Doi:10.1016/j.amjcard.2004.04.022

Brachial Artery Reactivity in
Asymptomatic Patients With Type 2
Diabetes Mellitus and Microalbuminuria
(from the Detection of Ischemia in
Asymptomatic Diabetics–Brachial Artery
Reactivity Study)
Georgios I. Papaioannou, MD, Richard L. Seip, PhD, Neil J. Grey, MD, Deborah Katten, RN, MPH, Amy Taylor, MS, Silvio E. Inzucchi, MD, Lawrence H. Young, MD, Deborah A. Chyun, PhD, Janice A. Davey, MSN, Frans J.Th. Wackers, MD, Ami E. Iskandrian, MD, Robert E. Ratner, MD, Evelyn C. Robinson, RN, Stella Carolan, RN, Samuel Engel, MD, and Microalbuminuria is a novel atherosclerotic risk factor in
DM (p ؍ 0.03). Endothelium-dependent vasodilation at 1
patients with type 2 diabetes mellitus (DM) and predicts
minute (p ؍ 0.01) and endothelium-independent vasodi-
future cardiovascular events. Endothelial dysfunction and
lation at 3 minutes (p ؍ 0.007) were significantly less in
systemic inflammation have been proposed as common
patients with microalbuminuria. In addition, 96% of pa-
links between microalbuminuria and cardiovascular dis-
tients with microalbuminuria and 76% of those with nor-
ease. However, no study has assessed the relation be-
moalbuminuria had impaired endothelium-dependent va-
tween microalbuminuria and vascular dysfunction as mea-
sodilation (<8%, p ؍ 0.01). Microalbuminuria was an
sured by brachial artery reactivity (BAR) in DM. We
independent predictor of endothelium-dependent vasodi-
evaluated 143 patients (85 men; mean age 60.0 ؎ 6.7
lation in the entire cohort (p ؍ 0.045) and after excluding
years) with DM (mean duration 8.2 ؎ 7.4 years) enrolled
patients on hormone replacement therapy (p ؍ 0.01).
in the Detection of Ischemia in Asymptomatic Diabetics
Levels of C-reactive protein were significantly higher in
study. Subjects were categorized as those with microalbu-
patients with microalbuminuria than in those with nor-
minuria (ratio of urinary albumin to creatinine 30 to 299
moalbuminuria (p ؍ 0.02). We conclude that in DM the
␮g/mg creatinine, n ؍ 28) and those with normoalbumin- presence of microalbuminuria is associated with impaired
uria (ratio of urinary albumin to creatinine 0 to 29.9
endothelium-dependent and endothelium-independent va-
␮g/mg creatinine, n ؍ 115). High-resolution ultrasound sodilations of the brachial artery and a higher degree of
BAR testing was used to measure endothelium-dependent
systemic inflammation. In addition, microalbuminuria is an
and endothelium-independent vasodilations. C-reactive
independent predictor of endothelial dysfunction in asymp-
protein was measured as a marker of systemic inflamma-
tomatic patients with DM, especially in the absence of
tion. Patients with microalbuminuria and normoalbumin-
hormone replacement therapy. ᮊ2004 by Excerpta
uria had similar baseline characteristics, with the exception
Medica, Inc.
that those with microalbuminuria had a longer duration of
(Am J Cardiol 2004;94:294–299)
Endothelial dysfunction occurs early in the athero- endothelium-dependent vasodilation (EDV) and endo-
thelium-independent vasodilation (EIV) of the bra- chial artery has been proposed as a method to nonin-vasively assess endothelial function and smoothmuscle function, Microalbuminuria has From the Cardiology Division, Henry Low Heart Center, Hartford been associated with increased cardiovascular events Hospital, Hartford and the University of Connecticut School of Medi- in the general population and in patients with type 2 cine, Farmington, Connecticut; the Sections of Endocrinology and diabetes mellitus Within the framework of Cardiovascular Medicine, Yale University School of Medicine, New the Detection of Ischemia in Asymptomatic Diabetics Haven, Connecticut; the Cardiovascular Division, University of Ala-bama at Birmingham School of Medicine, Birmingham, Alabama; the (DIAD) we investigated the relation between Medstar Clinical Research Institute, Washington, DC; and the Section impaired vascular reactivity and microalbuminuria in of Endocrinology, Norwalk Hospital, Norwalk, Connecticut. This study was supported by grants from the Hartford Hospital ResearchAdministration, Hartford, Connecticut, and Bristol-Myers Squibb Med-ical Imaging, North Billerica, Massachusetts. Manuscript received December 12, 2003; revised manuscript received and accepted April Patient population: The patient cohort consisted of
subjects enrolled in the DIAD study who underwent Address for reprints: Gary V. Heller, MD, PhD, Division of Cardi- ology, Hartford Hospital, 80 Seymour Street, Hartford, Connecticut assessment of BAR (DIAD-BAR substudy). The 06102. E-mail: gheller@harthosp.org.
DIAD study is a prospective, multicenter, randomized 2004 by Excerpta Medica, Inc. All rights reserved.
The American Journal of Cardiology Vol. 94 August 1, 2004 trial that evaluates the prevalence of adenosine tech- C-reactive protein had repeated measurements and netium-99m sestamibi myocardial perfusion imaging were excluded if there was evidence of infection or abnormalities in asymptomatic patients with type 2 systemic inflammation. Urinary albumin and creati- DM and its association with adverse clinical out- nine concentrations were determined by immunotur- comes. Eligible patients had asymptomatic type 2 bidimetric and kinetic methods, respectively. With DM, were 50 to 75 years old, and had no known these methods, the coefficients of variation for albu- coronary artery disease. Exclusion criteria were (1) min and creatinine were 2.7% and 3.5%, respectively.
known coronary artery disease, as documented by a Measurements of EDV and EIV: All measurements of
history of angina, myocardial infarction, percutaneous BAR were obtained in the morning after an overnight coronary intervention, or coronary artery bypass sur- fast, with medications withheld the morning of the gery; Q waves; new deep negative T waves or ST- study. Patients were also instructed to avoid caffeine- segment depression at rest on electrocardiogram; pos- containing products, smoking, and exercise Ն12 hours itive results on noninvasive stress test, or coronary before the test. Images were obtained with an Acuson angiography with documentation of coronary artery 10.0-MHz linear array transducer and an Aspen car- disease before recruitment; (2) signs or symptoms that diac ultrasound system (Acuson Corp, Elmwood Park, normally require cardiac evaluation, such as presence New Jersey) according to a standard technique for all of left bundle branch block, congestive heart failure, participating centers. After initial baseline measure- chest pain, or ventricular tachycardia; (3) any stress ments of brachial artery diameter, a blood pressure test or cardiac catheterization performed within 3 cuff was placed around the forearm distal to the seg- years before study enrollment; (4) significant nondia- ment of the artery scanned and inflated 60 mm Hg betic co-morbidity affecting life expectancy (e.g., ma- above the patient’s systolic blood pressure for 5 min- lignancy); (5) pregnancy; (6) known medical noncom- utes. After deflation, the brachial artery diameter was pliance with follow-up; and (7) history, clinical recorded at 1 minute and 3 minutes after occlusion.
findings, or treatment for significant bronchospasm.
After a 15-minute break, a second baseline measure- One hundred forty-nine patients from 5 participating ment of brachial artery diameter was recorded, and 0.4 centers were enrolled in the DIAD-BAR substudy.
mg of sublingual nitroglycerin was administered. Re- Institutional review committees from all centers ap- sponse of brachial artery diameter was recorded at 3 proved the study, and all patients gave informed and 5 minutes after administration of nitroglycerin.
Ten cardiac cycles were analyzed for each scan, and Study design: Baseline history, physical examina-
measurements were averaged. Brachial artery diame- tion, and laboratory tests were obtained at patient ter was measured at a fixed distance from an anatomic centers. Subsequently, the 149 patients who agreed to marker as the distance between the near and far in- participate underwent a baseline study of BAR. Sub- tima. EDV and EIV were calculated as percent max- jects were then categorized as those with microalbu- imal increases in artery diameter 1 minute and 3 minuria (ratio of urinary albumin to creatinine 30 to minutes after occlusion and 3 and 5 minutes after 299 ␮g/mg creatinine, n ϭ 28) and those with nor-moalbuminuria (ratio of urinary albumin to creatinine nitroglycerin administration, respectively. Examina- 0 to 29.9 ␮g/mg creatinine, n ϭ 115). Six subjects tion of BAR from Hartford Hospital and Yale Univer- were excluded from the final analysis (4 with mi- sity were recorded with CVI acquisition software croalbuminuria and 2 with unacceptable images of (version 2.1, Data Translation Inc., Marlboro, Massa- chusetts). Images from the other 3 participating cen- Clinical and laboratory measurements: The follow-
ters were recorded on tape. Two different interpreters ing clinical characteristics were collected for each at Hartford Hospital analyzed all scans independently patient: age, gender, smoking history, duration of DM, by using CVI Analysis software. The intra- and inter- type of DM therapy, history of hypertension and ther- observer variabilities in our laboratory were 1.1% and apy, history of hyperlipidemia and treatment, and body mass index. Systolic and diastolic blood pres- Statistical analysis: Results are presented as means
sures were defined as the mean of 3 measurements in Ϯ SD unless otherwise stated. Chi-square or Fisher’s the supine position. Blood and urine samples were exact test was used to compare categorical data. For obtained during morning hours after an overnight fast.
continuous variables, differences between patients This method is preferred because of the known diurnal with microalbuminuria and those with normal albu- variation in albumin Blood and urine spec- minuria were compared with a 2-sample independent imens were analyzed by a central laboratory (Lab- t test or Mann-Whitney U test. To study the correla- Corp, Raritan, New Jersey), with the exception of tion of urinary albumin:creatinine ratio with BAR, homocysteine and high-sensitivity C-reactive protein Spearman’s rank correlation test was used. Each vari- measurements, which were analyzed elsewhere (Lipo- able was evaluated for its association with EDV and Science, Raleigh, North Carolina). The Immulite 2000 EIV using univariate analysis. Variables with a p Homocysteine assay and high-sensitivity C-reactive value Ͻ0.1 were entered into a stepwise multivariate protein assay (Diagnostic Products Corporation, Hol- regression model to identify independent predictors of liston, Massachusetts) were used to determine total EDV and EIV. Results were considered statistically levels of plasma homocysteine and C-reactive protein, significant when p Ͻ0.05. Analysis was performed respectively. Patients with high levels (Ͼ10 mg/L) of with SPSS 10.1 (SPSS Inc., Chicago, Illinois).
CORONARY ARTERY DISEASE/BAR AND MICROALBUMINURIA IN TYPE 2 DM TABLE 1 Baseline Clinical and Biochemical Characteristics of Patients
Ϯ Values are means Ϯ SD.
*Data for C-reactive protein (mean Ϯ SEM) and homocysteine were available for 74 patients (microlbuminuria, n ϭ 12; normoalbuminuria, n ϭ 62).
TABLE 2 Baseline Medications of Patients
*Hormone replacement therapy data refer to the subpopulation of 53 postmenopausal women (microalbuminuria, n ϭ 11; normoalbuminuria, n ϭ 42).
minuria (p ϭ 0.06). The results of the BAR test are Baseline characteristics: The baseline clinical and
presented in EDV response was significantly biochemical characteristics of the entire study population less in patients with microalbuminuria than in those are listed in There were no significant differ- with normoalbuminuria at 1 minute (1.9 Ϯ 4.1% vs ences between groups regarding use of medications 4.9 Ϯ 5.7%, p ϭ 0.01) and 3 minutes (0.6 Ϯ 5.7% vs with the exceptions of thiazolidinediones being 2.8 Ϯ 5.1%, p ϭ 0.04) after occlusion. A similar used more frequently in patients with microalbuminuria significantly attenuated response was recorded for (p ϭ 0.05), and hormone replacement therapy was less EIV in patients with microalbuminuria compared with prevalent in patients with microalbuminuria than in those those with normoalbuminuria at 3 minutes (9.5 Ϯ 5.2% vs 13.4 Ϯ 7.2%, p ϭ 0.007) and 5 minutes (12.7 Relation between microalbuminuria and brachial ar-
Ϯ 5.5% vs 17.2 Ϯ 7.8%, p ϭ 0.005) after nitroglyc- tery reactivity: There was a trend for larger baseline
brachial artery diameter in patients with microalbu- Increasing levels of microalbuminuria had a sig- 296 THE AMERICAN JOURNAL OF CARDIOLOGYா
TABLE 3 Endothelium-dependent and Endothelium-independent Vasodilations of
the Brachial Artery in Patients With Microalbuminuria Versus Those With NormalAlbuminuria value, 27 of 28 patients with mi-croalbuminuria (96%) and 87 of 115 (76%) had abnormal EDV at 1minute (p ϭ 0.01).
Microalbuminuria and C-reactive
protein: For the subgroup of patients
with microalbuminuria (6.3 Ϯ 1.4mg/L) compared with those withnormoalbuminuria (3.9 Ϯ 0.5 mg/L, TABLE 4 Multivariate Linear Regression Model of Endothelium-dependent
Vasodilation as a Dependent Variable (n ϭ 143, R2 ϭ 0.258).
on hormone replacement therapy, be-cause hormone replacement therapy has been shown to increase C-reac-tive protein patients with †The standardized coefficient ␤ describes the unit-independent contribution of the independent vari- R2 ϭ square of the multivariate correlation coefficient.
patients with microalbuminuria, per-centages of subjects with low- (Ͻ1mg/L), moderate- (1 to 3 mg/L), and TABLE 5 Multivariate Linear Regression Model of Endothelium-independent
Vasodilation as a Dependent Variable (n ϭ 143, R2 ϭ 0.247) minuria. The percentage of high-riskpatients was significantly higher inthe group with microalbuminuria nificant inverse correlation with EDV at 1 minute (r ϭ than in the group with normoalbuminuria (82% vs Ϫ0.20, p ϭ 0.02) and EIV at 3 minutes (r ϭ Ϫ0.27, 40%, p ϭ 0.012), whereas the percentage of low-riskp ϭ 0.001). With stepwise multivariate regression patients was significantly lower (0% vs 28%, p ϭ analysis, significant independent predictors of EDV were baseline diameter, gender, current estrogen re-placement therapy, age, hemoglobin A1 , and pres- DISCUSSION
The aim of this study was to investigate the relation uria (standardized ␤ coefficient Ϫ0.22, p ϭ 0.01) and between microalbuminuria and BAR in asymptomatic age (standardized ␤ coefficient Ϫ0.20, p ϭ 0.02) were patients with DM. Our results demonstrated that pa- the only independent predictors of EDV after exclud- tients with microalbuminuria have significant impair- ing patients on hormone replacement therapy (n ϭ ment of EDV and EIV compared with those with 119). Data concerning independent predictors of EIV normoalbuminuria and a higher degree of systemic are presented in Baseline diameter (standard- inflammation. These data associated the presence of ized ␤ coefficient Ϫ0.30, p ϭ 0.001) and microalbu- microalbuminuria in patients with DM with impaired minuria (standardized ␤ coefficient Ϫ0.20, p ϭ 0.026) vascular reactivity secondary to smooth muscle dys- were independent predictors of EIV after excluding function rather than to endothelial dysfunction alone.
patients on hormone replacement therapy.
Microalbuminuria and brachial artery reactivity: In
The exact value of EDV that represents a normal patients with type 2 DM, the presence of microalbu- endothelium-dependent response has not been estab- minuria may necessitate screening for vascular disease and aggressive interventions to decrease cardiovascu- CORONARY ARTERY DISEASE/BAR AND MICROALBUMINURIA IN TYPE 2 DM lar In contrast, endothelial dysfunction, at populations, including the Framingham Heart Study least in the coronary circulation, predicts cardiovas- and the Cardiovascular Health Study, may determine whether testing of BAR identifies patients at risk for for microalbuminuria and cardiovascular disease in developing coronary artery disease and whether it is a Several studies have suggested that endothelial practical clinical tool. Despite a lack of large epide- dysfunction precedes the onset of microalbuminuria miologic studies, impaired endothelial function of bra- in patients with type 1 DIn type 2 DM, in- chial arteries is present in patients with coronary risk flammatory markers of endothelial dysfunction, mi- factors without evidence of coronary disease and im- croalbuminuria, and risk of death are parallel and progress with Consistent with these results, our study suggests that impaired vascular reactivity may study correlates endothelial dysfunction with increas- identify subjects with normoalbuminuria earlier in the ing levels of microalbuminuria. The attenuated EIV atherosclerotic process as candidates for more aggres- response in subjects with microalbuminuria also sug- sive medical therapy. However, such a recommenda- gests vascular smooth muscle dysfunction due to in- tion cannot be made at the moment because testing of creased inactivation of nitric oxide or decreased reac- BAR is a research rather than a clinically available tivity of the vascular smooth muscle to nitric tool and requires the care and precision of a well- The exact values of normal brachial EDV and EIV have not been established mainly due to differences in Study limitations: Our study is limited by its cross-
age, gender, and In patients with periph- sectional design and the absence of a control group. In eral vascular disease, an EDV Ͻ8.1% was associated addition, the interpretation of the results of EDV with ninefold increases in rates of morbidity and mor- should be viewed with caution because EIV was also tality compared with those with EDV of ՆIn impaired in subjects with microalbuminuria compared healthy nonsmokers, mean EDV at 1 minute has been with those with normoalbuminuria. However, the im- reported to be and in our laboratory, an EDV paired EDV in 76% of patients with normal albumin- 8% at 1 minute after lower arm occlusion was uria was associated with a “preserved” EIV response considered a normal response in healthy subjects of similar age and body size. With this cut-off point, 76%of patients with normal albuminuria had EDV of Ͻ8%, thus underscoring the concept that endothelial 1. Ludmer PL, Selwyn AP, Shook TL, Wayne RR, Mudge GH, Alexander RW,
dysfunction may precede the development of mi- Ganz P. Paradoxical coronary vasoconstriction induced by acetylcholine in ath- erosclerotic coronary arteries. N Engl J Med 1986;315:1046 –1051.
Coronary risk factors and their treatment may af- 2. Suwaidi JA, Hamasaki S, Higano ST, Nishimura RA, Holmes DR Jr, Lerman
A. Long-term follow up of patients with mild coronary artery disease and
fect endothelium-mediated response in the brachial endothelial dysfunction. Circulation 2000;101:948 –954.
3. Corretti MC, Anderson TJ, Benjamin EJ, Celermajer D, Charbonneau F,
differences, aside from a trend for higher body mass Creager MA, Deanfield J, Drexler H, Gerhard-Herman M, Herrington D, et al.
Guidelines for the ultrasound assessment of endothelium-dependent flow-medi- index and hemoglobin A1 in patients with microalbu- ated vasodilation of the brachial artery. A report of the International Brachial minuria. Active treatments that may favor endothelial Artery Reactivity Task Force. J Am Coll Cardiol 2002;39:257–265.
function were also similar in the 2 groups, with the 4. Borch-Johnsen K, Feldt-Rasmussen B, Strandgaard S, Schroll M, Jensen JS.
Urinary albumin excretion. An independent predictor of ischemic heart disease.
exception of thiazolidinediones and hormone replace- Arterioscler Thromb Vasc Biol 1999;19:1992–1997.
ment therapy. Hormone replacement therapy has been 5. Hillege HL, Fidler V, Diercks GF, van Gilst WH, de Zeeuw D, van Veldhuisen
reported to improve brachial artery endothelial func- DJ, Gans RO, Janssen WM, Grobbee DE, de Jong PE. Prevention of Renal andVascular End Stage Disease (PREVEND) Study Group. Urinary albumin excre- tion in postmenopausal women with or without coro- tion predicts cardiovascular and noncardiovascular mortality in general popula- nary artery Nevertheless, in multivariate tion. Circulation 2002;106:1777–1782.
6. Mattock MB, Morrish NJ, Viberti G, Keen H, Fitzgerald AP, Jackson G.
regression analysis, microalbuminuria was an inde- Prospective study of microalbuminuria as predictor of mortality in NIDDM.
pendent predictor of EDV, including or excluding subjects on hormone replacement therapy.
7. Wackers FJTh, Young LH, Inzucchi SE, Chyun DA, Davey JA, for the DIAD
Microalbuminuria and C-reactive protein:
Investigators. Detection of ischemia in asymptomatic diabetics: preliminary re- sults of the DIAD study. J Am Coll Cardiol 2003;41(suppl A):409A.
portant observation from our study was that subjects 8. American Diabetes Association. Diabetic nephropathy: position statement.
with microalbuminuria had significantly higher serum Diabetes Care 2003;26(suppl 1):S94 –S98.
9. Gokce N, Keaney JF Jr, Hunter LM, Watkins MT, Nedeljkovic ZS, Menzoian
C-reactive protein levels compared with subjects with JO, Vita JA. Predictive value of noninvasively determined endothelial dysfunc- normal albuminuria. Microalbuminuria is associated tion for long-term cardiovascular events in patients with peripheral vascular with a greater degree of systemic inflammation as disease. J Am Coll Cardiol 2003;41:1769 –1775.
10. Gaenzer H, Neumayr G, Marschang P, Sturm W, Kirchmair R, Patsch JR.
measured by various plasma Increased lev- Flow-mediated vasodilation of the femoral and brachial artery induced by exer- els of C-reactive protein are also associated with im- cise in healthy nonsmoking and smoking men. J Am Coll Cardiol 2001;38:1313– 1319.
11. Ridker PM. Clinical application of C-reactive protein for cardiovascular
our study with respect to C-reactive protein are con- disease detection and prevention. Circulation 2003;107:363–369.
sistent with the notion that the presence of microalbu- 12. Ridker PM, Buring JE, Cook NR, Rifai N. C-reactive protein, the metabolic
minuria correlates with higher states of systemic syndrome, and risk of incident cardiovascular events: an 8-year follow-up of14,719 initially healthy American women. Circulation 2003;107:391–397.
13. Yudkin JS. Coronary heart disease in diabetes mellitus: three new risk factors
Potential clinical applications: Assessment of BAR
and a unifying hypothesis. J Intern Med 1995;238:21–30.
14.
has the potential to be a preclinical marker of cardio- Stehouwer CD, Fischer HR, van Kuijk AW, Polak BC, Donker AJ. Endo- thelial dysfunction precedes development of microalbuminuria in IDDM. Diabe- vascular Ongoing studies in several large 298 THE AMERICAN JOURNAL OF CARDIOLOGYா
15. Lekakis J, Papamichael C, Anastasiou H, Alevizaki M, Desses N, Souvat-
19. Kuvin JT, Patel AR, Karas RH. Need for standardization of noninvasive
zoglou A, Stamatelopoulos S, Koutras DA. Endothelial dysfunction of conduit assessment of vascular endothelial function. Am Heart J 2001;141:327–328.
arteries in insulin-dependent diabetes mellitus without microalbuminuria. Car- 20. Anderson TJ. Assessment and treatment of endothelial dysfunction in hu-
diovasc Res 1997;34:164 –168.
mans. J Am Coll Cardiol 1999;34:631–638.
16. Dogra G, Rich L, Stanton K, Watts GF. Endothelium-dependent and inde-
21. Haines CJ, Yim SF, Sanderson JE. The effect of continuous combined
pendent vasodilation studied at normoglycemia in type 1 diabetes mellitus with hormone replacement on arterial reactivity in postmenopausal women with es- and without microalbuminuria. Diabetologia 2001;44:593–601.
tablished angina pectoris. Atherosclerosis 2001;59:467–470.
17. Stehouwer CD, Gall MA, Twisk JW, Knudsen E, Emeis JJ, Parving HH.
22. Fichtlscherer S, Rosenberger G, Walter DH, Breuer S, Dimmeler S, Zeiher
Increased urinary albumin excretion, endothelial dysfunction, and chronic low- AM. Elevated C-reactive protein levels and impaired endothelial vasoreactivity in grade inflammation in type 2 diabetes: progressive, interrelated, and indepen- patients with coronary artery disease. Circulation 2000;102:1000 –1006.
dently associated with risk of death. Diabetes 2002;51:1157–1165.
23. Vogel RA. Coronary risk factors, endothelial function and atherosclerosis: a
18. Williams SB, Cusco JA, Roddy MA, Johnstone MT, Creager MA. Impaired
review. Clin Cardiol 1997;20:426 –432.
nitric oxide-mediated vasodilation in patients with non–insulin-dependent diabe- 24. Bhagat K, Hingorani A, Vallance P. Flow associated or flow mediated
tes mellitus. J Am Coll Cardiol 1996;27:567–574.
dilatation? More than just semantics. Heart 1997;78:7–8.
CORONARY ARTERY DISEASE/BAR AND MICROALBUMINURIA IN TYPE 2 DM

Source: http://www.epemvatikikardiologia.gr/wp-content/uploads/2013/03/dimosiefseis/5.pdf