Study of 292,000 children finds screen use both predicts and follows emotional struggles

A new study provides the strongest evidence to date that screen use and socioemotional problems in children are linked in both directions. Published in the journal Psychological Bulletin, the comprehensive meta-analysis of 117 longitudinal studies involving more than 292,000 children found that increased screen use slightly predicted later emotional or behavioral difficulties—and that children with existing socioemotional problems were also slightly more likely to use screens over time.

Much of the earlier research on screen use and child mental health has been cross-sectional. That means it captures a snapshot in time, rather than tracking how behaviors and outcomes change over months or years. Cross-sectional studies are useful for identifying correlations, but they cannot establish which came first—screens or symptoms. Longitudinal studies, by contrast, follow the same children over time, offering stronger clues about directionality.

Even among longitudinal studies, individual results have varied, and many suffered from small sample sizes. To resolve this, the authors of the new meta-analysis aggregated data from 117 studies spanning decades and countries. They used advanced statistical techniques to evaluate whether screen use predicted socioemotional problems later on—and whether socioemotional problems at one point in time predicted greater screen use later.

They also explored how these associations were influenced by screen type, content, and purpose, and whether effects differed for internalizing problems (like anxiety and low self-esteem) versus externalizing ones (like aggression or defiance).

“I’ve got three kids under 7, and screens are everywhere in their lives,” said study author Michael Noetel, an associate professor at The University of Queensland Brisbane. “My wife and I sometimes disagree about screen rules at home, and I genuinely didn’t know if screens were causing problems—so my team spent 4 years trying to find out (though in a classic husband own-goal, the research said she was mostly right!)”

Free daily newsletter

The researchers searched five major academic databases and conducted extensive citation tracking to identify relevant studies. To qualify, a study had to measure screen use and socioemotional problems at two or more points in time, separated by at least six months, in children under age 10. The review excluded studies conducted during COVID-19 lockdowns due to unusual patterns in screen use and mental health during that period.

After screening more than 23,000 records, 132 longitudinal studies met inclusion criteria. Of those, 117 studies—comprising over 2,284 individual effects and nearly 293,000 participants—were included in the meta-analysis. Most studies used parent- or self-report questionnaires to assess screen use and socioemotional functioning, including well-known instruments like the Strengths and Difficulties Questionnaire.

A technique known as meta-analytic structural equation modeling allowed the researchers to estimate two core paths: from screen use to later socioemotional problems, and from socioemotional problems to later screen use. This method also accounts for the natural stability of each behavior across time, helping isolate the unique contributions of one variable to the other.

The study found statistically significant associations in both directions. On average, more screen time at an earlier time point slightly predicted greater socioemotional problems later (standardized coefficient b = 0.06), and children with more socioemotional difficulties also showed a small tendency to increase their screen use over time (b = 0.06).

These effects were modest, but consistent. They became stronger over longer time intervals—suggesting that the impact of screen use or distress may accumulate over time. For example, when the lag between time points exceeded four years, the effect of screen use on later socioemotional problems more than doubled.

Importantly, not all screen use was equal. Playing video games showed the strongest association with future socioemotional difficulties (b = 0.32). Gaming also appeared to be more strongly predicted by earlier emotional or behavioral issues (b = 0.44) than other types of screen activity, such as passive video watching. The researchers suggest that children might turn to gaming as a way to cope with distress—but that it may also reinforce or exacerbate emotional struggles over time.

“I was genuinely surprised that video games stood out as particularly problematic, especially since I’d just bought my kids a Nintendo Switch (which is now in the cupboard for special occasions),” Noetel told PsyPost. “Games are just incredibly well-designed to keep kids hooked, and maybe little kids aren’t quite ready for that level of engagement.”

Children who exceeded recommended screen time limits also showed stronger effects in both directions. In contrast, no significant associations were found among children who used screens within guideline thresholds. This supports the idea that excessive screen use may displace healthier activities, such as sleep, physical activity, and face-to-face interaction, all of which are important for emotional development.

The nature of socioemotional problems mattered, too. Internalizing issues like anxiety and peer problems were slightly more sensitive to screen use over time. Meanwhile, externalizing problems like inattention or aggression were more strongly predictive of later screen use, especially among boys.

“Screen use and emotional problems are a two-way street: too much screen time can cause anxiety and behaviour problems, but kids who are already struggling also turn to screens to cope,” Noetel said. “Parents should watch for heavy screen use as a possible red flag that their child needs emotional support. Don’t just add screen limits.”

The analysis also uncovered differences across age and gender groups. Overall, older children were more vulnerable to the effects of screen use on socioemotional problems than younger ones. One reason may be that older children have more autonomy over their media choices, and the content they access—including online games and social media—is more likely to include peer comparison and age-inappropriate material.

Girls were more likely to show socioemotional effects from screen use in early and middle childhood, but among older children, boys were more likely to use screens as a coping mechanism and to show screen-related behavioral difficulties. These patterns may reflect gendered differences in coping styles, emotional regulation, or media preferences.

There were also notable racial disparities. Samples with a lower proportion of White children showed stronger effects in both directions. While the causes remain unclear, these disparities may reflect differences in screen access, content exposure, or social supports. The authors note that more studies focused specifically on underrepresented racial and ethnic groups are needed.

While the findings provide some of the clearest longitudinal evidence to date, the researchers caution that the observed relationships were small and cannot rule out confounding variables. Many of the included studies relied on self-reported screen time, which can be inaccurate. Parental stress, family dynamics, or socioeconomic factors may influence both a child’s screen use and their emotional well-being. Some studies had high attrition rates, which may bias the results.

“We can’t completely rule out other factors like parenting style that might influence both screen use and emotional problems—it’s about as close as we can get to causal evidence without randomly cutting screens for thousands of kids,” Noetel said. “Also, most studies relied on parent or kid reports of screen time, which aren’t always accurate.”

The authors recommend future studies use objective screen tracking, consider more diverse populations, and test possible mediators such as parenting style or school engagement. More research is also needed on the effects of newer screen formats like smartphones and social media, which were less represented in earlier studies.

“We need studies that look at specific types of content and games rather than lumping all ‘screen time’ together—educational games might be very different from violent ones,” Noetel explained. “I’d also love to see more research using actual device tracking rather than parent estimates. My team’s working on both of these goals.”

“The key message for parents is that screens aren’t inherently evil, but they’re not neutral either. Like junk food, a little is fine but too much displaces the good stuff kids need—sleep, play, and real connections. If your child is glued to screens, ask why they’re there, not just how long they’ve been on.”

The study, “Electronic Screen Use and Children’s Socioemotional Problems: A Systematic Review and Meta-Analysis of Longitudinal Studies,” was authored by Roberta Pires Vasconcellos, Taren Sanders, Chris Lonsdale, Philip Parker, James Conigrave, Samantha Tang, Borja del Pozo Cruz, Stuart J. H. Biddle, Rachael Taylor, Christine Innes-Hughes, Katariina Salmela-Aro, Diego Vasconcellos, Katrina Wilhite, Ella Tremaine, Bridget Booker, and Michael Noetel.