Health & Medicine
What Is Depression?
What the diagnosis requires, how dangerous depression is, and how much treatment helps.
Health & Medicine
Is Ultra-Processed Food Bad for You?
Two feeding trials, the cohort studies of death and disease, and which kinds of ultra-processed food carry the clearest risk.
Health & Medicine
How to Help Someone With Depression
Which advice for supporters has been tested, and which is charity guidance.
Does Social Media Harm Teenagers’ Mental Health?
The claim is that social media is behind the rise in teenagers’ depression and anxiety since the early 2010s, especially among girls. The counter-claim is that this is a moral panic. This page sets out the trends, the studies that test cause, and where the leading researchers disagree.
Jump to the verdict ↓
The short answer
Probably a little, for some teenagers. Heavier social media use goes with slightly worse mental health, and the link is larger for girls. The studies best able to show cause, of Facebook’s arrival at U.S. colleges and of trials that cut use, find small harms, but in college students and adults. In 2023 the National Academies of Sciences, Engineering, and Medicine found the evidence did not support social media causing changes in adolescent health at the population level.
How teenagers’ mental health has changed
The decline in teenagers’ reported mental health is real. In the Centers for Disease Control and Prevention’s (CDC) 2023 Youth Risk Behavior Survey (YRBS), 39.7% of U.S. high school students reported persistent sadness or hopelessness. The survey counts a student who, in the past year, felt so sad or hopeless almost every day for at least two weeks running that they stopped some usual activities. Girls reported it far more often than boys, 52.6% against 27.7%.
These problems were worsening before COVID: the share of all students reporting persistent sadness or hopelessness rose from 30% in 2013 to 37% in 2019.
CDC YRBS 2023 — U.S. High School Students2023
Persistent sadness/hopelessness (overall)39.7%
Persistent sadness/hopelessness (girls)52.6%
Persistent sadness/hopelessness (boys)27.7%
Seriously considered suicide (overall)20.4%
Attempted suicide (overall)9.5%
Poor mental health most of the time or always (past 30 days)28.5%
Source: CDC, Youth Risk Behavior Survey 2023, a nationally representative survey of U.S. high school students, with 20,103 usable questionnaires.
England and elsewhereEngland shows a similar rise. NHS England’s survey found that 20.3% of 8 to 16 year olds had a probable mental disorder in 2023, against 12.5% in 2017. For this age group, the rise came between 2017 and 2020; in the survey waves after 2020 it found no statistically significant change. Among 17 to 19 year olds the figure was 23.3% in 2023. In that age group it was twice as high in young women as in young men (31.6% against 15.4%).
NHS England compares later years with 2017 only on a shorter screening questionnaire, because the 2017 survey used a fuller diagnostic assessment.
Other rich countries show a fall in well-being, though not everywhere. UNICEF’s 2025 report card on rich countries compared 2018 with 2022. The share of 15-year-olds satisfied with their lives fell by more than 5% in 14 of the 32 countries with data. The drop was larger among girls in nearly all countries, and Japan was the only country where it rose substantially.
Among 11, 13 and 15 year olds, the share showing problematic social media use, with addiction-like signs, rose from 7% in 2018 to 11% in 2022. The figures come from the World Health Organization’s Health Behaviour in School-aged Children (HBSC) survey. In 2022 the survey asked almost 280,000 of these young people in 44 countries and regions about social media.
Persistent sadness or hopelessness among U.S. high school students, by sex, 2013 to 2023
| Year | Girls | Boys |
|---|---|---|
| 2013 | 39 | 21 |
| 2015 | 40 | 20 |
| 2017 | 41 | 21 |
| 2019 | 47 | 27 |
| 2021 | 57 | 29 |
| 2023 | 53 | 28 |
What the evidence says about cause
Most studies survey people once or follow them over time, using their own reports of screen time, and they find small associations. A meta-analysis pools the data of earlier studies on one subject. In one from 2020, psychologists at the University of Oregon (Ivie and colleagues) pooled 12 studies of adolescents aged 11 to 18.
They found a correlation (r) of 0.11 between more social media use and depressive symptoms. A correlation measures how closely two things rise and fall together: 0 means no straight-line link, and 1 a perfect one. A correlation of 0.11 means social media use accounts for about 1.2% of the variation in depressive symptoms.
The Surgeon General’s 2023 advisory cites a U.S. cohort, a group observed over time, of 6,595 adolescents aged 12 to 15. Those using social media for more than 3 hours a day faced double the risk of poor mental health outcomes, after adjusting for their earlier mental health.
Orben and Przybylski (2019) analysed three large surveys of 355,358 adolescents in the U.S. and the UK in many alternative ways. Digital technology use, not only social media, explained at most 0.4% of the variation in well-being. In one of the surveys, regularly eating potatoes was nearly as strongly linked to lower well-being; in another, wearing glasses was more strongly linked to it.
Self-reported screen time is a rough measure. A 2021 meta-analysis of 66 comparisons found that people’s own estimates of their use correlated only moderately with the use their devices logged (r=0.38).
Stronger causal evidenceIn the Adolescent Brain Cognitive Development (ABCD) study, when a child’s social media use rose above their own usual level, their depressive symptoms rose the following year. The estimates, with 95% confidence intervals, were β=0.07 (0.01–0.12) from year one to two and β=0.09 (0.04–0.14) from year two to three. Each β is an effect size, and each bracket a confidence interval, a range that describes the uncertainty around the estimate.
This observational study, by Nagata and colleagues, followed 11,876 U.S. children, aged 9–10 at the start, through four yearly waves (JAMA Network Open, May 2025). Its lead author was in the department of pediatrics at the University of California, San Francisco.
The study’s authors call the effect medium, though the year-one-to-two interval only just clears zero, and the link in the study’s first year was not significant. The reverse path (depression → later use) was not significant, and neither was the between-person comparison of heavier against lighter users (β=−0.01).
Braghieri, Levy and Makarin (American Economic Review, 2022) used Facebook’s staggered arrival at 775 U.S. colleges between 2004 and 2006 as a natural experiment. They matched it to more than 430,000 responses to the National College Health Assessment. The authors were at Bocconi University, Tel Aviv University and the Massachusetts Institute of Technology’s Sloan School of Management.
Facebook’s arrival worsened the authors’ index of poor mental health by 0.085 standard deviations (a standard deviation measures how widely values spread around their average). They translate that into about 2 percentage points more students who would screen positive for depression, 9% above a pre-Facebook level of 25%. The same 2 points for generalized anxiety disorder is 12% above a level of 16%.
The study covers college students on early Facebook, not today’s teens on TikTok or Instagram.
Two 2025 meta-analyses pooled trials that cut use. Burnell and colleagues combined 32 randomised studies, with 5,544 people, that told participants to cut down on or stop social media for a time. Burnell is a research assistant professor in psychology and neuroscience at the University of North Carolina at Chapel Hill. Well-being improved (g=0.17), an effect the authors call small in magnitude, and every study was of college students or adults.
May and colleagues pooled 10 trials with 1,491 people, average age 24. They found depressive symptoms fell (g=0.25 after adjusting for publication bias). They use Hedges’ g: the difference between two groups’ averages, divided by their standard deviation and corrected for small samples.
A 2025 study of 30 secondary schools in England (SMART Schools) compared pupils aged 12 to 15 at two kinds of school. Some do not allow recreational phone use during the school day; the others do. Pupils at the restrictive schools spent 0.67 hours less on their phones during school time, but no less on weekdays or at weekends as a whole. Their mental wellbeing was no better. The study compared schools at one point in time, so it shows an association, not the effect of introducing a ban.
Where Haidt and Odgers disagree
In a 2023 essay, Jonathan Haidt concluded that “Social Media is a Major Cause of the Mental Illness Epidemic in Teen Girls.” He is a social psychologist at New York University. He calls social media the largest single factor, while saying it is not a one-factor story. He cites correlational, longitudinal and experimental studies, and natural experiments such as Facebook’s arrival at colleges. He notes that most experiments are done on college students or young adults.
Candice Odgers is a professor of psychological science and informatics at the University of California, Irvine. She reviewed Haidt’s 2024 book The Anxious Generation in Nature. She wrote that research has produced “a mix of no, small and mixed associations” and that most data are correlative. Where associations over time are found, she wrote, they suggest that young people who already have mental-health problems use the platforms more often or differently. She warned that blaming social media “might distract us from effectively responding to the real causes” of the crisis.
Amy Orben (Cambridge) and Andrew Przybylski (Oxford) argue that big surveys can be analysed in so many ways that small effects get reported as significant. Jean Twenge, of San Diego State University, reanalysed Orben and Przybylski’s three surveys with Haidt and two others. Among other changes, they looked at social media rather than all screen time and at girls separately, and found much larger associations among girls.
Mitch Prinstein, chief science officer of the American Psychological Association (APA), co-chaired its 2023 health advisory. The advisory says social media is “not inherently beneficial or harmful to young people.” Its effects, it says, likely depend on what teens can do and see online, their own strengths or vulnerabilities, and the contexts they grow up in.
Why they disagreeCausal StandardsHaidt counts these kinds of study together as enough to call social media a major cause. Odgers says such serious claims require serious evidence, and warns against reading cause into trend lines that rise together.
Effect Size InterpretationHaidt argues that studies of individuals probably understate the harm, because “much and perhaps most of the action is in the emergent network effects” on whole peer groups. Odgers reads the research as a mix of no, small and mixed associations.
Subgroup WeightingHaidt stresses that the associations are larger for girls. Odgers stresses the direction of the associations over time, described above.
Timing InterpretationHaidt argues that only social media explains why the rise began so suddenly, around 2012, in several countries. Odgers points to other pressures on this generation: the aftermath of the 2008 recession, child poverty, the opioid crisis, school shootings, and racial and sexual discrimination.
Threshold for ActionOdgers calls many of Haidt’s proposals reasonable, such as stricter content moderation, and agrees that platforms need considerable reform. She says others, such as age-based restrictions and bans on mobile devices, are unlikely to work and could backfire.
How girls and boys differ
The sex differences are large. In the U.S. survey, girls’ persistent sadness or hopelessness rose from 39% in 2013 to 53% in 2023, and boys’ from 21% to 28%. The World Health Organization’s HBSC survey found problematic social media use in 13% of girls and 9% of boys in 2022.
The link with social media use is also larger for girls. In the UK’s Millennium Cohort Study of 10,904 14-year-olds, heavier social media use went with higher depressive symptom scores in both sexes, and more so in girls (chart below). It is an observational study, so it shows an association, not a cause.
Twenge and colleagues’ reanalysis found that more use went with worse mental health among girls. Across their many versions of the analysis, the middle values of the link’s strength, its median betas, ran from −0.11 to −0.24. In Orben and Przybylski’s original analysis they were −0.01 to −0.04.
Depressive symptom scores of 14-year-olds by daily social media use, compared with 1 to 3 hours a day, UK
| Daily social media use | Girls | Boys |
|---|---|---|
| 3 to under 5 hours | 26 | 21 |
| 5 hours or more | 50 | 35 |
The Millennium Cohort Study traced the link through online harassment, poor sleep, low self-esteem and poor body image, each of which went with higher depressive symptom scores. The Surgeon General’s 2023 advisory says social media may perpetuate body dissatisfaction, disordered eating, social comparison and low self-esteem, especially among adolescent girls. It adds that girls and sexual minority youth are more likely to report being cyberbullied.
Orben and colleagues (2022), using UK data, found that higher social media use predicted lower life satisfaction a year later at particular ages: 11 to 13 for girls, 14 to 15 for boys, and 19 for both.
Mechanisms of harm and benefit
Sleep is one path with consistent evidence behind it. The Surgeon General’s advisory cites a review of 42 studies that found a consistent relationship between social media use and poor sleep quality, shorter sleep, sleep difficulties and depression among young people. On cyberbullying, it cites a review of 36 studies that found a consistent relationship between cyberbullying through social media and depression among children and adolescents. Body image and social comparison, covered in the section on girls and boys, are a third path.
Problematic use, with addiction-like signs, shows stronger associations than time spent. Shannon and colleagues’ 2022 meta-analysis of studies of adolescents and young adults found correlations of 0.27 between problematic use and depression and 0.35 with anxiety. The 2020 meta-analysis of time spent found 0.11 with depression.
BenefitsThe advisories and reviews also describe benefits. The APA advisory says online social support may benefit young people’s development, particularly during periods of social isolation, and perhaps especially for youth who experience adversity or isolation offline. The Surgeon General’s advisory says the support peers may offer online can be especially important for youth who are often marginalized, including racial, ethnic, and sexual and gender minorities.
The National Academies say social media can be a place of safety and support for lesbian, gay, bisexual, trans, questioning, and other (LGBTQ+) teens. The same holds, they say, for those dealing with serious illness, bereavement or mental health issues. In the 2018 HBSC survey, heavy users whose use was not problematic reported the most peer support.
The Surgeon General’s advisoryThe 2023 advisory does not claim proof that social media causes harm. It states: “At this time, we do not yet have enough evidence to determine if social media is sufficiently safe for children and adolescents.” It also says there are “ample indicators” that social media can carry “a profound risk of harm” to young people’s mental health. Its section on harms cites the Facebook college study, a cohort study, two trials in young adults and adults, and correlational research.
Its recommendations to policymakers include age-appropriate health and safety standards for platforms, data sharing with independent researchers, and policies that “further limit access” to social media for all children, including strengthening and enforcing age minimums. It does not call for a ban.
Steelman both sides
Trend TimingHaidt’s argument from the timing around 2012, set out above. Timing is suggestive even if not proof.
Causal EvidenceThe Facebook college study, the within-person ABCD findings and the trials that cut use all point the same way, though the trials were of college students and adults.
MechanismsThe paths described above, with girls reporting more cyberbullying and more problematic use than boys.
Population ScaleIn 2022 up to 95% of U.S. 13 to 17 year olds used social media, so a small average effect could reach many teenagers. The Surgeon General’s advisory names adolescent girls and those already experiencing poor mental health as of higher relative concern.
Tiny Effect SizesOrben and Przybylski suggest that effects of at most 0.4% of the variation in well-being are too small to warrant policy change. Orben and colleagues also found that lower life satisfaction predicted later increases in use.
Measurement ProblemsSelf-reports match logged use only moderately, and totals of “screen time” lump different activities together.
Alternative ExplanationsThe pressures Odgers lists bear on the same generation. Vuorre and Przybylski (2023), of the University of Oxford’s Oxford Internet Institute, used Gallup surveys of 946,798 people in 72 countries from 2008 to 2019. Facebook’s spread was not linked to lower well-being, and the link was generally more positive for people under 35.
Benefits for SubsetsMarginalized youth can find support online, as described above.
Historical PrecedentWorries about new technologies recur, from novels to radio to smartphones, and Orben (2020) argues that research seemingly restarts with each one. This does not prove current fears are wrong, but it counsels against overclaiming.
How sure each finding is
| Finding | Confidence | Basis |
|---|---|---|
| Teenagers’ reported mental health has worsened since the early 2010s, especially for girls | HIGH | CDC YRBS; NHS England; UNICEF |
| Social media use is associated with worse mental health | HIGH | A 12-study meta-analysis, r of 0.11; the National Academies |
| Social media is a major/dominant causal driver | LOW | The National Academies’ 2023 review |
| Social media is a modest contributing factor | MODERATE | ABCD, the Facebook college study, trials that cut use: converging, mostly small |
| The link between use and poor mental health is larger for girls than boys | HIGH | Kelly and colleagues (UK, age 14); Twenge and colleagues’ reanalysis of three surveys |
| Problematic use is more strongly linked to poor mental health than time spent | HIGH | r of 0.27 for problematic use and depression, against 0.11 for time spent |
| Poor sleep is one path from heavy use to poor mental health | MODERATE | A 42-study review cited by the Surgeon General; Kelly and colleagues; all observational |
| Social media benefits some teenagers | MODERATE | APA; the National Academies; HBSC |
The verdict: what the evidence shows
Probably a little, for some teenagers, and less than the strongest claims say. Across studies, more social media use goes with slightly worse symptoms, a correlation of about 0.11 with depressive symptoms. The designs best able to show cause find small harms: Facebook’s arrival at colleges and trials that cut use, which were in college students and adults.
None of these studies shows social media to be the main cause of the decline. The National Academies’ 2023 review found the evidence did not support social media causing changes in adolescent health at the population level. The links are larger for girls and for problematic, addiction-like use, so where harm occurs it looks concentrated rather than spread evenly across a generation.
Three common claims, against the evidence
Claim“Social media is destroying a generation”
OverstatedHeavier use goes with slightly worse mental health: a correlation of 0.11 with depressive symptoms, about 1.2% of their variation. The National Academies found the evidence did not support a population-level effect. The link is larger for girls.
Claim“This is just moral panic, there’s no real evidence”
Also overstatedFacebook’s arrival at colleges raised the share of students who would screen positive for depression from about 25% to 27%. Trials that cut use improved well-being (g of 0.17) and depressive symptoms (0.25), though in college students and adults.
Claim“The Surgeon General proved social media harms youth”
MisstatedThe 2023 advisory says there is not yet enough evidence to determine whether social media is sufficiently safe for children, while citing “ample indicators” of risk. It calls for limiting children’s access, not for a ban.
Download the conclusion
The verdict above as one image, with the share of U.S. high school students reporting persistent sadness or hopelessness: girls, 39% in 2013 and 53% in 2023; boys, 21% and 28%. The image is 1,200 by 630 pixels, the size social networks use for link previews, and anyone may download it and post it.
What would change this conclusion
This article concludes that: (1) teenagers’ reported mental health has worsened in the U.S. since the early 2010s, most of all for girls, and in England since 2017; (2) heavier social media use goes with slightly worse mental health, with larger links for girls and for problematic use; (3) the studies best able to show cause find small harms, in college students and adults, and do not show social media to be the main cause of the decline.
These conclusions would be falsified by:
• Randomised trials in adolescents, or natural experiments on today’s platforms, finding that cutting or removing access produces large, lasting falls in depression or anxiety. That would make social media a major cause
• Large preregistered panels using logged rather than self-reported use, finding no link even for heavy-using girls and for teenagers with problematic use. That would weaken the case for any harm
• Trials of school phone bans or age limits finding measurable improvements in pupils’ mental health. That would strengthen the case that cutting access helps
• Platform data on what teenagers see, opened to independent researchers, showing which content or features drive the links
Sources
Centers for Disease Control and Prevention. Youth Risk Behavior Survey Data Summary & Trends Report: 2013–2023. U.S. Department of Health and Human Services, 2024. Source for the share of all students reporting persistent sadness or hopelessness from 2013 to 2023, girls’ and boys’ rates in 2013, 2021 and 2023, the CDC’s reading of the change from 2021 to 2023, and the 20,103 usable questionnaires of 2023. cdc.gov
Verlenden JV, Fodeman A, Wilkins N, et al. Mental Health and Suicide Risk Among High School Students and Protective Factors — Youth Risk Behavior Survey, United States, 2023. MMWR Supplements, 2024;73(4):79–86. Source for the 2023 figures: persistent sadness or hopelessness among all students, girls and boys; seriously considered and attempted suicide; and poor mental health most of the time or always in the past 30 days. cdc.gov
Kann L, McManus T, Harris WA, et al. Youth Risk Behavior Surveillance — United States, 2015. MMWR Surveillance Summaries, 2016;65(6). Source for girls’ and boys’ persistent sadness in 2015 (39.8% and 20.3%), plotted rounded. cdc.gov
Kann L, McManus T, Harris WA, et al. Youth Risk Behavior Surveillance — United States, 2017. MMWR Surveillance Summaries, 2018;67(8). Source for girls’ and boys’ persistent sadness in 2017 (41.1% and 21.4%), plotted rounded. pmc.ncbi.nlm.nih.gov
Bitsko RH, Claussen AH, Lichstein J, et al. Mental Health Surveillance Among Children — United States, 2013–2019. MMWR Supplements, 2022;71(2). Source, from its Table 6, for girls’ and boys’ persistent sadness in 2019 (46.6% and 26.8%), plotted rounded. cdc.gov
NHS England. Mental Health of Children and Young People in England, 2023: wave 4 follow up to the 2017 survey. Newlove-Delgado T, Marcheselli F, Williams T, et al. 21 November 2023. Source for probable mental disorder among 8 to 16 year olds in 2017, 2020 and 2023, the absence of significant change after 2020, the 17 to 19 year olds’ figures by sex, and the use of the Strengths and Difficulties Questionnaire for comparisons with 2017. digital.nhs.uk
UNICEF Innocenti. Child Well-Being in an Unpredictable World. Report Card 19, May 2025. Source for the fall from 2018 to 2022 in the share of 15-year-olds satisfied with their lives, in 14 of the 32 OECD and European Union countries with data, its larger size among girls, and Japan as the only country with a substantial rise. unicef.org
WHO Regional Office for Europe. Teens, screens and mental health. News release, 25 September 2024, on the Health Behaviour in School-aged Children (HBSC) report “A focus on adolescent social media use and gaming in Europe, central Asia and Canada”, volume 6 of the 2021/2022 survey. Source for the almost 280,000 11, 13 and 15 year olds in 44 countries and regions, problematic social media use of 7% in 2018 and 11% in 2022, 13% of girls and 9% of boys, and, from the 2017/18 survey, heavy but non-problematic users’ greater peer support. who.int
Ivie EJ, Pettitt A, Moses LJ, Allen NB. A meta-analysis of the association between adolescent social media use and depressive symptoms. Journal of Affective Disorders, 2020;275:165–174. Source for the 12 studies of adolescents aged 11 to 18, the correlation of 0.11 between social media use and depressive symptoms, and the authors’ department of psychology at the University of Oregon (as Europe PMC’s record of the paper lists it). doi.org
Plain definitions of statistical terms. National Cancer Institute, Dictionary of Cancer Terms: meta-analysis, “a process that analyzes data from different studies done about the same subject”; cohort, as used in medicine, “a group that is part of a clinical trial or study and is observed over a period of time”. HtaGlossary.net: correlation coefficient, “a number between -1 and +1 that expresses the strength of the linear association between two numerical variables”, noted r in a sample, 0 meaning no linear relationship and +1 a perfect positive one. U.S. Census Bureau, A Basic Explanation of Confidence Intervals (page last revised 8 October 2021): “a range of values that describes the uncertainty surrounding an estimate”. EUPATI Toolbox glossary, standard deviation: “a measure of the amount of variation within a data set”, the values lying close to the mean (average) when it is near zero. HtaGlossary.net: standardised mean difference, “the difference between two means, divided by an estimate of the within-group standard deviation”. cancer.gov htaglossary.net census.gov eupati.eu htaglossary.net
Shannon H, Bush K, Villeneuve PJ, Hellemans KGC, Guimond S. Problematic Social Media Use in Adolescents and Young Adults: Systematic Review and Meta-analysis. JMIR Mental Health, 2022;9(4):e33450. Source for the correlations of problematic social media use with depression (0.27) and anxiety (0.35). doi.org
Orben A, Przybylski AK. The association between adolescent well-being and digital technology use. Nature Human Behaviour, 2019;3:173–182. Source for the three surveys of 355,358 adolescents, the at most 0.4% of variation in well-being, the potatoes and glasses comparisons, the argument about analytical flexibility, and the judgement that the effects are too small to warrant policy change. nature.com
Parry DA, Davidson BI, Sewall CJR, Fisher JT, Mieczkowski H, Quintana DS. A systematic review and meta-analysis of discrepancies between logged and self-reported digital media use. Nature Human Behaviour, 2021;5:1535–1547. Source for the 66 comparisons of self-reported with logged use and their moderate correlation (r=0.38). nature.com
Nagata JM, Otmar CD, Shim J, et al. Social Media Use and Depressive Symptoms During Early Adolescence. JAMA Network Open, 2025;8(5):e2511704. Source for the 11,876 children of the ABCD study, the four yearly waves, the within-person and between-person estimates, the absent reverse path, the authors’ reading of each β as an effect size, and the lead author’s department. jamanetwork.com
Braghieri L, Levy R, Makarin A. Social Media and Mental Health. American Economic Review, 2022;112(11):3660–93. Source for Facebook’s arrival at 775 colleges from 2004 to 2006, the more than 430,000 survey responses, the 0.085 standard deviation rise in the index of poor mental health, the authors’ translation into depression and anxiety, and, from Bocconi University’s news story of 18 October 2022, the authors’ universities. aeaweb.org unibocconi.it
Burnell K, Meter D, Andrade F, Slocum A, George M. The effects of social media restriction: meta-analytic evidence from randomized controlled trials. SSM – Mental Health, 2025;7:100459. Source for the 32 randomised studies of 5,544 college students and adults the pooled effect on well-being (g=0.17), small in the authors’ words, and, from the University of North Carolina’s Winston Center profile, Burnell’s post. doi.org winstoncenter.unc.edu
May W, Malouff JM, Meynadier J. Reducing Social Media Use Decreases Depression Symptoms: A Meta-Analysis of Randomised Controlled Trials. European Journal of Investigation in Health, Psychology and Education, 2025;15(11):222. Source for the 10 trials of 1,491 people, average age 24, the fall in depressive symptoms (g=0.25 after adjusting for publication bias), and their use of Hedges’ g, “the standardised mean difference corrected for bias in small samples”. pmc.ncbi.nlm.nih.gov
Goodyear VA, Randhawa A, Adab P, et al. School phone policies and their association with mental wellbeing, phone use, and social media use (SMART Schools): a cross-sectional observational study. The Lancet Regional Health – Europe, 2025;51:101211. Source for the 30 English secondary schools, the pupils aged 12 to 15, the lower phone use during school time at restrictive schools, and the absence of a difference in mental wellbeing or in overall use on weekdays and at weekends. doi.org
Haidt J. Social Media is a Major Cause of the Mental Illness Epidemic in Teen Girls. Here’s the Evidence. After Babel, 22 February 2023. Source for Haidt’s conclusion, his account of the timing around 2012, the kinds of study he cites, his remark that it is not a one-factor story, and his argument about network effects. afterbabel.com
Odgers CL. The great rewiring: is social media really behind an epidemic of teenage mental illness? Nature, 2024;628:29–30. Source for Odgers’s post and her reading of the research, the direction of the associations over time, the other pressures she names, and her view of Haidt’s proposals. nature.com
Twenge JM, Haidt J, Lozano J, Cummins KM. Specification curve analysis shows that social media use is linked to poor mental health, especially among girls. Acta Psychologica, 2022;224:103512. Source for the reanalysis of Orben and Przybylski’s three surveys and the girls’ median betas of −0.11 to −0.24. doi.org
Kelly Y, Zilanawala A, Booker C, Sacker A. Social Media Use and Adolescent Mental Health: Findings From the UK Millennium Cohort Study. EClinicalMedicine, 2019;6:59–68. Source for the 10,904 14-year-olds, the higher depressive symptom scores with heavier use for girls and boys, and the paths through online harassment, sleep, self-esteem and body image. doi.org
Orben A, Przybylski AK, Blakemore S-J, Kievit RA. Windows of developmental sensitivity to social media. Nature Communications, 2022;13:1649. Source for the ages at which higher use predicted lower life satisfaction a year later, the reverse prediction, and the authors’ universities. nature.com
American Psychological Association. Health Advisory on Social Media Use in Adolescence. May 2023. Source for the statement that social media is not inherently beneficial or harmful to young people, its account of benefits, and Mitch Prinstein’s post as co-chair. apa.org
U.S. Surgeon General. Social Media and Youth Mental Health: The U.S. Surgeon General’s Advisory. U.S. Department of Health and Human Services, 2023. Source for the quotation on safety, the “ample indicators” of risk, the studies it cites, the reviews of sleep and cyberbullying, the groups of higher concern, its account of benefits, and its recommendations to policymakers. hhs.gov
National Academies of Sciences, Engineering, and Medicine. Social Media and Adolescent Health. Report highlights, December 2023. Source for the committee’s finding on population-level effects, the small effects and weak associations in the research, and the support social media can give LGBTQ+ teens and others, with the highlights’ full form of LGBTQ+. nationalacademies.org
Vuorre M, Przybylski AK. Estimating the association between Facebook adoption and well-being in 72 countries. Royal Society Open Science, 2023. Source, as the Oxford Internet Institute’s release on the study reports it, for Gallup well-being data on 946,798 people in 72 countries from 2008 to 2019, the absence of a link between Facebook’s spread and lower well-being, and the more positive link for younger people; the authors’ Oxford Internet Institute affiliation is from the paper itself. oii.ox.ac.uk
Orben A. The Sisyphean Cycle of Technology Panics. Perspectives on Psychological Science, 2020. Source for the recurring worries about new technologies and the argument that research restarts with each one. repository.cam.ac.uk
Cite this article
TruthBased.org. “Does Social Media Harm Teenagers’ Mental Health?” Updated November 2026. https://www.truthbased.org/does-social-media-harm-teenagers-mental-health