Effect of Social Media Constraints on Mental Health: A Systematic Review and Meta-Analysis of Experiments

Background: Observational studies have consistently reported associations between social media use (SMU) and poorer mental health outcomes; however, such designs cannot establish causality. This study synthesised evidence from randomized experiments to estimate the effects of restricting SMU on mental health outcomes. Methods: A systematic search was conducted across MEDLINE, Embase, PsycINFO, and Cochrane CENTRAL to identify experimental trials evaluating interventions that constrained SMU for at least 24 hours and included an unconstrained control condition. Multilevel random-effects meta-analyses were used to synthesise effect estimates. Prespecified meta-regressions explored study-level moderators, and population-level impact fractions were estimated relative to global SMU prevalence. Results: From 7,784 screened records, 37 reports representing 35 distinct studies were included (pooled N = 7,160). Most interventions lasted one to three weeks and targeted college-aged youth. Pooled estimates favoured SMU constraints across outcomes, with magnitude and precision varying by domain. Confidence intervals were entirely above zero, consistent with a beneficial response for depressive symptoms (g = 0.22; 95% CI, 0.12 to 0.32), perceived stress (g = 0.15; 95% CI, 0.01 to 0.29), anxiety symptoms (g = 0.19; 95% CI, 0.05 to 0.34), fear of missing out/nomophobia (g = 0.14; 95% CI, 0.04 to 0.24), and well-being (g = 0.36; 95% CI, 0.10 to 0.63). Heterogeneity was substantial for several outcomes (I2 > 75%). In bivariate meta-regressions, higher baseline SMU was associated with larger effects for anxiety symptoms ({beta} = 0.13; 95% CI, 0.03 to 0.22), and longer interventions were associated with larger effects for depressive symptoms ({beta} = 0.16; 95% CI, 0.02 to 0.30). Inferences revealed that a short-term reduction in SMU globally could plausibly mitigate 17.5% and 15.4% of depressive and anxiety symptom cases, respectively. Conclusions: Experimental design-based evidence supports the causal case for an effect of SMU on mental health, with constraints producing improvements across multiple outcomes and no evidence of harm. Population-level inferences suggest that even individually modest effects may translate into meaningful public health benefits given the high prevalence of SMU exposure. These findings suggest that reducing SMU may represent a low-intensity, low-cost, scalable strategy to support mental health and improve well-being.