Introduction: Due to anatomical and functional similarities in microvascular beds, the brain and kidney share distinctive susceptibilities to vascular injury and common risk factors of small vessel disease. The aim of this updated meta-analysis is to explore the association between kidney function and the burden of cerebral small vessel disease (CSVD). Methods: PubMed, EMBASE, and Cochrane Library were systematically searched for observational studies that explored the association between the indicators of kidney function and CSVD neuroimaging markers. The highest-adjusted risk estimates and their 95% confidence intervals (CIs) were aggregated using random-effect models. Results: Twelve longitudinal studies and 51 cross-sectional studies with 57,030 subjects met the inclusion criteria of systematic review, of which 52 were included in quantitative synthesis. According to the pooled results, we found that low estimated glomerular filtration rate (eGFR <60 mL/min/1.73 m2) was associated with cerebral microbleeds (odds ratio (OR) = 1.55, 95% CI = 1.26–1.90), white matter hyperintensities (OR = 1.40, 95% CI = 1.05–1.86), and lacunar infarctions (OR = 1.50, 95% CI = 1.18–1.92), but not with severe perivascular spaces (OR = 1.20, 95% CI = 0.77–1.88). Likewise, patients with proteinuria (OR = 1.75, 95% CI = 1.47–2.09) or elevated serum cystatin C (OR = 1.51, 95% CI = 1.25–1.83) also had an increased risk of CSVD. Conclusion: The association between kidney function and CSVD has been comprehensively updated through this study, that kidney insufficiency manifested as low eGFR, proteinuria, and elevated serum cystatin C was independently associated with CSVD burden.

Cerebral small vessel disease (CSVD) is an essential component of cerebrovascular disease. Its neuroimaging findings include recent small subcortical infarcts, white matter hyperintensities (WMH), lacunes, cerebral microbleeds (CMB), perivascular spaces (PVS), and brain atrophy, which reflect intricate pathological changes affecting the cerebral microvessels [1, 2]. Though most SVD lesions are thought to be clinically silent, CSVD has been identified to commonly cause symptomatic stroke, cognitive impairment, and particularly vascular dementia [3‒5]. Furthermore, there has been increasing attention on other manifestations driven by CSVD such as a decline in executive function, gait disturbance, mood disorders, and atypical neurological symptoms. Despite the globally heavy health burden of CSVD, its underlying pathophysiological mechanism is poorly understood.

The brain and kidney share anatomical and functional characteristics which make them sensitive to common cardiovascular risk factors, suggesting cerebrovascular diseases and kidney diseases may have similar pathogenesis [6]. Since the common forms of chronic kidney disease (CKD) are featured by lipohyalinosis and endothelial dysfunction [7, 8], both of which are symbols of small vessel lesions, it could be assumed that kidney insufficiency might be closely related to systemic microvascular condition. Given that CSVD and kidney diseases share many traditional and nontraditional risk factors [9], there is likely to be a potential association between CSVD and kidney insufficiency. Furthermore, both of CSVD and kidney insufficiency were identified to promote the occurrence and progression of cognitive impairment or dementia [4, 10‒13], the non-negligible association between them has drawn much attention so far.

Previous meta-analyses have explored the associations of CSVD with low estimated glomerular filtration rate (eGFR) and proteinuria, showing that kidney impairment might be a risk factor for CSVD [14‒16]. Nonetheless, they only investigated a single traditional indicator (eGFR or proteinuria), without any investigation of cystatin C (CysC), a novel indicator of greater sensibility for kidney function [17]. Additionally, the number of included studies was limited in previous meta-analysis exploring low eGFR, which also hindered advanced analyses. Since an increasing number of studies concerning this topic have been emerging from then on, a more comprehensive meta-analysis referencing the latest evidence is warranted. Hence, after incorporating recent publications to enrich the evidence, we conducted this updated meta-analysis and systematic review with an additional exploration into CysC. The objective of our study is to provide the most up-to-date and comprehensive evaluations to further establish the associations between CSVD imaging markers with predominant indicators of kidney function.

Search Strategy

The study was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guideline (online suppl. Table 1; for all online suppl. material, see www.karger.com/doi/10.1159/000527069) [18], with the protocol prospectively registered on PROSPERO (CRD42021288721). Since our study is an updated meta-analysis of previous work, an updated literature search for studies published from 2013 to October 11, 2021, was performed in electronic databases (PubMed, Embase, and the Cochrane Library) with no language restrictions. The detailed search strategy is shown in online supplementary Method 1. Moreover, the reference lists of eligible studies and relevant reviews were also tracked for any pertinent studies that might be omitted.

Selection Criteria

Studies fulfilled the following criteria were considered eligible: (1) cohort, case-control, and cross-sectional studies enrolling only adult participants; (2) assessing kidney function by eGFR, urine albumin-to-creatinine ratio (ACR), urine protein dipstick measurements, or CysC; low eGFR was defined as eGFR <60 mL/min/1.73 m2; proteinuria was defined by dipstick test positive or albuminuria (ACR ≥30 mg/g or even lower cut-off points); (3) assessing neuroimaging markers of CSVD using magnetic resonance imaging (brain atrophy was excluded from our current analysis for lack of specificity); (4) exploring the associations between the above indicators of kidney function with CSVD markers; (5) reporting quantitative risk estimates via multivariable-adjusted odds ratios (ORs), relative risks, or hazard ratios (HRs) and their 95% confidence intervals (CIs). Animal studies, case reports, editorials, commentaries, hypothesis papers, review articles, as well as meta-analyses were excluded. To avoid complex effects in certain pathophysiological states, we also precluded studies that included patients with end-stage kidney disease (history of dialysis or eGFR <15 mL/min/1.73 m2), kidney transplant, atrial fibrillation, and inherited or genetic small vessel diseases (for example, Fabry’s disease).

Data Extraction and Quality Assessment

Two investigators used a unified format to extract the following characteristics of eligible studies independently: name of the first author, publication year, study design, country, sample size and the proportion of females, mean age of subjects, assessment of kidney function, markers of CSVD, covariates for adjustment, as well as highest-adjusted risk estimates and their 95% CIs (online suppl. Table 2). For studies that did not report risk estimates by OR, HR, or relative risk, if raw data available, effect sizes were then calculated. Additionally, considering that lacunar infarction (LI), lacunes, recent small subcortical infarcts, silent cerebral infarction, and silent brain infarction are different statements or periods referring to the same pathological change, we chose LI to represent all of the above. If there were studies investigating identical populations with the same assessment, we kept the most comprehensive or the latest report to avoid overlapping. As for studies included in quantitative syntheses, the Newcastle Ottawa Scale (NOS) [19] and the Joanna Briggs Institute Critical (JBI) Appraisal tools [20] were used to assess the quality of longitudinal and cross-sectional studies separately (online suppl. Table 3). Studies that scored more than 6 points for both scales were then considered to be of high quality. Disagreements were resolved via wider team discussion until consensuses were reached.

Statistical Analyses

In the first place, we used the fixed model to combine estimates reflecting risk for the same marker of CSVD within any single study, if necessary (for example, when CMB was reported by different locations respectively), and then pooled estimates from different studies by random-effects model for the sake of expected heterogeneity. We assessed heterogeneity via Cochran’s Q test and I2 estimation [21], using cut-offs for substantial (I2 >50%) and high (I2 >75%) heterogeneity referring to the Cochrane Handbook for Systematic Reviews of Interventions [22]. To explore the source of heterogeneity and the potential effects by special confounders, meta-regression was conducted for certain markers of CSVD that were reported in at least ten studies. We introduced the following confounders in meta-regression: region (Asia; other regions), population (general population; patients), publication year (before 2010; after 2010), study design (longitudinal study; cross-sectional study), sample size (≤200; >200), level of adjustment (adjusted estimates for: age and sex; age, sex, and cardiovascular risk factors; age, sex, cardiovascular factors, and other indicators of kidney function); study quality (≤ six points as low quality; > six points as high quality). If the p value <0.05 for any certain confounder, subgroup analysis stratified by this specific confounder would be subsequently performed. In addition, sub-analysis was also conducted for the association between low eGFR with different locations of CMB. Furthermore, sensitivity analyses by omitting a single study at a time were conducted to test the robustness of the pooled results, as well as explore the source of heterogeneity. Publication bias was estimated by Egger’s tests and funnel plots, and adjusted by the trim and fill method. Studies with different statistical approaches which were inappropriate for quantitative synthesis, for example, assessing total CSVD burden or using different cut-offs for eGFR, were then included in the narrative review. All reported probability values were two-tailed, and statistical significance was defined as the level of p value <0.05. Statistical analyses were all performed by STATA 16.0 (Stata Corporation, TX, USA).

Search Results

After removing duplicate records, the search strategy yielded 19,319 records to be initially screened. Among the 121 potentially eligible studies, we further read the full-text line by line and finally identified 38 new articles eligible for inclusion. Combined with 25 records from the previous meta-analyses, a total of 63 studies were ultimately included in this updated meta-analysis. Due to different evaluation measures of kidney function or CSVD, eleven studies inappropriate for quantitative synthesis were included in the systematic review, leaving 52 studies qualified for quantitative synthesis. The whole process of literature selection was summarized in Figure 1.

Fig. 1.

Flowchart of the literature search and study selection process.

Fig. 1.

Flowchart of the literature search and study selection process.

Close modal

Study Characteristics and Quality

Nine cohort studies, three case-control studies, and 51 cross-sectional studies, with a total of 57,030 subjects, met the inclusion criteria. Characteristics of the eligible studies were shown in online supplementary Table 2. To summarize, among the 63 included studies, forty-three studies enrolled participants from Asia (twenty-two from Japan, fifteen from China, and six from Korea), while the remaining 20 studies recruited subjects from European and American countries (nine from the USA, four from the Netherlands, two from England, two from Spain, and one each from Finland, Germany, and Ecuador). As for study population, twenty-three studies were based on general populations, and the rest 40 studies investigated patients with various cerebrovascular diseases, hypertension, diabetes, or CKD. Only 13 studies were reported to have less than 200 subjects, the sample sizes were appreciable in most studies.

Among the 52 studies qualified for quantitative synthesis, thirty-three of which comprising 30,193 subjects assessed the risk of CSVD markers in participants with low eGFR (eGFR <60 mL/min/1.73 m2 vs. eGFR ≥60 mL/min/1.73 m2); nine studies with 9,207 individuals evaluated the quantitative association with eGFR (increase) as a continuous variable; eighteen studies involving 15,323 subjects investigated proteinuria (presence vs. absence); twelve studies containing 13,910 subjects explored ACR (increase); and only six studies with 5,382 subjects studied serum CysC (increase). The scores of quality assessments ranged from five to seven for longitudinal studies (median = 6.25) and from six to eight for cross-sectional studies (median = 7.14) (online suppl. Table 3). Four of the eight longitudinal studies and 30 of the 44 cross-sectional studies were of high quality.

The Association between Kidney Function and CSVD

eGFR and CSVD

Seventeen eligible studies were identified exploring the association between low eGFR and CMB, with a total of 15,344 subjects. As revealed in Figure 2, low eGFR was associated with an increased risk of CMB (OR = 1.55, 95% CI = 1.26–1.90; I2 = 66.6%; Fig. 2a). Likewise, after pooling results from 11 studies and 13 studies separately, low eGFR was also found to be related to WMH (OR = 1.40, 95% CI = 1.05–1.86; I2 = 77.0%; 10,015 subjects; Fig. 2b) and LI (OR = 1.50, 95% CI = 1.18–1.92; I2 = 81.8%; 19,945 subjects; Fig. 2c). Nevertheless, no significant relationship was observed between low eGFR with both centrum semiovale PVS (OR = 1.65, 95% CI = 0.96–2.84; I2 = 51.1%) and basal ganglia PVS (OR = 1.05, 95% CI = 0.60–1.83; I2 = 77.0%; Fig. 2d) according to five studies with 4,269 subjects. Moreover, the pooled result of studies assessing eGFR as a continuous variable demonstrated a lower risk of CSVD in individuals with higher eGFR (OR = 0.97, 95% CI = 0.95–0.98; I2 = 63.9%; online suppl. Fig. 1), which indicated that decreased eGFR, might be a risk factor for incident CSVD.

Fig. 2.

Forest plot of the association between low eGFR and CSVD. Low eGFR (<60 mL/min/1.73 m2) was associated with the increased risk of cerebral microbleeds (a), white matter hyperintensity (b), and lacunar infarction (c). No significant associations between low eGFR with both centrum semiovale PVS and basal ganglia PVS were observed (d). CMB, cerebral microbleeds; CI, confidence interval; eGFR, estimated glomerular filtration; LI, lacunar infarctions; OR, odds ratio; PVS, perivascular spaces; WMH, white matter hyperintensities.

Fig. 2.

Forest plot of the association between low eGFR and CSVD. Low eGFR (<60 mL/min/1.73 m2) was associated with the increased risk of cerebral microbleeds (a), white matter hyperintensity (b), and lacunar infarction (c). No significant associations between low eGFR with both centrum semiovale PVS and basal ganglia PVS were observed (d). CMB, cerebral microbleeds; CI, confidence interval; eGFR, estimated glomerular filtration; LI, lacunar infarctions; OR, odds ratio; PVS, perivascular spaces; WMH, white matter hyperintensities.

Close modal

Proteinuria and CSVD

As illustrated in Figure 3, the overall risk of CSVD was greater in individuals with proteinuria than those without (OR = 1.75, 95% CI = 1.47–2.09; I2 = 69.3%). As for every single CSVD imaging marker, analysis of six studies on the association between proteinuria and CMB found that proteinuria was associated with CMB (OR = 1.75, 95% CI = 1.31–2.35; I2 = 34.1%; 4,927 subjects); nine studies involving 9,250 subjects explored the association between proteinuria and WMH, of which the pooled risk estimate showed an increased WMH risk among individuals with proteinuria (OR = 1.52, 95% CI = 1.14–2.04; I2 = 77.7%); we also observed that patients with proteinuria had 1.75 times the odds of LI (95% CI = 1.37–2.24; I2 = 0.0%, p = 0.814; six studies; 5,141 subjects), and 2.33 times the odds of PVS (95% CI = 1.45–3.74; I2 = 66.3%; three studies; 2,121 subjects) than those without. Furthermore, analyses of studies assessed ACR as a continuous variable also demonstrated that higher ACR level was related to an increased overall risk of CSVD (OR = 1.22, 95% CI = 1.13–1.32; I2 = 73.4%; online suppl. Fig. 2). With respect to unique CSVD imaging markers, increased ACR was found to be associated with WMH (OR = 1.24, 95% CI = 1.05–1.46; I2 = 76.7%), LI (OR = 1.40, 95% CI = 1.19–1.64; I2 = 65.1%), and PVS (OR = 1.19, 95% CI = 1.06–1.33; I2 = 2.5%), while its association with CMB was not significant (OR = 1.07, 95% CI = 0.97–1.18; I2 = 52.0%).

Fig. 3.

Forest plot of the association between proteinuria and CSVD. Proteinuria (dipstick test positive, or ACR ≥30 mg/g or even lower cut-off points) was associated with the burden of CSVD. CMB, cerebral microbleeds; CI, confidence interval; eGFR, estimated glomerular filtration; LI, lacunar infarctions; OR, odds ratio; PVS, perivascular spaces; WMH, white matter hyperintensities.

Fig. 3.

Forest plot of the association between proteinuria and CSVD. Proteinuria (dipstick test positive, or ACR ≥30 mg/g or even lower cut-off points) was associated with the burden of CSVD. CMB, cerebral microbleeds; CI, confidence interval; eGFR, estimated glomerular filtration; LI, lacunar infarctions; OR, odds ratio; PVS, perivascular spaces; WMH, white matter hyperintensities.

Close modal

CysC and CSVD

Among the six articles on the association between serum CysC (as a continuous variable) and CSVD, none of them investigated PVS. But the pooled estimates of other three markers still showed a relationship between elevated serum CysC and the risk of CSVD (OR = 1.51, 95% CI = 1.25–1.83; I2 = 73.9%; Fig. 4). The OR were 2.19 (95% CI = 1.27–3.76; I2 = 80.1%) for CMB after pooling risk estimates from four studies with 1,994 subjects, 1.25 (95% CI = 1.10–1.43; I2 = 38.7%) for LI according to two studies containing 3,391 subjects, and 1.31 (95% CI = 1.07–1.61) for WMH from only one study with 604 subjects, respectively.

Fig. 4.

Forest plot of the association between Cystatin C and CSVD. Cystatin C as continuous variable was associated with the increased risk of CSVD. CMB, cerebral microbleeds; CI, confidence interval; eGFR, estimated glomerular filtration; LI, lacunar infarctions; OR, odds ratio; WMH, white matter hyperintensities.

Fig. 4.

Forest plot of the association between Cystatin C and CSVD. Cystatin C as continuous variable was associated with the increased risk of CSVD. CMB, cerebral microbleeds; CI, confidence interval; eGFR, estimated glomerular filtration; LI, lacunar infarctions; OR, odds ratio; WMH, white matter hyperintensities.

Close modal

Meta-Regression and Subgroup Analyses

Since high heterogeneity has been observed in analyses of the associations between low eGFR with CSVD markers, meta-regression was undertaken to investigate the source of heterogeneity and the modification of main results by potential covariates. We discovered that the association of low eGFR with CMB was influenced by sample size and level of adjustment, and its association with LI was influenced by study quality (online suppl. Table 4). Thereby subgroup analyses by these three factors were consequently performed. For its association with CMB, substantial disparity was identified between studies with sample sizes greater than 200 (OR = 1.36, 95% CI = 1.12–1.64; I2 = 62.2%) and those with relatively small sample sizes (OR = 3.10, 95% CI = 1.97–4.88; I2 = 0%; online suppl. Fig. 3). When stratified by different levels of adjustment, the association between low eGFR and CMB remained significant in studies adjusting for age, sex, and cardiovascular risk factors (OR = 1.73, 95% CI = 1.39–2.16; I2 = 38.3%). Nevertheless, in studies additionally adjusting risk estimates for proteinuria or CysC, the association appeared to be absent (OR = 1.20, 95% CI = 0.93–1.55; I2 = 52.8%; online suppl. Fig. 4). For LI, the association remained significant in high-quality studies (OR = 1.80, 95% CI = 1.44–2.25; I2 = 50.9%) but was no longer significant in studies of low quality (OR = 1.05, 95% CI = 0.86–1.28; I2 = 51.7%; online suppl. Fig. 5). No substantial effects by other confounders were identified according to the results of meta-regression. In addition, according to its different locations, CMB was stratified into strictly lobar (LCMB) and nonlobar CMB (NLCMB, including deep and infratentorial CMB). Subgroup analyses indicated that low eGFR was related to NLCMB (OR = 1.51, 95% CI = 1.13–2.02; I2 = 47.4%; online suppl. Fig. 6) rather than LCMB (OR = 1.12, 95% CI = 0.87–1.42; I2 = 0%).

Publication Bias and Sensitivity Analyses

As suggested by funnel plots and Egger’s tests, there was an evident publication bias in studies reporting the association of low eGFR and CMB (Egger’s test, p = 0.003), whereas the estimate was attenuated slightly (OR = 1.43, 95% CI = 1.51–2.09) after adjusting for publication bias by the trim and fill method (online suppl. Fig. 7). No obvious publication bias was detected in other analyses. Sensitivity analyses further showed the robustness of our results, as the pooled effects did not vary substantially after excluding any single study in each analysis (online suppl. Fig. 8).

Systematic Review

Studies using different statistical approaches which did not allow us to include them in quantitative synthesis were only included in the systematic review. To be specific, one prospective cohort study with 4,083 type 1 diabetics defined proteinuria by urinary albumin excretion rate (UAE). It showed that the corresponding adjusted HRs for LI were 4.3 (95% CI = 1.8–10.0) for patients with microalbuminuria (30 ≤ UAE <300 mg/24 h), and 5.7 (95% CI = 2.5–12.9) for patients with macroalbuminuria (UAE ≥300 mg/24 h) compared with those with normal UAE after a 36,680 person-years’ follow-up [23]. One retrospective cohort study investigating 89 LI patients observed that per 10 mL/min/1.73 m2 decrease in eGFR was associated with higher odds of CMB progression (OR = 1.57, 95% CI = 1.03–2.38) rather than baseline CMB presence [24]. Another retrospective cohort study found that per unit increase of serum CysC was closely associated with higher degree of WMH (OR = 2.14) without providing a CI [25]. Two cross-sectional studies found that the total CSVD burden was higher in patients with low eGFR, proteinuria, and elevated serum CysC, separately [26, 27]. Another cross-sectional study based on general population of 1,937 Japanese found independent associations of CKD with silent brain infarctions (OR = 1.90, 95% CI = 1.18–3.05), WMH (OR = 2.06, 95% CI = 1.48–2.87), and CMB (OR = 3.30, 95% CI = 1.51–7.20) [28]. Nevertheless, other five studies reported inconsistent results on the associations between different measurements of kidney function and CSVD markers [29‒33].

Through this updated systematic review and meta-analysis, the noteworthy relationship between kidney function and CSVD has been further established. Our study suggested that as traditional indicators and gold standard for diagnosing impaired kidney function, low eGFR and proteinuria were both independently related to CSVD, though the association between low eGFR with PVS appeared to be nonsignificant. In addition, as continuous variables, not only eGFR and ACR but also serum CysC were associated with incident CSVD, thus highlighting the role of kidney insufficiency as a potential risk factor for CSVD.

With our search strategy covering a broader range and newly published studies additionally included, the pooled sample size in this current study was more appreciable. Through more comprehensive analyses including meta-regression, subgroup, and sensitivity analyses, along with additional investigation into ACR, and CysC, the main outcomes in our study were consistent with previous meta-analyses [14, 15]. As we know, hypertension is one of the most common risk factors for small vessel disease, particularly kidney disease. Since both hypertension and kidney insufficiency are significant independent risk factors of CSVD [2], our study provides support for the possible role of effective management of hypertensive nephropathy in the prevention of CSVD. However, it is worth noting that the association between low eGFR with PVS, which has not been analyzed in previous meta-analysis, appeared to be nonsignificant, though the relationship between proteinuria and PVS was observed. Since PVS is a relatively novel CSVD marker, the lack of relevant studies limits our understanding of its important role in CSVD [34], resulting in inadequate evidence to identify the association between kidney function and PVS. Unfortunately, studies on CysC are also insufficient, calling for future research to focus more on these important markers.

To investigate the source of high heterogeneity, subgroup analyses based on explanatory variables derived from meta-regression were performed. Despite that the heterogeneity in subgroups has decreased to varying degrees, moderate heterogeneity could still be detected in most subgroups. Nevertheless, there were evident disparities in heterogeneity and pooled estimates between subgroups, demonstrated that sample size, study quality, and the adjustment for other indicators of kidney function did have some modification effects on the main outcomes. After adjustment for other indicators of kidney function, the association was no longer significant, which is possibly due to overadjustment. It is worth noting that though the meta-regression showed that the region had no significant influence on the main results of the meta-analysis, the relationship between CVSD and renal insufficiency in different races is worth discussion. Due to differences in cardiovascular risk factors (especially hypertension) between Asians and Europeans, patients from Asian populations with a longer history of these risk factors would suffer more serious damage of small vessels both in kidney and brain [35, 36]. Previous meta-analysis has also shown that Asians with low eGFR have a higher risk of stroke than non-Asians [37]. These suggest that the influence of low eGFR on CSVD may be more likely and severe for Asians than other populations. In addition, sub-analysis showed that low eGFR was associated with NLCMB rather than LCMB, which was expected and consistent with the previous meta-analysis of proteinuria [15]. Since the leading cause of LCMB was assumed to be cerebral amyloid angiopathy while the presence of NLCMB is mostly due to perforating arteriolar vasculopathy, there were suggestions on distinguishing LCMB from NLCMB [1]. The absence of an association between ACR and CMB might be partly attributed to different main causes between LCMB and NLCMB.

The idea of using another territory where microvasculature is essential as a window to brain function is not new. For instance, retinal microvasculature function could be used to predict CKD [38]. Although there have been numerous epidemiologic and clinical investigations concerning the association between kidney impairment and CSVD, the underlying pathophysiological mechanism is poorly understood. As have been mentioned in Introduction, previous hypotheses presuming that both CSVD and kidney impairment were terminal organ damages mainly driven by traditional cardiovascular risk factors, such as hypertension, diabetes mellitus, hyperlipidemia, etc. However, in our analyses, all included studies adjusted estimates for age and sex, and the overwhelming majority of them further adjusted for traditional cardiovascular risk factors. Along with previous evidence from clinical trials [39], the above evidence indicated that there might be other underlying mechanisms independent of the conventional pattern. To date, the prevailing view is that the association was mainly mediated by a variety of nontraditional risk factors including inflammation, oxidative stress, uremic toxins, mineral bone disorder, etc. [40] These factors could contribute to variety of pathophysiological changes such as endothelial dysfunction and atherosclerosis [41, 42], which are shared injury mechanisms of the brain and kidney, and might be further strengthened through other potential effects by CKD such as uremic milieu and blood-brain barrier disruption [40, 43, 44]. Based on all the above proposed mechanisms, there was accumulating evidence illustrating a systemic feature of microvascular dysfunction, as well as the bidirectional relationship between kidney impairment and microvascular dysfunction [40, 45, 46]. Thus, the vital role for kidney functions to be a predictor as well as a risk factor for systemic microvascular dysfunction warrants further investigation.

Back to the present meta-analysis, there were also some limitations as well as implications. First, the vast majority of included studies were cross-sectional designed, making it less convincing to establish a specific causal association between kidney dysfunction and CSVD. Moreover, different methods in statistical processing of continuous variables together with insufficient studies investigating PVS and CysC also prevented us from reaching more precise conclusions on the associations. Second, due to a wide range of differences in various characteristics between studies, moderate to high heterogeneity has been observed in most analyses. And in most subgroup analyses, substantial heterogeneity still existed. Third, with a handful of studies estimating total CSVD burden by a proposed scoring system [47], a proportion of included studies used inconsistent evaluations of CSVD markers, which might exert a non-negligible influence on pooled results. Therefore, the findings should be interpreted cautiously. Further improvement and wide application of a standardized evaluation method are expected. Last but not least, eGFR, proteinuria, and serum CysC mainly reflect glomerular function, while tubular function was not studied in our meta-analysis. In summary, we suggest that future studies with large sample sizes should adhere to universally accepted evaluation criteria for CSVD, investigate different indicators of kidney function simultaneously, to provide convincing longitudinal evidence on the association between CSVD and kidney function.

This updated systematic review and meta-analysis comprehensively summarized the latest evidence regarding the association between kidney function and CSVD burden and highlighted the important role of different indicators of kidney function in identifying the high-risk population for CSVD. Based on the anatomical and functional similarity in microvascular beds and shared risk factors between the brain and kidney, kidney dysfunction may predict the presence, and promote the progression of CSVD through multiple plausible pathophysiological mechanisms that are independent of traditional cardiovascular risk factors and need to be further established. Our findings showed new directions for identifying and preventing SVD, facilitating decision-making in primary or secondary prevention of CSVD. High-quality longitudinal studies with large samples, rigorous design, and comprehensive evaluation are urgently required to provide more persuasive evidence.

An ethics statement is not applicable because this study is based on the published literature.

The authors declared no potential conflicts of interest with respect to this article.

This study was supported by grants from the National Natural Science Foundation of China (Grant No. 82071201, 81971032, 81901121).

Lan Tan contributed to the concept and design of this study. Chu-Yun Xiao, Ya-Hui Ma, and Ya-Nan Ou conducted the literature search and data extraction. Chu-Yun Xiao, Bing Zhao, He-Ying Hu, Zuo-Teng Wang, and Lan Tan analyzed and interpreted the data. Chu-Yun Xiao and Ya-Hui Ma wrote the draft manuscript. Lan Tan and Ya-Hui Ma critically revised the manuscript for important intellectual content. All authors reviewed and gave final approval of the submitted manuscript.

Additional information is provided in the supplementary material. For more details, the relevant data are available from the corresponding author with reasonable request.

1.
Wardlaw
JM
,
Smith
EE
,
Biessels
GJ
,
Cordonnier
C
,
Fazekas
F
,
Frayne
R
,
.
Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration
.
Lancet Neurol
.
2013
;
12
(
8
):
822
38
.
2.
Wardlaw
JM
,
Smith
C
,
Dichgans
M
.
Small vessel disease: mechanisms and clinical implications
.
Lancet Neurol
.
2019
;
18
(
7
):
684
96
.
3.
Rensma
SP
,
van Sloten
TT
,
Launer
LJ
,
Stehouwer
CDA
.
Cerebral small vessel disease and risk of incident stroke, dementia and depression, and all-cause mortality: a systematic review and meta-analysis
.
Neurosci Biobehavioral Rev
.
2018
;
90
:
164
73
.
4.
Hamilton
OKL
,
Backhouse
EV
,
Janssen
E
,
Jochems
ACC
,
Maher
C
,
Ritakari
TE
,
.
Cognitive impairment in sporadic cerebral small vessel disease: a systematic review and meta-analysis
.
Alzheimers Dement
.
2021
;
17
(
4
):
665
85
.
5.
Pantoni
L
.
Cerebral small vessel disease: from pathogenesis and clinical characteristics to therapeutic challenges
.
Lancet Neurol
.
2010
;
9
(
7
):
689
701
.
6.
Toyoda
K
.
Cerebral small vessel disease and chronic kidney disease
.
J Stroke
.
2015
;
17
(
1
):
31
7
.
7.
Kang
DH
,
Kanellis
J
,
Hugo
C
,
Truong
L
,
Anderson
S
,
Kerjaschki
D
,
.
Role of the microvascular endothelium in progressive renal disease
.
J Am Soc Nephrol
.
2002
;
13
(
3
):
806
16
.
8.
Jourde-Chiche
N
,
Fakhouri
F
,
Dou
L
,
Bellien
J
,
Burtey
S
,
Frimat
M
,
.
Endothelium structure and function in kidney health and disease
.
Nat Rev Nephrol
.
2019
;
15
(
2
):
87
108
.
9.
Toyoda
K
,
Ninomiya
T
.
Stroke and cerebrovascular diseases in patients with chronic kidney disease
.
Lancet Neurol
.
2014
;
13
(
8
):
823
33
.
10.
Deckers
K
,
Camerino
I
,
van Boxtel
MP
,
Verhey
FR
,
Irving
K
,
Brayne
C
,
.
Dementia risk in renal dysfunction: a systematic review and meta-analysis of prospective studies
.
Neurology
.
2017
;
88
(
2
):
198
208
.
11.
Drew
DA
,
Weiner
DE
,
Sarnak
MJ
.
Cognitive impairment in CKD: pathophysiology, management, and prevention
.
Am J Kidney Dis
.
2019
;
74
(
6
):
782
90
.
12.
Georgakis
MK
,
Dimitriou
NG
,
Karalexi
MA
,
Mihas
C
,
Nasothimiou
EG
,
Tousoulis
D
,
.
Albuminuria in association with cognitive function and dementia: a systematic review and meta-analysis
.
J Am Geriatr Soc
.
2017
;
65
(
6
):
1190
8
.
13.
Bos
D
,
Wolters
FJ
,
Darweesh
SKL
,
Vernooij
MW
,
Wolf
F
,
Ikram
MA
.
Cerebral small vessel disease and the risk of dementia: a systematic review and meta-analysis of population-based evidence
.
Alzheimers Dement
.
2018
;
14
(
11
):
1482
92
.
14.
Liu
Y
,
Lv
P
,
Jin
H
,
Cui
W
,
Niu
C
,
Zhao
M
,
.
Association between low estimated glomerular filtration rate and risk of cerebral small-vessel diseases: a meta-analysis
.
J Stroke Cerebrovasc Dis
.
2016
;
25
(
3
):
710
6
.
15.
Georgakis
MK
,
Chatzopoulou
D
,
Tsivgoulis
G
,
Petridou
ET
.
Albuminuria and cerebral small vessel disease: a systematic review and meta-analysis
.
J Am Geriatr Soc
.
2018
;
66
(
3
):
509
17
.
16.
Makin
SDJ
,
Cook
FAB
,
Dennis
MS
,
Wardlaw
JM
.
Cerebral small vessel disease and renal function: systematic review and meta-analysis
.
Cerebrovasc Dis
.
2015
;
39
(
1
):
39
52
.
17.
Randers
E
,
Erlandsen
EJ
.
Serum cystatin C as an endogenous marker of the renal function: a review
.
cclm
.
1999
;
37
(
4
):
389
95
.
18.
Page
MJ
,
McKenzie
JE
,
Bossuyt
PM
,
Boutron
I
,
Hoffmann
TC
,
Mulrow
CD
,
.
The PRISMA 2020 statement: an updated guideline for reporting systematic reviews
.
BMJ
.
2021
;
372
:
n71
.
19.
Zhang
L
,
Hu
P
,
Chen
X
,
Bie
P
.
Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomized trials
.
2014
.
20.
Zhou
Y
,
Ying
G
,
Yan
H
,
Xing
W
,
Nursing
SO
.
The joanna briggs institute critical appraisal tools for use in systematic review: prevalence study and analytical cross sectional study
.
J Nurses Train
.
2018
.
21.
Higgins
JPT
,
Thompson
SG
,
Deeks
JJ
,
Altman
DG
.
Measuring inconsistency in meta-analyses
.
BMJ
.
2003
;
327
(
7414
):
557
60
.
22.
Cumpston
M
,
Li
T
,
Page
MJ
,
Chandler
J
,
Welch
VA
,
Higgins
JP
,
.
Updated guidance for trusted systematic reviews: a new edition of the Cochrane Handbook for Systematic Reviews of Interventions
.
Cochrane Database Syst Rev
.
2019
;
10
:
Ed000142
.
23.
Hägg
S
,
Thorn
LM
,
Putaala
J
,
Liebkind
R
,
Harjutsalo
V
,
Forsblom
CM
,
.
Incidence of stroke according to presence of diabetic nephropathy and severe diabetic retinopathy in patients with type 1 diabetes
.
Diabetes Care
.
2013
;
36
(
12
):
4140
6
.
24.
van Overbeek
EC
,
Staals
J
,
van Oostenbrugge
RJ
.
Decreased kidney function relates to progression of cerebral microbleeds in lacunar stroke patients
.
Int J Stroke
.
2016
;
11
(
6
):
695
700
.
25.
Guoxiang
H
,
Hui
L
,
Yong
Z
,
Xunming
J
,
Zhuo
C
.
Association between cystatin C and SVD in Chinese population
.
Neurol Sci
.
2018
;
39
(
12
):
2197
202
.
26.
Yang
S
,
Cai
J
,
Lu
R
,
Wu
J
,
Zhang
M
,
Zhou
X
.
Association between serum cystatin C level and total magnetic resonance imaging burden of cerebral small vessel disease in patients with acute lacunar stroke
.
J Stroke Cerebrovasc Dis
.
2017
;
26
(
1
):
186
91
.
27.
Xu
XH
,
Ye
XH
,
Cai
JS
,
Gao
T
,
Zhao
GH
,
Zhang
WJ
,
.
Association of renal dysfunction with remote diffusion-weighted imaging lesions and total burden of cerebral small vessel disease in patients with primary intracerebral hemorrhage
.
Front Aging Neurosci
.
2018
;
10
:
171
.
28.
Toyoda
G
,
Bokura
H
,
Mitaki
S
,
Onoda
K
,
Oguro
H
,
Nagai
A
,
.
Association of mild kidney dysfunction with silent brain lesions in neurologically normal subjects
.
Cerebrovasc Dis Extra
.
2015
;
5
(
1
):
22
7
.
29.
Yao
H
,
Takashima
Y
,
Hashimoto
M
,
Uchino
A
,
Yuzuriha
T
.
Subclinical cerebral abnormalities in chronic kidney disease
.
Contrib Nephrol
.
2013
;
179
:
24
34
.
30.
Vemuri
P
,
Knopman
DS
,
Jack
CR
Jr
,
Lundt
ES
,
Weigand
SD
,
Zuk
SM
,
.
Association of kidney function biomarkers with brain MRI findings: the BRINK study
.
J Alzheimers Dis
.
2016
;
55
(
3
):
1069
82
.
31.
Tsai
YH
,
Lee
M
,
Lin
LC
,
Chang
SW
,
Weng
HH
,
Yang
JT
,
.
Association of chronic kidney disease with small vessel disease in patients with hypertensive intracerebral hemorrhage
.
Front Neurol
.
2018
;
9
:
284
.
32.
Ikram
MA
,
Vernooij
MW
,
Hofman
A
,
Niessen
WJ
,
van der Lugt
A
,
Breteler
MM
.
Kidney function is related to cerebral small vessel disease
.
Stroke
.
2008
;
39
(
1
):
55
61
.
33.
Henskens
LH
,
van Oostenbrugge
RJ
,
Kroon
AA
,
Hofman
PA
,
Lodder
J
,
de Leeuw
PW
.
Detection of silent cerebrovascular disease refines risk stratification of hypertensive patients
.
J Hypertens
.
2009
;
27
(
4
):
846
53
.
34.
Brown
R
,
Benveniste
H
,
Black
SE
,
Charpak
S
,
Dichgans
M
,
Joutel
A
,
.
Understanding the role of the perivascular space in cerebral small vessel disease
.
Cardiovasc Res
.
2018
;
114
(
11
):
1462
73
.
35.
Yusuf
S
,
Reddy
S
,
Ounpuu
S
,
Anand
S
.
Global burden of cardiovascular diseases: Part II: variations in cardiovascular disease by specific ethnic groups and geographic regions and prevention strategies
.
Circulation
.
2001
;
104
(
23
):
2855
64
.
36.
Nakamura
K
,
Barzi
F
,
Lam
TH
,
Huxley
R
,
Feigin
VL
,
Ueshima
H
,
.
Cigarette smoking, systolic blood pressure, and cardiovascular diseases in the Asia-Pacific region
.
Stroke
.
2008
;
39
(
6
):
1694
702
.
37.
Lee
M
,
Saver
JL
,
Chang
KH
,
Liao
HW
,
Chang
SC
,
Ovbiagele
B
.
Low glomerular filtration rate and risk of stroke: meta-analysis
.
BMJ
.
2010
;
341
:
c4249
.
38.
Theuerle
JD
,
Al-Fiadh
AH
,
Wong
E
,
Patel
SK
,
Ashraf
G
,
Nguyen
T
,
.
Retinal microvascular function predicts chronic kidney disease in patients with cardiovascular risk factors
.
Atherosclerosis
.
2022
;
341
:
63
70
.
39.
Lau
WL
,
Nunes
ACF
,
Vasilevko
V
,
Floriolli
D
,
Lertpanit
L
,
Savoj
J
,
.
Chronic kidney disease increases cerebral microbleeds in mouse and man
.
Transl Stroke Res
.
2020
;
11
(
1
):
122
34
.
40.
Querfeld
U
,
Mak
RH
,
Pries
AR
.
Microvascular disease in chronic kidney disease: the base of the iceberg in cardiovascular comorbidity
.
Clin Sci
.
2020
;
134
(
12
):
1333
56
.
41.
Bonetti
PO
,
Lerman
LO
,
Lerman
A
.
Endothelial dysfunction: a marker of atherosclerotic risk
.
Arterioscler Thromb Vasc Biol
.
2003
;
23
(
2
):
168
75
.
42.
O'Rourke
MF
,
Safar
ME
.
Relationship between aortic stiffening and microvascular disease in brain and kidney: cause and logic of therapy
.
Hypertension
.
2005
;
46
(
1
):
200
4
.
43.
Kelly
D
,
Rothwell
PM
.
Disentangling the multiple links between renal dysfunction and cerebrovascular disease
.
J Neurol Neurosurg Psychiatry
.
2020
;
91
(
1
):
88
97
.
44.
Lau
WL
,
Huisa
BN
,
Fisher
M
.
The cerebrovascular-chronic kidney disease connection: perspectives and mechanisms
.
Transl Stroke Res
.
2017
;
8
(
1
):
67
76
.
45.
Nowroozpoor
A
,
Gutterman
D
,
Safdar
B
.
Is microvascular dysfunction a systemic disorder with common biomarkers found in the heart, brain, and kidneys? - a scoping review
.
Microvasc Res
.
2021
;
134
:
104123
.
46.
Krishnan
S
,
Suarez-Martinez
AD
,
Bagher
P
,
Gonzalez
A
,
Liu
R
,
Murfee
WL
,
.
Microvascular dysfunction and kidney disease: challenges and opportunities
.
Microcirculation
.
2021
;
28
(
3
):
e12661
.
47.
Klarenbeek
P
,
van Oostenbrugge
RJ
,
Rouhl
RP
,
Knottnerus
IL
,
Staals
J
.
Ambulatory blood pressure in patients with lacunar stroke: association with total MRI burden of cerebral small vessel disease
.
Stroke
.
2013
;
44
(
11
):
2995
9
.