Voxel-based alterations in spontaneous brain activity among very-late-onset schizophrenia-like psychosis: A preliminary resting-state functional magnetic resonance imaging study

Abstract

BACKGROUND

Very late-onset schizophrenia-like psychosis (VLOSLP) is a subtype of schizophrenia spectrum disorders in which individuals experience psychotic symptoms for the first time after the age of 60. The incidence of VLOSLP shows a linear relationship with increasing age. However, no studies have reported alterations in spontaneous brain activity among VLOSLP patients and their correlation with cognitive function and clinical symptoms.

AIM

To explore VLOSLP brain activity and correlations with cognitive function and clinical symptoms using resting-state functional magnetic resonance imaging.

METHODS

This study included 33 VLOSLP patients and 34 healthy controls. The cognitive assessment utilized the Mini Mental State Examination, Montreal Cognitive Assessment, and the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). Clinical characteristic acquisition was performed via the Positive and Negative Syndrome Scale (PANSS). All participants were scanned via resting-state functional magnetic resonance imaging, and the data were processed using amplitude of low-frequency fluctuations (ALFF), fractional ALFF (fALFF), regional homogeneity, and voxel-mirrored homotopic connectivity (VMHC).

RESULTS

The VLOSLP group presented decreased ALFF values in the left cuneus, right precuneus, right precentral gyrus, and left paracentral lobule; increased fALFF values in the left caudate nucleus; decreased fALFF values in the right calcarine fissure and surrounding cortex (CAL) and right precuneus; increased regional homogeneity values in the right putamen; and decreased VMHC values in the bilateral CAL, bilateral superior temporal gyrus, and bilateral cuneus. In the VLOSLP group, ALFF values in the right precuneus were negatively correlated with Mini Mental State Examination score and PANSS positive subscale score, and VMHC values in the bilateral CAL were negatively correlated with the RBANS total score, RBANS delayed memory score, and PANSS positive subscale score.

CONCLUSION

The changes of brain activity in VLOSLP are concentrated in the right precuneus and bilateral CAL regions, which may be associated with cognitive impairment and clinically positive symptoms.

Key Words: Very-late-onset schizophrenia-like psychosis; Schizophrenia; Resting-state functional magnetic resonance imaging; Amplitude of low-frequency fluctuations; Fractional amplitude of low-frequency fluctuations; Regional homogeneity; Voxel-mirrored homotopic connectivity

Core Tip: The incidence of very-late-onset schizophrenia-like psychosis (VLOSLP) tends to increase linearly with age, and most of these patients exhibit prominent positive symptoms as well as cognitive impairments. This study explored the changes in spontaneous brain activity in VLOSLP patients and their correlations with cognitive functions and clinical symptoms by resting-state functional magnetic resonance imaging. The results indicated that the alterations in the brain activity indices of VLOSLP patients were mainly concentrated in specific brain regions, and that the changes in these regions may lead to cognitive impairment and the emergence of positive symptoms in VLOSLP patients.

INTRODUCTION

The onset of schizophrenia spectrum disorders is most commonly in late adolescence or adulthood, known as typical-onset schizophrenia (TOS), but more than 20 percent of patients experience their first psychotic symptoms in old age, without any prior emotional or neurological disorders[1]. Based on differences in the age of onset of schizophrenia, an international consensus has identified patients who experience psychotic symptoms for the first time after the age of 60 years as very-late-onset schizophrenia-like psychosis (VLOSLP) patients[2]. Unlike the different psychiatric symptoms of TOS, VLOSLP is distinguished by its prominent positive symptoms such as delusions and sensory hallucinations across multiple modalities, without the presence of formal thought disorders or negative symptoms[3-5]. The prevalence of VLOSLP in the community is also relatively lower than that of TOS, only ranging between 0.1% and 0.5%[5]. However, there is a linear relationship between the incidence of VLOSLP and increasing age, with the annual incidence increasing by 11 percent for every 5-year increase in age[4-6], which is higher in females than in males[2]. The mortality rate of VLOSLP patients is significantly higher than that of elderly TOS patients and the general population because of comorbidities and accidents, requiring more psychiatric hospitalization treatment[7]. Given that the elderly population is the fastest-growing demographic population nationally and that previous research on TOS may not always be applicable to VLOSLP[5], there is an urgent need for psychiatrists to explore and define the neurobiological mechanisms that lead to the onset of VLOSLP, especially in terms of cognitive impairment[8].

Consistent with findings in TOS, most individuals with VLOSLP exhibit mild cognitive impairment across various cognitive domains[3,9]. A systematic review by Yang et al[10] included 17 articles and reported that patients with VLOSLP experienced a decline in overall cognitive ability, particularly in reasoning, perception, language, memory, and attention. Van Assche et al[5] reported that VLOSLP patients exhibited slight declines in processing speed and executive function compared with normally aging individuals, with more pronounced declines in memory, language, psychomotor speed, and orientation. They also validated these functional impairments based on morphological measurements and neuropsychological scoring, and additionally found significant correlations between thalamic volumes and memory function in VLOSLP[8]. These declines in cognitive domains suggest that VLOSLP patients may exhibit abnormalities in both brain structural and functional aspects.

At present, only a few studies have focused on the pathophysiological mechanisms of VLOSLP, including genetic research, neuroinflammation research, protein biomarker research, and neuroimaging research[8,11-13]. Notably, neuroimaging research has been used to identify differences between VLOSLP and normal aging[8]. A large-scale nationwide study in Finland revealed that 58.2% of VLOSLP patients had brain imaging findings ranging from mild lesions to normal, with the most common brain structural changes being white matter or cerebrovascular lesions[7]. Van Assche et al[5] reported that, compared with elderly TOS patients, VLOSLP patients presented a larger cerebellar ventricle-to-brain ratio and more significant brain atrophy, and suggested that the underlying neurobiological changes in VLOSLP patients include atrophy and hypometabolism in the frontal, temporal lobes, and subcortical regions. Matsuoka et al[14] found that the pineal volume in VLOSLP samples was lower than that in cognitively normal samples, which may be one of the neuropathological mechanisms involved. Van Assche et al’s morphological study[8] reported that, compared with healthy controls (HCs), VLOSLP group presented lower gray matter volumes in the thalamus, left inferior frontal gyrus, left insula, and thalamo-temporal regions. These magnetic resonance imaging (MRI) studies demonstrated structural differences between VLOSLP patients and HCs through morphological analysis, but their perspectives were relatively narrow and research efforts have not yet explored VLOSLP from the perspective of brain functional activity.

In recent years, an increasing number of researchers have turned their attention to resting-state functional MRI (rs-fMRI), which uses blood oxygenation level-dependent (BOLD) signals to reflect changes in cerebral hemodynamics and explore alterations in spontaneous brain activity across various disease states[15]. The application of rs-fMRI has become increasingly widespread and the technique has been applied in the study of a variety of neuropsychiatric disorders, while there is a lack of research on the brain functional mechanisms of VLOSLP[16]. Therefore, this study conducted a voxel-based rs-fMRI analysis, which, for the first time assessed whole-brain functional indices in patients with VLOSLP and combined these indices with neuropsychological scores and clinical symptoms, to clarify the neurobiological mechanisms of VLOSLP and its associated cognitive impairment. In this study, the amplitude of low-frequency fluctuations (ALFF), fractional ALFF (fALFF), regional homogeneity (ReHo), and voxel-mirrored homotopic connectivity (VMHC) analyses were performed to explore abnormalities in resting-state functional activity among VLOSLP patients compared with HCs.

MATERIALS AND METHODS

Participants

In according with the consensus criteria proposed by the International Late-Onset Schizophrenia Group[2], this study recruited 33 individuals with VLOSLP from the Department of Geriatric Psychiatry, The Affiliated Mental Health Center of Nanjing Medical University, China. The inclusion criteria for VLOSLP patients were as follows: (1) Aged 60-90 years, Han ethnicity, and right-handed; (2) Age of onset of psychotic symptoms ≥ 60 years; and (3) Exclusion of organic, affective disorders, and dementia. 34 HCs were also recruited from the surrounding community. The inclusion criteria for HCs were as follows: (1) Aged 60-90 years, Han ethnicity, and right-handed; (2) Not meeting the diagnostic criteria for any mental disorders on Axis I of the Diagnostic and Statistical Manual of Mental Disorders, 5th edition; and (3) No severe physical illnesses. The exclusion criteria for both groups were as follows: (1) Diseases that could lead to cognitive impairment, such as cerebrovascular diseases, Parkinson’s disease, brain trauma, and other neurological diseases; (2) Other conditions such as pregnancy, or alcohol or drug abuse; (3) Cardiovascular diseases such as severe hypertension or arrhythmias, and metabolic diseases, such as severe diabetes or thyroid dysfunction; and (4) Contraindications for MRI. This study was approved by the Ethics Committee of the Wuxi Mental Health Center (Ethics approval No. WXMHCIRB2023 LLky001). All participants signed informed consent forms before participating in the study, and all procedures involving human participants in this study were in accordance with the Declaration of Helsinki.

MRI data acquisition

MRI data were collected at the Huadong Sanatorium using a 3.0 T Magnetom Skyra (Siemens Medical System, Germany) with a 12-channel phased-array head coil, including 3-dimensional (3D) T1-weighted and rs-fMRI images. During the MRI scan, participants were asked to lie still and stay awake in a supine position, with their head movements limited by foam pads. 3D T1-weighted images were acquired using the 3D magnetization-prepared rapid acquisition gradient-echo sequence with the following parameters[17]: Time repetition = 2530 ms; time echo = 2.98 ms; flip angle = 7°; field of view = 256 × 256 mm²; matrix size = 256 × 256; slice thickness = 1 mm; voxel size = 1 × 1 × 1 mm³; and 192 slices in total. Rs-fMRI data were collected via the following parameters[18]: Time repetition = 2050 ms; time echo = 30 ms; flip angle = 90°; field of view = 224 × 224 mm²; matrix size = 64 × 64; slice thickness = 3.5 mm; voxel size = 3.5 × 3.5 × 3.5 mm³; and 240 volumes in total.

FMRI preprocessing

The rs-fMRI image data were preprocessed by the DPABI 7.0 software package[19] in MATLAB 2020b (The Math Works, Natick, MA, United States). The main steps were as follows[20]: (1) Conversion of DICOM images to NIFIT format; (2) Discarding the data from the first 10 time points; (3) Correction of the data for slice timing; (4) Discarding data with a head motion distance greater than 2 mm or rotation angle greater than 2°; (5) Spatial normalization: Alignment of the original data with the Montreal Neurological Institute template, resampling the voxel size to 3 × 3 × 3 mm³; (6) Smoothing the data using a 4 mm full-width, half-maximum Gaussian kernel[21]; and (7) Detrending and low-frequency filtering.

ALFF, fALFF, ReHo, and VMHC analyses

ALFF, fALFF, ReHo, and VMHC values were calculated by Data Processing Assistant for rs-fMRI[15,21]. The ALFF value is determined by using the fast Fourier transform to shift the time series of a single voxel from the time domain to the frequency domain, and then averaging the square root of the power spectrum values within the low-frequency range of 0.01-0.1 Hz[17]. The fALFF value is the ratio of the ALFF value to the power of the entire detectable frequency range, which measures the low-frequency amplitude’s proportion in the full spectrum, providing enhanced sensitivity to spontaneous brain activity over ALFF due to reduced physiological noise impact[18]. The ReHo value is assessed for similarity between the time series of a single voxel and those of its 26 neighboring voxels by using Kendall’s coefficient of concordance (KCC), which is obtained by dividing the KCC value of a single voxel by the average KCC value of the whole brain, and then spatially smoothing by a Gaussian kernel with a full width at half maximum of 4 mm[22]. The VMHC value is obtained by calculating the Pearson correlation coefficient between the time series of each voxel in the cerebral hemisphere and the mirrored time series of the corresponding voxel in the contralateral hemisphere[17]. To improve the normal distribution of the data, the correlation coefficient is standardized through Fisher’s z-transform.

Neuropsychological and clinical data collection

In this study, the Activity of Daily Living (ADL) scale was used to assess the ability of all participants to care for themselves in daily life. Cognitive functions were assessed via the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA) scales. The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) scale was employed to assess the neurocognitive functions, with subscales including immediate memory, visuospatial skills, language, attention, and delayed memory. The severity of clinical psychiatric symptoms in individuals with VLOSLP was assessed via the Positive and Negative Syndrome Scale (PANSS), with subscales including positive symptoms, negative symptoms, and general psychopathology symptoms.

Statistical analysis

SPSS 25.0 software (Chicago, IL, United States) was used to compare the demographic and neuropsychological data of the VLOSLP patients and HCs. Two-sample t-tests were conducted on the age, education, ADL score, MMSE score, MoCA score, and RBANS total and subscale scores of the two groups, and a χ²-test was performed based on sex to assess the differences between the two groups. Using the DPABI toolbox, voxel-based two-sample t-tests were conducted to evaluate differences in ALFF, fALFF, ReHo, and VMHC values between VLOSLP patients and HCs, with age, sex, years of education, and head motion variables as covariates. Then Gaussian random field multiple comparison correction was performed on the statistical results (set P < 0.001 at the voxel level, P < 0.01 at the cluster level, two tailed). The clusters with significant differences between the two groups were defined as regions of interest (ROIs), and the average ALFF, fALFF, ReHo, and VMHC values of each participant were extracted from the ROIs. Exploratory partial correlation analysis was conducted between ALFF, fALFF, ReHo, and VMHC values of ROIs and cognitive function (MMSE score, MoCA score, RBANS total and subscale scores) and clinical symptoms (PANSS total and subscale scores), with age, sex, years of education, and head motion variables as covariates, and results with P < 0.05 were considered statistically significant.

RESULTS

Demographic, neuropsychological, and clinical data

There were no significant differences in age, sex, years of education, or ADL score between the two groups. Except for the RBANS visuospatial skills, there were significant differences in all the neuropsychological variables (Table 1).

Table 1 Demographic, neuropsychological, and clinical characteristics of two groups.

Variables	VLOSLP (n = 33)	HCs (n = 34)	t/χ2	P value
Age, years	74.18 ± 6.439	72.15 ± 4.432	1.502	0.139
Sex, male/female	11/22	16/18	1.311	0.252
Education, years	6.67 ± 3.079	7.94 ± 2.994	-1.718	0.091
ADL score	98.94 ± 3.699	100.00 ± 0.000	-1.647	0.109
MMSE score	24.70 ± 3.312	27.12 ± 1.572	-3.804	< 0.001
MoCA score	16.33 ± 4.456	20.62 ± 3.266	-4.498	< 0.001
RBANS total score	59.73 ± 11.555	73.00 ± 11.468	-4.719	< 0.001
RBANS immediate memory	54.39 ± 12.250	66.56 ± 15.315	-3.584	< 0.001
RBANS visuospatial skills	73.27 ± 12.350	73.68 ± 9.098	-0.152	0.880
RBANS language	79.48 ± 13.985	91.24 ± 11.594	-3.749	< 0.001
RBANS attention	63.36 ± 16.865	84.00 ± 18.055	-4.831	< 0.001
RBANS delayed memory	62.82 ± 17.782	80.68 ± 16.622	-4.248	< 0.001
PANSS total score	73.76 ± 9.779	NA	NA	NA
PANSS positive score	22.85 ± 4.452	NA	NA	NA
PANSS negative score	15.18 ± 3.860	NA	NA	NA
PANSS general score	36.03 ± 5.876	NA	NA	NA

VLOSLP: Very-late-onset schizophrenia-like psychosis; HCs: Healthy controls; ADL: Activity of Daily Living Scale; MMSE: Mini Mental State Examination; MoCA: Montreal Cognitive Assessment; RBANS: Repeatable Battery for the Assessment of Neuropsychological Status; PANSS: Positive and Negative Syndrome Scale; NA: Not available.

Brain regions with significant differences in ALFF, fALFF, ReHo, and VMHC values

Compared with HCs, the VLOSLP group presented decreased ALFF values in the left cuneus (CUN), right precuneus (PCUN), right precentral gyrus (PreCG), and left paracentral lobule (PCL); increased fALFF values in the left caudate nucleus (CAU); decreased fALFF values in the right calcarine fissure and surrounding cortex (CAL) and right PCUN; increased ReHo values in the right putamen (PUT); and decreased VMHC values in the bilateral CAL (clusters A, B, C, and D), bilateral superior temporal gyrus (STG), and bilateral CUN (Table 2, Figures 1 and 2).

Figure 1 Brain regions with significant differences in amplitude of low-frequency fluctuations, fractional amplitude of low-frequency fluctuations, regional homogeneity, and voxel-mirrored homotopic connectivity between the very-late-onset schizophrenia-like psychosis and healthy controls groups, corrected using Gaussian random field (voxel P < 0.001, cluster P < 0.01). A: Compared with healthy controls (HCs), the very-late-onset schizophrenia-like psychosis (VLOSLP) group showed decreased amplitude of low-frequency fluctuations (ALFF) values in the left cuneus, right precuneus, right precentral gyrus and left paracentral lobule; B: Compared with HCs, the VLOSLP group showed increased fractional ALFF values in the left caudate nucleus and decreased fractional ALFF values in the right calcarine fissure and surrounding cortex and right precuneus; C: Compared with HCs, the VLOSLP group showed increased regional homogeneity values in the right putamen; D: Compared with HCs, the VLOSLP group showed decreased voxel-mirrored homotopic connectivity values in the bilateral calcarine fissure and surrounding cortex (clusters A, B, C, and D), bilateral superior temporal gyrus, and bilateral cuneus. The colored bars show t values. ALFF: Amplitude of low-frequency fluctuations; fALFF: Fractional amplitude of low-frequency fluctuations; ReHo: Regional homogeneity; VMHC: Voxel-mirrored homotopic connectivity.

Figure 2 Boxplot with scattered points of amplitude of low-frequency fluctuations, fractional amplitude of low-frequency fluctuations, regional homogeneity, and voxel-mirrored homotopic connectivity values in brain regions with significant differences between the very-late-onset schizophrenia-like psychosis and healthy controls groups. A: Compared with healthy controls (HCs), the very-late-onset schizophrenia-like psychosis (VLOSLP) group showed decreased amplitude of low-frequency fluctuations (ALFF) values in the left cuneus, right precuneus, right precentral gyrus and left paracentral lobule; B: Compared with HCs, the VLOSLP group showed increased fractional ALFF values in the left caudate nucleus and decreased fractional ALFF values in the right calcarine fissure and surrounding cortex (CAL) and right precuneus; C: Compared with HCs, the VLOSLP group showed increased regional homogeneity values in the right putamen; D: Compared with HCs, the VLOSLP group showed decreased voxel-mirrored homotopic connectivity values in the bilateral CAL (clusters A, B, C, and D), bilateral superior temporal gyrus, bilateral cuneus. ^aP < 0.01. VL: Very-late-onset schizophrenia-like psychosis; HCs: Healthy controls; ALFF: Amplitude of low-frequency fluctuations; fALFF: Fractional amplitude of low-frequency fluctuations; ReHo: Regional homogeneity; VMHC: Voxel-mirrored homotopic connectivity; CUN.L: Left cuneus; CUN.R: Right cuneus; PCUN.R: Right precuneus; PreCG.R: Right precentral gyrus; PCL.L: Left paracentral lobule; CAU.L: Left caudate nucleus; PUT.R: Right putamen; CAL.R: Right calcarine fissure and surrounding cortex; CAL.L: Left calcarine fissure and surrounding cortex; STG.R: Right superior temporal gyrus; STG.L: Left superior temporal gyrus; A: Cluster A; B: Cluster B; C: Cluster C; D: Cluster D.

Table 2 Brain regions with significant differences in spontaneous brain activity between two groups.

Feature	Brain regions	Side	AAL	Cluster size	MNI coordinates, x	MNI coordinates, y	MNI coordinates, z	Peak t value
ALFF	CUN	L	45	38	-3	-78	30	-5.0145
	PCUN	R	68	75	9	-69	51	-5.7456
	PreCG	R	2	64	30	-24	60	-5.5530
	PCL	L	69	132	-12	-30	75	-5.0655
fALFF	CAL	R	44	75	30	-54	12	-4.5549
	CAU	L	71	40	-12	0	15	5.0065
	PCUN	R	68	62	3	-57	39	-5.0995
ReHo	PUT	R	74	60	24	6	15	6.8145
VMHC	CAL (A)	R	44	29	24	-60	12	-4.9374
	CAL (B)	L	43	24	-24	-60	12	-4.9374
	CAL (C)	R	44	19	9	-75	9	-4.5645
	CAL (D)	L	43	19	-9	-75	9	-4.5645
	STG	R	82	20	51	-27	15	-4.2929
	STG	L	81	20	-51	-27	15	-4.2929
	CUN	R	46	36	9	-93	27	-4.5842
	CUN	L	45	37	-9	-93	27	-4.5842

AAL: Anatomical automatic labeling; MNI: Montreal Neurologic Institute; ALFF: Amplitude of low-frequency fluctuations; fALFF: Fractional amplitude of low-frequency fluctuations; ReHo: Regional homogeneity; VMHC: Voxel-mirrored homotopic connectivity; CUN: Cuneus; PCUN: Precuneus; PreCG: Precental gyrus; PCL: Paracentral lobule; CAL: Calcarine fissure and surrounding cortex; CAU: Caudate nucleus; PUT: Putamen; STG: Superior temporal gyrus; A: Cluster A; B: Cluster B; C: Cluster C; D: Cluster D; L: Left; R: Right.

Exploratory partial correlation analysis of fMRI characteristics in the ROIs with neuropsychological and clinical scores

Exploratory partial correlation analysis revealed that, in the VLOSLP group, the ALFF values of the right PCUN were negatively correlated with the MMSE score (r = -0.426, P = 0.021) and PANSS positive subscale score (r = -0.400, P = 0.032); the VMHC values of clusters C and D in the bilateral CAL were identical, and both were negatively correlated with the RBANS total score (r = -0.395, P = 0.034), RBANS delayed memory score (r = -0.438, P = 0.017), and PANSS positive subscale score (r = -0.391, P = 0.036). No correlation was found between other ROI values and scales (P > 0.05) (Figure 3).

Figure 3 Scatter plots show partial correlations between indices of brain function and cognitive function and clinical symptom severity in very-late-onset schizophrenia-like psychosis, with age, sex, education, and head motion as covariates. A: The amplitude of low-frequency fluctuations values in the right precuneus were negatively correlated with the Mini Mental State Examination score; B: The amplitude of low-frequency fluctuations values in the right precuneus were negatively correlated with the Positive and Negative Syndrome Scale positive subscale score; C: The voxel-mirrored homotopic connectivity (VMHC) values in the bilateral calcarine fissure and surrounding cortex (CAL, clusters C and D, which have the same VMHC values) were negatively correlated with the Repeatable Battery for the Assessment of Neuropsychological Status total score; D: The VMHC values in the bilateral CAL (clusters C and D) were negatively correlated with the Repeatable Battery for the Assessment of Neuropsychological Status delayed memory score; E: The VMHC values in the bilateral CAL (clusters C and D) were negatively correlated with the Positive and Negative Syndrome Scale positive subscale score. MMSE: Mini Mental State Examination; ALFF: Amplitude of low-frequency fluctuations; PCUN.R: Right precuneus; PANSS: Positive and Negative Syndrome Scale; RBANS: Repeatable Battery for the Assessment of Neuropsychological Status; VMHC: Voxel-mirrored homotopic connectivity; CAL: Calcarine fissure and surrounding cortex.

DISCUSSION

To our knowledge, this is the first study to explore the brain functional characteristics of VLOSLP based on ALFF, fALFF, ReHo, and VMHC. Compared with HCs, the VLOSLP group exhibited cognitive impairment, and showed decreased ALFF values in the left CUN, right PCUN, right PreCG, and left PCL; increased fALFF values in the left CAU; decreased fALFF values in the right CAL and right PCUN; increased ReHo values in the right PUT; and decreased VMHC values in the bilateral CAL, bilateral STG, and bilateral CUN. Exploratory correlation analyses indicated that the ALFF values in the right PCUN and the VMHC values in the bilateral CAL were both correlated with cognitive impairment and the severity of clinically positive symptoms in individuals with VLOSLP. Additionally, the decreased VMHC values of the bilateral CAL were also correlated with memory impairment.

ALFF refers to the magnitude of low-frequency amplitude in BOLD signals, which is associated with neuronal activity, with a decrease in the ALFF representing a decrease in spontaneous neural activity[23]. In this study, compared with HCs, the VLOSLP group presented decreased ALFF values in the left CUN, right PCUN, right PreCG, and left PCL. These findings are consistent with previous TOS studies, where Wang et al[24] and Hoptman et al[25] both reported decreased ALFF values in these brain regions in TOS patients. Gao et al[26] further reported that the ALFF values in the bilateral PCUN of TOS patients were not only lower than those in HCs but also that decreased ALFF values could effectively distinguish TOS patients from healthy individuals, with accuracy and sensitivity specificity of up to 73.36% and 91.60%, respectively. This suggested that the decreased ALFF values in the PCUN may serve as a potential indicator for schizophrenia patients, suggesting the possibility of early identification and warning. Therefore, compared to other brain regions, the dysfunction of the PCUN may play a more significant role in the pathogenesis of both VLOSLP and TOS.

FALFF is the normalization index of ALFF, representing the proportion of low-frequency amplitude within the entire BOLD signal amplitude. An increase or decrease in fALFF values indicates escalated or reduced neural activity in that region, respectively[23]. In this study, compared with those of HCs, the fALFF values was increased in the left CAU and decreased in the right CAL and the right PCUN in the VLOSLP group. Consistent with this study, Hoptman et al[25] reported decreased fALFF values in the bilateral PCUN of TOS patients. The same results were also reported by Athanassiou et al[27] and Gao et al[28]. Liu et al[29] further found that TOS patients had increased fALFF values in the bilateral CAU and decreased fALFF values in the bilateral PCUN. Zhang et al’s study[30] revealed a decrease in fALFF values in the right CAL of TOS patients with tardive dyskinesia, which correlated with the RBANS delayed memory score, suggesting that reduced functional activity in the CAL is associated with memory impairment in TOS patients. Additionally, previous studies have shown that the functional stability of the CAL is reduced in TOS patients, suggesting the significant role of the functional abnormalities of the CAL in the pathogenesis of schizophrenia[31].

ReHo is a measure of the synchrony between the time series of a single voxel and its neighboring voxels, reflecting the consistency of neuronal activity in that brain region, and increased ReHo values indicate a state of near-synchronization of neuronal activity[18]. This study revealed increased ReHo values in the right PUT of the VLOSLP group, and the same abnormal ReHo values were also reported in TOS studies. Wu et al[32] reported increased ReHo values in the bilateral PUT of TOS patients. Zhou et al[33], using a newly developed method called dynamic regional phase synchronization analysis, reported that average dynamic regional phase synchronization and ReHo values in the right PUT were increased in first-episode TOS patients, indicating enhanced neuronal activity synchrony in the PUT of TOS patients. Yang et al[22] found that regardless of the age of onset of schizophrenia, whether in adolescence or adulthood, the ReHo values in the bilateral the PUT of patients were higher than those in normal controls. Combining the above research results of TOS with the results of this study, it indicates that the PUT plays a role in schizophrenia regardless of the age of onset.

VMHC is an indicator of functional connectivity between the two hemispheres of the brain and is calculated by comparing the activities of corresponding voxels in both hemispheres. A higher VMHC value implies a more balanced functional activity between the two hemispheres[18]. This study revealed that compared with those of HCs, the VMHC values of the bilateral CAL, bilateral STG, and bilateral CUN were decreased in the VLOSLP group. Notably, the bilateral CAL exhibited four clusters (A, B, C, and D) with decreased VMHC values. Although there is currently no direct research supporting the findings of decreased VMHC values in the CAL and CUN regions in this study, Xue et al[34] discovered through the study of dynamic functional connectivity that there were changes in dynamic functional connectivity between the lateral occipital cortex and multiple brain networks in TOS patients, indicating that the reduced stability of functional connectivity in the lateral occipital cortex may lead to the emergence of schizophrenia. Given that the CAL and CUN are anatomically part of the occipital lobe, this implies that the decreased functional activity in the CAL and CUN may be correlated with the onset of schizophrenia. In addition to abnormalities in the CAL and CUN, Chen et al[35] reported decreased VMHC values in the bilateral STG of TOS patients with auditory verbal hallucinations. Zhang et al[36] also reported a decrease in VMHC values in the STG of first-episode TOS patients, and suggested that the decreased values in the STG of TOS patients could serve as predictive markers of clinical features. Therefore, the findings of this study on VMHC values were generally consistent with previous TOS research findings.

Exploratory partial correlation analysis revealed that the ALFF values in the right PCUN were negatively correlated with the MMSE score and the PANSS positive subscale score. As a key component of the default mode network, changes in the structure and function of the PCUN may affect the integrity of its connections with other brain regions. Previous studies have shown that in schizophrenic patients with genetic polymorphisms of disrupted-in-schizophrenic-1, both the function and structure of the PCUN are abnormal, and disrupted-in-schizophrenic-1 is considered one of the genes that may be involved in the pathogenesis of schizophrenia, suggesting an important role of the PCUN in the development of schizophrenia[37]. Furthermore, the PCUN is instrumental in cognitive processes such as memory retrieval, self-awareness, and visuospatial functioning[38]. The decline in these functions may not only impair patients’ perception of the real world but also weaken their self-control over their own behavior and thoughts, leading to cognitive impairment and the development of positive symptoms. Echoed by the findings of this study, Zhou et al[39] reported a significant decrease in brain function activity in the left PCUN of deficit TOS patients, which was correlated with the Trail Making Test part B and spatial processing scores, indicating the role of the PCUN in cognitive processing and the impact of its reduced activity on cognitive function. This study revealed a correlation between the right PCUN and positive symptoms of the PANSS, but current TOS research has not reported the same finding. A possible explanation is that, although the pathogenesis of VLOSLP is similar to TOS, the pathogenesis of VLOSLP may be more similar to neurodegenerative diseases, thus causing structural and functional changes in the brain (including the PCUN region), which in turn affects its relationship with positive symptoms.

This study also revealed that the VMHC values of clusters C and D in the bilateral CAL were identical, and both were negatively correlated with RBANS total score, RBANS delayed memory score, and PANSS positive subscale score. As the primary visual area in the occipital lobe, the CAL includes the primary visual cortex (V1) and is involved mainly in the reception, processing, integration of visual information, and the formation of visual memory[40,41]. Although previous studies have paid less attention to the association between the occipital lobe and schizophrenia, Tohid et al[42] reported that schizophrenia is associated with structural and functional abnormalities in the occipital lobe. The CAL is not only involved in the visual processing of emotional stimuli in the occipital lobe, but is also associated with neurocognitive dysfunction[43]. The decline in emotional processing function may slow cognitive processing speed and lead to perceptual distortions and thought disorders in patients, thereby exacerbating positive symptoms. Zhou et al[39] reported that the neural activity of the CAL in non-deficit TOS patients was correlated with the Trail Making Test part B scores, whereas Zhang et al[30] reported a correlation between the CAL and RBANS delayed memory score in TOS patients, indicating the role of the CAL in cognitive function, particularly in memory function. Although previous TOS studies did not directly report the correlation between the CAL and the PANSS positive symptom subscale score, Tohid et al[42] reported that abnormalities in the visual cortex can lead to the occurrence and development of hallucinations, resulting in positive symptoms. However, this may also be a unique pathogenesis of VLOSLP and should be a focus for verification in future studies.

This study has several limitations. First, the low prevalence of VLOSLP among all psychiatric disorders makes it challenging to collect samples, and patients experiencing paranoid symptoms are frequently reluctant to participate in research, leading to a small sample size. Second, this study collected only cross-sectional data, lacking longitudinal data and follow-up data, and did not include TOS patients as an additional control group for comparison. Therefore, this study cannot determine the dynamic changes in the brain function and cognitive levels of VLOSLP patients during the course of the disease, nor can it determine the impact of different ages of onset on schizophrenia spectrum disorders. Third, in this study, although the exploratory correlation analysis results suggested that indices of the right PCUN and bilateral CAL in the VLOSLP were significantly correlated with cognitive impairment and clinically positive symptoms, it is not possible to rule out the possibility of false-positive results due to type I error. Future research should combine functional and structural MRI indices, collect longitudinal data, and compare schizophrenia spectrum disorder patients of different ages of onset to further investigate the neurobiological mechanisms of VLOSLP and the impact of age of onset on schizophrenia spectrum disorders.

CONCLUSION

This study utilized the ALFF, fALFF, ReHo, and VMHC to explore alterations in spontaneous brain activity among VLOSLP patients. The preliminary results of this study suggested that multiple significantly altered brain functional regions in VLOSLP patients were concentrated in the CUN, PCUN, and CAL, involving the PreCG, PCL, CAU, PUT, and STG. The decrease in the brain functional activity of the right PCUN and bilateral CAL may lead to cognitive impairment and clinically positive symptoms in VLOSLP patients. Additionally, the decreased VMHC values of the bilateral CAL may contribute to memory impairment. It was a useful attempt to explore the brain functional characteristics of VLOSLP, which could further clarify the neurobiological basis of VLOSLP and provide a beneficial approach to understanding the mechanisms by which VLOSLP leads to cognitive impairment from different perspectives.