Objective. We synthesized evidence on how remote work relates to mental health, distinguishing between Remote Work Use (RWU) and Remote Work Intensity (RWI). We also examined whether contextual and individual factors moderate these links across stress, anxiety, depression, and burnout.
Methods. Following PRISMA standards, we conducted a systematic review and two multilevel meta-analyses of studies available through 2024, drawing from Web of Science, Scopus, PubMed, ProQuest, PsycINFO, and MEDLINE. The dataset comprised 69 effect sizes (32 for RWI; 37 for RWU) from 50 studies totaling 54,311 participants. We applied random-effects models and robust variance estimation (RVE).
Results. Pooled associations between remote work and mental health were small and not statistically significant (RWI: r = 0.0018; RWU: r = 0.104). Moderator analyses indicated that the direction of association, educational level, and assessment period significantly shaped effect magnitudes—most notably in the RWU model. Additional moderators appeared significant under conventional random-effects estimation but did not remain significant with RVE.
Conclusions. Remote work does not emerge as a reliable predictor of mental health outcomes. Its psychological consequences seem to hinge on contextual conditions and person-level differences. Future work should prioritize moderator testing, strengthen control for confounding, and employ longitudinal or experimental designs to clarify causal mechanisms.