Tracking the impact of screen media activity on mental health, cognition and the brain from childhood to adolescence in the longitudinal ABCD study - Project Summary Today’s children spend an average of seven hours a day engaging with different types of Screen Media Activity (SMA), such as social media use, video gaming, watching TV or videos, and communicating with their peers. Mixed findings from small studies in adolescents and adults suggest that SMA is related to psychopathology, emotional regulation, and cognition. Importantly, there’s a lack of large longitudinal studies addressing if these effects are consequences of SMA or explaining how different types of SMA may impact child physical, emotional, cognitive and neural development. For example, do TV watching and video gaming have similar impact(s), if any, on children’s mental health and brain development? To address these matters, we propose to take advantage of the world’s largest multi-modal longitudinal neuroimaging dataset, the ABCD multi-site study, a landmark study following 11,880 demographically diverse children from ages 10 to 20 and is orders of magnitude larger than most neuroimaging studies. Our long-term goal is to advance understanding of how SMA combines with environment and polygenic variance to impact a broad range of outcomes such as child neurodevelopment, mental health, and cognitive function. The objective here is to identify the specific impact of active (e.g., video gaming), passive (e.g., watching TV and videos) and mixed (video chatting, social media use, texting) SMA types on mental health and neurocognition, and (2) identify probable causal mechanisms involving SMA with respect to moderating and potentially modifiable factors against detrimental effects of SMA. Our methodology includes applying robust non-parametric approaches and using twin studies and sophisticated analytic methods such as cross-lagged panels and Bayesian networks to identify probable causal mechanisms involving SMA across four ABCD data time points ranging from ages 9-10 to 15-16. We also propose using predictive machine learning models to publicly predict SMA outcomes in not-yet-gathered data from the ABCD release at age 18. The results of this proposal will not only be descriptive of the impacts of types of SMA, but will also be prescriptive, enabling practitioners to better understand the risks of various behaviors, as well as to enable the design of better interventions to mitigate or prevent the worst outcomes of SMA.
