Bayer CP, Klasen F, Adam H. Association of Trauma and PTSD Symptoms With Openness to Reconciliation and Feelings of Revenge Among Former Ugandan and Congolese Child Soldiers. JAMA. 2007;298(5):555-559. doi:10.1001/jama.298.5.555
Author Affiliations: Outpatient Clinic for Refugee Children and Their Families, Department of Child and Adolescent Psychiatry, University Clinic Hamburg, Germany (Mr Bayer, Ms Klasen, and Dr Adam); Children for Tomorrow Foundation, Hamburg, Germany (Ms Klasen and Dr Adam); and Hospital for Psychiatry, Neurology and Child and Adolescent Psychiatry, Eberswalde, Germany (Dr Adam).
Context Tens of thousands of the estimated 250 000 child soldiers worldwide are abused or have been abused during the last decade in Africa's Great Lakes Region. In the process of rebuilding the war-torn societies, it is important to understand how psychological trauma may shape the former child soldiers' ability to reconcile.
Objective To investigate the association of posttraumatic stress disorder (PTSD) symptoms and openness to reconciliation and feelings of revenge in former Ugandan and Congolese child soldiers.
Design, Setting, and Participants Cross-sectional field study of 169 former child soldiers (aged 11-18 years) in rehabilitation centers in Uganda and the Democratic Republic of the Congo, conducted in 2005.
Main Outcome Measures Potentially traumatic war-related experiences assessed via a sample-specific events scale; PTSD symptoms assessed using the Child Posttraumatic Stress Disorder Reaction Index (CPTSD-RI), with a score of 35 or higher indicating clinically important PTSD symptoms; and openness to reconciliation and feelings of revenge assessed via structured questionnaires.
Results Children participating in this study were a mean of 15.3 years old. These former child soldiers reported that they had been (violently) recruited by armed forces at a young age (mean [SD], 12.1  years), had served a mean of 38 months (SD, 24 months), and had been demobilized a mean of 2.3 months before data collection (SD, 2.4 months). The children were exposed to a high level of potentially traumatic events (mean [SD], 11.1 [2.99]). The most commonly reported traumatic experiences were having witnessed shooting (92.9%), having witnessed someone wounded (89.9%), and having been seriously beaten (84%). A total of 54.4% reported having killed someone, and 27.8% reported that they were forced to engage in sexual contact. Of the 169 interviewed, 59 (34.9%; 95% confidence interval, 34.4%-35.4%) had a PTSD symptom score higher than 35. Children who showed more PTSD symptoms had significantly less openness to reconciliation (ρ= −0.34, P < .001) and more feelings of revenge (ρ= 0.29, P < .001).
Conclusions PTSD symptoms are associated with less openness to reconciliation and more feelings of revenge among former Ugandan and Congolese child soldiers. The effect of psychological trauma should be considered when these children are rehabilitated and reintegrated into civilian society.
Approximately 250 000 children are being abused as child soldiers in the world today.1 Many of them, often influenced by ethnic tensions, are involved in conflicts in Africa's Great Lakes Region. While enduring their time as soldiers, they undergo and are forced to commit atrocities. Thus, child soldiers carry a special burden of simultaneously being the recipient and perpetrator of violence.2 They are blamed and stigmatized for the atrocities they have committed, and their psychological recovery and reintegration into civilian society is a difficult task.
Currently, Africa's Great Lakes Region is struggling to recover from war. Undoubtedly, a key issue to promote the process of reconciliation and reconstruction of society is the successful rehabilitation and social reintegration of the numerous former child soldiers. The psychological rehabilitation of child soldiers is an obligation, according to Article 39 of the United Nations Convention on the Rights of the Child.3
Despite the large number of children serving as soldiers in armed conflicts, the effect of being a child soldier on the children's mental health has been minimally researched. In 2004, Derluyn et al4 evaluated clinically significant symptoms of posttraumatic stress disorder (PTSD) in former Ugandan child soldiers, using the Impact of Event Scale–Revised.5 The authors reported that nearly all children (97%) showed posttraumatic stress reactions of clinical importance. The question of whether posttraumatic stress might influence the individual's attitude toward reconciliation has been investigated very little. In 2004, Pham et al6 examined this association in 2074 adult survivors of the Rwandan genocide, using the PTSD Checklist–Civilian Version7 and a newly developed reconciliation questionnaire. The investigators demonstrated that traumatic exposure and PTSD symptoms were associated with attitudes toward reconciliation.
Therefore, we sought to assess the prevalence of PTSD symptoms in former Ugandan and Congolese child soldiers and to explore how PTSD symptoms are associated with these children's openness to reconciliation and feelings of revenge on the person or group they consider their enemy.
The survey was conducted between May and July 2005 in Uganda and the Democratic Republic of the Congo. The study sample consisted of former child soldiers, living in rehabilitation centers run by aid organizations in northern Uganda (Gulu; World Vision) and eastern Congo (Goma, Bukavu; UNICEF). After demobilization or escape from the armed groups, the children stayed in these centers exclusively for former child soldiers to get medical and psychological support and to trace their family or to find a foster family. Approximately 20 000 former child soldiers in the Democratic Republic of the Congo and Uganda have passed through rehabilitation centers during the last 10 years.8,9 Because of the poor security situation with ongoing ambushes, traveling was almost impossible. Therefore, this study had to be restricted to the rehabilitation centers.
To be eligible, children had to be 18 years or younger and living in a rehabilitation center at the time of data collection. Study participants were identified by selecting every second child listed on an index of names of the former child soldiers living in the respective rehabilitation center. The children were asked to volunteer for the survey.
This cross-sectional field study was questionnaire based. Demographic variables assessed included age, sex, age when entering the armed forces, period served, time since demobilization, education level, and ethnicity. The ethnicity question aimed to convey the diversity of groups living in Eastern Congo. Ethnicity was defined independently by respondents. Sample-specific exposure to potentially traumatizing events was assessed by 17 yes/no questions, developed by considering interviews with local aid workers and literature reports4,10 on this topic.
Symptoms of PTSD were evaluated using the Child Posttraumatic Stress Disorder Reaction Index (CPTSD-RI).11 The CPTS-RI is one of the most frequently and transculturally used self-report measures to assess PTSD symptoms in children.12 The version of the scale used in this study includes 17 items, corresponding to the Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition) PTSD symptom criteria.13 The scale includes 3 subscales: intrusion (5 items), avoidance (7 items), and arousal (5 items). Items are rated on a 0 to 4 scale, assessing the presence of PTSD symptoms during the last month. A score higher than 35 was used to classify a child as having clinically important PTSD symptoms.13 Internal consistency of the scale in the present study was Cronbach α= 0.68.
The children's openness to reconciliation and their feelings of revenge were assessed separately by 2 questionnaires, each with 8 yes/no items (eg, “I am going to pay back the persons who harmed me for what they did”; “I am ready to forgive the persons who harmed me”). The scales were developed by one of the authors (H.A.) according to relevant literature, expert interviews, and the Transgression-Related Interpersonal Motivations Inventory (TRIM).14 Before this study, both scales were pilot tested among 112 war-affected children in Kosovo and 215 war-affected refugee children in Hamburg, Germany.15 Internal consistency of the scales in the present survey was high, with Cronbach α= 0.81 (child reconciliation questionnaire) and Cronbach α= 0.87 (child revenge questionnaire).
Questionnaires were translated into the children's native language (Luo in Uganda) or their lingua franca (Swahili in the Democratic Republic of the Congo) by professional linguists and verified for accuracy and comprehensibility by bilingual local mental health staff. All children were interviewed individually by trained local mental health staff in the presence of the first author (C.P.B.), who explained the purpose of the survey. The interviews lasted approximately 60 minutes and were conducted in private areas of the rehabilitation centers, after voluntary informed oral consent was obtained from the participants and their guardians. In addition, the interviewers were not involved in the wars, and study participants were assured that their answers would not affect the support they receive in the centers. A psychological counselor was available to provide support if necessary. This study was approved by the ethics committee of the Medical Association of Hamburg (Ethik-Kommission der Ärztekammer Hamburg; reference number 2671).
Exposure to trauma and PTSD symptom scores were normally distributed, whereas scores of openness to reconciliation (skewed left) and feelings of revenge (skewed right) were not. Hence, relationships between different measurement scores were tested by using Spearman ρ correlations and, if normally distributed, Pearson r correlation. Differences between subgroups were tested using Mann-Whitney U tests or t tests. Analyses were performed with SPSS version 15.0 (SPSS Inc, Chicago, Illinois).
Of the 296 former child soldiers living in the rehabilitation centers at the time of this study, 172 were invited to participate, and 3 refused participation (children were not asked to justify refusal). At least 52% of the children living in the rehabilitation centers were interviewed (62% [58/93] in Gulu [Uganda], 55% [88/159] in Goma [Democratic Republic of the Congo], and 52% [23/44] in Bukavu [Democratic Republic of the Congo]).
The 169 participants were a mean of 15.3 years old (SD, 1.6; range, 11-18 years) at the time of data collection. Of these, 141 (83.4%) were boys, 23.7% had no education, and 74.0% had education on an elementary level and 2.4% on a secondary level (Table 1). Children reported having been recruited by armed forces at a young age (mean [SD], 12.1 ; range, 5-18 years) and having served a mean of 38.3 months (SD, 24; range, 0-96 months). These former child soldiers had been demobilized a mean of 2.3 months before data collection (mean, 2.3; median, 2.0; SD = 2.4 months).
Table 2 lists the types and reported frequencies of war-related traumatic experiences of these former child soldiers. The mean number of potentially traumatic events experienced per child was 11.1 (SD, 2.99). The most commonly reported traumatic experiences were having witnessed shooting (92.9%), having witnessed someone wounded (89.9%), and having been seriously beaten (84%). A total of 54.4% reported having killed someone, and 27.8% reported that they were forced to engage in sexual contact.
The mean total PTSD symptom score was 29.0 (SD, 9.1; range, 19-51). The mean scores on the 3 subscales were intrusion, 9.19 (SD, 4.07; range, 0-18); arousal, 9.36 (SD, 3.44; range, 2-20); and avoidance, 10.59 (SD, 4.90; range, 0-24). Of these former child soldiers, 59 (34.9%; 95% CI, 34.4%-35.4%) had a PTSD symptom score higher than 35. Traumatic exposure score was not associated with PTSD symptom score (r = 0.07; P = .39; 95% CI, −0.03 to 0.07).
The more PTSD symptoms children had, the less they were willing to reconcile (ρ= −0.34; P < .001) and the more they demonstrated feelings of revenge (ρ= 0.29; P < .001) (Table 3). Children who met PTSD symptom criteria with scores higher than 35 differed significantly in their openness to reconciliation (Mann-Whitney U test, 2092.00; P < .001) and their feelings of revenge (Mann-Whitney U test, 2032.50; P < .001) from children who scored below the cutoff.
When the sample was split at the median for time since demobilization, children demobilized longer (>2 months) showed more openness to reconciliation (Mann-Whitney U test, 2773.00; P = .01) but no difference concerning feelings of revenge (Mann-Whitney U test, 3318.00; P = .49) from recently released children. Among children demobilized less than 2 months before this study was conducted, a stronger association between symptoms of PTSD and less openness to reconciliation was found than in children who had been demobilized longer than 2 months before this study (ρ= −0.48, P < .001 vs ρ= −0.15, P = .20). The association between PTSD and feelings of revenge in children demobilized less than 2 months before this study and longer than 2 months before this study differed only marginally (ρ= 0.30, P = .03 vs ρ= 0.28, P = .02). PTSD symptoms subscale scores for intrusion, arousal, and avoidance were significantly associated with children's openness to reconciliation and feelings of revenge (intrusion: ρ= −0.27, P < .001 and ρ= 0.21, P < .001; arousal: ρ= −0.34, P < .001 and ρ= 0.27, P < .001; avoidance: ρ= −0.21, P < .01 and ρ= 0.22, P < .005).
Children who were threatened with death or serious harm had a significantly lower openness to reconciliation (Mann-Whitney U test, 2056.50; P = .001) and slightly significantly more feelings of revenge (Mann-Whitney U test, 2466.00; P = .052) but no differences in PTSD symptoms (t167 = −1.25; P = .22). No other differences concerning particular kinds of exposure to traumatic events could be found. Duration of being a child soldier was not found to be associated with PTSD, openness to reconciliation, or feelings of revenge. In addition, sex, region, ethnicity, or education did not have a significant effect on the main outcome variables (number of potentially experienced traumatic events, PTSD symptoms, openness to reconciliation, feelings of revenge).
More than one-third (34.9%) of the 169 interviewed former child soldiers from Uganda and the Democratic Republic of the Congo met symptom criteria for PTSD (symptom score >35). This rate is low compared with the rate of PTSD symptoms in former child soldiers found in another study in the region.4 Variation in the scales and cutoffs used to assess PTSD symptoms among the selected populations in these studies may explain these different results.
Contrary to expectation and findings of other studies,16- 18 there was no significantly positive association between traumatic experiences and PTSD symptoms and no association between particular kinds of exposure and measures of PTSD symptoms. However, the former child soldiers in this study who showed clinically relevant symptoms of PTSD had significantly less openness to reconciliation and significantly more feelings of revenge than those with fewer symptoms. Hence, posttraumatic stress might hinder the children's ability to deal with and overcome emotions of hate and revenge. Likewise, the children with PTSD symptoms might regard acts of retaliation as an appropriate way to recover personal integrity and to overcome their traumatic experiences. Therefore, posttraumatic stress might be an important factor influencing postconflict situations and may contribute to the cycles of violence found in war-torn regions.
This study had some limitations. The results of this study may not be generalizable, because of the small sample size and the nonrandom selection of participants from selected rehabilitation centers in relatively safe areas. In addition, the number of girls in this study is low compared with that of a previous report that estimates the proportion of female child soldiers to be 40%,8 which limits any interpretations of results by sex. The reliance on self-reported scales, the potential that children responded in a socially desirable way, and the lack of sufficient validity of the scales used to assess children's feelings of revenge and openness to reconciliation among this population are additional limitations. Finally, language and other transcultural errors may occur when western psychological concepts and instruments are used with nonwestern populations.
The results of this study cannot determine whether openness to reconciliation and fewer feelings of revenge are inner personal characteristics that prevent PTSD symptoms or whether PTSD symptoms mediate the openness to reconciliation and feelings of revenge. However, our findings indicate that mental distress and mental illness, namely, symptoms of PTSD, are associated with war-affected children's attitudes toward reconciliation and could therefore impose barriers to sustainable and long-term peace building. Hence, the results of this study support the need to fulfill the obligation under the United Nations Convention on the Rights of the Child3 to promote psychological recovery for war-affected children, such as child soldiers.
Corresponding Author: Fionna Klasen, Dipl-Psych, Children for Tomorrow Foundation, Outpatient Clinic for Refugee Children and Their Families, University Clinics of Hamburg-Eppendorf Martinistraße 52, 20246 Hamburg, Germany (email@example.com).
Author Contributions: Mr Bayer had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Bayer, Klasen, Adam.
Acquisition of data: Bayer.
Analysis and interpretation of data: Bayer, Klasen.
Drafting of the manuscript: Bayer, Klasen.
Critical revision of the manuscript for important intellectual content: Bayer, Klasen, Adam.
Statistical analysis: Bayer, Klasen.
Obtained funding: Bayer, Klasen, Adam.
Administrative, technical, or material support: Bayer, Klasen, Adam.
Study supervision: Adam.
Financial Disclosures: None reported.
Funding/Support: Our work was supported by the Children for Tomorrow Foundation. Mr Bayer was supported by a doctoral thesis scholarship of the Werner Otto Foundation. The costs of travel for Mr Bayer were sponsored by the German Academic Exchange Service (DAAD). UNICEF and World Vision provided logistical support.
Role of the Sponsor: The funding organizations had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or the preparation, review, or approval of the manuscript.
Additional Contributions: We thank the team of UNICEF in Goma and Bukavu (Democratic Republic of the Congo), the team of UNICEF Germany, and the team of World Vision in Gulu (Uganda) for their strong support of this project. We also thank all colleagues from the Department of Child and Adolescent Psychiatry, University Clinic Hamburg, Germany, and particularly Peter Riedesser, MD, for their supervision and helpful comments. We are grateful to the Werner Otto Foundation, the DAAD, and the Children for Tomorrow Foundation for their generous support. We are especially grateful to all the children who made the study possible by sharing with us their experiences. Persons or organizations named in the acknowledgment did not receive any compensation.