Efficacy of EMDR Therapy for Children With PTSD: A Review of the Literature

Extent and Impact of Exposure to Traumaticand Adverse Events

The prevalence of being exposed to one or more traumatic events before the age of 18 years varies between 25% and 68% according to European and U.S. reports (Alisic et al., 2014; Cohen, 2010; Copeland, Keeler, Angold, & Costello, 2007). The Adverse Childhood Experiences (ACE) study (Felitti et al., 1998) demonstrated that being confronted with adverse experiences in childhood is both common and a risk factor for a range of medical and social problems later in life. More recently, Nemeroff (2016) identified the increased risk of several somatic illnesses and psychiatric disorders after exposure to childhood maltreatment. He summarized how neurobiological and clinical consequences, such as persistent changes in neuroendocrine systems and specific alterations in brain areas, associated with mood regulation, induce disease vulnerability. Many studies have indicated that childhood sexual abuse is a nonspecific risk factor for developing psychopathology (Maniglio, 2009). Thanks to the work of several researchers like Perry (2006), Teicher et al. (2003), Teicher and Samson (2013), we know that victims of childhood maltreatment and of sexual abuse are at a significant higher risk of exhibiting a wide range of medical, psychological, and sexual problems during adolescence and adulthood, including substance abuse, conduct disorder, and depression (Cohen, 2010).

The potential impact of exposure to adverse events—besides traumatic events—for children and adolescents was demonstrated in a study of Verlinden et al. (2013), where as many as 62% of a general population sample (n = 588) presented partial symptoms of PTSD, after having been exposed to adverse events (e.g., being bullied) not meeting criterion A of the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5). Interestingly, Mol et al. (2005) found in a study of adults in the general population (n = 1,498) that respondents who had experienced adverse life events (like chronic illness, marital discord, unemployment) reported stronger PTSD symptoms than people who reported having experienced traumatic events. These studies indicate that both adults and children may develop PTSD symptoms after having experienced adverse events not meeting criterion A of the DSM-5.

Trauma Treatments in the Guidelines

International practice guidelines recommend some form of trauma-focused cognitive behavioral therapy (TF-CBT) as a first-line treatment for PTSD in children based on empirical research (Cohen, 2010; Foa et al., 2008; Forbes et al., 2007; NCCMH, 2005; Potgieter Marks, Struik, & Sabau, 2017; WHO, 2013). The program of TF-CBT, developed by Cohen, Mannarino, and Deblinger (2006), is the most researched program (Cohen, 2010; Diehle, Opmeer, Boer, Mannarino, & Lindauer, 2015; Foa et al., 2008) for traumatized children. Besides reducing PTSD and other trauma-related symptoms, this program aspires to enhance relevant coping skills in parents and children for supporting better communication about what happened within the family and for ensuring future safety (building resilience).

EMDR therapy is mentioned in the guidelines mostly as “promising.” In the revised guidelines of the International Society for Traumatic Stress Studies (Foa et al., 2008), EMDR is rated for adults with an A status (meaning that empirical evidence is based on randomized and controlled studies) and for children with a B status (evidence based on controlled studies without randomization and without placebo condition).

Based on efficacy studies, we know now to a certain degree about the effectiveness of these recommended treatments. However, we do not know enough as yet which children are more at risk of developing psychopathology after traumatic or adverse events and what factors influence the effectiveness of treatment in a specific child or family. Additional research is needed to further investigate these important areas. This review is focused on efficacy studies.

Search Strategy

Key words in English and combinations, relevant to the research review questions were used to identify published studies: EMDR, research, children, adolescents, PTSD, trauma-related, and efficacy. The Google search engine was used, as was and the website of the EMDR Institute—the Francine Shapiro Library because most EMDR studies tend to be published there.

Inclusion and Exclusion Criteria

The studies cited were written in English; children and adolescents were participants, and effects were measured on the application of the EMDR standard protocol for treatment of PTSD and other trauma-related symptoms after having experienced traumatic or adverse events. Excluded studies were those written in other languages, in which adults were participants or which were non-evaluative in design; studies evaluating EMDR group treatment; studies in which EMDR was integrated with other interventions; and studies in which only a part of the EMDR standard protocol was used in combination with other interventions. Also excluded were studies not dealing with trauma-related symptoms (Muris, Merckelbach, Holdrinet, & Sijsenaar, 1998; Muris et al., 1997).

Criteria for Evaluation of Research Studies

Different criteria have been added since Foa and Meadows (1997) formulated seven methodological criteria for the evaluation of the quality of studies, known as the Golden Standards. See Table 1.

In order to address the divergence of results in previous studies, Maxfield and Hyer (2002) added three extra criteria: (a) no confounded conditions (i.e., concurrent psychotherapy), because confounded treatment conditions diminish construct validity; (b) multimodal measures, because these provide more accurate evaluation presumably than self-report measures only; and (c) length of treatment, since survivors of multiple trauma generally require more extensive treatment and insufficient treatment could interfere with good assessment of efficacy.

Three more criteria were supplemented 5 years later and together with the previous ones were called the Platinum Standards (Hertlein & Ricci, 2004): (d) reported level of therapist(s) training is relevant, not only for EMDR therapy, but also for the control conditions; (e) control or comparison group should be included, because without this one cannot exclude that reduction of symptoms can be attributed to spontaneous remission; and (f) report of effect size is necessary for better measurement and clear publication of the results. In this review the studies will be evaluated according to these criteria.²

Analysis of Studies

Studies are categorized in Table 2 in terms of design at three levels: (1) without control group; (2) with control group, but no randomization; or (3) RCTs, targeting PTSD, PTSD symptoms, or other trauma-related symptoms, including behavior problems.

Level 1: Studies Without Control Group

Level 1 contains two studies that were conducted without a control condition. Two field studies were excluded, because the intervention was either a group variant of the EMDR standard protocol (Fernandez, Gallinari, & Lorenzetti, 2004) or EMDR was combined with other interventions (Fernandez, 2007). Hence, these studies did not meet the inclusion criteria. Outcome measures of both field studies might suggest nevertheless that EMDR therapy can have an impact in disaster situations.

Hensel (2009) examined if young preschool children could benefit from treatment and if their results would be comparable to those of older children. He used an extended case series design and provided each child with the number of sessions needed until symptoms had disappeared or reduced sufficiently, according to the child and/or parent and therapist. He compared the results of 32 children who had developed a variety of trauma symptoms (e.g., separation anxiety, specific phobias, depression, behavioral problems) after a single traumatic event. These events were diverse in nature (e.g., car accident, witnessing domestic violence, sexual assault). The severity of the symptoms was assessed with parent-report measures (see Table 3: 24). There was no significant difference between symptom improvement of preschool (age 6 and younger) and school-aged children, although the latter had a greater benefit. Children who were exposed to a traumatic event longer ago, and who had remained untreated for a longer period, showed stronger symptoms of PTSD, but also benefited more from treatment. Limitations of this study were: no selection procedure for inclusion of participants (all children referred during a period of 5 years were included), the number of participants was small, there was no control group, progress was measured by one measure only, the author was both assessor and therapist, and data were not analyzed by an independent evaluator.

Ribchester, Yule, and Duncan (2010) explored in a case study design the effects of EMDR on symptoms of PTSD, anxiety, and depression in 11 children (8–16 years) diagnosed with PTSD who had been involved in a traffic accident. Additionally, the study investigated whether reductions in PTSD brought about reduction of attentional bias to emotionally laden information, because of the presumption that these measures could be valuable in clarifying the psychological process and outcome of EMDR. Multimodal measures were used including self-report questionnaires, structured diagnostic interviews, and computer tasks, testing attention, and memory functions (Table 3: 1, 10, 13, 16, 29). Significant results were found for reduction of PTSD, anxiety, depression, and attentional bias. However, the study could not answer the question whether the positive shift in attention was mediated by or only correlated with the effects of EMDR. Study limitations included the first author conducting all the treatments, although treatment fidelity was checked by independent evaluators.

Level 2: Studies With Control Group, but No Randomization

These studies did use a control group, but without randomized allocation of patients. Because results of EMDR therapy are compared to either a waiting list condition or to another treatment condition, it remains unclear to what we can attribute detected differences. The included studies assessed effects of EMDR on PTSD and other trauma-related symptoms. Hence, the studies of Muris et al., 1997 Muris et al., 1998) are excluded, because they tested the effects on spider phobia and did not target distressing memories or experiences.

Puffer, Greenwald, and Elrod (2000) studied the effects of a single EMDR therapy session in 20 children, 8–17 years, with PTSD symptoms, using a 1-month waiting list control. Formal diagnoses were not obtained. Criterion for eligibility was that there were disturbing memories of a specific event that was not being circumscribed. The EMDR session presumably targeted the memory of this event, though this is not mentioned in the publication. The PTSD scores reduced significantly from the clinical to the normal range in more than 50% of the children (for measure, see Table 3: 19). A Limitation was that the therapist was also the assessor and had only limited training in EMDR therapy, meaning that treatment fidelity was questionable.

Kemp, Drummond, and McDermott (2010) compared the effects of four EMDR sessions versus a waiting list condition in 27 children (6–12 years), who presented with PTSD symptoms after a motor vehicle accident. The children were screened with the Child Post Traumatic Stress Reaction Index (Table 3: 5) and included when having a score >12 or when meeting two or more criteria for PTSD (DSM-IV), rated by a clinician. The children in the waiting list condition received EMDR therapy after 6 weeks, the usual period for a waiting list in the institution. Outcome measures were taken pretreatment, posttreatment, 3 months posttreatment, and 12 months posttreatment (Table 3: 3, 11, 17, 18, 19, 33). The PTSD symptoms reduced to 25% in the EMDR group, while remaining 100% in the waiting list condition. Reported improvements remained stable at follow-up after 3 months and had improved further at 12 months. There was only one therapist, however, who also did the outcome assessments.

Wadaa, Zaharim, and Alqashan (2010) examined the effects of EMDR in a group of 12 children (7–12), who had been exposed to different kinds of war and violence (military, civil and sectarian war, and war of terrorism). All children suffered from severe PTSD according to the parents (Table 3: 35). The children had moved from Iraq to Malaysia after the war in 2003 with their parents. The parents were students at the university in Malaysia. None of the children had received any therapy previously. The parents of 12 children allowed their children to participate in the study; these children received 12 weekly sessions of EMDR within 3 months. The parents of 25 children did not give permission for participation; these children formed the control group and received no treatment during these 3 months. At pretreatment, there was no significant difference between the groups on mean PTSD scores, but posttreatment the mean scores of the experimental group had decreased significantly. In terms of limitations, the sample was very small and homogeneous (only children of postgraduate parents were included).

Tang, Yang, Yen, and Liu (2015) assessed in a group of 83 Taiwanese adolescents (12–15 years), who had experienced a typhoon 3 months earlier, the effects of four EMDR sessions on reduction of the severity of symptoms of anxiety and depression. All of them had been diagnosed with PTSD, depressive disorder, and/or an average to high suicide risk. The experimental group received four EMDR sessions and the control group received psychoeducation by weekly sessions. The results demonstrated that EMDR significantly alleviated symptoms of depression and general anxiety and reduced symptoms of disaster-related anxiety (Table 3: 2, 19, 22), all of which had developed after experiencing a major natural disaster. Limitations: the sample was small, allocation of treatment was not randomized (parents chose the treatment for their child), and there was no follow-up measurement.

Level 3: RCTs, Targeting PTSD, PTSD Symptoms, and/or Other Trauma-Related Symptoms

This level includes an overview of RCTs targeting PTSD and trauma-related symptoms, including behavioral problems. Studies in this category are closest to meeting all criteria.

The study of Farkas, Cyr, Lebeau, and Lemay (2010) is excluded because effects were measured of a combined intervention and not of EMDR alone.

Scheck, Schaeffer, and Gillette (1998) examined the effects of EMDR compared to active listening in a group of 60 adolescents and young adults (16–25 years). All participants had been confronted during their early childhood with interpersonal violence and all of them presented with severe behavioral problems besides PTSD. In the EMDR group, symptoms of PTSD, anxiety, and depression decreased significantly, measured by several self-report questionnaires. The effects on behavioral problems were not measured. So, this study showed efficacy of EMDR more than active listening for reduction of symptoms of PTSD, anxiety, and depression.

Chemtob, Nakashima, and Carlson (2002) examined the effects of EMDR on 32 children (6–12 years) who experienced intrusive memories of a hurricane that had taken place 3 years earlier. The children still suffered from PTSD a year after Care As Usual (psychoeducation) had taken place. The EMDR group received three sessions while the control group remained on the waiting list; treatment of the control group started 1 month later, after treatment of the experimental group had finished. Significant reduction of PTSD, anxiety, and depression was reported (Table 3: 4, 8, 29). Results remained stable at follow-up after 6 months. In the control group (waiting list) there was no improvement. So, apparently EMDR was effective for these children to recover from disaster-related PTSD.

Soberman, Greenwald, and Rule (2002) examined the effects of three EMDR sessions on top of Care As Usual—milieu treatment—in 29 boys (10–16 years) living in residential (day) care and diagnosed with PTSD, ADHD, and/or behavioral disorders. Study goals were to test whether EMDR would be effective for children with conduct problems, whether trauma treatment would result in reduction of conduct problems (besides PTSD symptoms), and whether this would result from the reduction of trauma-related distress. One key memory was targeted, and the measures were self-report and parent-report questionnaires (Table 3: 7, 19, 24). The measures showed a significant decrease of problem behavior and of reactivity on the targeted memory. Not all measures were convincing (significant) though, and the study had several methodological shortcomings (one single therapist and no blind assessors). The authors expressed doubts about the long-lasting effects of this brief intervention. Their intention was to demonstrate with this study that, by an effective trauma treatment, children with a history of multiple traumatization—reacting with conduct problems, and hence vulnerable to behavioral deterioration after new trauma—may become more open to respond to other interventions, which are appropriate for treatment of youth with conduct problems.

Jaberghaderi, Greenwald, Rubin, Zand, and Dolatabadi (2004) studied 14 Iranian girls (12–13 years) who had been sexually abused. They received either CBT (10–12 sessions) or EMDR (4–8: as many as needed until memories related to the abuse were fully processed). Three measures were used pretreatment and posttreatment (2 weeks after last session): self-report; parent-report, and teacher-report (Table 3: 7, 24, 31). Both groups showed statistically significant results, with big-effect sizes for reduction of PTSD symptoms and medium-effect sizes for reduction of problem behavior. EMDR appeared significantly more efficient because of fewer therapy sessions and less time for homework. Drop-out rate was equal for both conditions. After termination, three patients in the CBT condition were referred for further treatment versus none in the EMDR condition. Limitations were that the groups were too small to be conclusive, there was a single therapist in each condition, formal procedures for independent checking of treatment fidelity was lacking, and there was no follow-up measurement and no non-treatment condition.

Ahmad, Sundelin-Wahlsten, Sofi, Qahar, and von Knorring (2007) conducted a study with 33 children (6–16 years) who were diagnosed with PTSD, resulting from physical and sexual abuse (both single and chronic); 26 of them fulfilled criteria for an additional DSM diagnosis. All children lived currently in stress-inducing circumstances. For measure of symptoms and treatment outcome, see Table 3: item 27. They were randomly allocated to an EMDR condition or a waiting list condition and participants in the EMDR condition received eight sessions, irrespective of their progress. Immediately after treatment, The Post Traumatic Stress Symptom Scale for Children (Ahmad et al., 2000) demonstrated a significant symptom reduction, particularly in the re-experiencing symptoms, but not in hyperarousal symptoms. The persistence of the hyperarousal symptoms might be due to the fact that the children were still living in stressful circumstances. Follow-up was not measured; therefore, the long-term effects of EMDR are unknown. Other limitations of this study were again a small sample size and lack of independent verification of treatment fidelity.

Wanders, Serra, and de Jongh (2008) studied 26 children between 10 and 14 years who were referred to a child psychiatric (day) clinic because of behavioral problems resulting from exposure to varied traumatic experiences. The effects of four EMDR sessions were compared with an equal number of CBT sessions on measures of self-esteem, emotional, and behavioral problems. The children received four experimental sessions, either CBT or EMDR, shortly after admission to the (day) clinic, during a period in which only observation and assessment take place regularly, but not therapeutic interventions. After their four experimental sessions, they all received identical Care As Usual. The hypothesis was tested whether EMDR therapy–targeting meaningful memories of events that had damaged their self-esteem–would have a significant positive effect on their self-esteem and on reduction of behavioral problems related to this negative self-esteem. To test this hypothesis, parents and mentors completed a series of questionnaires and behavioral measures and the children completed self-assessment instruments pretreatment, immediately after treatment, and at 6 months follow-up (Table 3: 3, 15, 23, 25, 26, 32). Results revealed that both EMDR and CBT had significant positive effects on behavioral and self-esteem problems. The changes in target behaviors of children in the EMDR group seemed slightly larger. This supported the hypothesis that EMDR can produce significant and sustained effects on children’s self-esteem and related problems. Despite these encouraging results, the study also showed several limitations: There was no waiting list control group composed of children who received neither CBT nor EMDR, nor Care As Usual after the period of the experimental sessions. Nor was there a control group of children who received only the Care As Usual (and not the experimental sessions). Therefore, it is impossible to determine to what extent the reported outcomes might be attributed to the regular program (Care As Usual), rather than the experimental sessions, or even to passage of time rather than specific interventions. Another limitation was the small size of the sample, making it difficult to detect meaningful differences between the two groups.

de Roos et al. (2011) compared the effects of EMDR and CBT in a group of 52 children (4–18 years), who had experienced a disaster in their residential area (explosion of a fireworks factory) 1–3 years before, and who showed disaster-related symptoms and were not receiving treatment elsewhere. In this field study both conditions yielded significant results on self-report and parent-report measures. Outcome measures were: symptoms of PTSD, anxiety, depression, and problem behavior (Table 3: 3, 7, 16, 22, 35). No significant differences were found between the two groups. The posttreatment gains were maintained at 3 months follow-up. For efficiency though, a statistically significant difference was demonstrated in the EMDR condition: Results were achieved in fewer sessions (average 3.2 sessions EMDR versus 4 sessions CBT). Study limitations included a relatively small sample; no non-treatment condition; only short-term follow-up after 3 months; and treatment fidelity check was based on a checklist, as sessions were not recorded.

Diehle et al. (2015) compared the effects of eight sessions of EMDR and TF-CBT (Cohen et al., 2006), the treatment program being recommended in practice guidelines as first-line treatment (Cohen, 2010). Data were collected from 48 children (8–18 years) with various (single or chronic) traumatic experiences who presented varied forms of psychopathology besides partial or full PTSD. Results were assessed by both self-report and parent-report measures on reduction of PTSD and comorbid anxiety, depression, and behavior problems (Table 3: 1, 13, 14, 28, 34). The results for reduction of PTSD symptoms were strong in both conditions, statistically significant and equally effective. The results on comorbid problems suggest that parents of children in the TF-CBT condition reported more positive treatment effects than parents in the EMDR condition. Did the parents in the TFC-BT condition possibly learn to cope better with the emotions and the behavior of their child? Did they develop more effective behavior management skills than the parents in the EMDR condition, because in this condition they were more involved in the treatment of their child? The number of sessions, measured in both conditions, was equal; so, no differences were reported for efficiency. EMDR is compared here with an established treatment; therefore, these results are of great importance for EMDR. The study illustrates strong efficacy of both treatments. Limitations include: the sample size was small, there was no waiting list condition, and there was no long-term follow-up.

de Roos et al. (2017), compared three conditions in a large group of 103 children (8-18 years), recruited from multisite standard referrals: (a) EMDR, (b) Cognitive Behavioral Writing Therapy (CBWT) developed by Van der Oord, Lucassen, Van Emmerik, and Emmelkamp (2010), and (c) waiting list. All children had been exposed to a single traumatic event and were diagnosed with PTSD or subclinical PTSD. A statistically and clinically significant strong reduction was found in symptoms of PTSD, anxiety, depression, and behavior problems both in the EMDR and the CBWT conditions, measured by child- and parent-report measures and negative trauma-related appraisals reported by the child (Table 3: 1, 9, 12, 20, 21, 28, 35). Gains remained stable at follow-up 3 and 12 months posttreatment. Only in the EMDR group was a significant ongoing gain reported between the 3 and 12 months follow-up. Similar to the previous study of de Roos et al. (2011), the results in the EMDR condition were achieved in significantly less therapist contact time (mean = 4.1 sessions versus 5.4 sessions for the CBWT condition). Contrary to Diehle et al. (2015), De Roos did not find differences in results for the comorbid problems (anxiety, depression, problem behavior) between both conditions. This may be explained by the fact that CBWT does not include modules targeting skill-building for parents and child and therefore is closer to EMDR than TF-CBT (Cohen et al., 2006). This study illustrates that both CBWT and EMDR are highly effective for the reduction of trauma-related symptoms after a single-incident trauma.

Meta-Analyses

Three meta-analyses have been conducted, to my knowledge, since the first child studies on EMDR were published. The first meta-analysis was done by Rodenburg, Benjamin, de Roos, Meijer, and Stams (2009). The goal of this meta-analysis was to give an overview of studies that examined the effects of EMDR, while considering the criterion of incremental efficacy. At that point in time 22 studies with EMDR had been conducted, but only seven studies could be included in the meta-analysis while meeting the inclusion criteria. The conclusion from these seven studies was that effect sizes, based on comparisons between EMDR and non-established trauma treatment or no-treatment control groups, showed an indication for efficacy of EMDR therapy, plus incremental efficacy when effect sizes were based on comparisons between EMDR and established (CBT) trauma treatment.

The second meta-analysis (Brown et al., 2017) that included EMDR is a meta-analysis of psychosocial interventions for children after man-made and natural disasters. The goal was to examine their effectiveness and to explore moderating factors. The reason for conducting this analysis was the increasing numbers of refugees below 18 years of age, worldwide, and (child) victims of natural disasters like earthquakes and hurricanes, which demand effective treatments. Studies were included (36) that reported on treatment outcome regarding PTSD symptoms in children after natural or man-made disasters, published between 2002–2016. Four EMDR studies—with the standard protocol—were included (Chemtob et al., 2002; de Roos et al., 2011; Tang et al., 2015; Wadaa et al., 2010) as well as six studies with the EMDR Integrative Group Treatment Protocol (EMDR-IGTP; Jarero, Artigas, Montero, & Lena, 2008). Also included were classroom interventions, to be delivered by teachers or trained health care professionals so that larger numbers of children can be reached at the same time. The meta-analysis revealed that group interventions yielded lower effect sizes in pre-post comparisons, but in terms of cost-effectiveness group treatments with a smaller number of sessions could be considered, as they seem to yield similar effect sizes as individual treatments with a larger number of sessions. Other investigated treatment methods, to be delivered by well-trained health care specialists, were CBT, and KIDNET (narrative exposure therapy for children). EMDR, CBT, and KIDNET had all yielded high effect sizes in pre-post comparison and moderate effectsizes in controlled studies. The type of trauma (natural versus man-made disaster) did not moderate treatment effect sizes.

The aim of the third meta-analysis by Moreno-Alcázar et al. (2017) was to update the evidence for efficacy of EMDR therapy in the treatment of PTSD symptoms with children. Additionally, the authors wanted to assess whether EMDR is effective for improvement of comorbid depressive and anxious symptoms, because that was not done before. The meta-analysis is based on eight RCTs. See the studies listed in level 3, Table 2; only the study of de Roos et al. (2017) is lacking in this meta-analysis, because this was not published yet. The main result of the meta-analysis is that a significant reduction in trauma-associated symptoms and comorbid anxiety symptoms was established in patients treated with EMDR; a non-statistically-significant trend was observed for trauma-associated depressive symptoms. These findings are in line with what is found for adults. The incremental efficacy, which was observed by Rodenburg et al. (2009) when comparing EMDR to CBT, was not confirmed in this meta-analysis. Nonetheless, both these meta-analyses showed that EMDR is at least as effective as other techniques such as CBT for reducing PTSD symptoms (Cusack et al., 2016; Moreno-Alcázar et al., 2017, p. 7). No differences were detected in the number of drop-out patients between EMDR and control groups. In terms of limitations, only eight studies (RCTs) could be included in this meta-analysis; they used different control conditions, and the number of sessions that the patients received varied considerably.

Outcomes

Consistent in all 15 studies is the report of significant reductions in PTSD symptoms. In most studies, this is accompanied by significant reduction in measured comorbid symptoms. This effect was less clear in one study (Diehle et al., 2015), where parents reported less improvement on some of the subscales measuring symptoms of anxiety, conduct, and hyperactivity; however, these differences were not significant. The meta-analysis of Moreno-Alcázar et al. (2017) showed statistically significant reduction of anxiety symptoms, and not significant reduction of depressive symptoms, which may be due to the small number of studies measuring this (n = 5). So, for assessing the efficacy if EMDR for comorbid problems more studies are needed.

Four studies compared EMDR with various trauma-focused CBT interventions (de Roos et al., 2011, 2017; Jaberghaderi et al., 2004; Wanders et al., 2008) and one study (Diehle et al., 2015) compared EMDR with the program TF-CBT, developed by Cohen et al. (2006). All (five) studies found no significant difference between the two treatment conditions for PTSD symptom reduction. This is consistent with findings in adult studies.

Three studies found greater efficiency for EMDR and less time needed for therapy sessions (de Roos et al., 2011, 2017; Jaberghaderi et al., 2004), which was not established in one study (Diehle et al., 2015).

None of the studies reported a drop-out rate for EMDR being higher than for control conditions.

Potential Research Questions

EMDR therapy has been found to be an effective treatment for a variety of symptoms, wherever a relation exists between symptoms in the present and distressing memories of past events, or indeed, fantasies about anticipated future events (Engelhard et al., 2011; Logie & de Jongh, 2014). The presumption of the Adaptive Information Processing (AIP) model is that these memories or fantasies cause symptoms because they are not stored adaptively. This model evokes the following questions: Which symptoms/disorders in children and adolescents besides PTSD can EMDR therapy effectively address? Can EMDR address all symptoms, that are elicited or maintained by intrusive memories or fantasies evoking arousal by negative emotions? Are the effects (dis)similar to PTSD when the negative emotion is anger or disgust instead of fear?

Within a political and economic societal context that wants good value for money invested, is the question of cost-effectiveness valid? Do the results achieved with EMDR require equal, less, or more therapist hours compared to other effective treatments? And if so, why? Such comparisons are less than clear cut, however, since the other recommended evidence-based treatments for children (e.g., TF-CBT, developed by Cohen et al., 2006) do not necessarily share identical goals with EMDR (Beer, 2014).

Future research could also examine if there are certain child characteristics or symptom presentations that require modifications to the standard EMDR procedure. This would allow the therapist to modify treatment according to the child’s needs. Research could also investigate whether children benefit equally from a variety of attention demanding tasks,⁵ or whether different tasks have different effects.

Apart from future research regarding the standard protocol, the efficacy of other EMDR protocols should also be tested in controlled studies, such as the group protocol (developed by Jarero, Artigas, Mauer, Alcala, & Lopez, 1999; Jarero et al., 2008) and protocols based on the storytelling method (developed by Lovett 1999, 2015).