JMIR Publications

ABSTRACT

Background:

Virtual reality–based mindfulness interventions (VRbMIs) increasingly populate studies as scalable tools for stress and emotion regulation. However, findings across psychological outcomes are heterogeneous, and methodological variation, including reporting practices and AI-assisted evidence synthesis, complicates interpretation.

Objective:

I conducted a systematic review and meta-analysis to evaluate the psychological and physiological effects of VR-based mindfulness intervention, examining methodological quality, outcome consistency, and the role of AI-assisted tools in evidence synthesis.

Methods:

I conducted a systematic search across major databases and AI-assisted retrieval tools in accordance with PRISMA 2020 guidelines. Thirty-five empirical studies met the inclusion criteria. I performed random-effects meta-analyses using Hedges’ g to estimate pooled effects for psychological outcomes where sufficient data were available. I assessed heterogeneity using Cochran’s Q and between-study variance (τ²). Supplementary analytic outputs are available via the project’s Open Science Framework (OSF) repository.

Results:

VR-based mindfulness interventions were associated with consistent short-term reductions in stress, depression, and negative affect, with pooled effects in the small to moderate range. State mindfulness increased reliably following the intervention, while the impact on trait mindfulness and positive affect was negligible and less consistent. For anxiety, five studies provided sufficient data for quantitative synthesis; the pooled random-effects estimate indicated a slight, non-significant reduction in anxiety (g = −0.25, 95% CI −1.14 to 0.64), with substantial heterogeneity across studies (Q(4) = 27.72, p < .001; τ² = 0.44). Physiological outcomes generally aligned with self-reported psychological changes; however, inconsistent measurement and reporting precluded a quantitative synthesis.

Conclusions:

VR-based mindfulness interventions demonstrate reliable benefits for stress-related outcomes and negative affect, with more variable effects for anxiety and positive psychological states. High heterogeneity highlights the influence of intervention design, population, and the operationalization of outcomes. Greater standardization in reporting and transparent documentation of AI-assisted review processes is needed to strengthen the evidentiary foundation of this rapidly evolving field. Clinical Trial: N/A