JRC social media and youth mental health review
This 2026 systematic review by the European Commission’s Joint Research Centre (JRC) examines longitudinal and experimental evidence on social-media use and mental health and well-being among young people.
The authors included 55 studies from 4,188 screened records: 41 longitudinal studies and 14 experimental studies. They distinguish time or frequency of use, particular behaviors and experiences, and problematic or dysregulated use. This avoids treating “social media use” as one homogeneous exposure.
Main findings
The review does not find one uniform association between time or frequency of use and poor mental health. Results are mixed, often small, and sensitive to the population, outcome, measure, and statistical model. Reverse and bidirectional pathways remain plausible: distress can alter social-media use as well as follow it.
More consistent risks concern particular mechanisms and experiences, including problematic use, social comparison and appearance pressure, sleep disruption, and encounters with harmful content or interaction. Experimental evidence is clearest for some short-run effects on appearance-related self-evaluation; broad changes in symptoms or well-being are inconsistent.
The literature also contains heterogeneous and positive experiences. Connection, support, expression, and active participation can coexist with harms, and average effects conceal substantial differences between individuals. The report therefore frames the policy question as the conditions under which social media supports or harms well-being, not whether every unit of use has the same effect.
Limits and policy significance
The evidence base relies heavily on self-report, uses varying definitions and outcomes, and often has short follow-up periods. Experimental samples lean toward older adolescents and university students. Most randomized studies had at least some risk-of-bias concern, and broad restrictions can displace activity to other communication channels.
The review supports interventions aimed at a specified mechanism more strongly than generic screen-time reduction or exclusion. It does not establish that every proposed design change is effective; those changes still require outcome-based testing. It is the principal evidence review used in Less harmful social media.