Background Numerous factors are thought to connect violence to mental illness, including civil conflict, yet very few studies of this relationship have been conducted using large, population-based samples in developing nations.

Aim To test the association between prior violent events and mental health outcomes for Guatemala.

Methods We collected and analysed a national cluster random sample of 1452 participants as part of the 2009 Guatemalan National Survey of Mental Health. Our exposure of interest was the most stressful or upsetting violent event experienced by participants; ICD-10 criteria classified mental illness. Analyses of the 36-year Guatemalan civil war as an effect modifier were also completed.

Results In total, 22.51% participants experienced ≥1 prior violent event. The proportions of subjects who were diagnosed for PTSD, depression, and anxiety were 5.45%, 3.66%, and 1.85%, respectively. The unadjusted risk ratios associated with prior violent event for PTSD, depression, and anxiety were 18.21 (95% CI 6.40 to 51.81, p<0.001), 1.29 (95% CI 0.60 to 2.76, p=0.504), and 1.50 (95% CI 0.55 to 4.08, p=0.417), respectively. After adjusting for various confounders, risk ratios for PTSD, depression, and anxiety were 21.73 (95% CI 7.23 to 65.23, p<0.001), 1.36 (95% CI 0.58 to 3.20, p=0.469), and 1.59 (95% CI 0.51 to 4.97, p=0.411), respectively. These risk ratios were larger and more significant during the Guatemala civil war.

Significance Guatemalans exposed to prior violent events had a significantly higher risk of developing PTSD. Mental illness treatment and prevention programmes that identify individuals with prior exposure to violent events could be of great value to developing nations, many of which have a recent history of civil war or are actively involved in civil war.