Impact of increased educational screen exposure on adolescent overweight and the mediating role of physical activity and sleep quality: a natural experiment study from China.
Educational screen exposure may contribute to increased overweight in adolescents by reducing physical exercise and impairing sleep quality, with the increase in overweight prevalence 3.1% greater in the exposure group than the control group.
Key Findings
Results
Increased educational screen exposure was associated with a 3.1% greater increase in adolescent overweight prevalence compared to the control group.
Difference-in-differences coefficient: 0.03 (95% CI: 0.01 to 0.05, p < 0.01)
Data drawn from 41,157 adolescents across 16 cities in China from the Database of Youth Health
Two survey waves used: 2015 and 2017, coinciding with China's online education policy implementation since 2015
Adolescents were allocated to exposure and control groups based on local Internet penetration rate as a proxy for policy intensity
Results
Educational screen exposure was associated with significant reductions in physical activity frequency among adolescents in the exposure group.
Difference-in-differences coefficient for physical activity frequency: -0.49 (95% CI: -0.52 to -0.45)
Physical activity and sleep quality data were collected via self-administered questionnaire
The natural experiment design leveraged variation in local Internet infrastructure to establish the exposure condition
Results
Educational screen exposure was associated with significant reductions in sleep quality among adolescents in the exposure group.
Difference-in-differences coefficient for sleep quality: -0.06 (95% CI: -0.09 to -0.02)
Sleep quality was assessed alongside physical activity using a self-administered questionnaire
Both physical activity and sleep quality reductions were observed in the same exposure group that showed greater overweight increases
Results
Reduced physical activity and impaired sleep quality were identified as plausible mediating variables linking educational screen exposure to increased overweight prevalence.
Mediation analyses were conducted to examine pathways between educational screen exposure and overweight
Both reduced physical activity frequency and reduced sleep quality were identified as mediators in the relationship
The study design used a difference-in-differences approach within a natural experiment framework to support causal inference
Results
Male adolescents experienced a greater increase in overweight associated with educational screen exposure compared to female peers.
Difference-in-differences coefficient for male adolescents: 0.026 (95% CI: 0.01 to 0.04)
Sex-stratified analyses were conducted to identify subgroup differences in the association between educational screen exposure and overweight
The sex difference suggests that the impact of educational screen exposure on overweight may not be uniform across adolescent subgroups
Methods
China's 2015 online education policy provided a natural experiment context in which local Internet penetration rate determined intensity of educational screen exposure.
The policy implementation resulted in 'a substantial increase in adolescents' educational screen exposure'
The intensity of policy implementation was influenced by local Internet infrastructure, creating quasi-random variation in exposure
This variation was used to allocate adolescents into exposure and control groups for difference-in-differences analysis
The study sample included 41,157 adolescents from 16 cities, with data from two waves (2015 and 2017)
What This Means
This research suggests that increased use of digital screens for educational purposes is linked to higher rates of overweight among teenagers. The study took advantage of a natural experiment in China: when the government introduced online education policies in 2015, some cities had better internet infrastructure and therefore experienced a larger increase in educational screen time. By comparing teenagers in high-internet-access cities (more screen exposure) to those in low-access cities (less exposure) before and after the policy, researchers found that the overweight rate rose 3.1 percentage points more in the high-exposure group over two years. The study analyzed data from over 41,000 adolescents across 16 Chinese cities.
The research also found that teenagers with greater educational screen exposure exercised less frequently and had poorer sleep quality. Statistical analyses suggested that these behavioral changes — less physical activity and worse sleep — help explain why more screen time for education was linked to more overweight. In other words, screens don't just replace exercise time; they may also disrupt sleep in ways that contribute to weight gain. Male adolescents appeared more affected than female adolescents in terms of weight gain associated with screen exposure.
This research suggests that simply increasing digital learning tools without addressing their side effects could unintentionally harm adolescent health. Policymakers, schools, and parents may want to consider balancing educational screen use with structured physical activity and strategies to protect sleep, particularly as online and hybrid learning continues to expand globally.
Li J, Wu T, Song F, Li S, Wang W, Zhang Q, et al.. (2026). Impact of increased educational screen exposure on adolescent overweight and the mediating role of physical activity and sleep quality: a natural experiment study from China.. BMC public health. https://doi.org/10.1186/s12889-026-26600-4