Schizophrenia is a heterogeneous mental disorder characterized by exacerbating and remitting positive symptoms, coexisting with the so-called negative symptoms, which are relatively unremitting and resistant to drug treatment, and general symptoms and cognitive deficits, usually chronic (1, 2). Although challenging for clinical and research settings, investigating hostility and aggression in schizophrenia is difficult due to multi-factorial correlations and inconsistencies in the definition and assessment of these dimensions (3). Impaired disease awareness and impaired perception of the need for treatment are standard features of these patients’ clinical presentation.

Hostility implies the recurrent and persistent tendency to feel anger toward and a wish to injure a person or group. It can be related to a negative cognitive bias of strong dissatisfaction toward others with thoughts and feelings of antagonism, resentment, and alienation (3, 4) with a predisposition to attack one or more people in a connection. Patients with schizophrenia can show hostility symptoms both during the acute episodes and in stable phases of the disorder (3, 5). In some cases, hostility can lead to verbal or physical aggressiveness, and extensive population studies have shown a significant association between schizophrenia and violent behaviors (6–8). Furthermore, in patients with stable schizophrenia, excitement symptoms and prefrontal dysfunction correlated with elevated aggressiveness (9). These aspects are generally associated with poor compliance with treatment and involuntary hospitalization (10), which occurs more frequently in patients with schizophrenia than in patients with other mental disorders (11). Hostility-related dimensions in schizophrenia were related to structural and functional changes in different brain areas, including the prefrontal, anterior cingulate, and insular cortices, and amygdala, striatum, and hippocampus. These changes were involved in modulating hostile thoughts and behaviors, mainly in patients affected by schizophrenia with high urgency, impulsivity, and aggressiveness (3).

In recent decades, an increasing interest in identifying different aspects of mental and neurodevelopmental disorders through reliable magnetic resonance imaging (MRI) correlates has been observed, including the possible definition of their clinical implications (12). Gyrification is the result of the fetal and infantile developmental processes of gyri and sulci folding that transform the anatomy of the cortex from a smooth (“lissencephalic”) structure to a highly convoluted surface (13–15). The gyrification index (GI) is the ratio between the total contour of the pial surface and the contour of the smooth external surface (16–18) and increases with the complexity of the folds. It was initially calculated on two-dimensional coronal sections (16), but newer three-dimensional computations were developed and allowed for greater precision (19), representing an improvement over older methods (20).

The Local GI (LGI) is a measure based on voxel-based morphometry (VBM) techniques. LGI measures are currently widely accepted and involved in psychiatric studies, especially early psychotic episodes, depression, intellectual disabilities, 22q11 deletion, and schizophrenia (21, 22). Different studies demonstrated increased gyrification patterns (hypergyria) in individuals with autism (23–26), reduced gyrification in psychosis risk syndrome (27), and variable changes in gyrification patterns during the course of schizophrenia (16, 28, 29), which lead to the hypothesis that this is a morphological trace of early developmental anomalies that eventually persist throughout adulthood.

Though research has dedicated to the study of gyrification in the context of first-episode schizophrenia, chronic schizophrenia, siblings and at-risk individuals, findings are conflicting, and still need to clarify whether LGI can correlate with different aspects of the disease as well. To summarize the results of the LGI studies, it is thought that a characteristic of cortical gyrification in schizophrenia is a decreased GI and hypogyria. Abnormal cortical gyrification is also found in siblings of patients with schizophrenia and “high-risk” (HR) individuals who have family members with schizophrenia (21).

Gyrification is a potent marker of early neurodevelopment. It can be hypothesized that most of the gyrification process occurs during the third trimester of pregnancy, a period of considerable brain growth (30), and its pattern remains stable after birth (18, 31). Therefore, gyrification is a significant indicator of neurodevelopment, which proceeds according to an ordered topographic model (32–34).

Considering the neurodevelopmental hypothesis of schizophrenia (2, 35–37), early hypergyria would fit well with the abnormal mechanical developmental patterns of the brain. Furthermore, aberrant cortical curvature could be associated with reduced cortical thickness due to centripetal forces generated by cortico-cortical connections in cerebral maturation (13). Since the brain is believed to operate through intricate networks of neuronal connections, it has been hypothesized that the destruction of connectivity within a network or between two or more neural networks can be a major biological correlate of schizophrenia (38–41).

Differences in gray matter neuroanatomy and white matter connectivity are closely related. They may have common etiological pathways (18, 42): therefore, alterations of gyrification pattern could play a role in cognitive and psychiatric outcomes (43) and schizophrenia onset (29). There is a direct relationship between neural migration disorders (such as lissencephaly) and aberrant neural connectivity (44).

Hostility in schizophrenia is a critical clinical aspect, and there is a need for further neuroimaging studies on these dimensions. In this sense, hostility and/or aggressive behavior can manifest both as a psychopathological marker (lack of insight, medication non-adherence) and/or as a consequence of the disease (due to substance abuse and/or violence abuse) (45). Research still needs to clarify whether LGI can be a marker of hostility as a comorbid behavioral change in schizophrenia.

This study hypothesizes that patients affected by schizophrenia might show abnormal gyrification of specific cortical brain areas correlated with the symptom severity. Furthermore, LGI might show differences in patients with schizophrenia and high hostility symptoms (HHS) (in terms of BPRS hostility scores and history of compulsory treatment) when compared to patients with low hostility symptoms (LHS). We compared subjects dividing them into the two groups assessing them on clinical and neuroradiological bases, i.e., the LGI.

This is an observational, retrospective, non-interventional, single-center study focusing on the neuropathological correlates of LGI in patients with schizophrenia spectrum disorders, dividing them into HHS and LHS. The local ethical committee approved this study (prot. N. 470/2012). All subjects gave their informed consent before their inclusion in the study, which followed the generally accepted ethical research standards of the Declaration of Helsinki (46).

Patients admitted to the Psychiatric Diagnosis and Treatment Service of the “Sant’Andrea” University Hospital of Rome were included with the following criteria: diagnosis of non-affective psychosis at onset or schizophrenia, age range 18–65 years, and voluntary adherence to the research protocol. The treating psychiatrists established diagnoses according to the DSM-IV-TR criteria (47) and DSM-5 (48), which we confirmed according to DSM-5 criteria through medical record examination. Exclusion criteria were as follows: inability to undergo MRI examination, any neurological disease involving the central nervous system, history of traumatic brain injuries, radiological signs of neurological and neurovascular disease, untreated systemic illnesses, substance use disorder, and pregnancy.

Our correlation analyses of the whole sample showed that the severity of symptoms significantly directly correlated with the LGI in different cortical areas, including bilateral frontal, cingulate, parietal, temporal cortices, and right occipital cortex. This is partly in line with a recent study that showed a symptom-LGI direct correlation in patients with a first episode of schizophrenia in the left temporal cortex (57). The involvement of more cortical brain areas in our study could be due to the differences in the sample compositions. The picture of a diffuse hypergyria associated with schizophrenia, regardless of the correlation with the severity of symptoms, is reported by several studies, and our results are to be considered in the context of this evidence, mainly during the illness onset (58–60).

Considering the lack of MRI correlation studies on more advanced stages of schizophrenia, an issue that needs more investigation relates to whether the correlation between LGI and symptoms severity is a constant aspect of the disease or whether it is more typical of the onset and early course of schizophrenia (29). Further longitudinal, case-control, and symptom correlation neuroimaging studies are needed. These may shed light on the nature and clinical correlates of gyrification changes in schizophrenia, also considering that future studies could take advantage of higher magnetic induction fields (≥3 T), which could reinforce existing evidence or lead to more accurate findings.

We found that LHS showed significantly lower LGI related to the severity of symptoms in the bilateral frontal lobe, including the right pars triangularis, opercularis, and orbitofrontal cortex, and left precentral, postcentral, and caudal middle frontal cortices. Furthermore, we observed the same reduction in LHS in bilateral temporal areas, including the bilateral insula, bilateral superior temporal cortices, left transverse and middle temporal cortices.

The insular cortex plays an essential role in the consciousness processes of one’s body and its properties (61–64), emotions (65), empathy (66), sense of agency (67), and language (68, 69). The insula is thought to be involved in the violation of rules and social conventions as well (70), together with processing convergent information to produce a relevant emotional context in response to sensory experience and in a general salience system that monitors the surrounding environment and selects appropriate responses and behaviors (71).

Other studies have reported an association between destructive, aggressive, and antisocial behaviors and increased gyrification of the insula and superior temporal gyrus (72, 73). These clinical factors have been associated with involuntary psychiatric hospitalization (74–76). Dysfunction of neural connectivity in the left insula has been hypothesized to represent neural correlates of clinical insight impairment in schizophrenia spectrum disorders (77). Our results can be linked with existing evidence of insular structural and functional changes mirroring poor illness insight in patients with schizophrenia (78, 79).

Temporal regions also play an essential role in different dimensions of insight (80), and insight deficits have been previously associated with structural alterations of the temporal cortex (81, 82). Another important study showed that gyrification patterns in schizophrenia spectrum disorders in frontal and temporal areas are related both to early neurodevelopmental abnormalities (vulnerability) and active brain pathology, especially in the early stages of illness (83).

We showed gyrification changes in the right orbitofrontal cortex in patients affected by schizophrenia with hostility/aggressive symptoms. This area has been involved in the development of severe negative symptoms (84), deficits in abstract thinking (85), and could be interpreted in the context of the evidence of a structural deficit in the corticothalamic systems, especially in the orbitofrontal-thalamic system in schizophrenia (86). We also discovered a lower LGI/symptom correlation of the right pars triangularis in the LHS vs. HHS group, therefore confirming the existing evidence of structural and functional deficits in this area in patients with schizophrenia and states of chronic illness with a lack of medication (87), cognitive deterioration (88), lack of cognitive insight (89), and symptom misattribution (90). Furthermore, a significantly impaired brain activity of the right pars triangularis in response to auditory words presented with negative valence has been crucially involved in the neuropathophysiology of positive symptoms, including the persecutory delusion and delusional behavior (91). These findings are in line with our results, suggesting an involvement of the right pars triangularis in the neuropathophysiology of severe symptoms of schizophrenia.

HHS patients showed significantly lower dosages of antipsychotics on admission than their counterparts. This result might also be interpreted in the light of poor adherence to treatments before hospitalization, i.e., one of the most important factors that usually lead to compulsory hospitalization (87). This is in line with other studies highlighting the risk of aggression toward others (74, 75) and psychomotor agitation (76) as factors strongly associated with involuntary hospitalization.

This study showed a correlation between the severity of psychopathology and cortical gyrification in schizophrenia, both in LHS and HHS. One explanation for this finding is that a higher level of gyrification correlates with neurodevelopmental impairment, which predisposes patients to manifest more severe psychotic episodes during schizophrenia. Poor insight and aspects that may arise from it, including involuntary hospitalization, treatment non-adherence, lower psychosocial functioning, poor prognosis, and higher utilization of emergency services, are all clinical issues hypothesized to have neurobiological bases (92). Current evidence identified specific neural correlates for insight types and dimensions, such as anatomical and functional changes in the prefrontal cortex, cingulate cortex, and regions of the temporal and parietal lobe (precuneus, inferior parietal lobule), and hippocampus (80). Another point of reflection for future research is provided by the possible association between gyrification and aggressiveness in schizophrenia. An element that could be analyzed would be gender differences in gyrification, considering that male patients with schizophrenia could manifest aggressiveness more often than female counterparts (93). Greater severity of psychopathology, also related to LGI changes, can predispose patients to manifest symptoms that are risk factors for compulsory treatment, for example, acute worsening, aggressive behaviors, social maladjustment, poor therapeutic compliance, and prolonged hospitalizations (94). The results of this study, even if of an exclusively biological nature, may pave the way for further studies and the acquisition of big data, which may in the future have a clinical implication using clinical applications, which may help to guide the diagnostic and treatment pathways in patients with acute schizophrenia (95). Other important clinical perspectives regard the possibility to overcome to LGI impairment through the restoration of brain activity via brain stimulation. Initial evidence is now available regarding the possibility to modulate aggression/impulsivity features in several psychiatric disorder thanks to Brain stimulation techniques (96–98).

Acute inpatients with schizophrenia showed increased LGI correlated to symptom severity in bilateral frontal, cingulate, parietal, temporal cortices, and right occipital cortex. The comparison between LHS and HHS showed in the first group a significantly lower LGI related to the severity of symptoms in the bilateral frontal and temporal cortices.

These changes in LGI involve abnormalities in areas that have been already associated with a lack of insight and awareness of the illness, severity of psychotic symptoms, and aggressive and antisocial behavior, which are frequently related to involuntary hospitalization. Consequently, the LGI differences found between HHS and LHS could be mediated by the severity of psychopathology. Gyrification changes can be a measurable correlate, possibly linking neurodevelopment alterations to specific clinical dimensions in adults who have schizophrenia. The modifications in LGI we found in HHS might represent a new possible biological component for severe clinical behavioral disturbance. Further longitudinal studies are needed to better understand the nature of these changes and their possible correlation in the context of severe schizophrenia.

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 Sapienza University Ethics Committee prot. N. 470/2012. The patients provided their written informed consent to participate in this study.

SF and ADC ideated and designed the study. ADC, AnB, EP, GM, GP, IG, LM, PG, and TZ performed the clinical data analyses. AR and MR-E performed the MRI acquisition and data collection. ML, AN, ADC, and MR-E conducted the statistical analyses. SF, ADC, IG, MNM, and GDK provided the first draft. ADC, GDK, SF, MP, and AlB supervised the writing of the manuscript. All authors contributed to the manuscript writing and approved the final draft.