| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of Cardiology B (H.M.S., M.B., M.M.M., T.T.N., H.E.B.), Clinical Pharmacology (H.M.S., O.S.), and Endocrinology (O.S.) and The PET Center (S.B.H.), Aarhus University Hospital, Aarhus University, DK-8200 Aarhus, Denmark
Address all correspondence and requests for reprints to: Hans Erik Bøtker M.D., Ph.D., D.M.Sc., Department of Cardiology B, Aarhus University Hospital (SKS), Brendstrupgaardsvej 100, DK-8200 Aarhus N, Denmark. E-mail: heb{at}dadlnet.dk.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Methods: We studied 27 CAD patients (LVEF > 50%): 12 with T2DM (CAD+DM), 15 without T2DM (CAD-NoDM). Regional myocardial and skeletal glucose uptake, myocardial and skeletal muscle perfusion were measured with positron emission tomography. Myocardial muscle perfusion was measured at rest and during hyperemia in nonstenotic and stenotic regions with and without acute hyperinsulinemia.
Results: Myocardial glucose uptake was similar in CAD+DM and CAD-NoDM in both nonstenotic and stenotic regions [0.38 ± 0.08 and 0.36 ± 0.11 µmol/g·min; P value nonsignificant (NS)] and (0.35 ± 0.09 and 0.37 ± 0.13 µmol/g·min; P = NS). Skeletal glucose uptake was reduced in CAD+DM (0.05 ± 0.04 vs. 0.10 ± 0.05 µmol/g·min; P = 0.02), and likewise, whole-body glucose uptake was reduced in CAD+DM (4.0 ± 2.8 vs. 7.0 ± 2.4 mg/kg·min; P = 0.01). Insulin did not alter myocardial muscle perfusion at rest or during hyperemia. Insulin increased skeletal muscle perfusion in CAD-NoDM (0.11 ± 0.03 vs. 0.06 ± 0.03 ml/g·min; P = 0.02), but not in CAD+DM (0.08 ± 0.04 and 0.09 ± 0.05 ml/g·min; P = NS).
Conclusion: Myocardial IR to glucose uptake is not an inherent feature in T2DM patients with preserved LVEF. Acute physiological insulin exposure exerts no coronary vasodilation in CAD patients irrespective of T2DM.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Impaired insulin-dependent glucose uptake may involve metabolic pathways responsible for cellular glucose uptake as well as endothelium-dependent vascular systems responsible for vasodilatation. Insulin dilates the microvasculature in skeletal muscle, whereas this effect is less pronounced in the heart (8, 9, 10, 11, 12). In patients with CAD, intracellular glucose availability is likely to play a pivotal role for the recovery of contractile function after ischemia (13). Insulin promotes glucose metabolism in the ischemically injured myocardium, which retains the ability to increase glucose uptake after insulin stimulation regardless of any myocardial IR (7) and contractile dysfunction (13). Because IR is associated with an increased incidence of congestive heart failure (14), we hypothesized that patients with T2DM and CAD have impaired insulin-stimulated myocardial perfusion and glucose uptake in regions with preserved myocardial contractility supplied by nonstenotic coronary arteries.
Consequently, the present study was undertaken to compare regional basal and insulin-stimulated perfusion and glucose uptake in myocardial regions supplied by stenotic (S) and non-S coronary arteries in CAD patients with and without T2DM and preserved global left ventricular function.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
We studied 27 patients with angiographically verified CAD recruited from a percutaneous coronary intervention/coronary artery bypass grafting waiting list. All patients were required to have at least one area supplied by a S coronary artery (>70% stenosis) and one area supplied by a non-S coronary artery (<20% stenosis). Twelve patients had T2DM (fasting plasma glucose 
7 mmol/liter) according to the American Diabetes Association criteria (CAD+DM), duration 6 ± 5 yr. Fifteen patients had CAD but no diabetes (CAD-NoDM). The CAD-NoDM patients were matched according to the degree of coronary disease, age, and gender to a similar group of CAD+DM patients. None of the patients had left ventricle hypertrophy, evaluated by echocardiography, and all patients had preserved left ventricular function defined as left ventricular ejection fraction more than 50%. All medication was paused for 12 h before the examination on both study days. Three patients did not complete the full protocol.
The study was conducted in accordance with the Declaration of Helsinki and approved by the local ethics committee. All participants signed an approved consent form before entering the study.
Study design
After an overnight fast, myocardial and skeletal glucose uptake, as well as myocardial and skeletal perfusion was measured by positron emission tomography (PET). All patients refrained from smoking and intake of methylxanthines including caffeine for at least 24 h before the study. The patients were examined twice in random order. On d 1, one myocardial muscle perfusion (MBF) baseline protocol was done, consisting of one measurement at rest and one during dipyridamole-induced hyperemia (0.56 mg/kg). Simultaneously with the MBF rest measurement, we measured skeletal muscle perfusion (SBF). On d 2, one MBF insulin protocol and one MGU protocol was done. During steady-state hyperinsulinemic euglycemic clamp, we performed the same perfusion sequence as d 1, and one simultaneous MGU and skeletal muscle glucose uptake (SGU). Three venous cannulas were inserted: 1) one in a hand vein for sampling of arterialized venous blood, 2) one in the left forearm used for infusion of insulin (Actrapid 100 IE/ml; Novo Nordisk, Gentofte, Denmark) and 20% glucose, and finally 3) one in the right forearm for the administration of dipyridamole and radioactive tracers.
Hyperinsulinemic euglycemic clamp
Insulin infusion rate was 2 mU/kg·min during the first hour, followed by 1 mU/kg·min for the rest of the examination. Plasma glucose concentration was clamped at 5 mmol/liter by a variable infusion of glucose. Plasma glucose concentration was measured every 5 min in blood drawn from an arterialized venous cannula. Whole-body glucose uptake (M-value) was assessed during the final 30-min period of glucose infusion, when the rate of glucose infusion approximates whole-body glucose disposal. The first perfusion scan was initiated after 120 min of insulin infusion and the [18F]fluorodeoxyglucose (FDG) scan was initiated after 220 min of insulin infusion.
Regions of interest (ROI)
The entire left ventricle was delineated according to a semiautomatic delineation technique that included any defect in the myocardium as described earlier (15). A bull’s eye plot was generated from the delineation where the three 90° sectorial ROIs were assigned and corresponded to the distribution of the left anterior descending, left circumflex, and right coronary arteries (16). The input function was obtained from the center of the left ventricular blood pool.
Myocardial glucose uptake
Exogenous myocardial glucose use was quantified by PET (model EXACT HR 961; Siemens/CTI, Knoxville, TN) using iv FDG as glucose tracer [produced by adaptation of a standard procedure using a prefabricated kit (FDG Microlab) from GE Medical Systems, Uppsala, Sweden]. FDG was injected (200 MBq diluted in 10 ml saline) over 20 sec, and a 69-min dynamic scan was acquired consisting of 24 frames (8 x 15, 6 x 30, 4 x 60, and 6 x 600 sec).
The parameters required for calculation of myocardial glucose uptake from FDG retention were generated by graphical analysis as described by Gjedde (17) and Patlak and Blasberg (18). The rate of MGU was obtained by multiplying the net uptake rate of the glucose analog by the plasma glucose concentration divided by the lumped constant (LC). We used a variable LC calculated individually as described earlier (19, 20). MGU was also calculated using a fixed LC of 1.0. Correction for partial volume effect and spillover were performed as described earlier (21).
Myocardial perfusion
Myocardial perfusion was quantified by PET using iv [13N]ammonia as perfusion tracer (22). For each perfusion scan, 740 MBq of [13N]ammonia diluted in 10 ml saline was injected over 20 sec. At the time of injection, acquisition of a dynamic sequence of images (12 x 10, 2 x 30, 1 x 60, and 1 x 900 sec) was started to obtain time-activity curves from the blood pool and from the myocardium. The last frame was acquired to obtain a high-contrast image that was used for ROI allocation. Myocardial blood flow was quantified by fitting the tissue and blood pool time-activity curves to a validated two-compartment model for [13N]ammonia (23, 24).
SGU and perfusion
SGU and SBF were quantified in the brachial muscle in a subset of patients. For both SGU and SBF, an image-derived input function was sampled in the ascending aorta because activities obtained with PET correlates with the arterial plasma time-activity curve (25). Images were not resliced, but ROI were drawn on the medial and lateral brachial muscles on the contralateral arm to the injection site. A LC of 1.2 was used to derive SGU (26). Perfusion was quantified using a two-compartment model for [13N]ammonia (23, 24). The parameters required for calculation of SGU and SBF were generated by graphical analysis as described by Gjedde (17) and Patlak and Blasberg (18).
Analytical methods
Plasma insulin levels were analyzed by ELISA using a two-site immunoassay (Dako Diagnostics Ltd., Cambridge, UK). Free fatty acids (FFA) were measured by ELISA (Wako Chemicals, Dusseldorf, Germany) in all patients at 30-min intervals throughout the study. Plasma glucose levels were measured immediately by a glucose analyzer (Beckman Instruments, Palo Alto, CA). All measurements were made in duplicate, and the mean values are reported.
Calculations and statistical analysis
We report data from one nonstenosed and one stenosed region for each patient. Data are presented as mean ± SD. We report the absolute values of perfusion, mean arterial pressure (MAP), and the minimal vascular resistance index [MVRindex= (MAP/perfusiondipyridamole)/(MAP/perfusionbaseline)]. MAP was calculated as diastolic blood pressure + 1/3 pulse pressure. MVR was calculated as MAP/perfusion. Paired and unpaired Student’s t tests or Wilcoxon signed rank sum tests were used to compare two mean values. The choice of parametric or nonparametric methods was based on normality plots. One-way ANOVA was performed to compare more than two values and followed by a pairwise post hoc t test modified according to Bonferroni when statistically significant differences were detected. We used multiple linear regression analysis to study the relation between metabolic variables, body mass index (BMI), and SGU. The group size was determined from a calculation of strength with a desired power of 0.8 and a level of significance of 0.05.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Age, BMI, and smoking habits were similar in CAD+DM and CAD-NoDM patients (Table 1
). There were no differences between CAD+DM and CAD-NoDM patients regarding localization of atherosclerotic lesions or history of hypertension, previous myocardial infarction, or hypercholesterolemia. We found no difference in medication between the study groups except for antidiabetic drugs.
|
Plasma glucose levels were significantly increased in CAD+DM at baseline (9.4 ± 0.4 vs. 5.6 ± 0.5 mmol/liter; P < 0.001), and in the CAD+DM group, plasma glucose level declined during the first hour of hyperinsulinemia. Steady state was reached after approximately 2 h of insulin infusion when plasma glucose levels were comparable in the two groups (Fig. 1A). The glucose infusion rate required to maintain euglycemia during the clamp was significantly increased in the CAD-NoDM patients who required between 1.5 and 6.8 times more glucose than the CAD+DM patients throughout the study (P < 0.05) (Fig. 1A). There was no statistically significant difference in serum FFA concentrations at baseline between the two groups. After the initiation of the clamp, serum FFA levels were almost completely suppressed with a decline from 0.6 ± 0.3 to 0.13 ± 0.15 mmol/liter in the CAD+DM group and 0.53 ± 0.2 to 0.04 ± 0.03 mmol/liter in the CAD-NoDM group. In the CAD-NoDM group, serum FFA level remained suppressed to a lower level than the CAD+DM group throughout the study period (P < 0.05). Baseline insulin levels were similar in the CAD+DM and CAD-NoDM patients. During insulin infusion, plasma insulin levels increased similarly and were stable during the whole study (Fig. 1B). Whole-body glucose uptake defined as M-value was reduced in CAD+DM patients, 4.0 ± 2.8 vs. 7.0 ± 2.4 mg/kg·min (P = 0.01), compared with CAD-NoDM patients. The difference in M-value was independent of BMI (ANOVA P = 0.63).
|
Blood pressure, heart rate, MAP, and MVRindex were similar in CAD+DM and CAD-NoDM patients at rest and increased similarly during dipyridamole-induced hyperemia. The hemodynamic parameters did not change during insulin infusion (Table 2).
|
Insulin-stimulated MGU was similar in non-S and S regions in CAD+DM (0.38 ± 0.08 and 0.35 ± 0.09 µmol/g·min; P = 0.30) and CAD-NoDM patients (0.36 ± 0.11 and 0.37 ± 0.13 µmol/g·min; P = 0.80). MGU was also similar in CAD+DM and CAD-NoDM patients in non-S regions (0.38 ± 0.08 and 0.36 ± 0.11 µmol/g·min; P = 0.56) and S regions (0.35 ± 0.09 and 0.37 ± 0.13 µmol/g·min; P = 0.56) (Fig. 2A). We found no correlation between whole-body glucose uptake and MGU in either non-S (r2 = 0.001; P = 0.89) or S (r2 = 0.01; P = 0.67) regions.
|
Insulin-stimulated SGU was significantly lower in CAD+DM patients than in CAD-NoDM patients (0.05 ± 0.04 vs. 0.10 ± 0.05 µmol/g·min; P = 0.02), corresponding to a 50% reduction in SGU (Fig. 2B
). We observed a significant correlation between whole-body glucose uptake and SGU (r2 = 0.7; P < 0.001). The difference in SGU was independent of BMI (ANOVA P = 0.58).
MBF
Baseline perfusion was similar in non-S and S regions in CAD+DM (0.94 ± 0.26 and 0.88 ± 0.27 ml/g·min; P = 0.60) and CAD-NoDM patients (0.80 ± 0.19 and 0.82 ± 0.14 ml/g·min; P = 0.78). Dipyridamole increased MBF similarly in non-S and S regions in CAD+DM (2.28 ± 1.05 and 1.88 ± 0.95 ml/g·min; P = 0.36) and CAD-NoDM patients (2.38 ± 1.15 and 1.88 ± 0.75 ml/g·min; P = 0.20). Flow reserve (MBFdipyridamole/MBPrest) was reduced in S regions in CAD+DM (2.61 ± 1.38 vs. 2.19 ± 1.00; P = 0.02) as well as in CAD-NoDM, (2.98 ± 1.33 vs. 2.35 ± 0.94; P = 0.006) compared with non-S regions. Insulin stimulation did not alter flow reserve in both study groups: CAD+DM non-S regions (2.61 ± 1.38 and 2.62 ± 0.86; P = 0.98), CAD+DM S regions (2.19 ± 1.0 and 2.43 ± 1.0; P = 0.57), CAD-NoDM non-S regions (2.98 ± 1.33 and 2.93 ± 1.29; P = 0.92), and CAD-NoDM S regions (2.35 ± 0.94 and 2.48 ± 1.09, P = 0.74). Insulin stimulation did not alter MBF at rest in CAD+DM patients [non-S regions, (0.94 ± 0.26 and 0.99 ± 0.19 ml/g·min (P = 0.29); S regions, 0.88 ± 0.26 and 0.91 ± 0.27 ml/g·min (P = 0.49)] (Fig. 3A) or in CAD-NoDM patients [non-S regions, 0.80 ± 0.19 and 0.81 ± 0.24 ml/g·min (P = 0.88); S regions, 0.82 ± 0.13 and 0.85 ± 0.21 ml/g·min (P = 0.67)] (Fig. 3B). Similar findings were observed during dipyridamole-induced hyperemia: CAD+DM patients [non-S regions, 2.28 ± 1.05 and 2.53 ± 0.82 ml/g·min (P = 0.54); S regions, 1.89 ± 0.95 and 2.13 ± 0.70 ml/g·min (P = 0.53)] (Fig. 3C) and CAD-NoDM patients [non-S regions, 2.38 ± 1.15 and 2.27 ± 0.88 ml/g·min (P = 0.78); S regions, 1.88 ± 0.75 and 1.98 ± 0.70 ml/g·min (P = 0.71)] (Fig. 3D).
|
In skeletal muscle, insulin increased perfusion in CAD-NoDM patients (0.06 ± 0.03 vs. 0.11 ± 0.03 ml/g·min; P = 0.02) (Fig. 4). In contrast, we found no significant difference between basal and insulin-stimulated perfusion in CAD+DM patients (0.08 ± 0.04 and 0.09 ± 0.05 ml/g·min; P = 0.75) (Fig. 4).
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
We undertook the present study because previous studies have revealed conflicting results about T2DM and myocardial IR to glucose uptake. Myocardial IR to glucose uptake may be a rational pathophysiological mechanism behind the increased prevalence of congestive heart failure in patients with CAD and T2DM. If the whole-body IR to glucose uptake associated with congestive heart failure (14) also involves the heart, the metabolic disarray may compromise myocardial energy metabolism (27). Vascular IR may further alter perfusion in ischemic and nonischemic regions (27) such that both mechanisms may lead to myocardial dysfunction. Part of the discrepancy regarding myocardial IR in T2DM has been raised by the presence of concomitant CAD (2, 7, 28). Coronary and systemic atherosclerosis is known to be associated with a generalized state of IR. However, the important finding in our study is that in patients with CAD, T2DM does not per se generate additional myocardial IR to glucose uptake despite whole-body IR. These findings may also support the idea that hyperglycemia as observed in diabetics may not impair vascular function (29).
Because our results may add further controversy about myocardial IR to glucose uptake in patients with T2DM, we sought to optimize patient selection. We studied only patients with preserved left ventricular function to avoid the influence of reduced contractility, which may in itself compromise MGU (30). We carefully matched patients according to gender, age, and CAD location. We ensured stable and similar metabolic conditions by preparing all patients with the same dose of insulin. Consequently, all patients were euglycemic with the same degree of hyperinsulinemia and an almost similar FFA suppression at the time of examination. We examined perfusion at rest and during hyperemia and MGU in regions supplied by S and non-S coronary arteries, allowing characterization of MGU in ischemic as well as nonischemic myocardial regions. Finally, we used a variable LC in the assessment of MGU because this is the physiologically correct approach. However, the use of a fixed LC did not alter our findings.
In healthy humans, a small increase in resting myocardial perfusion occurs early during physiological insulin stimulation (31). However, studies of the forearm have shown that insulin increases glucose extraction more rapidly than perfusion and that supraphysiological insulin levels are required to achieve stimulation of perfusion in skeletal muscle (32). Consequently, we intended to obtain circulating insulin levels in the high physiological but not truly pharmacological range for a relatively long period of time to allow a potential insulin-stimulated vasodilation. Measurement of perfusion demonstrated an insulin-stimulated increment in skeletal muscle in CAD-NoDM but not in the CAD+DM patients, confirming that whole-body IR to glucose uptake in CAD+DM is associated with IR at the vascular level in skeletal muscle. This pattern was different in the heart. Not only insulin-stimulated glucose uptake but also resting as well as hyperemic perfusion did not differ between patients with and without T2DM in either S or non-S regions, respectively. Because diabetes cannot further deteriorate the measured responses compared with the state that is caused by CAD, these findings argue against vascular IR specifically associated with T2DM as a predominant factor accounting for the increased morbidity in this disorder. However, our findings do not argue against the idea that patients with CAD are characterized by coronary vascular IR independently of myocardial insulin sensitivity to glucose uptake (33). Despite similar perfusion as in another recent study of T2DM patients with CAD using 15O-labeled H2O as flow tracer, we did not find any significant vasodilating effect of insulin at rest or during hyperemia (34). During hyperemia, metabolic control of flow is uncoupled in contrast to the resting conditions. Hyperemia obtained by dipyridamole stimulation consequently allows testing of the independent effect of insulin on coronary vasodilation. Insulin is predominantly an endothelium-dependent vasodilator. Patients with CAD and T2DM suffer from endothelial dysfunction and are frequently characterized by a blunted vascular response to insulin. This vascular IR, consistent with endothelial dysfunction, appears to originate in an imbalance between endothelium-derived vasoconstrictor and vasodilator effects (27). Because insulin seems to have important implications for normal vascular function, our results are in accordance with the suggestion that vascular IR may provide a mechanism involved in the development of atherosclerosis in the coronary arteries. The facts that the vascular response to insulin was similar in regions supplied by S and non-S arteries and that it did not differ between the patient groups indicate that the degree of IR does not account for any difference in insulin-dependent vascular reactivity in the coronary arteries in patients with and without T2DM. A recent study suggests that postprandial hyperglycemia may be a more important determinant for CAD than IR and hyperinsulinemia (35). Thus, the biological background for the development of CAD may be an extended period of IR that characterizes diabetes from the prediabetic into the hyperglycemic stage. Whether differences in IR in skeletal muscle is pathogenically involved in the more prevalent peripheral atherosclerosis observed among patients with diabetes and CAD than among CAD patients without concomitant diabetes (20) remains unknown.
Limitations
Because we did not compare our findings with healthy subjects, we cannot define to which extent patients with CAD are whole-body IR and IR in the myocardium. Because it has already been established that patients with CAD are IR, we merely aimed at defining any additional degree of IR in T2DM and its impact on MGU and coronary vascular function.
Our patients were recruited from a percutaneous coronary intervention/coronary artery bypass grafting waiting list and suffered from varying degrees of CAD. The degree of coronary stenosis varied from 70–90% and was located on the three coronary arteries or major side branches. It was, however, not possible to make a full match of the angiograms.
Circulating FFA levels were not strictly matched in the study groups when suppression was delayed because of elevated baseline values in T2DM patients. Although MGU is influenced by FFA, we consider the differences between the study groups of minor importance because no difference in myocardial insulin sensitivity was detected.
We studied glucose disposal under high physiological hyperinsulinemia and euglycemia. This may mask the effect of diabetes on myocardial function.
The limitation of our PET equipment allows detection of glucose uptake and perfusion differences between the two groups of 20% or more.
Skeletal muscle perfusion measurements were performed using 13NH3 as perfusion tracer. In contrast to 15O-labeled H2O, 13NH3 has not been validated as a tracer for skeletal muscle perfusion. Skeletal muscle perfusion is 10% of myocardial perfusion, and quantification of perfusion using tracer kinetic principles requires accurate determination of 13NH3 in the blood because 13NH3 is rapidly metabolized (36). We sought to avoid contamination by metabolites using identical input functions for MBF and SBF measurements. Our findings also relied on the fact that 13NH3 is valid for quantification of cerebral perfusion, which is in the range of 50–100% of MBF. We obtained excellent image quality useful for proper quantitative analysis in a subgroup of our patients. Our findings are within the perfusion range that has previously been measured with a validated perfusion tracer in skeletal muscle.
Conclusion
This study demonstrates that myocardial IR to glucose uptake is not an inherent feature of T2DM irrespective of the degree of epicardial disease and whole-body IR when myocardial contractility is preserved. When administered in high physiological doses, insulin does not seem to act as a vasodilator in CAD patients with and without T2DM in regions of the myocardium supplied by S and non-S coronary arteries.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
First Published Online September 19, 2006
Abbreviations: BMI, Body mass index; CAD, coronary artery disease; IR, insulin resistance; LC, lumped constant; MAP, mean arterial pressure; MBF, myocardial muscle perfusion; MGU, myocardial glucose uptake; MVR, minimal vascular resistance; PET, positron emission tomography; S, stenotic; SBF, skeletal muscle perfusion; SGU, skeletal muscle glucose uptake; T2DM, type 2 diabetes mellitus.
Received July 3, 2006.
Accepted September 12, 2006.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. J. Gibbons, P. A. Araoz, and E. E. Williamson The Year in Cardiac Imaging J. Am. Coll. Cardiol., September 4, 2007; 50(10): 988 - 1003. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) and CAD-NoDM (
) and mean glucose infusion rate for CAD+DM (
) and CAD-NoDM (
) during the study period; B, serum insulin (S-insulin) values for CAD+DM (
), CAD-NoDM non-S regions (
), and CAD-NoDM S regions (
). For all comparisons pairwise, P value is not significant. B, Insulin-stimulated SGU in CAD+DM patients (•) and CAD-NoDM patients (
