Introduction

Intima media thickness (IMT) of the carotid arteries is associated with coronary atherosclerosis and is a predictor of coronary events [1]. Improved blood glucose control obtained by intensive insulin treatment of type 1 diabetes is associated with lower carotid artery IMT than that found with conventional insulin treatment [2, 3].

We aimed to evaluate the IMT and its association with mean HbA1c measured longitudinally over the previous 18 years in people with type 1 diabetes. The diabetic patients in the present study were also examined for early coronary atherosclerosis by intravascular ultrasound (IVUS) [4]. As carotid IMT measurements are easier to perform than IVUS examinations of the coronary arteries and as, to our knowledge, a comparison of these methods has never been done, we assessed any association between carotid IMT and early coronary atherosclerosis as observed by IVUS.

Subjects and methods

Subjects

Thirty-nine patients from the Oslo study on type 1 diabetes participated in the present study. They were followed prospectively for 18 years (1982–2000). The protocol and results are given in detail elsewhere [5]. In the original study 45 patients were included. Two of these patients died, one of lung disease and the other of breast cancer, and four declined further participation. The Governmental Regional Ethics Committee approved the study. Twenty-nine patients agreed to coronary IVUS examination. All patients gave their written consent. The reference population is the fourth survey of the Tromsø Study [6].

Ultrasonography

Ultrasound studies were performed according to a standard protocol. High-resolution B-mode ultrasonography of the right carotid artery was performed with a 7-MHz ultrasound scanner (Acuson 128 Computed Sonography System, Mountain View, CA, USA) equipped with a linear array transducer. All examinations were performed by the same sonographer. Patients were examined in the supine position with the head turned 45° contralaterally with the help of a cushion that had been specially designed for this purpose. IMT of the far wall of the CCA and the carotid bulbous was examined. The images were stored on high-resolution S-VHS videotape for offline analysis. For repeated measurements in individual patients the mean coefficient of variation was 5.6% for CCA IMT.

The ultrasonic images were analysed offline by a computerised technique for automated ultrasonic image analysis. This was undertaken by an experienced investigator from the Tromsø study who had no knowledge of the patients [7]. The different layers with their interfaces of the carotid artery wall were defined automatically. A 10-mm wall segment on both sides of the demarcation arrow separating the CCA and the bifurcation of the carotid artery was measured. The computer program estimated the maximum, minimum and average IMT along the 10-mm segment. The mean of the average IMT and the mean of the maximum IMT of three to four measurements were calculated and used in the analyses.

Intravascular coronary ultrasound examinations

The IVUS system used in this study consisted of the Clear View Ultra with Automatic Pullback Device and the Ultra Cross 3.2 F30-MHz catheter (Boston Scientific, Sunnyvale, CA, USA). The ultrasound images were recorded on S-VHS videotape for offline analysis. Percentage of vessel area stenosis, defined as plaque area divided by vessel area multiplied by 100 [4], was used to study the statistical association between IMT and HbA1c.

Laboratory methods

Lipid profiles were measured by conventional methods in the fasting state. The first HbA1c measurement of each year was used for analyses. Further details are described elsewhere [4]. Microalbuminuria was defined as urinary albumin excretion above 30 mg/24 h in at least two of three samples, and overt nephropathy as albumin excretion above 300 mg/24 h in two of three samples [4].

Statistical methods

To allow for adjustment for age when studying the relation between HbA1c and IMT in the CCA, multivariate linear regression analysis was performed. All significance tests were two tailed, and p values of less than 0.05 were considered significant. All calculations were performed using SPSS version 10.0 (SPSS, Chicago, IL, USA).

Results

Demographic characteristics of the 39 patients at the time of the IMT measurements are presented in Table 1. Mean age was 43 (range: 35–58) years, and mean duration of disease was 30 (23–39) years. The mean HbA1c over 18 years was 8.2% (6.6–11.3%). Table 2 shows the mean IMT values according to age and sex for the average and maximal common carotid and the carotid bulb. A comparison with a large age-matched reference population is also shown. The mean CCA IMT of the far wall was 0. 86 (0.62–1.35) mm for men and 0.83 (0.65–1.09) mm for women. Compared with the control population, IMT values in the patients were at the same level as those in people more than 20 years older.

Table 1 Demographic and putative risk factors for atherosclerosis according to sex
Full size table
Table 2 IMT in diabetic and reference populations according to sex and age
Full size table

Carotid IMT was significantly correlated with age in females (r=0.68, p<0.004) but not in males (r=0.21, p=0.33). Furthermore, multivariate linear regression analysis with age and HbA1c as independent variables demonstrated that in females, both age and HbA1c were significantly related to mean IMT (r 2=0.77; age: p<0.0001, HbA1c: p<0.0001) whilst in males, no significant associations were found (r 2=0.07; age: p=0.71, HbA1c: p=0.32). According to the regression analyses, an increase of HbA1c of 1% over 18 years corresponded in women to a 0.10-mm increase in mean IMT, whereas a 10-year increase in age resulted in a rise of 0.15 mm. We found no association between HbA1c and IMT in the carotid bifurcation. The crude associations between IMT and the factors: family history of cardiovascular disease, family history of type 2 diabetes, BMI, microalbuminuria, total cholesterol, smoking or systolic blood pressure were not statistically significant. For the whole group there was a significant association between mean CCA IMT and the percentage of vessel area stenosis as found by IVUS (r=0.43, p=0.034). When the men and women were examined separately, this correlation was high for women (r=0.65, p=0.032) and low for men (r=0.07, p=0.78). Similar results were obtained when mean CCA IMT was replaced with maximal IMT.

Discussion

In the present study we found a strong association between carotid IMT and HbA1c in women but not in men. This result indicates an important role of glycaemic control in the development of preclinical atherosclerosis in women with type 1 diabetes. Considering the low p values in this analysis for women, our results suggest that age and HbA1c have a strong impact on CCA IMT in diabetic females.

In the Epidemiology of Diabetes Interventions and Complications (EDIC) study, where IMT was measured 18 months after the Diabetes Control and Complications Trial (DCCT), mean HbA1c during the DCCT did not significantly correlate with carotid IMT. However, when the measurements in the EDIC were repeated 6 years later, IMT for both men and women was associated with mean HbA1c during the DCCT [3]. Considering that our results refer to 18 years of follow-up, the strong association found between HbA1c and IMT for diabetic women in our study is in accordance with the findings of the EDIC study in women. The implication of these studies might be that elevated long-term plasma glucose is not only a risk factor for the progression of atherosclerosis in type 1 diabetes but may be causally related to the atherosclerotic process.

The mean CCA IMT in the present study is significantly higher than in the sex- and age-matched reference groups and corresponds to what was found in subjects around 20 years older, even though mean blood pressure, mean cholesterol and BMI were lower in the diabetic group [6]. In middle-aged non-diabetic subjects the risk of coronary artery disease is two to five times higher in men than in women [8]. Diabetes is an independent risk factor for cardiovascular disease, greater in women than in men; the usual lower risk in women does not seem to be the case here [9]. Interestingly, an association between HbA1c and cardiovascular disease was found in women but not in men in the Pittsburgh Epidemiology of Diabetes Complications Study although the researchers were not convinced by this result [10]. In earlier studies, poor glycaemic control did not explain the increased risk of the early development of coronary artery disease in either sex [11, 12]. However, IMT measurements reveal atherosclerosis at an earlier stage than coronary angiography and our findings indicate that long-term hyperglycaemia is a stronger risk factor for the development of atherosclerosis in women than in men. We do not know why this is the case but it is plausible that diabetes-related risk factors have a stronger atherosclerotic effect in women.

We also identified an association between IMT and preclinical coronary atherosclerosis (by IVUS) in women but not in men. This suggests that IMT is an indirect measure of coronary atherosclerosis in asymptomatic type 1 diabetes, at least in women. However, considering that the low sample size of this study does not allow adjustment for several confounders, our results must be interpreted with caution until further long-term results from larger studies are available.

Concluding remarks

In women with type 1 diabetes there is an association between long-term HbA1c and carotid IMT. Individuals with type 1 diabetes have higher carotid IMT than age- and sex-matched controls. Carotid IMT in women with type 1 diabetes reflects preclinical coronary atherosclerosis as shown by IVUS.