Insulin Resistance in Essential Hypertension Is Related to Plasma Renin Activity

Home , Plasma renin activity, Renin

Journal of Human Hypertension (1998) 12, 379–382  1998 Stockton Press. All rights reserved 0950-9240/98 $12.00 http://www.stockton-press.co.uk/jhh ORIGINAL ARTICLE Insulin resistance in essential hypertension is related to plasma renin activity

L Lind, R Reneland, P-E Andersson, A Haenni and H Lithell Department of Geriatrics, University Hospital, Uppsala, Sweden

A high plasma renin activity (PRA) has previously been well as to fasting insulin (r = 0.32, P Ͻ 0.05) and to insu- related to several cardiovascular risk factors as well as lin at 60 min at the IVGTT (r = 0.30, P Ͻ 0.05), but not to to later cardiovascular events. As insulin resistance has other risk factors. Serum aldosterone was not related to been suggested as the unifying factor in the insulin any of the metabolic risk factors. resistance metabolic syndrome, insulin resistance was In conclusion, the present investigation showed that evaluated by the euglycaemic hyperinsulinaemic clamp insulin resistance is associated with elevated levels of technique in 50 untreated hypertensive subjects in PRA in patients with untreated essential hypertension. whom PRA and serum aldosterone were measured It thus seems as if a high activity in the renin system together with lipids and an intravenous glucose toler- should be included in the disturbances included in the ance test (IVGTT). insulin resistance metabolic syndrome, a syndrome PRA was inversely related to insulin-mediated glu- with a major impact on future cardiovascular events. cose disposal during the clamp (r =−0.31, P Ͻ 0.05), as

Keywords: hypertension; insulin; insulin resistance; lipids; renin; aldosterone

Introduction essential hypertension in whom both the euglycaemic hyperinsulinaemic clamp procedure to evaluate During the last decade much attention has been paid insulin resistance and an intravenous glucose toler- to the fact that several risk factors for the develop- ance test were performed together with determi- ment of cardiovascular diseases such as hyperten- nations of PRA and serum aldosterone. sion, dyslipidaemia with low levels of HDL-cholesterol and hypertriglyceridaemia, insulin resistance with hyperinsulinaemia and abdominal obesity, tend to cluster in the same individuals.1–3 This clus- Materials and methods tering of risk factors has been named ‘syndrome X’ or ‘the insulin resistance metabolic syndrome’, as Patients were recruited by advertisements in the insulin resistance has been suggested to be a com- 4 local newspapers. At an initial screening visit mon link in this syndrome. patients received oral and written information about For more than two decades it has been obvious the study and gave informed consent. Secondary that patients with essential hypertension could hypertension was excluded by medical history, present with a very divergent activity of the renin- physical examination and laboratory tests. Patients aldosterone system.5 It has also been shown that a then received single-blind placebo for 4–6 weeks high plasma renin activity (PRA) is a powerful car- with blood pressure (BP) measurements every week diovascular risk factor independently of other for later inclusion in a drug treatment trial. If during known risk factors.6,7 Although other studies have shown associations this period they had a diastolic BP (DBP) of 95–114 between the renin-aldosterone system and metabolic mm Hg on two consecutive occasions they were cardiovascular risk factors,8–12 no study has evalu- included in the present study. However, at the time ated if a relationship exists between direct measure- of collection of blood samples for PRA and other ments of insulin sensitivity and this hormonal sys- metabolic variables, five of the patients showed a tem in hypertensive patients. The present study was DBP between 90 and 95 mm Hg. therefore undertaken in a sample of patients with To reduce the risk of carry-over effects from previous anti-hypertensive medication, no investi- gations were performed until 4 weeks after discon- tinuation of the patient’s regular medication in the Correspondence: Lars Lind, Department of Medicine, University Hospital, S-751 85 Uppsala, Sweden 50 patients enrolled in the study (35 men and 15 Received 5 January 1998; revised 27 February 1998; accepted 13 women). March 1998 Insulin resistance and PRA L Lind et al 380 Blood pressure measurements ments of PRA were based upon radioimmunoassay of angiotensin-I generated by endogenous renin and Office BP and heart rate were measured twice in the renin substrate. Serum aldosterone was measured by supine as well as standing position by trained per- a radioimmunoassay. sonnel. Cuff sizes of 12×35 or 15×45 cm were used depending on the arm circumference. Ambulatory BP was measured during 24 h Anthropometric measurements (Accutracker II, Suntech Medical Instr, USA) and Height was measured to the nearest whole centi- the mean of all recordings during that period was metre and body weight was measured to the nearest used. 100 g. The body mass index (BMI) was calculated as the ratio of the weight in kg to the square of the Insulin sensitivity measurements height in metres. Whole body sensitivity to insulin was measured by the euglycaemic hyperinsulinaemic clamp pro- Statistics cedure according to DeFronzo et al13 with minor  Pearson’s correlation coefficient was used when modifications. Insulin (Actrapid Human , Novo, PRA or serum aldosterone were related to the car- Copenhagen, Denmark) was infused at a rate of diovascular risk factors. Log transformation was per- × 2 56 mU/(min m body surface area) for 2 h. The formed if variables were not normally distributed amount of glucose infused to maintain the target glu- (PRA, serum triglycerides and different measure- cose level during the last 60 min was defined as glu- ments of insulin). Multiple regression analyses in a × −1 × −1 cose uptake (M; mg kg min ). Adjustment for forward stepwise fashion were performed to evalu- the steady-state insulin concentration defined the − − ate were the influence of confounding variables, insulin sensitivity index (M/I; mg × min 1 × kg 1 × Ͻ −1 such as age, sex or BMI. P 0.05 was used as entry (mU/l) ). criterion in the models. P Ͻ 0.05 was regarded as significant and two-tailed significant levels are Intravenous glucose tolerance test (IVGTT) given. Each patient’s response to an intravenous glucose load was assessed by a 90-min intravenous glucose Results tolerance test. The glucose load was an intravenous Means and standard deviation for studied variables injection within 1.5 min of 300 mg glucose per kg of and basic characteristics are given in Table 1. body weight in a 50% glucose solution. Insulin was As shown in Table 2, PRA was significantly corre- assayed in EDTA plasma in duplicate using an enzy-  lated with insulin sensitivity, as measured by glu- matic-immunological assay (Enzymmun , Boehr- cose uptake during the euglycaemic hyperinsulinae- inger Mannheim, Germany) performed in an ES300 mic clamp, both when expressed as the M-value and automatic analyser (Boehringer Mannheim, as M/I-value (Figure 1). PRA was furthermore sig- Germany). The mean fasting plasma glucose and nificantly correlated to fasting insulin (Figure 2) and insulin values were calculated from three blood insulin at 60 min at the IVGTT. These relationships samples drawn 5 min apart prior to the injection of were still significant even when the patient with the glucose. Peak insulin response was defined as the mean of insulin values measured in the samples Table 1 Means ± s.d. for basic characteristics and studied vari- drawn at 4, 6, and 8 min. Insulin was also determ- ables ined after 60 min. Mean s.d.

Lipid and lipoprotein measurement Age (years) 58 9.7 2 Cholesterol and triglyceride concentrations in serum BMI (kg/m ) 27.6 3.3 Waist-to-hip ratio 0.93 0.06 were assayed by enzymatic techniques (Instru- Office SBP (mm Hg) 158 16 mentation Laboratories, Lexington, MA, USA) in a Office DBP (mm Hg) 98 7.0 Monarch 2000 centrifugal analyser. The lipoprotein Daytime ambulatory SBP (mm Hg) 160 15 fractions were separated by a combination of ultra- Daytime ambulatory DBP (mm Hg) 98 7.8 M-value at clamp (mg/kg/min) 5.1 1.7 centrifugation and precipitation. Very-low-density M/I-value at clamp (mg/kg/min/mU/l 5.2 2.2 lipoproteins (VLDL) were separated at a density of × 100) 1.006 kg/l. The bottom fraction, containing low-den- Fasting insulin (mU/l) 12.7 5.6 sity lipoproteins (LDL) and high-density lipopro- Fasting blood glucose (mmol/l) 5.5 0.4 teins (HDL) were separated by precipitation with Insulin peak at IVGTT (mU/l) 40 22 Insulin 60 min at IVGTT (mU/l) 29 18 magnesium chloride/phosphotungstate. K-value at IVGTT 1.09 0.35 Serum triglycerides (mmol/l) 1.75 1.4 Serum cholesterol (mmol/l) 6.03 0.98 PRA and serum aldosterone HDL-cholesterol (mmol/l) 1.18 0.25 VLDL-triglycerides (mmol/l) 1.21 1.4 Blood samples for measurement of PRA and aldos- Serum aldosterone (pmol/l) 80 28 terone were drawn in the morning after an overnight PRA (␮g/l/h) 0.90 0.68 fast after 1 h of rest in the supine position. The diet- Free fatty acids (mmol/l) 0.55 0.19 ary intake of sodium was not standardised. Measure- Insulin resistance and PRA L Lind et al 381 Table 2 Relationships between plasma renin activity and serum aldosterone and cardiovascular risk factors. Pearson’s correlation coefficient given

PRA Serum aldosterone

SBP (mm Hg) 0.03 0.28 DBP (mm Hg) 0.04 0.34* BMI (kg/m2) 0.14 −0.15 Waist-to-hip ratio 0.16 0.01 Fasting glucose (mmol/l) 0.13 0.10 Fasting insulin (mU/l) 0.32* 0.23 Insulin peak at IVGTT (mU/l) 0.19 0.13 Insulin 60 min at IVGTT (mU/l) 0.30* 0.08 K-value at IVGTT 0.06 0.02 M-value at clamp (mg/kg/min) −0.31* −0.03 M/I-value at clamp 0.29* −0.03 (mg/kg/min/mU/1×100) Serum triglycerides (mmol/l) 0.15 0.03 − Serum cholesterol (mmol/l) 0.01 0.09 Figure 2 Relationship between PRA and fasting insulin (r = 0.32, Serum free fatty acids (mmol/l) 0.05 0.04 P Ͻ 0.05). HDL-cholesterol (mmol/l) −0.14 −0.22 VLDL-triglycerides (mmol/l) 0.13 0.00 Discussion *P Ͻ0.05. The present study showed that insulin resistance was related to elevated levels of PRA, when evaluated by the euglycaemic hyperinsulinaemic clamp. ␮ Furthermore, similar relationships were obtained highest PRA value (3.5 g/l/h) was excluded from when indirect measurements of insulin resistance the analysis. No other significant relationship were used, such as fasting hyperinsulinaemia and between PRA and other cardiovascular risk factors elevated insulin levels during the IVGTT. was seen. Although it has previously been shown by our- Except for a relationship between DBP and serum selves,8 and other investigators,9–12 that high levels aldosterone no significant relationships between the of PRA are related to hyperinsulinaemia, this is the circulating levels of this hormone and other risk fac- first study to establish this relationship in untreated tors were seen. patients with essential hypertension with direct The relationships between PRA and insulin sensi- measurements of insulin sensitivity. This approach tivity (P Ͻ 0.05), fasting insulin (P Ͻ 0.05) and insu- Ͻ has previously been used in healthy normotensive lin during the IVGTT (P 0.05) were still significant volunteers with similar results.14 In the study perfor- after the influence of age, sex, BMI and BP were med in normotensive subjects PRA was furthermore taken into account in the multiple regression analy- related to serum levels of triglycerides. Only a weak ses. The relationship between PRA and insulin sen- tendency for such a relationship could be seen in sitivity was very similar in men (r =−0.31) and =− the present study, as well as in our previous study women (r 0.29). in a separate sample of untreated hypertensive subjects.8 In a previous study, Goodfriend and co-workers10 suggested that high aldosterone levels may be a link between dyslipidaemia, insulin resistance and hypertension based on findings of an inverse relationship between plasma aldosterone and HDL- cholesterol and positive correlations between plasma aldosterone and insulin and triglycerides. In the present study no relationships between serum levels of aldosterone and any of the metabolic cardiovascular risk factors could be seen, although a tendency towards an inverse relationship between serum aldosterone and HDL-cholesterol was found in accordance with our previous study.8 The fact that salt intake was not controlled in the present study might explain why the correlations between serum aldosterone and metabolic cardiovascular risk factors did not reach statistical significance. The mechanisms connecting high PRA and insulin resistance are not known. It has been shown by Figure 1 Relationship between PRA and insulin sensitivity, several authors that during acute hyperinsulinae- expressed as the M-value during hyperinsulinaemic euglycaemic mia, as induced by the euglycaemic hyperinsulinae- clamp (r =−0.31, P Ͻ 0.05). mic clamp, PRA, but not serum levels of aldosterone Insulin resistance and PRA L Lind et al 382 is increased.15,16 The reason for this elevation in aldosterone, heart attack and stroke. N Engl J Med PRA might be the concomitant rise in circulating 1972; 9: 441–449. levels of catecholamines during hyperinsulinaemia. 7 Alderman MH et al. Association of the renin-sodium It seems less likely that an increased activity of profile with the risk of myocardial infarction in plasma renin directly would impair insulin sensi- patients with hypertension. N Engl J Med 1991; 324: 1098–1104. tivity, as infusion of angiotensin-II in healthy volun- 8 Lind L, Lithell H, Wide L, Ljunghall S. Metabolic car- teers resulted in an increase in insulin sensi- diovascular risk factors and the renin-aldosterone sys- 17,18 tivity. On the other hand, long-term treatment tem in essential hypertension. J Hum Hypertens 1992; with the ACE inhibitor captopril in hypertensive 6: 27–29. subjects resulted in a sustained improvement in 9 Allikmets K, Parik T, Teesalu R. Association between insulin sensitivity.19 plasma renin activity and metabolic cardiovascular Another putative factor might be the magnesium risk factors in essential hypertension. J Intern Med ion. It has been shown that hypertensive subjects 1996; 239: 49–55. with high PRA have low levels of serum magnes- 10 Goodfriend TL, Egan B, Stepniakowski K, Ball DL. ium,20 while on the other hand intracellular magnes- Relationships among plasma aldosterone, high-density lipoprotein cholesterol, and insulin in humans. Hyper- ium levels have been found to be inversely related 21 tension 1995; 25: 30–36. to insulin sensitivity. 11 Phillips GB, Jing T-Y, Laragh JH, Sealey JE. Serum sex In the present study a large number of univariate hormone levels and renin-sodium profile in men with correlations have been studied and some of them hypertension. Am J Hypertens 1995; 8: 626–629. might therefore have turned out to be significant by 12 Egan BM, Stepniakowski K, Goodfriend TL. Renin and chance only. However, as the major aim of this study aldosterone are higher and the hyperinsulinemic effect was to confirm our previous finding of an associ- of salt restriction greater in subjects with risk factors ation between high PRA levels and insulin resist- clustering. Am J Hypertens 1994; 7: 886–893. ance with a more accurate technique, we believe 13 DeFronzo RA, Tobin JD, Andres R. Glucose clamp that the main finding in this study is not due to mul- technique: a method for quantifying insulin secretion and resistance. Am J Physiol 1979; 6: 214–223. tiple testing. 14 Townsend RR, Zhao H. Plasma renin activity and insu- In conclusion, the present investigation showed lin sensitivity in normotensive subjects. Am J Hyper- that insulin resistance is associated with elevated tens 1994; 7: 894–898. levels of PRA in patients with untreated essential 15 Stenvinkel P, Bolinder J, Alvestrand A. Effects of insu- hypertension. It thus seems as if a high activity in lin on renal haemodynamics and the proximal and dis- the renin system should be included in the disturb- tal tubular sodium handling in healthy subjects. Dia- ances included in the insulin resistance metabolic betologia 1992; 35: 1042–1048. syndrome, a syndrome with a major impact on 16 Trovati M et al. Insulin influences the renin-angioten- future cardiovascular events. sin-aldosteron system in humans. Metabolism 1989; 38: 501–503. 17 Townsend RR, Dipette DJ. Pressor doses of angioten- References sin-II increase insulin-mediated glucose uptake in normotensive man. Am J Physiol 1993; 265: E362–E366. 1 Pollare T, Lithell H, Berne C. Insulin resistance is a 18 Buchana TA et al. Angiotensin II increases glucose uti- characteristic feature of primary hypertension inde- lization during acute hyperinsulinemia via a hemody- pendently of obesity. Metabolism 1990; 39: 167–174. namic mechanism. J Clin Invest 1993; 92: 720–726. 2 Ferrannini E et al. Insulin resistance in essential 19 Lind L, Berne C, Pollare T, Lithell H. Long-term meta- hypertension. N Engl J Med 1987; 317: 350–357. bolic effects of different antihypertensive treatments. 3 Modan M et al. Hyperinsulinemia: a link between Am Heart J 1994; 128: 1177–1183. hypertension obesity and glucose intolerance. J Clin 20 Resnick LM, Laragh JH, Sealey JE, Alderman MH. Invest 1985; 75: 809–817. Divalent cations in essential hypertension. Relations 4 Reaven GM. Role of insulin resistance in human dis- between serum ionized calcium, magnesium, and ease. Diabetes 1988; 37: 1595–1607. plasma renin activity. N Engl J Med 1983; 309: 888– 5 Laragh JH et al. Renin, angiotensin and aldosterone 891. system in pathogenesis and management of hyperten- 21 Resnick LM et al. Hypertension and peripheral insulin sive vascular diseases. Am J Med 1972; 52: 633–652. resistance. Possible mediating role of intracellular free 6 Brunner HR et al. Essential hypertension: Renin and magnesium. Am J Hypertens 1990; 3: 373–379.