Introduction

Alzheimer’s disease (AD) is the most common neurodegenerative disease causing cognitive decline among elderly individuals [1]. Extracellular neuritic plaques consisting of β-amyloid (Aβ) and intracellular neurofibrillary tangles formed by phosphorylated tau are the two most characteristic pathological hallmarks of AD [2]. Pathological proteins can efflux into the periphery via the blood‒brain barrier (BBB) and glymphatic system [3]. The peripheral system shows physiological abilities of scavenging Aβ and tau [4, 5]. A growing number of studies suggests that abnormal metabolism of peripheral Aβ and tau plays a crucial role in AD pathogenesis [6,7,8,9,10]. In vitro and in vivo studies indicate that the kidneys may be vital organs involved in the peripheral clearance of Aβ and tau [5, 11]. Previous studies have shown that a decrease in the estimated glomerular filtration rate (eGFR) is associated with an increase in blood Aβ levels [12, 13]. In addition, patients with chronic kidney disease (CKD) are reported to be more prone to developing cognitive impairment and AD [14, 15]. However, the effects of deficiency in the clearance of blood AD-related proteins by kidney on brain AD-type pathologies in humans remain largely unknown. Therefore, in the present study, we aimed to investigate the associations of eGFR with Aβ and tau levels in blood as well as cerebrospinal fluid (CSF).

Materials and Methods

Study Participants

To analyze the associations of eGFR with plasma Aβ and tau levels, we recruited 41 cognitively normal CKD patients and 40 age- and sex-matched cognitively normal controls with normal renal function from Daping Hospital between January and October 2018. All CKD patients were recently diagnosed and had not initiated any type of dialysis. In addition, 42 cognitively normal patients (≥ 60 y) with CKD and 150 healthy aged controls who had undergone surgical treatment for noninflammatory disorders (e.g., benign prostatic hyperplasia, stress incontinence) were recruited in the same hospital from January 2019 to December 2022. Therefore, we were able to collect CSF samples during lumbar anesthesia before surgery to analyze the associations of eGFR with CSF AD biomarkers.

Subjects in the above three groups were excluded if they met any of the following criteria: (1) abnormal cognition assessed by the Chinese version of the Mini-Mental State Examination (MMSE); (2) a family history of AD or other dementia; (3) severe cardiac, pulmonary, hepatic, or neurological diseases; (4) cancers; and (5) unwillingness to participate in the present study. Written consent was obtained from all participants or their legal representatives. This study was approved by the Institutional Review Board of Daping Hospital.

CKD Diagnosis

The diagnosis of CKD was based on the 2012 clinical practice guideline published by the Kidney Disease: Improving Global Outcomes (KDIGO) organization [16]. In brief, the patients had suffered a continuing abnormality in kidney structure or function for more than 3 months, including eGFR ≤ 60 mL/min/1.73 m2 or urine albumin-to-creatinine ratio ≥ 30 mg/g. eGFR was calculated with the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation based on serum creatinine levels [17]. We collected the demographic characteristics and medical history data of all participants.

CSF and Plasma Sampling and Processing

Fasting blood was collected between 06:00 and 07:00 to avoid potential circadian rhythm effects. The blood samples were centrifuged at 2,000 × g for 10 min within one hour after standing and then stored at − 80 °C until use. CSF was also stored at -80 °C after centrifugation at 2,000 × g for 10 min.

Measurements of Aβ and tau Levels

The plasma Aβ40, Aβ42 and total tau (T-tau) levels were measured by the commercially available single-molecule array (SIMOA) Human Neurology 3-Plex A assay kit (Quanterix, Lexington, Massachusetts) on the automated SIMOA HD-1 analyzer. The CSF levels of Aβ, total tau (T-tau) and phosphorylated tau-181 (P-tau) were determined by human Aβ and tau enzyme-linked immunosorbent assay (ELISA) kits (INNOTEST, Fujirebio, Belgium) according to the manufacturer’s protocol.

Statistical Analysis

Data are presented as the mean ± standard deviation (SD) unless otherwise stated. The normality of the data was evaluated by the Kolmogorov–Smirnov test. Then, a two-tailed independent t test was used to compare data with a normal distribution, while nonnormal data were compared using the Mann‒Whitney U test. The chi-squared test was used to compare categorical variable data. Correlations of eGFR with AD biomarkers were tested by partial correlation analyses adjusted by age, sex and APOE ε4 genotype, and r represents the partial correlation coefficient. We defined two-sided p values less than 0.05 as statistically significant. The statistical analyses were performed with Statistical Product Service Solutions (SPSS), version 25.0 (SPSS Software, USA).

Results

Increased Plasma Aβ and tau Levels in CKD Patients

First, 41 clinically diagnosed CKD patients and 40 age- and sex-matched cognitively normal controls with normal renal function were recruited to investigate the associations of plasma Aβ and tau levels with renal function. As shown in Table 1, there was no significant difference in the comorbidities of hypertension, diabetes, and hyperlipidemia between groups. As expected, the eGFR of CKD patients was lower than that of controls (23.57 ± 17.56 mL/min/1.73 m2 vs. 124.1 ± 19.95 mL/min/1.73 m2, p < 0.001). CKD patients had significantly higher plasma levels of Aβ40 (245.9 ± 209.5 pg/ml vs. 65.44 ± 66.51 pg/ml, p < 0.001) and Aβ42 (15.57 ± 14.86 pg/ml vs. 3.01 ± 3.06 pg/ml, p < 0.001) than controls (Fig. 1A and B). Moreover, plasma T-tau levels were also increased in CKD patients (14.59 ± 10.54 pg/ml vs. 6.37 ± 4.59 pg/ml, p < 0.001) (Fig. 1C).

Table 1 Characteristics of the participants with plasma samples
Full size table
Fig. 1
figure 1

Comparisons and correlations of plasma Aβ and tau levels in the participants. A-C Comparisons of plasma Aβ40, Aβ42 and T-tau levels between the controls and patients with CKD. D-F Correlations of eGFR with plasma Aβ40, Aβ42 and T-tau levels in subjects with CKD, controls, and all subjects. eGFR, estimated glomerular filtration rate; CKD, chronic kidney disease; CON, controls; T-tau, total tau. *** denotes p < 0.001. Partial correlation analyses were adjusted by age and sex. The shaded areas represent the 95% confidence intervals

Full size image

Correlations of Plasma Aβ and tau Levels with Renal Function

Then, we analyzed the correlations of plasma Aβ and tau levels with eGFR adjusted for age and sex. As shown in Fig. 1D-F, plasma Aβ levels were negatively correlated with eGFR in both the CKD group (Aβ40: r = -0.570, p < 0.001; Aβ42: r = -0.560, p < 0.001) and all subjects (Aβ40: r = -0.607, p < 0.001; Aβ42: r = -0.577, p < 0.001) but not in the control group (Aβ40: r = -0.102, p = 0.540; Aβ42: r = 0.048, p = 0.786). In addition, plasma T-tau levels were also negatively correlated with eGFR in CKD patients (r = -0.461, p = 0.003), controls (r = -0.376, p = 0.020) and all subjects (r = -0.533, p < 0.001).

Decrease in CSF Aβ and Increase in CSF Tau/Aβ42 Levels in CKD Patients

To further reveal the relationships between renal function and brain AD-type pathologies, we evaluated another group of CKD patients and controls and collected CSF samples. The demographic characteristics of the participants are shown in Table 2. There were no significant differences in age, sex, education level, comorbidities, APOE ε4 genotype or MMSE scores between groups. CKD patients had significantly lower levels of eGFR than controls (73.81 ± 15.95 mL/min/1.73 m2 vs. 138.5 ± 26.91 mL/min/1.73 m2, p < 0.001).

Table 2 Characteristics of the participants with CSF samples
Full size table

As shown in Fig. 2A-B, CSF levels of Aβ40 (11,595 ± 5629 pg/ml vs. 13,362 ± 3734 pg/ml, p = 0.020) and Aβ42 (1248 ± 518.9 pg/ml vs. 1557 ± 420.7 pg/ml, p < 0.001) were significantly decreased in CKD patients compared with controls. Although there was no significant difference in CSF P-tau levels between groups (48.78 ± 17.34 pg/ml vs. 46.36 ± 15.06 pg/ml, p = 0.374), the CSF T-tau levels showed an increasing trend in patients with CKD (225.9 ± 95.71 pg/ml vs. 194.1 ± 78.17 pg/ml, p = 0.063) (Fig. 2C-D). Additionally, CKD patients had significantly higher levels of CSF T-tau/Aβ42 (0.22 ± 0.15 vs. 0.14 ± 0.074, p < 0.001) and P-tau/Aβ42 (0.046 ± 0.027 vs. 0.032 ± 0.016, p = 0.001) than controls (Fig. 2E-F).

Fig. 2
figure 2

Comparisons of CSF Aβ40, Aβ42, T-tau, P-tau, T-tau/Aβ42 and P-tau/Aβ42 levels between subjects with CKD and controls. eGFR, estimated glomerular filtration rate; CSF, cerebrospinal fluid; T-tau, total tau; P-tau, phosphorylated tau-181. * denotes p < 0.05, ** denotes p < 0.01, *** denotes p < 0.001. NS denotes no statistical significance

Full size image

Correlations of CSF AD Biomarkers with Renal Function

Partial correlation analyses adjusted for age, sex and APOE ε4 genotype were used to analyze the correlations of CSF AD biomarkers with eGFR. No correlations of eGFR with CSF Aβ40 (CKD patients: r = 0.179, p = 0.297; controls: r = -0.063, p = 0.457; all participants: r = 0.107, p = 0.151) or Aβ42 (CKD patients: r = -0.126, p = 0.486; controls: r = -0.095, p = 0.255; all participants: r = 0.103, p = 0.170) levels were found (Fig. 3A-B). CSF T-tau levels were negatively correlated with eGFR in all subjects (r = -0.145, p = 0.048), and CSF P-tau levels were negatively correlated with eGFR in controls (r = -0.188, p = 0.022) (Fig. 3C-D). As previous studies indicate that the ratio of P-tau to Aβ42 can reflect cerebral Aβ pathology to some extent [18], we further analyzed the correlations of CSF tau/Aβ42 with eGFR. eGFR was negatively correlated with CSF T-tau/Aβ42 (r = -0.185, p = 0.013) and P-tau/Aβ42 (r = -0.152, p = 0.041) in all subjects (Fig. 3E-F). However, there was no correlation between CSF tau/Aβ42 levels and eGFR in CKD patients (T-tau/Aβ42: r = -0.028, p = 0.880; P-tau/Aβ42: r = 0.175, p = 0.329) or controls (T-tau/Aβ42: r = 0.001, p = 0.990; P-tau/Aβ42: r = -0.038, p = 0.651).

Fig. 3
figure 3

Correlations of eGFR with the CSF Aβ40, Aβ42, T-tau, P-tau, T-tau/Aβ42 and P-tau/Aβ42 levels in subjects with CKD, controls, and all subjects. eGFR, estimated glomerular filtration rate; CSF, cerebrospinal fluid; T-tau, total tau; P-tau, phosphorylated tau-181. Partial correlation analyses were adjusted by age, sex and APOE ε4 genotype. The shaded areas represent the 95% confidence intervals

Full size image

Correlations of MMSE Scores with Renal Function

Finally, we examined the correlations of eGFR with cognitive function. As shown in Table 3, MMSE scores were positively correlated with eGFR in controls (r = 0.202, p = 0.013) and all participants (r = 0.211, p = 0.003). These correlations remained significant after adjusting for age, sex, APOE ε4 genotype and education level. No correlation of the MMSE scores with eGFR was found in CKD patients regardless of adjusting for confounders. These findings further indicate that cognitive status may decrease with renal insufficiency.

Table 3 Correlations of MMSE scores with eGFR in subjects with CKD, controls, and all subjects
Full size table

Discussion

Our study showed that CKD patients had higher levels of Aβ and T-tau in plasma and lower levels of Aβ and higher levels of T-tau/Aβ42 and P-tau/Aβ42 in CSF. Both plasma Aβ and T-tau levels were negatively correlated with eGFR. In addition, we found that eGFR was negatively correlated with CSF levels of T-tau, T-tau/Aβ42 and P-tau/Aβ42 but positively correlated with cognitive function. These results suggest that renal function may be involved in the pathogenesis of AD.

Previous investigations have indicated that the clearance of Aβ and tau beyond the brain plays critical roles in the progression of AD [19,20,21]. It is estimated that 40–60% of Aβ and approximately 19% of tau from the brain is cleared through transport to the periphery [5, 8, 22, 23]. As the major excretory organs, the kidneys may play an important role in the clearance of AD pathogenic peptides. Radioactivity can be detected in the kidneys after injecting radioactive iodine-labeled Aβ or tau into the brain [5, 8]. In addition, an animal study showed that the serum Aβ levels in the renal artery are higher than those in the renal vein, which further supports the critical role of the kidneys in clearing Aβ [11]. Thus, CKD patients have increased Aβ levels in the blood compared with healthy patients [12]. Previous studies have shown that tau can be cleared by peripheral organs, including the kidneys, in animals [5], but the detailed mechanism remains unknown. Our study provides clinical evidence for the roles of the kidneys in the clearance of peripheral Aβ as well as tau.

Although several studies have demonstrated the associations of renal function with blood Aβ and tau in CKD [12, 24], it remains unknown whether renal function are related to CSF levels of Aβ and tau, which could represent brain AD pathology. According to a systemic view of AD pathogenesis, abnormal metabolism of Aβ in the periphery can aggravate brain AD pathology [6, 25, 26]. In this study, we found that renal function may be related to brain Aβ deposition, as reflected by decreased CSF Aβ and increased CSF P-tau/Aβ42 in CKD patients. This further confirmed the systemic view of AD. In light of the Aβ cascade hypothesis, the increase in cerebral Aβ levels caused by renal failure can further induce the hyperphosphorylation of tau [11, 27, 28]. In addition, the decrease in GFR can impair aluminum excretion, which could lead to increased levels of p-tau in the CSF and AD-related pathological changes [29, 30]. Consequently, renal insufficiency may affect AD-related Aβ and tau pathology in the brain.

There are several theories linking chronic kidney failure to cognitive impairment, including the kidney-brain-axis theory [31]. Our study showed that MMSE scores decreased with lower eGFR values. Previous studies have also shown that a metabolic imbalance of homocysteine caused by CKD plays an important role in the pathogenesis of AD through the activation of oxidative stress [32, 33]. Massive accumulation of uremic toxins occurs due to renal insufficiency, and these toxins exert neurotoxic effects and aggravate cognitive impairment [34]. Moreover, the disturbance of angiotensin II in chronic kidney failure also promotes brain damage in AD [35, 36]. As a result, patients with CKD are more prone to hypomnesia and AD comorbidities [14]. Epidemiological investigation reported that lower eGFR was associated with a higher risk of dementia [15, 37, 38]. A postmortem study showed that patients with impaired kidney function had a higher risk of AD dementia and a higher burden of cerebral amyloid angiopathy [39]. However, a cohort study in Germany noted that impaired kidney function was associated with higher blood neurofilament and p-tau181 levels but not with AD or all-cause dementia risk [24]. As the author discussed in the limitations, the participants in this study were younger at baseline, the incidence of dementia was comparatively low, and the clinical diagnoses of dementia in this study may have resulted in misdiagnosis or underdiagnosis, which may partially explain the inconsistent results compared with previous studies. It showed that lower eGFR was associated with a reduction in cortical brain volume, although eGFR-related brain atrophy was not selective for regions typically affected by AD [40]. This study suggests that AD may not be a leading factor in the development of brain pathologies related to CKD but may coexist with vascular etiologies of reduced brain volume, but this needs to be further investigated in additional studies. Therefore, our study not only confirms the relationships of renal dysfunction with cognitive impairment in elderly individuals but also provides clinical evidence that renal dysfunction is related to brain AD-type pathological changes.

Our results also bring attention to the diagnostic efficiency of p-tau in the periphery. Recent studies have found that pathological changes in the brains of AD patients can be reflected by plasma P-tau217 and P-tau181 levels, which show excellent diagnostic performance [41, 42]. However, the above research noted that renal insufficiency may affect AD diagnosis based on levels of plasma P-tau. We further provided evidence for this issue. Thus, adjusting for renal function is needed for early diagnosis of AD based on plasma P-tau levels in the future.

It is noteworthy that we were unable to determine causality for the relationship between renal function and changes in AD biomarkers due to the cross-sectional design of our study. Longitudinal studies including more CKD patients with eGFR < 30 mL/min/1.73 m2 should be carried out to better clarify the relationships between renal failure and brain AD-type pathologies and cognitive impairment in the future. Dialysis is an effective measure to treat renal failure. The peripheral levels of Aβ in patients with CKD can be effectively reduced after hemodialysis or peritoneal dialysis [12, 43]. Hemodialysis was also found to reduce Aβ deposition in the brains of AD patients [44]. However, whether dialysis decreases the burden of cerebral Aβ and tau and improves symptoms of cognitive decline in CKD and AD patients remains to be further studied.

In summary, our study found that the decline in renal function was correlated with cognitive decline and abnormal AD biomarker levels in plasma as well as CSF. This study provides human evidence that renal function may be involved in the pathogenesis of AD, further indicating that AD may be a systemic disease.