Serum Resistin Is Associated with High Risk in Patients with Congestive Heart Failure a Novel Link Between Metabolic Signals and Heart Failure

Circ J 2007; 71: 460–464

Serum Resistin is Associated With High Risk in Patients With Congestive Heart Failure A Novel Link Between Metabolic Signals and Heart Failure

Yasuchika Takeishi, MD; Takeshi Niizeki, MD; Takanori Arimoto, MD; Naoki Nozaki, MD; Osamu Hirono, MD; Joji Nitobe, MD; Tetsu Watanabe, MD; Noriaki Takabatake, MD; Isao Kubota, MD

Background Resistin is derived from fat tissue in rodents, and serum levels are elevated in animal models of obesity and insulin resistance. Recent studies have reported that resistin is correlated with markers of inflammation and oxidative stress and is predictive of coronary atherosclerosis in humans. However, clinical significance of serum resistin has not been examined in heart failure. Therefore, the purpose of this study was to examine whether: (1) resistin is correlated with the severity of heart failure; and (2) resistin can predict clinical outcomes of patients with heart failure. Methods and Results Serum levels of resistin in 126 patients hospitalized for heart failure and 18 control subjects were measured. The patients were followed up with end-points of cardiac death and re-hospitalization caused by worsening of heart failure. The serum resistin level was higher in patients with heart failure than in control subjects and increased with advancing New York Heart Association functional class. The normal upper limit of the resistin level was determined as the mean +2standard deviation value of control subjects (14.1ng/ml). In heart failure patients, the cardiac event rate was higher in patients with a high resistin level than in those with a normal level. Among age, body mass index, serum levels of resistin, brain natriuretic peptide, loop diuretics selected by the univariate Cox regression hazard analysis, age and resistin were significant predictors of future cardiac events by multivariate Cox analysis. Conclusion Serum resistin was related to the severity of heart failure and associated with a high risk for adverse cardiac events in patients with heart failure. (Circ J 2007; 71: 460–464) Key Words: Adipocytokine; Heart failure; Resistin; Prognosis

dipose tissue secretes bioactive peptides, called “adi- health problem that causes a high mortality rate because of pocytokines”, which act locally and distally through the increasing aging population and high prevalence of A autocrine, paracrine and endocrine effects.1–4 An heart failure in the elderly.11–13 However, the relationship increased production of adipocytokines impacts on multi- between adipocytokines and heart failure has not been ple functions such as energy balance, immunity, insulin previously examined. Despite the significant reduction in sensitivity, angiogenesis, blood pressure, lipid metabolism mortality achieved in recent clinical trials, heart failure and haemostasis, all of which are linked to cardiovascular patients still have a poor prognosis.14 Thus, the role of diseases. Resistin has been recently identified as a novel cardiac biomarkers in the evaluation and risk stratification adipose-specific cysteine-rich protein.5–8 It is a 12.5-kDa of patients with heart failure continues to increase in polypeptide, whose functions are not clearly established, importance. although it has been associated with insulin resistance in Therefore, the purpose of the present study was to ex- rodents. It has been reported that resistin potentially links amine whether: (1) serum levels of resistin, 1 of adipo- obesity to diabetes in mice with diet-induced obesity and in cytokines, were related to the severity of heart failure; and genetically obese mice.9 A recent study has demonstrated (2) serum resistin was useful to predict adverse clinical that plasma resistin levels are correlated with markers of outcomes in patients with heart failure. inflammation and are predictive of coronary atherosclerosis in humans.10 Congestive heart failure (CHF) is a major and increasing Methods Study Population (Received November 9, 2006; revised manuscript received December Between January 2001 and April 2004, we prospectively 26, 2006; accepted January 16, 2007) enrolled 126 consecutive CHF patients (76 men and 50 Department of Cardiology, Pulmonology, and Nephrology, Yamagata women, mean age 67±13 years, range 21–92 years) who University School of Medicine, Yamagata, Japan Mailing address: Yasuchika Takeishi, MD, Department of Cardiology, had been admitted to the Yamagata University Hospital for Pulmonology, and Nephrology, Yamagata University School of Medi- treatment because of worsening CHF, for the diagnosis and cine, 2-2-2 Iida-nishi, Yamagata 990-9585, Japan. E-mail: takeishi pathophysiological investigations, or for therapeutic evalua- @med.id.yamagata-u.ac.jp tions of CHF. Baseline characteristics of the study popula-

Circulation Journal Vol.71, April 2007 Serum Resistin in Heart Failure 461 tion are shown in Table1. Exclusion criteria in the present Table 1 Clinical Characteristics of 126 Patients With Chronic Heart study included: patients with clinical or electrocardio- Failure graphic evidence suggestive of acute coronary syndrome n=126 within 3 months preceding admission, those with renal failure characterized by a serum creatinine concentration Age (years) 67±13 ≥ M/F 76/50 1.5mg/dl, and those with active hepatic or pulmonary dis- NYHA class (I/II/III/IV) 33/44/43/6 15–19 ease. Informed consent was obtained from all patients Hypertension (%) 71 (56%) before participation in this study, and the study protocol DM (%) 32 (16%) was approved by the Human Investigations Committee of Hyperlipidemia (%) 24 (19%) our institution. Smoking (%) 31 (29%) Blood samples were obtained at admission for measure- BMI (kg/m2) 23.5±3.8 Etiology of heart failure (%) ments of serum resistin levels. A 2-dimensional echocar- Dilated cardiomyopathy 48 (38%) diography was performed by an experienced cardiologist, Ischemic heart disease 34 (27%) who did not have knowledge of the biochemical data, Valvular heart disease 20 (15%) within 1 week after admission. Clinical data, including age, Hypertensive heart disease 12 (10%) sex and New York Heart Association (NYHA) functional Others 12 (10%) class at admission were obtained from patient interviews Blood examinations Glucose (mg/dl) 131±62 and hospital medical records. The diagnosis of CHF was HbA1c (%) 6.2±1.3 based on a history of dyspnea and symptomatic exercise in- Hb (g/dl) 12.8±2.1 tolerance with signs of pulmonary congestion or peripheral Total cholesterol (mg/dl) 187±44 edema or documentation of left ventricular enlargement or Triglyceride (mg/dl) 113±63 dysfunction by chest X-ray film, echocardiography, or HDL-C (mg/dl) 50±15 radionuclide ventriculography.15–19 The etiologies of CHF LDL-C (mg/dl) 113±38 Creatinine (mg/dl) 0.9±0.3 were dilated cardiomyopathy in 48 patients, ischemic heart Resistin (ng/ml) 12.1±8.5 disease in 34 patients, valvular heart disease in 20 patients, BNP (pg/ml) 495±647 and hypertensive heart disease in 12 patients. A diagnosis of Echocardiography hypertension, diabetes, and hyperlipidemia were obtained LVEDD (mm) 53±10 from medical records or patient history of currently or pre- LVEF (%) 49±20 viously received medical therapy. E/A ratio 2.2±7.8 DcT (ms) 203±85 Physicians were kept blinded to the results of the bio- Medications (%) chemical markers, and optimal medical therapy was per- ACEI/ARB 86 (68%) formed independently based on improvement of symptoms, β-blockers 50 (40%) physical examination findings, and pulmonary congestion Ca channel blockers 40 (32%) on chest X-rays. Loop diuretics 74 (59%) Spironolactone 31 (25%) Digoxin 40 (32%) End-Points and Follow up Statins 22 (17%) No patients were lost to follow up (mean follow up 645±644 days, range 29–1,080 days) after discharge. The NYHA, New York Heart Association; DM, diabetes mellitus; BMI, body mass end-points were: (1) cardiac death, defined as death from index; Hb, hemoglobin; HDL-C, high-density lipoprotein-cholesterol; LDL- worsening CHF or sudden cardiac death; and (2) worsening C, low-density lipoprotein-cholesterol; BNP, brain natriuretic peptide; 15–19 LVEDD, left ventricular end-diastolic dimension; LVEF, left ventricular ejec- CHF requiring readmission. Sudden cardiac death was tion fraction; DcT, deceleration time; ACEI, angiotensin-converting enzyme- defined as death without definite premonitory symptoms or inhibitors; ARB, angiotensin II receptor blocker. signs and was confirmed by the attending physician. A re- view of medical records and follow-up telephone interviews were conducted to survey cardiac events by 2 cardiologists, value below the lower detection limit of the assay was who were blinded to blood examination data. Cardiac defined as zero. A Cox proportional hazard analysis was events were adjudicated using medical records, electrocar- performed to determine the independent predictor of car- diograms, chest X-ray reports, autopsy reports, death cer- diac events for the entire population. Independent pre- tificates and witness statements. dictors selected in the univariate analysis were entered into the multivariate analysis. A cardiac event-free curve was Measurement of Serum Resistin Levels computed according to the Kaplan – Meier method and Blood samples obtained at admission for measurement compared by a log-rank test. A p value of less than 0.05 was of serum levels of resistin were drawn and centrifuged at considered as statistically significant. Statistical analysis 2,500G for 15min at 4°C within 30min of collection, and was performed with a standard statistical program package the serum obtained was stored at –70°C until analysis. (StatView, version 5.0, SPSS version 10.0 Institute Inc). Serum resistin concentration was measured by using a sandwich enzyme-linked immunosorbent assay (ELISA, Phoenix Pharmaceutical, Inc, Belmont, CA, USA) as pre- Results viously reported.7 Intra- and inter-assay coefficients of varia- Serum Resistin Levels in Study Subjects tion were 3.4 and 6.3%, respectively. Serum resistin levels were measured in 126 patients with heart failure and in 18 control subjects. As shown in Fig1, Statistical Analysis serum resistin levels increased with advancing NYHA Data are presented as mean±standard deviation (SD). functional class. The normal upper limit of resistin levels The t-test or chi-square test was used for comparisons of was defined as mean +2SD value, and for 18 control sub- continuous and categorical variables, respectively. The jects, it was 14.1ng/ml. There were 32 patients with a high

Circulation Journal Vol.71, April 2007 462 TAKEISHI Y et al.

Fig1. Serum resistin levels in patients with chronic heart failure and control subjects. *p<0.01 vs control, $p<0.01 vs New York Heart Fig 2. Comparison of cardiac event rates between high and low Association (NYHA) I, #p<0.01 vs NYHA II, †p<0.05 vs NYHA III. resistin groups. Dashed line indicates normal upper limit of resistin determined as mean +2SD value of control subjects (14.1ng/ml).

Table 2 Comparisons of Clinical Characteristics Between Patients With and Without Cardiac Events

Event (+) Event (–) p value (n=31) (n=95) Age (years) 73±11 65±13 0.0026 M/F19/12 57/38 0.8984 NYHA class (I/II/III/IV) 3/7/18/3 30/37/25/3 0.0017 Hypertension (%) 16 (52%) 55 (58%) 0.5412 DM (%) 5 (16%) 27 (28%) 0.1577 Hyperlipidemia (%) 6 (19%) 18 (19%) 0.9600 Smoking (%) 9 (29%) 22 (23%) 0.5148 BMI (kg/m2) 22.5±3.9 23.8±3.7 0.1405 Etiology of heart failure (%) Dilated cardiomyopathy 13 (42%) 35 (37%) Ischemic heart disease 6 (19%) 28 (29%) Valvular heart disease 5 (16%) 15 (16%) Fig3. Kaplan–Meier survival curves between patients with high Hypertensive heart disease 7 (23%) 5 (5%) and low resistin levels. Others 0 (0%) 12 (13%) 0.0151 Blood examinations Glucose (mg/dl) 116±37 136±68 0.1546 HbA1c (%) 5.8±0.9 6.3±1.3 0.1291 resistin level (>14.1 ng/ml) and 94 patients with a low Hb (g/dl) 12.1±2.2 13.1±1.9 0.0262 resistin level (≤14.1ng/ml). Total cholesterol (mg/dl) 181±43 189±44 0.4044 Serum resistin levels were not correlated with body mass Triglyceride (mg/dl) 97±48 118±66 0.1267 index (r=0.118, p=0.224) or blood glucose (r=0.103, p= HDL-C (mg/dl) 45±11 52±16 0.0403 0.280). Serum resistin levels were not different between LDL-C (mg/dl) 116±39 112±37 0.6574 patients with and without diabetes mellitus (10.1±6.9 vs Creatinine (mg/dl) 1.0±0.3 0.8±0.2 0.0005 Resistin (ng/ml) 14.8±7.7 11.2±8.6 0.0368 12.8±8.9ng/ml, p=0.117). BNP (pg/ml) 808±716 393±592 0.0017 Echocardiography Resistin Levels and Cardiac Event Rates LVEDD (mm) 54±10 52±9 0.4003 There were 31 cardiac events including 10 cardiac deaths LVEF (%) 45±22 50±18 0.2760 and 21 re-hospitalizations because of worsening heart fail- E/A ratio 1.2±1.4 1.1±0.9 0.0222 ure during the follow-up period. Comparisons of clinical DcT (ms) 199±99 205±81 0.8274 Medications (%) characteristics between patients with and without cardiac ACEI/ARB 24 (77%) 62 (65%) 0.2068 events are shown in Table2. Patients with cardiac events β-blockers 15 (50%) 35 (37%) 0.1997 were older, and had more severe NYHA functional class, Ca channel blockers 7 (23%) 33 (35%) 0.2431 lower hemoglobin levels, lower high-density lipoprotein- Loop diuretics 24 (80%) 50 (53%) 0.0078 cholesterol levels, higher creatinine levels, higher resistin Spironolactone 12 (40%) 19 (20%) 0.0270 levels, higher brain natriuretic peptide (BNP) levels, and a Digoxin 12 (40%) 28 (29%) 0.2813 Statins 6 (20%) 16 (17%) 0.6921 higher E/A ratio than those without cardiac events. Patients with cardiac events were given loop diuretics and spirono- Abbreviations as in Table1. lactone more frequently than event-free patients. The cardiac event rate was compared between high and low resistin groups (Fig2). The cardiac event rate was sig- niﬁcantly higher in patients with high resistin levels than those with low resistin levels (p<0.0036 by chi-square test).

Circulation Journal Vol.71, April 2007 Serum Resistin in Heart Failure 463

Table 3 Predictors of Cardiac Events in the Univariate Cox Regression Analysis

Hazard ratio 95%CI p value Age (per 1SD increase) 2.358 1.450–3.838 0.0006 Gender (F vs M) 1.041 0.505–2.145 0.9138 NYHA class 1.012 0.936–1.094 0.7725 Hypertension 0.740 0.365–1.499 0.4027 DM 0.601 0.231–1.567 0.2984 Hyperlipidemia 0.979 0.401–2.398 0.9643 BMI (per 1SD increase) 0.587 0.354–0.977 0.0398 Resistin (per 1SD increase) 1.357 1.061–1.746 0.0145 BNP (per 1SD increase) 1.910 1.000–1.909 0.0001 LVEDD (per 1SD increase) 1.209 0.809–1.808 0.3605 LVEF (per 1SD increase) 0.803 0.544–1.196 0.2994 ACEI/ARB 1.812 0.779–4.219 0.1678 -blockers 1.876 0.914–3.846 0.0862 Fig4. Comparisons of cardiac event rates among 4 groups. Patients β Ca channel blockers 0.465 0.197–1.095 0.0800 were divided into 4 groups (1st to 4th quartile) based on the serum Loop diuretics 3.367 1.372–8.264 0.0080 resistine levels. Spironolactone 2.695 1.292–5.650 0.0831 Digoxin 1.098 0.843–3.704 0.1317 Statins 1.098 0.447–2.695 0.8386 Kaplan–Meier survival analysis also demonstrated that the high resistin group showed a significantly lower event-free CI, confidence interval; SD, standard deviation. Other abbreviations as in rate than the low resistin group did, as shown in Fig 3 Table 1. (p=0.0041 by log rank test). Next, patients were divided into 4 groups based on the Table 4 Predictors of Cardiac Events in the Multivariate Cox resistin levels: 1st quartile (1.7–6.6ng/ml, n=31), 2nd quar- Regression Analysis tile (6.7–10.1ng/ml, n=31), 3rd quartile (10.2–14.0ng/ml, n=31), and 4th quartile (14.1–60.2ng/ml, n=33). As shown Hazard ratio 95%CI p value in Fig4, the highest 4th quartile of resistin was associated Age (per 1SD increase) 2.583 1.361–4.842 0.0035 with the highest risk for cardiac events (42.4%) compared BMI (per 1SD increase) 0.902 0.527–1.538 0.7046 to the other 3 groups. Resistin (per 1SD increase) 1.439 1.017–2.059 0.0414 BNP (per 1SD increase) 1.910 1.000–1.909 0.0942 Univariate and Multivariate Cox Regression Analyses Loop diuretics 2.183 0.794–5.988 0.1305 To determine the risk factors to predict cardiac events, Abbreviations as in Tables1,3. we performed the univariate Cox proportional hazard regression analysis (Table3). In the univariate analysis, serum resistin level was associated with cardiac death and re- the apM1 gene, and is highly expressed in human adipose hospitalization (per 1SD increase, hazard ratio 1.357, 95% tissue. Adiponectin has anti-atherogenic and anti-inflam- confidence interval 1.061–1.746, p<0.0145). Furthermore, matory properties. Resistin is a novel adipose-specific age, body mass index, BNP, and the use of loop diuretics cysteine-rich secretory protein.5–9 Resistin is derived almost were significantly associated with cardiac death and re- exclusively from fat tissue in rodents, and adipose expres- hospitalization. sion and serum levels are elevated in animal models of Then, those variables with a p value of less than 0.05 in obesity and insulin resistance. High serum levels of resistin the univariate analysis were entered into the multivariate impair glucose tolerance and induce insulin resistance in Cox proportional hazard regression analysis. As shown in rats.20 In contrast, mice deficient in resistin are protected Table4, resistin and age were independent predictors for from obesity-associated insulin resistance.21 cardiac events in patients with heart failure. However, serum resistin was unexpectedly not correlated with body mass index and glucose levels in the present study. In addition, serum resistin levels were not different Discussion between patients with and without diabetes mellitus. Possi- We demonstrated that serum levels of resistin were ble reasons for this include: (1) the body mass index was increased in patients with heart failure, especially in those relatively low (23.5±3.8kg/m2) in the present study popu- with severe heart failure (NYHA class III and IV). High lation compared to Western countries; (2) patients who resistin levels were associated with high cardiac event were obese and who had diabetes mellitus were not com- rates, and resistin was an independent predictor of future mon sufferers of heart failure compared to coronary heart cardiac events by multivariate Cox regression hazard disease; and (3) the incidence of ischemic cardiomyopathy analysis. caused by advanced coronary atherosclerosis was only Adipose tissue is not only a store for excess energy but 27% in the present study (Table1). also a highly active endocrine organ.1–4 The adipocyte- Serum resistin levels increased with advancing NYHA derived molecules, called adipocytokines, influence the functional class and high serum resistin was associated function and structural integrity of various tissues. Leptin, with high cardiac event rates including cardiac death and a 167-amino acid protein, is expressed exclusively by progressive heart failure requiring re-hospitalization. Car- adipose tissue. Leptin is a fundamental signal of satiety to diac event rates increased from 1st to 4th quartile of resistin, the brain and a regulator of insulin and glucose metabolism. and the highest 4th quartile was associated with the highest Adiponectin, one of the most abundant adipocytokines re- risk compared to the other 3 groups. The univariate Cox cently discovered, is a 244-amino acid protein produced by regression analysis showed that age, body mass index,

Circulation Journal Vol.71, April 2007 464 TAKEISHI Y et al. resistin, BNP and loop diuretics were significant predictors and resistin, and their relationship with cardiovascular disease in pa- of future cardiac events in patients with heart failure. tients with end-stage renal disease. Clin Endocrinol 2005; 62: 242– 249. Among them, age and resistin were significant variables to 8. Burnett MS, Devaney JM, Adenika RJ, Lindsay R, Howard BV. predict future cardiac events by the multivariate analysis in Cross-sectional associations of resistin, coronary heart disease, and the present study, although a number of studies have previ- insulin resistance. J Clin Endocrinol Metab 2006; 91: 64–68. ously demonstrated that BNP is a useful tool to monitor 9. Steppan CM, Bailey ST, Bhat S, Brown EJ, Banerjee RR, Wright 22,23 CM, et al. The hormone resistin links obesity to diabetes. Nature disease outcome in patients with heart failure. 2001; 409: 307–312. An experimental study showed that resistin impairs car- 10. Reilly MP, Lehrke M, Wolfe ML, Rohatgi A, Lazar MA, Rader DJ. diac recovery following ischemia through stimulation of Resistin is an inflammatory marker of atherosclerosis in humans. cardiac tumor necrosis factor (TNF)-αsecretion.24 Because Circulation 2005; 111: 932–939. a number of studies have demonstrated that increased 11. Cowburn PJ, Cleland JG, Coats AJ, Komajda M. Risk stratification in chronic heart failure. Eur Heart J 1998; 19: 696–710. levels of inflammatory cytokines including TNF-α play a 12. Bassenge E, Fink B, Schwemmer M. Oxidative stress, vascular dys- pathological role in heart failure, it is possible that high function and heart failure. Heart Fail Rev 1999; 4: 133–145. serum levels of resistin might cause TNF-αsecretion and 13. Tsutsui H. Novel pathophysiological insight and treatment strategies be related to adverse clinical outcomes in the present study. for heart failure: Lessons from mice and patients. Circ J 2004; 68: 1095–1103. We did not measure levels of TNF-αand other inflamma- 14. Bettencourt P, Ferreira A, Dias P. Predictors of prognosis in patients tory cytokines in this study population, and so further study with stable mild to moderate heart failure. J Card Fail 2000; 6: 306– is needed to support this hypothesis in humans. 313. 15. Arimoto T, Takeishi Y, Shiga R, Fukui A, Tachibana H, Nozaki N, et al. Prognostic value of elevated circulating heart-type fatty acid Conclusion binding protein in patients with congestive heart failure. J Card Fail 2005; 11: 56–60. These data suggest that resistin is useful for risk stratifi- 16. Niizeki T, Takeishi Y, Arimoto T, Okuyama H, Takabatake N, Fukui cation of patients with heart failure and might represent a A, et al. Serum heart-type fatty acid binding protein predicts cardiac events in the elderly chronic heart failure patients. J Cardiol 2005; novel link between metabolic signals and heart failure. 46: 9–15. 17. Niizeki T, Takeishi Y, Arimoto T, Takahashi T, Okuyama H, Acknowledgement Takabatake N, et al. A combination of heart-type fatty acid binding protein and brain natriuretic peptide can reliably risk stratify patients This study was supported, in part, by a grant-in-aid for Scientific Re- hospitalized for chronic heart failure. Circ J 2005; 69: 922–927. search (No. 17590702) from the Ministry of Education, Culture, Sports, 18. Arimoto T, Takeishi Y, Niizeki T, Takabatake N, Okuyama H, Fukui Science and Technology, a grant-in-aid from the 21th century center of A, et al. Cystatin C is a novel predictor of cardiac events in patients excellence (COE) program of the Japan Society for the Promotion of with chronic heart failure. J Card Fail 2005; 11: 595–601. Science, and grants from Takeda Science Foundation and Fukuda 19. Arimoto T, Takeishi Y, Niizeki T, Okuyama H, Nozaki N, Hirono O, Foundation for Medical Technology. et al. Cardiac sympathetic denervation and ongoing myocardial dam- age for prognosis in heart failure. J Card Fail 2007 (in press). References 20. Rajala MW, Obici S, Scherer PE, Rossetti L. Adipose-derived resistin and gut-derived resistin-like molecule-beta selectively impair 1. Yi-Hao Y, Ginsberg HN. Adipocyte signaling and lipid homeostasis. insulin action on glucose production. J Clin Invest 2003; 111: 225– Circ Res 2005; 96: 1042–1052. 230. 2. Petersen KF, Shulman GI. Etiology of insulin resistance. Am J Med 21. Benerjee RR, Rangwala SM, Shapiro JS, Rich AS, Rhoades B, Qi Y, 2006; 119(5 Suppl 1): S10–S16. et al. Regulation of fasted blood glucose by resistin. Science 2004; 3. Rondinone CM. Adipocyte-derived hormones, cytokines, and media- 303: 1195–1198. tors. Endocrine 2006; 29: 81–90. 22. Taniguchi R, Sato Y, Yamada T, Ooba M, Higuchi H, Matsumori A, 4. Ronti T, Lupattelli G, Mannarino E. The endocrine function of et al. Combined measurements of cardiac troponin T and N-terminal adipose tissue: An update. Clin Endocrinol (Oxf) 2006; 64: 355– pro-brain natriuretic peptide in patients with heart failure. Circ J 365. 2004; 68: 1160–1164. 5. Beltowski J. Adiponectin and resistin-new hormones of white adipose 23. Ishikawa C, Tsutamoto T, Fujii M, Sakai H, Tanaka T, Horie M. Pre- tissue. Med Sci Monit 2003; 9: RA55–RA61. diction of mortality by high-sensitive C-reactive protein and brain 6. Kunnari A, Ukkola O, Kesaeniemi YA. Resistin polymorphisms are natriuretic peptide in patients with dilated cardiomyopathy. Circ J associated with cerebrovascular disease in Finnish type 2 diabetic 2006; 70: 857–863. patients. Diabet Med 2005; 22: 583–589. 24. Rothwell SE, Richards AM, Pemberton CJ. Resistin worsens cardiac 7. Diez JJ, Iglesias P, Fernandez-Reyes MJ, Aguilera A, Bajo MA, ischaemia-reperfusion injury. Biochem Biophys Res Commun 2006; Alvarez-Fidalgo P, et al. Serum concentration of leptin, adiponectin 349: 400–407.

Circulation Journal Vol.71, April 2007