Cardiovascular and Metabolic Risk ORIGINAL ARTICLE

Insulin Resistance and Hyperinsulinemia Are Related to Plasma Levels in Hypertensive Patients

GIANLUCA COLUSSI, MD ELISA NADALINI, MD and maintenance of high CRISTIANA CATENA, MD, PHD ALESSANDRA CHIUCH, MD in different ethnic groups (5), with a rela- ROBERTA LAPENNA, MD LEONARDO A. SECHI, MD tionship that is stronger in blacks (5) and in obese subjects (6). In the Framingham Offspring Study, normotensive subjects OBJECTIVE — An association between aldosterone and resistance has been demon- with elevated plasma aldosterone levels, strated in obesity and and in blacks with the metabolic syndrome. The albeit within the normal range, were at aim of this study was to evaluate the relationship of plasma aldosterone with insulin sensitivity high risk of blood pressure elevation and in white subjects. subsequent development of (7). Moreover, recent evidence indicates RESEARCH DESIGN AND METHODS — In 356 patients with that chronic exposure to elevated aldoste- and 102 normotensive control subjects of comparable age and BMI, we measured, after discon- tinuation of treatment, plasma active renin, aldosterone, , glucose, insulin, and C- rone levels might result in substantial levels and calculated markers of insulin sensitivity. Direct assessment of insulin sensitivity was damage of the and blood vessels. obtained in a subset of 64 hypertensive patients by a hyperinsulinemic clamp. This damage appears to be independent of the blood pressure level and might con- RESULTS — Hypertensive patients had significantly greater fasting plasma insulin and C- tribute to an increased risk of cardiovas- peptide concentrations and model assessment (HOMA) indexes than normotensive cular events (8). control subjects. A positive association with increasing plasma aldosterone concentrations was A relationship between aldosterone demonstrated for plasma glucose, insulin, C-, and HOMA. Assessment of insulin sen- and insulin resistance has been consis- sitivity by clamp showed a significant decrease of the metabolic rate of glucose with tently demonstrated in obesity (9) and increasing aldosterone levels. Significant correlations were found between plasma aldosterone, primary aldosteronism (10). Two recent plasma insulin, and C-peptide levels, HOMA, and glucose metabolic clearance rate. Blood pressure and plasma potassium, plasma cortisol, and renin levels, but not BMI, were also directly studies that have been conducted in fam- correlated with plasma aldosterone. Multiple regression analysis showed that HOMA, together ilies of African descent in the Seychelles with plasma potassium, cortisol, and renin levels, was independently correlated with plasma (11) and in African Americans (12) have aldosterone. demonstrated that plasma aldosterone, but not plasma renin, is associated with CONCLUSIONS — This study demonstrates a direct relationship between aldosterone, in- the metabolic syndrome and with mark- sulin resistance, and hyperinsulinemia in white subjects. In patients with hypertension, this ers of insulin resistance. The present relationship might contribute to maintenance of high blood pressure and increased cardiovas- study has evaluated the relationship of cular risk. plasma aldosterone with glucose metabo- lism and insulin sensitivity in white pa- Diabetes Care 30:2349–2354, 2007 tients, the majority of whom had essential hypertension. eminal studies that were published in blacks (3). Population-based studies Ͼ20 years ago demonstrated an as- have subsequently suggested that insulin RESEARCH DESIGN AND sociation between hyperinsulin- resistance and hyperinsulinemia might S METHODS — A total of 356 patients emia, insulin resistance, and arterial contribute to progression of cardiovascu- with mild to moderate essential hyperten- hypertension (1,2). This association was lar disease (4). sion who were referred to the hyperten- confirmed even after adjustment for body Elevated plasma aldosterone levels sion clinic of our department were weight and was present in whites but not have been implicated in the development included in a cross-sectional study. High ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● blood pressure (systolic blood pressure From the Hypertension and Diabetes Unit, Division of Internal Medicine, Department of Experimental and Ն140 mmHg and/or diastolic blood pres- Clinical Pathology and Medicine, University of Udine, Udine, Italy. sure Ն90 mmHg) was measured at least Address correspondence and reprint requests to Leonardo A. Sechi, MD, Clinica Medica, University of Udine, Department of Experimental and Clinical Pathology and Medicine, Piazzale S. Maria della Misericor- twice on two different occasions and sub- dia, 1 Udine 33100, Italy. E-mail: [email protected]. sequently confirmed on at least two more Received for publication 16 March 2007 and accepted in revised form 8 June 2007. visits during the next 4 weeks. Blood Published ahead of print at http://care.diabetesjournals.org on 15 June 2007. DOI: 10.2337/dc07-0525. pressure was measured by a mercury Additional information for this article can be found in an online appendix at http://dx.doi.org/10.2337/ dc07-0525. sphygmomanometer after each subject Abbreviations: HOMA, homeostasis model assessment; QUICKI, quantitative insulin sensitivity check had been supine for 15 min. The average index. of three readings obtained in 5 min was A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion recorded. The study patients seen at our factors for many substances. clinic are white, include individuals with © 2007 by the American Diabetes Association. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby all grades of hypertension living in north- marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. east Italy, and are representative of hyper-

DIABETES CARE, VOLUME 30, NUMBER 9, SEPTEMBER 2007 2349 Aldosterone and insulin resistance tensive patients in this geographic area. Table 1—Clinical characteristics, laboratory variables, and glucose metabolism parameters Patients with secondary hypertension, se- of the study subjects vere hypertension (as defined by diastolic blood pressure Ն120 mmHg), renal fail- Ͻ Normotensive Hypertensive ure with clearance 30 ml/min group group P per 1.73 m2 of body surface area, urinary excretion Ն1.0 g/day, pregnancy, Clinical characteristics chronic debilitating illness, and recent n 102 356 — (within 6 months) myocardial infarction, Age (years) 51 Ϯ 14 49 Ϯ 12 0.154 unstable angina, or stroke were excluded. Sex (male) 71 (70) 195 (55) 0.028 Secondary causes of hypertension were SBP (mmHg) 129 Ϯ 11 160 Ϯ 19 Ͻ0.001 identified on the basis of extensive labo- DBP (mmHg) 79 Ϯ 7 100 Ϯ 11 Ͻ0.001 ratory testing (13). Primary aldosteron- BMI (kg/m2) 28.3 Ϯ 3.6 27.9 Ϯ 4.8 0.435 ism was screened by the demonstration of Laboratory variables an increased plasma aldosterone–to– Plasma (mmol/l) 141 Ϯ 2 141 Ϯ 3 1.000 renin ratio in the presence of a plasma Plasma potassium (mmol/l) 4.3 Ϯ 0.3 4.0 Ϯ 0.4 Ͻ0.001 aldosterone concentration Ͼ150 pg/ml Plasma creatinine (␮mol/l) 84 Ϯ 25 88 Ϯ 17 0.062 and confirmed by the lack of aldosterone Urinary sodium (mmol/24 h) 120 Ϯ 52 132 Ϯ 67 0.098 suppression following an intravenous sa- Urinary potassium (mmol/24 h) 46 Ϯ 23 56 Ϯ 23 Ͻ0.001 line load (10,14). All measurements were Plasma active renin (pg/ml) 9.2 Ϯ 10.7 10.8 Ϯ 17.0 0.575 performed under a normal sodium diet, Plasma aldosterone (pg/ml) 131 Ϯ 77 167 Ϯ 123 0.005 and 24-h urinary sodium excretion was Plasma cortisol (nmol/l) 429 Ϯ 107 420 Ϯ 236 0.709 assessed in all patients. Patients treated with Triglycerides (mmol/l) 1.25 Ϯ 0.56 1.40 Ϯ 0.92 0.118 antihypertensive drugs were withdrawn Total cholesterol (mmol/l) 5.30 Ϯ 1.04 5.58 Ϯ 1.13 0.025 from treatment a minimum of 2 weeks be- HDL cholesterol (mmol/l) 1.43 Ϯ 0.41 1.42 Ϯ 0.42 0.831 fore diagnostic assessment. No patient was LDL cholesterol (mmol/l) 3.31 Ϯ 0.98 3.55 Ϯ 1.05 0.039 taking aldosterone antagonists. Glucose metabolism parameters Patients with essential hypertension Plasma glucose (mmol/l) 4.8 Ϯ 0.9 5.1 Ϯ 1.2 0.020 were compared with 102 normotensive Plasma insulin (pmol/l) 55.9 Ϯ 21.7 70.0 Ϯ 35.1 Ͻ0.001 subjects who were selected from the gen- Plasma C-peptide (nmol/l) 0.53 Ϯ 0.20 0.69 Ϯ 0.30 Ͻ0.001 eral population of the same geographic HOMA index 1.65 Ϯ 0.64 2.29 Ϯ 1.55 Ͻ0.001 area as the hypertensive patients after QUICKI 0.354 Ϯ 0.013 0.348 Ϯ 0.030 0.050 specification of inclusion criteria to avoid Data are means Ϯ SD or n (%) unless otherwise indicated. Comparisons were done by Student’s t test for age and BMI as potential confounding unpaired data. DBP, diastolic blood pressure; SBP, systolic blood pressure. variables. Normotensive control subjects were not taking any regular medications and did not have any concomitant dis- (microunits per millimeter) using the fol- pressed as the glucose metabolic clear- ease. Informed consent was obtained lowing formula: [(glucose ϫ insulin)/ ance rate (milliliters per kg body weight from the study participants, and the study 22.5]. Logarithmic values of fasting per minute) during 60 min of the clamp. protocol received approval by the local plasma glucose (milligrams/deciliter) and Sodium, potassium, and creatinine review committee. insulin (microunits per millimeter) con- were measured in plasma obtained after centrations were obtained to calculate the fasting for 12–14 h by automated analyz- Glucose metabolism evaluation and QUICKI using the following formula: ers. Plasma glucose was assayed using the laboratory measurements [1/(log glucose ϩ log insulin)]. glucose oxidase method. Plasma insulin Assessment of glucose metabolism pa- Insulin sensitivity was further and di- and C-peptides were measured by radio- rameters and insulin sensitivity was done rectly assessed in a subgroup of 64 patients immunoassay (16). Plasma active renin at the same time as diagnostic screening with hypertension by a hyperinsuline- and aldosterone concentrations were mea- after appropriate antihypertensive drugs mic-euglycemic clamp that was per- sured by radioimmunoassay in plasma wash-out, as previously described (15). formed as previously described (15). samples obtained with patients in the sit- At the time of the study, patients main- Briefly, a priming dose of 100 mU/kg ting position. Both renin and aldosterone tained their usual unrestricted diet. A body wt of rapidly acting insulin was ad- values were referred to the urinary so- sample of venous blood was obtained af- ministered intravenously over a period of dium excretion of a 24-h collection com- ter fasting for 12–14 h and after the pa- 10 min, and then a sustained infusion of pleted on the day of sampling (17). Ϫ tients were in the sitting position for 10 insulin at a rate of 2 mU kg body wt 1 Ϫ min for analysis of glucose, insulin, and min 1 was started to maintain serum in- Statistical analysis C-peptide. The homeostasis model as- sulin concentrations at ϳ700 pmol/l. All values are expressed as means Ϯ SD. sessment (HOMA) index and the quanti- Concomitantly, an intravenous infusion Variables with skewed distribution were tative insulin sensitivity check index of a 20% glucose solution was started to analyzed after logarithmic transforma- (QUICKI) were calculated as markers of stabilize blood glucose values at 5.0 tion. Student’s t test was used for compar- sensitivity to insulin (15). The HOMA in- mmol/l. For this purpose, plasma glucose isons between normotensive and dex was calculated from fasting plasma was determined every 10 min during the hypertensive subjects. One-way ANOVA glucose (millimoles per liter) and insulin clamp. Sensitivity to insulin was ex- was used for comparisons of values when

2350 DIABETES CARE, VOLUME 30, NUMBER 9, SEPTEMBER 2007 Colussi and Associates

Table 2—Clinical characteristics, laboratory variables, and glucose metabolism parameters of hypertensive patients according to plasma aldosterone tertiles

Tertile I Tertile II Tertile III P Clinical characteristics n 119 118 119 — Age (years) 50 Ϯ 12 49 Ϯ 12 49 Ϯ 11 0.655 Sex (male) 62 (52) 73 (62) 60 (50) 0.161 SBP (mmHg) 157 Ϯ 18 160 Ϯ 18 162 Ϯ 20 0.088 DBP (mmHg) 98 Ϯ 10 100 Ϯ 10 101 Ϯ 11 0.026 BMI (kg/m2) 27.3 Ϯ 4.6 28.2 Ϯ 4.7 28.3 Ϯ 5.0 0.193 Laboratory variables Plasma potassium (mmol/l) 3.8 Ϯ 0.4 4.0 Ϯ 0.4 4.1 Ϯ 0.4 Ͻ0.001 Urinary sodium (mmol/24 h) 133 Ϯ 70 131 Ϯ 68 132 Ϯ 64 0.963 Urinary potassium (mmol/24 h) 52 Ϯ 23 59 Ϯ 27 58 Ϯ 23 0.113 Plasma creatinine (␮mol/l) 86 Ϯ 15 91 Ϯ 19 89 Ϯ 17 0.048 Creatinine clearance (ml/min per 1.73 m2) 94 Ϯ 27 90 Ϯ 24 91 Ϯ 22 0.503 Plasma active renin (pg/ml) 9.7 Ϯ 17.0 11.7 Ϯ 18.9 11.1 Ϯ 15.1 0.721 Plasma aldosterone (pg/ml) 68 Ϯ 22 139 Ϯ 24 293 Ϯ 132 Ͻ0.001 Plasma cortisol (nmol/l) 356 Ϯ 264 408 Ϯ 213 494 Ϯ 209 Ͻ0.001 Glucose metabolism parameters Plasma glucose (mmol/l) 4.8 Ϯ 0.9 5.2 Ϯ 1.2 5.3 Ϯ 1.3 0.002 Plasma insulin (pmol/l) 60.3 Ϯ 31.2 73.1 Ϯ 35.6 76.8 Ϯ 36.1 0.001 Plasma C-peptide (nmol/l) 0.66 Ϯ 0.29 0.67 Ϯ 0.32 0.75 Ϯ 0.28 0.043 HOMA index 1.84 Ϯ 1.28 2.40 Ϯ 1.45 2.63 Ϯ 1.77 Ͻ0.001 QUICKI 0.359 Ϯ 0.029 0.345 Ϯ 0.030 0.341 Ϯ 0.030 Ͻ0.001 Data are means Ϯ SD or n (%) unless otherwise indicated. Comparisons were done by one-way ANOVA. DBP, diastolic blood pressure; SBP, systolic blood pressure. the patients were subdivided in aldoste- pressed plasma renin (Ͻ2.5 ng/ml) was tabolism when compared with each rone tertiles. Pearson’s ␹2 test was used to 19.1%. Table 2 summarizes the intra- other. BMI was significantly and directly compare frequency distributions. The re- group comparison of patients with hyper- correlated with systolic (r ϭ 0.110; P ϭ lationship between continuously distrib- tension and demonstrates that increasing 0.038) and diastolic (r ϭ 0.152; P ϭ uted variables was examined by linear plasma aldosterone levels were associated 0.024) blood pressure and with HOMA regression analysis, and the correlation with higher diastolic blood pressure and (r ϭ 0.334; P Ͻ 0.001) but not with al- was expressed by Pearson’s correlation higher potassium, creatinine, and cortisol dosterone and renin levels. Additional coefficient. Stepwise multiple regression levels. Significant association with in- correlations were found between plasma analysis was used to ascertain which vari- creasing plasma aldosterone was demon- potassium and HOMA (r ϭ 0.285; P Ͻ ables were independently associated. strated for fasting plasma glucose, insulin, 0.001) and QUICKI (r ϭϪ0.239; P Ͻ Two-tailed probability values Ͻ0.05 were and C-peptide levels, HOMA index, and 0.001). Multiple regression analysis was considered to indicate statistical signifi- QUICKI. performed with a forward stepwise ap- cance. In patients with hypertension, uni- proach in which variables that were sig- variate analysis showed that plasma aldo- nificantly related to aldosterone in RESULTS — The clinical, laboratory, sterone concentrations were directly univariate analysis were included follow- and metabolic measurements of the study correlated with fasting plasma insulin ing the strength of the relationship. Anal- subjects are shown in Table 1. Plasma al- (r ϭ 0.214; P Ͻ 0.001), C-peptide (r ϭ ysis showed that plasma potassium (P Ͻ dosterone levels and urinary potassium 0.138; P ϭ 0.009), and HOMA (r ϭ 0.001), plasma cortisol (P Ͻ 0.001), excretion were greater and plasma potas- 0.228; P Ͻ 0.001) and inversely corre- HOMA (P ϭ 0.009), and plasma active sium lower in the hypertensive patients lated with QUICKI (r ϭϪ0.223; P Ͻ renin (P ϭ 0.013) were independently than in the normotensive control sub- 0.001) (supplementary Figure [viewable correlated with plasma aldosterone levels jects, with no difference in plasma active in an online appendix, available at http:// (Table 3). The relationship of aldosterone renin or cortisol levels. In the hyperten- dx.doi.org/10.2337/dc07-0525]). Plasma with insulin and HOMA was independent sive patients, fasting plasma glucose, in- aldosterone was also positively correlated (P Ͻ 0.001) of blood pressure levels, and sulin, and C-peptide concentrations and with systolic (r ϭ 0.108; P ϭ 0.041) and the relationship between aldosterone and HOMA index were significantly different diastolic (r ϭ 0.152; P ϭ 0.004) blood blood pressure was independent (P ϭ from those in normotensive control sub- pressure and plasma potassium (r ϭ 0.013) of cortisol levels. jects, indicating the presence of insulin 0.279; P Ͻ 0.001), cortisol (r ϭ 0.255; To further explore the relationship resistance. The percentage of hyperten- P Ͻ 0.001), and active renin (r ϭ 0.138; between plasma aldosterone and sensitiv- sive patients with plasma aldosterone P ϭ 0.023) levels. No correlations were ity to insulin, we measured the metabolic above the normal range was 14.6%, and observed among plasma renin, blood clearance rate of glucose in a subgroup of the percentage of patients with sup- pressure, and parameters of glucose me- 64 patients with hypertension who un-

DIABETES CARE, VOLUME 30, NUMBER 9, SEPTEMBER 2007 2351 Aldosterone and insulin resistance

Table 3—Stepwise linear regression analysis of variables associated with plasma aldosterone ectomy or treatment with aldosterone an- tagonists (10) restored normal sensitivity (356 ؍ levels in hypertensive patients (n to insulin. More recently, two large stud- Plasma active ies have reported that plasma aldosterone, Plasma potassium Plasma cortisol HOMA index renin but not plasma renin levels, are associated with the metabolic syndrome (11,12) and SC P SC P SC P SC P markers of insulin resistance (12) in nor- 0.331 Ͻ0.001 — — ————motensive and hypertensive blacks. In 0.287 Ͻ0.001 0.222 Ͻ0.001 ————contrast, in a subanalysis of the Trial of 0.257 Ͻ0.001 0.211 Ͻ0.001 0.153 0.010 — — Preventing Hypertension Study (19), no 0.259 Ͻ0.001 0.200 Ͻ0.001 0.155 0.009 0.144 0.013 evidence for elevated aldosterone was found in individuals with high normal Calculations were done with log-transformed values. SC, standard coefficient. blood pressure and the metabolic syn- drome when compared with that in con- derwent a hyperinsulinemic-euglycemic patients, the majority of whom had high trol subjects without the syndrome. In clamp. In these patients, we observed a blood pressure. Multivariate analysis that study, however, 82% of patients were significant decrease of glucose metabolic demonstrates that this association is inde- white, raising the issue of a race-specific clearance rate with increasing aldosterone pendent of plasma potassium and cortisol effect. Our findings extend the evidence levels (Fig. 1), with a highly significant levels. In patients with hypertension, the of a significant association between aldo- inverse correlation between the rate of relationship between aldosterone and de- sterone, hyperinsulinemia, and insulin glucose disposal and plasma aldosterone creased sensitivity to insulin was con- resistance to white subjects. Because the (r ϭϪ0.586; P Ͻ 0.001). firmed by direct assessment of insulin- majority of the individuals included in Analysis of correlations that included mediated glucose disposal rates under a our study had hypertension, it could be both healthy control subjects and hyper- hyperinsulinemic-euglycemic clamp. speculated that aldosterone and insulin tensive patients demonstrated a positive The issue of a possible relationship resistance might together contribute to and highly significant relationship be- between plasma aldosterone and insulin rising blood pressure and, eventually, in- tween plasma aldosterone and fasting resistance is important because aldoste- creased cardiovascular risk. plasma insulin (r ϭ 0.208; P Ͻ 0.001), rone has been shown to contribute, inde- The methodology required to mea- fasting plasma C-peptide (r ϭ 0.161; P Ͻ pendent of blood pressure, to the sure insulin sensitivity is complex, and 0.001), and HOMA (r ϭ 0.243; P Ͻ development of cardiovascular damage this makes the translation of research 0.001). Multivariate analysis showed that (8) and because insulin resistance and hy- findings on insulin resistance into clinical the relationship of plasma aldosterone perinsulinemia are predictors of cardio- practice rather difficult. Previous studies with HOMA (P ϭ 0.008) was indepen- vascular events in hypertensive patients (11,12,19) have defined insulin resis- dent of blood pressure, active renin, and (18), as in the general population (4). Ini- tance by use of fasting plasma glucose and cortisol. tial demonstrations of an association be- insulin values rather than by the gold tween plasma aldosterone levels and standard, the euglycemic-hyperinsuline- CONCLUSIONS — The results of the insulin resistance were obtained in obese mic clamp. Our study is the first to pro- present study demonstrate that plasma al- subjects (9) and patients with primary al- vide direct assessment of insulin- dosterone levels are associated with dosteronism (10). In these patients, mediated glucose disposal rate and to plasma markers of insulin resistance and weight loss (9) and removal of the effects demonstrate, with this technique, a hyperinsulinemia in a white population of of excess aldosterone with either adrenal- strong association between elevated plasma aldosterone and decreased sensi- tivity to insulin. Although an association does not necessarily imply causality, the strength of the inverse relationship be- tween aldosterone and the metabolic clearance rate of glucose clearly suggests the possibility that elevated aldosterone might cause or, alternatively, be the result of insulin resistance. The interaction between mineralo- corticoid and insulin that is suggested by the present findings is sup- ported by substantial experimental evi- dence (rev. in 20,21). It was initially thought that the cause leading to glucose intolerance in conditions characterized by increased plasma aldosterone, such as Figure 1— Bar graph showing the glucose metabolic clearance rate, as assessed during a eugly- primary aldosteronism, is potassium de- cemic-hyperinsulinemic clamp, across plasma aldosterone tertiles in patients with essential hy- pletion, which could modulate both pan- pertension (n ϭ 356). Comparisons were done by one-way ANOVA (P Ͻ 0.001) followed by creatic insulin secretion and insulin group-to-group comparisons. MCR, metabolic clearance rate. receptor function (22,23). In this study,

2352 DIABETES CARE, VOLUME 30, NUMBER 9, SEPTEMBER 2007 Colussi and Associates plasma aldosterone was correlated with sociations, such as those between aldoste- factor for ischemic heart disease. N Engl hyperinsulinemia and markers of insulin rone and BMI, that have been reported in J Med 334:952–957, 1996 resistance independent of plasma potas- previous studies (12,33,34) might have 5. Grim CE, Cowley AW Jr, Hamet P, sium, ruling out a possible role for this been missed because the average BMI of Gaudet D, Kaldunski ML, Kotchen JM, electrolyte in mediating the relationship. our hypertensive patients was relatively Krishnaswami S, Pausova Z, Roman R, Tremblay J, Kotchen TA: Hyperaldoste- On the other hand, aldosterone might ex- low, compared with that of hypertensive ronism and hypertension: ethnic differ- ert direct effects on insulin receptors (22), patients included in those studies. In this ences. Hypertension 45:766–772, 2005 and recent experiments indicate that al- context, measurement of waist circumfer- 6. Goodfriend TL, Calhoun DA: Resistant dosterone might decrease insulin sensi- ence, as a more specific indicator of vis- hypertension, obesity, sleep apnea, and tivity in human adipocytes (24). Finally, it ceral , would have been aldosterone: theory and therapy. Hyper- is possible that greater aldosterone levels useful. tension 43:518–524, 2004 might result from hyperinsulinemia This study demonstrates a significant 7. Vasan RS, Evans JC, Larson MG, Wilson (25,26) or might be related to the associ- relationship between aldosterone and in- PW, Meigs JB, Rifai N, Benjamin EJ, Levy ation between insulin resistance and body sulin resistance with use of the clamp and D: Serum aldosterone and the incidence fat content. Fatty acids and adipokines re- extends to whites the relationship previ- of hypertension in nonhypertensive per- sons. N Engl J Med 351:33–41, 2004 leased from adipose tissue play a key role ously demonstrated in studies performed 8. Rossi GP, Boscaro M, Ronconi V, Funder in the development of insulin resistance in blacks. The interaction between in- JW: Aldosterone as a cardiovascular risk (27) and have been shown to stimulate creased aldosterone and insulin resistance factor. Trends Endocrinol Metab 16:104– aldosterone production (9,28). More- could contribute to maintenance of hy- 107, 2005 over, experimental observations indicate pertension and increase the risk of cardio- 9. Goodfriend TL, Egan BM, Kelley DE: that fat cells can directly stimulate aldo- vascular events in these subjects. Further Plasma aldosterone, plasma lipoproteins, sterone secretion by adrenal glands in studies will be necessary to establish obesity and insulin resistance in humans. vitro (29). whether increased aldosterone decreases Leukot Essent Fatty Acids 60: Plasma aldosterone (5–7), hyperinsu- peripheral sensitivity to insulin or, alter- 401–405, 1999 linemia, and insulin resistance (1–3) can natively, whether insulin resistance with 10. Catena C, Lapenna R, Baroselli S, Nad- contribute to maintenance of increased ensuing hyperinsulinemia stimulates al- alini E, Colussi GL, Novello M, Favret G, Melis A, Cavarape A, Sechi LA: Insulin blood pressure in the hypertensive popu- dosterone production. It would be also sensitivity in patients with primary aldo- lation through several mechanism. In this worth testing the possibility that pharma- steronism: a follow-up study. J Clin Endo- study, we have observed significant recip- cological interventions with aldosterone crinol Metab 91:3457–3463, 2006 rocal and independent correlations be- antagonists would be particularly benefi- 11. Bochud M, Nussberger J, Bovet P, Mail- tween aldosterone, insulin, and blood cial on the clinical outcome of patients lard MR, Elston RC, Paccaud F, Shamlaye pressure, suggesting the possibility that with insulin resistance and that insulin C, Burnier M: Plasma aldosterone is inde- interactions between these hormones sensitizers would favorably affect the pendently associated with the metabolic might affect regulation of blood pressure. course of clinical conditions characterized syndrome. Hypertension 48:239–245, 2006 For instance, insulin has been shown to by high circulating levels of aldosterone. 12. Kidambi S, Kotchen JM, Grim CE, Raff H, modulate the blood pressure response to Mao J, Singh RJ, Kotchen TA: Association of adrenal steroids with hypertension and aldosterone (30), and decrease in plasma the metabolic syndrome in blacks. Hyper- aldosterone could contribute to the Acknowledgments— This work was sup- tension 49:1–8, 2007 change of blood pressure with weight loss ported by grants from the Italian Ministry of 13. Sechi LA, Kronenberg F, De Carli S, Fal- and resulting plasma insulin reduction the University to L.A.S. and C.C. and by grants leti E, Zingaro L, Catena C, Utermann G, (31). Also, in studies conducted in hyper- from the Italian Society of Hypertension to Bartoli E: Association of serum lipopro- insulinemic rodents, a lack of aldosterone G.L.C. and E.N. tein (a) levels and apolipoprotein (a) size suppression by salt has been demon- polymorphism with target-organ damage strated (32). Relevant to this issue, it in arterial hypertension. JAMA 277: should be noticed that, in our study, sys- References 1689–1695, 1997 tolic blood pressure had a weaker rela- 1. Modan M, Halkin H, Almog S, Lusky A, 14. Mulatero P, Stowasser M, Loh KC, Fardella CE, Gordon RD, Mosso L, Go- tionship with plasma aldosterone than Eshkol A, Shefi M, Shitrit A, Fuchs Z: Hy- perinsulinemia: a link between hyperten- mez-Sanchez CE, Veglio F, Young WF Jr: diastolic blood pressure, as a likely result sion, obesity and glucose intolerance. Increased diagnosis of primary aldoste- of its greater intrinsic variability. Plasma J Clin Invest 75:809–817, 1985 ronism, including surgically correctable aldosterone was also positively correlated 2. Ferrannini E, Buzzigoli G, Bonadonna R, forms, in centers from five continents. with plasma cortisol, but this relationship Giorico MA, Oleggini M, Graziadei L, J Clin Endocrinol Metab 89:1045–1050, did not appear to be relevant for either Pedrinelli R, Brandi L, Bevilacqua S: Insu- 2004 blood pressure or insulin sensitivity. lin resistance in essential hypertension. 15. Sechi LA, Catena C, Zingaro L, Melis A, Some limitations of this study need to N Engl J Med 317:350–357, 1987 De Marchi S: Abnormalities of glucose be highlighted. First, the cross-sectional 3. Saad MF, Lillioja S, Nyomba BL, Castillo metabolism in patients with early renal design does not permit the establishment C, Ferraro R, De Gregorio M, Ravussin R, failure. Diabetes 51:1226–1232, 2002 of clear evidence of a causal relationship Knowler WC, Bennet PH, Howard BV: 16. Sechi LA, Melis A, Tedde R: Insulin hyper- Racial differences in the relation between secretion: a distinctive feature between es- between aldosterone and insulin resis- blood pressure and insulin resistance. sential and secondary hypertension. tance, nor the establishment of which of N Engl J Med 324:733–739, 1991 Metabolism 41:1261–1266, 1992 the two is the causative factor. Second, 4. Despres JP, Lamarche B, Mauriege P, Can- 17. Sealey JE, Gordon RD, Mantero F: Plasma although this study was designed to have tin B, Dagenais GR, Moorjani S, Lupien PJ: renin and aldosterone measurements in high statistical power, some additional as- Hyperinsulinemia as an independent risk low renin hypertensive states. Trends En-

DIABETES CARE, VOLUME 30, NUMBER 9, SEPTEMBER 2007 2353 Aldosterone and insulin resistance

docrinol Metab 16:86–91, 2005 24. Kraus D, Jager J, Meier B, Fasshauer M, releasing factors. Proc Natl Acad SciUSA 18. Reaven GM: Insulin resistance/compensa- Klein J: Aldosterone inhibits uncoupling 100:14211–14216, 2003 tory hyperinsulinemia, essential hyper- protein-1, induces insulin resistance, and 30. Rocchini AP, Moorehead C, DeRemer S, tension, and cardiovascular disease. J Clin stimulates proinflammatory adipokines Goodfriend TL, Ball DL: Hyperinsulin- Endocrinol Metab 88:2399–2403, 2003 in adipocytes. Horm Metab Res 37:455– emia and the aldosterone and pressor 19. Egan BM, Papademetriou V, Wofford M, 459, 2005 responses to II. Hypertension Calhoun D, Fernandes J, Riehle JE, Nes- 25. Petrasek D, Jensen G, Tuck M, Stern N: In 15:861–866, 1990 bitt S, Michelson E, Julius S: Metabolic vitro effects of insulin on aldosterone pro- 31. Rocchini AP, Key J, Bondie D, Chico R, syndrome and insulin resistance in the duction in rat cells. Life Moorehead C, Katch V, Martin M: The TROPHY sub-study: contrasting views in Sci 50:1781–1787, 1992 effect of weight loss on the sensitivity of patients with high-normal blood pres- 26. Haenni A, Reneland R, Lind L, Lithell H: blood pressure to sodium in obese adoles- sure. Am J Hypertens 18:3–12, 2005 Serum aldosterone changes during hyper- cents. N Engl J Med 321:580–585, 1989 20. Giacchetti G, Sechi LA, Rilli S, Carey RM: insulinemia are correlated to body mass 32. Huang DY, Boini KM, Osswald H, The renin-angiotensin-aldosterone sys- index and insulin sensitivity in patients Friedrich B, Artunc F, Ullrich S, Rajaman- tem, glucose metabolism and diabetes. with essential hypertension. J Hypertens ickam J, Palmada M, Wulff P, Kuhl D, Trends Endocrinol Metab 3:120–126, 19:107–112, 2001 Vallon V, Lang F: Resistance of mice lack- 2005 27. Eckel RH, Grundy SM, Zimmet PZ: The ing the serum- and glucocorticoid-induc- 21. Corry DB, Tuck M: The effect of aldoste- metabolic syndrome. Lancet 365:1415– ible kinase SGK1 against salt-sensitive rone on glucose metabolism. Curr Hyper- 1428, 2005 hypertension induced by a high-fat diet. tens Rep 5:106–109, 2003 28. Goodfriend TL, Ball DL, Egan BM, Camp- Am J Physiol Renal Physiol 29:F1264– 22. Henquin JC: Triggering and amplifying bell WB, Nithipatikom K: Epoxy-keto F1273, 2006 pathways of regulation of insulin secre- derivative of linolenic acid stimulates al- 33. Goodfriend TL, Egan BM, Kelley DE: Al- tion by glucose. Diabetes 49:1751–1760, dosteronesecretion.Hypertension43:358– dosterone in obesity. Endocr Res 24:789– 2000 363, 2004 796, 1998 23. Plavinik FL, Rodrigues CI, Zanella MT, 29. Ehrhart-Bornstein M, Lamounier-Zepter 34. Rocchini AP, Katch VL, Grekin R, Moore- Ribeiro AB: Hypokalemia, glucose intol- V, Schraven A, Lagenbach J, Willenberg head C, Anderson J: Role for aldosterone erance, and hyperinsulinemia during di- HS, Barthel A, Hauner H, McCann SM, in blood pressure regulation of obese uretic therapy. Hypertension 19(Suppl. 2): Scherbaum WA, Bornstein SR: Human adolescents. Am J Cardiol 57:613–618, 26–29, 1992 adipocytes secrete mineralocorticoid- 1986

2354 DIABETES CARE, VOLUME 30, NUMBER 9, SEPTEMBER 2007