White-matter functional network dysfunction associated with cognitive deficits and clinical phenotypes in patients with end-stage renal disease

White-matter functional network dysfunction associated with cognitive deficits and clinical phenotypes in patients with end-stage renal disease | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article White-matter functional network dysfunction associated with cognitive deficits and clinical phenotypes in patients with end-stage renal disease Ming Zhang, Peng Li, Yu-Xuan Shang, junya mu, Xinyi Zhu, Zhaoyao Luo, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7702413/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Among patients with end-stage renal disease (ESRD), the white matter (WM) is a particularly vulnerable area that is susceptible to various clinical risk factors. However, whether WM function is disrupted in ESRD patients and how this disruption provides valuable information for cognitive deficits and potential clinical phenotypes remain unknown. We prospectively enrolled 78 ESRD patients and 50 healthy controls. Using resting-state functional magnetic resonance imaging, we studied ESRD-related WM functional networks alterations. Functional connectivity, functional covariance connectivity, and coefficient Granger causality analysis were probed interactions among WM functional networks. The machine-learning models with leave-one-out cross-validation were applied. ESRD patients exhibited extensively disrupted interactions among WM functional networks, which correlated with cognitive deficits and ESRD-specific clinical risk factors, including uremic toxin accumulation, dysregulation of calcium-phosphorus homeostasis, and anemia. A random forest classifier achieved a maximum performance of 95.31% accuracy and 0.982 area under the ROC curve (AUC). Our results emphasized the imbalances of WM functional networks in ESRD patients, which might be used as potential neuroimaging markers for cognitive deficits and potential clinical phenotypes. Biological sciences/Neuroscience/Cognitive neuroscience Health sciences/Neurology/Neurological disorders/White matter disease Health sciences/Diseases/Kidney diseases/Chronic kidney disease/End-stage renal disease Biological sciences/Neuroscience/Computational neuroscience/Network models End-stage renal disease White matter Functional magnetic resonance imaging Functional connectivity Machine learning Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction End-stage renal disease (ESRD) has emerged as an increasingly severe global public health issue 1 . Cerebral white matter (WM) is a preferentially vulnerable region in the brain of patients with ESRD 2 – 4 , susceptible to various clinical risk factors including uremic toxins, calcium-phosphate metabolism disorders, renal anemia, and dialysis therapy 5 . Such patients often exhibit cognitive deficits 6 , which exerts a substantial impact on their self-management capabilities, dietary adherence, treatment compliance, and long-term survival outcomes 7 . Notwithstanding, the association between WM function and clinical phenotypes in ESRD patients remains largely undefined. The rapid advancement of MRI-based neuroimaging has provided critical evidence for exploring the neuropathological mechanisms of cognitive deficits in ESRD patients. Resting-state functional MRI (rs-fMRI) identifies cortical functional abnormalities in ESRD-related cognitive deficits by analyzing blood oxygen level-dependent (BOLD) signals, manifested as reduced spontaneous neural activity and functional connectivity in the default mode network 8 , 9 [medial prefrontal cortex, middle temporal gyrus, precuneus, and posterior cingulate gyrus], decreased functional connectivity (FC) in the sensorimotor network 10 , and disruption of large-scale brain connectome functional gradients 11 . Evidence from diffusion tensor imaging and diffusion kurtosis imaging has also revealed WM demyelination and microstructural abnormalities in ESRD patients with cognitive deficits 2 – 4 , including damage to critical projection fibers, association fibers, and commissural fibers such as the corona radiata, corpus callosum, superior longitudinal fasciculus, and corticospinal tract. Notably, prior studies have neither elucidated the existence and characteristics of the WM functional dysfunction in ESRD patients, nor elucidated the complex relationships among WM functional networks, cognitive deficits, and ESRD-related risk factors. Recent advancements in rs-fMRI have garnered increasing attention for investigating the WM neural activation and functional organization. During the resting-state, BOLD signals in WM encapsulate functional signatures of brain activity and connectivity 12 , which can be differentially evoked during various task paradigms—including perceptual, linguistic, and motor domains 13 , 14 . Peer et al 15 . further unveiled the intrinsic functional organization of WM, deriving discrete functional network parcellations through unsupervised clustering of BOLD signals. Moreover, recent studies have linked the FC of WM functional networks to pathological mechanisms underlying schizophrenia 16 , 17 , subcortical vascular cognitive impairment 18 , epilepsy 19 , Parkinson's disease 20 , and bipolar disorder 21 , highlighting the feasibility of using rs-fMRI to investigate WM networks dysfunctions. Thus, elucidating the interactions among WM functional networks may facilitate the interpretation of neuropathological mechanisms underlying cognitive deficits in patients with ESRD. To address these gaps, we collected rs-fMRI and clinical data from 78 ESRD patients and 50 healthy controls (HCs). In addition to traditional FC method, which quantifies direct temporal synchrony between two WM networks, we employed a novel functional covariance connectivity (FCC) approach 19 , 22 to estimate covariant relationships between WM networks—based on their correlations with multiple gray matter regions. We further probed interactions among WM functional networks using coefficient Granger causality analysis (cGCA). To further quantify the overall influence of each WM functional network within the interaction model, we examined the causal in/out strength of each network using classic graph-theoretic metrics 23 , 24 . Additionally, we employed multiple machine learning classification models to assess the discriminative power of WM functional network features extracted from ESRD patients and HCs. Specifically, we hypothesized that: (i) patients with ESRD would demonstrate aberrant interactions of WM functional networks, as indexed by altered metrics of FC, FCC, and Granger causality (GC) strength; (ii) these WM network dysfunctions would correlate with cognitive deficits and ESRD-specific clinical risk factors, including uremic toxin burden, calcium-phosphorus homeostasis dysregulation, and anemia. Our study linked the WM functional abnormalities with the pathophysiology mechanisms underlying cognitive deficits in ESRD patients and demonstrated that the dysfunctional interactions of WM networks as neurological biomarkers for ESRD patients. The entire experimental workflow and analytical methods are illustrated in Fig. 1 . Results Demographic and neuropsychological results Table 1 presents demographic and clinical characteristics for each group, with no significant between-group differences observed in age, sex, or educational level. Relative to HCs, ESRD patients demonstrated significantly poorer performance on multiple neuropsychological assessments, including the MoCA(total score [ p < 0.001], visuospatial [ p < 0.001], language [ p < 0.001], name [ p = 0.023], delayed memory [ p < 0.001], abstraction [ p < 0.001] ), AVLT-H (immediate recall [ p < 0.001], short-term recall [ p < 0.01], long-term recall [ p < 0.001], recognition [ p < 0.001]), TMT ( p < 0.001), BAI ( p < 0.001), and BDI ( p < 0.001). Table 1 Demographic, clinical characteristics and neuropsychological assessment in ESRD patients and HCs Variable HC(n = 50) ESRD(n = 78) t/χ² p -value Age(years) 34.76 (8.56) 35.15 (10.18) 0.227 0.821§ Sex (M/F) 38 (12) 53 (25) 0.609 0.435† Education(years) 12.32 (2.11) 12.46 (1.91) 0.392 0.696§ Dialysis vintage (months) / 38.58 (29.33) / / Creatinine(µmol/L) / 919.06 (218.10) / / Urea(mmol/L) / 23.69 (7.21) / / Kt/V / 1.46 (0.18) / / Haemoglobin(g/L) / 107.03 (20.31) / / Hematocrit(%) / 32.89 (5.98) / / Cystatin C(µg/mL) / 6.07 (3.17) / / Potassium(mmol/L) / 4.82 (0.74) / / Sodium(mmol/L) / 142.31 (3.28) / / Phosphorus(mmol/L) / 1.80 (0.51) / / Calcium(mmol/L) / 2.13 (0.34) / / Parathormone(ng/L) / 626.10 (453.38) / / Medication N % AT1-blocker 68 87 / / Beta-blocker 56 72 / / EPO 76 97 / / Antidepressants 0 0 / / Antihistamines 0 0 / / Analgesics 0 0 / / Vitamin D 42 54 / / Calcium antagonists 44 56 / / MoCA Total score 27.52(2.23) 23.85(2.87) -7.688 < 0.001 §* Visuospatial 4(1) 4(2) -5.631 < 0.001 ‡* Name 3(0) 3(0) -2.276 0.023 ‡* Attention 6(1) 6(1) -1.827 0.068 ‡ Language 3(1) 2(0) -5.420 < 0.001 ‡* Abstraction 2(0) 1(1) -5.392 < 0.001 ‡* Orientation 6(0) 6(1) -1.435 0.151 ‡ Delayed memory 5(1) 4(1) -5.410 < 0.001 ‡* AVLT-H IR-S 27.84(3.62) 25.22(4.45) -3.489 < 0.001 §* SR-S 10.30(1.20) 9.59(1.65) -2.624 0.010 §* LR-S 10.24(1.27) 9.00(1.73) -4.370 < 0.001 §* REC-S 11.84(0.47) 11.32(1.00) -3.437 < 0.001 §* TMT 41.76(13.26) 60.34(28.22) 4.355 < 0.001 ‡* BDI 7.90(5.20) 16.10(10.14) 5.282 < 0.001 §* BAI 25.18(2.83) 28.60(6.04) 3.750 < 0.001 §* Note.—Unless otherwise indicated, data are mean (standard deviation). § Analyzed with the independent two-sample t -test; data in parentheses have a 95% confidence interval. data are mean (standard deviation). ‡ Analyzed with the Mann-Whitney u -test,data are median (range interquartile).† Analyzed with the chi-square test. ESRD, end-stage renal disease; HCs, health controls; AVLT-H, auditory verbal learning test–Huashan version; IR-S, immediate recall score; SR-S, short-term recall score; LR-S, long-term recall score; REC-S, recognition score; MoCA, Montreal cognitive assessment; TMT, trail-making test; BAI, Beck anxiety inventory; BDI, Beck depression inventory. * Indicates a statistically significant difference after controlling age, sex, and education level. WM functional networks Clustering analysis identified k = 11 as the optimal cluster number, achieving both fine-grained resolution and high reproducibility (Dice coefficient > 0.85, Supplement Fig. 1 ). Consequently, these 11 WM functional networks were retained for subsequent analyses. Network nomenclature was assigned based on spatial localization. Consistent with prior literature 15 , K-means clustering revealed a symmetrical, interlaced architecture of functional networks within the tri-layer WM functional network (Fig. 2 ): superficial, middle, and deep layers. The WM network-tract correspondences are shown in Supplementary Table 1. Between-group differences of the FC and FCC in WM functional networks Compared with HCs, ESRD patients exhibited widespread reductions in FC among 11 WM functional networks (Fig. 3 a and Supplementary Table 2, p < 0.05, FDR corrected). ESRD patients also exhibited widespread reductions in FCC among WM functional networks (Fig. 3 b and Supplementary Table 3, p < 0.05, FDR corrected). In addition, compared with HCs, ESRD patients also exhibited localized increased in FCC among WM functional networks ( p < 0.05, FDR corrected), including the middle temporal network and occipital network, as well as between cerebellar network and occipital network. For details, see Supplementary Materals. Within-group GC patterns in WM functional networks Within-group patterns of influence among WM functional networks in ESRD patients and HCs were identified using one-sample t -tests (Fig. 4 a and b, p < 0.05, FDR-corrected). Greater total excitatory and inhibitory influence strengths were observed in middle and deep networks relative to superficial networks, which partially aligns with previously reported unique characteristics of middle and deep networks 20 . In the Circos plot for ESRD patients, effective connections exhibited a sparser pattern compared with HCs (Fig. 4 a and b). Between-group differences in GC patterns in WM functional networks Differences in GC patterns are summarized as the following tri-layer network-level findings (Fig. 4 c and Supplementary Table 4, p < 0.05, FDR-corrected). Compared with HCs, in excitatory interaction difference patterns, ESRD patients showed significantly lower influence from the middle→superficial network(corona radiate→frontal; corona radiate→pre/post-central; corona radiate→cerebellar), superficial→deep network (middle temporal→deep), superficial→middle network (middle temporal→corona radiate), superficial→superficial network (middle temporal→pre/post-central; middle temporal→cerebellar; middle temporal→occipital; middle temporal→orbitofrontal; frontoparietal→occipital), and deep→superficial network (deep→occipital). In inhibitory interaction differences pattern, compared with HCs, ESRD showed significantly greater influence from the superficial→deep network (orbitofrontal→deep), superficial→middle network (pre/post-central→corona radiate), and superficial→superficial network (pre/post-central→middle temporal; orbitofrontal→frontal; orbitofrontal→pre/post-central; cerebellar→middle temporal). As shown in Fig. 4 d and Supplementary Table 5, compared with HCs, ESRD patients exhibited significantly lower in-strength in the superficial networks ( t = -2.490, p = 0.014), middle networks ( t = -2.385, p = 0.028), and deep networks ( t = -1.985, p = 0.049). Additionally, the out-strength of the superficial networks was significantly lower in ESRD patients relative to HCs ( t = -2.557, p = 0.012) . Correlation analyses As shown in Fig. 5 , we found that neuropsychological scores were significantly correlated with GC strength of WM network interactions in ESRD patients ( p < 0.05, FDR corrected). Specifically, the greater the cognitive deficits and mood disorder (anxiety and depression), the weaker the excitatory influences (corona radiate→pre/post-central network, middle temporal→occipital network, and middle temporal→orbitofrontal network) and the stronger the inhibitory influences (orbitofrontal→deep network and pre/post-central→corona radiate network). Similarly, the clinical indictors in ESRD patients were significantly correlated with GC strength of WM network interactions. Specifically, the lower the haemoglobin level and the higher the parathormone level, the weaker the excitatory influences (middle temporal→corona radiate and middle temporal→deep network). The higher the cystatin C level, the stronger the inhibitory influences (pre/post-central→middle temporal network and pre/post-central→corona radiate network). The clinical and neuropsychological variables associated with FC and FCC alterations in ESRD patients were shown in Supplementary Materials. Feature selection and classification based on ESRD-related WM functional network alterations Eighty-nine WM feature metrics from between-group differences of the WM functional networks as input into six classifiers to develop the corresponding classification models. We used the leave-one-out (LOOCV) cross-validation to obtain the accuracy, sensitivity, specificity, and area under the curve (AUC) of the seven models. Among them, the accuracy and AUC of random forest were 95.31% and 0.982, respectively, which was better than those of the other classifier algorithms (Fig. 6 a, Supplementary Table 6). Figure 6 b and c revealed features correlation heatmap and weights map. Figure 6 d revealed the permutation test of random forest, while the permutation test results for the remaining five classification models were shown in Supplementary Fig. 4. The top five contributing features were the FC of the deep-deep network, FC of the deep-frontoparietal network, FC of the corona radiate-middle temporal network, FCC of the cerebellar-orbitofrontal network, and FC of the corona radiate-frontal network. Validation analyses The results of validation analyses supported the robustness of our findings. Firstly, both the main results of 70% and 80% group-level masks were consistent with the main results of 60% masks (see Supplementary Fig. 5). Thus, the main results were still stable even with stricter masks. Secondly, no significant differences in the mean FD values between-group ( t = -1.12, p = 0.904, Supplementary Fig. 6). Discussion To our knowledge, this is the first study to reveal WM functional networks dysfunction in ESRD patients and its association with clinical phenotype and cognitive deficits. We delineated 11 distinct WM functional networks organized into three hierarchical layers (superficial, middle, and deep) - across both ESRD patients and HCs based on resting-state correlation matrices. At both the network and hierarchical levels, we observed widespread reductions in FC and FCC within these WM functional networks in ESRD patients, alongside decreased excitatory influence and increased inhibitory influence. These alterations were mainly localized to the corona radiata, middle temporal, frontal, precentral/postcentral, and deep networks, which were associated with multiple cognitive deficits. Additionally, serum creatinine, hemoglobin, cystatin C, parathormone, and calcium levels were associated with WM functional networks dysfunction in ESRD patients. Machine learning classification results revealed that the correlations among three hierarchical layers of WM functional networks could be used to discriminate ESRD patients from HCs. Consistent with our hypothesis, ESRD patients exhibited extensively disrupted interactions among WM functional networks, which correlated with cognitive deficits and ESRD-specific clinical risk factors, including uremic toxin accumulation, dysregulation of calcium-phosphorus homeostasis, and anemia. Our results emphasized the imbalances in the WM functional networks in ESRD patients, which might be used as potential neuroimaging markers for clinical symptoms. In the present study, a total of 11 stable WM functional networks were identified, consistent with previous research 16 and further validating the feasibility of investigating WM networks in ESRD. Relative to HCs, ESRD patients demonstrated extensive decreases in both FC and FCC across WM functional networks. More critically, ESRD patients exhibited diminished excitatory influence and enhanced inhibitory influence among several critical WM networks, including the corona radiata→frontal networks, middle temporal→corona radiata networks, and pre/post-central→corona radiata networks. These alterations were associated with ESRD patients' performance across global cognition, short- and long-term memory, attention, and mood disorder. The corona radiata is a critical WM structure, defined by a fan-shaped array of projection fibers radiating from the cerebral cortex to subcortical regions. As a pivotal hub for ascending and descending pathways, it forms bidirectional communication networks integral to sustaining global brain connectivity 25 . Moreover, the corona radiata plays a central role in integrating motor, sensory, higher-order cognitive, and executive functions 26 , 27 . The middle temporal network is primarily composed of the uncinate fasciculus and temporal U-Fibers. The uncinate fasciculus, connecting the orbitofrontal cortex to the medial temporal lobe, underpins emotional regulation, episodic memory consolidation, and social cognition 28 . Meanwhile, temporal U-Fibers are posited to expedite visual-semantic integration along the temporal lobe, enabling real-time alignment of sensory inputs with semantic knowledge stores. Previous studies have revealed WM demyelination and microstructural abnormalities in ESRD patients with cognitive deficits 2 – 4 , 29 , including damage to association fibers, projection fibers, and commissural fibers such as the corona radiata, corpus callosum, superior longitudinal fasciculus, and corticospinal tract. Notably, from the perspective of WM functional networks, this study identified multiple lines of evidence demonstrating widespread connectivity disruptions in WM functional networks among ESRD patients, as well as their close association with cognitive deficits. These findings offer novel insights into elucidating the underlying mechanisms driving cognitive deficits in ESRD patients. Network hierarchy has been widely known as a key principle of human brain organization. Anatomically, superficial WM tracts connect distant cortical neuronal cell bodies with distinct functions, while middle and deep tracts are less encased by gray matter 15 . Functionally, superficial WM networks and cortical gray-matter networks show synchronous neural activity, whereas middle and deep WM networks exhibit barely any correlation with gray-matter networks 17 . Thus, superficial WM networks likely interact indirectly via gray-matter networks, while middle and deep ones tend to communicate directly through axon-to-axon interactions 16 . From the perspective of hierarchical interactions in the WM functional network, the HCs in our study exhibited significant and consistent strengths of excitatory and inhibitory influences in the middle and deep networks, suggesting that the middle and deep networks play a crucial role in WM function. However, ESRD patients showed weaker within-group interaction effects, indicating that the middle and deep networks were disrupted in their functionality. More importantly, we found that ESRD patients had reduced bidirectional excitatory influences in both superficial-middle and superficial-deep networks, along with enhanced downward inhibitory influences from the superficial to the middle network and from the superficial to the deep network. Additionally, there was a reduction in excitatory influence and an enhancement in inhibitory influence within the intrinsic interactions of superficial networks in ESRD patients. Previous neuroimaging studies have identified abnormal changes in gray matter functional networks in patients with ESRD 8 – 11 , 30 , predominantly involving the default mode network and sensorimotor network. These abnormalities are closely linked to multidimensional cognitive deficits and sensorimotor disorder. Notably, we explored the potential pathogenic mechanisms underlying cognitive deficits in ESRD patients from the perspective of hierarchical interaction dysregulation in WM functional networks. The underlying brain injury mechanism of cognitive deficits in ESRD patients remains unclear 5 . Cerebral WM is a preferentially vulnerable brain region in patients with ESRD 2 , 31 , 32 , susceptible to various clinical risk factors associated with the disease itself and dialysis treatment 5 —including calcium-phosphate metabolic disorders, renal anemia, uremic toxins, dialysis adequacy, and dialysis vintage. Anemia is one of the most common complications in patients with ESRD, characterized by reduced haemoglobin and hematocrit levels, stemming from inadequate erythropoietin (due to impaired renal function) or factors like iron metabolism disorders, inflammation, and nutrient deficiencies 33 . In patients with ESRD, reduced hemoglobin and hematocrit levels directly diminish cerebral oxygen supply and trigger subsequent neuroinflammatory responses 5 , 34 . Long-standing anemia can exert detrimental effects on neurons and myelin sheaths, leading to brain atrophy and white matter hyperintensities. The present study revealed that anemia in ESRD patients was closely linked to a weakened descending excitatory influence from the middle temporal network to the corona radiata network. This association suggests that correcting anemia in ESRD patients may hold potential value in preserving the normal interactive functions of networks such as the middle temporal network and corona radiata network, as well as in improving cognitive function. We also found that elevated parathormone in ESRD patients is closely associated with reduced descending excitatory influence from the middle temporal network to the deep network, while calcium-phosphate metabolism disorders are related to decreased FCC strength in the pre/postcentral-cerebellar network and the corona radiata-deep network. Calcium-phosphate imbalance and elevated parathormone activate L-type calcium channels in neurons 35 , leading to intracellular calcium overload and excitotoxicity, which impairs synaptic plasticity and disrupts cortico-subcortical connectivity 36 , 37 . As a small molecule water-soluble uremic toxin 38 , serum creatinine level in ESRD patients was significantly negnitively correlated with FCC strength between middle temporal and deep network, indicating the role of uremic toxin in ESRD-related WM functional network dysfunction. For patients with ESRD, dialysis stands as a pivotal life-sustaining intervention. Nevertheless, long-term dialysis elicits cerebral hemodynamic fluctuations and even precipitate intradialytic hypotension 39 , thereby inducing hypoxia-ischemia in cerebral WM tracts 40 —most notably within deep WM regions 3 , which are inherently vulnerable owing to their reliance on end-arteries with sparse collaterals. This pathogenic cascade activates microglia 41 , thereby triggering myelin impairment and axonal degeneration. Accordingly, our study revealed that reduced FC between the middle temporal network and orbitofrontal network in ESRD patients exhibits a significant negative correlation with dialysis vintage. This finding implicates dialysis-associated disruptions in WM network connectivity among ESRD patients, though whether this phenomenon can serve as a potential biomarker for monitoring dialysis-induced cerebral WM injury remains to be further explored. Moreover, various machine classification models were used to capture the WM functional netwrok patterns information underlying ESRD. Based on between-group differences in WM functional networks, most interaction features with high discriminative power were mainly located in deep-deep network, deep-superficial network, and middle-superficial network. Deep WM constitutes a critical vulnerability region for WM lesions in patients with ESRD 2 , 42 – 45 . The deep WM functional networks encompasses the posterior thalamic radiation 43 , internal capsule 2 , inferior longitudinal fasciculus 2 , 44 , and superior longitudinal fasciculus 45 , has been consistently implicated in prior neuroimaging studies, which have identified microstructural disruptions within these tracts in ESRD populations. More importantly, from the perspective of WM functional networks, the present study identified that metrics of dysfunctional interactions among these critical WM functional networks can distinguish ESRD patients from HCs. Our results suggest that the dysfunction of these core WM networks may serve as neurological biomarkers of ESRD. This finding underscores their significance in understanding the functional architecture and hierarchical characteristics of WM that underlie ESRD, along with its associated cognitive and clinical phenotypes. Limitations and future research Several limitations of the present study should be acknowledged. First, we failed to exclude the potential impact of medications on cognition, such as antihypertensive drugs 46 , erythropoietin 47 , and vitamin D 48 , which are commonly prescribed for patients with ESRD. Second, although ESRD patients typically present with frailty 49 and fatigue 50 , the current study did not assess the severity of these symptoms. Such variables deserve to be documented in future research to explore their effects on cognition and WM functional networks in ESRD patients. Third, we only collected uremic toxins routinely measured in clinical practice for ESRD patients, such as creatinine and urea. However, other uremic toxins, including p-cresyl sulfate, indoxyl sulfate, and methylglyoxal with direct neurotoxic effects in ESRD patients 51 . Future studies should prioritize investigating the potential relationships between these uremic toxins and WM functional networks. Conclusions ESRD patients exhibited extensively disrupted interactions among WM functional networks, which correlated with cognitive deficits and ESRD-specific clinical risk factors, including uremic toxin accumulation, dysregulation of calcium-phosphorus homeostasis, and anemia. This finding underscores their significance in understanding the functional architecture and hierarchical characteristics of WM that underlie ESRD, along with its associated cognitive and clinical phenotypes. Materials and Methods Participants This study was approved by the local Research Ethics Committee (Approval No. 2020G64) and registered at ClinicalTrials.gov (https://clinicaltrials.gov/ct2/show/NCT03191409). All procedures adhered to the Declaration of Helsinki, and written informed consent was obtained from all participants. Between July 2020 and June 2024, 78 right-handed ESRD patients (53 males, 25 females; mean age: 35 ± 10.18 years) undergoing maintenance hemodialysis were enrolled (Table 1). All patients had a dialysis duration of >3 months. Exclusion criteria included: (1) age 60 years; (2) brain lesions (hemorrhage, head trauma, tumor, stroke, or encephalomalacia) confirmed by conventional MRI or medical history; (3) psychiatric or neurodegenerative disorders; (4) diabetic nephropathy or lupus nephritis; (5) Hepatitis B, hepatitis C, syphilis, or acquired immunodeficiency syndrome; (6) smoking, alcohol, or drug abuse; (7) visual/auditory impairments (blurred vision, hearing loss, or other symptoms precluding neuropsychological assessment); and (8) claustrophobia or MR contraindications. Fifty right-handed HCs demographically matched to patients were recruited from the local community via advertisements (38 males, 12 females; mean age: 35 ± 8.56 years). HCs inclusion criteria: age 18–60 years with no history of neurologic, systemic, or psychiatric disorders. Clinical characteristics and neuropsychological assessment Demographic data and clinical parameters were extracted from medical records. The underlying etiologies of ESRD were as follows: glomerulonephritis (n=58), immunoglobulin A nephropathy (n=12), and membranous nephropathy (n=8). All patients underwent maintenance hemodialysis three times weekly with a short midweek interval, lasting approximately 4 hours per session. Hemodialysis adequacy was quantified by a mean kinetic urea clearance/volume ratio (Kt/V) >1.2 52 . On the day prior to hemodialysis, patients completed neuropsychological testing in a quiet environment, followed by blood sampling. Cognitive assessments included the following: the Auditory Verbal Learning Test-Huashan Version (AVLT-H), which assesses short- and long-term memory; the Montreal Cognitive Assessment (MoCA), which evaluates global cognition, visuospatial function, attention, orientation, language, abstract thinking, and executive function; and the Trail-Making Test (TMT), which measures visual attention ability. Mood disorders were evaluated using the Beck Depression Inventory (BDI) and Beck Anxiety Inventory (BAI). MRI data acquisition and preprocessing MRI data were acquired using a Discovery MR750 3.0 T scanner with an eight-channel phased-array head coil. All participants underwent three sequences: T1-weighted fluid-attenuated inversion recovery (T1-FLAIR), high-resolution T1-weighted structural imaging, and rs-fMRI. Participants were instructed to lie quietly with eyes closed, remain awake, and minimize head movement; foam padding and earplugs were used to reduce motion artifacts and scanner noise. The scan was terminated if participants reported discomfort or inability to complete the procedure, and post-scan verification confirmed cooperation. The rs-fMRI data were preprocessed using the Data Processing Assistant for Resting-State fMRI (DPARSF, Advanced Edition, V4.3; www.restfmri.net), Statistical Parametric Mapping 8 (SPM8; www.fil.ion.ucl.ac.uk/spm), and custom MATLAB scripts 15 (https://mind.huji.ac.il/white-matter.aspx).The T1-weighted anatomical images were segmented into gray matter, WM, and cerebrospinal fluid (CSF) using SPM8’s New Segment algorithm, then normalized to the Montreal Neurological Institute (MNI) template. The first 10 volumes were discarded to account for T1 equilibration effects. Remaining volumes underwent slice-time correction and motion correction (threshold: <3 mm translation or 3° rotation) relative to the mean functional image, followed by coregistration to the corresponding anatomical image. Framewise displacement (FD) was calculated to identify motion "spikes" (FD > 0.5 mm), which were included as separate regressors for data censoring without altering correlation values 17,53,54 . Functional images were detrended to remove linear drifts. Nuisance covariates—including 24 head motion parameters 55 and mean CSF signals—were regressed out. WM and global brain signals were retained to avoid eliminating signals of interest. A temporal band-pass filter (0.01–0.15 Hz) was applied to reduce non-neuronal contributions to BOLD fluctuations 15,16 . For white matter functional images: Individual T1 segmentation images were coregistered to functional space to define WM masks (threshold = 0.5, as validated in previous studies 15,16,20 ). To avoid the mixture of the WM and gray-matter signals, images were spatially smoothed within the WM or gray-matter masks using a 4 mm full-width at half-maximum (FWHM) kernel. Smoothed white matter functional images were normalized to the standard EPI template and resampled to 3 × 3 × 3 mm³ voxels for subsequent analyses. Five ESRD patients with larger mean FD > 0.3 mm were discarded. No controls exceeded the maximum thresholds for translation, rotation, or mean FD. WM functional networks clustering Clustering analysis of rs-fMRI data was adapted from Peer et al 15 . First, a unified group-level WM mask was generated from T1 segmentation results. Voxels with a percentage of participants >60% were identified as the group-level WM mask. Deep brain structures 56,57 —including the thalamus, caudate, putamen, globus pallidus, and nucleus accumbens (defined using the Harvard-Oxford Atlas 58 )—were excluded to refine the mask. The final group WM mask (17064 voxels) was coregistered to functional space and resampled to match rs-fMRI voxel dimensions. For each subject, Pearson correlation coefficients were computed between every WM voxel and all voxels in the group mask. To reduce computational load, the mask was subsampled to 4245 nodes using an interchanging grid strategy 59,60 . Individual matrices were averaged to generate a group-level mean correlation matrix. K-means clustering was applied to the group matrix, testing cluster counts from 2 to 22 59,61 . Stability was evaluated via four-fold cross-validation: the connectivity matrix was randomly split into four subsets (1061 voxels per fold), clustering was performed on each fold, and similarity between results was quantified using Dice's coefficient for adjacency matrices 15,19 . The optimal cluster number was selected based on the highest stability and fine-grained network differentiation. According to the Dice's coefficient of each number of clusters, the most stable and detailed WM functional network number was 11 (see Supplementary Figure 1). Similar to past studies 15,16,20 , Each WM functional network was co-registered to the JHU White Matter Tractography Atlas (20 tracts) and the ICBM DTI workgroup (48 tracts) to validate anatomical correspondence via spatial overlay. The WM functional networks were subsequently classified into superficial, middle, and deep layers. FC and FCC of WM networks We first used conventional FC to quantify the relationship between WM functional networks. For each participant, Pearson's correlation coefficient was computed between the average time courses of all pairs of WM functional networks, then transformed to Fisher z-scores. In addition, we employed FCC 19,22 , a method based on the "correlation of correlations" framework, to estimate covariant relationships between WM functional networks by leveraging their correlations with multiple gray matter regions. First, 96 gray matter regions were defined using the Harvard-Oxford Gray Matter Atlas 58 . A gray matter mask was applied to restrict analysis to gray matter voxels, and time series were extracted from voxels in the intersection of this mask and the Harvard-Oxford atlas. Second, Pearson correlations were computed between each WM network and each gray matter region, yielding a K × 96 correlation matrix (where K denotes the number of WM networks). Third, FCC between all pairs of WM networks was estimated to generate a K × K FCC matrix; FCC values were subsequently transformed to Fisher z-scores. Coefficient GCA To quantify signed and directional influences among the 11WM functional networks, we performed bivariate cGCA 24 , accounting for region-specific hemodynamic delays in WM functional networks. Unlike delay correlation methods that quantify temporal dependencies, cGCA leverages a multivariate autoregressive model to estimate direct causal effects—discerning excitatory/inhibitory influences and making it ideal for mapping signed, directional interactions in WM functional networks. For each functional network, individual preprocessed BOLD time series were extracted by averaging time series across all voxels within the network. Granger causal strength between networks was computed using the REST toolbox (v1.8; www.restfmri.net ). Consistent with previous studies 16,20,23 , the signed strength and direction of pairwise network relationships were characterized by regression coefficients: positive/negative coefficients indicated excitatory/inhibitory pathways (i.e., source network activity predicted subsequent increases/decreases in target network activity). A directed asymmetric matrix (11×11 regression coefficient matrix) was generated for each participant. Graph-theoretic metrics for cGCA were used to quantify incoming and outgoing influence strengths of each network 24 : 1) In-strength: Sum of absolute regression coefficients representing incoming connections, where the network acts as a target (significantly predicted by other networks). 2) Out-strength: Sum of absolute regression coefficients representing outgoing connections, where the network acts as a source (significantly predicts other networks). Statistics analyses Demographic, clinical characteristics, and neuropsychological variables The between-group differences in demographic characteristics (age and education level) and neuropsychological variables were compared using the Kolmogorov–Smirnov test (for normality), Levene’s test (for equality of variances), and independent two-sample t -tests (for equality of means) via the Statistical Package for the Social Sciences (SPSS Statistics, version 22.0; IBM, Armonk, NY). A Chi-square test examined group differences in sex distribution. Multiple linear regression analyses adjusted for the effects of age, sex, education level, and mood disorders on between-group in neuropsychological results. Network statistical analysis To examine within-group patterns of FC and FCC in WM functional networks, we performed one-sample t -tests on correlation coefficient matrices across all participants, respectively. Group differences in network-level patterns were assessed using two-sample t -tests, with age, sex, mean FD, and Euclidean distance between each pair of networks included as confounding variables 62 . Statistical significance was defined as p < 0.05 after FDR correction for multiple comparisons. To verify whether edges between network pairs differed from random results, we conducted a permutation test 63 (5000 iterations) by randomly assigning participants to two groups matching the sample sizes of the original data, respectively, followed by t -tests with the same set of confounding variables. Permutation p -values were calculated as the proportion of permutations where the group difference exceeded the observed difference from the original t -test. The within-group GC patterns of WM functional networks were assessed for each directed edge across subjects in each group with one-sample t -test. The network-level between-group difference patterns for each directed edge were obtained using two-sample t -tests. Age, sex, mean FD, and Euclidean distance between each two networks were controlled for as confounding variables 62 ( p < 0.05, FDR correction). A permutation test of difference distribution (5000 iterations) was performed with the same confounding variables ( p < 0.05, FDR corrected). The 11 WM functional networks were categorized as superficial, middle, and deep network layers. The in/out strength of tri-layer networks in the ESRD patients and HCs were compared via two-sample t -tests ( p < 0.05, FDR corrected). Correlation analyses Pearson or Spearman correlation analyses were used to assess relationships between alterations in WM functional network metrics (FC, FCC, and GC patterns) and both clinical indicators and neuropsychological variables. Statistical significance was defined as p < 0.05 following FDR correction. Feature extraction and machine learning classification This study employed six commonly used classifiers implemented in Python 3.11 (sklearn), namely k-nearest neighbor, logistic regression, support vector machine, adaptive boost, categorical naive bayes, and random forest models, to discern whether between-group alterations in WM functional network metrics could effectively distinguish ESRD patients from HCs. A leave-one-out (LOOCV) cross-validation strategy was rigorously adopted to ensure robust estimation of classifier generalizability 64,65 . To assess the statistical significance of each classification model's performance, a permutation test was executed, wherein the label vectors of all participants were randomly shuffled 5,000 times. For each permutation, the identical training and testing protocol was meticulously applied. Following 5,000 permutations, null distributions were derived for each performance metric—encompassing accuracy, sensitivity, specificity, and AUC with 95% confidence intervals. The p -value for each metric was computed by dividing the number of permutations yielding values exceeding the non-permuted model's actual result by the total number of permutations, with statistical significance defined as p < 0.05. Finally, the discriminative weight values assigned to each feature were leveraged to quantify the relative contribution of individual features to the classification process. Validation analyses Several validation analyses were conducted to verify the robustness of our results. First, considering that potential leakage of gray matter signals might affect the clustering pattern of WM signals, we applied two relatively stricter masks (with thresholds set at >70% and >80%) to define the group-level WM mask. We then re-conducted WM network clustering and functional metric analyses using these stricter masks to verify the consistency of our primary findings. Second, we calculated and compared the mean FD values between-group. Declarations Acknowledgments The authors would like to thank all the participants and their families who supported them. Authors contributions WW and MZ drafted the work and revised it critically for important intellectual content. PL. and YXS. made substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work. JYM, XYZ, ZYL, QGZ, HJY, TG, and CXW revised the article and interpreted the data. All authors approved the version submitted for publication. Data availability The discovery data that support the findings are publicly available in the study’s Open Science Framework repository (https://osf.io/qkbj5/). Code availability Main analysis codes are publicly available in the study’s Open Science Framework repository (https://osf.io/qkbj5/). Competing interests All authors report no biomedical financial interests or potential conflicts of interest. Funding This work is supported by National Natural Science Foundation of China (Grant No. 82202121), the Hovering Program of Fourth Military Medical University (axjhww), the Talent Foundation of Tangdu Hospital (2018BJ003), 7T MRI Precision Neurology Platform of Shaanxi Province (2025PT-08) and Innovative Team for Early Warning and Rehabilitation of Mental Fatigue, the Health Research and Innovation Capacity Strengthening Platform Program of Shaanxi Province (Grant No.2023PT-09), the Science and Technology Research Project of Shaanxi Nuclear Industry Group Co., Ltd (Grant No. 61240302), the Clinical Research Award of the First Affiliated Hospital of Xi’an Jiaotong University (No. XJTU1AF-CRF-2023-021). References Collaboration, G.B.D.C.K.D.: Global, regional, and national burden of chronic kidney disease, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 395 , 709–733 (2020). 10.1016/S0140-6736(20)30045-3 Mu, J., et al.: Altered white matter microstructure mediates the relationship between hemoglobin levels and cognitive control deficits in end-stage renal disease patients. Hum. Brain Mapp. 39 , 4766–4775 (2018). 10.1002/hbm.24321 Mu, J., et al.: Neurological effects of hemodialysis on white matter microstructure in end-stage renal disease. Neuroimage Clin. 31 , 102743 (2021). 10.1016/j.nicl.2021.102743 Zheng, J., et al.: Brain Micro-Structural and Functional Alterations for Cognitive Function Prediction in the End-Stage Renal Disease Patients Undergoing Maintenance Hemodialysis. Acad. Radiol. 30 , 1047–1055 (2023). 10.1016/j.acra.2022.06.019 Viggiano, D., et al.: Mechanisms of cognitive dysfunction in CKD. Nat. Rev. Nephrol. 16 , 452–469 (2020). 10.1038/s41581-020-0266-9 Capasso, G., et al.: Drivers and mechanisms of cognitive decline in chronic kidney disease. Nat. Rev. Nephrol. (2025). 10.1038/s41581-025-00963-0 Bobot, M., Burtey, S.: New insights into mechanisms underlying cognitive impairment in chronic kidney disease. Kidney Int. 106 , 1020–1022 (2024). 10.1016/j.kint.2024.09.008 Ni, L., et al.: Aberrant default-mode functional connectivity in patients with end-stage renal disease: a resting-state functional MR imaging study. Radiology. 271 , 543–552 (2014). 10.1148/radiol.13130816 Luo, S., et al.: Abnormal Intrinsic Brain Activity Patterns in Patients with End-Stage Renal Disease Undergoing Peritoneal Dialysis: A Resting-State Functional MR Imaging Study. Radiology. 278 , 181–189 (2016). 10.1148/radiol.2015141913 Ding, D., et al.: The relationship between putamen-SMA functional connectivity and sensorimotor abnormality in ESRD patients. Brain Imaging Behav. 12 , 1346–1354 (2018). 10.1007/s11682-017-9808-6 Li, P., et al.: Brain connectome gradient dysfunction in patients with end-stage renal disease and its association with clinical phenotype and cognitive deficits. Commun. Biol. 8 , 701 (2025). 10.1038/s42003-025-08132-6 Makedonov, I., Chen, J.J., Masellis, M., MacIntosh, B.J.: Physiological fluctuations in white matter are increased in Alzheimer's disease and correlate with neuroimaging and cognitive biomarkers. Neurobiol. Aging. 37 , 12–18 (2016). 10.1016/j.neurobiolaging.2015.09.010 Gawryluk, J.R., Mazerolle, E.L., D'Arcy, R.C.: N. Does functional MRI detect activation in white matter? A review of emerging evidence, issues, and future directions. Front. Neurosci. 8 , 239 (2014). 10.3389/fnins.2014.00239 Fabri, M., Polonara, G.: Functional topography of human corpus callosum: an FMRI mapping study. Neural Plast 251308, (2013). 10.1155/2013/251308 (2013) Peer, M., Nitzan, M., Bick, A.S., Levin, N., Arzy, S.: Evidence for Functional Networks within the Human Brain's White Matter. J. Neurosci. 37 , 6394–6407 (2017). 10.1523/JNEUROSCI.3872-16.2017 Fan, Y.-S., et al.: Impaired interactions among white-matter functional networks in antipsychotic-naive first-episode schizophrenia. Hum. Brain Mapp. 41 , 230–240 (2020). 10.1002/hbm.24801 Jiang, Y., et al.: White-matter functional networks changes in patients with schizophrenia. Neuroimage. 190 , 172–181 (2019). 10.1016/j.neuroimage.2018.04.018 Ma, J., et al.: Frequency-dependent white-matter functional network changes associated with cognitive deficits in subcortical vascular cognitive impairment. Neuroimage Clin. 36 , 103245 (2022). 10.1016/j.nicl.2022.103245 Jiang, Y., et al.: Dysfunctional white-matter networks in medicated and unmedicated benign epilepsy with centrotemporal spikes. Hum. Brain Mapp. 40 , 3113–3124 (2019). 10.1002/hbm.24584 Meng, L., et al.: Attenuated brain white matter functional network interactions in Parkinson's disease. Hum. Brain Mapp. 43 , 4567–4579 (2022). 10.1002/hbm.25973 Lu, F., et al.: Superficial white-matter functional networks changes in bipolar disorder patients during depressive episodes. J. Affect. Disord. 289 , 151–159 (2021). 10.1016/j.jad.2021.04.029 Zhang, H., et al.: Topographical Information-Based High-Order Functional Connectivity and Its Application in Abnormality Detection for Mild Cognitive Impairment. J. Alzheimers Dis. 54 , 1095–1112 (2016) Liao, W., et al.: Preservation Effect: Cigarette Smoking Acts on the Dynamic of Influences Among Unifying Neuropsychiatric Triple Networks in Schizophrenia. Schizophr Bull. 45 , 1242–1250 (2019). 10.1093/schbul/sby184 Liao, W., et al.: Small-world directed networks in the human brain: multivariate Granger causality analysis of resting-state fMRI. Neuroimage. 54 , 2683–2694 (2011). 10.1016/j.neuroimage.2010.11.007 Catani, M., de Thiebaut, M.: A diffusion tensor imaging tractography atlas for virtual in vivo dissections. Cortex. 44 , 1105–1132 (2008). 10.1016/j.cortex.2008.05.004 Stave, E.A., et al.: Dimensions of Attention Associated With the Microstructure of Corona Radiata White Matter. J. Child. Neurol. 32 , 458–466 (2017). 10.1177/0883073816685652 Burke, T., et al.: Bilateral anterior corona radiata microstructure organisation relates to impaired social cognition in schizophrenia. Schizophr Res. 262 , 87–94 (2023). 10.1016/j.schres.2023.10.035 Von Heide, D., Skipper, R.J., Klobusicky, L.M., E., Olson, I.R.: Dissecting the uncinate fasciculus: disorders, controversies and a hypothesis. Brain. 136 , 1692–1707 (2013). 10.1093/brain/awt094 Zhang, R., et al.: Reduced white matter integrity and cognitive deficits in maintenance hemodialysis ESRD patients: a diffusion-tensor study. Eur. Radiol. 25 , 661–668 (2015). 10.1007/s00330-014-3466-5 Wang, Y.F., et al.: The gut microbiota-inflammation-brain axis in end-stage renal disease: perspectives from default mode network. Theranostics. 9 , 8171–8181 (2019). 10.7150/thno.35387 Tamura, M.K., et al.: Chronic kidney disease, cerebral blood flow, and white matter volume in hypertensive adults. Neurology. 86 , 1208–1216 (2016). 10.1212/WNL.0000000000002527 Martinez-Vea, A., et al.: Silent cerebral white matter lesions and their relationship with vascular risk factors in middle-aged predialysis patients with CKD. Am. J. Kidney Dis. 47 , 241–250 (2006) van der Veen, P.H., et al.: Hemoglobin, hematocrit, and changes in cerebral blood flow: the Second Manifestations of ARTerial disease-Magnetic Resonance study. Neurobiol. Aging. 36 , 1417–1423 (2015). 10.1016/j.neurobiolaging.2014.12.019 Park, S.E., et al.: Decreased hemoglobin levels, cerebral small-vessel disease, and cortical atrophy: among cognitively normal elderly women and men. Int. Psychogeriatr. 28 , 147–156 (2016). 10.1017/S1041610215000733 Hanna, R.M., Ahdoot, R.S., Kalantar-Zadeh, K., Ghobry, L., Kurtz, I.: Calcium Transport in the Kidney and Disease Processes. Front. Endocrinol. (Lausanne). 12 , 762130 (2021). 10.3389/fendo.2021.762130 Hirasawa, T., et al.: Adverse effects of an active fragment of parathyroid hormone on rat hippocampal organotypic cultures. Br. J. Pharmacol. 129 , 21–28 (2000) Rosner, M.H., Husain-Syed, F., Reis, T., Ronco, C., Vanholder, R.: Uremic encephalopathy. Kidney Int. 101 , 227–241 (2022). 10.1016/j.kint.2021.09.025 De Deyn, P.P., D'Hooge, R., Van Bogaert, P.P., Marescau, B.: Endogenous guanidino compounds as uremic neurotoxins. Kidney Int. Suppl. 78 , S77–S83 (2001) Keane, D.F., et al.: The time of onset of intradialytic hypotension during a hemodialysis session associates with clinical parameters and mortality. Kidney Int. 99 , 1408–1417 (2021). 10.1016/j.kint.2021.01.018 Eldehni, M.T., Odudu, A., McIntyre, C.W.: Randomized clinical trial of dialysate cooling and effects on brain white matter. J. Am. Soc. Nephrol. 26 , 957–965 (2015). 10.1681/ASN.2013101086 Chagas, Y.W., Vaz de Castro, P.A.S.: Simões-E-Silva, A. C. Neuroinflammation in kidney disease and dialysis. Behav. Brain Res. 483 , 115465 (2025). 10.1016/j.bbr.2025.115465 Park, B.S., et al.: The effects of the dialysis on the white matter tracts in patients with end-stage renal disease using differential tractography study. Sci. Rep. 13 , 20064 (2023). 10.1038/s41598-023-47533-7 Pilato, F., Norata, D., Rossi, M.G., Di Lazzaro, V., Calandrelli, R.: Consciousness disturbance in patients with chronic kidney disease: Rare but potentially treatable complication. Clinical and neuroradiological review. Behav. Brain Res. 480 , 115393 (2025). 10.1016/j.bbr.2024.115393 Liu, M., et al.: White Matter Microstructure Changes and Cognitive Impairment in the Progression of Chronic Kidney Disease. Front. Neurosci. 14 , 559117 (2020). 10.3389/fnins.2020.559117 Jiang, Y., et al.: Reduced White Matter Integrity in Patients With End-Stage and Non-end-Stage Chronic Kidney Disease: A Tract-Based Spatial Statistics Study. Front. Hum. Neurosci. 15 , 774236 (2021). 10.3389/fnhum.2021.774236 Birkenhager, W.H., Staessen, J.A.: Antihypertensives for prevention of Alzheimer's disease. Lancet Neurol. 5 , 466–468 (2006). 10.1016/S1474-4422(06)70453-7 Lee, S.T., et al.: Erythropoietin improves memory function with reducing endothelial dysfunction and amyloid-beta burden in Alzheimer's disease models. J. Neurochem. 120 , 115–124 (2012). 10.1111/j.1471-4159.2011.07534.x Latimer, C.S., et al.: Vitamin D prevents cognitive decline and enhances hippocampal synaptic function in aging rats. Proc. Natl. Acad. Sci. U S A. 111 , E4359–4366 (2014). 10.1073/pnas.1404477111 Chan, G.C.-K., et al.: Frailty in patients on dialysis. Kidney Int. 106 , 35–49 (2024). 10.1016/j.kint.2024.02.026 Bossola, M., Tazza, L., Postdialysis Fatigue: A Frequent and Debilitating Symptom. Semin Dial. 29 , 222–227 (2016). 10.1111/sdi.12468 Andrews, T.D., Day, G.S., Irani, S.R., Kanekiyo, T., Hickson, L.J.: Uremic Toxins, CKD, and Cognitive Dysfunction. J. Am. Soc. Nephrol. 36 , 1208–1226 (2025). 10.1681/ASN.0000000675 Daugirdas, J.T.: Kt/V (and especially its modifications) remains a useful measure of hemodialysis dose. Kidney Int. 88 , 466–473 (2015). 10.1038/ki.2015.204 Ji, G.-J., Liao, W., Chen, F.-F., Zhang, L., Wang, K.: Low-frequency blood oxygen level-dependent fluctuations in the brain white matter: more than just noise. Sci. Bull. (Beijing). 62 , 656–657 (2017). 10.1016/j.scib.2017.03.021 Power, J.D., Barnes, K.A., Snyder, A.Z., Schlaggar, B.L., Petersen, S.E.: Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. Neuroimage. 59 , 2142–2154 (2012). 10.1016/j.neuroimage.2011.10.018 Friston, K.J., Williams, S., Howard, R., Frackowiak, R.S., Turner, R.: Movement-related effects in fMRI time-series. Magn. Reson. Med. 35 , 346–355 (1996) Wonderlick, J.S., et al.: Reliability of MRI-derived cortical and subcortical morphometric measures: effects of pulse sequence, voxel geometry, and parallel imaging. Neuroimage. 44 , 1324–1333 (2009). 10.1016/j.neuroimage.2008.10.037 Lorio, S., et al.: New tissue priors for improved automated classification of subcortical brain structures on MRI. Neuroimage. 130 , 157–166 (2016). 10.1016/j.neuroimage.2016.01.062 Desikan, R.S., et al.: An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage. 31 , 968–980 (2006) Yeo, B.T.T., et al.: The organization of the human cerebral cortex estimated by intrinsic functional connectivity. J. Neurophysiol. 106 , 1125–1165 (2011). 10.1152/jn.00338.2011 Craddock, R.C., James, G.A., Holtzheimer, P.E., Hu, X.P., Mayberg, H.: S. A whole brain fMRI atlas generated via spatially constrained spectral clustering. Hum. Brain Mapp. 33 , 1914–1928 (2012). 10.1002/hbm.21333 Lange, T., Roth, V., Braun, M.L., Buhmann, J.M.: Stability-based validation of clustering solutions. Neural Comput. 16 , 1299–1323 (2004) Li, J., et al.: Exploring the functional connectome in white matter. Hum. Brain Mapp. 40 , 4331–4344 (2019). 10.1002/hbm.24705 Zhang, Z., et al.: Altered functional-structural coupling of large-scale brain networks in idiopathic generalized epilepsy. Brain. 134 , 2912–2928 (2011). 10.1093/brain/awr223 Varoquaux, G., et al.: Assessing and tuning brain decoders: Cross-validation, caveats, and guidelines. Neuroimage. 145 , 166–179 (2017). 10.1016/j.neuroimage.2016.10.038 Varoquaux, G.: Cross-validation failure: Small sample sizes lead to large error bars. Neuroimage. 180 , 68–77 (2018). 10.1016/j.neuroimage.2017.06.061 Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryMaterials.pdf Supplementary Materials Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7702413","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":529733714,"identity":"33679917-31e3-4b01-8448-cd25c3e1d239","order_by":0,"name":"Ming Zhang","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0002-6546-2762","institution":"The First Affiliated Hospital of Xi'an Jiaotong University","correspondingAuthor":true,"prefix":"","firstName":"Ming","middleName":"","lastName":"Zhang","suffix":""},{"id":529733715,"identity":"7b5c541c-d905-47d5-b0e1-9cbe8c94a7dc","order_by":1,"name":"Peng Li","email":"","orcid":"","institution":"Tangdu Hospital, Fourth Military Medical University","correspondingAuthor":false,"prefix":"","firstName":"Peng","middleName":"","lastName":"Li","suffix":""},{"id":529733716,"identity":"4b6e821f-a5d8-41a0-b347-b93567e9e48e","order_by":2,"name":"Yu-Xuan Shang","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Yu-Xuan","middleName":"","lastName":"Shang","suffix":""},{"id":529733717,"identity":"a5a8bc9d-ebea-4bb3-a4ce-10b98f238cc4","order_by":3,"name":"junya mu","email":"","orcid":"https://orcid.org/0000-0002-5135-1156","institution":"the First Affiliated Hospital of Xi’an Jiaotong University","correspondingAuthor":false,"prefix":"","firstName":"junya","middleName":"","lastName":"mu","suffix":""},{"id":529733718,"identity":"9ab75741-1418-4765-8e27-4865d6ba6c38","order_by":4,"name":"Xinyi Zhu","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Xinyi","middleName":"","lastName":"Zhu","suffix":""},{"id":529733719,"identity":"42f04185-2f61-4dfa-b2b6-edf14041a5c3","order_by":5,"name":"Zhaoyao Luo","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Zhaoyao","middleName":"","lastName":"Luo","suffix":""},{"id":529733720,"identity":"4ba70f3a-5ff2-4c47-88ab-b62777dfacb2","order_by":6,"name":"Qiange Zhu","email":"","orcid":"","institution":"the First Affiliated Hospital of Xi’an Jiaotong University","correspondingAuthor":false,"prefix":"","firstName":"Qiange","middleName":"","lastName":"Zhu","suffix":""},{"id":529733721,"identity":"a139a2e7-c3b0-4adb-9d54-eed5e38b73ee","order_by":7,"name":"Huijie Yuan","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Huijie","middleName":"","lastName":"Yuan","suffix":""},{"id":529733722,"identity":"871af78b-34cc-461d-ac88-d4ffed17a985","order_by":8,"name":"Ting Ge","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Ting","middleName":"","lastName":"Ge","suffix":""},{"id":529733723,"identity":"6922f6e6-bb67-4508-9ae0-b1050b909717","order_by":9,"name":"Chen-Xi Wang","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Chen-Xi","middleName":"","lastName":"Wang","suffix":""},{"id":529733724,"identity":"ae3d8789-2006-44af-91ba-61a04fd9d440","order_by":10,"name":"Wen Wang","email":"","orcid":"https://orcid.org/0000-0001-6473-4888","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Wen","middleName":"","lastName":"Wang","suffix":""}],"badges":[],"createdAt":"2025-09-24 10:11:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7702413/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7702413/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":94709212,"identity":"846259f3-d820-4a32-9d41-5a153a1293e7","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"jpg","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4671777,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/62878719ace1edf0061bf18f.jpg"},{"id":94729713,"identity":"1ac7a168-98a9-4b27-a476-4e933439e3b6","added_by":"auto","created_at":"2025-10-30 07:05:19","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":231950,"visible":true,"origin":"","legend":"","description":"","filename":"ArticleFile.docx","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/a85201ebdcd06dffc6993083.docx"},{"id":94709203,"identity":"bf48d95a-39f0-48d0-aef4-7e26797e46ad","added_by":"auto","created_at":"2025-10-30 01:06:36","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":50724,"visible":true,"origin":"","legend":"","description":"","filename":"Tables.docx","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/3327d03814ab9bb372b40929.docx"},{"id":94709213,"identity":"34b8e277-aef3-4f6d-b4bf-a3b0a4733853","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"jpg","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7483607,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/3c1a8a231062bfcee1f508f1.jpg"},{"id":94728701,"identity":"1ae50408-0639-4941-91c4-f4c31a80a269","added_by":"auto","created_at":"2025-10-30 07:04:10","extension":"jpg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1999618,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/b97d2e6ce0811e56a8f44cb3.jpg"},{"id":94709215,"identity":"07a5cb17-f275-4fff-864a-761aadcb7c59","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"jpg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7133509,"visible":true,"origin":"","legend":"","description":"","filename":"Figure4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/ef2de72ddd6de7f665245ae6.jpg"},{"id":94729427,"identity":"b017ec95-f311-4df5-be02-e24cfff55be1","added_by":"auto","created_at":"2025-10-30 07:04:57","extension":"jpg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1959657,"visible":true,"origin":"","legend":"","description":"","filename":"Figure5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/fa49d157ada68727d739188a.jpg"},{"id":94729734,"identity":"292969c3-9157-4a53-a89b-44fa82ff945d","added_by":"auto","created_at":"2025-10-30 07:05:20","extension":"jpg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4211619,"visible":true,"origin":"","legend":"","description":"","filename":"Figure6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/a52b2fef4da76c0198e637e4.jpg"},{"id":94709220,"identity":"5d32c6af-c04c-45ac-8091-97d766cfb9e5","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"json","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11874,"visible":true,"origin":"","legend":"","description":"","filename":"COMMSBIO259356.json","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/8498a70f8e2b57c9bf8ae2af.json"},{"id":94729907,"identity":"c3609cc5-5c85-432b-99a8-6c771f3932e8","added_by":"auto","created_at":"2025-10-30 07:05:29","extension":"pdf","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":974921,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterials.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/bd00cd8c3b6be76a74706e57.pdf"},{"id":94709225,"identity":"1b515610-c631-4f21-928b-817974520d10","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"xml","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":200950,"visible":true,"origin":"","legend":"","description":"","filename":"COMMSBIO2593560enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/43b4791db5a8ed8429b30a79.xml"},{"id":94729757,"identity":"d2ac18f2-f9a8-489b-a83a-c0efc9ab115d","added_by":"auto","created_at":"2025-10-30 07:05:22","extension":"jpg","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4671777,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/ef32513371ca0016e8be7e2a.jpg"},{"id":94709233,"identity":"147909b9-22d9-492e-97bf-93481ca9b596","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"jpg","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7483607,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/49a6c09cb49804089e0a46f5.jpg"},{"id":94709228,"identity":"1d2503dc-68a6-48fe-8bdc-ad45353f36e9","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"jpg","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1999618,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/9d3236a9112bcb8d0542fc44.jpg"},{"id":94729441,"identity":"32500687-5387-48b9-bde6-aff12a36efeb","added_by":"auto","created_at":"2025-10-30 07:04:58","extension":"jpg","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7133509,"visible":true,"origin":"","legend":"","description":"","filename":"Figure4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/522618a11c53624b7f11bcd0.jpg"},{"id":94729767,"identity":"aa5b820a-929f-4a11-94c8-bb66829424df","added_by":"auto","created_at":"2025-10-30 07:05:23","extension":"jpg","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1959657,"visible":true,"origin":"","legend":"","description":"","filename":"Figure5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/0b411456607827ad0cbd5f73.jpg"},{"id":94729329,"identity":"d60a0859-259b-4a40-9db5-456bd5933942","added_by":"auto","created_at":"2025-10-30 07:04:49","extension":"jpg","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4211619,"visible":true,"origin":"","legend":"","description":"","filename":"Figure6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/c74e4515abe16a9981b72941.jpg"},{"id":94709217,"identity":"13ee24d4-e530-4a63-82b1-5cbaa99ed1b6","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":853690,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure1.png","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/f8e948f4f115bc0c01f493bb.png"},{"id":94709219,"identity":"d65987c8-7164-49a9-9a14-c2f2f0ed8794","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1565408,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure2.png","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/5bf2e12953d57e7905c9d0c5.png"},{"id":94709227,"identity":"49087f2c-2aa5-41bf-8c67-a7cfe93153c2","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":266544,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure3.png","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/a67fd4b17a4881ba3026815f.png"},{"id":94729948,"identity":"fdd507e2-0230-417e-8d08-dc90fb90c8a0","added_by":"auto","created_at":"2025-10-30 07:05:31","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1194888,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure4.png","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/68dc83cd0a417c1085634b44.png"},{"id":94709221,"identity":"727b5767-d4be-4646-a1c0-cac80f0e4d5a","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":260460,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure5.png","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/ea23ad226ef3046504a86cbf.png"},{"id":94709230,"identity":"298a7b82-3750-472b-aa6e-47c89b051dae","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":546684,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure6.png","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/3c01d0c9640b95aa7b9ffd21.png"},{"id":94709231,"identity":"0e6be5df-734d-41cd-89c6-9afe1722a1f8","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"xml","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":198845,"visible":true,"origin":"","legend":"","description":"","filename":"COMMSBIO2593560structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/fb241062784acbad314b5c28.xml"},{"id":94709223,"identity":"a7e8fead-abe4-47f5-8408-b13448091eb3","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"html","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":217071,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/d31be19e995b24548108c880.html"},{"id":94729692,"identity":"2b826b5a-cf3a-48cd-8a16-1817ff776cf1","added_by":"auto","created_at":"2025-10-30 07:05:17","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":3019703,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSchematic flowchart of this study. \u003c/strong\u003eESRD, end-stage renal disease; HCs, healthy controls; rs-fMRI, resting-state functional magnetic resonance imaging; WM, white matter; FC, functional connectivity; FCC, functional covariance connectivity; cGCA, coefficient Granger causality analysis.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/fe2135e7bd6e60fbb211fbca.png"},{"id":94709208,"identity":"6e6d5a8b-8953-4940-9997-73bf4a7f1852","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":3690808,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBrain WM functional networks. \u003c/strong\u003eA total of 11 clusters were identified by K-means clustering algorithm, which can be organized in superficial (frontal, pre/post-central, middle temporal, cerebellum, occipital, orbitofrontal, and frontoparietal network), middle (corona radiate network), and deep (deep networks) layers.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/e28e2ba84bc479bd033a931d.png"},{"id":94709205,"identity":"569aa3f8-0bf7-448e-8635-14ff6a5e05e6","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1485374,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBetween-group differences of FC and FCC among WM functional networks in ESRD patients and HCs (\u003c/strong\u003e\u003cem\u003e\u003cstrong\u003ep\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e \u0026lt; 0.05, FDR-corrected).\u003c/strong\u003e (a) and (b) respectively show the between-group differences in FC and FCC of 11 WM networks. As illustrated, compared with HCs, ESRD patients exhibited widespread reductions in both FC and FCC among 11 WM functional networks. These reductions were primarily observed in the connectivity between the middle temporal network, corona radiata network, and frontal lobe network. ESRD, end-stage renal disease; HCs, healthy controls; WM, white matter; FC, functional connectivity; FCC, functional covariance connectivity.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/b1c5ddf2be861c256b1fe6b6.png"},{"id":94709210,"identity":"c3a4b0e9-4b16-43b5-901f-80d8d2776bb0","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":4277081,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWithin-group patterns of causal influence and between-group differences among WM functional networks in ESRD patients and HCs (\u003c/strong\u003e\u003cem\u003e\u003cstrong\u003ep\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e \u0026lt; 0.05, FDR-corrected)\u003c/strong\u003e. (a) Within-group interactions among 11 WM functional networks in HCs; (b) Within-group interactions among 11 WM functional networks in ESRD patients. (a)-(b): On the left, details of the one-sample \u003cem\u003et\u003c/em\u003e-test matrix are displayed: positive \u003cem\u003et\u003c/em\u003e-values indicate significant excitatory influences, while negative \u003cem\u003et\u003c/em\u003e-values denote significant inhibitory influences; On the right, the Circos plot illustrates these patterns: red lines represent significant excitatory influences, blue lines represent significant inhibitory influences, and line darkness increases with connection strength. (c)-(d), cold lines denote significantly less influence, warm lines denote significantly greater influence; relative to the HCs. (c) Details of the two-sample \u003cem\u003et\u003c/em\u003e-test are shown in the directed connection differences Circos. (d) Interaction pattern diagram: Compared with HCs, in excitatory interaction difference patterns, ESRD patients showed significantly lower influence from the superficial→middle network, superficial→deep network, superficial→superficial network, middle→superficial network, and deep→superficial network; in inhibitory interaction differences pattern, ESRD showed significantly greater influence from the superficial→deep network, superficial→middle network, and superficial→superficial network.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/34ec53c2f665e90e72aeceee.png"},{"id":94729677,"identity":"aa0b60aa-286a-4ae9-9101-249cf7f120b3","added_by":"auto","created_at":"2025-10-30 07:05:16","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1198643,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe associations between GC patterns alterations and clinical and neuropsychological variables in ESRD patients.\u003c/strong\u003e Spearman (§) correlation analyses was used to assess relationships between GC patterns alterations and both clinical indicators and neuropsychological variables. Green/blue circles denote correlations between the GC strength of excitatory/inhibitory interactions and neuropsychological variables, respectively. Purple/yellow circles denote correlations between the GC strength of excitatory/inhibitory interactions and clinical indicators, respectively. The significance threshold was set to \u003cem\u003ep\u003c/em\u003e values of \u0026lt; 0.05 after FDR correction. GC, Granger causality; ESRD, end-stage renal disease; MoCA, Montreal cognitive assessment; MoCA, Montreal cognitive assessment; BAI, Beck anxiety inventory; BDI, Beck depression inventory.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/dfed14891eb466a6d4c85e1e.png"},{"id":94729110,"identity":"c2af9c08-863e-4bb8-942a-6747dd85eb9c","added_by":"auto","created_at":"2025-10-30 07:04:37","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":3409581,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePerformance of machine learning algorithms for classification.\u003c/strong\u003e \u003cstrong\u003ea.\u003c/strong\u003e ROC curve of the classifiers. The accuracy and AUC of random forest were 95.31% and 0.982, respectively, which was better than those of the other classifier algorithms. \u003cstrong\u003eb-c.\u003c/strong\u003e The top 10 contributing features correlation heatmap and weights map of the random forest. \u003cstrong\u003ed.\u003c/strong\u003e Permutation test of the random forest (repeated 5,000 times). ROC: receiver operating characteristic, AUC: area under the curve.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/2bb22f204abdcc2c7581825a.png"},{"id":94822719,"identity":"6a04f25a-3753-4b29-aae8-116bd1fdcf39","added_by":"auto","created_at":"2025-10-31 06:43:06","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":18673417,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/bf12d387-420f-425f-8dd9-8bc811320cd7.pdf"},{"id":94709204,"identity":"7da33a9b-c766-414d-9ab5-98f867e88b06","added_by":"auto","created_at":"2025-10-30 01:06:37","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":974921,"visible":true,"origin":"","legend":"Supplementary Materials","description":"","filename":"SupplementaryMaterials.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7702413/v1/bcc51b9c99c0f0afd4aae57b.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"White-matter functional network dysfunction associated with cognitive deficits and clinical phenotypes in patients with end-stage renal disease","fulltext":[{"header":"Introduction","content":"\u003cp\u003eEnd-stage renal disease (ESRD) has emerged as an increasingly severe global public health issue\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Cerebral white matter (WM) is a preferentially vulnerable region in the brain of patients with ESRD\u003csup\u003e\u003cspan additionalcitationids=\"CR3\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, susceptible to various clinical risk factors including uremic toxins, calcium-phosphate metabolism disorders, renal anemia, and dialysis therapy\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Such patients often exhibit cognitive deficits\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e, which exerts a substantial impact on their self-management capabilities, dietary adherence, treatment compliance, and long-term survival outcomes\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Notwithstanding, the association between WM function and clinical phenotypes in ESRD patients remains largely undefined.\u003c/p\u003e\u003cp\u003eThe rapid advancement of MRI-based neuroimaging has provided critical evidence for exploring the neuropathological mechanisms of cognitive deficits in ESRD patients. Resting-state functional MRI (rs-fMRI) identifies cortical functional abnormalities in ESRD-related cognitive deficits by analyzing blood oxygen level-dependent (BOLD) signals, manifested as reduced spontaneous neural activity and functional connectivity in the default mode network\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e [medial prefrontal cortex, middle temporal gyrus, precuneus, and posterior cingulate gyrus], decreased functional connectivity (FC) in the sensorimotor network\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e, and disruption of large-scale brain connectome functional gradients\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Evidence from diffusion tensor imaging and diffusion kurtosis imaging has also revealed WM demyelination and microstructural abnormalities in ESRD patients with cognitive deficits\u003csup\u003e\u003cspan additionalcitationids=\"CR3\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, including damage to critical projection fibers, association fibers, and commissural fibers such as the corona radiata, corpus callosum, superior longitudinal fasciculus, and corticospinal tract. Notably, prior studies have neither elucidated the existence and characteristics of the WM functional dysfunction in ESRD patients, nor elucidated the complex relationships among WM functional networks, cognitive deficits, and ESRD-related risk factors.\u003c/p\u003e\u003cp\u003eRecent advancements in rs-fMRI have garnered increasing attention for investigating the WM neural activation and functional organization. During the resting-state, BOLD signals in WM encapsulate functional signatures of brain activity and connectivity\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e, which can be differentially evoked during various task paradigms\u0026mdash;including perceptual, linguistic, and motor domains\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Peer et al\u003csup\u003e15\u003c/sup\u003e. further unveiled the intrinsic functional organization of WM, deriving discrete functional network parcellations through unsupervised clustering of BOLD signals. Moreover, recent studies have linked the FC of WM functional networks to pathological mechanisms underlying schizophrenia\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, subcortical vascular cognitive impairment\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e, epilepsy\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, Parkinson's disease\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, and bipolar disorder\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e, highlighting the feasibility of using rs-fMRI to investigate WM networks dysfunctions. Thus, elucidating the interactions among WM functional networks may facilitate the interpretation of neuropathological mechanisms underlying cognitive deficits in patients with ESRD.\u003c/p\u003e\u003cp\u003eTo address these gaps, we collected rs-fMRI and clinical data from 78 ESRD patients and 50 healthy controls (HCs). In addition to traditional FC method, which quantifies direct temporal synchrony between two WM networks, we employed a novel functional covariance connectivity (FCC) approach\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e to estimate covariant relationships between WM networks\u0026mdash;based on their correlations with multiple gray matter regions. We further probed interactions among WM functional networks using coefficient Granger causality analysis (cGCA). To further quantify the overall influence of each WM functional network within the interaction model, we examined the causal in/out strength of each network using classic graph-theoretic metrics\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. Additionally, we employed multiple machine learning classification models to assess the discriminative power of WM functional network features extracted from ESRD patients and HCs. Specifically, we hypothesized that: (i) patients with ESRD would demonstrate aberrant interactions of WM functional networks, as indexed by altered metrics of FC, FCC, and Granger causality (GC) strength; (ii) these WM network dysfunctions would correlate with cognitive deficits and ESRD-specific clinical risk factors, including uremic toxin burden, calcium-phosphorus homeostasis dysregulation, and anemia. Our study linked the WM functional abnormalities with the pathophysiology mechanisms underlying cognitive deficits in ESRD patients and demonstrated that the dysfunctional interactions of WM networks as neurological biomarkers for ESRD patients. The entire experimental workflow and analytical methods are illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eDemographic and neuropsychological results\u003c/h2\u003e\u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents demographic and clinical characteristics for each group, with no significant between-group differences observed in age, sex, or educational level. Relative to HCs, ESRD patients demonstrated significantly poorer performance on multiple neuropsychological assessments, including the MoCA(total score [\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001], visuospatial [\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001], language [\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001], name [\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.023], delayed memory [\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001], abstraction [\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001] ), AVLT-H (immediate recall [\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001], short-term recall [\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01], long-term recall [\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001], recognition [\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001]), TMT (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), BAI (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and BDI (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eDemographic, clinical characteristics and neuropsychological assessment in ESRD patients and HCs\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"6\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariable\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHC(n\u0026thinsp;=\u0026thinsp;50)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003eESRD(n\u0026thinsp;=\u0026thinsp;78)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003et/χ\u0026sup2;\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eAge(years)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e34.76 (8.56)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e35.15 (10.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.227\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.821\u0026sect;\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eSex (M/F)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e38 (12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e53 (25)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.609\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.435\u0026dagger;\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eEducation(years)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e12.32 (2.11)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e12.46 (1.91)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.392\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.696\u0026sect;\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eDialysis vintage (months)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e38.58 (29.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eCreatinine(\u0026micro;mol/L)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e919.06 (218.10)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eUrea(mmol/L)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e23.69 (7.21)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eKt/V\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e1.46 (0.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eHaemoglobin(g/L)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e107.03 (20.31)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eHematocrit(%)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e32.89 (5.98)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eCystatin C(\u0026micro;g/mL)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e6.07 (3.17)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003ePotassium(mmol/L)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e4.82 (0.74)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eSodium(mmol/L)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e142.31 (3.28)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003ePhosphorus(mmol/L)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e1.80 (0.51)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eCalcium(mmol/L)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e2.13 (0.34)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eParathormone(ng/L)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e626.10 (453.38)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eMedication\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e%\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eAT1-blocker\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e68\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e87\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eBeta-blocker\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e56\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e72\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eEPO\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e76\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e97\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eAntidepressants\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eAntihistamines\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eAnalgesics\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eVitamin D\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e42\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e54\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eCalcium antagonists\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e44\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e56\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e/\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eMoCA\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eTotal score\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e27.52(2.23)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e23.85(2.87)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-7.688\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003csup\u003e\u0026sect;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eVisuospatial\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e4(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-5.631\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003csup\u003e\u0026Dagger;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eName\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3(0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e3(0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-2.276\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.023\u003csup\u003e\u0026Dagger;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eAttention\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e6(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-1.827\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.068\u003csup\u003e\u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eLanguage\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e2(0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-5.420\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003csup\u003e\u0026Dagger;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eAbstraction\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2(0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e1(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-5.392\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003csup\u003e\u0026Dagger;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eOrientation\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e6(0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e6(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-1.435\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.151\u003csup\u003e\u0026Dagger;\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eDelayed memory\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-5.410\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003csup\u003e\u0026Dagger;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eAVLT-H\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eIR-S\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e27.84(3.62)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e25.22(4.45)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-3.489\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003csup\u003e\u0026sect;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eSR-S\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e10.30(1.20)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e9.59(1.65)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-2.624\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.010\u003csup\u003e\u0026sect;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eLR-S\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e10.24(1.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e9.00(1.73)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-4.370\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003csup\u003e\u0026sect;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eREC-S\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e11.84(0.47)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e11.32(1.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-3.437\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003csup\u003e\u0026sect;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eTMT\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e41.76(13.26)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e60.34(28.22)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e4.355\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003csup\u003e\u0026Dagger;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eBDI\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e7.90(5.20)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e16.10(10.14)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5.282\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003csup\u003e\u0026sect;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eBAI\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e25.18(2.83)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003e28.60(6.04)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.750\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003csup\u003e\u0026sect;*\u003c/sup\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e\u003cp\u003eNote.\u0026mdash;Unless otherwise indicated, data are mean (standard deviation). \u0026sect; Analyzed with the independent two-sample \u003cem\u003et\u003c/em\u003e-test; data in parentheses have a 95% confidence interval. data are mean (standard deviation). \u0026Dagger; Analyzed with the Mann-Whitney \u003cem\u003eu\u003c/em\u003e-test,data are median (range interquartile).\u0026dagger; Analyzed with the chi-square test. ESRD, end-stage renal disease; HCs, health controls; AVLT-H, auditory verbal learning test\u0026ndash;Huashan version; IR-S, immediate recall score; SR-S, short-term recall score; LR-S, long-term recall score; REC-S, recognition score; MoCA, Montreal cognitive assessment; TMT, trail-making test; BAI, Beck anxiety inventory; BDI, Beck depression inventory. \u003c/p\u003e\u003cp\u003e* Indicates a statistically significant difference after controlling age, sex, and education level.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eWM functional networks\u003c/h3\u003e\n\u003cp\u003eClustering analysis identified k\u0026thinsp;=\u0026thinsp;11 as the optimal cluster number, achieving both fine-grained resolution and high reproducibility (Dice coefficient\u0026thinsp;\u0026gt;\u0026thinsp;0.85, Supplement Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Consequently, these 11 WM functional networks were retained for subsequent analyses. Network nomenclature was assigned based on spatial localization. Consistent with prior literature\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e, K-means clustering revealed a symmetrical, interlaced architecture of functional networks within the tri-layer WM functional network (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e): superficial, middle, and deep layers. The WM network-tract correspondences are shown in Supplementary Table\u0026nbsp;1.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\n\u003ch3\u003eBetween-group differences of the FC and FCC in WM functional networks\u003c/h3\u003e\n\u003cp\u003eCompared with HCs, ESRD patients exhibited widespread reductions in FC among 11 WM functional networks (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea and Supplementary Table\u0026nbsp;2, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05, FDR corrected). ESRD patients also exhibited widespread reductions in FCC among WM functional networks (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb and Supplementary Table\u0026nbsp;3, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05, FDR corrected). In addition, compared with HCs, ESRD patients also exhibited localized increased in FCC among WM functional networks (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05, FDR corrected), including the middle temporal network and occipital network, as well as between cerebellar network and occipital network. For details, see Supplementary Materals.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\n\u003ch3\u003eWithin-group GC patterns in WM functional networks\u003c/h3\u003e\n\u003cp\u003eWithin-group patterns of influence among WM functional networks in ESRD patients and HCs were identified using one-sample \u003cem\u003et\u003c/em\u003e-tests (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea and b, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05, FDR-corrected). Greater total excitatory and inhibitory influence strengths were observed in middle and deep networks relative to superficial networks, which partially aligns with previously reported unique characteristics of middle and deep networks\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. In the Circos plot for ESRD patients, effective connections exhibited a sparser pattern compared with HCs (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea and b).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\n\u003ch3\u003eBetween-group differences in GC patterns in WM functional networks\u003c/h3\u003e\n\u003cp\u003eDifferences in GC patterns are summarized as the following tri-layer network-level findings (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec and Supplementary Table\u0026nbsp;4, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05, FDR-corrected). Compared with HCs, in excitatory interaction difference patterns, ESRD patients showed significantly lower influence from the middle\u0026rarr;superficial network(corona radiate\u0026rarr;frontal; corona radiate\u0026rarr;pre/post-central; corona radiate\u0026rarr;cerebellar), superficial\u0026rarr;deep network (middle temporal\u0026rarr;deep), superficial\u0026rarr;middle network (middle temporal\u0026rarr;corona radiate), superficial\u0026rarr;superficial network (middle temporal\u0026rarr;pre/post-central; middle temporal\u0026rarr;cerebellar; middle temporal\u0026rarr;occipital; middle temporal\u0026rarr;orbitofrontal; frontoparietal\u0026rarr;occipital), and deep\u0026rarr;superficial network (deep\u0026rarr;occipital).\u003c/p\u003e\u003cp\u003eIn inhibitory interaction differences pattern, compared with HCs, ESRD showed significantly greater influence from the superficial\u0026rarr;deep network (orbitofrontal\u0026rarr;deep), superficial\u0026rarr;middle network (pre/post-central\u0026rarr;corona radiate), and superficial\u0026rarr;superficial network (pre/post-central\u0026rarr;middle temporal; orbitofrontal\u0026rarr;frontal; orbitofrontal\u0026rarr;pre/post-central; cerebellar\u0026rarr;middle temporal).\u003c/p\u003e\u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ed and Supplementary Table\u0026nbsp;5, compared with HCs, ESRD patients exhibited significantly lower in-strength in the superficial networks (\u003cem\u003et\u003c/em\u003e = -2.490, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.014), middle networks (\u003cem\u003et\u003c/em\u003e = -2.385, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.028), and deep networks (\u003cem\u003et\u003c/em\u003e = -1.985, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.049). Additionally, the out-strength of the superficial networks was significantly lower in ESRD patients relative to HCs (\u003cem\u003et\u003c/em\u003e = -2.557, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.012) .\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eCorrelation analyses\u003c/h2\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, we found that neuropsychological scores were significantly correlated with GC strength of WM network interactions in ESRD patients (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05, FDR corrected). Specifically, the greater the cognitive deficits and mood disorder (anxiety and depression), the weaker the excitatory influences (corona radiate\u0026rarr;pre/post-central network, middle temporal\u0026rarr;occipital network, and middle temporal\u0026rarr;orbitofrontal network) and the stronger the inhibitory influences (orbitofrontal\u0026rarr;deep network and pre/post-central\u0026rarr;corona radiate network). Similarly, the clinical indictors in ESRD patients were significantly correlated with GC strength of WM network interactions. Specifically, the lower the haemoglobin level and the higher the parathormone level, the weaker the excitatory influences (middle temporal\u0026rarr;corona radiate and middle temporal\u0026rarr;deep network). The higher the cystatin C level, the stronger the inhibitory influences (pre/post-central\u0026rarr;middle temporal network and pre/post-central\u0026rarr;corona radiate network). The clinical and neuropsychological variables associated with FC and FCC alterations in ESRD patients were shown in Supplementary Materials.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eFeature selection and classification based on ESRD-related WM functional network alterations\u003c/h3\u003e\n\u003cp\u003eEighty-nine WM feature metrics from between-group differences of the WM functional networks as input into six classifiers to develop the corresponding classification models. We used the leave-one-out (LOOCV) cross-validation to obtain the accuracy, sensitivity, specificity, and area under the curve (AUC) of the seven models. Among them, the accuracy and AUC of random forest were 95.31% and 0.982, respectively, which was better than those of the other classifier algorithms (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea, Supplementary Table\u0026nbsp;6). Figure\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb and c revealed features correlation heatmap and weights map. Figure\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ed revealed the permutation test of random forest, while the permutation test results for the remaining five classification models were shown in Supplementary Fig.\u0026nbsp;4. The top five contributing features were the FC of the deep-deep network, FC of the deep-frontoparietal network, FC of the corona radiate-middle temporal network, FCC of the cerebellar-orbitofrontal network, and FC of the corona radiate-frontal network.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\n\u003ch3\u003eValidation analyses\u003c/h3\u003e\n\u003cp\u003eThe results of validation analyses supported the robustness of our findings. Firstly, both the main results of 70% and 80% group-level masks were consistent with the main results of 60% masks (see Supplementary Fig.\u0026nbsp;5). Thus, the main results were still stable even with stricter masks. Secondly, no significant differences in the mean FD values between-group (\u003cem\u003et\u003c/em\u003e = -1.12, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.904, Supplementary Fig.\u0026nbsp;6).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eTo our knowledge, this is the first study to reveal WM functional networks dysfunction in ESRD patients and its association with clinical phenotype and cognitive deficits. We delineated 11 distinct WM functional networks organized into three hierarchical layers (superficial, middle, and deep) - across both ESRD patients and HCs based on resting-state correlation matrices. At both the network and hierarchical levels, we observed widespread reductions in FC and FCC within these WM functional networks in ESRD patients, alongside decreased excitatory influence and increased inhibitory influence. These alterations were mainly localized to the corona radiata, middle temporal, frontal, precentral/postcentral, and deep networks, which were associated with multiple cognitive deficits. Additionally, serum creatinine, hemoglobin, cystatin C, parathormone, and calcium levels were associated with WM functional networks dysfunction in ESRD patients. Machine learning classification results revealed that the correlations among three hierarchical layers of WM functional networks could be used to discriminate ESRD patients from HCs. Consistent with our hypothesis, ESRD patients exhibited extensively disrupted interactions among WM functional networks, which correlated with cognitive deficits and ESRD-specific clinical risk factors, including uremic toxin accumulation, dysregulation of calcium-phosphorus homeostasis, and anemia. Our results emphasized the imbalances in the WM functional networks in ESRD patients, which might be used as potential neuroimaging markers for clinical symptoms.\u003c/p\u003e\u003cp\u003eIn the present study, a total of 11 stable WM functional networks were identified, consistent with previous research\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e and further validating the feasibility of investigating WM networks in ESRD. Relative to HCs, ESRD patients demonstrated extensive decreases in both FC and FCC across WM functional networks. More critically, ESRD patients exhibited diminished excitatory influence and enhanced inhibitory influence among several critical WM networks, including the corona radiata\u0026rarr;frontal networks, middle temporal\u0026rarr;corona radiata networks, and pre/post-central\u0026rarr;corona radiata networks. These alterations were associated with ESRD patients' performance across global cognition, short- and long-term memory, attention, and mood disorder. The corona radiata is a critical WM structure, defined by a fan-shaped array of projection fibers radiating from the cerebral cortex to subcortical regions. As a pivotal hub for ascending and descending pathways, it forms bidirectional communication networks integral to sustaining global brain connectivity\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Moreover, the corona radiata plays a central role in integrating motor, sensory, higher-order cognitive, and executive functions\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. The middle temporal network is primarily composed of the uncinate fasciculus and temporal U-Fibers. The uncinate fasciculus, connecting the orbitofrontal cortex to the medial temporal lobe, underpins emotional regulation, episodic memory consolidation, and social cognition\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. Meanwhile, temporal U-Fibers are posited to expedite visual-semantic integration along the temporal lobe, enabling real-time alignment of sensory inputs with semantic knowledge stores. Previous studies have revealed WM demyelination and microstructural abnormalities in ESRD patients with cognitive deficits\u003csup\u003e\u003cspan additionalcitationids=\"CR3\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e, including damage to association fibers, projection fibers, and commissural fibers such as the corona radiata, corpus callosum, superior longitudinal fasciculus, and corticospinal tract. Notably, from the perspective of WM functional networks, this study identified multiple lines of evidence demonstrating widespread connectivity disruptions in WM functional networks among ESRD patients, as well as their close association with cognitive deficits. These findings offer novel insights into elucidating the underlying mechanisms driving cognitive deficits in ESRD patients.\u003c/p\u003e\u003cp\u003eNetwork hierarchy has been widely known as a key principle of human brain organization. Anatomically, superficial WM tracts connect distant cortical neuronal cell bodies with distinct functions, while middle and deep tracts are less encased by gray matter\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Functionally, superficial WM networks and cortical gray-matter networks show synchronous neural activity, whereas middle and deep WM networks exhibit barely any correlation with gray-matter networks\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Thus, superficial WM networks likely interact indirectly via gray-matter networks, while middle and deep ones tend to communicate directly through axon-to-axon interactions\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. From the perspective of hierarchical interactions in the WM functional network, the HCs in our study exhibited significant and consistent strengths of excitatory and inhibitory influences in the middle and deep networks, suggesting that the middle and deep networks play a crucial role in WM function. However, ESRD patients showed weaker within-group interaction effects, indicating that the middle and deep networks were disrupted in their functionality. More importantly, we found that ESRD patients had reduced bidirectional excitatory influences in both superficial-middle and superficial-deep networks, along with enhanced downward inhibitory influences from the superficial to the middle network and from the superficial to the deep network. Additionally, there was a reduction in excitatory influence and an enhancement in inhibitory influence within the intrinsic interactions of superficial networks in ESRD patients. Previous neuroimaging studies have identified abnormal changes in gray matter functional networks in patients with ESRD\u003csup\u003e\u003cspan additionalcitationids=\"CR9 CR10\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, predominantly involving the default mode network and sensorimotor network. These abnormalities are closely linked to multidimensional cognitive deficits and sensorimotor disorder. Notably, we explored the potential pathogenic mechanisms underlying cognitive deficits in ESRD patients from the perspective of hierarchical interaction dysregulation in WM functional networks.\u003c/p\u003e\u003cp\u003eThe underlying brain injury mechanism of cognitive deficits in ESRD patients remains unclear\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Cerebral WM is a preferentially vulnerable brain region in patients with ESRD\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e, susceptible to various clinical risk factors associated with the disease itself and dialysis treatment\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e\u0026mdash;including calcium-phosphate metabolic disorders, renal anemia, uremic toxins, dialysis adequacy, and dialysis vintage. Anemia is one of the most common complications in patients with ESRD, characterized by reduced haemoglobin and hematocrit levels, stemming from inadequate erythropoietin (due to impaired renal function) or factors like iron metabolism disorders, inflammation, and nutrient deficiencies\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. In patients with ESRD, reduced hemoglobin and hematocrit levels directly diminish cerebral oxygen supply and trigger subsequent neuroinflammatory responses\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e. Long-standing anemia can exert detrimental effects on neurons and myelin sheaths, leading to brain atrophy and white matter hyperintensities. The present study revealed that anemia in ESRD patients was closely linked to a weakened descending excitatory influence from the middle temporal network to the corona radiata network. This association suggests that correcting anemia in ESRD patients may hold potential value in preserving the normal interactive functions of networks such as the middle temporal network and corona radiata network, as well as in improving cognitive function. We also found that elevated parathormone in ESRD patients is closely associated with reduced descending excitatory influence from the middle temporal network to the deep network, while calcium-phosphate metabolism disorders are related to decreased FCC strength in the pre/postcentral-cerebellar network and the corona radiata-deep network. Calcium-phosphate imbalance and elevated parathormone activate L-type calcium channels in neurons\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e, leading to intracellular calcium overload and excitotoxicity, which impairs synaptic plasticity and disrupts cortico-subcortical connectivity\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e,\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. As a small molecule water-soluble uremic toxin\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e, serum creatinine level in ESRD patients was significantly negnitively correlated with FCC strength between middle temporal and deep network, indicating the role of uremic toxin in ESRD-related WM functional network dysfunction. For patients with ESRD, dialysis stands as a pivotal life-sustaining intervention. Nevertheless, long-term dialysis elicits cerebral hemodynamic fluctuations and even precipitate intradialytic hypotension\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e, thereby inducing hypoxia-ischemia in cerebral WM tracts\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e\u0026mdash;most notably within deep WM regions\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e, which are inherently vulnerable owing to their reliance on end-arteries with sparse collaterals. This pathogenic cascade activates microglia\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e, thereby triggering myelin impairment and axonal degeneration. Accordingly, our study revealed that reduced FC between the middle temporal network and orbitofrontal network in ESRD patients exhibits a significant negative correlation with dialysis vintage. This finding implicates dialysis-associated disruptions in WM network connectivity among ESRD patients, though whether this phenomenon can serve as a potential biomarker for monitoring dialysis-induced cerebral WM injury remains to be further explored.\u003c/p\u003e\u003cp\u003eMoreover, various machine classification models were used to capture the WM functional netwrok patterns information underlying ESRD. Based on between-group differences in WM functional networks, most interaction features with high discriminative power were mainly located in deep-deep network, deep-superficial network, and middle-superficial network. Deep WM constitutes a critical vulnerability region for WM lesions in patients with ESRD\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan additionalcitationids=\"CR43 CR44\" citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. The deep WM functional networks encompasses the posterior thalamic radiation\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e, internal capsule\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, inferior longitudinal fasciculus\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e, and superior longitudinal fasciculus\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e, has been consistently implicated in prior neuroimaging studies, which have identified microstructural disruptions within these tracts in ESRD populations. More importantly, from the perspective of WM functional networks, the present study identified that metrics of dysfunctional interactions among these critical WM functional networks can distinguish ESRD patients from HCs. Our results suggest that the dysfunction of these core WM networks may serve as neurological biomarkers of ESRD. This finding underscores their significance in understanding the functional architecture and hierarchical characteristics of WM that underlie ESRD, along with its associated cognitive and clinical phenotypes.\u003c/p\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eLimitations and future research\u003c/h2\u003e\u003cp\u003eSeveral limitations of the present study should be acknowledged. First, we failed to exclude the potential impact of medications on cognition, such as antihypertensive drugs\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e, erythropoietin\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e, and vitamin D\u003csup\u003e48\u003c/sup\u003e, which are commonly prescribed for patients with ESRD. Second, although ESRD patients typically present with frailty\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e and fatigue\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e, the current study did not assess the severity of these symptoms. Such variables deserve to be documented in future research to explore their effects on cognition and WM functional networks in ESRD patients. Third, we only collected uremic toxins routinely measured in clinical practice for ESRD patients, such as creatinine and urea. However, other uremic toxins, including p-cresyl sulfate, indoxyl sulfate, and methylglyoxal with direct neurotoxic effects in ESRD patients\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. Future studies should prioritize investigating the potential relationships between these uremic toxins and WM functional networks.\u003c/p\u003e\u003c/div\u003e"},{"header":"Conclusions","content":"\u003cp\u003eESRD patients exhibited extensively disrupted interactions among WM functional networks, which correlated with cognitive deficits and ESRD-specific clinical risk factors, including uremic toxin accumulation, dysregulation of calcium-phosphorus homeostasis, and anemia. This finding underscores their significance in understanding the functional architecture and hierarchical characteristics of WM that underlie ESRD, along with its associated cognitive and clinical phenotypes.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003eParticipants\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the local Research Ethics Committee (Approval No. 2020G64) and registered at ClinicalTrials.gov (https://clinicaltrials.gov/ct2/show/NCT03191409). All procedures adhered to the Declaration of Helsinki, and written informed consent was obtained from all participants. Between July 2020 and June 2024, 78 right-handed ESRD patients (53 males, 25 females; mean age: 35 \u0026plusmn; 10.18 years) undergoing maintenance hemodialysis were enrolled (Table 1). All patients had a dialysis duration of \u0026gt;3 months. Exclusion criteria included: (1) age \u0026lt;18 or \u0026gt;60 years; (2) brain lesions (hemorrhage, head trauma, tumor, stroke, or encephalomalacia) confirmed by conventional MRI or medical history; (3) psychiatric or neurodegenerative disorders; (4) diabetic nephropathy or lupus nephritis; (5) Hepatitis B, hepatitis C, syphilis, or acquired immunodeficiency syndrome; (6) smoking, alcohol, or drug abuse; (7) visual/auditory impairments (blurred vision, hearing loss, or other symptoms precluding neuropsychological assessment); and (8) claustrophobia or MR contraindications. Fifty right-handed HCs demographically matched to patients were recruited from the local community via advertisements (38 males, 12 females; mean age: 35 \u0026plusmn; 8.56 years). HCs inclusion criteria: age 18\u0026ndash;60 years with no history of neurologic, systemic, or psychiatric disorders.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eClinical characteristics and neuropsychological assessment\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDemographic data and clinical parameters were extracted from medical records. The underlying etiologies of ESRD were as follows: glomerulonephritis (n=58), immunoglobulin A nephropathy (n=12), and membranous nephropathy (n=8). All patients underwent maintenance hemodialysis three times weekly with a short midweek interval, lasting approximately 4 hours per session. Hemodialysis adequacy was quantified by a mean kinetic urea clearance/volume ratio (Kt/V) \u0026gt;1.2\u003csup\u003e52\u003c/sup\u003e. On the day prior to hemodialysis, patients completed neuropsychological testing in a quiet environment, followed by blood sampling. Cognitive assessments included the following: the Auditory Verbal Learning Test-Huashan Version (AVLT-H), which assesses short- and long-term memory; the Montreal Cognitive Assessment (MoCA), which evaluates global cognition, visuospatial function, attention, orientation, language, abstract thinking, and executive function; and the Trail-Making Test (TMT), which measures visual attention ability. Mood disorders were evaluated using the\u0026nbsp;Beck Depression Inventory (BDI)\u0026nbsp;and\u0026nbsp;Beck Anxiety Inventory (BAI).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eMRI data acquisition and preprocessing\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMRI data were acquired using a Discovery MR750 3.0 T scanner with an eight-channel phased-array head coil. All participants underwent three sequences: T1-weighted fluid-attenuated inversion recovery (T1-FLAIR), high-resolution T1-weighted structural imaging, and rs-fMRI. Participants were instructed to lie quietly with eyes closed, remain awake, and minimize head movement; foam padding and earplugs were used to reduce motion artifacts and scanner noise. The scan was terminated if participants reported discomfort or inability to complete the procedure, and post-scan verification confirmed cooperation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe rs-fMRI data were preprocessed using the Data Processing Assistant for Resting-State fMRI (DPARSF, Advanced Edition, V4.3; www.restfmri.net), Statistical Parametric Mapping 8 (SPM8; www.fil.ion.ucl.ac.uk/spm), and custom MATLAB scripts\u003csup\u003e15\u003c/sup\u003e (https://mind.huji.ac.il/white-matter.aspx).The T1-weighted anatomical images were segmented into gray matter, WM, and cerebrospinal fluid (CSF) using SPM8\u0026rsquo;s New Segment algorithm, then normalized to the Montreal Neurological Institute (MNI) template. The first 10 volumes were discarded to account for T1 equilibration effects. Remaining volumes underwent slice-time correction and motion correction (threshold: \u0026lt;3 mm translation or 3\u0026deg; rotation) relative to the mean functional image, followed by coregistration to the corresponding anatomical image. Framewise displacement (FD) was calculated to identify motion \u0026quot;spikes\u0026quot; (FD \u0026gt; 0.5 mm), which were included as separate regressors for data censoring without altering correlation values\u003csup\u003e17,53,54\u003c/sup\u003e.\u0026nbsp;Functional images were detrended to remove linear drifts. Nuisance covariates\u0026mdash;including 24 head motion parameters\u003csup\u003e55\u003c/sup\u003e and mean CSF signals\u0026mdash;were regressed out.\u0026nbsp;WM\u0026nbsp;and global brain signals were retained\u0026nbsp;to avoid eliminating signals of interest. A temporal band-pass filter (0.01\u0026ndash;0.15 Hz) was applied to reduce non-neuronal contributions to BOLD fluctuations\u003csup\u003e15,16\u003c/sup\u003e. For white matter functional images: Individual T1 segmentation images were coregistered to functional space to define\u0026nbsp;WM\u0026nbsp;masks (threshold = 0.5, as validated in previous studies\u003csup\u003e15,16,20\u003c/sup\u003e).\u0026nbsp;To avoid the mixture of the WM and gray-matter signals, images were spatially smoothed within the\u0026nbsp;WM\u0026nbsp;or gray-matter masks using a 4 mm full-width at half-maximum (FWHM) kernel.\u0026nbsp;Smoothed white matter functional images were normalized to the standard EPI template and resampled to 3 \u0026times; 3 \u0026times; 3 mm\u0026sup3; voxels for subsequent analyses.\u0026nbsp;Five ESRD patients with larger mean FD \u0026gt; 0.3 mm were discarded.\u0026nbsp;No controls exceeded the maximum thresholds for translation, rotation, or mean FD.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eWM functional networks clustering\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eClustering analysis of rs-fMRI data was adapted from Peer et al\u003csup\u003e15\u003c/sup\u003e. First, a unified group-level WM mask was generated from T1 segmentation results. Voxels with a percentage of participants \u0026gt;60% were identified as the group-level WM mask. Deep brain structures\u003csup\u003e56,57\u003c/sup\u003e\u0026mdash;including the thalamus, caudate, putamen, globus pallidus, and nucleus accumbens (defined using the Harvard-Oxford Atlas\u003csup\u003e58\u003c/sup\u003e)\u0026mdash;were excluded to refine the mask. The final group WM mask (17064 voxels) was coregistered to functional space and resampled to match rs-fMRI voxel dimensions. For each subject, Pearson correlation coefficients were computed between every WM voxel and all voxels in the group mask. To reduce computational load, the mask was subsampled to 4245 nodes using an interchanging grid strategy\u003csup\u003e59,60\u003c/sup\u003e. Individual matrices were averaged to generate a group-level mean correlation matrix. K-means clustering was applied to the group matrix, testing cluster counts from 2 to 22\u003csup\u003e59,61\u003c/sup\u003e. Stability was evaluated via four-fold cross-validation: the connectivity matrix was randomly split into four subsets (1061 voxels per fold), clustering was performed on each fold, and similarity between results was quantified using Dice\u0026apos;s coefficient for adjacency matrices\u003csup\u003e15,19\u003c/sup\u003e. The optimal cluster number was selected based on the highest stability and fine-grained network differentiation. According to the Dice\u0026apos;s coefficient of each number of clusters, the most stable and detailed WM functional network number was 11 (see Supplementary Figure 1). Similar to past studies\u003csup\u003e15,16,20\u003c/sup\u003e, Each WM functional network was co-registered to the JHU White Matter Tractography Atlas (20 tracts) and the ICBM DTI workgroup (48 tracts) to validate anatomical correspondence via spatial overlay. The WM functional networks were subsequently classified into superficial, middle, and deep layers.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eFC and FCC of WM networks\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe first used conventional FC to quantify the relationship between WM functional networks. For each participant, Pearson\u0026apos;s correlation coefficient was computed between the average time courses of all pairs of WM functional networks, then transformed to Fisher z-scores. In addition, we employed FCC\u003csup\u003e19,22\u003c/sup\u003e, a method based on the \u0026quot;correlation of correlations\u0026quot; framework, to estimate covariant relationships between WM functional networks by leveraging their correlations with multiple gray matter regions. First, 96 gray matter regions were defined using the Harvard-Oxford Gray Matter Atlas\u003csup\u003e58\u003c/sup\u003e. A gray matter mask was applied to restrict analysis to gray matter voxels, and time series were extracted from voxels in the intersection of this mask and the Harvard-Oxford atlas. Second, Pearson correlations were computed between each WM network and each gray matter region, yielding a K \u0026times; 96 correlation matrix (where K denotes the number of WM networks). Third, FCC between all pairs of WM networks was estimated to generate a K \u0026times; K FCC matrix; FCC values were subsequently transformed to Fisher z-scores.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eCoefficient GCA\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo quantify signed and directional influences among the 11WM functional networks, we performed bivariate cGCA\u003csup\u003e24\u003c/sup\u003e, accounting for region-specific hemodynamic delays in WM functional networks.\u0026nbsp;Unlike delay correlation methods that quantify temporal dependencies, cGCA leverages a multivariate autoregressive model to estimate direct causal effects\u0026mdash;discerning excitatory/inhibitory influences and making it ideal for mapping signed, directional interactions in WM functional networks.\u0026nbsp;For each functional network, individual preprocessed BOLD time series were extracted by averaging time series across all voxels within the network. Granger causal strength between networks was computed using the REST toolbox (v1.8;\u0026nbsp;\u003cu\u003ewww.restfmri.net\u003c/u\u003e).\u0026nbsp;Consistent with previous studies\u003csup\u003e16,20,23\u003c/sup\u003e, the signed strength and direction of pairwise network relationships were characterized by regression coefficients: positive/negative coefficients indicated excitatory/inhibitory pathways (i.e., source network activity predicted subsequent increases/decreases in target network activity). A directed asymmetric matrix (11\u0026times;11 regression coefficient matrix) was generated for each participant. Graph-theoretic metrics for cGCA were used to quantify incoming and outgoing influence strengths of each network\u003csup\u003e24\u003c/sup\u003e:\u003c/p\u003e\n\u003cp\u003e1) In-strength: Sum of absolute regression coefficients representing incoming connections, where the network acts as a target (significantly predicted by other networks).\u003c/p\u003e\n\u003cp\u003e2) Out-strength: Sum of absolute regression coefficients representing outgoing connections, where the network acts as a source (significantly predicts other networks).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eStatistics analyses\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003col\u003e\n \u003cli\u003e\u003cstrong\u003e\u003cem\u003eDemographic, clinical characteristics, and neuropsychological variables\u003c/em\u003e\u003c/strong\u003e\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eThe between-group differences in demographic characteristics (age and education level) and neuropsychological variables were compared using the Kolmogorov\u0026ndash;Smirnov test (for normality), Levene\u0026rsquo;s test (for equality of variances), and independent two-sample \u003cem\u003et\u003c/em\u003e-tests (for equality of means) via the Statistical Package for the Social Sciences (SPSS Statistics, version 22.0; IBM, Armonk, NY). A Chi-square test examined group differences in sex distribution. Multiple linear regression analyses adjusted for the effects of age, sex, education level, and mood disorders on between-group in neuropsychological results.\u0026nbsp;\u003c/p\u003e\n\u003col start=\"2\"\u003e\n \u003cli\u003e\u003cstrong\u003e\u003cem\u003eNetwork statistical analysis\u003c/em\u003e\u003c/strong\u003e\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eTo examine within-group patterns of FC and FCC in WM functional networks, we performed one-sample \u003cem\u003et\u003c/em\u003e-tests on correlation coefficient matrices across all participants, respectively. Group differences in network-level patterns were assessed using two-sample \u003cem\u003et\u003c/em\u003e-tests, with age, sex, mean FD, and Euclidean distance between each pair of networks included as confounding variables\u003csup\u003e62\u003c/sup\u003e. Statistical significance was defined as \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05 after FDR correction for multiple comparisons. To verify whether edges between network pairs differed from random results, we conducted a permutation test\u003csup\u003e63\u003c/sup\u003e (5000 iterations) by randomly assigning participants to two groups matching the sample sizes of the original data, respectively, followed by \u003cem\u003et\u003c/em\u003e-tests with the same set of confounding variables. Permutation \u003cem\u003ep\u003c/em\u003e-values were calculated as the proportion of permutations where the group difference exceeded the observed difference from the original \u003cem\u003et\u003c/em\u003e-test.\u003c/p\u003e\n\u003cp\u003eThe within-group GC patterns of WM functional networks were assessed for each directed edge across subjects in each group with one-sample \u003cem\u003et\u003c/em\u003e-test. The network-level between-group difference patterns for each directed edge were obtained using two-sample \u003cem\u003et\u003c/em\u003e-tests. Age, sex, mean FD, and Euclidean distance between each two networks were controlled for as confounding variables\u003csup\u003e62\u003c/sup\u003e(\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05, FDR correction). A permutation test of difference distribution\u0026nbsp;(5000 iterations)\u0026nbsp;was performed with the same confounding variables (\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05, FDR corrected). The 11 WM functional networks were categorized as superficial, middle, and deep network layers. The in/out strength of tri-layer networks in the ESRD patients and HCs were compared via two-sample \u003cem\u003et\u003c/em\u003e-tests (\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05, FDR corrected).\u003c/p\u003e\n\u003col start=\"3\"\u003e\n \u003cli\u003e\u003cstrong\u003e\u003cem\u003eCorrelation analyses\u003c/em\u003e\u003c/strong\u003e\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003ePearson or Spearman correlation analyses were used to assess relationships between alterations in WM functional network metrics (FC, FCC, and GC patterns) and both clinical indicators and neuropsychological variables. Statistical significance was defined as \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05 following FDR correction.\u003c/p\u003e\n\u003col start=\"4\"\u003e\n \u003cli\u003e\u003cstrong\u003e\u003cem\u003eFeature extraction and machine learning classification\u003c/em\u003e\u003c/strong\u003e\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eThis study employed six commonly used classifiers\u0026nbsp;implemented in Python 3.11 (sklearn), namely k-nearest neighbor, logistic regression, support vector machine, adaptive boost, categorical naive bayes, and random forest models, to discern whether between-group alterations in WM functional network metrics could effectively distinguish ESRD patients from HCs. A leave-one-out (LOOCV) cross-validation strategy was rigorously adopted to ensure robust estimation of classifier generalizability\u003csup\u003e64,65\u003c/sup\u003e. To assess the statistical significance of each classification model\u0026apos;s performance, a permutation test was executed, wherein the label vectors of all participants were randomly shuffled 5,000 times. For each permutation, the identical training and testing protocol was meticulously applied. Following 5,000 permutations, null distributions were derived for each performance metric\u0026mdash;encompassing accuracy, sensitivity, specificity, and AUC with 95% confidence intervals. The \u003cem\u003ep\u003c/em\u003e-value for each metric was computed by dividing the number of permutations yielding values exceeding the non-permuted model\u0026apos;s actual result by the total number of permutations, with statistical significance defined as \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05. Finally, the discriminative weight values assigned to each feature were leveraged to quantify the relative contribution of individual features to the classification process.\u003c/p\u003e\n\u003col start=\"5\"\u003e\n \u003cli\u003e\u003cstrong\u003e\u003cem\u003eValidation analyses\u003c/em\u003e\u003c/strong\u003e\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eSeveral validation analyses were conducted to verify the robustness of our results. First, considering that potential leakage of gray matter signals might affect the clustering pattern of WM signals, we applied two relatively stricter masks (with thresholds set at \u0026gt;70% and \u0026gt;80%) to define the group-level WM mask. We then re-conducted WM network clustering and functional metric analyses using these stricter masks to verify the consistency of our primary findings. Second, we calculated and compared the mean FD values between-group.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors would like to thank all the participants and their families who supported them.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWW and MZ drafted the work and revised it critically for important intellectual content. PL. and\u0026nbsp;YXS. made substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work. JYM, XYZ, ZYL, QGZ, HJY, TG, and CXW revised the article and interpreted the data. All authors approved the version submitted for publication.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe discovery data that support the findings are publicly available in the study’s Open Science Framework repository (https://osf.io/qkbj5/).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMain analysis codes are publicly available in the study’s Open Science Framework repository (https://osf.io/qkbj5/).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors report no biomedical financial interests or potential conflicts of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work is supported by National Natural Science Foundation of China (Grant No. 82202121), the Hovering Program of Fourth Military Medical University (axjhww), the Talent Foundation of Tangdu Hospital (2018BJ003), 7T MRI Precision Neurology Platform of Shaanxi Province (2025PT-08) and Innovative Team for Early Warning and Rehabilitation of Mental Fatigue, the Health Research and Innovation Capacity Strengthening Platform Program of Shaanxi Province (Grant No.2023PT-09), the Science and Technology Research Project of Shaanxi Nuclear Industry Group Co., Ltd (Grant No. 61240302), the Clinical Research Award of the First Affiliated Hospital of Xi’an Jiaotong University (No. XJTU1AF-CRF-2023-021).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eCollaboration, G.B.D.C.K.D.: Global, regional, and national burden of chronic kidney disease, 1990\u0026ndash;2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. \u003cb\u003e395\u003c/b\u003e, 709\u0026ndash;733 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S0140-6736(20)30045-3\u003c/span\u003e\u003cspan address=\"10.1016/S0140-6736(20)30045-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMu, J., et al.: Altered white matter microstructure mediates the relationship between hemoglobin levels and cognitive control deficits in end-stage renal disease patients. Hum. Brain Mapp. \u003cb\u003e39\u003c/b\u003e, 4766\u0026ndash;4775 (2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/hbm.24321\u003c/span\u003e\u003cspan address=\"10.1002/hbm.24321\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMu, J., et al.: Neurological effects of hemodialysis on white matter microstructure in end-stage renal disease. Neuroimage Clin. \u003cb\u003e31\u003c/b\u003e, 102743 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.nicl.2021.102743\u003c/span\u003e\u003cspan address=\"10.1016/j.nicl.2021.102743\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZheng, J., et al.: Brain Micro-Structural and Functional Alterations for Cognitive Function Prediction in the End-Stage Renal Disease Patients Undergoing Maintenance Hemodialysis. Acad. Radiol. \u003cb\u003e30\u003c/b\u003e, 1047\u0026ndash;1055 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.acra.2022.06.019\u003c/span\u003e\u003cspan address=\"10.1016/j.acra.2022.06.019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eViggiano, D., et al.: Mechanisms of cognitive dysfunction in CKD. Nat. Rev. Nephrol. \u003cb\u003e16\u003c/b\u003e, 452\u0026ndash;469 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41581-020-0266-9\u003c/span\u003e\u003cspan address=\"10.1038/s41581-020-0266-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCapasso, G., et al.: Drivers and mechanisms of cognitive decline in chronic kidney disease. Nat. Rev. Nephrol. (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41581-025-00963-0\u003c/span\u003e\u003cspan address=\"10.1038/s41581-025-00963-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBobot, M., Burtey, S.: New insights into mechanisms underlying cognitive impairment in chronic kidney disease. Kidney Int. \u003cb\u003e106\u003c/b\u003e, 1020\u0026ndash;1022 (2024). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.kint.2024.09.008\u003c/span\u003e\u003cspan address=\"10.1016/j.kint.2024.09.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNi, L., et al.: Aberrant default-mode functional connectivity in patients with end-stage renal disease: a resting-state functional MR imaging study. Radiology. \u003cb\u003e271\u003c/b\u003e, 543\u0026ndash;552 (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1148/radiol.13130816\u003c/span\u003e\u003cspan address=\"10.1148/radiol.13130816\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLuo, S., et al.: Abnormal Intrinsic Brain Activity Patterns in Patients with End-Stage Renal Disease Undergoing Peritoneal Dialysis: A Resting-State Functional MR Imaging Study. Radiology. \u003cb\u003e278\u003c/b\u003e, 181\u0026ndash;189 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1148/radiol.2015141913\u003c/span\u003e\u003cspan address=\"10.1148/radiol.2015141913\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDing, D., et al.: The relationship between putamen-SMA functional connectivity and sensorimotor abnormality in ESRD patients. Brain Imaging Behav. \u003cb\u003e12\u003c/b\u003e, 1346\u0026ndash;1354 (2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s11682-017-9808-6\u003c/span\u003e\u003cspan address=\"10.1007/s11682-017-9808-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, P., et al.: Brain connectome gradient dysfunction in patients with end-stage renal disease and its association with clinical phenotype and cognitive deficits. Commun. Biol. \u003cb\u003e8\u003c/b\u003e, 701 (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s42003-025-08132-6\u003c/span\u003e\u003cspan address=\"10.1038/s42003-025-08132-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMakedonov, I., Chen, J.J., Masellis, M., MacIntosh, B.J.: Physiological fluctuations in white matter are increased in Alzheimer's disease and correlate with neuroimaging and cognitive biomarkers. Neurobiol. Aging. \u003cb\u003e37\u003c/b\u003e, 12\u0026ndash;18 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neurobiolaging.2015.09.010\u003c/span\u003e\u003cspan address=\"10.1016/j.neurobiolaging.2015.09.010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGawryluk, J.R., Mazerolle, E.L., D'Arcy, R.C.: N. Does functional MRI detect activation in white matter? A review of emerging evidence, issues, and future directions. Front. Neurosci. \u003cb\u003e8\u003c/b\u003e, 239 (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fnins.2014.00239\u003c/span\u003e\u003cspan address=\"10.3389/fnins.2014.00239\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFabri, M., Polonara, G.: Functional topography of human corpus callosum: an FMRI mapping study. \u003cem\u003eNeural Plast\u003c/em\u003e 251308, (2013). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1155/2013/251308\u003c/span\u003e\u003cspan address=\"10.1155/2013/251308\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePeer, M., Nitzan, M., Bick, A.S., Levin, N., Arzy, S.: Evidence for Functional Networks within the Human Brain's White Matter. J. Neurosci. \u003cb\u003e37\u003c/b\u003e, 6394\u0026ndash;6407 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1523/JNEUROSCI.3872-16.2017\u003c/span\u003e\u003cspan address=\"10.1523/JNEUROSCI.3872-16.2017\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFan, Y.-S., et al.: Impaired interactions among white-matter functional networks in antipsychotic-naive first-episode schizophrenia. Hum. Brain Mapp. \u003cb\u003e41\u003c/b\u003e, 230\u0026ndash;240 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/hbm.24801\u003c/span\u003e\u003cspan address=\"10.1002/hbm.24801\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJiang, Y., et al.: White-matter functional networks changes in patients with schizophrenia. Neuroimage. \u003cb\u003e190\u003c/b\u003e, 172\u0026ndash;181 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neuroimage.2018.04.018\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2018.04.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMa, J., et al.: Frequency-dependent white-matter functional network changes associated with cognitive deficits in subcortical vascular cognitive impairment. Neuroimage Clin. \u003cb\u003e36\u003c/b\u003e, 103245 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.nicl.2022.103245\u003c/span\u003e\u003cspan address=\"10.1016/j.nicl.2022.103245\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJiang, Y., et al.: Dysfunctional white-matter networks in medicated and unmedicated benign epilepsy with centrotemporal spikes. Hum. Brain Mapp. \u003cb\u003e40\u003c/b\u003e, 3113\u0026ndash;3124 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/hbm.24584\u003c/span\u003e\u003cspan address=\"10.1002/hbm.24584\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMeng, L., et al.: Attenuated brain white matter functional network interactions in Parkinson's disease. Hum. Brain Mapp. \u003cb\u003e43\u003c/b\u003e, 4567\u0026ndash;4579 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/hbm.25973\u003c/span\u003e\u003cspan address=\"10.1002/hbm.25973\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLu, F., et al.: Superficial white-matter functional networks changes in bipolar disorder patients during depressive episodes. J. Affect. Disord. \u003cb\u003e289\u003c/b\u003e, 151\u0026ndash;159 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jad.2021.04.029\u003c/span\u003e\u003cspan address=\"10.1016/j.jad.2021.04.029\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang, H., et al.: Topographical Information-Based High-Order Functional Connectivity and Its Application in Abnormality Detection for Mild Cognitive Impairment. J. Alzheimers Dis. \u003cb\u003e54\u003c/b\u003e, 1095\u0026ndash;1112 (2016)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiao, W., et al.: Preservation Effect: Cigarette Smoking Acts on the Dynamic of Influences Among Unifying Neuropsychiatric Triple Networks in Schizophrenia. Schizophr Bull. \u003cb\u003e45\u003c/b\u003e, 1242\u0026ndash;1250 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/schbul/sby184\u003c/span\u003e\u003cspan address=\"10.1093/schbul/sby184\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiao, W., et al.: Small-world directed networks in the human brain: multivariate Granger causality analysis of resting-state fMRI. Neuroimage. \u003cb\u003e54\u003c/b\u003e, 2683\u0026ndash;2694 (2011). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neuroimage.2010.11.007\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2010.11.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCatani, M., de Thiebaut, M.: A diffusion tensor imaging tractography atlas for virtual in vivo dissections. Cortex. \u003cb\u003e44\u003c/b\u003e, 1105\u0026ndash;1132 (2008). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cortex.2008.05.004\u003c/span\u003e\u003cspan address=\"10.1016/j.cortex.2008.05.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eStave, E.A., et al.: Dimensions of Attention Associated With the Microstructure of Corona Radiata White Matter. J. Child. Neurol. \u003cb\u003e32\u003c/b\u003e, 458\u0026ndash;466 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/0883073816685652\u003c/span\u003e\u003cspan address=\"10.1177/0883073816685652\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBurke, T., et al.: Bilateral anterior corona radiata microstructure organisation relates to impaired social cognition in schizophrenia. Schizophr Res. \u003cb\u003e262\u003c/b\u003e, 87\u0026ndash;94 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.schres.2023.10.035\u003c/span\u003e\u003cspan address=\"10.1016/j.schres.2023.10.035\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVon Heide, D., Skipper, R.J., Klobusicky, L.M., E., Olson, I.R.: Dissecting the uncinate fasciculus: disorders, controversies and a hypothesis. Brain. \u003cb\u003e136\u003c/b\u003e, 1692\u0026ndash;1707 (2013). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/brain/awt094\u003c/span\u003e\u003cspan address=\"10.1093/brain/awt094\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang, R., et al.: Reduced white matter integrity and cognitive deficits in maintenance hemodialysis ESRD patients: a diffusion-tensor study. Eur. Radiol. \u003cb\u003e25\u003c/b\u003e, 661\u0026ndash;668 (2015). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00330-014-3466-5\u003c/span\u003e\u003cspan address=\"10.1007/s00330-014-3466-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang, Y.F., et al.: The gut microbiota-inflammation-brain axis in end-stage renal disease: perspectives from default mode network. Theranostics. \u003cb\u003e9\u003c/b\u003e, 8171\u0026ndash;8181 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7150/thno.35387\u003c/span\u003e\u003cspan address=\"10.7150/thno.35387\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTamura, M.K., et al.: Chronic kidney disease, cerebral blood flow, and white matter volume in hypertensive adults. Neurology. \u003cb\u003e86\u003c/b\u003e, 1208\u0026ndash;1216 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1212/WNL.0000000000002527\u003c/span\u003e\u003cspan address=\"10.1212/WNL.0000000000002527\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMartinez-Vea, A., et al.: Silent cerebral white matter lesions and their relationship with vascular risk factors in middle-aged predialysis patients with CKD. Am. J. Kidney Dis. \u003cb\u003e47\u003c/b\u003e, 241\u0026ndash;250 (2006)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003evan der Veen, P.H., et al.: Hemoglobin, hematocrit, and changes in cerebral blood flow: the Second Manifestations of ARTerial disease-Magnetic Resonance study. Neurobiol. Aging. \u003cb\u003e36\u003c/b\u003e, 1417\u0026ndash;1423 (2015). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neurobiolaging.2014.12.019\u003c/span\u003e\u003cspan address=\"10.1016/j.neurobiolaging.2014.12.019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePark, S.E., et al.: Decreased hemoglobin levels, cerebral small-vessel disease, and cortical atrophy: among cognitively normal elderly women and men. Int. Psychogeriatr. \u003cb\u003e28\u003c/b\u003e, 147\u0026ndash;156 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1017/S1041610215000733\u003c/span\u003e\u003cspan address=\"10.1017/S1041610215000733\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHanna, R.M., Ahdoot, R.S., Kalantar-Zadeh, K., Ghobry, L., Kurtz, I.: Calcium Transport in the Kidney and Disease Processes. Front. Endocrinol. (Lausanne). \u003cb\u003e12\u003c/b\u003e, 762130 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fendo.2021.762130\u003c/span\u003e\u003cspan address=\"10.3389/fendo.2021.762130\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHirasawa, T., et al.: Adverse effects of an active fragment of parathyroid hormone on rat hippocampal organotypic cultures. Br. J. Pharmacol. \u003cb\u003e129\u003c/b\u003e, 21\u0026ndash;28 (2000)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRosner, M.H., Husain-Syed, F., Reis, T., Ronco, C., Vanholder, R.: Uremic encephalopathy. Kidney Int. \u003cb\u003e101\u003c/b\u003e, 227\u0026ndash;241 (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.kint.2021.09.025\u003c/span\u003e\u003cspan address=\"10.1016/j.kint.2021.09.025\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDe Deyn, P.P., D'Hooge, R., Van Bogaert, P.P., Marescau, B.: Endogenous guanidino compounds as uremic neurotoxins. Kidney Int. Suppl. \u003cb\u003e78\u003c/b\u003e, S77\u0026ndash;S83 (2001)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKeane, D.F., et al.: The time of onset of intradialytic hypotension during a hemodialysis session associates with clinical parameters and mortality. Kidney Int. \u003cb\u003e99\u003c/b\u003e, 1408\u0026ndash;1417 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.kint.2021.01.018\u003c/span\u003e\u003cspan address=\"10.1016/j.kint.2021.01.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEldehni, M.T., Odudu, A., McIntyre, C.W.: Randomized clinical trial of dialysate cooling and effects on brain white matter. J. Am. Soc. Nephrol. \u003cb\u003e26\u003c/b\u003e, 957\u0026ndash;965 (2015). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1681/ASN.2013101086\u003c/span\u003e\u003cspan address=\"10.1681/ASN.2013101086\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChagas, Y.W., Vaz de Castro, P.A.S.: Sim\u0026otilde;es-E-Silva, A. C. Neuroinflammation in kidney disease and dialysis. Behav. Brain Res. \u003cb\u003e483\u003c/b\u003e, 115465 (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.bbr.2025.115465\u003c/span\u003e\u003cspan address=\"10.1016/j.bbr.2025.115465\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePark, B.S., et al.: The effects of the dialysis on the white matter tracts in patients with end-stage renal disease using differential tractography study. Sci. Rep. \u003cb\u003e13\u003c/b\u003e, 20064 (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-023-47533-7\u003c/span\u003e\u003cspan address=\"10.1038/s41598-023-47533-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePilato, F., Norata, D., Rossi, M.G., Di Lazzaro, V., Calandrelli, R.: Consciousness disturbance in patients with chronic kidney disease: Rare but potentially treatable complication. Clinical and neuroradiological review. Behav. Brain Res. \u003cb\u003e480\u003c/b\u003e, 115393 (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.bbr.2024.115393\u003c/span\u003e\u003cspan address=\"10.1016/j.bbr.2024.115393\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu, M., et al.: White Matter Microstructure Changes and Cognitive Impairment in the Progression of Chronic Kidney Disease. Front. Neurosci. \u003cb\u003e14\u003c/b\u003e, 559117 (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fnins.2020.559117\u003c/span\u003e\u003cspan address=\"10.3389/fnins.2020.559117\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJiang, Y., et al.: Reduced White Matter Integrity in Patients With End-Stage and Non-end-Stage Chronic Kidney Disease: A Tract-Based Spatial Statistics Study. Front. Hum. Neurosci. \u003cb\u003e15\u003c/b\u003e, 774236 (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fnhum.2021.774236\u003c/span\u003e\u003cspan address=\"10.3389/fnhum.2021.774236\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBirkenhager, W.H., Staessen, J.A.: Antihypertensives for prevention of Alzheimer's disease. Lancet Neurol. \u003cb\u003e5\u003c/b\u003e, 466\u0026ndash;468 (2006). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S1474-4422(06)70453-7\u003c/span\u003e\u003cspan address=\"10.1016/S1474-4422(06)70453-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLee, S.T., et al.: Erythropoietin improves memory function with reducing endothelial dysfunction and amyloid-beta burden in Alzheimer's disease models. J. Neurochem. \u003cb\u003e120\u003c/b\u003e, 115\u0026ndash;124 (2012). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1471-4159.2011.07534.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1471-4159.2011.07534.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLatimer, C.S., et al.: Vitamin D prevents cognitive decline and enhances hippocampal synaptic function in aging rats. Proc. Natl. Acad. Sci. U S A. \u003cb\u003e111\u003c/b\u003e, E4359\u0026ndash;4366 (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1073/pnas.1404477111\u003c/span\u003e\u003cspan address=\"10.1073/pnas.1404477111\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChan, G.C.-K., et al.: Frailty in patients on dialysis. Kidney Int. \u003cb\u003e106\u003c/b\u003e, 35\u0026ndash;49 (2024). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.kint.2024.02.026\u003c/span\u003e\u003cspan address=\"10.1016/j.kint.2024.02.026\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBossola, M., Tazza, L., Postdialysis Fatigue: A Frequent and Debilitating Symptom. Semin Dial. \u003cb\u003e29\u003c/b\u003e, 222\u0026ndash;227 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/sdi.12468\u003c/span\u003e\u003cspan address=\"10.1111/sdi.12468\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAndrews, T.D., Day, G.S., Irani, S.R., Kanekiyo, T., Hickson, L.J.: Uremic Toxins, CKD, and Cognitive Dysfunction. J. Am. Soc. Nephrol. \u003cb\u003e36\u003c/b\u003e, 1208\u0026ndash;1226 (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1681/ASN.0000000675\u003c/span\u003e\u003cspan address=\"10.1681/ASN.0000000675\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDaugirdas, J.T.: Kt/V (and especially its modifications) remains a useful measure of hemodialysis dose. Kidney Int. \u003cb\u003e88\u003c/b\u003e, 466\u0026ndash;473 (2015). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/ki.2015.204\u003c/span\u003e\u003cspan address=\"10.1038/ki.2015.204\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJi, G.-J., Liao, W., Chen, F.-F., Zhang, L., Wang, K.: Low-frequency blood oxygen level-dependent fluctuations in the brain white matter: more than just noise. Sci. Bull. (Beijing). \u003cb\u003e62\u003c/b\u003e, 656\u0026ndash;657 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.scib.2017.03.021\u003c/span\u003e\u003cspan address=\"10.1016/j.scib.2017.03.021\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePower, J.D., Barnes, K.A., Snyder, A.Z., Schlaggar, B.L., Petersen, S.E.: Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion. Neuroimage. \u003cb\u003e59\u003c/b\u003e, 2142\u0026ndash;2154 (2012). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neuroimage.2011.10.018\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2011.10.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFriston, K.J., Williams, S., Howard, R., Frackowiak, R.S., Turner, R.: Movement-related effects in fMRI time-series. Magn. Reson. Med. \u003cb\u003e35\u003c/b\u003e, 346\u0026ndash;355 (1996)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWonderlick, J.S., et al.: Reliability of MRI-derived cortical and subcortical morphometric measures: effects of pulse sequence, voxel geometry, and parallel imaging. Neuroimage. \u003cb\u003e44\u003c/b\u003e, 1324\u0026ndash;1333 (2009). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neuroimage.2008.10.037\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2008.10.037\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLorio, S., et al.: New tissue priors for improved automated classification of subcortical brain structures on MRI. Neuroimage. \u003cb\u003e130\u003c/b\u003e, 157\u0026ndash;166 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neuroimage.2016.01.062\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2016.01.062\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDesikan, R.S., et al.: An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage. \u003cb\u003e31\u003c/b\u003e, 968\u0026ndash;980 (2006)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYeo, B.T.T., et al.: The organization of the human cerebral cortex estimated by intrinsic functional connectivity. J. Neurophysiol. \u003cb\u003e106\u003c/b\u003e, 1125\u0026ndash;1165 (2011). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1152/jn.00338.2011\u003c/span\u003e\u003cspan address=\"10.1152/jn.00338.2011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCraddock, R.C., James, G.A., Holtzheimer, P.E., Hu, X.P., Mayberg, H.: S. A whole brain fMRI atlas generated via spatially constrained spectral clustering. Hum. Brain Mapp. \u003cb\u003e33\u003c/b\u003e, 1914\u0026ndash;1928 (2012). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/hbm.21333\u003c/span\u003e\u003cspan address=\"10.1002/hbm.21333\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLange, T., Roth, V., Braun, M.L., Buhmann, J.M.: Stability-based validation of clustering solutions. Neural Comput. \u003cb\u003e16\u003c/b\u003e, 1299\u0026ndash;1323 (2004)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, J., et al.: Exploring the functional connectome in white matter. Hum. Brain Mapp. \u003cb\u003e40\u003c/b\u003e, 4331\u0026ndash;4344 (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/hbm.24705\u003c/span\u003e\u003cspan address=\"10.1002/hbm.24705\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang, Z., et al.: Altered functional-structural coupling of large-scale brain networks in idiopathic generalized epilepsy. Brain. \u003cb\u003e134\u003c/b\u003e, 2912\u0026ndash;2928 (2011). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/brain/awr223\u003c/span\u003e\u003cspan address=\"10.1093/brain/awr223\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVaroquaux, G., et al.: Assessing and tuning brain decoders: Cross-validation, caveats, and guidelines. Neuroimage. \u003cb\u003e145\u003c/b\u003e, 166\u0026ndash;179 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neuroimage.2016.10.038\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2016.10.038\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVaroquaux, G.: Cross-validation failure: Small sample sizes lead to large error bars. Neuroimage. \u003cb\u003e180\u003c/b\u003e, 68\u0026ndash;77 (2018). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neuroimage.2017.06.061\u003c/span\u003e\u003cspan address=\"10.1016/j.neuroimage.2017.06.061\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"End-stage renal disease, White matter, Functional magnetic resonance imaging, Functional connectivity, Machine learning","lastPublishedDoi":"10.21203/rs.3.rs-7702413/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7702413/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAmong patients with end-stage renal disease (ESRD), the white matter (WM) is a particularly vulnerable area that is susceptible to various clinical risk factors. However, whether WM function is disrupted in ESRD patients and how this disruption provides valuable information for cognitive deficits and potential clinical phenotypes remain unknown. We prospectively enrolled 78 ESRD patients and 50 healthy controls. Using resting-state functional magnetic resonance imaging, we studied ESRD-related WM functional networks alterations. Functional connectivity, functional covariance connectivity, and coefficient Granger causality analysis were probed interactions among WM functional networks. The machine-learning models with leave-one-out cross-validation were applied. ESRD patients exhibited extensively disrupted interactions among WM functional networks, which correlated with cognitive deficits and ESRD-specific clinical risk factors, including uremic toxin accumulation, dysregulation of calcium-phosphorus homeostasis, and anemia. A random forest classifier achieved a maximum performance of 95.31% accuracy and 0.982 area under the ROC curve (AUC). Our results emphasized the imbalances of WM functional networks in ESRD patients, which might be used as potential neuroimaging markers for cognitive deficits and potential clinical phenotypes.\u003c/p\u003e","manuscriptTitle":"White-matter functional network dysfunction associated with cognitive deficits and clinical phenotypes in patients with end-stage renal disease","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-30 01:06:32","doi":"10.21203/rs.3.rs-7702413/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"communications-biology","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"commsbio","sideBox":"Learn more about [Communications Biology](http://www.nature.com/commsbio/)","snPcode":"","submissionUrl":"","title":"Communications Biology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Communications Series","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"c0310219-a09a-41bc-819c-7d6099823857","owner":[],"postedDate":"October 30th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":56306767,"name":"Biological sciences/Neuroscience/Cognitive neuroscience"},{"id":56306768,"name":"Health sciences/Neurology/Neurological disorders/White matter disease"},{"id":56306769,"name":"Health sciences/Diseases/Kidney diseases/Chronic kidney disease/End-stage renal disease"},{"id":56306770,"name":"Biological sciences/Neuroscience/Computational neuroscience/Network models"}],"tags":[],"updatedAt":"2026-02-24T05:25:34+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-30 01:06:32","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7702413","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7702413","identity":"rs-7702413","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}