Jump to
Abstract
Essential hypertension is associated with multiple metabolic abnormalities, among them, hyperinsulinemia. This hyperinsulinemia is attributed to the presence of decreased insulin sensitivity (insulin resistance) with consequent compensatory insulin secretion. We tested the hypothesis that decreased insulin clearance is present in hypertensive subjects and contributes to hyperinsulinemia independently of the degree of insulin resistance. Seventy-five subjects were studied (48 hypertensive and 27 normotensive). Both groups were comparable in terms of age, body fat content, waist-to-hip ratio, and sex distribution. A primed continuous insulin infusion at 40 mU/m2 per minute was performed. Glucose was maintained at baseline levels with the euglycemic clamp technique. Hypertensive subjects were characterized by decreased insulin sensitivity (insulin-mediated glucose uptake: 5.14±0.28 versus 7.26±0.61 mg glucose/kg fat-free mass per minute, hypertensive versus normotensive, P=.002), increased insulin levels during the insulin infusions (804±36 versus 510±38 pmol/L, hypertensive versus normotensive, P<.001), and decreased insulin metabolic clearance rate (328±15 versus 521±30 mL/min per meter squared, hypertensive versus normotensive, P<.001). In an ANCOVA (including sex, degree of obesity, waist-to-hip ratio, and insulin sensitivity as covariates) the differences in insulin metabolic clearance rate between normotensive and hypertensive subjects remained highly significant (P<.001). Insulin metabolic clearance rate was significantly associated with fasting insulin levels. We conclude that essential hypertension is independently associated with decreased insulin metabolic clearance rate in addition to insulin resistance. A low insulin metabolic clearance rate may be a contributory factor to the hyperinsulinemia observed in essential hypertension.
Essential hypertension greatly contributes to the development of atherosclerosis.1 2 The contribution of essential hypertension to the development of coronary artery disease is thought to be due to the direct effect of elevated blood pressure on the arterial wall as well as the indirect effects of multiple metabolic abnormalities frequently present in hypertensive subjects. Epidemiological studies have identified an association between hyperinsulinemia (elevated fasting and postprandial insulin levels) and hypertension. This correlation is at least partly independent of the degree of obesity.3 4 Elevated insulin levels may stimulate the proliferation of the smooth muscle cells in the arterial wall5 and are also associated with atherogenic lipid patterns6 and increased sympathetic nervous system activity.7 Population-based studies suggest that hyperinsulinemia is an independent risk factor for the development of coronary artery disease.8 9
The etiology of hyperinsulinemia in essential hypertension is thought to be the consequence of a compensatory increase in pancreatic β cell secretion as a response to decreased peripheral (ie, muscular) insulin-mediated glucose utilization (insulin resistance).10 In vivo insulin action has been reported to be decreased in hypertensive subjects.11 12 13 14 15 16 An alternative (or concurrent) explanation for hyperinsulinemia is the presence of reduced insulin clearance, which would result in insulin remaining in the circulation for a longer period of time and thus higher plasma insulin levels. Recent reports evaluating insulin clearance in hypertension have provided controversial results, and none has examined the relationship between insulin clearance and insulin sensitivity.17 18 19 In the present study, we tested the hypothesis that essential hypertension is associated with decreased insulin clearance in addition to and independently of insulin resistance.
Methods
Selection of Study Subjects
A total of seventy-five white volunteers (48 hypertensive and 27 normotensive) were studied. Investigators explained the purpose and methods of the study, and all participants gave their informed, written consent. The study was approved by the Institutional Review Board for Human Studies of the University of Texas Southwestern Medical Center at Dallas. The inclusion criteria for the hypertensive subjects were as follows: (1) presence of chronic hypertension with an average sitting diastolic pressure between 95 and 114 mm Hg on two separate readings 1 week apart 3 weeks after discontinuation of all antihypertensive medications; (2) age 21 years or older; (3) weight less than 40% over ideal body weight; (4) absence of angina pectoris, previous myocardial infarction, cardiac failure, arrhythmia, and renal dysfunction; (5) absence of the use of other concomitant medication that could affect insulin sensitivity; and (6) fasting glucose level less than 110 mg/dL. The normotensive subjects were all 21 years of age or older, less than 40% over ideal body weight, and free of any significant disease and took no concomitant medications. Systolic pressure levels were less than 140 mm Hg and diastolic pressure levels less than 85 mm Hg. All subjects had an initial laboratory screening panel that included a complete blood count and a chemistry panel that included determination of liver function and fasting glucose and lipid, electrolyte, serum creatinine, and uric acid levels.
Experimental Protocol
The study participants were admitted to the General Clinical Research Center at Parkland Memorial Hospital the day before the studies, where they were assured of a nonstressful environment. Duplicate samples for fasting insulin and glucose levels were obtained on the day of admission. On the following morning, an insulin infusion study was performed as follows: A small plastic catheter was placed in a hand vein in a retrograde fashion to obtain blood samples. This hand was placed in a warm box at 65°C to obtain arterialized venous blood. A long, thin plastic catheter was placed in an antegrade fashion in an antecubital vein for infusion of insulin and 20% dextrose (17.7% glucose). Plasma insulin level was raised acutely, and a continuous infusion of insulin was maintained at 40 mU per meter squared of body surface area per minute during 2 hours. Dextrose was infused at a variable rate according to the results of plasma glucose levels measured at 5-minute intervals on samples with the use of an enzymatic glucose analyzer. Glucose infusion rates were determined by the protocol of DeFronzo et al.20 With this methodology, plasma glucose levels were maintained at basal levels with less than 5% variation from this level. Infusion pumps (model 2205, Harvard Apparatus) were used during the procedure. Insulin levels were measured in blood samples obtained every 10 minutes during the procedure by a sensitive radioimmunoassay (Diagnostic Products Corp), with intra-assay and interassay coefficients of variation of 2% to 5% and 5% to 10%, respectively. Cross-reactivity with proinsulin with this assay is 40%. We do not expect this cross-reactivity to significantly interfere with the calculation of insulin clearance because during hyperinsulinemic insulin infusions, the contribution of endogenously secreted insulin and proinsulin to the total insulin measured in the assay is in the range of 5%. The mean insulin level between minutes 90 and 120 was defined as the steady-state insulin level. Insulin-mediated glucose uptake (M value) was calculated with the use of the glucose infusion rates during minutes 90 to 120 of the clamp and was expressed in milligrams glucose per kilogram fat-free mass per minute. During insulin infusions, at the levels used in this study, hepatic glucose output has previously been shown to be completely suppressed in nondiabetic subjects.11 C peptide levels were also measured at time 0 and at minutes 90 and 120 during the clamps with a sensitive radioimmunoassay.
Body density was obtained in all subjects using underwater weight. Percentage of body fat was derived from body density, and fat-free mass was calculated by subtracting fat mass from total body weight. Waist-to-hip ratio was calculated by dividing the minimum circumference at the level of the umbilicus and the maximum circumference at the level of the gluteus.
Forty-six of the study subjects underwent two studies at identical insulin infusion rates (40 mU/m2 per minute) as part of a different protocol. The intrasubject variability in steady-state insulin levels during these two insulin infusions was 13%.
The metabolic clearance rate of insulin (MCRi) was calculated as previously published21 by dividing the insulin infusion rate by the steady-state insulin concentration, which is obtained by subtracting the estimated residual insulin secretion from the mean insulin concentration between minutes 90 and 120 of the study. The estimated residual insulin secretion is the product of the basal (preinfusion) insulin level multiplied by the ratio of C peptide at baseline to C peptide at minutes 90 to 120 of the insulin infusion. Thus, the formula for MCRi is as follows21 : MCRi=Insulin Infusion Rate/{[Insulin90-120]−[Insulin0]×(C Peptide90-120/C Peptide0)}.
Statistical Analysis
Statistical analysis was performed with BMDP software (BMDP Statistical Software). Results are expressed as mean±SE. Blood pressure was treated as a categorical variable (hypertensive, normotensive). The means of MCRi, M value, steady-state insulin level, and fasting insulin level were compared between the normotensive and hypertensive groups with a two-sample Student's t test. To evaluate the potential influences of obesity, sex, and insulin sensitivity on MCRi, we used ANCOVA and multiple regression models to analyze MCRi according to hypertension status with sex, percentage of body fat, M value, and waist-to-hip ratio as covariates. We also assessed the effect of MCRi, adiposity, M value, sex, and hypertension on fasting insulin levels using multiple regression analysis. Variables for the regression models were selected to minimize multicolinearity. Subsequently, stepwise regression was performed. A value of P<.05 was considered statistically significant.
Results
The baseline characteristics of the study subjects are presented in Table 1⇓. Normotensive and hypertensive subjects (both men and women) were comparable in terms of age, adiposity, waist-to-hip ratio, and fasting glucose levels. Fasting insulin levels were slightly higher in the hypertensive group, but no statistically significant difference was found. During the insulin infusion studies, plasma glucose and insulin levels were stable in all study subjects. The mean glucose level was 6.1 mmol/L (110 mg/dL), with a coefficient of variation of 3%. The mean insulin level of all the studies was 714±36 pmol/L (119±6 μU/mL). Hypertensive individuals had decreased insulin sensitivity as evidenced by a low M value compared with normotensive individuals (5.14±0.28 versus 7.26±0.61 mg/kg fat-free mass per minute, hypertensive versus normotensive, P=.002). Insulin levels obtained during the infusions were significantly higher in the hypertensive group despite a lower M value (804±36 [134±6] versus 510±38 pmol/L [85±6 μU/mL], hypertensive versus normotensive, P<.001). MCRi was lower in the hypertensive group (328±15 versus 521±30 mL/min per meter squared, hypertensive versus normotensive, P<.001) (Figure⇓).
Insulin-mediated glucose uptake (IMGU) (a), steady-state insulin (SSI) levels (b), and insulin metabolic clearance rate (MCR i) (c) in normotensive (Normo) and hypertensive (Hyper) subjects during insulin infusion studies. Hypertensive subjects were insulin resistant (a) (*P=.002 vs normotensive) and developed higher insulin levels during insulin infusion studies (b) (+P<.001). Insulin metabolic clearance rate was significantly lower in the hypertensive group (**P<.001). FFM indicates fat-free mass.
Characteristics of Study Subjects
ANCOVA showed that the lower MCRi observed in the hypertensive group was independent of sex, degree of adiposity, waist-to-hip ratio, and insulin sensitivity as measured by the M value. The differences remained statistically significant after adjustment for these variables (P<.001). In regression analysis, a model including hypertension status, sex, percentage of body fat, and M value explained 40% of the variability in MCRi levels; of these variables, hypertension status was the most significant predictor of MCRi (Table⇑s 2 and 3).⇓⇓ Sex and M value were not independent predictors of MCRi. Adiposity as percentage of body fat was of borderline significance.
Multiple Regression Analysis: Contribution to Variability in MCRi (Dependent Variable, MCRi)
Multiple Regression Model (Dependent Variable, MCRi)
MCRi, percentage of fat, hypertension status, and sex together explained 50% of the variability in fasting insulin levels. The variable with the greatest independent contribution to the model was MCRi (Table⇑s 4 and 5).⇓⇓ The M value was not an independent predictor of fasting insulin levels.
Multiple Regression Analysis: Contribution to Variability in Fasting Plasma Insulin (Dependent Variable, Fasting Plasma Insulin)
Multiple Regression Model (Dependent Variable, Fasting Plasma Insulin)
Discussion
This study clearly shows marked metabolic differences between hypertensive and normotensive individuals. The two main variables studied (insulin clearance and insulin sensitivity) varied markedly and independently in the two groups. Since both groups were comparable in terms of potential confounding factors, such as age, adiposity, sex distribution, and waist-to-hip ratio, the differences observed are likely to be caused by the hypertension status per se. Furthermore, adjustment for confounding variables with ANCOVA indicated that hypertension was independently associated with low MCRi. Since the majority of the study subjects (hypertensive and normotensive) were mildly to moderately obese, it is not prudent to generalize these results to nonobese populations.
The mechanism of the reduced MCRi in hypertension is not known, and the hypotheses presented here are merely speculative. Insulin clearance is a complex phenomenon21 and depends on the distribution22 and elimination of insulin. Several studies have shown that insulin elimination occurs mainly through metabolic degradation,23 which involves mainly the liver and kidney and is thought to be largely receptor mediated. Endogenously secreted insulin suffers from first pass metabolism in the liver, and it is estimated that 50% of secreted insulin is cleared during this step. Peripherally administered insulin is distributed between the intravascular and interstitial fluid spaces before it reaches its receptors in insulin-sensitive tissues. The transcapillary transport of insulin may be rate limiting, and a lag of approximately 15 to 30 minutes exists between concentrations achieved in the intravascular compartment and interstitial fluid.24 Impaired capillary transport (caused by either decreased flow or decreased permeability) could conceivably result in higher plasma insulin concentrations caused by decreased efflux of insulin from the intravascular space. Hypertension has been associated with capillary thinning in several vascular beds.25 26 If this phenomenon occurs in the tissues responsible for insulin degradation, it could explain the observation of decreased insulin clearance. Capillary permeability is increased in hypertensive individuals.27 This would accelerate (rather than slow) the disappearance of insulin from the intravascular space.
Impaired receptor function could also affect insulin clearance. Studies by Nijs et al28 in normotensive subjects indicate that a good correlation exists between insulin clearance and in vivo insulin action. In the present study, we found only a weak correlation between the M value (a reflection of insulin action) and MCRi in the hypertensive and normotensive groups. The reasons for the differences between the study of Nijs et al and the present results are not clear. Since the defect in in vivo insulin action in hypertension is thought to reflect a postreceptor rather than receptor-binding abnormality, the absence of a relationship between insulin clearance and insulin action is not surprising.
The existence of an association between MCRi and fasting insulin levels suggests that insulin clearance is an important determinant of basal insulin levels. Hyperinsulinemia in the basal state has been described in several epidemiological studies in hypertensive groups.8 9 Decreased insulin clearance may play a role in this abnormality. In the present study, however, the fasting insulin level was only marginally elevated in hypertensive subjects.
In summary, essential hypertension is associated with a markedly diminished insulin clearance. This abnormality is independent of the degree of insulin sensitivity and may contribute independently to the hyperinsulinemia observed in essential hypertension. The pathogenetic roles of decreased insulin clearance in the development or maintenance of the hypertensive state and in the development of coronary artery disease in hypertensive subjects need to be explored further.
Acknowledgments
This work was supported in part by National Institutes of Health Grant M01RR00633 to the General Clinical Research Center.
Footnotes
-
A preliminary report was presented at the 1995 Clinical Research Meeting, San Diego, Calif, May 18-21, 1995.
- Received June 28, 1996.
- Revision received July 26, 1996.
- Revision received August 16, 1996.
References
This Issue
Jump to
Article Tools
Share this Article
- Essential Hypertension Is Associated With Decreased Insulin Clearance and Insulin Resistance
