Sedentary Video Gaming and Body Mass: A Meta-Analysis

Background: Video gaming has been widely discussed as one leisure activity that is positively associated with body mass and overweight (e.g., Borland, 2011; Inchley, Currie, Jewell, Breda, & Barnekow, 2017; Mazur et al., 2018). Empirical findings on the popular form of non-active video games (i.e., games that are played while sitting in front of a screen, sedentary video games), however, have been mixed. While some studies found positive associations between the intensity of playing sedentary games and indicators of overweight, such as the body mass index (BMI; e.g., Martinovic et al., 2015; Siervo, Cameron, Wells, & Lara, 2014), others found no relationships (Bickham, Blood, Walls, Shrier, & Rich, 2013; Scharrer & Zeller, 2014). Objectives and research questions: The current meta-analysis had two goals. First, we wanted to provide an estimate of the average effect size of the relationship between body mass and video gaming that includes recent research from the last one and a half decades, and we acknowledged several context variables to gauge the stability of the average effect. Second, to provide additional evidence on processes, we tested the displacement effect of physical activity by video gaming time with the help of a meta-analytic structural equation model (MASEM; Cheung & Hong, 2017). Method: Meta-Analytic Database: Relevant studies published until June 2018 were identified through databases (PsychINFO, MEDLINE, ProQuest), gray literature (e.g., unpublished reports, conference proceedings, or theses); Google Scholar, and from the references of all relevant articles. This resulted in 753 potentially relevant studies. The studies were included in the meta-analysis if they met the following criteria: The study contained (a) a measure of body mass (i.e., body mass index, body fat percentage, waist circumference, or subscapular skinfold thickness), (b) a measure of video game use (e.g., frequency or duration of video game sessions), (c) data on their zero-order relationship (or respective statistics that could be used to approximate this relationship), and (d) the sample size. After applying all eligibility criteria, 20 publications met our criteria and were included in the meta-analysis. Meta-Analytic Procedure: The meta-analysis was conducted following the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA, Moher et al., 2015) and standard procedures and recommendations for the social and medical sciences (Lipsey & Wilson, 2001). The focal effects concerned the zero-order relationship between video gaming and body mass. For studies that did not report respective correlation coefficients, we extracted any relevant statistic (e.g., odds ratio) that could be transformed into correlation coefficients. Inter-coder reliability between the two coders for the coded effect sizes showed an excellent Krippendorff’s (1970) α = 1.00. The effect sizes were pooled using a random effects model with a restricted maximum likelihood estimator (Viechtbauer, 2005). To account for sampling error, the effect sizes were weighted by the inverse of their variances. Because some studies reported multiple effect sizes for two or more eligible associations (e.g., scores for two video gaming measures were each correlated with BMI), theses dependencies were accounted for by fitting a three-level meta-analysis to the data (Moeyaert et al., 2017; Van den Noortgate, López-López, Marín-Martínez, & Sánchez-Meca, 2013). Analyses of the heterogeneity as well as analyses of possible publication bias were conducted. The meta-analytic models were estimated in R version 3.5.0 using the metafor package version 2.0-0 (Viechtbauer, 2010). Sensitivity analyses and structural equation model: Sensitivity analyses were conducted for (1) publication year, (2) age groups, (3) gender ratio in the sample, (4) a sample-wise estimate of gender differences in body mass, (5) body mass measure, (6) continuous vs. dichotomous body mass measures, and (7) a study quality index. A possible mediating effect of physical activity was examined using MASEM following two steps (see Cheung & Hong, 2017). Results: Across k = 24 samples and 32 effect sizes (total N = 38,097), the mean effect (corrected for sampling error) of the relationship between video gaming and body mass was , 95% CI [.03, .14]. Hence, higher video gaming was positively associated with higher body mass. This relationship was significant, but there remained significant total heterogeneity, Q (31) = 593.03, p < .001, I² = 95.13. In the sensitivity analyses, we found a significant moderation for the age groups; the omnibus test for age was χ2 (df = 2) = 6.56, p = .038. Compared to adults, children and adolescents showed a significantly lower relationship between video gaming and body mass. The estimated mediation model is presented in Figure 1. The relationship between body mass and physical activity was significant with B = -.07, 95% CI [-.14, -.00]. Higher physical activity was associated with lower body mass. The average relationship between video gaming and physical activity was only marginally significant with B = -.08, 95% CI [-0.16, 0.00]. The respective indirect effect was significant B = .01, 95% CI [.00, .02]; it explained 7 percent of the total effect of video gaming on body mass. However, this result should be interpreted with caution because of the small sample of primary studies. Figure 1. Meta-analytic structural equation model. Standardized regression parameters (*p < 05) are presented. Conclusions and implications: This meta-analysis investigated the relationship between non-active (sedentary) video gaming and body mass, contributing to the research base on the behavioral correlates of overweight and obesity. We identified a small significant correlation between video gaming and body mass overall. This relationship was qualified by participants’ age. The focal association was identified for adult samples, but there was no significant association for samples of children or adolescents. Based on a smaller subset of primary studies we found a small indirect effect on body mass, indicating a displacement of physical activity by video gaming. In summary, sedentary video gaming is only weakly associated with overweight and obesity, physical activity might play a mediating role, and the effects vary with participants’ age. References: Bickham, D. S., Blood, E. A., Walls, C. E., Shrier, L. A., & Rich, M. (2013). Characteristics of screen media use associated with higher BMI in young adolescents. Pediatrics, 131, 935-941. doi: 10.1542/peds.2012-1197 Borland, S. (2011). Playing computer games increases obesity risk in teens by making them hungry. Daily Mail. Retrieved from: http://www.dailymail.co.uk/health/article-1389096/Playing-games-encourages-obesity-teens-making-hungry.html Cheung, M. W. L., & Hong, R. Y. (2017). Applications of meta-analytic structural equation modeling in health psychology: Examples, issues, and recommendations. Health Psychology Review, 11, 265-279. doi:10.1080/17437199.2017.134 Inchley, J., Currie, D., Jewell, J., Breda, J., & Barnekow, V. (2017). Adolescent obesity and related behaviours: trends and inequalities in the WHO European Region, 2002–2014. Observations from the Health Behaviour in School-aged Children (HBSC) WHO collaborative cross-national study. Copenhagen, Denmark: World Health Organisation. Krippendorff, K. (1970). Estimating the reliability, systematic error and random error of interval data. Educational and Psychological Measurement, 30, 61-70. doi:10.1177/001316447003000105 Martinovic, M., Belojevic, G., Evans, G. W., Lausevic, D., ... & Boljevic, J. (2015). Prevalence of and contributing factors for overweight and obesity among Montenegrin schoolchildren. The European Journal of Public Health, 25, 833-839. doi:10.1093/eurpub/ckv071 Mazur, A., Caroli, M., Radziewicz-Winnicki, I., .... & Hadjipanayis, A. (2018). Reviewing and addressing the link between mass media and the increase in obesity among European children. Acta Paediatrica, 107, 568-576. doi: 10.1111/apa.14136 Moher, D., Shamseer, L., Clarke, M., Ghersi, D., Liberati, A., Petticrew, M., ... Stewart, L. A. (2015). Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Systematic Reviews, 4, 1. doi:10.1186/2046-4053-4-1 Moeyaert, M., Ugille, M., Beretvas, S. N., Ferron, J., Bunuan, R., & Van den Noortgate, W. (2017). Methods for dealing with multiple outcomes in meta-analysis: a comparison between averaging effect sizes, robust variance estimation and multilevel meta-analysis. International Journal of Social Research Methodology, 20, 559-572. doi:10.1080/13645579.2016.1252189 Scharrer, E., & Zeller, A. (2014). Active and sedentary video game time: Testing associations with adolescents’ BMI. Journal of Media Psychology, 26, 39-49. doi:0.1027/1864-1105/a000109 Siervo, M., Cameron, H., Wells, J. C., & Lara, J. (2014). Frequent video-game playing in young males is associated with central adiposity and high-sugar, low-fibre dietary consumption. Eating and Weight Disorders-Studies on Anorexia, Bulimia and Obesity, 19, 515-520. doi:10.1007/s40519-014-0128-1 Van den Noortgate, W., López-López, J. A., Marín-Martínez, F., & Sánchez-Meca, J. (2013). Three-level meta-analysis of dependent effect sizes. Behavior Research Methods, 45, 576-594. doi:10.3758/s13428-012-0261-6 Viechtbauer, W. (2005). Bias and efficiency of meta-analytic variance estimators in the random-effects model. Journal of Educational and Behavioral Statistics, 30, 261-293. doi:10.3102/10769986030003261 Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of Statistical Software, 36, 1-48. doi:10.18637/jss.v036.i03/