Cognitive profile of psychiatric patients with coexisting enlarged ventricles and cognitive impairment: A rare cross‐diagnostic phenotype predominantly observed in schizophrenia

While psychiatric diagnoses are conventionally symptom‐based, recent studies have aimed to integrate objective biological findings. A previous study classified brain biotypes based on subcortical volumes in 5602 individuals, including 2525 patients with major psychiatric disorders and 3077 healthy controls (HCs). ¹ Further analyses revealed that the biotype with severe symptoms could be stratified by cognitive impairment (CI) and enlarged ventricle (EV). ² This subgroup, termed EVCI because of the coexistence of EV and CI, was characterized by frequent abnormal electroencephalography findings, schizophrenia diagnoses, and rare copy number variations. The objective of this study was to stratify major psychiatric disorders into four groups based on the presence or absence of EV or CI, and to elucidate the roles and interactions of EV and CI in psychiatric disorders, with a particular focus on the EVCI.

Lateral ventricle volume (LVV), for the diagnosis of EV was measured using 3DT1 images, and the MRI protocol is described in the Supporting Information. LVV was analyzed using FreeSurfer. ¹ , ³ , ⁴ Scanner differences were harmonized according to the method proposed by Nemoto et al., ⁵ , ⁶ by using residuals derived from linear regression. For each scanner, regression coefficients were estimated with LVV as the dependent variable and age, sex, and intracranial volume as independent variables for covariates. ⁵ , ⁶ The residual of lateral ventricle volume for each participant was then calculated and used for the statistical analyses. EV was defined as LVV exceeding three standard deviations (SDs) above the HC mean. The mean values and SDs of these 1168 HCs were used to convert the LVV residuals into Z scores. CI was classified as moderate or severe when the difference between the current IQ measured by the Wechsler Intelligence Scales and the estimated premorbid IQ measured by the Japanese Adult Reading Test (JART) was 20 points or more. The data from nine patients with EVCI, 134 subjects without EV and with CI (nonEV‐CI), 31 subjects with EV and without CI (EV‐nonCI), and 1442 subjects without EV and CI (nonEV‐nonCI) were compared with each other. HCs were included in the initial stratification based on the presence or absence of EV and CI; nevertheless, none of the HCs fulfilled the criteria for the EVCI subgroup. This study was carried out in accordance with the World Medical Association's Declaration of Helsinki and approved by the institutional review board of the National Center of Neurology and Psychiatry (approval number: B2023‐080). Written informed consent was obtained from all participants. Demographic characteristics for each group are presented in Table S1, and the distribution of HCs and psychiatric diagnoses across groups is provided in Table S2. No significant age difference was observed among the four groups, whereas, years of education differed significantly between the nonEV‐CI and nonEV‐nonCI groups. Sex distribution also differed significantly among the groups. Cognitive profile was assessed across 18 cognitive measures, including CI value (current IQ – estimated premorbid IQ), estimated premorbid IQ, current IQ, Wechsler Adult Intelligence Scale–Third Edition (WAIS–III) indices, the Wechsler Memory Scale–Revised (WMS) indices, the Auditory Verbal Learning Test (AVLT), Verbal Fluency Task (VF), the Continuous Performance Test d Prime Mean (CPT), and the Wisconsin Card Sorting Test Categories Completed (WCST). A one‐way Welch's ANOVA was conducted because the assumption of homogeneity of variances was violated according to Levene's test, followed by the Games–Howell post hoc test. As this procedure was repeated for 18 cognitive measures, Bonferroni correction was applied to the significance level (α), which was adjusted to 0.05/18 to control for multiple comparisons. Table S3 presents the number of subjects analyzed for each cognitive measure after excluding missing data.

As shown in the Fig. 1a, cognitive dysfunction followed a graded pattern across the four strata (nonEV‐nonCI < EV‐nonCI < nonEV‐CI and EVCI). Among non‐CI participants, the EV effect (EV‐nonCI vs. nonEV‐nonCI) corresponded to a mean effect size of −0.60 across 18 cognitive measures (Fig. 1b). Within the EV stratum, adding CI (EVCI vs. EV‐nonCI) was associated with an additional mean effect size of −1.86 (Fig. 1b).

Fig. 1.

(a) Cognitive profile across four subgroups stratified by the enlarged ventricle (EV) and cognitive impairment (CI): nonEV‐nonCI, nonEV‐CI, EV‐nonCI, and EVCI. The y‐axis shows Z‐scores. Full names of the cognitive measures and their abbreviations are provided in the Supporting Information. (b) Mean standardized cognition (the average Z‐scores across the 18 cognitive measures) for each subgroup, with healthy control (HC) set to 0, nonEV‐nonCI to −0.16, EV‐nonCI to −0.75, and EVCI to −2.61. The difference in mean standardized cognition between EV‐nonCI and nonEV‐nonCI is −0.60, and that between EVCI and EV‐nonCI is −1.86.

This study revealed a detailed cognitive profile in EVCI. An association between EV and CI has been reported in patients with schizophrenia, ⁷ , ⁸ , ⁹ and recently, this association has also been reported in patients with first‐episode schizophrenia. ¹⁰ Although most previously reported EVCI cases were diagnosed with schizophrenia, only a small proportion of individuals with schizophrenia meet the criteria for EVCI. The most severe or extreme cases of schizophrenia may, in fact, correspond to EVCI. The large effect size observed in the EVCI group appears to reflect additive contributions from EV pathology and CI pathology, suggesting that overlapping mechanisms underlie severe cognitive dysfunction in EVCI. A limitation of this study is that the EVCI group included only nine individuals, making the results less generalizable. Replication in larger cohorts is warranted.

Funding Information

This work was supported by the Japan Agency for Medical Research and Development (AMED) under Grant Numbers JP18dm0307002 (RH), JP21dk0307103 (RH), JP21wm0425012 (JM and RH), JP22tm0424222 (RH), JP24dk0307132 (RH), JP25wm0625521 (JM and RH); JSPS KAKENHI Grant Numbers JP19H05467: Grant‐in‐Aid for Specially Promoted Research (RH), JP20H03611: Scientific Research B (JM, NH, and RH), JP20K07945 and JP23K07001: Scientific Research C (JM), JP23H00395: Scientific Research A (RH), and JP25K17172: Early‐Career Scientists (SI); and Intramural Research Grants for Neurological and Psychiatric Disorders of NCNP (Grants 3‐1, RH; 6‐1, RH; and 7‐5, JM and RH); Multilayered Stress Diseases (JPMXP1323015483), Science Tokyo.

Disclosure statement

The authors declare that they have no competing interests.

Supporting information

Data S1 Supporting Information.

Data availability statement

The data that support the findings of this study are available upon request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.