In a forensic hospital study, violence clustered in the periods when treatment had stopped

Comparing 55 patients with schizophrenia who had killed with 49 who had been violent without killing, Korean researchers found most acts occurred during treatment discontinuation, driven by persecutory delusions.

The overall crime rate among people with schizophrenia is not meaningfully higher than the general population's, which makes the question of which specific circumstances carry risk more useful than the question of whether the diagnosis does. Researchers led by Yujin Lee reviewed the records of patients admitted to Korea's National Forensic Hospital between 2020 and 2024, comparing 55 who had committed homicide with 49 who had been violent without killing. They extracted clinical history, trauma history, treatment adherence, cognitive testing, symptom severity on the PANSS, and personality assessment. The finding that applies to both groups is the important one: most violent acts occurred during periods when treatment had been discontinued, and were most commonly driven by persecutory delusions and hallucinations. Both groups also frequently had limited social support and a history of childhood trauma or victimisation in adulthood. The groups differed in pattern rather than in kind. The homicide group more often committed premeditated, organised acts in private settings against family members, and scored higher on measures of antisocial traits and suicidal ideation. The non-homicide group more often acted impulsively, in public, against strangers, without a weapon, and had higher positive-symptom scores, more frequent violent incidents and more rehospitalisations. This is a single forensic hospital, retrospectively reviewed, with around a hundred patients, and everyone in it had already been violent, so it cannot estimate risk in the wider population and should not be read as doing so. What it describes is where violence sat relative to treatment among people who had already come to forensic attention.