Abstract
Background
Emerging evidence suggests a potential role of endothelial activation and oxidative stress in the pathogenesis of diabetic kidney disease (DKD). Endothelial activation and stress index (EASIX) is a reliable alternative biomarker of endothelial dysfunction. We aimed to investigate the relationship between the EASIX and both all-cause and cardiovascular mortality in individuals with DKD.
Methods
Data were extracted from the National Health and Nutrition Examination Survey database. Multivariable Cox proportional hazards models were applied to assess the relationship between EASIX and mortality. Kaplan-Meier survival curves, Subgroup analyses, receiver operating characteristic curves, and sensitivity analyses were conducted. Additionally, restricted cubic spline analysis was employed to elucidate the nonlinear relationships between EASIX and hazard ratio in DKD patients.
Rsults
A total of 1700 participants were included in this study. Over a median follow-up of 87.6 months, 571 all-cause deaths were recorded. The highest EASIX group showed higher hazard ratios for all-cause and cardiovascular mortality in both crude and multivariable-adjusted models. In the fully adjusted model, the hazard ratios for the highest EASIX group were 1.59 (95% CI: 1.24–2.03; p < 0.001) for all-cause mortality and 1.67 (95% CI: 1.09–2.54; p = 0.018) for cardiovascular mortality, compared to the lowest tertile. Furthermore, a nonlinear relationship was observed, with a sharp increase in all-cause mortality risk beyond an EASIX threshold of 0.619.
Conclusion
The results showed EASIX as an independent predictor of all-cause and cardiovascular mortality in DKD patients and further revealed that a non-linear positive correlation exists between EASIX and both all-cause mortality and a linear relationship with cardiovascular mortality in DKD patients.
Keywords: Diabetic kidney disease, Endothelial activation and stress index, National health and nutrition examination survey, All-cause mortality, Cardiovascular mortality
Introduction
Diabetic kidney disease (DKD) represents a critical global health challenge, driven by the escalating prevalence of diabetes mellitus (DM). As a common complication affecting around 40% of DM patients [1], DKD is a leading risk factor for cardiovascular disease and end-stage renal disease. More critically, the coexistence of chronic kidney disease and diabetes mellitus is associated with a synergistic, multiplicative increase in the risk of adverse cardiovascular outcomes, including mortality [2]. Further corroborating this risk, the Silesia Diabetes-Heart Project reported a 2% increase in cardiovascular event risk among individuals with DKD compared to those without [3]. Despite therapeutic advances, cardiovascular events remain the predominant cause of death among DKD patients. Therefore, elucidating the factors influencing mortality in DKD is crucial for optimizing clinical management and developing effective interventions.
Glomerular endothelial dysfunction is well recognized as contributing to the pathogenesis of DKD. Accumulating evidence indicates that aggravated injury and dysfunction of glomerular endothelial cells accelerate the progression of DKD [4, 5]. Among the underlying mechanisms, oxidative stress is widely recognized as a key contributor to cellular damage in DKD, and has been implicated as a plausible driver of glomerular endothelial injury [6]. Furthermore, oxidative stress and chronic inflammation can synergistically exacerbate endothelial damage, promoting the development of unstable atherosclerotic plaques [7, 8]. These interconnected pathways contribute to the substantially elevated cardiovascular risk observed in patients with DKD.
The Endothelial Activation and Stress Index (EASIX)—a composite biomarker based on three routinely measured laboratory parameters—has gained attention as a practical tool for clinical and epidemiological applications. Originally developed to predict survival following graft-versus-host disease [9], EASIX has since been validated for risk stratification in diverse settings, including before chimeric antigen receptor (CAR) T-cell therapy with idecabtagene vicleucel [10]. Recent work also supports its role as a long-term prognostic marker in coronary artery disease [11]. Furthermore, elevated EASIX has been consistently associated with adverse outcomes across multiple conditions, such as acute kidney injury, myelodysplastic syndromes, sepsis, CAR T-cell-related toxicities, and stem cell transplantation [12–16]. These observations strongly suggest that EASIX may serve as an independent predictor of mortality in varied clinical populations. Despite this broad evidence, the association between EASIX and the risks of all-cause and cardiovascular mortality, specifically in individuals with DKD, remains unexplored.
Therefore, utilizing data from the National Health and Nutrition Examination Survey (NHANES), this study aims to investigate the association between EASIX and cardiovascular and all-cause mortality in patients with DKD. The results are anticipated to establish a foundational basis for future mechanistic and prognostic investigations in this high-risk population.
Materials and methods
Data and study participants
The present analysis utilized data from the 2001–2018 cycles of the National Health and Nutrition Examination Survey (NHANES), a nationally representative, cross-sectional survey directed by the National Center for Health Statistics (NCHS). The survey employs a complex, stratified, multistage probability sampling design to offer a comprehensive overview of health status and lifestyle characteristics in the U.S. population. All phases of data collection—comprising household interviews, standardized physical examinations, and laboratory tests—were conducted under stringent NCHS quality control protocols. This integrated approach provides comprehensive data on demographics, clinical measures, laboratory biomarkers, dietary patterns, and health histories, thus offering a holistic profile of population health and determinants. The NCHS Research Ethics Review Board approved the study protocol, and written informed consent was secured from all participants before data collection.
The study initially included 91,351 participants from the NHANES cycles between 2001 and 2018. Participants were sequentially excluded for the following reasons: being under 18 years of age (n = 37,595); lacking follow-up data (n = 137); missing EASIX data (n = 6004); having insufficient information (n = 13854); not having diabetes mellitus (DM) (n = 28,479); without chronic kidney disease (CKD) (n = 3,248); participants who were pregnant or had cancer(n = 334). After these exclusions, 1700 participants remained for the final analysis (Fig. 1).
Fig. 1.
The flow chart of individual inclusion and exclusion in this study
Exposure variable
The biomarkers lactate dehydrogenase, creatinine, and platelet count were measured according to standardized NHANES protocols. Subsequently, the EASIX score was derived from these parameters using the established formula: lactate dehydrogenase (LDH, U/L) × creatinine (mg/dL)/platelet count (×10⁹/L) [9]. For analytical purposes, the study population was divided into tertiles based on the calculated EASIX values. The tertile thresholds were defined by the 33.3rd and 66.7th percentiles (0.488 and 0.798, respectively). Accordingly, the cohort was categorized into groups T1, T2, and T3, with T1 designated as the reference category in subsequent analyses.
Definition of diabetic kidney disease
Diabetes mellitus (DM) was defined according to established American Diabetes Association criteria and prior studies. Individuals meeting any of the following criteria were classified as having DM: (1) fasting plasma glucose (FPG) ≥ 7.0 mmol/L; (2) glycosylated hemoglobin (HbA1c) ≥ 6.5%; (3) a self-reported prior physician diagnosis; or (4) current use of glucose-lowering medication (American Diabetes Association).
Diabetic kidney disease (DKD) was diagnosed in participants with diabetes who also met the criteria for CKD, according to the KDIGO 2021 Guidelines [17]. CKD was defined as a urinary albumin-to-creatinine ratio (UACR) ≥ 30 mg/g and/or an estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m². The eGFR was calculated using the 2009 CKD-EPI creatinine equation [18]. Subsequently, staging of CKD (Stages I-V) was performed based on the eGFR value in accordance with established international guidelines. Albuminuria was classified into three stages based on UACR levels: A1 (< 30 mg/g), A2 (30–300 mg/g), and A3 (> 300 mg/g).
Covariates
The analysis adjusted for a comprehensive set of covariates. Demographic factors included age, gender, race (categorized as Mexican American, Non-Hispanic White, Non-Hispanic Black, and Other Races), educational attainment (classified as high school or less, above high school, or other), and the poverty-income ratio (PIR). Anthropometric measures comprised body mass index (BMI) and waist circumference. Laboratory parameters encompassed aspartate aminotransferase (AST), alanine aminotransferase (ALT), serum creatinine, uric acid, triglycerides, high-density lipoprotein cholesterol (HDL-C), haemoglobin A1c (HbA1c), and fasting plasma glucose. Behavioral covariates included smoking status (never, former, now) and drinking status, with participants classified as drinkers if they consumed ≥ 12 drinks in the previous year(Yes, No). Furthermore, hypertension was defined as a self-reported diagnosis, current use of antihypertensive medication, or an average systolic/diastolic blood pressure(SBP/DBP) > 140/90 mmHg across three separate measurements. Clinical characteristics included cardiovascular disease, defined as a self-reported diagnosis of one or more of the following conditions: heart failure, heart attack, coronary heart disease, angina, or stroke. This variable was categorized dichotomously as “Yes” or “No”. Additionally, prescription medication use was assessed based on self-reported use of any prescribed medication within the past month (categorized as Yes/No). Dialysis status (including hemodialysis or peritoneal dialysis) was assessed based on the participant’s response to the question, “In the past 12 months, have you received dialysis?” and was categorized as a binary variable (Yes/No).
Outcome assessment
The primary outcomes of this study were all-cause and cardiovascular mortality. Mortality data were ascertained through linkage with the National Death Index (NDI) (https://ftp.cdc.gov/pub/Health_Statistics/NCHS/datalinkage/linked_mortality/). The duration of follow-up was calculated from the date of the baseline survey until the date of death or December 31, 2019, whichever occurred first. Cardiovascular mortality was defined as deaths attributable to diseases of the heart and cerebrovascular system, identified using underlying cause-of-death codes I00-I09, I11, I13, I20-I51, and I60-I69 from the International Classification of Diseases, tenth revision (ICD-10).
Statistical analysis
All statistical analyses were performed using R (version 4.5.1). Continuous variables are expressed as mean ± standard deviation, and categorical variables as frequencies (percentages). The association between the EASIX score and mortality was evaluated using a series of Cox proportional hazards models: a crude model (Model 1), a model adjusted for demographic factors (Model 2: gender, age, race, education, PIR), and a fully adjusted model [Model 3 further adjusted for clinical and laboratory factors (BMI, waist circumference, hypertension, smoking and drinking status, medication use, dialysis status, cardiovascular diseases, SBP, DBP, ALT, AST, UACR, HbA1c.)]. The Kaplan-Meier method was used to visualize survival differences across EASIX groups. The predictive performance of EASIX was assessed by the area under the receiver operating characteristic (ROC) curve. Furthermore, the dose-response relationship was examined using restricted cubic splines within the Cox model. Interaction analyses were conducted to explore effect modifications across pre-specified subgroups. To further validate the robustness of our findings, we performed a series of sensitivity analyses. First, we constructed a model adjusting for estimated glomerular filtration rate (eGFR) alone to assess the stability of the association between the EASIX score and the outcomes, thereby testing the independent prognostic value of the composite metric. Second, to evaluate the potential influence of end-stage renal disease management, we excluded participants undergoing dialysis treatment and replicated the primary Cox regression analyses within this restricted sub-cohort. Third, to quantitatively assess the potential impact of unmeasured or residual confounding on the observed associations, we calculated E-values using the dedicated ‘EValue’ package in R software [19]. The E-value quantifies the minimum strength of association an unmeasured confounder would need to have with both the exposure and the outcome to fully explain away the observed hazard ratio. A two-sided p-value < 0.05 was considered statistically significant.
Results
Population characteristics
This study comprised a cohort of 1700 participants, of whom 935 were male, with a mean age of 63.4 years. Participants were stratified into three groups according to EASIX tertiles: T1 (< 0.488), T2 (0.488–0.798), and T3 (≥ 0.798). Significant differences across EASIX groups were observed in several baseline characteristics, including age, sex, race, comorbid hypertension and cardiovascular disease, smoking status, chronic kidney disease (CKD) stage, prescription medication use, dialysis status, systolic blood pressure (SBP) and diastolic blood pressure (DBP), estimated glomerular filtration rate (eGFR), serum uric acid (SUA), aspartate aminotransferase (AST), serum creatinine, urinary albumin to creatinine ratio (UACR), albuminuria grade, triglycerides (TG), lactate dehydrogenase (LDH), platelet count (PC) and haemoglobin A1c. In contrast, no statistically significant differences were found in poverty-income ratio(PIR), educational attainment, drinking habits, body mass index (BMI), waist circumference(WC), alanine aminotransferase (ALT), or high-density lipoprotein cholesterol (HDL-C) (Table 1).
Table 1.
Demographic and clinic characteristics
| Total (N = 1700) | T1 (N = 567) |
T2 (N = 566) |
T3 (N = 567) | p value | |
|---|---|---|---|---|---|
| Age, years | 63.4 (13.1) | 57.1 (13.6) | 64.1 (12.4) | 69.0 (10.1) | <0.001 |
| Gender, % | <0.001 | ||||
| Male | 935 (55.0%) | 217 (38.3%) | 333 (58.8%) | 385 (67.9%) | |
| Female | 765 (45.0%) | 350 (61.7%) | 233 (41.2%) | 182 (32.1%) | |
| Race, % | <0.001 | ||||
| Mexican American | 362 (21.3%) | 167 (29.5%) | 115 (20.3%) | 80 (14.1%) | |
| Non-Hispanic White | 605 (35.6%) | 161 (28.4%) | 224 (39.6%) | 220 (38.8%) | |
| Non-Hispanic Black | 426 (25.1%) | 114 (20.1%) | 135 (23.9%) | 177 (31.2%) | |
| Other races | 307 (18.1%) | 125 (22.0%) | 92 (16.3%) | 90 (15.9%) | |
| PIR | 2.13 (1.47) | 2.05 (1.48) | 2.18 (1.47) | 2.18 (1.46) | 0.227 |
| Education, % | 0.233 | ||||
| < High School | 346 (20.4%) | 109 (19.2%) | 124 (21.9%) | 113 (19.9%) | |
| Completed high school | 305 (17.9%) | 118 (20.8%) | 91 (16.1%) | 96 (16.9%) | |
| > High school | 1049 (61.7%) | 340 (60.0%) | 351 (62.0%) | 358 (63.1%) | |
| BMI, kg/m2 | 32.1 (7.05) | 32.5 (7.37) | 32.0 (6.95) | 31.7 (6.80) | 0.164 |
| WC, cm | 109 (15.8) | 109 (16.3) | 109 (15.8) | 110 (15.2) | 0.773 |
| Hypertension, % | <0.001 | ||||
| Yes | 1379 (81.1%) | 418 (73.7%) | 474 (83.7%) | 487 (85.9%) | |
| No | 321 (18.9%) | 149 (26.3%) | 92 (16.3%) | 80 (14.1%) | |
| Smoking, % | <0.001 | ||||
| Never | 801 (47.1%) | 269 (47.4%) | 261 (46.1%) | 271 (47.8%) | |
| Former | 623 (36.6%) | 180 (31.7%) | 210 (37.1%) | 233 (41.1%) | |
| Now | 276 (16.2%) | 118 (20.8%) | 95 (16.8%) | 63 (11.1%) | |
| Drinking, % | 0.404 | ||||
| Yes | 1133 (66.6%) | 366 (64.6%) | 386 (68.2%) | 381 (67.2%) | |
| No | 567 (33.4%) | 201 (35.4%) | 180 (31.8%) | 186 (32.8%) | |
| CVD, % | <0.001 | ||||
| Yes | 551(32.4%) | 109(19.2%) | 162(28.6%) | 280(49.4%) | |
| No | 1149(67.6%) | 458(80.8%) | 404(71.4%) | 287(50.6%) | |
| Medication use, % | <0.001 | ||||
| Yes | 1524 (89.6%) | 471 (83.1%) | 510 (90.1%) | 543 (95.8%) | |
| No | 176 (10.4%) | 96 (16.9%) | 56 (9.9%) | 24 (4.2%) | |
| CKD stage, % | <0.001 | ||||
| I-II | 958 (56.4%) | 505 (89.1%) | 324 (57.2%) | 129 (22.8%) | |
| III | 644 (37.9%) | 61 (10.8%) | 236 (41.7%) | 347 (61.2%) | |
| IV-V | 98 (5.76%) | 1 (0.18%) | 6 (1.06%) | 91 (16.0%) | |
| Dialysis, % | <0.001 | ||||
| Yes | 21 (1.24%) | 0 (0.00%) | 1 (0.18%) | 20 (3.53%) | |
| No | 1679 (98.8%) | 567 (100%) | 565 (99.8%) | 547 (96.5%) | |
| SBP, mmHg | 137 (22.0) | 135 (20.6) | 137 (21.1) | 139 (24.1) | 0.047 |
| DBP, mmHg | 69.9 (13.9) | 72.1 (13.4) | 70.2 (13.9) | 67.3 (14.0) | <0.001 |
| ALT, IU/L | 27.0 (41.5) | 26.7 (19.8) | 26.0 (15.9) | 28.2 (67.3) | 0.633 |
| AST, IU/L | 26.7 (22.2) | 25.0 (14.7) | 26.2 (13.1) | 28.9 (32.9) | 0.029 |
| eGFR, mL/min/1.73 m2 | 71.7 (28.5) | 94.8 (22.4) | 70.7 (22.1) | 49.4 (20.4) | <0.001 |
| SCR, µmol/L | 104 (63.3) | 69.0 (20.0) | 96.1 (24.0) | 147 (88.8) | <0.001 |
| SUA, µmol/L | 361 (99.8) | 317 (84.9) | 367 (90.9) | 399 (105) | <0.001 |
| UACR, mg/g | 343 (992) | 276 (807) | 255 (728) | 499 (1319) | <0.001 |
| Albuminuria grade, % | <0.001 | ||||
| A1 | 396 (23.3%) | 49 (8.64%) | 150 (26.5%) | 197 (34.7%) | |
| A2 | 969 (57.0%) | 424 (74.8%) | 315 (55.7%) | 230 (40.6%) | |
| A3 | 335 (19.7%) | 94 (16.6%) | 101 (17.8%) | 140 (24.7%) | |
| TG, mmol/L | 2.39 (2.40) | 2.64 (2.78) | 2.45 (2.60) | 2.07 (1.62) | <0.001 |
| HDL-C, mmol/L | 1.23 (0.38) | 1.22 (0.33) | 1.24 (0.42) | 1.23 (0.39) | 0.547 |
| LDH, IU/L | 144 (36.2) | 127 (25.1) | 140 (28.9) | 164 (41.9) | <0.001 |
| PC, 109/L | 242 (76.2) | 297 (80.3) | 237 (52.0) | 193 (52.6) | <0.001 |
| HbA1c, % | 7.70 (2.02) | 8.29 (2.36) | 7.47 (1.82) | 7.34 (1.70) | <0.001 |
Median (IQR) for continuous; n (%) for categorical
EASIX endothelial activation and stress index, PIR poverty-income ratio, BMI body mass index, WC Waist circumference, CVD Cardiovascular diseases, SBP systolic blood pressure, DBP diastolic blood pressure, AST aspartate transaminase, ALT alanine transaminase, eGFR estimated glomerular filtration rate, SCR serum creatinine, SUA serum uric acid, UACR urinary albumin to creatinine ratio, TG triglycerides, HDL-C high-density lipoprotein cholesterol, LDH Lactate dehydrogenase, PC platelet count, HbA1C haemoglobin A1c. Bold values indicate a statistically significant difference at the p < 0.05 level
Association between EASIX and mortality of adults with DKD
To evaluate the independent association between the EASIX and mortality, we constructed three multivariate Cox regression models with progressive adjustment for covariates. In the crude model (Model 1), participants in the highest EASIX tertile (T3) had significantly higher risks of all-cause (HR = 2.99, 95% CI 2.42–3.70, p < 0.001) and cardiovascular mortality (HR = 3.31, 95% CI 2.29–4.77, p < 0.001) compared to those in the lowest tertile (T1). Model 2 was adjusted for sociodemographic factors (age, gender, race, education level, and PIR). Model 3 was further adjusted for clinical characteristics, lifestyle factors, and laboratory parameters, including BMI, waist circumference, hypertension status, systolic and diastolic blood pressure, smoking status, alcohol and prescription medication use, dialysis, comorbid cardiovascular diseases, ALT, AST, UACR, and HbA1c.
The analysis revealed that the highest EASIX tertile (T3) consistently predicted elevated mortality risks across models. Notably, after accounting for all potential confounders in Model 3, T3 remained significantly associated with a 59% increased risk of all-cause death (HR 1.59, 95% CI 1.24–2.03, p < 0.001) and a 67% increased risk of CVD death (HR 1.67, 95% CI 1.09–2.54, p = 0.018) (Table 2), confirming EASIX as an independent risk factor.
Table 2.
Multivariate Cox regression analysis of EASIX and mortality
| Characteristic EASIX for All-cause mortality | HR | 95% CI | p-value | Characteristic EASIX for CVD mortality | HR | 95% CI | p-value | |
|---|---|---|---|---|---|---|---|---|
| Model 1 | T1 | Reference | — | T1 | Reference | — | ||
| T2 | 1.52 | 1.22, 1.90 | < 0.001 | T2 | 1.89 | 1.30, 2.75 | < 0.001 | |
| T3 | 2.99 | 2.42, 3.70 | < 0.001 | T3 | 3.31 | 2.29, 4.77 | < 0.001 | |
| Model 2 | T1 | Reference | — | T1 | Reference | — | ||
| T2 | 1.03 | 0.82, 1.30 | 0.78 | T2 | 1.23 |
0.83, 1.82 |
0.303 | |
| T3 | 1.71 | 1.35, 2.16 | < 0.001 | T3 | 1.76 | 1.18, 2.64 | 0.006 | |
| Model 3 | T1 | Reference | — | T1 | Reference | — | ||
| T2 | 1.05 |
0.83, 1.33 |
0.667 | T2 | 1.29 |
0.87, 1.92 |
0.207 | |
| T3 | 1.59 | 1.24, 2.03 | < 0.001 | T3 | 1.67 | 1.09, 2.54 | 0.018 |
95% CI, 95% confidence interval; HR, hazard ratio. CVD mortality, cardiovascular mortality
Model 1: No confounding factors were adjusted
Model 2:Adjusted for Age, Gender, Race, Education level, PIR
Model 3:Adjusted for Age, Gender, Race, Education level, PIR, BMI, hypertension, smoking and drinking status, SBP, DBP, ALT, AST, prescription medication use, dialysis, cardiovascular diseases, UACR, waist circumference, HbA1c
Bold values indicate a statistically significant difference at the p < 0.05 level
Kaplan-Meier survival curves demonstrated statistically significant differences across EASIX tertiles for both all-cause (log-rank p < 0.001) and cardiovascular mortality (log-rank p < 0.001) (Fig. 2). Visual inspection of the curves and the accompanying number-at-risk table confirms that participants in the highest tertile (T3) experienced a markedly steeper decline in survival probability over time compared to those in the lower tertiles (T1, T2), corroborating the findings from our Cox regression analysis.
Fig. 2.
Kaplan-Meier survival curve. (A) Kaplan-Meier survival curve for all-cause mortality. (B) Kaplan-Meier survival curve for cardiovascular mortality
Subgroup analysis and interaction analysis
Based on comprehensive subgroup analyses, EASIX demonstrated consistent associations with both all-cause mortality and cardiovascular mortality across most demographic and clinical strata. Notably, no significant effect modifications were observed by age, gender, race, education level, CKD stage, or smoking and drinking status (all p for interaction > 0.05) (Fig. 3).
Fig. 3.
Subgroup analysis and interaction analysis. (A) Subgroup analysis for the association between EASIX and all-cause mortality. (B) Subgroup analysis for the association between EASIX and cardiovascular mortality
Sensitivity and specificity analysis
The predictive performance of EASIX for all-cause and cardiovascular mortality was assessed using receiver operating characteristic (ROC) curve analysis (Fig. 4). EASIX demonstrated a discriminatory ability with an area under the curve (AUC) of 0.6290 (95% CI: 0.601, 0.657, p < 0.0001) for all-cause mortality and an AUC of 0.6041 (95% CI: 0.566, 0.643, p < 0.0001) for cardiovascular mortality. Although the Neutrophil-Percentage-to-Albumin Ratio (NPAR), Neutrophil-to-Lymphocyte Ratio (NLR), and Systemic Immune-inflammation Index (SII) have been implicated in predicting all-cause and cardiovascular mortality in diabetic nephropathy [20–22], a direct comparison of their prognostic performance was lacking. Therefore, we performed a synchronous comparative analysis. This analysis revealed that EASIX significantly outperformed NPAR, NLR, and SII (Tables 3, 4).
Fig. 4.
Sensitivity and specificity analysis. The ROC value of EASIX, NLR, NPAR and SII in predicting outcomes in DKD patients. (A) ROC curve for EASIX in predicting all-cause mortality. (B) ROC curve for EASIX in predicting cardiovascular mortality. AUC area under the curve, NPAR Neutrophil-Percentage-to-Albumin Ratio, NLR Neutrophil-to-Lymphocyte Ratio, SII Systemic Immune-inflammation Index
Table 3.
ROC analysis of the comparative ability of EASIX and parameters to predict all-cause mortality
| Parameters | AUC(95%CI) | Best threshold | Sensitivity(%) | Specificity(%) | p-value |
|---|---|---|---|---|---|
| EASIX | 0.6290(0.601–0.657) | 0.6552 | 61.38 | 60.39 | < 0.0001 |
| NLR | 0.5884(0.559–0.618) | 2.447 | 50.44 | 65.63 | < 0.0001 |
| NPAR | 0.5715(0.542–0.601) | 15.26 | 49.38 | 62.79 | < 0.0001 |
| SII | 0.5429(0.513–0.573) | 612.2 | 41.09 | 68.29 | 0.004 |
ROC receiver operating characteristic, AUC area under the curve. Bold values indicate a statistically significant difference at the p < 0.05 level
Table 4.
ROC analysis of the comparative ability of EASIX and parameters to predict cardiovascular mortality
| Parameters | AUC(95%CI) | Best threshold | Sensitivity(%) | Specificity(%) | p-value |
|---|---|---|---|---|---|
| EASIX | 0.6041(0.566–0.643) | 0.6023 | 69.9 | 49.7 | < 0.0001 |
| NLR | 0.5457(0.503–0.589) | 3.491 | 21.36 | 86.89 | 0.03 |
| NPAR | 0.5414(0.499–0.584) | 14.41 | 62.62 | 45.73 | 0.05 |
| SII | 0.5089(0.466–0.552) | 381.7 | 36.41 | 70.14 | 0.68 |
ROC receiver operating characteristic, AUC area under the curve. Bold values indicate a statistically significant difference at the p < 0.05 level
Restricted cubic spline regression analysis
To delineate the precise relationship between EASIX and mortality hazard ratios, we performed restricted cubic spline analyses with full adjustment for potential confounders (Fig. 5). After adjusting for multiple potential confounders, the nonlinear associations between EASIX and all-cause mortality were statistically significant (p for overall < 0.001; p for nonlinearity = 0.028), characterized by a sharp increase in the hazard ratio once EASIX exceeds a threshold of 0.619. In contrast, EASIX was significantly associated with cardiovascular mortality (p for overall = 0.001); after adjusting for all covariates, the RCS result revealed a linear association between EASIX and cardiovascular mortality (p for nonlinear = 0.077).
Fig. 5.
Restricted cubic spline regression analysis. (A) Non-linear relationship between EASIX and all-cause mortality. (B) Non-linear relationship between EASIX and cardiovascular mortality. All models were adjusted for demographic characteristics (age, gender, race, education level, PIR), anthropometric measurements (BMI, waist circumference), clinical parameters (SBP and DBP), laboratory indicators (AST, ALT, UACR, HbA1c), behavioral factors (smoking and drinking status), treatment profiles (medication use, dialysis) and clinical comorbidities (hypertension, cardiovascular disease)
Sensitivity analyses
The mortality risks were slightly attenuated in some sensitivity analyses. However, most of the results remained robustly consistent in all the sensitivity analyses (Table 5). The significant positive association for the T3 group persisted in all models. For all-cause mortality, hazard ratios (HRs) ranged from 1.40 to 1.59; for cardiovascular mortality, from 1.67 to 1.77. These analyses demonstrate that the association between high EASIX levels and increased mortality risk is robust to alternative model specifications, adjustments for key confounders like eGFR, and restriction to the non-dialysis subpopulation.
Table 5.
Sensitivity analysis of the associations between EASIX and mortality
| Mortality outcomes | HR (95% CI) Sensitivity−1 |
HR (95% CI) Sensitivity−2 |
HR (95% CI) Sensitivity−3 |
|
|---|---|---|---|---|
| All-cause mortality | ||||
| EASIX | T1 | Reference | Reference | Reference |
| T2 | 1.05(0.83–1.33) | 0.99(0.77–1.27) | 1.06(0.84–1.34) | |
| T3 | 1.59(1.25–2.04) | 1.40(1.03–1.90) | 1.57(1.23–2.01) | |
| CVD mortality | ||||
| EASIX | T1 | Reference | Reference | Reference |
| T2 | 1.29(0.87–1.92) | 1.33(0.87–2.04) | 1.32(0.88–1.96) | |
| T3 | 1.67(1.09–2.54) | 1.77(1.05–2.98) | 1.70(1.12–2.60) | |
95% CI, 95% confidence interval; HR, hazard ratio; CVD mortality, cardiovascular mortality
Sensitivity−1: Adjusted for Gender, Age, Race, Education level, PIR, BMI, Hypertension, Smoking and drinking status, Prescription medication use, Dialysis, Cardiovascular diseases, uacr, WC, HbA1c, ALT, AST;
Sensitivity−2: Adjusted for Gender, Age, Race, Education level, PIR, BMI, Hypertension, Smoking and drinking status, Prescription medication use, Dialysis, Cardiovascular diseases, uacr, WC, HbA1c, ALT, AST, eGFR;
Sensitivity−3: excluded participants with baseline dialysis and adjusted for confounding factors (Gender, Age, Race, Education level, PIR, BMI, Hypertension, Smoking and drinking status, Prescription medication use, Cardiovascular diseases, uacr, WC, HbA1c, ALT, AST)
Furthermore, the E value quantifying the impact of potential confounding factors was 2.20 for cardiovascular mortality and 2.10 for all-cause mortality. This indicates that an unmeasured confounder would need to be associated with both the EASIX and cardiovascular death by 2.2 and 2.1-fold, respectively, to render the lower bound of the HR confidence interval for the association between the EASIX and death risk < 1. Given that many known confounders have already been adjusted for, it is unlikely that a single unmeasured residual confounder could achieve such a magnitude of association.
Discussion
In this representative cohort study based on a general population sample, we present the first investigation into the association between the EASIX and mortality in the general population. Our findings reveal the following: After adjusting for potential confounding factors, EASIX is an independent risk factor for both all-cause mortality (HR = 1.59, 95% CI 1.24–2.03, p < 0.001) and cardiovascular mortality (HR = 1.67, 95% CI 1.09–2.54, p = 0.018); Furthermore, the relationship between EASIX and mortality exhibited distinct patterns: a non-linear association was identified for all-cause mortality, whereas a linear relationship was observed for cardiovascular mortality. EASIX demonstrated a modest but statistically significant predictive accuracy. These findings support that EASIX should be viewed as a practical and informative composite clinical indicator rather than a definitive diagnostic test. Its clinical utility stems from its ability to synthesize endothelial and inflammatory stress signals into a single metric, offering a complementary perspective that may enhance conventional risk stratification in DKD management. Collectively, EASIX is a robust prognostic biomarker for mortality risk in DKD and suggests its potential as a therapeutic target for improving clinical outcomes.
Epidemiological studies have consistently established that diabetic kidney disease portends a grave prognosis, with affected individuals facing a threefold increased risk of all-cause mortality and an average loss of 16 years in life expectancy [23]. Despite advances in standard-of-care treatment, a substantial residual mortality risk persists in a subset of patients, highlighting a critical clinical challenge: the current lack of reliable biomarkers to identify those at the highest risk for fatal outcomes. This unmet need underscores the imperative to elucidate the underlying pathophysiological drivers of this excess mortality. In this context, emerging evidence points to endothelial activation and oxidative stress as pivotal mechanisms in the pathogenesis and progression of DKD [24, 25], offering a promising avenue for novel risk stratification and therapeutic intervention. The accumulation of uremic toxins initiates a cascade of endothelial dysfunction, a central driver of cardiovascular disease and systemic inflammation. This impairment is characterized by increased oxidative stress, the upregulation of proinflammatory and prothrombotic molecules, and compromised endothelial repair mechanisms. A key manifestation is the toxin-mediated alteration in the expression of adhesion molecules—including E-selectin, P-selectin, intercellular adhesion molecule-1, and vascular cell adhesion molecule-1—which promotes the infiltration of monocytes and macrophages into the activated endothelium [26, 27]. Furthermore, uremic toxins directly disrupt vascular homeostasis by inhibiting nitric oxide production, a critical regulator of vascular tone, while simultaneously promoting the generation of reactive oxygen species. This imbalance exacerbates oxidative stress, leading to profound alterations in the regulatory properties of the vascular endothelium [28, 29]. Ultimately, these pathological changes facilitate diapedesis and heightened immune cell activity, particularly in advanced CKD, thereby accelerating the progression of CVD and inflammatory states [27, 30, 31].
Accumulating evidence underscores LDH as a robust prognostic variable across various patient populations. In the general context of critical illness, elevated serum LDH, a routinely available but non-specific marker of cellular injury, has been consistently linked to adverse outcomes. This is corroborated by machine learning analyses identifying LDH as a significant contributor to predicting mortality and ICU stay length [32]. Specifically, in end-stage renal disease, a linear and incremental association exists between higher LDH levels (> 280 U/L) and increased all-cause and cardiovascular mortality in incident dialysis patients [33]. The prognostic significance of LDH extends even to vulnerable, immunocompromised cohorts, as demonstrated by its association with 90-day mortality in renal transplant recipients with severe pneumonia [34]. Collectively, these findings affirm that LDH serves as a valuable and pervasive prognostic indicator in critically ill patients, irrespective of the underlying condition.
Platelets are pivotal in maintaining vascular homeostasis. Upon activation, they not only release extracellular vesicles but also enhance interactions with neutrophils, mast cells, and macrophages, thereby orchestrating inflammation-associated vascular permeability [35, 36]. This role is exemplified by research on dengue virus infection, where platelet-derived exosomes from infected patients have been shown to directly impair endothelial cell barrier integrity [37]. This disruption enhances vascular permeability and promotes the secretion of inflammatory markers. Paradoxically, a subsequent reduction in platelet count, often observed in severe disease, further aggravates vascular inflammation and disease severity, highlighting the complex and critical role of platelets in dysregulating vascular function.
The EASIX score integrates markers of renal dysfunction (creatinine), endothelial injury (LDH), and microvascular thrombosis (platelets). This pathophysiological breadth underpins its superior ability to predict mortality, independent of conventional risk factors, in endothelial activation-driven cardiovascular outcomes. Finke et al. observed that a similar hazard ratio (2–3 fold increased risk of death) for high EASIX values in both the training and validation cohorts with coronary artery disease [11]. The prognostic value of EASIX has been established in prior studies, particularly in relation to kidney-related outcomes. For instance, in the study of Zhang et al., EASIX is a reliable prognostic indicator for short-term mortality in sepsis-associated acute kidney injury patients [38]. These findings are consistent with our study’s conclusion. The elevated EASIX level signifies that such endothelial injury has reached a critical threshold, subsequently leading to microcirculatory dysfunction, tissue ischemia, and a substantially increased risk of atherosclerotic cardiovascular disease. This pathophysiological mechanism well explains the strong relationship observed between EASIX and cardiovascular mortality. Our study identifies a critical non-linear relationship between EASIX and all-cause mortality in DKD, marked by a sharply accelerated risk beyond the threshold of 0.619. We propose that this threshold marks a pivotal transition to irreversible endothelial dysfunction, where compensatory mechanisms are overwhelmed, precipitating a cascade of microcirculatory failure and multi-organ dysfunction that drives mortality. Thus, extreme EASIX elevation defines a high-risk DKD phenotype, burdened by an aggravated uremic, inflammatory, and comorbid state. Consequently, DKD patients with elevated EASIX face a significantly heightened risk of both all-cause and cardiovascular mortality (as evidenced by HRs of 1.59 and 1.67, respectively), enabling clinicians to identify this high-risk subgroup for more intensive management and proactive intervention.
The EASIX score holds notable practical advantages for clinical implementation. EASIX is calculated from three routinely measured laboratory parameters. These tests are standard components of hematologic panels in most hospital and outpatient settings. Therefore, the derivation of EASIX does not incur any additional direct laboratory costs, making it an economically viable tool in both resource-abundant and resource-limited healthcare systems. This would allow clinicians to identify high-risk DKD patients without adding to clinical workflow burdens. However, widespread adoption requires addressing several considerations. First, although the component tests are common, standardization of assay methods across laboratories (especially for LDH) is necessary to ensure score consistency. Second, prospective validation in diverse, multi-ethnic cohorts is needed to confirm generalizability. Therefore, incorporating EASIX into established risk prediction models (e.g., those based on age and blood pressure) holds promise for significantly enhancing model discrimination accuracy. Finally, serial monitoring of EASIX may be valuable in assessing the efficacy of interventions, such as sodium-glucose cotransporter-2 inhibitors or glucagon-like peptide-1 receptor agonists, which are known to confer cardiorenal benefits [39–41]. A decline in EASIX following treatment initiation could potentially serve as an early biomarker for improved vascular endothelial function and treatment effectiveness.
Our study possesses several strengths, including a substantial cohort of 1700 participants with comprehensive clinical and follow-up data over a median of 87.6 months. Nevertheless, some limitations merit consideration. First, the observational and cross-sectional design at baseline precludes the establishment of causality. All clinical and laboratory parameters, including the components of the EASIX score, were assessed at a single time point. Consequently, while associations with mortality can be identified, definitive causal relationships cannot be inferred. Second, despite extensive multivariable adjustment, residual confounding remains possible. We rigorously adjusted for key demographic, lifestyle, and clinical factors, including a history of cardiovascular disease, hypertension, dialysis status, and smoking. Although we adjusted for any prescription medication use as a proxy for overall pharmacologic burden, the publicly available data did not support a granular analysis of specific drug classes (e.g., specific antihypertensive drugs, antiplatelet agents) that might directly modulate EASIX components such as lactate dehydrogenase or platelet count. Other unmeasured factors (e.g., peripheral artery disease, acute kidney injury, severe infection, or hospitalization) may also influence the results. We calculated E-values to assess the potential impact of such unmeasured confounding, but its influence cannot be entirely excluded. Third, the generalizability of our findings may be limited. Our cohort comprised exclusively patients with diabetic kidney disease from the United States. Variations in genetic background, anthropometric characteristics, cardiovascular risk profiles, and healthcare systems across different ethnicities and geographic regions may affect the external validity of the EASIX score. Its predictive performance in non-Western populations requires independent validation. Finally, the precise biological mechanism linking the EASIX composite score to cardiovascular mortality warrants further elucidation. While EASIX integrates markers of endothelial and renal stress, the exact pathways through which these components interact to confer prognostic risk are not fully understood. Future translational studies are needed to enhance the pathophysiological interpretability and clinical utility of this biomarker. Despite these limitations, the EASIX score emerged as a robust predictor of mortality in DKD, underscoring its clinical relevance as a low-cost, readily available prognostic tool with utility in both routine practice and research on endothelial dysfunction. Further validation through large-scale, multi-center prospective studies is strongly warranted.
Conclusion
The purpose of this study was to explore the relationship between EASIX and the risk of mortality in patients with diabetic kidney disease. After adjusting for confounding variables, the level of EASIX was positively correlated with the risk of all-cause mortality and cardiovascular mortality. As a convenient and cost-effective marker, EASIX may play an important role in predicting the risk of death in DKD.
Acknowledgements
We thank the investigators, the staff, and the participants of NHANES for their valuable contribution.
Abbreviations
- EASIX
Endothelial activation and stress index
- DKD
Diabetic kidney disease
- NHANES
National health and nutrition examination survey
- NLR
Neutrophil-to-Lymphocyte Ratio
- SII
Systemic immune-inflammation Index
- NPAR
Neutrophil-percentage-to-albumin ratio
- CKD
Chronic kidney disease
- DM
Diabetes mellitus
- NCHS
National Center for Health Statistics
- eGFR
Estimated glomerular filtration rate
- UACR
Urinary albumin to creatinine ratio
- PC
Platelet count
- LDH
Lactate dehydrogenase
- ROC
Receiver operating characteristic
- CVD
Cardiovascular diseases
- WC
Waist circumference
- PIR
poverty income ratio
- BMI
Body mass index
- SUA
Serum uric acid
- TG
Triglycerides
- HDL-C
High-density lipoprotein cholesterol
- FPG
Fasting plasma glucose
- Scr
Serum creatinine
- BUN
Blood urea nitrogen
- AST
Aspartate aminotransferase
- ALT
Alanine aminotransferase
- HbA1c
Haemoglobin A1c
- SBP
Systolic blood pressure
- DBP
Diastolic blood pressure
- AUC
Area under the curve
Author contributions
WHY: Research conceptualization, statistical analysis and manuscript drafting. JYZ, QC, MX, BL: Data collation and visualization. LLW, CXL, XZT, XYF and MXH: Data cleansing and statistical analyses. CYT, GCC, YL, FY and HL: Manuscript revisions. HL: Research conceptualization and research designs. All authors collaborated on the article and cleared the presented version.
Funding
This work was supported by the National Natural Science Foundation of China (No. 82470741, 82270752, 82270733), Natural Science Foundation of Hunan Province (No. 2025JJ30042), Scientific Research Project of the Health Commission of Hunan Province (No. A202303057091) and 2023 Hunan Provincial Health High-Level Talents Program - Leading Talent Cultivation Funding (Liu Hong).
Data availability
All data analyzed in this study were derived from the National Health and Nutrition Examination Survey (NHANES) (https://www.cdc.gov/nchs/nhanes/index.html).
Declarations
Ethics approval and consent to participate
This study used anonymous data from the National Health and Nutrition Examination Survey and complied with the ethical guidelines and regulations of the Declaration of Helsinki. The study was approved by the National Center for Health Statistics Ethics Review Board, and all participants provided written informed consent before the study.
Consent for publication
All participants signed the informed consent before participating in the study. All methods were carried out in accordance with relevant guidelines and regulations. All the authors agreed to submit the manuscript to this journal for consideration.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Martinez Leon V, Hilburg R, Susztak K. Mechanisms of diabetic kidney disease and established and emerging treatments. Nat Rev Endocrinol. 2025. 10.1038/s41574-025-01171-3. [DOI] [PubMed] [Google Scholar]
- 2.Ruhe J, Nadal J, Barthlein B, Meiselbach H, Schultheiss UT, Kotsis F, et al. Cardiovascular risk due to diabetes mellitus in patients with chronic kidney disease-prospective data from the German Chronic Kidney Disease cohort. Clin Kidney J. 2023;16:2032–40. 10.1093/ckj/sfad194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Janota-Sosinska O, Mantovani M, Irlik K, Kwiendacz H, Olejarz A, Pabis P, et al. Diabetic kidney disease phenotypes and the risk of cardiovascular events: the Silesia Diabetes-Heart project. Cardiovasc Diabetol. 2025;24:305. 10.1186/s12933-025-02852-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Zhang J, Wu Y, Zhang J, Zhang R, Wang Y, Liu F. ABCA1 deficiency-mediated glomerular cholesterol accumulation exacerbates glomerular endothelial injury and dysfunction in diabetic kidney disease. Metabolism. 2023;139:155377. 10.1016/j.metabol.2022.155377. [DOI] [PubMed] [Google Scholar]
- 5.Hong Q, Zhang L, Fu J, Verghese DA, Chauhan K, Nadkarni GN, et al. LRG1 promotes diabetic kidney disease progression by enhancing TGF-beta-Induced angiogenesis. J Am Soc Nephrol. 2019;30:546–62. 10.1681/ASN.2018060599. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hu S, Hang X, Wei Y, Wang H, Zhang L, Zhao L. Crosstalk among podocytes, glomerular endothelial cells and mesangial cells in diabetic kidney disease: an updated review. Cell Commun Signal. 2024;22:136. 10.1186/s12964-024-01502-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Yang C, Wang H, Zhao X, Matsushita K, Coresh J, Zhang L, et al. CKD in China: evolving spectrum and public health implications. Am J Kidney Dis. 2020;76:258–64. 10.1053/j.ajkd.2019.05.032. [DOI] [PubMed] [Google Scholar]
- 8.Yang Z, Huang Y, Zhu L, Yang K, Liang K, Tan J, et al. <article-title update="added">SIRT6 promotes angiogenesis and hemorrhage of carotid plaque via regulating HIF-1α and reactive oxygen species. Cell Death Dis. 2021;12(1):77. 10.1038/s41419-020-03372-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Luft T, Benner A, Jodele S, Dandoy CE, Storb R, Gooley T, et al. EASIX in patients with acute graft-versus-host disease: a retrospective cohort analysis. Lancet Haematol. 2017;4:e414–23. 10.1016/S2352-3026(17)30108-4. [DOI] [PubMed] [Google Scholar]
- 10.Frenking JH, Zhou X, Wagner V, Hielscher T, Kauer J, Mai EK, et al. EASIX-guided risk stratification for complications and outcome after CAR T-cell therapy with ide-cel in relapsed/refractory multiple myeloma. J Immunother Cancer. 2024. 10.1136/jitc-2024-009220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Finke D, Hund H, Frey N, Luft T, Lehmann LH. EASIX (endothelial activation and stress index) predicts mortality in patients with coronary artery disease. Clin Res Cardiol. 2025;114:1008–18. 10.1007/s00392-024-02534-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Merz A, Germing U, Kobbe G, Kaivers J, Jauch A, Radujkovic A, et al. EASIX for prediction of survival in lower-risk myelodysplastic syndromes. Blood Cancer J. 2019;9:85. 10.1038/s41408-019-0247-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Xu HB, Ye Y, Xue F, Wu J, Suo Z, Zhang H. Association between endothelial activation and stress index and 28-day mortality in septic ICU patients: a retrospective cohort study. Int J Med Sci. 2023;20:1165–73. 10.7150/ijms.85870. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kordelas L, Terzer T, Gooley T, Davis C, Sandmaier BM, Sorror M, et al. EASIX-1year and late mortality after allogeneic stem cell transplantation. Blood Adv. 2023;7:5374–81. 10.1182/bloodadvances.2022008617. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Penack O, Luft T, Peczynski C, Benner A, Sica S, Arat M, et al. Endothelial activation and stress index (EASIX) to predict mortality after allogeneic stem cell transplantation: a prospective study. J Immunother Cancer. 2024. 10.1136/jitc-2023-007635. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wan DY, Luo XY, He M, Zhang ZW. Exploring trajectories of acute kidney injury in the intensive care unit: a population-based cohort study. Sci Rep. 2025;15:33056. 10.1038/s41598-025-11587-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Kidney Disease: Improving Global Outcomes Glomerular Diseases Work G. KDIGO 2021 clinical practice guideline for the management of glomerular diseases. Kidney Int. 2021;100:S1–276. 10.1016/j.kint.2021.05.021. [DOI] [PubMed] [Google Scholar]
- 18.Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150:604–12. 10.7326/0003-4819-150-9-200905050-00006. [DOI] [PMC free article] [PubMed]
- 19.VanderWeele TJ, Ding P. Sensitivity analysis in observational research: introducing the E-value. Ann Intern Med. 2017;167:268–74. 10.7326/M16-2607. [DOI] [PubMed] [Google Scholar]
- 20.Zhang M, Ye S, Li J, Zhang M, Tan L, Wang Y, et al. Association of systemic immune-inflammation index with all-cause and cardio-cerebrovascular mortality in individuals with diabetic kidney disease: evidence from NHANES 1999–2018. Front Endocrinol (Lausanne). 2024;15:1399832. 10.3389/fendo.2024.1399832. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Zeng G, Lin Y, Xie P, Lin J, He Y, Wei J. Relationship of the neutrophil-lymphocyte ratio with all-cause and cardiovascular mortality in patients with diabetic kidney disease: a prospective cohort study of NHANES study. J Multidiscip Healthc. 2024;17:2461–73. 10.2147/JMDH.S465317. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Tan J, Du J, Liu J, Zhao W, Liu Y. Prognostic effect of neutrophil percentage-to-albumin ratio (NPAR) on all-cause and cardiovascular mortality in diabetic kidney disease (DKD): NHANES 1999–2018. Diabetol Metab Syndr. 2025;17:105. 10.1186/s13098-025-01674-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Naaman SC, Bakris GL. Diabetic nephropathy: update on pillars of therapy slowing progression. Diabetes Care. 2023;46:1574–86. 10.2337/dci23-0030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Wang N, Zhang C. Oxidative stress: a culprit in the progression of diabetic kidney disease. Antioxidants (Basel). 2024. 10.3390/antiox13040455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Jandeleit-Dahm KAM, Kankanamalage HR, Dai A, Meister J, Lopez-Trevino S, Cooper ME, et al. Endothelial NOX5 obliterates the reno-protective effect of Nox4 deletion by promoting renal fibrosis via activation of EMT and ROS-sensitive pathways in diabetes. Antioxidants. 2024. 10.3390/antiox13040396. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Cunha RSD, Santos AF, Barreto FC, Stinghen AEM. How do uremic toxins affect the endothelium? Toxins (Basel). 2020. 10.3390/toxins12060412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Eloueyk A, Osta B, Alameldinne R, Awad D. Uremic serum induces inflammation in cultured human endothelial cells and triggers vascular repair mechanisms. Inflammation. 2019;42:2003–10. 10.1007/s10753-019-01061-7. [DOI] [PubMed] [Google Scholar]
- 28.Masai N, Tatebe J, Yoshino G, Morita T. Indoxyl sulfate stimulates monocyte chemoattractant protein-1 expression in human umbilical vein endothelial cells by inducing oxidative stress through activation of the NADPH oxidase-nuclear factor-kappaB pathway. Circ J. 2010;74:2216–24. 10.1253/circj.cj-10-0117. [DOI] [PubMed] [Google Scholar]
- 29.Huang M, Wei R, Wang Y, Su T, Li P, Chen X. The uremic toxin hippurate promotes endothelial dysfunction via the activation of Drp1-mediated mitochondrial fission. Redox Biol. 2018;16:303–13. 10.1016/j.redox.2018.03.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Jourde-Chiche N, Fakhouri F, Dou L, Bellien J, Burtey S, Frimat M, et al. Endothelium structure and function in kidney health and disease. Nat Rev Nephrol. 2019;15:87–108. 10.1038/s41581-018-0098-z. [DOI] [PubMed] [Google Scholar]
- 31.Kim HY, Yoo TH, Hwang Y, Lee GH, Kim B, Jang J, et al. Indoxyl sulfate (IS)-mediated immune dysfunction provokes endothelial damage in patients with end-stage renal disease (ESRD). Sci Rep. 2017;7:3057. 10.1038/s41598-017-03130-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Iwase S, Nakada TA, Shimada T, Oami T, Shimazui T, Takahashi N, et al. Prediction algorithm for ICU mortality and length of stay using machine learning. Sci Rep. 2022;12:12912. 10.1038/s41598-022-17091-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Ryu SY, Kleine CE, Hsiung JT, Park C, Rhee CM, Moradi H, et al. Association of lactate dehydrogenase with mortality in incident hemodialysis patients. Nephrol Dial Transplant. 2021;36:704–12. 10.1093/ndt/gfaa277. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Su Y, Ju MJ, Ma JF, Tu GW, He HY, Gu ZY, et al. Lactate dehydrogenase as a prognostic marker of renal transplant recipients with severe community-acquired pneumonia: a 10-year retrospective study. Ann Transl Med. 2019;7:660. 10.21037/atm.2019.10.75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Sung PS, Huang TF, Hsieh SL. Extracellular vesicles from CLEC2-activated platelets enhance dengue virus-induced lethality via CLEC5A/TLR2. Nat Commun. 2019;10:2402. 10.1038/s41467-019-10360-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Rathore AP, Mantri CK, Aman SA, Syenina A, Ooi J, Jagaraj CJ, et al. Dengue virus-elicited tryptase induces endothelial permeability and shock. J Clin Invest. 2019;129:4180–93. 10.1172/JCI128426. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Vedpathak S, Sharma A, Palkar S, Bhatt VR, Patil VC, Kakrani AL, et al. Platelet derived exosomes disrupt endothelial cell monolayer integrity and enhance vascular inflammation in dengue patients. Front Immunol. 2023;14:1285162. 10.3389/fimmu.2023.1285162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Zhang Z, Wang C, Wang J, Luo Z, Zou K, Xu Q. Endothelial activation and stress index elevation associated with adverse prognosis in Sepsis-associated acute kidney injury patients: validation based on two critical care cohorts. Shock. 2025. 10.1097/SHK.0000000000002715. [DOI] [PubMed] [Google Scholar]
- 39.Kaze AD, Zhuo M, Kim SC, Patorno E, Paik JM. Association of SGLT2 inhibitors with cardiovascular, kidney, and safety outcomes among patients with diabetic kidney disease: a meta-analysis. Cardiovasc Diabetol. 2022;21:47. 10.1186/s12933-022-01476-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Perkovic V, Jardine MJ, Neal B, Bompoint S, Heerspink HJL, Charytan DM, et al. Canagliflozin and renal outcomes in type 2 diabetes and nephropathy. N Engl J Med. 2019;380:2295–306. 10.1056/NEJMoa1811744. [DOI] [PubMed] [Google Scholar]
- 41.Shaman AM, Bain SC, Bakris GL, Buse JB, Idorn T, Mahaffey KW, et al. Effect of the glucagon-like peptide-1 receptor agonists Semaglutide and Liraglutide on kidney outcomes in patients with type 2 diabetes: pooled analysis of SUSTAIN 6 and LEADER. Circulation. 2022;145:575–85. 10.1161/CIRCULATIONAHA.121.055459. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All data analyzed in this study were derived from the National Health and Nutrition Examination Survey (NHANES) (https://www.cdc.gov/nchs/nhanes/index.html).