ABSTRACT
Background
Hippocampal alterations have been implicated in the pathophysiology of cognitive impairment in hemodialysis patients. The hippocampus consists of several distinct subfields, and the molecular mechanisms underlying cognition might be associated with specific hippocampal subfield volume changes. However, this has not yet been investigated in hemodialysis patients. This study aimed to explore volumetric abnormalities in hippocampal subfields in regular hemodialysis patients.
Methods
High-resolution T1-weighted structural images were collected in 61 subjects including 36 hemodialysis patients and 25 healthy controls. A state-of-the-art hippocampal segmentation approach was adopted to segment the hippocampal subfields. Group differences in hippocampal subfield volumes were assessed in Python with a statsmodels module using an ordinary least squares regression with age and sex as nuisance effects.
Results
Hemodialysis patients had significantly smaller volumes in the bilateral hippocampus (P < .05/2, Bonferroni corrected), cornu ammonis 1 (CA1), CA4, granule cell and molecular layer of the dentate gyrus, hippocampus–amygdala transition area and molecular layer of the hippocampus than healthy controls (P < .05/24, Bonferroni corrected). Hemodialysis patients also had lower volumes in the left hippocampal tail and right fimbria than healthy controls (P < .05/24, Bonferroni corrected). Hippocampal subfield volumes were associated with neuropsychological test scores, the duration of disease and hemoglobin levels.
Conclusions
We found smaller hippocampal subfield volumes in hemodialysis patients, which were associated with impaired cognition, supporting their role in memory disturbance in the hemodialysis population. However, multiple clinical factors may have confounded the results, and therefore, the interpretation of these results needs to be cautious.
Keywords: anemia, cognition, hemodialysis, hippocampal subfields, memory
KEY LEARNING POINTS.
What is already known about this subject?
Hippocampal alterations have been implicated in the pathophysiology of cognitive impairment in hemodialysis patients.
The hippocampus consists of several distinct subfields and the molecular mechanisms underlying cognition might be associated with specific hippocampal subfield volume changes. However, this possibility has not yet been investigated in hemodialysis patients.
What this study adds?
Hemodialysis patients had significantly smaller volumes of the bilateral hippocampus, cornu ammonis 1 (CA1), CA4, granule cell and molecular layer of the dentate gyrus, hippocampus–amygdala transition area, molecular layer of the hippocampus left hippocampal tail and right fimbria than healthy controls.
Hippocampal subfield volumes were associated with neuropsychological test scores, duration of disease and hemoglobin levels.
What impact this may have on practice or policy?
The present findings revealed smaller hippocampal subfield volumes in hemodialysis patients, which were associated with impaired cognition, supporting their role in memory disturbance in the hemodialysis population.
Treating anemia might be beneficial for cognitive improvement in hemodialysis patients.
INTRODUCTION
End-stage renal disease (ESRD) patients are at high risk of developing cognitive deficits, with prevalence rates of 30%–70% [1]. For ESRD patients on hemodialysis, the incidence rate of cognitive impairment could be even higher. Cognitive deficits have an adverse effect on patients’ daily life, social adaptation, employment options, medication adherence and medical decision making [2]. Moreover, these cognitive deficits contribute to increased morbidity and mortality [2, 3]. Thus, it is crucial to clarify the pathophysiological mechanisms underlying cognitive impairment to reduce its harmful influence on hemodialysis patients.
The hippocampus plays a vital role in multiple cognitive functions, such as learning and memory acquisition, consolidation and retrieval, stress and emotional responsivity, sensorimotor processes, and goal-directed activity [4]. Recently, altered function in the hippocampus has been found in hemodialysis patients. For example, Chen et al. [5, 6] demonstrated that hemodialysis patients had an increased amplitude of low-frequency fluctuation and regional homogeneity in the left hippocampus. In contrast, Wu et al. revealed decreased nodal centralities in the bilateral hippocampus in hemodialysis patients in comparison with nondialysis patients [7]. Li et al. found that the cerebral blood flow in the hippocampus significantly decreased after hemodialysis initiation [8]. A structural study also revealed decreased volume in the hippocampus in dialysis patients [9]. The hippocampus plays a vital role in cognition [10]. However, as far as we know, the structural subfield alterations in the hippocampus in hemodialysis patients have not yet been investigated.
It has been recognized that the hippocampus consists of distinct subfields: the cornu ammonis (CA) subfields CA1–4, the dentate gyrus, the fimbria, and the adjacent subiculum and presubiculum [11]. Different subfields are responsible for different functions. Importantly, subfield alterations in hemodialysis patients have not been comprehensively explored.
Recent technological advancements have made it possible to automatically segment hippocampal subfields with high-resolution atlases. Ex vivo hippocampal tissues with ultrahigh field strength can achieve accurate hippocampal subfield segmentation. When this new method was applied in vivo, the segmentation results were found to be more accurate than those of previous in vivo templates [12, 13]. With this new method, we can better understand the regional alterations within the hippocampus associated with hemodialysis and its disease development.
The purpose of the present study was to compare the hippocampal subfield volumes in regular hemodialysis patients and healthy controls. We hypothesized that the hippocampal subfield volumes in hemodialysis patients would be smaller than those of healthy controls. Furthermore, we explored the relationship between clinical indices and hippocampal subfield volumes in hemodialysis patients. We hypothesized smaller volumes to be associated with memory disorders.
MATERIALS AND METHODS
Participants
This study was performed in line with the Declaration of Helsinki and was approved by the ethics committee of Hainan General Hospital. All participants were provided with the details of the experiment and they gave written informed consent before the experiment. Thirty-six ESRD patients on hemodialysis were enrolled from June 2019 to September 2021, including 17 males and 19 females. The average age was 43.7 ± 11.1 years (range 19–67 years). The exclusion criteria were as follows: (i) drug or alcohol abuse history; (ii) noticeable cerebral abnormalities detected by magnetic resonance imaging (MRI); (iii) past or present mental illness and brain trauma; (iv) hearing or visual deficiencies; and (v) organ transplantation history. Meanwhile, 25 age-, sex- and education-matched healthy controls were recruited via advertisement as normal controls, including 13 males and 12 females. The average age was 43.9 ± 10.8 years (range 23–57 years). All healthy controls had no history of systemic diseases or neuropsychiatric diseases; they were right-handed with normal vision. Other exclusion criteria were consistent with those in the patient group. All individuals who completed all the procedures were paid 200 CNY for their participation.
Ethics statement
The studies involving human subjects were reviewed and approved by the Medical Research Ethics Committee of Hainan General Hospital in line with the Helsinki Declaration. Written informed consent was acquired by all the participants before the study.
Neuropsychological assessment and laboratory results
All subjects completed the neuropsychological scales before MRI scanning. The neuropsychological scales included the Montreal Cognitive Assessment (MoCA), number connection test type A (NCT-A), California Verbal Learning Test-Ⅱ (CVLT-Ⅱ), digital symbol substitution test (DSST) and digit span test (DS). The MoCA was used to screening for mild cognitive impairment. The MoCA assesses the following domains including visuospatial/executive abilities, naming, memory, attention, abstraction and orientation. The maximal score for the MoCA was 30, and higher scores suggested better cognitive performance [14]. The NCT-A and DSST evaluated psychomotor speed, attention and visual-associated memory. Shorter time on the NCT-A and higher scores on the DSST indicate better performance, respectively. The DS evaluates the domains of attention function. Higher scores are associated with better performance. CVLT performance includes subscales such as immediate recall (trials 1–5), short delay cued and free recall (SDCR and SDFR), and long delay cued and free recall (LDCR and LDFR). The CVLT-II includes two word inventories. List A contains 16 items, including 4 words in each category (vegetables, animals, house furnishings and vehicles). After presentation of each word in List A, the subjects were asked to recall the words, which provided an immediate free recall score. After the five trials, List B was presented as a form of interference. SDCR and SDFR of List A are examined. After 20 min of nonverbal testing, LDCR and LDFR of List A was evaluated.
All subjects underwent laboratory tests, including determination of creatinine, serum urea, uric acid, red blood cells (RBCs) and hemoglobin levels, before MR imaging.
Data acquisition
All MRI data were acquired with a 3T MR scanner (TIM Skyra, Siemens Medical Solutions, Erlangen, Germany). To minimize head movement, foam padding was used for support. Earplugs were used to decrease noise. All participants were instructed to hold still, to keep their eyes open and to try not to do intentional thinking during scanning. Conventional T1-weighted images, T2-weighted images and T2-FLAIR images were obtained to exclude any obvious lesions. The parameters for T1-weighted image acquisition were as follows: 20 axial scans; slice thickness = 6 mm; interlayer spacing = 0.4 mm; in-plane resolution = 320 × 256; field of view (FOV) = 240 × 240 mm2; repetition time (TR) = 2500 ms; and echo time (TE) = 9 ms. The parameters for T2-FLAIR image acquisition were as follows: 20 axial slices; slice thickness = 6 mm; in-plane resolution = 232 × 256; FOV = 240 × 240 mm2; and flip angle = 150°. Then, 192 sagittal scans of high-resolution 3D T1-weighted structural images were acquired. The parameters were as follows: TR/TE = 2530/2.98 ms; FOV = 256 × 256 mm2; and matrix size = 256 × 256. The slice thickness was 1 mm.
Structural data processing
Cortical surface reconstruction and subcortical structure segmentation were performed with FreeSurfer v6.0.1 [15, 16]. Briefly, this processing includes motion correction removal of nonbrain tissue, automated Talairach transformation, segmentation of the subcortical white matter and deep gray matter volumetric structures (including the hippocampus, amygdala, caudate, putamen, ventricles) [17, 18], intensity normalization [19], tessellation of the gray matter–white matter boundary, automated topology correction [20, 21] and surface deformation. A number of deformable procedures were performed for further data processing and analysis including surface inflation [16], registration to a spherical atlas [16], parcellation of the cerebral cortex into gyral and sulcal structures [18, 22–25], and creation of a variety of surface-based data including maps of curvature, sulcal depth and cortical thickness. Furthermore, the estimated total intracranial volume (eTIV) representing head size was automatically calculated for each subject using an atlas-based spatial normalization procedure [26]. An automated segmentation of hippocampal subfields [12] based on a statistical atlas built primarily upon ultrahigh resolution ex vivo MRI data was applied in the present study. The whole hippocampus was identified, and we obtained 12 subfields in the present study: hippocampal tail, hippocampal fissure, fimbria, hippocampus–amygdala transition area (HATA), molecular layer, granule cell and molecular layer of the dentate gyrus (GC-ML-DG), CA4, CA3, CA1, subiculum, presubiculum and parasubiculum (Fig. 1).
Figure 1:
Representative examples of left hippocampal subfield segmentations from a hemodialysis patient and healthy control. HD, hemodialysis patients; HC, healthy controls; GC-DG, granule cell and molecular layer of the dentate gyrus; molecular_layer_HP, molecular layer of the hippocampus; HP, hippocampus.
The whole analysis was accelerated and simplified through a cloud platform (http://www.humanbrain.cn, Beijing Intelligent Brain Cloud, Inc.).
Hippocampal subfield volume analysis
The hippocampal subfield volumes between the two groups were compared in Python with a statsmodels module using an ordinary least squares regression with age and sex as nuisance effects. We first assessed volumetric differences for the left and right whole hippocampal volumes using a statistical threshold with Bonferroni correction (P < .05/2). Then, pairwise comparisons were performed for all left and right hippocampal subfield volumes (24 comparisons for the left and right subfields). The statistical P-values were corrected using Bonferroni correction across all 24 subfield tests (P < .05/24). In addition, we included age, sex and eTIV as nuisance effects for group comparisons. The implementation scripts of the above statistical analysis were written in Python with the statsmodels module.
Statistical analysis
The analysis of demographic and clinical data was performed with SPSS version 21 (IBM Corp., Armonk, NY, USA). Independent samples t tests or the Mann–Whitney tests were used to analyze the quantitative data. Data are expressed as the means ± standard deviations. The chi-square test was applied to evaluate the sex difference between groups. Additionally, Pearson correlation analyses were used evaluate the relationship between the hippocampus/hippocampal subfield volumes that demonstrated significant differences and clinical parameters in the hemodialysis group. The statistical threshold was set at P < .05.
RESULTS
Demographic and clinical characteristics
Table 1 shows the demographic and clinical characteristics of the hemodialysis patients and healthy controls. Nine patients were receiving hemodialysis. Twenty-seven patients underwent hemodiafiltration. Nine patients did not have a Kv/t test. Kv/t was available in 27 patients. The average Kv/t was 1.55 ± 0.43. The blood pressures for the patients were 137.4 ± 23.1 mmHg and 83.9 ± 19.7 mmHg for the systolic blood pressure and diastolic blood pressure, respectively. Among these patients, three patients had comorbid diabetes. No patient had a stroke. Ten patients had cardiovascular disease. Twenty-six patients were using erythropoietin. Three patients were using roxadustat. Seven patients were using neither erythropoietin nor roxadustat. Thirty-two patients were using fistulas (29 with arteriovenous fistula, 3 with arteriovenous graft) and 4 were using central catheters. The hemodialysis patients did not significantly differ regarding age, sex, education, NCT-A, trials 1–5 scores, and SDFR and LDCR scores compared with the healthy controls (all P > .05). The hemodialysis patients had significantly lower scores in MoCA total scores, visual/executive function, orientation, DS, DSST, SDCR and LDFR scores than the healthy controls (P = .036, P = .019, P = .044, P = .019, P = .045, P = .046 and P = .024, respectively), indicating that the hemodialysis patients performed worse on these neuropsychiatric tests. The patients on hemodialysis had significantly higher creatinine (P < .001) and urea and uric acid levels (P < .001) than healthy controls. The hemodialysis patients had significantly lower levels of RBCs and hemoglobin (P < .001).
Table 1:
Demographic and clinical data of hemodialysis patients and healthy controls.
| Parameters | HD (n = 36) | HC (n = 25) | t value or χ2 value | P-value |
|---|---|---|---|---|
| Gender (males/females) | 17/19 | 13/12 | 0.135 | .714a |
| Age (years) | 43.7 ± 11.1 | 43.9 ± 10.8 | −0.089 | .930b |
| Education (years) | 10.6 ± 2.6 | 12.1 ± 2.8 | .052c | |
| Hemodialysis type | ||||
| Hemodialysis | 27 | |||
| Hemodiafiltration | 9 | |||
| Kv/t | 1.55 ± 0.43d | |||
| Systolic blood pressure (mmHg) | 137.4 ± 23.1 | |||
| Diastolic pressure (mmHg) | 83.9 ± 19.7 | |||
| Diabetes | 3 | |||
| Stroke | 0 | |||
| Cardiovascular disease | 10 | |||
| Erythropoietin | 26 | |||
| Roxadustat | 3 | |||
| Fistulas | 32 | |||
| Central catheter | 4 | |||
| MoCA (score) | 24.8 ± 3.3 | 26.5 ± 2.5 | −2.146 | .036 b |
| Visual/executive function | 4.0 ± 1.2 | 4.6 ± 0.7 | −2.420 | .019 b |
| Naming | 2.6 ± 0.8 | 2.8 ± 0.4 | −1.546 | .128b |
| Memory | 1.7 ± 1.7 | 2.2 ± 1.4 | −1.170 | .247b |
| Attention | 5.6 ± 0.9 | 5.8 ± 0.7 | −0.831 | .409b |
| Language | 2.5 ± 0.7 | 2.7 ± 0.9 | −0.920 | .361b |
| Abstraction | 1.9 ± 0.4 | 1.9 ± 0.6 | 0.07 | .945b |
| Orientation | 5.9 ± 0.4 | 6.0 ± 0.0 | −2.092 | .044 b |
| DS (score) | 11.9 ± 2.0 | 13.4 ± 2.7 | −2.409 | .019 b |
| DSST (score) | 34.5 ± 13.6 | 43.6 ± 18.9 | −2.071 | .045 b |
| NCT-A (s) | 51.2 ± 25.8 | 49.5 ± 30.5 | 0.241 | .811b |
| Trials 1–5 (score) | 38.1 ± 10.6 | 44.0 ± 12.2 | −2.005 | .050b |
| SDFR (score) | 8.2 ± 2.8 | 9.6 ± 3.3 | −1.810 | .075b |
| SDCR (score) | 8.7 ± 2.7 | 10.3 ± 3.4 | −2.039 | .046 b |
| LDFR (score) | 7.9 ± 3.8 | 10.2 ± 3.8 | −2.317 | .024 b |
| LDCR (score) | 8.8 ± 3.0 | 10.0 ± 3.7 | −1.408 | .164b |
| Duration of disease (months) | 110.7 ± 77.8 | |||
| Hemodialysis duration (months) | 70.8 ± 53.6 | |||
| Creatinine (μmol/L) | 942.5 ± 232.2 | 67.1 ± 15.0 | 22.555 | <.001 b |
| Urea (μmol/L) | 21.9 ± 7.0 | 4.4 ± 1.1 | 14.661 | <.001 b |
| Urid acid (μmol/L) | 423.8 ± 108.0 | 319.2 ± 88.0 | 4.001 | <.001 b |
| RBC (×1012/L) | 4.0 ± 1.1 | 4.9 ± 0.4 | −4.347 | <.001 b |
| Hemoglobin (g/L) | 116.4 ± 28.5 | 146.4 ± 18.3 | −5.000 | <.001 b |
P-value obtained with Chi-square test.
b P value obtained with independent t test for continuous variables.
c P-value obtained with independent t test for non-continuous variables.
dKv/t was available in 27 patients. Values are presented as mean ± standard deviation except for gender, which is presented as a number. Bold values indicate statistical significance between groups. HD, hemodialysis; HC, healthy control.
Findings between hemodialysis patients and healthy controls
Hemodialysis patients had significantly smaller subfield volumes than the healthy controls in the left CA1 (t = –4.32; P < .01), CA4 (t = –3.51; P < .01), GC-ML-DG (t = –3.79, P < .01), HATA (t = –3.38, P < .01), hippocampal tail (t = –3.86, P < .01), molecular layer of the hippocampus (t = –4.72, P < .01) and left whole hippocampus (t = –4.45, P < .01) (Table 2, Fig. 2).
Table 2:
Statistical significance of hippocampal subfield volumes between hemodialysis patients and healthy controls.
| Hippocampal subfields | HD | HC | t-value | P-value |
|---|---|---|---|---|
| left_Whole_hippocampus | 3252.8 ± 269.6 | 3573.7 ± 289.3 | −4.45 | <.01 |
| left_Hippocampal_tail | 512.4 ± 63.5 | 583.7 ± 78.6 | −3.86 | <.01 |
| left_subiculum | 431.5 ± 43.0 | 462.9 ± 48.6 | −2.58 | .012 |
| left_CA1 | 582.0 ± 56.2 | 646.4 ± 70.0 | −4.32 | <.01 |
| left_hippocampal_fissure | 169.5 ± 32.9 | 159.9 ± 28.9 | 1.33 | .19 |
| left_presubiculum | 307.1 ± 38.2 | 323.1 ± 46.7 | −1.39 | .17 |
| left_parasubiculum | 59.8 ± 12.7 | 55.9 ± 12.8 | 1.12 | .27 |
| left_molecular_layer_HP | 529.3 ± 47.5 | 588.0 ± 50.1 | −4.72 | <.01 |
| left_GC_ML_DG | 274.9 ± 30.5 | 303.8 ± 27.2 | −3.79 | <.01 |
| left_CA3 | 188.8 ± 26.6 | 200.3 ± 22.2 | −1.79 | .08 |
| left_CA4 | 239.9 ± 26.2 | 262.8 ± 23.3 | −3.51 | <.01 |
| left_fimbria | 76.8 ± 19.8 | 90.6 ± 17.1 | −2.75 | <.01 |
| left_HATA | 50.3 ± 6.0 | 56.2 ± 7.3 | −3.38 | <.01 |
| right_Whole_hippocampus | 3361.2 ± 309.7 | 3683.3 ± 298.2 | −4.25 | <.01 |
| right_Hippocampal_tail | 553.1 ± 61.9 | 600.6 ± 82.7 | −2.55 | .01 |
| right_subiculum | 440.5 ± 46.7 | 473.1 ± 47.5 | −2.70 | .01 |
| right_CA1 | 617.6 ± 69.9 | 679.1 ± 72.9 | −3.52 | <.01 |
| right_hippocampal_fissure | 181.1 ± 36.6 | 161.5 ± 34.6 | 2.28 | .03 |
| right_presubiculum | 294.3 ± 32.6 | 309.2 ± 33.7 | −1.67 | .10 |
| right_parasubiculum | 55.5 ± 10.7 | 52.5 ± 8.0 | 1.09 | .28 |
| right_molecular_layer_HP | 547.2 ± 55.1 | 611.8 ± 53.0 | −4.73 | <.01 |
| right_GC_ML_DG | 284.1 ± 36.6 | 318.3 ± 31.9 | −3.93 | <.01 |
| right_CA3 | 201.1 ± 32.3 | 220.9 ± 28.0 | −2.57 | .01 |
| right_CA4 | 247.6 ± 30.6 | 274.5 ± 29.0 | −3.62 | <.01 |
| right_fimbria | 68.1 ± 19.4 | 83.7 ± 16.1 | −3.24 | <.01 |
| right_HATA | 52.1 ± 8.2 | 59.7 ± 8.6 | −3.47 | <.01 |
Subfields in bold denote significance.
HD, hemodialysis; HC, healthy control; HP, hippocampus.
Figure 2:
Hippocampal subfield volumes in HD and HC. **P < .005; ***P < .0005; ****P < .0001. The error bar indicates a 95% confidence interval. HD, hemodialysis patients; HC, healthy controls; GC-DG, granule cell and molecular layer of the dentate gyrus; molecular_layer_HP, molecular layer of the hippocampus; HP, hippocampus.
The hemodialysis patients showed significantly smaller volumes than the healthy controls in hippocampal subfields, including the right CA1 (t = –3.52; P < .01), CA4 (t = –3.62; P < .01), fimbria (t = –3.24, P < .01), GC-ML-DG (t = –3.79, P < .01), HATA (t = –3.47, P < .01), molecular layer of the hippocampus (t = –4.73, P < .01) and right whole hippocampus (t = –4.25, P < .01) (Table 2, Fig. 3). No other subfields were found to be significantly different between these two groups.
Figure 3:
Results of correlation analyses between hippocampal subfields volume and clinical indices in hemodialysis patients. The correlation results showed that the volume of the left CA1 was negatively correlated with DS, DSST, LDFR and SDCR scores (r = −0.368, P = .027; r = −0.367, P = .030; r = –0.417, P = .011; r = –0.365, P = .029, respectively). The volume of the left HATA volume was negatively correlated with the duration of disease (r = –0.505, P = .002). The volume of right fimbria volume was negatively correlated with the duration of disease (r = –0.376, P = .024) and positively correlated with DS scores, and RBC and Hb levels (r = 0.339, P = .043; r = 0.438, P = .008; r = 0.342, P = .041, respectively). Hb, hemoglobin.
After including eTIV as a covariate, the hemodialysis patients showed significantly smaller subfield volumes than the healthy controls in the left molecular layer of the hippocampus, left whole hippocampus, right hippocampus and increased subfield volumes in the right hippocampal fissure (Supplementary data, Fig. S1).
Relationships between hippocampal subfield volumes and clinical parameters
The correlation results showed that the volume of the left CA1 was negatively correlated with DS, DSST, LDFR and SDCR scores (r = −0.368, P = .027; r = −0.367, P = .030; r = −0.417, P = .011; r = −0.365, P = .029, respectively) (Fig. 3). The volume of the left HATA was negatively correlated with the duration of disease (r = −0.505, P = .002) (Fig. 3). The volume of right fimbria was negatively correlated with the duration of disease (r = −0.376, P = .024) and positively correlated with DS scores and RBC and hemoglobin levels (r = 0.339, P = .043; r = 0.438, P = .008; r = 0.342, P = .041, respectively) (Fig. 3). Other subfield volumes were not statistically correlated with clinical indices.
After including eTIV as a covariate, the correlation results showed that the volume of the left CA1 was negatively correlated with DS scores and LDFR scores (r = −0.404, P = .020; r = −0.386, P = .027, respectively); the volume of the left CA3 was negatively correlated with LDFR scores (r = −0.393, P = .024); the volume of the left CA3 was positively correlated with uric acid and urea (r = 0.349, P = .046, r = 0.491; P = .004, respectively); and the volume of left HATA was negatively correlated with the duration of disease (r = −0.541, P = .001), but positively correlated with uric acid levels (r = 0.362, P = .038). The volume of the left hippocampal fissure was positively correlated with dialysis duration (r = 0.359, P = .040). The volumes of the left presubiculum and left subiculum were positively correlated with creatinine levels (r = 0.382, P = .028; r = 0.352, P = .044, respectively). The volume of the right CA3 was positively correlated with urea levels (r = 0.361, P = .039). The volume of the right fimbria was negatively correlated with the duration of disease (r = −0.350, P = .049) and positively correlated with hemoglobin levels (r = 0.398, P = .022) and RBC levels (r = 0.469, P = .006). The volume of the right presubiculum was positively correlated with DSST scores (r = 0.431, P = .014) (Supplementary data, Fig. S2).
DISCUSSION
The present study demonstrated that hemodialysis patients, in comparison with healthy controls, had statistically significantly smaller volumes in multiple subfields of the hippocampus such as the right CA3, bilateral CA1, CA4, fimbria, GC-ML-DG, HATA, hippocampal tail, molecular layer of the hippocampus, subiculum and bilateral whole hippocampus. Correlation analysis revealed that the volume of the left HATA was also associated with disease duration, indicating that the longer the patient had ESRD, the smaller the HATA might be. Meanwhile, the volume of right fimbria was positively correlated with RBC levels.
The present study found that hemodialysis patients had a smaller bilateral hippocampal volume than healthy controls. The hippocampus plays a vital role in cognitive processing, particularly episodic memory [27]. A previous functional MRI study revealed aberrant spontaneous brain activity in the hippocampus of hemodialysis patients, which was associated with cognition [5]. Li et al. [8] found that cerebral blood flow decreased globally in the hippocampus after hemodialysis initiation. We deduced that the volume decreases in hemodialysis patients might be associated with reduced cerebral blood flow. This should be validated in future studies.
The volume of the CA1 was also found to be smaller in hemodialysis patients. This is supported by the evidence that uremia induces cell death in rat hippocampal CA1 by the increases in aCaMKII/NR2A synaptic complex and subsequent activation of phosphorylated GluR1-containing AMPA [28]. Our results also supported the idea that hippocampal CA1 was vulnerable to neuronal cell death in the uremia state. The CA1 subfield is mainly involved in spatial working memory and pattern separation [29]. CA1 atrophy was related to neuropsychological scores in the present study, which might be partly caused by an increased glutamatergic transmission in the hippocampus [30]. Thus, the decreased volume in the CA1 might be responsible for the impaired working memory in hemodialysis patients.
The molecular layer is located above the subiculum and below the fissure, including parts of the subiculum and the CA region. Recent longitudinal research on the development of hippocampal subregions across adolescence revealed that global cognitive ability also correlated with the developmental trajectory of molecular layer capacity, with higher scoring individuals showing more nonlinear development [31]. Similarly, Wan et al. [32] found that the volume of the molecular layer was associated with both cognitive function and negatively correlated with the frailty index in older adults with cognitive frailty. Total synaptic numbers in the molecular layer were significantly correlated with cognition in individuals with early Alzheimer's disease and mild cognitive dysfunction [33]. Therefore, we deduced that the decreased number of synapses in the molecular layer of the hippocampus might be an important event related to cognitive alterations.
The molecular layer and CA1 are important components for hippocampal information communication: the molecular layer builds a bridge between the subiculum and the CA fields, while the subiculum receives a large amount of input from pyramidal neurons in the CA1 area [34]. The smaller volume in all three subfields indicated abnormalities in hippocampal connectivity. The disruption of CA1 and the molecular layer may play an important role in the development of hemodialysis-related alterations in cognition. Similar findings have been found in children and adolescents with mood disorders [35]. It might be valuable to explore the white matter tracts between these regions to further clarify the role of hippocampal connectivity in hemodialysis.
The CA4 acts as an important brain region for building connections between the hippocampus and other brain areas. The decreased CA4 volume observed here might affect the information communication between the hippocampus and other cerebral regions in hemodialysis patients.
The HATA is a bridge for information communication between the hippocampus and amygdala. A reduced volume of the HATA has been reported in Alzheimer's disease patients [36, 37]. The volume loss of the HATA might lead to a disruption in the hippocampus–amygdala network, which is responsible for information processing [38]. These studies indicated that the HATA might play a crucial role in cognition. Furthermore, Zheng et al. [39] also found that memory performance was associated with the volume of the HATA. The decreased volume in the HATA might be associated with cognitive alterations in hemodialysis patients, but more research should be performed to verify this hypothesis. Moreover, the volumes of the left HATA and right fimbria were negatively correlated with the duration of disease, suggesting progressive atrophy. Interestingly, after including eTIV as a covariate, volumes of the left CA3 and left HATA were found to be positively correlated with uric acid levels, indicating that uric acid may be neuroprotective [40].
Importantly, after including eTIV as a covariate, the hemodialysis patients showed significantly smaller subfield volumes than the healthy controls only in the left molecular layer of the hippocampus and bilateral hippocampus but higher volume in the right hippocampal fissure. A recent study revealed that the dilation in the hippocampal fissure was closely associated with reduced volumes of hippocampal subfields and can be used as a sensitive marker for hippocampal atrophy in mice [41]. Widening hippocampal fissures have also been found in Parkinson's disease [42] and Alzheimer's disease [43]. Increased volume in the hippocampal fissure may be used as an indicator of hippocampal atrophy.
The current study has several limitations that should not be neglected. First, it was a cross-sectional study with a relatively small sample size. A larger cohort and longitudinal design should be considered in the future. Second, ESRD patients without hemodialysis were not included in the present study. Whether the volume loss in these patients was specific to hemodialysis or the ESRD disease per se or other confounding factors remains to be elucidated. Moreover, the hemodialysis patients differed in many parameters from the healthy controls, such as the presence of a fistula, chronic anemia and multidrug treatments, which limits the determination of particular causes of any brain changes.
In summary, the present study revealed smaller hippocampal subfield volumes in hemodialysis patients, which were associated with cognition, disease duration and RBC levels. The present findings improve our understanding of cognition in hemodialysis patients and may be helpful in clarifying the neurobiological mechanisms underlying memory disturbance in the hemodialysis population.
Supplementary Material
ACKNOWLEDGEMENTS
We thank all patients and volunteers for their participation.
Contributor Information
Hui Juan Chen, Department of Radiology, Hainan General Hospital (Hainan Affiliated Hospital of Hainan Medical University), No. 19, Xiuhua St, Xiuying Dis, Haikou, Hainan, P. R. China.
Jie Qiu, Department of Ultrasound, Hainan General Hospital (Hainan Affiliated Hospital of Hainan Medical University), No. 19, Xiuhua St, Xiuying Dis, Haikou, Hainan, P. R. China.
Yonghui Qi, Blood Purification Center, Hainan General Hospital (Hainan Affiliated Hospital of Hainan Medical University), No. 19, Xiuhua St, Xiuying Dis, Haikou, Hainan, P. R. China.
Lili Fu, Department of Radiology, Hainan General Hospital (Hainan Affiliated Hospital of Hainan Medical University), No. 19, Xiuhua St, Xiuying Dis, Haikou, Hainan, P. R. China.
Qingqing Fu, Department of Radiology, Hainan General Hospital (Hainan Affiliated Hospital of Hainan Medical University), No. 19, Xiuhua St, Xiuying Dis, Haikou, Hainan, P. R. China.
Wenyuan Wu, Department of Radiology, Hainan General Hospital (Hainan Affiliated Hospital of Hainan Medical University), No. 19, Xiuhua St, Xiuying Dis, Haikou, Hainan, P. R. China.
Ganmian Dai, Department of Radiology, Hainan General Hospital (Hainan Affiliated Hospital of Hainan Medical University), No. 19, Xiuhua St, Xiuying Dis, Haikou, Hainan, P. R. China.
Feng Chen, Department of Radiology, Hainan General Hospital (Hainan Affiliated Hospital of Hainan Medical University), No. 19, Xiuhua St, Xiuying Dis, Haikou, Hainan, P. R. China.
FUNDING
This work was supported by the Program of Hainan Association for Science and Technology Plans to Youth R & D Innovation (QCXM201919), the National Nature Science Foundation of China (grant number 81801684, 82160327, 81971602), Hainan Provincial Natural Science Foundation of China (grant number 818MS124), Key Science and Technology Project of Hainan Province (ZDYF2021SHFZ239, ZDKJ2021042, ZDYF2021SHFZ246) and Hainan Province Postgraduate Innovative Research Projects (Hys2020-347). This project was supported by the Hainan Province Clinical Medical Center and the Hainan Academician Innovation Platform Fund. This work was also supported by Hainan Provincial Natural Science Foundation of China (819MS118).
DATA AVAILABILITY STATEMENT
The datasets in this study are not available currently because the present data are part of an ongoing longitudinal study and most data are still in collection and ought to be protected. Reasonable requests to obtain the data can be e-mailed to F.C. (fenger0802@163.com).
CONSENT FOR PUBLICATION
Not applicable.
CONFLICT OF INTEREST STATEMENT
The authors declare that there is no conflict of interest.
REFERENCES
- 1. Sharma A, Yabes J, Al Mawed Set al. Impact of cognitive function change on mortality in renal transplant and end-stage renal disease patients. Am J Nephrol 2016;44:462–72. 10.1159/000451059. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Sehgal AR, Grey SF, DeOreo PBet al. Prevalence, recognition, and implications of mental impairment among hemodialysis patients. Am J Kidney Dis 1997;30:41–9. 10.1016/S0272-6386(97)90563-1. [DOI] [PubMed] [Google Scholar]
- 3. Drew DA, Weiner DE, Sarnak MJ.. Cognitive impairment in CKD: pathophysiology, management, and prevention. Am J Kidney Dis 2019;74:782–90. 10.1053/j.ajkd.2019.05.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Eichenbaum H. A cortical-hippocampal system for declarative memory. Nat Rev Neurosci 2000;1: 41–50. 10.1038/35036213. [DOI] [PubMed] [Google Scholar]
- 5. Chen HJ, Qiu J, Fu Qet al. Alterations of spontaneous brain activity in hemodialysis patients. Front Hum Neurosci 2020;14:278. 10.3389/fnhum.2020.00278. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Li P, Ding D, Ma XYet al. Altered intrinsic brain activity and memory performance improvement in patients with end-stage renal disease during a single dialysis session. Brain Imaging Behav 2018;12:1640–9. 10.1007/s11682-018-9828-x. [DOI] [PubMed] [Google Scholar]
- 7. Wu B, Li X, Zhang Met al. Disrupted brain functional networks in patients with end-stage renal disease undergoing hemodialysis. J Neurosci Res 2020;98:2566–78. 10.1002/jnr.24725. [DOI] [PubMed] [Google Scholar]
- 8. Li X, Slinin YX, Zhang Let al. Cerebral blood flow characteristics following hemodialysis initiation in older adults: a prospective longitudinal pilot study using arterial spin labeling imaging. Neuroimage Clin 2020;28:102434. 10.1016/j.nicl.2020.102434. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Chiu YL, Tsai HH, Lai YJet al. Cognitive impairment in patients with end-stage renal disease: accelerated brain aging? J Formos Med Assoc 2019;118:867–75. 10.1016/j.jfma.2019.01.011. [DOI] [PubMed] [Google Scholar]
- 10. Cao B, Passos IC, Mwangi Bet al. Hippocampal subfield volumes in mood disorders. Mol Psychiatry 2017;22:1352–8. 10.1038/mp.2016.262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Small SA, Schobel SA, Buxton RBet al. A pathophysiological framework of hippocampal dysfunction in ageing and disease. Nat Rev Neurosci 2011;12:585–601. 10.1038/nrn3085. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Iglesias JE, Augustinack JC, Nguyen Ket al. A computational atlas of the hippocampal formation using ex vivo, ultra-high resolution MRI: application to adaptive segmentation of in vivo MRI. Neuroimage 2015;115:117–37. 10.1016/j.neuroimage.2015.04.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Ho NF, Iglesias JE, Sum MYet al. Progression from selective to general involvement of hippocampal subfields in schizophrenia. Mol Psychiatry 2017;22:142–52. 10.1038/mp.2016.4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Nasreddine ZS, Phillips NA, Bédirian Vet al. J Am Geriatr Soc MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc 2005;53:695–9. 10.1111/j.1532-5415.2005.53221.x. [DOI] [PubMed] [Google Scholar]
- 15. Dale AM, Fischl B, Sereno MI.. Cortical surface-based analysis. I. Segmentation and surface reconstruction. Neuroimage 1999;9:179–94. 10.1006/nimg.1998.0395. [DOI] [PubMed] [Google Scholar]
- 16. Fischl B, Sereno MI, Dale AM.. Cortical surface-based analysis. II: inflation, flattening, and a surface-based coordinate system. Neuroimage 1999;9:195–207. 10.1006/nimg.1998.0396. [DOI] [PubMed] [Google Scholar]
- 17. Fischl B, Salat DH, Busa Eet al. Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron 2002;33:341–55. 10.1016/S0896-6273(02)00569-X. [DOI] [PubMed] [Google Scholar]
- 18. Fischl B, van der Kouwe A, Destrieux Cet al. Automatically parcellating the human cerebral cortex. Cereb Cortex 2004;14:11–22. 10.1093/cercor/bhg087. [DOI] [PubMed] [Google Scholar]
- 19. Sled JG, Zijdenbos AP, Evans AC.. A nonparametric method for automatic correction of intensity nonuniformity in MRI data. IEEE Trans Med Imaging 1998;17:87–97. 10.1109/42.668698. [DOI] [PubMed] [Google Scholar]
- 20. Fischl B, Liu A, Dale AM.. Automated manifold surgery: constructing geometrically accurate and topologically correct models of the human cerebral cortex. IEEE Trans Med Imaging 2001;20:70–80. 10.1109/42.906426. [DOI] [PubMed] [Google Scholar]
- 21. Ségonne F, Pacheco J, Fischl B.. Geometrically accurate topology-correction of cortical surfaces using nonseparating loops. IEEE Trans Med Imaging 2007;26:518–29. 10.1109/TMI.2006.887364. [DOI] [PubMed] [Google Scholar]
- 22. Desikan RS, Ségonne F, Fischl Bet al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage 2006;31:968–80. 10.1016/j.neuroimage.2006.01.021. [DOI] [PubMed] [Google Scholar]
- 23. Destrieux C, Fischl B, Dale Aet al. Automatic parcellation of human cortical gyri and sulci using standard anatomical nomenclature. Neuroimage 2010;53:1–15. 10.1016/j.neuroimage.2010.06.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Klein A, Tourville J.. 101 labeled brain images and a consistent human cortical labeling protocol. Front Neurosci 2012;6:171. 10.3389/fnins.2012.00171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Fan L, Li H, Zhuo Jet al. The human brainnetome atlas: a new brain atlas based on connectional architecture. Cereb Cortex 2016;26:3508–26. 10.1093/cercor/bhw157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Buckner RL, Head D, Parker Jet al. A unified approach for morphometric and functional data analysis in young, old, and demented adults using automated atlas-based head size normalization: reliability and validation against manual measurement of total intracranial volume. Neuroimage 2004;23:724–38. 10.1016/j.neuroimage.2004.06.018. [DOI] [PubMed] [Google Scholar]
- 27. Kurkela KA, Dennis NA.. Event-related fMRI studies of false memory: an activation likelihood estimation meta-analysis. Neuropsychologia 2016;81:149–67. 10.1016/j.neuropsychologia.2015.12.006. [DOI] [PubMed] [Google Scholar]
- 28. Kim JW, Ha GY, Jung YW.. Chronic renal failure induces cell death in rat hippocampal CA1 via upregulation of αCaMKII/NR2A synaptic complex and phosphorylated GluR1-containing AMPA receptor cascades. Kidney Res Clin Pract 2014;33:132–8. 10.1016/j.krcp.2014.07.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Kesner RP, Gilbert PE, Wallenstein GV.. Testing neural network models of memory with behavioral experiments. Curr Opin Neurobiol 2000;10:260–5. 10.1016/S0959-4388(00)00067-2. [DOI] [PubMed] [Google Scholar]
- 30. Natale G, Calabrese V, Marino Get al. Effects of uremic toxins on hippocampal synaptic transmission: implication for neurodegeneration in chronic kidney disease. Cell Death Discov 2021;7:295. 10.1038/s41420-021-00685-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Tamnes CK, Bos MGN, van de Kamp FCet al. Longitudinal development of hippocampal subregions from childhood to adulthood. Dev Cogn Neurosci 2018;30:212–22. 10.1016/j.dcn.2018.03.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Wan M, Ye Y, Lin Het al. Deviations in hippocampal subregion in older adults with cognitive frailty. Front Aging Neurosci 2020;12:615852. 10.3389/fnagi.2020.615852. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Scheff SW, Price DA, Schmitt FAet al. Hippocampal synaptic loss in early Alzheimer's disease and mild cognitive impairment. Neurobiol Aging 2006;27:1372–84. 10.1016/j.neurobiolaging.2005.09.012. [DOI] [PubMed] [Google Scholar]
- 34. Behr J, Wozny C, Fidzinski Pet al. Synaptic plasticity in the subiculum. Prog Neurobiol 2009;89:334–42. 10.1016/j.pneurobio.2009.09.002. [DOI] [PubMed] [Google Scholar]
- 35. Tannous J, Amaral-Silva H, Cao Bet al. Hippocampal subfield volumes in children and adolescents with mood disorders. J Psychiatr Res 2018;101:57–62. 10.1016/j.jpsychires.2018.03.003. [DOI] [PubMed] [Google Scholar]
- 36. Gabere M, Thu Pham NT, Graff-Radford Jet al. Automated hippocampal subfield volumetric analyses in atypical Alzheimer's disease. J Alzheimers Dis 2020;78:927–37. 10.3233/JAD-200625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Parker TD, Slattery CF, Yong KXXet al. Differences in hippocampal subfield volume are seen in phenotypic variants of early onset Alzheimer's disease. Neuroimage Clin 2019;21:101632. 10.1016/j.nicl.2018.101632. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Foo H, Mak E, Chander RJet al. Associations of hippocampal subfields in the progression of cognitive decline related to Parkinson's disease. Neuroimage Clin 2017;14:37–42. 10.1016/j.nicl.2016.12.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Zheng F, Cui D, Zhang Let al. The volume of hippocampal subfields in relation to decline of memory recall across the adult lifespan. Front Aging Neurosci 2018;10:320. 10.3389/fnagi.2018.00320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Viggiano D, Wagner CA, Martino Get al. Mechanisms of cognitive dysfunction in CKD. Nat Rev Nephrol 2020;16:452–69. 10.1038/s41581-020-0266-9. [DOI] [PubMed] [Google Scholar]
- 41. Li Y, Yan J, Zhu Xet al. Increased hippocampal fissure width is a sensitive indicator of rat hippocampal atrophy. Brain Res Bull 2018;137:91–7. 10.1016/j.brainresbull.2017.11.014. [DOI] [PubMed] [Google Scholar]
- 42. Lenka A, Ingalhalikar M, Shah Aet al. Hippocampal subfield atrophy in patients with Parkinson's disease and psychosis. J Neural Transm 2018;125:1361–72. 10.1007/s00702-018-1891-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Bastos-Leite AJ, van Waesberghe JH, Oen ALet al. Hippocampal sulcus width and cavities: comparison between patients with Alzheimer disease and nondemented elderly subjects. AJNR Am J Neuroradiol 2006;27:2141–5. [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets in this study are not available currently because the present data are part of an ongoing longitudinal study and most data are still in collection and ought to be protected. Reasonable requests to obtain the data can be e-mailed to F.C. (fenger0802@163.com).