The study used a case-control design and was conducted during the period of October and November 2019 in a medical college in the Trichy District of Tamil Nadu, India. Ethical clearance was obtained from the Institutional Ethical Committee of Trichy SRM Medical College (1007/TSRMMCH&RC/ME-1/2019-IEC no:039). Informed written consent was obtained from all the participants. If the enrolled participants were not interviewed on a specified date, they were interviewed subsequently at a time and place of their convenience. The purpose of the study was explained in detail and assured that the data collected would be used only for scientific purposes. Ethical principles such as respect for the person and confidentiality of their data were strictly adhered to.

A total of 448 undergraduate medical students in their first to prefinal year were included as participants using the universal sampling technique (Figure 1). The study preliminarily surveyed the entire study population of 448 students using personal interviews. All 448 participants were asked only 1 question: “Have you been gaming in the last 6 months?” From this preliminary sample of 448 surveyed students, 91 students replied affirmatively and were recruited as cases in the study. Following this, the investigator used purposive sampling to select 91 controls from the remaining 357 students who had never indulged in gaming in the last 6 months, and matched both cases and controls for age and gender. The controls selected were matched for age and sex using a 1:1 ratio. The frequency of gaming hours per week was assessed among the cases. The internal assessment scores of the two midterm examinations held in the last 6 months were accessed from the students’ records kept by the institution after obtaining written permission from the students and the Institutional Ethical Committee. The internal assessment scores of cases and controls were then compared and recorded in percentages.

To assess the prevalence of gaming disorder, the Gaming Addiction Scale (GAS) by Lemmens et al [8] was used. The GAS is a pretested, prevalidated scale with a Cronbach alpha of .82 to .87 [8]. It has 7 items: salience, tolerance, mood modification, relapse, withdrawal, conflict, and problems. Each item has three questions with a score range of 0 to 5 with all the components scored on a Likert scale: 1=never, 2=rarely, 3=sometimes, 4=often, and 5=very often. The investigators used the monothetic format in the study, that is, a score of >3 for all items being indicative of gaming addiction. Lemmens himself hypothesized that the monothetic format would lead to a better estimate of the prevalence of addiction than the polythetic format [8]. Therefore, the investigators used the GAS according to protocol, but for the convenience of analysis, the investigators summed up the total score of all 7 items, and classified participants with a score of ≥63 as having a gaming disorder.

The data entry and analysis were done using SPSS software (version 21, IBM Corp). Descriptive statistics were used for analyzing sociodemographic details, frequency, and type of gaming. The Snedecor F test and the Student t test were used to determine the association between the hours spent gaming and scholastic performance, and gaming and internal assessment scores, respectively. Odds ratios were used to calculate the risk of low internal assessment scores among cases and controls. The correlation coefficient (r) was used to determine the correlation between the GAS scores and internal assessment scores.

To the best of our knowledge, this study is the first to use a case-control design to examine the association between gaming and scholastic performance in medical students. Since the availability of literature on internet gaming among medical students is sparse, it is difficult to draw meaningful comparisons.

This study observed that smartphones were the most commonly used gaming platform by medical students. The study observed no significant association between the frequency of gaming and internal assessment scores. Gamers (cases) showed a significantly lower score than nongamers (controls). Male gamers showed significantly lower scores compared to male nongamers, whereas the difference between scores of female gamers and nongamers was not statistically significant. The study found a negative correlation between GAS scores and internal assessment scores. Further, there was a higher risk of lower scores among those who played games compared to those who did not.

There was no significant differences observed in the internal assessment scores and the number of hours spent gaming. This finding differs from a study by Ip et al [9], where frequent gamers (both males and females) scored less than nonfrequent gamers in examinations, with the average grades of nongamers being 9.4% higher than those of frequent gamers. A study conducted by Dumrique and Castillo [10] observed no significant relationships between the number of hours of playing and the social behavior of the respondents. The reason for this difference may be because they included assessments from the whole academic year, whereas we have included only assessments from the last 6 months. In addition, the scale of measurements differs between those studies and our study.

In this study, the mean scores of the first and second midterms of those who played games were 5.2% and 4.1% lower than those who did not play games, respectively. Male nongamers had 9.5% and 8.4% higher scores than male gamers for the first and second assessments, respectively. This is somewhat similar to the finding of Ip et al [9], where the examination grades of infrequent male gamers were on average 7.2% higher compared to regular male gamers. In our study, we observed no significant difference in internal assessment scores between female gamers and female nongamers. We also observed that female gamers had higher internal assessment scores compared to male gamers despite having higher GAS scores. This indicates that although there is a greater incidence of gaming disorder among females, this is not associated with poor scholastic performance. This is similar to the findings of Ip et al [9] on gaming frequency and academic performance, where female students performed better than male students in all disciplines even though they were gaming. Contrary to our finding was Dumrique and Castillo’s [10] observation that the academic performance of students was not affected even if they played online games. This difference is because their participants had better self-control, played games preferably during the weekends, and socialized more. This finding is useful in the context of gaming as a leisure activity that is not done in excess.

In this study, gaming addiction, as assessed by the GAS, was found to be prevalent in 6.2% of the study population and 31% of those who played games. The prevalence of gaming disorder using different scales in various prior studies ranged from 2.0% to 22.7% [8,11-23]. This variation may be due to differences in study populations and measurement scales used.

In terms of specific studies, Lemmens et al [8] found the prevalence of the gaming addiction to be 2.3% using the monothetic format and 9.3% using the polythetic format [8]. Mentzoni et al [24], who used the GAS, observed the prevalence of problematic users (score of ≥4 out of 7 on the GAS) to be 4.1%. Wang et al [21] in Hong Kong identified 15.6% of study participants as having a gaming addiction. In a study conducted in Germany by Festl et al [25], 3.7% of the respondents were considered to be problematic gamers.

We found a negative correlation between GAS scores and mean internal assessment scores—greater gaming disorder scores were associated with lower internal assessment scores, emphasizing the fact that gaming negatively affects scholastic performance. A review by Mihara and Higuchi [26] showed that many studies reported lower grades and career attainment in students indulging in excessive gaming.

Our novel study quantifies the risk of poor scholastic scores associated with excessive gaming, with gamers at higher risk than nongamers (odds ratio 1.9, 95% CI 1.04-3.44). This finding is useful in the context of restricting gaming as a leisure activity than indulging in it excessively. This observation also helps in the early identification and treatment of students who are gaming excessively to prevent poor academic performance.

The study comes with the inherent limitations of the case-control design. The retrospective nature of the study can be used to establish an association between gaming and scholastic performance, but cannot establish causation. Additionally, it should be noted that cases and controls were matched only for age and gender since matching for other potential confounders would have led to overmatching and fewer control participants. Further, the findings of this study pertain to a single educational setting, which could limit its generalizability.

We conclude that gaming adversely affects scholastic performance among male students compared to female students. Awareness needs to be created among medical students about the negative effects of gaming, which can have a detrimental effect on their scholastic performance. Students, parents, teachers, and institutions should be advised on the early detection of gaming disorder. Treatment services should be made available to those with gaming disorder in medical institutions. The study also opens new avenues for further exploration in different educational settings using a cohort study design to examine the long-term impact of gaming on the scholastic performance of students.