Abstract
Objective
This study aimed to investigate the independent association between plasma Lipoprotein(a) [Lp(a)] levels and the occurrence, anatomical distribution, and severity of cerebral microbleeds (CMBs) in patients with Chronic kidney disease (CKD).
Methods
A single-center retrospective observational cohort enrolled 137 CKD patients admitted from January 2022 to January 2026. All subjects completed 3.0T MRI with susceptibility-weighted imaging (SWI) for CMBs identification and regional stratification (cortical, deep, subcortical). Clinical demographics, and serum lipid levels including Lp(a) were collected. Univariate comparisons, binary logistic regression, ROC curve analysis, Spearman and partial correlation analyses were performed to assess associations between Lp(a) and CMBs.
Results
The overall occurrence of CMBs in the cohort was 58.39%. Patients with CMBs had significantly higher plasma Lp(a) levels than those without CMBs (40.23±113.56 mg/dL vs 8.83±11.88 mg/dL, P=0.040). After adjustment, binary logistic regression analysis indicated that elevated Lp(a) was independently associated with global CMBs (Odds Ratio [OR]=1.052, 95% Confidence Interval [CI] = 1.020–1.086, P = 0.001) and deep CMBs (Odds Ratio [OR] = 1.034, 95% Confidence Interval [CI] =1.015–1.054, P < 0.001), but not with cortical or subcortical CMBs. The area under the ROC curve (AUC) of Lp(a) was 0.728, with an optimal cutoff value of 14.055 mg/dL (sensitivity = 0.590, specificity = 0.829). Lp(a) was positively related to deep CMBs severity across all CKD patients (partial correlation coefficient = 0.253, P=0.003), while this association was not significant among patients with CMBs.
Conclusion
Elevated plasma Lp(a) is an independent risk factor for CMBs in patients with CKD, with a specific anatomical predilection for deep cerebral lesions. It may serve as a potential serologic marker for cerebrovascular risk stratification among patients with CKD.
Keywords: lipoprotein,a; chronic kidney disease; cerebral microbleeds; risk factor analysis
Introduction
Chronic kidney disease (CKD) is a chronic progressive disorder characterized by persistent structural or functional abnormalities of the kidney lasting for more than 3 months due to multiple etiologies.1 Its global prevalence has been rising steadily,2 making it a major public health concern. In addition to the risk of renal failure, accumulating clinical evidence indicates that patients with CKD face higher risks of cerebrovascular events.3 Pathological alterations in CKD, including metabolic disturbances, vascular endothelial injury, chronic inflammation and oxidative stress, can induce cerebral small vessel disease (CSVD), further elevating the risk of ischemic stroke, hemorrhagic stroke and cognitive impairment.4
Cerebral microbleeds (CMBs) are important imaging markers of CSVD, which can objectively reflect potential microvascular lesions in the brain. They play a vital role in the diagnosis, severity assessment and prognostic evaluation of cerebral amyloid angiopathy and stroke.4,5 Previous studies have confirmed that CKD patients are a high-risk population for CMBs.6 The presence of CMBs further increases the risks of recurrent stroke and cognitive impairment in CKD patients.5,7,8 Therefore, exploring the risk factors for CMBs in CKD patients is of great significance for disease diagnosis, treatment and prognostic improvement.
At present, the risk factors for CMBs in CKD patients remain unclear. Lipoprotein(a) [Lp(a)] is a unique plasma lipoprotein whose concentration is mainly regulated by genetic factors and largely genetically determined, with minimal influence from lifestyle factors.9,10 The uremic microenvironment in CKD patients can lead to elevated plasma Lp(a) levels.11 Increased Lp(a) concentrations may impair vascular wall stability via lipid deposition, inflammatory activation and endothelial dysfunction, resulting in vascular fragility and an elevated risk of microvascular rupture and bleeding, which might contribute to CMB development.12–15
Nevertheless, studies focusing on the relationship between Lp(a) and CMBs in CKD patients are still limited. This study will explore the association between LP(a) and CMBs, as well as its distribution across different brain regions, to provide evidence for early screening and targeted intervention in CKD patients with CMBs.
Materials and Methods
Study Design and Participants
This was a single-center retrospective observational study. Patients diagnosed with CKD and admitted to the Neurology Department of Hangzhou Traditional Chinese Medicine (TCM) Hospital from January 2022 to January 2026 were consecutively enrolled. This study was approved by the Ethics Committee of Hangzhou Hospital of Traditional Chinese Medicine (Approval No.:2025KLL197) and conducted in accordance with the 2013 revision of the Declaration of Helsinki. Written informed consent was obtained from all participants or their legal guardians.
Inclusion Criteria
(1) Diagnosis of Chronic Kidney Disease(CKD) consistent with the 2024 Clinical Practice Guideline: Assessment and Management of Chronic Kidney Disease issued by Kidney Disease: Improving Global Outcomes (KDIGO);1 (2) Aged≥40 years; (3) Complete all examinations and assessments; (4) Voluntary participation in this study with signed informed consent form.
Exclusion Criteria
(1) Severe organic diseases of the heart, liver, or lungs (eg, New York Heart Association (NYHA) Class III–IV heart failure, active hepatitis, decompensated cirrhosis, severe respiratory failure); (2) Malignant tumors; (3) Cerebrovascular events within the 3 months (cerebral infarction or cerebral hemorrhage); (4) CMBs caused by other known factors (eg, cerebral amyloid angiopathy, hereditary cerebral small-vessel disease, traumatic brain injury, cerebral infection, history of cranial radiation therapy); (5) Magnetic resonance imaging (MRI) contraindications; (6) Incomplete clinical, laboratory or imaging datasets.
Clinical Data Collection
General demographic data and medical histories were collected, including age, gender, years of education, hypertension history, diabetes mellitus history, coronary heart disease history, smoking history and alcohol consumption history. Smoking history was defined as smoking at least one cigarette per day for more than 1 year. Alcohol consumption history was defined as drinking alcohol no less than three times per week for more than 6 months. The detailed screening process of participants based on inclusion and exclusion criteria is shown in Figure 1.
Figure 1.
Flow chart of participant screening in this study.
Laboratory Data Collection
Fasting venous blood samples were collected from all patients after an 8-hour overnight fast upon admission. All samples were analyzed via immunoturbidimetric and enzymatic assays on an automatic biochemistry analyzer in the clinical laboratory of Hangzhou TCM Hospital. Detected indicators included plasma Lp(a), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), apolipoproteinA-I (ApoA-I), apolipoprotein B (ApoB), apolipoprotein E (ApoE) and estimated glomerular filtration rate (eGFR). The Cockcroft-Gault formula was used for the calculation of eGFR:16 For male, eGFR=[(140-age) * weight * 1.23]/[Serum Creatinine (μmol/L)]; for female,eGFR=[(140-age)*weight*1.04]/[Serum Creatinine (μmol/L)]. Of note, CKD diagnosis and staging were strictly performed according to the KDIGO 2024 guideline; the Cockcroft-Gault formula was only applied for secondary creatinine-clearance estimation and was not used for CKD diagnosis or staging.
Imaging Assessment of Cerebral Microbleeds in CKD Patients
All patients underwent 3.0T MRI. Magnetic susceptibility-weighted imaging (SWI), T1-weighted imaging (T1WI) and T2-weighted imaging (T2WI) sequences were acquired. The diagnostic criteria for CMBs referred to the 2020 Chinese Guidelines for the Diagnosis and Treatment of Cerebral Small Vessel Disease and the 2023 Standards for Reporting Vascular Lesions on Neuroimaging (STRIVE2).17,18 CMBs were defined as round or oval hypointense lesions on SWI sequences, with a diameter of 2–5 mm (maximum diameter≤10 mm) and no hyperintense signals on T1WI or T2WI. Lesions were differentiated from blood vessels, iron or calcium deposition, bone artifacts and diffuse axonal injury caused by craniocerebral trauma. Two experienced neuroimaging physicians (more than 10 years of clinical experience) independently evaluated all images in a double-blind manner and recorded the number of CMBs in cortical, deep and subcortical regions. Discrepancies were resolved through mutual discussion. CMB severity was graded according to lesion count: Grade 0 (no CMBs), Grade 1 (1–2 CMBs), Grade 2 (3–9CMBs), Grade 3 (≥10 CMBs).18 All patients were divided into the CMB group and non-CMB group according to the presence of CMBs. The CMB group was further divided into cortical, deep and subcortical subgroups based on lesion locations.
Statistical Methods
SPSS 25.0 software was used for statistical analysis. Categorical variables were described as counts and percentages (%), with between-group comparisons conducted via the chi-square test. Continuous variables were tested for normality: normally distributed data were presented as mean ± standard deviation (SD) and compared using the t-test, while non-normally distributed data were reported as median (interquartile range) and compared with the Mann–Whitney U-test.
Primary Analysis
We focused on evaluating the association between plasma Lp(a) and deep CMBs. Binary logistic regression was used to identify independent risk factors for the presence of global CMBs and deep CMBs. For regression modeling, age, sex and variables with univariate P < 0.1 were considered for multivariate models.
Exploratory Analyses
Pearson, Spearman correlation and partial correlation analyses were performed to explore relationships between clinical-laboratory indicators and CMB severity.
Results
Baseline Characteristics of CKD Patients
A total of 137 CKD patients were enrolled, including 66 males (48.2%) and 71 females (51.8%), with a mean age of 72.84±9.80 years and a mean years of education of 8.21±4.24 years. The distribution of CKD stages was as follows: Stage 1 (9 cases, 6.6%), Stage 2 (57 cases, 41.6%), Stage 3 (56 cases, 40.9%), Stage 4 (11 cases, 8.0%), Stage 5 (4 cases, 2.9%). Hypertension was the most common comorbidity (73.0%), followed by diabetes mellitus (27.0%) and coronary heart disease (13.9%). The mean plasma Lp(a) level of all patients was 27.16±88.26 mg/dL. Imaging results showed the mean number of cortical CMBs, deep CMBs and subcortical CMBs was 1.14±4.30, 1.70±3.08 and 0.81±3.82, respectively, and the total mean CMBs count was 3.65±9.67. Among all subjects, 121 patients (88%) had plasma Lp(a) levels below 50 mg/dL, and 16 patients (12%) had Lp(a) levels higher than 50 mg/dL, including one patient with an extremely high Lp(a) level of 1000.99 mg/dL. The mean eGFR among participants was 58.17 ± 19.92 μmol/L. Detailed baseline data are shown in Table 1.
Table 1.
Baseline Characteristics of Enrolled Patients
| Variable | [n(%)/Mean/Median (Interquartile Range)] |
|---|---|
| Demographic Characteristics | |
| Age | 72.84±9.80 |
| Male | 66 (48.2%) |
| Years of education | 8.21±4.24 |
| Traditional vascular risk factors | |
| History of hypertension | 100 (73.0%) |
| History of diabetes | 37 (27.0%) |
| History of coronary heart disease | 19 (13.9%) |
| History of smoking | 19 (13.9%) |
| History of alcohol consumption CKD stage |
16 (11.7%) |
| Stage 1 | 9 (6.6%) |
| Stage 2 | 57 (41.6%) |
| Stage 3 | 56 (40.9%) |
| Stage 4 | 11 (8.0%) |
| Stage 5 | 4 (2.9%) |
| Laboratory parameters | |
| eGFR (mg/dL) | 58.17 ± 19.92 |
| LP(a) (mg/dL) | 27.16 ± 88.26 |
| HDL-C (mmol/L) | 1.06 ± 0.23 |
| LDL-C (mmol/L) | 2.29 ± 0.83 |
| ApoA-I (g/L) | 1.30 ± 0.22 |
| ApoB (g/L) | 0.67 ± 0.22 |
| ApoE (mg/dL) | 4.02 ± 1.41 |
| Distribution of CMBs | |
| Number of cortical CMBs | 1.14 ± 4.30 |
| Number of deep CMBs | 1.70 ± 3.08 |
| Number of subcortical CMBs | 0.81 ± 3.82 |
| Total number of CMBs | 3.65 ± 9.67 |
Comparison of Clinical Data Between the CMBs and Non-CMBs Groups
Univariate Analysis
Univariate analysis revealed that plasma Lp(a) levels were significantly higher in the CMBs group than in the non-CMBs group (40.23±113.56 mg/dL vs 8.83±11.88 mg/dL, P=0.040). The non-CMBs group had higher eGFR levels than the CMBs group (61.50±18.24 mg/dL vs 55.80±20.83 mg/dL, P=0.099). The prevalence of hypertension was higher in the CMBs group (80.0% vs 63.2%, P=0.033). There were no statistically significant differences between the groups in terms of age, gender, education duration, diabetes mellitus history, coronary heart disease history, smoking history, alcohol consumption history, HDL-C, LDL-C, ApoA-I, ApoB and ApoE (all P>0.05). Detailed results are presented in Table 2.
Table 2.
Comparison of Clinical Characteristics Between the CMBs and Non-CMBs Groups
| CMBs Group (Number of CMBs > 0, n = 80) |
Non-CMBs Group (Number of CMBs = 0, n = 57) |
P | |
|---|---|---|---|
| Demographic characteristics | |||
| Age | 73.58 ± 9.73 | 71.81 ± 9.90 | 0.300 |
| Male | 38 (47.5%) | 28 (49.1%) | 0.864 |
| Years of education | 8.14 ± 4.28 | 8.32 ± 4.21 | 0.809 |
| Traditional vascular risk factors | |||
| History of hypertension | 64 (80.0%) | 36 (63.2%) | 0.033 |
| History of diabetes | 20 (25.0%) | 17 (29.8%) | 0.562 |
| History of coronary heart disease | 10 (12.5%) | 9 (15.8%) | 0.622 |
| History of smoking | 12 (15.0%) | 7 (12.3%) | 0.803 |
| History of alcohol consumption | 10 (12.5%) | 6 (10.5%) | 0.793 |
| Laboratory parameters | |||
| eGFR | 55.80 ± 20.83 | 61.50 ± 18.24 | 0.099 |
| LP(a) | 40.23 ± 113.56 | 8.83 ± 11.88 | 0.040 |
| HDL-C | 1.07 ± 0.22 | 1.05 ± 0.24 | 0.601 |
| LDL-C | 2.34 ± 0.81 | 2.21 ± 0.85 | 0.341 |
| ApoA-I | 1.30 ± 0.23 | 1.30 ± 0.21 | 0.956 |
| ApoB | 0.69 ± 0.21 | 0.64 ± 0.23 | 0.167 |
| ApoE | 4.18 ± 1.44 | 3.81 ± 1.35 | 0.134 |
Binary Logistic Regression Analysis for Independent Risk Factors of CMBs
After adjusting for confounding factors including gender, age, hypertension and eGFR, binary logistic regression analysis confirmed that Lp(a) was an independent risk factor for CMBs in CKD patients (OR=1.052, 95% CI: 1.020–1.086, P=0.001). The corresponding forest plot is shown in Figure 2.
Figure 2.
Forest plot of the binary logistic regression analyzing the association between LP(a) and the risk of CMBs in CKD patients.
Stratified Analysis of CMBs by Lesion Location
Cortical CMBs Group vs Non-Cortical CMBs Group
No significant differences in all clinical and laboratory indicators were found between the two groups (P>0.05) (Table 3).
Table 3.
Comparison of Indicators Among CMB Subgroups with Different Lesion Locations
| Model 1 | Model 2 | Model 3 | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Cortical CMBs Group (Number of CMBs in cerebral lobes > 0, n = 41) | Non-cortical CMBs group (Number of CMBs in the cerebral lobe region = 0, n=96) | P | Deep CMBs group (Number of CMBs in deep regions > 0, n = 61) | Non-deep CMBs Group (Number of CMBs in deep regions = 0, n=76) | P | Subcortical CMBs Group (Number of CMBs in the subcortical region > 0, n = 32) | Non-subcortical CMBs Group (Number of CMBs in the Infratemporal Region = 0, n=105) | P | |
| Demographic Characteristic | |||||||||
| Age | 72.29±9.32 | 73.07±10.04 | 0.671 | 73.72±9.52 | 72.13±10.79 | 0.347 | 72.59±10.12 | 72.91±9.75 | 0.872 |
| Male | 23 (56.1%) | 43 (44.8%) | 0.265 | 28 (45.9%) | 38 (50.0%) | 0.731 | 18 (56.3%) | 48 (45.7%) | 0.319 |
| Years of education | 8.32±4.70 | 8.17 ± 4.05 | 0.850 | 8.00±4.29 | 8.38±4.22 | 0.602 | 7.50±4.08 | 8.43±4.28 | 0.279 |
| Traditional vascular risk factors | |||||||||
| History of hypertension | 33 (80.5%) | 67 (69.8%) | 0.216 | 48 (78.7%) | 52 (68.4%) | 0.245 | 24 (75.0%) | 76 (72.4%) | 0.824 |
| History of diabetes | 10 (24.4%) | 27 (28.1%) | 0.834 | 14 (23.0%) | 23 (30.3%) | 0.439 | 10 (31.3%) | 23 (25.7%) | 0.650 |
| History of coronary heart disease | 4 (9.8%) | 15 (15.6%) | 0.430 | 7 (11.5%) | 12 (15.8%) | 0.620 | 5 (15.6%) | 14 (13.3%) | 0.772 |
| History of smoking | 8 (19.5%) | 11 (11.5%) | 0.280 | 5 (8.2%) | 14 (18.4%) | 0.134 | 8 (25.0%) | 11 (10.5%) | 0.075 |
| History of alcohol consumption | 7 (17.1%) | 9 (9.4%) | 0.246 | 4 (6.6%) | 12 (15.8%) | 0.114 | 5 (15.6%) | 11 (10.5%) | 0.529 |
| Laboratory parameters | |||||||||
| eGFR | 56.1±23.38 | 59.0±18.32 | 0.435 | 53.94±20.25 | 61.57±19.12 | 0.025 | 59.94±23.96 | 57.63±18.62 | 0.568 |
| LP(a) | 26.93±32.47 | 27.27±103.48 | 0.984 | 46.75±128.54 | 11.44±18.59 | 0.019 | 50.35±175.90 | 20.10±27.37 | 0.090 |
| HDL-C | 1.09±0.21 | 1.05±0.24 | 0.432 | 1.08±0.23 | 1.06±0.23 | 0.609 | 1.05±0.21 | 1.07±0.24 | 0.594 |
| LDL-C | 2.27±0.64 | 2.29±0.90 | 0.876 | 2.32±0.87 | 2.25±0.80 | 0.597 | 2.30±0.61 | 2.28±0.88 | 0.908 |
| ApoA-I | 1.32±0.19 | 1.29±0.23 | 0.414 | 1.29±0.23 | 1.30±0.22 | 0.781 | 1.24±0.23 | 1.32±0.22 | 0.091 |
| ApoB | 0.68 ±0.18 | 0.67±0.23 | 0.842 | 0.69±0.22 | 0.66±0.21 | 0.415 | 0.69±0.18 | 0.66±0.23 | 0.558 |
| ApoE | 3.97±1.31 | 4.05±1.46 | 0.768 | 4.27±1.52 | 3.83±1.30 | 0.071 | 4.31±1.63 | 3.94±1.33 | 0.192 |
Deep CMBs Group vs Non-Deep CMBs Group
Univariate analysis showed that plasma Lp(a) levels were significantly higher in the deep CMBs group than in the non-deep CMBs group (46.75±128.54 mg/dL vs 11.44±18.59 mg/dL, P=0.019). The non-deep CMBs group had higher eGFR levels (61.57±19.12 mg/dL vs 53.94±20.25 mg/dL, P=0.025). The ApoE level was slightly higher in the deep CMBs group (4.27±1.52 mg/dL vs 3.83±1.30 mg/dL, P=0.071). Detailed data are shown in Table 3.
Subcortical CMBs Group vs Non-Subcortical CMBs Group
No statistically significant differences in demographic characteristics, vascular risk factors and laboratory indicators were detected between the two groups (all P>0.05) (Table 3).
Binary Logistic Regression Analysis for Deep CMBs
After adjustment for ApoE, eGFR, age and gender, binary logistic regression analysis indicated that Lp(a) was an independent risk factor for deep CMBs (OR = 1.034, 95% CI: 1.015–1.054, P < 0.001). The corresponding forest plot is shown in Figure 3.
Figure 3.
Forest plot of the binary logistic regression analyzing the association between LP(a) and the risk of deep CMBs in patients with CKD.
ROC Curve Analysis for Deep CMBs
ROC curves were plotted to evaluate the predictive performance of Lp(a) for deep CMBs. The AUC was 0.728, with a cut-off value of 14.055 mg/dL, a sensitivity of 0.590 and a specificity of 0.829 (Figure 4, Table 4).
Figure 4.
ROC curve of Lp(a) for predicting deep CMBs in CKD patients.
Table 4.
ROC Curve Analysis of Lp(a) for Predicting Deep CMBs in CKD Patients
| AUC | 95% CI | P | Cut-off Value | Sensitivity | Specificity | |
|---|---|---|---|---|---|---|
| LP(a) | 0.728 | 0.641~0.815 | 0.000 | 14.055 | 0.590 | 0.829 |
Correlation Between Lp(a) and Severity of Deep CMBs
Spearman’s rank correlation analysis revealed a significant positive correlation between plasma LP(a) levels and the severity of deep CMBs in the overall CKD patient population (r = 0.396, P = 0.000). After adjusting for age, gender, history of smoking and eGFR in a partial correlation analysis, this positive correlation remained significant (r=0.251, P=0.004).
Spearman correlation indicated a positive correlation between Lp(a) levels and deep CMBs severity (r= 0.270, P=0.013). After adjusting for age, gender, history of coronary heart disease in a partial correlation analysis, the results indicated no significant correlation between Lp(a) and the severity of deep CMBs severity (r=0.185, P=0.100). These findings indicated a threshold effect of Lp(a). Detailed data are shown in Table 5.
Table 5.
Correlation Analysis Between Indicators and Deep CMBs Severity
| CKD Population CKD with CMBs Population | ||||
|---|---|---|---|---|
| Spearman Correlation Analysis | Correlation Coefficient | P | Correlation Coefficient | P |
| Demographic Characteristics | ||||
| Age | 0.077 | 0.372 | 0.050 | 0.655 |
| Male | 0.038 | 0.660 | 0.026 | 0.815 |
| Years of education | −0.069 | 0.426 | −0.132 | 0.235 |
| Traditional vascular risk factors | ||||
| History of hypertension | 0.102 | 0.235 | −0.089 | 0.421 |
| History of diabetes | −0.098 | 0.254 | 0.015 | 0.890 |
| History of coronary heart disease | −0.106 | 0.216 | −0.231 | 0.036 |
| History of smoking | −0.152 | 0.076 | −0.085 | 0.447 |
| History of alcohol consumption | −0.139 | 0.105 | −0.086 | 0.438 |
| Laboratory parameters | ||||
| eGFR | −0.199 | 0.020 | −0.185 | 0.153 |
| LP(a) | 0.396 | 0.000 | 0.270 | 0.013 |
| HDL-C | 0.069 | 0.425 | 0.023 | 0.838 |
| LDL-C | 0.165 | 0.853 | −0.147 | 0.185 |
| ApoA-I | 0.019 | 0.824 | −0.051 | 0.649 |
| ApoB | 0.055 | 0.522 | −0.095 | 0.393 |
| ApoE | 0.125 | 0.146 | −0.020 | 0.854 |
Discussion
In this retrospective cross-sectional study, we found that elevated plasma Lp(a) was independently associated with the presence of CMBs in CKD patients, and this statistically significant independent association was specifically observed for deep CMBs. Besides, Lp(a) was also correlated with deep CMBs burden in the full cohort; this association was no longer statistically significant following multivariable adjustment among patients with confirmed CMBs.
Several mechanisms may explain the association between Lp(a) and CMBs formation in CKD patients. Lp(a) carries oxidized phospholipids that promote vascular endothelial inflammation, macrophage infiltration, vascular smooth muscle cell proliferation, and subsequent vascular fibrosis, thereby increasing vessel fragility.19 These adverse effects might be further enhanced in CKD patients who generally suffer from persistent inflammation, oxidative stress, and endothelial dysfunction.20,21 In addition, renal clearance dysfunction in CKD patients may contribute to progressive accumulation of plasma Lp(a),22 resulting in persistent Lp(a)-mediated inflammatory responses and vascular remodeling cause continuous damage to cerebral microvessels. These pathological changes synergize with Lp(a)-induced vascular injury, damage the integrity of cerebral microvessels, and may raise the potential likelihood of CMB development, which may provide one potential mechanistic perspective for the high prevalence of CMBs in CKD populations. Previous studies exploring the association between Lp(a) and CMBs have largely yielded consistent findings. Multiple observational cohort studies conducted in community-dwelling populations have demonstrated that elevated Lp(a) is independently associated with CMBs,23 which is in line with our observations among patients with CKD.
What’s more, the association with deep rather than cortical or subcortical CMBs is biologically plausible. This difference is attributed to the different pathogenesis of CMBs in different brain regions. Deep CMBs are closely associated with hypertensive small-vessel hyalinosis and arteriosclerosis,23,24 while cortical CMBs are mainly attributable to cerebral amyloid angiopathy characterized by β-amyloid deposition.24,25 The abundant collateral circulation in subcortical regions can compensate for hemodynamic abnormalities attributable to Lp(a)-related vascular damage and reduce bleeding risk.26 Lp(a)-induced vascular damage mainly affects small arteries rich in smooth muscle cells: it may promotes collagen deposition and vascular wall calcification, weakens the vascular resistance to intravascular pressure, and may ultimately leads to vascular rupture and hemorrhage.19 These pathological changes are highly consistent with the pathogenesis of small artery sclerosis, which may be the main reason why Lp(a) is only associated with deep CMBs in CKD patients. This brain region-specific finding provides a new clinical insight into the mechanism by which Lp(a) may relate to CSVD in CKD patients.
Notably, partial correlation analysis showed that Lp(a) was no longer significantly correlated with deep CMBs severity in patients with established CMBs after adjusting for multiple confounding factors (r=0.175, P>0.05). This finding might support an threshold effect of Lp(a): Lp(a) may plays a role in the pathogenesis of deep CMBs, but it may not further promote the progression of existing lesions. Therefore, these data suggest Lp(a) may carry potential utility for early risk screening of CMBs rather than evaluating the progression of established lesions.
At present, clinical lipid management for CKD patients mainly focuses on lowering LDL-C, while the role of Lp(a) is often ignored. Even when LDL-C is well-controlled, CKD patients still face a high risk of cerebrovascular events.27 ROC analysis confirmed that Lp(a) has moderate predictive efficiency for deep CMBs (AUC=0.728), indicating that Lp(a) can be participated as one of serum biomarkers for risk stratification of CMBs. When plasma Lp(a) exceeds the corresponding cut-off value, clinicians may consider arrange cranial MRI for timely diagnosis and intervention. Although targeted drugs for Lp(a) have not been widely applied clinically, recent studies have demonstrated that antisense oligonucleotides can effectively reduce Lp(a) levels.28 Lp(a) represents a promising potential intervention target for protecting cerebral microvessels in CKD patients.
This study has several strengths. We investigated Lp(a) in a high-risk CKD cohort with stratified anatomical‑subgroup analyses of CMBs. With adjustment for clinical confounders, we examined both binary CMB status and severity-related correlations, yielding multi‑dimensional evidence indicating Lp(a) may represent a potential biomarker of CKD-related cerebral small-vessel injury.
This study has several limitations. First, the sample size of this study is relatively limited, and larger -scale multicenter surveys are still needed to further validate our findings. Second, this cross-sectional study cannot confirm the causal relationship between Lp(a) and CMBs. Long-term prospective follow-up studies are required to clarify the dynamic influence of Lp(a) on CMBs occurrence and progression. Third, our dataset does not include a control group of non-CKD participants with elevated Lp(a). Future observational studies should include non-CKD participants with elevated Lp(a) to permit comparisons of deep-CMB prevalence and further clarify whether CKD status modifies the association between Lp (a) and deep CMBs. Fourth, Lp(a) isoforms and related genetic factors were not detected in this study, and their roles remain to be explored. Fifth, subgroup analysis based on CKD stages was not performed. Sixth, the use of lipid-lowering and antihypertensive drugs was not collected, which may act as potential confounding factors.
Conclusion
In summary, in this selected cohort of CKD patients, higher circulating Lp(a) was independently associated with CMBs presence, particularly deep CMBs. These findings suggest that Lp(a) may serve as a potential serological candidate biomarker for risk stratification of CKD-related cerebral small-vessel injury.
Funding Statement
Zhejiang Province Hangzhou Medical and Health Science and Technology Project (Grant No.: ZD20250085); Zhejiang Province Medical and Health Science and Technology Program (Grant No.: 2025ky1142).
Disclosure
The authors report no conflicts of interest in this work.
References
- 1.Stevens PE, Ahmed SB, Carrero JJ, et al. KDIGO 2024 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int. 2024;105(4):S117–11. [DOI] [PubMed] [Google Scholar]
- 2.Couser WG, Remuzzi G, Mendis S, et al. The contribution of chronic kidney disease to the global burden of major noncommunicable diseases. Kidney Int. 2011;80(12):1258–1270. doi: 10.1038/ki.2011.368 [DOI] [PubMed] [Google Scholar]
- 3.Hanna RM, Ferrey A, Rhee CM, et al. Renal-cerebral pathophysiology: the interplay between chronic kidney disease and cerebrovascular disease. J Stroke Cerebrovasc Dis. 2021;30(9):105461. doi: 10.1016/j.jstrokecerebrovasdis.2020.105461 [DOI] [PubMed] [Google Scholar]
- 4.Nannoni S, Ohlmeier L, Brown RB, et al. Cognitive impact of cerebral microbleeds in patients with symptomatic small vessel disease. Int J Stroke. 2021;17(4):415–424. doi: 10.1177/17474930211012837 [DOI] [PubMed] [Google Scholar]
- 5.Haller S, Vernooij MW, Kuijer JPA, et al. Cerebral Microbleeds: imaging and Clinical Significance. Radiology. 2018;287(1):11–28. doi: 10.1148/radiol.2018170803 [DOI] [PubMed] [Google Scholar]
- 6.Miwa K, Toyoda K. Covert vascular brain injury in chronic kidney disease. Front Neurol. 2022;13:824503. doi: 10.3389/fneur.2022.824503 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ma Y, Yin X, Wang C. Diagnosis of cerebral microbleeds and assessment of prognosis in acute cerebral infarction using magnetic susceptibility-weighted angiography. Chin J Med Imaging. 2019;(08):584–588. [Google Scholar]
- 8.Tang J, Zhang M, Liu N, et al. The association between glymphatic system dysfunction and cognitive impairment in cerebral small vessel disease. Front Aging Neurosci. 2022;14:916633. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Reyes-soffer G, Ginsberg HN, Berglund L, et al. Lipoprotein(a): a genetically determined, causal, and prevalent risk factor for atherosclerotic cardiovascular disease: a scientific statement from the American Heart Association. Arterioscler Thromb Vasc Biol. 2022;42(1). doi: 10.1161/ATV.0000000000000147 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Tsioulos G, Kounatidis D, Vallianou NG, et al. Lipoprotein(a) and atherosclerotic cardiovascular disease: where do we stand? Int J Mol Sci. 2024;25(6):3537. doi: 10.3390/ijms25063537 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wang Y, Gao L. Inflammation and cardiovascular disease associated with hemodialysis for end-stage renal disease. Front Pharmacol. 2022;13:800950. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Nurmohamed NS, Gaillard EL, Malkasian S, et al. Lipoprotein(a) and long-term plaque progression, low-density plaque, and pericoronary inflammation. JAMA Cardiol. 2024;9(9):826. doi: 10.1001/jamacardio.2024.1874 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Xie DF, Fang C, Crouzet C, et al. Development of cerebral microhemorrhages in a mouse model of hypertension. J Neuroinflammation. 2025;22(1). doi: 10.1186/s12974-025-03378-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hayden MR. Cerebral microbleeds are associated with brain endothelial cell activation-dysfunction and blood-brain barrier dysfunction/disruption, with an increased risk of hemorrhagic and ischemic stroke. Biomedicines. 2024;12(7):1463. doi: 10.3390/biomedicines12071463 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wang H, Zhang C, Qiu Y, et al. Blood-brain barrier dysfunction in cerebral microbleeds: from bedside to bench. Aging Dis. 2021;12(8):1898. doi: 10.14336/AD.2021.0514 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron. 1976;16(1):31–41. doi: 10.1159/000180580 [DOI] [PubMed] [Google Scholar]
- 17.Branch of Neurology, Chinese Medical Association; Cerebrovascular Disease Group, Branch of Neurology, Chinese Medical Association. 2020 Chinese guidelines for the diagnosis and treatment of cerebral small vessel disease. Chin J Neurol. 2022;55(08):807–818. [Google Scholar]
- 18.Gregoire SM, Chaudhary UJ, Brown MM, et al. The Microbleed Anatomical Rating Scale (Mars): reliability of a tool to map brain microbleeds. Neurology. 2009;73(21):1759–1766. doi: 10.1212/WNL.0b013e3181c34a7d [DOI] [PubMed] [Google Scholar]
- 19.Tsimikas S, Witztum JL. The role of oxidized phospholipids in mediating lipoprotein (a) atherogenicity. Curr Opin Lipidol. 2008;19(4):369–377. doi: 10.1097/MOL.0b013e328308b622 [DOI] [PubMed] [Google Scholar]
- 20.Hahn KM, Strutz F. The early diagnosis and treatment of chronic renal insufficiency. Dtsch Arztebl Int. 2024;121(13):428–435. doi: 10.3238/arztebl.m2024.0072 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Nyúl-Tóth Á, Patai R, Csiszar A, et al. Linking peripheral atherosclerosis to blood-brain barrier disruption: elucidating its role as a manifestation of cerebral small vessel disease in vascular cognitive impairment. GeroScience. 2024;46(6):6511–6536. doi: 10.1007/s11357-024-01194-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Dieplinger H, Lackner C, Kronenherg F, et al. Elevated plasma concentrations of lipoprotein(a) in patients with end-stage renal disease are not related to the size polymorphism of apolipoprotein(a). J Clin Invest. 1993;91(2):397–401. doi: 10.1172/JCI116213 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Lau WL, Nunes ACF, Vasilevko V, et al. Chronic kidney disease increases cerebral microbleeds in mouse and man. Transl Stroke Res. 2020;11(1):122–134. doi: 10.1007/s12975-019-00698-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Liu Y, Qiu X, Wang L. A study on the relationship between cerebral microbleeds and hypertension using magnetic susceptibility-weighted imaging. J Bengbu Medical College. 2016;41(04):518–521. [Google Scholar]
- 25.Duering M, Biessels GJ, Brodtmann A, et al. Neuroimaging standards for research into small vessel disease—advances since 2013. Lancet Neurol. 2023;22(7):602–618. doi: 10.1016/S1474-4422(23)00131-X [DOI] [PubMed] [Google Scholar]
- 26.Zhang J, Zhao X, Cao X, et al. A study on the relationship between collateral circulation and cerebral microbleeds in patients with acute cerebral infarction. Hebei Medicine. 2018;40(24):3777–3780. [Google Scholar]
- 27.Li Y. A Study on the Correlation Between Low LDL-C Levels and Cerebral Microhemorrhages in Patients with Ischemic Stroke. Hebei Medical University; 2016. [Google Scholar]
- 28.Mäkinen P, Ruotsalainen AK, Ylä-Herttuala S. Nucleic acid-based therapies for atherosclerosis. Curr Atheroscler Rep. 2020;22(2):10. doi: 10.1007/s11883-020-0826-2 [DOI] [PMC free article] [PubMed] [Google Scholar]