A Data-Driven Psychiatric Disorder Subtype Defined by Significant Enlarged Ventricles and Cognitive Impairment: A Replication Study.

Abstract

We previously performed data-driven classification based on large-scale neuroimaging datasets to identify a potential novel subtype of psychiatric disorders characterized by the pronounced enlarged ventricles and cognitive impairment (EVCI) in one cohort. However, these findings have yet to be validated across independent cohorts. Herein, we investigated the availability of cognitive data in other cohorts in the Cognitive Genetics Collaborative Research Organization (COCORO) project to assess the presence of cognitive impairment (CI). Subsequently, we compared the prevalence of CI, lateral ventricular volume, and demographic and clinical characteristics between individuals with EVCI and those with EV but without CI (EV-nonCI). Cognitive data were available for 27 individuals with EVCI from the other 8 cohorts. Among the 27 individuals in the EV group, 5 (18.5%) exhibited EVCI, all of whom were diagnosed with schizophrenia. Three individuals (60%) in the EVCI group were male. Compared with the EV-non-CI group, the EVCI group presented significantly lower current intellectual quotient (IQ) scores (p = 5.8 × 10^-4), but there was no significant difference in premorbid IQ (p = 0.56). Furthermore, there was no significant difference in the Z scores for lateral ventricular volume between the groups (p = 0.26). This study validated the existence of the EVCI subgroup in independent cohorts, thus supporting our previous findings.