Sequel of war : Investigating appetitive aggression, violence perpetration, PTSD symptoms and treatment effects using psychological and epigenetic variables of the doctoral thesis

Abstract

Aggression and traumatic experiences are common in individuals affected by war. Persons who are actively involved in organised violence such as combatants or soldiers have previously reported developing an attraction towards the use of violence. These positive feelings, the excitement and thrill when exercising violence, are referred to as Appetitive Aggression (AA) and can be awakened in all of us (Elbert et al., 2017a). In particular, individuals who enter a violent environment early in life, such as gang members or child soldiers, are known to experience more traumatic event types (traumatic load), to perpetrate more violence and to exhibit higher levels of AA than their adult counterparts (Hermenau et al., 2013). Additionally, specific sensitive periods between the age of 16 and 17 have been proposed to develop a robust trait of AA (Köbach & Elbert, 2015). The present thesis investigates the interplay of psychological and epigenetic variables related to AA, violent behaviour, posttraumatic stress disorder (PTSD) symptoms, and therapy effects. Specifically, this thesis addresses the following objectives in four chapters: 1. To model AA and perpetrated violent acts in a sample of survivors of the civil war in Northern Uganda; 2. To identify epigenetic associations for AA in women and men affected by war; 3. To detect longitudinal DNA methylation (DNAm) patterns linked to the course of PTSD in a cohort of active-duty soldiers after deployment; 4. To investigate epigenetic changes related to psychotherapy in a sample of former female child soldiers. Chapter 2 of the present thesis looks at crucial predictors in modelling AA and perpetrated violent acts. Analysing data of a sample of survivors of the civil war in Northern Uganda, the combination of early abduction into an armed group and a high traumatic load were significant predictors of AA and perpetrated violence. Additionally, in individuals with combat experience, this relationship between the predictors and AA and perpetrated violent acts was even more pronounced. Moreover, for AA there were no sex differences while for perpetrated violence, men reported a higher number of events than women. This study underlines the significance of combat experience when it comes to AA and violence perpetration. Furthermore, it shows that in particular persons who entered an armed group at a young age and have experienced a high number of traumatic events are at risk and thus should receive special attention in reintegration processes. So far, there is only one study investigating epigenetic patterns for AA (Xulu et al., 2021). This study analysed promoter regions of 22 genes and not the whole epigenome. Hence, epigenome-wide research on AA has been – up to this point – non-existent. Therefore, chapter 3 investigates DNA methylation (DNAm) patterns for AA in women and men using epigenomewide analyses. For the model in men, three different cohorts from Burundian soldiers, Ugandan survivors of the civil war, and Congolese former combatants were combined to increase the sample size. For the model in women, two cohorts were combined, a sample of Ugandan survivors of the civil war, and a sample of Congolese former child soldiers. Elastic net regression was used to cope with the methodological challenge of having more predictors than observations. For men, the final model selected combat experience, traumatic load, PTSD symptoms and cg04359295 on the intersection of the SEMA4G and MRPL43 genes. The predictors selected for the female model were combat experience and cg18081666 located on the RNF182 gene. The findings suggest that there might be different underlying biological mechanisms in women and men for AA. Moreover, soldiers are at risk to experience traumatic events during war deployment (Hoge et al., 2004; Wittchen et al., 2012). As the likelihood of developing PTSD increases in a dose-response manner with every new experienced traumatic event type (Kolassa & Elbert, 2007; M. Schauer et al., 2003; Wilker et al., 2015), a significant number of soldiers suffer from PTSD or symptoms of it (Levin-Rector et al., 2018; Thomas et al., 2010). There is a growing body of research on PTSD and epigenetic associations in civilian and military cohorts. Longitudinal studies found associations with genes involved in memory processes, the immune system, and synaptic plasticity (Katrinli et al., 2022; Rutten et al., 2018). For military samples, the research focus so far was mostly placed on PTSD symptoms before and during deployment. Hence, chapter 4 of this thesis presents findings on DNAm patterns related to PTSD symptoms in a longitudinal study after deployment. For this study, data of Burundian soldiers after their African Union Mission in Somalia (AMISOM) deployment and 9 to 14 months after their return were collected. The focus of this study was to identify epigenetic patterns that explain why some soldiers showed an improvement in PTSD symptoms after their return, while for others, their mental health deteriorated. Findings for gen enrichment analysis showed a link between DNAm and linoleic acid pathways for deterioration of PTSD symptoms. Linoleic and alphalinolenic administration has previously been shown to buffer against stress reactions (cortisol and cholesterol blood levels) and learning deficits in rats (S. Yehuda et al., 2000). This suggests memory related epigenetic effects for the deterioration of PTSD symptoms in active soldiers. Because combatants are at high risk to develop PTSD and show elevated levels of aggression, their reintegration into civil society can be challenging (Hermenau et al., 2013; Koebach et al., 2021). Also, different social rules between armed groups and civil society and stigmatisation experiences exacerbate these difficulties often resulting in mental health problems (Schmitt et al., 2021). Narrative Exposure Therapy for Forensic Offender Rehabilitation (FORNET) has been developed to reduce PTSD symptoms and AA levels (Elbert et al., 2012). Studies in gang members, former children in street situations, and former combatants show successful PTSD symptom reduction and decreased AA levels (Crombach & Elbert, 2015; Hinsberger et al., 2020; Koebach et al., 2021; Robjant et al., 2019). The topic of chapter 5 is epigenetic changes related to FORNET in a sample of former female child soldiers in Eastern Democratic Republic of the Congo (DRC). To this end, longitudinal DNAm changes that are associated to treatment effects and changes of DNAm in Cytosine-phosphate-guanine (CpG) sites linked to clinical (PTSD, depression, AA) and social (social acknowledgment, feelings of guilt, current violent behaviour) outcomes were investigated. Additionally, gene ontology analyses were performed to further investigate functional relationships between found CpG sites and genes. For replication of PTSD, depression, and AA outcomes an independent sample of former female Ugandan child soldiers was used. Main results showed a link between treatment and CpG sites at genes involved in memory formation and long-term memory and the immune system which were partially replicated in the Ugandan sample. For the clinical and social outcomes, dopamine and cytokine related genes, amongst others, were found. Gene ontology analyses identified pathways implicated in Hypothalamic-pituitary-adrenal (HPA) axis functioning, glutaminergic signalling, and aldosterone synthesis and secretion. Overall, the results of this study suggest that trauma-focused treatment has far-reaching effects on the epigenome that exceed a simple reversal of DNAm changes to their corresponding pre-treatment state. In summary, this thesis provides novel insights into the interplay of psychological and epigenetic predictors related to AA, violence perpetration, PTSD symptoms, and psychotherapy. AA and violence perpetration are prevalent characteristics found in former combatants and active soldiers, a pattern that is amplified by combat experience. Considering psychological variables, study 1 did not indicate sex differences for AA. Including epigenetic predictors, a different image presents itself. Here, as indicated by study 2, a CpG site on MRPL43/SEMA4G genes is a predictor in men, while in women, a CpG on RNF182 predicts AA. These results suggest different biological processes at play for AA in women and men. Thus, future research should investigate these sex differences in more detail to verify these findings. Additionally, study 1 suggests that in particular persons abducted at a young age into a rebel group and presenting a high traumatic load are at risk to develop high levels of AA and perpetrate violence. This indicates the need to address these challenges in reintegration programs for combat returnees to ensure a successful social reintegration. Approaches that can be useful are FORNET and NETfacts, an intervention which processes traumatic events and stigmatisation experiences on individual and community level. Moreover, this thesis offers novel perspectives on deterioration of PTSD symptoms. The linoleic acid pathway was significantly enriched in the gene enrichment analysis in study 3, suggesting a relationship between DNAm of memory-related genes and deteriorating of PTSD symptoms. Furthermore, study 4 shows that trauma-focused psychotherapy impacts DNAm and changes in accordance with treatment. Important associations with treatment were HPA related pathways and genes involved in memory processes and the immune system. The results indicate that treatment effects are not necessarily reversing DNAm to pre-treatment status; more complex processes seem to be at play. In conclusion, this thesis presents first evidence on epigenome-wide DNAm associations for AA and related concepts and offers thus novel starting points for future research.