WHY LOOK UP
Young people, social media and mental health.
16 documented findings, with their geography, dates and limitations. Evidence to help explain what motivates our work.
Suicide statistics describe a public health context. They do not measure harm caused by social media. Association does not establish causation; an odds ratio (OR) is not a direct risk multiplier.
Digital habits
11 %
Signs of problematic social media use, up from 7% in 2018.
Ages 11, 13 and 15; Europe, central Asia and Canada. Not global prevalence.
Read the source13 % / 9 %
Girls more often reported problematic use.
Same HBSC coverage; a measure of control and negative consequences.
Read the source45 %
Reported nervousness or anxiety without their device nearby.
15-year-old students; a reported feeling, not an addiction diagnosis.
Read the source28 %
Reported digital distraction in most or every science lesson.
Students’ reported perception of their classroom.
Read the sourceOnline experiences
15 %
Around 15% reported experiencing cyberbullying.
Ages 11, 13 and 15; not representative of every young person worldwide.
Read the source29 %
Reported an offensive or upsetting online experience.
A category broader than cyberbullying.
Read the sourceOR 2,35
Association between cyberbullying victimisation and self-harm.
95% CI [1.65–3.34]; high heterogeneity and confounding.
Read the sourceOR 2,57
Association between cyberbullying victimisation and suicide attempts.
95% CI [1.69–3.90]; not an estimate of attributable deaths.
Read the sourceSleep and wellbeing
+88 %
Among girls, relative difference between passive use ≥7 h/day and <1 h/day in strongly endorsing becoming nervous easily.
RR 1.88 [1.79–1.97]; self-reported and cross-sectional, not a diagnosis or proof of causation.
Read the sourceOR 2,17
Association between bedtime device use and insufficient sleep.
95% CI [1.42–3.32]; heterogeneous observational studies, not 2.17 times the risk.
Read the sourceOR 2,72
Association with excessive daytime sleepiness.
95% CI [1.32–5.61]; historical, non-causal finding.
Read the source−1,67
Improvement in difficulties scores after two weeks of intensive recreational screen reduction.
95% CI [−2.68 to −0.67]; 181 participants, mostly children. No proof of lasting effects.
Read the sourcePublic health context
1 / 7
One in seven adolescents experiences a mental disorder.
Mental health context; not attributed to screens.
Read the source3rd
Suicide was the third leading cause of death in this age group.
A ranking of causes of death, not a measure of digital causation.
Read the source−35 %
The age-standardised suicide mortality rate declined.
This global improvement must also be visible.
Read the source+17 %
The same measure increased in the Americas.
Regional trajectories differ; this is not youth-specific.
Read the sourceRead the broader evidence
Harmful experiences, patterns of use and individual circumstances matter. Time spent alone does not describe every risk. The following syntheses help put striking findings into context.
- UNICEF · Childhood in a Digital World (2025)
- European Union · longitudinal and experimental evidence (2026)
- Fassi et al. · Systematic review and meta-analysis (2024)
- National Academies · Social Media and Adolescent Health (2024)
- GSHS · school surveys from 90 countries (2003–2017)
Method and transparency
This is a curated evidence resource, not an exhaustive systematic review. We prioritise public institutions, original research and scientific syntheses. Findings cannot be added together or directly compared. Different publications may use the same surveys and are not independent confirmations.
The UNICEF report finds no clear evidence that screen time alone causes the mental health problems it examines. The EU synthesis reports mixed, often small associations for general use, with more consistent findings for dysregulated use and harmful experiences.
Our ambition of life without a permanent smartphone remains a design and social choice. These studies do not establish the effectiveness of Look Up or imply that screen-free technology is beneficial by nature.