Researchers at the University of Jyväskylä, the University of Eastern Finland and other institutions have published results from a study that followed the relationship between screen time and cognitive function over eight years, from early childhood to adolescence. In the study, children with longer cumulative screen time, estimated from questionnaires, tended to respond faster on cognitive tests at around age 15 and to have higher overall scores. The finding runs counter to the widespread concern that long hours in front of screens and a sedentary lifestyle harm children's development. However, the associations differed by type of physical activity, measurement method and sex, and the results do not support the conclusion that screen time improved cognitive ability.
Longer screen time was linked to better cognitive test performance, but the associations were small
The research team, led by Petri Jalanko, published its findings in the peer-reviewed journal Pediatric Exercise Science (Vol. 38, No. 3, pp. 268–283). The paper was published online on May 29, 2026, and the University of Jyväskylä also described the results in an official news release on August 11.
In the analysis, longer cumulative screen time since childhood, estimated from questionnaires, was associated with a higher overall score on the CogState cognitive test administered in adolescence. The standardized regression coefficient was (95% confidence interval $0.070$ to $0.305$). On individual tasks, longer screen time was associated with shorter reaction times on the One Back task (, 95% confidence interval $-0.316$ to $-0.071$) and the Two Back task (, 95% confidence interval $-0.355$ to $-0.112$), both of which assess working memory. Because a smaller reaction-time value means a faster response, the negative regression coefficients indicate an association with quicker responses.
The authors classify associations with standardized regression coefficients of absolute value $0.10$ to $0.29$ as "small," and the associations between screen time and the overall score and reaction times fall within this range. These figures show statistical associations with cognitive test performance; they are not direct measurements of improvements in IQ or academic achievement, or of structural changes in the brain. All three screen-time associations remained statistically significant after false discovery rate (FDR) correction using the Benjamini-Hochberg method. The study set the FDR threshold at $0.2$.
Jalanko, a doctoral researcher at the University of Jyväskylä, said: "Rather than viewing screen time only as harmful, we should look for a balance between physical activity and screen time that encourages active thinking."
260 participants analyzed over eight years, with screen time estimated from questionnaires
The study used data from the Physical Activity and Nutrition in Children (PANIC) study conducted in Kuopio, Finland. PANIC is an intervention study on physical activity and nutrition, but in this analysis participants were not randomly assigned to different amounts of screen time. Instead, the researchers examined the association as a prospective cohort study.
The PANIC study originally invited 736 children aged 6 to 9 from 16 primary schools in Kuopio, and 512 took part in baseline testing. Of these, 506 provided baseline analysis data. The eight-year follow-up included 277 participants, or 54% of the original group. After excluding 17 participants with missing exposure data, such as screen time, or cognitive test data, 260 were included in the final analysis: 136 boys and 124 girls, with a mean age of 15.8 years (standard deviation 0.4 years).
Physical activity and screen time were assessed by questionnaire at three points: at baseline, after two years and after eight years. Parents answered the first two questionnaires, and at eight years the adolescent participants answered for themselves. Screen time was the sum of daily time spent watching television or videos, using a computer, playing video games, using a mobile phone and playing mobile games. The study did not compare activities such as gaming, studying and watching videos.
Using the measurements from the three time points, the team estimated each participant's change over time and used the area under the curve (AUC) to calculate cumulative screen time and physical activity over the eight years. Cognitive function was assessed with the CogState test battery, administered only at the eight-year follow-up. The tests covered psychomotor function, attention, working memory (One Back and Two Back) and paired-associate learning, among others, and an overall cognitive score was calculated by averaging the standardized scores of each measure.
The statistical analysis adjusted for age in adolescence, sex, parental education and the score on the Raven's Coloured Progressive Matrices test given at baseline. Additional analyses also included body fat percentage, pubertal stage and whether participants were in the PANIC intervention or control group. However, because CogState measured cognitive function only once, at eight years, the study could not directly assess how much each participant's cognitive function changed over the eight years.
Associations between physical activity and cognition varied by activity type and sex
For measures other than screen time, results differed depending on the type of physical activity and the measurement method. Longer questionnaire-reported non-screen sedentary time was associated with shorter Two Back reaction times (, 95% confidence interval $-0.299$ to $-0.053$). At the same time, it was also associated with lower accuracy on the One Back task (, 95% confidence interval $-0.274$ to $-0.024$). Even for the same sedentary behavior, the direction of the association differed between speed and accuracy.
More physical activity done without supervision from an instructor was also associated with lower accuracy on the Two Back task (, 95% confidence interval $-0.247$ to $-0.008$). This association was no longer statistically significant after FDR correction. Among boys, more participation in organized sports tended to be associated with shorter One Back reaction times, but this too was not significant after FDR correction.
In an analysis of 171 participants who provided valid data from a chest-worn heart rate and movement sensor, no statistically significant overall associations were found between device-measured cumulative physical activity or sedentary time and cognitive function. In sex-specific analyses, however, more low-intensity physical activity was associated with higher One Back accuracy in girls. The value for the interaction indicating a sex difference was $0.025$, and this association remained after FDR correction.
The association between screen time and One Back reaction time also differed by sex. It was (95% confidence interval $-0.503$ to $-0.167$) in girls, compared with (95% confidence interval $-0.224$ to $0.196$) in boys, with an interaction value of $0.008$. However, most of the sex interactions tested were not statistically significant, so there is no consistent difference between boys and girls in the relationship between cognitive function and physical activity.
The authors point out that while heart rate and movement sensors can measure the intensity of physical activity, they cannot show what cognitive or social stimulation accompanied that activity. In addition, the chest-worn sensor used here may underestimate sedentary time compared with thigh-worn sensors.
The main results are summarized below. The regression coefficients are standardized coefficients adjusted for factors such as age and parental education.
| Activity / measurement method | Cognitive measure | Size of association (, 95% CI) | After FDR correction | Population |
|---|---|---|---|---|
| Cumulative screen time (questionnaire) | Overall cognitive score | $0.187$ ($0.070$ to $0.305$) | Significant | All () |
| Cumulative screen time (questionnaire) | One Back reaction time | $-0.194-0.316$ to $-0.071$) | Significant | All () |
| Cumulative screen time (questionnaire) | Two Back reaction time | $-0.233-0.355$ to $-0.112$) | Significant | All () |
| Cumulative screen time (questionnaire) | One Back reaction time | Girls: $-0.335-0.503$ to $-0.167$); boys: $-0.069-0.224$ to $0.196$) | Interaction | Sex-specific analysis |
| Non-screen sedentary time (questionnaire) | Two Back reaction time | $-0.176-0.299$ to $-0.053$) | Significant | All () |
| Non-screen sedentary time (questionnaire) | One Back accuracy | $-0.149-0.274$ to $-0.024$) | Significant | All () |
| Unsupervised physical activity (questionnaire) | Two Back accuracy | $-0.127-0.247$ to $-0.008$) | Not significant | All () |
| Cumulative physical activity and sedentary time (device) | Cognitive measures | No significant association in overall analysis | Not significant | Valid-data group () |
| Low-intensity physical activity (device) | One Back accuracy | Positive association (interaction ) | Significant | Girls |
| Organized sports participation (questionnaire) | One Back reaction time | Association with shorter reaction time | Not significant | Boys |
These results show that the relationship between physical activity or sedentary behavior and cognitive function cannot be explained by activity time alone. Two points stand out: even the same sedentary behavior showed different associations with reaction speed and accuracy, and the questionnaire and device measurements did not agree.
Whether screen time boosts cognitive ability, and whether it causes any effect, remains unresolved
In the paper, the authors suggest that activities such as reading, writing, homework, video games and board games can provide cognitive stimulation. In the University of Jyväskylä news release, Jalanko likewise said children should be encouraged to use digital devices in ways that involve actively thinking, solving problems and using creativity and learning ability. However, the study assessed screen time as a total amount of time and did not examine which activities explain the association with cognitive function.
Reverse causation is also possible: children with higher cognitive processing ability to begin with may tend to use digital devices for longer. The team included baseline nonverbal reasoning ability in its statistical models, but it cannot fully rule out that differences in existing cognitive ability influenced later behavior. In addition, measuring screen time just three times over eight years cannot adequately capture changes in usage habits in between. Questionnaire responses from parents or the participants themselves may also be affected by recall errors or a tendency to give socially desirable answers.
Previous findings have not been consistent either. A 2014 cross-sectional study of 224 children conducted in Finland in spring 2011 found no significant association between self-reported total screen time and cognitive function. It did, however, report that more time playing computer or video games was associated with poorer visuospatial working memory performance (, 95% confidence interval $-0.333$ to $-0.024$).
The World Health Organization's (WHO) 2020 guidelines on physical activity and sedentary behaviour recommend that children and adolescents aged 5 to 17 do an average of at least 60 minutes a day of moderate- to vigorous-intensity physical activity. This recommendation is based on moderate-certainty evidence. The guidelines also strongly recommend limiting sedentary time, particularly recreational screen time, though the certainty of that evidence is low. The guidelines do not set a uniform cap, such as "up to two hours a day," for ages 5 to 17.
This study reports a favorable association between screen time and some aspects of cognitive function; it does not provide grounds for recommending more screen time. The physical health effects that the WHO guidelines address are also a separate matter from the cognitive tests used here. Jalanko noted that intervention studies are needed to clarify causality and sex differences. Future research will need to distinguish activities such as gaming, watching videos and studying, measure cognitive function at multiple time points, and examine in more detail how the content of digital device use relates to cognitive processing.
