Thickness of the intima media as a new correlate for atherosclerosis risk factors in Indian type 2 diabetes patients

1. Introduction 

The measurement of the Common Carotid Intima Media Thickness (CCIMT) is the state of the art assessment of pan atherosclerosis in any group of patients, for its role as the surrogate reflector of progress or regress of atherosclerosis in the body. The assessment of atherosclerosis through direct visualization at the anatomical vascular level, CCIMT provides an immense opportunity to assess the modulation of in situ pan atherosclerosis via control of risk factors for atherosclerosis. Diabetes mellitus itself accelerates atherosclerosis but the huge impetus comes from the in concert interaction with the various risk factors for atherosclerosis. Once the strongest correlates and associates of the CCIMT are uniformly unraveled in patients with diabetes, we will have immense modifying strength in our hands to prevent atherosclerotic vascular events. The key word and beacon behind the purpose of this study is the word “uniformly” because the astute reviewer of this topic will realize that more the studies, more the variability of risk factors correlating with CCIMT are found.

It has been suggested by the international atherosclerosis project that the atherosclerotic process occurs at the same time in carotid, cerebral and coronary arteries. The assessment of carotid atherosclerosis by the ultrasonography measurement of Common Carotid Intima Media Thickness (CCIMT) took over as being the marker of atherosclerosis [1], [2]. A highly accurate, reproducible, reliable and valid estimate of arterial wall thickness and a useful tool for detecting and monitoring changes in intima media thickness, and assessing subclinical atherosclerosis [3], is now available. The progression of CCIMT is an independent predictor of atherosclerotic events and useful surrogate marker for coronary and other atherosclerotic events [4], [5], [6], [7] and numerous studies have shown that CCIMT is higher in type 2 diabetes mellitus patients than in non-diabetic patients [8], [9], [10].

As the value of CCIMT in predicting atherosclerosis and its association with diabetes have become clearer, the next step is to evaluate the determinants of CCIMT and to evaluate the association of risk factors with pan atherosclerosis reflected by CCIMT.

Therefore, the present study was undertaken to know the relationship between Common Carotid Intima Media Thickness (CCIMT) and the risk factors for atherosclerosis in type 2 diabetes mellitus patients and to assess the risk factors determining the point status of atherosclerosis with its attendant events and reduction in the mortality and morbidity of these patients and also of the fiscal burden on both the patients and health institutions.

2. Subjects and methods 

The present study was conducted in the Department of Medicine, Kasturba Hospital, Manipal, India, between August 2001 and August 2004. Ninety Indian type 2 diabetes patients [67 male and 23 female] with an age range from 32 to 78 years, were recruited for this study. Demographic data of test subjects were collected including age, sex, duration of diabetes and, hypertension. The following hemato-logical and biochemical tests were taken: fasting, post parandial blood glucose estimation, urine analysis, renal function tests and fasting lipid profile glycosylated haemoglobin levels. Chest X-ray and ECG were also conducted for each subject. The subjects underwent measurement of Common Carotid Intima Media Thickness. This was done through B-mode ultrasonography using high frequency linear transducer on a “GE LogiQ 700” machine with the help of a specialist radiologist. Intima plus media thickness (IMT) was measured as the distance from the leading edge of the first echogenic line (lumen intima interface) to the second echogenic line (the collagen containing upper layer of the intimal adventitia) [1]. The common carotid arteries of both sides were visualized in the neck and at each longitudinal projection, the point of maximum thickness by visual inspection was measured. At two more points, one at 1cm upstream and the other 1cm downstream from the point of maximum thickness, the CCIMT was measured. The total of six values (three from right and three from left common carotid artery) were averaged to give a mean CCIMT value for the patient. Ninety type 2 diabetes mellitus patients, after complete evaluation, were classified into two groups. Group 1 (n=45) consisted of type 2 diabetes mellitus patients without atherosclerotic events. Group 2 (n=45) were those type 2 diabetes mellitus patients with atherosclerotic events. Various risk factors were then correlated with CCIMT values in both the groups to get a clearer idea of the determinants of CCIMT actually translating into events.

3. Statistical analysis 

SPSS package was used, and Student’s t-test was applied for independent groups to compare the significant difference between the means. The chi-square test and the fisher exact test were used to compare proportions. Statistical significance was assumed to be p<0.05. Pearson’s correlation coefficient (r) analysis was carried out to determine the correlation of CCIMT with different variables. Multivariate regression analysis was also performed to study the independent effect of the various risk factors and co-morbidities on CCIMT.

4. Results 

All the type 2 diabetic patients were divided into two groups; group 1 included 45 diabetic patients who had no atherosclerotic events, while group 2 included another 45 diabetic patients who had atherosclerotic events. Table 1 shows that there was no significant difference in the mean age, sex distribution and physicochemical characteristics of the two groups except for duration of diabetes mellitus and duration of hypertension. Comparing the distribution of risk factors between the two groups, the number of patients having hypertension and dyslipidemia was significantly higher in group 2 as compared to group 1 (p=0.0013, 0.0052, respectively) whereas there was no significant difference in the number of smokers or obese individuals in the two groups. On comparing CCIMT patients of group 1 (no events) and group 2 (events), the result (Table 2) showed that group 2 patients had significantly higher values of CCIMT as compared to group 1 patients i.e. (1.005±0.17 mm) versus (0.798±0.12 mm) (p<0.0001). A correlation of different variables with Common Carotid Intima Media Thickness (CCIMT) groups 1 & 2 (Table 3) showed that group 1 patients’ age had a statistically significant correlation with CCIMT, while in group 2 patients, proteinuria had a statistically significant correlation with CCIMT (p<0.05). In both groups, waist/hip ratio, the duration of diabetes mellitus, systolic blood pressure, glycosylated hemoglobin, triglyceride, total cholesterol/HDL cholesterol ratio had a positive correlation with CCIMT but could not assume statistical significance (p>0.05). On multivariate regression analysis with CCIMT as dependent variable, the only variables that were found to have a statistically significant association with CCIMT were Age in group 1 and Proteinuria in group 2 patients.

Table 1. Demographic, clinical and biochemical characteristics of the subjects.

	Characteristic	Group 1	Group 2	P-Value
	Age (years)	57.7±10.4	60.9±+9.7	0.1348
	Male sex	32 (71.1%)	35 (77.7%)	0.5381
	BMI (Kg/m2)	23.8±4.4	23.4±4.7	0.6779
	Waist/Hip ratio	0.97±0.07	0.98±0.07	0.4998
	Duration of diabetes (years)⁎	4.5±5.1	9.1±6.9	0.005
	Systolic BP (mm Hg.)	133.8±16.4	139.4±15.7	0.1016
	Diastolic BP (mm Hg.)	81.7±8.4	84.8±8.0	0.0765
	Duration of Hypertension (Years)⁎	1.7+3.7	4.3+5.4	0.0092
	Fasting blood glucose (mg %)	161.6+77.7	158.5+72.1	0.84
	Postprandial glucose (mg %)	237.1+102.	233.1+81.2	0.8382
	Glycosylated hemoglobin (%)	9.5+3.5	9.7+2.6	0.795
	Total cholesterol (mg/dl)	190.8+57.0	198.5+51.3	0.051
	Triglyceride (mg/dl)	143.2+79.4	156.8+82.4	0.4274
	LDL cholesterol (mg/dl)	118.4+42.7	126.1+44.5	0.5486
	HDL cholesterol (mg/dl)	42.1+13.5	41.1+12.5	0.72
	T. cholesterol/HDL cholesterol	4.72+1.3	5.09+1.61	0.2336
	Smoking	14 (31%)	21 (47%)	0.1183
	Body mass index>25	16 (36%)	9 (20%)	0.0894
	Hypertension⁎	16 (36%)	31 (69%)	0.0013
	Dyslipidemia⁎	20 (48%)	28 (62.2%)	0.0052
	Proteinuria	12 (26.7%)	22 (48.9%)	0.2951

All values are mean±SD.

⁎ p<0.05.

Table 2. CCIMT in patients with risk factors for atherosclerosis but no events (group 1) and atherosclerotic events (group 2).

		Group 1	Group 2
	Range of CCIMT (mm)	0.56–1.1	0.6–1.7
	Mean CCIMT (mm)	0.798	1.005∗
	Standard deviation (SD)	0.12	0.17

∗ Statistically highly significant (p<0.0001) as compared with group 1.

Table 3. Correlation coefficients (r) and multivariate regression analysis of different variables with common carotid intima media thickness (CCIMT) in groups 1 & 2 using CCIMT as dependent variable.

	Variable	Group 1		Group 2
		r-Values	p-Value	r-Values	p-Value
	Age	0.2948	0.0493∗	−0.126	0.4085
	Proteinuria	0.0616	0.6876	0.4083	0.005∗
	Regression coefficient/standard error of coefficient ratio
	Age	2.4474∗		0.2513
	Proteinuria	1.7166		2.2205∗

∗ Statistically significant (p<0.05) β/SE (regression coefficient/standard error of coefficient ratio ⩾2, indicates statistical significance).

5. Discussion 

Common Carotid Intima Media Thickness (CCIMT) is an established indicator of atherosclerosis and an important functional predictor of cardiovascular system [3], [9] which can be used to measure progress of atherosclerosis and also to assess the success of interventions [2], [8]. It is known that measurement of CCIMT by non-invasive B-mode ultrasonography can detect atherosclerosis at the earliest preclinical stage and help in the prediction and diagnosis of asymptomatic vascular disease [3], [10], [11]. This study demonstrates the factors which have the strongest correlation with CCIMT and hence with pan atherosclerosis.

In the present study, on comparing group 1 with group 2, there was no significant difference in the mean age or sex distribution in the two groups. The two groups were found to be comparable with respect to other characteristics. However, a higher duration of diabetes, hypertension and dyslipidemia were present in group 2, even though none of these was found to have a significant correlation or association with CCIMT. These patients also had a significantly higher mean CCIMT value.

In the present study, we observed that age has significant positive correlation and independent association (by multivariate regression analysis) with CCIMT of group 1 patients who did not have atherosclerotic events. Such a positive correlation of intima media thickness with age has also been reported by Kraml et al. [12] and Guvener et al. [13]. A linear positive correlation of age with intima media thickness was also observed by Mohan et al. [8]. In fact, in their study, the correlation of age with IMT was true for diabetic, as well as non-diabetic patients. The fact that age did not correlate with CCIMT in group 2 was explained by the presence of certain skew observations, which made statistical impact. We preserved these cases as it not only minimised bias, but also pointed towards a very important and novel concept of the genetic predisposition to higher CCIMTs irrespective of risk factors. Such studies proposing the importance of genetic factors have been reported by Lange et al. [14], Diamontopoulos et al. [15] and also, very recently by Moskau et al. [16].

Apart from age, proteinuria also had a correlation with increased CCIMT in this study. Proteinuria showed a positive correlation with CCIMT in group 2 patients (those with atherosclerotic events) also by multivariate regression analysis using CCIMT as dependent variable, proteinuria showed an independent association with CCIMT. Similar correlation of proteinuria with increased CCIMT was reported by Visona et al. [17] and Mykkanen et al. [18].

This correlation in group 2 patients is understandable because of an overall longer duration of the brunt of disease and co-morbidities in this group. Moreover, in diabetic patients who had atherosclerotic events, another interesting finding was that there was a statistically significant positive correlation between proteinuria and systolic blood pressure. This is another example of the ever-puzzling link between hypertension, proteinuria (microvascular event), carotid atherosclerosis and attendant macrovascular events in diabetic patients. It also reflects the role of hypertension possibly both directly and indirectly in influencing CCIMT. The same may hold good for other risk factors that were significantly higher in group 2 patients.

In both groups, waist/hip ratio, duration of diabetes mellitus, systolic blood pressure, glycosylated hemoglobin, triglyceride, total cholesterol/HDL cholesterol ratio had positive correlation with CCIMT but could not assume statistical significance. However, the higher incidence and duration of hypertension, the duration of diabetes and hyperlipidemia were present in group 2. In addition the mean value of CCIMT in group 2 was significantly higher. This strongly suggests an effect of these parameters on CCIMT.

The risk factors for increased CCIMT in patients with diabetes seem to be variable in various studies. While certain risk factors correlate with CCIMT in one study, the others do not have similar observations. Some studies showed that none of the variables of their study were associated with CCIMT in type 2 diabetics. Geroulakos et al. [19] also found that none of the potential risk factors were associated with IMT in patients with diabetes. Considering the above mentioned parameters, our findings are similar to these studies, except for age and proteinuria. On the other hand, in the study conducted by Guvener et al. [13] even though age, body mass index (BMI), duration of diabetes, smoking, lipid profile, fasting insulin levels, serum fibrinogen, hypertension and coronary artery disease were all assessed as determinants of carotid artery intima media thickness, multivariate analysis showed that age and BMI were the most important independent determinants of carotid intima media thickness for both sides.

Temelkova-Kurktschiev et al. [20] noticed increased CCIMT in diabetic patients with hyperlipidaemia, Jadhav and Kadam [21] with hypertension and very recently Karim et al. [22] pronounced that the association between carotid IMT and the duration of diabetes increases with both the frequency and duration of smoking. Moreover, in another recent study Kablak-Ziembicka et al. [23] concluded that hypertension, hyperlipidemia and non-insulin-dependent diabetes mellitus are related to a greater IMT, whereas other risk factors did not reveal that correlation. The findings of Folsom et al. [24] showed that carotid artery intima media thickness was also positively associated with coronary calcification, and reaffirmed the established role of traditional risk factors in the etiology of coronary artery disease, as assessed by computed tomography, but did not identify any important nontraditional risk factors.

McDonald et al. [25] remarked that in a multivariate model age, smoking, waist circumference and the highest quartile of CRP concentrations (⩾14mg/l) remained significant predictors of IMT⩾0.72mm. Our findings differ from the findings of other workers with the exception of age and proteinuria. Furthermore, according to Yokoyama et al. [26] a slight elevation of albuminuria is a significant determinant of IMT and PWV (pulse wave velocity) independent of conventional cardiovascular risk factors in type 2 diabetic patients with no clinical nephropathy or any vascular diseases. Tatsukawa et al. [27] concluded that in a section of their subjects, BMI was not a cardiovascular risk factor, although LDL cholesterol was a common important risk factor.

Considering all these observations, it is clear that there is much variability in the correlation of risk factors with CCIMT. Once a uniform set of the strongest correlates and independent associates is established, a tight control of the risk factors will enable a tight control over the atherosclerotic process. This, in turn will ensure a good quality of life and freedom from the morbidity and mortality associated with atherosclerotic events in patients with diabetes mellitus.

6. Conclusion 

This study shows that the value of CCIMT has a significant positive correlation and independent association with age and proteinuria in Indian type 2 diabetes mellitus patients. Moreover, hypertension, duration of diabetes and dyslipidemias may actually have a correlation with CCIMT either directly or by influencing other parameters like proteinuria, even though these risk factors did not attain the statistical significance. Further studies with large sample size are needed to reconfirm this statistically and to propose a uniform set of the strongest correlates of CCIMT which may then be dealt with in a most thorough manner.