Predictive value of multimodal functional magnetic resonance imaging for cognitive impairment in patients with non-dialysis chronic kidney disease

Introduction

Chronic kidney disease (CKD) is a progressive condition that affects hundreds of millions of individuals worldwide and ranks among the leading causes of mortality globally (1). Cognitive impairment (CI) is relatively common in individuals with CKD (2). Micro-structural and functional brain alterations in patients with end-stage renal disease (ESRD) undergoing hemodialysis constitute pathophysiological substrates of CI and may serve as early predictive markers for cognitive decline in this population (3). However, the development of predictive and evaluative frameworks for CI in patients with non-dialysis CKD remains critically understudied, impeding early intervention efforts.

The pathophysiological mechanisms underlying cognitive decline in CKD are multifactorial. They include vascular injury, uremic toxin accumulation, inflammatory changes, and oxidative stress, all of which may contribute to neuronal damage (4). CI in patients with CKD typically manifests insidiously and progresses gradually, making early detection during routine clinical assessment challenging. Therefore, identifying effective methods for early prediction and evaluation of cognitive decline in this patient population is critical.

Multimodal functional magnetic resonance imaging (fMRI) integrates multiple MRI techniques to provide a comprehensive assessment of brain structure, function, activity, and connectivity. Voxel-based morphometry (VBM), a neuroimaging technique, analyzes focal differences in the anatomical structure of the brain. This enables the quantitative assessment of gray matter, white matter, and cerebrospinal fluid volumes and densities, thereby indicating the relationship between brain structure and cognitive function (5). Prior neuroimaging studies have demonstrated that mild CI (MCI) in older adults is strongly associated with reduced gray matter volume (GMV) in the bilateral parahippocampal gyri, hippocampi, and fusiform gyri compared to cognitively normal individuals (P<0.05) (6). In patients with Alzheimer’s disease, the decline in executive function is strongly associated with reduced GMV in the frontal and occipital lobes (7).

Additionally, blood oxygenation level-dependent fMRI (BOLD-fMRI) offers a non-invasive technique for evaluating brain tissue function by assessing neural activity based on hemodynamic responses (8). Of the resting-state fMRI (rs-fMRI) metrics, amplitude of low-frequency fluctuations (ALFF) and regional homogeneity (ReHo) are commonly used measures of spontaneous brain activity. Increased ALFF indicates increased excitability in the brain region, which has been correlated with cognitive performance (9). Conversely, reduced ReHo suggests impaired local neural synchronization, indicative of abnormal activity in the affected brain regions (9).

Given these insights, in the present study, we investigated the use of multimodal fMRI for the early prediction of CI in patients with non-dialysis CKD. By combining structural MRI and rs-fMRI to identify brain network abnormalities and structural changes associated with cognitive decline, this study provides a novel, neuroimaging-based approach for the early detection of cognitive dysfunction in patients with CKD. Such an approach may contribute to an improved prognosis and enhanced quality of life for individuals affected by this condition. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1771/rc).

Methods

Participants and clinical evaluation

Patients were recruited from the Department of Nephrology, Northern Jiangsu People’s Hospital, between September 2022 and September 2023. Patients with non-dialysis CKD were included based on the following criteria: (I) age between 18 and 75 years; (II) not receiving dialysis; (III) diagnosis of CKD based on the established diagnostic criteria.

Exclusion criteria were: (I) previous renal replacement therapy or kidney transplantation; (II) presence of acutely-active lesions, including but not limited to acute infection, acute kidney injury, or unstable conditions in the acute phase following stroke; (III) contraindications to MRI, such as pacemakers or metallic implants; (IV) visual or hearing impairment, or inability to complete MRI examination or cognitive function assessments; (V) history of mental illness or long-term use of psychotropic medications; and (VI) congenital brain lesions.

MRI data acquisition and processing

All MRI scans were performed using a 3.0T MR750W system (GE Healthcare, Chalfont St Giles, UK). The imaging protocol comprised both structural and functional sequences.

Image acquisition

High-resolution 3D T1-weighted images (3D T1WI), which provide a detailed anatomical map of the brain with excellent contrast between gray matter, white matter, and cerebrospinal fluid, were acquired using a brain volume (BRAVO) sequence with the following parameters: repetition time (TR) =8.5 ms, echo time (TE) =3.2 ms, inversion time (TI) =450 ms, flip angle =12°, field of view (FOV) =256×256 mm², matrix size =256×256, and slice thickness =1.2 mm, no gap. Conventional sequences, which are standard clinical scans used to rule out incidental neurological findings such as strokes, tumors, or white matter lesions, including axial T2-weighted fluid-attenuated inversion recovery, T2-weighted, and T2*-weighted imaging, were obtained for clinical evaluation and exclusion of incidental findings.

Image processing

Structural T1-weighted images were processed with FreeSurfer for automated cortical reconstruction and subcortical segmentation. VBM pipeline included spatial normalization to the Montreal Neurological Institute space; tissue segmentation into gray matter, white matter, and cerebrospinal fluid; modulation; and smoothing with an 8-mm full-width at half-maximum Gaussian kernel. Total intracranial volume (TIV) was calculated for covariance adjustment.

Resting-state functional MRI data were preprocessed using the Statistical Parametric Mapping (SPM12) implemented in the MathWorks, Inc., Natick, MA, USA (MATLAB) R2021a, incorporating motion correction, spatial normalization, smoothing with a 6-mm kernel, and nuisance covariate regression. Subsequent analyses generated amplitude of ALFF and ReHo maps following established methodologies.

Definition of CI

All participants underwent standardized cognitive function assessments before MRI scanning. Overall cognitive function was evaluated using the Montreal Cognitive Assessment (MoCA) scale, which has a maximum score of 30. Lower scores indicate worse cognitive function, with a score <26 defined as CI. A +1 point correction was applied for participants with ≤12 years of education to adjust for educational bias. The MoCA assessment lasts approximately 10 min and evaluates the following cognitive domains: visuospatial/executive function (5 points), naming (3 points), attention (6 points), language (3 points), abstraction (2 points), delayed recall (5 points), and orientation (6 points). Based on MoCA scores, patients with CKD were categorized into two groups: the CI group (MoCA score <26) and non-cognitive impairment (NCI) group (MoCA score ≥26).

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the Northern Jiangsu People’s Hospital (No. 2023ky183), and informed consent was taken from all individual participants.

Statistical analysis

Clinical data were analyzed using descriptive statistics. Continuous variables with a normal distribution are expressed as mean ± standard deviation (SD), whereas non-normally distributed variables are presented as median [interquartile range (IQR)]. Categorical variables are expressed as proportions. Group comparisons were performed using the t-test for normally distributed continuous variables, Mann-Whitney U test for non-normally distributed variables, and chi-squared test for categorical variables. The primary neuroimaging metrics were global GMV and gray matter volume fraction (GMVF). GMVF was calculated as GMV divided by TIV. We used binary logistic regression to assess its association with global GMV, adjusting for age, education, and eGFR. Model performance was evaluated using the area under the curve (AUC), with the optimal GMV cut-off determined using the Youden index.

To examine domain-specific associations, each MoCA domain was dichotomized (‘impaired’/‘normal’). The relationship between GMV and each domain was tested using adjusted binary logistic regression (same covariates). For GMVF, owing to model in multivariable regression, we evaluated its discriminative power for each domain using univariate ROC analysis and reported the AUC [95% confidence interval (95% CI)]. Additionally, Spearman’s correlation was used to assess the bivariate relationship between global GMV and eGFR. To further address potential confounders, we conducted supplementary and sensitivity analyses: (I) a sensitivity analysis was performed by excluding participants with extreme eGFR values (below the 10th and above the 90th percentile) to re-test the primary group difference in global GMV; and (II) the cohort was stratified by a pre-specified eGFR threshold (60 mL/min/1.73 m²) to explore the modulation of renal function on the GMV-CI relationship. All statistical analyses were conducted using SPSS version 27.0, with P<0.05 considered statistically significant.

Results

Demographic, clinical, and brain structure characteristics

Among the 60 initially-included patients, VBM imaging data from 53 were ultimately included in the analysis. Three patients were excluded due to dyspnea during the MRI examination and four declined to undergo the MRI scan. The demographic and clinical characteristics of the patients are presented in Table 1. Patients with CKD with CI were significantly older {58 [53, 67] years) than those without CI (36 [31, 48] years; P=0.000038} and had fewer years of education (8.68±4.22 vs. 13.73±3.58 years; P=0.000157). Notably, cerebral GMV was substantially lower in the CI group (572.56±39.70 cm³) compared to the NCI group (621.30±62.12 cm³; P=0.001). Consistent with the absolute GMV results, the GMVF was significantly reduced in the CI group (0.410±0.032) compared to the NCI group (0.433±0.034, P=0.024).

Association between GMV and renal function, and its confounding role

We first examined the direct bivariate relationship between brain structure and renal function. As shown in Figure S1, there was a positive but non-significant correlation between global GMV and eGFR in the overall cohort (r=0.247, P=0.074).

To investigate the confounding effect of renal function, we included eGFR as a covariate along with age and education in an ANCOVA model. After controlling for these confounding effects, the group differences in both GMV [F(1, 49) =1.101, P=0.299] and GMVF [F(1, 49) =0.004, P=0.95] were no longer statistically significant (Table S1). This indicates that renal function, together with age and education, substantially explains the observed association between brain structure and cognitive status.

We further explored whether the severity of renal dysfunction moderates the relationship between GMV and CI through two sensitivity analyses. First, after excluding participants with extreme eGFR values (below the 5th and above the 95th percentiles), the reduction in normalized GMV in the CI group remained significant [572.79±39.91 vs. 621.30±62.12 cm³ in NCI; t(47)=–3.362, P=0.002; Table S2]. Second, stratifying the cohort by the clinical eGFR cut-off of 60 mL/min/1.73 m² revealed a clear moderating effect: the GMV reduction in the CI group was not significant in patients with eGFR ≥60 mL/min/1.73 m² (P=0.148) but was highly significant in those with eGFR <60 mL/min/1.73 m² (P=0.006; Table S3).

Binary logistic regression and predictive accuracy of GMV and GMVF for domain-specific CI. To explore the associations between brain structure and specific cognitive domains, we performed binary logistic regression for GMV and univariate ROC analysis for GMVF. After adjusting for age, education, and eGFR, GMV was significantly associated only with visuospatial/executive impairment [odds ratio (OR) per 1 cm³ decrease = 0.970, % CI: 0.975–0.997, P=0.028]. Associations with other cognitive domains were not statistically significant (Table 2). The univariate discriminative ability of GMVF was generally modest, with AUC ranging from 0.49 to 0.713 across domains (Table 2).

Differences in the VBM-derived GMV between the groups

VBM analysis using gender and total cranial brain volume as covariates indicated that the GMVs in the right amygdala (AMYG.R), left insula (INS.L), and right middle temporal gyrus (MTG.R) were smaller in the CI group than those in the NCI group [voxel-wise P<0.001, cluster-level P<0.05, family-wise error (FWE)-corrected; Table 3, Figure 1].

Figure 1 Comparison of gray matter volume between groups based on voxel-based morphometrics. The color scale represents the t-values from the two-sample t-test comparing the NCI and CI groups. The maps are thresholded at a voxel-wise significance of P<0.001 and a cluster-level significance of P<0.05, with family-wise error correction for multiple comparisons. The red areas indicate brain regions with differences between groups. The red areas indicate brain regions with differences between groups. CI, cognitive impairment; NCI, non-cognitive impairment.

Differences in the amplitude of low-frequency fluctuations between the groups

Among the 60 patients included initially, BOLD imaging data from 51 were ultimately included in the analysis. Three patients were excluded owing to dyspnea during the MRI examination, four declined to undergo MRI, and two had incomplete imaging data that could not be preprocessed. Using sex as a covariate, BOLD signal analysis indicated statistically significant differences in ALFF between the two groups. Specifically, ALFF values were lower in the CI group in the left postcentral gyrus (PoCG.L) and right supplementary motor area (SMA.R) (voxel-wise P<0.001, cluster-level P<0.05, FWE-corrected multiple comparisons). The NCI group exhibited substantially higher ALFF than the CI group in these two brain regions (Table 4, Figure 2).

Figure 2 Comparison of low-frequency fluctuating values between groups based on the oxygenation level-dependent-functional magnetic resonance imaging analysis. The color scale represents the t-values from the two-sample t-test. The maps are thresholded at a voxel-wise significance of P<0.001 and a cluster-level significance of P<0.05, with family-wise error correction for multiple comparisons. The red areas indicate brain regions with differences between groups.

Differences in ReHo between groups

Using sex as a covariate, BOLD processing and analysis indicated differences in ReHo values between the groups in both the PoCG.L and right postcentral gyrus (voxel-wise P<0.001, cluster-level P<0.05, corrected for multiple comparisons using FWE). Specifically, the NCI group exhibited statistically higher ReHo values than the CI group in these two brain regions (Table 5, Figure 3).

Figure 3 Comparison of regional homogeneity values between groups based on the oxygenation level-dependent-functional magnetic resonance imaging analysis. The maps are thresholded at a voxel-wise significance of P<0.001 and a cluster-level significance of P<0.05, with family-wise error correction for multiple comparisons. The red areas indicate brain regions with differences between groups.

Predictive value of brain GMV for CI

The AUC for brain GMV in predicting CI was 0.729, with a cut-off value of 619.9 cm³ (95% CI: 0.561–0.879). At this threshold, the sensitivity was 94.7%, and the specificity was 53.3% (P=0.01; Figure 4).

Figure 4 Receiver operating characteristic curve for brain gray matter volume in predicting cognitive impairment. 95% CI, 95% confidence interval; AUC, area under the curve.

Discussion

CI is a common complication among patients with non-dialysis CKD. CKD is an established independent risk factor for early cognitive decline and dementia (10). Compared to the general population, patients with CKD are at higher risk of developing dementia and its precursor, mild CI (11). Although cognitive assessment scales offer valuable clinical measures of cognitive function, they can be influenced by subjective factors such as educational level. In contrast, multimodal fMRI provides complementary objective evidence by revealing structural and functional brain alterations associated with CI, thereby offering neurobiological insights beyond behavioral assessment alone.

In this study, VBM analysis indicated that the GMV was substantially reduced in the CI group compared with the NCI group (572.56±39.70 vs. 621.30±62.12 cm³). This finding was further substantiated by analyzing GMVF, which accounts for individual differences in TIV. The significant reduction in GMVF confirms that the observed grey matter loss represents specific pathological atrophy, rather than a mere consequence of smaller overall brain size. The AUC for brain GMV in predicting CI was 0.729, with a cut-off value of 619.9 cm³. Gray matter, primarily comprising neuronal cell bodies, dendrites, and synapses, reflects the integrity of neurons. Its reduction typically indicates neuronal loss or degeneration, directly affecting cognitive function. Patients with Alzheimer’s disease exhibit specific atrophy in olfactory regions of the limbic/internal temporal lobes during mild CI, which can predict early cognitive decline (12).

We examined the relationship between brain structure and renal function. A positive, albeit non-significant, correlation was found between global GMV and eGFR (Spearman’s r=0.247, P=0.074). Crucially, when age, education, and eGFR were included as covariates in an ANCOVA model, the group differences in both GMV and GMVF were no longer statistically significant. This indicates that renal function, along with age and education, is a substantial confounder of the observed association between global brain structure and cognitive status. Sensitivity analyses further clarified this relationship. After excluding participants with extreme eGFR values, the reduction in normalized GMV remained significant in the CI group. More importantly, stratification by a clinical eGFR cut-off of 60 mL/min/1.73 m² revealed a clear moderating effect: the GMV reduction in the CI group was highly significant only in the subgroup with eGFR <60 mL/min/1.73 m², but not in those with eGFR ≥60 mL/min/1.73 m². This suggests that the pathophysiological mechanisms linking CKD to cerebral atrophy become a dominant, detectable factor influencing the brain-cognition relationship primarily in advanced stages of renal dysfunction (13). This stratification underscores a key clinical insight from our multimodal data: the predictive utility of imaging biomarkers such as GMV is context-dependent and is significantly modulated by the stage of renal disease (13), highlighting the necessity of an integrated assessment that considers both brain and systemic health metrics (14,15).

After adjusting for sex and total brain volume, the results of the VBM analysis indicated differences in GMV between the CI and NCI groups in the right amygdala, left insula, and right middle temporal gyrus (voxel-wise P<0.001, cluster-level P<0.05, FWE-corrected). The NCI group exhibited substantially larger GMVs in these regions, suggesting that region-specific atrophy may serve as the basis of cognitive deficits in patients with CKD. The amygdala plays a central role in emotion processing and memory integration, and reduced GMV is associated with impaired emotion regulation and memory function (16). The role of the amygdala in neuropsychiatric disorders, such as anxiety, depression, aggression, and temporal lobe epilepsy, has been reported (17). Beyond its role in cognition, the amygdala is a core component of the neural circuitry underlying anxiety. Our finding of reduced GMV in this region may therefore extend beyond CI to help explain the high prevalence of anxiety disorders in the CKD populations. This interpretation is supported by neuroimaging studies linking structural and functional alterations in the amygdala to anxiety symptoms (18) The pathophysiological link between CKD and amygdala atrophy may involve uremic toxin accumulation. Serum urea is a uremic toxin suggested to accumulate in the brains of patients with ESRD, potentially contributing to cognitive decline (19). Recent experimental evidence further indicates that elevated urea can directly induce pathological changes in the amygdala, such as promoting abnormal oligodendrocyte progenitor cell proliferation (20). A recent animal study suggested that CKD induced anxiety by altering corticotropin-releasing hormone gene expression and tryptophan metabolism in the amygdala (21). The current finding of markedly reduced GMV in the right amygdala in patients with CKD and CI may explain the high prevalence of mood disorders, including depression and anxiety, as well as memory decline. However, as this study did not assess emotional states, further research is required to clarify this association.

The insula is involved in higher cognitive functions, including emotional regulation, self-awareness, and executive function (22). Reduced GMV in the left and right insular cortices may be associated with cognitive deficits (23). Reduced functional connectivity between the posterior insula and language/auditory cortex has been linked to CI and poor social functioning (24). In the current study, patients with CKD and CI had lower GMV in the left insula compared to the NCI group in patients with CKD (P<0.01), supporting the role of insula in CI associated with CKD.

The temporal cortex, incorporating regions critical for auditory, visual, and language processing, also supports emotion and memory (25). The medial temporal lobe is crucial for working memory and perception (26). A strong association exists between structural changes in the medial temporal lobe and cognitive decline (27). However, an interesting finding in our study was the spatial dissociation between regions showing structural atrophy and those exhibiting functional alterations. This pattern, although initially challenging to interpret, may reflect network-level disruptions rather than focal pathology. The affected regions are key components of distributed brain networks: the amygdala and insula are hubs of the salience network, whereas the postcentral gyrus is involved in somatosensory processing. It is plausible that structural damage in network hubs produces remote functional effects in connected regions—a phenomenon known as “network diaschisis”. This interpretation is consistent with recent multimodal neuroimaging studies (28) and suggests that CKD-related CI may involve large-scale network disruption rather than isolated regional damage. For instance, patients with Parkinson’s disease and mild CI exhibit increased cortical thinning over time in the right middle temporal gyrus, insula, and precuneus (29), aligning with our findings in CKD. Future studies should investigate the mechanisms underlying GMV reduction in these regions and explore whether targeted interventions can improve cognitive outcomes in CKD.

Spontaneous brain activity was investigated using ALFF and ReHo measures from BOLD-fMRI. Compared to the NCI group, the CI group exhibited substantially lower ALFF in the left postcentral gyrus and right supplementary motor area, with reduced ReHo in the bilateral postcentral gyrus. These findings suggest that the cognitive deficits in patients with CKD may be associated with abnormal spontaneous neural activity and altered functional connectivity in specific brain regions. The ALFF and ReHo reflect the intensity of spontaneous brain activity and regional coordination, respectively, and are important markers for brain dysfunction. Voxel-wise whole-brain rs-fMRI metrics, including ReHo, degree of centrality (DC), and ALFF, are commonly used, and these studies can be used for coordinate-based meta-analysis (30).

The postcentral gyrus, a key component of the primary somatosensory cortex, processes sensory inputs and plays an important role in emotional processing, including identification, generation, and regulation (31). Modulating activities in the right postcentral gyrus via transcranial direct current stimulation improves emotional recognition and enhances empathy, guilt, and shame. This enhanced emotional and cognitive ability strongly inhibits malicious creative performance, likely due to increased emotional empathy, which reduces the propensity for harmful behaviors (32). The current study found that patients with CKD and CI exhibited substantially lower ALFF and ReHo in the postcentral gyrus, indicating reduced spontaneous neural activity and local synchronization in this region, which may impair sensory perception, attention, and cognitive function.

Domain-specific analyses provided further nuance. After adjusting for confounders, global GMV was independently associated only with visuospatial/executive impairment, but not with other cognitive domains. Concurrently, GMVF showed limited univariate discriminative ability across most domains (AUCs <0.7). These findings collectively indicate that global gray matter measures are neither robust nor independent biomarkers of domain-specific CI in non-dialysis CKD, a pattern consistent with observations in other populations where diffuse atrophy poorly captures domain-specific deficits (33). This underscores the predominant confounding role of demographic factors and renal function in driving the observed cognitive decline in this population. Consequently, our results emphasize the need to systematically assess and control for these potent confounders in both clinical evaluation and research, as exemplified in contemporary cohort studies (34). Furthermore, the limited predictive value of global metrics underscores the potential superiority of regional brain measures. The focal atrophy in regions such as the amygdala and insula, identified by VBM in this study, may offer more specific insights into the neural mechanisms of CKD-related CI than whole-brain volumetric indices. Therefore, the predictive value of multimodal MRI may be best realized not by global metrics alone, but by integrating them with regional structural and functional signatures to create a more precise and mechanistically informed predictive profile.

Limitations

Despite these promising findings, there are some limitations in this study. First, the relatively small sample size may restrict the generalizability of our results. Second, the CI and NCI groups differed at baseline in age, education, and eGFR. Although our sensitivity analysis supports the main finding, this imbalance should be considered when interpreting the results. Third, the difference in education levels between the groups, likely a cohort effect related to the age difference, exemplifies the challenge of completely disentangling these intertwined factors. Additionally, we observed a spatial dissociation between structural and functional alterations that we could not fully explore with connectivity analyses. Future research should address these issues by expanding sample size, incorporating broader confounding variables, and adopting multivariate models to validate and improve the predictive value of the findings.

Conclusions

This study demonstrated that multimodal fMRI can identify brain structural and functional alterations associated patients with non-dialysis CKD with CI. These alterations, including reduced GMV in specific regions and abnormal spontaneous neural activity, are linked to cognitive decline. The findings support the predictive value of multimodal MRI, particularly when regional measures are combined with clinical factors such as eGFR, for the early identification of CI in this population.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1771/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1771/dss

Funding: This work was supported by the Young Scientists Fund of the National Natural Science Foundation of China (No. 82302133); Scientific Research Project of Jiangsu Provincial Healthcare Commission (No. Z2022068); Yangzhou Municipal Science and Technology Program Projects (No. YZ2022169); and Social Development Projects under Yangzhou Municipal Science and Technology Bureau’s Science and Technology Plan (No. YZ2022094).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1771/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the Northern Jiangsu People’s Hospital (No. 2023ky183), and informed consent was taken from all individual participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Kovesdy CP. Epidemiology of chronic kidney disease: an update 2022. Kidney Int Suppl (2011) 2022;12:7-11. [Crossref] [PubMed]
2. Raphael KL, Wei G, Greene T, Baird BC, Beddhu S. Cognitive function and the risk of death in chronic kidney disease. Am J Nephrol 2012;35:49-57. [Crossref] [PubMed]
3. Zheng J, Sun Q, Wu X, Dou W, Pan J, Jiao Z, Liu T, Shi H. Brain Micro-Structural and Functional Alterations for Cognitive Function Prediction in the End-Stage Renal Disease Patients Undergoing Maintenance Hemodialysis. Acad Radiol 2023;30:1047-55. [Crossref] [PubMed]
4. Xie Z, Tong S, Chu X, Feng T, Geng M. Chronic Kidney Disease and Cognitive Impairment: The Kidney-Brain Axis. Kidney Dis (Basel) 2022;8:275-85. [Crossref] [PubMed]
5. Nemoto K. Understanding Voxel-Based Morphometry. Brain Nerve 2017;69:505-11. [Crossref] [PubMed]
6. Chen Z, Xie Q, Wang J, Wang Y, Zhang H, Li C, Wang Y, Cong L, Tang S, Hou T, Song L, Du Y, Qiu C. Mapping grey matter and cortical thickness alterations associated with subjective cognitive decline and mild cognitive impairment among rural-dwelling older adults in China: A population-based study. Neuroimage Clin 2024;44:103691. [Crossref] [PubMed]
7. Riverol M, Ríos-Rivera MM, Imaz-Aguayo L, Solis-Barquero SM, Arrondo C, Montoya-Murillo G, Villino-Rodríguez R, García-Eulate R, Domínguez P, Fernández-Seara MA. Structural neuroimaging changes associated with subjective cognitive decline from a clinical sample. Neuroimage Clin 2024;42:103615. [Crossref] [PubMed]
8. Woodward OB, Driver I, Schwarz ST, Hart E, Wise R. Assessment of brainstem function and haemodynamics by MRI: challenges and clinical prospects. Br J Radiol 2023;96:20220940. [Crossref] [PubMed]
9. Wang R, Liu N, Tao YY, Gong XQ, Zheng J, Yang C, Yang L, Zhang XM. The Application of rs-fMRI in Vascular Cognitive Impairment. Front Neurol 2020;11:951. [Crossref] [PubMed]
10. Lipnicki DM, Crawford J, Kochan NA, Trollor JN, Draper B, Reppermund S, Maston K, Mather KA, Brodaty H, Sachdev PSSydney Memory and Ageing Study Team. Risk Factors for Mild Cognitive Impairment, Dementia and Mortality: The Sydney Memory and Ageing Study. J Am Med Dir Assoc 2017;18:388-95. [Crossref] [PubMed]
11. Giannakou K, Golenia A, Liabeuf S, Malyszko J, Mattace-Raso F, Farinha A, Spasovski G, Hafez G, Wiecek A, Capolongo G, Capasso G, Massy ZA, Pépin M. Methodological challenges and biases in the field of cognitive function among patients with chronic kidney disease. Front Med (Lausanne) 2023;10:1215583. [Crossref] [PubMed]
12. Jobin B, Boller B, Frasnelli JCIMA-Q group. Smaller grey matter volume in the central olfactory system in mild cognitive impairment. Exp Gerontol 2023;183:112325. [Crossref] [PubMed]
13. Drew DA, Weiner DE, Tighiouart H, Duncan S, Gupta A, Scott T, Sarnak MJ. Cognitive Decline and Its Risk Factors in Prevalent Hemodialysis Patients. Am J Kidney Dis 2017;69:780-7. [Crossref] [PubMed]
14. Jack CR Jr, Bennett DA, Blennow K, Carrillo MC, Feldman HH, Frisoni GB, Hampel H, Jagust WJ, Johnson KA, Knopman DS, Petersen RC, Scheltens P, Sperling RA, Dubois B. A/T/N: An unbiased descriptive classification scheme for Alzheimer disease biomarkers. Neurology 2016;87:539-47. [Crossref] [PubMed]
15. Almgren H, Camacho M, Hanganu A, Kibreab M, Camicioli R, Ismail Z, Forkert ND, Monchi O. Machine learning-based prediction of longitudinal cognitive decline in early Parkinson's disease using multimodal features. Sci Rep 2023;13:13193. [Crossref] [PubMed]
16. LeDoux JE. Emotion circuits in the brain. Annu Rev Neurosci 2000;23:155-84. [Crossref] [PubMed]
17. Dalla Corte A, Pinzetta G, Ruwel AG, Maia TFA, Leal T, Frizon LA, Isolan GR. Anatomical Organization of the Amygdala: A Brief Visual Review. Cogn Behav Neurol 2024;37:13-22. [Crossref] [PubMed]
18. Viggiano A, Cacciola G, Widmer DA, Viggiano D. Anxiety as a neurodevelopmental disorder in a neuronal subpopulation: Evidence from gene expression data. Psychiatry Res 2015;228:729-40. [Crossref] [PubMed]
19. Chen HJ, Qiu J, Qi Y, Guo Y, Zhang Z, Qin H, Wu F, Chen F. Regional cortical thinning and area reduction are associated with cognitive impairment in hemodialysis patients. Brain Res Bull 2025;229:111446. [Crossref] [PubMed]
20. Huang B, Huang Z, Wang H, Zhu G, Liao H, Wang Z, Yang B, Ran J. High urea induces anxiety disorders associated with chronic kidney disease by promoting abnormal proliferation of OPC in amygdala. Eur J Pharmacol 2023;957:175905. [Crossref] [PubMed]
21. Ibos KE, Bodnár É, Dinh H, Kis M, Márványkövi F, Kovács ZZA, Siska A, Földesi I, Galla Z, Monostori P, Szatmári I, Simon P, Sárközy M, Csabafi K. Chronic kidney disease may evoke anxiety by altering CRH expression in the amygdala and tryptophan metabolism in rats. Pflugers Arch 2024;476:179-96. [Crossref] [PubMed]
22. Craig AD. How do you feel--now? The anterior insula and human awareness. Nat Rev Neurosci 2009;10:59-70. [Crossref] [PubMed]
23. Liao J, Yan H, Liu Q, Yan J, Zhang L, Jiang S, Zhang X, Dong Z, Yang W, Cai L, Guo H, Wang Y, Li Z, Tian L, Zhang D, Wang F. Reduced paralimbic system gray matter volume in schizophrenia: Correlations with clinical variables, symptomatology and cognitive function. J Psychiatr Res 2015;65:80-6. [Crossref] [PubMed]
24. Tian Y, Zalesky A, Bousman C, Everall I, Pantelis C. Insula Functional Connectivity in Schizophrenia: Subregions, Gradients, and Symptoms. Biol Psychiatry Cogn Neurosci Neuroimaging 2019;4:399-408. [Crossref] [PubMed]
25. Zachlod D, Kedo O, Amunts K. Anatomy of the temporal lobe: From macro to micro. Handb Clin Neurol 2022;187:17-51. [Crossref] [PubMed]
26. Wu Z, Buckley MJ. Prefrontal and Medial Temporal Lobe Cortical Contributions to Visual Short-Term Memory. J Cogn Neurosci 2022;35:27-43. [Crossref] [PubMed]
27. Bonarota S, Caruso G, Domenico CD, Sperati S, Tamigi FM, Giulietti G, Giove F, Caltagirone C, Serra L. Integration of automatic MRI segmentation techniques with neuropsychological assessments for early diagnosis and prognosis of Alzheimer's disease. A systematic review. Neuroimage 2025;314:121264. [Crossref] [PubMed]
28. Buonincontri V, Viggiano D, Gigliotti G. The brain extracellular space in chronic kidney disease. Behav Brain Res 2025;476:115271. [Crossref] [PubMed]
29. Sokołowski A, Bhagwat N, Chatelain Y, Dugré M, Hanganu A, Monchi O, McPherson B, Wang M, Poline JB, Sharp M, Glatard T. Longitudinal brain structure changes in Parkinson's disease: A replication study. PLoS One 2024;19:e0295069. [Crossref] [PubMed]
30. Zheng YX, Huai YY, Qiao Y, Zang YF, Luo H, Zhao N. Neural correlates of psychotherapy in mental disorders: A meta-analysis of longitudinal resting-state fMRI studies. Psychiatry Res 2025;348:116495. [Crossref] [PubMed]
31. Kropf E, Syan SK, Minuzzi L, Frey BN. From anatomy to function: the role of the somatosensory cortex in emotional regulation. Braz J Psychiatry 2019;41:261-9. [Crossref] [PubMed]
32. Gao Z, Lu K, Hao N. Transcranial direct current stimulation (tDCS) targeting the postcentral gyrus reduces malevolent creative ideation. Soc Cogn Affect Neurosci 2023;18:nsad019. [Crossref] [PubMed]
33. Pini L, Pievani M, Bocchetta M, Altomare D, Bosco P, Cavedo E, Galluzzi S, Marizzoni M, Frisoni GB. Brain atrophy in Alzheimer's Disease and aging. Ageing Res Rev 2016;30:25-48. [Crossref] [PubMed]
34. Tang X, Han YP, Chai YH, Gong HJ, Xu H, Patel I, Qiao YS, Zhang JY, Cardoso MA, Zhou JB. Association of kidney function and brain health: A systematic review and meta-analysis of cohort studies. Ageing Res Rev 2022;82:101762. [Crossref] [PubMed]