Abnormal Functional Connectivity Pattern of the Reward Circuitry is Associated with Negative Symptoms and Cognitive Functions in First-Episode Schizophrenia

Abstract

Background:

This study aimed to investigate the relationship between abnormal functional connectivity (FC) patterns in the reward circuitry of the brain and negative symptoms and cognitive impairment in individuals with first-episode schizophrenia (FES).

Methods:

Fifty-two FES patients and 59 healthy controls (HCs) were recruited for this cross-sectional study. Thirteen brain regions associated with the reward circuitry were defined as regions of interest (ROIs), and FCs between each ROI and the whole brain were analyzed. Cognitive function was assessed by the MATRICS Consensus Cognitive Battery.

Results:

Within-network analyses indicated that, compared to HCs, FES patients exhibited increased FCs between the left ventrolateral prefrontal cortex and left thalamus, which negatively correlated with negative symptoms. Whole-brain analyses revealed some weakened FCs in FES patients compared to those in HCs. The FCs between the right nucleus accumbens and right insular lobe and between the right putamen and both the left anterior cingulate cortex and left precentral gyrus positively correlated with attention/vigilance only in HCs. Additionally, the FC between the left putamen nucleus and left inferior frontal gyrus positively correlated with verbal and visual learning only in HCs.

Conclusion:

These findings highlight the differential FC patterns in the reward circuitry in FES patients and indicate that the enhanced within-network FC observed in these patients may contribute to their negative symptoms. The absence of correlations between certain FCs and attention, verbal learning, and visual learning can be explained by decoupling of the reward circuitry from the cognitive control brain regions.

Main Points

• Study Objective: The research aimed to explore the relationship between abnormal functional connectivity (FC) patterns in the brain’s reward circuitry and their association with negative symptoms and cognitive functions in individuals with first-episode schizophrenia (FES).

• Key Findings: The study observed that altered FC patterns in the reward system were specifically linked to negative symptoms and cognitive impairments, indicating the potential importance of this brain circuitry in schizophrenia pathophysiology.

• Implications for Treatment: The results suggest that targeting the reward circuitry through therapeutic interventions may provide a new avenue for addressing negative symptoms and cognitive dysfunction in FES patients.

Introduction

Schizophrenia, impacting roughly 1% of the global population,¹ is marked by positive symptoms, negative symptoms, and cognitive impairment. The latter 2 symptoms significantly impact long-term recovery and social functioning and are linked to abnormalities within the neural reward systems of the brain.

First-episode schizophrenia (FES) is a critical period that defines the disease course. Early identification of abnormalities in the reward circuitry may guide prognosis and treatment. Key regions of this circuitry include the amygdala, anterior cingulate cortex (ACC), dorsolateral prefrontal cortex (DLPFC), ventrolateral prefrontal cortex (VLPFC), dorsomedial prefrontal cortex (DMPFC), orbitofrontal cortex (OFC), ventromedial prefrontal cortex (VMPFC), nucleus accumbens (NAc), ventral caudate, lateral habenula (LHb), thalamus, putamen, and ventral tegmental area (VTA).^2,3 Aberrant activations of the reward circuitry, linked to negative symptomatology and cognitive impairments, have been observed in several schizophrenia resting-state functional magnetic resonance imaging (rs-fMRI) studies.⁴

Resting-state functional magnetic resonance imaging studies have indicated possible associations between aberrant functional connectivities (FCs) in the reward circuitry and negative symptoms and cognitive functions in schizophrenia.⁵ However, recent studies have yielded inconsistent findings regarding connectivity within the reward system in patients with schizophrenia. Some studies have revealed reduced FC between the VTA, a core region of dopaminergic projections, and the ACC and DLPFC,⁶ as well as decreased connectivity with the associative striatum,⁷ while others find no abnormalities.⁸ Studies on the OFC have also shown varied results. Some studies observed weakened FC between the amygdala and right OFC, which was linked with impaired social function,⁹ and lower FC linking the OFC and medial prefrontal cortex exhibited a positive relationship with visual-motor processing speed and attentional performance.¹⁰ Further research on the ACC has revealed abnormal FCs in different ACC subregions, with reduced FCs between the dorsal ACC and medial prefrontal cortex, right posterior OFC, and precuneus.¹¹ However, another study revealed abnormally heightened connectivity between the ACC and the left OFC in individuals with schizophrenia, which predicted violent behavior.¹²

Although these investigations examined various reward-related brain regions, the majority predominantly emphasized the influence of abnormal functional connectivity in just one specific area of the reward network. Furthermore, these variations may stem from heterogeneity across factors such as symptomatology, illness duration, and treatment regimens. Comprehensive evaluations of abnormal FC patterns of the reward circuitry in FES patients are still lacking. Identifying these FC patterns is crucial for better understanding their effects on clinical symptoms and cognitive dysfunction in FES. This investigation aimed to systematically examine how abnormal FC patterns within the reward network, as well as the seed-based whole-brain voxel-wise connectivity of each regions of interest (ROI), influence clinical symptoms and cognitive functioning in FES.

Material and Methods

In this study, 52 patients with FES and 59 healthy controls (HCs) were recruited at Beijing Anding Hospital affiliated to Capital Medical University. University for a cross-sectional study. This cross-sectional investigation complied with the ethical standards set forth by the Declaration of Helsinki and received approval from the Institutional Review Board of Beijing Anding Hospital, Capital Medical University (Approval Number: 2023 Research No. 58 - 202391FS-2). Each participant received an in-depth, detailed account of the study procedure and provided written informed consent. Informed assent was secured from participants aged <18 years, with additional informed consent provided by their parents or legal guardians.

Participants

Individuals diagnosed with FES were enrolled from the Early Psychosis Cohort Program of the hospital between August 2020 and November 2023. Healthy controls aligned with the patients in terms of gender, age, and education level were enlisted through advertisements. Individuals with FES were included if they fulfilled the following criteria: (a) fulfilled the diagnostic requirements for schizophrenia or schizophreniform disorder as determined by the Mini International Neuropsychiatric Interview (MINI 7.0); (b) duration of disease of no more than 3 years; (c) aged between 16 and 50 years; (d) right-handed; (e) intelligence quotient (IQ) ≥80; and (f) no contraindications for MRI. Healthy controls also underwent the MINI 7.0 to ensure that they had neither a past nor present psychiatric diagnosis.

Clinical Symptoms and Neurocognitive Assessment

Psychopathology was evaluated using the Positive and Negative Syndrome Scale (PANSS),¹³ while neurocognitive functioning was measured with the Chinese version of the MATRICS Consensus Cognitive Battery (MCCB).¹⁴ For detailed information on MCCB, refer to the Supplementary Materials. All assessments were conducted by 2 trained professionals.

Magnetic Resonance Imaging Data

Brain imaging was conducted at the Hospital, Capital Medical University University, using a 3.0-T Prisma MR scanner (Siemens Healthineers, Erlangen, Germany) equipped with a 64-channel radio frequency head coil. The participants were guided to remain still with their eyes closed and refrain from concentrating on any particular thoughts during the scan. Earplugs were provided to mitigate the noise from the scanner, and cotton pads were placed to reduce any head movement. Head motion parameters served as a criterion for assessing image quality (mean FD_Jenkinson <3 mm).

Functional imaging was conducted using a gradient-echo echo-planar imaging sequence, capturing 200 volumes over a period of 6 minutes and 40 seconds. The functional scan parameters were as follows: repetition time (TR), 2000 ms; echo time (TE), 30 ms; interleaved axial slices, 33; slice thickness/gap, 3.5/0.7 mm; flip angle (FA), 90°; matrix, 64 × 64; field of view (FOV), 200 × 200 mm²; and voxel size, 3.13 × 3.13 × 4.2 mm³. Additionally, T1-weighted structural images were acquired using a magnetization-prepared rapid acquisition gradient-echo sequence with the following parameters: TR, 2530 ms; TE, 1.85 ms; sagittal slices, 192; slice thickness, 1 mm; FA, 15°; matrix, 256 × 256; FOV, 256 × 256 mm²; and voxel size, 1 × 1 × 1 mm³.

Resting-State Functional Magnetic Resonance Imaging Preprocessing

Resting-state functional magnetic resonance imaging data were pre-processed using DPARSF version 5.4 (http://rfmri.org/dpabi), an integral component of the DPABI toolbox. The initial 10 data volumes were omitted to ensure stabilization of the functional signal. The preprocessing pipeline encompassed the following key procedures: correction for slice timing to address acquisition delays, realignment to correct for head motions, and segmentation of T1-weighted structural images; regression of covariates to mitigate noise, encompassing both linear and quadratic trends, the 24-parameter head motion model by Friston, and the principal components of signals derived from the white matter and cerebrospinal fluid; normalization of T1 images to the Montreal Neurological Institute (MNI) space through unified segmentation; resampling to achieve 2 mm isotropic voxel dimensions; application of spatial smoothing using a Gaussian kernel set to a 4 mm full width at half maximum; and temporal filtering within the range of 0.01-0.1 Hz to retain relevant frequencies.

Definition of Regions of Interest

In this study, ROIs were primarily defined using the Brainnetome Atlas,¹⁵ which delineates cortical and subcortical areas by integrating various neuroimaging techniques, such as diffusion tensor imaging, fMRI, and structural MRI, ensuring a rational parcellation of brain regions from both structural and functional perspectives. Thirteen brain regions were included in the reward circuit: DLPFC, VLPFC, VMPFC, DMPFC, OFC, ACC, amygdala, ventral caudate, putamen, NAc, thalamus, LHb, and VTA. Note that the VTA and LHb were defined using MNI coordinates and existing literature.¹⁶ Specifically, the VTA was defined within a 3-mm radius sphere using the DPABI and the coordinates (MNI: 0, −16, −7).¹⁶ The LHb was defined based on a study.¹⁷ The remaining eleven ROIs were defined by the Brainnetome Atlas. The specific areas selected are presented in Table 1. Excluding the VTA, each seed region was defined in both hemispheres, resulting in 25 seed ROIs. The rs-FC patterns within these 25 seed regions, as well as between the seeds and whole-brain voxels, were identified.

Table 1.

Reward System ROIs and Corresponding Brainnetome Subregions

ROIs	Brainnetome Subregions, MNI Coordinates, or Source of ROI
Left DLPFC	A8dl_l, A9l_l, A9_46d_l, A46_l, A9_46v_l, A8vl_l
Right DLPFC	A8dl_r, A9l_r, A9_46d_r, A46_r, A9_46v_r, A8vl_r
Left VLPFC	A44d_l, IFS_l, A45c_l, A45r_l, A44op_l, A44v_l
Right VLPFC	A44d_r, IFS_r, A45c_r, A45r_r, A44op_r, A44v_r
Left VMPFC	A8m_l, A6m_l, A9m_l, A10m_l
Right VMPFC	A8m_r, A6m_r, A9m_r, A10m_r
Left DMPFC	A6vl_l, A10l_l
Right DMPFC	A6vl_r, A10l_r
Left OFC	A14m_l, A12_47o_l, A11l_l, A11m_l, A13_l, A12_47l_l
Right OFC	A14m_r, A12_47o_r, A11l_r, A11m_r, A13_r, A12_47l_r
Left ACC	A23d_l, A24rv_l, A32p_l, A23v_l, A24cd_l, A23c_l, A32sg_l
Right ACC	A23d_r, A24rv_r, A32p_r, A23v_r, A24cd_r, A23c_r, A32sg_r
Left amygdala	mAmyg_l, lAmyg_l
Right amygdala	mAmyg_r, lAmyg_r
Left ventral caudate	vCa_l
Right ventral caudate	vCa_r
Left NAc	NAc_l
Right NAc	NAc_r
Left thalamus	mPFtha_l, mPMtha_l, Stha_l, rTtha_l, PPtha_l, Otha_l, cTtha_l, lPFtha_l
Right thalamus	mPFtha_r, mPMtha_r, Stha_r, rTtha_r, PPtha_r, Otha_r, cTtha_r, lPFtha_r
Left putamen	vmPu_l, dlPu_l
Right putamen	vmPu_r, dlPu_r
Left lateral habenula	Kim et al., 2016
Right lateral habenula	Kim et al., 2016
VTA	(0, −16, −7)

ACC, anterior cingulate cortex; DLPFC, dorsolateral prefrontal cortex; DMPFC, dorsomedial prefrontal cortex; MNI, Montreal Neurological Institute; NAc, nucleus accumbens; OFC, orbitofrontal cortex; ROI, region of interest; VLPFC, ventrolateral prefrontal cortex; VMPFC, ventromedial prefrontal cortex; VTA, ventral tegmental area.

Statistical Analysis

Data analyses were conducted using SPSS, version 25.0 (IBM SPSS Corp.; Armonk, NY, USA), SPM12 (https://www.fil.ion.ucl.ac.uk/spm/software/spm12/), and GRETNA (https://www.nitrc.org/frs/download.php/10440/GRETNA_2.0.0_release_1121.zip). The 2-sample t-test and χ² test were used to analyze differences in demographic data (sex, age, years of education, IQ) and cognitive function between individuals with FES and HCs. Statistical significance was defined as P < .05. Sex, age, duration of formal education, and head motion were included as covariates in the statistical analysis. Within-network FC analysis was performed using 2-sample t-tests in GRETNA. The results were corrected for multiple comparisons using family-wise error (FWE) correction (corrected P < .05). Whole-brain FC analysis, including analysis of between-group differences in whole-brain FC, was performed using 2-sample t-tests in SPM12. Multiple comparison corrections were performed using Gaussian random field correction. The voxel-level threshold was set at P < .001, whereas the cluster-level threshold was set at P < .05. Pearson’s correlation analysis with false discovery rate (FDR) correction was performed to investigate the association between symptom severity, cognitive function, and FC values.

Within-Network Functional Connectivity Analysis:

Resting-state data for each ROI were obtained by averaging the voxel-wise time series across the region. Pearson’s correlation coefficients for the time series of each ROI pair were then computed, yielding a set of 24 × 25/2 coefficients per participant. These coefficients were then converted into z-scores using the Fisher r-to-z transformation to measure the FC within each network.

Whole-Brain Functional Connectivity Analysis:

For each participant, voxel-wise whole-brain FC maps were produced by computing Pearson’s correlation coefficients between the average time series from each ROI and the time series of every voxel in the brain. Thereafter, the FC maps were transformed using the Fisher r-to-z method to achieve normality.

Association Between Altered Functional Connectivity and Clinical Symptoms in First-Episode Schizophrenia:

In the subsequent correlation analysis, altered ROI- and voxel-level FCs were represented by mean z-scores derived from the voxels within each surviving cluster. The clinical symptom was assessed using the PANSS. The normality of the clinical and cognitive scores was assessed using the Kolmogorov–Smirnov test. All scale scores were normally distributed. The relationship between aberrant FCs, clinical scores, and cognitive functions was examined by the Pearson correlation coefficient in FES patients and HCs.

Results

Demographic, Clinical, and Cognitive Function Data of the Participants

Fifty-two FES patients (medicated <3 days = 24, drug-naïve = 28) and 59 HCs were recruited for this study. To rule out the potential effects of medication use on the results, demographic information, clinical symptoms (Supplementary Table 1), and rs-fMRI data were compared between drug-naïve and medicated FES patients. No significant differences in demographic characteristics, clinical symptoms, or FCs within the reward circuit and between its nodes and the whole brain were observed. These findings suggest that short-term medication use did not significantly affect resting-state brain function. Therefore, in subsequent analyses, these 2 subgroups were combined and collectively referred to them as the FES group.

Table 2 summarizes the demographic and clinical characteristics of the FES and HC cohorts. The 2 groups did not differ significantly in age, sex, educational background, IQ, or mean FD. However, according to the MCCB, the FES group displayed significantly lower T-scores across all measured domains compared to the HCs (Table 3).

Table 2.

Demographic and Clinical Characteristics of FES (n = 52) and HC groups (n = 59)

Variable	Mean (SD)		P
	FES (n = 52)	HC (n = 59)
Age (years)	29.50 (±8.91)	27.10 (±6.99)	.115
Sex (M/F)	26/26	24/35	.345
Education (years)	14.75 (±3.47)	14.84 (±2.64)	.867
IQ	105.42 (±10.72)	108.45 (±9.76)	.121
Use of antipsychotics	24	-	-
Mean FD	0.085 (±0.058)	0.077 (±0.038)	.389
PANSS score
Positive symptoms	19.92 (±3.92)	-	-
Negative symptoms	14.21 (±5.54)	-	-
General symptoms	33.02 (±5.98)	-	-
Total score	68.04 (±13.12)	-	-

Table 3.

MCCB Domain Scores and Overall Composite Score of FES and HC Groups

	Mean (SD)		P
	FES (n = 52)	HC (n = 59)
SOP	44.10 (±10.77)	50.92 (±8.08)	<.001 b
AV	45.38 (±10.93)	50.34 (±9.23)	.011 b
WM	43.44 (±10.38)	48.47 (±10.46)	.013 b
VISL	45.58 (±11.52)	51.03 (±9.07)	.006 b
VERL	45.88 (±11.52)	50.64 (±9.07)	.021 b
RPS	45.10 (±10.84)	52.19 (±8.38)	<.001 b
SC	44.13 (±12.16)	52.53 (±7.78)	<.001 b
Composite score	42.25 (±11.82)	51.25 (±7.77)	<.001 b

P < .05.

Within-Network Functional Connectivity

After applying FWE correction (P < .05), only 1 within-network FC remained significant (Figure 1A). More specifically, compared to the HC group, the FES group demonstrated increased connectivity between the left VLPFC and left thalamus (P = .001, Cohen’s d = −0.64, 95% CI = [−1.01, −0.25]; Figure 1B).

Figure 1.

Aberrant within-network FC and its correlation with clinical symptoms. (A) FCs between ROIs showing significant differences between groups. The red line denotes increased connectivity (FES > HC). (B) Significant ROI-wise FC that survived FWE correction. (C) Analysis of the correlation between the aberrant ROI-wise FC and the severity of negative symptoms. **Indicates a statistically significant difference between the FES and HC groups (corrected P = .001). The violin plot shows the kernel probability density.

Whole-Brain Functional Connectivity

Table 4 and Figure 2 present the FC findings. In particular, relative to the HC group, the FES group exhibited reduced connectivity between the right NAc and right insular lobe, as well as between the left OFC and both the left medial cingulate gyrus and right lingual gyrus. Additional reductions in FC involved the right OFC and left cingulate gyrus; the left DMPFC and left precuneus; the left putamen nucleus and the left cingulate gyrus, left precentral gyrus, and left inferior frontal gyrus (IFG); and the right putamen and the left ACC, left precentral gyrus, and right insula.

Table 4.

ROIs with Altered Functional Connectivity with Whole Brain

ROIs	Region of Altered Connectivity	MNI Coordinates	Cluster Size (Voxels)	Cluster P
HC > FES
R. NAc	R. insula	(30, −12, 0)	57	<.001
L. OFC	L. lingual gyrus	(0, −75, −6)	66	<.001
	L. medial cingulate gyrus	(−3, 12, 39)	65	<.001
R. OFC	L. medial cingulate gyrus	(6, 12, 42)	41	<.001
L. DMPFC	L. precuneus	(−9, −42, 66)	51	<.001
L. putamen	L. IFG	(−54, 33, 0)	46	<.001
	L. medial cingulate gyrus	(−6, 21, 48)	54	<.001
	L. precentral gyrus	(−45, 3, 30)	85	<.001
R. putamen	R. insula	(48, 6, −9)	63	<.001
	L. ACC	(−3, 39, 18)	80	<.001
	L. precentral gyrus	(−48, 0, 33)	84	<.001

ACC, anterior cingulate cortex; DMPFC, dorsomedial prefrontal cortex; FES, first-episode schizophrenia; HC, healthy control; IFG, inferior frontal gyrus; L., left; MNI, Montreal Neurological Institute; NAc, nucleus accumbens; OFC, orbitofrontal cortex; R., right; ROI, region of interest.

Multiple comparison correction was performed using Gaussian Random Field Correction, the voxel-level threshold was set at P < .001, and the cluster-level threshold was set at P < .05. a means P < .001.

Figure 2.

Aberrant whole-brain FC in the reward circuit. The axial maps depict significant differences between the FES and HC groups for each ROI in the reward circuitry. Red to yellow indicates weaker FC in the FES group than in the HC group (Gaussian random field theory correction with voxel-wise P < .001 and cluster-wise P < .05).

Correlation Analysis

In terms of clinical symptoms, the within-network FC analysis identified a significant negative association between the connectivity of the left VLPFC and left thalamus and the negative symptom subscale of the PANSS (r = −0.33, 95% CI = [−0.55, −0.06], P = .018; Figure 1C). However, whole-brain FC analysis indicated that no aberrant FC was correlated with any domain of the PANSS.

Multiple FCs positively correlated with various domains of cognitive functions in the HC group. However, these positive correlations were absent in the FES group. Specifically, after FDR correction in the HC group, attention/vigilance (AV) was significantly positively correlated with FCs between the right NAc and right insular lobe (r = 0.34, P = .039) and between the right putamen and both the left ACC (r = 0.45, P = .002) and left precentral gyrus (r = 0.32, P = .04), (Figure 3A-C). In addition, FC between the left putamen and left IFG positively correlated with both verbal learning and memory (VERL) (r = 0.32, P = .03) and visual learning and memory (VISL) (r = 0.32, P = .04), (Figure 3D-E).

Figure 3.

Relationship between aberrant whole-brain FC and cognitive function in HCs. (A-C) Scatterplots showing the relationship between FCs and the AV domain in the MCCB. (D) Relationship between the FC and VISL. (E) Relationship between the FC and VERL. The dashed curves represent 95% CIs.

Discussion

In this study, the correlation between abnormal FC patterns within reward-related neural pathways and negative symptoms and cognitive impairment in FES patients was analyzed. An abnormally increased FC between the left thalamus and left VLPFC within the reward circuitry was revealed by the analyses and was negatively associated with negative symptoms. Furthermore, multiple core nodes within the reward circuitry exhibited widespread abnormal FCs throughout the brain. Notably, these correlations between FCs and the AV, VISL, and VERL domains were only found in HCs but were absent in the FES group.

In the present study, individuals with FES showed enhanced FC between the left VLPFC and left thalamus within the reward circuit. The enhanced FC suggested a dysfunctional activation pattern within the reward circuitry. It showed a significant inverse relationship with negative symptoms, similar to a study that also discovered a negative correlation between thalamo-frontal FC and negative symptoms.¹⁸ The VLPFC is crucial for the modulation of emotional responses, the process of making choices, and social interactions,¹⁹ while the thalamus functions as a pivotal relay hub involved in various functions, including cognition, memory, emotions, and language.²⁰ Therefore, the observed enhanced FC could reflect overactive engagement in processing reward and emotional information. This aligns with the theory of cerebellothalamocortical circuitry hyperconnectivity.²¹ In a ketamine intervention study, increased global connectivity in the thalamus was associated with a reduction in negative symptoms after ketamine injection.²² Therefore, the increased FC between the thalamus and VLPFC may enhance the brain’s capacity to process and regulate emotional responses, temporarily mitigating negative symptoms.

Additionally, the study revealed disrupted FCs between key reward circuit nodes (DMPFC, OFC, putamen, and NAc) and multiple brain regions. Specifically, FES patients showed reduced FC between these reward-related regions and several areas, including the insula, lingual gyrus, precuneus, ACC, medial cingulate gyrus, and prefrontal cortex. These affected regions are primarily components of the default mode and executive control networks. Previous research has demonstrated that weakened putamen-ACC connectivity can predict antipsychotic treatment response.²³ Additionally, disrupted insular regulation of default mode and executive networks has been linked to cognitive deficits in schizophrenia patients.²⁴ The lingual gyrus and precuneus, as key nodes in the default network, are crucial for episodic memory and attention processing.²⁵ Similarly, the precentral gyrus, an important component of the frontoparietal attention network, shows a correlation between its functional decline and attention deficits.²⁶ These findings suggest that the observed FC reductions may contribute to cognitive impairment in FES patients.

Positive correlations between weakened FCs and cognitive functions were observed only in the HC group. Specifically, FC between the right NAc and right insular lobe, as well as between the right putamen and both the left ACC and left precentral gyrus, showed positive associations with AV. Additionally, FC between the left putamen nucleus and left IFG positively correlated with both verbal and visual learning abilities. The absence of correlations between certain FCs and cognitive functions (attention, VISL, VERL) can be explained by decoupling of the reward circuitry from the cognitive control brain regions. This finding aligns with previous research showing that weakened FC between the VLPFC and IPL positively correlated with working memory only in HCs, but not in schizophrenia patients.²⁷ This decoupling of brain regions indicates that cognitive impairments in schizophrenia are linked to a reduction in connectivity.

Additionally, the NAc can direct behavior and recruit attention,²⁸ particularly when processing competing stimuli.²⁹ Working in conjunction with the insula, a key hub in the salience network, these regions collectively maintain cognitive function and attention.³⁰ Therefore, diminished FC between the NAc and insula may explain the attention deficits observed in FES through the decoupling of reward and attention circuits. The ACC and precentral gyrus, as part of the frontoparietal network,³¹ are essential for sustained attention and behavioral control.³² Additionally, the putamen, a core component of the dopamine system, contributes significantly to cognitive functions including reward prediction and decision-making.³³ Disrupted connectivity among these regions likely impairs information processing speed and attention.

The findings also showed that FC between the left putamen nucleus and left IFG was positively correlated with both VERL and VISL in HCs. The IFG, crucial for language processing, comprehension, and attention control,³⁴ also contributes to visual detection despite not being directly associated with visual cortical areas.³⁵ These results suggest that the decoupling between reward circuits and frontoparietal networks significantly impacts learning abilities.

This study involved a comprehensive evaluation of the reward circuitry in FES patients compared with HCs. However, it has some limitations. First, some patients received pharmacological therapy for a relatively short period. Although rs-fMRI analysis did not reveal any significant difference between medicated and unmedicated patients, the influence of medication cannot be completely ruled out. Moreover, the inclusion criterion of FES patients with an illness duration less than 3 years was relatively lenient, which might have affected the assessment of the initial disease state. To address this limitation, longitudinal studies were planned to be conducted with drug-naïve FES patients whose illness duration is less than 1 year and track their brain connectivity changes before and after medication initiation. This approach would help better understand the effects of medication effects on reward circuitry and potentially identify optimal treatment strategies. Second, the relatively small sample size highlights the need for additional studies with larger cohorts. Future multi-center collaborative studies with standardized protocols would not only increase statistical power but also help validate the findings across different populations and clinical settings. Additionally, larger sample sizes would enable more sophisticated analyses, such as machine learning approaches, to identify potential biomarkers for early diagnosis and treatment response prediction. Third, the cross-sectional design limited the ability to ascertain causality or the directionality of the relationships observed. To overcome this limitation, future studies should employ longitudinal designs following patients from the prodromal phase through different stages of the illness. This would help elucidate the temporal dynamics of reward circuit alterations and their relationship with symptom progression. Furthermore, incorporating multiple assessment time points would allow for the investigation of whether the observed connectivity changes are state-dependent or represent stable trait markers of the disease.

In conclusion, this study demonstrated that abnormalities in FC within the reward circuitry correlate with negative symptoms in individuals with FES. Furthermore, the absence of a correlation between certain FCs and cognitive functions in FES patients suggests a decoupling of the reward circuitry from cognitive-related brain regions. These findings have important clinical implications, as they suggest that targeting the reward circuitry through novel therapeutic approaches, such as neuromodulation techniques or targeted cognitive interventions, might be a promising strategy for treating negative symptoms and cognitive impairments in FES patients. Additionally, the identified FC patterns could potentially serve as neuroimaging biomarkers for early diagnosis and treatment response monitoring in clinical practice.

Funding Statement

This study was supported by the Project supported by the Beijing Natural Science Foundation (Grant No. 7242073), the National Natural Science Foundation of China (Grant No. 82171501), the Beijing Anding Hospital, Capital Medical University (Grant No. ADDL-03), the Beijing Hospitals Authority Clinical Medicine Development of special funding support, code: ZLRK202335.

Data Availability Statement:

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