Elevated Plasma Levels of Nt-Probnp in Patients with Type 2 Diabetes Without Overt Cardiovascular Disease

Emerging Treatments and Technologies ORIGINAL ARTICLE

Elevated Plasma Levels of Nt-proBNP in Patients With Type 2 Diabetes Without Overt Cardiovascular Disease

1 3 MARTIN MAGNUSSON, MD ANDERS GRUBB, MD, PHD day, but there is no comparable 2 2 OLLE MELANDER, MD, PHD LEIF GROOP, MD, PHD reoccurring screening for cardiac compli- 1 1,4,5 BO ISRAELSSON, MD, PHD STEFAN JOVINGE, MD, PHD cations of diabetes. This may simply be due to the lack of cost-effective methods; an echocardiographical examination is both expensive and time consuming and, therefore, not suited for screening pur- OBJECTIVE — The NH2-terminal portion of the precursor of brain natriuretic peptide (Nt- poses. The most evident cardiac compli- proBNP) has been reported to be elevated in left ventricular dysfunction. This peptide is a split cation is coronary atherosclerosis. Not product from the proBNP molecule, and its level in the circulation is not, as the mature BNP only is the extent of coronary atheroscle- peptide, dependent on the peripheral number of BNP receptors. We aimed to test the hypothesis that asymptomatic left ventricular dysfunction (ALVD), as estimated by Nt-proBNP, would be rosis increased, the disease becomes clin- more prevalent in patients with type 2 diabetes without overt cardiovascular disease in compar- ical earlier and is more generalized in the ison with matched control subjects. coronary tree compared with the subjects without diabetes (3). Diabetes is also RESEARCH DESIGN AND METHODS — The study population consisted of 253 pa- more prevalent among patients with heart tients with type 2 diabetes and 230 matched control subjects aged 40–70 years without any overt failure. In the Framingham study, male heart disease from primary care centers in Western Finland and Southern Sweden. Nt-proBNP patients with diabetes had twice the risk was measured in plasma by competitive enzyme immunosorbent assay. and female patients five times the risk of a control population to develop heart fail- RESULTS — Patients with type 2 diabetes were shown to have higher Nt-proBNP values ure (4). At least partially, this could be (360.9 pmol/l [262.6–467.9]) than control subjects (302.7 pmol/l [215.4–419.2]) (P Ͻ 0.001). Nt-proBNP levels were independently related to diabetes after adjustment for age, sex, systolic explained by the increase in severity and and diastolic blood pressure, BMI, heart rate, drug treatment, serum creatinine, and cystatin C. incidence of ischemic heart disease among patients with diabetes. However, CONCLUSIONS — Our data suggest that the secretion of Nt-proBNP is increased in type 2 data from autopsy studies have suggested diabetic patients with no overt heart disease, suggesting that type 2 diabetes is associated with a that hearts from patients with diabetes higher prevalence of ALVD than hitherto thought. Nt-proBNP may thus serve as a screening also have an increased collagen content instrument to select patients with type 2 diabetes who could benefit from an echocardiographical (5). Moreover, patients with diabetes examination. have a disproportional increase in left ventricular mass independent of blood Diabetes Care 27:1929–1935, 2004 pressure (6–8). All of these factors may contribute to increased myocardial stiff- ness. This is especially important because he incidence and prevalence of type resents a severe threat to the population left ventricular hypertrophy in a meta- 2 diabetes increases worldwide. In with diabetes, macroangiopathy and sub- analysis has been associated with a 1.5- to T the adult population all over the sequent cardiovascular disease are the 3.5-fold increased risk of future cardio- world, the average prevalence for diabetes major causes of morbidity and mortality vascular morbidity and a 1.5- to 6.8-fold is estimated to be at least 4.0% (1). This in these patients. Screening for kidney increase of all-cause mortality (9). Thus, figure is predicted to double until the year and retinal complications is already an es- taken together, there are several mecha- 2015 (2). Although microangiopathy rep- tablished part of routine diabetes care to- nisms beside the more aggressive athero- ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● sclerosis that could explain why patients From the 1Department of Cardiology, University Hospital MAS, Lund University, Sweden; the 2Department with diabetes have a higher cardiac mor- of Endocrinology, University Hospital MAS, Lund University, Lund, Sweden; the 3Department of Clinical bidity and mortality. Chemistry, University Hospital in Lund, Lund University, Lund, Sweden; the 4Cardiovascular Research Brain natriuretic peptide (BNP) is a Group, Wallenberg Laboratory University Hospital MAS, Lund University, Lund, Sweden; and the 5Lund 32–amino acid peptide (10). It is synthe- Strategic Research Center for Stem Cell Biology and Cell Therapy, Lund, Sweden. sized predominantly in the left ventricle Address correspondence to Dr. Stefan Jovinge, Department of Cardiology, Ing 35 Univ Hosp MAS, S-205 02 Malmo¨, Sweden. E-mail: [email protected]. of the heart as the 108–amino acid pro- Received for publication 1 December 2003 and accepted in revised form 25 March 2004. hormone preproBNP (␥-BNP) (11–13). Abbreviations: ALVD, asymptomatic left ventricular dysfunction; BNP, brain natriuretic peptide; LVD, The hormone is a potent vasodilator and a left ventricular dysfunction; Nt-proBNP, NH2-terminal portion of the precursor of BNP; SBP, systolic blood natriuretic factor regulating salt and water pressure. A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion homeostasis. BNP is stored in the human factors for many substances. cardiac tissue mainly as BNP-32 with a © 2004 by the American Diabetes Association. lesser amount of the precursor pre-

DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 1929 Nt-proBNP elevation in type 2 diabetes

Table 1—Demographic data for the study population in accordance to most of the levels cited in the literature. Patients with type 2 diabetes Control subjects P Cystatin C assay Plasma cystatin C was measured by a fully n 253 230 automated particle-enhanced turbidimet- Age (years) 59.6 (54.3–63.9) 58.1 (51.7–63.4) 0.256 ric assay on undiluted samples (23,24) Sex (% women) 51.4 60.4 0.072 using a Behring BN ProSpec analyzer BMI (kg/m2) 28.3 (25.5–32.4) 25.8 (24.1–28.1) Ͻ0.001 (Dade Behring, Deerfield, IL) and a cali- SBP (mmHg) 140.0 (130.0–158.0) 130.0 (120.0–144.0) Ͻ0.001 brator (24) obtained from DakoCytoma- Diastolic blood pressure (mmHg) 84.0 (78.0–90.0) 80.0 (74.0–90.0) Ͻ0.003 tion (Glostrup, Denmark). Pulse (sϪ1) 70 (64–78) 66 (60–72) Ͻ0.001 Creatinine (␮mol/l) 84.0 (76.0–95.3) 86.0 (77.0–95.3) 0.352 Statistics Cystatin C (g/l) 1.05 (0.980–1.17) 1.09 (1.00–1.24) 0.034 Analyses were performed in statistical Cholesterol (mmol/l) 5.61 (5.03–6.48) 5.84 (5.21–6.73) 0.061 software package SigmaStat 2.0. Demo- LDL cholesterol (mmol/l) 3.56 (2.95–4.26) 3.86 (3.22–4.56) 0.007 graphic data were initially described as HDL cholesterol (mmol/l) 1.18 (0.99–1.41) 1.35 (1.16–1.67) Ͻ0.001 the median value of groups and the 25th Triglycerides (mmol/l) 1.70 (1.20–2.32) 1.20 (0.90–1.63) Ͻ0.001 and 75th percentiles. Significance of dif- Data are means (SD range). ferences between groups was tested by a Mann-Whitney rank-sum test with P Ͻ 0.05 considered as statistically signifi- proBNP. The circulating plasma forms of ganization definition from 1985 (21). cant. Due to the skewed distribution of BNP are BNP-32 and the NH2-terminal Kidney function was normal, and none of Nt-proBNP values, ln(Nt-proBNP) was portion proBNP (Nt-proBNP) (1-76) the patients had any nephropathy either used in a multiple logistic regression (14,15). Increased secretion of BNP and by report or by creatinine/cystatin C val- model with diabetic phenotype as out- Nt-proBNP occurs mainly with increased ues. No subjects were under drug treat- come. A ␹2 analysis was performed to test tension in the ventricular walls, decreased ment with digitalis or nitrates. The significance of frequency differences. oxygen supply, acute myocardial infarc- characteristics of the study population are tion, chronic cardiac heart failure, and in presented in Table 1. RESULTS — Patients with diabetes hypertrophy of the heart (16,17). Patient recruitment and inclusion had a higher median value of Nt-proBNP In a head-to-head comparison study visit was performed as described earlier (360.9 pmol/l [262.6–467.9]) than the by Hammerer-Lercher et al. (18), BNP for the Botnia study. Briefly, all patients population without diabetes (302.7 and Nt-proBNP were found to be superior underwent medical examination by a pmol/l [215.4–419.2]) (P Ͻ 0.001) (Fig. markers to Nt-pro atrial natriuretic pep- physician. A careful medical history was 1). The proportion of individuals with a tide in detecting left ventricular dysfunc- taken to obtain information about other Nt-proBNP value above the cutoff value tion (LVD). diseases (particularly hypertension, coro- (350 pmol/l) was significantly higher Cardiovascular death accounts for nary heart disease, myocardial infarction, (61.3 vs. 45.1%) among the diabetic pa- ϳ70% of the deaths among subjects with stroke, peripheral vascular disease, and tients than in the control group, accord- diabetes. The treatment of LVD, even endocrine disorders) and medication. ing to a ␹2 analysis (P Ͻ 0.001). when asymptomatic, is associated with a Body weight and height were measured First, a multiple logistic regression better prognostic outcome (19). This with subjects in light clothing without analysis model taking sex, age, pulse, study was performed to establish whether shoes. Three blood pressure recordings BMI, cystatin C, and systolic and diastolic asymptomatic LVD (ALVD), as estimated were obtained from the right arm while in blood pressure into account was used. by Nt-proBNP, is overrepresented in pa- the supine position after 30 min of rest at Then, the analysis was reperformed, tak- tients with type 2 diabetes compared with 5-min intervals, and their mean value was ing only age and sex into account. nondiabetic control subjects for both calculated. Blood samples were drawn The odds ratio (OR) for diabetes and groups without overt cardiovascular into Vacutainer tubes containing EDTA. 1 SD change in parameters was calcu- disease. Plasma was frozen at Ϫ80°C for the mea- lated. Systolic blood pressure (SBP), surement creatinine, cystatin C, and Nt- pulse, BMI, cystatin C, and ln(BNP) were RESEARCH DESIGN AND proBNP. all independently influencing the risk of METHODS — We compared plasma diabetic phenotype for 1 SD of change in concentrations of Nt-proBNP between parameters. For ln(Nt-proBNP), the OR 253 subjects with and 230 without type 2 Nt-proBNP assay was 1.60 (95% CI 1.26–2.03). This OR diabetes. None of the subjects had any Nt-proBNP was analyzed using a compet- was only mildly changed when the mul- known cardiovascular disease. The pa- itive enzyme immunosorbent assay de- tiple regression analysis was reperformed, tients were recruited from health care signed to measure the immunoreactive taking only sex and age into account 1.54 centers in the Botnia region in Western Nt-proBNP (Biomedica Laboratories, Vi- (1.25–1.90) (Table 2). Finland, Southern Finland, and Southern enna, Austria). The cutoff value for LVD Nt-proBNP values showed no corre- Sweden (20). Type 2 diabetes was diag- was set to 350 pmol/l, according to an lation to HbA1c, age, or SBP in the patients nosed according to the World Health Or- earlier study by Hughes et al. (22). This is with diabetes (data not shown). However,

1930 DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 Magnusson and Associates

function (25). Nt-proBNP is a split product from the BNP. It is more stable, and the circulating concentration is not dependent on the receptor population in the individual patient. An increase in BNP might also be dependent on a downregu- lation of the receptor population, as has been suggested for patients with nephropathy (26). This should not be the case for Nt-proBNP, which is solely elim- inated through glomerular filtration. In our study, the Nt-proBNP level was shown to be significantly elevated in the cohort of patients with diabetes. How- ever, patients with diabetes have a higher BMI, heart rate, and systolic and diastolic blood pressure than the control subjects, which might confound our results. In a multiple logistic regression model taking diabetes status as an outcome, ln(Nt- proBNP) was identified as an independent variable, even when the previously Figure 1—Distribution of Nt-proBNP values in the patient (n ϭ 253) and control (n ϭ 230) described possible confounders for BNP cohorts. P Ͻ 0.001 for difference between groups. levels, systolic and diastolic blood pressure, sex, age, pulse, BMI, and cystatin C high blood pressure has been implicated and age as independent. Relation of were taken into account. In addition, the OR was only mildly changed (and even as being a confounder to Nt-proBNP val- HbA1c, BMI, and SBP were for each factor ues. Therefore, we reperformed analyses individually analyzed by a linear regres- lower) if the multiple regression analysis of Nt-proBNP values between individuals sion in the patient group, with Nt- was reperformed with only sex and age above and below an SBP of 135 mmHg. proBNP as the dependent variable and taken into account. There was no statistical difference be- each of the factors used as independent There was in our material a consider- tween the two groups in the control pop- variables. able overlap between the two groups. ulation (P ϭ 0.483). There was, however, Nt-proBNP change over age in the This might partially be due to ALVD a statistical difference within the diabetes control population followed a regression among the control population. In the Rot- group, such that the patients with SBP curve Nt-proBNP ϭ 90.099 ϩ (4.044 ϫ terdam study, systolic dysfunction was re- Ͻ135 mmHg had higher Nt-proBNP val- age). Thus, for each year of increase in ported in 4% of the population aged ues (401.0 [290.5–504.9]) than their hy- age, the Nt-proBNP-level increased 4.04 55–95 years. However, 60% of these were pertensive (SBP Ͼ135 mmHg) peers (P Ͻ pmol/l. Interestingly enough, no such asymptomatic (27). Diastolic abnormal- 0.027). trend was seen in patients with diabetes. ity, as defined by the European Study The trend of change for Nt-proBNP to No differences in Nt-proBNP levels Group on Diastolic Heart Failure, is much age was estimated in a linear regression were detected between individuals on more prevalent, 11.1%, according to a with Nt-proBNP as a dependent variable ␤-blockers, diuretics, and angiotensino- substudy of the MONICA project. How- gen inhibitors compared with untreated ever, it differed between different age- subjects. However, individuals on Ca- groups, with 2.8% in those aged 25–35 Table 2—OR for diabetes outcome Ͼ blocker treatment had significantly years and 15.8% in subjects 65 years of higher Nt-proBNP values (417.4 [277.8– age (28). In previous studies, BNP has Parameters OR for 1 SD (95% CI) 530.1]) versus nontreated subjects (333.7 been shown to be an early determinant of ϭ both diastolic and systolic dysfunction. In SBP 1.86 (1.40–2.49) [229.3–444.8]) (P 0.015). However, if Mann-Whitney’s rank-sum test was animal models, BNP gene expression has DBP 0.791 (0.600–1.05) been shown to reflect ventricular and Pulse 1.53 (1.21–1.93) reperformed on Nt-proBNP values between the control subjects and patients atrial pressures (29). In that study, even BMI 1.86 (1.44–2.39) the animals with a mild compensated Cystatin C 0.809 (0.644–1.02) with diabetes with the omission of subjects on Ca-blockers, the differences be- heart failure had elevated BNP levels. ln(BNP) 1.60 (1.26–2.03) Thus, the Nt-ProBNP measurement, in ln(BNP)* 1.54 (1.25–1.90) tween groups were consistent and still highly significant (P Ͻ 0.001). addition to being an estimation of the Data are OR (95% CI) for 1 SD change in a multiple combined diastolic and systolic perfor- regression model with SBP, diastolic blood pressure mance, might also be a more sensitive (DBP), pulse, BMI, cystatin C, ln(BNP), age, and sex taken into account. *Representing estimation of OR CONCLUSIONS — BNP has been measurement of ventricular disturbances for 1 SD change only taking age and sex into shown to be elevated in early left ventric- than the echocardiographical examina- account. ular systolic as well as in diastolic dys- tion. This might explain why, in our ma-

DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 1931 Nt-proBNP elevation in type 2 diabetes

Figure 2—Levels of Nt-proBNP in relation to age in the control (A) and the patient (B) cohorts. The linear regression curve for the nondiabetic subjects is Nt-proBNP ϭ 90.099 ϩ (4.044 ϫ age). terial, 45% of the nondiabetic subjects Nt-proBNP (Fig. 1). The previously de- jects with mildly elevated Nt-proBNP lev- had elevated levels of Nt-proBNP. Despite ﬁned cutoff for our assay of 350 pmol/l els is warranted. this suggested sensitivity for ventricular might either be too sensitive or the true A linear regression model for age as an dysfunction with elevated levels among prevalence might be as high as depicted independent and Nt-proBNP as a depen- 45% of control subjects, the patients with (22,30–36). Further echocardiographical dent variable suggests an age-related in- diabetes had signiﬁcantly higher levels of characterization of healthy control sub- crease in Nt-proBNP of a 4.04-pmol/l

1932 DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 Magnusson and Associates increase per year in the control popula- diabetic group was 360.9 pmol/l as com- In conclusion, our data suggest that tion. This effect was not seen in the pared with 302.7 pmol/l in the control the secretion of Nt-proBNP, if paired with subjects with diabetes (Fig. 2). Several group. This 58.2-pmol/l difference, ac- an estimation of glomerular filtration rate, pathophysiological mechanisms might cording to the Nt-proBNP versus age re- is increased in patients with type 2 diabe- explain why an early deterioration of left gression equation in control subjects with tes compared with control subjects with- ventricular function could overrun the a 4.04 pmol/l per year coefficient, would out overt heart disease. Therefore, age effect on Nt-proBNP in the patients correspond to an age effect of 14.4 years. measurement of Nt-proBNP paired with with diabetes. Hearts from patients with Rakowski et al. (42) estimated that the cystatin C might be a simple screening diabetes have an increased collagen con- E/A value, as a functional measurement of tool to identify patients with diabetes at tent, as have been verified in autopsy diastolic function, passed 1.0 in patients risk for ventricular dysfunction requiring studies (5). Another possible mechanism with diabetes at the age of 56 years as further examination with echocardiogra- working from the very start of diabetes compared with 78 years for the control phy. However, this is a very early study could be decreased relaxation, and thus population. that needs confirmation in larger-scale diastolic dysfunction, of the myocardium Blood lipids of traditional risk value studies. because of ATP deficiency. The intracel- differed between groups as expected, lular glucose deficiency among patients with diabetic patients having higher trig- with diabetes leads to a higher use of free lycerides and lower HDL cholesterol than Acknowledgments— This study was sup- fatty acids through ␤-oxidation in the control subjects. In addition to the risk of ported by Swedish Research Council Grants myocardium. A sufficient amount of car- having diabetes, this gives an increased K2003-71x-13498-04A and K2002-71X- bohydrate breakdown is of great impor- risk of having coronary atheromatosis. 14042-02A, Swedish National Heart-Lung Foundation Grant 200141781, the Cornell tance for assuring an adequate function of However, in our material, diabetic pa- ϩ ϩ Research Foundation and Craaford Research the ion pumps, meaning Na /K -ATPase tients did have lower LDL cholesterol. 2ϩ Foundation, Swedish National Heart-Lung and Ca -ATPase, which maintains the This is probably due to the fact that dia- Foundation Grant 20020681, the Påhlssons right cardiomyocyte membrane potential betic subjects were selected to not have any Foundation, and the Malmo¨ University Hospi- and intracellular calcium transport that overt cardiovascular disease (Table 1). tal. triggers relaxation. In the diabetic heart, The testing for Nt-proBNP accompanied We thank Jan-Åke Nilsson (Department of this balance is disturbed, proposing a with cystatin C is, however, not designed Medicine, University Hospital MAS, Lund functional explanation to the impaired re- to differ between the atherosclerosis- University) for helpful statistical advice and laxation in the myocardium (37– 40). dependent and -independent decreased Prof. Jan Nilsson (Department of Medicine, These effects could be so strong that they cardiac performance. This is a strength University Hospital MAS, Lund University) for discussions and helpful consultation. would overrun the age effect in the dia- rather than a weakness of the assay. betic group. Nt-proBNP has been sug- Of all pharmacological compounds gested to be more age sensitive than BNP tested, Nt-proBNP values did not differ (41), but in our material, the patients with among treated and untreated subjects, References diabetes show no relation between Nt- with the exception of calcium channel 1. Ryde`n L, Malmberg K: Reducing the im- pact of the diabetics heart’s increased vul- proBNP levels and their age. The lack of blockers. In the Antihypertensive and nerability to cardiovascular disease. Dial age effect on the levels of Nt-proBNP seen Lipid-Lowering Treatment to Prevent Cardiovasc Med 5:5–20, 2000 in patients with diabetes speaks against Heart Attack study, Ca-channel blockers 2. King H, Aubert RE, Herman WH: Global this supposed disadvantage of Nt-proBNP have been suggested to increase the inci- burden of diabetes, 1995–2025: preva- when it is used for screening in this pop- dence of heart failure as compared with lence, numerical estimates, and projec- ulation. Thus, Nt-proBNP might be espe- diuretics. However, the diagnosis of heart tions. Diabetes Care 21:1414–1431, 1998 cially useful for screening among patients failure was not based strictly on scientific 3. Steiner G: Diabetes and atherosclerosis: with diabetes (Fig. 2). definitions in this study but was taken to- an overview. Diabetes 30:1–7, 1981 Taken together, both BNP and Nt- gether with our findings. Further studies 4. Kannel WB, McGee DL: Diabetes and car- proBNP serve as sensitive markers of are warranted to examine this possible re- diovascular disease: the Framingham study. JAMA 214:2035–2038, 1979 LVD, but the levels of both markers are lationship (43). 5. Butler R, MacDonald TM, Struthers AD, influenced not only by their rate of syn- This cross-sectional study has its in- Morris AD: The clinical implications of di- thesis but also their respective clearance herent disadvantages as being only a one- abetic heart disease. Eur Heart J 19:1617– rate. The clearance of BNP is more com- time observation of each individual. Still, 1627,1998 plex through glomerular filtration as well the study was able to detect a difference 6. Vanninen E, Mustonen J, Vainio P, Lan- as receptor-dependent mechanisms, between the two groups. Moreover, this is simies E, Uusitupa M: Left ventricular whereas Nt-proBNP is cleared solely by how it would be used under a screening function and dimensions in newly diag- glomerular filtration. Because the glomer- setting for the individual physician. Even nosed non-insulin-dependent diabetes ular filtration rate is easily controlled for, if the groups differ significantly, there is a mellitus. Am J Cardiol 70:371–378, 1992 e.g., by cystatin C, and receptor clearance considerable overlap. Therefore, the test 7. Palmieri V, Bella JN, Arnett DK, Liu JE, Oberman A, Schuck MY, Kitzman DW, rate cannot be estimated by today’s meth- would not be usable for a general screen- Hopkins PN, Morgan D, Rao DC, De- ods, a paired screening with Nt-proBNP ing. However, it could be used as a screen- vereux RB: Effect of type 2 diabetes mel- and cystatin C might be a preferred ing tool to separate patients with diabetes litus on left ventricular geometry and screening model in patients with diabetes. eligible for an echocardiographical exam- systolic function in hypertensive subjects: The median Nt-proBNP value in the ination. Hypertension Genetic Epidemiology Net-

DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 1933 Nt-proBNP elevation in type 2 diabetes

work (HyperGEN) study. Circulation 103: 19. Johnstone D, Limacher M, Rousseau M, 24:320–328, 2003 102–107, 2001 Liang CS, Ekelund L, Herman M, Stewart 29. Langenickel T, Pagel I, Hohnel K, Dietz R, 8. Grossman E, Shemesh J, Shamiss A, Tha- D, Guillotte M, Bjerken G, Gaasch W, et Willenbrock R: Differential regulation of ler M, Caroll J, Rosenthal T: Left Ventric- al.: Clinical characteristics of patients in cardiac ANP and BNP mRNA in different ular mass in diabetes-hypertension. Arch studies of left ventricular dysfunction stages of experimental heart failure. Am J Intern Med 152:1001–1004, 1992 (SOLVD). Am J Cardiol 70:894–900, Physiol Heart Circ Physiol 278:H1500– 9. Vakili BA, Okin PM, Devereux RB: Prog- 1992 H1506, 2000 nostic implications of left ventricular hy- 20. Groop L, Forsblom C, Lehtovirta M, 30. Hunt PJ, Yandle TG, Nicholls MG, Rich- pertrophy. Am Heart J 141:334–341, Tuomi T, Karanko S, Nissen M, Ehrn- ards AM, Espiner EA: The amino-terminal 2001 strom BO, Forsen B, Isomaa B, Snickars B, portion of pro-brain natriuretic peptide 10. Tateyama H, Hino J, Minamino N, Kan- Taskinen MR: Metabolic consequences of (Pro-BNP) circulates in the human gawa K, Ogihara T, Matsuo H: Character- a family history of NIDDM (the Botnia plasma. Biochem Biophys Res Commun 214: ization of immunoreactive brain natri- study): evidence for sex-specific parental 1175–1183, 2000 uretic peptide in human cardiac atrium. effects. Diabetes 45:1585–1593, 1996 31. Hunt PJ, Richards AM, Nicholls MG, Biochem Biophys Res Commun 166:1080– 21. World Health Organization: Diabetes Mel- Yandle TG, Doughty RN, Espiner EA: Im- 1087, 1990 litus: Report of a WHO Study Group. Ge- munoreactive amino-terminal pro-brain 11. Yandle TG, Richards AM, Gilbert A, neva, World Health Org., 1985 (Tech. natriuretic peptide (NT-PROBNP): a new Fisher S, Holmes S, Espiner EA: Assay of Rep. Ser., no. 727), p. 1–113. marker of cardiac impairment. Clin Endo- brain natriuretic peptide (BNP) in hu- 22. Hughes D, Talwar S, Squire IB, Davies JE, crinol (Oxf) 47:287–296, 1997 man plasma: evidence for high molecular Ng LL: An immunoluminometric assay 32. Richards AM, Nicholls MG, Yandle TG, weight BNP as a major plasma component for N-terminal pro-brain natriuretic pep- Frampton C, Espiner EA, Turner JG, in heart failure. J Clin Endocrinol Metab 76: tide: development of a test for left ventric- Buttimore RC, Lainchbury JG, Elliott JM, 832–838, 1993 ular dysfunction. Clin Sci (Lond) 96:373– Ikram H, Crozier IG, Smyth DW: Plasma 12. Mukoyama M, Nakao K, Hosoda K, Suga 380, 1999 N-terminal pro-brain natriuretic peptide S, Saito Y, Ogawa Y, Shirakami G, Jou- 23. Mussap M, Ruzzante N, Varagnolo M, Ple- and adrenomedullin: new neurohormo- gasaki M, Obata K, Yasue H, et al: Brain bani M: Quantitative automated particle- nal predictors of left ventricular function natriuretic peptide as a novel cardiac hor- enhanced immunonephelometric assay and prognosis after myocardial infarction. mone in humans: evidence for an exquis- for the routinary measurment of human Circulation 97:1921–1929, 1998 ite dual natriuretic peptide system, atrial cystatin C. Clin Chem Lab Med 36:859– 33. Richards AM, Doughty R, Nicholls MG, natriuretic peptide and brain natriuretic 865,1998 MacMahon S, Sharpe N, Murphy J, Es- peptide. J Clin Invest 87:1402–1412, 1991 24. Khyse-Andersen J, Schmidt C, Nordin G, piner EA, Frampton C, Yandle TG: 13. Tateyama H, Hino J, Minamino N, Kan- Andersson B, Nilsson-Ehle P, Lindstrom Plasma N-terminal pro-brain natriuretic gawa, Minamino T, Sakai K, Ogihara T, V, Grubb A: Serum cystatin C, deter- peptide and adrenomedullin: prognostic Matsuo H: Concentrations and molecular mined by a rapid, automated particle-en- utility and prediction of benefit from forms of brain natriuretic peptide in hanced turbidimetric method, is a better carvedilol in chronic ischemic left ventric- plasma. Biochem Biophys Res Commun 185: marker than serum creatinine for glomerular dysfunction: Australia-New Zealand 760–767, 1992 ular filtration rate. Clin Chem 40:1921– Heart Failure Group. J Am Coll Cardiol 37: 14. Hino J, Tateyama H, Minamino N, Kan- 1926, 1994 1781–1787, 2001 gawa K, Matsuo H: Isolation and iden- 25. Yu CM, Sanderson JE, Shum IO, Chan S, 34. Troughton RW, Frampton CM, Yandle tification of human brain natriuretic Yeung LY, Hung YT, Cockram CS, Woo TG, Espiner EA, Nicholls MG, Richards peptides in cardiac atrium. Biochem Bio- KS: Diastolic dysfunction and natriuretic AM: Treatment of heart failure guided by phys Res Commun 167:693–700, 1990 peptides in systolic heart failure: higher plasma aminoterminal brain natriuretic 15. Kambayashi Y, Nakao K, Mukoyama M, ANP and BNP levels are associated with peptide (N-BNP) concentrations. Lancet Saito Y, Ogawa Y, Shiono S, Inouye K, the restrictive filling pattern. Eur Heart J 355:1126–1130, 2000 Yoshida N, Imura H: Isolation and se- 17:1694–1702, 1996 35. Clerico A, Del Ry S, Maffei S, Prontera C, quence determination of human brain na- 26. Yano Y, Katsuki A, Gabazza EC, Ito K, Emdin M, Giannessi D: The circulating triuretic peptide in human atrium. FEBS Fujii M, Furuta M, Tuchihashi K, Goto H, levels of cardiac natriuretic hormones in Lett 259:341–345, 1990 Nakatani K, Hori Y, Sumida Y, Adachi Y: healthy adults: effects of age and sex. Clin 16. Omland T, Aakvaag A, Vik-Mo H: Plama Plasma brain natriuretic peptide levels Chem Lab Med 40:371–377, 2002 cardiac natriuretic peptide determination in normotensive noninsulin-dependent 36. Karl J, Borgya A, Gallusser A, Huber E, as ascreening test for the detection of pa- diabetic patients with microalbumineria. Krueger K, Rollinger W, Schenk J: Devel- tients with mild left ventricular impair- J Clin Endocrinol Metab 84:2353–2356, opment of a novel, N-terminal-proBNP ment. Heart 76:232–237, 1996 1999 (NT-proBNP) assay with low detection 17. Davis M, Espiner E, Richards G, Billings J, 27. Mosterd A, Hoes A, de Bruyne MC, Deck- limit. Scand J Clin Lab Invest Suppl 230: Town I, Neill A, Drennan C, Richards M, ers JW, Linker DT, Hofman A, Grobbee 177–181, 1999 Turner J, Yandle T: Plasma brain natri- DE: Prevalence of heart failure and left 37. Taegtmeyer H, McNulty P, Young ME: uretic peptide in assessment of acute dys- ventricular dysfunction in the general Adaption and maladaption of the heart in pnoea. Lancet 343:440–444, 1994 population: the Rotterdam Study. Eur diabetes. Part I. General concepts. Circu- 18. Hammerer-Lercher A, Neubauer E, Mul- Heart J 20:447–455,1999 lation 105:1727–1733, 2002 ler S, Pachinger O, Puschendorf B, Mair J: 28. Fischer M, Baessler A, Hense HW, Heng- 38. Young ME, McNulty P, Taegtmeyer H: Head-to-head comparison of N-terminal stenberg C, Muscholl M, Holmer S, Dor- Adaption and maladaption of the heart in pro-brain natriuretic peptide, brain natri- ing A, Broeckel U, Riegger G, Schunkert diabetes. Part II. Potential mechanisms. uretic peptide and N-terminal pro-atrial H: Prevalence of left ventricular diastolic Circulation 105:1861–1870, 2002 natriuretic peptide in diagnosing left ven- dysfunction in the community: results 39. Braunwald E, Bristow MR: Congestive tricular dysfunction. Clin Chim Acta 310: from a Doppler echocardiographic-based heart failure: fifty years of progress. Circu- 193–197, 2001 survey of a population sample. Eur Heart J lation 102:IV14–IV23, 2000

1934 DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 Magnusson and Associates

40. King LM, Opie LH: Glucose delivery is a mesnil JG, Honos G, Jue J, Koilpillai C, Echocardiography. J Am Soc Echocardiogr major determinant of glucose utilisation in Lepage S, Martin RP, Mercier LA, O’Kelly 9:736–760, 1996 the ischemic myocardium with a residual B, Prieur T, Sanﬁlippo A, Sasson Z, 43. Major outcomes in high-risk hypertensive ﬂow. Cardiovasc Res 39:381–392, 1998 Alvarez N, Pruitt R, Thompson C, Tom- patients randomized to angiotensin- 41. McCullough PA, Omland T, Maisel AS: linson C: Canadian consensus recom- converting enzym inhibitor or calcium B-type natriuretic peptides: a diagnostic mendations for the measurement and channel blocker vs diuretic: the Antihy- breakthrough for clinicians. Rev Cardio- reporting of diastolic dysfunction by pertensive and Lipid-Lowering Treatment vasc Med 4:72–80, 2003 0 echocardiography: from the Investigators to Prevent Heart Attack Trial (ALLHAT). 42. Rakowski H, Appleton C, Chan KL, Du- of Consensus on Diastolic Dysfunction by JAMA 288:2981–2997, 2002

DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 1935