- Email: [email protected]
Effect of estrogen on endothelial dysfunction in postmenopausal women with diabetes
Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81– S92 www.elsevier.com/locate/diabres
Effect of estrogen on endothelial dysfunction in postmenopausal women with diabetes Sang Jun Lee, Dong Wook Lee, Kee Sik Kim, In Kyu Lee * Department of Internal Medicine, Dongsan Medical Center, Keimyung Uni6ersity, Taegu, Republic of Korea
Abstract Background: Treatment with estrogen causes increased release of nitric oxide and this appears to be an important mechanism involved in the cardioprotective effect of estrogen. Previous studies have clearly shown that estrogen has an acute vasodilatory effect. Recently, abnormality of flow mediated endothelial dilation, which is defined as an endothelial dysfunction, has been proposed as an early manifestation of atherogenesis. However, its effect in type 2 diabetes is unknown. Stimulus-provoked endothelium dependent dilatation, using high-resolution external vascular ultrasound, is a useful method in the measurement of endothelial function. The main advantages of this method are that it is completely non-invasive, accurate and reproducible. This classical method for the measurement of endothelial dysfunction, utilizing non-invasive, high-resolution ultrasonography, has some limitations especially in terms of reproducibility. To improve the accuracy and reproducibility of this method, measurements made for this study involved the use of an additional attachment. This device was an ultrasound probe that was attached to the arm of patient. In addition, this study evaluated not only a change in diameter in terms of flow mediated vasodilation, but also initial and peak response times during this test. The aims of this study are: (1) to evaluate the flow mediated peak vasoreactivity (FMD), the lag time from base line to initiation of vasodilation (IRT) and the peak response of endothelial dependent vasodilation in a control group and in young diabetic patients; and (2) to evaluate the effect of estrogen treatment on flow mediated vasodilation in postmenopausal women with diabetes. Method: Measurements were taken for flow mediated vasodilation (endothelial dependent vasodilation: FMD), endothelial independent vasodilation (EID) and lag time from base line to initial reaction of FMD. The aforementioned non-invasive method and attachment were used. Subjects for the experiment were 12 young people with diabetes (six male and six female, mean age 26.3) and 12 age-matched healthy controls (six male and six female, mean age 25.6). In addition, measurements were taken for FMD, EID and lag time in the brachial artery before and after estrogen supplementation (Premarin 0.625mg for 7 days). Subjects for the experiment were 16 normal postmenopausal women and 18 age-matched postmenopausal women with type 2 diabetes. Result: There is no significant difference in BMI (body mass index), mean age, or blood pressure between the control group and the young diabetic group. There were no differences in age, total and LDL cholesterol levels or body mass index between the groups of postmenopausal women (PB 0.05). However, the HDL cholesterol level was found to be significantly lower in postmenopausal women with diabetes than in normal postmenopausal women (respectively, a mean 9 SD; 38.95 911.96 mg/dl vs. 52.20 9 9.55 mg/dl, PB0.05). The FMD of young diabetic patients is slightly lower than that of age-matched young healthy controls (healthy control: 8.9 92.7%, young diabetic patients: 6.3 9 2.1%, P B0.05). There was no difference in * Corresponding author. Tel.: + 82-53-250-7421. E-mail address: [email protected] (I.K. Lee). 0168-8227/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 8 - 8 2 2 7 ( 0 1 ) 0 0 3 3 9 - 4
S82
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
endothelial independent vasodilation between the control group and the young diabetic group. The IRT of FMD was significantly shorter in the healthy control group than in the young diabetic group (healthy control: 20.4 9 2.8 s, diabetic group: 29.5 95.4 s, PB0.0001). In contrast, the IRT of EID was no different between the two groups. The IRT of FMD showed a significant negative correlation with FMD (r= − 0.61, P B 0.01) and a significant negative correlation with high-density lipoprotein (HDL) cholesterol (r= − 0.68, PB 0.0001). The basal endothelium dependent vascular reactivity was significantly decreased in postmenopausal women with diabetes compared with normal postmenopausal women (8.0 93.9 vs. 13.7 96.2%, P B 0.05). Estrogen supplementation increased endothelium dependent vasodilation not only in postmenopausal women with diabetes (from 8.0 9 3.9 to 15.1 9 4.0%, P B 0.05) but also in normal postmenopausal women (from 13.7 9 6.2 to 20.1 9 4.7%, P B 0.05). In contrast, the responses to sublingual nitroglycerin were comparable between postmenopausal women with diabetes (from 21.1 9 6.0 to 22.1 9 4.1%, P\0.05) and normal postmenopausal women (from 25.8 9 7.8 to 25.2 9 4.5%, P \ 0.05), both before and after estrogen supplementation. Conclusion: These findings indicate that in addition to the percentage change in FMD, lag time from base line to initial response time (IRT) of FMD is a good indicator for evaluation of endothelial dysfunction. Also, this new parameter may be a more sensitive parameter for differentiation between the diseased state and the normal state of the endothelium than percentage changes in FMD. In terms of postmenopausal women, endothelial dysfunction was prominent in women with diabetes and was significantly improved by estrogen but not reversed. These results suggest that other factors in addition to estrogen deficiency play a role in endothelial dysfunction in postmenopausal women with diabetes mellitus. © 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Initial response time (IRT); Endothelial dependent vasodilation; Endothelial dysfunction
1. Introduction The endothelium, the inner cell layer that lines the blood vessels, has been regarded as a semi-permeable membrane between blood and vessel wall, facilitating the exchange of macromolecules. In 1980, Furchgott [1] and others first reported that the endothelium produces important vasodilatory substances called endothelium-derived relaxing factors (EDRFs). This EDRF is now recognized as being nitric oxide (NO) [2] or a NO-containing donor that mediates vasodilation by activating vascular smooth muscle guanylate cyclase. The endothelium also releases numerous cytokines and other substances, such as prostacyclin [3] (PGI2), endothelium-derived hyperpolarizing factors [4], vasoconstrictor prostaglandins [4], reactive oxygen species (ROS) [5] and endothelin [6], which produce vasodilation or vasoconstriction. A number of studies have shown that the endothelium has vascular homeostatic functions, such as inhibition of platelet aggregation [7], leukocyte adhesion [8] and vascular smooth muscle growth [9]. Recently, endothelial dysfunction has been proposed as an early manifestation of atherogenesis. To elucidate the relationship between endothelial dysfunction and the development of atherosclero-
sis, this and other studies are concerned with whether the patients with risk factors for coronary heart disease have impaired flow-mediated, endothelium-dependent vasodilation (FMD). These risk factors include diabetes [10,11], impaired glucose tolerance [12], hypertension [13], and hyperlipidemia [14]. Many studies have demonstrated that impaired endothelial function, early manifestation of atherosclerosis, is often found in the early life of patients with hypertension, non-insulin dependent diabetes mellitus (NIDDM) and insulin dependent diabetes mellitus (IDDM) [10–13]. To evaluate the endothelial dysfunction, several invasive techniques [15] have been used, such as coronary artery catheterization and blood flow measurement with infusion of vasoactive drugs. However, these invasive methods have limitations, they involve complicated techniques and there is difficulty in the follow-up assessment of endothelial function after treatment [16]. In place of these invasive methods, a non-invasive technique using external vascular ultrasound has been developed to detect endothelial dysfunction [17,18]. The advantages of this method are that it is completely non-invasive, accurate and the results are reproducible.
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
More recently, some groups have reported poor reproducibility for this technique, however this is mainly due to peak time variability and change of measurement site (Fig. 1). To overcome these problems, this study made use of a special device, which can keep the probe fixed at the same position during the whole testing period. Furthermore, this study involved continuously recording the change in artery diameter in response to reactive hyperemia after occlusion and release of the brachial artery. It is hypothesized that the lag time measurement, from base line to initiation of vasodilation after cuff deflation, is an important parameter to detect endothelial dysfunction. The aims of this study are firstly, to evaluate the flow mediated peak vasoreactivity (FMD), its lag time from base line to initiation of vasodilation (IRT) and the peak response of endothelial dependent vasodilation in young diabetic patients in comparison to a control group. Secondly, to evaluate the effect of estrogen treatment on flow mediated vasodilation and free radical generation of neutrophils in postmenopausal women with diabetes.
S83
2. Method
2.1. Subjects Twelve healthy volunteers, six male and six female, aged between 21 and 31 (mean 25.6) years and 12 patients with IDDM, six male and six female, aged between 23 and 30 (mean 26.3) years, participated in this study. In addition, 16 normal postmenopausal women (mean age 53.4 years) and 18 age-matched postmenopausal women (mean age 57.5 years) with NIDDM were enrolled for the study of estrogen treatment. The definitions of IDDM and NIDDM were based on the National Diabetes Data Group (NDDG) criteria [19]. The mean duration of IDDM was 12 months (range of diabetes: 1–12 months). The IDDM patients have received insulin (at least twice daily injection) for glycemic control. All subjects were questioned on their full medical history and underwent physical examination and laboratory testing. Subjects were excluded from this study on the basis of a history of ischemic heart disease or peripheral vascular disease, hypertension (\140/90 mmHg), impairment
Fig. 1. The limitation of classical method in diagnosis of endothelial dysfunction.
S84
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
Table 1 Clinical characteristics of healthy control and young diabetic patients
Number (% of male) Age (year) BMI (kg/m2) DM duration (month) FBS (mg/dl) CH (mg/dl) TG (mg/dl) H DL(mg/dl)
Control
DM
12 (50) 25.69 1.8 20.59 1.4 0 80.09 6.0 143.6915.1 48.29 18.6 54.69 90
12 (50) 26.3 9 3.5 20.79 5.2 13.79 11.1** 176.6983.1 166.5 956.7 146.9 9124.7* 43.3 912.4**
CH, total cholesterol; TG, triglyceride; HDL, HDL-cholesterol. ** PB0.005 vs control * PB0.05 vs control.
of renal function (serum Cr\ 1.5 mg/dL urine protein \0.3 g/24 h urine), or a history of smoking, taking of antioxidant supplements (Vitamin E, Vitamin C, estrogen) or cardiovascular drugs. Diabetic patients with chronic complications, such as diabetic retinopathy, diabetic neuropathy or diabetic vasculopathy, were also excluded from this study. The clinical characteristics of the patients enrolled for this study are shown in Tables 1 and 2. In these subjects measurements were taken for flow-mediated vasodilation (endothelial dependent vasodilation: FMD), endothelial indeTable 2 Clinical characteristics of estrogen study subjects Controls (n= 16) Age 53.496.3 Postmenopausal 7.1 9 7.4 duration (years) BMI (kg/m2) 23.293.4 Total cholesterol 4.89 0.7 (mmol/L) LDL cholesterol 2.79 0.2 (mmol/L) HDL cholesterol 1.4 90.2 (mmol/L) Triglyceride (mmol/L) 1.2 9 0.4 FBS (mmol/L) 4.69 0.7 HbAlc (%) * PB0.05.
Diabetes (n = 18) 57.59 7.4 8.49 4.5 22.093.6 4.879 1.1 2.759 0.8 1.0 9 0.3* 2.3 9 1.6 12.2 94.5* 11.19 2.7
pendent vasodilation (EID) and lag time from base line to initial reaction of FMD, using a non-invasive method and a special device. In addition measurements were taken and examined for FMD, EID and lag time from base line to initiation of FMD in the brachial artery, before and after estrogen supplementation (Premarin 0.625 mg for 7 days) in 16 normal postmenopausal women and 18 age-matched postmenopausal women with NIDDM.
2.2. Experimental protocol All subjects were studied in the morning after overnight fasting and before insulin injection. To improve the accuracy and reproducibility of the test, the authors’ designed a special device (Fig. 2), which can attach and fix the probe at the same site in the brachial artery during the whole testing period. In comparison with the classical non-invasive methods, using this device, it is possible to hold the transducer more consistently at the same site without movement. Arterial endothelium-dependent and endothelium-independent vasodilation were studied noninvasively in a quiet and temperature-controlled (room temperature; 21–23 °C) dark room. Brachial artery responses to endothelium-dependent and independent-stimuli were examined according to a slightly modified method designed by Celermajer et al. [17]. A SONOS 5500 (HewlettPackard, USA) echocardiograph and 7.5 MHz linear array transducer (Hewlett-Packard, USA) was used, with the addition of the positioning device. The scanning point on the brachial artery was 2 –6 cm above the antecubital fossa in a longitudinal array of transducers. All parameters for scanning, such as depth and gain setting, were optimized at the beginning of the study and were kept constant throughout the examination period. When a satisfactory scanning image was found, the transducer position was marked on the upper arm to allow easy positioning of a later transducer at a same position. During the examination, blood pressure was monitored in the right arm every 2 min by an automatic pressure recorder. A brief description of the study protocols follows. After resting for 30 min in a supine position,
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
S85
Fig. 2. Special device designed for hold transducer and arm in the same position without movement.
the base line luminal diameter of the brachial arteries was measured. Following this a BP cuff was placed around the proximal forearm (proximal upper arm in the estrogen group) and inflated with a pressure of 300 mmHg. Five minutes later, the BP cuff was deflated rapidly and the change in the luminal diameter of the brachial arteries in response to increased blood flow (shear stress) was measured. Following this the patients lay for 15 min and the change in the luminal diameter of the brachial arteries in response to nitroglycerin 0.6 mg (exogenous NO-donor) was measured. A continuous record of the B-mode scanning image of the brachial artery was recorded on super-VHS videocassette tape during the examination. After the examination, the tape was analyzed at a 3 s intervals with an ultrasonic caliper for a 120 s period starting 30 s prior to the deflation of the BP cuff in the off line state. Measurements of the lumen diameter were taken from the anterior to the posterior interface between media and adventitia at the end diastole, incident with the R wave on a continuously monitored electrocardiogram. Using this analysis, evaluation of endothelium-dependent vasodilation was more accurate
and reliable than using the classical method. In addition, it was possible to measure the lag time from base line to initial reaction of FMD and the lag time from base line to peak reaction of FMD.
2.3. Blood sampling and analysis Ten millilitre of fresh venous blood was taken after measurement of FMD at each study and the blood specimens were anticoagulated with 0.1% EDTA. Plasma was recovered from fresh venous blood specimens after low-speed centrifugation at 1000 rpm for 7 min and immediately frozen at − 70 °C for chemical analysis. The plasma glucose concentration was determined with an autoanalyzer using a glucose oxidase method (747 autoanalyzer, Hitachi, Japan). The plasma insulin and c-peptide concentration were measured by RIA method (RIA kit, TechnoGenetics, Italy). The plasma total cholesterol, triglyceride, highdensity lipoprotein (HDL) cholesterol and lowdensity lipoprotein (LDL) cholesterol concentrations were measured by enzymatic method (COBAS INTEGRA 700®, Roche, Swiss).
S86
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
2.4. Statistical analysis The data are expressed as the mean9 SD. Statistical differences of mean values between young diabetic and age-matched non-diabetic subjects were determined using the Student’s t-test.
3. Results
3.1. Clinical characteristics The base line characteristics of the young groups are shown in Table 1 (young IDDM and age-matched healthy controls) and estrogen groups in Table 2 (normal postmenopausal women and age-matched postmenopausal women with type 2 diabetes mellitus). Age and gender were matched between groups. No significant difference in mean arterial pressure, heart rate and body mass index (BMI) was found between the young groups. The estrogen groups also showed no significant difference in mean arterial pressure, heart rate and BMI. Fasting blood glucose was higher in the young IDDM subjects than in the healthy control subjects (P B0.005) (Table 1). In the estrogen groups, fasting blood glucose was also higher in postmenopausal women with NIDDM (PB0.05) (Table 2). There was no difference in the plasma concentration of total cholesterol between the young groups. The plasma level of triglyceride, low-density lipoprotein (LDL) cholesterol was higher in the young IDDM subjects than in the healthy control subjects (PB0.05). Also young IDDM subjects exhibited lower plasma concentrations of high-density lipoprotein (HDL) cholesterol than that of healthy control subjects (P B0.05). In the same way, postmenopausal women with NIDDM have lower plasma concentrations of high-density lipoprotein (HDL) cholesterol than normal postmenopausal women (P B0.05).
Table 3 Outcome of FMD, EID and IRT in young group
FMD L. d (cm) VA (%) IRT(s) EID VA (%) IRT(s)
Normal (n = 12)
DM (n =12)
0.376 9 0.068 8.9 9 2.7 20.4 9 2.8
0.366 90.057 6.3 92.1* 29.5 95.4**
19.7 9 3.8 83.4 9 12.9
18.5 95.5 86.7 915.7
FMD, endothelial dependent vasodilation; EID, endothelial independent vasodilation; L.d, lumen diameter; VA, vasoactivity; IRT, initial reaction time. ** PB0.0005 vs control. * PB0.05 vs control.
young groups. The endothelial dependent vasodilation was significantly decreased in young IDDM subjects compared with healthy control subjects (PB0.05) (Table 3) (Fig. 3). There was no significant difference in the endothelium-independent vasodilation seen between the young groups (Table 3). The basal endothelium-dependent vasodilation was significantly decreased in postmenopausal women with NIDDM compared with normal postmenopausal women (PB 0.05). Estrogen supplementation increased endothelium dependent vasodilation not only in postmenopausal women with NIDDM (PB 0.05) but also in normal postmenopausal women (PB 0.05) (Fig. 4). No significant difference in the endothelium-independent vasodilation was seen between the two estrogen groups. Fig. 5 shows the percent change
3.2. Endothelium-dependent, flow-mediated 6asodilation There are no significant differences in the luminal diameter of the brachial arteries between the
Fig. 3. Compare FMD between healthy controls and IDDM patients (mean 9SD).
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
S87
oxide donor, nitroglycerin, exhibited no significant difference between the two young groups.
3.3. Correlation between FMD and IRT
Fig. 4. The percent increase in arterial diameter during reactive hyperemia between postmenopausal controls and NIDDM (*PB0.05).
(%) in the luminal diameter of the brachial arteries during hyperemic phase, recorded continuously. This figure shows the significant difference in lag time from base line to initial reaction of FMD. The lag time from base line to initial reaction of FMD was increased in diabetic subjects (29.59 5.4 s) compared with control subjects (20.4 9 2.8 s) (PB 0.0001) (Fig. 4). The lag time to peak reaction of FMD, the duration to maximal change in the brachial arteries’ luminal diameter in response to increased blood flow (shear stress), exhibited no significant difference between the two groups. The lag time to initial reaction of EID, the duration to initial change in the brachial arteries’ luminal diameter in response to exogenous nitric
The endothelium-dependent vasodilation showed a significant positive correlation with plasma levels of high-density lipoprotein (HDL) cholesterol (r= 0.58, PB0.005) (Fig. 6). In contrast, there was no significant correlation with plasma levels of triglyceride (r= −0.31, P\ 0.05). The IRT showed a significant negative correlation with FMD (r= − 0.74, PB0.001) (Fig. 7A) and a significant negative correlation with plasma levels of high-density lipoprotein (HDL) cholesterol (r= − 0.68, PB 0.0001) (Fig. 7B). There was also a significant positive correlation between IRT and plasma levels of triglyceride (r= 0.44, PB 0.05) (Fig. 7C).
4. Discussion The results of this study showed that endothelium-dependent vasodilation is impaired in young diabetic patients without chronic complications. This means that endothelial dysfunction can occur in young children with risk factors of coronary heart disease, such as hypertension, diabetes, hyperlipidemia and smoking. Celermajer et al. [17] also reported that these physiological changes of vascular endothelium occurred in hypercholes-
Fig. 5. Continuous analysis of endothelial dependent vasodialtion (FMD). *, P B 0.05 vs control; **, P B0.0001 vs control; ‡, IRT, lag time of FMD in each group (control: healthy young volunteers, diabetes: IDDM).
S88
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
Fig. 6. The correlation between FMD and HDL-cholesterol.
terolemic children. Thus Celermajer et al. [17] suggest that physiologically important changes of the vascular endothelium may start as early as the first decade of life in children with risks factors. Physiological change of vascular endothelium causes decreased bioactivity of EDRFs [20]. Endothelial dysfunction may result in the abnormal adherence of neutrophils, monocytes and platelets to the endothelium, and thus may contribute to the progression of atherogenesis [21]. Early detection of endothelial dysfunction, before anatomical evidence of atherosclerotic lesions, such as wall thickening and plaque, may contribute to a decrease in the development of complications related to atherosclerosis. The results of the present study have shown that endothelial dysfunction was prominent in postmenopausal women with diabetes and was significantly improved by estrogen supplement but not corrected. Furthermore, these results suggest that other factors as well as estrogen deficiency may play a role in endothelial dysfunction in postmenopausal women with diabetes. Available studies have clearly shown that estrogen has some
effect on vasodilation. Several hypotheses exist as to the mechanism of vasodilation after estrogen treatment. These hypotheses include release of nitric oxide, prostaglandin release and activation of potassium or calcium channels [22–24]. Atherosclerosis, the major cause of ischemic heart disease, is known to be the major cause of mortality and mobidity in the patients with risk factors for atherosclerosis and in people who eat a westernized diet. However, the reasons behind this increased risk are still unclear. Studies conducted in this decade suggest that early detection of atherosclerosis is important in reducing the risk of complications including ischemic heart disease [25]. Recently, endothelial dysfunction has been regarded as the early process of atherogenesis. Therefore, some non-invasive techniques for assessing the vascular state are becoming more available and more popular than invasive techniques. Non-invasive techniques are simpler and give more accurate results and are without the problems of invasive techniques in terms of follow-up assessment. Non-invasive techniques in-
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92 Fig. 7. (A) The correlation between IRT and FMD; (B) the correlation between IRT and HDL-cholesterol; and (C) the correlation between IRT and triglyceride.
S89
S90
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
clude carotid artery intima-media thickness (IMT) [26], aortic pulse wave velocity (PWV), arterial stiffness [27] and flow-mediated [17], endothelial dependent vasodilation of the brachial artery during reactive hyperemia. The requirements of screening methods for atherosclerosis are that they must be simple, easy and have high levels of sensitivity and specificity. The classical non-invasive methods require significant technical skill. Measurement of carotid artery intima-media thickness by high-resolution ultrasound B-mode has shown histopathological change in the arterial wall. Although IMT reflects the early stages of atherosclerotic disease [28,29], an increase in IMT may result from actors other than atherosclerosis, and precaution is required in the interpretation of IMT of the carotid artery [30]. Many studies have shown a good association between IMT and risk factors for coronary heart disease [28,29]. Despite these reports, the correlation of IMT with the extent and severity of coronary disease as assessed by coronary artery angiography is not always high [31]. So, B-mode measurement of IMT of common carotid artery gives us only anatomical information about the vascular wall. Non-invasive aortic pulse wave velocity measurements in diabetic patients have shown a positive relationship with atherosclerosis. The PWV is increased in stiffer arteries. So, an increase in non-invasive PWV measurements is regarded as a marker for the risk of coronary artery disease [32]. However, PWV measurement may offer less significant clinical benefits because the results of PWV are different for different vessels and there is limited data on the reproducibility of the test [33]. Celermajer et al. [17] originally designed flowmediated vasodilation of the superficial artery using high-resolution ultrasound B-mode measurement. Flow-mediated vasodilation may be induced by increased blood flow (reactive hyperemia) after vessel occlusion and release. Shear stress induced by increased blood flow is the major stimuli for the release of nitric oxide (endothelial dependent relaxing factors) from the endothelium. Nitric oxide plays a major role in the control of vascular homeostasis. Decreased FMD is regarded as endothelial dysfunction. Measurement of brachial artery FMD
using high-resolution ultrasound is widely available in clinical settings due to its high reproducibility, repeatability and easy access to the superficial artery. For the last decade many studies have reported that decreased FMD in the brachial artery is associated with risk factors for coronary artery disease and has good correlation with the FMD of the coronary artery [16]. Therefore, early detection of decreased FMD value in patients with risk factors for coronary artery disease with no symptoms is clinically important because endothelial dysfunction in these groups may contribute to the progression of atherosclerosis. However, recently inconsistencies in the results of non-invasive measurements of FMD in the brachial artery have been reported. There are limitations [30] to the use of non-invasive measurement of FMD for screening or early detection of endothelial dysfunction. The gold standard for assessment of vascular function is angiographic technique, which requires arterial cannulation with infusion of pharmacokinetic drugs. The limitation is that serial follow-up using angiographic technique is difficult in the clinical setting due to its invasiveness. In addition, the technique has a high difficulty level as the examiner must hold the probe without movement during the measurement of FMD in peripheral artery. One report has suggested that the correlation between brachial artery FMD and coronary artery endothelial function is not strong. A further limitation is that FMD results occur over a wide range due to the variability of endothelial reactions to EDRFs. According to the results of this study, although the difference in endothelium-dependent vasodilation was significant between the two groups, the range of FMD results was wide. Therefore, the use of FMD as screening tools for the early detection of endothelial dysfunction has some difficulties. Using non-invasive methods of flow-mediated, endothelium-dependent vasodilation, for evaluation of endothelial dysfunction in adult patients with risk factors, such as hypertension, diabetes and obesity, there was a clearly defined difference in FMD results compared to those of the healthy controls. Using the same method for early evalua-
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
tion of endothelial dysfunction in healthy children and young adults with or without risk factors for atherosclerosis has no clinical benefit. These patients have mild endothelial dysfunction, such as physiological change however, some other factor such as sex, arm length, occlusion pressure of the BP cuff or autonomic neuropathy may also affect endothelium-dependent vasodilation. The widely ranging FMD results in young people with no symptoms are also regarded as a problem in the evaluation of impaired endothelial function [30]. In comparison, the results for lag time to initial reaction of flow-mediated, endothelial dependent vasodilation (IRT) show a narrower range (relative dispersion) than those for endothelium-dependent vasodilation. Therefore, additional measurements of lag time to initial reaction of FMD for early detection of endothelial dysfunction using a special device increase the reliability of the results of the non-invasive technique. Then, the observer overcomes the limitation of classical methods, such as peak time variability and change of measurement site. This novel approach in detection of endothelial dysfunction is simple and highly reproducible. In conclusion, these findings firstly indicate that in addition to the percentage change in FMD, lag time from base line to initial response time (IRT) of FMD is a good indicator to evaluate endothelial dysfunction. Also, that this new parameter may be a more sensitive parameter for differentiation between the diseased state and the normal state of the endothelium than the percentage changes in FMD. Secondly, this study found that endothelial dysfunction was prominent in postmenopausal women with diabetes and was significantly improved by estrogen but not sufficiently. These results suggest that other factors in addition to estrogen deficiency play a role in endothelial dysfunction in postmenopausal women with diabetes mellitus.
Acknowledgements This study was supported by a grant from the Institute for Medical genetics, Keimyung University School of Medicine (97– 1) and by a grant
S91
from the Korea Health 21 R&D project, Ministry of Health & Welfare, Republic of Korea (HMP99-M-08-0004).
References [1] R. Furchgott, D. Zawadzki, The obligatory role of endothelial cells in the relaxation of arterial smooth muscle cell by acetylcholine, Nature 288 (1980) 373 – 376. [2] R.M.J. Palmer, D.S. Ashton, S. Moncada, Vascular endothelial cells synthesize nitric oxide from L-arginine, Nature 333 (1988) 664 – 666. [3] S. Moncada, Biological importance of prostacyclin, Br. J. Pharmacol. 76 (1982) 3 – 31. [4] P.M. Vanhoutte, Endothelial dysfunction and atherosclerosis, Eur. Heart J. 18 Suppl. E (1997) E19 – E29. [5] B. Tesfamariam, Free radicals in diabetic endothelial cell dysfunction, Free Radic. Biol. Med. 16 (1994) 383 – 391. [6] M. Yanagisawa, H. Kurihara, S. Kimura, A novel potent vasoconstrictor peptide produced by vascular endothelial cells, Nature 332 (1988) 411 –415. [7] M.W. Radomski, R.M.J. Palmer, S. Moncada, An Larginine/nitric oxide pathway present in human platelets regulates aggregation, Proc. Natl. Acad. Sci. 87 (1990) 5193 – 5197. [8] P. Kubes, M. Suzuki, N. Granger, Nitric oxide: an endogenous modulator of leukocyte adhesion, Proc. Natl. Acad. Sci. 88 (1991) 4651 – 4655. [9] U.C. Grag, A. Hassid, Nitric oxide-generating vasodilators and 8-bromo-cyclic guanosine monophosphate inhibit mitogenesis and proliferation of cultures rat vascular smooth muscle cells, J. Clin. Invest. 83 (1989) 1774 – 1777. [10] S.B. Williams, J.A. Cusco, M.A. Roddy, M.T. Johnstone, M.A. Creager, Impaired nitric oxide-mediated vasodilation in patients with non-insulin-dependent diabetes mellitus, J. Am. Coll. Cardiol. 27 (1996) 567 – 574. [11] M.T. Johnstone, S.J. Creager, K.M. Scales, J.A. Cusco, B.K. Lee, M.A. Creager, Impaired endothelium-dependent vasodilation in patients with insulin dependent diabetes mellitus, Circulation 88 (1993) 2510 – 2516. [12] H.O. Steinberg, H. Chaker, R. Leaming, A. Johnson, G. Brechtel, A. Baron, Obesity/insulin resistance is associated with endothelial dysfunction, J. Clin. Invest. 97 (1996) 2601 – 2610. [13] M. Barenbrock, M. Hausberg, M. Kosch, S.A. Golubev, K. Kisters, K. Rahn, Flow-mediated vasodilation and distensibility in relation to intima-media thickness of large arteries in mild essential hypertension, Am. J. Hypertens. 12 (1999) 973 – 979. [14] K.E. Sorensen, D.S. Celermajer, D. Georgakopoulos, G. Hatcher, Impairment of endothelium-dependent dilation is an early event in children with familial hypercholesterolemia and is related to the lipoprotein (a) level, J. Clin. Invest. 93 (1994) 50 – 55.
S92
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
[15] P.L. Ludmer, A.P. Selwyn, T.L. Shook, et al., Paradoxical vasoconstriction induced by acetylcholine in atherosclerotic coronary arteries, N. Engl. J. Med. 315 (1986) 1046 – 1051. [16] E.L. Nabel, A.P. Selwyn, P. Ganz, Large coronary arteries in humans are responsive to changing blood flow: an endothelium dependent mechanism that fails in patients with atherosclerosis, J. Am. Coll. Cardiol. 16 (1990) 349– 356. [17] D.S. Celermajer, K.E. Sorensen, V.M. Gooch, et al., Non-invasive detection of endothelial dysfunction in children and adults at risk of atherosclerosis, Lancet 340 (1992) 1111 – 1115. [18] R.D. Fish, E.G. Nabel, A.P. Selwyn, et al., Responses of coronary arteries of cardiac transplant patients to acetylcholine, J. Clin. Invest. 81 (1988) 21 –31. [19] National Diabetes Data Group, Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance, Diabetes 28 (1979) 1039 –1057. [20] R.J. Gryglewski, R.M. Palmer, S. Moncada, Superoxide anion is involved in the breakdown of endotheliumderived vascular relaxing factor, Nature 320 (1986) 454 – 456. [21] R. Ross, The pathogenesis of atherosclerosis — an update, N. Engl. J. Med. 314 (1986) 488 – 500. [22] E.H. Lieberman, M.D. Gerhard, A. Uehata, et al., Estrogen improves endothelium-dependent, flow-mediated vasodilation in postmenopausal women, Annals Int. Med. 121 (1994) 936 – 941. [23] H. Tagawa, H. Shimokawa, T. Tagawa, M. KuroiwaMatsumoto, Y. Hirooka, A. Takeshita, Short-term estrogen augments both nitric oxide-mediated and non-nitric oxide-mediated endothelium-dependent vasodilation in postmenopausal women, J. Cardiov. Pharmacol. 30 (1997) 481 – 488. [24] D.E. Bush, C.E. Jones, K.M. Bass, G.K. Walters, J.M. Bruza, P. Ouyang, Estrogen replacement reverses en-
[25] [26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
dothelial dysfunction in postmenopausal women, Am. J. Med. 104 (1998) 552 – 558. H. Shimokawa, Primary endothelial dysfunction: atherosclerosis, J. Mol. Cell. Cardiol. 31 (1999) 23 – 37. P. Pignoli, E. Tremoli, A. Poli, P. Oreste, R. Paoletti, Intimal plus medial thickness of the arterial wall: a direct measurement with ultrasound imaging, Circulation 74 (1986) 1399 – 1406. D.J. Farrar, M.G. Bond, W.A. Riley, J.K. Sawyer, Anatomic correlates of aortic pulse wave velocity and carotid artery elasticity during atherosclerosis progression and regression in monkeys, Circulation 83 (1991) 1754 – 1763. J.L. Wofford, F.R. Kahl, G.R. Howard, W.M. McKinney, J.F. Toole, J.R. Crouse, Relation of extent of extracranial carotid artery atherosclerosis as measured by B-mode ultrasound to the extent of coronary atherosclerosis, Arterioscler. Thromb. 11 (1991) 1786 – 1794. L. Niskanen, R. Rauramaa, H. Miettinen, S.M. Haffner, M. Mercuri, M. Uusitupa, Carotid artery intima-media thickness in elderly patients with NIDDM and in non-diabetic subjects, Stroke 27 (1996) 1986 – 1992. E.D. Lehmann, W.A. Riley, P. Clarkson, R.G. Gosling, Non-invasive assessment of cardiovascular disease in diabetes mellitus, Lancet 350 (1) (1997) 14 – 19. M.R. Adams, A. Nakagomi, A. Keech, et al., Carotid intima-media thickness is only weakly correlated with the extent and severity of coronary artery, Circulation 92 (1995) 2127 – 2134. M. Karamanoglu, M.F. O’Rourke, A.P. Avolio, R.P. Kelly, An analysis of the relationship between central aortic and peripheral upper limb pressure waves in man, Eur. Heart J. 14 (1993) 160 – 167. A. Uehata, A.M.D. Gerhard, I.T. Meredith, et al., Close relationship of endothelial dysfunction in coronary and brachial artery (Abstract), Circulation 88 (1983) I-618.
Effect of estrogen on endothelial dysfunction in postmenopausal women with diabetes Sang Jun Lee, Dong Wook Lee, Kee Sik Kim, In Kyu Lee * Department of Internal Medicine, Dongsan Medical Center, Keimyung Uni6ersity, Taegu, Republic of Korea
Abstract Background: Treatment with estrogen causes increased release of nitric oxide and this appears to be an important mechanism involved in the cardioprotective effect of estrogen. Previous studies have clearly shown that estrogen has an acute vasodilatory effect. Recently, abnormality of flow mediated endothelial dilation, which is defined as an endothelial dysfunction, has been proposed as an early manifestation of atherogenesis. However, its effect in type 2 diabetes is unknown. Stimulus-provoked endothelium dependent dilatation, using high-resolution external vascular ultrasound, is a useful method in the measurement of endothelial function. The main advantages of this method are that it is completely non-invasive, accurate and reproducible. This classical method for the measurement of endothelial dysfunction, utilizing non-invasive, high-resolution ultrasonography, has some limitations especially in terms of reproducibility. To improve the accuracy and reproducibility of this method, measurements made for this study involved the use of an additional attachment. This device was an ultrasound probe that was attached to the arm of patient. In addition, this study evaluated not only a change in diameter in terms of flow mediated vasodilation, but also initial and peak response times during this test. The aims of this study are: (1) to evaluate the flow mediated peak vasoreactivity (FMD), the lag time from base line to initiation of vasodilation (IRT) and the peak response of endothelial dependent vasodilation in a control group and in young diabetic patients; and (2) to evaluate the effect of estrogen treatment on flow mediated vasodilation in postmenopausal women with diabetes. Method: Measurements were taken for flow mediated vasodilation (endothelial dependent vasodilation: FMD), endothelial independent vasodilation (EID) and lag time from base line to initial reaction of FMD. The aforementioned non-invasive method and attachment were used. Subjects for the experiment were 12 young people with diabetes (six male and six female, mean age 26.3) and 12 age-matched healthy controls (six male and six female, mean age 25.6). In addition, measurements were taken for FMD, EID and lag time in the brachial artery before and after estrogen supplementation (Premarin 0.625mg for 7 days). Subjects for the experiment were 16 normal postmenopausal women and 18 age-matched postmenopausal women with type 2 diabetes. Result: There is no significant difference in BMI (body mass index), mean age, or blood pressure between the control group and the young diabetic group. There were no differences in age, total and LDL cholesterol levels or body mass index between the groups of postmenopausal women (PB 0.05). However, the HDL cholesterol level was found to be significantly lower in postmenopausal women with diabetes than in normal postmenopausal women (respectively, a mean 9 SD; 38.95 911.96 mg/dl vs. 52.20 9 9.55 mg/dl, PB0.05). The FMD of young diabetic patients is slightly lower than that of age-matched young healthy controls (healthy control: 8.9 92.7%, young diabetic patients: 6.3 9 2.1%, P B0.05). There was no difference in * Corresponding author. Tel.: + 82-53-250-7421. E-mail address: [email protected] (I.K. Lee). 0168-8227/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 1 6 8 - 8 2 2 7 ( 0 1 ) 0 0 3 3 9 - 4
S82
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
endothelial independent vasodilation between the control group and the young diabetic group. The IRT of FMD was significantly shorter in the healthy control group than in the young diabetic group (healthy control: 20.4 9 2.8 s, diabetic group: 29.5 95.4 s, PB0.0001). In contrast, the IRT of EID was no different between the two groups. The IRT of FMD showed a significant negative correlation with FMD (r= − 0.61, P B 0.01) and a significant negative correlation with high-density lipoprotein (HDL) cholesterol (r= − 0.68, PB 0.0001). The basal endothelium dependent vascular reactivity was significantly decreased in postmenopausal women with diabetes compared with normal postmenopausal women (8.0 93.9 vs. 13.7 96.2%, P B 0.05). Estrogen supplementation increased endothelium dependent vasodilation not only in postmenopausal women with diabetes (from 8.0 9 3.9 to 15.1 9 4.0%, P B 0.05) but also in normal postmenopausal women (from 13.7 9 6.2 to 20.1 9 4.7%, P B 0.05). In contrast, the responses to sublingual nitroglycerin were comparable between postmenopausal women with diabetes (from 21.1 9 6.0 to 22.1 9 4.1%, P\0.05) and normal postmenopausal women (from 25.8 9 7.8 to 25.2 9 4.5%, P \ 0.05), both before and after estrogen supplementation. Conclusion: These findings indicate that in addition to the percentage change in FMD, lag time from base line to initial response time (IRT) of FMD is a good indicator for evaluation of endothelial dysfunction. Also, this new parameter may be a more sensitive parameter for differentiation between the diseased state and the normal state of the endothelium than percentage changes in FMD. In terms of postmenopausal women, endothelial dysfunction was prominent in women with diabetes and was significantly improved by estrogen but not reversed. These results suggest that other factors in addition to estrogen deficiency play a role in endothelial dysfunction in postmenopausal women with diabetes mellitus. © 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Initial response time (IRT); Endothelial dependent vasodilation; Endothelial dysfunction
1. Introduction The endothelium, the inner cell layer that lines the blood vessels, has been regarded as a semi-permeable membrane between blood and vessel wall, facilitating the exchange of macromolecules. In 1980, Furchgott [1] and others first reported that the endothelium produces important vasodilatory substances called endothelium-derived relaxing factors (EDRFs). This EDRF is now recognized as being nitric oxide (NO) [2] or a NO-containing donor that mediates vasodilation by activating vascular smooth muscle guanylate cyclase. The endothelium also releases numerous cytokines and other substances, such as prostacyclin [3] (PGI2), endothelium-derived hyperpolarizing factors [4], vasoconstrictor prostaglandins [4], reactive oxygen species (ROS) [5] and endothelin [6], which produce vasodilation or vasoconstriction. A number of studies have shown that the endothelium has vascular homeostatic functions, such as inhibition of platelet aggregation [7], leukocyte adhesion [8] and vascular smooth muscle growth [9]. Recently, endothelial dysfunction has been proposed as an early manifestation of atherogenesis. To elucidate the relationship between endothelial dysfunction and the development of atherosclero-
sis, this and other studies are concerned with whether the patients with risk factors for coronary heart disease have impaired flow-mediated, endothelium-dependent vasodilation (FMD). These risk factors include diabetes [10,11], impaired glucose tolerance [12], hypertension [13], and hyperlipidemia [14]. Many studies have demonstrated that impaired endothelial function, early manifestation of atherosclerosis, is often found in the early life of patients with hypertension, non-insulin dependent diabetes mellitus (NIDDM) and insulin dependent diabetes mellitus (IDDM) [10–13]. To evaluate the endothelial dysfunction, several invasive techniques [15] have been used, such as coronary artery catheterization and blood flow measurement with infusion of vasoactive drugs. However, these invasive methods have limitations, they involve complicated techniques and there is difficulty in the follow-up assessment of endothelial function after treatment [16]. In place of these invasive methods, a non-invasive technique using external vascular ultrasound has been developed to detect endothelial dysfunction [17,18]. The advantages of this method are that it is completely non-invasive, accurate and the results are reproducible.
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
More recently, some groups have reported poor reproducibility for this technique, however this is mainly due to peak time variability and change of measurement site (Fig. 1). To overcome these problems, this study made use of a special device, which can keep the probe fixed at the same position during the whole testing period. Furthermore, this study involved continuously recording the change in artery diameter in response to reactive hyperemia after occlusion and release of the brachial artery. It is hypothesized that the lag time measurement, from base line to initiation of vasodilation after cuff deflation, is an important parameter to detect endothelial dysfunction. The aims of this study are firstly, to evaluate the flow mediated peak vasoreactivity (FMD), its lag time from base line to initiation of vasodilation (IRT) and the peak response of endothelial dependent vasodilation in young diabetic patients in comparison to a control group. Secondly, to evaluate the effect of estrogen treatment on flow mediated vasodilation and free radical generation of neutrophils in postmenopausal women with diabetes.
S83
2. Method
2.1. Subjects Twelve healthy volunteers, six male and six female, aged between 21 and 31 (mean 25.6) years and 12 patients with IDDM, six male and six female, aged between 23 and 30 (mean 26.3) years, participated in this study. In addition, 16 normal postmenopausal women (mean age 53.4 years) and 18 age-matched postmenopausal women (mean age 57.5 years) with NIDDM were enrolled for the study of estrogen treatment. The definitions of IDDM and NIDDM were based on the National Diabetes Data Group (NDDG) criteria [19]. The mean duration of IDDM was 12 months (range of diabetes: 1–12 months). The IDDM patients have received insulin (at least twice daily injection) for glycemic control. All subjects were questioned on their full medical history and underwent physical examination and laboratory testing. Subjects were excluded from this study on the basis of a history of ischemic heart disease or peripheral vascular disease, hypertension (\140/90 mmHg), impairment
Fig. 1. The limitation of classical method in diagnosis of endothelial dysfunction.
S84
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
Table 1 Clinical characteristics of healthy control and young diabetic patients
Number (% of male) Age (year) BMI (kg/m2) DM duration (month) FBS (mg/dl) CH (mg/dl) TG (mg/dl) H DL(mg/dl)
Control
DM
12 (50) 25.69 1.8 20.59 1.4 0 80.09 6.0 143.6915.1 48.29 18.6 54.69 90
12 (50) 26.3 9 3.5 20.79 5.2 13.79 11.1** 176.6983.1 166.5 956.7 146.9 9124.7* 43.3 912.4**
CH, total cholesterol; TG, triglyceride; HDL, HDL-cholesterol. ** PB0.005 vs control * PB0.05 vs control.
of renal function (serum Cr\ 1.5 mg/dL urine protein \0.3 g/24 h urine), or a history of smoking, taking of antioxidant supplements (Vitamin E, Vitamin C, estrogen) or cardiovascular drugs. Diabetic patients with chronic complications, such as diabetic retinopathy, diabetic neuropathy or diabetic vasculopathy, were also excluded from this study. The clinical characteristics of the patients enrolled for this study are shown in Tables 1 and 2. In these subjects measurements were taken for flow-mediated vasodilation (endothelial dependent vasodilation: FMD), endothelial indeTable 2 Clinical characteristics of estrogen study subjects Controls (n= 16) Age 53.496.3 Postmenopausal 7.1 9 7.4 duration (years) BMI (kg/m2) 23.293.4 Total cholesterol 4.89 0.7 (mmol/L) LDL cholesterol 2.79 0.2 (mmol/L) HDL cholesterol 1.4 90.2 (mmol/L) Triglyceride (mmol/L) 1.2 9 0.4 FBS (mmol/L) 4.69 0.7 HbAlc (%) * PB0.05.
Diabetes (n = 18) 57.59 7.4 8.49 4.5 22.093.6 4.879 1.1 2.759 0.8 1.0 9 0.3* 2.3 9 1.6 12.2 94.5* 11.19 2.7
pendent vasodilation (EID) and lag time from base line to initial reaction of FMD, using a non-invasive method and a special device. In addition measurements were taken and examined for FMD, EID and lag time from base line to initiation of FMD in the brachial artery, before and after estrogen supplementation (Premarin 0.625 mg for 7 days) in 16 normal postmenopausal women and 18 age-matched postmenopausal women with NIDDM.
2.2. Experimental protocol All subjects were studied in the morning after overnight fasting and before insulin injection. To improve the accuracy and reproducibility of the test, the authors’ designed a special device (Fig. 2), which can attach and fix the probe at the same site in the brachial artery during the whole testing period. In comparison with the classical non-invasive methods, using this device, it is possible to hold the transducer more consistently at the same site without movement. Arterial endothelium-dependent and endothelium-independent vasodilation were studied noninvasively in a quiet and temperature-controlled (room temperature; 21–23 °C) dark room. Brachial artery responses to endothelium-dependent and independent-stimuli were examined according to a slightly modified method designed by Celermajer et al. [17]. A SONOS 5500 (HewlettPackard, USA) echocardiograph and 7.5 MHz linear array transducer (Hewlett-Packard, USA) was used, with the addition of the positioning device. The scanning point on the brachial artery was 2 –6 cm above the antecubital fossa in a longitudinal array of transducers. All parameters for scanning, such as depth and gain setting, were optimized at the beginning of the study and were kept constant throughout the examination period. When a satisfactory scanning image was found, the transducer position was marked on the upper arm to allow easy positioning of a later transducer at a same position. During the examination, blood pressure was monitored in the right arm every 2 min by an automatic pressure recorder. A brief description of the study protocols follows. After resting for 30 min in a supine position,
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
S85
Fig. 2. Special device designed for hold transducer and arm in the same position without movement.
the base line luminal diameter of the brachial arteries was measured. Following this a BP cuff was placed around the proximal forearm (proximal upper arm in the estrogen group) and inflated with a pressure of 300 mmHg. Five minutes later, the BP cuff was deflated rapidly and the change in the luminal diameter of the brachial arteries in response to increased blood flow (shear stress) was measured. Following this the patients lay for 15 min and the change in the luminal diameter of the brachial arteries in response to nitroglycerin 0.6 mg (exogenous NO-donor) was measured. A continuous record of the B-mode scanning image of the brachial artery was recorded on super-VHS videocassette tape during the examination. After the examination, the tape was analyzed at a 3 s intervals with an ultrasonic caliper for a 120 s period starting 30 s prior to the deflation of the BP cuff in the off line state. Measurements of the lumen diameter were taken from the anterior to the posterior interface between media and adventitia at the end diastole, incident with the R wave on a continuously monitored electrocardiogram. Using this analysis, evaluation of endothelium-dependent vasodilation was more accurate
and reliable than using the classical method. In addition, it was possible to measure the lag time from base line to initial reaction of FMD and the lag time from base line to peak reaction of FMD.
2.3. Blood sampling and analysis Ten millilitre of fresh venous blood was taken after measurement of FMD at each study and the blood specimens were anticoagulated with 0.1% EDTA. Plasma was recovered from fresh venous blood specimens after low-speed centrifugation at 1000 rpm for 7 min and immediately frozen at − 70 °C for chemical analysis. The plasma glucose concentration was determined with an autoanalyzer using a glucose oxidase method (747 autoanalyzer, Hitachi, Japan). The plasma insulin and c-peptide concentration were measured by RIA method (RIA kit, TechnoGenetics, Italy). The plasma total cholesterol, triglyceride, highdensity lipoprotein (HDL) cholesterol and lowdensity lipoprotein (LDL) cholesterol concentrations were measured by enzymatic method (COBAS INTEGRA 700®, Roche, Swiss).
S86
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
2.4. Statistical analysis The data are expressed as the mean9 SD. Statistical differences of mean values between young diabetic and age-matched non-diabetic subjects were determined using the Student’s t-test.
3. Results
3.1. Clinical characteristics The base line characteristics of the young groups are shown in Table 1 (young IDDM and age-matched healthy controls) and estrogen groups in Table 2 (normal postmenopausal women and age-matched postmenopausal women with type 2 diabetes mellitus). Age and gender were matched between groups. No significant difference in mean arterial pressure, heart rate and body mass index (BMI) was found between the young groups. The estrogen groups also showed no significant difference in mean arterial pressure, heart rate and BMI. Fasting blood glucose was higher in the young IDDM subjects than in the healthy control subjects (P B0.005) (Table 1). In the estrogen groups, fasting blood glucose was also higher in postmenopausal women with NIDDM (PB0.05) (Table 2). There was no difference in the plasma concentration of total cholesterol between the young groups. The plasma level of triglyceride, low-density lipoprotein (LDL) cholesterol was higher in the young IDDM subjects than in the healthy control subjects (PB0.05). Also young IDDM subjects exhibited lower plasma concentrations of high-density lipoprotein (HDL) cholesterol than that of healthy control subjects (P B0.05). In the same way, postmenopausal women with NIDDM have lower plasma concentrations of high-density lipoprotein (HDL) cholesterol than normal postmenopausal women (P B0.05).
Table 3 Outcome of FMD, EID and IRT in young group
FMD L. d (cm) VA (%) IRT(s) EID VA (%) IRT(s)
Normal (n = 12)
DM (n =12)
0.376 9 0.068 8.9 9 2.7 20.4 9 2.8
0.366 90.057 6.3 92.1* 29.5 95.4**
19.7 9 3.8 83.4 9 12.9
18.5 95.5 86.7 915.7
FMD, endothelial dependent vasodilation; EID, endothelial independent vasodilation; L.d, lumen diameter; VA, vasoactivity; IRT, initial reaction time. ** PB0.0005 vs control. * PB0.05 vs control.
young groups. The endothelial dependent vasodilation was significantly decreased in young IDDM subjects compared with healthy control subjects (PB0.05) (Table 3) (Fig. 3). There was no significant difference in the endothelium-independent vasodilation seen between the young groups (Table 3). The basal endothelium-dependent vasodilation was significantly decreased in postmenopausal women with NIDDM compared with normal postmenopausal women (PB 0.05). Estrogen supplementation increased endothelium dependent vasodilation not only in postmenopausal women with NIDDM (PB 0.05) but also in normal postmenopausal women (PB 0.05) (Fig. 4). No significant difference in the endothelium-independent vasodilation was seen between the two estrogen groups. Fig. 5 shows the percent change
3.2. Endothelium-dependent, flow-mediated 6asodilation There are no significant differences in the luminal diameter of the brachial arteries between the
Fig. 3. Compare FMD between healthy controls and IDDM patients (mean 9SD).
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
S87
oxide donor, nitroglycerin, exhibited no significant difference between the two young groups.
3.3. Correlation between FMD and IRT
Fig. 4. The percent increase in arterial diameter during reactive hyperemia between postmenopausal controls and NIDDM (*PB0.05).
(%) in the luminal diameter of the brachial arteries during hyperemic phase, recorded continuously. This figure shows the significant difference in lag time from base line to initial reaction of FMD. The lag time from base line to initial reaction of FMD was increased in diabetic subjects (29.59 5.4 s) compared with control subjects (20.4 9 2.8 s) (PB 0.0001) (Fig. 4). The lag time to peak reaction of FMD, the duration to maximal change in the brachial arteries’ luminal diameter in response to increased blood flow (shear stress), exhibited no significant difference between the two groups. The lag time to initial reaction of EID, the duration to initial change in the brachial arteries’ luminal diameter in response to exogenous nitric
The endothelium-dependent vasodilation showed a significant positive correlation with plasma levels of high-density lipoprotein (HDL) cholesterol (r= 0.58, PB0.005) (Fig. 6). In contrast, there was no significant correlation with plasma levels of triglyceride (r= −0.31, P\ 0.05). The IRT showed a significant negative correlation with FMD (r= − 0.74, PB0.001) (Fig. 7A) and a significant negative correlation with plasma levels of high-density lipoprotein (HDL) cholesterol (r= − 0.68, PB 0.0001) (Fig. 7B). There was also a significant positive correlation between IRT and plasma levels of triglyceride (r= 0.44, PB 0.05) (Fig. 7C).
4. Discussion The results of this study showed that endothelium-dependent vasodilation is impaired in young diabetic patients without chronic complications. This means that endothelial dysfunction can occur in young children with risk factors of coronary heart disease, such as hypertension, diabetes, hyperlipidemia and smoking. Celermajer et al. [17] also reported that these physiological changes of vascular endothelium occurred in hypercholes-
Fig. 5. Continuous analysis of endothelial dependent vasodialtion (FMD). *, P B 0.05 vs control; **, P B0.0001 vs control; ‡, IRT, lag time of FMD in each group (control: healthy young volunteers, diabetes: IDDM).
S88
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
Fig. 6. The correlation between FMD and HDL-cholesterol.
terolemic children. Thus Celermajer et al. [17] suggest that physiologically important changes of the vascular endothelium may start as early as the first decade of life in children with risks factors. Physiological change of vascular endothelium causes decreased bioactivity of EDRFs [20]. Endothelial dysfunction may result in the abnormal adherence of neutrophils, monocytes and platelets to the endothelium, and thus may contribute to the progression of atherogenesis [21]. Early detection of endothelial dysfunction, before anatomical evidence of atherosclerotic lesions, such as wall thickening and plaque, may contribute to a decrease in the development of complications related to atherosclerosis. The results of the present study have shown that endothelial dysfunction was prominent in postmenopausal women with diabetes and was significantly improved by estrogen supplement but not corrected. Furthermore, these results suggest that other factors as well as estrogen deficiency may play a role in endothelial dysfunction in postmenopausal women with diabetes. Available studies have clearly shown that estrogen has some
effect on vasodilation. Several hypotheses exist as to the mechanism of vasodilation after estrogen treatment. These hypotheses include release of nitric oxide, prostaglandin release and activation of potassium or calcium channels [22–24]. Atherosclerosis, the major cause of ischemic heart disease, is known to be the major cause of mortality and mobidity in the patients with risk factors for atherosclerosis and in people who eat a westernized diet. However, the reasons behind this increased risk are still unclear. Studies conducted in this decade suggest that early detection of atherosclerosis is important in reducing the risk of complications including ischemic heart disease [25]. Recently, endothelial dysfunction has been regarded as the early process of atherogenesis. Therefore, some non-invasive techniques for assessing the vascular state are becoming more available and more popular than invasive techniques. Non-invasive techniques are simpler and give more accurate results and are without the problems of invasive techniques in terms of follow-up assessment. Non-invasive techniques in-
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92 Fig. 7. (A) The correlation between IRT and FMD; (B) the correlation between IRT and HDL-cholesterol; and (C) the correlation between IRT and triglyceride.
S89
S90
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
clude carotid artery intima-media thickness (IMT) [26], aortic pulse wave velocity (PWV), arterial stiffness [27] and flow-mediated [17], endothelial dependent vasodilation of the brachial artery during reactive hyperemia. The requirements of screening methods for atherosclerosis are that they must be simple, easy and have high levels of sensitivity and specificity. The classical non-invasive methods require significant technical skill. Measurement of carotid artery intima-media thickness by high-resolution ultrasound B-mode has shown histopathological change in the arterial wall. Although IMT reflects the early stages of atherosclerotic disease [28,29], an increase in IMT may result from actors other than atherosclerosis, and precaution is required in the interpretation of IMT of the carotid artery [30]. Many studies have shown a good association between IMT and risk factors for coronary heart disease [28,29]. Despite these reports, the correlation of IMT with the extent and severity of coronary disease as assessed by coronary artery angiography is not always high [31]. So, B-mode measurement of IMT of common carotid artery gives us only anatomical information about the vascular wall. Non-invasive aortic pulse wave velocity measurements in diabetic patients have shown a positive relationship with atherosclerosis. The PWV is increased in stiffer arteries. So, an increase in non-invasive PWV measurements is regarded as a marker for the risk of coronary artery disease [32]. However, PWV measurement may offer less significant clinical benefits because the results of PWV are different for different vessels and there is limited data on the reproducibility of the test [33]. Celermajer et al. [17] originally designed flowmediated vasodilation of the superficial artery using high-resolution ultrasound B-mode measurement. Flow-mediated vasodilation may be induced by increased blood flow (reactive hyperemia) after vessel occlusion and release. Shear stress induced by increased blood flow is the major stimuli for the release of nitric oxide (endothelial dependent relaxing factors) from the endothelium. Nitric oxide plays a major role in the control of vascular homeostasis. Decreased FMD is regarded as endothelial dysfunction. Measurement of brachial artery FMD
using high-resolution ultrasound is widely available in clinical settings due to its high reproducibility, repeatability and easy access to the superficial artery. For the last decade many studies have reported that decreased FMD in the brachial artery is associated with risk factors for coronary artery disease and has good correlation with the FMD of the coronary artery [16]. Therefore, early detection of decreased FMD value in patients with risk factors for coronary artery disease with no symptoms is clinically important because endothelial dysfunction in these groups may contribute to the progression of atherosclerosis. However, recently inconsistencies in the results of non-invasive measurements of FMD in the brachial artery have been reported. There are limitations [30] to the use of non-invasive measurement of FMD for screening or early detection of endothelial dysfunction. The gold standard for assessment of vascular function is angiographic technique, which requires arterial cannulation with infusion of pharmacokinetic drugs. The limitation is that serial follow-up using angiographic technique is difficult in the clinical setting due to its invasiveness. In addition, the technique has a high difficulty level as the examiner must hold the probe without movement during the measurement of FMD in peripheral artery. One report has suggested that the correlation between brachial artery FMD and coronary artery endothelial function is not strong. A further limitation is that FMD results occur over a wide range due to the variability of endothelial reactions to EDRFs. According to the results of this study, although the difference in endothelium-dependent vasodilation was significant between the two groups, the range of FMD results was wide. Therefore, the use of FMD as screening tools for the early detection of endothelial dysfunction has some difficulties. Using non-invasive methods of flow-mediated, endothelium-dependent vasodilation, for evaluation of endothelial dysfunction in adult patients with risk factors, such as hypertension, diabetes and obesity, there was a clearly defined difference in FMD results compared to those of the healthy controls. Using the same method for early evalua-
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
tion of endothelial dysfunction in healthy children and young adults with or without risk factors for atherosclerosis has no clinical benefit. These patients have mild endothelial dysfunction, such as physiological change however, some other factor such as sex, arm length, occlusion pressure of the BP cuff or autonomic neuropathy may also affect endothelium-dependent vasodilation. The widely ranging FMD results in young people with no symptoms are also regarded as a problem in the evaluation of impaired endothelial function [30]. In comparison, the results for lag time to initial reaction of flow-mediated, endothelial dependent vasodilation (IRT) show a narrower range (relative dispersion) than those for endothelium-dependent vasodilation. Therefore, additional measurements of lag time to initial reaction of FMD for early detection of endothelial dysfunction using a special device increase the reliability of the results of the non-invasive technique. Then, the observer overcomes the limitation of classical methods, such as peak time variability and change of measurement site. This novel approach in detection of endothelial dysfunction is simple and highly reproducible. In conclusion, these findings firstly indicate that in addition to the percentage change in FMD, lag time from base line to initial response time (IRT) of FMD is a good indicator to evaluate endothelial dysfunction. Also, that this new parameter may be a more sensitive parameter for differentiation between the diseased state and the normal state of the endothelium than the percentage changes in FMD. Secondly, this study found that endothelial dysfunction was prominent in postmenopausal women with diabetes and was significantly improved by estrogen but not sufficiently. These results suggest that other factors in addition to estrogen deficiency play a role in endothelial dysfunction in postmenopausal women with diabetes mellitus.
Acknowledgements This study was supported by a grant from the Institute for Medical genetics, Keimyung University School of Medicine (97– 1) and by a grant
S91
from the Korea Health 21 R&D project, Ministry of Health & Welfare, Republic of Korea (HMP99-M-08-0004).
References [1] R. Furchgott, D. Zawadzki, The obligatory role of endothelial cells in the relaxation of arterial smooth muscle cell by acetylcholine, Nature 288 (1980) 373 – 376. [2] R.M.J. Palmer, D.S. Ashton, S. Moncada, Vascular endothelial cells synthesize nitric oxide from L-arginine, Nature 333 (1988) 664 – 666. [3] S. Moncada, Biological importance of prostacyclin, Br. J. Pharmacol. 76 (1982) 3 – 31. [4] P.M. Vanhoutte, Endothelial dysfunction and atherosclerosis, Eur. Heart J. 18 Suppl. E (1997) E19 – E29. [5] B. Tesfamariam, Free radicals in diabetic endothelial cell dysfunction, Free Radic. Biol. Med. 16 (1994) 383 – 391. [6] M. Yanagisawa, H. Kurihara, S. Kimura, A novel potent vasoconstrictor peptide produced by vascular endothelial cells, Nature 332 (1988) 411 –415. [7] M.W. Radomski, R.M.J. Palmer, S. Moncada, An Larginine/nitric oxide pathway present in human platelets regulates aggregation, Proc. Natl. Acad. Sci. 87 (1990) 5193 – 5197. [8] P. Kubes, M. Suzuki, N. Granger, Nitric oxide: an endogenous modulator of leukocyte adhesion, Proc. Natl. Acad. Sci. 88 (1991) 4651 – 4655. [9] U.C. Grag, A. Hassid, Nitric oxide-generating vasodilators and 8-bromo-cyclic guanosine monophosphate inhibit mitogenesis and proliferation of cultures rat vascular smooth muscle cells, J. Clin. Invest. 83 (1989) 1774 – 1777. [10] S.B. Williams, J.A. Cusco, M.A. Roddy, M.T. Johnstone, M.A. Creager, Impaired nitric oxide-mediated vasodilation in patients with non-insulin-dependent diabetes mellitus, J. Am. Coll. Cardiol. 27 (1996) 567 – 574. [11] M.T. Johnstone, S.J. Creager, K.M. Scales, J.A. Cusco, B.K. Lee, M.A. Creager, Impaired endothelium-dependent vasodilation in patients with insulin dependent diabetes mellitus, Circulation 88 (1993) 2510 – 2516. [12] H.O. Steinberg, H. Chaker, R. Leaming, A. Johnson, G. Brechtel, A. Baron, Obesity/insulin resistance is associated with endothelial dysfunction, J. Clin. Invest. 97 (1996) 2601 – 2610. [13] M. Barenbrock, M. Hausberg, M. Kosch, S.A. Golubev, K. Kisters, K. Rahn, Flow-mediated vasodilation and distensibility in relation to intima-media thickness of large arteries in mild essential hypertension, Am. J. Hypertens. 12 (1999) 973 – 979. [14] K.E. Sorensen, D.S. Celermajer, D. Georgakopoulos, G. Hatcher, Impairment of endothelium-dependent dilation is an early event in children with familial hypercholesterolemia and is related to the lipoprotein (a) level, J. Clin. Invest. 93 (1994) 50 – 55.
S92
S.J. Lee et al. / Diabetes Research and Clinical Practice 54 Suppl. 2 (2001) S81 – S92
[15] P.L. Ludmer, A.P. Selwyn, T.L. Shook, et al., Paradoxical vasoconstriction induced by acetylcholine in atherosclerotic coronary arteries, N. Engl. J. Med. 315 (1986) 1046 – 1051. [16] E.L. Nabel, A.P. Selwyn, P. Ganz, Large coronary arteries in humans are responsive to changing blood flow: an endothelium dependent mechanism that fails in patients with atherosclerosis, J. Am. Coll. Cardiol. 16 (1990) 349– 356. [17] D.S. Celermajer, K.E. Sorensen, V.M. Gooch, et al., Non-invasive detection of endothelial dysfunction in children and adults at risk of atherosclerosis, Lancet 340 (1992) 1111 – 1115. [18] R.D. Fish, E.G. Nabel, A.P. Selwyn, et al., Responses of coronary arteries of cardiac transplant patients to acetylcholine, J. Clin. Invest. 81 (1988) 21 –31. [19] National Diabetes Data Group, Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance, Diabetes 28 (1979) 1039 –1057. [20] R.J. Gryglewski, R.M. Palmer, S. Moncada, Superoxide anion is involved in the breakdown of endotheliumderived vascular relaxing factor, Nature 320 (1986) 454 – 456. [21] R. Ross, The pathogenesis of atherosclerosis — an update, N. Engl. J. Med. 314 (1986) 488 – 500. [22] E.H. Lieberman, M.D. Gerhard, A. Uehata, et al., Estrogen improves endothelium-dependent, flow-mediated vasodilation in postmenopausal women, Annals Int. Med. 121 (1994) 936 – 941. [23] H. Tagawa, H. Shimokawa, T. Tagawa, M. KuroiwaMatsumoto, Y. Hirooka, A. Takeshita, Short-term estrogen augments both nitric oxide-mediated and non-nitric oxide-mediated endothelium-dependent vasodilation in postmenopausal women, J. Cardiov. Pharmacol. 30 (1997) 481 – 488. [24] D.E. Bush, C.E. Jones, K.M. Bass, G.K. Walters, J.M. Bruza, P. Ouyang, Estrogen replacement reverses en-
[25] [26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
dothelial dysfunction in postmenopausal women, Am. J. Med. 104 (1998) 552 – 558. H. Shimokawa, Primary endothelial dysfunction: atherosclerosis, J. Mol. Cell. Cardiol. 31 (1999) 23 – 37. P. Pignoli, E. Tremoli, A. Poli, P. Oreste, R. Paoletti, Intimal plus medial thickness of the arterial wall: a direct measurement with ultrasound imaging, Circulation 74 (1986) 1399 – 1406. D.J. Farrar, M.G. Bond, W.A. Riley, J.K. Sawyer, Anatomic correlates of aortic pulse wave velocity and carotid artery elasticity during atherosclerosis progression and regression in monkeys, Circulation 83 (1991) 1754 – 1763. J.L. Wofford, F.R. Kahl, G.R. Howard, W.M. McKinney, J.F. Toole, J.R. Crouse, Relation of extent of extracranial carotid artery atherosclerosis as measured by B-mode ultrasound to the extent of coronary atherosclerosis, Arterioscler. Thromb. 11 (1991) 1786 – 1794. L. Niskanen, R. Rauramaa, H. Miettinen, S.M. Haffner, M. Mercuri, M. Uusitupa, Carotid artery intima-media thickness in elderly patients with NIDDM and in non-diabetic subjects, Stroke 27 (1996) 1986 – 1992. E.D. Lehmann, W.A. Riley, P. Clarkson, R.G. Gosling, Non-invasive assessment of cardiovascular disease in diabetes mellitus, Lancet 350 (1) (1997) 14 – 19. M.R. Adams, A. Nakagomi, A. Keech, et al., Carotid intima-media thickness is only weakly correlated with the extent and severity of coronary artery, Circulation 92 (1995) 2127 – 2134. M. Karamanoglu, M.F. O’Rourke, A.P. Avolio, R.P. Kelly, An analysis of the relationship between central aortic and peripheral upper limb pressure waves in man, Eur. Heart J. 14 (1993) 160 – 167. A. Uehata, A.M.D. Gerhard, I.T. Meredith, et al., Close relationship of endothelial dysfunction in coronary and brachial artery (Abstract), Circulation 88 (1983) I-618.