Coronary Artery Disease and Carotid Artery Intima-Media Thickness in Japanese Type 2 Diabetic Patients Free

During the period from February 1998 to January 1999, high-resolution ultrasound examination of the carotid arteries was performed in 620 consecutive type 2 diabetic patients (diagnosed by the World Health Organization criteria) among the outpatients attending clinics at Juntendo University Hospital. Among them, none had carotid endarterectomies or other therapeutic procedures. Among these patients, 18 complained of chest pain. Of these 18 patients, 14 underwent coronary angiography for suspected CAD that was based on combinations of symptoms, resting electrocardiogram (ECG) changes, and abnormal cardiac enzyme levels. Of these patients, 12 were diagnosed with CAD by coronary angiography. Screening for silent myocardial ischemia was performed on the remaining clinically asymptomatic patients. Exercise ECG was the first choice for screening method. A double Master two-step test, a form of exercise ECG, was performed in 290 randomly selected asymptomatic patients. Among them, 48 had ischemic changes on the ECG, and 32 underwent coronary angiography, with 28 eventually being confirmed to have CAD by angiography. These 40 diabetic patients with CAD were investigated in the present study. There were no significant differences in IMT values among 12 symptomatic patients and 28 asymptomatic patients (1.30 ± 0.09 vs. 1.26 ± 0.07 mm, respectively). No significant difference between the two groups was found for other relevant variables (age, sex, and duration of diabetes).

There were 242 subjects who had normal findings on the double Master two-step test. Control subjects were selected from the diabetic patients. A subject was eligible as a control if he or she 1) was 40–80 years old, 2) had normal findings on the double Master two-step test, and 3) remained free from myocardial infarction or stroke during follow-up. We excluded subjects with previously diagnosed CAD and whose HbA_1c levels were >12%. There were 223 subjects who fulfilled the inclusion criteria. We stratified them for sex, age, and HbA_1c level and randomly selected 40 control subjects who closely matched case subjects for baseline characteristics.

At the time of enrollment, baseline laboratory data, blood pressure, BMI, resting 12-lead ECG, double Master two-step test results (16), and IMT measurement were collated for each patient. Fasting blood was obtained by standard laboratory techniques for the measurement of serum lipids [total cholesterol, HDL cholesterol, triglycerides, and lipoprotein(a)], serum uric acid, serum creatinine, blood urea nitrogen, and HbA_1c.

Blood pressure was measured with a standard mercury sphygmomanometer placed on the arm after the subject had been seated for at least 5 min. Among the 40 patients with CAD, 7 were treated with diet only, 22 received insulin injections, and 8 were on sulfonylureas. Among the patients without CAD, 10 were on diet alone, 18 received insulin, and 7 were treated with sulfonylureas. Thus, treatment of diabetes did not differ between the two groups (NS). There were no significant differences regarding the treatment of hypertension between the patients with and without CAD (calcium-channel blockers: 13 and 10, respectively; ACE inhibitors: 9 and 5; α-blockers: 3 and 2; β-blockers, 2 and 1; P = 0.21). The treatment of hyperlipidemia also did not differ between patients with CAD and those without CAD (hydroxymethylglutaryl [HMG]-CoA reductase inhibitors: 11 and 8, respectively; clofibrates: 2 and 2; NS). Hypertension was defined as a systolic blood pressure >150 mmHg, a diastolic blood pressure >90 mmHg, and/or the use of antihypertensive medication. Hyperlipidemia was considered present when the patient had a serum total cholesterol level >220 mg/dl, had a triglyceride level >150 mg/dl, and/or was on lipid-lowering medications. Hyperuricemia was considered present when the patient had a serum uric acid level >6.0 mg/ dl and/or uricosuric therapy. Informed consent was obtained from all subjects.

Ultrasonographic scanning of the carotid arteries was performed using an tomographic ultrasound system (EUB-555; Hitachi Medico, Tokyo) with a linear transducer (midfrequency range 7.5–10 MHz). Scanning of the extracranial carotid arteries in the neck was performed bilaterally in three different longitudinal projections (i.e., anterior-oblique, lateral, and posterior-oblique) and in the transverse projection, as reported previously (3). This method allowed the common carotid artery, the carotid bulb, and the internal and external carotid arteries to be scanned. All images were photographed.

Carotid IMT was defined as the distance from the leading edge of the first echogenic line to the leading edge of the second echogenic line on the scans, with the first line representing the lumen-intimal interface and the second line representing the collagen-containing upper layer of the adventitia. Plaque was defined as a localized thickening lesion. In each longitudinal projection, the site with the greatest thickness (including plaque) was detected along the vessel from the common carotid artery to the internal carotid artery. Three measurements of the IMT were made at the site of the greatest thickness by visual examination and at two other points (1 cm proximal and 1 cm distal to this site). These three determinations were averaged. The greatest value among the six averaged IMTs (three from the left and three from the right) was used as the representative value (the absolute millimeter of IMT) for each individual. The scanning period averaged 30 min in each patient, and scanning was performed by a trained physician (N.M.). The reproducibility of IMT measurement was examined by repeating the procedure after 2 weeks in eight subjects with IMT values of 0.74–1.36. The mean absolute difference between the two measurements was 0.05 mm, and the SD was 0.07 mm, demonstrating good reproducibility.

Coronary angiography was performed in diabetic subjects with abnormal findings in the double Master two-step test or who complained of chest pain. CAD was diagnosed by detection of >50% stenosis in one of the three major coronary arteries.

Data are presented as the means ± SE. To perform two group comparisons closely matched for baseline characteristics, the Student’s unpaired t test was used to compare the mean values of variables between diabetic patients with and without CAD, whereas the χ² test was used to compare proportions. ANCOVA was used to compare mean IMT values in the diabetic patients with and without CAD after adjustment for age, sex, duration of diabetes, and antidiabetic therapy. Multivariate analysis was performed using logistic regression analysis with stepwise forward selection method. In these models, the reported parameter estimates are odds ratios for prevalent CAD. We used the StatView program. Statistical significance was defined at the P < 0.05 level.

The clinical characteristics of the 40 diabetic patients with CAD and 40 diabetic control subjects without CAD are shown in Table 1. The number of men and women in the two groups were identical. CAD patients had a significantly higher BMI and uric acid level and a lower HDL cholesterol level than the control subjects. In contrast, HbA_1c, serum total cholesterol, and serum triglyceride levels were similar in the two groups.

The IMT values and the prevalence of plaque are shown in Table 1. IMT was significantly greater in the CAD patients than in the control subjects (1.27 ± 0.07 vs. 1.03 ± 0.04 mm, P < 0.05), but the prevalence of plaque did not differ between the two groups.

Logistic regression analysis was used to assess the relationship between the IMT and the prevalence of CAD (Tables 2 and 3). CAD patients had a significantly higher prevalence of hypertension, hyperlipidemia, and hyperuricemia than the control subjects (85 vs. 50%, P < 0.001; 70 vs. 30%, P < 0.001; and 25 vs. 7.5%, P < 0.05, respectively). Logistic regression analysis using the stepwise forward selection method showed that the prevalence of hypertension (Wald’s χ² = 7.2430), hyperlipidemia (Wald’s χ² = 5.9046), and hyperuricemia (Wald’s χ ² = 4.8790) as well as the IMT (Wald’s χ ² = 4.7087) was positively related to the likelihood of prevalent CAD (Table 2). IMT was also strongly but not significantly associated with CAD by logistic regression analysis using all independent variables (P = 0.062; Table 3).

The 40 patients with CAD were divided into 13 patients with two-vessel disease and 27 patients with three-vessel disease based on the results of coronary angiography. However, IMT was not significantly associated with the number of involved vessels.

Among the 40 patients with CAD, percutaneous transluminal coronary angioplasty (PTCA) was performed in 11 patients, coronary artery bypass grafting (CABG) in 20, and conservative treatment by medication in only 9. The sex ratio was identical in the two groups with PTCA (8 men and 3 women in the post-PTCA group and 6 men and 3 women in the non-PTCA group). The CAD patients were divided into a post-CABG group (n = 20) and a non-CABG group (n = 20). The sex ratio was also identical in the two groups with CABG. The treatment of diabetes did not differ between the two groups with CABG. Cardiovascular risk factors such as hypertension, hyperlipidemia, and hyperuricemia did not differ between the two groups, but the IMT was significantly greater in the post-CABG group than in the non-CABG group (Table 4).

Many studies have shown that the incidence of CAD in diabetic patients is two or three times higher than that in healthy control subjects (16–18). We previously reported that IMT was significantly greater in Japanese subjects with type 2 diabetes when compared with healthy control subjects. In the present study, we showed that IMT of type 2 diabetic patients with CAD was significantly greater than that of diabetic patients without CAD. We also found that the prevalence of CAD related to increased IMT. There have been reports on the relationship between IMT and CAD prevalence and incidence in diabetic and nondiabetic subjects (1, 2, 7, 10–12, 19, 20). During a 3.1-year follow-up in Japanese type 2 diabetic patients (14), it was reported that initial IMT could be a predictor of nonfatal coronary heart disease (angina pectoris or nonfatal myocardial infarction). In that study, however, the occurrence of coronary heart disease was diagnosed from the ECG or by symptoms of myocardial infarction. In contrast, CAD was defined by coronary angiography in our study. Most diabetic subjects with CAD have multivessel disease and diffuse stenosis of the coronary arteries, and some of them suffer from asymptomatic myocardial infarction. Burke et al. (12) reported that silent myocardial infarction was defined as a diagnostic Q wave detected on the baseline resting ECG. Our study included patients with asymptomatic CAD that was diagnosed by the double Master two-step test. The non-CAD patients came from subjects with normal findings on the step test. However, without coronary angiography, one cannot be completely sure that these subjects were free of CAD. This point could be a limitation of our study.

In general, our subjects with CAD had more cardiovascular risk factors than those without CAD. We used multivariate analysis to examine whether IMT was an independent contributor to the prevalence of CAD. IMT was independently associated with CAD prevalence, although the association of IMT was not so strong when compared with the other cardiovascular risk factors: hypertension, hyperlipidemia, and hyperuricemia. The results indicated that IMT might be an independent predictor of CAD prevalence.

IMT was not associated with the number of coronary vessels in our subjects showing >50% stenosis. However, most other studies have suggested that IMT may be associated with the number of stenosed vessels in nondiabetic subjects (21–23), although a few studies have also not found an association between IMT and the number of diseased vessels (24). It is well known that most diabetic patients with CAD have multivessel disease and diffuse stenotic regions of the coronary arteries. In this study, all our diabetic subjects with CAD had multivessel disease (two- or three-vessel disease), and none had single-vessel disease. This finding may have been the cause of the discrepancy between the association of IMT with the number of stenoses in our study and in other studies. However, IMT was significantly greater in post-CABG patients than in non-CABG patients, suggesting an association between IMT and the severity of multivessel disease. Carotid IMT was measured within 6 months after CABG to avoid not only the influence of the immediate postsurgical complications, but also the changes of IMT. However, it is possible that IMT may have become thicker after the CABG because a temporal sequence cannot be established with cross-sectional data.

In conclusion, our results indicate that early atherosclerosis in the carotid arteries suggests a high probability of coronary involvement in Japanese patients with type 2 diabetes. If carotid IMT is increased in diabetic patients, further screening for CAD should be performed, and treatment should be initiated to prevent the progression of CAD.