Massive Review of 9 Million Young People Reveals Screen Time Harms Depend on Age, Content and Context

One of the most exhaustive syntheses of digital media research ever conducted has concluded that the question dominating public debate about screens and young people is framed too narrowly. The central issue, according to the analysis, is not simply how much time children and adolescents spend on devices, but what they watch, how they engage, and with whom. The umbrella review and meta-analysis, published in eClinicalMedicine, pooled evidence from 121 systematic reviews covering roughly 2,790 unique primary studies and more than nine million participants aged 0 to 24 years, making it the most comprehensive assessment to date of how digital media use relates to health, behaviour and development.

The research team, led by Natarajan Padmapriya and Falk Müller-Riemenschneider at the National University of Singapore, systematically searched PubMed/MEDLINE, EMBASE, PsycINFO and the Cochrane Database for systematic reviews published between January 2021 and December 2025. Rather than treating screen time as a single homogeneous exposure, the investigators classified digital media use into seven categories: total or general screen time, social media use, problematic use, smartphone or tablet use, television viewing, video gaming, and screen-media content or context. Outcomes were organised into five domains spanning mental health and well-being, behaviour, physical health, neurocognitive and academic development, and social and psychosocial functioning. In total, the synthesis captured 316 distinct exposure–outcome associations.

A striking pattern emerged when the findings were stratified by developmental stage. In childhood, the evidence base is dominated by research on general screen time and media content, whereas in adolescence the focus shifts sharply toward social media use and problematic patterns of engagement, characterised by impaired control, preoccupation, and continued use despite adverse consequences. This developmental transition, the authors argue, is biologically and socially plausible: young children’s media use is largely caregiver-mediated, while adolescents gain autonomy, personal devices, and socially oriented digital lives as peer interaction and identity exploration become central to their development.

For children aged 0 to 10 years, the quantitative evidence linked higher screen time with poorer social-emotional development, increased hyperactivity and emotional problems, and shorter sleep. Each additional hour of smartphone or tablet use was associated with roughly eleven minutes less sleep, and a television in the child’s bedroom with about thirty minutes less sleep and worse sleep quality. Yet the synthesis also uncovered a genuinely hopeful counterpoint: educational screen content consumed with caregiver involvement and co-viewing was associated with substantially better language development, with a pooled odds ratio of 2.45. Conversely, fantastical or fast-paced content was associated with poorer cognitive and academic outcomes, and background television with weaker language and executive function skills.

The adolescent findings were considerably darker, particularly for problematic use. Across the 10 to 24 age range, problematic digital media use showed the strongest associations of any exposure: pooled odds ratios reached 3.26 for negative emotions and conduct problems, 2.45 for poorer quality of life, 2.30 for loneliness, 2.27 for common mental health problems, and 2.20 for self-harm, suicidal ideation and suicidal behaviour. Problematic use was also linked to sleep problems, reduced physical activity, and increased alcohol, tobacco and other substance use. The magnitude of the association between problematic use and mental health problems grew with age, rising from an odds ratio of 1.60 in those aged 10 to 18 to 2.40 in those aged 18 to 24, a group the authors note is largely absent from existing guidelines.

Ordinary social media use, distinct from problematic engagement, was associated with increased depression, with a pooled odds ratio of 1.60 that was stronger in girls, alongside anxiety, body dissatisfaction and disordered eating. Notably, the analysis found no association between general social media use and overall well-being, self-esteem or sleep. Content mattered enormously: body-positive content was associated with better well-being, whereas appearance-focused body-ideal content was associated with poorer well-being. Passive use was linked to mood disturbance and increased suicidality in girls, while online social capital and positive one-to-one interaction were associated with better well-being and reduced loneliness. Cybervictimisation was associated with self-harm and suicidal behaviours.

Physical health outcomes, including obesity, myopia, cardiometabolic risk and low back pain, were reported mainly in relation to overall screen time across age groups. Pooled estimates showed increased odds of myopia, overweight and obesity, metabolic syndrome risk, and earlier pubertal timing among heavier screen users. Some findings on video gaming ran against stereotype: action video games were associated with improved attention and processing speed, and active video games with reduced depression, anxiety and stress, improved sleep quality and greater physical activity. Gaming used as a coping strategy was associated with reduced depression and anxiety, and multiplayer games with reduced loneliness, although violent content and excessive duration were linked to aggression and addictive behaviour.

The authors are candid about the limitations of the evidence base. On the AMSTAR-2 quality assessment, only 18 percent of the included reviews were rated high quality, while 39 percent were rated critically low. Under the GRADE framework, 92 percent of reviews reached only low or very low certainty of evidence. The literature is dominated by cross-sectional observational designs, meaning reverse causation remains plausible for several findings, including the reported association between screen time and autism spectrum disorder, which may partly reflect greater screen exposure among children with pre-existing developmental characteristics. The authors also note that fewer than 5 percent of studies came from low- and middle-income countries, despite rapid growth in digital media use in those regions, and that self-reported measures of media use introduce measurement error.

Despite these caveats, the review concludes that the convergence of suggestive pooled associations, predominantly adverse and consistent narrative findings, and plausible developmental mechanisms provides a sufficient basis for precautionary public health action. For early childhood, the evidence supports minimising screen exposure and prioritising age-appropriate educational content with caregiver involvement. For adolescents, guidance should extend beyond raw screen time to address social media use and problematic engagement patterns. Because digital media exposures operate partly through platform-level mechanisms, the authors argue that policy responses should include transparency around algorithmic systems, restrictions on engagement-maximising design features, and regulation of advertising targeting young people, building on recent legislation such as the European Union’s Digital Services Act, the United Kingdom’s Online Safety Act 2023 and Australia’s social media minimum age law.

The review also issues a pointed challenge to the research community. Thousands of studies involving millions of participants have not translated into clear answers, the authors write, because of shortcomings in study design, measurement and reporting. Continued research using similar approaches may add volume without improving usefulness. Future work should prioritise longitudinal and experimental designs with objective, device- or app-based measures of exposure, broaden outcome assessment beyond mental health to include physical health, development and beneficial outcomes such as social connection, and expand evidence from under-studied regions and age groups. Emerging technologies, including virtual reality, AI-generated content and increasingly sophisticated recommendation algorithms, warrant prospective study before widespread adoption by young people. The pace of technological change, the authors acknowledge, will continue to outstrip the evidence, but the developmental patterns identified in this synthesis provide a foundation for age-appropriate, context-specific action that can be refined as better data emerge.

Subject of Research: Associations between digital media use and health, behavioural and developmental outcomes in children and adolescents

Article Title: Associations of digital media use with health, behavioural and developmental outcomes in children and adolescents: an umbrella review and meta-analysis

Article References: Associations of digital media use with health, behavioural and developmental outcomes in children and adolescents: an umbrella review and meta-analysis. (n.d.). https://doi.org/10.1016/j.eclinm.2026.104225

Image Credits: AI Generated

DOI: 10.1016/j.eclinm.2026.104225

Keywords: digital media, screen time, children, adolescents, mental health, social media, problematic use, umbrella review, meta-analysis, public health, child development, sleep