Criminogenic risk factors are proving more reliable than psychiatric symptoms at predicting whether individuals with mental illness will encounter the criminal justice system, a new study has found. Researchers compared three groups of justice-involved psychiatric patients in the United States to see what distinguishes those touching the legal system from those who have not.
Historically, intervention programmes aimed at reducing reoffending among people with mental illness focused on treating psychiatric symptoms alone. The latest findings underscore that such medical approaches often fell short of preventing jail time, arguing that many offenders with mental illness do not commit crimes because of hallucinatory or delusional experiences.
The study reiterates a well-established conclusion in criminology: non-clinical factors—collectively known as criminogenic risk factors—play a decisive role in criminal behaviour. The central eight factors include a history of antisocial behaviour, antisocial personality traits, antisocial friends, substance abuse, family or marital problems, school or work difficulties, lack of positive recreational activities, and criminal attitudes.
Across decades of research, these factors have consistently outperformed mental health variables in predicting criminal activity. In other words, mental illness by itself does not automatically translate into a higher risk of offending; some individuals with mental illness possess co-occurring antisocial tendencies that raise their chances of legal trouble.
The current investigation focused on a particularly nuanced subgroup: patients who have been found not guilty by reason of insanity. In such cases, a court has determined that a mental disease or defect diminished the person’s responsibility for the crime at the time it was committed. The question was whether these insanity acquittees show the same criminogenic risk profile as other offenders with mental illness.
To explore this, researchers Angelea Bolaños, Faith Scanlon, and colleagues compared three groups comprising 349 participants. The first consisted of 74 psychiatric patients with a history of misdemeanor or felony convictions. The second comprised 68 psychiatric patients with no criminal history, both groups drawn from a private inpatient hospital in Texas.
The third group included 207 forensic patients in the California state hospital system who had been found not guilty by reason of insanity and were undergoing compulsory psychiatric treatment. Participants completed self-report questionnaires measuring attitudes toward crime, social support, and current mental health, along with an inventory assessing psychiatric symptoms from the prior week.
The researchers then used statistical models to determine how well these measures could distinguish the three groups based solely on survey responses. They found that the group with mental illness and a criminal history scored highest on nearly all indicators of criminal risk, reporting more criminal associates and stronger antisocial attitudes, along with the lowest perceived social support from family and friends.
Insanity acquittees presented a more mixed picture. They showed higher general criminal tendencies than the no-history group but did not display the same elevated antisocial attitudes as the typical criminal group with mental illness.
In terms of social networks, the insanity acquittees resembled the no-criminal-history group, with similar levels of perceived social support. The researchers suggest that support from friends and family may be more robust when a crime is attributed to a mental health crisis rather than to ordinary criminal behaviour.
Remarkably, the insanity acquittees reported the lowest severity of current psychiatric distress among the three groups. The authors note this could reflect the contexts of their care: Texas patients were in short-term crisis facilities, while California patients were more likely to have experienced long-term treatment, potentially stabilising their symptoms.
Overall, the study found that criminogenic risk factors and social-support measures correctly classified roughly two-thirds to three-quarters of participants into their respective groups. The findings reinforce the view that addressing criminal attitudes and social environments is essential in treatment for justice-involved individuals with mental illness, alongside standard psychiatric care.
The authors advocate for dual-focused rehabilitation programmes that integrate mental health care with targeted efforts to reduce criminal risk factors, helping forensic patients acquire life skills needed for recovery and eventual release. The study argues for a holistic approach that treats the whole person rather than viewing their situation solely through a psychiatric lens.
However, the research design is cross-sectional, capturing data at one point in time, which limits conclusions about how mental illness and criminal behaviour evolve. Self-report questionnaires may also be subject to response bias, and corroboration with official records or clinician observations would strengthen future work.
Additionally, the sample has limitations: the Texas group was predominantly White and male, raising questions about broader applicability. The California sample’s high rate of prior convictions further blurs the boundary between psychiatric and criminal profiles. The researchers recommend separating insanity acquittees by past criminal histories in future studies to better understand this distinct population.
The study, titled Psychiatric Symptoms and Criminogenic Risk in People With Mental Illness: Comparing Patients Across Forensic and Nonforensic Settings, was authored by Angelea D. Bolaños, Faith Scanlon, Robert D. Morgan, Sean M. Mitchell, and Darci Delgado.
