Journal Logo

Article

Insulin Sensitivity in a Long-Term Crossover Trial with Celiprolol and Other Antihypertensive Agents

Malminiemi, K.*†; Lahtela, J.; Malminiemi, O.*; Ala-Kaila, K.§; Huupponen, R.

Author Information
Journal of Cardiovascular Pharmacology: January 1998 - Volume 31 - Issue 1 - p 140-145
  • Free

Abstract

Cardiovascular risk factors are clustered (1). Impaired insulin sensitivity and hyperinsulinemia are often seen in patients with hypertension (2), and antihypertensive treatment may further aggravate metabolic changes (3). First- and second-generation β-blockers (e.g., propranolol, atenolol, metoprolol, and pindolol) have been reported to deteriorate insulin sensitivity (4). β-Blockers with other ancillary pharmacodynamic properties may overcome this harmful effect. Preliminary studies with celiprolol, a β1-selective adrenoceptor antagonist with partial β2-adrenoceptor-agonist properties (5), revealed beneficial effects on carbohydrate metabolism in hypertensive patients with dyslipidemia (6) and in patients with hypertension and non-insulin-dependent diabetes mellitus (NIDDM) (7). We carried out a long-term, crossover trial to clarify the underlying mechanisms and to exclude the possibility that the observed beneficial effects might be the result of the withdrawal of a disadvantageous antihypertensive agent.

METHODS

Study design

The trial was a randomized, investigator-masked, two-period crossover study with an active run-in period. The previous antihypertensive monotherapy with a β-blocker, a calcium channel blocker, or an angiotensin-converting enzyme (ACE) inhibitor was used as the reference therapy for celiprolol. The study consisted of three phases: a run-in period of ≥3 months, and treatment periods I and II of 12 and 6 months' duration, respectively. On the prestudy visit, patients were informed about healthful living habits and encouraged to maintain them throughout the trial. During the run-in period, all patients received their previous antihypertensive monotherapy with constant dosage. Then, after the baseline tests, the patients were randomized in two groups. During period I, group A received celiprolol 200-400 mg/day, and group B continued on their previous monotherapy. In period II, group A was changed back to the prestudy medication, and group B received celiprolol. Metabolic studies were carried out at the end of the run-in period and then every 6 months. Exercise, dietary, and sedentary habits of the patients were assessed at each visit by using a selfadministered questionnaire with 40 items.

Patients

The study population consisted of 25 hypertensive patients with dyslipidemia [grades 1-2 (8); age range from 42 to 64 years] who had been receiving antihypertensive therapy for ≥6 months before the baseline tests. Five patients had a β1-selective adrenoceptor blocker (metoprolol), nine had a calcium channel blocker of dihydropyridine type (five nifedipine and four felodipine), and 11 had an ACE inhibitor (11 enalapril and two captopril) as the prestudy and control therapy. Pregnant women, alcoholics, diabetics, excessively obese persons (BMI 35 kg/m2), and patients with other endocrine, liver, or kidney disease, or heart failure were excluded.

Seventeen patients of 25 had serum low-density lipoprotein (LDL) cholesterol 4.0 mM after the run-in period at randomization. Twenty patients had a high-density lipoprotein (HDL) cholesterol/total cholesterol ratio < 0.2, and 16 had a fasting serum triglyceride level 2.0 mM. Classification of hyperlipidemia (9) and demographic characteristics are presented in Table 1. At the beginning of the study, 20 patients had a reduced insulin sensitivity on the previous antihypertensive therapy when the cut-off point was set to the mean value (< 5.8 mg/kg/min) of middle-aged healthy men (10).

TABLE 1
TABLE 1:
Patient demographics and clinical characteristics at entry

Because of the moderate hypertension and antihypertensive medication for >2 years, no washout period was included in the protocol for ethical reasons. Good Clinical Trial Practice (Nordic Guidelines, 1989) was followed in the study. The study was approved by the Ethics Committees of Tampere and Turku University Hospitals.

Metabolic studies

Metabolic tests were carried out by different laboratories ≥1 week apart. The patients arrived in the laboratory at ∼8 a.m. after an overnight (12-h) fast. They were not allowed to take their study medication on the test day before the measurements. Studies were performed on average 25 h after the last administration of the antihypertensive medication.

Insulin sensitivity was determined by using a euglycemic hyperinsulinemic clamp (EC) technique (11) in Tampere University Hospital and in the University of Turku. The patients received insulin (Actrapid; Novo Nordisk A/S, Copenhagen, Denmark) with the infusion rate of 40 mU (288 pmol)/kg/m2 body surface area. Euglycemia (target blood concentration of 5 mM) was maintained by adjusting the rate of 20% dextran infusion according to whole blood glucose concentration measured from arterialized venous blood; the patient kept his or her right arm in a box containing heated air (60°C). Insulin and glucose were infused in the left arm. In healthy subjects, hepatic glucose production is completely suppressed when the serum insulin level is >60 mU/L (12). Here the expected insulin level in serum was 80 mU/L. Blood samples were drawn at 10- to 15-min intervals for the determination of blood glucose, serum insulin, free fatty acids, urate, and lactate. The insulin-sensitivity index (ISI) was calculated by dividing the average glucose-infusion rate by the mean steady-state blood glucose and serum insulin levels during a period of 60-120 min.

The oral glucose-tolerance tests (OGTTs) were carried out in Tampere City Hospital and in Meditori Medical Center, Turku. In the OGTT, the patient took 1.2 g of glucose/kg body weight in water (Glucodyn, Leiras, Finland). Blood was sampled after a 15-min rest before the glucose administration, and 60 min and 120 min thereafter for glucose and insulin measurements. The insulin secretory response and glucose appearance were calculated in terms of the incremental area under the respective time-concentration curve (AUC). Area up to 2 h was estimated by using linear interpolation between the concentration points, with the fasting value as the baseline (13). Negative glucose differences were zeroed.

Apolipoprotein E (ApoE) phenotypes were determined in delipidated plasma after isoelectric focusing and cysteamine treatment by using the immunoblotting technique (14). Serum insulin concentration was determined by using a commercial radioimmunoassay kit (Phadeseph Insulin RIA; Pharmacia, Uppsala, Sweden). Plasma fibrinogen was determined by using a turbidimetric technique (ACL 2000; Instrumentation Laboratories, Milan, Italy), and plasma sex hormone-binding globulin (SHBG) by using a modified Delfia method (Wallac, Turku, Finland).

Statistical analysis

Only the results of 6 months' treatment are included in the comparisons. This is because of the varying length of the treatment period (Fig. 1). The results of group A at 6 months and those of group B at 18 months (6 months taking celiprolol) were pooled to represent a 6-month celiprolol effect. The control data consisted of the results of group B at 6 months and group A at 18 months. Comparisons were carried out by using two-way analysis of variance with repeated measurements (RANOVA) and by using treatment and treatment order as categoric factors. A period effect was declared if the probability for similarity between the different periods was <0.1. All metabolic variables were normally distributed (p > 0.05 in D'Agostino's test) and homoscedastic in the two pooled groups. The analyses were carried out as intention-to-treat (25 patients). The mean, standard error of mean (SEM) and probabilities (two-sided) to accept the null hypothesis of an equal treatment effect during the two treatments are presented unless otherwise stated. In intention-to-treat analysis, the metabolic variables of the discontinued patients were carried forward from the point at 12-18 months.

FIG. 1
FIG. 1:
Insulin-sensitivity index (ISI) determined by a hyperinsulinemic (40 mU/min/m2) euglycemic clamp test. Mean ± SEM (n = 25) of intention-to-treat analysis. Solid symbols, celiprolol; open symbols, control antihypertensive monotherapy; triangles, group A; circles, group B. *p < 0.05 compared with previous medication.

RESULTS

Twenty-five hypertensive patients with dyslipidemia started the study medication. Two patients in both groups discontinued after 12 months of therapy before crossing the treatments. No statistically significant difference was observed in demographic variables or fasting plasma glucose, serum insulin, or lipid concentrations between the groups A and B at baseline before randomization. No statistically significant change occurred in the body mass index, smoking (five smokers), exercise, or dietary habits, or alcohol consumption during the trial. Body weight increased on average 0.4 kg during the celiprolol period and 0.1 kg during the control treatment (NS). No severe adverse effects were reported.

The average steady-state (60-120 min) blood glucose levels were between 4.9 and 5.1 mM in all EC tests for both groups; the range of all tests was 4.8-5.3 mM. The range of the mean steady-state serum insulin levels in the different EC tests varied between 71 and 77 mU/L with the maximal SEM of 3 mU/L. There was no statistically significant difference in steady-state values between the groups or between the periods. Celiprolol treatment significantly improved the ISI from 71.9 ± 7.2 to 94.7 ± 7.9 ml/min/kg/(U/L) (p = 0.0023; Table 2), corresponding to an increase of 32% (intention-to-treat analysis). In the control group, the mean of the ISI increased by 3%. If expressed as the relative individual change, celiprolol increased the ISI on average by 47% (range, −16-+169%). The average steady-state glucose-infusion rate (M expressed in mg/min/kg) increased on average by 35% during the 6-month celiprolol treatment and 6% when the patient was receiving the control antihypertensive therapy (Table 2). The increase of mean ISI for group A during celiprolol therapy was 39%, and for the group B, it was only 18% (Fig. 1). The beneficial effect of celiprolol was reversible. ISI decreased 24% in group A in 6 months after returning to the previous antihypertensive medication. In absolute terms, this decrease was on average 22.6 ml/kg/min/(U/L). ISI after 6 months' monotherapy with different antihypertensive agents is presented in Fig. 2.

TABLE 2
TABLE 2:
Metabolic variables after 6-month antihypertensive monotherapy
FIG. 2
FIG. 2:
Insulin-sensitivity index after 6-month therapy with different antihypertensive medications. Mean ± SEM of per protocol (measured data) analysis. The column of previous monotherapy represents the average insulin sensitivity after the open run-in period. *p < 0.05.

According to the OGTTs (Table 2 and Fig. 3), glucose tolerance increased significantly (p = 0.0087) during celiprolol treatment. The mean area of incremental serum glucose during the 2-h OGTT decreased by 36%, from 4.50 ± 0.59 to 2.88 ± 0.56 h × mM, after 6-month celiprolol treatment, and was practically unchanged during the control treatment. The serum insulin area during the 2-h OGTT decreased by 26%, from 138 ± 12 to 102 ± 12 h × mU/L during the 6-month celiprolol treatment (p = 0.0016). No change occurred during the control treatment. A tendency for a beneficial metabolic effect also was seen in fasting serum insulin, which decreased by 13% during celiprolol therapy (p = 0.066).

FIG. 3
FIG. 3:
Serum glucose and insulin during the oral glucose tolerance test. Glucose dose was 1.2 g/kg body weight. Mean ± SEM (n = 25). Solid circles, baseline at the end of the open run-in period; triangles, celiprolol; open circles, control treatment after 6-month therapy. Analysis of variance with repeated measurements was used for between-treatment comparisons. **p < 0.01; *p < 0.05.

Celiprolol had only a modest effect on serum cholesterol: LDL cholesterol decreased on average by 7% (NS), and HDL cholesterol increased by 7% (NS). A statistically significant beneficial effect was seen only in the HDL/LDL ratio, which increased by 15% (p = 0.012) during celiprolol treatment. Among all the lipid variables, only fasting serum triglycerides correlated significantly (r = 0.32; p < 0.01) with the ISI. The average decrease in triglycerides during the celiprolol period was 11% (NS). No statistically significant change was seen in plasma fibrinogen (p = 0.092), serum creatine kinase (p = 0.074), urate, or SHBG (Table 2). Serum potassium increased on average by 6% during celiprolol treatment (p = 0.015). Other electrolytes (Ca2+, Na+, Cl), liver enzymes (AST, ALT, GGT), or creatinine in serum did not change statistically significantly during the trial. In this study, the baseline values or study effect of the patients with apolipoprotein E phenotype E3/3 did not differ statistically significantly from those of the patients having the E4 allele (E2/4, E4/3, E4/4).

After the run-in period with the previous antihypertensive monotherapy, the mean ± SEM of supine systolic blood pressure was 142 ± 3 mm Hg, supine diastolic blood pressure, 93 ± 1 mm Hg, and supine heart rate, 69 ± 2 beats/min. All patients maintained acceptable blood pressure < 160/100 mm Hg) during the trial. After 6 months' therapy, the systolic and diastolic blood pressures were 145 ± 3 and 93 ± 1 mm Hg in the celiprolol group and 143 ± 4 and 94 ± 2 mm Hg in the control drug group, respectively. The daily celiprolol dose was increased to 400 mg for nine of the 23 patients. Celiprolol did not affect the orthostatic reaction. After 2 min standing, systolic blood pressure was in average 2 mm Hg less, diastolic blood pressure, 3 mm Hg greater, and heart rate, 5 beats/min higher than those in the supine position after a 10-min rest.

DISCUSSION

The preliminary findings that celiprolol has a favorable effect on insulin sensitivity and glucose tolerance in patients with dyslipidemia (7) are supported by this study. The conclusion is based on parallel results from two independent tests, EC and OGTT. The ISI increased during the celiprolol period, but this was less pronounced in group B than in group A. This may be because of a difference in group size and the mean of the ISI at the start, and intention-to-treat analysis. The magnitude of the change in ISI (32%) was about the same as in the preliminary study with nondiabetic hypertensive patients (37%; 7) and that found with patients with NIDDM (26%; 8). In lean healthy humans, β-receptor modulation with celiprolol was neutral in regard to insulin sensitivity and lipoprotein metabolism (15). The results of published animal studies agree with our findings. Celiprolol has been reported to reduce insulin resistance in obese Zucker rats by two groups (16,17). The magnitude of the ISI change may be slightly overestimated in our study because of a rather low serum insulin level during steady state (74 mU/L). This may not suppress hepatic glucose production completely, especially with those patients more resistant to insulin. Serum insulin level during steady state of the EC was, however, higher than the average serum insulin level in the OGTT (during 2 h after a high-glucose dose) and may thus be considered to be greater than the physiological postprandial level for these patients. Infusion rates of insulin were kept the same for each patient in all EC tests. Endogenous glucose production could have been differentiated with infusion of a glucose tracer (12), but that was not used in our study.

Vasodilation, β3-adrenoceptor agonism, and modulation of various enzymes of lipid metabolism have been proposed as mediators of the beneficial lipid effects of celiprolol (6,18-20). In our study, a tendency toward normalization of fasting serum lipids was observed, but a statistically significant effect between the celiprolol and the control group was seen only in the HDL/LDL ratio. The causal relation between chronic hyperinsulinemia and increased very low density lipoprotein (VLDL) production or hypertriglyceridemia is unclear (21,22). A decrease in serum triglycerides and accelerated reverse cholesterol transport (increased HDL/LDL ratio) induced by celiprolol may be connected to improved insulin sensitivity.

Enhanced peripheral perfusion increases insulin sensitivity (12). However, possibly improved perfusion does not fully explain the finding of increased insulin sensitivity during celiprolol treatment, because in our trial, the ISI increased slightly also among those patients whose medication was changed from a vasodilatory ACE inhibitor to celiprolol. It remains to be elucidated whether the difference lies in a more complete inhibition of the sympathetic nervous action by celiprolol than by an ACE inhibitor. The latter may cause a reflex activation of the sympathetic system (23) that powerfully inhibits the actions of insulin at both the pre- (24,25) and postreceptor levels (26).

Endothelial dysfunction has been linked to insulin resistance (27,28). Celiprolol inhibited hypertrophy of vascular intima-media in spontaneously hypertensive rats (29). The authors postulated that this may be caused by stimulation of nitric oxide (NO) synthase. In hypertensive humans, neutrophil NO synthase activity increased by >100% during an 8-week celiprolol treatment (30). Restoration of endothelial function in essential hypertension via the NO system might be one explanation for the better insulin sensitivity during celiprolol treatment. This mechanism could also explain the nonadrenergic vaso/bronchodilative action of celiprolol (31).

The increase in serum potassium level during celiprolol treatment is probably caused by decreased aldosterone secretion. β1-antagonists inhibit the renin-angiotensin-aldosterone system directly by blocking the effect of catecholamines on β-adrenoceptors located on the juxtaglomerular cells (32,33). Increased urate and decreased SHBG in serum have been proposed to be indicators of the metabolic syndrome (34,35). We observed only a modest beneficial effect of celiprolol treatment on serum urate, and no change in SHBG. Plasma fibrinogen, another proposed cardiovascular risk factor (36), has been reported to decrease during celiprolol therapy (37). Fibrinogen levels of the patients in our study remained within normal levels during the whole study.

If antihypertensive treatment deteriorates insulin sensitivity, the net effect on cardiovascular morbidity may be negative, even though arterial blood pressure decreases (3,38,39). Our study supports the preliminary finding that celiprolol has a favorable effect on insulin sensitivity. Thus it is a recommendable drug for the hypertensive patients showing signs of the metabolic syndrome.

Acknowledgment: We appreciate the expert assistance of Mrs. Leena Lahtela and Elina Kahra in the clamp tests, and Mirja Ala-Kaila in scheduling the control visits. We thank Drs. Pertti Himanen, Kristiina Hyytiäinen, Thomas Ilvonen, Jyrki Perttilä, Jussi Ripsaluoma, and Heikki Tulento for the patient recruitment. This study was supported by The Academy of Finland and Leiras Oy.

REFERENCES

1. Genest J Jr, Cohn JS. Clustering of cardiovascular risk factors: targeting high-risk individuals. Am J Cardiol 1995;76:8-20A.
2. Reaven GM, Lithell H, Landsberg L. Hypertension and associated metabolic abnormalities: the role of insulin resistance and sympathoadrenal system. N Engl J Med 1996;334:374-81.
3. Black HR. The coronary artery disease paradox: the role of hyperinsulinemia and insulin resistance and implications for therapy. J Cardiovasc Pharmacother 1990;15(suppl 5):S26-38.
4. Mediratta S, Fozailoff A, Frishman WH. Insulin resistance in systemic hypertension: pharmacotherapeutic implications. J Clin Pharmacol 1995;35:943-56.
5. Milne RJ, Buckley MMT. Celiprolol: an updated review of its pharmacodynamic and pharmacokinetic properties, and therapeutic efficacy in cardiovascular disease. Drugs 1991;41:941-69.
6. Malminiemi K, Lahtela J, Huupponen R. Effects of celiprolol on insulin sensitivity and glucose tolerance in dyslipidemic hypertension. Int J Clin Pharmacol Ther 1995;32:156-63.
7. Gelding SV, Wadsworth J, Ali K, Niththyananthan R, Henderson AD, Johnston DG. Celiprolol improves insulin sensitivity and decreases serum cholesterol levels in hypertensive NIDDM patients. Diabetologia 1996;39(suppl 1):A289.
8. JNC V. The Fifth Report of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure (JNC V). Arch Intern Med 1993;153:154-83.
9. Beaumont JL, Carlson LA, Cooper GR, Fejfar Z, Fredrickson DS, Strasser T. Classification of hyperlipidaemias and hyperlipoproteinaemias. Bull World Health Org 1970;43:891-915.
10. Boden G, Chen X, DeSantis RA, Kendrick Z. Effects of age and body fat on insulin resistance in healthy men. Diabetes Care 1993;16:728-33.
11. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol 1979;237:E214-23.
12. Yki-Järvinen H. Action of insulin on glucose metabolism in vivo. Baillieres Clin Endocrinol Metab 1993;7:903-28.
13. Gottesman I, Mandarino L, Gerich J. Use of glucose uptake and glucose clearance for the evaluation of insulin action in vivo. Diabetes 1984;33:184-91.
14. Lehtimäki T, Moilanen T, Viikari J, et al. Apolipoprotein E phenotypes in Finnish youths: a cross-sectional and 6-year follow-up study. J Lipid Res 1990;31:487-95.
15. Bohlen LM, de Courten M, Hafezi F, Shaw S, Riesen W, Weidmann P. Insulin sensitivity and atrial natriuretic factor during betareceptor modulation with celiprolol in normal subjects. J Cardiovasc Pharmacol 1994;23:877-83.
16. Santti E, Rouru J, Koulu M, et al. Effect of chronic celiprolol treatment on brown adipose tissue thermogenesis on genetically obese Zucker rats. Int J Obesity 1995;19(suppl 2):146.
17. Jacob S, Henriksen EJ, Fogt DL, Dal Ponte D, Dietze GJ. The betaadrenergic modulator celiprolol reduces insulin resistance in obese Zucker rats. Diabetologia 1996;39(suppl 1):A289.
18. Hunninghake DB. The effects of cardioselective vasodilating betablockers on lipids. Am Heart J 1991;121:1029-32.
19. Lijnen P. Biochemical mechanisms involved in the beta-blocker-induced changes in serum lipoproteins. Am Heart J 1992;124:549-56.
20. Fitzgerald JD. Do partial agonist beta-blockers have improved clinical utility? Cardiovasc Drugs Ther 1993;7:303-10.
21. Jeppesen J, Hollenbeck CB, Zhou MY, et al. Relation between insulin resistance, hyperinsulinemia, postheparin plasma lipoprotein lipase activity, and postprandial lipemia. Arterioscler Thromb Biol 1995;15:320-4.
22. Steiner G, Lewis GF. Hyperinsulinemia and triglyceride-rich lipoproteins. Diabetes 1996;45(suppl 3):S24-6.
23. Jackson EK, Garrison JC. Renin and angiotensin. In: Hardman JG, Limbird LE, Molinoff PB, et al., eds. Goodman & Gilman's the pharmacological basis of therapeutics. New York: McGraw-Hill, 1996;733-58.
24. Webber J, Macdonald IA. Metabolic actions of catecholamines in man. Baillieres Clin Endocrinol Metab 1993;7:393-413.
25. Jamerson KA, Smith SD, Amerena JV, Grant E, Julius S. Vasoconstriction with norepinephrine causes less forearm insulin resistance than a reflex sympathetic vasoconstriction. Hypertension 1994;23:1006-11.
26. Connolly CC, Steiner KE, Stevenson RW, et al. Regulation of glucose metabolism by norepinephrine in conscious dogs. Am J Physiol 1991;261:E764-72.
27. Steinberg HO, Chaker H, Leaming R, Johnson A, Brechtel G, Baron AD. Obesity/insulin resistance is associated with endothelial dysfunction: implications for the syndrome of insulin resistance. J Clin Invest 1996;97:2601-10.
28. Baron AD. The coupling of glucose metabolism and perfusion in human skeletal muscle: the potential role of endothelium-derived nitric oxide. Diabetes 1996;45(suppl 1):S105-9.
29. Tolvanen J-P, Wu X, Kähönen M, et al. Effect of celiprolol therapy on arterial dilatation in experimental hypertension. Br J Pharmacol 1996;119:1137-44.
30. Mehta JL, Lopez LM, Chen L, Cox OE. Alterations in nitric oxide synthase activity, superoxide anion generation, and platelet aggregation in systemic hypertension, and effects of celiprolol. Am J Cardiol 1994;74:901-5.
31. Van Inwegen RG, Khandwala A, Gordon RJ, Sutherland CA, Wolf P, Pruss TP. Mechanistic studies on the bronchosparing activity of celiprolol, a cardioselective beta adrenoceptor antagonist. Arch Int Pharmacodyn Ther 1986;284:127-44.
32. Keeton TK, Campbell WB. The pharmacologic alteration of renin release. Pharmacol Rev 1980;32:81-227.
33. Conway J, Cruickshank J. Beta adrenoceptor blockers and the renin-angiotensin system. In: Robertson J, Nicholls M, eds. The renin-angiotensin system. London: Gower Medical Publishing, 1993;84.1-84.9.
34. Bengtsson C, Lapidus L, Stendahl C, Waldenstrom J. Hyperuricaemia and risk of cardiovascular disease and overall death: a 12-year follow-up of participants in the population study of women in Gothenburg, Sweden. Acta Med Scand 1988;224:549-55.
35. Haffner SM, Katz MS, Stern MP, Dunn JF, Association of decreased sex hormone binding globulin and cardiovascular risk factors. Arteriosclerosis 1989;9:136-43.
36. Heinrich J, Assmann G. Fibrinogen and cardiovascular risk. J Cardiovasc Risk 1995;2:197-205.
37. Herrmann JM, Mayer EO. A long-term study of the effects of celiprolol on blood pressure and lipid-associated risk factors. Am Heart J 1988;116:1416-21.
38. Despres JP, Marette A. Relation of components of insulin resistance syndrome to coronary disease risk. Curr Opin Lipidol 1994;5:274-89.
39. Wingard DL, Barrett-Connor EL, Ferrara A. Is insulin really a heart disease risk factor? Diabetes Care 1995;18:1299-304.
Keywords:

Insulin resistance; Celiprolol; Euglycemic clamp; Cardiovascular risk factors; Crossover trial

© Lippincott-Raven Publishers