第二代正念介入對憂鬱與焦慮的功效：系統性回顧與統合分析
Liucan Xu, et al. Clinical Psychology Review. Volume 123, February 2026, 102693
1. 研究背景
第二代正念介入（second-generation mindfulness-based interventions, SG-MBIs）在傳統正念介入的基礎上，明確整合倫理與慈悲相關的實踐，近年被視為在改善憂鬱與焦慮症狀方面具有潛力的心理介入模式。相較於僅聚焦於注意力與當下覺察的第一代正念介入，SG-MBIs 強調以倫理導向與慈悲動機來引導正念的運用，以促進更全面的情緒調節與心理健康。
本研究旨在透過系統性回顧與統合分析，評估 SG-MBIs 在減少成人憂鬱與焦慮症狀方面的整體效果，並進一步檢視影響介入成效的潛在調節因素。
2. 方法
研究設計
本研究納入隨機對照試驗（randomized controlled trials, RCTs），研究對象為成人樣本，包括臨床族群、健康成人，以及具有心理困擾或身體疾病但未正式診斷之混合族群。
資料來源
系統性搜尋 PubMed、Web of Science 與 EBSCOhost 資料庫，檢索截至 2025 年 4 月發表之相關研究。
效應大小計算
採用 Hedges’ g 作為效應大小指標，比較介入組與對照組在介入後憂鬱與焦慮症狀的差異，並使用隨機效應模型進行統合分析。
3. 主要發現
- SG-MBIs 能顯著降低成人的憂鬱症狀（g = 0.59）與焦慮症狀（g = 0.61）。
- 介入效果在臨床族群中最為顯著，且以自我慈悲為核心的介入方案呈現較佳成效。
- 憂鬱症狀的改善在追蹤期間仍可維持，而焦慮症狀的長期效果相對不穩定。
- 受試者類型與介入類型為顯著調節因子，其餘調節變項（如控制組型態、介入時長與偏誤風險）未達顯著。
4. 討論
本研究結果顯示，SG-MBIs 在減少憂鬱與焦慮症狀方面具備穩健效果，特別適用於臨床族群與以自我慈悲為核心的介入模式。相較於僅強調注意力訓練的正念介入，SG-MBIs 可能透過培養自我關係的溫和態度與倫理導向的心理調節機制，帶來更深層的情緒改善。
然而，現有 SG-MBIs 多以倫理與慈悲成分為主，對智慧（wisdom, prajñā）層面的直接訓練仍相對不足。未來研究有必要發展更清晰的操作型定義、標準化的介入架構，並進一步探索不同正念世代在心理治療中的角色與分工。
5. 限制
- 多數納入研究被評為「有一些擔憂」的偏誤風險，顯示研究品質仍有提升空間。
- 然而，在排除高偏誤風險研究後，主要結果仍維持一致，支持本研究結論的穩定性。
結論
整體而言，第二代正念介入在降低成人憂鬱與焦慮症狀方面具有顯著效果。未來研究應進一步釐清倫理與慈悲成分的關鍵作用機制，並探索智慧導向正念介入的臨床潛力，以促進更全面且具文化深度的正念治療模式。
二、第一代正念（FG-MBIs） vs 第二代正念（SG-MBIs）的差異
| 面向 | 第一代正念（FG-MBIs） | 第二代正念（SG-MBIs） |
|---|---|---|
| 代表介入 | MBSR、MBCT | MSC、CCT、CBCT、MAT、MBPBS、MBPP |
| 理論核心 | 正念作為注意力與覺察技術 | 正念作為具倫理導向的修習 |
| 三學架構定位 | 以「定（samādhi）」為主 | 「戒（sīla）＋定（samādhi）」 |
| 核心成分 | 當下覺察、非評價接受 | 慈悲、自我慈悲、倫理反思 |
| 是否納入空性／智慧 | ❌ | ❌（屬於第三代） |
| 主要治療目標 | 壓力與情緒調節 | 情緒調節＋自我關係修復 |
| 心理作用機制 | 去中心化、減少反芻 | 降低羞愧、自責，提升安全感 |
| 應用重點 | 焦慮、壓力、復發預防 | 臨床憂鬱、焦慮、心理脆弱族群 |
📌 空性、無我、非二元等智慧導向實踐，依現行文獻分類，屬於第三代正念介入（TG-MBIs）
Review of
“The Effectiveness of Second-Generation Mindfulness Interventions on Anxiety and Depression: A Systematic Review and Meta-Analysis”
This systematic review and meta-analysis by Xu et al. offers a timely and methodologically rigorous synthesis of the evidence on second-generation mindfulness-based interventions (SG-MBIs) for anxiety and depression in adults. By focusing on interventions that explicitly integrate ethical and moral dimensions alongside attentional mindfulness practices, the authors address an important and underexplored development in the mindfulness literature.
A major strength of this study lies in its comprehensive scope and adherence to high methodological standards. The inclusion of randomized controlled trials across clinical, non-clinical, and mixed adult populations strengthens the robustness and generalizability of the findings. The meta-analytic results demonstrate moderate and statistically significant reductions in depressive (g = 0.59) and anxiety symptoms (g = 0.61), with particularly strong effects observed in clinical samples. These findings provide compelling evidence that SG-MBIs constitute a clinically meaningful extension of first-generation mindfulness-based interventions.
The moderator analyses further enhance the contribution of this review. By identifying participant type and intervention type as significant moderators, the study highlights that SG-MBIs are not uniformly effective across populations or formats. In particular, the stronger effects associated with self-compassion–focused interventions suggest that targeting self-related processes—such as shame, self-criticism, and emotional safety—may be a key mechanism underlying symptom improvement. Moreover, the finding that reductions in depressive symptoms were sustained at follow-up underscores the potential durability of SG-MBIs, even though long-term effects on anxiety appear less stable.
Despite these strengths, several conceptual and methodological issues warrant closer consideration. First, although the authors define SG-MBIs as interventions integrating ethical and moral practices, the operationalization of these components varies substantially across included studies. Programs such as Mindful Self-Compassion, Compassion Cultivation Training, and Cognitively-Based Compassion Training differ in theoretical emphasis, pedagogical structure, and practice intensity. Treating these interventions as a relatively homogeneous category may obscure meaningful differences in underlying mechanisms and limits interpretability of the pooled estimates.
Second, while the study reports that many potential moderators (e.g., control condition, intervention duration, outcome measure, and risk of bias status) were not statistically significant, these null findings are not fully theorized. For example, the lack of a dose–response relationship may suggest that SG-MBIs operate through qualitative shifts in self–emotion relations rather than through incremental practice time alone. Greater engagement with theoretical models of compassion-based and ethics-oriented regulation could strengthen interpretation of these results.
Third, the authors appropriately note the need to explore wisdom-based practices in future research, but this recommendation would benefit from clearer conceptual boundaries. Within emerging frameworks that distinguish first-, second-, and third-generation mindfulness-based interventions, wisdom-oriented practices (e.g., insight into non-attachment, emptiness, or non-duality) are increasingly viewed as conceptually distinct from ethics- and compassion-based training. Explicitly situating SG-MBIs within this broader generational and theoretical landscape would help clarify what SG-MBIs currently achieve—and what remains beyond their scope.
Finally, although most included trials were rated as having “some concerns” regarding risk of bias, the authors’ sensitivity analyses suggest that the main findings are robust. Nonetheless, the overall quality of evidence underscores the need for more rigorously designed and adequately powered randomized controlled trials, as well as clearer reporting standards and standardized intervention protocols.
In conclusion, this meta-analysis makes a valuable contribution to the mindfulness and mental health literature by demonstrating that SG-MBIs are effective in reducing depression and anxiety, particularly among clinical populations. By foregrounding ethical and compassion-based dimensions of mindfulness practice, the study advances understanding beyond attentional models alone. Future research would benefit from sharper conceptual definitions, theoretically informed moderator analyses, and systematic investigation of how ethics-based and wisdom-based components differentially contribute to psychological change. As such, this review provides a solid empirical foundation while also pointing toward important directions for the next phase of mindfulness intervention research.
Translator and Reviewer: PI-Union Medical Science Ltd.
- E-mail: piunion@pi-union.com
- Official Website: https://pi-union.com/
- Facebook: www.facebook.com/piunion2020
- Youtube: www.youtube.com/@pi-union
- Instagram: www.instagram.com/piunion2020
- LINE: @654eukag