Abstract:
Background: The Kurdistan Region has experienced major human rights violations, including the 1988 chemical attacks during the Anfal Campaign carried out by the Iraqi regime, and the 2014 genocide of Yazidis perpetrated by the Islamic State (IS). These events have had long-term effects on survivors’ mental and physical health.
Objective: This study examined genetic, trauma-related, and psychosocial factors associated with post-traumatic stress disorder (PTSD) and related symptoms in 892 Kurdish trauma survivors: chemical gas survivors (n = 534) and Yazidi survivors of IS violence (n = 358). We assessed whether PTSD polygenic risk scores (PRS) are associated with symptom severity, whether this varies with cumulative trauma exposure, and whether trauma-related factors explain additional variance beyond demographic and psychosocial variables. Differences between the two cohorts were also examined.
Methods: Data were collected between March and September 2023 from survivors of the chemical gas attacks and between May and September 2024 from Yazidi survivors of ISIS attacks. Participants completed the PTSD Checklist for DSM-5 (PCL-5), the Patient Health Questionnaire-15 (PHQ-15), the Perceived Stress Scale-14 (PSS-14), and the Hopkins Symptom Checklist-25 (HSCL-25). Deoxyribonucleic acid (DNA) was extracted from saliva samples and genotyped using genome-wide single nucleotide polymorphism (SNP) arrays to calculate PTSD polygenic risk scores (PRS). Hierarchical regression models were used to examine demographic, psychosocial, and trauma-related predictors of psychopathology. Linear regression and mixed-effects models were subsequently applied to assess the main effects of PRS and PRS × trauma interactions, adjusting for age, gender, education, population structure, and relatedness. In addition, this study contributed to the translation and cultural adaptation of several psychometric instruments (PHQ-15, PCL-5, HSCL-25, and PSS-14) into Kurdish. Internal consistency and reliability were evaluated using Cronbach’s alpha and McDonald’s omega.
Results: Demographic variables explained 10-38% of the variance in mental health outcomes. In the chemical gas cohort, cumulative trauma (β = 0.21-0.24), lower education (β = −0.17 to
−0.21), and female gender (β = 0.03-0.35) were consistently associated with higher symptom levels. Psychotropic medication use was also positively associated with symptoms. In the Yazidi cohort, cumulative trauma (β = 0.19-0.37), lower education, and female gender were the main predictors, while social support and medication had limited additional effects beyond demographics and cumulative trauma. Across both cohorts, cumulative trauma exposure and education were the most stable predictors of PTSD, depression/anxiety, somatic symptoms, and perceived stress.
In chemical gas survivors, PTSD PRS showed a small but significant main effect (B = 1.76, p
= .010) and a significant interaction with trauma exposure (B = 1.44, p = .032), indicating that genetic risk became more apparent at higher levels of cumulative trauma. In Yazidi survivors, PRS showed no significant association with symptoms and no interaction with trauma exposure.
Conclusions: The results suggest that PTSD PRS is modestly associated with symptom severity and interacts with trauma exposure among chemical gas survivors, but not among Yazidi survivors. Across both cohorts, cumulative trauma exposure, female gender, and lower education consistently emerged as key predictors, highlighting the central role of trauma and socioeconomic context. Psychotropic medication and psychosocial factors showed stronger associations in the chemical gas cohort, whereas social support appeared more protective among Yazidi survivors, underscoring the importance of context-specific environmental and structural factors. Overall, the findings highlight the importance of considering ancestry and population structure in genomic prediction models and support the need for context-sensitive interventions in long-term trauma-exposed Kurdish populations.