Psychoticlike symptoms not tied to risk of violent recidivism



Psychoticlike symptoms were not positively associated with violent recidivism in a population of male adolescents, even when these symptoms were comorbid with substance abuse disorders, a study has shown.

In addition, having paranoid delusions and threat/control override delusions was actually linked to a lower risk of reoffending with violence, reported Olivier F. Colins, Ph.D., of Leiden (the Netherlands) University Medical Center and his associates in the June issue of the Journal of Nervous and Mental Disease.

"By identifying detained youths with [paranoid delusions] or [threat/control override delusions], clinicians are likely to identify youths with a low risk for future violent crimes," Dr. Colins and his associates wrote (J. Nerv. Ment. Dis. 2013;201:478-83).

The researchers assessed 224 male adolescents, aged 12-17, with the Diagnostic Interview Schedule for Children (DISC-IV), specifically analyzing the results of the schizophrenia module and the substance use disorder module. The boys had all been recently detained at youth detention centers in Flanders, Belgium, and were recruited between January 2005 and February 2007.

All boys were of Belgian origin except 22.3% who were of Moroccan origin. All of them had been placed in detention for at least 1 month. A quarter (25%) of the boys had a criminal history of violence. Violent recidivism included arrests, after the initial clinical interviews, for murder, manslaughter, assault, sexual offenses, and theft with violence.

The schizophrenia module of the DISC-IV investigates 22 psychoticlike symptoms within the past year that are grouped into four subsets: paranoid delusions, nonparanoid delusions, hallucinations, and threat/control override delusions. The substance use disorder module measured the existence of alcohol use disorder, marijuana use disorder, and other substance abuse disorders, including cocaine and amphetamines.

A total of 79% of the boys had at least one psychoticlike symptom, including 43% with at least one hallucination, 67% with at least one paranoid delusion, 26% with at least one nonparanoid delusion, and 53% with at least one threat/control override delusion. Substance abuse disorders were evenly distributed: 26% of the boys had no substance abuse disorder, 23% had one, 24% had two, and 27% had three. Just over half the boys (54%) had substance use disorder, 64% had marijuana use disorder, and a third (33%) had other substance use disorders.

After a mean follow-up of 1,219 days, two-thirds (66%) of the boys were rearrested for violent crimes during their time at risk, defined as the number of days between the DISC-IV and Feb. 1, 2009, minus the participants’ days in detention.

In both the univariate and multivariate analyses, odds ratios for all psychoticlike symptoms and total substance use disorders were not significant for predicting violent recidivism or violent rearrest frequency – except total delusions, paranoid delusions, and threat/control override delusions, which showed several significant negative associations. In the univariate analysis for violent rearrest frequency, total delusions had an odds ratio of 0.84 (CI, 0.73-0.98; P = .03), paranoid delusions had an OR of 0.78 (CI, 0.63-0.96; P = .02), and threat/control override delusions had an OR of 0.76 (CI, 0.61-0.95; P = .01).

In the multivariate analysis, the number of threat/control override delusions was negatively associated with violent recidivism even after the investigators controlled for hallucinations and the number of substance use disorders. The negative association for total delusions and threat/control override delusions for violent rearrest frequency also remained in the multivariate analysis even after hallucinations and the number of substance use disorders were controlled for.

Aside from the different psychoticlike symptoms and substance use disorders, the confounders in the multivariate analysis included age, origin, and parental/caretaker’s occupation.

"The absence of a positive relation between [psychoticlike symptoms] and violent recidivism may simply suggest that detained male adolescents with psychoticlike symptoms do not have a higher risk for future violent crimes than detained male adolescents without psychoticlike symptoms and, consequently, that findings from community samples are not generalizable to juvenile justice settings," the authors wrote. They suggested that the negative association between violent recidivism and delusions might result from fewer opportunities for interpersonal violence if delusional individuals are less inclined to engage in interpersonal interactions in the first place.

The authors noted that larger populations would be needed to assess alcohol, marijuana, and other substance use disorders as independent predictors. They also noted the study’s limitation of not accounting for the clinical significance, frequency, or intensity of any psychoticlike symptoms. The study also did not assess actual schizophrenia diagnoses; other symptoms of schizophrenia might independently increase the risk of violence.

The authors reported no relevant financial disclosures.

Next Article: