The effects of mindfulness-based interventions on anxiety, depression, stress, and mindfulness in menopausal women: A systematic review and meta-analysis

Menopause, characterized by declining ovarian function and cessation of menstruation, is accompanied by neuropsychological symptoms linked to hormonal fluctuations. Anxiety and depression are prevalent in menopausal women and can be up to three times higher than pre-menopause, adversely affecting quality of life. Psychological interventions are recommended first-line for mood symptoms during menopause and are generally preferred by women due to fewer side effects and better long-term outcomes than pharmacotherapy. Mindfulness-based interventions (MBIs)—including MBSR, MBCT, brief mindfulness meditation, and mindfulness components within DBT and ACT—aim to cultivate nonjudgmental present-moment awareness and enhance emotion regulation. Although RCTs of MBIs in menopausal women have increased, findings on anxiety, depression, stress, and mindfulness remain inconsistent. This review aimed to synthesize RCT evidence on the effectiveness of MBIs for these outcomes in menopausal women to inform clinical practice and community-based programs across cultural contexts.

Prior research indicates MBIs can alleviate negative emotions and stress and have benefits in various clinical and nonclinical populations. In menopausal women, some trials reported significant reductions in anxiety, depression, and stress, and increases in mindfulness, while others found no significant effects. Differences in intervention types (e.g., MBSR, MBCT, yoga-based approaches, ACT), durations, control conditions, and cultural contexts may contribute to mixed results. Broader literature suggests mindfulness training improves emotion regulation, reduces rumination and worry, and may modulate neurobiological pathways (e.g., amygdala and hippocampal structure/function, HPA-axis activity). However, the specific efficacy of MBIs on menopausal mood and stress outcomes had not been consistently quantified prior to this meta-analysis.

Design: Systematic review and meta-analysis following PRISMA and Cochrane Handbook guidance; registered in PROSPERO (CRD42022319349). Searches: PubMed, Cochrane Library, Embase, Web of Science, CNKI, and Wanfang through March 13, 2022; references of relevant reviews were also screened. Eligibility (PICOS): P—women meeting diagnostic criteria for menopause; I—MBIs (e.g., MBSR, MBCT, DBT, ACT, mindfulness yoga/meditation; no restriction on duration); C—non-MBI controls (e.g., wait-list, routine health care, psychoeducation, general conversation, hormone therapy, or other active comparators); O—anxiety, depression, stress, or mindfulness (any validated instrument); S—randomized controlled trials. Exclusions: duplicates, unavailable full text, incomplete data, non-English/Chinese publications. Study selection and data extraction: Two reviewers independently screened titles/abstracts and full texts; disagreements resolved by a third reviewer. Extracted data included study characteristics, sample sizes, interventions and controls, duration and intensity, and outcome instruments. Risk of bias: Assessed using the Cochrane Risk of Bias tool (sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, other bias) by two reviewers with adjudication by a third. Statistical analysis: RevMan 5.4 was used. Standardized mean differences (SMDs) with 95% CIs summarized continuous outcomes across different instruments. Heterogeneity assessed via chi-square and I2. Fixed-effects model applied when p ≥ 0.10 and I2 < 50%; otherwise random-effects models were used. Given variability in samples, interventions, and measures, random-effects models were often chosen. Statistical significance set at p < 0.05. Sensitivity analyses removed one study at a time to probe heterogeneity and robustness. Planned subgroup analyses: continent and intervention duration (weeks). Publication bias: Funnel plots planned if ≥10 studies per outcome.

Included studies: 13 RCTs (published ≤2022), total N = 1,138 menopausal women (560 intervention, 578 control), ages 40–70, across Asia, Europe, and the Americas. Interventions ranged 8–16 weeks, 0.5–5 hours/week, including MBSR, ACT, yoga/meditation-based programs, and general mindfulness training, versus controls such as wait-list, routine care, psychoeducation, or hormone therapy. Outcomes used validated scales (e.g., STAI, GAD-7, SAS, HADS, DASS-21, bPOMS for anxiety; SDS, PHQ-9, HADS, CES-D, DASS-21 for depression; PSS and DASS-21 for stress; MAAS and FFMQ for mindfulness). Risk of bias: Many studies had unclear sequence generation or allocation concealment; blinding was limited due to intervention nature; outcome data completion generally adequate; one study had >20% attrition. Pooled effects with heterogeneity and sensitivity analyses: - Anxiety (7 studies; I2 up to 96%): Initial pooled SMD favored MBIs, but after removing the largest outlier, the effect was not statistically significant (SMD = −0.40, 95% CI −0.81 to 0.01, p = 0.06; I2 = 76%). - Depression (9 studies; I2 up to 95%): After removing the largest outlier, the pooled effect was not statistically significant (SMD = −0.19, 95% CI −0.45 to 0.07, p = 0.16; I2 = 59%). - Stress (4 studies; I2 up to 98%): MBIs significantly reduced stress (random-effects after removing the largest outlier: SMD = −0.84, 95% CI −1.64 to −0.05, p = 0.04; I2 = 92%). Fixed-effects analysis also showed a significant reduction (SMD = −0.60, 95% CI −0.81 to −0.40, p < 0.01), suggesting robustness. - Mindfulness (4 studies; I2 up to 98%): Initial pooled SMD suggested an increase, but after removing the largest outlier, the effect was not statistically significant (SMD = 0.37, 95% CI −0.06 to 0.81, p = 0.09; I2 = 68%). Subgroup analyses: - Continent: For anxiety, MBIs showed significant effects in Asian studies (SMD = −0.55, 95% CI −0.91 to −0.18, p = 0.003) but not in Europe (SMD = 0.39, 95% CI −0.24 to 1.02, p = 0.23). For depression, no significant subgroup differences by continent. - Intervention duration: For mindfulness, 16-week interventions showed significant improvement (SMD = 0.51, 95% CI 0.15 to 0.88, p = 0.006), while 8-week interventions did not (SMD = 0.34, 95% CI −0.41 to 1.09, p = 0.37). Publication bias: Not assessed with funnel plots due to <10 studies per outcome; potential bias cannot be excluded.

This meta-analysis indicates that MBIs can significantly reduce stress in menopausal women, aligning with proposed mechanisms whereby mindfulness training enhances present-moment, nonjudgmental awareness, reduces rumination and worry, and modulates neurobiological pathways implicated in stress reactivity (e.g., HPA-axis, amygdala, hippocampus, autonomic balance). In contrast, pooled effects on anxiety, depression, and mindfulness were not statistically significant after accounting for heterogeneity via sensitivity analyses. High heterogeneity likely reflects differences in sample characteristics, intervention types and intensities, control conditions, cultural contexts, adherence and attrition, and measurement instruments. Subgroup results suggest potential cultural modulation (greater anxiety reduction in Asian contexts) and a possible dose-response effect for mindfulness with longer interventions (16 weeks). Compliance and missing data may attenuate detectable effects. Overall, findings suggest MBIs may be particularly beneficial for stress reduction in this population, while effects on anxiety, depression, and mindfulness require confirmation through larger, well-controlled, and culturally sensitive trials with standardized measures and strategies to enhance adherence.

MBIs significantly reduced stress in menopausal women but did not significantly improve anxiety, depression, or mindfulness scores in pooled analyses after sensitivity checks. Future research should employ high-quality, adequately powered RCTs, standardize interventions and outcome measures, consider cultural adaptations, and explore optimal intervention duration to clarify effects on mood outcomes and mindfulness.

- Many trials had unclear randomization/allocation concealment; only a minority implemented blinding, introducing potential bias. - Limited number of studies per outcome constrained subgroup analyses and increased heterogeneity. - Potential moderators (e.g., participant characteristics, intervention type and intensity, control condition, attrition) were not fully explored via stratified analyses. - Possible publication bias cannot be ruled out due to few studies per outcome. - Nonuniform outcome instruments necessitated SMD aggregation; interpretability is limited. - High heterogeneity across pooled analyses reduces confidence in some estimates.