Systematic review and meta-analysis: the effectiveness of cognitive behavioral therapy and eye movement desensitization and reprocessing in children and adolescents

ABSTRACT

Background: Post-traumatic stress disorder (PTSD) is prevalent among children and adolescents, with cognitive behavioral therapy (CBT) and eye movement desensitization and reprocessing (EMDR) widely used treatment options. This study compared the effects of CBT and EMDR in children and adolescents using a meta-analysis approach.

Methods: PubMed, EMBASE, CENTRAL, CINAHL, ProQuest Dissertations & Theses, and PsycInfo were searched in March 2025. Randomized controlled trials that compared the effectiveness of CBT and EMDR on PTSD symptoms in children and adolescents aged 6–18 years who had been exposed to traumatic events were included. The Cochrane Risk of Bias version 2 was used to assess the methodological quality of the included studies. A random-effects model was applied to calculate the standardized mean difference (SMD) along with 95% confidence interval (CI), and the I² statistic was used to quantify statistical heterogeneity.

Results: Five trials (248 participants) were included. There was no statistically significant difference in PTSD symptoms between the CBT and EMDR groups, whether reported by children (SMD: 0.18; 95% CI: –0.07–0.43; I²= 0%) or by parents (SMD: 0.23; 95% CI: –0.02–0.48; I²= 0%). For anxiety, the pooled estimate did not reach statistical significance (SMD: 0.44; 95% CI: –0.19–1.07; I²= 69%). Depression symptoms were higher in the CBT group than in the EMDR group (SMD: 0.44; 95% CI: 0.13–0.76; I²= 0%).

Conclusion: The comparative effects of CBT and EMDR in reducing PTSD and anxiety symptoms remain inconclusive. Although CBT was associated with higher depression symptoms compared with EMDR, the current evidence is based on a small number of trials, and the pooled estimates should be interpreted with caution.

HIGHLIGHTS

• This updated meta-analysis clarified the differential effects of cognitive behavioral therapy and eye movement desensitization and reprocessing on post-traumatic stress disorder and comorbid symptoms in children and adolescents.

• No statistically significant differences between CBT and EMDR were observed for child- or parent-reported PTSD symptoms or for anxiety symptoms.

• EMDR was associated with lower depression symptoms than CBT in children and adolescents.

List of abbreviations

PTSD

Post-traumatic stress disorder

CBT

Cognitive Behavioral Therapy

EMDR

Eye Movement Desensitization and Reprocessing

SMD

Standardized mean differences

1. Background

A substantial body of research indicates that children exposed to traumatic events are at increased risk of experiencing adverse emotional and psychological outcomes (Downey & Crummy, 2022; Tamir et al., 2025; Woolgar et al., 2022). Post-Traumatic Stress Disorder is highly prevalent among children and adolescents, with rates estimated at 25% and rising from 17% to 28% over the past five years (Tamir et al., 2025). Experiencing PTSD at a young age significantly disrupts emotional and cognitive development, often resulting in chronic functional impairment, increased susceptibility to comorbid psychiatric disorders, and a heightened risk of suicide (TT & MB, 2025).

Psychological interventions are recommended as the first-line treatment for PTSD in children and adolescents (Torrico et al., 2026). Several psychological interventions have demonstrated effectiveness in reducing PTSD symptoms in this population. Among these, CBT and EMDR have consistently shown favorable treatment effects and are recommended in clinical practice guidelines (Gillies et al., 2016; Syros et al., 2022). An earlier meta-analysis found no statistically significant difference between CBT and EMDR in reducing PTSD symptoms (Gillies et al., 2016). In contrast, more recent network meta-analyses have suggested that CBT incorporating trauma exposure is among the most effective psychological interventions for reducing PTSD symptoms in children and adolescents (Alkærsig et al., 2026; Gkintoni et al., 2024; Hoppen et al., 2025). However, Alkærsig et al. (2026) emphasized the need for caution when interpreting these findings because the evidence base for CBT was substantially larger than that for EMDR.

Another network meta-analysis comparing CBT and EMDR suggested that EMDR may be associated with greater effectiveness than CBT (Xie et al., 2024). In summary, CBT and EMDR are both effective and recommended treatments for PTSD in children and adolescents. However, their comparative effectiveness remains uncertain, underscoring the need for further high-quality studies that directly compare these two interventions.

Children exposed to traumatic events frequently experience comorbid anxiety and depression in addition to PTSD. A recent study reported that 40.8% of trauma-exposed children developed PTSD symptoms, of whom 72.6% also experienced anxiety symptoms and 49.6%–74.4% experienced depressive symptoms (Lawrence-Sidebottom et al., 2024; Zhang et al., 2022). Furthermore, post-traumatic anxiety was associated with an increased risk of subsequent depression (RR = 1.20, 95% CI: 1.09–1.34) (Lawrence-Sidebottom et al., 2024). The presence of comorbid anxiety and depression may complicate clinical management and adversely affect treatment outcomes in children with PTSD symptoms.

Psychological interventions for pediatric PTSD aim not only to reduce PTSD symptoms but also to improve common comorbid conditions, particularly anxiety and depression. Previous evidence from conventional meta-analysis suggested that CBT significantly reduced anxiety symptoms but showed no significant effect on depression, whereas EMDR did not demonstrate statistically significant effects on either anxiety or depression (Gillies et al., 2016). In contrast, subsequent meta-analyses reported that EMDR was associated with greater reductions in anxiety and depression than CBT in children, adolescents, and adults (Hudays et al., 2022; Khan et al., 2018). These inconsistent findings indicate that the comparative effects of CBT and EMDR on anxiety and depression remain inconclusive, highlighting the need for further evidence from direct comparative studies.

Therefore, this systematic review and meta-analysis aimed to compare the effectiveness of CBT and EMDR for children and adolescents exposed to traumatic events. The primary outcome was PTSD symptom severity, while anxiety and depression were evaluated as secondary outcomes. Since CBT comprises several related treatment protocols used in clinical practice and evaluated in randomized controlled trials, the present review synthesized the available evidence by comparing EMDR with CBT interventions evaluated across eligible studies, rather than restricting the comparison to a single CBT protocol. To account for the clinical heterogeneity of CBT interventions, predefined subgroup analyses were conducted according to CBT therapeutic components. We hypothesized that CBT and EMDR would differ in their effectiveness in reducing PTSD symptoms, anxiety, and depression among children and adolescents exposed to traumatic events.

2. Methods

We conducted this systematic review according to the Cochrane Collaboration guidelines for a systematic review of interventions (Higgins et al., 2024) and following the PRISMA 2020 statement (Page et al., 2021). The protocol of this study was registered in PROSPERO (CRD420251013825).

2.1. Search strategies

We searched PubMed, EMBASE, CENTRAL, CINAHL, ProQuest Dissertations & Theses, and PsycInfo from the inception of each database to March 2025. There were two main keywords for database searching, i.e. Adolescent. AND cognitive behavioral therapy, post-traumatic stress disorder. A complete list of search terms that were used in each database is presented in Appendix 1. Additionally, we scanned the references of trials that met the inclusion criteria for full-text review. We also conducted citation tracking to screen articles that cited the trials included in our study via Google Scholar.

2.2. Selection criteria

Randomized controlled trials (RCTs) were included if they met the following criteria: (1) studied children or adolescents aged 6–18 years who were exposed to a traumatic event; (2) compared the effect of CBT and EMDR; and (3) reported any of the following outcomes: PTSD, anxiety, or depression symptoms. IM and AK independently screened the titles and abstracts of the search results to determine whether the studies assessed the effects of CBT and EMDR on PTSD symptoms. Full-text articles of potentially eligible studies were subsequently assessed independently by IM and AK. Disagreements were resolved through discussions between IM and AK. If disagreement persisted, TD was consulted to reach a consensus.

2.3. Data extraction

IM and AK independently extracted data using a pre-specified data extraction form. Extracted data included study setting; participant characteristics (e.g. age, gender, type of traumatic event, baseline PTSD symptoms); study duration; and intervention details such as the number and duration of sessions. The outcomes of interest were PTSD symptoms (reported by either parents or children), as well as symptoms of anxiety and depression. All extracted data were verified for accuracy by TD.

2.4. Quality assessment

IM and AK independently assessed the quality of the included RCTs using the Cochrane Risk of Bias 2 (RoB 2) tool (Sterne et al., 2019). The risk of bias was categorized as high, some concerns, or low. Disagreements were resolved through discussion between IM and AK, with TD acting as the third reviewer to reach consensus. A web-application visualization tool (RobVis 10) was used to generate a traffic light plot summarizing the risk of bias for each RCT (McGuinness & Higgins, 2021).

2.5. Analysis

We conducted a pairwise meta-analysis to estimate the standardized mean differences (SMD) and corresponding 95% confidence intervals (95% CI). The standardized mean differences were calculated in RevMan 5.4.1, which applies Hedges’ g for small-sample correction.

A random-effects model, using the DerSimonian and Laird method, was employed for all analyses (DerSimonian & Laird, 1986). Heterogeneity was assessed primarily by considering clinical and methodological diversity across included trials, while the Chi-squared (Q) test and the I² value were reported as descriptive information only. Because the number of studies contributing to each comparison was small, both statistics are imprecise and the low I² value was not interpreted as evidence of homogeneity (Deeks et al., n.d.). Time from treatment to outcome measurement, parent involvement in therapy, and components of CBT were used as criteria for subgroup analysis. However, subgroup analysis based on CBT components could only be conducted for one outcome, as intervention characteristics were largely similar across studies for the other outcomes. All analyses were performed using RevMan 5.4.

3. Results

3.1. Search results

Out of 2119 articles identified from electronic database searches, six studies were initially selected based on title and abstract screening, with three of these meeting the inclusion criteria. The full list of studies excluded from the review is available in Appendix 2. Additionally, 767 studies were identified through citation tracking, of which two met the inclusion criteria. Consequently, a total of five studies (de Roos et al., 2011; de Roos et al., 2017; Diehle et al., 2015; Jaberghaderi et al., 2004; Jaberghaderi et al., 2019) were eligible for inclusion in this systematic review (Figure 1), comprising 248 participants.

Figure 1.

PRISMA flow diagram of selected articles.

3.2. Study characteristics and details of interventions

All included studies (de Roos et al., 2011; de Roos et al., 2017; Diehle et al., 2015; Jaberghaderi et al., 2004; Jaberghaderi et al., 2019) were open-label randomized controlled trials (RCTs). Of these, three (de Roos et al., 2011; de Roos et al., 2017; Diehle et al., 2015) were conducted in the Netherlands and two (Jaberghaderi et al., 2004; Jaberghaderi et al., 2019) in Iran. The duration of the studies ranged from 8 weeks to 3 months, with sample sizes varying from 7 to 43 participants. Across the studies, 26.61% of participants were male, and the mean age ranged from 10.00 to 13.41 years (Table 1). The included studies (de Roos et al., 2011; de Roos et al., 2017; Diehle et al., 2015; Jaberghaderi et al., 2004; Jaberghaderi et al., 2019) investigated the effects of CBT compared with EMDR for treating PTSD in children and adolescents. CBT typically involved psychoeducation, trauma exposure, cognitive restructuring, and coping strategies. Sessions ranged from 4 to 12, lasting 45–227 minutes. EMDR followed Shapiro's standard protocol with age-appropriate modifications, including bilateral stimulation and positive closure, delivered over 4–12 sessions (45–140 minutes each). Therapists were licensed or EMDR-trained. Parental involvement varied. All interventions aimed to process trauma and reduce symptoms (Table 2).

Table 1.

Study characteristics.

Study	Setting, country	Number of participants	Age (years) Mean (SD)	Male (N)	Type of traumatic event	Baseline PTSD	Study duration	Outcome	Funding
de Roos, 2011	Primary care, Netherlands	CBT = 26, EMDR = 26	CBT = 10 (4.1), EMDR = 10.2 (4.0)	CBT = 16, EMDR = 13	Disaster	CBT,CROPS = 22.7 (9.6) EMDR, CROPS = 23.3 (9.9)	3 months	PTSD symptoms Anxiety Depression	NR
de Roos, 2017	Primary care, Netherlands	CBT = 42, EMDR = 43	CBT = 13.41 (2.76), EMDR = 12.96 (3.05)	CBT = 17, EMDR = 20	Various types of traumatic events	CBT,CRTI Child = 90.93 (23.64) EMDR, CRTI Child = 83.27(20.74)	12 months	PTSD symptoms Anxiety Depression	NR
Diehle, 2015	Tertiary care, Netherlands	CBT = 23, EMDR = 25	CBT = 12.8 (3.2), EMDR = 13.0 (3.7)	CBT = 9, EMDR = 9	Various types of traumatic events	CBT,CRTI CAPS-CA =  42.3 (15.2) EMDR, CAPS-CA 44.5(19.4)	8 weeks	PTSD symptoms Anxiety Depression	NR
Jaberghadari, 2004	School, Iran	CBT = 7, EMDR = 7	NR	NR	Sexual abuse	CBT,CROPS = 30.0 (6.4) EMDR, CROPS = 34.86 (5.8)	NR	PTSD symptoms	NR
Jaberghadari, 2019	School, Iran	CBT = 25, EMDR = 24	NR	NR	Domestic violence	CBT,CROPS = 22.20 (10.36) EMDR, CROPS = 22.83 (9.60)	NR	PTSD symptoms	NR

CBT, cognitive behavioral therapy; EMDR, eye movement desensitization and reprocessing; CROPS, The Child Report of Post-traumatic Symptoms; CRTI, Children's Responses to Trauma Inventory; CAPS-CA Clinician-Administered PTSD Scale for Children and Adolescents; NR, no report.

Table 2.

Details of intervention.

Study	CBT			EMDR
	Session content	Intensity (Session)	Duration (minutes)	Session content	Intensity (Session)	Duration (minutes)
de Roos, 2011	Psychoeducation Exposure to trauma memory Cognitive restructuring Examining maladaptive coping behaviors Relapse prevention	4	60	Taking history and planning treatment Preparation of target memory Desensitization of memory Guiding the client to adopt a positive belief regarding the event Identification and processing any residual disturbing body sensations	4	60
de Roos, 2017	Psychoeducation Exposure to trauma memory Cognitive restructuring Promoting healthy coping strategies Engaging loved ones for support (social sharing)	Up to 6	45	History taking Treatment planning Preparation Reprocessing Installation of a positive cognition. Check for and process any residual disturbing body sensations Positive closure and evaluation	Up to 6	45
Diehle, 2015	Psychoeducation Relaxation, effective expression, regulation, and cognitive coping Child's trauma narrative Parent management skill Conjoint child–parent session Enhancing safety and development	Maximum 8	60	Psychoeducation Preparation of the target memory Desensitization of memory Identification and processing of body sensations Re-evaluation of the target	Maximum 8	60
Jaberghadari, 2004	Skill development Identify trauma memory Homework (checklists, drawings, activities, and listening to rapes of the exposure narrative)	10–12	45	Skill development Identify trauma memory Homework (Drawing a safe place)	12	30–45
Jaberghadari, 2019	Psychoeducation for both the parent and child Parents ‘commitment to stop physically abusing Skill development Cognitive, behavioral, social, and affect–focused intervention Homework (checklists, drawings, activities, and listening to rapes of the exposure narrative)	6–12	45–60	Psychoeducation for both the parent and child Parents’ commitment to stop physically abusing The details of the EMDR procedure were not reported	Maximum 12	45–60

3.3. Quality of included studies

According to Cochrane risk of bias assessment version 2.0, all studies (de Roos et al., 2011; de Roos et al., 2017; Diehle et al., 2015; Jaberghaderi et al., 2004; Jaberghaderi et al., 2019) were assessed as having some concerns regarding risk of bias (Figure 2). This judgment reflects issues across multiple domains, including potential bias from the randomization process, deviations from the intended intervention, missing outcome data, and selective reporting of the results.

Figure 2.

Risk of bias of included randomized controlled trial.

3.4. Effects on PTSD symptoms

All five studies (de Roos et al., 2011; de Roos et al., 2017; Diehle et al., 2015; Jaberghaderi et al., 2004; Jaberghaderi et al., 2019) reported on PTSD symptoms, with assessments primarily conducted using self-report measures. In all studies, children assessed their own PTSD symptoms, while four studies also included parent-reported assessments. Regarding the timing of assessments, all five studies measured PTSD symptoms immediately after treatment; two studies conducted follow-up assessments at 3 months post-treatment, and only one study included a 12-month follow-up (Figure 3).

Figure 3.

Forest plots of the effect of CBT and EMDR on PTSD symptoms, anxiety, and depression: (A) the effect of CBT and EMDR on PTSD symptoms (Child reported), (B) the effect of CBT and EMDR on PTSD symptoms (Parent reported), (C) the effect of CBT and EMDR on Anxiety, (D) the effect of CBT and EMDR on Depression.

The pooled SMD of PTSD symptoms of CBT compared to EMDR was 0.18 (95%CI: –0.07–0.43), with no observed heterogeneity (I² = 0.0%) (Figure 3A). Similarly, for studies that reported parent-rated PTSD symptoms, the pooled SMD between CBT and EMDR was 0.23 (95% CI: -0.02–0.48), with no observed heterogeneity (I² =   0.0%) (Figure 3B). The pooled analyses did not demonstrate a statistically significant difference between CBT and EMDR in reducing PTSD symptoms based on either child- or parent-reported measure.

3.5. Effects on anxiety

A total of three studies (Khan et al., 2018; Lawrence-Sidebottom et al., 2024; Zhang et al., 2022) reported anxiety symptoms in children and adolescents receiving treatment for PTSD (Figure 3C). The effect estimate showed no statistically significant difference between CBT and EMDR in anxiety symptoms (SMD: 0.44; 95%CI: -0.19, 1.07; I² = 69%).

3.6. Effects on depression

A total of three studies (de Roos et al., 2011; de Roos et al., 2017; Diehle et al., 2015) reported depression symptoms in children and adolescents treated for PTSD (Figure 3D). The pooled effect estimate (SMD 0.44; 95%CI: 0.13–0.76; I² = 0%) showed that CBT is associated with higher depression symptoms, compared with EMDR.

3.7. Subgroup analysis

Based on the subgroup analyses, there were no statistically significant differences among subgroups for any of the reported outcomes, including PTSD symptoms, anxiety, and depression. The effect estimates remained consistent across different time points (immediately after treatment, 3 months, and 12 months), suggesting that the timing of outcome measurement did not influence the effectiveness of either intervention. However, in the subgroup analysis based on the inclusion of parent–child session, EMDR was found to be more effective than CBT in reducing PTSD symptoms when the intervention did not include parent–child sessions.

In addition, subgroup analyses based on different components of CBT showed no significant differences compared with EMDR in reducing PTSD symptoms (Appendix 4). For depression symptoms, no subgroups demonstrated significant differences (Appendix 4).

4. Discussion

This meta-analysis provides an up-to-date and comprehensive comparison of CBT and EMDR in treating PTSD among children and adolescents, a population where evidence remains inconclusive. Based on a systematic review of recent randomized controlled trials, the present meta-analysis did not demonstrate a statistically significant difference between CBT and EMDR in reducing PTSD symptoms, whether assessed by children or their parents. This finding is consistent with previous meta-analyses conducted across broader age groups (Page et al., 2021). Moreover, a previous network meta-analysis reported that EMDR was effective in reducing PTSD symptoms in children exposed to disasters when compared with a waitlist or no treatment. However, the difference between EMDR and CBT was not significant (Xie et al., 2024). This suggests that the type of trauma may not be a mediating factor in the effectiveness of these two interventions. Both previous studies and our findings indicate that evidence remains inconclusive regarding the comparative effects of CBT and EMDR in alleviating PTSD symptoms.

For the secondary outcomes, CBT was associated with higher depression symptoms, compared with EMDR, whereas no statistically significant difference between CBT and EMDR was observed for anxiety symptoms. The observed effects of EMDR on depression may be explained by the Adaptive Information Processing model, which posits that symptoms arise from maladaptively stored traumatic memories (Shapiro 2014). EMDR relies on implicit processing to facilitate the reprocessing of these memories without extensive cognitive engagement. In contrast, CBT involves explicit learning processes, requiring active cognitive engagement such as cognitive reappraisal, and directly targets maladaptive cognitions associated with emotional (Syros et al., 2022). Although both approaches involve activation of distressing material, their differing processing pathways may contribute to variability in outcomes. However, these findings should be interpreted with caution because the evidence is based on only five trials overall and three trials for anxiety and depression.

This study has several limitations. First, the heterogeneity of CBT interventions across the included studies may have influenced the findings. A low I² in a meta-analysis with few studies indicates limited power to detect heterogeneity and does not indicate that the studies are homogeneous. Moreover, only three studies comparing CBT and EMDR reported outcomes for depression and anxiety, limiting the robustness of the findings. Given the small number of studies, the confidence intervals may be imprecise. Therefore, these findings should be considered preliminary and confirmed through additional trials. Second, the diversity of trauma types represented in the included trials may have affected the severity of PTSD symptoms and contributed to variability in treatment outcomes. Moreover, the methodological quality of the included studies raised some concerns regarding the risk of bias.

5. Conclusion

Our study did not identify a statistically significant difference between CBT and EMDR in reducing PTSD and anxiety symptoms in children and adolescents. However, the current evidence suggests that CBT was associated with higher depression symptoms, compared with EMDR. Although these findings should be interpreted cautiously because the analysis included only five trials overall, of which only three reported anxiety and depression outcomes.

In addition, the methodological quality of the included studies warrants careful consideration. Further high-quality randomized controlled trials are needed to confirm these findings and to evaluate the effects of both interventions on anxiety and depression in children and adolescents.

Human ethics and consent to participate declarations

Not applicable.

Consent for publication

Not applicable

Funding Statement

No funding.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Supplemental Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/20008066.2026.2731855.