HM Medical Clinic


Coronary anatomy and left ventricular ejection fraction in patients with type 2 diabetes admitted for elective coronary angiography

Catheterization and Cardiovascular Interventions 62:432– 438 (2004)
Coronary Anatomy and Left Ventricular Ejection Fraction
in Patients With Type 2 Diabetes Admitted for Elective
Peter Ammann,1,2* MD, Hanspeter Brunner-La Rocca,2 MD, Thomas Fehr,3 MD,
Thomas Mu¨nzer,4 MD, Markus Sagmeister,3 MD, Walter Angehrn,1 MD, and Hans Rickli,1 MD
Patients with diabetes mellitus (DM) have more severe coronary artery disease and a two- to
fourfold higher risk for myocardial infarction and death as compared to patients without DM.
In this study, we analyzed coronary anatomy, left ventricular ejection fraction, and cardiac
risk factors in patients with DM referred for coronary angiography and compared them with
findings in nondiabetic patients. Coronary anatomy was assessed in a total of 6,234 patients
and left ventricular ejection fraction in a subset of 4,767 (76.5%) patients. Diabetic patients
641) were older (60.8 9.6 vs. 58.5 10.5 years; P < 0.0001) and had higher rates of
hypertension (65% vs. 47%; P
< 0.0001). Three-vessel disease (DM 44.7% vs. no DM 25.4%; P <
0.0001) and reduced left ventricular ejection fraction (DM 58.4%
15.2 vs. no DM 63.9% 13.2;
< 0.0001) were significantly associated with DM. After adjustment for age and other vascular
risk factors, the presence of DM was associated with a higher atherosclerotic burden. We
conclude that advanced coronary heart disease and left ventricular dysfunction are highly
prevalent in diabetic patients, independent of age and other cardiovascular risk factors.
Thus, cardiac assessment in diabetic patients should, in addition to optimal diabetic control,
involve screening for left ventricular dysfunction. Catheter Cardiovasc Interv 2004;62:
432– 438.

2004 Wiley-Liss, Inc.
Key words: diabetes; coronary angiography; left ventricular ejection fraction; sex; athero-
sclerosis; coronary artery disease

yet studied in comparison with nondiabetic patients,consecutively referred to our center between 1990 and An epidemic increase of patients suffering from type 2 2000. Additionally, we proposed that diabetic patients diabetes mellitus (DM) has been observed in the United might have an impaired left ventricular ejection frac- States and most of the Western countries over the last tion as compared to nondiabetics as an important rea- decade [1,2]. Diabetes is associated with a two- to fourfold son for their higher mortality rate beyond the risk higher risk for myocardial infarction and death [3]. The factors for coronary artery disease.
higher mortality rate observed in patients with DM cannotonly be explained by a higher incidence of other risk factorsfor coronary artery disease such as smoking, hypertension,and hypercholesterolemia [3], with an incidence of macro- vascular complications twice that of microvascular disease Division of Cardiology, Kantonsspital St. Gallen, St. Gallen, [4]. On the other hand, many studies of patients with cor- 2Division of Cardiology, University Hospital Basel, Basel, onary artery disease, including a high percentage of DM patients, have shown improved outcome with optimal med- 3Department of Internal Medicine, Kantonsspital St. Gallen, ical therapy [5–11]. Taken together, these studies underline St. Gallen, Switzerland4 the importance of DM as a major risk factor for coronary Center for Geriatric Medicine and Rehabilitation Bu¨rgerspital, St. Gallen, Switzerland artery disease and the potential for therapeutic interventionsin these patients.
*Correspondence to: Dr. Peter Ammann, Department of Internal Med- However, previous studies have provided controver- icine, Division of Cardiology, Kantonsspital St. Gallen, CH-9007 sial results regarding the association between the se- St. Gallen, Switzerland. E-mail: [email protected] verity of coronary artery disease and the presence of Received 18 September 2003; Revision accepted 2 March 2004 DM [12–15]. Therefore, the aim of the present studywas to assess the prevalence and severity of coronary DOI 10.1002/ccd.20135 artery disease in the largest cohort of diabetic patients Published online in Wiley InterScience (
2004 Wiley-Liss, Inc.
Coronary Anatomy and LVEF
TABLE I. Characteristics of the Study Population
Study Population and Technical Procedures
Between January 1990 and December 1999, data from a total of 6,234 patients were collected with a computer- assisted registry of all patients referred for coronary angiography to our hospital. Coronary angiography was performed by nine trained cardiologists in all patients Body mass index,a kg/m2 using right and/or left Judkins angiographic procedures.
History of smoking Left ventricular angiograms were performed in a subset Arterial hypertensionc of 4,767 (76.5%) patients in the right anterior oblique Family history of CAD projection. In the remaining patients, angiograms were Angina prior to angiography not done due to unstable clinical conditions such as acute No angina pectoris coronary syndrome. All patients provided written in- Stable angina pectoris Unstable angina pectoris formed consent for coronary angiographies and the local History of myocardial infarction ethics committee approved the analysis of registry data.
aThe body mass index is calculated as weight in kg divided by the squareof the height in m; it is indicated as mean ⫾ SD.
Quantitative Coronary Study
bDefined as total cholesterol ⬎ 200 mg/dl.
Coronary anatomy was assessed by at least two senior cDefined as blood pressure ⬎ 140/90 mm Hg.
cardiologists using a simplified scoring system derived fromDortimeter et al. [16]. Nine coronary segments were ana-lyzed: the left main artery, four segments of the left descen- square of height in m, was used as a measure of central dent artery (proximal, middle, distal segment, and side obesity. Hypercholesterolemia was diagnosed if fasting branches), two segments of the circumflex artery, and two blood samples of total cholesterol were above 200 mg/dl segments of the right coronary artery. Relevant coronary (5.0 mmol/L); hypertension was defined as repeated blood artery stenosis was defined as stenosis ⬎ 50%. In addition, pressure measurements above 140/90 mm Hg.
a more sophisticated scoring system was used to assess theatherosclerotic burden of each individual patient as de-scribed previously [14,17]: each of the nine segments was graded as follows: grade 0, angiographically normal seg- All data are expressed as mean ⫾ standard deviation ment; grade 1, ⬍ 50% stenosis; grade 2, 50–75% stenosis;grade 3, 75–99% stenosis; and grade 4, occlusion. Three or median with 25th and 75th percentile and as per- scores were used to describe coronary atherosclerosis. The centages or frequencies as appropriate. The Fisher's coronary score was defined by the number of coronary exact test was used to compare categorical data, num- arteries exhibiting a stenosis greater than 75% (grades 3 and ber of vessels involved, and gender differences in 4). Stenosis of the left main coronary artery ⬎ 50% was diabetic and nondiabetic patients (Tables I and II, Fig.
considered a two-vessel disease. The extent score was de- 1). Mann-Whitney statistics were used to compare fined as the number of segments exhibiting stenoses coronary, extent, and atherosclerotic scores, left ven- greater ⬎ 50% (grade 2) adjusted to the nine-segment tricular ejection fraction, as well as for calculation of coronary model. The atherosclerotic score was calculated as continuous data in diabetic and nondiabetic patients the average severity of all analyzable segments. Left ven- (Tables I and III, Fig. 2).
tricular function was assessed using computerized calcula- Ordinal multivariate regression analysis was used to tions of angiographic data. According to the left ventricular test independence of risk factors for the coronary artery ejection fraction, four groups of patients were identified: disease (CAD) scores in multivariate analysis. Ordinal patients with normal ejection fraction (⬎ 55%), patients regression was used instead of linear regression because with slightly impaired ejection fraction (41–55%), patients the number of categories and differences between the with moderately impaired left ventricular ejection fraction categories were not equivalent throughout the whole (30 – 40%), and patients with severely impaired ejection range of the CAD scores. Multivariate logistic regression fraction (⬍ 30%).
analysis was used to calculate the impact of different riskfactors and angiographic findings on impaired left ven- Assessment of Risk Factors
tricular ejection fraction (Table V). A two-tailed P of Clinical diagnosis of type 2 diabetes was established by ⬍ 0.05 was considered statistically significant. A com- reviewing the medical records of all patients. Body mass mercially available statistical program was used for anal- index (BMI), defined as body weight in kg divided by the ysis (SPSS 11.0 for Windows).
Ammann et al.
TABLE II. Percentages of Segments With Stenoses > 50% in Patients With Type 2 Diabetes Compared to Patients Without
Diabetes in Respect to the Coronary Artery Anatomy

P between male, female, Diabetes (n ⫽ 641) No diabetes (n ⫽ 5,593) aP values compare the distribution of coronary artery stenoses in each vessel between diabetic and nondiabetic patients overall and in each genderseparately.
the presence of clinical symptoms prior to coronary an-giography nor the history of a prior myocardial infarctiondiffered between patients with or without diabetes.
Results of coronary anatomy summarized by sex and coronary stenoses ⬎ 50% are presented in Figure 1. Inde-pendent of sex, diabetic patients tended to have more oftentwo-vessel (DM 24.5% vs. no DM 22.4%; P ⫽ 0.09) andsignificantly more often three-vessel disease (DM 44.7% vs.
no DM 25.4%; P ⫽ 0.002). In contrast, one-vessel diseasewas more frequent in patients without diabetes (no DM24.6% vs. DM 17.1%; P ⫽ 0.02). In the angiographicsubgroups (zero-vessel, one-vessel, two-vessel, and three-vessel disease), we found no gender differences between Coronary artery anatomy in respect to sex and the pres-
diabetic and nondiabetic patients (Fig. 1).
ence of one-, two-, and three-vessel disease or normal coronary
Distribution of segments with stenoses ⬎ 50% in the left arteries. In the zero-, one-, two-, and three-vessel disease groups,
anterior descending artery (LAD), the circumflex (RCx), statistical differences between diabetic and nondiabetic patients
were as follows: for zero-vessel disease, P
0.91; one-vessel, P
and right coronary artery (RCA) is presented in Table II.
0.02; two-vessel, P 0.09; and three-vessel, P 0.002. Gender
Male patients with diabetes presented significantly more differences between diabetic and nondiabetic patients in the ze-
often with ⬎ 50% stenoses in the LAD as compared to ro-, one-, two-, and three-vessel disease groups were as follows:
patients without diabetes, whereas in women, no significant for female, zero-vessel disease, P 0.0001; one-vessel, P 0.76;
difference could be demonstrated. Furthermore, men had two-vessel, P 0.09; and three-vessel P 0.0001; for male, zero-
vessel, P
0.0001; one-vessel, P 0.0001; two-vessel, P 0.57;
significantly higher rates of RCx stenosis. A similar trend and three-vessel, P 0.0001.
was seen regarding RCA stenoses; however, the differencefailed to reach statistical significance.
Differences in coronary atherosclerosis scores be- tween patients with and without diabetes are shown in Baseline Patient Data and Frequencies of Risk
Table III. Diabetic patients presented significantly higher coronary, extent, and atherosclerotic scores as compared Baseline clinical data of the study population are pre- to patients without diabetes. In patients without diabetes, sented in Table I. Six hundred forty-one (10.3%) of the men had a significantly higher atherosclerotic burden 6,234 patients were diabetic. Patients with diabetes were than women. However, in diabetics we could not dem- older, had a higher body mass index, and were more onstrate such a difference except for a slightly higher frequently hypertensive, but smoked less often. Neither atherosclerotic score (Table III). After adjustment for Coronary Anatomy and LVEF
TABLE III. Differences in Coronary Atherosclerosis Scores Between Women and Men With and Without Diabetes*
Diabetes (n ⫽ 641) No Diabetes (n ⫽ 5593) Atherosclerotic score 1.4 (0.67–1.78) 1.4 (0.78–1.89) 0.89 (0.33–1.44) *For each patient, a coronary, extent, and atherosclerotic score was calculated. Afterward, the median value and interquartile ranges for each score werecalculated and presented. Note the highly significant gender differences in the coronary, extent, and atherosclerotic scores in nondiabetic patients.
Concerning coronary and extent score, these differences are blunted by the presence of diabetes. All score differences between patients with and withoutdiabetes are highly significant (P ⬍ 0.0001). Interquartile range in parentheses.
TABLE IV. Multivariate Ordinal Regression Analyses of Risk
Factors for Prediction of Various Scores of
Atherosclerosis (n
Overall mean left ventricular ejection fraction is signif-
icantly worse in diabetic as compared to nondiabetic patients
(58.4% 15.2 vs. 63.9% 13.2; P < 0.0001). Patients are
divided into different risk groups of normal ( 50%), slightly
Atherosclerotic score impaired (41– 49%), moderately impaired (30 – 40%), and se-
verely impaired (< 30%) left ventricular ejection fraction.
age, gender, and other cardiovascular risk factors, these differences remained statistically significant (Table IV).
*Independent predictors of scores are shown. Predictors not statistically In fact, the effect of diabetes on all the scores was not significant in multivariate analysis are not depicted. B denotes parameterestimate of each variable in ordinal regression model. Wald denotes im- significantly influenced by the covariates.
portance of each variable as predictor of the scores (i.e., severity ofcoronary artery disease).
Left Ventricular Ejection Fraction
Diabetic patients referred to coronary angiography had a significantly reduced left ventricular ejection fraction as ejection fraction between patients with and without diabetes compared to patients without diabetes (58.4% ⫾ 15.2 vs.
(data not presented).
63.9% ⫾ 13.2; P ⬍ 0.0001). Dividing patients into differentrisk groups of sudden cardiac death according to the studies of implantable defibrillators in coronary artery disease [5,6],we found a significantly impaired left ventricular ejection To our knowledge, this is the largest angiographic study fraction in the quartiles below 50% ejection fraction in that systematically compared left ventricular ejection frac- diabetic compared to nondiabetic patients (P ⫽ 0.009; tion and coronary artery anatomy in more than 6,000 pa- Fig. 2). Additionally, the presence of diabetes mellitus was tients with and without diabetes. The present study shows the second most independent risk factor for decreased left that, after myocardial infarction, the presence of diabetes ventricular ejection fraction (OR ⫽ 2.19; CI ⫽ 1.75–2.75; itself is the strongest independent risk factor for left ven- P ⬍ 0.0001) after myocardial infarction (OR ⫽ 3.68; CI ⫽ tricular dysfunction. Additionally, we found an increased 3.06 – 4.42; P ⬍ 0.0001) in multivariate analysis (Table V).
rate of advanced atherosclerosis in patients with DM, inde- No gender differences could be observed in left ventricular pendent of age, sex, and other cardiovascular risk factors Ammann et al.
TABLE V. Multivariate Logistic Regression Analysis of
phy in patients with diabetes are confirmed in two re- Different Risk Factors and Angiographic Findings Associated
cently published investigations by Ledru et al. [14] and With Impaired Left Ventricular Ejection Fraction (< 50%)*
Natali et al. [22], as well as two epidemiological studies [23,24] that also found a more severe and diffuse coro- angiographic finding nary atherosclerosis in diabetic patients. The higher rates of diffuse disease found in these studies is in line with the higher frequency of three-vessel disease and the presence 3.68 (3.06–4.42) of peripheral coronary artery disease in our study popu- Diabetes mellitus 2.19 (1.75–2.75) 0.94 (0.92–0.96) lation. In contrast to the study of Natali et al. [22], we History of smoking 1.26 (1.07–1.49) were not able to detect a sex-specific pattern in coronary Atherosclerotic score 1.12 (1.00–1.25) atherosclerosis in diabetic patients. The most likely ex- *Wald denotes the importance of each variable as predictor of the risk planation for this difference is the much higher number factors for left ventricular dysfunction. Predictors not statistically signifi- of diabetic patients investigated in the present study. This cant in multivariate analysis are not depicted.
is supported by a larger angiographic study from Meli-donis et al. [25], who also found no sex differences indiabetic patients concerning coronary artery anatomy.
such as hypercholesterolemia, hypertension, positive family Our finding of such a remarkable amount of left ventric- history, and smoking. Patients with diabetes presented with ular dysfunction in patients with diabetes is astonishing. The a higher percentage of two- and three-vessel disease defined significantly higher rate of coronary artery disease in dia- as coronary artery stenoses ⬎ 50% when compared to betic patients suggests that decreased left ventricular ejec- nondiabetics. Interestingly, in diabetic patients, atheroscle- tion fraction in diabetics may be due to more severe coro- rotic burden and severity scores were considerably less nary artery disease. The fact that myocardial infarction was different between sexes when compared to patients without the strongest predictor for left ventricular dysfunction in diabetes. This finding indicates that the protective effects of multivariate analysis and the results of a large autopsy study the development of atherogenesis of female sex are consid- [20], which found more myocardial lesions in diabetic as erably reduced by the presence of diabetes. Still, male compared to matched nondiabetic patients, support this hy- diabetics demonstrated more frequently peripheral coronary pothesis. However, beyond myocardial infarction, the pres- artery disease, while these differences could not be demon- ence of diabetes alone was the second most important strated in women. Taken together, our study demonstrates independent risk factor for decreased left ventricular ejec- the paramount importance of DM as a risk factor for devel- tion fraction. This suggests that diabetes itself leads to left opment and progression of coronary artery disease.
ventricular dysfunction. Decreased left ventricular ejection Multifactorial analyses of several prospective popula- fraction has also been shown in a small group of patients tion studies have shown that the high incidence of cor- with insulin-dependent diabetes mellitus without coronary onary artery disease in DM patients cannot be explained artery disease, and an impaired cardiac sympathetic inner- by a higher incidence of other known risk factors as vation associated with a defective blunted recruitment of compared to the general population [18,19]. Our findings myocardial contractility leading to left ventricular dysfunc- indicate that the severity of coronary artery disease is tion has been postulated [26]. Microvascular disease and the also notably influenced by the presence of DM, irrespec- significantly higher presence of hypertension in diabetics as tive of other cardiovascular risk factors, age, and gender.
compared to nondiabetic patients may add to this condition.
Even after adjustment for these factors, DM was signif- Whether poor control of blood sugar and associated glyco- icantly associated with a more severe atherosclerotic sylation of cardiac myocytes add an additional burden is burden as compared to patients without DM. However, currently undetermined [27]. The overproportionally high there is some controversy in this regard in the literature.
percentage of DM patients (33%) with severely impaired Whereas Ledru et al. [14] found similar results as in our left ventricular ejection fraction included to the recently registry, there are other reports that could not demon- published trials of implantable cardioverter and defibrilla- strate a significant difference in the severity of coronary tors [5,6] as compared to the routine prevalence of about artery disease between diabetic patients and patients 10% of diabetic patients referred to coronary angiography without diabetes [15,20,21]. The much smaller number (10.3% in our hospital) supports our data of a higher of patients investigated compared to the present study prevalence of left ventricular dysfunction in patients may explain this discrepancy. Our study is the first that with diabetes admitted for elective coronary angiog- used established scores of the atherosclerotic burden raphy. This suggests that diabetes is an important risk determined by angiography in a large study population.
factor not only for the presence of coronary artery The present findings of more severe coronary artery disease, but also for left ventricular dysfunction with a disease with higher atherosclerotic scores on angiogra- poor outcome by yet unknown mechanisms. Our find- Coronary Anatomy and LVEF
ings are supported by other investigators who found quate method to detect coronary artery disease and ath- that diabetes and arterial hypertension have adverse erosclerotic burden in patients with type 2 diabetes.
effects on left ventricular geometry and function, and Patients with DM have significantly more severe coro- that the combination of hypertension and diabetes nary artery disease even after adjustment for age and other resulted in the greatest degree of left ventricular hy- coronary atherosclerotic risk factors. In particular, DM is pertrophy and myocardial dysfunction [28]. However, the most important risk factor for the severity of the disease.
coronary angiographies were not routinely performed Furthermore, diabetic patients more often show left ventric- in this study. In this context, it is noteworthy that 50% ular dysfunction as compared to nondiabetics, which is of patients with left ventricular dysfunction are related to more severe and more diffuse atherosclerotic asymptomatic [29]. Therefore, especially diabetic pa- disease and to the presence of diabetes itself. Thus, diabetic tients should be screened for left ventricular dysfunc- patients with and without coronary artery disease may need tion even in the absence of clinical symptoms.
more aggressive diagnostic and therapeutic efforts, irrespec- Previous studies using myocardial scintigraphy, tive of other cardiac risk factors and gender. Since the more treadmill stress test, or 24-hr heart rate variability have diffuse nature of coronary disease in diabetic patients may demonstrated a higher percentage of silent angina in make revascularization difficult or even impossible, slowing patients with diabetes [30 –32]. Autonomic impair- down progression of the disease by means of secondary ment and abnormalities in pain perception have been prevention is particularly important in these high-risk pa- advocated for it in diabetics. Interestingly, we could tients. Future prospective studies are warranted to assess the not demonstrate differences between diabetic and non- long-term course of left ventricular function in newly diag- diabetic patients concerning their clinical symptoms nosed diabetics in relation to several strategies to modify when referred for coronary angiography, although cor- additional risk factors.
onary artery disease was significantly more severe indiabetics. This supports theories of an abnormal painperception in patients with DM. It is tempting tospeculate if abnormal pain perception was the main reason for the significantly higher age of diabetic ascompared to nondiabetic patients referred for coronary 1. Mokdad AH, Bowman BA, Ford ES, Vinicor F, Marks JS, Koplan angiography. However, due to the retrospective design JP. The continuing epidemics of obesity and diabetes in the United of this study, a negative referral bias for coronary States JAMA 2001;286:1195–2000.
angiography in diabetic patients due to physicians' 2. Diamond J. The double puzzle of diabetes: why is the prevalence decisions cannot be excluded.
of type 2 diabetes mellitus now exploding in most populations, butnot in Europeans? the genetic and evolutionary consequences of geographic differences in food history may provide the answer.
Nature 2003;423:599 – 602.
In our study population, diagnosis of type 2 diabetes 3. Stamler J, Vaccaro O, Neaton JD, Wentwoth D. Diabetes, other mellitus was based on information of patients, medical risk factors, and 12-yr cardiovascular mortality for men screened records, and blood glucose measurements in the majority of in the Multiple Risk Factor Intervention Trial. Diabetes Care patients prior to coronary angiography. Information on the 1993;16:434 – 444.
type of antidiabetic therapy and other cardiac medication 4. Turner R, Cull C, Holman R. United Kingdom Prospective Dia- betes Study 17: a 9-year update of a randomized, controlled trial was not systematically collected in our registry. Information on the effect of improved metabolic control on complications in on the duration of the disease, the glycemic control, other non-insulin-dependent diabetes mellitus. Ann Intern Med 1996; end-organ damage, or systematical follow-up investigations 124:136 –145.
after coronary angiography is lacking.
5. Moss AJ, Hall WJ, Cannom DS, Daubert JP, Higgins SL, Klein H, In addition, coronary anatomy was assessed by an- Levine JH, Saksena S, Waldo AL, Wilber D, Brown MW, Heo M.
giography and not by intravascular ultrasound. Until Improved survival with an implanted defibrillator in patients withcoronary disease at high risk for ventricular arrhythmia: Multi- now, there are only few studies giving insight into the center Automatic Defibrillator Implantation Trial investigators.
role of diabetes on vascular remodeling and plaque ac- N Engl J Med 1996;335:1933–1940.
cumulation in stenotic and nonstenotic coronary vessels 6. Moss AJ, Zareba W, Hall WJ, Klein H, Wilber DJ, Cannom DS, [33,34]. In these studies, limited vessel remodeling could Daubert JP, Higgins SL, Brown MW, Andrews ML. Prophylactic be demonstrated, but seemed to be confined to patients implantation of a defibrillator in patients with myocardial infarc- with type 1 diabetes. However, patients with type 2 tion and reduced ejection fraction. N Engl J Med 2002;346:877– diabetes showed a very similar plaque burden and vas- 7. Garg R, Yusuf S. Overview of randomized trials of angiotensin- cular remodeling as compared to patients without diabe- converting enzyme inhibitors on mortality and morbidity in pa- tes at least in obstructive (⬎ 50% stenosis) lesions [33].
tients with heart failure: Collaborative Group on ACE Inhibitor Therefore, coronary angiography seems to be an ade- Trials. JAMA 1995;273:1450 –1456.
Ammann et al.
8. Packer M, Fowler MB, Roecker EB, Coats AJ, Katus HA, Krum Association of diabetes mellitus with coronary atherosclerosis and H, Mohacsi P, Rouleau JL, Tendera M, Staiger C, Holcslaw TL, myocardial lesions: an autopsy study from the Honolulu Heart Amann-Zalan I, DeMets DL. Effect of carvedilol on the mortality Program. Am J Epidemiol 1993;137:1328 –1340.
of patients with severe chronic heart failure: results of the carve- 21. Wilson CS, Gau GT, Fulton RE, Davis GD. Coronary artery dilol prospective randomized cumulative survival (COPERNI- disease in diabetic and nondiabetic patients: a clinical and angio- CUS) study. Circulation 2002;106:2194 –2199.
graphic comparison. Clin Cardiol 1983;6:440 – 446.
9. CIBIS investigators and committees. A randomized trial of beta- 22. Natali A, Vichi S, Landi P, Severi S, L'Abbate A, Ferrannini E.
blockade in heart failure: the Cardiac Insufficiency Bisporolol Coronary atherosclerosis in type II diabetes: angiographic findings Study (CIBIS). Circulation 1994;90:1765–1773.
and clinical outcome. Diabetologia 2000;43:632– 641.
10. Sacks FM, Pfeffer MA, Moye LA, Rouleau JL, Rutherford JD, 23. Kannel WB, McGee DL. Diabetes and cardiovascular disease: the Cole TG, Brown L, Warnica JW, Arnold JM, Wun CC, Davis BR, Framingham Study. JAMA 1979;241:2035–2038.
Braunwald E. The effect of pravastatin on coronary events after 24. Krolewski AS, Warram JH, Valsania P, Martin BC, Laffel LM, myocardial infarction in patients with average cholesterol levels: Christlieb AR. Evolving natural history of coronary artery disease Cholesterol and Recurrent Events Trial investigators. N Engl in diabetes mellitus. Am J Med 1991;90:56S– 61S.
J Med 1996;335:1001–1009.
25. Melidonis A, Dimopoulos V, Lempidakis E, et al. Angiographic 11. Anonymous. Prevention of cardiovascular events and death with study of coronary artery disease in diabetic patients in comparison pravastatin in patients with coronary heart disease and a broad with nondiabetic patients. Angiology 1999;50:997–1006.
range of initial cholesterol levels: the Long-Term Intervention 26. Scognamiglio R, Avogaro A, Casara D, et al. Myocardial dys- with Pravastatin in Ischemic Disease (LIPID) study group. N Engl function and adrenergic innervation in patients with insulin-de- J Med 1998;339:1349 –1357.
pendent diabetes mellitus. J Am Coll Cardiol 1998;31:404 – 412.
12. Lemp GF, Vander Zwaag R, Hughes JP, Maddock V, Kroetz F, 27. Petrova R, Yamamoto Y, Muraki K, et al. Advanced glycation Ramanathan KB, Mirvis DM, Sullivan JM. Association between endproduct-induced calcium handling impairment in mouse car- the severity of diabetes mellitus and coronary arterial atheroscle- diac myocytes. J Mol Cell Cardiol 2002;34:1425–1431.
rosis. Am J Cardiol 1987;60:1015–1019.
28. Bella, JN, Devereux RB, Roman MJ, et al., for the Heart Study 13. Henry P, Makowski S, Richard P, Beverelli F, Casanova S, Louali investigators. Separate and joint effects of systemic hypertension and A, Boughalem K, Battaglia S, Guize L, Guermonprez JL. In- diabetes mellitus on left ventricular structure and function in Amer- creased incidence of moderate stenosis among patients with dia- ican Indians (the Strong Heart Study) Am J Cardiol 2001;87:1260 – betes: substrate for myocardial infarction? Am Heart J 1997;134: 29. McDonagh TA, Morrison CE, Lawrence A, et al. Symptomatic 14. Ledru F, Ducimetiere P, Battaglia S, Courbon D, Beverelli F, and asymptomatic left-ventricular systolic dysfunction in an urban Guize L, Guermonprez JL, Diebold B. New diagnostic criteria for population. Lancet 1997;350:829 – 833.
diabetes and coronary artery disease: insights from an angio-graphic study. J Am Coll Cardiol 2001;37:1543–1550.
30. Marchant B, Umachandran V, Stevenson R, Kopelman PG, Tim- 15. Pajunen P, Nieminen MS, Taskinen MR, Syvanne M. Quantitative mis AD. Silent myocardial ischemia: role of subclinical neurop- comparison of angiographic characteristics of coronary artery athy in patients with and without diabetes. J Am Coll Cardiol disease in patients with noninsulin-dependent diabetes mellitus compared with matched nondiabetic control subjects. Am J Car- 31. Langer A, Freeman MR, Josse RG, Armstrong PW. Metaiodo- diol 1997;80:550 –556.
benzylguanidine imaging in diabetes mellitus: assessment of car- 16. Dortimeter AC, Shenoy PN, Shiroff RA, Leaman DM, Babb JD, diac sympathetic denervation and its relation to autonomic dys- Liedtke AJ, Zelis R. Diffuse coronary artery disease in diabetic function and silent myocardial ischemia. J Am Coll Cardiol 1995; patients: fact or fiction? Circulation 1978;57:133–136.
25:610 – 618.
17. CASS principal investigators. The National Heart, Lung, and 32. Valensi P, Sachs RN, Harfouche B, et al. Predictive value of Blood Institute coronary artery surgery study. Circulation 1981; cardiac autonomic neuropathy in diabetic patients with or without silent myocardial ischemia. Diabetes Care 2001;24:339 –343.
18. Manson JE, Colditz GA, Stampfer MJ, Willett WC, Krolewski AS, 33. Vavuranakis M, Stefanadis C, Toutouzas K, Pitsavos C, Spanos Rosner B, Arky RA Speizer FE, Hennekens CH. A prospective study V, Toutouzas P. Impaired compensatory coronary artery enlarge- of maturity-onset diabetes mellitus and risk of coronary heart disease ment in atherosclerosis contributes to the development of coronary and stroke in women. Arch Intern Med 1991;151:1141–1147.
artery stenosis in diabetic patients: an in vivo intravascular ultra- 19. Rosengren A, Welin L, Tsipogianni A, Wilhelmsen L. Impact of sound study. Eur Heart J 1997;18:1090 –1094.
cardiovascular risk factors on coronary heart disease and mortality 34. Kornowski R, Mintz GS, Kent KM, et al. Increased restenosis in among middle aged diabetic men: a general population study. Br diabetes mellitus after coronary interventions is due to exagger- Med J 1989;299:1127–1131.
ated intimal hyperplasia: a serial intravascular ultrasound study.
20. Burchfiel CM, Reed DM, Marcus EB, Strong JP, Hayashi T.
Circulation 1997;95:1366 –1369.


Recommandations de la Société française de cardiologie concernant les indications et la surveillance du traitement anticoagulant oral Les recommandations suivantes sont indicatives et pro v i s o i res. Dans la mesure du possible, les niveaux de preuve sur lesquels elles sont fondées etle degré de validité des recommandations ont été indiquées, en utilisant laclassification décrite par Cook et coll. [1] (Annexe).

Persistence Patterns with Oral HypoglycemicAgents in Type 2 DiabetesJennifer M. Stephens, PharmD, Betina T. Blak, MSc, Karen F. Gold, PhD, Marc F. Botteman, MSc, MA, Chris L. Pashos, PhD, Caroline S. Walkinshaw, MBA, and R. Keith Campbell, RPh, MBA • Objectives: To determine persistence with therapy in affects more than 20% of the population older than age