The Long-Term Impact of Digital Media on Brain Development in Children ====================================================================== * Samson Nivins * Bruno Sauce * Magnus Liebherr * Nicholas Judd * Torkel Klingberg ## Abstract Digital media takes an increasingly large part of children’s time, however, the long-term effect on brain development is unclear. Here, we investigated the individual effects of digital media use (watching television and videos, using social media, or playing video games, respectively), while controlling for genetic predisposition and socioeconomic status. We included 4502 children with two MR-scans, aged 9-11 years at baseline, from the Adolescent Brain and Cognitive Development Study. Children, on average, spent 2.2h/day watching videos, 1.2h/day using social media, and 1.2h/day playing video games during these two years. At the two-year follow-up MR-scan, digital media use had no effect on the change in the cortical surface or volume of the cerebellum. However, girls, but not boys, who spent more time playing video games showed a smaller change in the volumes of the striatum (β = +0.07 (+0.11 – +0.02); *P*=0.03), and there was a significant interaction effect of sex and playing video games (*P*<0.001). Overall, our findings suggest that digital media usage, particularly playing video games, does not affect cortical brain development but has a weak effect on the development of the striatum in a sex-specific manner in children aged 9-11 years. More studies, with longer follow-ups and studies of other age groups, are warranted. Keywords * Videogames * gaming * digital media use * MRI * brain * children * polygenic score ## Introduction Children are spending more time with digital media than ever before; however, the effects of this on psychological health, cognition, and brain maturation are still unclear. There are growing concerns among parents, caregivers, and policymakers about the negative impact of digital media use on the developing brain of children. Therefore, studies focusing on the long-term effect of digital media use on children are warranted. In 2016, the American Academy of Paediatrics recommended healthy media use (i.e., less than 2h/day) for children aged 6-10 years. However, children of this age range do not meet this recommendation globally. For example, in the USA, children aged 8-12 years on average spent 4h and 44mins per day on digital media for entertainment purposes1, aside from use during school and homework. Behavioural studies have reported benefits along with adverse consequences to the amount of time spent on digital media activities amongst school-aged children2–5. For example, a cross-sectional analysis of the Adolescent Brain Cognitive Development (ABCD)-cohort suggested that longer exposure to digital media use was associated with poor mental health, lower academic performance, and behavioural problems in a US sample of children aged 9-11 years5. However, a longitudinal analysis of our group showed that there is significant self-selection for the usage of different types of digital media use, and that there is a positive effect of playing video games (but not watching television and videos, or social media use) on intelligence after two years, when accounting for both socioeconomic status (SES) and genetics4. Cross-sectional neuroimaging studies have associated digital media use mainly with smaller grey matter volumes in various cortical regions6–9. However, these studies cannot account for possible self-selection. SES and genetics need to be considered, as childhood brain development could be affected by environmental characteristics such as upbringing, parental education, parental income, and genetics10. Short- and long-term effects of video game training have been investigated using structural and functional magnetic resonance imaging (MRI) on brain regions in young and healthy adults11–13, and reported brain changes mainly in the prefrontal and parietal cortex and hippocampus. However, in children, these findings might differ because of the critical phase of brain development. To our knowledge, only one longitudinal study has investigated the associations between weekly videogame use and brain changes one-year later in children aged 7-11 years and found significant changes in functional connectivity between the dorsal putamen and motor cortex14, but did not control for SES or genetic effects. To overcome the above limitations, here we aimed to investigate the cumulative effect of digital media use on brain development over two years in children aged 9-11 years at baseline (T), accounting for age, sex, handedness (to avoid lateralization of the brain structures associated with handedness), SES, scanner sites, and genetics. Our main outcome measure was a global cortical surface area, which is known to be the measure most highly correlated to intelligence in both children and young adults15, 16. We used cortical surface area rather than cortical thickness because studies have consistently found an effect of environmental variables, such as SES, on surface area rather than the thickness17. As secondary outcomes, we further investigated the individual brain regions, i.e., the volumes of the striatum and cerebellum, which have been implicated in cross-sectional studies of digital media use13, 14. Based on our existing findings4, we hypothesised that digital media use, particularly playing video games, would accelerate brain development globally. Since digital media use differs between sexes18, sub-group analysis was carried out separately for boys and girls. ## Methods ### Participants The neuroimaging and behavioural data used in this study were obtained from the ABCD Study (data release 4.0 [https://abcdstudy.org/](https://abcdstudy.org/); [http://doi.org/10.15154/1523041](http://doi.org/10.15154/1523041)), a longitudinal cohort of 11,875 children born between 2005 and 2009, enrolled at ages 9-11 years from 21 research sites across the United States between 2016-2018.19 Children were excluded if they were born extremely preterm (<28 weeks) or with birth weight (<1200g); were not proficient in English; had any neurological problems; had a history of seizures; or had a contraindication to undergo brain MRI scans. All children and their parents/guardians provided informed written consent/assent for participation, and the central Institutional Review Board at the University of California, San Diego approved the study protocols. ### Neuroimaging Children underwent brain MRI scans on 3-Tesla scanner platforms (Siemens Prisma, Philips, or General Electric 750) using a standard adult-sized head coil across two different time points, i.e., baseline (T) and Two-year later (T2). A unified protocol for the scanning was used to harmonise between scanning sites and MRI scanners. Three-dimensional T1-weighted images (1-mm isotropic) were acquired using magnetization prepared rapid acquisition gradient-echo (MP-RAGE) sequence and processed using the FreeSurfer (version 5.3.0) software as described elsewhere.20 All the pre-processed data were quality-checked according to the ABCD protocol, and children with excessive head motion or poor image quality were excluded from the data release. In this study, we used the Destrieux atlas as opposed to the Desikan-Killiany atlas, as it provides a finer granulated parcellation (i.e., parcellates each hemisphere into 74 regions of interest). Children who did not have relevant data on either SES, genetics, or digital media use were excluded from the present analysis. In addition, the ABCD study cohort included twins and siblings within families, and we randomly selected one child per family to eliminate the potential bias in the analysis. ### Socioeconomic status SES was defined as the first principal component from a probabilistic principal component analysis (PCA) of total household income, parental education, and neighbourhood quality. Children missing more than one of these SES measures were excluded. Parental education was categorised into middle school or less (1), some high school (2), high school graduate (3), some college/associates degree (4), bachelor’s degree (5), a master’s degree (6), or professional degree (7). The neighbourhood quality measure was the area deprivation index calculated from the American Community Survey using the address of the primary residency.21 The SES composite and each subcomponent were normalised (mean=0, SD=1). ### Genotyping, quality control, and polygenic score Saliva samples were collected from all the children at the baseline visit, and genotyping was performed using the Smokescreen array consisting of 646,247 genetic variants by the ABCD Study team.22 Quality control and imputation were performed by the National Bioinformatics Infrastructure Sweden (NBIS), as a service contracted by our lab (i.e., Klingberg Laboratory). Single Nucleotide Polymorphisms (SNPs) with call rates <98% or minor allele frequencies (MAF) <1% were excluded before imputation. Individuals with higher rates of missingness (>2%) or absolute autosomal heterozygosity >0.2 were also excluded, resulting in 10,069 children and 430,622 genetic variants. Next, haplotypes were pre-phased with SHAPEIT2, and genetic markers were imputed using the IMPUTE4 software and the 1000 Genomes References Panel (phase 3, build 37). After that, genotypes with an INFO score <0.3 or MAF <0.001% were excluded, which yielded 40,637,119 SNPs in a total of 10,069 children. The PCA module, as implemented in RICOPILI, was used to check for outliers and control population structure. We used the first 20 principal components from that genetic PCA, as discussed below. We created polygenic scores for cognitive performance for each child using PRSice-2, which was calculated by summing the effect sizes of thousands of SNPs, discovered by a large genome-wide association study (GWAS) on educational attainment, mathematical ability, and general cognitive ability.23 The large GWAS has all effect sizes and p values of their SNPs on the Social Science Genetics Association Consortium ([https://www.thessgac.org/data](https://www.thessgac.org/data)). Then, we used the data available by the consortium from a multi-trait analysis of GWAS, which, in our case, represents a joint polygenic score focused on a GWAS of cognitive performance and complemented by information from a GWAS on educational attainment, a GWAS on the highest-level math class completed, and a GWAS on self-reported math ability. This joint analysis is ideal because pairwise genetic correlations of these traits were high, and these GWAS had hundreds of thousands of individuals. Such a large sample size allows new studies to detect effects in samples of a few hundred individuals with 80% statistical power. To create the polygenic scores in our study, we performed clumping and pruning to remove nearby SNPs that are correlated with each another. The clumping sliding window was 250kb, with the linkage disequilibrium clumping set to r2 >0.25. We included the weightings of all SNPs, regardless of their p-value from the GWAS (p=1.00 threshold), resulting in 5255 SNPs. Finally, we normalised (mean=0, SD=1) the polygenic scores to fairly compare their effects on different phenotypes. ### Exposures Digital media use was measured across all annual visits (i.e., baseline (T), One-year later (T1), and Two-year later (T2)) using the Youth Screen Time Survey. Children were asked to report the number of hours spent on a typical weekday and weekend days by device, media platform, or activity for the following: (1) watching television or movies, (2) watching videos (e.g., YouTube), (3) playing video games on a computer, console, phone, or another device (e.g., Xbox, PlayStation, iPad), (4) Texting on a cell phone, tablet, or computer (e.g., Google Chat, WhatsApp), (5) Visiting social networking sites (e.g., Facebook, Twitter, Instagram), (6) Using video chat (e.g., Skype, FaceTime). Digital media activities were categorised as (a) watching television and videos (1+2), (b) using social media (4+5+6), and (c) playing video games (3). The response option included were none -‘0’, < 30min - ‘0.25’, 30min – ‘0.5’, 1h – ‘1’, 2h – ‘2’, 3h – ‘3’, or > 4h – ‘4’. To calculate the average hours spent per day for each type of digital media use, we used the following: (total number of hours spent on a weekday * 5 + the total number of hours on a weekend day * 2)/7. During both T and T1visits, details regarding digital media usage were collected using the same categorical scale as described above. However, from T2 onwards, the Youth Screen Time Survey was slightly modified due to the rise in digital media usage among children. The time spent watching television and videos was changed into ‘watching or streaming videos or movies’, and watching videos (such as YouTube) was changed into ‘watching or streaming videos or live streaming (such as YouTube, Twitch)’; we combined them into a single category as ‘watching television and videos’. ‘Video chatting, visiting social media apps, and texting cell phone’ was combined as ‘using social media’. ‘Editing photos and videos’ and ‘searching or browsing the internet was excluded as it does not co-exist with the T details. Playing video games was further divided into two sub-categories, i.e., ‘time spent on Single-player’ and ‘time spent on Multi-player’; we combined them as ‘playing video games’. Furthermore, the response format was also changed from categorical to continuous; the response options included were 0min, 15min, 30mins, 45min, 1h, 1.5h, 2h, 2.5h, 3h and every additional hour up until 24h. To keep the data consistent across all time points, we harmonised the T2 visit data according to baseline and T1 visits data and re-coded the T2 visit into (none -‘0’, < 30min - ‘0.25’, 30min – ‘0.5’, 1h – ‘1’, 1.15h – ‘1.25’, 1.30h – ‘1.5’, 2h – ‘2’, 2.15h – ‘2.25’, 2.30h – ‘2.5’, 3h – ‘3’, 3.15h – ‘3.25’, 3.30h – ‘3.5’, and >4h – ‘4’). ### Outcomes We prespecified our primary outcome as a global cortical surface area. The prespecified secondary outcomes were the volumes of the striatum and cerebellum. We defined the striatum by combining the volumes of the caudate nucleus, putamen, and accumbens. For the cerebellum, we combined the volumes of both grey and white matter structures of the cerebellum. We explored the individual regions of the cerebral cortex as exploratory outcomes adjusted for the total cortical surface area. For all the analyses, we collapsed both the left and right hemispheres. ### Data Analysis A linear mixed effect model was used to investigate the association between individual digital media use (i.e., watching television and videos, using social media, and playing video games; separately) and brain regions, both cross-sectionally and longitudinally. Age at baseline MRI scan visit, sex, 20 principal components from the genetic data, polygenic scores, SES, and handedness were used as fixed effects, and MRI scanner sites were used as random effects for the cross-sectional analyses (Model 1). Due to covid lockdown, there is more likely that these children could spent more time on digital media than anticipated at T. Therefore, to account for the rise in digital media use amongst children over two years, we used cumulative effects of digital media use rather than considering only T or just T2. The cumulative effects of digital media usage were calculated by averaging digital media across all three-time points (i.e., an average = T + T1 + T2 /3). For the longitudinal analyses, we investigated the cumulative effects of digital media use on changes in the brain (i.e., brain measures at T2 – brain measure at T) (Model 2), controlling for differences in age of the child (i.e., age at T2 visit – the age at T visit) as an additional fixed effect covariate, in addition to all the previously mentioned covariates in the Model 1. We used absolute values for the surface area (mm2) and subcortical volumes (mm3) measures for all the analyses. To limit the number of statistical tests, we computed bilateral estimates by combining both the left and right sides of the regions. For the longitudinal analyses, we used the differences in absolute values of brain measures (i.e., T2 – T visits). Analysis of the primary and secondary outcomes was unadjusted for multiple comparisons, as they were predefined. The subgroup analysis was carried out only for the primary and secondary outcomes and was corrected for multiple comparisons. Given the number of statistical tests examined for the primary and secondary outcomes in the subgroups, the Benjamini-Hochberg corrections were applied to the whole sub-group analysis to control for Type-1 error24. In total, three brain regions for three digital media measures (i.e., watching television, using social media, and playing video games), were tested at two time points (T and T2), resulting in 36 tests altogether for both boys and girls. In addition, for the exploratory outcomes, separate Benjamini-Hochberg corrections were applied to each of the digital media use across two-time points. In total, 75 separate tests were performed for each digital media use across two-time points. The alpha value of less than 0.05 was considered significant only if it survived the Benjamini-Hochberg corrections. All analyses were carried out using the R package (version 4.2.0). ## Results Of the 11,875 children recruited into the ABCD study cohort, 6492 children (age in months, mean (SD)=118.6 (7.2), i.e., 9.9 years; boys, n (%)=3382 (52%); right-handed, n (%)=5176 (80%)) fulfilled our inclusion criteria were included at the T visit (i.e., 9-11 years of age). Of these, 4502 children (age in months=142.6 (7.6), i.e., 11.9 years; boys=2427 (54%)) were included at the T2 visit. The number of children followed up at the T2 visit was slightly lower compared to the T visit because the COVID-19 pandemic had significantly affected the data collection. The average time spent by children on digital media use at T was 2.2h/day for watching television and videos, 0.4h/day for using social media, and 0.9h/day for playing video games **(Table 1)**. View this table: [Table 1](http://medrxiv.org/content/early/2022/10/19/2022.07.01.22277142/T1) Table 1 Usage of digital media by children aged 9-11 over three-time points Data are presented as mean (standard deviation). Abbreviations - a, significant differences between T and T1; b, significant differences between T1 and T2. Cumulative effects over two years are calculated by the average across all three annual visits. Compared to the T visit, the average time spent on digital media use significantly increased over two years in the overall cohort and in boys and girls (**Table 1)**. During the two years follow-up period (i.e., all three annual visits), children, on average, spent 2.2h/day watching television and videos, 1.2h/day using social media, and 1.2h/day playing video games. Moreover, boys, during this period, on average, spent 2.3h/day watching television and videos, 1.2h/day using social media, and 1.5h/day playing video games, whereas girls spent 2.1h/day watching television and videos, 1.3h/day using social media, and 0.8h/day for playing video games. ### Baseline associations At T, children who spent more time watching television and videos had a smaller global cortical surface area (standardised beta, β=+0.03; 95% standardised CI, (+0.06 – +0.02); *P*=0.001) at 9-11 years of age **(Table 2)**. Children who spent more time playing video games (β=+0.03 (+0.05 – +0.008); *P*=0.007) but not using social media (β = +0.002 (+0.02 – 0.02); *P*=0.84) also had a smaller global cortical surface area. View this table: [Table 2](http://medrxiv.org/content/early/2022/10/19/2022.07.01.22277142/T2) Table 2 Association between digital media use and brain regions in children aged 9-11 years at baseline (T) in the overall cohort and in boys and girls separately (Model - 1) Data are presented as absolute beta coefficients (95% confidence interval). Model 1, adjusted for the age at baseline MRI scan visit, sex, 20 principal components from the genetic data, polygenic scores, SES, handedness, and MRI scanner sites. \***| *P*<0.001, ** *P*<0.01, * *P*<0.05. # Interaction effect was significant (*P*<0.001). Findings that survive corrections for multiple comparisons in boys and girls using false discovery rates are bolded. At T, children who spent more time watching television and videos (β=+0.03 (+0.06 – +0.01); *P*=0.002) or playing video games (β=+0.03 (+0.05 – +0.006); *P*=0.01) but not using social media, had a smaller volume of the striatum at 9-11 years **(Table 2)**. However, watching television and videos, using social media, or playing video games, were not associated with the volume of the cerebellum. ### Test-retest reliability To assess the reliability of the measurements, we calculated the interclass correlation coefficients (ICC) between the global surface area measures and found them to be 0.97 (0.97 – 0.98). ### Primary outcome – cortical surface area changes Between T and T2, the average cortical surface area increased by ∼2%, reflecting the normal pattern of cortical development in these selected children. At T2, children who cumulatively spent more time watching television and videos (β=0.01 (0.02 – 0.03); *P*=0.72), using social media (β=0.01 (0.02 – 0.03); *P*=0.70), or playing video games (β=0.03 (0.005 – 0.06); *P*= 0.09) showed no changes (i.e., neither increase nor decrease) in size of the global cortical surface area **(Table 3)**. View this table: [Table 3](http://medrxiv.org/content/early/2022/10/19/2022.07.01.22277142/T3) Table 3 Association between cumulative effects of digital media use over two years and the corresponding change in brain regions (T2) in the overall cohort and in boys and girls separately (Model – 2) Data are presented as absolute beta coefficients (95% confidence interval). Model 2, adjusted for the age at baseline MRI scan visit, sex, 20 principal components from the genetic data, polygenic scores, SES, handedness, differences in age (age at two years of follow-up – the age at baseline MRI scan visit), and MRI scanner sites. \***| *P*<0.001, ** *P*<0.01, * *P*<0.05; # represent the significant interaction of sex and playing video games. Findings that survive corrections for multiple comparisons in boys and girls using false discovery rates are bolded. ### Secondary outcome – striatum and cerebellum At T2, the children who spent more time watching television and videos or playing video games showed no difference in changes in volumes of the cerebellum or striatum **(Table 3)**. However, those who spent more time using social media showed a decreasing trend in volumes of the striatum (β=+0.03 (+0.06 – 0.0007); *P*= 0.05). ### Subgroup analysis – the effect of sex At T, boys who spent more time watching television and videos (β=+0.05 (+0.08 – +0.01); *Punadj*=0.005) or playing video games (β=+0.04 (+0.07 – +0.003); *Punadj*=0.03) had a smaller global surface area at 9-11 years **(Table 2)**. Boys who spent more time watching television and videos also had smaller volumes of the striatum (β=+0.06 (+0.09 – +0.02); *Punadj*=0.001). There was no association found between using social media and any brain regions studied in boys. Conversely, girls (β=0.05 (0.02 – 0.09); *Punadj*=0.004) who spent more time using social media had larger volumes of the cerebellum. However, after correcting for multiple comparisons, only the association between watching television and videos and global cortical surface area and striatum in boys; and the association between social media and cerebellum volumes in girls, remained significant (all FDR-corrected *P*<0 .05). At T2, boys who spent more time playing video games showed a significant increase in sizes of the global cortical surface area (β=0.05 (0.008 – 0.09); *Punadj*=0.01), whereas girls showed a significant decrease in the volumes of the striatum (β=+0.07 (+0.11 – +0.02); *Punadj*=0.003) **(Table 3)**. However, the interaction effect of sex and playing video games was significant only for the volumes of the striatum (*P*<0.001) (**Figure 1)**. Girls who also cumulatively spent more time using social media showed a significant decrease in the volumes of the striatum (β=+0.05 (+0.13 – +0.03); *Punadj*=0.03). However, after correcting for multiple comparisons, the only longitudinal effect that remained significant was that of playing video games on striatum volumes in girls (FDR-corrected *P*=0.03). ![Figure 1](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/10/19/2022.07.01.22277142/F1.medium.gif) [Figure 1](http://medrxiv.org/content/early/2022/10/19/2022.07.01.22277142/F1) Figure 1 Box and whisker plots showing the effect of using social media or playing video games (25th, 75th quartile) on striatal volume (mm3) at two time points (T and T2). The length of the box represents the interquartile range (IQR), the horizontal line in the box interior represents the median, and the whiskers represent the 1.5 IQR of the 25th quartile or 1.5 IQR of the 75th quartile, and the dots represent outliers. ### Exploratory outcomes We investigated both the models (i.e., Model 1 and Model 2) in each of the individual regions (collapsed across hemispheres) and found no association between any of the digital media use and individual brain regions after correcting for multiple comparisons **(eSupplement)**. ### Post-hoc analysis Although we found a negative effect of playing video games on the striatum volumes in girls, along with a significant interaction effect. However, in the cross-sectional analysis, a negative trend toward significance (*Punadj*=0.06) was found between playing video games and striatum volumes at T in girls. Since there were significant differences in brain volumes at baseline, we also made a post-hoc analysis using brain volume at T as a covariate and brain volume at T2 as the dependent variable. Additionally, we included the interaction effect of sex and playing video games in the revised model to reconfirm the subgroup findings of the effect of sex on striatum volumes. We proceeded to the separate analysis only when the interaction effect of sex was significant. At T2, children who cumulatively spent more time using social media showed a significant decrease in the volumes of the striatum (β=+0.009 (+0.02 – +0.002); *Punadj*= 0.001); but the interaction effect of sex and using social media on striatum was not significant. Further, children who cumulatively spent more time playing video games also showed a significant decrease in the volumes of the striatum (β=+0.01 (+0.02 – +0.005); *Punadj*=0.001) and the interaction effect of sex and playing video games on striatum were significant. On investigating the sex separately, girls (β=+0.01 (+0.02 – +0.004); *Punadj*=0.002) but not boys (β=+0.002 (+0.009 – 0.004); *Punadj*=0.51) who spent more time playing video games showed a significant decrease in the volumes of the striatum. ## Discussion Here, we investigated the association between digital media use over two years and changes in cortical surface area and striatal and cerebellar volumes in children aged 9-11 years. No type of digital media use was associated with the development of the cerebellum or cortical surface area, neither globally nor regionally, although there was a trend for an increased surface global area with gaming in boys. However, we found a weak (β=+0.07) negative effect of using social media or playing video games on change in striatal volume in girls. These associations were found in the analysis of subtracted brain volumes (i.e., T2 - T) and remained significant in the *post-hoc* analysis of relative changes, using the T volume of the striatum as a covariate. Interestingly, the negative effect of playing video games was found only in girls, even though the average time spent playing video games was lower in girls than in boys. One possible explanation for this sex difference could be related to the genres of video games that were played, which are known to differ between boys and girls25. Alternatively, this could be due to developmental differences. In general, the onset of puberty in girls begins earlier than in boys and includes hormonal and biological changes, which could have affected brain development. Children aged 11-13 years are in a dynamic phase of brain development, which includes myelination, strengthening of synapses, and selective pruning of neurons and connections26, all of which could be affected by hormonal levels27, 28. Furthermore, the age range of 11-13 years in girls could also be a psychologically and socially sensitive period. This is illustrated by a longitudinal study showing age and sex-specific associations between social media use on psychological well-being in adolescents aged 10-21 years, where they found a negative association for girls at 11-13 years and for boys at age 14-15 years29, possibly relating to sensitive periods due to hormonal, social, and psychological changes. It will therefore be of particular interest to follow up with this cohort again later to analyze the effect of digital media use when boys reach adolescence. Decreased striatal volume has not, to our knowledge, previously been found in association with playing video games. But, one experimental, functional MRI study in young adults reported decreased activity in the ventral striatum after eight weeks of video game training30. In the present study, we found that playing video games was associated with smaller striatal volumes at T **(Table 2)**, with similar trends for both boys and girls. This could be due to self-selection but is also consistent with a causal effect of playing video games on striatal development over two years. The striatum is important for cognitive functions, including both inhibitory ability and working memory31. But given that a previous study from our group using the same cohort found a positive impact of playing video games on intelligence4, it is unlikely that the striatal effect in this study might link to any negative cognitive effect. However, there is an association between smaller striatal volume and attention deficit hyperactivity disorder in children and adolescence32, 33. Therefore, the present finding is a potentially serious observation that needs to be closely followed with longer periods of follow-up and associations to cognitive function and behavioural symptoms. It should also be noted that the effect size was very small (β=0.07 in model 2, and β=0.01 in *post-hoc*), and this may have no functional consequence. In our earlier study, we found that playing video games had a positive impact on intelligence in boys but not girls. Given the association between intelligence and global cortical surface area17, we expected a positive impact on this measure. This was also found **(Table 3)**, although the effect was weak (β=0.05), and the test did not survive corrections for multiple comparisons. Although weak, this effect is in line with the complex demands of video game playing, including the processing of rapidly changing visual stimuli and complex motor demands based on the specific genre of the game. Additionally, it requires children to concentrate on those stimuli carefully, process the related information promptly, and act quickly using a keyboard/mouse/joysticks. A limitation is that the ABCD study consortium collected the daily digital media use from children using the Youth Screen Time Survey, which is not a quantitative measurement of screen time. It also fails to capture any information about the genre of video games. It is possible that different types of gaming content can differently alter children’s brain development. In summary, digital media usage, particularly gaming, does not affect cortical brain development but affects the volumes of the striatum in a sex-specific manner in children aged 9-11 years. Further, our findings are particularly suggestive of considering physiological and critical window periods in adolescent brain development. Longer follow-ups and other stages of development are warranted. ## Authors contribution **Samson Nivins:** Conceptualization, Writing – original draft, Investigation, Methodology, Formal analysis, Visualization. **Bruno Sauce, Nicholas Judd:** Conceptualization, Methodology, Writing – review & editing. **Magnus Liebherr:** Conceptualization, Writing – review & editing. **Torkel Klingberg:** Conceptualization, Methodology, Writing – critical review, insights, & editing, Funding acquisition. ## Conflict of interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ## Funding This study was supported by the Swedish Research Council to Torkel Klingberg ## Supporting information Supplements [[supplements/277142_file03.docx]](pending:yes) Supplements [[supplements/277142_file04.docx]](pending:yes) ## Data Availability All data produced in the present study are available upon reasonable request to the authors ## Acknowledgements Data used in the preparation of this article were obtained from the Adolescent Brain Cognitive Development (ABCD) Study ([https://abcdstudy.org](https://abcdstudy.org)), held in the NIMH Data Archive (NDA). The ABCD study was supported by the National Institutes of Health and additional federal partners under award numbers U01DA041022, U01DA041025, U01DA041028, U01DA041048, U01DA041089, U01DA041093, U01DA041106, U01DA041117, U01DA041120, U01DA041134, U01DA041148, U01DA041156, U01DA041174, U24DA041123, and U24DA041147. A full list of supporters is available at [https://abcdstudy.org/?s=nIH-collaborators](https://abcdstudy.org/?s=nIH-collaborators). A listing of participating sites and a complete listing of the study investigators can be found at [https://abcdstudy.org/principal-investigators.html](https://abcdstudy.org/principal-investigators.html). The ABCD consortium investigators designed and implemented the study and/or provided data but did not necessarily participate in the analysis or writing of this report. This manuscript reflects the views of the authors and may not reflect the opinions or views of the NIH or ABCD consortium investigators. ## Footnotes * The manuscript has been revised along with the title and a few methodology as per reviewers suggestion * Received July 1, 2022. * Revision received October 19, 2022. * Accepted October 19, 2022. * © 2022, Posted by Cold Spring Harbor Laboratory The copyright holder for this pre-print is the author. All rights reserved. The material may not be redistributed, re-used or adapted without the author's permission. ## Reference 1. 1.Media CS. The Common Sense Census: Media Use by Tweens and Teens, 2019.) (2019). 2. 2.Lissak G. Adverse physiological and psychological effects of screen time on children and adolescents: Literature review and case study. Environmental research 164, 149–157 (2018). [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F19%2F2022.07.01.22277142.atom) 3. 3.Kostyrka-Allchorne K, Cooper NR, Simpson A. The relationship between television exposure and children’s cognition and behaviour: A systematic review. Developmental review 44, 19–58 (2017). 4. 4.Sauce B, Liebherr M, Judd N, Klingberg T. The impact of digital media on children’s intelligence while controlling for genetic differences in cognition and socioeconomic background. Scientific reports 12, 7720–7720 (2022). 5. 5.Paulich KN, Ross JM, Lessem JM, Hewitt JK. Screen time and early adolescent mental health, academic, and social outcomes in 9- and 10-year old children: Utilizing the Adolescent Brain Cognitive Development _ (ABCD) Study. PloS one 16, e0256591–e0256591 (2021). 6. 6.Bailey K, West R. The effects of an action video game on visual and affective information processing. Brain research 1504, 35–46 (2013). 7. 7.Altbäcker A, et al. Problematic internet use is associated with structural alterations in the brain reward system in females. Brain imaging and behavior 10, 953–959 (2015). 8. 8.Wang H, et al. The alteration of gray matter volume and cognitive control in adolescents with internet gaming disorder. Frontiers in behavioral neuroscience 9, 64–64 (2015). 9. 9.Paulus MP, et al. Screen media activity and brain structure in youth: Evidence for diverse structural correlation networks from the ABCD study. NeuroImage (Orlando, Fla) 185, 140–153 (2019). 10. 10.Fjell AM, et al. Development and aging of cortical thickness correspond to genetic organization patterns. Proceedings of the National Academy of Sciences - PNAS 112, 15462–15467 (2015). 11. 11.Gleich T, Lorenz RC, Gallinat J, Kühn S. Functional changes in the reward circuit in response to gaming-related cues after training with a commercial video game. Neuroimage 152, 467–475 (2017). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.neuroimage.2017.03.032&link_type=DOI) 12. 12.Haier RJ, Karama S, Leyba L, Jung RE. MRI assessment of cortical thickness and functional activity changes in adolescent girls following three months of practice on a visual-spatial task. BMC research notes 2, 1–7 (2009). 13. 13.Kühn S, Gallinat J, Mascherek A. Effects of computer gaming on cognition, brain structure, and function: a critical reflection on existing literature. Dialogues in clinical neuroscience 21, 319–330 (2019). 14. 14.Pujol J, et al. Video gaming in school children: How much is enough? Annals of neurology 80, 424–433 (2016). 15. 15.Schnack HG, et al. Changes in Thickness and Surface Area of the Human Cortex and Their Relationship with Intelligence. Cerebral cortex (New York, NY 1991) 25, 1608–1617 (2015). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/cercor/bht357&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=24408955&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F19%2F2022.07.01.22277142.atom) 16. 16.Skranes J, Løhaugen GCC, Martinussen M, Håberg A, Brubakk A-M, Dale AM. Cortical surface area and IQ in very-low-birth-weight (VLBW) young adults. Cortex 49, 2264–2271 (2013). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.cortex.2013.06.001&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23845237&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F19%2F2022.07.01.22277142.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000324351800024&link_type=ISI) 17. 17.Judd N, et al. Cognitive and brain development is independently influenced by socioeconomic status and polygenic scores for educational attainment. Proceedings of the National Academy of Sciences - PNAS 117, 12411–12418 (2020). 18. 18.Bagot KS, et al. Youth screen use in the ABCD® study. Dev Cogn Neurosci 57, 101150 (2022). 19. 19.Casey BJ, et al. The Adolescent Brain Cognitive Development (ABCD) study: Imaging acquisition across 21 sites. Dev Cogn Neurosci 32, 43–54 (2018). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/J.DCN.2018.03.001&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F19%2F2022.07.01.22277142.atom) 20. 20.Cornejo MD, et al. Image processing and analysis methods for the Adolescent Brain Cognitive Development Study. NeuroImage (Orlando, Fla) 202, 116091–116091 (2019). 21. 21.Kind AJH, et al. Neighborhood socioeconomic disadvantage and 30-day rehospitalization: a retrospective cohort study. Annals of internal medicine 161, 765–774 (2014). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.7326/M13-2946&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=25437404&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F19%2F2022.07.01.22277142.atom) 22. 22.Baurley JW, Edlund CK, Pardamean CI, Conti DV, Bergen AW. Smokescreen: a targeted genotyping array for addiction research. BMC genomics 17, 145–145 (2016). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1186/s12864-016-2495-7&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=26921259&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F19%2F2022.07.01.22277142.atom) 23. 23.Choi SW, O’Reilly PF. PRSice-2: Polygenic Risk Score software for biobank-scale data. Gigascience 8, (2019). 24. 24.Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal statistical society: series B (Methodological) 57, 289–300 (1995). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.2307/2346101&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=WOS:A1995QE4&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F19%2F2022.07.01.22277142.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1995QE45300017&link_type=ISI) 25. 25.Leonhardt M, Overå S. Are There Differences in Video Gaming and Use of Social Media among Boys and Girls?-A Mixed Methods Approach. Int J Environ Res Public Health 18, (2021). 26. 26.Giedd JN, Rapoport JL. Structural MRI of pediatric brain development: what have we learned and where are we going? Neuron 67, 728–734 (2010). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.neuron.2010.08.040&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=20826305&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F19%2F2022.07.01.22277142.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000281938900008&link_type=ISI) 27. 27.Schulz KM, Molenda-Figueira HA, Sisk CL. Back to the future: The organizational-activational hypothesis adapted to puberty and adolescence. Horm Behav 55, 597–604 (2009). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.yhbeh.2009.03.010&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=19446076&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F19%2F2022.07.01.22277142.atom) 28. 28.Juraska JM, Willing J. Pubertal onset as a critical transition for neural development and cognition. Brain Res 1654, 87–94 (2017). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.brainres.2016.04.012&link_type=DOI) 29. 29.Orben A, Przybylski AK, Blakemore S-J, Kievit RA. Windows of developmental sensitivity to social media. Nature Communications 13, 1649 (2022). 30. 30.Lee H, et al. Videogame training strategy-induced change in brain function during a complex visuomotor task. Behavioural Brain Research 232, 348–357 (2012). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.bbr.2012.03.043&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=22504276&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F19%2F2022.07.01.22277142.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000305855100004&link_type=ISI) 31. 31.McNab F, Klingberg T. Prefrontal cortex and basal ganglia control access to working memory. Nature neuroscience 11, 103–107 (2008). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/nn2024&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=18066057&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F19%2F2022.07.01.22277142.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000252117900018&link_type=ISI) 32. 32.Cupertino RB, et al. Reduced fronto-striatal volume in attention-deficit/hyperactivity disorder in two cohorts across the lifespan. NeuroImage: Clinical 28, 102403 (2020). 33. 33.Qiu A, et al. Basal ganglia volume and shape in children with attention deficit hyperactivity disorder. Am J Psychiatry 166, 74–82 (2009). [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1176/appi.ajp.2008.08030426&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=19015232&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F10%2F19%2F2022.07.01.22277142.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000262173300015&link_type=ISI)