Brain–kidney connection: Impact on cerebral blood flow with estimated glomerular filtration rates in healthy brains

To the Editor: Brain health is essential for overall well-being, as the brain regulates bodily functions. Cerebral blood flow (CBF) plays a crucial role in sustaining brain function and metabolism, making it a vital parameter for evaluating brain health and diagnosing and monitoring various neurological disorders.^[1] Magnetic resonance imaging (MRI) supports the diagnosis and monitoring of neurological disorders by quantifying CBF.^[1] Arterial spin labeling (ASL) perfusion MRI non-invasively labels arterial blood water magnetically, using it as an endogenous tracer.^[1] In clinical settings, the pseudo-continuous ASL (pCASL) technique represents the standard ASL approach to labeling inflowing blood.^[1]

The estimated glomerular filtration rate (eGFR) measures kidney function and determines the stage of kidney disease.^[2] A normal eGFR is ≥60 mL/min per 1.73m², whereas an eGFR <60 mL/min per 1.73m² for 3 months or more, may suggest chronic kidney disease. The eGFR is a useful tool for screening, diagnosing, and monitoring renal disease and adjusting the dosage of some medications cleared by the kidneys. The association between CBF and eGFR may indicate the influence of renal function on brain health and the risk of neurological disorders, such as stroke, dementia, and cognitive impairment.^[3] Measuring CBF and eGFR may help assess brain and kidney health. Investigating the relationship between CBF and eGFR can provide insights into the interplay between the brain and the kidney and the impact of renal disease on brain function and health. The objectives of this study are as follows: (1) to investigate differences in CBF and brain tissue volume among three different groups of eGFR levels, and (2) to investigate the correlation between CBF values or brain tissue volume loss and eGFR values.

This cross-sectional retrospective study obtained approval from the Institutional Review Board (IRB khnmc 2023-06-008). All participants provided written informed consent for the brain MRI examination. Detailed inclusion and exclusion criteria are outlined in the Supplementary Methods, http://links.lww.com/CM9/C987. Supplementary Figure 1, http://links.lww.com/CM9/C987 provides a flowchart detailing the participant inclusion and exclusion process. There are 368 participants in Stage 1, 287 in Stage 2, and 23 in Stage 3. The specific approach for eGFR calculation is also described in the Supplementary Methods, http://links.lww.com/CM9/C987.

Detailed information concerning imaging acquisition and processing can be found in the Supplementary Methods, http://links.lww.com/CM9/C987. Briefly, brain MRI was conducted on a 3T MRI system (Ingenia or Achieva, Philips Healthcare, Best, The Netherlands). Three-dimensional (3D) sagittal T1-weighted (T1W), axial 2-dimensional turbo spin-echo, fluid attenuation inversion recovery, and pCASL images were acquired from each participant. Image post-processing was conducted using MATLAB software (MathWorks, Natick, MA, USA) and Statistical Parametric Mapping version 12 (SPM12) software (Wellcome Department of Imaging Neuroscience, University College, London, UK).

The following statistical analyses were conducted. First, the analysis of variance (ANOVA) test was employed to evaluate continuous variables among the three participant groups, using the Bonferroni post hoc test for further comparisons. The chi-squared test was utilized to assess differences in sex between the groups. Second, a voxel-based analysis of covariance (ANCOVA) analysis was carried out to compare CBF, gray matter volume (GMV), and white matter (WMV) maps across the three groups, with participants’ age included as a covariate. Additionally, a voxel-based multiple regression analysis was performed to evaluate the relationship between CBF, GMV, or WMV maps and eGRF or creatinine values, controlling for participants’ age without separating the groups. Finally, the region-of-interest (ROI) areas were defined, and ROI values were obtained in several brain areas. Detailed information on defining ROIs can be found in the Supplementary Methods, http://links.lww.com/CM9/C987. An ANCOVA was conducted to assess group differences, again controlling for participants’ age with Bonferroni correction applied. Furthermore, the Pearson partial correlation test was used to examine the correlation between ROI values and eGRF or creatinine values across all participants, with age as a control factor.

The results of the statistical analysis of participants’ demographic data, creatinine levels, and eGFR are summarized in Supplementary Table 1, http://links.lww.com/CM9/C987. Age did not differ significantly among the three groups (F = 0.182, P = 0.834). As expected, creatinine and eGFR levels showed significant differences among the groups, as they were categorized based on eGFR levels. Additionally, sex did not differ significantly among the three groups (all P >0.05).

Figure 1 shows the results of the voxel-based ANCOVA [Figure 1A] and multiple regression [Figure 1B] analyses of CBF. For the ANCOVA analysis, CBF was significantly higher in Stage 1 than in Stages 2 and 3 in most of the brain areas shown in one large cluster [Supplementary Table 2, http://links.lww.com/CM9/C987]. Furthermore, CBF was significantly higher in Stage 2 than in Stage 3 in most of the brain areas shown in one large cluster [Supplementary Table 2, http://links.lww.com/CM9/C987]. Both GMV and WMV were not significantly different between Stages 1 and 2, but were significantly higher in both Stages 1 and 2 than in Stage 3 in some brain areas [Supplementary Figure 2, http://links.lww.com/CM9/C987]. For the multiple regression analysis, we found that CBF was positively associated with eGFR levels but was significantly negatively associated with creatinine levels [Supplementary Table 3, http://links.lww.com/CM9/C987]. Both GMV and WMV were not significantly associated with eGFR but were significantly negatively associated with creatinine levels in small brain areas [Supplementary Figure 2, http://links.lww.com/CM9/C987].

Figure 1.

Participant information and results from the voxel-based ANCOVA and multiple regression analyses of CBF. (A) Results of the voxel-based ANCOVA analyses of CBF. CBF was significantly higher in Stage 1 than in Stages 2 and 3 in the most of brain areas shown in one large cluster. Furthermore, CBF was significantly higher in Stage 2 than in Stage 3 in the most of brain areas shown in one large cluster. (B) Results of the voxel-based multiple regression test of CBF maps with eGFR or creatinine levels with the participant’s age as a covariate for all participants together. CBF was positively associated with eGFR levels but was negatively associated with creatinine levels. Stage 1 included participants with the normal range of the eGFR value between 90 to less than 120 mL/min per 1.73m², Stage 2 is for the participants with the normal range of the eGFR value between 60 to less than 90 mL/min per 1.73m², and Stage 3 is for the participants with the abnormal rangeof the eGFR value between 30 to less than 60 mL/min per 1.73m². ANCOVA: Analysis of covariance; CBF: Cerebral blood flow; eGFR: Estimated glomerular filtration rate.

Supplementary Table 4, http://links.lww.com/CM9/C987 summarizes the mean values and standard deviations of left and right CBF, GMV, and WMV for each ROI, along with the results of the ANCOVA and independent t-test analyses for these values. The ANCOVA analysis revealed significant differences in CBF values across the 3 stages in several brain regions. Specifically, CBF values differed significantly between Stage 1 and Stage 2 in the left hippocampus (P = 0.026) and both the right (P = 0.037) and left (P = 0.041) parahippocampal gyrus. Supplementary Table 5, http://links.lww.com/CM9/C987 presents the results of the Pearson partial correlation analysis between MRI ROI values and eGFR levels. CBF values showed a significantly positive correlation with eGFR levels in all defined ROIs, while GMV and WMV values exhibited significant positive correlations with eGFR levels in some ROIs. Supplementary Table 6, http://links.lww.com/CM9/C987 details the results of the Pearson partial correlation analysis between MRI ROI values and creatinine levels. CBF values demonstrated a significant negative correlation with creatinine levels across all defined ROIs, and GMV and WMV values were also significantly negatively correlated with creatinine levels in several defined ROIs.

Investigating simple blood tests that reflect CBF is crucial for assessing brain health in healthy individuals. This study’s key findings are as follows: First, CBF was significantly greater in individuals with high and moderate eGFRs compared to those with low eGFRs, particularly in the hippocampus and parahippocampal gyrus. This indicates a decline in brain perfusion as renal function worsens. Second, CBF was positively correlated with eGFR and negatively correlated with creatinine across all defined brain regions, suggesting that better kidney function is associated with improved cerebral vascular health. An important finding was that CBF values were sensitive to group differences [Figure 1A], decreasing with worsening renal function even before changes in brain tissue volumes occurred. This suggests that CBF could serve as a potential biomarker for early cerebral vascular or metabolic disturbances. Both CBF and eGFR are key indicators of vascular health. A lowered eGFR may indicate systemic vascular changes affecting the brain, with altered CBF signaling, early brain vascular changes without affecting tissue volumes.

The second important finding was that CBF values were significantly positively correlated with eGFR levels in all defined ROI brain areas [Figure 1B], but GMV and WMV were positively correlated with eGFR levels in limited brain areas [Supplementary Figure 2, http://links.lww.com/CM9/C987]. This suggests that CBF is a more sensitive indicator of changes in renal function and vascular health than GMV and WMV, which reflect more chronic changes. Higher eGFR indicates better vascular conditions, supporting stable cerebral perfusion and leading to higher CBF. Lower eGFR is linked to inflammation and vascular damage, reducing blood flow to organs, including the brain. CBF is dynamic and responds to physiological changes like renal function, while GMV and WMV reflect structural changes from chronic disturbances.

In summary, this study utilized MRI metrics and blood markers (creatinine and eGFR) to investigate brain integrity in a non-diseased population stratified by eGFR levels. It found that CBF decreases with worsening renal function and is positively correlated with eGFR and negatively with creatinine, indicating its potential for monitoring early brain perfusion changes. The hippocampus and parahippocampal gyrus showed particular sensitivity to CBF changes. Significant GMV and WMV differences appeared mainly between normal and reduced renal function stages, suggesting these metrics reflect advanced disturbances compared to CBF, which detects earlier changes. The study highlights the link between renal and brain health, suggesting CBF as a marker for renal function-related brain health. Further research could enhance MRI’s role in evaluating brain health concerning renal function and guide early interventions.

Acknowledgments

The authors appreciate Miss Seon Hwa Lee (Clinical Research Institute, Kyung Hee University Hospital at Gangdong, Seoul, Korea) for providing advice on the statistical analyzes. They also thank Dr. Munbae Lee at Konkuk University (Seoul, Republic of Korea) for the technical support to map CBF.

Funding

This research was supported by a grant of the Korea Dementia Research Project through the Korea Dementia Research Center (KDRC), funded by the Ministry of Health & Welfare and Ministry of Science and ICT (RS-2021-KH112574, G.H.J.), Republic of Korea.

Conflicts of interest

None.

Supplementary Material

cm9-139-3000-s001.pdf ^{(1.4MB, pdf)}