Analysis of early psychological symptoms sharpens diagnosis of teens at high risk for a brain disorder that affects millions worldwide.
Schizophrenia remains one of the most challenging mental illnesses to detect in its earliest phase — the period when subtle symptoms first appear but full disorder has not yet developed. A new analysis led by researchers at the UNC School of Medicine and the Renaissance Computing Institute (RENCI) sheds light on which early symptoms most reliably predict later development of psychosis and schizophrenia. The study, published in Schizophrenia Research, finds that disturbed thought content and suspiciousness are the strongest early indicators, while perceptual disturbances that resemble mild hallucinations are less predictive than previously believed.
Early identification is critical: the sooner at-risk individuals receive appropriate care, the better their long-term outcomes. Diana Perkins, MD, MPH, a clinician and psychiatry professor at UNC and a first author of the study, emphasizes that improving early detection could enable preventive interventions that reduce the likelihood of progression to full-blown schizophrenia and limit associated functional decline.
Schizophrenia is a chronic mental disorder that typically appears in late adolescence or early adulthood. It affects millions worldwide and can cause persistent disability for many people. Psychosis — which includes paranoia, delusions, hallucinations, and disorganized thought and behavior — is a core feature of schizophrenia but also occurs in other conditions such as bipolar disorder. Many individuals experience mild psychosis-like symptoms before any formal diagnosis, yet only a minority of those people (roughly 15–20 percent) go on to develop a psychotic disorder within a few years.
Current criteria for identifying attenuated psychosis focus on three symptom categories: illogical or disorganized thinking, and perceptual disturbances of a severity that impairs functioning. To refine these criteria, Perkins and co-author Clark Jeffries, PhD, a scientist at RENCI, analyzed symptom profiles from a cohort of 296 individuals who met clinical high-risk criteria for psychosis. Each participant was followed for two years to determine which symptoms best predicted conversion to psychosis.
The analysis identified two core symptoms — suspiciousness and unusual thought content — as the most predictive of later psychosis. Adding two additional symptoms improved prediction: reduced ideational richness (difficulty generating and connecting ideas) and trouble with focus or concentration. These four symptoms formed a brief risk scale that reliably distinguished individuals who later developed psychosis from those who did not.
To ensure the findings were robust, the researchers applied rigorous validation methods. They used randomized permutation tests and then tested the symptom-based classifier in an independent validation cohort of 592 individuals at clinical high risk. The same pattern held: unusual thought content and suspiciousness were consistently linked to higher risk, and the four-item subscale (including reduced ideational richness and concentration problems) offered stronger discrimination than a two-item scale.
In practical terms, suspiciousness and unusual thought content often present as a persistent sense of being watched, worries that others are talking about the person despite knowing this is unlikely, over-interpretation of coincidences, seeing meaning or “signs” where none exist, or experiencing distorted timing of events. Difficulty with focus or concentration appears as distractibility and short-term memory problems, while reduced ideational richness shows up as trouble following conversations or thinking abstractly.
Interestingly, perceptual disturbances — such as seeing fleeting shadows or hearing indistinct noises that feel “not real” — were not uniquely predictive of later psychosis. Although such experiences were common among those who later developed psychosis, they were equally common among those who did not, making them poor discriminators for risk assessment.
Perkins notes that these findings should shift clinical focus toward assessing thought processes rather than relying primarily on perceptual complaints. “For clinicians and researchers working on early detection, this study suggests we should prioritize measures of unusual thought content and suspiciousness, and consider attention and ideational richness when estimating risk,” she said. Refining diagnostic tools in this way could improve accuracy and guide the development of targeted interventions for adolescents and young adults at elevated risk.
The research team included investigators from multiple institutions involved in the North American Prodrome Longitudinal Study, including specialists from Zucker Hillside Hospital, Yale University, the University of Calgary, UCLA, UC San Diego, the National Institute of Mental Health, UCSF, and Harvard Medical School. The study was funded by the National Institute of Mental Health.
Funding: This research was supported by the National Institute of Mental Health.
Original study: Abstract for “Severity of thought disorder predicts psychosis in persons at clinical high-risk” by Diana O. Perkins, Clark D. Jeffries, Barbara A. Cornblatt, Scott W. Woods, Jean Addington, Carrie E. Bearden, Kristin S. Cadenhead, Tyrone D. Cannon, Robert Heinssen, Daniel H. Mathalon, Larry J. Seidman, Ming T. Tsuang, Elaine F. Walker, and Thomas H. McGlashan, published in Schizophrenia Research.
Abstract
Severity of thought disorder predicts psychosis in persons at clinical high-risk
Background
Enhancing predictive accuracy for early psychosis is essential for prevention and early intervention. This study aimed to develop a symptom-based severity classifier to improve prediction of psychosis conversion in at-risk individuals.
Methods
Data came from two cohorts in the North American Prodrome Longitudinal Study. All participants met criteria for psychosis-risk states. In Cohort 1 (n = 296), the investigators identified items from the Scale of Psychosis-Risk Symptoms that best distinguished converters from nonconverters and validated initial performance with randomization tests. Cohort 2 (n = 592) served as an independent test set.
Results
Two- and four-item subscales were derived. Both included unusual thought content and suspiciousness; the four-item subscale added reduced ideational richness and difficulties with focus/concentration. Across cohorts, the four-item subscale showed stronger discrimination. Calibration indicated proportional differences between cohorts related to their differing two-year conversion rates.
Conclusion
Severity of unusual thought content, suspiciousness, reduced ideational richness, and impaired concentration meaningfully informed psychosis risk prediction. Scales based on these symptoms may be useful in research and, with further validation, in clinical practice.