Scholarlink

<p>Evidence regarding the long-term efficacy and optimal dosage of school-based Cognitive Behavioral Therapy (CBT) remains equivocal. Prior meta-analyses have frequently overlooked the non-independence of effect sizes inherent in school-based research. This study employed multilevel modeling to rigorously re-assess outcomes for internalizing symptoms and, crucially, psychological resilience. Methods: A systematic search of five databases until November 28, 2025, identified 49 randomized and quasi-randomized controlled trials (N = 12,702; 164 effect sizes). Variance was partitioned using a three-level random-effects model to separate sampling (Level 1), within-study (Level 2), and between-study (Level 3) components. Restricted cubic splines (RCS) were utilized to model non-linear dose–response relationships, and evidence certainty was appraised via the GRADE framework. Results: Three-level analysis demonstrated that school-based CBT significantly reduced depressive (Hedges’ g = 0.29) and anxiety symptoms (g = 0.30), while yielding a substantial improvement in resilience (g = 0.88). Crucially, sensitivity analysis using follow-up score differences confirmed the robustness of these findings, albeit with slightly more conservative estimates (Depression g = 0.27; Anxiety g = 0.26; Resilience g = 0.66). Therapeutic gains for depression and anxiety were sustained beyond the six-month follow-up. Between-study characteristics (Level 3) accounted for 74.59%–94.76% of total heterogeneity, suggesting that implementation context, alongside other study-level properties (e.g., sample characteristics and measurement tools), plays a substantial role in intervention efficacy. Non-linear dose–response modeling for resilience did not yield a statistically significant non-linear trend (P<sub>nonlinear</sub> = 0.055). Conclusions: School-based CBT is an effective modality for mitigating internalizing symptoms and bolstering youth resilience, with targeted protocols yielding the most distinct benefits. Given the high between-study heterogeneity, future research must pivot toward implementation science to identify how ecological factors modulate outcomes. Resource-constrained educational systems should prioritize targeted interventions for high-risk populations to maximize therapeutic efficiency.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Efficacy and Dose–Response of School-Based CBT for Internalizing Symptoms and Resilience: A Three-Level Meta-analysis and Non-linear Modeling

  • 期刊论文

  • 发表日期：

    2026年6月18日

查看全文
  • Mao Zhou,
  • Yizheng Zhu,
  • Zhiyuan Xu,
  • Yuanyue Zhang,
  • Jianxiu Liu,
  • Ruidong Liu

摘要

Evidence regarding the long-term efficacy and optimal dosage of school-based Cognitive Behavioral Therapy (CBT) remains equivocal. Prior meta-analyses have frequently overlooked the non-independence of effect sizes inherent in school-based research. This study employed multilevel modeling to rigorously re-assess outcomes for internalizing symptoms and, crucially, psychological resilience. Methods: A systematic search of five databases until November 28, 2025, identified 49 randomized and quasi-randomized controlled trials (N = 12,702; 164 effect sizes). Variance was partitioned using a three-level random-effects model to separate sampling (Level 1), within-study (Level 2), and between-study (Level 3) components. Restricted cubic splines (RCS) were utilized to model non-linear dose–response relationships, and evidence certainty was appraised via the GRADE framework. Results: Three-level analysis demonstrated that school-based CBT significantly reduced depressive (Hedges’ g = 0.29) and anxiety symptoms (g = 0.30), while yielding a substantial improvement in resilience (g = 0.88). Crucially, sensitivity analysis using follow-up score differences confirmed the robustness of these findings, albeit with slightly more conservative estimates (Depression g = 0.27; Anxiety g = 0.26; Resilience g = 0.66). Therapeutic gains for depression and anxiety were sustained beyond the six-month follow-up. Between-study characteristics (Level 3) accounted for 74.59%–94.76% of total heterogeneity, suggesting that implementation context, alongside other study-level properties (e.g., sample characteristics and measurement tools), plays a substantial role in intervention efficacy. Non-linear dose–response modeling for resilience did not yield a statistically significant non-linear trend (Pnonlinear = 0.055). Conclusions: School-based CBT is an effective modality for mitigating internalizing symptoms and bolstering youth resilience, with targeted protocols yielding the most distinct benefits. Given the high between-study heterogeneity, future research must pivot toward implementation science to identify how ecological factors modulate outcomes. Resource-constrained educational systems should prioritize targeted interventions for high-risk populations to maximize therapeutic efficiency.