In this study, we examined the relationship between screen time use, sleep characteristics, daytime somnolence, and academic performance in school-aged adolescents. We surveyed 1,257 12- to 18-year-old adolescents attending 52 schools in urban or suburban areas of Argentina. We recorded the daily exposure to various screen-based activities, including video- and online-gaming, social media, TV or streaming. Screen time and device type in the hour before bedtime, sleep patterns during weekdays and weekends, somnolence (Pediatric Daytime Sleepiness Scale score), and grades in language and mathematics were also assessed. Structural Equation Modelling was used to identify a path connecting the latent variables. Results are expressed as standardized regression weights (srw). Missing data were present in 393 subjects, and thus the final sample consisted of 864 complete responses. Daytime somnolence (i.e., PDSS score ≥ 15) was observed in 614 participants (71%), and academic failure (i.e., grades < 7/10) in 352 of them (41%). Time spent using video gaming consoles was negatively associated with sleep duration (srw = -0.22, p<0.01) and positively connected with daytime somnolence (srw = 0.11, p<0.01). Use of mobile devices was associated with lower academic performance (srw = -0.11, p<0.01). Sleep duration was inversely related to daytime somnolence (srw = -0.27, p<0.01), which was in turn negatively associated with academic performance (srw = -0.18, p<0.05). Bedtime computer use did not influence any outcome. In summary, among adolescents, screen use adversely affected nighttime sleep, daytime somnolence, and academic performance. These findings call for the implementation of educational public campaigns aimed at promoting healthy sleep and reducing screen exposure among adolescents.
Citation: Pérez-Chada D, Bioch SA, Schönfeld D, Gozal D, Perez-Lloret S, On behalf of the Sleep in Adolescents Collaborative study group (2023) Screen use, sleep duration, daytime somnolence, and academic failure in school-aged adolescents. PLoS ONE 18(2): e0281379. https://doi.org/10.1371/journal.pone.0281379
Editor: Fakir Md Yunus, Dalhousie University, CANADA
Received: March 16, 2022; Accepted: January 22, 2023; Published: February 14, 2023
Copyright: © 2023 Pérez-Chada et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The dataset used in the study can be accessed at: https://data.mendeley.com/datasets/pj9hzmp7ym/1. It can be cited as: Perez-Lloret, Santiago; Perez Chada, Daniel (2022), “Screen use, sleep duration, daytime somnolence, and academic failure in school-aged adolescents”, Mendeley Data, V1, doi: 10.17632/pj9hzmp7ym.1.
Funding: This study was supported by the “Dr Gustavo Amestoy” grant (Universidad Austral, Facultad de Ciencias Biomedicas) awarded to DPC. DG is supported in part by the Leda Sears Foundation, by NIH grant AG061824, and by Tier 2 and TRIUMPH grants from the University of Missouri. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Technological advances over the last decades have markedly enhanced the use of screen devices for both work and leisure activities (e.g., gaming, social networking, or watching multimedia contents) at all ages. For example, the proportion of 0 to 8-year-old children that had access to mobile devices grew from 52% in 2011 to 75% in 2013 . Some studies suggest that school-aged children and adolescents spend about 7 hours per day in front of a screen (i.e., about 43% of the waking day assuming 8 hours of sleep per night) [2,3]. In many instances, screens are used as “babysitters”, allowing parents to conduct work activities at home or perform their domestic chores .
Excessive screen use may affect overall health, disrupt sleep, and reduce learning and academic performance. One of the most obvious effects of screen use is the reduction in physical activity, thereby leading to sedentarism and facilitating weight gain [5–7]. This is further aggravated by changes in diet related to advertising campaigns, and also spurred by behaviors associated with screen use, such as snacking [8–11]. Obesity and sedentarism, in turn, increase the risk of hypertension and other metabolic disorders, while further perpetuating the unhealthy behaviors that promoted their occurrence .
Sleep time and quality are also affected by screen use [2,12]. Time spent in front of the screen usually comes at the expense of sleep duration . In addition, bright light and electromagnetic radiation, when used during the evening, inhibit melatonin production [13,14], thereby altering circadian rhythms. Furthermore, psychophysiological arousal incited by video gaming or social media use, particularly before bedtime, may impair pre-sleep relaxation, and further delay sleep onset [13,15].
Sleep is essential for learning , as it participates in the consolidation of declarative and procedural memories [17,18]. The reactivation during sleep of memories acquired during wakefulness is fundamental for consolidation of such memories in brain regions such as the hippocampus [19,20]. Regular sleep facilitates key aspects of synaptic plasticity, whereas extended wakefulness tends to reduce it . It is thus not surprising that academic performance may be impaired among children and adolescents with poor sleep quality, as shown in a recent meta-analysis involving 13,631 school-aged children and adolescents participating in 16 different studies . Daytime sleepiness resulting from insufficient sleep also imposes a profound adverse impact on learning and academic performance [22–24].
A recent meta-analysis involving 480,479 participants aged 4 to 18 years, showed that television viewing and video gaming were associated with reduced academic performance . It is tempting to hypothesize that sleep disturbances and daytime somnolence mediate the deleterious effects of screen use on academic performance. A recent study in more than 10,000 children showed that adequate sleep time and screen use ≤ 2-h daily independently predicted improved academic performance . Notwithstanding, the interaction(s) between these factors remains to be explored.
In this study, we explored the relationship between screen time use, sleep characteristics, daytime somnolence, and academic performance in a sample of school-aged adolescents from Argentina.
Twelve- to eighteen-year-old adolescents attending the morning shift of participating schools were invited to answer a survey between May and November 2019. Fifty-two schools, both urban and suburban, participated in the survey.
The protocol was approved by the Institutional Review Board of the Austral University (Buenos Aires, Argentina, CIE17-044-B). Participants were requested to complete the questionnaire in their smartphones or in computers provided by the schools after written consent was obtained from the parents or caretakers, as well as from the school authorities. Students were free to participate or not at their own discretion.
After the survey was completed, we approached a subset of the participants and their corresponding schools (i.e., those that indicated the ability to provide grades), and asked for authorization to collect grades for language and mathematics achieved by each student in the trimester corresponding to the survey responses.
Daytime somnolence was assessed using the Pediatric Daytime Sleepiness scale [23,27]. Briefly, the PDSS is an 8-item questionnaire assessing the weekly frequency of episodes of inattention in classroom due to sleepiness, of sleep episodes while doing homework, sustaining adequate level of attention in the classroom, feeling tired or being in a bad mood during the day, problems for getting up from bed during the morning, falling asleep immediately after being awakened, needing someone to wake you up in the morning, or feeling the need of more sleep. It has a Chronbach alpha = 0.74 . PDSS scores > 15 indicate the presence of daytime somnolence [23,27]. Sleep patterns were obtained by collecting information on bedtimes and wake up times during weekdays and weekends, whether any naps were obtained during weekdays and weekends and their duration, and dinner times during weekdays. In addition, distance from home to school, and information about the composition and characteristics of the household were collected.
Weight and height were inquired, and Body Mass Index was calculated and then transformed to age- and gender-adjusted Z-scores. Overweight and obesity were defined as Z-scores > 1 or > 2, respectively .
Exposure times spent with various types of screens as related to specific activities were recorded, including video gaming devices, tablets/smartphones for online gaming, social media networking, and watching TV or video streaming. Participants were instructed to document the average time spent per day engaged in each of these activities and deemed representative of the whole week. Answers were capped at 12-h, with higher values treated as missing data. Participants were also requested to indicate how many days per week computers, smartphones, tablets, gaming devices, or TV/streaming apps were used in the hour before bedtime.
Academic failure was defined as language and/or math grades < 7/10, as this is the cut-off point usually used in Argentina’s educational system for passing an exam.
Numerical and categorical variables were compared between participants with or without daytime somnolence or according to academic results (i.e., failure or success) using t-tests (with an adjustment in cases where variances were non-homogeneous among the groups) or Chi-square tests, respectively. Data are presented as mean ±SD or frequencies as appropriate. Multivariate models were built using logistic regression. Only variables attaining statistical significance in bivariate comparisons, or those deemed of clinical relevance, were included in the multivariate models. Multicolinearity was absent from all models. A likelihood ratio test was used to test whether the relationship between screen time variables and outcomes (i.e., daytime somnolence and academic failure) was of a linear or non-linear nature. All statistical tests were conducted using statistical software SPSS v.23 (IBM®, Chicago, Ill). A two-tailed p value designating statistical significance was set at <0.05.
Structural equation modelling analysis
A complex multi-step pathway connecting screen use, sleep time, daytime somnolence, and academic performance was modelled using Structural Equation Modelling (SEM) . We followed the usual procedures . Briefly, the model was initially fitted by using the Maximum Likelihood estimation (ML) method. Univariate and multivariate normality was assessed by computing Z-tests on variable’s kurtosis. Variable’s kurtosis > 5 or multivariate kurtosis critical value > 7 were considered as indicators of non-normality . Variables with kurtosis > 5 were log10-transformed and multivariate kurtosis was rechecked. If non-normality was still detected, the model was refitted by using the ML method with 100-sample bootstrapping. The factor structure of the PDSS and of the latent variable “academic failure” were assessed in first place by Confirmatory Factor Analysis (CFA). Latent variables connected with screen use and sleep characteristics were explored, first by exploratory factor analysis, and then tested by CFA. Model’s validity was assessed by the Chi-square (χ2) statistic, the root mean square error of approximation (RMSEA), the Comparative Fit Index (CFI), and Normed Fit Index (NFI). Only after a satisfactory measurement model was built (i.e., RMSEA < 0.08 and CFI/NFI>0.70), we fitted the path model connecting the latent variables (i.e. sleep duration, daytime somnolence, 0-to-10 math and language grades to assess academic performance, and screen use) between each other.
We used Modification Indices (MI) to modify the model by including significant associations not modelled before, if they had a biological or clinical significance. Results from SEM are presented in terms of standardized regression weights (srw) between pairs of variables. Such coefficients represent the strength of the relationship between any pair of variables independently from confounding factors. P-values < 0.05 were considered statistically significant. SEM analysis was performed with AMOS v.23 (IBM®, Chicago, Ill).
A total of 1,257 subjects participated in the survey and provided academic grades. Of these, 393 participant responses had missing data (299 had missing data on sleep variables and 132 had missing data on screen use). Such that the final sample was comprised of 864 subjects. The characteristics of the participants are shown in Table 1. Compared to those that were included in the final sample, participants with missing data had lower average grades (7.40±1.37 vs 7.09±1.56, respectively, p<0.01), and had a higher frequency of academic failure (40.7% vs 48.1%, respectively, p = 0.01).
Screen time, sleep characteristics, and daytime sleepiness
Mean PDSS score was 18.4±5.1, daytime somnolence (i.e., PDSS > 15) was present among in 614 (71%) of the responders. Statistically significant differences between somnolent and non-somnolent groups emerged for the frequency of male gender, weekday bedtime, sleep time, awakening time, and nap time, as well as weekend sleep patterns (Table 1). Among subjects with daytime somnolence, time spent on social networks was higher than in those without somnolence (3.3±2.6 vs. 2.6±2.3 h/day, p<0.01), whereas bedtime smartphone use was higher (6.1±1.8 vs 5.4±2.4 days/week). Likelihood ratio tests results supported a linear relationship between screen times and daytime somnolence scores (results not shown). A logistic regression analysis identified female gender, longer weekend sleep time, higher time spent video gaming, and bedtime smartphone use as factors significantly and independently associated with daytime somnolence (Table 1).
Screen time and other factors related to academic performance
Overall language and math grades were 7.51±1.39 (min-max: 3–10) and 7.18±1.74 (2–10). Out of the 864 participants, 352 (40.7%) failed in language and/or math. Participants who failed in their exams were more frequently males, suffered more frequently from daytime somnolence, slept less during weekdays, made longer naps, and went to sleep later on weekends (Table 2). Regarding screen time, participants who failed their exams spent longer times playing videogames (0.9±1.6 vs. 0.6±1.3 h/day in those who succeeded, p<0.01), spent longer time watching TV or streamed contents (2.4±1.8 vs. 2.1±1.8 h/day, p = 0.03), and used gaming devices at bedtime more frequently (0.7±1.6 vs. 0.5±1.3, p = 0.02). The likelihood ratio suggested that a non-linear relationship existed between academic failure and using mobile devices for gaming (data not shown). A logistic regression analysis identified that the following factors were significantly and independently connected with academic failure: male gender, higher PDSS score, shorter weekdays sleep time, and time spent using mobile devices for gaming (Table 2).
A multi-step pathway connecting screen time, sleep patterns, daytime somnolence, and academic performance
The characteristics of the final measurement model are shown in Table 3. To achieve univariate and multivariate normality, several variables were log10-transformed, including weekly computer bedtime use, bedtime tablet use, and time spent video gaming. Sleep latency was initially included, but later left out, as it increased the multivariate non-normality ratio, and did not contribute to the predictive power of the model. In addition, 9 outlier cases were excluded based on the calculation of Mahalanobis squared distances. In the final model, all variables were normally distributed, and multivariate non-normality critical ratio was 7.9, slightly above the required level. Therefore, bootstrapping results are also provided in Table 3. MIs did not suggest any significant modification. The model included the following latent variables, which were included in the path diagram: Bedtime computer use (days/week), time spent using of video gaming consoles (h/day), Time spent using mobile devices (h/day), sleep duration, male gender, PDSS Somnolence score, and academic performance. The fit was fair, with CFI = 0.753, NFI = 0.728, RMSEA = 0.089, and AIC = 1637.27.
The characteristics of the final path model are shown in Table 4. The fit was still fair, with a CFI = 0.702, NFI = 0.679, RMSEA = 0.096, AIC = 1896.65. It should be kept in mind that the measurement model is more saturated as it allows for the correlation between all latent variables within each other, and thus may result in higher, yet spurious fit.
The summary of the most relevant association between variables in the final path model is shown in Fig 1. Time spent using video gaming consoles was negatively related sleep duration (srw = -0.22, p<0.01) and positively connected with daytime somnolence (srw = 0.11, p<0.01). Use of mobile devices was associated with lower academic performance (srw = -0.11, p<0.01). Sleep duration was inversely related to daytime somnolence (srw = -0.27, p<0.01), which was in turn negatively associated with academic performance (srw = -0.18, p<0.05). Bedtime computer use did not influence any outcome. Sleep duration did not show a significant direct relationship with academic performance.
Paths were fitted by Structured Equation modelling. Shown are Standardized Regression Weights which represent the strength of the relationship between any pair of variables independently from all other variables included in the model. * p<0.05, ** p<0.01.
The present study including a large sample of 12- to 18-years old adolescents from large and small cities in Argentina, explores the direct and indirect burden imposed by screen use on daytime somnolence, and academic performance. The concurrent assessment of these two variables enabled improved modeling of the complex interrelationships between these outcomes. One of the findings is that the effects of screen time on these outcomes depended on the type of screen activity in which the participants were engaged and demonstrate that time spent video gaming affects sleep and increase daytime somnolence, independently from the effects of the former on the later. Using mobile devices was associated with worse academic performance, without effects on other outcomes. In agreement with our results, a recent meta-analysis involving 58 studies with a total of 480,479 participants showed that academic performance was significantly associated with watching TV or playing videogames, but not with the total daily time spent in front of screens.
Our results suggest that the effects of screen time on sleep duration and academic performance are related to the characteristics of each activity performed in front of screens. Use of mobile devices for watching online contents, social networking, or gaming, was associated with lower academic performance. The impact of screen use on learning is further stressed by the findings of altered brain structure in children and adolescents spending increased time in front of screens [30,31]. While it is tempting to conclude that withdrawing screens will directly result in academic performance improvements, further studies are needed to test the efficacy of this intervention.
Bedtime use of screens also adversely affected sleep, increased daytime somnolence, and reduced academic performance. Blue light emitted by screens suppresses melatonin production, causing circadian disruption and impairing sleep onset and continuity [13,14,32]. Impaired sleep is, in turn, the leading cause of daytime somnolence [23,24], which increases significantly the risk of academic failure [23,24]. Interestingly, it has been recently shown that an intervention directed at reducing the use of screens before bedtime was accompanied by earlier sleep onset times, increased total sleep duration, and improved daytime vigilance .
The strengths of our paper included a large sample size, in which screen use, sleep patterns, daytime somnolence, and academic performance were evaluated simultaneously, and a rigorous modelling of the interrelationship between these variables. Implementation of Structural Equation Modelling adds further insights since this analytical approach allows considering measurement errors when assessing the relationships between pairs of variables, which can be combined in a multi-step path. Our study has some limitations. Its cross-sectional nature and the absence of targeted interventions aimed at reducing specific screen times preclude the assessment of causality between screen time, somnolence, and academic performance. Our results are based on self-reported screen use. Participants may not recall such exposures with precision or may be unwilling to disclose them with accuracy. Indeed, some participants reported screen use beyond 12 hours per day, likely an exaggeration, and therefore were excluded from the data analysis. Errors in reporting might have increased data variability, thus obscuring some relationships, but bias is unlikely. Of note, a small but significant difference in academic performance was noted between participants with or without missing data, suggesting the possibility of a selection bias. Regarding the SEM model, caution is advised when interpreting the regression weights of the log10-transformed variables. However, the direction of the association and its statistical significance is still adequately represented.
Adolescents who play video games for longer periods of time exhibit shorter sleep duration and higher probability of reporting daytime somnolence, whereas those that used mobile devices more extensively showed increased risk for reduced academic performance. In addition, reduced sleep duration was associated with daytime somnolence, which in turn adversely affected academic performance. Our findings highlight the importance of maintaining an adequate sleep hygiene and of limiting the use of screens to preserve learning among adolescents. Future studies should be designed while accounting for the findings reported herein.
We are grateful to all the teachers, students and parents who helped make this research possible. In addition, we would like to thanks “Austral Salud” for helping us design the electronic case report form.
Sleep in Adolescents Collaborative study group:
Joaquín Diez (Instituto Panamericano de Medicina del Sueño y Cronobiología, Buenos Aires, Argentina)
Guido Simonelli MD (Center for Advanced Research in Sleep Medicine, Hôpital du Sacré-Coeur de Montréal, Centre Intégré Universitaire de Santé et de Services Sociaux du Nord de l’Île-de-Montréal, Montreal, Quebec, Canada. Department of Medicine, Faculty of Medicine, Université de Montréal, Montreal, Quebec, Canada).
María D Perez Chada (Universidad del Salvador, Buenos Aires, Argentina)
Alan Gomez (Escuela Secundaria Antonio Toro, Pilar, Buenos Aires, Argentina)
Fabiana Fernandez (Equipo Interdisciplinario de Sueño Saludable, Chubut, Argentina)
Elsa Reyes (Universidad Austral, Buenos Aires, Argentina)
Carlos Elías (INSARES, Mendoza, Argentina)
Roque Baldasaria (Hospital Centro de Salud "Zenón J. Santillán", Tucumán, Argentina)
Miguel A. Penizzotto (Sanatorio San Roque. Curuzú Cuatiá. Corrientes, Argentina)
Ana Diez (Instituto Cardiovascular de Rosario, Santa Fe, Argentina)
Daniel E. Vigo (Instituto de Investigaciones BIOMED UCA-CONICET, Buenos Aires, Argentina)
Roberto Abouzed (Asociación Española de S.M., Chubut, Argentina)
Carlos Franceschini (Unidad de sueño y ventilación mecánica. Hospital Cosme Argerich, Buenos Aires, Argentina)
Mariano Garcia Carrillo (Clinica Davila, Santiago de Chile, Chile)
Daniel Mattar (Servicio de Tisioneumonología, Hospital Marcial Quiroga, San Juan, Argentina)
Luciano Busi (Hospital de Trelew, Chubut, Argentina)
Gustavo Douglas Nazareno (Sección Neumonologia del Hospital Vera Barros, La Rioja, Argentina)
Mireya Rodas (Universidad Espíritu Santo, Laboratorio Neumos, Guayaquil, Ecuador)
Michelle Alemán Rodas (Ministerio de Salud Pública, Guayaquil, Ecuador)
María Elide Blanc (Instituto Cardiovascular de Rosario, Rosario, Santa Fé, Argentina)
Daniela Mara Visentini (Hospital Antonio Cetrangolo, Vicente López, Argentina)
Jorge Sap (Tierra del Fuego, Argentina)
Gabriela Manonelles (Ministerio de Salud y Ambiente Santa Cruz y Policlínico Medisur, Santa Cruz, Argentina)
Luis Larrateguy (Centro Privado de Medicina Respiratoria de Paraná, Entre Ríos, Argentina)
Pedro Valiensi (Escuelas EETPySO 363 y EESOPI 8125, Santa Fe, Argentina)
Stella M Valiensi (Hospital Italiano de Buenos Aires, Buenos Aires, Argentina)
Patricia Maggio (Hospital de Rehabilitación María Ferrer y Sanatorio Güemes, Buenos Aires, Argentina)
Marcelino de Vega (Hospital Privado Santa Clara De Asis, Salta, Argentina)
- 1. COUNCIL ON COMMUNICATIONS AND MEDIA. Media Use in School-Aged Children and Adolescents. Pediatrics 2016;138(5). pmid:27940794.
- 2. Hale L, Guan S. Screen time and sleep among school-aged children and adolescents: a systematic literature review. Sleep Med Rev 2015;21:50–8. pmid:25193149.
- 3. Magee CA, Lee JK, Vella SA. Bidirectional relationships between sleep duration and screen time in early childhood. JAMA Pediatr 2014;168(5):465–70. pmid:24589672.
- 4. Kılıç AO, Sari E, Yucel H, et al. Exposure to and use of mobile devices in children aged 1–60 months. Eur J Pediatr 2019;178(2):221–7. pmid:30402717.
- 5. Fang K, Mu M, Liu K, He Y. Screen time and childhood overweight/obesity: A systematic review and meta-analysis. Child Care Health Dev 2019;45(5):744–53. pmid:31270831.
- 6. Goldfield GS, Kenny GP, Hadjiyannakis S, et al. Video game playing is independently associated with blood pressure and lipids in overweight and obese adolescents. PLoS One 2011;6(11):e26643. pmid:22069461.
- 7. Merghani A, Malhotra A, Sharma S. The U-shaped relationship between exercise and cardiac morbidity. Trends Cardiovasc Med 2016;26(3):232–40. pmid:26187713.
- 8. Biddle SJ, García Bengoechea E, Wiesner G. Sedentary behaviour and adiposity in youth: a systematic review of reviews and analysis of causality. Int J Behav Nutr Phys Act 2017;14(1):43. pmid:28351363.
- 9. Chahal H, Fung C, Kuhle S, Veugelers PJ. Availability and night-time use of electronic entertainment and communication devices are associated with short sleep duration and obesity among Canadian children. Pediatr Obes 2013;8(1):42–51. pmid:22962067.
- 10. Mihrshahi S, Drayton BA, Bauman AE, Hardy LL. Associations between childhood overweight, obesity, abdominal obesity and obesogenic behaviors and practices in Australian homes. BMC Public Health 2017;18(1):44. pmid:28732475.
- 11. Pérez-Farinós N, Villar-Villalba C, López Sobaler AM, et al. The relationship between hours of sleep, screen time and frequency of food and drink consumption in Spain in the 2011 and 2013 ALADINO: a cross-sectional study. BMC Public Health 2017;17(1):33. pmid:28056890.
- 12. Lissak G. Adverse physiological and psychological effects of screen time on children and adolescents: Literature review and case study. Environ Res 2018;164:149–57. pmid:29499467.
- 13. Cheung CH, Bedford R, Saez De Urabain IR, Karmiloff-Smith A, Smith TJ. Daily touchscreen use in infants and toddlers is associated with reduced sleep and delayed sleep onset. Sci Rep 2017;7:46104. pmid:28406474.
- 14. Wood B, Rea MS, Plitnick B, Figueiro MG. Light level and duration of exposure determine the impact of self-luminous tablets on melatonin suppression. Appl Ergon 2013;44(2):237–40. pmid:22850476.
- 15. Woods HC, Scott H. #Sleepyteens: Social media use in adolescence is associated with poor sleep quality, anxiety, depression and low self-esteem. J Adolesc 2016;51:41–9. pmid:27294324.
- 16. Curcio G, Ferrara M, De Gennaro L. Sleep loss, learning capacity and academic performance. Sleep Med Rev 2006;10(5):323–37. pmid:16564189.
- 17. Smith C. Sleep states and memory processes in humans: procedural versus declarative memory systems. Sleep Med Rev 2001;5(6):491–506. pmid:12531156.
- 18. Plihal W, Born J. Effects of early and late nocturnal sleep on declarative and procedural memory. J Cogn Neurosci 1997;9(4):534–47. pmid:23968216.
- 19. Maier JG, Nissen C. Sleep and memory: mechanisms and implications for psychiatry. Curr Opin Psychiatry 2017;30(6):480–4. pmid:28858009.
- 20. Peigneux P, Laureys S, Fuchs S, et al. Are spatial memories strengthened in the human hippocampus during slow wave sleep? Neuron 2004;44(3):535–45. pmid:15504332.
- 21. Kuhn M, Wolf E, Maier JG, et al. Sleep recalibrates homeostatic and associative synaptic plasticity in the human cortex. Nat Commun 2016;7:12455. pmid:27551934.
- 22. Dewald JF, Meijer AM, Oort FJ, Kerkhof GA, Bögels SM. The influence of sleep quality, sleep duration and sleepiness on school performance in children and adolescents: A meta-analytic review. Sleep Med Rev 2010;14(3):179–89. pmid:20093054.
- 23. Perez-Chada D, Perez-Lloret S, Videla AJ, et al. Sleep disordered breathing and daytime sleepiness are associated with poor academic performance in teenagers. A study using the Pediatric Daytime Sleepiness Scale (PDSS). Sleep 2007;30(12):1698–703. pmid:18246979.
- 24. Perez-Lloret S, Videla AJ, Richaudeau A, et al. A multi-step pathway connecting short sleep duration to daytime somnolence, reduced attention, and poor academic performance: an exploratory cross-sectional study in teenagers. J Clin Sleep Med 2013;9(5):469–73. pmid:23674938.
- 25. Adelantado-Renau M, Moliner-Urdiales D, Cavero-Redondo I, Beltran-Valls MR, Martínez-Vizcaíno V, Álvarez-Bueno C. Association Between Screen Media Use and Academic Performance Among Children and Adolescents: A Systematic Review and Meta-analysis. JAMA pediatrics 2019;173(11):1058–67. pmid:31545344.
- 26. Lien A, Sampasa-Kanyinga H, Colman I, Hamilton HA, Chaput JP. Adherence to 24-hour movement guidelines and academic performance in adolescents. Public Health 2020;183:8–14. pmid:32402739.
- 27. Drake C, Nickel C, Burduvali E, Roth T, Jefferson C, Pietro B. The pediatric daytime sleepiness scale (PDSS): sleep habits and school outcomes in middle-school children. Sleep 2003;26(4):455–8. pmid:12841372.
- 28. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. Bmj 2000;320(7244):1240–3. pmid:10797032.
- 29. Byrne BM. Structural Equation Modeling With AMOS: Basic Concepts, Applications, and Programming. Third ed. London, UK: Routledge; 2016.
- 30. Paulus MP, Squeglia LM, Bagot K, et al. Screen media activity and brain structure in youth: Evidence for diverse structural correlation networks from the ABCD study. Neuroimage 2019;185:140–53. pmid:30339913.
- 31. Hutton JS, Dudley J, Horowitz-Kraus T, DeWitt T, Holland SK. Associations Between Screen-Based Media Use and Brain White Matter Integrity in Preschool-Aged Children. JAMA Pediatr 2020;174(1):e193869. pmid:31682712.
- 32. Falbe J, Davison KK, Franckle RL, et al. Sleep duration, restfulness, and screens in the sleep environment. Pediatrics 2015;135(2):e367–75. pmid:25560435.
- 33. Perrault AA, Bayer L, Peuvrier M, et al. Reducing the use of screen electronic devices in the evening is associated with improved sleep and daytime vigilance in adolescents. Sleep 2019;42(9). pmid:31260534.