UC San Diego

Abstract Background: Contrary to popular belief the majority of patients with schizophrenia will never commit an act of severe violence. Highest risk for violent offending appears to be during the first episode of psychosis (FEP), a meta-analysis by Large and Nielssen (2010) reporting that about a third of patients in the FEP exhibited some violent behavior before initial treatment. However, most acts involve minor violence and fewer than 1 in 100 patients committed assaults resulting in serious injury. Risk of violence in clinical high risk populations, those with attenuated psychosis symptoms (APS) remains unexplored although recent findings by Marshall and colleagues (2016) suggest that a significant amount of APS involve violent content although most is self-directed violence. The aim of the current study was to explore risk of violence in individuals with attenuated psychosis symptom syndrome (APSS). We were also interested to investigate whether risk of violence as assessed was associated with symptomology. Methods: The Structured Assessment of Violence Risk in Youth (SAVRY) was completed for 285 individuals who met criteria for APSS as well as 44 Healthy Controls. The SAVRY is a clinician rated assessment tool and provides a clinical risk rating for each individual by assessing multiple domains including Historical Risk Factors, Social/Contextual Risk Factors, and Individual/Clinical Factors. Results: Violence risk was significantly different between the two groups with healthy controls assessed to be at a lower risk than individuals in the attenuated psychosis symptoms group, χ2(2) = 13.03, P < .001. Only 2 of the APSS group and none of the healthy controls were rated to be at high level of risk for violence. Violence risk was not different between men and women, χ2(2) = 4.58, P = 1.01. A between-group ANOVA conducted to compare severity of attenuated psychotic symptoms and risk of violence within the APSS group showed that symptom severity was significantly different across levels of violence risk, F(2, 235) = 6.32; P < .002. This was largely driven by negative symptoms severity. Individuals with low risk for violence as compared to moderate level of violence risk had lower negative symptoms, F(2, 235) = 13.42; P < .000. Severity of positive, disorganized, or general symptoms independently did not differ across levels of violence risk. This relationship remained unchanged when income level was adjusted. Level of income, age, ethnicity, or parental education was not associated with level of risk. Conclusion: To our knowledge this is the first study to assess violence risk in individuals with APSS. While, the APSS group was assessed to be at a higher risk for violence as compared to healthy controls, the majority were judged to be at a moderate risk and high risk ratings were rare. In the APSS group higher risk was associated with symptoms. Specifically, results suggest that negative symptoms uniquely contribute to risk of violence.