| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Division of Endocrinology and Metabolic Diseases, University of Verona Medical School, Verona, Italy
Address all correspondence and requests for reprints to: Prof. Enzo Bonora, Endocrinologia e Malattie del Metabolismo, Ospedale Civile Maggiore, Piazzale Stefani 1, 37126, Verona, Italy. E-mail: enbonor{at}tin.it
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We quantitated the rates of total glucose disposal, glycogen synthesis
(GS), glycolysis, glucose oxidation, endogenous glucose production, and
LOX in the basal state and during a 4-h euglycemic (
5
mM) hyperinsulinemic (300 pM) clamp carried
out in combination with a dual-tracer infusion ([3H]-3-
and [14C]-U-D-glucose) and indirect
calorimetry in 42 nonobese noninsulin-treated type 2 diabetic subjects
(22 hypertensive and 20 normotensive) and 23 nonobese nondiabetic
subjects (9 without and 14 with essential hypertension).
Compared with normotensive controls, both groups of diabetic subjects were markedly insulin resistant. In the basal state, all glucose fluxes were similar in diabetic subjects with or without hypertension. During insulin infusion, total glucose disposal was significantly reduced in hypertensive diabetic subjects, compared with their normotensive counterparts (18.7 ± 1.0 vs. 28.6 ± 3.0 µmol/min·kg lean body mass; P < 0.01). This difference was almost entirely explained by a marked reduction in GS (4.5 ± 2.0 vs. 12.5 ± 3.3 µmol/min·kg lean body mass; P < 0.01). Endogenous glucose production was not different in the two diabetic groups during insulin infusion and was significantly higher than in normotensive controls. Lipid oxidation was less suppressed by hyperinsulinemia in hypertensive than in normotensive diabetic subjects (1.46 ± 0.1 vs. 0.91 ± 0.1 µmol/min·kg lean body mass; P < 0.01). Glucose fluxes were not significantly different in nondiabetic subjects with essential hypertension and in normotensive diabetic individuals.
These results indicate that hypertension markedly aggravates insulin resistance featuring type 2 diabetes mellitus. The molecular defects underlying this phenomenon involve primarily GS.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Insulin resistance is known to substantially contribute to the development of hyperglycemia in type 2 diabetes (11), whereas its contribution to the genesis of hypertension is still under dispute (12). However, up to 80% of subjects with type 2 diabetes, most of whom are insulin resistant, have high blood pressure (13). For this reason, several authors speculated that insulin resistance is an underlying defect contributing to hypertension in type 2 diabetes (14, 15, 16, 17). If this hypothesis is correct, one should find that hypertensive type 2 diabetic subjects are more insulin resistant than normotensive diabetic individuals. Therefore, it is of interest to examine whether the combination of hypertension and type 2 diabetes mellitus confers a degree of insulin resistance that is more severe than that observed in subjects with diabetes but normal blood pressure. Moreover, if hypertensive diabetic subjects were actually more severely insulin resistant, it would be interesting to understand which biochemical pathways are involved.
Only few studies measured accurately glucose metabolism by an isotopic tracer dilution technique and compared it in type 2 diabetic patients with and without hypertension (18, 19, 20, 21). However, in none of these studies the multiple pathways of intracellular glucose metabolism were explored.
In the present study, we employed a dual-tracer technique, associated with the euglycemic hyperinsulinemic clamp and indirect calorimetry, to quantitate the intracellular fate of plasma glucose in the basal state and during insulin infusion in type 2 diabetic subjects with or without hypertension. The aim was to address the following questions: 1) Does hypertension worsen insulin resistance of type 2 diabetes? 2) Do normotensive diabetic subjects have a normal insulin sensitivity? 3) If hypertension aggravates insulin resistance of type 2 diabetes, which tissues and/or which biochemical pathways are involved?
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The study sample consisted of 42 nonobese, noninsulin-treated patients with type 2 diabetes mellitus regularly attending the Diabetes Clinic of Verona, most of them recruited over a period of 6 months to participate in a clinical trial on the effects of antihypertensive treatment on insulin sensitivity (22). Inclusion criteria were the following: age 30 to 70 yr; body mass index (BMI) <30 kg/m2; no treatment with insulin; no recent acute illness or clinical evidence suggestive of kidney, liver, or endocrine diseases; and no severe chronic diabetic complications (proliferative retinopathy, macroalbuminuria, symptomatic neuropathy, clinically manifest coronary heart disease, or other vascular diseases). Nine patients were treated with diet only, and the remaining 33 with diet and oral hypoglycemic agents (sulfonylurea alone, n = 14; sulfonylurea plus metformin, n = 19). Hypoglycemic medications were withheld on the morning of the study.
Twenty diabetic patients had normal blood pressure, and 22 had hypertension according to conventional criteria (23). Seven hypertensive subjects were treated with antihypertensive drugs (four with angiotensin-converting enzyme inhibitors and three with calcium channel blockers). None of the subjects was taking ß-blockers or diuretics. In all subjects the antihypertensive treatment was discontinued at least 4 weeks before the study (22). None of the subjects was taking other medications known to interfere with glucose metabolism.
Nine healthy nondiabetic normotensive subjects and 14 nondiabetic hypertensive subjects of the same age range were recruited by a community advertisement and served as control groups. In these subjects, glucose tolerance was confirmed to be normal by a standard oral glucose tolerance test. In hypertensive controls, the antihypertensive medications, if any, were discontinued at least 4 weeks before the study.
The purpose and potential risks of the study were explained to all subjects, and their informed, written, voluntary consent was obtained before their participation. The protocol was reviewed and approved by the Ethical Committee of the University of Verona Medical School.
Blood pressure
Blood pressure was measured to the nearest millimeter with a mercury sphygmomanometer in the right arm after at least 10 min of rest in the sitting position. Systolic and diastolic blood pressures were defined as Korotkoff phases I and V. The mean of three measurements taken at 2-min intervals were averaged and used for the analysis.
Anthropometric measurements
Height (to the nearest 0.5 cm) and weight (to the nearest 0.5 kg) were recorded while subjects were wearing only underwear garments. BMI was calculated as weight/height (2). Waist circumference (widest between the lower rib margin and the iliac crest) was measured in duplicate, and the values were averaged for statistical purposes. Higher values of waist circumference indicate a predominantly central fat distribution (24).
Based on the principle that resistance to a mild electrical current is related to total body water and that the latter is highly correlated to fat-free mass, a tetrapolar impedance analyzer (BIA-103, Akern, Florence, Italy) was used to measure body electrical resistance and to derive an estimate of total body water, fat-free mass, and body fat (25). The measures achieved with this technique are highly correlated with those generated by more sophisticated methods, including isotopic tracer dilution technique (26).
Metabolic studies
In the morning, after an overnight fast, a 20-gauge Teflon catheter was inserted into an antecubital vein for the infusion of all test substances. A second catheter was inserted retrogradely into a wrist vein of the contralateral hand for blood sampling, and the hand was inserted into a hot (60 C) box to achieve the arterialization of venous blood. Both catheters were kept patent with the infusion of a normal saline solution.
Glucose fluxes in the basal state and during insulin infusion were measured by the combination of euglycemic hyperinsulinemic clamp, radioisotopic technique, and indirect calorimetry (27).
Glucose fluxes in the basal state. At 0800–0830 h
(time = -150 min) a prime-constant infusion of
[3H]-3-D-glucose (0.15 µCi/min)
and a prime-constant infusion of
[14C]-U-D-glucose (0.075 µ
Ci/min) (NEN Life Science Products, Boston, MA) were
started and continued for 150 min. The prime of
[3H]-3-D-glucose and
[14C]-U-D-glucose was administered
together with a bolus injection of
NaH14CO3 (8 µCi). The
prime for tracer glucose was calculated dividing the glucose pool
(plasma glucose concentration times glucose distribution volume,
assumed to be 25% of body weight) by the estimated basal glucose
turnover (11 µmol/min·kg lean body mass) when fasting plasma
glucose was below 11.1 mmol/L, and 14 µmol/min·kg lean body mass
when it was >11.1 mmol/L and then multiplying the result by the tracer
infusion rate.
Glucose fluxes during insulin clamp. At the end of the basal
state (time = 0) the dual glucose tracer infusion was discontinued
and a euglycemic hyperinsulinemic (20 mU/min·square meter of surface
area) clamp was started and continued for the subsequent 240 min.
Plasma glucose was left to drop until euglycemia (5 mmol/L) was
reached and then maintained at that level by a glucose infusion
adjusted every 5–10 min according to the changes in plasma glucose.
Two hours after the beginning of the insulin clamp (time = +120
min) the prime constant infusions of
[3H]-3-D-glucose and
[14C]-U-D-glucose were resumed at
the rates of 0.30 and 0.15 µCi/min, respectively, and continued until
the end of the study. A second bolus of
NaH14CO3 was administered
(8 µCi) with the prime. The prime dose of labeled glucose was
calculated dividing the glucose pool by the product of 1.1 by
GIR100–120 and then multiplying the result by
the tracer infusion rate. GIR100–120 was the
glucose infusion rate during the time interval 100–120 min of the
glucose clamp. It was multiplied by 1.1 to take into account the
expected 10% average increase in glucose infusion from 100–120 min to
180–240 min (10, 21, 22, 27, 28). The rationale for
discontinuing the tracer administration during the first 2 h of
the insulin clamp was that, by resuming the tracer infusion in a
near-steady-state, the time of equilibration for labeled precursor
(glucose) and products (water and CO2) are
significantly shortened. Indeed, with this methodological approach a
good steady-state specific activity was obtained during the last 60 min
of the clamp (CV < 10%) (10, 21, 22, 27, 28).
Expired air samples for the determination of CO2-specific activity were collected in the last hour of the basal period and the insulin clamp period. At the same time, blood was withdrawn for the determination of plasma [3H]-3-D-glucose and [14C]-U-D-glucose-specific activities and plasma 3H2O and insulin and FFA concentrations. Blood was collected in heparinized tubes and promptly centrifuged and the plasma decanted and stored at -20 C until analyzed. Expired air was bubbled through a CO2-trapping solution (hyamine hydroxide/absolute ethanol/0.1% phenolphtalein; 3:5:1). The solution was titrated with 1 N HCl to trap 1 mM CO2 per 3 mL of solution. Then 0.5 mL of the saturated solution was added to 5 mL of scintillation liquid and 14C radioactivity was measured using a ß-scintillation counter (Beckman Coulter, Inc., Fullerton, CA).
Computerized open-circuit continuous indirect calorimetry with a canopy system (Deltatrac, Sensormedics, Anaheim, CA) was employed for the determination of gaseous exchanges in the last hour of the basal period and the insulin clamp period. Data were used to compute the rates of O2 consumption and CO2 production. Protein oxidation was calculated from the urinary nitrogen excretion measured before and during the insulin clamp (29). The rate of lipid oxidation (LOX) in the last hour of the basal state and insulin clamp were measured according to standard equations (29).
Analytical determinations
Plasma total cholesterol, HDL cholesterol, and triglycerides were assessed by standard enzymatic methods. Plasma LDL cholesterol was calculated with the equation of Friedewald et al. (30), except when triglycerides exceeded 5 mmol/L. Glycated hemoglobin was measured by HPLC. Albumin excretion rate was measured on a timed (24 h) urine collection by immunonephelometric method. Patients were classified as normoalbuminuric (albumin excretion rate <30 mg/day) or microalbuminuric (30–300 mg/day). None was macroalbuminuric. Plasma glucose was measured by the glucose-oxidase method on a glucose analyzer (Beckman Coulter, Inc., Fullerton, CA). Insulin was measured by double-antibody RIA, with an antibody not cross-reacting with proinsulin (Linco Research, Inc., St. Louis, MO). Plasma FFA concentration was determined by an enzymatic spectrophotometric method. Urinary nitrogen concentration, on samples collected during the basal and insulin-stimulated periods, was measured by the method of Kjeldhal (31).
[3H]-3-D-glucose, [14C]-U-D-glucose, and 3H2O-specific activities in the plasma and 14CO2-specific activity in the expired air were determined as described in detail elsewhere (27).
Calculations
Total glucose disposal (TGD), endogenous glucose production (EGP), glycolysis (GLYC), glycogen synthesis (GS), and glucose oxidation (GOX) were computed as previously reported (27). Briefly, TGD was calculated by dividing the [3H]-3-D-glucose infusion rate by the steady-state [3H]-3-D-glucose-specific activity. EGP in the basal state was equal to TGD, while during insulin clamp it was computed as the difference between TGD and exogenous glucose infused to maintain euglycemia. GLYC was calculated by dividing the 3H2O production rate by the plasma [3H]-3-D-glucose-specific activity. GS was computed as the difference between TGD and GLYC. GOX was calculated from the 14CO2 production rate divided by the steady-state plasma [14C]-U-D-glucose-specific activity.
All fluxes were expressed in µmol/min·kg lean body mass.
Statistical analysis
All data are presented as means ± SE
throughout the paper. The following statistical tests were carried out
following standard procedures (32): unpaired Student
t test, one-way ANOVA, analysis of covariance,
2 test (for categorical variables), Pearson’s
correlation, and multiple regression analysis. Because nonparametric
statistical tests (i.e., Mann-Whitney U test,
Kruskal Wallis rank test, Spearman’s rank correlation) yielded very
similar results to parametric tests, only the results achieved by the
latter ones are presented. Data were logarithmically transformed, when
appropriate, and then back-transformed to their natural units for
presentation in tables. P values less than 0.05 were
considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. The four groups were comparable for
sex, age, BMI, and percent body fat. Waist circumference was slightly
but significantly higher in hypertensive subjects. The two groups of
diabetic subjects were similar for plasma lipids, prevalence of
microalbuminuria, diabetes duration, degree of metabolic control, and
treatment.
|
).
|
, glucose fluxes and
LOX in the basal state were not significantly different in diabetic and
nondiabetic subjects.
|
). Also, hypertensive nondiabetic subjects had a lower-than-normal
TGD. From the quantitative standpoint, GS was the more severely
impaired biochemical pathway in the two diabetic groups and in subjects
with essential hypertension. GOX was significantly lower than normal
only in hypertensive diabetic patients. The inhibitory effect of
insulin on EGP also was significantly impaired in the two diabetic
groups. An impaired suppression of LOX was found only in hypertensive
diabetic subjects.
When compared with their normotensive counterparts, diabetic subjects
with hypertension showed a significantly lower TGD during insulin
clamp, which was almost entirely explained by a reduction in GS (Table 3). GOX was moderately lower (P < 0.05) in
hypertensive vs. normotensive diabetic patients. On the
contrary, GLYC was quite similar in normotensive and hypertensive
diabetic subjects. The inhibitory effect of insulin on EGP also was
similar in the two groups of diabetic subjects, while suppression of
LOX was blunted in hypertensive diabetic subjects (Table 3). The
results were virtually unchanged after adjustment (analysis of
covariance) for the small differences in age, BMI, percent body fat,
and waist circumference. Also, the adjustment for hemoglobin
A1c did not change the differences between
normotensive and hypertensive diabetic subjects in TGD and GS during
clamp (data not shown). Only the difference in GOX during
hyperinsulinemia between normotensive and hypertensive diabetic
subjects was lost after adjustment for the above-mentioned covariates
(data not shown).
As compared with nondiabetic hypertensive subjects, type 2 diabetic patients with normal blood pressure had a slightly but not significantly lower TGD.
In diabetic subjects, TGD was inversely related to both systolic
(r = -0.47, P < 0.01) and diastolic blood
pressure (r = -0.44, P < 0.01) (Fig. 1). In a multiple linear regression
analysis including also sex, age, percent body fat (or BMI), and waist
circumference, only the presence of hypertension and the degree of
metabolic control (hemoglobin A1c) were
independent predictors of TGD during clamp in diabetic subjects (Table 4).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Previous studies have documented that skeletal muscle is the primary site of insulin resistance in essential hypertension (6, 7) and that impaired nonoxidative glucose disposal, which reflects GS, is the intracellular metabolic pathway primarily responsible for insulin resistance in patients with essential hypertension (8). In these subjects, GOX is generally normal (4, 8). These findings (i.e., normal GOX with impaired nonoxidative glucose disposal) suggest that the mechanism(s) responsible for insulin resistance in essential hypertension may be distal to the insulin receptor and the glucose transport/phosphorylation system and may specifically impair a single intracellular pathway of glucose metabolism. Conversely, in type 2 diabetic individuals, the impairment in glucose metabolism involves both nonoxidative and oxidative pathways of glucose metabolism (10). Accordingly, defects in early steps of glucose metabolism (i.e., glucose transport and glucose phosphorylation) were documented (33, 34). Thus, the biochemical and molecular features of insulin resistance might be different in essential hypertension and type 2 diabetes mellitus. Therefore, it is of interest to examine whether hypertension confers some added features to the insulin resistance of type 2 diabetes mellitus.
In this study, we used a 4-h euglycemic hyperinsulinemic clamp technique in combination with indirect calorimetry and a dual glucose tracer infusion to compare basal and insulin-mediated glucose metabolism in nonobese type 2 diabetic individuals with or without hypertension. This approach dissects out accurately the metabolic fates of intracellular glucose (27).
Compared with healthy normotensive nondiabetic subjects (controls), both groups of diabetic patients were markedly insulin resistant. Insulin-mediated whole-body glucose disposal was reduced by >50% in diabetic subjects when compared with controls, and this was primarily accounted for by a decrease in GS. Insulin-mediated suppression of both EGP and LOX also was significantly impaired in diabetic subjects. When hypertension was associated with diabetes, the defect in insulin-mediated glucose disposal was more severe than that observed in normotensive diabetic subjects, and this was accounted for primarily by GS, which was about 80% lower in hypertensive than in normotensive diabetic subjects. GOX was involved in this phenomenon to a lesser extent (reduction of about 30% in hypertensive vs. normotensive diabetic subjects). The impaired suppression of EGP observed in normotensive diabetic subjects was not aggravated by the concomitance of hypertension, whereas LOX rate was suppressed to a lesser extent in hypertensive than in normotensive diabetic subjects.
Subjects with essential hypertension but without diabetes showed an impairment of TGD similar to that found in normotensive diabetic individuals. In these subjects, GS was the only defective biochemical pathway.
From the general standpoint, the present results agree with those of previous studies (18, 19, 20), suggesting the presence of more severe degree of insulin resistance in hypertensive, compared with normotensive nonobese type 2 diabetic subjects. In two of these studies (18, 19), however, only whole-body glucose disposal was measured, with no evaluation of the intracellular partition of plasma glucose disposal. In one study indirect calorimetry (20) showed a predominant defect in nonoxidative glucose disposal, suggesting an impairment in GS. No study was devoted to examine intracellular glucose disposal in detail. Moreover, in one study the number of normotensive nonobese diabetic subjects, who displayed an unusually high rate of insulin-mediated glucose disposal, was small (n = 6), and the antihypertensive treatment was not discontinued until the day preceding the study (19). In another study, hypertensive and normotensive diabetic subjects were not well matched for age, BMI, diabetes duration, and treatment (18). For these reasons, the results of these two studies should be interpreted with some degree of caution.
The present results look at variance with those we reported a few years ago (21). However, in that report we studied a group of type 2 diabetic patients, most of whom were obese and of Mexican-American ancestry. It is conceivable that in type 2 diabetic subjects with obesity, the defect in nonoxidative glucose disposal (GS) is almost maximal and that hypertension adds little, if any, to it. Moreover, differences in ethnicity might contribute to this apparent discordance. The overall evidence, from this and other studies (18, 19, 20), therefore, suggests that hypertension, at least in nonobese Caucasian subjects with type 2 diabetes, exerts an additive effect on insulin resistance. As to normotensive type 2 diabetic subjects, their insulin resistance is only slightly more severe than that found in essential hypertension and is confined essentially to GS.
This study gauged the metabolic pathways but could not define the basic cellular mechanisms responsible for the more severe insulin resistance associated with hypertension in type 2 diabetes mellitus. Nevertheless, two hypotheses might be formulated: 1) insulin resistance of essential hypertension and type 2 diabetes mellitus rests on distinct molecular defects that are not mutually exclusive but, indeed, additive when the two clinical conditions cluster in the same individual; and 2) essential hypertension and type 2 diabetes share a common molecular defect responsible for insulin resistance that, when maximally expressed, leads to increased blood pressure. The latter hypothesis would imply that insulin resistance could alter the physiological functions involved in the control of blood pressure (e.g., transmembrane traffic of cations and endothelial function) but that these functions are disturbed only when insulin resistance is particularly severe. The inverse relationship we found between blood pressure and TGD in diabetic subjects supports this hypothesis (i.e., blood pressure actually increases only when insulin resistance is particularly severe). However, if this were the case, the degree of insulin resistance should be similar in nondiabetic hypertensive and hypertensive diabetic subjects and should be more severe in these individuals than in normotensive diabetic patients. On the contrary, our present results show that insulin resistance in nondiabetic hypertensive subjects is less pronounced than in hypertensive diabetic patients and is quite similar to that measured in normotensive type 2 diabetic patients.
A third hypothesis is that high blood pressure per se, and not a molecular defect typical of hypertensive subjects, could promote insulin resistance, so that hypertensive diabetic patients are more insulin resistant than normotensive diabetic individuals. High blood pressure can impair the endothelium (35) as well as change the capillarization of skeletal muscle (36). Both phenomena were implicated in a lower-than-normal delivery of glucose/insulin to insulin-sensitive cells, leading to an impaired action of the hormone on glucose utilization (37). If this were the case, lowering blood pressure should result in increased insulin sensitivity. Unfortunately, this is true only for few antihypertensive medications (22, 38). Thus, this third hypothesis may be untenable. So the most likely explanation is that distinct molecular defects responsible for insulin resistance coexist in type 2 diabetic subjects with hypertension.
Although a further impairment in the glycogen synthetic pathway was observed in diabetic subjects with hypertension, these results cannot be used to prove the existence of a primary defect in glycogen synthase or in any molecular process regulating glycogen synthase enzymatic complex. A more proximal defect in insulin-mediated glucose disposal (i.e., insulin receptor signal transduction, glucose transport, or glucose phosphorylation) might account for the defect in glycogen formation simply because this is the main biochemical pathway in which glucose is channeled during hyperinsulinemia and/or because the defect in GOX is virtually maximal.
The relationships among hypertension, type 2 diabetes mellitus, and insulin resistance have important therapeutic implications for the treatment of hypertensive diabetic individuals, because many antihypertensive agents (e.g. diuretics, ß-blockers) have potentially adverse effects on glucose metabolism (38). However, it is still controversial how these considerations should affect the choice of antihypertensive treatment in diabetic patients.
In conclusion, this study shows that in nonobese type 2 diabetic subjects, the presence of hypertension is associated with a more severe degree of insulin resistance, which is mainly the consequence of a more pronounced defect in GS. Nevertheless, type 2 diabetic patients are insulin resistant even when blood pressure is not elevated.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received April 17, 2000.
Revised January 8, 2001.
Accepted January 12, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  E. Bonora, G. Formentini, F. Calcaterra, S. Lombardi, F. Marini, L. Zenari, F. Saggiani, M. Poli, S. Perbellini, A. Raffaelli, et al. HOMA-Estimated Insulin Resistance Is an Independent Predictor of Cardiovascular Disease in Type 2 Diabetic Subjects: Prospective data from the Verona Diabetes Complications Study Diabetes Care, July 1, 2002; 25(7): 1135 - 1141. [Abstract] [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 |
