Screen Time and Autism Spectrum Disorder: A Systematic Review and Meta-Analysis

Yaakov Ophir,Hananel Rosenberg,Refael Tikochinski,Shani Dalyot,Yuliya Lipshits-Braziler

DOI: https://doi.org/10.1001/jamanetworkopen.2023.46775

2023-12-01

Abstract:Importance: Contemporary studies raise concerns regarding the implications of excessive screen time on the development of autism spectrum disorder (ASD). However, the existing literature consists of mixed and unquantified findings. Objective: To conduct a systematic review and meta-analyis of the association between screen time and ASD. Data sources: A search was conducted in the PubMed, PsycNET, and ProQuest Dissertation & Theses Global databases for studies published up to May 1, 2023. Study selection: The search was conducted independently by 2 authors. Included studies comprised empirical, peer-reviewed articles or dissertations published in English with statistics from which relevant effect sizes could be calculated. Discrepancies were resolved by consensus. Data extraction and synthesis: This study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) reporting guideline. Two authors independently coded all titles and abstracts, reviewed full-text articles against the inclusion and exclusion criteria, and resolved all discrepancies by consensus. Effect sizes were transformed into log odds ratios (ORs) and analyzed using a random-effects meta-analysis and mixed-effects meta-regression. Study quality was assessed using the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) approach. Publication bias was tested via the Egger z test for funnel plot asymmetry. Data analysis was performed in June 2023. Main outcomes and measures: The 2 main variables of interest in this study were screen time and ASD. Screen time was defined as hours of screen use per day or per week, and ASD was defined as an ASD clinical diagnosis (yes or no) or ASD symptoms. The meta-regression considered screen type (ie, general use of screens, television, video games, computers, smartphones, and social media), age group (children vs adults or heterogenous age groups), and type of ASD measure (clinical diagnosis vs ASD symptoms). Results: Of the 4682 records identified, 46 studies with a total of 562 131 participants met the inclusion criteria. The studies were observational (5 were longitudinal and 41 were cross-sectional) and included 66 relevant effect sizes. The meta-analysis resulted in a positive summary effect size (log OR, 0.54 [95% CI, 0.34 to 0.74]). A trim-and-fill correction for a significant publication bias (Egger z = 2.15; P = .03) resulted in a substantially decreased and nonsignificant effect size (log OR, 0.22 [95% CI, -0.004 to 0.44]). The meta-regression results suggested that the positive summary effect size was only significant in studies targeting general screen use (β [SE] = 0.73 [0.34]; t58 = 2.10; P = .03). This effect size was most dominant in studies of children (log OR, 0.98 [95% CI, 0.66 to 1.29]). Interestingly, a negative summary effect size was observed in studies investigating associations between social media and ASD (log OR, -1.24 [95% CI, -1.51 to -0.96]). Conclusions and relevance: The findings of this systematic review and meta-analysis suggest that the proclaimed association between screen use and ASD is not sufficiently supported in the existing literature. Although excessive screen use may pose developmental risks, the mixed findings, the small effect sizes (especially when considering the observed publication bias), and the correlational nature of the available research require further scientific investigation. These findings also do not rule out the complementary hypothesis that children with ASD may prioritize screen activities to avoid social challenges.