Schizophrenia is characterized by substantial clinical and biological heterogeneity, where negative symptoms are a core component that mainly account for long-term disability and poor functional outcomes in patients with the disorder (1, 2). The deficit form of schizophrenia (D-Sz), a well-defined clinical subgroup independent of the DSM (3) or ICD (4) subtype classification, is characterized by primary (i.e., not secondary to other factors, such as positive symptoms, comorbid depression, and extrapyramidal side effects) and enduring negative symptoms (5–7). The DSM/ICD subtypes of schizophrenia based on symptom profiles (e.g., paranoid, disorganized, and undifferentiated) have been eliminated because the subtype classification changes with time and cannot estimate their outcomes (8). On the other hand, the D-Sz defined by established classification tools {i.e., the Schedule for the Deficit Syndrome (9) or the Proxy for the Deficit Syndrome [PDS; (10)]} has a subtype stability over time and rather homogeneous outcome (7). While the etiological factors associated with D-Sz have not yet been identified, neuroimaging evidence of enhanced interregional cortical coupling (11) and altered gross brain morphology (e.g., gyrification patterns) (12, 13) specifically in D-Sz appear to support its prominent neurodevelopmental pathology. Therefore, brain morphological characteristics associated with early neurodevelopmental abnormalities, which likely exist at illness onset (14), may be a prognostic biomarker of worse long-term functioning in schizophrenia.

Recent magnetic resonance imaging (MRI) studies on schizophrenia revealed an increased prevalence of a duplicated Heschl’s gyrus (HG) pattern (15–17), which may reflect the development of an anomalous cytoarchitecture in utero (18, 19). This alteration in the gross brain morphology has been detected irrespective of illness stages [i.e., high-risk status (17), both first-episode (15) and chronic (16) stages] and the medication status (15–17) of schizophrenia. Furthermore, these studies suggested that the duplicated HG pattern was associated with severe prodromal symptomatology (17), but rather mild positive psychotic symptoms after onset (15, 16), as well as prominent verbal fluency deficit (17). These findings likely support the possibility that the HG gyrification pattern may contribute to specific clinical syndrome in schizophrenia. In addition, it has been suggested that the duplicated pattern of the HG, which participates in emotional processing (20), is associated with regional brain hypoactivity (21). It may be thus hypothesized that the patients with D-Sz, who are characterized by persistent negative symptoms, have an increased prevalence of HG duplication, but no MRI studies to date have specifically examined the HG duplication patterns in D-Sz.

Therefore, we herein used MRI to investigate and compare duplicated HG patterns in the D-Sz and non-deficit subtype of schizophrenia (ND-Sz) with those in matched healthy controls. We had previously explored brain characteristics of our D-Sz cohort and found gross morphological changes associated with early neurodevelopment [e.g., small adhesio interthalamica, altered distribution of the orbitofrontal sulcogyral pattern; (12)] and gray matter reduction in the insular cortex (unpublished data). The present study aimed to further expand these studies to HG duplication pattern, a recently identified early neurodevelopmental marker. Based on previous findings showing HG duplication in schizophrenia as a stable neurodevelopmental marker (15–17) and a putative prominent neurodevelopmental pathology in its deficit subtype (11–13), we predicted that the prevalence of HG duplication may be higher in patients with schizophrenia, particularly those with D-Sz. Furthermore, we investigated the relationships between the HG pattern and the clinical characteristics of patients in the different subgroups.

While its preliminary nature with a small number of subjects, we demonstrated for the first time that prevalence of right HG duplication was higher in patients with D-Sz than in those with ND-Sz. This difference was not explained by demographic or clinical differences in these subgroups, except for the PDS score, which reflects the trait characteristics of deficit syndrome. These results suggest that prominent early neurodevelopmental abnormalities associated with sulcal formation during gestation may contribute to the characteristic clinical manifestations of D-Sz, such as enduring negative symptomatology and poor functional outcomes.

The present study demonstrated that the prevalence of the right HG duplication patterns was significantly higher in patients with schizophrenia, particularly those with the deficit subtype, than in healthy controls, while the HG duplication itself has been reported in approximately 30–50[percentage] of healthy adults (28–30). The functional significance of this anatomical variation currently remains unclear, but may reflect inter-individual differences in the cytoarchitectonic development of the primary auditory cortex in utero (18, 19); duplicated HG has been implicated in the development of learning disabilities (31, 32) and the inhibition of HG activity in auditory processes (21) after birth, even in non-clinical populations. HG is a primary brain region of auditory processing, but also participates in emotional processing (20) and social communication (33). Therefore, the core clinical features of D-Sz, such as persistent blunted affect and prominent cognitive deficits, particularly in social and verbal domains (7), may be partly attributed to neurodevelopmental abnormalities associated with HG sulcus formation. Indeed, our previous study on the early stages of psychosis (17) revealed a correlation between the HG duplication pattern and verbal fluency deficits, although it was evident between different duplication patterns (i.e., CSD vs. CPD) in the left hemisphere. Since brain gyrification may reflect regional neural connectivity (34), our hypothesis needs to be tested in a cohort with more detailed clinical assessments using the multimodal neuroimaging of brain connectivity/function. Potential clinical significance of different HG duplication patterns (CSD, CPD) should be also tested in such future studies.

The present results on HG duplication patterns were not influenced by illness chronicity or medication, supporting previous neuroimaging findings suggesting a prominent early neurodevelopmental pathology in D-Sz [e.g., small adhesio interthalamica (12) and gyrification pattern changes (12, 13)]. We failed to replicate our previous findings in larger and less confounded (i.e., first-episode) schizophrenia groups to show that the right CPD pattern specifically contributed to less severe positive symptomatology (15, 16); however, the right CPD pattern in the present cohort correlated with the trait characteristics of deficit syndrome (i.e., PDS score). In addition, the right lateralized group difference in the HG pattern in the present study may support right hemisphere dominance for emotion processing, particularly for negative emotional information (35), as well as D-Sz having severe and prolonged neurodevelopmental abnormalities. Right HG generally develops 1–2 weeks earlier than left HG during mid-to-late gestation (18) and is more complex (18, 29); therefore, gyral formation of HG in D-Sz may be more affected on the right hemisphere. Since a meta-analysis of gray and white matter volumes across various brain regions found no significant differences between D- and ND-Sz (36) and these volumetric data are affected by various confounding factors (e.g., illness chronicity and medication), a better predictive biomarker of the clinical subtype and course of schizophrenia may be gross brain morphology, which is strongly related to early neurodevelopment.

Several potential confounding factors in the present study need to be addressed. First, the present study was partly limited by its small sample size for each schizophrenia subgroup. Due to its reliability and stability, the PDS score has been widely used to classify patients into the D and ND subgroups in biological (11, 37–39) and clinical (40–42) studies (10, 25, 41, 43); however, a cross-sectional PDS rating does not have the capacity to directly assess enduring deficit-like features. Therefore, we excluded the ambiguous middle group for the PDS score (−8 to −3) from the current schizophrenia sample in an attempt to reduce potential false classifications, as in previous imaging studies (11–13), which further decreased the number of schizophrenic patients examined. Therefore, the present results need to be confirmed in a well-defined cohort of a large number of patients with D-Sz with clinical follow-up data and/or a semi-structured interview [e.g., the Schedule for the Deficit Syndrome (9)]. Second, the sex ratio significantly differed between the D- and ND-Sz subgroups in the present group, potentially reflecting the general characteristics of D-Sz [more males than females (27)]. However, this difference did not appear to significantly affect the results obtained because there was no significant sex effect on HG patterns. Third, we did not systematically assess the cognitive function of our cohort. Since HG duplication patterns have been implicated in prominent cognitive impairments in D-Sz, as described above, future studies on the gyrification–cognition relationship may provide a more detailed understanding of the pathophysiology of this specific schizophrenia subtype. Finally, although the present HG pattern classification by manual delineation on 1.5T MRI data was reliable (Cronbach’s α > 0.83), our results should be replicated using unbiased automatic analysis on higher resolution images.

In summary, this preliminary MRI study demonstrated that the prevalence of HG duplication was higher in patients with schizophrenia, particularly those with the deficit syndrome subtype who typically exhibit enduring negative symptomatology and poor functional outcomes. Although the HG duplication itself is observed in healthy subjects and there is considerable overlap on the HG pattern distribution between D- and ND-Sz subgroups, it is possible that, in combination with other brain characteristics, the HG pattern in the early stages of schizophrenia may have potential as a prognostic biomarker of worse long-term functioning.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Committee on Medical Ethics of the University of Toyama. The patients/participants provided their written informed consent to participate in this study.

MS and TT conceived the idea and methodology of this study. TT conducted the statistical analyses and wrote the manuscript. DS and HK recruited the participants and were involved in clinical and diagnostic assessments. TT and DS analyzed the MRI data. KN provided the technical support for MRI scanning and data processing. AF managed the MRI and clinical data. MS and YT contributed to the writing and editing of the manuscript. All authors contributed to and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.