Clinical and Criminal Characteristics of Patients With Schizophrenia Exhibiting Homicidal and Non-Homicidal Behavior: A Comparative Analysis.

BACKGROUND: Public concern regarding homicides committed by individuals with schizophrenia has intensified in recent years. Although the overall crime rate in this population is not significantly higher than that of the general population, certain subgroups may present an elevated homicide risk. Identifying associated risk factors and distinguishing between impulsive and premeditated violence are crucial for developing effective forensic psychiatric assessment and prevention strategies. METHODS: Medical records of patients with schizophrenia admitted to the National Forensic Hospital in Korea from 2020 to 2024 were retrospectively reviewed. Patients were classified into the homicide group (HG; n = 55) and non-homicide violent group (VG; n = 49). Demographic and clinical data, trauma history, treatment adherence, cognitive function (Korean Wechsler Adult Intelligence Scale), symptom severity (Positive and Negative Syndrome Scale [PANSS]), personality traits (Minnesota Multiphasic Personality Inventory [MMPI] and Personality Assessment Inventory [PAI]), and crime-related characteristics were extracted from psychiatric evaluations, inpatient records, and psychological test results. The χ² tests, t-tests, and Mann-Whitney U tests were used for statistical analysis. RESULTS: Most violent acts occurred during periods of treatment discontinuation and were most commonly driven by persecutory delusions and hallucinations. Both groups frequently experienced limited social support and a history of childhood trauma or adult victimization. The HG was more likely to commit premeditated, organized crimes using sharp weapons against family members in private settings. Conversely, the VG tended to exhibit impulsive violence in public settings, often against strangers and without weapons. Persecutory delusions predominated in both groups, whereas grandiose delusions were more prevalent in the VG. The HG had higher MMPI Psychopathic Deviate and Psychasthenia subscale and PAI suicidal ideation scale scores, whereas the VG showed higher PANSS positive symptom scores. The VG also demonstrated more frequent violent incidents and rehospitalizations. CONCLUSION: Although psychotic symptoms underpinned violence in both groups, the HG showed antisocial traits, suicidal tendencies, and premeditated violence. Conversely, the VG exhibited more severe positive symptoms, grandiosity, and recurrent impulsive violence. These findings highlight the importance of comprehensive violence risk assessments in schizophrenia that incorporate symptomatology, personality pathologies, and violence typologies.