Abstract
INTRODUCTION: While imagery practice is effective for performance enhancement, its impact on mental health is inconclusive due to mixed findings and heterogeneous athlete populations. This study aims to evaluate the effectiveness and optimal dosage of imagery practice on athletes' mental health outcomes. METHODS: A total of 24 randomized controlled trials (RCTs), encompassing 1,294 athletes, were synthesized using a Bayesian multilevel meta-analysis in accordance with the PRISMA 2020 guidelines. Among them, 623 were male and 375 were female; the remaining 296 participants were from studies that did not report sex-specific data. The included trials spanned diverse continents (e.g., America, Asia, Europe), covered a wide age range (from adolescents to adults), and involved both individual and team sports (e.g., gymnastics, soccer, swimming). A Bayesian multilevel approach was selected to account for potential clustering within studies and to provide full posterior distributions for effect estimates, allowing for more robust inferences under uncertainty. RESULTS: Our research indicate that imagery practice may improve athletes' mental health [μ(SMD): 0.5, 95% CI: 0.34 to 0.56; HDI: 0.22 to 0.89; BF: 17.16], including reducing anxiety levels [μ(SMD): 0.52, 95% CI: 0.11 to 0.96; HDI: 0.11 to 0.96; BF: 2.33], strengthening self-confidence [μ(SMD): 0.62, 95% CI: 0.1 to 1.13; HDI: 0.12 to 1.15; BF: 2.2], and improving self-efficacy [μ(SMD): 1.36, 95% CI: 0.26 to 2.47; HDI: 0.24 to 2.45; BF: 5.38]. In the athlete model, statistically significant effects of imagery practice were found only among tennis players [μ(SMD): 1.16, 95% CI: 0.19 to 2.48; HDI: 0.13 to 2.39; BF: 4.06]. In moderation analysis, a dosage of 45-min sessions once a week for 100 days may be associated with more favorable mental health outcomes. DISCUSSION: To our knowledge, limited prior meta-analytic research has employed Bayesian multilevel modeling to examine the effects and moderators of imagery practice on mental health outcomes. While the findings suggest that imagery practice may offer psychological benefits for athletes, these effects appear to vary across contexts and athlete populations. Therefore, imagery should be applied with caution, considering individual differences and potential limitations. This study contributes to sport psychology by offering preliminary empirical guidance for tailoring imagery practice to support athletes' mental health. SYSTEMATIC REVIEW REGISTRATION: https://www.crd.york.ac.uk/PROSPERO/view/CRD420251050005.