Are Addictive Screen Use Patterns Linked to Suicidal Behaviors and Mental Health Problems in U.S. Youth?

This study asked whether different trajectories of addictive screen use—specifically social media, mobile phone, and video game use—were associated with suicidal behaviors, suicidal ideation, and mental health outcomes in U.S. youths. Researchers analyzed longitudinal data from 4,285 participants in the Adolescent Brain Cognitive Development (ABCD) Study, ages 9–15, over four years. High or increasing trajectories of addictive screen use were significantly associated with elevated risks of suicidal behaviors (risk ratios up to 2.39), suicidal ideation (up to 1.53), and higher internalizing (T-score difference up to 2.03) and externalizing symptoms (up to 1.25) compared to low-use trajectories. Total screen time alone showed no such associations.

Full Citation and Link to Article

Xiao Y, Meng Y, Brown TT, Keyes KM, Mann JJ. Addictive Screen Use Trajectories and Suicidal Behaviors, Suicidal Ideation, and Mental Health in US Youths. JAMA. Published online June 18, 2025. doi:10.1001/jama.2025.7829 

Extended Summary

Central Research Question

This study investigated whether different trajectories of addictive screen use—specifically through social media, mobile phones, and video games—were associated with suicidal behaviors, suicidal ideation, and mental health outcomes in U.S. youths. The researchers aimed to determine whether patterns of escalating or persistently high addictive use, rather than total screen time, could predict elevated risks of suicidality and psychological symptoms across a four-year span. The study also sought to determine whether these associations persisted after controlling for total screen time and baseline mental health characteristics.

Previous Literature

Prior research has repeatedly raised concerns about the mental health implications of screen time in adolescents. However, much of the existing literature has focused on the overall amount of screen time, without accounting for how that use evolves over time or becomes addictive. Meta-analyses and longitudinal studies have linked elevated screen use to internalizing problems (like depression and anxiety) and externalizing behaviors (such as aggression), but these associations have often been small or inconsistent. Notably, emerging work suggests that problematic or addictive use—characterized by compulsive behavior, distress during disengagement, and functional impairment—may better explain poor outcomes than duration alone.

Additionally, existing studies have often failed to disaggregate the effects of different platforms (social media vs. video games, for example) or track patterns of use across developmental stages. The U.S. Surgeon General has recently emphasized the need for more rigorous, longitudinal evidence assessing addictive use patterns and their consequences. This study directly addresses those gaps by using a large, nationally representative dataset and validated measures of addictive use.

Data

Data were drawn from the Adolescent Brain Cognitive Development (ABCD) Study, a longitudinal cohort study of over 11,000 children across 21 U.S. sites. For this analysis, the sample was restricted to 4,285 participants (mean age 10.0 years at baseline) who had complete data from years 2 through 4 of follow-up (roughly ages 11–15). The sample was diverse in terms of sex (47.9% female), race (58.7% White, 19.4% Hispanic, 9.9% Black, and others), and socioeconomic background.

Addictive use data were collected annually using validated self-report instruments tailored to each screen type: the Social Media Addiction Questionnaire (SMAQ), the Mobile Phone Involvement Questionnaire (MPIQ), and the Video Game Addiction Questionnaire (VGAQ). Mental health outcomes, including suicidal behaviors and ideation, were assessed using the Kiddie Schedule for Affective Disorders and Schizophrenia (KSADS), based on both child and parent reports. Internalizing and externalizing symptoms were measured using the Child Behavior Checklist (CBCL), a widely used and validated parent-report instrument.

Methods

The authors used latent class linear mixed models to identify distinct trajectories of addictive use for each platform (social media, mobile phones, and video games) over the study period. These models allowed the researchers to capture developmental patterns in screen use that might otherwise be obscured in cross-sectional analyses. Each child was assigned to one of several trajectory groups depending on how their addictive use scores evolved over time.

For each screen type, the following addictive use trajectories were identified:

• Social media: high-peaking, increasing, and low (three classes)

• Mobile phone: high, increasing, and low (three classes)

• Video game: high and low (two classes)

Once these trajectories were established, Poisson regression models (for categorical outcomes like suicidal behaviors and ideation) and generalized linear models (for continuous CBCL symptom scores) were used to estimate the associations between trajectory group and outcomes. All models adjusted for baseline demographics (age, sex, race/ethnicity, household income, parental education and marital status) as well as baseline mental health indicators (prior suicidal ideation, suicidal behavior, internalizing and externalizing symptoms).

In addition, total screen time at baseline (reported separately for weekdays and weekends) was included as a covariate to test whether it attenuated the associations between trajectory group and mental health outcomes.

Findings/Size Effects

Nearly one-third of youth (31.3%) followed an increasing trajectory of addictive social media use, while 24.6% followed an increasing trajectory of mobile phone use. A striking 49.2% fell into the high-use mobile phone trajectory, and 41.1% were in the high-use video game trajectory. These results suggest that high or increasing levels of addictive screen use are common in early adolescence.

Key findings include:

• Social media: Youths in the high-peaking and increasing addictive use trajectories had more than twice the risk of suicidal behaviors (risk ratios of 2.39 and 2.14, respectively) compared with those in the low-use group. They also had elevated risks of suicidal ideation (risk ratios of 1.51 and 1.46). Internalizing and externalizing symptom scores were modestly but significantly higher for both high and increasing groups (T-score differences ranging from 1.05 to 1.27).

• Mobile phones: The high-use trajectory group had a risk ratio of 2.17 for suicidal behaviors and 1.50 for suicidal ideation, relative to the low group. Internalizing symptoms were slightly elevated (T-score difference of 0.66), but externalizing symptoms did not differ significantly.

• Video games: The high-use trajectory group had a risk ratio of 1.54 for suicidal behaviors and 1.53 for suicidal ideation. They also exhibited the highest increase in internalizing symptoms (T-score difference of 2.03) and elevated externalizing symptoms (0.94-point difference).

Importantly, total screen time at baseline was not associated with any of the suicide-related or mental health outcomes, either in bivariate models or when included as a covariate in the adjusted trajectory models. This suggests that the quality and pattern of screen use, rather than the quantity, is what matters most in predicting poor outcomes.

Effect sizes for mental health symptoms were generally small to moderate (Cohen’s d < 0.2), but consistent and statistically significant across domains. E-values indicated that moderate to strong unmeasured confounding would be required to nullify the observed associations, supporting their robustness.

Conclusion

This study offers some of the strongest longitudinal evidence to date that addictive screen use trajectories—particularly for social media, mobile phones, and video games—are significantly associated with suicidal behaviors, suicidal ideation, and poorer mental health outcomes in adolescents. The associations were consistent across platforms and persisted even after accounting for total screen time and baseline mental health.

The key takeaway is that the pattern of screen use—specifically increasing or consistently high levels of addictive use—is a more meaningful predictor of youth mental health risk than total time spent using screens. This finding has direct implications for clinicians, educators, and parents, suggesting that interventions should focus on monitoring problematic use behaviors rather than simply restricting screen time.

Given that many of these addictive use patterns began at age 10 and escalated over time, the study also underscores the importance of early identification and longitudinal tracking of screen-related behaviors. Future research should explore the role of moderating and mediating factors—such as sleep, peer dynamics, family context, and social support—and evaluate whether clinical or school-based screening tools can incorporate measures of addictive screen use to better identify at-risk youth.

In conclusion, this study moves the field beyond simplistic screen time metrics and provides a rigorous, developmentally informed framework for understanding how evolving patterns of digital engagement can impact youth mental health.