Comparison of carotid intima-media thickness, arterial stiffness, and brachial artery flow mediated dilatation in diabetic and nondiabetic subjects (The Chennai Urban Population Study [CUPS-9])

Comparison of Carotid Intima-Media Thickness, Arterial Stiffness, and Brachial Artery Flow Mediated Dilatation in Diabetic and Nondiabetic Subjects (The Chennai Urban Population Study [CUPS-9]) Radhakrishnan Ravikumar, MBBS, DNB, Raj Deepa, MPhil, PhD, Coimbatore Subramaniam Shanthirani, MSc, and Viswanathan Mohan, MD,

PhD, DSc

This study compares flow-mediated dilation (FMD) and the augmentation index (AI) in diabetic and nondiabetic subjects and correlates these measurements with carotid intima-media thickness (IMT). Fifty diabetic subjects and 50 age- and sex-matched nondiabetic control subjects were recruited from the Chennai Urban Population Study. IMT of the common carotid artery and FMD of the brachial artery were determined using high-resolution B-mode ultrasonography. AI was measured using the Sphygmocor apparatus. The mean AI of diabetic subjects was significantly higher than the nondiabetic subjects (27.48 ⴞ 7.41% vs 19.10 ⴞ 8.19%, p <0.0001). The FMD values were significantly lower among diabetic subjects compared with the nondiabetic subjects (2.1 ⴞ 2.95% vs 6.64 ⴞ 4.38%, p <0.0001). At any given age point, diabetic subjects had significantly higher AI and lower FMD values compared with nondiabetic subjects (p <0.05). In the total population, AI and FMD showed a correlation with age (p <0.001), fasting plasma glu-

cose (p <0.01), glycosylated hemoglobin (p ⴝ 0.001), and IMT (p ⴝ 0.001). Among the nondiabetic subjects, FMD and AI showed a strong correlation with IMT. FMD also showed a strong correlation with age and systolic blood pressure, whereas AI showed a correlation with fasting plasma glucose in diabetic subjects. AI and FMD values showed a strong correlation with age. AI values increased and FMD values decreased with an increase in quartiles of IMT both in diabetic and nondiabetic subjects. Multivariate linear regression analyses in the total study population showed that age and glycosylated hemoglobin were the risk factors associated with AI and FMD, in addition to diastolic blood pressure with AI. Diabetic patients have decreased FMD and increased arterial stiffness compared with age- and sex-matched nondiabetic subjects. These functional changes correlate well with the structural changes of the arteries measured by IMT. 䊚2002 by Excerpta Medica, Inc. (Am J Cardiol 2002;90:702–707)

recent report from the World Health Organization (WHO) shows that India has the largest A number of diabetic subjects in any given country in

(IMT) among diabetic subjects compared with nondiabetic subjects.10 In this study, we assessed the functional changes of the artery by measuring arterial stiffness and flow-mediated dilation (FMD) in diabetic and nondiabetic subjects in CUPS. We also tried to correlate the structural changes in the arteries (carotid IMT) with the functional changes (FMD and augmentation index [AI]). This study is the first to assess all the 3 preclinical atherosclerotic markers in the same group of subjects.

the world.1 Migrant Asian Indians are known to have a greater propensity for premature coronary artery disease.2–5 To elucidate the prevalence and risk factors for coronary artery disease and the metabolic syndrome in native Indians within the Indian subcontinent, we took up the Chennai Urban Population Study (CUPS).6 Earlier publications from this study have reported on the prevalence of the metabolic syndrome7 and peripheral vascular disease in this population.8 In a recent publication, we showed a tenfold increase in the prevalence of coronary artery disease in urban India in the last 40 years.9 Similarly, we also documented increased carotid intima-media thickness From the Madras Diabetes Research Foundation, Gopalapuram, Chennai, India. Manuscript received March 11, 2002; revised manuscript received and accepted May 29, 2002. Address for reprints: Viswanathan Mohan, MD, PhD, DSc, Madras Diabetes Research Foundation, 4 (Old No. 36), Conran Smith Road, Gopalapuram, Chennai, 600 086, India. E-mail: mvdsc@ vsnl.com.

702

©2002 by Excerpta Medica, Inc. All rights reserved. The American Journal of Cardiology Vol. 90 October 1, 2002

METHODS The CUPS is an ongoing epidemiologic study in Chennai (formerly Madras), the fourth largest city in India with a population of about 6 million people. The methods of the study have been previously described.6 –10 Briefly, 2 residential colonies in Chennai were selected for the study. Of the total of 1,339 persons ⬎20 years old living in the 2 colonies, 1,262 individuals (90%) participated in the initial screening program. Of these, 152 had diabetes (91 had known diabetes and 61 were newly detected after the screening program) based on the WHO consulting group 0002-9149/02/$–see front matter PII S0002-9149(02)02593-6

sin) with an electrical linear transducer midfrequency of 7.5 MHz. The Nondiabetic Subjects Diabetic Subjects axial resolution of the system was Variables (n ⫽ 50) (n ⫽ 50) p Value 0.3 mm. The images were recorded, as well as photographed. Scanning Age (yrs) 54 ⫾ 11 54 ⫾ 10 — Men 23 (46%) 23 (46%) — was performed for an average of 20 23.9 ⫾ 3.7 24.8 ⫾ 4.0 0.236 Body mass index (kg/m2) minutes. Waist–hip ratio 0.88 ⫾ 0.09 0.90 ⫾ 0.1 0.461 IMT as defined by Pignoli and Systolic blood pressure (mm Hg) 121 ⫾ 12 126 ⫾ 11 0.03 Longo13 was measured as the disDiastolic blood pressure (mm Hg) 77 ⫾ 8 81 ⫾ 5 0.04 Duration of diabetes (years) — 8.4 ⫾ 6.2 — tance from the leading edge of the Fasting plasma glucose (mmol/L) 4.8 ⫾ 0.83 8.6 ⫾ 3.4 ⬍0.0001 first echogenic line to the second HbA1c(%) 5.74 ⫾ 0.43 8.2 ⫾ 2.0 ⬍0.0001 echogenic line. The first echogenic Serum cholesterol (mmol/L) 4.81 ⫾ 0.96 4.91 ⫾ 1.07 0.684 line represents the lumen intimal inHDL cholesterol (mmol/L) 1.22 ⫾ 0.31 1.14 ⫾ 0.26 0.178 terface and the second line is proLDL cholesterol (mmol/L) 2.89 ⫾ 0.83 2.96 ⫾ 0.96 0.648 Serum triglycerides (mmol/L) 1.58 ⫾ 1.107 1.74 ⫾ 1.09 0.480 duced by the collagen-containing upAI (%) 19.10 ⫾ 8.19 28.56 ⫾ 7.38 ⬍0.0001 per layer of the tunica adventitia. At FMD (%) 6.64 ⫾ 4.38 1.72 ⫾ 2.82 ⬍0.0001 each longitudinal projection, deterIMT (mm) 0.77 ⫾ 0.13 0.86 ⫾ 0.23 0.010 minations of IMT were conducted at the site of greatest thickness and at 2 points 1 cm upstream and 1 cm criteria.11 Seventy-four of the study subjects (5.9%) downstream from the site of greatest thickness as had impaired glucose tolerance, whereas the rest (n ⫽ described by Yamasaki et al.14 The mean of the 6 IMT 1,036) had normal glucose tolerance. measurements (3 from the far wall and 3 from the near Of the 152 diabetic subjects, 140 diabetic subjects wall) was used as the representative value for each (92.1%) agreed to participate in our earlier study on subject. The scanning was done using fine manipulacarotid IMT.10 For the present study, as we were study- tions of the transducer, to visualize, with maximum ing functional changes that could be altered in a variety clarity, the double-line pattern of the IMT both at the of situations, we excluded 88 subjects who were aged near and far wall of the artery. The reproducibility of ⬎70 years (n ⫽ 11), smokers (n ⫽ 11), hypertensives (n IMT measurements has been previously described.10 ⫽ 44), and subjects with known coronary artery disease Arterial stiffness: Arterial stiffness was measured us(n ⫽ 29). This left 52 diabetic subjects suitable for the ing the Sphygmocor apparatus (Sphygmocor BPAS-1; present study, 50 of whom agreed to participate. Fifty PWV Medical, Sydney, Australia). In brief, a high-fidelage- (frequency matched) and sex-matched controls ity micromanometer (SPC-301; Millar Instruments, were selected from the group of subjects with normal Houston, Texas) was used to flatten but not occlude the glucose tolerance. The control subjects were also non- right radial artery, using gentle pressure. When the 2 smokers, had no evidence of coronary artery disease surfaces are flattened, circumferential pressures are (clinically and based on electrocardiographs at rest), and equalized and an accurate pressure waveform can be were normotensive. recorded. Data were collected directly into a portable Physical examination included height and weight microcomputer. The system software allowed on-line measurements and calculation of the body mass index. recording of the peripheral waveform, which was asBlood pressure was recorded in the right arm with a sessed visually to ensure that the best possible recording mercury sphygmomanometer (Diamond Deluxe was obtained and that artifacts from movement were Blood pressure apparatus, Pune, India) while patients minimized. After 20 sequential waveforms had been were seated. Two readings were taken 5 minutes apart acquired, the integral software was used to generate an and the mean of the 2 was taken as the blood pressure. averaged peripheral and corresponding central waveform Other tests included a complete lipid profile, including that was used for the determination of the AI. AI was total serum cholesterol, serum triglyceride, and serum defined as the difference between the first and second high-density lipoprotein cholesterol. Low-density li- peaks of the central arterial waveform, expressed as a poprotein cholesterol was calculated using the Friede- percentage of the pulse pressure.15 wald formula.12 Biochemical analysis were done on AI is a measure of the contribution that the wave Ciba Corning Express Plus Auto Analyser (Corning, reflection makes to the arterial pressure waveform. Medfield, Massachusetts) using kits supplied by The amplitude and timing of the reflected wave ultiBoehringer Mannheim (Mannheim, Germany). Glyco- mately depends on the stiffness of the small vessels sylated hemoglobin (HbA1c) was estimated by high- and large arteries, and thus, AI provides a measure of pressure liquid chromatography using the Variant ma- systemic arterial stiffness.16 Pulse-wave analysis has chine (Bio Rad, Hercules, California). been demonstrated as a reproducible, noninvasive Measurement of intima-media thickness: The method for assessing AI.15,16 method used for measurement of carotid IMT has To check for the reproducibility of AI, 2 measurebeen previously described.10 The intima plus medial ments were performed on 20 subjects on consecutive thickness of the right common carotid artery was days by the same observer. The mean difference in AI determined using a high-resolution B-mode ultrasono- between the first and second measurements was 1.58, graphic system (Logic 400 GE, Milwaukee, Wiscon- and the SD was 2.54. TABLE 1 Clinical Features of the Study Groups

CORONARY ARTERY DISEASE/FMD AND ARTERIAL STIFFNESS IN SOUTH INDIANS

703

ables such as IMT, age, body mass index, systolic blood pressure, diastolic blood pressure, fasting plasma glucose, glycosylated hemoglobin, serum cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, and serum triglycerides. Multivariate linear regression analysis was carried out using FMD and AI as the dependent variables and age, body mass index, systolic blood pressure, diastolic blood pressure, glycosylated hemoglobin, serum cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, and serum triglycerides as independent variables. A p value ⬍0.05 was considered significant. All statistical analyses were performed using SPSSC version 4.0.1 (SPSS Inc., Chicago, Illinois). FIGURE 1. Duration distribution of AI in diabetic subjects compared with nondiabetic subjects. *p <0.05; **p <0.001 compared with nondiabetic subjects.

Flow-mediated dilation: FMD of the brachial artery was determined using a high-resolution B-mode ultrasonographic system (Logic 400 GE) with an electrical linear transducer midfrequency of 7.5 MHz, using the technique described by Celermajer et al.17 Briefly, each subject was requested to lie at rest for ⱖ10 minutes before the procedure began and the first scan at rest was then taken. This was followed by inflation of pneumatic tourniquet of the standard sphygmomanometer (Diamond BP Apparatus) placed around the forearm to a pressure of 300 mm Hg followed by deflation after 4.5 minutes. The second scan was taken 30 seconds before and 90 seconds after cuff deflation. Fifteen minutes was then allowed for vessel recovery and a further scan at rest was then recorded. Sublingual glyceryl trinitrate spray (400 ␮g) was administered and 3 to 4 minutes later the last scan was performed. Electrocardiography was monitored continuously throughout the study. FMD was calculated using the ratio: diameter of brachial artery after cuff deflation to the diameter measured at rest.17 The reproducibility of the FMD measurements was determined by repeating the measurements on 20 subjects on consecutive days by the same observer. The mean difference in FMD between the first and second measurements was 0.05, and the SD was 1.23. All 3 preclinical atherosclerotic markers were assessed on the same day on the right side (right common carotid artery, right brachial artery, right radial artery) by the same observer (RR). Statistical analysis: Student’s t test or analysis of variance as appropriate was used for comparing mean values of selected variables in diabetic and nondiabetic subjects and the chi-square test was used for comparing proportions. Pearson’s correlation analysis was carried out to determine the correlation of AI and FMD with vari704 THE AMERICAN JOURNAL OF CARDIOLOGY姞

VOL. 90

RESULTS

Table 1 lists the baseline characteristics of the study groups. Systolic and diastolic blood pressures were higher among the diabetic subjects compared with their nondiabetic counterparts. Mean values of IMT and AI values were significantly higher among diabetic subjects compared with nondiabetic subjects. The mean FMD value among the diabetic subjects was significantly lower compared with nondiabetic subjects. None of the other parameters showed any significant difference between the study groups. The AI values were significantly higher among the diabetic subjects at all age points compared with the nondiabetic subjects. At age ⱕ50 years, mean AI of nondiabetic subjects was 16.7 ⫾ 10.1% and AI of diabetic subjects was 25.05 ⫾ 7.51% (p ⫽ 0.005); at age 51 to 60 years, nondiabetic subjects had an AI of 21.67 ⫾ 5.71% and diabetic subjects had an AI of 27.58 ⫾ . 6.17% (p ⫽ 0.026). At 61 to 70 years of age, nondiabetics subjects’ AI was 21.70 ⫾ 6.59% and diabetic subjects’ AI was 33.11 ⫾ 5.69% (p ⬍0.001). A reverse trend was noted in the distribution of FMD values (i.e., lower in diabetic subjects) with a significant difference at all age points between the diabetic and nondiabetic groups. In subjects with age ⱕ50 years, mean FMD among nondiabetic subjects was 9.15 ⫾ 3.27%, and in diabetic subjects, the mean FMD was 2.9 ⫾ 3.60% (p ⬍0.001). At 51 to 60 years of age, nondiabetic subjects had a mean FMD of 7.92 ⫾ 3.73%; for diabetic subjects, the mean FMD was 1.50 ⫾ 2.07% (p ⬍0.001). At 61 to 70 years of age, nondiabetic subjects had a mean FMD of 3.00 ⫾ 3.45%, whereas diabetic subjects had a mean FMD of 0.56 ⫾ 1.62% (p ⫽ 0.012). Figure 1 shows the distribution of AI according to the duration of diabetes in comparison with the control group. Diabetic subjects had significantly higher values of AI compared with nondiabetic subjects (19.10 ⫾ 8.19%) even with diabetes duration of ⬍5 years (p ⬍0.001). The mean AI values were as follows: in OCTOBER 1, 2002

tive correlation with IMT in the total population (p ⫽ 0.001) and nondiabetic subjects (p ⫽ 0.001) but not in diabetic subjects. Figure 3 presents the correlation of AI and FMD with the IMT quartiles in the total study population. AI values increased with an increase in IMT quartiles (quartile I 21.41 ⫾ 8.60%; quartile II 23.13 ⫾ 9.42%; quartile III 26.16 ⫾ 9.12%; and quartile IV 29.69 ⫾ 7.40%). The AI values in the fourth quartile (p ⫽ 0.003) were significantly higher compared with the first quartile. The FMD values decreased with each IMT quartile (quartile I 5.68 ⫾ 4.41%; quartile II 4.26 ⫾ 4.95%; quartile III 3.26 ⫾ 3.53%; quartile IV 0.62 ⫾ 1.56%) The FMD values in the third (p ⫽ 0.040) and fourth quartiles (p ⬍0.001) were significantly lower (p ⫽ 0.024) compared FIGURE 2. Duration distribution of FMD in diabetic subjects compared with nondiabetic subjects. *p <0.05; **p <0.001 compared with nondiabetic subjects. with the first quartile. When we analyzed the data separately in diabetic and nondiabetic subjects, similar subjects with diabetes, at duration of ⬍5 years, the trends were seen (data not shown), but the values did value was 27.95 ⫾ 8.36%, at 6- to 10-year duration, not reach statistical significance, perhaps due to small the AI value was 28.00 ⫾ 7.34%, at 11- to 15-year numbers. duration, the AI value was 30.50 ⫾ 7.64%, and at ⬎15 Multivariate linear regression analysis was done years duration, the value was 31.60 ⫾ 1.97%. using AI and FMD as dependent variables; the results Figure 2 shows that the mean FMD values were are listed in Table 3. In the total population, age also lower in diabetic subjects at all durations of showed a positive association with AI (p ⬍0.0001) disease compared with controls who had a mean FMD and a negative association with FMD (p ⬍0.0001). of 6.64 ⫾ 4.38%. The mean FMD values in subjects HbA1c also showed an association with AI (p with diabetes were as follows: at duration of ⬍5 years, ⬍0.001) and FMD (p ⬍0.0001), whereas diastolic the mean FMD was 2.14 ⫾ 3.18%; at 6- to 10-year blood pressure showed an association only with AI (p duration, the mean FMD was 1.83 ⫾ 2.58%; at 11- to ⫽ 0.029). Among nondiabetic subjects, FMD showed 15-year duration, the mean FMD value was 1.60 ⫾ an association with age (p ⫽ 0.002), whereas AI 2.30%; and at ⬎15 years duration, the mean FMD showed no association with any of the parameters value was 1.10 ⫾ 2.86%. studied. Among diabetic subjects, FMD showed an Table 2 lists the results of the Pearson correlation association with age, diastolic blood pressure, and analysis of AI and FMD with risk factors of coronary serum triglycerides, whereas AI only showed an asartery disease. In the total population, AI showed a sociation with age. positive association with age, diastolic blood pressure, fasting plasma glucose, and HbA1c. FMD showed a negative correlation with age, systolic and diastolic DISCUSSION The results of the study suggest that diabetic pablood pressures, fasting plasma glucose, and glycosytients have decreased FMD and increased arterial stifflated hemoglobin. In nondiabetic subjects, AI showed a positive cor- ness compared with age- and sex-matched nondiabetic relation with fasting plasma glucose, whereas FMD subjects. Furthermore, the functional changes of athshowed a negative correlation with age and systolic erosclerosis showed a strong correlation with strucblood pressure. In diabetic subjects, only age corre- tural changes. This study is unique as it compares 3 preclinical atherosclerotic markers in diabetic and lated with AI and FMD. AI and FMD were inversely correlated to each nondiabetic subjects. We observed a strong correlation between AI and other in the total population (p ⬍0.001) and among diabetic (p ⫽ 0.012) patients, whereas among nondi- FMD. We also observed that the mean values of AI abetic subjects, the correlation did not reach statistical and FMD were significantly different among the diabetic compared with nondiabetic subjects, even at a significance. AI showed a correlation with IMT in the total younger age. While categorizing the diabetic patients population (p ⫽ 0.001) and nondiabetic subjects (p ⫽ according to the duration of diabetes, we observed AI 0.038) but not in diabetic subjects. FMD had a nega- values to be higher and FMD values to be lower even CORONARY ARTERY DISEASE/FMD AND ARTERIAL STIFFNESS IN SOUTH INDIANS

705

TABLE 2 Pearson’s Correlation of Augmentation Index (AI) and Flow-mediated Dilation (FMD) With Risk Factors of Coronary Artery Disease Total Population (n ⫽ 100) Risk factors Age Body mass index Waist hip ratio Systolic blood pressure Diastolic blood pressure Fasting plasma glucose HbA1c Serum cholesterol Serum triglycerides HDL cholesterol LDL cholesterol IMT AI

Nondiabetic Subjects (n ⫽ 50)

Diabetic Subjects (n ⫽ 50)

AI

FMD

AI

FMD

AI

FMD

r ⫽ 0.335 p ⬍0.001 r ⫽ 0.135 p ⫽ 0.182 r ⫽ 0.035 p ⫽ 0.728 r ⫽ 0.176 p ⫽ 0.079 r ⫽ 0.225 p ⫽ 0.024 r ⫽ 0.313 p ⫽ 0.002 r ⫽ 0.335 p ⫽ 0.001 r ⫽ 0.003 p ⫽ 0.980 r ⫽ 0.111 p ⫽ 0.273 r ⫽ ⫺0.077 p ⫽ 0.448 r ⫽ 0.033 p ⫽ 0.741 r ⫽ 0.334 p ⫽ 0.001 —

r ⫽ ⫺0.409 p ⬍0.001 r ⫽ 0.079 p ⫽ 0.434 r ⫽ ⫺0.176 p ⫽ 0.079 r ⫽ ⫺0.245 p ⫽ 0.014 r ⫽ ⫺0.243 p ⫽ 0.015 r ⫽ ⫺0.339 p ⫽ 0.001 r ⫽ ⫺0.335 p ⫽ 0.001 r ⫽ 0.044 p ⫽ 0.665 r ⫽ 0.027 p ⫽ 0.790 r ⫽ 0.056 p ⫽ 0.581 r ⫽ 0.018 p ⫽ 0.863 r ⫽ ⫺0.336 p ⫽ 0.001 r ⫽ ⫺0.388 p ⬍0.001

r ⫽ 0.216 p ⫽ 0.132 r ⫽ 0.253 p ⫽ 0.059 r ⫽ ⫺0.133 p ⫽ 0.356 r ⫽ 0.103 p ⫽ 0.476 r ⫽ 0.158 p ⫽ 0.273 r ⫽ 0.417 p ⫽ 0.003 r ⫽ ⫺0.164 p ⫽ 0.255 r ⫽ ⫺0.052 p ⫽ 0.722 r ⫽ ⫺0.164 p ⫽ 0.254 r ⫽ 0.271 p ⫽ 0.057 r ⫽ ⫺0.060 p ⫽ 0.681 r ⫽ 0.294 p ⫽ 0.038 —

r ⫽ ⫺0.569 p ⬍0.001 r ⫽ ⫺0.262 p ⫽ 0.053 r ⫽ ⫺0.206 p ⫽ 0.150 r ⫽ ⫺0.341 p ⫽ 0.015 r ⫽ ⫺0.127 p ⫽ 0.380 r ⫽ ⫺0.028 p ⫽ 0.845 r ⫽ 0.050 p ⫽ 0.728 r ⫽ 0.085 p ⫽ 0.557 r ⫽ ⫺0.036 p ⫽ 0.805 r ⫽ 0.046 p ⫽ 0.752 r ⫽ 0.104 p ⫽ 0.471 r ⫽ ⫺0.471 p ⫽ 0.001 r ⫽ ⫺0.011 p ⫽ 0.937

r ⫽ 0.589 p ⬍0.001 r ⫽ ⫺0.183 p ⫽ 0.203 r ⫽ ⫺0.039 p ⫽ 0.790 r ⫽ 0.203 p ⫽ 0.157 r ⫽ 0.009 p ⫽ 0.950 r ⫽ ⫺0.119 p ⫽ 0.412 r ⫽ 0.023 p ⫽ 0.728 r ⫽ ⫺0.005 p ⫽ 0.970 r ⫽ ⫺0.199 p ⫽ 0.198 r ⫽ 0.043 p ⫽ 0.767 r ⫽ 0.079 p ⫽ 0.587 r ⫽ 0.227 p ⫽ 0.112 —

r ⫽ ⫺0.395 p ⫽ 0.005 r ⫽ 0.037 p ⫽ 0.801 r ⫽ ⫺0.110 p ⫽ 0.448 r ⫽ ⫺0.080 p ⫽ 0.579 r ⫽ ⫺0.119 p ⫽ 0.411 r ⫽ 0.003 p ⫽ 0.985 r ⫽ 0.077 p ⫽ 0.597 r ⫽ 0.082 p ⫽ 0.572 r ⫽ 0.997 p ⫽ 0.061 r ⫽ ⫺0.136 p ⫽ 0.345 r ⫽ ⫺0.012 p ⫽ 0.933 r ⫽ ⫺0.084 p ⫽ 0.564 r ⫽ ⫺0.354 p ⫽ 0.012

HDL ⫽ high-density lipoprotein; LDL ⫽ low-density lipoprotein.

Although AI and FMD represent “functional atherosclerosis,” IMT represents the early stages of “structural atherosclerosis.”19 In this study, the functional changes in the artery, namely, endothelial dysfunction and arterial stiffness, showed a strong correlation with structural changes in the artery (IMT), which in turn has been shown to be a risk factor for myocardial infarction and stroke.20 The Second Manifestations of ARTerial disease (SMART) Study21 examined IMT and arterial stiffness in 570 patients and suggested that IMT and arterial stiffness are clear markers of cardiovascular risk. A study on Japanese subjects has also shown a strong correlation between structural (IMT) and functional changes (arterial FIGURE 3. Correlation of endothelial function and arterial stiffness with IMT in the stiffness) of the artery,20 but endothelial total study population. *p ⴝ 0.04; **p ⴝ 0.003; ***p <0.001 compared with the first quartile. dysfunction was not studied. In summary, our study suggests among subjects with short duration of diabetes. This that increased arterial stiffness and impaired endothesupports the view that the clock starts “ticking” for lial function are seen among diabetic subjects comatherosclerosis even before the stage of overt diabetes pared with age- and sex-matched nondiabetic subjects. is reached.18 However, statistically significant differ- Age and HbA1c are the predominant risk factors for ences for AI and FMD values were not observed when both arterial stiffness and endothelial dysfunction. subjects with long-term diabetes were compared with Early screening for diabetes and correction of risk those with shorter durations of diabetes. This could factors, perhaps in the prediabetic stage itself, may help to delay or postpone the onset of atherosclerosis. probably be due to the small study numbers. 706 THE AMERICAN JOURNAL OF CARDIOLOGY姞

VOL. 90

OCTOBER 1, 2002

TABLE 3 Multivariate Regression Analysis Using Augmentation Index (AI) and Flow-mediated Dilation (FMD) as Dependent Variables Total Population (n ⫽ 100) Risk factors Age Body mass index Systolic blood pressure Diastolic blood pressure HbA1c Serum cholesterol Serum triglycerides HDL cholesterol

Nondiabetic Subjects (n ⫽ 50)

Diabetic Subjects (n ⫽ 50)

AI

FMD

AI

FMD

AI

FMD

␤ ⫽ 0.316 p ⬍0.0001 ␤ ⫽ 0.177 p ⫽ 0.419 ␤ ⫽ 0.09 p ⫽ 0.336 ␤ ⫽ 0.351 p ⫽ 0.029 ␤ ⫽ 1.850 p ⬍0.001 ␤ ⫽ 0.005 p ⫽ 0.810 ␤ ⫽ 0.067 p ⫽ 0.923 ␤ ⫽ 0.015 p ⫽ 0.111

␤ ⫽ ⫺0.192 p ⬍0.0001 ␤ ⫽ 0.027 p ⫽ 0.790 ␤ ⫽ ⫺0.023 p ⫽ 0.598 ␤ ⫽ ⫺0.132 p ⫽ 0.069 ␤ ⫽ ⫺0.913 p ⬍0.0001 ␤ ⫽ 0.0075 p ⫽ 0.501 ␤ ⫽ 0.043 p ⫽ 0.289 ␤ ⫽ 0.004 p ⫽ 0.352

␤ ⫽ 0.183 p ⫽ 0.237 ␤ ⫽ ⫺0.368 p ⫽ 0.318 ␤ ⫽ ⫺0.139 p ⫽ 0.421 ␤ ⫽ 0.331 p ⫽ 0.137 ␤ ⫽ ⫺1.725 p ⫽ 0.550 ␤ ⫽ 0.013 p ⫽ 0.750 ␤ ⫽ 0.006 p ⫽ 0.694 ␤ ⫽ 0.131 p ⫽ 0.303

␤ ⫽ ⫺0.216 p ⫽ 0.002 ␤ ⫽ 0.214 p ⫽ 0.216 ␤ ⫽ 0.025 p ⫽ 0.761 ␤ ⫽ 0.039 p ⫽ 0.706 ␤ ⫽ ⫺0.339 p ⫽ 0.802 ␤ ⫽ 0.008 p ⫽ 0.681 ␤ ⫽ 0.002 p ⫽ 0.759 ␤ ⫽ 0.069 p ⫽ 0.242

␤ ⫽ 0.475 p ⬍0.001 ␤ ⫽ ⫺0.133 p ⫽ 0.572 ␤ ⫽ 0.065 p ⫽ 0.467 ␤ ⫽ 0.114 p ⫽ 0.676 ␤ ⫽ 0.608 p ⫽ 0.208 ␤ ⫽ 0.025 p ⫽ 0.320 ␤ ⫽ 0.013 p ⫽ 0.226 ␤ ⫽ 0.041 p ⫽ 0.669

␤ ⫽ ⫺0.134 p ⫽ 0.003 ␤ ⫽ 0.023 p ⫽ 0.816 ␤ ⫽ 0.072 p ⫽ 0.057 ␤ ⫽ ⫺0.158 p ⫽ 0.029 ␤ ⫽ 0.043 p ⫽ 0.829 ␤ ⫽ 0.004 p ⫽ 0.689 ␤ ⫽ 0.009 p ⫽ 0.039 ␤ ⫽ 0.045 p ⫽ 0.253

Abbreviation as in Table 2.

Acknowledgment: We are grateful to John R. Cockcroft, FRCP, Department of Cardiology, University of Wales College of Medicine, Cardiff, United Kingdom, and David S. Celermajer, PhD, Department of Cardiology, Royal Prince Alfred Hospital, Sydney, Australia, for their help in setting up the arterial stiffness and endothelial function studies at our center.

1. King H, Aubert RE, Herman WH. Global burden of diabetes, 1995–2025— prevalence;numerical estimates and projections. Diabetes Care 1998;2:1414 – 1431. 2. Mckeigue PM, Miller GJ, Marmot MG. Coronary artery disease in south Asian overseas: a review. J Clin Epidemiol 1989;42:597–609. 3. Reddy KS, Yusuf S. Emerging epidemic of coronary artery disease in developing countries. Circulation 1997;97:596 –601. 4. Enas EA, Yusuf S, Mehta J. Prevalence of coronary artery disease in Asian Indians. Am J Cardiol 1992;70:945–949. 5. Anand SS, Yusuf S, Vuksan V, Devanesen S, Teo KK, Montague PA, Kelemen L, Yi C, Lonn E, Gerstein H, Hegele RA, McQueen M. Differences in risk factors, atherosclerosis, and cardiovascular disease between ethnic groups in Canada: the Study of Health Assessment and Risk in Ethnic groups (SHARE). Lancet 2000;356:279 –284. 6. Shanthirani CS, Rema M, Deepa R, Premalatha G, Ravikumar R, Mohan A, Sastry NG, Ramu M, Saroja R, Kayalvizhi G, Mohan V. The Chennai Urban Population Study (CUPS)—methodological details—(CUPS Paper No.1). Int J Diabet Develop Countries 1999;19:149 –157. 7. Mohan V, Shanthirani CS, Deepa R, Premalatha G, Sastry NG, Saroja R. Intra-urban differences in the prevalence of the metabolic syndrome in southern India —the Chennai Urban Population Study (CUPS No. 4). Diabet Med 2001; 18:280 –287. 8. Premalatha G, Shanthirani CS, Deepa R, Markovitz J, Mohan V. Prevalence and risk factors of peripheral vascular disease in a selected South Indian population: the Chennai Urban Population Study (CUPS). Diabetes Care 2000;23: 1295–1300. 9. Mohan V, Deepa R, Shanthirani CS, Premalatha G. Prevalence of coronary artery disease and its relationship to lipids in a selected population in South India—The Chennai Urban population Study (CUPS No. 5). J Am Coll Cardiol 2001;38:682–687.

10. Mohan V, Ravikumar R, Shanthirani CS, Deepa R. Intimal medial thickness

of the carotid artery in South Indian diabetic and non-diabetic subjects: the Chennai Urban Population Study (CUPS). Diabetologia 2000;43:494 –99. 11. Alberti KGMM, Zimmet PZ. Definition diagnosis and classification of diabetes mellitus and its complications. Part 1: Diagnosis and classification of diabetes mellitus, provisional report of a WHO Consultation. Diabet Med 1998; 15:539 –553. 12. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low density lipoprotein cholesterol in plasma without use of the preparative ultracentrifuge. Clin Chem 1972;18:499 –502. 13. Pignoli P, Longo T. Ultrasound evaluation of atherosclerosis. Methodological problems and technological developments. Eur Surg Res 1986;18:238 –253. 14. Yamasaki Y, Kawamori R, Matsushima H, Nishizawa H, Kodama M, Kajimoto Y, Morishima T, Kamada T. Atherosclerosis in carotid artery of young IDDM patients monitored by ultrasound high resolution B-mode imaging. Diabetes 1994;43:634 –639. 15. Wilkinson IB, Fuchs SA, Jansen IM, Spratt JC, Murray GD, Cockcroft JR, Webb DJ. Reproducibility of pulse wave velocity and augmentation index measured by pulse wave analysis. J Hypertension 1998;16:2079 –2084. 16. Wilkinson JB, Prasad K, Hall IR, Thomas A, MacCallum H, Webb DJ, Frenneaus MP, Cockcroft JR. Increased central pulse pressure and augmentation index in subjects with hypercholesterolemia. J Am Coll Cardiol 2002;39:1005– 1011. 17. Celermajer DS, Sorensen KE, Gooch VM, Spiegelhalter DJ, Miller OI, Sullivan ID, Lloyd JK, Deanfield JE. Non invasive detection of endothelial dysfunction in children and adults at risk of atherosclerosis. Lancet 1992;340: 1111–1115. 18. Haffner SM, Stern MP, Hazuda HP, Mitchell BD, Patterson JK. Cardiovascular risk factors in confirmed prediabetic individuals. Does the clock for coronary heart disease start ticking before the onset of clinical diabetes? JAMA 1990;263:2893–2898. 19. Taniwaki H, Kawagishi T, Emoto M, Shoji T, Kanda H, Maekawa K, Nishizawa Y, Morii H. Correlation between the intima-media thickness of the carotid artery and aortic pulse wave velocity in patients with type 2 diabetes. Vessel wall properties in type 2 diabetes. Diabetes Care 1999;22:1851–1857. 20. O’Leary DH, Polak JF, Kronmal RA, Manolio TA, Burke GL, Wolfson SK Jr. Carotid artery intima and media thickness as a risk factor for myocardial infarction and stroke in older adults. N Engl J Med 1999;340:14 –22. 21. Simons PC, Algra A, Bots ML, Grobbee DE, van der Graaf Y. Common carotidintima-media thickness and arterial stiffness: indicators of cardiovascular risk in high-risk patients. The SMART Study (Second Manifestations of ARTerial disease). Circulation 1999;100:951–957.

CORONARY ARTERY DISEASE/FMD AND ARTERIAL STIFFNESS IN SOUTH INDIANS

707