Assessment of oxidative stress and endothelial dysfunction in Asian Indians with type 2 diabetes mellitus with and without macroangiopathy

Transcription

1 Q J Med 2008; 101: doi: /qjmed/hcn020 Advance Access published on 1 April 2008 Assessment of oxidative stress and endothelial dysfunction in Asian Indians with type 2 diabetes mellitus with and without macroangiopathy N. SINGHANIA 1, D. PURI 1, S.V. MADHU 2 and S.B. SHARMA 1 From the 1 Department of Biochemistry and 2 Department of Medicine, University College of Medical Sciences, Shahdara, Delhi, India Received 12 September 2007 and in revised form 11 January 2008 Summary Background: Enhanced oxidative stress coupled with increased expression of adhesion molecules (e.g. VCAM-1, ICAM-1) and decreased nitric oxide (NO) levels are implicated in development of atheromatous vascular lesion in diabetes. The present study addresses the correlation between oxidative stress, vascular cell adhesion molecules-1 (VCAM- 1), NO end products and macroangiopathic complications in type 2 diabetes mellitus (DM). Design and Methods: The study population consisted of three groups (i) diabetic patients with macroangiopathy (Group I); (ii) diabetic patients without macroangiopathy (Group II) and (iii) healthy controls (Group III) (n = 30, each group). Results: Serum malondialdehyde(mda) concentration was significantly higher in diabetic patients as compared to controls. Group I had significantly higher malondialdehyde level as compared to Group II (P < 0.05) ( mmol/l vs mmol/l), suggesting higher oxidative stress in patients with macroangiopathy. Significant reduction in NO end products was observed in diabetic patients compared to controls. Levels of serum NO end products levels were further reduced in patients with macroangiopathy compared to those without macroangiopathy. Group I ( ng/ml) and Group II ( ng/ml) had significantly higher level of svcam-1 compared to healthy controls ( ng/ml) (P < 0.001). Also, patients with macroangiopathy had significantly higher levels of svcam-1 compared to those without macroangiopathy (P < 0.05). Multiple regression analysis indicated that post-prandial blood glucose, GSH and MDA were independent predictors of svcam-1 level (R = 0.690, P = 0.000). Conclusion: It can be concluded from the present study that an enhanced oxidative stress coupled with endothelial dysfunction as indicated by reduced activity of NO pathway and enhanced expression of svcam-1 play an important intermediary role in the pathogenesis of macrovascular complications in type 2 DM. Introduction Macroangiopathic complications such as coronary artery disease (CAD), peripheral vascular disease (PVD) and cerebrovascular disease (CVD) are important causes of morbidity and mortality in type 2 diabetes mellitus (DM). Conventional risk factors contribute similarly to macrovascular complications in patients with type 2 diabetes and non-diabetic subjects; therefore, other explanations have been sought for enhanced atherosclerosis in type 2 diabetes. Cardiovascular complications correlate only weakly with duration and severity of diabetes 1 suggesting hyperglycaemia per se is not responsible for development of diabetic macroangiopathy. It has been suggested that hyperglycaemia Address correspondence to Dr D. Puri, B-163, Yojana Vihar, Delhi , India. dineshpuri21@yahoo.co.in! The Author Published by Oxford University Press on behalf of the Association of Physicians. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 450 N. Singhania et al. can increase free radical activity through various pathways strictly associated with hyperglycaemia (polyol pathway, protein glycation, glucose autooxidation and prostanoid synthesis). 2 Vascular endothelial cells are the major targets of hyperglycaemic damage in which oxygen free radicals (OFRs) are thought to play an important role. These OFRs can in turn shift endothelial functions from antiatherogenic to pro-atherogenic. Vascular endothelial dysfunction represents an early potentially reversible event in the pathology of atherosclerosis. Nitric oxide (NO) synthesized by endothelium mediates vasorelaxation and inhibits platelet adhesion. Earlier studies conducted for elucidating the cellular and molecular mechanisms of macroangiopathic diabetic complications suggest that enhanced generation of free radicals in diabetes leads to structural and functional abnormalities of diabetic macroangiopathy. 2 Reactive oxygen species (ROS) may modify endothelial functions by a variety of mechanisms such as peroxidation of membrane lipids, activation of transcription factor (NF-kB) leading to upregulation of adhesion molecules and interference with availability of NO. 3 Adhesion molecules promote the earliest stage of atherosclerosis. Very high levels of soluble vascular cellular adhesion molecules-1 (svcam-1) have been reported in diabetic patients with atherosclerosis and it has been suggested that they may be used as an index of endothelial activation and even a molecular marker of atherosclerosis. 4 Earlier studies have shown that mortality and morbidity from CAD are higher in people of South Asian descent settled overseas than in other groups. 5,6 While there are several studies in literature that compare oxidative stress and endothelial function parameters in diabetic and non-diabetic subjects, very few studies 7,8 have attempted to specifically address these issues in type 2 diabetic patients with macroangiopathy and none, to the best of our knowledge, in carefully selected patients with diabetic macroanigopathy. Also, data on the role of oxidative stress and svcam-1 in diabetic macroangiopathy is scant in patients of Asian Indian ethnic origin. Therefore, the present study aims to undertake a comprehensive comparative study of oxidative stress, anti-oxidant defense system parameters, as well as biochemical markers of endothelial function (NO end products and svcam-1) in patients of type 2 DM with and without macroangiopathy. This would help in better understanding of the relationship between oxidative stress and diabetic macroangiopathy in type 2 DM and the role of svcam-1 concentration as an early marker of diabetic atherosclerotic vascular disease. Such studies may form the basis for development of novel therapeutic approach for arresting the progression of macroangiopathy in type 2 DM. Materials and methods Patients Study was carried out in 90 subjects above the age of 30 years, which included 30 type 2 diabetic patients with one or more of the following macroangiopathic complications (Group I), i.e. CAD, PVD and CVD; 30 type 2 diabetic patients without any evidence of macroangiopathy (Group II) and 30 healthy controls (Group III). Diabetic patients were selected from the diabetic clinic in Guru Teg Bhadur (GTB) Hospital, Delhi and controls were selected from the staff working in GTB Hospital or University College of Medical Sciences. All the subjects were matched for age, sex and body mass index (BMI). Duration of diagnosed diabetes exceeded more than 1 year. Diagnosis of diabetes was made according to revised American Diabetes Association criteria. CAD was defined on the basis of history, clinical examination, ECG findings, positive tread meal test (TMT) or coronary angiography. PVD defined on the basis of history, clinical examination and ankle brachial index (ABI), determined by hand held vascular Doppler, of <0.90. CVD was defined on the basis of history of stroke. Patients with history of episode of cerebrovascular accidents (CVA) or CAD within last 3 months; patients having acute infection anywhere in body, with nephropathy as evidenced by microalbuminuria (>20 mg/l) and/or raised blood urea, serum creatinine levels, with proliferative retinopathy and those on insulin therapy, anti-oxidant drugs or hypolipidaemic drugs were excluded. All the subjects gave informed consent and the study was approved by local Ethical Committee. Biochemical assays The presence or absence microalbuminuria was determined using commercially available dipstick (Roche diagnostics, Mannheim, Germany). Blood samples were collected after an overnight fast of 12 h and the sample for 2 h plasma glucose (2hPG) was collected after their usual breakfast meal. Plasma glucose fasting and 2 h post-prandial (2hPG), renal profile and lipid profile was estimated by using kits from Accurex Biomedical Pvt. Ltd. Mumbai, India. Glycated haemoglobin was estimated by ion exchange chromatography using kits from Transasia, Daman, India. Serum insulin was estimated by ELISA technique using commercial kits from Biochem Immunosystems, Italy.

3 Oxidative stress and endothelial dysfunction in type 2 DM 451 Table 1 Clinical characteristics in three study groups (in each group, n = 30) Group I Group II Group III Age (years) Sex (male/female) 12/18 11/19 12/18 BMI (kg/m 2 ) Duration (year) Waist to hip ratio a b ABI Hypertension H/O Smoking 4 4 Background diabetic retinopathy 1 2 Oral antidiabetic agents Sulfonylurea Biguanides 2 5 Sulfonylurea + Biguanides Value are expressed as mean SD. a Group I vs. Group III (P < 0.05, S). b Group II vs. Group III (P < 0.05, S). Table 2 Baseline biochemical profile in three study groups (in each group, n = 30) Serum MDA levels were measured as an index of lipid peroxidation using the colorimetric methods described by Satoh. 9 Reduced glutathione (GSH) was estimated by the method of Beutler et al. 10 Activity of superoxide dismutase (SOD) in erythrocytes was measured by method described by Marklund and Marklund 11 as modified by Nandi and Chatterjee. 12 Erythrocyte catalase activity was determined by method of Sinha. 13 Serum nitrite levels were determined by a colorimetric method based on Griess reaction. 14 Nitrate was measured as nitrite after enzymatic reduction by an improved method described by Guevara et al. 15 svcam-1 was estimated by ELISA, using kits from Diaclone Research, France. Statistical analysis All the results expressed as mean SD. All parameters among the three groups were compared by one way analysis of variance. Pearson correlation Group I Group II Group III Fasting plasma glucose (mmol/l) a b Post-prandial plasma glucose (mmol/l) a b Glycosylated haemoglobin (%) a b Insulin (miu/ml) a Total serum cholesterol (mmol/l) a b Triglyceride (mmol/l) a b HDL cholesterol (mmol/l) a b LDL cholesterol (mmol/l) a b Value are expressed as mean SD. a Group I vs. Group III (P < 0.05, S). b Group II vs. Group III (P < 0.05, S). coefficients were determined and multiple regression analysis was performed where stated. Results Demographic and baseline biochemical parameters in type 2 diabetic patients with macroangiopathy (Group I), type 2 diabetic patients without macroangiopathy (Group II) and healthy controls (Group III), are shown in Tables 1 and 2, respectively. The subjects in the three groups matched for age, sex and BMI. Duration of diabetes was similar in both the diabetic groups as was the level of glycaemia. In macroangiopathy group, 13 patients had only PVD, three patients had CVD and 12 patients had CAD. In the CAD group, six patients also had PVD and two also had CVD. Two patients had all the three complications. All 100% patients in macroangiopathy group (Group 1) and 27/30 (90%) patients in the Group 2 were receiving oral

4 452 N. Singhania et al. Table 3 Comparison of levels of markers of oxidative stress and endothelial dysfunction in three study groups (in each group, n = 30) Group I Group II Group III SOD (U/gmHb) a b Catalase (mmolh 2 O 2 /min/ghb) a b GSH (mg/ghb) a b MDA (nmol/ml) a,c b Serum nitrite (mmol/l) a Serum nitrate (mmol/l) a,c b svcam-1 (ngm/ml) a,c b Values are expressed as mean SD. a Group I vs. Group III (P < 0.05). b group II vs. Group III (P < 0.05). c Group I vs. group II (P < 0.05). anti-diabetic drugs. Fifteen vs. 10 patients were on sulfonyureas only, two vs. five on biguanides only and 13 vs. 12 on a combination of sulfonylurea and biguanides in both the groups, respectively. Similarly, 18/30 (60%) patients in the Group 1 and 20/30 (66.6%) patients in Group 2 were taking antihypertensive treatment. None of the diabetic patients belonging to either Group 1 or Group 2 was actually taking lipid lowering agents, even where prescribed as they could not afford them. Serum malondialdehyde (MDA) concentration was significantly higher in the diabetic patients of Group I ( mmol/l) and Group II ( mmol/l) as compared to healthy controls ( mmol/l). Between the two diabetic groups, the Group I had significantly higher serum MDA than Group II (P < 0.05), suggesting higher oxidative stress in the patients with macroangiopathy. Activity of erythrocyte SOD was significantly reduced in both the diabetic groups ( U/gHb) in Group I and ( U/gHb) in Group II compared to controls ( U/gHb) (Table 3). Difference in the SOD activity between Group I and Group II, however, was not significant. Activity of erythrocyte catalase was likewise reduced in diabetic patients with and without macroangiopathy compared to control (P < 0.05). But statistically, no significant difference in levels of catalase between Group I ( mmol H 2 O 2 / min/ghb) and Group II ( mmol H 2 O 2 / min/ghb) was observed. Erythrocyte GSH content was significantly lower in the diabetic patients of Group I ( mg/ ghb) and Group II ( mg/ghb) compared to controls ( mg/ghb), P < But no significant difference in the GSH levels between the two diabetic groups was observed. Serum nitrite + nitrate levels were significantly decreased in Group I Table 4 Correlation of glycated hemoglobin and svcam-1 with other parameters (n = 90) Parameters Glycated hemoglobin svcam-1 r P r P Fasting plasma <0.001 glucose Post-prandial <0.001 plasma glucose Glycosylated haemoglobin SOD Catalase GSH MDA <0.001 Serum <0.001 nitrite + nitrate svcam ( mmol/l) and Group II ( mmol/l) as compared to controls ( mmol/l), P < Levels of nitrite + nitrate were further reduced in patients with macroangiopathy compared to those without macroangiopathy. Group I ( ngm/ml) and Group II ( ngm/ml) had significantly higher level of svcam-1 compared to healthy controls ( ngm/ml) (P < 0.001). Group I had further higher levels of svcam-1 compared to those without macroangiopathy (P < 0.05) as shown in Table 3. Correlation of glycosylated haemoglobin and VCAM-1 with other parameters is shown in Table 4. Multiple regression analysis indicated that 2hPG, GSH and MDA were independently related to VCAM-1 levels (R = 0.690, P = 0.000). In multiple regression analysis no association of glycosylated

5 Oxidative stress and endothelial dysfunction in type 2 DM 453 haemoglobin with oxidative stress, VCAM-1 and nitrite was observed. Discussion The current study found significantly higher levels of lipid peroxidation (MDA) and a greater endothelial dysfunction (high svcam-1 and low NO end products) in Asian Indian patients with diabetic macroangiopathy compared to those without macroangiopathy and healthy controls. The anti-oxidant defense system was found to be significantly impaired in all patients with type 2 DM with or without macroangiopathy compared to controls. Lipid peroxidation and endothelial dysfunction showed a graded increase from healthy controls to diabetic patients without macroangiopathy to those with macroangiopathy. These findings support the view that increased oxidative load in the face of a compromised anti-oxidant defense system triggers endothelial dysfunction that could mediate atherogeneis in type 2 DM. Migdalis et al. 7 also reported high MDA level in type 2 diabetic patients with macroangiopathy than those without macroangiopathy. But in this study, duration of diabetes, fasting glucose levels and HbA1c were significantly higher in diabetic patients with macroangiopathy than those without macroangiopathy. In contrast, in our study, these parameters were comparable in both the diabetic groups (Table 1), still only Group 1 developed macroangiopathy. It would thus appear that in type 2 diabetics the glycaemic status is not the only factor that determines development of macroangiopathy. Wolff 16 had earlier suggested oxidative stress is often responsible for diabetic complication irrespective of glycaemic status due to variability in individual resistance to the stress. Thus, it can be suggested from the study that subgroup of diabetic patients with greater net oxidative stress are more at risk for development of macroangiopathy. In diabetes, activity of anti-oxidant enzymes has been reported to be increased, 17,18 decreased 19 or unaltered, 20 in both experimental and clinical studies. These differences are explained by the time of estimation of these enzymes in natural history of diabetes. Initially, the anti-oxidative defenses are activated to compensate for oxidative load, which may be measured as high. Later, enzymes get progressively depleted resulting in normal or reduced levels. In the present study, the finding of low antioxidant defense in diabetic subjects indicates that such depletion would make them vulnerable to oxidative insult. In addition, our study found a progressive increase in oxidative stress from control to diabetic patients without macroangiopathy to those with macroangiopathy. This suggests that against a background of weakened anti-oxidant defense system common to all diabetic patients, higher oxidative stress determines progression to macroangiopathy. Endothelial dysfunction is an early and potentially reversible event in pathology of atherosclerosis. Abnormalities of NO/O 2. pathway occur in diabetes and are important cause of endothelial dysfunction. Serum nitrate and Nitrite + nitrate levels (end products of NO) that reflected NO production, were found to be decreased in type 2 diabetic patients with and without macroangiopathy compared to controls (P < 0.05). Similar results were reported by Vanizor et al. 21 and there was further reduction in the patients with macroangiopathy compared to those without macroangiopathy. Cause of decreased availability of NO in diabetes are: (i) direct quenching of NO by advanced glycation end products; (ii) increased activity of aldose reductase pathway leading to depletion of NADPH, which is a co-factor for NO synthase (NOS) and (iii) O 2. anion can directly quench NO by forming peroxynitrite ion, which can reduce activity of endothelial NOS by oxidizing its co-factor, tetrahydrobiopterin. 22 As observed in the present study, production of free radicals is increased in patients with diabetic macroangiopathy, which result in a greater interaction between NO and ROS, resulting in inactivation of NO. Various known effects of NO (inhibition of platelet aggregation, vascular smooth muscle cell proliferation, inhibition of expression of adhesion molecules and abrogation of the ability of monocytes to oxidize LDL) suggest that if its availability were reduced this could predispose to atherogenesis. VCAM-1 is a transmembrane glycoprotein, which is a member of the immunoglobulin gene superfamily. VCAM-1 is not expressed on resting endothelium but its expression is upregulated by cytokines. These molecules play important role in adhesion of circulating leucocytes to endothelium, which is the first step in initiation of atherosclerosis. In the present study, circulating level of svcam-1 were significantly high in diabetic patients compared to controls (P < 0.001). Group III had significantly higher concentration of svcam-1 compared to Group II (P < 0.05). These results are in agreement with those of Matsumoto et al. 23 Contrary to our results, Fasching et al. 24 did not find any difference in svcam-1 levels between type 2 diabetic patients with and without macroangiopathy. However, in that study as, suggested by the selection criteria, a number of patients with asymptomatic atherosclerosis might have been placed in without macroangiopathy

6 454 N. Singhania et al. group. While in the present study only those patients who had ABI >1.0 were included in Group II, therefore chances of asymptomatic atherosclerosis was minimal in this group. ABI is one of the important methods of preclinical assessment of atherosclerosis in body. It can detect not only asymptomatic PVD but can also indicate presence of subclinical and asymptomatic atherosclerosis elsewhere in body including cardiovascular disease. 25 Jager et al. 26 reported that levels of svcam-1 were significantly and independently associated with the risk of cardiovascular mortality in type 2 diabetic subjects and therefore might be useful marker for identifying subjects at increased cardiovascular risk. What links diabetes to increased VCAM-1 level is not clear but following mechanisms has been proposed by Schmidt et al. 27 Interaction of AGES with their receptor RAGE may increase superoxide anion production that can activate NF-kB, which in turn can induce VCAM-1 expression. Intracellular thiol redox state has also been observed to be a major regulator of NF-kB response. 28 Molecular cloning of VCAM-1 has provided evidence for NF-kB binding domains in the promoter region of this gene and NF-kB has been reported to respond to OFRs. The effect of NO in reducing generation of OFRs may explain its repression of NF-kB-mediated gene expression. 29 Also, immunoprecipitation studies indicated that NO stabilized the NF-kB/I kba. Oxidized LDL can also induce VCAM-1 expression by modulating NF-kB activity. 30 Thus reduced NO availability, enhanced oxidative stress and altered thiol redox status as evidenced by decreased NO end products level, increased MDA levels reduced GSH contents in diabetic patients in the present study can increase VCAM-1 expression by increasing NF-kB activity. In multiple regression analysis, 2hPG, GSH and MDA were independently related to svcam-1 level but FPG was not independent predictor of SVCAM- 1. To the best of over knowledge, no previous studies have shown significant association between 2hPG and VCAM-1 concentration. Post-meal metabolic excursion and insulin resistance can lead to oxidative stress and decreased nitric oxide availability that can increase VCAM-1 expression (maker of endothelial dysfunction and low grade inflammation), which in turn will lead to initiation of atherogenesis. This would suggest that the strong association of post-prandial hyperglycaemia and cardiovascular disease may be mediated through enhanced expression of svcam-1. Glycosylated haemoglobin had negative correlation with anti-oxidant activity and NO end products and it had positive correlation with svcam-1 and MDA. No correlation between duration of diabetes and these parameters was observed suggesting that prolonged hyperglycaemia per se is not responsible for development of diabetic macroangiopathy but can do so through increased oxidative stress. It may be observed from the above results that though the duration of diabetes and glycaemic control were comparable in both the diabetic group, only Group I developed macroangipathy suggesting that a subgroup of diabetic patients having higher oxidative stress is more prone to develop macroangiopathy due to reduced nitric oxide and enhanced VCAM-1 expression. In conclusion, our study finds that patients of type 2 diabetes with macroangiopathy display higher oxidative stress, reduced NO availability and increased VCAM-1 expression. Acknowledgements This study was supported by a grant from University college of Medical Sciences, Delhi University References 1. Gray RP, Yudkin JS. Cardiovascular diseases in diabetes mellitus. In: Pickup JC, William G, Textbook of Diabetes, vol 2, 2nd edn. Blackwell sciences, Inc, London, Giugliano D, Paolisso G, Ceriello A. Oxidative stress and diabetic vascular complications. Diab care 1996; 19: Stehouwer CDA, Lambert J, Donker AJM, van Hinsbergh VWM. Endothelial dysfunction and pathogenesis of diabetic angiopathy. Cardiovas Res 1997; 34: Otsuki M, Hashimoto K, Morimoto Y, Kishimoto T, Kasayama S. Circulating vascular cell adhesion molecule-1 (VCAM-1) in atherosclerotic NIDDM patients. Diabetes 1997; 46: Mckeigue PM, Miller GJ, Marmot MG. Coronary heart disease in South Asian overseas a review. J Clin Epidemiol 1989; 42: Chaturvedi N, Fuller JH. Ethric differences in mortality from cardiovascular disease in UK: do they persist in people with diabetes? J Epidemiol Commun Health 1996; 50: Migdalis IN, Kalogeropoulou K, Hiopoulou V, Triantafilou P, Charalabides J, Mertzos G, et al. Plasma levels of endothelin, lipid peroxides and prostacyclin in diabetic patients with macroangiopathy. Diab Res Clin Pract 2001; 54: Dyer RG, Stewart MW, Mitcheson J, George K, Alberti MM, Laker MF. 7-Ketocholestrol, a specific indicator of lipoprotein oxidation, and malondialdehyde in non-insulin dependent diabetes and peripheral vascular disease. Clin Chim Acta 1997; 260: Satoh K. Serum lipid peroxides in cerebrovascular disorders determined by new colorimetric method. Clin Chim Acta 1978; 90: Beutler E, Duron O, Kelley B. Improved method for determination of blood glutathione. J Lab Clin Med 1963; 61:882.

7 Oxidative stress and endothelial dysfunction in type 2 DM Marklund S, Marklund G. Human copper containing superoxide dismutase of high molecular weight. Proc Natl Acad Sci USA 1982; 79: Nandi A, Chatterjee IB. Assay of superoxide dismutase activity in animal tissues. J Biosci 1988; 13: Sinha AK. Colorimetric assay of catalase. Anal Biochem 1972; 47: Moshage H, Kok B, Huizenga JR, Jansen PL. Nitrite and nitrate determination in plasma: a critical evaluation. Clin Chem 1995; 41: Guevara I, Iwanejko J, Dembinska Kiec A. Determination of nitrite/nitrate in human biological material by simple Griess reaction. Clin Chim Acta 1998; 274: Wolff SP. Diabetes mellitus and free radicals. Free radicals, transition metals, and oxidative stress in the etiology of diabetes mellitus and complications. Br Med Bull 1993; 49: Giri R, Kesavulu MM, Rao BK, Ramana V, Appa Rao CH. Hyperlipidemia, increased lipid peroxidation and changes in antioxidant enzymes, Na + -K + ATPase in erthythrocytes of type 2 diabetes patients in Andhra Pradesh. Indian J Clin Biochem 1999; 14: Kumar AP, Rajagopal G. Lipid peroxidation in erythrocytes of patients with type 2 diabetes mellitus. Indian J Clin Biochem 2003; 18: Sundaram K, Bhaskar A, Vijayalingam S, Viswanathan M, Mohan R, Shanmugasundaram K. Antioxidant status and lipid peroxidation in type 2 diabetes mellitus with and without complications. Clin Sci 1996; 90: Akkus I, Kalak S, Vural H, Caglayan O, Menekse E, Can G, et al. Leukocyte lipid peroxidation, Superoxide dismutase, glutathione peroxidase and serum and leukocyte vitamin C levels of patients with type 2 diabetes mellitus. Clin Chim Acta 1996; 244: Vanizor B, Orem A, Karahan SC. Decreased nitric oxide end products and its relationship with high density lipoprotein and oxidative stress in people with type 2 diabetes without complications. Diab Res Clin Pract 2001; 54: Milstien S, Katusic Z. Oxidation of tetrahydrobiopterin by peroxynitrite: Implication for vascular endothelial function. Biochem Biophys Res Commun 1999; 263: Matsumoto K, Sera Y, Ueki Y, Inukait G, Niirot E, Miyake S. Comparison of serum concentration of soluble adhesion molecules in diabetic microangiopathy and macroangiopathy. Diab Med 2002; 19: Fasching P, Waldhausl W, Wagner OF. Elevated circulating adhesion molecules in NIDDM potential mediators in diabetic macroangiopathy. Diabetologia 1996; 39: Redberg RF, Vogel RA, Criqui MH, Herringron DM, Lima JAC, Roman MJ. What is the spectrum of current and emerging technique for non-invasive measurement of atherosclerosis? JACC 2003; 41: Jager A, van Hinsbergh VWM, Kostense PJ, Emeis JJ, Nijpels G, Dekker JM, et al. Increased level of soluble vascular cell adhesion molecules are associated with risk of cardiovascular mortality in type 2 diabetes. The Hoorn study. Diabetes 2000; 49: Schmidt AM, Hori O, Chen JX, Crandall J, Zhang J, Cao R, et al. Advanced glycation end products interacting with their endothelial receptor induce expression of vascular cell adhesion molecule 1 (VCAM-1) in cultured human endothelial cells in mice. J Clin Invest 1995; 96: Mustafa G, Fatma ZK, Temocin S, Hanninen O. Cellular and clinical implications of glutathione. Indian J Exp Biol 2000; 38: Tsao PS, Buitrago R, Chan JR, Cooke JP. Fluid flow inhibits endothelial adhesiveness. NO and transcriptional regulation of VCAM-1. Circulation 1996; 94: Khan BV, Parthasarthy SS, Alexander RW, Medford RM. Modified low density lipoprotein and its constituents augment cytokine-activated vascular cell adhesion molecule-1 gene expression in human vascular endothelial cells. J Clin Invest 1995; 95:

8