The findings of this study do not support the widely held view that adolescent technology use is a major causal factor in their mental health difficulties.
Key Findings
Results
Social media use did not predict later internalizing symptoms in either girls or boys at the within-person level.
A random-intercept cross-lagged panel model (RI-CLPM) was used to separate within-person effects from between-person effects across three annual waves
Sample included N = 25,629 adolescents (51% girls, average age 12 years 7 months, SD = 3.58 months at baseline) in Greater Manchester, England
No evidence was found that time spent on social media predicted later internalizing symptoms among girls or boys
Sensitivity analyses distinguishing active versus passive social media use replicated these null findings
Results
Gaming frequency did not predict later internalizing symptoms in either girls or boys at the within-person level.
No within-person cross-lagged effect of gaming frequency at T1 or T2 on internalizing symptoms at subsequent waves was found for either gender
This null finding held for both girls and boys
Results were based on three annual waves of data (T1, T2, T3)
These findings were replicated in sensitivity analyses
Results
More frequent gaming at T2 predicted less time spent on social media at T3 in girls but not boys.
This gender-differentiated cross-lagged effect was identified at the within-person level using the RI-CLPM
The effect was specific to the T2 to T3 interval
No equivalent effect was observed for boys
This suggests a possible substitution or displacement effect between gaming and social media use among girls
Results
More frequent internalizing symptoms at T2 predicted reductions in gaming frequency at T3 in boys but not girls.
This within-person cross-lagged effect was specific to boys and the T2 to T3 interval
The direction of the effect was that higher internalizing symptoms predicted less subsequent gaming, not more
No equivalent effect was observed for girls
This finding suggests that boys with more internalizing symptoms may withdraw from gaming over time
Results
Longitudinal relationships between digital technology use and internalizing symptoms varied by gender.
Gender moderated the cross-lagged associations between gaming, social media use, and internalizing symptoms
Two gender-specific effects were identified: gaming predicting social media use in girls, and internalizing symptoms predicting gaming in boys
Neither effect was observed in the opposite gender group
Overall, the pattern of findings suggests gender is an important moderator in technology-mental health research
Methods
The study used a random-intercept cross-lagged panel model to isolate within-person effects from stable between-person differences.
RI-CLPM was applied to control for time-invariant between-person confounding, which is a key methodological limitation of standard cross-lagged panel models
Three annual waves of data were used (T1, T2, T3)
The large sample of N = 25,629 adolescents provided substantial statistical power
Sensitivity analyses distinguished active versus passive social media use to test robustness of findings
What This Means
This research suggests that the popular belief that social media and gaming cause mental health problems in teenagers is not well supported by evidence. Using data from over 25,000 young adolescents in Greater Manchester followed over three years, the researchers found no evidence that how much time teenagers spent on social media or gaming predicted the development of depression or anxiety-like symptoms (called 'internalizing symptoms') over time. Importantly, the study used a sophisticated statistical method that controls for pre-existing differences between individuals, giving a cleaner picture of whether changes in technology use within a person over time relate to changes in their mental health.
The study did find some gender-specific patterns: girls who gamed more at one time point tended to use social media less the following year, suggesting gaming and social media may substitute for each other in girls. For boys, having more internalizing symptoms at one time point predicted less gaming the following year, which is the opposite of what a 'gaming causes harm' narrative would predict. Neither of these effects was seen in the other gender.
This research matters because public debate and policy decisions have often assumed that digital technology use is a primary driver of the adolescent mental health crisis. This large, methodologically rigorous study suggests that the relationship is far more complex and that technology use may not be a major direct cause of mental health difficulties in early-to-middle adolescence. This does not mean technology use is entirely without effect, but it does suggest that focusing policy efforts solely on reducing screen time may not be the most effective approach to improving adolescent mental health.
Cheng Q, Panayiotou M, Finserås T, Andersen A, Humphrey N. (2026). How do social media use, gaming frequency, and internalizing symptoms predict each other over time in early-to-middle adolescence?. Journal of public health (Oxford, England). https://doi.org/10.1093/pubmed/fdaf150