The growing epidemic of overweight and obesity is the leading driving force of an increased prevalence of the metabolic syndrome (MS) worldwide, a cluster of cardiovascular risk factors including impaired glucose homeostasis, atherogenic dyslipidaemia, obesity and hypertension (HTN). In turn, the presence MS results in an increased risk for type 2 diabetes mellitus (T2DM), cardiovascular disease (CVD), stroke, chronic kidney disease (CKD) and end stage renal disease.
Central to the pathophysiology of the MS is insulin resistance, and the mechanisms implicated in its development are a field of intense research. Activation of the Renin–Angiotensin–Aldosterone–system (RAAS) is linked to impaired insulin signalling and insulin resistance in classic target tissues such as skeletal muscle and liver, as well as in ‘non-classic’ target tissues for insulin action, including cardiac, vascular and renal tissues.1 Evidence of a close relationship between insulin resistance and an inappropriately overactive RAAS in the pathogenesis of HTN, one of the main components of the MS, has been recognized for a long time. Earlier studies in humans reported a significant association between insulin resistance, resulting compensatory hyperinsulinaemia and HTN.2 Subsequent clinical studies have also demonstrated a relationship between hyperinsulinaemia and ischemic heart disease in non-diabetic men, independently of other CVD risk factors such as hyperlipidaemia or HTN.3
RAAS activation results in increased production of angiotensin II (Ang II) and aldosterone. The influence of aldosterone on the development of HTN has been classically attributed to sodium retention and expansion of intravascular volume balance. In addition, new evidence suggests that CVD pathogenesis is mediated via events involving signalling through the mineralocorticoid receptor (MR). This is supported by the clinical benefits elicited by MR antagonism in patients with impaired cardiovascular function,4 as well as by increased cardiac mass and fibrosis found in patients with primary aldosteronism (PA) relative to individuals with essential HTN with comparable blood pressures.5
Aldosterone and glucose Homeostasis
Numerous epidemiologic studies have described the association between impaired glucose tolerance (IGT) and T2DM with PA, but the underlying mechanisms largely remain to be uncovered. Available literature indicates a prevalence of IGT in patients with PA that can reach up to 50% of patients with PA,6 leading to the Expert Committee for the Diagnosis and Classification of Diabetes Mellitus of the American Diabetes Association to include PA as a unique cause of T2DM.7
Some hypotheses to explain this relationship have implicated aldosterone-induced hypokalaemia resulting in impaired insulin secretion. There is indeed experimental evidence suggesting a role for potassium in the regulation of insulin receptor function and glucose-stimulated insulin secretion by β-cells.8 A recent meta-analysis reported a significant correlation between diuretics-induced hypokalaemia and abnormally increased blood glucose.9 Interestingly, a recent study in humans demonstrated a direct correlation between adiponectin, an adipokine with insulin-sensitizing properties, and potassium in patients with PA and with low renin essential HTN. This association was significantly stronger in PA patients relative to low renin hypertensive individuals,10 suggesting the possibility that chronic hypokalaemia in PA could contribute to low adiponectin levels and hence to insulin resistance.
Another clinical study reported lower glucose utilization during euglycaemia hyperinsulinaemic clamp in patients with PA compared to normotensive individuals.11 During and OGTT, these patients also exhibited higher areas under the curve for glucose, which improved after adrenalectomy. In addition, there was a significant positive correlation between serum potassium levels and insulin resistance in PA patients. Recently, Sindelka demonstrated increased insulin resistance in PA patients, which also improved after adrenalectomy.12 However, no significant improvement of glucose homeostasis parameters was observed in individuals with PA resulting from idiopathic adrenal hyperplasia treated with spironolactone. As plasma aldosterone concentrations were normalized only in surgically treated patients, the findings of this study support the concept that aldosterone per se has deleterious effects on insulin sensitivity.
Aldosterone, insulin resistance and adipose tissue RAAS
From an experimental standpoint, available literature supports several interactions between mineralocorticoids and insulin (Fig. 1). Numerous studies demonstrate an influence of insulin on aldosterone production and secretion. In rodents increased aldosterone production mediated by insulin in a dose-dependent manner in glomerulosa zone cells has been reported.13 Another study14 demonstrated that aldosterone suppresses the expression of glucose transporters in avian animal colonic cells, whereas in human promonocytes aldosterone promotes down-regulation of the insulin receptor as well as reduced subcutaneous adipose tissue insulin receptor expression by approximately 54%.15
Clinical studies have confirmed insulin resistance, evaluated through Homeostatic Model Assessment (HOMA), as well as impaired insulin-stimulated glucose utilization (measured by euglycaemic hyperinsulinaemic clamp) in PA patients but not in essential hypertensive individuals.16 However, a study by Shamiss, also using clamp techniques, found increased insulin sensitivity in patients with PA, which adds to previous observations that pharmacologic treatment does not consistently result in improved glucose tolerance.17 Other studies in humans support a strong association between aldosterone, insulin resistance and impaired glucose homeostasis. Indeed, Fallo et al. recently reported a 41·1% prevalence of MS in PA patients, compared to 29·6% in essential hypertensives (P < 0·05) in 466 hypertensive patients with PA.18
Importantly, plasma aldosterone levels are increased in obese hypertensive patients, especially in patients with excess visceral fat. A recent report from the primary aldosteronism prevalence in HTN (PAPY) Study, a large multicentre prospective study of newly diagnosed hypertensive patients referred to specialized HTN reference centres, addressed the relationship between body mass index (BMI), aldosterone, plasma renin activity and aldosterone renin ratio.19 Of 1125 hypertensive participants, 11·2% had biochemical evidence of PA, and in 5·4% of them a diagnosis of aldosterone producing adenoma was reached. There was a positive correlation between BMI and plasma aldosterone, independently of age, blood pressure, gender and sodium intake in essential hypertensive patients but not in participants with PA. The association was stronger in patients with overweight and obesity, thus suggesting a pathophysiological link between aldosterone production and fat deposition.
Interestingly, a recent experimental study demonstrates that complement-C1q TNF-related protein 1 (CTRP1), a novel adipokine, that shares structural homology with adiponectin, stimulates aldosterone production in Zucker diabetic fatty rats.20 Conversely aldosterone and MR may be involved in adipose tissue regulation, as aldosterone promotes achievement of the white adipose phenotype in a MR-dependent manner, possibly via activation of Peroxisome Proliferator Activated Receptor gamma in experimental conditions.21
On the other hand, the role of an active local RAAS appears to be of paramount importance in the relationship between obesity, insulin resistance and HTN, as there is abundant evidence demonstrating that adipose tissue expresses a functional RAAS. RAAS triggers the activation of the NADPH oxidase enzymatic complex, a highly regulated membrane-bound enzyme complex that catalyses the production of reactive oxygen species (ROS), which in turn regulate numerous tyrosine kinases, mitogen-activated protein kinases (MAPKs) and cysteine-based phosphatases such as tyrosine phosphatases and lipidic phosphatases. These enzymes, subsequently can potentially impact several intracellular signalling pathways, as well as cellular proliferation, migration, secretion of inflammatory cytokines, activation of extracellular metalloproteinases, contraction, differentiation and apoptosis of vascular smooth muscle cells (VSMCs) (Fig. 2).
Non-genomic actions of aldosterone and the Metabolic Syndrome (MS)
The receptors and pathways implicated in non-genomic actions of mineralocorticoids remain to be fully elucidated. The existence of a membrane receptor in non-epithelial tissues, able to bind aldosterone has been hypothesized.22 However, it is widely suggested that non-genomic actions of aldosterone are largely mediated through the classic MR. Also, non-genomic actions of mineralocorticoids appear to be mediated via activation of the protein kinase C (PKC) pathway, increased sodium/hydrogen interchange activity and intracellular calcium.23–25 In addition, mineralocorticoid-induced stimulation of c-Src induces activation MAPKs (P38 MAPK, JNK, ERK1/2) associated with cellular growth, apoptosis and collagen deposition.26 Also, Ang II induces cardiovascular tissue remodelling, proliferation, migration and hypertrophy through activation of several small G proteins including Ras, Rho and Rac.27 RAAS-mediated oxidative stress results in activation of the JAK/STAT, Akt (protein Kinase B), and P38 MAPK pathways, which are implicated in the regulation of gene transcription and cell migration. In addition, RAAS activation results in increased signalling through the Rho and Rho Kinase pathways, which also participate in the development of hypertrophy, VSMCs migration, proliferation, inflammation, and hyperplasia in cardiovascular tissue.28
On the other hand, it has been demonstrated that insulin resistance related to aldosterone results in c-Src and ROS-mediated increase in proteasomal degradation of Insulin Substrate 1 (IRS-1) in VSMCs,29 demonstrating a novel mechanism of aldosterone-induced insulin resistance and further implicating mineralocorticoids in the pathogenesis of insulin resistance in the vasculature. Similar experiments30 have also confirmed that Ang II reduces IRS-1 levels in VSMCs through Src activation, and also through tyrosine phosphorylation of PDK1 and increased production of ROS that also result in abnormal serine phosphorylation and increased proteasomal degradation of IRS-1. In turn, decreased IRS-1 activity leads to impaired Akt activity and reduced glucose utilization in VSMCs.
Collectively, these studies demonstrate the involvement of RAAS, oxidative stress and aldosterone in vascular insulin resistance, and explain some of the beneficial actions of AT1R blockers and MR antagonists on insulin sensitivity and CVD (Fig. 3). Similar results have been obtained in type 2 diabetic patients, in whom increased ROS production, along with low levels of endogenous antioxidants such as glutathione and increased activity of NADPH oxidase have all been reported. Moreover, hyperglycaemia and insulin resistance are associated with impaired vascular vasodilatation and abnormal VSMC function that contribute to macro and micro vascular diabetic complications, including retinopathy, nephropathy, and constrictive remodelling of small arteries and atherosclerosis.
RAAS inhibition in the clinical setting
RAAS blockade through ACE inhibitors and Ang II receptor blockers (ARBs) have been studied extensively in HTN, congestive heart failure, coronary artery disease and CKD and are recommended to prevent CVD and nephropathy in patients with T2DM.31–33 In addition, some studies suggest a beneficial influence of RAAS blockade on insulin resistance and glucose homeostasis.31,34 Possible mechanisms responsible for the reduced incidence of diabetes in these trials include improvement in insulin mediated glucose uptake, enhanced endothelial function, increased nitric oxide activation, reduced inflammatory response, and increased bradykinin levels.35
On the other hand, it has been established that MR antagonism favourably affects cardiovascular outcomes in the clinical setting, as has been established in the RALES (Randomized Aldactone Evaluation Study) and EPHESUS (Eplerenone Post-Acute Myocardial Infarction Heart Failure Efficacy and Survival Study) trials.4,36
However, the impact of MR antagonism on insulin resistance and glucose homeostasis remains to be fully elucidated. From an experimental standpoint, it has been demonstrated that in vivo MR antagonism results in decreased NADPH oxidase activity and oxidative stress, in concert with improved insulin-stimulated glucose uptake as well as attenuated whole-body insulin resistance in skeletal muscle in a transgenic rodent model of inappropriately active tissue RAAS.37 Nonetheless, clinical studies have not always confirmed previous experimental and clinical observations suggesting improved insulin sensitivity through aldosterone reduction and/or MR blockade. In the EPHESUS trial, MR blockade significantly reduced hypoglycaemia by approximately 43%, which could indicate impaired glycaemic control.4 A recent randomized double-blind study assessing endothelial function and heart rate variability in 42 normotensive type 2 diabetic patients following 1 month of treatment with spironolactone reported a significantly higher glycohemoglobin A1c and plasma Ang II concentrations relative to placebo (0·26 ± 0·07% and 8·12 ± 1·94 pg/m, P < 0·05 respectively).38 In addition, in this trial forearm blood flow response to acetylcholine, a surrogate of endothelial function, and heart rate variability were significantly decreased in spironolactone-treated patients. It is possible that impaired glycaemic control and increased Ang II in patients treated with spironolactone contributed to worsening of endothelial function in this study.
In another clinical trial including 50 obese type 2 diabetic patients with poorly controlled HTN (> 140/80 mmHg), participants were randomized to spironolactone 50 mg daily or matching placebo.39 Average blood pressure at baseline was 162·7/88·9 mmHg, and was significantly reduced by spironolactone compared to placebo (mean reductions in systolic and diastolic blood pressures were 14·2 and 5·3 mmHg respectively). However, endothelium-dependent vasodilatation in response to acetylcholine was not significantly improved by spironolactone. Also in this study, spironolactone use was associated with worsened glycaemic control as demonstrated by a significantly higher glycohaemoglobin A1c (0·21% increment), as well as increased plasma Ang II (12·2 pmol/l, P < 0·05) and cortisol (92·35 nmol/l, P < 0·05). Finally, a small but significant increment in glycohemoglobin A1c consistent with impaired glycaemic control has also been reported in type 2 diabetics complicated by diabetic nephropathy despite a positive effect on systolic blood pressure.40
The discrepancy between clinical trials and observational studies in which there is a beneficial effect of MR blockade on glucose homeostasis can have several reasons. In the above mentioned studies, Davis et al. as well as Swaminathan et al. reported a significant increase in plasma Ang II concentrations,38 which is associated with insulin resistance and impaired glucose homeostasis.41,42 In addition, in these studies MR blockade was also associated with significantly increased plasma cortisol. As glucocorticoids and mineralocorticoids have similar affinity for MR, it is possible that MR blockade decreases clearance of cortisol, leading to impaired insulin action in numerous tissues such as skeletal muscle, adipose, hepatic and cardiovascular tissues,43 whole-body insulin resistance and impaired glucose homeostasis. Glucocorticoids in turn are known to cause insulin resistance through numerous mechanisms, including increased fatty acids oxidation,43 decreased insulin-mediated glucose uptake in skeletal muscle,44 decreased serine phosphorylation of Akt/protein kinase B (PKB),45 impaired insulin receptor tyrosine phosphorylation and insulin receptor substrate 1 (IRS-1) expression,46 as well as increased hepatic glucose output.47 The findings of augmented Ang II and cortisol during MR block might explain not only the deleterious effects on insulin sensitivity, but also the lack of benefit in terms of endothelial function observed in these studies.