No Evidence of Cognitive Subtypes in People at Clinical High Risk for Psychosis

There is no evidence of distinct cognitive subtypes among individuals at clinical high risk for psychosis (CHR-P), according to study findings published in the Schizophrenia Bulletin.

Cognitive performance has been associated with social and occupational function in CHR-P and can be used to inform clinical care. However, it remains unclear as to whether cognition in CHR-P exhibits a cluster structure and whether individuals in these cognitive clusters display consistent patterns of clinical outcomes.

To address this uncertainty, investigators conducted an observational study using data from the multicenter EUropean network of national schizophrenia networks studying Gene-Environment Interactions (EU-GEI) high-risk study. The investigators used the Comprehensive Assessment of At-Risk Mental States (CAARMS) criteria to define CHR-P status and individuals who did not meet CAARMs criteria were included as healthy controls. Participants were excluded if they had a history of drug/alcohol dependency, neurological disorder, psychotic disorder that would explain relevant CHR-P symptoms, or an intelligence quotient (IQ) estimate below 60. The investigators used k-means clustering and density-based spatial clustering to identify clusters within the cognition data. The cognitive subtypes identified were then validated using participant data on functioning, psychosis symptoms, psychosis transition status, and grey matter volume.

The investigators included 291 CHR-P participants and 53 healthy controls in the analysis and the mean duration of follow-up for CHR-P was 643 days. The CHR-P participants and the healthy control group were balanced for age and sex. The CHR-P group had significantly lower functioning scores, estimated IQ, and years of education.

Using k-means clustering, the investigators found a 2-cluster solution to be optimal, largely replicating previous reports of “spared” and “impaired” cognitive subtypes. However, the clusters were poorly formed given the low silhouette score (0.16) and there were no differences between the subtypes in terms of functioning, rate of transition to psychosis, negative symptoms, or basic symptoms.

Additionally, the results of regression models indicated that cognitive subtypes were not significantly associated with disability scores (F_{16, 206}=1.34; P =.174; R²=0.02) or transition to psychosis (χ²₁₆=19.50; P =.243). Further, the spared and impaired cognition clusters were not associated with differences in grey matter volume (P <.05).

“Stratifying CHR-P patients in terms of cognitive function could facilitate a more personalized approach to clinical care,” the investigators noted. However, “[W]e suggest care needs to be taken in inferring the existence of distinct cognitive sub­types within patient groups from such studies.”

Study limitations include bias toward participants with higher IQ among those with follow-up data, and bias due to a substantial proportion of the original sample being removed.

Disclosure: Some study authors declared affiliations with biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of authors’ disclosures.

This article originally appeared on Psychiatry Advisor