Cuijpers Pim, Harrer Mathias, Miguel Clara, Ciharova Marketa, Papola Davide, Basic Djordje, Botella Cristina, Cristea Ioana, de Ponti Nino, Donker Tara, Driessen Ellen, Franco Pamela, Gómez-Gómez Irene, Hamblen Jessica, Jiménez-Orenga Noelia, Karyotaki Eirini, Keshen Aaron, Linardon Jake, Motrico Emma, Matbouriahi Minoo, Panagiotopoulou Olga M, Pfund Rory A, Plessen Constantin Yves, Riper Heleen, Schnurr Paula P, Sijbrandij Marit, Toffolo Marieke B J, Tong Lingyao, van Ballegooijen Wouter, van der Ven Els, van Straten Annemieke, Wang Yingying, Furukawa Toshi A
Department of Clinical, Neuro and Developmental Psychology, Amsterdam Public Health Research Institute, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands.
WHO Collaborating Center for Research and Dissemination of Psychological Interventions, Amsterdam Public Health Research Institute, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands.
JAMA Psychiatry. 2025 Apr 16. doi: 10.1001/jamapsychiatry.2025.0482.
Cognitive behavior therapy (CBT) is a first-line treatment for most mental disorders. However, no meta-analytic study has yet integrated the results of randomized clinical trials on CBT across different disorders, using uniform methodologies and providing a complete overview of the field.
To examine the effect sizes of CBT for 4 anxiety disorders, 2 eating disorders, major depression, obsessive-compulsive disorder (OCD), posttraumatic stress disorder (PTSD), and psychotic and bipolar disorders on symptoms of the respective disorders using uniform methodologies for data extraction, risk of bias (RoB) assessment, and meta-analytic techniques.
Major bibliographical databases (PubMed, PsycINFO, and Embase for all disorders) were searched up to January 1, 2024, for each disorder separately. Data analysis was performed from August 2024 to January 2025.
Randomized clinical trials comparing CBT with inactive control conditions in adults with 1 of the mental disorders established through a clinical interview were included.
Basic characteristics of patients, CBT, and studies were extracted. RoB was assessed with the Cochrane RoB tool 2. Meta-analyses were conducted using random-effects models.
The primary outcome was the standardized mean difference (Hedges g) indicating the difference between CBT and controls at posttreatment on symptoms of the respective disorders.
A total of 375 trials (423 comparisons) between CBT and controls were included among 32 968 patients. The overall mean (SD) patient age was 43.4 (13.7) years, and the mean (SD) proportion of women was 0.68 (0.24). Effect sizes for CBT compared to all control conditions (g) were lower than 0.5 for bipolar and psychotic disorder; between 0.5 and 1.0 for panic, social anxiety, and generalized anxiety disorders, bulimia nervosa, binge eating disorders, depression, and OCD; and larger than 1.0 for PTSD and specific phobias (range of effect sizes: 0.31 for bipolar disorder to 1.27 for PTSD). Large effect sizes (g > 0.94) were observed in waitlist-controlled trials, a control condition mostly used in anxiety and eating disorders, PTSD, and OCD. Trials using care as usual showed more modest effect sizes (0.22-1.13). Study dropout rates within the CBT conditions ranged from 8% for specific phobia to 24% for PTSD.
In this unified series of meta-analyses, CBT was probably effective in the treatment of mental disorders, including major depression, anxiety disorders, PTSD, OCD, and eating disorders, and possibly effective in psychotic and bipolar disorders. However, the effect sizes depended on the type of control condition.
认知行为疗法（CBT）是大多数精神障碍的一线治疗方法。然而，尚无荟萃分析研究使用统一方法整合不同障碍的CBT随机临床试验结果并全面概述该领域。
使用统一的数据提取方法、偏倚风险（RoB）评估和荟萃分析技术，研究CBT对4种焦虑症、2种饮食失调症、重度抑郁症、强迫症（OCD）、创伤后应激障碍（PTSD）以及精神病性和双相情感障碍在各自障碍症状方面的效应大小。
截至2024年1月1日，分别针对每种障碍搜索主要书目数据库（所有障碍均检索PubMed、PsycINFO和Embase）。数据分析于2024年8月至2025年1月进行。
纳入通过临床访谈确诊患有1种精神障碍的成人中比较CBT与无活性对照条件的随机临床试验。
提取患者、CBT和研究的基本特征。使用Cochrane RoB工具2评估RoB。采用随机效应模型进行荟萃分析。
主要结局是标准化均数差（Hedges g），表明治疗后CBT与对照组在各自障碍症状上的差异。
32968例患者中，共纳入375项CBT与对照组之间的试验（423项比较）。患者总体平均（标准差）年龄为43.4（13.7）岁，女性平均（标准差）比例为0.68（0.24）。与所有对照条件相比，双相情感障碍和精神病性障碍的CBT效应大小（g）低于0.5；惊恐障碍、社交焦虑障碍、广泛性焦虑障碍、神经性贪食症、暴饮暴食症、抑郁症和强迫症的效应大小在0.5至1.0之间；PTSD和特定恐惧症的效应大小大于1.0（效应大小范围：双相情感障碍为0.31至PTSD为1.27）。在等待名单对照试验（一种主要用于焦虑症、饮食失调症、PTSD和OCD的对照条件）中观察到较大的效应大小（g>0.94）。采用常规护理的试验显示效应大小较为适中（0.22 - 1.13）。CBT组内的研究脱落率从特定恐惧症的8%到PTSD的24%不等。
在这一系列统一的荟萃分析中，CBT可能对包括重度抑郁症、焦虑症、PTSD、OCD和饮食失调症在内的精神障碍治疗有效，对精神病性和双相情感障碍可能有效。然而，效应大小取决于对照条件的类型。
