Strengthening foster children’s skills: a descriptive study on a trauma-informed group intervention

ABSTRACT

Background: Foster children have an increased risk of psychological distress because they often experienced multiple traumatizing events. Untreated trauma and behavioural problems are important risk factors for breakdown of foster placements. Unfortunately, the application of first-choice trauma treatment is often complicated due to avoidance, dysregulation and motivational problems.

Objective: The present study investigates whether the group intervention Tame your Dragon improves behavioural functioning and emotion regulation skills in children with trauma-related problems, who are unable or unwilling to start trauma treatment. It also examines whether this results in facilitating subsequent evidence-based trauma treatment.

Method: In total, 7 intervention groups with 29 foster children participated in pre-test (T1) and post-test (T2) measurements. Questionnaires on child post-traumatic stress symptoms (PTSS), behavioural problems and emotion regulation strategies were administered. Treatment advices for a subsequent intervention after completing the group were collected by file research.

Results: Foster parents reported a significant increase in child adaptive emotion regulation strategies, but no changes in children’s behaviour, PTSS and maladaptive emotion regulation strategies. Self-report showed no change in PTSS, adaptive – and maladaptive emotion regulation strategies. After group intervention, 45% of the children was able to proceed with an evidence-based trauma treatment.

Conclusion: This study provides initial evidence that a group intervention for foster children who are not willing or not able to enter evidence-based trauma treatment can be effective in strengthening adaptive emotion regulation strategies according to foster parents. Also there are indications that the group intervention enables almost half of the participants to proceed to trauma treatment.

HIGHLIGHTS

• Twenty nine foster children with trauma-related symptoms who were not able to engage in trauma treatment, participated in a group intervention aimed at strengthening skills.

• Foster parents report a significant increase in child’s adaptive emotion regulation strategies.

• The group intervention enabled almost half of the children to enter evidence-based trauma treatment.

1. Introduction

Internationally, a substantial number of children grow up in foster care: 175,282 in the United States and 17,700 in the Netherlands (National Data Archive on Child Abuse and Neglect, 2025; Pleegzorg Nederland, 2024). Many have experienced chronic and multiple forms of maltreatment, particularly neglect, physical abuse, and sexual abuse, which are common reasons for foster placement (Oswald et al., 2010). Maltreated children frequently present with mental health problems, including behavioural and emotional disorders (Burns et al., 2004), posttraumatic stress disorder (PTSD; Dubner & Motta, 1999; Visser et al., 2025), and attachment disorders (Zeanah et al., 2004). These conditions often impair functioning within foster families, at school, and in peer relationships.

Although foster care can provide a safer environment for children, untreated trauma and behavioural problems are major risk factors for placement breakdown (Konijn et al., 2019). The number of adverse childhood experiences (ACEs; Felitti et al., 1998) is strongly associated with mental health problems and increased placement instability (Knipschild et al., 2024; Liming et al., 2021). Children with 6–9 ACEs have a 52% higher risk of placement disruption compared to those with 1–5 ACEs (Liming et al., 2021). Behavioural problems act both as a cause and consequence of breakdowns (Strijker et al., 2008). Early and effective treatment of trauma- and stressor-related symptoms is essential to improve individual functioning, reduce placement disruptions, and prevent the intergenerational transmission of these problems and their societal impact (Kooij et al., 2022).

Some existing interventions applied in foster care aim at improving parenting strategies. A recent meta-analysis (Teculeasa et al., 2023) showed that foster parents receiving trauma-informed interventions reported significantly better outcomes regarding trauma-related problems compared to control conditions and treatment as usual (Lotty et al., 2020; Pace et al., 2024). In the Netherlands the training Caring for Children who Have Experienced Trauma for foster parents showed an increase in knowledge on child trauma, but no changes in children’s behaviour as reported by the foster parents (Gigengack et al., 2017). However, these interventions are not aimed at improving foster children’s mental health problems directly. In addition to interventions targeting foster parents’ caregiving strategies, effective trauma treatments for children are available. Trauma-Focused Cognitive Behavioural Therapy (TF-CBT) and Eye Movement Desensitization and Reprocessing (EMDR) are recommended in international guidelines (ISTSS, 2019; NICE, 2018). However, their application in chronically traumatized foster children can be challenging. Overall, 25–30% of youth discontinue trauma therapy prematurely (Diehle et al., 2015), and not living with a parent is a significant static risk factor for dropout (van der Hoeven et al., 2023a). Avoidance, fear, shame, guilt, and loyalty often hinder disclosure of traumatic memories (Holt et al., 2014; Yasinski et al., 2018). Caregivers and professionals may also fear destabilizing the child when trauma surfaces (Ormhaug & Jensen, 2018). Furthermore, children may worry that disclosure will affect parental contact or reunification (Struik, 2018). Low perceived parental approval for treatment often leads children to mirror parental ambivalence, resulting in discomfort and negativity toward therapy, which increases dropout risk (Ormhaug & Jensen, 2016). To reduce barriers for foster children in accessing and maintaining trauma treatment, a specialized group intervention was developed. This intervention targets children aged 7–12 who are not yet ready or willing to engage in trauma-focused therapy, aiming to enhance motivation and preparedness by building skills. By this, the intervention seeks to alleviate some behavioural and emotional difficulties and equip them with some helpful strategies that facilitate subsequent trauma processing if needed. Initially implemented in 2012, the programme has since been refined into a standardized protocol called Tame your Dragon (TyD) (Schlattmann et al., 2023a).

The TyD group intervention integrates psychoeducation and relaxation, recently identified as common elements in evidence-based trauma therapies (Kooij et al., 2022), to help children understand their reactions and develop coping strategies. It also incorporates techniques from the initial TF-CBT modules, such as affect regulation and cognitive restructuring (Cohen et al., 2021), enabling children to recognize emotions and make sense of their experiences. Additionally, the metaphor of a ‘dragon’ representing trauma, which can be defeated with support from others, adds a playful and engaging dimension (Greenwald, 2013). TyD is delivered in groups of children with similar out-of-home placement experiences, involving both biological and foster parents. This group setting fosters shared experiences, reduces feelings of isolation, and addresses attachment and loyalty issues, thereby mitigating shame and stigma. Finally, the programme provides adaptive information often lacking in these children’s narratives, such as examples of parents who care but are unable to provide adequate support, to reshape perceptions of self, others, and the world, and to reduce guilt (Schlattmann et al., 2023b). Both (foster-) parents, children and professionals learn by experience that breaking the avoidance does not lead to an increase in symptoms.

The current exploratory study examines changes in PTSS, behavioural problems and emotion regulation strategies in traumatized (foster) children aged 7–12 years who participated in TyD. In addition, this study will rate whether these foster children will enter a subsequent evidence-based trauma-focused treatment after completing the group intervention.

2. Materials and methods

2.1. Participants

From October 2022 to December 2024 a total of 32 children (age 7–12 years) were eligible and enrolled in TyD. Inclusion criteria were: trauma- and stressor-related symptoms inventoried by clinical judgment by the professional at intake; unable to verbally disclose their traumatic experiences; able to function in a group setting.

Children currently living in an acutely unsafe environment and with insufficient reading and writing skills were excluded. Efforts were made, when feasible, to ensure that intervention groups were balanced in terms of age composition and the proportion of children with internalizing versus externalizing problems. Children did not receive any other form of treatment at the time of participation, which was confirmed by their foster care worker. Of the 32 eligible children for TyD, only one participant was excluded from the current study because parental consent was not obtained. Two participants were excluded because they missed more than one session. The TyD groups were executed at specialized centres for foster care, 6 at Levvel Amsterdam and 1 at Parlan, Alkmaar, in the Netherlands.

2.2. Procedure

The TyD group is a regular treatment programme at Levvel and Parlan. Participants were referred for the TyD programme by their foster care worker. Written and verbal information was provided regarding the purpose and procedure of the research, including the child, biological parents or other legal guardians and written informed consent was obtained. If not possible because of limited involvement, verbal consent from the legal guardians was asked (by phone), witnessed by a second professional, and documented. Also, verbal (or written) assent was obtained from the child. Ethical approval was granted from the Ethics Review Board of the University of Amsterdam [n. 2022-W22_280 # 22.340).

Demographic data, including age, gender, education, living situation, PTSD DSM-5 classification and former treatment, was extracted from participants’ files. The number of occurrences of changes in living situation and ACEs was quantified, with each instance being systematically tallied. Data from questionnaires were collected at two time points: before the start of TyD (T1 Pre-Test) and immediately after the last session (T2 Post-Test).

After the final meeting an advice was formulated regarding a successive intervention and discussed with the child, his/her foster parents and (when possible) biological parent(s). These advices could entail a form of evidence-based trauma treatment (EMDR or TF-CBT), another type of (preparatory) intervention, another- or no specific form of treatment. This advice was based on various factors: was the child able and willing to talk about experienced events, clinical observations during TyD (including emotion regulation skills), scores on pre- and post- measurements, current physical and emotional safety, perspective of the foster care placement and the availability of the foster family regarding further (trauma) treatment. The advice was agreed upon with all stakeholders and a referral was made, however, we were not able to follow-up on treatment uptake.

2.3. The group intervention

The TyD group has been developed at Levvel, a specialized centre for youth mental health care, Amsterdam, the Netherlands by N. Schlattman, M. Goris, S. Regoli-Bakker and R.J.L. Lindauer in 2022. The group can be used complementary to the training for foster parents ‘Caring for children who have experienced trauma’ (Coppens & Kregten, 2018), and offered in parallel.

Each session is clearly written out in a manual and consists of different recurring elements (Table 1). Besides the first components of TF-CBT (Psychoeducation, Relaxation Skills, Affect Modulation Skills and Cognitive Coping Skills), additional components were added to the different sessions: motivation, reducing shame, guilt and isolation, providing adaptive information, social support and assessment of the behavioural functioning. The intervention consists of 8 weekly sessions of 2 h, with groups of 4–6 foster children.

Table 1.

Tame your Dragon: topics per session.

Session number	Session topic
1	Get to know each other, establishing group rules, introduction of reward system, psychoeducation (Greenwald fairytale), introduction toolbox and relaxation technique.
2	Psychoeducation on trauma, symptoms and triggers, mindfulness technique, crafting toolbox, feelings.
3	Psychoeducation on stress reactions and alarm system, good and bad memories (metaphor: chest of drawers), psychoeducation on tension and relaxation, relaxation technique.
4	Safety, sources of support, relaxation techniques, psychoeducation on attachment.
5	Where is what feeling, iceberg of feelings and thoughts, relaxation technique.
6	Helping and non-helping thoughts, psychoeducation on core beliefs (metaphor: invisible suitcase), relaxation technique, game of tag with compliments.
7	Compliments, cognitive triangle (thoughts, feelings, behaviour), psychoeducation loyalty problems, relaxation technique, preparing for goodbye.
8	Repeating of relaxation techniques, positive qualities, completing questionnaires, play and treat, diploma and goodbye.
9	Individual reflection (how is your dragon?), evaluation of the group and psychoeducation on different kind of therapies.

Each group was conducted by two trained youth care professionals. Training consists of a one day course by the developers of the intervention. Youth care professionals were social workers specialized in foster care or psychologists To maintain treatment fidelity, trainers received ongoing supervision by the developers and feedback during the intervention period to ensure adherence to the treatment protocol.

2.4. Instruments

2.4.1. Kind en Jeugd Trauma Screener (KJTS)

The symptoms of the child’s PTSD according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (APA, 2013) were assessed with the total score of the child version and parental version (age 7–17 years) of the Kind en Jeugd Trauma Screener (KJTS: Kooij et al., 2025; translation of the Child and Adolescent Trauma Screen (CATS; Sachser et al., 2017)). Parents were, for example, asked how often the following things have bothered the child in the last two weeks: upsetting thoughts or images about a stressful event. Or re-enacting a stressful event in play. Children were asked similar questions, for example, how often the following things have bothered you in the last two weeks: not being able to remember part of what happened. Or negative thoughts about yourself or others. Answers were measured with scores on 4-points scale: ‘Never’ ( = 0), ‘Once in a while’ (1), ‘Sometimes’ (2) and ‘Almost always’ (3). The KJTS contains 20 items and offers a total score (range 0–60) and four subscale scores on the symptom cluster intrusion (range 0–15), avoidance (0–6), negative mood and cognition (0–21) and hyperarousal (0–18). A total score of < 15 refers to normal, 15–20 refers to possible trauma-related stress symptoms, ≥ 21 refers to elevated trauma related stress symptoms and ≥ 25 refers to probable PTSD.

A recent study demonstrates the KJTS reflects adequately the dimensionality of PTSD as described in the DSM-5, with a good fit for self-reports, an acceptable fit for caregiver reports, excellent reliability and sufficient validity (Kooij et al., 2025). In the present study, the Cronbach’s alphas of the parental version pretest and posttest were good respectively, (α = .86) and (α = .87). The Cronbach’s alphas of the child version pretest and posttest were both good (α = .84).

2.4.2. Strengths and difficulties questionnaire (SDQ)

To assess the children’s behaviour problems, we used the parental version (age 4–17 years) of the ‘Total difficulties’ score of the Strengths and Difficulties Questionnaire (SDQ; Van Widenfelt et al., 2003). This scale is based on the subscales ‘emotional symptoms’, ‘conduct problems’, ‘hyperactivity/inattention’ and ‘peer relationship problems’. Examples of items are ‘Restless, overactive, cannot stay still for long’, ‘Nervous or clingy in new situations, easily loses confidence’, and ‘Often fights with other children or bullies them’. Answers were measured on a 3-point scale: (0) not true, (1) somewhat true, and (2) certainly true.

The SDQ contains 25 items and offers a total score (0–40) and four subscale scores (0–10). The higher the score on ‘Total difficulties’, the stronger the indication for the presence of psychosocial problems in a young person. The internal consistency, test–retest stability, and parent-youth agreement of the various SDQ scales have been shown to be acceptable, and the concurrent validity of the SDQ was good (Muris et al., 2003).

In the present study, the Cronbach’s alphas of the parental version pretest and posttest on ‘Total difficulties’ were questionable, respectively (α = .67) and (α = .69). Research on the Dutch Translation of the SDQ showed similar internal consistency (mean α for subscales 0.66) (Van Widenfelt et al., 2003).

2.4.3. FEEL-KJ

Emotion regulation strategies are assessed with the FEEL-KJ. The FEEL-KJ consists of a self-report list and a parent questionnaire that asks how children and adolescents aged 8–18 deal with anger, fear and sadness (Grob & Smolenski, 2005; Dutch translation by Braet et al., 2013). We used the subscales ‘Total Adaptive Strategies’ and ‘Total Maladaptive Strategies’. Parents were, for example, asked ‘When my child is angry … he/she doesn’t want to see anybody’ or ‘When my child is scared … he/she argues with other people’ or ‘When my child is sad … he/she accepts what makes him/her sad’.

Children were asked similar questions like ‘When I am angry … I try to change what makes me angry’ or ‘When I am scared … I think it’s my own problem’, or ‘When I am sad … I try to do something fun’. Answers were measured on a 5-point scale: (1) almost never, (2) rarely, (3) occasionally, (4) often, and (5) almost always. It provides scores for 15 emotion regulation strategies divided into three categories: adaptive strategies, maladaptive strategies and external regulation strategies.

The FEEL-KJ contains 90 items, for each emotion 30 questions, and offers a score on ‘Total Adaptive Strategies’ (42–210), ‘Total Maladaptive Strategies’ (30–150) and ‘Total External Regulation strategies’ (18–90). The higher the score, the stronger the indication for the use of these strategies.

Scientific research (Braet et al., 2013; Cracco et al., 2015) has shown that the FEEL-KJ generally has sufficient to very good reliability with an internal consistency and Cronbach’s alpha ranging from 0.79 to 0.95. Construct validity is considered sufficient and criterion validity is considered good (Braet et al., 2013; Cracco et al., 2015).

In the present study, the Cronbach’s alphas of the parental version pretest and posttest on ‘Total Adaptive Strategies’ were excellent, respectively (α = .96) and (α = .92). Cronbach’s alpha’s of the self-report pretest and posttest on ‘Total Adaptive Strategies’ were both excellent (α = .90).

On ‘Total Maladaptive Strategies’ the Cronbach’s alphas of the parental version pretests and posttest were good respectively (α = .86) and (α = .89). Cronbach’s alphas of the self-report pretest and posttest on ‘Total Maladaptive Strategies’ were good (α = .87) and (α = .84).

2.4.4. Adverse childhood experiences

At baseline, children’s ACEs were screened using the 10 ACEs: physical abuse, psychological abuse, sexual abuse, physical neglect, psychological neglect, witness violence between parents, divorce of parents, parent with mental illness, parent with addiction issues, family member in prison (Felitti et al., 1998). Three commonly used items were added, including societal ones: death of a parent, growing up in structural poverty and chronic bullying.

2.5. Statistic approach

Participants were removed from further analysis if they had completed the pre- but not post-test, or vice versa on every instrument. Before data analysis began, all variables were screened for normality by visual inspection of the histograms, Q-Q plots, and box plots. Also Shapiro–Wilk test was conducted to investigate possible violation of skewness and kurtosis. Differences between the pre- and posttest were mostly analysed with a paired sample t-test because the assumption of normality was not violated. Only the pretest SDQ parental report showed a significant result (p = .03) on the Shapiro–Wilk test. In this specific case, a Wilcoxon signed rank test for two paired samples was used. An alpha confidence level of 0.05 was set for all statistical tests. Statistics were performed using IBM SPSS (V28). Since a few outliers were found, a sensitivity analysis with the removal of the outliers was conducted (see Supplementary Material).

3. Results

3.1. Descriptive statistics

Of the 29 foster children who were included at first, 26 (89.7%) foster children completed pre- and post-KJTS and 21 (72.4%) completed pre- and post-FEEL-KJ. Furthermore, 25 foster parents completed pre- and post-KJTS (86.2%) and 24 (82.7%) completed pre- and post-SDQ. Another 24 (86.2%) foster parents completed pre- and post-FEEL-KJ. The included group consisted of 11 (37.9%) boys and 18 (62.1%) girls, aged 7–12 years (M = 9.4; SD = 1.6) (Table 2). The majority, around three quarters (75.9%), attended a regular form of education. An equal number of foster children, both 14 (48.3%) were living in kinship–, and non-kinship foster care and 1 child lived in a family home.

Table 2.

Characteristics of the foster children at baseline (N = 29).

Variables	N	%
Age
7	4	13.8
8	6	20.7
9	6	20.7
10	3	10.3
11	7	24.1
12	3	10.3
Gender
Male	11	37.9
Female	18	62.1
Other	0	0.0
Education
Regular	22	75.9
Special education	7	24.1
Living situation
Kinship foster care	14	48.3
Non-kinship foster care	14	48.3
Family home	1	3.4
Number of changes in living situation
0–2	20	69.0
3–5	8	27.6
6–8	1	3.4
Number of ACEs
0–3	0	0.0
4–7	15	51.7
8–11	14	48.3
PTSD DSM-5 classification
Yes	4	13.8
No	25	86.2
Former trauma treatment
Yes	4	13.8
No	25	86.2

About two thirds (69.0%) of the foster children experienced 0–2 previous changes in living situation. About a quarter (27.6) had 3–5 changes in living situation. There was 1 child in the category 6–8 changes in living situation. Half of the population (51.7%) was exposed to 4–7 ACEs. The other half (48.3%) was exposed to 8–11 ACEs. Only 4 children (13.8%) were known with a PTSD DSM-5 classification or received trauma treatment previously, with no successful result.

3.2. Results of the group intervention

3.2.1. PTSS

A paired samples t-test was performed on the KJTS pre- and post-test parental version and self-report. The results showed that the pre-test of the parental version of the KJTS (M = 19.32, SD = 9.39) (t(24) = 1.79, p = .09) did not differ significantly from the post-test of the parental version of the KJTS (M = 16.88, SD = 9.42). The effect size was small, Cohen’s d = −0.35, indicating a modest reduction in scores. In addition, self-report showed no significant difference between the pre- (M = 19.81, SD = 10.23) (t(25) = 0.21, p = .84) and post-test (M = 19.46, SD = 10.06). The effect size was very small, Cohen’s d = −0.04, indicating a negligible reduction in scores. This means that, both foster parents and foster children did not report significant changes in PTSS after group intervention.

3.2.2. Behavioural problems

A Wilcoxon signed-rank test was performed on the foster parents SDQ pre- and post-test. The analysis showed no significant change in the scores between pre- and post-test, Z = 0.31, p = .75. A minimal increase in the median was observed from 15.9 at pre-test to 16.2 at post-test, r = 0.01, suggesting a negligible effect. This means that foster parents did not report significant differences regarding the behavioural problems of their foster children after group intervention.

3.2.3. Adaptive emotion regulation strategies

A paired samples t-test was performed on the FEEL-KJ pre- and post-test on adaptive strategies for both parental version and self-report. The pre-test of the FEEL-KJ parental version on adaptive strategies (M = 111.13, SD = 32.37) (t(23) = −3.36, p = .003) differed significantly from the post-test of the FEEL-KJ parental version on adaptive strategies (M = 124.58, SD = 20.8). The effect size was medium, Cohen’s d = 0.69, indicating a noticeable improvement. In line with the formulated hypothesis, foster parents reported a significant increase in adaptive emotion regulation strategies in their foster children after following group intervention.

On the contrary, the pre-test of the FEEL-KJ self-report on adaptive strategies (M = 102.57, SD = 28.21) (t(20) = −1.53, p = .14) did not differ significantly from the post-test of the FEEL-KJ self-report on adaptive strategies (M = 112.31, SD = 22.85). The effect size was small, Cohen’s d = 0.28, suggesting a modest improvement. This means foster children did not report a significant increase in adaptive emotion regulation strategies after following group intervention.

3.2.4. Maladaptive emotion regulation strategies

A paired samples t-test was performed on the FEEL-KJ pre- and post-test on maladaptive strategies for both parental version and self-report. The pre-test of the FEEL-KJ parental version on maladaptive strategies (M = 82.58, SD = 15.45) (t(23) = 1.28, p = .21) did not differ significantly from the post-test of the FEEL-KJ parental version on maladaptive strategies (M = 78.54, SD = 15.82). Effect size was small, Cohen’s d = −0.26, suggesting a modest reduction in scores. Contrary to the formulated hypothesis foster parents did not report a significant decrease in maladaptive emotion regulation strategies in their foster children after following group intervention. Also, the pre-test of the FEEL-KJ self-report on maladaptive strategies (M = 72.14, SD = 20.95) (t(20) = −0.11, p = .92) did not differ significantly from the post-test of the FEEL-KJ self-report on maladaptive strategies (M = 72.71, SD = 16.39). Effect size was very small, Cohen’s d = −0.02, suggesting a negligible reduction in scores. This means, contrary to the formulated hypothesis, that foster children did not report a significant decrease in maladaptive emotion regulation strategies after following group intervention.

3.2.5. Advice after completing group intervention

After the group intervention, 13 foster children (44.8%) were referred to subsequent evidence based trauma treatment (EMDR or TF-CBT). A referral to a different form of preparatory intervention (lifebook, trauma sensitive parenting training or attachment based intervention) was made for 10 foster children (34.5%). Another form of intervention (mediation therapy or pharmacotherapy) was indicated for 4 foster children (13.8%). No follow-up intervention was initiated for 2 foster children (6.9%) due to a lack of motivation.

3.2.6. Outlier assessment

A few outliers were detected from examination of the boxplots. However, as none of the z-scores for the effect sizes exceeded the threshold values of −3.29 or +3.29, we retained the outliers in all our analyses and ran a separate sensitivity analysis excluding the outliers (see Supplementary Material) (Table 3).

Table 3.

Clinical ratings before (Pre) and after (Post) group intervention Tame your Dragon.

Questionnaire	N	M	SD	t	df	p
Foster parents
KJTS (pre) – KJTS (post)	25	19.32 16.88	9.39 9.42	1.79	24	.09¹
SDQ (pre) – SDQ (post)	24	15.93 16.20	5.25 5.51			.75²
Adaptive (pre) – Adaptive (post)	24	111.13 124.58	32.37 20.80	−3.36	23	.003*¹
Maladaptive (pre) – Maladaptive (post)	24	82.58 78.54	15.45 15.82	1.28	23	.21¹
Foster children
KJTS (pre) – KJTS (post)	26	19.81 19.46	10.23 10.06	0.21	25	.84¹
Adaptive (pre) – Adaptive (post)	21	102.57 112.31	28.21 22.85	−1,53	20	.14¹
Maladaptive (pre) – Maladaptive (post)	21	72.14 72.71	20.95 16.39	−0.11	20	.92¹

Note: Results of paired samples t-test and non-parametric test.

M = mean; SD = standard deviation; df = degrees of freedom * p < .05 ¹ = paired samples t-test ² = Wilcoxon signed-rank test.

4. Discussion

This study is the first prospective study to evaluate the group intervention Tame your Dragon (TyD, Schlattmann et al., 2023a, 2023b). This intervention designed for foster children (age 7–12) who experienced chronic trauma and suffer from trauma-related symptoms aims to alleviate some behavioural and emotional difficulties and equip them with some helpful strategies that facilitate subsequent trauma processing if needed. Overall, TyD improved adaptive emotion regulation per caregiver report but did not affect PTSS or behavioural outcomes. After the group intervention, the proportion of foster children referred to evidence based trauma treatment increased by 44.8%.

As expected, this group intervention aiming at preparing children who need it for further trauma treatment, did not extinct PTSS, because it does not directly address the processing of the trauma. Research on TF-CBT shows that children assigned to a trauma narrative group compared to children assigned to a non-trauma narrative group reported significantly less anxiety (Deblinger et al., 2011). This finding is in line with a systematic review on Skills Training and Interpersonal Regulation (STAIR), which also focuses on strengthening emotion regulation but with greater emphasis on training interpersonal skills compared to TyD. This review indicates that reductions in PTSS occurred primarily in the phase of trauma-focused therapy, but not in the preparatory phase (Lorbeer et al., 2023). Similarly, Ehring et al. (2014) reported in their meta-analysis that trauma-focused treatments were superior to non-trauma-focused treatments for PTSS. Nevertheless, the preparation phases of STAIR prove to be effective in reducing affective dysregulation and interpersonal problems (Lorbeer et al., 2023).

In addition, various factors might have influenced the reporting of PTSS and therefore the lack of change in our study. KJTS pre-test scores for both parental- and self-report were in the range of moderate trauma-related distress and not in the expected (sub)clinical range. A systematic review of brief psychoeducational interventions after trauma demonstrated that participants experienced an immediate increase in knowledge following the intervention (Brooks et al., 2021). Possibly, this increase in knowledge contributes to a higher rate of PTSS reporting in the post-test, which makes it more difficult to demonstrate (significant) change. Furthermore, TyD does not include trauma-processing aimed at reducing PTSS, as exposing children in a group to traumatic memories of other children is not desirable (Dorsey et al., 2016).

In children, it is known that PTSD often co-occurs with behavioural problems (Lindauer & De Boer, 2020; Struik, 2021; van der Hoeven et al., 2023b). More specific, PTSD may induce behavioural problems (DeNigris, 2008). Building upon this association, it is comprehensible why no reduction in behavioural problems was observed in our study. Furthermore, reduction might have been undetectable because moderate pre-test levels may have limited potential for change.

As hypothesized, foster parents did report a significant increase in children’s adaptive emotion regulation strategies. TyD teaches children and their parents a wide range of emotions, arousal reduction strategies (e.g. deep breathing, progressive muscle relaxation), and cognitive strategies for targeting unhelpful thoughts (Cohen et al., 2021). This noticeable increase is clinically relevant, as adaptive strategies such as acceptance, problem-solving, and cognitive reevaluation serve as protective factors against the development of psychopathology (Aldao et al., 2010). On the contrary, self-report did not show a direct change in adaptive emotion regulation strategies. It is to be noted that the clinical use of the FEEL-KJ self-report is challenging in this group of children. A limitation of the FEEL-KJ is that it has not been evaluated in clinical groups (Cracco et al., 2015). The questionnaire is quite lengthy and repetitive (Junghänel et al., 2022), while participants frequently struggle with attention- and concentration problems as a result of chronic trauma (Lindauer & De Boer, 2020). It is also possible that self-report is limited by children’s awareness and ability to differentiate between different emotions, monitor their response and communicate this information (Thornback & Muller, 2015). Further evaluation of the feasibility of FEEL-KJ in the target population seems needed.

Contrary to our hypothesis, children’s use of maladaptive strategies as reported by both foster children and foster parents did not change. Aldao et al. (2010) differentiate between three maladaptive emotion regulation strategies: avoidance, suppression and rumination. There is an overlap between these maladaptive emotion regulation strategies and PTSS, concerning avoidance. Possibly, children with PTSS might therefore also score positive on the maladaptive emotion regulation strategy avoidance. Individuals who are highly reliant on avoidant coping strategies and highly reactive to trauma reminders may be at greatest risk of maintaining their PTSS within the first few months following the trauma (Pineles et al., 2011). As discussed before, in TyD no form of exposure to the traumatic material takes place. Thereby, possibly, avoidance stays in place as a maladaptive strategy. In comparison, when exposure was included over the course of a TF-CBT treatment, children did reduce their use of maladaptive emotion regulation strategies (Thornback & Muller, 2015).

The strengths of the current study include the focus on an understudied, difficult-to-reach population of chronically traumatized foster children and the use of a multi-informant approach that includes both foster parents and foster children as respondents. A few limitations need to be mentioned. Firstly, this study had a small sample size caused by a lack of referral, drop-out and missing data, which reduces the power to detect medium effects. Secondly, there were no follow-up measurements because almost all respondents were enrolled in another type of intervention by then which limits the ability to assess long-term outcomes. Thirdly, the present study design did not include a control group, which limits causal inference. Fourthly, we only analysed cumulative effect of ACEs without discriminating which ACEs had higher or lower impact. Finally, in clinical practice, in parallel to the TyD group, a training for foster parents: ‘caring for children who have experienced trauma’ is regularly offered, which might have affected the results.

Future research should aim at increasing the sample size and the use of a control group. Furthermore, concepts such as self-image, feelings of isolation, and loyalty or attachment issues could be valuable variables for future research due to their significance for this specific group of children. Because of the limitations of self- en parent report, the development of observational methods for emotion regulation strategies may complement a future research design (Thornback & Muller, 2015). Although the awareness and regulation of emotions are individual child factors, they are also outcomes of social and relational processes (Morris et al., 2007). The inclusion of external emotion regulation strategies in future research could build upon our findings. More practical, greater emphasize within the group intervention on teaching various emotion regulation strategies to enable flexible application, depending on the type of situation, might further strengthen this group of children (Berking & Whitley, 2014).

Concluding, a valuable implication for clinical practice is that the group intervention in a considerable amount of cases motivates and prepares the foster children who need it for subsequent evidence-based trauma treatment. In our sample, almost half of the participating children were subsequently referred to EMDR or TF-CBT. This finding is significant given the well-established association between untreated PTSS and the development of subsequent psychopathology (Felitti et al., 1998; Petruccelli et al., 2019), as well as the recognized risk that untreated trauma poses for the breakdown of foster family placements (Konijn et al., 2019). Recommendations for clinical practice thereby include upscaling and increasing the accessibility of this group intervention in the field of foster care.

Author contributions

S.B., N.S., M.G., and I.H. conceptualized and designed the study. S.B. performed the literature search and created the data file. S.B. conducted the analyses, and I.H. supervised the analyses. S.B. wrote the first draft of the manuscript. All authors reviewed and revised the manuscript.

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, S.B., upon reasonable request.

Supplemental Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/20008066.2025.2612414.