A randomized trial of 1,012 teenagers and young adults found that health warnings displayed on mock social media platforms increased perceived awareness of social media harms and perceived discouragement from use compared to neutral messages . However, the study measured only self-reported perceptions in a controlled setting, not real-world behavior change.
Researchers at JAMA Health Forum tested whether social media health warnings could influence how teenagers and young adults perceive the risks of platform use. The team conducted a within-participant randomized trial in December 2025 where 1,012 US participants aged 13 to 29 (mean age 19.3 years) viewed 15 different messages displayed on a mock social media page. These included 13 health warnings about specific harms (such as negative body image, depression and anxiety, sleep disruption, and addiction), one screen time break warning similar to existing voluntary features, and one neutral control message.
Participants rated each message on a 1 to 5 scale for two outcomes: how effective they perceived the message to be at discouraging social media use, and how much it increased their awareness of social media harms. The results showed consistent patterns across both measures. All health warnings were rated as significantly more effective than the neutral control message, with effect sizes ranging from 0.32 to 0.62 on the 5-point scale. Health warnings also outperformed the screen time break warning on perceived effectiveness for discouragement, though the addiction-focused warning showed a smaller advantage (effect size 0.07, with confidence intervals that crossed zero, suggesting less certainty).
For awareness of harms, the pattern held: all warnings, both health-specific and screen time focused, increased perceived awareness compared to control, with effect sizes ranging from 0.19 to 0.63. Health warnings again outperformed the screen time break message on this secondary outcome, with effect sizes ranging from 0.21 to 0.44. Notably, warnings about negative body image, sleep disruption, and mental health effects showed the largest perceived effects on both outcomes, while the addiction warning performed most weakly.
The trial's strength lies in its rigorous design, blinded randomization, and large sample size with balanced representation by age and gender. However, a critical limitation must be emphasized: the study measured only perceived effectiveness and stated awareness in a controlled laboratory environment with mock social media platforms. Participants rated their intentions and perceptions, not their actual social media use, behavior change, or long-term adherence. Whether displaying these warnings on actual social media platforms would produce measurable reductions in time spent on platforms, or sustained changes in mental health outcomes, remains unknown.
If you are a teenager or young adult, this research suggests that viewing warnings about social media harms could increase your conscious awareness of those risks. The findings support proposed state-level policies requiring social media warnings, at least as a mechanism for raising awareness. However, awareness alone does not guarantee behavior change. This study does not show that warnings would actually reduce your time on social media or prevent harms like depression or anxiety.
For parents and educators, the results indicate that explicit messaging about social media risks may be more effective than generic reminders to take breaks. The strongest perceived effects came from warnings about mental health and body image effects, suggesting these may be the most salient concerns for young people.
Policymakers considering warning requirements should view these findings as promising but incomplete evidence. A perceived increase in awareness of harms is a necessary first step toward behavior change, but real-world testing would be needed to determine whether warnings translate to actual reductions in social media use or improvements in mental health outcomes. Complementary interventions addressing habit formation, peer influence, and platform design may also be necessary.
| Attribute | Details |
|---|---|
| Study design | Within-participant randomized controlled trial |
| Sample size | 1,012 participants (509 women/girls, 491 men/boys, 12 nonbinary/other) |
| Age range | 13-29 years (mean 19.3 years) |
| Intervention | 15 different warning/control messages on mock social media page |
| Primary outcome | Self-reported perceived effectiveness of warnings for discouraging use (1-5 scale) |
| Secondary outcome | Self-reported increased awareness of harms (1-5 scale) |
| Key finding | All health warnings outperformed control; health warnings outperformed screen time break warning |
| Effect sizes | Primary outcome ADEs ranged 0.32-0.62; secondary outcome ADEs ranged 0.19-0.63 |
| Evidence tier | A tier (large RCT with pre-registration) |
| Limitations | Measured self-report in controlled setting; no measurement of actual behavior change; short-term assessment only |
| Journal | JAMA Health Forum |
| Publication date |
Designing Social Media Health Warnings for Teenagers and Young Adults: A Randomized Clinical Trial. JAMA Health Forum. 2025.
ProtocolEngine provides general health information based on published research. This is not medical advice. Consult a healthcare professional before starting any supplement or health protocol.
| Trial registration | NCT07199660 |