The so-called visceral obesity is now well-known to be associated to a higher mortality than the peripheral obesity, since it is linked to a higher prevalence of dysmetabolism and hypertension and endothelial dysfunction[3]. This evidence highlighted the importance of adipose tissue location in determining its unfavourable effects.

Obesity is characterized by an expanded adipose tissue mass[10] and, interestingly, as noted above, in overweight subjects a typical pattern of adipokines is present, which negatively affect metabolic situation and cardiovascular function[11].

Typically, the source of these substances are organs (the liver), and immune cells[12-14]. Most notably, several studies indicate that fat could either regulate the synthesis of these molecules or could be itself an immediate source.

The complexity of adipose tissue as an endocrine organ has been highlighted in several studies which have reported important differences among adipocytes from different depots. These characteristics may account for a differential contribution to obesity-related disorders[15].

Adipocytes frome brown adipose tissue (BAT) are mainly represented in fetuses and newborn and are implicated in thermogenesis[16,17]. A small amount of BAT persists in adults human AT[17].

On the other hand, white adipose tissue (WAT) represents a main kind of adipose tissue in adults[15], largely present in the subcutaneous region (SAT) or close and within the abdominal viscera (VAT).

Vague[18] first described a link between increased VAT and atherosclerosis, diabetes, and other diseases. This association might be explained by the increased production of mediators, acting with endocrine, autocrine and paracrine mechanisms[3]. Bigger VAT adipocytes in obese subjects are related to a higher mediators release, comparing to SAT[19,20]. These factors, adipocytokines-include molecules like tumor necrosis factor-α (TNF-α), leptin, adiponectin, resistin, plasminogen activator inhibitor-1 (PAI-1), apelin, interleukin-6 (IL-6), resistin, angiotensinogen, serum amyloid A (SAA), and C-reactive protein (CRP)[21].

Moreover, several inflammatory cells[19] are largely represented in the WAT stroma. These cells play an important role in tissue homeostasis, such as the clearance of necrotic adipocytes[18], increased in obesity. In particular, macrophages produce the majority of TNF-α and increase IL-6 and inducible-Nitric Oxide synthase expression. Then they were thought to be the primary source of the cytokines in adipose tissue[19,22,23]. However, we have demonstrated with an in vitro model, that the mature adipocytes fraction isolated from human adipose tissue is directly involved in both CRP and SAA release[24].

Interestingly, among visceral fat depots, another specific local fat depot has been studied in the last decade for its relationship with MetS and coronary artery disease (CAD), i.e., epicardial adipose tissue (EAT). EAT surrounds the heart, within the pericardium, it is in contact with coronary vessels (i.e., perivascular) and the surface of ventricles. It shows a higher rate of lipolysis and lipogenesis comparing to other fat depots[25]. EAT is involved in myocardial energy supply, thermoregulation, and protection of the cardiac autonomic nervous system as well as in the regulation of coronary vessel motion and lumen diameter[26].

In post-mortem series, epicardial fat have been reported to account for 20% of total heart weight[27]. Obesity is associated with an increase in the amount of EAT. Both total weight of epicardial fat and epicardial fat cells size correlate with body weight[28]. Moreover, epicardial fat assessed by echocardiography correlated significantly in multivariable analysis with key components of the MetS, including waist circumference, which is known to parallel the increase in VAT seen in MetS[29,30]. Then, the cardioprotective features of EAT might be somewhat blunted by the increase in cardiovascular risk carried by its pathological enlargement in obesity[31].

From a clinical point of view, several clinical and epidemiological studies have found an association between EAT and cardiometabolic risk factors and early stages of atherogenesis. The echocardiographic EAT evaluation was found to be associated with arterial stiffness and intima-media thickness (IMT), two indexes of subclinical atherosclerosis, in hypertensive patients in a study from our group[32]. Among the 459 patients examined, subjects with epicardial fat > 7 mm were older, had higher systolic, diastolic, and pulse pressure, increased left ventricular mass index, carotid IMT, diastolic and stiffness parameters compared with those with epicardial fat < 7 mm. Age, carotid IMT, and stiffness parameters were independently related to epicardial fat. These findings have been confirmed in a more recent study by Choi et al[33].

Importantly, the Framingham Heart Study[34] and Multi-Ethnic Study of Atherosclerosis[35] identified EAT volume as independent risk markers for CVD. Moreover, recently results of the Heinz Nixdorf Recall study, a population-based prospective cohort study, have been published[36]. Among the 4093 participants incidence of coronary events increased by quartile of EAT. Doubling of the EAT, measured by computed tomography (CT) scan, carried a 1.5-fold risk of coronary events after adjusting for cardiovascular risk factors and coronary artery calcium score.

Comparing to subcutaneous fat, EAT shows a more dense inflammatory cell infiltrate, predominantly represented by macrophages[37], and it produces highly atherogenic and inflammatory adipocytokines in patients with CAD[38].

EAT might then participate also in the inflammatory process within the atherosclerotic plaque. The incidence and severity of CAD and coronary calcification have been in fact associated with EAT thickness and volume[39].

A recent study assessed the relationship between EAT volume and plaque vulnerability in significant coronary stenosis using intravascular ultrasound. Authors found a positive correlation between EAT volume, measured by CT scan and the percentage of necrotic plaque tissue, and an inverse correlation between with the percentage of fibrous tissue[40]. Low-density lipoprotein (LDL) cholesterol level and EAT volume were independently associated with the percentage of necrotic plaque tissue. These findings are consistent with previous reports[41].

Moreover, EAT is associated also with microvascular dysfunction in the absence of obstructive CAD. In a recent study, patients underwent Rb-82 positron emission tomography to obtain standard myocardial perfusion index and myocardial flow reserve (MFR). EAT thickness, EAT volume and coronary calcium score values were higher in patients with impaired flow reserve, with EAT thickness showing the strongest negative correlation with MFR[42].

Leptin has been the first adipocytokine identified as the product of the ob gene in obese ob/ob mice[49], which participate in the signalling of fat stores[50].

Then, early studies on leptin focused on its role in obesity and potential therapies to control it. However, soon it became evident a broader biological role for leptin, including its potential implication in leading to cardiovascular complications in obese patients (Table 1). However, conflicting results on leptin role in CVD have been reported.

In animal and in vitro models, leptin promotes atherogenesis, through an increase in oxidative stress in endothelial cells[51], an increased platelet reactivity and thrombosis[52]. On the other hand, several studies reported opposite effects such as induction of nitric oxide production, with anti atherogenic properties[53].

Most in vivo studies in humans demonstrated that leptin levels are linked with cardiovascular risk factors, like hyperlipidemia or hypertension[54-57], and markers of endothelial dysfunction, thrombophilia and inflammation. Finally, several prospective trials described a link of leptin levels with atherosclerosis and myocardial infarction (MI)[58-61].

In the West of Scotland Coronary Prevention Study, including 377 male participants and 738 controls[59], a 20% increase in the incidence of CAD was associated with each standard deviation increase in leptin levels, even after adjustment for potential confouders. Sattar et al[54] found a moderate association, although non significant, between leptin levels and CAD risk. This association was attenuated by BMI, but continued to be significant with other validated risk factors, including several markers of inflammation. Also in female population no significant link between leptin plasma levels and CAD was noted[61]. On the other hand, Kappelle et al[62], in healthy men, demonstrated a positive association of leptinemia and leptin/adiponectin ratio with incident CVD, even after additional adjustment for several potential confounders, such as clinical risk factors, lipid and microalbuminuria.

In patients with known CAD, leptin might predict future cardiovascular events independent of other risk factors, including lipid status and CRP, according to several reports. Wolk et al[60] followed up 504 patients who had undergone coronary angiography for both stable angina (SA) and ACS for up to 4 years. In the field of ACS, admission plasma leptin levels is associated with the success of thrombolytic therapy in patients with STEMI presenting < 6 h[63]. In particular, authors found higher rates of thrombolysis failure in patients with basal plasma leptin levels ≥ 14 ng/mL, in comparison with patients with levels less than 14 ng/mL.

The association between plasma leptin and adiponectin levels and recurrent cardiovascular events after an ACS has been investigate in The Long-Term Intervention with Pravastatin in Ischaemic Disease study[64]. Leptin was a significant and independent predictor of recurrent cardiovascular events.

Consistent with this report, Rajendran et al[65] measured leptin levels, high-sensitivity C Reactive protein (hs-CRP) and IL-6 levels in patients with acute myocardial infarction (AMI), showing higher levels of leptin than control subjects.

The relationship between serial serum leptin levels measurement after thrombolysis for AMI and the degree of coronary atherosclerosis, coronary reperfusion, echocardiographic findings, and clinical outcome was investigated by Khafaji et al[66] in a small study. Leptin concentrations peaked 36 h after admission. Significant correlation of mean serum leptin with reduced ejection fraction and a trend for an increase in the mean serum leptin levels with increasing number of diseased vessels were found. However, there was no correlation between serum leptin levels and outcome or myocardial reperfusion.

Karakas et al[67] in a population-based case-cohort study within the MONICA/KORA studies. After adjustment for various confounding factors, neither increased leptin levels or low adiponectin were associated with the incidence of coronary events in healthy subjects. Moreover, the leptin/adiponectin ratio didn’t improve the ability of the single adipocytokine to predict incident CAD. In another study conducted among patients with ACS and controls, lipid profiles, leptin, pregnancy associated plamsa protein A (PAPP-A) and CRP levels were assessed as markers of plaque vulnerability[68]. Significantly higher levels of leptin, PAPP-A and high-sensitivity CRP (hs-CRP) were observed in cases. At the multivariable analysis, leptin was not independently associated with ACS, while a positive correlation between CRP and leptin concentrations was noted.

Higher adiponectin and lower leptin levels were found to be associated with a high incidence of adverse events also in a Japanese cohort after successful emergency percutaneous coronary intervetion for AMI[69]. A low leptin to adiponectin ratio remained a significant independent predictor of adverse events during long-term follow-up at the multivariable analysis.

Similiar observations came earlier from Ku et al[70] They found that subjects with low baseline leptin levels had higher subsequent CV events and death. Interestingly, although subjects with low leptin had fewer co-morbidities and more favorable metabolic and inflammatory profiles, they showed a worse prognosis than subjects with higher leptin levels. This could be an example of “reverse epidemiology”[71,72], whereby a predictor of disease becomes inversely associated with prognosis once the disease has developed. According with this idea, a second paper described an association between low leptin levels and cardiovascular death in patients with chronic kidney disease[73]. Moreover, Leptin is elevated in chronic CAD. Multiple reports have shown that leptin causes coronary vasodilatation, activates endothelial progenitor cells, prevents lipid accumulation, and protects against ischemia-reperfusion injury[53,60,74,75]. Then a relative leptin deficiency might explain poorer prognosis seen in subjects with established CAD. Finally, the lack of association between leptin and mortality, especially in patients with higher BMI, could be otherwise explained by leptin resistance[76].

Adiponectin is a well-described adipocytokine, traditionally reported as a protective factor with an anti inflammatory effect (Table 1)[11]. It is clear that its circulating levels decrease with weight gain and are inversely correlated to the amount of VAT, as illustrated in CT scan studies[11]. Interestingly, decreased adiponectin was associated with enhanced pro-inflammatory phenotype in EAT in patients with CAD[77]. As notet above about leptin, growing conflicting data on adiponectin levels are emerging, suggesting higher complexity of its role, than previously thought. This is particularly evident in the balance between obesity, cardiovascular effects, and prognosis.

Consistent with a putative protective role, Ouchi et al[78] first detected lower adiponectin levels in subjects with established CAD. Following this experience, several cross-sectional and prospective studies have confirmed an inverse correlation between plasma adiponectin levels and incidence, severity and outcome of CAD[79-82].

Recent studies, however, failed to demonstrate this correlation[83,84], or even showed a paradoxical link between higher adiponectin levels and negative events, especially in patients with known CAD or at high cardiovascular risk[85]. Moreover, Zhang et al[86] demonstrated higher adiponectin levels inpatients with stable CAD and inducible ischemia.

In the Pravastatin Or atorVastatin Evaluation and Infection Trial-Thrombolysis in Myocardial Infarction 22[87], plasma adiponectin was measured in 3931 subjects with Acute Coronary Syndrome. Adiponectin was negatively associated with age, diabetes, BMI, and triglycerides, while a positive link was noted with the risk of death, MI, and heart failure.

Seven hundred and thirty-five consecutive patients with STEMI treated with primary percutaneous coronary intervention were included in a report by Lindberg et al[88]. Plasma adiponectin was measured immediately before the procedure. Patients with highest adiponectin quartile had increased mortality compared to patients with low adiponectin. After adjustment for conventional risk factors high adiponectin concentration remained an independent predictor of all-cause mortality.

Also in a cohort from the Jackson Heart Study, conducted among African Americans, Adiponectin was associated with a higher risk of incident stroke in women[89].

This conflicting evidence has highlighted the complex role of adiponectin in the pathogenesis of CVD. Interestingly, the paradoxical association of adiponectin with a worse outcome was found in a population with a more advanced disease status, which might trigger a compensatory increase in adiponectin levels[71]. Otherwise, the worse long-term outcome could be the result of the more advanced disease status per se. Another possible explanation of these findings might be a condition of relative resistance to adiponectin, as suggested by animal models[85].

In two recent metanalysis of prospective studies, it has been shown that plasma adiponectin levels are not related to the risk of CAD or stroke in apparently healthy[90] and diabetic patients[91]. Another reason why this association could not be found might lie in adiponectin isoforms. In particular, several studies now consider the high molecular weight (HMW) form to be biologically active[92]. However, only few cohort studies have prospectively evaluated the association of HMW adiponectin with CAD[92-94].

Among 30111 women from the Nurses’ Health Study, high levels of total and HMW adiponectin, and HMW/total adiponectin ratio were associated with a lower risk of CAD. These associations were largely mediated by parameters related to glucose and lipid metabolism and inflammation, especially HDL-cholesterol levels[95]. In a study by Kunita et al[96], in 394 patients referred for computed tomography angiography (CTA), levels of plasma HMW adiponectin were evaluated. In patients with obstructive CAD HMW adiponectin was significantly lower than that in patients without. Furthermore, it was significantly associated with the disease extent and with characteristics of plaque instability, such as positive remodeling, low CT density and adjacent spotty calcium. In a recent nested case-control study conducted among 15566 free of CVD subjects, baseline total and HMW adiponectin and their ratio were examined[97]. After adjustment for matched variables and traditional risk factors, total and HMW adiponectin and their ratio were not associated with overall risk of CVD. However, the highest quartile for HMW adiponectin and HMW/total adiponectin ratio decreased risk of CVD compared with the lowest quartile among middle-aged individuals with high blood glucose, while this association was not seen among the elderly.

Resistin was discovered as a fat-derived molecule in obese mice, for its capacity of inducing insulin resistance[98,99]. The main source in animals were adipocytes, in particular, from abdominal depots[98]. The substantial lack of homology between human and mouse resistin genes made difficult to confirm these observation in humans. In particular, in humans resistin is produced mainly by stromal vascular cells, rather than adipocytes[100]. However, resistin expression has been reported in human WAT by preadipocytes[101]. Even the relationship between resistin and insulin resistance, overweight and DM type 2 was extensively reported in human with non consistent conslcusions[99,102-104].

Considering that macrophages are its main source in humans, a main pro-inflammatory role has been hypothesized for Resistin. Thus, several in vitro studies was conducted (Table 2), which illustrated that Resistin levels increased in response to endotoxin and proinflammatory cytokines administration[104]. Moreover, Chen et al[105] recently reported that ox-LDL induced resistin mRNA expression in cultured adipocytes. In a recent study from our group[106], resistin induced prothrombotic phenotype of human coronary artery endothelial cells (HCAECs). HCAECs incubated with resistin showed an upregulation of TF expression, and its activity was induced in a dose-dependent manner through the activation of NF-κB pathway.

In light of these evidences, resistin has been studied for its implications in CAD.

In apparently healty individuals from the European Investigation into Cancer and Nutrition-Potsdam Study, individuals in the highest quartile of resistin levels, compared with the lowest quartile, had a relevant increased risk of myocardial infarction but not of ischemic stroke[107]. This association persisted even after adjustment for CRP levels. In contrast, subsequent study described the independent link between resistin and the incidence of ischemic stroke within post menopausal women[108].

In a recent study of 6636 adults recruited from general population, after 3.5 years of follow-up, the group in the highest quintile of resistin plasma levels had a higher incidence of AMI[109]. The serum resistin concentrations were higher in women, and the associated increase in the risk of AMI based on the resistin level was also higher in women than in men.

In patients with known CAD, some cross-sectional and case-control studies showed higher plasma resistin levels than controls. Subject referring angina which had CAD at the coronary angiography had higher Resistin levels than patients without CAD[110]. Besides, resistin was also associated with coronary artery calcification at the CT scan[111]. Pischon et al[112] documented higher resistin levels in women with CAD compared with healty subjects from the CORA study. This link remained significant even after adjustment for several risk factors except the hs-CRP levels.

In contrast, no relationship between Resistin levels and CAD was found among 1161 subjects in the LURIC study[113]. Moreover, Resistin was not associated with cardiovascular mortality. Then, high resistin levels could simply mirror the presence of other established cardiovascular risk factors. However in the same study, an enhanced expression of adhesion molecules was found in association with increased resistin levels, highlighting a pathophysiological role in atherogenesis.

In a perspective study[114] comparing subjects with stable angina and subjects with unstable angina (UA), NSTEMI, and STEMI, higher resistin levels were found within subjects with ACS. Interestingly, an early rise in resistin levels was reported, at 3-6 h after symptoms onset. This increase lasted 6 and 12 h after.

CRP is an acute phase protein, member of the pentraxin family. Since it is a well-known marker of systemic inflammation[45], CRP was one of the first studied protein from its potential role in both pathogenesis and risk prediction of atherosclerosis. Subsequent studies showed that CRP is other than an innocent bystander of the inflammatory response associated with atherogenesis[45,115] (Table 2). In particular, together with the Adipocytokines, CRP charachterizes the chronic inflammatory status associated with obesity and MetS.

Interestingly, the adipose tissue has been described as producer of CRP[23,116]. In particular, we found that mature adipocytes are able to produce CRP, under inflammatory stimuli[23]. This finding was confirmed later in the experience by Anty et al[117], which demonstrated the expression of the CRP gene in adipocytes of obese subjects.

From a clinical point of view, high sensitivity assays are available to detect even low CRP concentration. Then high-sensitivity (hs) CRP has been largely evaluated as a suitable candidate for cardiovascular risk prediction. This idea was first supported by pioneering studies by Ridker et al[118], which demonstrated higher hs-CRP levels in apparently healthy subject who developed CV events during follow-up. In light of these results, Ridker et al[119] evaluated several risk prediction algorithms to improve the cardiovascular risk classification in apparently healthy American women. In particular, a simplified score, including hs-CRP (Reynolds risk score), was validated in this study, and subsequently in a male population[120].

Then the American Heart Association (AHA) and the CDC Consensus incorporated hs-CRP into the risk prediction strategy of cardiovascular diseases[121]. Measurement of hs-CRP is considered reasonable in the assessment of absolute risk for CAD in intermediate-risk individuals, with a Framingham risk score of 10% to 20%. This recommendation was confirmed in ACCF/AHA Guidelines in 2010[122] and in the recently published European Society of Cardiology Guidelines on Prevention[123].

Moreover, in a metanalysis[124] confirmed a role of hs-CRP for a better risk stratification of subjects at intermediate risk for CVD. In particular, for every 400 to 500 people screened for hs-CRP or fibrinogen level, one additional cardiovascular event could be prevented over a period of 10 years.

Results from the Justification for the Use of Statins in Primary Prevention: an International Trial Evaluating Rosuvastatin (JUPITER)[125] provided robust evidence of the association between inflammation and cardiovascular risk. subjects with LDL cholesterol below 130 mg/dL were treated with rosuvastatin vs placebo; patients at higher cardiovascular risk were identified by a hs-CRP level of 2.0 mg/L or higher. The Steering Committee stopped the trial after a median follow-up of 1.9 years due to striking benefit in patients treated with rosuvastatin (44% relative risk reduction of the primary end point and of hard outcomes).

Tehrani et al[126], recently investigate wheter inflammatory markers had an impact on the association of high density lipoprotein (HDL) cholesterol with CVD. In 3888 older adults without known CVD, authors evaluated CRP, IL-6, and lipoprotein-associated phospholipase A2 levels. CAD incidence was higher for higher levels of CRP, IL-6, and lower for higher levels of HDL-C. Compared to high HDL-C/low-inflammation categories, incident CAD was increased for those with high HDL-C and high CRP or highest IL-6 tertile. Then the protective relation of high HDL-C for incident CAD appears to be attenuated by greater inflammation.

Hs-CRP has also been studied for its potential role in the prediction of adverse outcome in patients with established CVD. Several studies clearly demonstrated an associationbetween hs-CRP and future acute coronary events in patients with SA[42].

However, conflicting report exsists about the additive benefit of measurement of hs-CRP. While data from Sinning et al[127] suggest that, in patients with established CVD, traditional risk factors are the most powerful predictors with only little information added by inflammatory markers (including CRP), on the other hand several studies[45,128] showed that hs-CRP independently predicts cardiac events in patients with ACS[45].

Moreover, patients with higher hs-CRP on admission for ACS had higher rate of impaired myocardial perfusion[129] and death.

Nakachi et al[130] reported that an hs-CRP elevation at admission and increase independently predicted 30-d events. In contrast, Bogaty et al[131] found that serial measurements of hs-CRP in ACS patients have only a modest predictive ability, which disappeared after adjustment for common clinical variables. However authors did not exclude patients with acute or chronic inflammatory diseases.

Among ACS patients of the FAST-MI, authors found that low fetuin-A and high hs-CRP concentrations were associated with cardiovascular death, even after adjustment for GRACE risk score[132]. In another study by Schaub et al[133] in 398 consecutive patients presenting with acute chest pain novel biomarkers like myeloperoxidase, MRP-8/14 and hs-CRP, provided incremental value in the risk stratification of these patients.

Apelin was first discovered in 1998[134], as the ligand of the so-called APJ receptor, a G-protein-coupled receptor (GPCR) identified in 1993 from a human genomic library. It is produced as preproapelin, then cleaved by an AT-converting enzyme to form several shorter C-terminal active peptides, i.e., apelin-13, -16, -17, -19, -36 and a pyroglutamate form (Pyr1 apelin-13)[135]. Since the absence of an immediately apparent ligand, APJ was first classified as an orphan GPCR. It shares 31% sequence homology with the human angiotensin II (AT II) type 1 receptor, which led to further characterization of the Apelin-APJ system.

Overall, the apelin system has several physiological roles, most notably in the cardiovascular system, hypothalamus and the adipo-insular axis[136], such as fluid homoeostasis, glucose homeostasis, feeding behaviour, regulation of vascular tone, cardiac inotropism and immunity. First studies about apelin-APJ system found both similar and opposite functions to those of the AT system[137]. The distribution of both receptors and peptides overlaps in the hypothalamus and vasculature[138]. Moreover, Apelin has been detected in adipose tissue[139] and it was found that it was both produced and secreted by adipocytes. Apelin has been then considered as a novel adipokine. Also APJ is present in human and mouse adipose tissue, both in isolated adipocytes and in the stromal vascular fraction[140].

Apelin expression in adipose tissue is regulated by nutritional status. In obese subjects, APJ-apelin expression is increased and this up-regulation could be reversed after diet or surgery-induced weight loss[141]. Moreover, changes in insulin levels might be involved for both apelin and APJ regulations in adipose tissue, according to the severity of insulin resistance[140]. A close relationship between apelin and insulin has been demonstrated both in vivo and in vitro. In cultured adipocytes, insulin treatment increased expression and secretion of apelin. Apelin expression in adipocytes is increased in various mouse models of obesity associated with hyperinsulinemia[139,142]. Interestingly, in highly insulin-resistant mice, such as db/db ones, APJ expression isn’t increased[143] and in studies conducted in type 2 diabetic subjects, the effect of insulin resulted completely blunted in adipose tissue[140].

Moreover, Apelin expression in adipose tissue is regulated also by TNF-α, gastro-intestinal inflammation, peroxisome proliferator-activated receptor gamma (PPARc) coactivator-1 a (PGC1a), Eicosapentaenoic acid-a polyunsaturated fatty acid from the omega-3 family, which all increase the apelin expression and secretion[142,144-146]. ATII exerts different effects on the expression of apelin, depending on the receptor involved: type 1 AT receptor mediates an increase of the apelin secretion, while type 2 receptor may reduce its production[147]. Interestingly, glucocorticoids modulate the production of apelin and its secretion from fat cells, simultaneously increasing ATII and suppressing apelin expression, suggesting a possible pathogenetic mechanism in obesity-related hypertension[148].

Since the increase in vascular density is essential for adipose tissue expansion, with endothelial cells actively promoting the development of preadipocytes and growth of mature adipocytes, apelin has been proposed to contribute to the development of new vasculature in expanding fat depot[149]. Several studies have demonstrated that apelin is a potent angiogenic factor[149], grossly equivalent to vascular endothelial cell growth factor and, like other angiogenic factors, its gene is upregulated under hypoxia condition[150]. Hypoxia induces expression and secretion of apelin in both human and murine adipocytes, through the hypoxia-inducible transcriptor factor 1a.

The first report in humans of plasma apelin concentrations was shown in obese and hyperinsulinemic subjects[139,141] where plasma apelin levels are increased. In morbidly obese patients with or without diabetes, apelin levels were only higher in the diabetic morbidly obese subjects[151]. However, reduced plasma apelin levels were described in obese subjects with untreated type 2 diabetes, compared to non-diabetic subjects and anti-diabetic treatment (rosiglitazone and metformin) was found to increase apelin concentration, with the improvement of glycemic profile[152,153].

Changes in apelin levels after weight loss or bariatric surgery in obese individuals were also investigated. Diet-induced weight loss decreases apelin levels in moderate obese women[154] but not significantly in patients with the MetS[155] or in obese children[156]. Bariatric surgery resulted in a significant decrease in apelin levels only in morbidly obese patients exhibiting impaired fasting glucose or type 2 diabetes before surgery[151]. Probably, obesity per se is not the main determinant of increased plasma apelin concentrations since circulating apelin levels are not necessary significantly correlated to the BMI in all the published studies[140].

The possible role of Apelin in the atherosclerotic process has been investigated. In studies by Liu et al[157], Apelin-13, the predominant circulating apelin isoform, significantly promoted intracellular cholesterol efflux and reduces macrophage foam cell formation, indicating a potential antiatherogenic function. Moreover, Kadoglou et al[158] have demonstrated lower apelin levels in patients with carotid atherosclerosis as compared to healthy controls, and that apelin increment is independently associated with atorvastatin-related carotid plaque stabilization. The same group reported considerably lower apelin concentrations in CAD patients in comparison with healthy controls[159]. This finding was confirmed in other studies conducted among subjects with stable angina (SA), and the plasma apelin levels were found to be negatively correlated with the severity of the disease[160].

A decrease of Apelin levels early after AMI has also been reported, with a progressive elevation over time, however reaching values lower than control subjects at 24 wk[161]. These observation were confirmed in patients with a first STEMI, where the reduction in apelin levels was independent from left ventricular dysfunction and outcome[162]. In comparison to asymptomatic CAD patients, plasmatic apelin were lower in ACS patients on admission, with a negative correlation with the severity of CAD[163].

A myocardial protective effect have been suggested from studies on the possible therapeutic use of apelin in CAD in animal models. Azizi et al[164] demonstrated in rat models of MI that post-infarct treatment with [Pyr1]-apelin-13 significantly attenuates myocardial damage, via the reduction of oxidative injury and enhancement of NO production. In addition, apelin-13 has been found to promote angiogenesis and ameliorate cardiac repair after AMI by a mechanism involving vascular progenitor cells[165].

However from a pathophysiological point of view, some conflicting data have been reported. For example, Rittig et al[166] observed that plasma apelin levels are not associated with early stages of atherosclerosis in young subjects prone to atherosclerosis and type 2 diabetes. Interestingly, other studies in animal models suggested a role of apelin-APJ system in vasculature oxidative stress. Furthermore, Apelin is upregulated in human atherosclerotic coronary artery and potently constricts human coronary artery[167]. Data by Jin et al[168] show that genetic defects in apelin/APJ pathway may confer a potential risk for CAD in Chinese hypertensive patients. These evidences underline the complex role of Apelin and its receptor in atherosclerosis.