Bidirectional Causal Pathways Between Addictive Screen Use and Youth Suicide: Risk, Protection, and Multilevel Mechanisms

In the US, 95% of teens have access to smartphones, 45% are online 'almost constantly,' and 85% play video games. Increases in social media use coincide with a youth mental health (MH) crisis: around 1 in 6 US youth aged 6-17 have MH problems, with nearly a 60% rise in suicidal ideation (SI) and suicide attempts (SA) since 2007, and widening disparities by race/ethnicity, sex, and gender. However, current research typically measures crude 'screen time' rather than addictive use patterns, relies on cross-sectional data precluding causal inference, overlooks early-life predisposing factors and protective mechanisms, and treats MH, SI, and suicidal behaviors (SB) as interchangeable outcomes. Our JAMA (2025) study demonstrated that youth following addictive screen use trajectories—Traj-1 (High/already addicted) and Traj-2 (Increasing)—had 2–3 times the risk for SB, whereas total screen time showed no significant association. Using updated ABCD 6.1 data (7 waves, ages 9–16), we replicated these trajectories with even larger risk ratios (RR>3.0) for SB. This project is organized around two sequential causal stages. Aim 1 determines how familial adversity (ACEs, parental SB/MH history), heritable neurodevelopmental diatheses (ADHD, polygenic risk scores from PGC GWAS), and early psychopathology predict high-risk addictive screen use trajectories across social media, mobile phone, and video game platforms, with moderation by SDOH patterns, school context, pubertal timing, sex, and race/ethnicity. Aim 2 estimates the potential causal effects of addictive trajectories on SB using instrumental variable methods (leave-one-out peer instruments, Mendelian randomization sensitivity analyses), decomposes total effects into risk mediators (depression, cyberbullying, sleep disturbance, family conflict) and protective mediators (parental monitoring, emotional regulation, peer prosocial connection) via VanderWeele four-way decomposition with g-computation, and tests bidirectionality using RI-CLPM across all seven ABCD waves. A Youth Advisory Board (YAB) of 12–15 members (ages 14–18) co-designs measures, co-interprets findings, and co-produces policy recommendations throughout the project lifecycle. The key innovation is applying a two-stage causal architecture—from adversity/diathesis through psychopathology to addiction trajectories, and from addiction trajectories through decomposed risk and protective mechanisms to SB—using the Adolescent Brain Cognitive Development (ABCD) Study (N=11,878; 7 waves, ages 9–18). Our findings will address urgent public health needs to prevent youth suicide, closely aligned with the NIMH Strategic Plan.