Recent studies have suggested that soluble urokinase plasminogen activator receptor (suPAR), a biomarker of subclinical levels of inflammation, is significantly correlated with cardiovascular events.
We investigated the association between suPAR and left ventricular ejection fraction (LVEF), left ventricular mass index (LVMI), and plasma B-type natriuretic peptide (BNP) among cardiac inpatients.
Methods and Results
In total, 242 patients (mean age 71.3 ± 9.8 years; 70 women) admitted to the cardiology department were enrolled in the study. suPAR was significantly correlated with LVEF (R = -0.24, P<0.001), LVMI (R = 0.16, P = 0.014) and BNP (R = 0.46, P<0.001). In logistic regression analysis, the highest suPAR tertile (> 3236 pg/mL) was associated with low LVEF (< 50%) and elevated BNP (> 300 pg/mL) with an odds ratio of 3.84 (95% confidence interval [CI], 1.22–12.1) and 5.36 (95% CI, 1.32–21.8), respectively, after adjusting for age, sex, log-transformed estimated glomerular filtration rate (log(eGFR)), C-reactive protein, and diuretic use. The association between suPAR and LVMI was not statistically significant. In multivariate receiver operating characteristic analysis, addition of log(suPAR) to the combination of age, sex, log(eGFR) and CRP incrementally improved the prediction of low LVEF (area under the curve [AUC], 0.827 to 0.852, P = 0.046) and BNP ≥ 300 pg/mL (AUC, 0.869 to 0.906; P = 0.029).
Citation: Fujita S-i, Tanaka S, Maeda D, Morita H, Fujisaka T, Takeda Y, et al. (2017) Serum Soluble Urokinase-Type Plasminogen Activator Receptor Is Associated with Low Left Ventricular Ejection Fraction and Elevated Plasma Brain-Type Natriuretic Peptide Level. PLoS ONE 12(1): e0170546. https://doi.org/10.1371/journal.pone.0170546
Editor: Vincenzo Lionetti, Scuola Superiore Sant'Anna, ITALY
Received: April 11, 2016; Accepted: January 6, 2017; Published: January 30, 2017
Copyright: © 2017 Fujita et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data are ethically restricted and cannot be shared publicly. Data are available from the corresponding author by request, and subject to ethical considerations. Future interested researchers may request data access to Ethics Committee in Osaka Medical College. The URL of the Ethics Committee at Osaka Medical College is http://office.osaka-med.ac.jp/rco/rinric/. The Ethics Committee can be contacted at: TEL: +81-72-683-1221 (extension 2815) Email: firstname.lastname@example.org
Funding: This work was supported in part by Grants in Aid for Scientific Research from the Ministry of Education, Science, and Culture of Japan (No. 15K09106).
Competing interests: The authors have declared that no competing interests exist.
The receptor for urokinase-type plasminogen activator (uPAR), a membrane-linked protein, may mediate immune and inflammatory activation and cancer cell progression [1,2,3,4]. Soluble uPAR (suPAR), which is formed by the cleavage and release of uPAR, has been gathering increasing attention owing to its potential as a biomarker for the presence or progression of various diseased conditions. For example, elevated suPAR levels have been shown to be associated with chronic kidney disease (CKD) and cardiovascular abnormalities, including coronary artery disease, early cardiac systolic and diastolic myocardial impairment, heart failure, and incident cardiovascular events [5,6,7,8,9,10,11,12,13]. Recent cohort studies showed that elevated suPAR levels were independently associated with incident chronic kidney disease, a decline in the renal function  and hospitalization due to impaired kidney function .
Despite the observed association between suPAR and several aspects of cardiovascular diseases, it remains unclear whether suPAR plays a causal role in the disease process, whether suPAR levels increase as a resultant of the disease process, or whether suPAR is a mere bystander .
Left ventricular systolic dysfunction and hypertrophy are presumed to have an association with low-grade inflammation [17,18,19]; however, only a few studies have investigated the possible association between cardiac function and left ventricular hypertrophy (LVH) and suPAR. In the current study, we retrospectively examined whether serum suPAR is associated with left ventricular ejection fraction (LVEF) and left ventricular mass (LVM) among cardiac inpatients who were not undergoing chronic hemodialysis.
The current retrospective study was approved by the Ethics Committee at the Osaka Medical College and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all patients or their guardians.
Between April 2014 and February 2015, 1289 patients were admitted to the cardiology department; among them, suPAR was measured in 286 consecutive patients after obtaining written informed consent. Of 286 patients, 33 for whom echocardiographic data were not sufficient for the current study, were excluded from the study population. A further 11 patients whose B-type natriuretic peptide (BNP) levels were not available were also excluded. Thus, 242 patients were enrolled as the study population, which included 6 patients who were undergoing chronic hemodialysis (Fig 1).
Blood samples were collected in the morning after an overnight fast. Aliquots of serum and plasma were immediately obtained and stored at -80 degrees until analysis. Serum levels of suPAR were measured by a kit (R&D Systems, Minneapolis, MN) according to the manufacturer’s instructions. High-sensitivity C-reactive protein (CRP) and BNP levels were measured by routine laboratory methods. The estimated glomerular filtration rate (eGFR) was calculated by the following Modification of Diet in Renal Disease equation for Japanese subjects: eGFR = 194 × (serum creatinine)-1.094 × (age)-0.287 (× 0.739, when female) .
Echocardiographic examinations were performed with a Vivid 7 Dimension equipped with a multi-frequency transducer (GE Healthcare, Vingmed, Norway). Left ventricular (LV) end-diastolic dimension (LVDd), interventricular septal thickness (IVST) and posterior wall thickness (PWT) were measured at end diastole. LV volumes were calculated by the modified Simpson method using the apical 4-chamber view. The LVEF was defined as low when < 50%. LVM was calculated by the formula proposed by Devereux et al.  with the following modification: 0.8 x 1.04 x [(LVDd + IVST + PWT)3—LVDd3] + 0.6 . Body surface area (BSA) was calculated by using the following formula: (body weight)0.425 × (height)0.725 × 0.007184, and the LVM index (LVMI) was calculated as the ratio of LVM to BSA. When the LVMI was greater than 118 g/m2 (men) or 108 g/m2 (women), LV hypertrophy was defined as present .
Baseline characteristics were assessed with standard descriptive statistics. Data were expressed either as mean ± standard deviation, number (percentage) or median and interquartile range (IQR). Spearman rank correlation test was used to assess the correlation between two variables. Multivariate logistic regression analysis was performed by SPSS statistics version 21.0 (IBM, Armonk, NY). Multivariate receiver operating characteristic (ROC) analysis was performed by STATA 12 (StataCorp LP, College Station, TX). For the multivariate analyses, only those who were not undergoing chronic hemodialysis were included, because eGFR was used as a covariate for the analysis.
The demographic data, laboratory values, and echocardiographic parameters of the study subjects are summarized by suPAR tertile in Tables 1 and 2. Those with higher suPAR were older, but gender did not significantly differ significantly across the tertiles (Table 1). Moderate or severe heart failure (New York Heart Association functional class III or IV) was more than 10 times more prevalent among subjects in the highest suPAR tertile than among those in the lowest tertile. Patients with higher suPAR were more likely to be taking loop diuretics and thiazides.
Patients with higher suPAR had greater CRP values and lower eGFR values (Table 2). Fifty-four patients (22%) had an eGFR of ≥ 60 mL/min/1.73m2. By Spearman correlation analysis, suPAR was significantly correlated with eGFR, age, hemoglobin serum albumin, BNP, and CRP with a coefficient of -0.495, 0.30, -0.45, -0.47, 0.46, and 0.40, respectively (all P < 0.001). In addition, suPAR also showed a significant correlation with LVEF (R = -0.24, P < 0.001), and with LVMI (R = 0.16, P = 0.014).
Relationship between admission diagnosis and suPAR levels
Next, we examined whether certain cardiovascular condition on admission affected suPAR levels. Prevalence of neither acute myocardial infarction nor unstable angina pectoris did not significantly differ across the suPAR tertile. On the other hand, prevalence of worsening heart failure was significantly greater among the higher suPAR tertile. Prevalence of other admission diagnosis, including arrhythmic diseases, follow-up coronary angiography, stable angina pectoris, pre-operative cardiovascular screening for cardiovascular or non-cardiovascular surgery, aortic dissection, arteriosclerosis obliterans, or silent myocardial ischemia did not significantly differ across the suPAR tertile group (Table 3).
Multivariate logistic regression analysis
By univariate logistic regression analysis, log(suPAR) was significantly associated with low LVEF (< 30%) and elevated BNP (≥ 300 pg/mL), but not with LVH (Table 4). The associations between suPAR and low LVEF and elevated BNP remained statistically significant after adjusting for sex, age, log(eGFR) (model 2), CRP (model 3), and diuretic use (model 4). When subjects with an eGFR of ≥ 60 mL/min/m2 (n = 54) and those with an eGFR of < 60 mL/min/m2 (n = 182) who were not undergoing hemodialysis were analyzed separately in model 3, the odds ratio of the highest suPAR tertile for low LVEF and elevated BNP was 6.95 (95% CI 1.91–25.35, P = 0.003) and 5.84 (95% CI 1.34–25.40, P = 0.019), respectively (eGFR of ≥ 60 mL/min/m2), and 7.38 (95% CI 0.76–71.45) and 23.20 (95% CI 0.98–551), respectively (eGFR of < 60 mL/min/m2).
When we excluded the patients who were admitted due to the worsening heart failure or undergoing hemodialysis from the statistical analysis, it was found that, in model 4, middle and the highest suPAR was associated with low LVEF with an odds ratio of 1.64 (95% CI 0.44–6.14, P = 0.464) and 5.20 (95% CI 1.25–21.66 P = 0.023). In this analysis, 188 patients were analyzed. On the other hand, in this same model 4, neither the middle (odds ratio, 0.63; 95% CI 0.05–8.20, P = 0.721) or the highest (odds ratio, 1.80; 95% CI 0.20–15.81, P = 0.598) suPAR was not significantly associated with elevated BNP.
When we analyzed subjects with ischemic heart disease (n = 168) and those without (n = 188) separately in model 3, the odds ratio of the highest suPAR tertile for low LVEF and BNP ≥ 300 pg/mL was 15.5 (95% CI 3.33–72.2, P<0.001) and 9.66 (95% CI 0.94–99.5, P = 0.057), respectively (ischemic heart disease), and 1.77 (95% CI 0.30–10.30, P = 0.53) and 13.45 (95% CI 2.05–88.23, P = 0.007), respectively (no ischemic heart disease).
Multivariate ROC analysis
In multivariate ROC analysis, the area under the curve (AUC) to predict low LVEF, for the combination of age, sex, log(eGFR), CRP, and diuretic use was 0.827 (standard error [SE], 0.033), and further addition of log(suPAR) incrementally increased the prediction (AUC, 0.852; SE, 0.029, P = 0.046, Fig 2). The AUC to predict BNP ≥ 300 pg/mL, for the combination of age, sex, log(eGFR), CRP, and diuretic use was 0.869 (SE, 0.035), and further addition of log(suPAR) incrementally increased the prediction (AUC, 0.906; SE, 0.026, P = 0.029).
The purple line shows the ROC curve to predict low LVEF, for the combination of age, sex, log(eGFR), CRP, and diuretic use (model 1). The green line shows the ROC curve to predict low LVEF for model 1 plus log-transformed soluble urokinase-type plasminogen activator receptor (suPAR) (model 2). The area under the ROC curve was significantly greater in model 2 than in model 1 (0.827 versus 0.852, P = 0.046). In this analysis, only data from patients who were not undergoing chronic hemodialysis were included.
We herein demonstrated that suPAR was associated positively with LVEF and negatively with plasma BNP levels among cardiac patients. These associations were found to be independent of eGFR, CRP, and diuretic use. On the other hand, the association between suPAR and LVH was not significant after adjusting for various confounders.
Several previous studies have reported a relationship between suPAR and N-terminal prohormone BNP (NT-proBNP). For example, Kruger et al. reported that NT-proBNP was significantly associated with suPAR in black African subjects, but not in Caucasian subjects . By analyzing data from the Malmö Diet and Cancer Study, a prospective cohort study conducted Malmö, Sweden, Borne et al. found that suPAR was significantly associated with increased plasma levels of NT-proBNP . Although the mechanisms underlying the relationship between suPAR and BNP remain unclear, there are several possibilities. Subjects with increased suPAR may have enhanced systemic immune and inflammatory conditions [6,24] that may also be associated with the development of heart failure [25,26] and left ventricular function [27,28].
Mekonnen et al. also showed that coronary flow reserve was negatively associated with suPAR in patients with non-obstructive coronary artery disease . Furthermore, Theilade et al. reported that, among patient with type 1 diabetes, subjects with increased suPAR were more likely to have a high degree of arterial stiffness . These observations suggest that impaired coronary microcirculation and increased arterial stiffness may lead to left ventricular diastolic dysfunction [30,31]. In the current study, we could not assess the relationship between suPAR and diastolic dysfunction because of the small sample size-only 12 patients were judged to have diastolic dysfunction among patients with preserved LVEF (>50%, n = 193). This point should be investigated in the future studies.
We also found that patients with higher suPAR levels had lower LVEF (Table 2). To date, only a few studies have reported a relationship between suPAR and LVEF. By analyzing 318 patients with type 1 diabetes without known heart disease, Theilade et al. found that subjects with higher suPAR tended to have lower LVEF by univariate analysis, although this relationship was not significant after multivariate adjustment . Theilade et al.’s population did not include patients with known heart disease or end-stage renal disease. On the other hand, our study population included both those who had more than moderately impaired renal function and those with ischemic heart disease. In the subgroup analysis in the current study, the association between high suPAR and low LVEF was more pronounced and significant among those with low eGFR and those with ischemic heart disease, respectively; therefore, the difference in the observation between Theilade et al.’s study and ours might be attributed to the different study population. Fewer studies seem to have investigated the relationship between suPAR and cardiac hypertrophy. In the above-mentioned study of Taheilade et al., suPAR did not have a significant association with LVMI after multivariate adjustment , in agreement with our study.
Interestingly, after excluding patients who were admitted due to the worsening heart failure from the analysis, the highest suPAR was still significantly associated with low LVEF with an odds ratio of 5.20 (95% CI 1.25–21.66 P = 0.023) after adjusting for age, sex, log(eGFR), CRP, and diuretic use. It was suggested, therefore, that suPAR may be independently associated with decreased LV function, although those who were admitted due to worsening heart failure had significantly higher suPAR levels compared with those who were admitted due to other reasons. It was shown by a recent experimental study that bone marrow Gr-1lo immature myeloid cells may be responsible for the elevated, pathological levels of suPAR, and when these cells were transferred to healthy mice, it efficiently transmitted proteinuria when transferred to healthy animals . Which cells were responsible for the increased suPAR among patients with low LVEF, and whether suPAR per ce play a role in promoting cardiac systolic dysfunction await further investigation.
Although suPAR was found to be associated with decreased LVEF and elevated of BNP independent of renal function and CRP, we do not propose routine measurement of suPAR in clinical practice. The utility of suPAR, in addition to its biomarker properties , may lie in its ability to increase our understanding of the pathogenesis of the observed cardiovascular abnormalities.
The current study has several limitations. First, owing to its cross-sectional nature, the study cannot provide information on the causal or resultant nature of the relationship. Second, the study subjects had various cardiovascular disorders because we enrolled patients who were admitted to the cardiology department. Although the data are relevant in real-world clinical practice; however, the possibility that the relationship between suPAR and left ventricular dysfunction differs according to certain specific cardiovascular conditions requires further evaluation. Third, the number of subjects with preserved LVEF was small; therefore, we could not examine the potential relationship between suPAR and cardiac diastolic dysfunction, which has been suggested in a previous study .
In conclusion, serum suPAR concentrations were associated with low LVEF (< 50%) and elevated plasma BNP (>300 pg/mL), but not with left ventricular hypertrophy among cardiac patients. The association between suPAR and low LVEF and elevated BNP remained significant after adjusting for age, sex, eGFR, CRP, and diuretic use. Whether suPAR represents a useful guiding biomarker for the treatment of cardiac dysfunction and heart failure, and whether it is involved in the progression of cardiac disorders await further investigation.
This work was supported in part by Grants in Aid for Scientific Research from the Ministry of Education, Science, and Culture of Japan (No. 15K09106) but there are no relationships with the company relating to employment, consultancy, patents, products in development or marketed products, and other categories that may potently our adherence to all the PLOS ONE policies. We are highly appreciative of Chieko Ohta, Yumiko Ohgami, and Megumi Hashimoto for their excellent technical assistance.
- Conceptualization: NI.
- Data curation: SF ST HM TF.
- Formal analysis: YT DM.
- Investigation: TI DM.
- Methodology: NI.
- Supervision: NI.
- Writing – original draft: SF.
- Writing – review & editing: NI.
- 1. Gyetko MR, Sud S, Chen GH, Fuller JA, Chensue SW, Toews GB (2002) Urokinase-type plasminogen activator is required for the generation of a type 1 immune response to pulmonary Cryptococcus neoformans infection. J Immunol 168: 801–809. pmid:11777975
- 2. Cook AD, De Nardo CM, Braine EL, Turner AL, Vlahos R, Way KJ, et al. (2010) Urokinase-type plasminogen activator and arthritis progression: role in systemic disease with immune complex involvement. Arthritis Res Ther 12: R37. pmid:20196869
- 3. Kaneko T, Konno H, Baba M, Tanaka T, Nakamura S (2003) Urokinase-type plasminogen activator expression correlates with tumor angiogenesis and poor outcome in gastric cancer. Cancer Sci 94: 43–49. pmid:12708473
- 4. Han B, Nakamura M, Mori I, Nakamura Y, Kakudo K (2005) Urokinase-type plasminogen activator system and breast cancer (Review). Oncol Rep 14: 105–112. pmid:15944776
- 5. Lyngbaek S, Marott JL, Sehestedt T, Hansen TW, Olsen MH, Andersen O, et al. (2013) Cardiovascular risk prediction in the general population with use of suPAR, CRP, and Framingham Risk Score. Int J Cardiol 167: 2904–2911. pmid:22909410
- 6. Edsfeldt A, Nitulescu M, Grufman H, Gronberg C, Persson A, Nilsson M, et al. (2012) Soluble urokinase plasminogen activator receptor is associated with inflammation in the vulnerable human atherosclerotic plaque. Stroke 43: 3305–3312. pmid:23150653
- 7. Sorensen MH, Gerke O, Eugen-Olsen J, Munkholm H, Lambrechtsen J, Sand NP, et al. (2014) Soluble urokinase plasminogen activator receptor is in contrast to high-sensitive C-reactive-protein associated with coronary artery calcifications in healthy middle-aged subjects. Atherosclerosis 237: 60–66. pmid:25222341
- 8. Eapen DJ, Manocha P, Ghasemzadeh N, Patel RS, Al Kassem H, Hammadah M, et al. (2014) 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 3: e001118. pmid:25341887
- 9. Persson M, Engstrom G, Bjorkbacka H, Hedblad B (2012) Soluble urokinase plasminogen activator receptor in plasma is associated with incidence of CVD. Results from the Malmo Diet and Cancer Study. Atherosclerosis 220: 502–505. pmid:22119508
- 10. Steins MB, Padro T, Schwaenen C, Ruiz S, Mesters RM, Berdel WE, et al. (2004) Overexpression of urokinase receptor and cell surface urokinase-type plasminogen activator in the human vessel wall with different types of atherosclerotic lesions. Blood Coagul Fibrinolysis 15: 383–391. pmid:15205586
- 11. Theilade S, Rossing P, Eugen-Olsen J, Jensen JS, Jensen MT (2016) suPAR level is associated with myocardial impairment assessed with advanced echocardiography in patients with type 1 diabetes with normal ejection fraction and without known heart disease or end-stage renal disease. Eur J Endocrinol 174: 745–753. pmid:26951602
- 12. Borne Y, Persson M, Melander O, Smith JG, Engstrom G (2014) Increased plasma level of soluble urokinase plasminogen activator receptor is associated with incidence of heart failure but not atrial fibrillation. Eur J Heart Fail 16: 377–383. pmid:24464777
- 13. Kruger R, Schutte R, Huisman HW, Hindersson P, Olsen MH, Eugen-Olsen J, et al. (2013) NT-proBNP, C-reactive protein and soluble uPAR in a bi-ethnic male population: the SAfrEIC study. PLoS One 8: e58506. pmid:23516493
- 14. Hayek SS, Sever S, Ko YA, Trachtman H, Awad M, Wadhwani S, et al. (2015) Soluble Urokinase Receptor and Chronic Kidney Disease. N Engl J Med 373: 1916–1925. pmid:26539835
- 15. Schulz CA, Persson M, Christensson A, Hindy G, Almgren P, Nilsson PM, et al. (2016) Soluble Urokinase-type Plasminogen Activator Receptor (suPAR) and Impaired Kidney Function in the Population-based Malmö Diet and Cancer Study. KI Rep.
- 16. Hodges GW, Bang CN, Wachtell K, Eugen-Olsen J, Jeppesen JL (2015) suPAR: A New Biomarker for Cardiovascular Disease? Can J Cardiol 31: 1293–1302. pmid:26118447
- 17. Sola S, Mir MQ, Lerakis S, Tandon N, Khan BV (2006) Atorvastatin improves left ventricular systolic function and serum markers of inflammation in nonischemic heart failure. J Am Coll Cardiol 47: 332–337. pmid:16412856
- 18. van der Walt C, Malan L, Uys AS, Malan NT (2013) Low grade inflammation and ECG left ventricular hypertrophy in urban African males: The SABPA study. Heart Lung Circ 22: 924–929. pmid:23711691
- 19. Mehta SK, Rame JE, Khera A, Murphy SA, Canham RM, Peshock RM, et al. (2007) Left ventricular hypertrophy, subclinical atherosclerosis, and inflammation. Hypertension 49: 1385–1391. pmid:17404181
- 20. Matsuo S, Imai E, Horio M, Yasuda Y, Tomita K, Nitta K, et al. (2009) Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis 53: 982–992. pmid:19339088
- 21. Devereux RB, Reichek N (1977) Echocardiographic determination of left ventricular mass in man. Anatomic validation of the method. Circulation 55: 613–618. pmid:138494
- 22. Wachtell K, Bella JN, Liebson PR, Gerdts E, Dahlof B, Aalto T, et al. (2000) Impact of different partition values on prevalences of left ventricular hypertrophy and concentric geometry in a large hypertensive population: the LIFE study. Hypertension 35: 6–12. pmid:10642267
- 23. Roman MJ, Pickering TG, Schwartz JE, Pini R, Devereux RB (1996) Relation of arterial structure and function to left ventricular geometric patterns in hypertensive adults. J Am Coll Cardiol 28: 751–756. pmid:8772767
- 24. Pliyev BK, Menshikov MY (2010) Release of the soluble urokinase-type plasminogen activator receptor (suPAR) by activated neutrophils in rheumatoid arthritis. Inflammation 33: 1–9. pmid:19756998
- 25. Yndestad A, Damas JK, Oie E, Ueland T, Gullestad L, Aukrust P (2006) Systemic inflammation in heart failure—the whys and wherefores. Heart Fail Rev 11: 83–92. pmid:16819581
- 26. Briasoulis A, Androulakis E, Christophides T, Tousoulis D (2016) The role of inflammation and cell death in the pathogenesis, progression and treatment of heart failure. Heart Fail Rev 21: 169–176. pmid:26872673
- 27. Li L, Zhao X, Lu Y, Huang W, Wen W (2010) Altered expression of pro- and anti-inflammatory cytokines is associated with reduced cardiac function in rats following coronary microembolization. Mol Cell Biochem 342: 183–190. pmid:20625799
- 28. Nakagomi A, Seino Y, Kohashi K, Kosugi M, Endoh Y, Kusama Y, et al. (2012) Effects of statin therapy on the production of monocyte pro-inflammatory cytokines, cardiac function, and long-term prognosis in chronic heart failure patients with dyslipidemia. Circ J 76: 2130–2138. pmid:22785006
- 29. Mekonnen G, Corban MT, Hung OY, Eshtehardi P, Eapen DJ, Al-Kassem H, et al. (2015) Plasma soluble urokinase-type plasminogen activator receptor level is independently associated with coronary microvascular function in patients with non-obstructive coronary artery disease. Atherosclerosis 239: 55–60. pmid:25574858
- 30. Marechaux S, Samson R, van Belle E, Breyne J, de Monte J, Dedrie C, et al. (2016) Vascular and Microvascular Endothelial Function in Heart Failure With Preserved Ejection Fraction. J Card Fail 22: 3–11. pmid:26386451
- 31. Hsu PC, Tsai WC, Lin TH, Su HM, Voon WC, Lai WT, et al. (2012) Association of arterial stiffness and electrocardiography-determined left ventricular hypertrophy with left ventricular diastolic dysfunction. PLoS One 7: e49100, pmid:23145083
- 32. Hahm E, Wei C, Fernandez I, Li J, Tardi NJ, Tracy M, et al. (2016) Bone marrow-derived immature myeloid cells are a main source of circulating suPAR contributing to proteinuric kidney disease. Nat Med.