Skip to main content
NCBI home page
Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease logoLink to Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease
. 2026 Jul 3;15(13):e047057. doi: 10.1161/JAHA.125.047057

Proteomics‐Based Soluble Urokinase Plasminogen Activator Receptor Levels Are Associated With Adverse Cardiovascular Outcomes in the General Population: Insights From the UK Biobank

Shaimaa M Sakr 1, Jose R Medina‐Inojosa 1, Chang Liu 2, Ishwar D Chuckaree 3, Veena Agusala 3, Rafia Lodhi 1, Zara Mecklai 1, Muhammad Owais 1, Mahmoud Al Kasem 1, Wesam Hritani 1, Muhammet Hasan 1, Salim Hayek 4,5, Yan V Sun 2, Arshed A Quyyumi 1,✉
PMCID: PMC13477334  PMID: 42396825

Abstract

Background

Elevated soluble urokinase plasminogen activator receptor (suPAR) levels are associated with inflammation, immune activation, and major adverse cardiovascular events in coronary artery disease. Encoded by the PLAUR gene, suPAR levels are influenced by the rs4760 genetic variant. Whether proteomics‐based suPAR levels predict adverse outcomes in the general population remains unknown.

Methods

Proteomics‐based suPAR levels were measured using the Olink Immunoassay in 33 963 UK Biobank participants without known coronary artery disease. Fine‐Gray and Cox proportional hazards models assessed associations between suPAR and major adverse cardiovascular events (primary outcome: cardiovascular mortality, nonfatal myocardial infarction, or stroke), cardiovascular mortality, and all‐cause mortality (secondary outcomes), after adjustment for demographic and clinical risk factors, hs‐CRP (high‐sensitivity C‐reactive protein), and the rs4760 variant. Incremental discrimination was evaluated using C‐statistics.

Results

Participants were aged 56.4 (SD, 8.2) years; 45% were men, and 93.4% were White. Over a median follow‐up of 14 years (476 177 person‐years), 10.7% experienced major adverse cardiovascular events, 2.6% experienced cardiovascular mortality, and 9.4% experienced all‐cause mortality. Each 1‐SD increment in proteomics‐based suPAR was associated with significantly higher risk of major adverse cardiovascular events (hazard ratio [HR], 3.2 [95% CI, 2.9–4.5]), cardiovascular mortality (HR, 5.9 [95% CI, 5.1–6.9]), and all‐cause mortality (HR, 5.0 [95% CI, 4.6–5.5]), independent of clinical risk factors and hs‐CRP. Additional adjustment for rs4760 did not attenuate these associations. Proteomics‐based suPAR significantly improved discrimination beyond clinical risk factors (C‐statistic: 0.719 versus 0.732; P<0.001).

Conclusions

Proteomics‐based suPAR independently predicts adverse cardiovascular outcomes in the general population, beyond conventional risk factors, hs‐CRP, and genetic predisposition to elevated suPAR levels.

Keywords: coronary artery disease, hs‐CRP (high‐sensitivity C‐reactive protein), inflammation, plasminogen activator urokinase receptor, urokinase plasminogen activator

Subject Categories: Biomarkers, Clinical Studies, Cardiovascular Disease, Risk Factors

Nonstandard Abbreviations and Acronyms

IDI

integrated discrimination improvement

MACE

major adverse cardiovascular event

NRI

net reclassification index

PLAUR

plasminogen activator urokinase receptor

suPAR

soluble urokinase plasminogen activator receptor

UKB

UK Biobank

Clinical Perspective.

What Is New?

  • In a large general population cohort without known coronary artery disease, elevated proteomics‐based soluble urokinase plasminogen activator receptor levels were strongly associated with major adverse cardiovascular events, cardiovascular mortality, and all‐cause mortality, independent of traditional risk factors, hs‐CRP (high‐sensitivity C‐reactive protein) levels, and the plasminogen activator urokinase receptor rs4760 genetic variant.

  • Proteomics‐based soluble urokinase plasminogen activator receptor significantly improved risk discrimination beyond established clinical models, highlighting its incremental prognostic value in cardiovascular risk assessment.

What Are the Clinical Implications?

  • Soluble urokinase plasminogen activator receptor levels may serve as a robust biomarker to identify individuals in the general population at heightened long‐term cardiovascular risk who may also benefit from earlier preventive strategies and closer clinical surveillance.

Atherosclerosis is a complex process, with inflammation and immune system activation being significant risk enhancers, leading to adverse cardiovascular events and mortality. 1 , 2 The urokinase system has emerged as a significant signal and potential driver of chronic inflammation and immune activation 3 and plays a central role in the development of atherosclerosis and its associated cardiovascular complications. 4 The urokinase plasminogen activator receptor is encoded by the plasminogen activator urokinase receptor (PLAUR) gene, and on cleavage, its soluble form, the soluble urokinase plasminogen activator receptor (suPAR) is released into the circulation. 3 suPAR contributes to atherosclerosis by promoting the adhesion, migration, and proliferation of inflammatory cells, facilitates interactions between immune cells and endothelial cells, and promotes the formation and progression of atherosclerotic plaque. 3 , 4 Elevated suPAR levels have been linked to the presence and severity of coronary artery disease 5 and with adverse cardiovascular outcomes in both healthy individuals and those with coronary artery disease. 5 , 6 , 7 Additionally, suPAR has been identified as a key driver of kidney function decline. 8 , 9 , 10 The rs4760 variant is a missense mutation in the PLAUR gene that is associated with higher circulating suPAR levels in experimental settings. 11 , 12

Although previous studies have established the relationship between circulating suPAR levels and cardiovascular outcomes in the general population, they have been constrained by smaller sample sizes and limited follow‐up periods. 13 , 14 This investigation addresses these gaps by using a substantially larger cohort of >33 000 individuals enrolled in the UK Biobank (UKB), with a median follow‐up of 14 years that provides a unique opportunity to assess the long‐term prognostic value of proteomics‐based suPAR levels, on predicting major adverse cardiovascular events (MACEs). Previous studies have confirmed a modest correlation between suPAR measured by ELISA and Olink‐based measurements. 15 This modest correlation likely reflects differences in detected suPAR isoforms, posttranslational processing, and assay‐specific characteristics, including cross‐reactivity and variable sensitivity to distinct circulating suPAR forms. 15 We therefore sought to evaluate whether circulating proteomics‐based suPAR, measured using the Olink proteomics immunoassay, provides biologically relevant information related to urokinase plasminogen activator–uPAR pathway activity and serves as a scalable approach for risk assessment where ELISA‐based suPAR platforms are unavailable. Importantly, we also investigated the impact of the traditional marker of inflammation, hs‐CRP (high‐sensitivity C‐reactive protein) and the rs4760 variant in the PLAUR gene on the relationship between proteomics‐based suPAR and adverse outcomes. 12

METHODS

The research was conducted using data from the UK Biobank Resource under application number 34031. The UKB will make the data available to all bona fide researchers for all types of health‐related research that is in the public interest, without preferential or exclusive access for any individuals. All researchers will be subject to the same application process and approval criteria as specified by UKB. For more details on the access procedure, see the UKB website: www.ukbiobank.ac.uk.

The study received approval from the Institutional Review Board of the Northwest Multicenter Research Ethics Committee in the United Kingdom. All participants provided electronic signed informed consent at the time of their assessment visit. The UKB cohort has been previously described. 16 Briefly, the UKB is a large‐scale biomedical biobank and research initiative in the United Kingdom, involving ≈500 000 participants aged 40 to 69 years, who were recruited between 2006 and 2010. The UKB cohort is linked to Hospital Episode Statistics data, which include all hospital admissions, operations, and procedures, with records dating back to 1997 for England, 1998 for Wales, and 1981 for Scotland. Additionally, the cohort was linked to primary care data and the death registry, which provided information on the date of death and both primary and secondary causes. The first occurrence of disease diagnosis and adverse outcomes was mapped to International Classification of Diseases, Tenth Revision (ICD‐10) codes, integrating primary care data, Hospital Episode Statistics data, death registry, and self‐reported medical conditions. Disease prevalence and incidence of outcomes were defined by comparing the onset date and enrollment date: diseases and outcomes with onset on or before the enrollment date were classified as prevalent, whereas those with onset after enrollment were considered incident.

Patients with prevalent coronary artery disease (ICD‐10 codes I20 and I25), myocardial infarction (ICD‐10 codes I21–I23 and I25), and heart failure (ICD‐10 code I50) were excluded from the study to evaluate proteomics‐based suPAR's predictive value in individuals without existing cardiovascular disease (CVD). This ensured that proteomics‐based suPAR was assessed as an independent risk predictor, rather than a marker of established CVD, while still accounting for common risk factors, such as hypertension and diabetes.

The primary MACE outcome was defined as a composite event of cardiovascular mortality, nonfatal myocardial infarction, and ischemic stroke. Cardiovascular mortality was determined based on ICD‐10 codes corresponding to diseases of the heart (I00–I09, I11, I13, and I20–I51), essential hypertension and hypertensive renal disease (I10, I12, and I15), and cerebrovascular diseases (I60–I69). 17 Secondary outcomes included cardiovascular mortality and all‐cause mortality, assessed independently to evaluate the broader prognostic relevance of proteomics‐based suPAR.

Follow‐up time accrued from the UKB baseline assessment (enrollment) date until the earliest of the first occurrence of the incident outcome of interest, death from any cause, loss to follow‐up, or censoring at the end of available linked follow‐up data as of September 2023. At enrollment, demographic information and risk factors collected included age, sex, race and ethnicity, body mass index (BMI), history of hypertension, diabetes, and smoking, total cholesterol, triglycerides, low‐density lipoprotein cholesterol and high‐density lipoprotein cholesterol levels, and blood pressure and cholesterol medication use. 16 , 18

Medication use was collected at baseline and treated as a fixed covariate in the models. Estimated glomerular filtration rate (eGFR) was calculated using the 2021 Chronic Kidney Disease Epidemiology Collaboration equation. 19 , 20

The genotyping of the UKB is described elsewhere. 16 , 18 Plasma proteomic profiling of 2923 unique proteins was performed on 54 219 participants from the UKB. 21 Proteomics‐based suPAR levels, measured using the Olink immunoassay, 15 , 21 , 22 were used for this study. 3 Serum CRP was measured using a high‐sensitivity immunoturbidimetric assay on a Beckman Coulter AU5800, with a minimum detection limit of 0.08 mg/L. 22

Statistical Analysis

Continuous variables are reported as mean (SD), whereas categorical variables are reported as counts and proportions and were compared using a 2‐sample t‐test for normally distributed continuous variables or a Wilcoxon test for nonnormally distributed continuous variables and a χ2 test for categorical variables.

Both proteomics‐based suPAR and hs‐CRP values underwent rank‐based inverse normal transformation to ensure normality before analysis. The median value was used as the cutoff for proteomics‐based suPAR and hs‐CRP when analyzed as binary variables. The relationships between proteomics‐based suPAR levels and MACEs and cardiovascular mortality were assessed using Fine and Gray subdistribution hazard models to account for noncardiovascular mortality as a competing risk. 23 All‐cause mortality was analyzed using Cox proportional hazards models.

Stepwise adjustments were made, starting from the unadjusted model, to assess the association between proteomics‐based suPAR levels and outcomes. Model 1 adjusted for age, sex, race (White versus other), history of smoking, hypertension, prevalent diabetes, BMI, low‐density lipoprotein cholesterol, triglycerides, high‐density lipoprotein cholesterol, and eGFR. Model 2 adjusted for covariates in model 1 and the use of blood pressure and cholesterol‐lowering medications. Model 3 further adjusted for hs‐CRP levels, in addition to model 2. Model 4 additionally adjusted for the rs4760 variant, while also adjusting for population structure using the top 10 principal components derived from the full genomic data set, in addition to model 3.

We used linear regression to evaluate the multivariate associations between proteomics‐based suPAR levels and the aforementioned demographic, clinical characteristics, and the genetic variant rs4760. We investigated the association of proteomics‐based suPAR and hs‐CRP with MACE risk using restricted cubic splines in fully adjusted Cox models (model 4; the suPAR model excluded hs‐CRP, and the hs‐CRP model excluded suPAR). Proteomics‐based suPAR levels were stratified into quartiles to evaluate their predictive value for the outcomes of interest. Cumulative incidence curves were used to illustrate these associations. Participants were further categorized into 4 groups based on proteomics‐based suPAR levels (greater than or equal to or less than the median) and hs‐CRP levels (greater than or equal to or less than the median). Stratified sensitivity analyses were conducted to assess whether the relationship between proteomics‐based suPAR levels and adverse outcomes varied by sex (male/female), age (≥65 or <65 years), hypertension (yes/no), diabetes (yes/no), eGFR (≥60 or <60 mL/min per 1.73 m2), BMI (≥30 or <30 kg/m2), hs‐CRP (greater than or equal to or less than the median), and genetic variant rs4760 (AG, AA, and GG genotypes), while adjusting for the aforementioned covariates. Formal interaction terms between proteomics‐based suPAR levels and each stratification variable were tested separately in the regression models. No multiple testing correction was applied, as each interaction was evaluated individually.

Additionally, Harrell c‐statistics were calculated to assess the discriminatory power of the risk prediction models before and after the inclusion of proteomics‐based suPAR levels. To further evaluate the incremental predictive utility of proteomics‐based suPAR, we computed the continuous net reclassification index (NRI) and integrated discrimination improvement (IDI). These metrics quantify improvements in risk classification and overall model performance, offering a more nuanced assessment of proteomics‐based suPAR's added value beyond traditional risk factors. IDI was estimated as the improvement in average predicted risk separation between participants who did and did not experience events. In addition, discrimination (ΔC‐statistic) and NRI were reported to provide complementary, clinically interpretable measures of model performance and reclassification. Statistical analyses were performed using R, version 4.0.2 (https://www.R‐project.org/). Statistical significance was defined as a 2‐sided P< 0.05.

RESULTS

Of 33 963 UKB participants enrolled, 45% were men, 93.4% were White race, 2.2% had diabetes, and 25.4% had hypertension (Figure S1). Participants with higher proteomics‐based suPAR levels were older, more likely to be women, and more likely to be White race, with a greater prevalence of hypertension, diabetes, history of smoking, higher BMI and triglyceride levels, lower high‐density lipoprotein cholesterol levels, and a lower eGFR (Table 1). There was no association between the rs4760 allele frequencies and proteomics‐based suPAR levels. After multivariable adjustments, women, older individuals, and those with diabetes, history of smoking, a lower eGFR, a lower high‐density lipoprotein cholesterol or low‐density lipoprotein cholesterol level, or higher triglyceride levels had higher proteomics‐based suPAR levels (Table S1). Each 1‐SD increase in hs‐CRP was associated with a 7.3% higher proteomics‐based suPAR level after adjustment.

Table 1.

Demographic and Clinical Characteristics Stratified by Proteomics‐Based suPAR Levels

Characteristic All cohort High suPAR Low suPAR P value
Total, N (%) 33 963 16 982 (50) 16 981 (50) …
Age, y 56.4±8.2 58.2±8.0 54.6±8.1 <0.001†
Male sex 15 266 (44.9) 6663 (39.2) 8603 (50.7) <0.001†
White race 31 715 (93.4) 16 036 (94.4) 15 679 (92.3) <0.001†
Hypertension 8643 (25.4) 5171 (30.4) 3472 (20.4) <0.001†
Diabetes 742 (2.2) 539 (3.2) 203 (1.2) <0.001†
Smoking history 15 091 (44.4) 8442 (49.7) 6649 (39.2) <0.001†
BMI, kg/m2 27.3±4.72 28±5.2 26.6±4.0 <0.001†
HDL‐C, mg/dL 56.5±14.8 55.4±14.8 57.5±14.7 <0.001†
LDL‐C, mg/dL 138.3±33.5 138±34.5 138.6±32.4 0.10
Triglycerides, mg/dL 152.8±89.3 160.6±92.9 145±84.9 <0.001†
eGFR, mL/min per 1.73 m2 94.6±13.6 91.8±14.9 97.5±11.4 <0.001†
rs4760 genotype frequencies 0.11
GG 799 (2.4) 373 (2.2) 426 (2.5)
AG 8469 (24.9) 4279 (25.2) 4190 (24.7)
AA 24 695 (72.7) 12 330 (72.6) 12 365 (72.8)
Medications
Blood pressure medications 6509 (19.2) 4163 (24.5) 2346 (13.8) <0.001†
Cholesterol medications 5022 (14.8) 3008 (17.7) 2014 (11.9) <0.001†
Biomarkers
Proteomics‐based suPAR levels * −0.013 (−0.21 to 0.2) 0.20 (0.08 to 0.37) −0.21 (−0.35 to −0.11) <0.001†
hs‐CRP, mg/L* 1.31 (0.65 to 2.74) 1.74 (0.85 to 3.63) 1.02 (0.53 to 2.00) <0.001†
Events
All‐cause mortality 3197 (9.4) 2271 (13.4) 926 (5.5) <0.001†
Cardiovascular mortality 882 (2.6) 666 (3.9) 216 (1.3) <0.001†
MACEs 3626 (10.7) 2331 (13.7) 1295 (7.6) <0.001†
Open in a new tab

BMI indicates body mass index; eGFR, estimated glomerular filtration rate; HDL‐C, high‐density lipoprotein cholesterol; hs‐CRP, high‐sensitivity C‐reactive protein; LDL‐C, low‐density lipoprotein cholesterol; MACE, major adverse cardiovascular event; and suPAR, soluble urokinase plasminogen activator receptor.

Data are presented as mean±SD or number (percentage) unless otherwise indicated. For continuous normally distributed variables, a 2‐sample t‐test was used. For continuous nonnormally distributed variables, the Wilcoxon rank‐sum test was used. For categorical variables, the χ2 test was used.

*

Median (interquartile range).

†

Statistically significant results (P < 0.001).

Proteomics‐Based suPAR Levels and Outcomes

During a median follow‐up of 14 years (476 177 person‐years), participants with high proteomics‐based suPAR levels had significantly worse outcomes (P<0.001) compared with those with low levels in unadjusted and adjusted models (Tables 1 and 2; Figures 1 and 2). After adjusting for demographic, clinical, and behavioral risk factors, and medication use, proteomics‐based suPAR levels remained independent predictors of MACEs (hazard ratio [HR], 2.2 [95% CI, 2.0–2.5]; P<0.001), cardiovascular mortality (HR, 4.0 [95% CI, 3.3–4.8]; P<0.001), and all‐cause mortality (HR, 3.8 [95% CI, 3.4–4.2]) (Table 2, model 2; Figure 1).

Table 2.

Relationship Between Proteomics‐Based suPAR Levels and Incident Adverse Outcomes

Outcomes Model HR (95% CI) P value
MACEs Unadjusted 3.2 (2.9–4.5) <0.001*
Model 1 2.2 (2.0–2.4) <0.001*
Model 2 2.2 (2.0–2.5) <0.001*
Model 3 2.1 (1.8–2.3) <0.001*
Model 4 2.0 (1.8–2.3) <0.001*
Cardiovascular mortality Unadjusted 5.9 (5.1–6.9) <0.001*
Model 1 4.0 (3.3–4.9) <0.001*
Model 2 4.0 (3.3–4.8) <0.001*
Model 3 3.6 (3.0–4.4) <0.001*
Model 4 3.6 (3.0–4.4) <0.001*
All‐cause mortality Unadjusted 5.0 (4.6–5.5) <0.001*
Model 1 3.8 (3.4–4.2) <0.001*
Model 2 3.8 (3.4–4.2) <0.001*
Model 3 3.4 (3.0–3.8) <0.001*
Model 4 3.4 (3.0–3.7) <0.001*
Open in a new tab

HR indicates hazard ratio; MACE, major adverse cardiovascular event; and suPAR, soluble urokinase plasminogen activator receptor.

Model 1: Adjusted for age, sex, race (White vs other), history of smoking, hypertension, prevalent diabetes, body mass index, low‐density lipoprotein cholesterol, triglycerides, high‐density lipoprotein cholesterol, and estimated glomerular filtration rate. Model 2: Adjusted for covariates in model 1 and use of blood pressure– and cholesterol‐lowering medications. Model 3: Adjusted for covariates in model 2 and hs‐CRP (high‐sensitivity C‐reactive protein). Model 4: Adjusted for covariates in model 3, the rs4760 variant, and population structure using the top 10 principal components derived from the full genomic data set. Associations between proteomics‐based suPAR and MACEs and cardiovascular mortality were evaluated using Fine‐Gray subdistribution hazard models to account for competing risks; associations with all‐cause mortality were assessed using Cox proportional hazards regression. Exact 2‐sided P values are reported; values <0.001 are shown as <0.001.

*

Statistically significant results (P < 0.001).

Figure 1. Associations between proteomics‐based suPAR and hs‐CRP levels and the risk of MACEs: restricted cubic spline analysis.

Figure 1

Open in a new tab

The association between major adverse cardiovascular events and proteomics‐based suPAR (A) and hs‐CRP levels (B) is shown using restricted cubic spline analysis. Solid lines represent the estimated hazard ratio, and shaded areas indicate 95% CIs. Models were fully adjusted for age, sex, race, history of smoking, hypertension, diabetes, body mass index, low‐density lipoprotein cholesterol, triglycerides, high‐density lipoprotein cholesterol, estimated glomerular filtration rate, use of blood pressure– or cholesterol‐lowering medications, the rs4760 variant, and population structure using the top 10 principal components derived from the full genomic data set. The suPAR model excluded hs‐CRP, and the hs‐CRP model excluded proteomics‐based suPAR. P for nonlinearity <0.001 for proteomics‐based suPAR and P=0.063 for hs‐CRP. hs‐CRP indicates high‐sensitivity C‐reactive protein; MACE, major adverse cardiovascular event; and suPAR, soluble urokinase plasminogen activator receptor.

Figure 2. Cumulative risk for MACEs (A), cardiovascular mortality (B), and all‐cause mortality (C) by proteomics‐based suPAR level quartiles.

Figure 2

Open in a new tab

Cumulative incidence curves are shown for each quartile of proteomics‐based suPAR levels. The red line corresponds to Q4 (highest suPAR), the purple line to Q3, the blue line to Q2, and the green line to Q1 (lowest suPAR). MACE indicates major adverse cardiovascular events; Q, quartile; and suPAR, soluble urokinase plasminogen activator receptor.

Proteomics‐Based suPAR, hs‐CRP Levels, and Outcomes

As expected, higher hs‐CRP levels were associated with higher risk of MACEs, cardiovascular mortality, and all‐cause mortality, even after adjustment for demographic and clinical risk factors. After further adjustment for proteomics‐based suPAR levels, the association was attenuated but was still significant for all outcomes (Table S2; Figure 1).

Participants with elevated levels of both proteomics‐based suPAR and hs‐CRP (greater than or equal to the median) levels had the highest risk of MACEs, with a 1.7‐fold increase in risk (P<0.001) compared with the reference group with low levels of both markers (less than the median) (Table S3; Figure 3). Notably, individuals with high proteomics‐based suPAR levels (greater than or equal to the median) but low hs‐CRP (less than the median) exhibited a 1.4‐fold increase in MACE risk (P<0.001) relative to the reference group. Similar associations were observed for cardiovascular and all‐cause mortality (Table S3).

Figure 3. Cumulative risk for MACEs (A), cardiovascular mortality (B), and all‐cause mortality (C) by proteomics‐based suPAR level status (greater than or less than the median) and hs‐CRP status (greater than or less than the median).

Figure 3

Open in a new tab

Cumulative incidence curves are shown for 4 groups defined by proteomics‐based suPAR and hs‐CRP levels relative to their cohort‐specific medians. The red line represents participants with elevated suPAR and elevated hs‐CRP (both greater than or equal to the median); the purple line represents those with elevated suPAR but low hs‐CRP (suPAR greater than or equal to the median, hs‐CRP less than the median); the blue line represents those with low suPAR but elevated hs‐CRP (suPAR less than the median, hs‐CRP greater than or equal to the median); and the green line represents participants with low levels of both biomarkers (both less than the median). CRP indicates C‐reactive protein; hs‐CRP, high‐sensitivity CRP; MACE, major adverse cardiovascular event; and suPAR, soluble urokinase plasminogen activator receptor.

rs4760 Variant, Proteomics‐Based suPAR, and Outcomes

There was no association between rs4760 genotypes and proteomics‐based suPAR levels (Table 1; Table S1). After adjusting for rs4760 genotypes (AG, AA, and GG) and accounting for population structure using the top 10 principal components, proteomics‐based suPAR levels remained significantly associated with all clinical outcomes (Table 2, model 4).

Sensitivity Analysis

Sensitivity analyses assessed the heterogeneity in the predictive value of proteomics‐based suPAR levels for adverse events (Figure S2). For MACEs, statistically significant interactions were observed for sex (P=0.04), age (P=0.01), and BMI (P=0.02). The association between proteomics‐based suPAR levels and MACEs was stronger in women (HR, 2.38 [95% CI, 2.00–2.82]) compared with men (HR, 2.09 [95% CI, 1.84–2.38]), and participants aged <65 years had a higher hazard (HR, 2.27 [95% CI, 2.01–2.57]) compared with those aged ≥65 years (HR, 1.94 [95% CI, 1.61–2.33]). Similarly, the association was more pronounced in individuals with BMI <30 kg/m2 (HR, 2.30 [95% CI, 2.03–2.60]) than in those with BMI ≥30 kg/m2 (HR, 1.93 [95% CI, 1.60–2.34]) (Figure S2).

Discrimination and Reclassification Analyses

We assessed the incremental predictive value of proteomics‐based suPAR levels for MACEs in comparison with standard demographic, clinical, and behavioral risk factors using the Harrell C‐statistic, NRI, and IDI (Table 3). Compared with the baseline clinical model (model 1) that included demographic and cardiometabolic risk factors, the inclusion of proteomics‐based suPAR (model 2 versus model 1) significantly increased the C‐statistic by 0.012 (95% CI, 0.008–0.015; P<0.001), NRI was 0.518 (95% CI, 0.506–0.529; P<0.001), and IDI was 0.020 (95% CI, 0.016–0.024; P<0.001), indicating meaningful enhancement in both discrimination and reclassification.

Table 3.

Comparative Predictive Performance of Models With and Without Proteomics‐Based suPAR Levels Using C‐Statistic, NRI, and IDI

Comparison C‐statistics (C1, C2) ΔC (95% CI) P value NRI (95% CI) P value IDI (95% CI) P value
Model 1 vs model 2 0.719 vs 0.731 0.012 (0.008–0.015) <0.001 0.518 (0.506–0.529) <0.001 0.020 (0.016–0.024) <0.001
Model 3 vs model 2 0.723 vs 0.731 0.008 (0.004–0.011) <0.001 0.525 (0.509–0.534) <0.001 0.028 (0.022–0.033) <0.001
Open in a new tab

Model 1 includes demographic and clinical variables: age at enrollment, male sex, White race, smoking history, low‐density lipoprotein cholesterol, triglycerides, high‐density lipoprotein cholesterol, body mass index, prevalent diabetes, prevalent hypertension, estimated glomerular filtration rate, and use of cholesterol‐ and blood pressure–lowering medications. Model 2 includes all model 1 variables plus proteomics‐based suPAR. Model 3 includes all model 1 variables plus hs‐CRP. C‐statistic reflects the discriminatory ability of each model; ΔC represents the change in C‐statistic between compared models. NRI and IDI quantify improvements in risk classification and overall model discrimination, respectively, with the addition of the novel biomarker. Exact 2‐sided P values are reported; values <0.001 are shown as <0.001. ΔC indicates change in C‐statistic; IDI, integrated discrimination improvement; NRI, net reclassification index; and suPAR, soluble urokinase plasminogen activator receptor.

Importantly, proteomics‐based suPAR also demonstrated added predictive value beyond hs‐CRP levels. When model 2 (including proteomics‐based suPAR) was compared with model 3 (including hs‐CRP), there was a significant improvement in the C‐statistic: ΔC=0.008 (95% CI, 0.004–0.011; P<0.001), NRI=0.525 (95% CI, 0.509–0.534; P<0.001), and IDI=0.028 (95% CI, 0.022–0.033; P<0.001) (Table 3).

DISCUSSION

Novel findings of this study are that proteomics‐based suPAR level is a predictor of adverse cardiovascular events in individuals without known CVD, where it importantly reclassifies risk compared with conventional risk measures as well as hs‐CRP levels. 24 These findings are notable particularly in light of the findings using Mendelian randomization studies suggesting that suPAR may contribute to atherosclerotic CVD and not merely be a biomarker of chronic inflammation and immune dysregulation.

suPAR has emerged as a robust measure of both chronic inflammation and immune system activation. 3 , 25 , 26 It is secreted by various inflammatory and immune cells, including monocytes, endothelial cells, and fibroblasts, in response to immune triggers. 3 , 25 , 26 This is particularly significant in chronic diseases in comparison to acute‐phase reactants, such as hs‐CRP, that are primarily influenced by acute inflammation. 3 , 24 These differences may at least partly be attributed to the longer half‐life and stability suPAR, making it a more consistent indicator of sustained inflammation and immune activation. 3 , 26 , 27 Notably, suPAR has also been linked to the endothelial dysfunction 28 , 29 that precedes plaque formation and macrovascular complications. 30 This may explain why proteomics‐based suPAR levels are particularly valuable in the general population, as they can help identify at‐risk individuals at an early stage and its use as a screening tool 15 for earlier detection of cardiovascular risk could result in earlier implementation of preventive strategies.

suPAR levels measured using ELISA immunoassays predict adverse cardiovascular events in both patients with coronary artery disease 5 , 31 , 32 and the general population, 6 , 7 and to mediate the relationship between diabetes and cardiovascular outcomes. 33 In parallel, studies using proteomics‐based approaches have reported associations between proteomics‐based suPAR and heart failure, 34 dilated cardiomyopathy, 35 as well as depression and anxiety‐related disorders. 36 , 37

Importantly, although prior studies have demonstrated a modest 15 correlation between proteomics‐based suPAR and full‐length suPAR measured by ELISA, 15 these observations collectively highlight the biological relevance of proteomics‐based suPAR in systemic immune activation, vascular inflammation, and broader cardiometabolic and neuropsychiatric pathways, supporting its potential utility as a versatile biomarker across diverse disease processes. Building on this evidence, the present study uniquely evaluates the relationship between proteomics‐based suPAR, including alternative circulating forms, and adverse cardiovascular events in a large population‐based cohort.

Elevated levels of both hs‐CRP and proteomics‐based suPAR remained independent predictors of adverse CVD outcomes after full adjustment. Thus, individuals with elevated levels of both hs‐CRP and proteomics‐based suPAR had the worst outcomes compared with those without elevation of either biomarker level, consistent with previous findings with suPAR in high‐risk populations. 6 Those with elevated levels of either proteomics‐based suPAR or hs‐CRP alone had an intermediate level of added risk. Importantly, adjusting for hs‐CRP levels did not significantly impact the relationship between proteomics‐based suPAR levels and adverse outcomes, highlighting that the risk associated with proteomics‐based suPAR is not fully explained by hs‐CRP levels as a marker of systemic inflammation.

We observed significant interactions between certain demographic parameters, including sex, age, and BMI, and the prognostic impact of proteomics‐based suPAR levels; however, proteomics‐based suPAR levels were predictive in all individuals. Previous studies have established that suPAR levels tend to be higher in women, in older individuals, and in the presence of obesity and renal disease. 3 , 8 , 9 , 10 , 38

We also examined whether the rs4760 variant, which has been previously reported to be associated with a 0.11‐SD increase in suPAR levels, 12 would impact the association between proteomics‐based suPAR levels and outcomes. We found no relationship between proteomics‐based suPAR levels and the rs4760 polymorphisms and no impact of the polymorphism on events. This could be because ELISA‐based and proteomics‐based suPAR levels are only modestly correlated, 15 and several other polymorphisms in the proteomics‐based suPAR gene also influence its levels. 12 These findings underscore the need for further studies directly comparing ELISA‐ and Olink‐based suPAR measurements to determine whether the lack of association with rs4760 reflects a true predominance of environmental influences and independence from genetic regulation, or instead results from technical differences in the proteomics platform, particularly in light of our observation that higher proteomics‐based suPAR levels are associated with smoking and diabetes.

Our findings potentially have significant implications for clinical practice. Integrating proteomics‐based suPAR measurements into routine cardiovascular risk assessments may aid in risk discrimination, as shown by the significant improvement in reclassification and discrimination indexes. Furthermore, monitoring proteomics‐based suPAR levels could be valuable for tracking disease progression and assessing therapeutic responses to interventions. Clinical trials are investigating novel suPAR antagonists that may offer targeted strategies to modulate suPAR signaling and improve outcomes in inflammatory and CVDs.

Strengths and Limitations

A key strength of this study is the large population studied that enhances generalizability and robustness of our findings. Inclusion of women allowed understanding of sex‐specific effects and the extended follow‐up offer valuable insights into long‐term cardiovascular outcomes. Limitations include the fact that proteomics‐ and not ELISA‐based suPAR levels were available in the UKB data, but despite the modest correlation between the 2, our findings demonstrate significant predictive impact of proteomics‐based suPAR of adverse outcomes. 15 Second, the UKB is generally healthier and more affluent than the general population and is predominantly individuals of European ancestry, limiting the applicability to more ethnically diverse populations. Furthermore, biomarker measurements, including proteomics‐based suPAR and hs‐CRP, were obtained at a single time point, which does not account for intraindividual variability over time. Residual confounding is also possible, as observational data, even when adjusted for multiple covariates, cannot fully eliminate unmeasured or unknown confounders. Furthermore, the observational nature of the study precludes causal inference, and prospective validation in independent, diverse cohorts will be essential to confirm the predictive utility of proteomics‐based suPAR in clinical setting. Finally, reclassification‐based metrics, including net reclassification improvement (NRI) and integrated discrimination improvement (IDI), are sensitive to model specification and may overestimate the incremental predictive value of novel biomarkers. Therefore, these measures should be interpreted cautiously and in conjunction with established discrimination metrics, such as the C‐statistic. 39 , 40

CONCLUSIONS

Circulating levels of proteomics‐based suPAR predict future adverse cardiovascular outcomes in the general population, regardless of traditional risk factors, traditional inflammatory markers (such as hs‐CRP), and PLAUR gene expression profile. Whether earlier therapeutic intervention will help reduce risk needs further study.

Sources of Funding

This work was supported partially by Abbott Labs. Arshed A. Quyyumi, MD, is supported by National Institutes of Health through the following grants: 1P01HL154996‐04, 5R01HL157311‐04, 1R01HL166004‐03, R01 AG089759‐01, 3P30DK111024‐07S2, and 5T32 HL130025.

Disclosures

None.

Supporting information

Tables S1–S3

Figures S1–S2

JAH3-15-e047057-s001.pdf (620.7KB, pdf)

STROBE Statement

JAH3-15-e047057-s002.pdf (162.5KB, pdf)

This manuscript was sent to June‐Wha Rhee, MD, Associate Editor, for review by expert referees, editorial decision, and final disposition.

For Sources of Funding and Disclosures, see page 10.

References

  • 1. Hansson GK. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med. 2005;352:1685–1695. doi: 10.1056/NEJMra043430 [DOI] [PubMed] [Google Scholar]
  • 2. Taleb S. Inflammation in atherosclerosis. Arch Cardiovasc Dis. 2016;109:708–715. doi: 10.1016/j.acvd.2016.04.002 [DOI] [PubMed] [Google Scholar]
  • 3. Rasmussen LJH, Petersen JEV, Eugen‐Olsen J. Soluble urokinase plasminogen activator receptor (suPAR) as a biomarker of systemic chronic inflammation. Front Immunol. 2021;12:780641. doi: 10.3389/fimmu.2021.780641 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Fuhrman B. The urokinase system in the pathogenesis of atherosclerosis. Atherosclerosis. 2012;222:8–14. doi: 10.1016/j.atherosclerosis.2011.10.044 [DOI] [PubMed] [Google Scholar]
  • 5. Eapen DJ, Manocha P, Ghasemzadeh N, Patel RS, al Kassem H, Hammadah M, Veledar E, le NA, Pielak T, Thorball CW, et al. Soluble urokinase plasminogen activator receptor level is an independent predictor of the presence and severity of coronary artery disease and of future adverse events. J Am Heart Assoc. 2014;3:e001118. doi: 10.1161/JAHA.114.001118 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Lyngbæk S, Marott JL, Sehestedt T, Hansen TW, Olsen MH, Andersen O, Linneberg A, Haugaard SB, Eugen‐Olsen J, Hansen PR, et al. Cardiovascular risk prediction in the general population with use of suPAR, CRP, and Framingham risk score. Int J Cardiol. 2013;167:2904–2911. doi: 10.1016/j.ijcard.2012.07.018 [DOI] [PubMed] [Google Scholar]
  • 7. Eugen‐Olsen J, Andersen O, Linneberg A, Ladelund S, Hansen TW, Langkilde A, Petersen J, Pielak T, Møller LN, Jeppesen J, et al. Circulating soluble urokinase plasminogen activator receptor predicts cancer, cardiovascular disease, diabetes and mortality in the general population. J Intern Med. 2010;268:296–308. doi: 10.1111/j.1365-2796.2010.02252.x [DOI] [PubMed] [Google Scholar]
  • 8. Hayek SS, Ko YA, Awad M, Ahmed H, Gray B, Hosny KM, Aida H, Tracy MJ, Wei C, Sever S, et al. Cardiovascular disease biomarkers and suPAR in predicting decline in renal function: a prospective cohort study. Kidney Int Rep. 2017;2:425–432. doi: 10.1016/j.ekir.2017.02.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Hayek SS, Koh KH, Grams ME, Wei C, Ko YA, Li J, Samelko B, Lee H, Dande RR, Lee HW, et al. A tripartite complex of suPAR, APOL1 risk variants and α(v)β(3) integrin on podocytes mediates chronic kidney disease. Nat Med. 2017;23:945–953. doi: 10.1038/nm.4362 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Hayek SS, Sever S, Ko YA, Trachtman H, Awad M, Wadhwani S, Altintas MM, Wei C, Hotton AL, French AL, et al. Soluble urokinase receptor and chronic kidney disease. N Engl J Med. 2015;373:1916–1925. doi: 10.1056/NEJMoa1506362 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Goodchild TT, Li Z, Lefer DJ. Soluble urokinase plasminogen activator receptor: from biomarker to active participant in atherosclerosis and cardiovascular disease. J Clin Invest. 2022;132:e165868. doi: 10.1172/JCI165868 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Hindy G, Tyrrell DJ, Vasbinder A, Wei C, Presswalla F, Wang H, Blakely P, Ozel AB, Graham S, Holton GH, et al. Increased soluble urokinase plasminogen activator levels modulate monocyte function to promote atherosclerosis. J Clin Invest. 2022;132:e158788. doi: 10.1172/JCI158788 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Diederichsen MZ, Diederichsen SZ, Mickley H, Steffensen FH, Lambrechtsen J, Sand NPR, Christensen KL, Olsen MH, Diederichsen A, Grønhøj MH. Prognostic value of suPAR and hs‐CRP on cardiovascular disease. Atherosclerosis. 2018;271:245–251. doi: 10.1016/j.atherosclerosis.2018.01.029 [DOI] [PubMed] [Google Scholar]
  • 14. Wohlwend NF, Grossmann K, Aeschbacher S, Weideli OC, Telser J, Risch M, Conen D, Risch L. The association of suPAR with cardiovascular risk factors in young and healthy adults. Diagnostics (Basel). 2023;13:2938. doi: 10.3390/diagnostics13182938 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Vasbinder A, Raffield LM, Gao Y, Engstrom G, Quyyumi AA, Reiner AP, Reiser J, Hayek SS. Assay‐related differences in SuPAR levels: implications for measurement and data interpretation. J Nephrol. 2023;36:157–159. doi: 10.1007/s40620-022-01344-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, Downey P, Elliott P, Green J, Landray M, et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 2015;12:e1001779. doi: 10.1371/journal.pmed.1001779 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. World Health Organization . International Statistical Classification of Diseases and related health problems (ICD). World Health Organization; 1990. Accessed February 17, 2026. https://www.who.int/standards/classifications/classification‐of‐diseases [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp K, Motyer A, Vukcevic D, Delaneau O, O'Connell J, et al. The UK biobank resource with deep phenotyping and genomic data. Nature. 2018;562:203–209. doi: 10.1038/s41586-018-0579-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. National Kidney Foundation . CKD‐EPI creatinine equation. National Kidney Foundation; 2021. Accessed February 17, 2026. https://www.kidney.org/ckd‐epi‐creatinine‐equation‐2021‐0 [Google Scholar]
  • 20. Inker LA, Eneanya ND, Coresh J, Tighiouart H, Wang D, Sang Y, Crews DC, Doria A, Estrella MM, Froissart M, et al. New creatinine‐ and cystatin C‐based equations to estimate GFR without race. N Engl J Med. 2021;385:1737–1749. doi: 10.1056/NEJMoa2102953 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Sun BB, Chiou J, Traylor M, Benner C, Hsu YH, Richardson TG, Surendran P, Mahajan A, Robins C, Vasquez‐Grinnell SG, et al. Plasma proteomic associations with genetics and health in the UK biobank. Nature. 2023;622:329–338. doi: 10.1038/s41586-023-06592-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. UK Biobank . UK Biobank Biomarker Project: companion document to accompany serum biomarker data. UK Biobank; 2019. Accessed February 17, 2026. https://biobank.ndph.ox.ac.uk/showcase/showcase/docs/serum_biochemistry.pdf [Google Scholar]
  • 23. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496–509. doi: 10.1080/01621459.1999.10474144 [DOI] [Google Scholar]
  • 24. Plebani M. Why C‐reactive protein is one of the most requested tests in clinical laboratories? Clin Chem Lab Med. 2023;61:1540–1545. doi: 10.1515/cclm-2023-0086 [DOI] [PubMed] [Google Scholar]
  • 25. Desmedt S, Desmedt V, Delanghe JR, Speeckaert R, Speeckaert MM. The intriguing role of soluble urokinase receptor in inflammatory diseases. Crit Rev Clin Lab Sci. 2017;54:117–133. doi: 10.1080/10408363.2016.1269310 [DOI] [PubMed] [Google Scholar]
  • 26. Thunø M, Macho B, Eugen‐Olsen J. suPAR: the molecular crystal ball. Dis Markers. 2009;27:157–172. doi: 10.1155/2009/504294 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Chew‐Harris J, Appleby S, Richards AM, Troughton RW, Pemberton CJ. Analytical, biochemical and clearance considerations of soluble urokinase plasminogen activator receptor (suPAR) in healthy individuals. Clin Biochem. 2019;69:36–44. doi: 10.1016/j.clinbiochem.2019.05.010 [DOI] [PubMed] [Google Scholar]
  • 28. Mekonnen G, Corban MT, Hung OY, Eshtehardi P, Eapen DJ, al‐Kassem H, Rasoul‐Arzrumly E, Gogas BD, McDaniel MC, Pielak T, et al. Plasma soluble urokinase‐type plasminogen activator receptor level is independently associated with coronary microvascular function in patients with non‐obstructive coronary artery disease. Atherosclerosis. 2015;239:55–60. doi: 10.1016/j.atherosclerosis.2014.12.025 [DOI] [PubMed] [Google Scholar]
  • 29. Corban MT, Prasad A, Nesbitt L, Loeffler D, Herrmann J, Lerman LO, Lerman A. Local production of soluble urokinase plasminogen activator receptor and plasminogen activator inhibitor‐1 in the coronary circulation is associated with coronary endothelial dysfunction in humans. J Am Heart Assoc. 2018;7:e009881. doi: 10.1161/JAHA.118.009881 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Davignon J, Ganz P. Role of endothelial dysfunction in atherosclerosis. Circulation. 2004;109:III27–III32. doi: 10.1161/01.CIR.0000131515.03336.f8 [DOI] [PubMed] [Google Scholar]
  • 31. Lyngbæk S, Marott JL, Møller DV, Christiansen M, Iversen KK, Clemmensen PM, Eugen‐Olsen J, Jeppesen JL, Hansen PR. Usefulness of soluble urokinase plasminogen activator receptor to predict repeat myocardial infarction and mortality in patients with ST‐segment elevation myocardial infarction undergoing primary percutaneous intervention. Am J Cardiol. 2012;110:1756–1763. doi: 10.1016/j.amjcard.2012.08.008 [DOI] [PubMed] [Google Scholar]
  • 32. Nikorowitsch J, Borchardt T, Appelbaum S, Ojeda F, Lackner KJ, Schnabel RB, Blankenberg S, Zeller T, Karakas M. Cardio‐renal biomarker soluble urokinase‐type plasminogen activator receptor is associated with cardiovascular death and myocardial infarction in patients with coronary artery disease independent of troponin, C‐reactive protein, and renal function. J Am Heart Assoc. 2020;9:e015452. doi: 10.1161/JAHA.119.015452 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Sakr SM, Desai S, Medina‐Inojosa J, Liu C, Razavi AC, Yadalam AK, Ko YA, Ahmed T, Harris KA, Huang J, et al. Soluble urokinase plasminogen activator receptor (suPAR) mediates the impact of diabetes on adverse outcomes in coronary artery disease. Diabetes Care. 2026;49:450–459. doi: 10.2337/dc25-0940 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Yadalam AK, Gold ME, Patel KJ, Liu C, Razavi AC, Jain V, Vatsa N, Gold D, Owais M, Haroun N, et al. Proteomics‐based soluble urokinase plasminogen activator receptor levels are associated with incident heart failure risk. JACC Adv. 2025;4:101442. doi: 10.1016/j.jacadv.2024.101442 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Xu S, Zhang G, Tan X, Zeng Y, Jiang H, Jiang Y, Wang X, Song Y, Fan H, Zhou Y. Plasma Olink proteomics reveals novel biomarkers for prediction and diagnosis in dilated cardiomyopathy with heart failure. J Proteome Res. 2024;23:4139–4150. doi: 10.1021/acs.jproteome.4c00522 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Kang J, Yang L, Jia T, Zhang W, Wang LB, Zhao YJ, You J, Deng YT, Ge YJ, Liu WS, et al. Plasma proteomics identifies proteins and pathways associated with incident depression in 46,165 adults. Sci Bull (Beijing). 2025;70:573–586. doi: 10.1016/j.scib.2024.09.041 [DOI] [PubMed] [Google Scholar]
  • 37. Ding C, Qi W, Tu H, Wang Y, Zhang T, Sun H. Proteomic profiling of human plasma for anxiety and depression: discovery of potential biomarkers and mechanistic insights. J Affect Disord. 2025;391:120067. doi: 10.1016/j.jad.2025.120067 [DOI] [PubMed] [Google Scholar]
  • 38. Mehta A, Desai SR, Ko YA, Liu C, Dhindsa DS, Nayak A, Hooda A, Martini MA, Ejaz K, Sperling LS, et al. Sex differences in circulating soluble urokinase‐type plasminogen activator receptor (suPAR) levels and adverse outcomes in coronary artery disease. J Am Heart Assoc. 2020;9:e015457. doi: 10.1161/JAHA.119.015457 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Pepe MS, Kerr KF, Longton G, Wang Z. Testing for improvement in prediction model performance. Stat Med. 2013;32:1467–1482. doi: 10.1002/sim.5727 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Xanthakis V, Sullivan LM, Vasan RS, Benjamin EJ, Massaro JM, D'Agostino RB Sr, Pencina MJ. Assessing the incremental predictive performance of novel biomarkers over standard predictors. Stat Med. 2014;33:2577–2584. doi: 10.1002/sim.6165 [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.

Supplementary Materials

Tables S1–S3

Figures S1–S2

JAH3-15-e047057-s001.pdf (620.7KB, pdf)

STROBE Statement

JAH3-15-e047057-s002.pdf (162.5KB, pdf)

Articles from Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease are provided here courtesy of Wiley

RESOURCES