Escalating violence is omnipresent in South African townships and can be traced back to two core mechanisms: a trauma-related hyper-arousal and a positive rewarding perception of violence. In the past, there was no therapeutic intervention available addressing both, trauma and the so-called appetitive aggression. The University of Konstanz has developed a culturally sensitive and scientifically based short-term intervention for the treatment of traumatized patients, the Narrative Exposure Therapy (NET), that has proven its effectiveness in various randomized clinical studies in different war-affected populations. Recently, the NET has been adapted for the forensic offender rehabilitation (FORNET) by also addressing the perpetration of violence related to a self-rewarding perception of the exposure with violence. It has shown to be effective in reducing the number of committed offenses in a perpetrator sample in Burundi and to reduce PTSD symptoms in a perpetrator sample in the Democratic Republic of the Congo. In this study, we investigated the therapeutic efficiency of FORNET in a randomized clinical control trial with a sample of former offenders of the townships of Cape Town. In addition to the previous studies, we specifically addressed the context of ongoing stress and linked our findings to epigenetic markers of stress and violence. Participants were followed over a period of up to 25-months post-treatment. The FORNET was also disseminated to local staff of our collaboration partners from the South African Universities and an organization working in the townships to warrant sustainability of the therapeutic intervention.

A group of 290 former young male offenders and young males at risk for perpetrating or becoming victims of crime (age 10-40) from Cape Town, recruited via a community based skill-training program has been interviewed in a first step. Participants with neurological or psychiatric disorders other than those resulting from exposure to traumatic stress, like chronic psychosis, were excluded from participation in the study but received referrals in the professional health care system, if necessary.

We applied standardised clinical questionnaires in a semi-structured interview for the assessment of exposure to traumatic stress (Children Exposed to Community Violence (CEVC)), trauma symptomatology (PSS-I for diagnosis of post-traumatic stress disorder), aggression (Appetitive Aggression Scale (AAS); Buss & Perry Aggression Questionnaire), offense types checklist (from the AAS) and related drug abuse, psychosocial functioning (Work and Social Adjustment Scale, WSAS) and suicidality (M.I.N.I. 6.0.0., shortened version). Additionally, we used a questionnaire about social acknowledgement (Social Acknowledgement Questionnaire) and substance dependence (M.I.N.I. version 6.0.0).

We set up a randomised clinical intervention trial for the treatment of former offenders that covers violence perpetration on the one hand and traumatization on the other. 90 participants were chosen by the highest aggression scale scores and a certain threshold of trauma symptoms. 17 participants received FORNET, while 14 matched participants received CBT and another 7 matched participants treatment as usual (Waiting List camp) in a camp. Matching took place according to the severity of trauma symptomatology, aggressive behavior and suicidality. 36 participants didn't receive intervention and weren't part of the camp (Waiting List no camp) but were included in the follow-ups. In order to compare the treatment effect with the two control groups and follow the long-term effects, post-treatment assessments with the initial survey were performed about 8, 18 and 25 months post-treatment.

After the follow-ups, 17 FORNET participants and 11 CBT participants were included in the analysis. Two CBT clients had been excluded since they had not come to the follow-ups, one CBT client died a relatively sudden death due to a severe disease. The camp waiting list had been reduced to only six persons since one participant of the waiting list has been stabbed to death in a gang fight. Therefore a combined waiting list has been emerged from the two waiting lists ("camp" and "no camp"). 13 men from the "no camp"-waiting list were excluded since they have not been to the follow ups. From the residual 21 clients from this waiting list, 11 have been matched (together with the six participants from the "camp"-waiting list) with the participants from the FORNET and CBT group in terms of post traumatic stress symptom severity, level of appetitive aggression and suicidality. In the end 17 FORNET-, 11CBT- and 17 waiting list-attendees have been included into the analysis, hence a total sample size of 45 participants.

The 2 subjects of the control group, that still fulfilled the criteria for the therapy study and were interested in it, received FORNET by trained ZA staff for ethical reasons and to examine whether it is possible to effectively disseminate FORNET.

In addition to the subjective data, we will corroborate the results with the assessment of epigenetic consequences of appetitive aggression and trauma and investigate its significance and persistence in the long-term course of the study. We propose that specific DNA methylation patterns are important markers of environmentally driven mental health pathologies, such as chronic aggression/violence and traumatic stress in such a sample of perpetrators or young males at risk for perpetration. Further, we will test the stability of DNA methylation and its potential modification over a period of up to 25 months by behavioural intervention (FORNET).

Saliva samples (2ml) were collected through a non-invasive method using Oragene-Discover (OGR- 500) Collection Kit (DNA genotek, Ontario, Canada). We will collect two DNA saliva samples and one RNA saliva sample. Methylation patterns will be assessed at the University of Stellenbosch (Prof. Soraya Seedat), focussing on markers of trauma and violence.

Inclusion Criteria:

• High levels of PTSD symptoms (i.e. Scores on the PSS-I > 7 points)
• High levels of appetitive aggression (i.e. Scores on the Appetitive Aggression Scale (AAS) > 8 points)
• participants are recruited through the community based re-integration program

Exclusion Criteria:

• acute psychotic episode