OBJECTIVE: Video game playing may negatively impact youth. However, the existing literature on gaming is inconsistent and often has focused on aggression rather than the health correlates of gaming and the prevalence and correlates of problematic gaming.
METHODS: We anonymously surveyed 4028 adolescents about gaming and reported problems with gaming and other health behaviors. A total of 51.2% of the sample reported gaming (76.3% of boys and 29.2% of girls).
RESULTS: There were no negative health correlates of gaming in boys and lower odds of smoking regularly; however, girls who reported gaming were less likely to report depression and more likely to report getting into serious fights and carrying a weapon to school. Among gamers, 4.9% reported problematic gaming, defined as reporting trying to cut back, experiencing an irresistible urge to play, and experiencing a growing tension that could only be relieved by playing. Boys were more likely to report these problems (5.8%) than girls (3.0%). Correlates of problematic gaming included regular cigarette smoking, drug use, depression, and serious fights. Results suggest that gaming is largely normative in boys and not associated with many health factors. In girls, however, gaming seems to be associated with more externalizing behaviors and fewer internalizing symptoms.
CONCLUSIONS: The prevalence of problematic gaming is low but not insignificant, and problematic gaming may be contained within a larger spectrum of externalizing behaviors. More research is needed to define safe levels of gaming, refine the definition of problematic gaming, and evaluate effective prevention and intervention strategies.
WHAT'S KNOWN ON THIS SUBJECT:
The effects of video-gaming have been found to be mixed. There is little evidence that recreational play has poor consequences; however, problematic gaming has been recognized as a potential area of concern.
WHAT THIS STUDY ADDS:
This study examined the correlates of recreational gaming and revealed few poor consequences but important health consequences of problematic gaming, particularly externalizing behaviors such as smoking, drug use, and fighting.
The popularity of video, computer, online, and virtual reality games has raised concern in both the popular media1 and the research community regarding the potential for negative health effects of gaming, including the potential for addiction.2,–,10 Gaming has been associated with both positive and negative clinical correlates; however, the evidence for a negative impact of gaming has been inconsistent. Experienced gamers may exhibit superior visual, spatial, and attention skills,11,–,13 and video-game formats have been successfully used to deliver health interventions to children and adolescents.14 Depending on the specific game, high levels of motor skill and often problem-solving are required to advance through levels of play.15 Survey research has indicated that children who play video games often do so in social groups, such as with friends or family members,16,17 and that frequency of gaming is positively associated with more peer interaction outside of school,16 although there is not enough evidence to conclude that gaming is a positive contributor to social development in children.18 Research on the association between gaming and aggression, in both laboratory19,–,21 and survey22,–,25 settings, and several meta-analyses,26,27 has generally indicated a lack of a strong association between playing games (even violent ones) and aggressive behavior,18 although some studies have revealed such an association.4,24,28,–,37
Although the research to support concern about gaming as a negative health behavior in general remains controversial, there is clear evidence that some individuals develop a pattern of gaming behavior that is problematic.1,5,6,8,38,–,42 Because video-gaming is a nondrug behavior with hedonic components, problematic gaming might be best conceptualized as a nonsubstance addiction or an impulse-control disorder (ICD). The formal ICDs, such as pathologic gambling, are characterized by appetitive urges or growing tension before participation, relief or pleasure after engagement in the behavior, and repeated behavioral engagement despite negative consequences.43 It is important to note that they are not defined by concern or complaints about the behavior from family, which is important when evaluating behaviors in adolescents who may be in conflict with parents around a wide variety of issues. Problematic gaming has typically been defined in the literature on the basis of measures of pathologic gambling5,6,40,44,45; however, a variety of definitions and criteria have been used, which has led to inconsistent estimates of the prevalence of problematic gaming.
Individuals with problematic gaming behavior have been shown to exhibit inattention, hyperactivity, and poor self-control46,47; experience time distortion while playing48; and demonstrate increased aggression and diminished empathy if their games of choice contain aggressive content.28,–,30,32,37 Those with gaming problems who are engaged primarily in large online gaming communities often are shy,49 have an external locus of control,49 and use gaming to deal with negative emotions.50
Multiple research gaps presently exist. First, not all adolescents play video games, and the health correlates of gaming (eg, relationships with drug and alcohol use or depression) are incompletely understood, because much of the published research has focused on school performance, obesity, and aggression. Second, the likelihood of gaming, and the clinical correlates thereof, are likely to differ by gender, and gender differences in gaming and its health correlates are poorly understood. Third, much of the research on gaming has used online samples of gamers or small groups of individuals identified as having problematic gaming behaviors, which presents problems for generalizability. This study used survey data from a large sample of adolescents and examined the prevalence and clinical correlates of gaming, reported problems associated with video games, and the prevalence and correlates of problematic gaming.
Data were derived from a cross-sectional anonymous survey of students in public high schools, as described previously.51 Schools were first recruited into the study, and then students at each enrolled school were invited to participate. Invitation letters were sent to all public 4-year and nonvocational or special-education high schools in the state of Connecticut. After the initial round of letters were mailed, the response from schools was not yet sufficient to ensure that all geographic regions of the state were sufficiently represented. Therefore, targeted contacts were made to schools that were in areas that would ensure a more representative sample. The final sample was representative of 14- to 18-year-old adolescents in Connecticut according to the most recent US census.51
In most cases, the entire student body was targeted for administration of the survey. Students were told that they could voluntarily refuse to complete the survey if they wished and were reminded to keep surveys anonymous. The survey took ∼50 minutes to complete.
The measures used in this analysis included self-reported gender, race, ethnicity, grade, and family structure (living with 1 parent, living with 2 parents, or other family structure).
Health and functioning measures were categorized as presented in the tables and included grade average; extracurricular activities; lifetime smoking history; lifetime marijuana use; lifetime history of a sip of alcohol; current alcohol use among those with a history of any alcohol consumption, categorized as none, light (1–2 days of drinking in a month), moderate (3–9 days of drinking in a month), and heavy (≥10 days of drinking in a month); lifetime use of other drugs; caffeine use; report of being sad or hopeless for 2 weeks or more in the previous year; getting into fights and requiring medical attention in the previous year; carrying a weapon such as a knife, club, or gun to school in the previous year; and BMI calculated from self-reported height and weight.
Respondents were asked to report how much time they spent playing video or computer games in a typical week. Respondents who reported “none” were classified as non–game players. For those who played games, the frequency of play was categorized as less than 7, 7 to 14, 15 to 20, and 21 or more hours/week.
Those who played any video or computer games were asked if they had ever tried to cut back on playing; whether a family member had expressed concern about the amount of time they spend playing games; whether they missed school, work, or an important social activity because they were playing video or computer games; whether they thought they had a problem with excessive video or computer game use; whether they experienced an irresistible urge or uncontrollable need to play video or computer games; and whether they experienced a growing tension or anxiety that could only be relieved by playing video or computer games. The 3 items of unsuccessfully trying to cut back, experiencing an irresistible urge to play, and experiencing growing tension only relieved by playing were modeled after the Minnesota Impulse Disorder Inventory52 and are considered the core features of an ICD. Students who endorsed all 3 items were categorized as “problematic video gamers.” The coefficient α value for these items was .76.
First, demographic characteristics and health correlates were compared between respondents who played video or computer games and those who did not, stratified according to gender, by using χ2 tests for categorical variables and t tests for continuous variables. Second, logistic regression models were fit to assess the association between each health correlate and playing video games, adjusting for demographic differences. Interaction terms between gender and health correlates determined whether the associations were significantly different for girls and boys.
Next, among the subsample of respondents who reported any game-playing, those with problematic game-playing were compared with those with nonproblematic gaming, stratified according to gender for bivariate analyses. Logistic regression models were fit to examine associations between health correlates and problematic gaming, adjusting for gender and race. There was insufficient statistical power to test gender interactions in multivariable models.
Of 4028 respondents with data on video-gaming, 2064 (51.2%) reported playing at least 1 hour of video games per week. This proportion was higher in boys (76.3%) than in girls (29.2%) (P < .0001) (Table 1). In the total sample, gaming was significantly more prevalent in Asian students and lower grades. Among boys, playing video or computer games was associated with lower grades and with living in a 2-parent household. Among girls, significant positive associations were found with Asian race and lower grades.
Among boys, gaming was associated with higher grade average, never smoking, never having used marijuana, and high caffeine consumption (Table 2). In girls, gaming was associated with occasional smoking, never having used marijuana, never having a sip of alcohol, high caffeine use, no history of depression, getting into serious fights, and carrying a weapon. Gaming was associated with slightly higher BMI in girls (mean [SD] BMI: 22.35 [4.44] for gamers, 21.94 [3.61] for nongamers; P = .03) but not in boys.
Table 3 presents results of adjusted logistic regression analyses with interaction terms to identify significant differences across gender groups. Boys who reported gaming were less likely to be regular smokers, whereas there was no association between smoking and gaming in girls. Boys were also more likely to drink 1 to 2 servings of caffeinated drinks per day, whereas girls who reported gaming were more likely to drink 3 or more caffeine drinks per day. Girls who reported gaming were less likely to report depression, whereas there was no such association among boys. Similarly, girls were more likely to get into serious fights and carry a weapon, but no such association was seen among the boys. Finally, girls who reported gaming had slightly higher average BMI measures (odds ratio [OR]: 1.03; P = .01), whereas there was no association in boys (OR: 1.0001; P = .98).
Table 4 lists the frequency of gaming, along with reported problems associated with gaming, among the sample of 2196 gamers. In the total sample, the majority of respondents reported playing less than 7 hours/week (61.1%); however, 10.9% reported playing 20 hours or more in a typical week. The most commonly endorsed items related to problems with gaming were family members expressing concern about gaming, trying to cut back on gaming, and experiencing an irresistible urge to play. Although the majority of respondents endorsed none of the problematic symptoms, 4.9% endorsed all 3 of the items indicative of an ICD.
There were significant gender differences in patterns of gaming and problems with gaming as well. Girls, compared with boys, more frequently reported playing fewer than 7 hours/week, and 14% of boys reported playing 20 or more hours/week. Girls also reported problematic gaming less often than boys. There were 84 (5.9%) boys who endorsed the 3 problem measures, whereas only 22 (3.0%) girls did so.
Table 5 compares demographic characteristics of those in the problematic gaming group with all others who reported gaming, stratified according to gender. Among boys, problematic gaming was associated with nonwhite and Asian race. There were no significant associations among girls.
Among health correlates (Table 6), problematic gaming was associated with boys who smoked regularly, were depressed, and got into serious fights or carried a weapon. Among girls, problematic gaming was associated with other drug use, depression, and serious fights. There was no association with BMI in either boys (P = .20) or girls (P = .33).
Logistic regression models presented in Table 7 indicated that, adjusted for race and gender, problematic gaming was associated with higher odds of smoking regularly, other drug use, lower caffeine consumption, depression, serious fights, and carrying a weapon to school. However, given effect sizes and the number of models, the most robust of these findings are for an increase in regular smoking (OR: 2.12; P = .007), depression (OR: 3.62; P < .0001), and serious fights (OR: 2.97; P < .0001).
This study is among the first and largest to examine clinical correlates of video-gaming and problematic gaming in a community sample of adolescents. We found that approximately half of the students reported gaming, concentrated among younger students and more common in boys.
There were no significant negative health correlates of gaming in boys, which likely reflects the popularity and normative nature of such games for this group. In addition, boys who reported gaming were significantly less likely to report being a regular smoker. However, among the girls, gaming was associated with modestly lower risk of depression and moderate increases in serious fights (OR: 1.7) and carrying a weapon (OR: 2.1).
We also found that among boys who reported gaming, 5.9% endorsed problematic gaming, compared with 3.0% of girls who reported gaming, which suggests that male gamers may be at higher risk for developing a gaming problem but that overall the risk of developing a problem is relatively low. Problematic gaming was associated with some important risk behaviors (moderate effect sizes were found for depression [OR: 3.6] and fighting [OR: 3.0]) but was not associated with grade averages, extracurricular activities, marijuana use, or alcohol use.
Gender Differences in Correlates of Gaming
The gender differences observed between gamers and nongamers, coupled with the contrast in frequencies of gaming across gender, are suggestive of a gender-specific self-selection process. That is, whereas gaming may be more appealing to boys in general, it may be particularly attractive to girls with certain characteristics. Although causality cannot be examined in these cross-sectional data and competing hypotheses cannot be eliminated, this finding may suggest not that gaming leads to aggression but that more aggressive girls are attracted to gaming as a recreational activity. The finding also may reflect cultural differences in socioeconomic conditions at home and in their communities: it is possible that girls who live in more violent neighborhoods, where they are more likely to experience fights and to carry a weapon for protection, also are more likely to be attracted to gaming, may be more likely to spend time with male peers who are themselves gaming, or may prefer to stay home and play games rather than be outside in dangerous communities.
The finding also may be reflective of personality characteristics that are more externalizing; girls who reported gaming were not only more likely to get into fights and carry weapons, behaviors generally considered to be externalizing, but also were somewhat less likely to report feelings of depression, which would reflect more internalizing patterns. An additional possibility is that gaming may exert a positive effect on mood in girls; however, this hypothesis and the precise nature of the relationship warrant additional study in longitudinal investigations.
Among boys, the finding that gaming is associated with more prosocial or beneficial behaviors, such as less cigarette use, may reflect a different peer group among gamers. As noted above, the general lack of associations between health behaviors and gaming in boys also may reflect the normative nature of gaming for boys in the current US culture.
Although there are no uniformly agreed-on thresholds for excessive game playing, we found higher frequencies of playing among boys than girls, and ∼14% of the boys reported playing an average of 3 hours/day or more. This, coupled with the possibility that frequencies may be underreported because of the known phenomenon of time lapse,48 in which gamers seem unaware of how much time has elapsed while playing, may reflect a strong appeal of these games, particularly to boys. However, it also suggests a need to further characterize the potential risks and benefits associated with high-frequency play and to balance such risks and benefits against those of other activities favored by adolescents, including watching television or experimenting with substances.53
Boys were significantly more likely than girls to report problems with gaming. When examining the 3 measures of an ICD, a relatively low but important proportion of both boys and girls have gaming problems. Additional research is needed to examine the accuracy of self-reports of such problems and whether these or other questions are the best way to assess impairment related to gaming. It is possible, for example, that some of the items may be endorsed as a “badge of honor” (eg, claiming to have a problem or an irresistible urge to play as a sign of a serious or experienced player).
The results of this study suggest that adolescents with problematic gaming are more likely to also be engaging in other risk behaviors such as smoking, drug use, and violence and are more likely to report depression. It is not possible with these data to determine whether problematic gaming leads to experimentation, aggression, or depression or vice versa, or the factors develop in conjunction, perhaps related to common etiologic factors such as the violence content in games or other common underlying traits such as introversion/extraversion, impulsivity, or sensation-seeking. Future longitudinal research is needed to examine the onset of risk behaviors in temporal relation to gaming and their potential roles in the development of health problems.
Those who reported problems with gaming also were significantly more likely to report depression. Additional research is needed to understand the role that video games play in brain function,54,–,56 including those pathways also associated with depression. However, some research has suggested that adolescents who play excessive amounts of video games, in part, do so to deal with negative affect.50,57 Conversely, excessive playing may alter brain function in such a way as to increase depressed affect and the risk of depression.
Strengths and Limitations
This study is among the largest to examine gaming in high school students, and this sample size allowed for the examination of gender differences in health correlates. Data were examined by using regression models in a manner consistent with previous work that investigated youth impulse-control behaviors, which allows for comparability across studies.58 However, the data are cross-sectional, so that temporal associations cannot be elucidated, and this limits our ability to suggest causal pathways or theoretical models for problematic gaming. Also, the data were self-reported and the psychometric properties of the questions have not been directly evaluated, although they were based on validated questions used to assess other impulsive behaviors in the Minnesota Impulse Disorder Inventory.52 Additional research is needed to determine whether these questions are accurate, reliable, and appropriate for assessing problem gaming. We selected a relatively stringent threshold for defining problematic gaming, and additional research is needed to determine the optimal threshold for this definition. Despite a substantial sample size, the low prevalence of problematic gaming did not allow us to fully investigate gender interactions in multivariable models. Given the suggestion of differences in bivariate analyses, future research should examine potential gender differences in the correlates of problematic gaming. In addition, other factors, such as depression, that are found in association with other impulse-control behaviors, may represent important variables of consequence (perhaps particularly for girls59,60) in the development of problematic video-gaming, and future studies should investigate directly and systematically the relationships between gender, mood, and problematic video-gaming. Additional factors such as temperament warrant similar consideration as potential vulnerability factors.61 Such analyses could help identify how additional factors might be mediating the relationship between problem video-game playing and negative measures of health and functioning. Finally, we were unable to distinguish between computer games, handheld video games, physically active games such as Wii, and online gaming. It is possible that each of these types of games attracts different types of players and correlates differentially with specific health measures, particularly because the types of games played may be directly associated with socioeconomic status, which we were unable to assess reliably.
We found no significant negative health correlates of video game use in boys, modest correlations with higher aggression and reduced depression in girls, a relatively low but important percentage of reported problems with gaming, and important associations between problematic gaming and smoking, drug use, aggressive behavior and depression, although there were no associations with grade averages, extracurricular activities, marijuana use, or alcohol use. Additional research is needed to examine recreational and problematic levels of video-gaming, to determine safe levels of gaming, and to identify risk factors and potential points of intervention and prevention. In addition, more research is needed into beneficial uses of video games given their popularity among youth.62
This study was supported by National Institutes of Health grants PSOAA15632, RLl AA017539, and ULl DE19586; the National Institutes of Health Roadmap for Medical Research/Common Fund; and the National Center for Responsible Gaming and its affiliated Institute for Research on Gambling Disorders and the State of Connecticut, Department of Mental Health and Addiction Services. This work was also supported by the Yale University Transdisciplinary Tobacco Use Research Center and the Yale University Psychotherapy Development Research Center.
- Accepted August 12, 2010.
- Address correspondence to Rani A. Desai, PhD, MPH, NEPEC/182, VA Connecticut Healthcare System, 950 Campbell Ave, West Haven, CT 06546. E-mail:
The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of any of the funding agencies.
FINANCIAL DISCLOSURE: Over the past 3 years, Dr Potenza has received financial support or compensation as a consultant and advisor for Boehringer Ingelheim, has had financial interests in Somaxon, has received research support from the National Institutes of Health (NIH), the Veteran's Administration, the Mohegan Sun Casino, the National Center for Responsible Gaming and its affiliated Institute for Research on Gambling Disorders, and Forest Laboratories pharmaceuticals, has participated in surveys, mailings, or telephone consultations related to drug addiction, ICDs, or other health topics, has consulted for law offices on issues related to addictions or ICDs, has provided clinical care in the Connecticut Department of Mental Health and Addiction Services Problem Gambling Services Program, has performed grant reviews for the NIH and other agencies, has guest-edited sections of journals, has given academic lectures in grand rounds, continuing medical education events, and other clinical or scientific venues, and has generated books or book chapters for publishers of mental health texts; the other authors have indicated they have no financial relationships relevant to this article to disclose.
- ICD =
- impulse-control disorder •
- OR =
- odds ratio
- Feng J,
- Spence I,
- Pratt J
- Ferguson CJ,
- Rueda SM,
- Cruz AM,
- Ferguson DE,
- Fritz S,
- Smith SM
- Savage J,
- Yancey C
- Anderson CA,
- Bushman BJ
- Anderson CA,
- Sakamoto A,
- Gentile DA,
- et al
- Kutner LA,
- Olson CK
- 43.↵American Psychiatric Association. Impulse-control disorders not elsewhere classified. In: Diagnostic and Statsitical Manual of Mental Disorders. 4th ed.Washington, DC: American Psychiatric Association; 1994:609–621
- Liu TC,
- Desai RA,
- Krishnan-Sarin S,
- Cavallo DA,
- Potenza MN
- Copyright © 2010 by the American Academy of Pediatrics