Association of Screen Time With Physical and Mental Health in University Students: Systematic Review and Meta-Analysis
Journal of Medical Internet Research · · g7260219
Background: Screen use is almost universal, and numerous studies have highlighted both the benefits and the risks of this digital integration. However, evidence on the association between screen time and health outcomes among university students remains inconsistent. Substantial methodological variation has limited the synthesis and interpretation of existing findings. Objective: This study aimed to (1) quantify the associations between screen time and health outcomes as primary analysis; (2) quantify the association between screen time and physical (sleep quality, eye health, weight-related) and mental health outcomes (depression, psychological distress/stress, combined mental health, and well-being); and (3) identify study-level characteristics that explain between-study heterogeneity by using subgroup and meta-regression analyses. Methods: PubMed, PsycINFO, CINAHL, and Web of Science were searched, with final searches conducted on July 11, 2026, and reference lists were additionally screened. Eligible studies included university students aged 17 years or older, assessed screen time as duration, and reported physical or mental health outcomes with sufficient statistical data for effect-size calculation. We followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist. Data were synthesized in RStudio using the restricted maximum likelihood estimator and Hartung-Knapp-Sidik-Jonkman adjustments. Subgroup analyses and meta-regressions was examined. Small-study effects, leave-one-out, and influence diagnostics were also conducted. Results: A total of 82 studies (83 independent samples; 161,340 participants) were included. Longer screen time duration was significantly associated with poorer health outcomes overall (k=83; r =0.175, 95% CI 0.133-0.217; P <.001; 95% prediction interval –0.184 to 0.493), with substantial between-study heterogeneity ( I 2 =98.35%; Q 82 =4744.45; P <.001). Pooled associations were similar for mental health (k=38; r =0.176, 95% CI 0.124-0.227) and physical health outcomes (k=45; r =0.177, 95% CI 0.108-0.243) and broad health domain did not significantly moderate the association ( P =.96). Within mental health, outcome type significantly moderated the association ( P =.03). The association was the strongest for depression (k=21; r =0.230, 95% CI 0.167-0.291) and was weaker for well-being (k=7; r =0.040, 95% CI –0.146 to 0.223). Combined mental health outcomes were also significant (k=8; r =0.170, 95% CI 0.076-0.261), whereas stress was not (k=2; r =0.103, 95% CI –0.222 to 0.407). Within physical health, screen time was significantly associated with poorer sleep quality (k=26; r =0.120, 95% CI 0.083-0.158) and eye health outcomes (k=16; r =0.277, 95% CI 0.089-0.446), but not weight-related outcomes (k=3; r =0.071, 95% CI –0.222 to 0.351); differences across physical outcome types were not significant. Study-level moderators did not meaningfully explain the remaining heterogeneity. Conclusions: Greater screen time was associated with poorer health outcomes among university students, but the magnitude of association varied across specific outcomes and substantial unexplained heterogeneity remained. The predominance of cross-sectional designs, self-reported screen time, and sparse evidence for some outcomes limits causal inference and precision. Future research should prioritize longitudinal designs and standardized measurement approaches. Trial Registration: PROSPERO CRD42024539034; https://www.crd.york.ac.uk/PROSPERO/view/CRD42024539034