can be used among native speakers in future studies.

12
The prevalence of IGD in our study was 5.06%, which is greater than the global pooled prevalence of 13 3.05%, 27 but is lower than the pooled South-East Asian prevalence of 10.1%. 30 It is also higher than the 14 prevalence reported from India, i.e., 3.5-3.6%, 31,32 although this may be explained by the strict diagnostic 15 criteria adopted in these studies. Our prevalence resembles that among Thai high-school students which is 16 5.4%. 37

17
Male gamers were significantly more likely to have IGD than females which is consistent with the extant 18 research. 29 This is partly explained by the fact that males spend more time gaming and prefer high risk genres, whereas most female gamers favor low-risk casual games. Maladaptive cognitions such as the 20 overvaluation of virtual rewards may also explain the higher prevalence of IGD among males. 38 Of the 21 academic characteristics explored, the only significant association was the language of study, with Sinhala medium students being more likely to have IGD. Poor sleep is a known association of IGD. 20 In our study 23 however, reduced sleep hours was not linked to IGD. In fact, some studies report higher sleep hours among 24 IGD gamers in high-school student populations. 37 This could be explained by the regular sleep schedules 25 students are compelled to adhere to, owing to early school commencement hours. Although low exercise is 26 cited as a risk factor, 22 a low involvement in sports was not associated with IGD in our study, which may be 27 explained by the lack of opportunities for sports activities owing to lockdowns amid the COVID-19 pandemic.

28
A lack of interest in recreational activities is a known correlate 17 and our findings reflect this as IGD gamers 29 were more likely to report low involvement in student societies. Disordered gamers also had poor inter-30 personal relationships with their parents and teachers. These interindividual factors which are associated with 31 Internet Addiction, 11 appear to be linked to IGD as well. We did not find the number of friends to be a 32 significant association which supports the notion that despite real-life social exclusion, online gaming can in 33 fact facilitate social interactions, 39 with IGD gamers reporting more friends than non-IGD counterparts in some 34 studies. 16 We also found that IGD gamers were significantly more likely to be dissatisfied with their self-35 appearance which may explain the tendency to idolize in-game avatars with idealized physical forms. 40

36
Although low self-esteem is commonly described among IGD gamers, 12,13 this was not a significant 37 association in our study, and so, it is conceivable that the feeling of virtual achievement may have a positive 38 influence on a gamer's self-esteem.
Daily weekday gaming for ≥6h daily was retained as a significant association in multivariable analyses, IJMS tantamount to disordered gaming, with certain studies failing to show increased gameplay as a significant 1 association. 12 Although, MMORPG and FPS games are traditionally implicated with disordered gaming, 18 in 2 our study, most of the gamers (27.0%) preferred Multiplayer Battle Royale games such as "PUBG" and 3 "Fortnite" which was also linked to IGD. Most gamers in our study (64.0%) preferred mobile phones among 4 traditional gaming devices such as personal computers and consoles. 20 The mobile phone is indeed an 5 emerging gaming device in middle-income countries like Sri Lanka with the number of mobile connections 6 increasing by 2.2 million from January 2019 to January 2020, 41 although a preference for mobile gaming per 7 se was not associated with disordered gaming.

8
Escape and Fantasy were positively predictive of IGD, which echoes extant work on gaming motivations. 12 9 However, Competitiveness had a negative effect in our model, which is in fact described as a positive 10 predictor among women in the cited study. Escape refers to "escaping from problems in the real world" and 11 Fantasy entails "stepping out of one's usual identity, trying new identities in a different fantasy world", while 12 Competition represents the motive of "competing with and defeating others to feel a sense of achievement". 26

14
Limitations 15 Our findings should be interpreted with caution in light of certain limitations. First, we acknowledge the limited 16 generalizability of our results which may be applicable to the four selected national schools. The cross-17 sectional design of the study does not allow temporal interpretation of associations as true predictors or 18 consequences of disordered gaming. IGD was diagnosed using the IGDS9-SF scale, which is a validated 19 psychometric tool. This is less accurate than the structured clinical interview which is the gold standard of 20 diagnosis. We also admit the possibility of recall bias and reduced answering accuracy as participants were 21 expected to report gaming practices over a period of 12 months. Furthermore, the study was conducted amid 22 the COVID-19 pandemic and so the data presented here may not reflect the "pre-COVID normal" where 23 students were not compelled to follow online lessons owing to countrywide lockdowns. Despite these 24 limitations, our study has set an important precedent for future studies both nationally and regionally.

27
Gaming addiction is an emerging public health challenge in our student population. The Sinhala IGDS9-SF 28 was found to be a potentially valid and reliable tool. IGD was significantly higher among male gamers and was 29 associated with a poor parent-child relationship, low extra-curricular involvement, increased daily weekday 30 gaming and specific gaming motives. These associations should be considered when planning preventive 31 interventions and awareness programs targeting high-risk groups. The inclusion of gaming motivations in the 32 psychological assessment of at-risk gamers will facilitate screening efforts, as gaming time alone is insufficient 33 in distinguishing disordered gaming from non-pathological high involvement. IGD should be considered as a 34 serious public health problem that is emergent in school-going adolescent populations and we recommend 35 further studies that explore this condition in nationally/ regionally representative samples.

2
These questions will ask you about your gaming activity during the past year (i.e., last 12 months). By gaming 3 activity, we understand any gaming-related activity that has been played either from a computer/laptop or from 4 a gaming console or any other kind of device (e.g., mobile phone, tablet, etc.) both online and/or offline. There 5 are no right or wrong answers.

7
Read each of the given statements and put an "X" on the appropriate number.

2.
Do you feel more irritability, anxiety or even sadness when you try to either reduce or stop your gaming activity? 1 2 3 4 5

3.
Do you feel the need to spend increasing amounts of time engaged gaming in order to achieve satisfaction or pleasure? 1 2 3 4 5

4.
Do you systematically fail when trying to control or stop your gaming activity? 1 2 3 4 5

5.
Have you lost interests in previous hobbies and other entertainment activities as a result of your engagement with the game? 1 2 3 4 5

6.
Have you continued your gaming activity despite knowing it was causing problems between you and other people? 1 2 3 4 5

7.
Have you deceived any of your family members or others because of the amount of your gaming activity? 1 2 3 4 5

9.
Have you jeopardized or lost an important relationship, job or an educational opportunity because of your gaming activity?