The metabolic syndrome and cardiovascular disease

Abstract

The metabolic syndrome, which is very common in the general population, is defined by the clustering of several classic cardiovascular risk factors, such as type 2 diabetes, hypertension, high triglycerides and low high‐density lipoprotein cholesterol (HDL). Central obesity and insulin resistance, which are the two underlying disorders of the syndrome, are further risk factors for cardiovascular disease. Moreover, a panel of novel (non‐traditional) risk factors are ancillary features of the metabolic syndrome. They include biomarkers of chronic mild inflammation (e.g. C‐reactive protein, CRP), increased oxidant stress (e.g. oxidized low density lipoprotein, LDL), thrombophilia (e.g. plasminogen activator inhibitor‐1, PAI‐1) and endothelial dysfunction (e.g. E‐selectin). Therefore, subjects with the metabolic syndrome are potentially at high risk of developing atherosclerosis and clinical cardiovascular events.

In recent years several longitudinal studies have confirmed that subjects with the metabolic syndrome present with atherosclerosis and suffer from myocardial infarction and stroke at rates higher than subjects without the syndrome. The risk of cardiovascular disease (CVD) is particularly high in women with the syndrome and in subjects with pre‐existing diabetes, CVD and/or high CRP. However, an increased risk is already present in subjects with a cluster of multiple mild abnormalities. The risk related to the metabolic syndrome is definitely higher when subjects affected are compared to subjects free of any metabolic abnormality.

• Cardiovascular disease
• central obesity
• dyslipidaemia
• hypertension
• insulin resistance
• metabolic syndrome
• microalbuminuria
• type 2 diabetes mellitus

Historical background

It has been known for a long time that metabolic disorders such as adult‐onset diabetes (nowadays type 2 diabetes, T2DM), dyslipidaemia, gout and obesity often occur together and cluster with hypertension and cardiovascular disease (CVD). This notion is deeply rooted in the history of pioneers of medicine. However, the first descriptions in modern literature of what today is called metabolic syndrome were made by Kylin in Sweden 1, Vague in France 2, and Avogaro and co‐workers in Italy 3. A giant step in the development of the concepts surrounding the metabolic syndrome and a strong impulse to the research in the field came from Reaven 4, who was the first to provide a nosologic dignity to the syndrome and pointed out the importance of insulin resistance in its pathogenesis. Interestingly, Reaven excluded obesity from the features of the syndrome, whereas central (visceral) obesity became a common denominator of the syndrome in more recent years 5. In fact, in the diagnosis of the syndrome, a methodological switch occurred from a pathophysiologically oriented approach, requiring the intriguing and more or less accurate assessment of insulin resistance 6, to a clinically oriented approach, based upon the assessment of parameters easily available for any physician and easier to standardize 5,7.

Since the beginning of the story of the metabolic syndrome it seemed that CVD could be a potential complication in subjects affected. In fact, sparse clinical observations and, later, many epidemiological data clearly indicated that the metabolic disorders featuring the syndrome (e.g. diabetes and hypertension) were factors strongly predisposing to CVD. Therefore, it was postulated that subjects with the syndrome have a high cardiovascular risk. This concept was clearly expressed by Vague 2 in the 1940s and by Crepaldi and colleagues 3 in the 1960s. However, the unquestionable evidence came many years later, when largely accepted sets of diagnostic criteria for the syndrome were established, and when these criteria were used in large scale longitudinal studies. In fact, only in the last few years there was a clear demonstration of an increased incidence of cardiovascular events in the clinical condition that now is universally called metabolic syndrome but in the past has been called syndrome X 4, insulin resistance syndrome 8, deadly quartet 9 or with a variety of other names 10–12.

Key messages

• The metabolic syndrome, which is a very common clinical condition, is strongly associated with cardiovascular morbidity and mortality.

• The cardiovascular risk related to the syndrome is higher in women (about 4‐fold) than in men (about 2‐fold), and in subjects with pre‐existing diabetes or cardiovascular disease.

• The increased cardiovascular risk is well evident also when abnormalities composing the syndrome are mild, and is definitely higher (several‐fold) if it is calculated using as the reference category subjects without any of the disorders composing the syndrome and not subjects without the metabolic syndrome.

Classic features of the metabolic syndrome and cardiovascular disease

According to a document issued by a committee of experts of the World Health Organization in 1999, the diagnostic criteria of the metabolic syndrome are: impaired glucose regulation (impaired fasting glucose, IFG, or impaired glucose tolerance, IGT, or T2DM), insulin resistance, dyslipidaemia (high triglycerides and/or low HDL cholesterol), hypertension, obesity or central fat distribution, and microalbuminuria 6 (Table I). According to the experts of the National Cholesterol Education Program‐Adult Treatment Program III (NCEP‐ATPIII) the diagnostic criteria are: fasting hyperglycemia (IFG or T2DM), high triglycerides, low HDL cholesterol, hypertension and excess of central fat 7 (Table II). The latter criteria were recently considered valid, with few changes, by a group of experts gathered by the International Diabetes Federation 8 (Table III). In the next paragraphs, the most relevant data supporting the conclusion that each of these abnormalities predisposes to CVD are briefly reviewed. As a corollary, any cluster of these abnormalities is logically a clinical condition strongly predisposing to CVD.

Table I. Diagnostic criteria of the metabolic syndrome according to WHO 6.

Impaired glucose regulation (impaired fasting glucose, IFG; impaired glucose tolerance, IGT; type 2 diabetes mellitus) and/or

insulin resistance (under hyperinsulinaemic euglycaemic conditions, glucose uptake below lowest quartile for background population under investigation), with

two or more of the components listed below:

• impaired glucose regulation

• insulin resistance

• hypertension (systolic ⩾140 mmHg and/or diastolic ⩾90 mmHg and/or anti‐hypertensive treatment)

• dyslipidaemia (triglycerides ⩾1.7 mmol/L and/or HDL cholesterol <0.9 mmol/L in men and <1.0 mmol/L in women)

• central obesity (WHR >0.90 in men and >0.85 in women and/or BMI >30)

• microalbuminuria (urinary albumin excretion >20 µg/min or urinary albumin/creatinine ratio >30 mg/g)

Table II. Diagnostic criteria of the metabolic syndrome according to NCEP‐ATP III 7.

Any combination of three or more of the following clinical/biochemical abnormalities:

• fasting plasma glucose ⩾6.1 mmol/L or known diabetes

• systolic blood pressure ⩾130 and/or diastolic blood pressure ⩾85 mmHg and/or treatment

• plasma triglycerides ⩾1.7 mmol/L

• plasma HDL‐cholesterol <1.03 mmol/L in men and <1.29 mmol/L in women

• waist circumference >102 cm in men or >88 cm in women.

Table III. Diagnostic criteria of the metabolic syndrome according to IDF 8.

Central obesity (waist ⩾94 cm in Europid men and ⩾80 cm in Europid women; cut‐offs are different for other ethnic groups), plus

any two of the following four abnormalities:

• raised triglycerides (⩾1.7 mmol/L) or specific treatment for this abnormality

• reduced HDL cholesterol (<1.03 mmol/L in men; <1.29 mmol/L in women) or specific treatment for this abnormality

• raised blood pressure (systolic ⩾130 and/or diastolic ⩾85 mmHg) or treatment for previously diagnosed hypertension

• raised fasting plasma glucose (⩾5.6 mmol/L) or previously diagnosed type 2 diabetes

Impaired glucose regulation and cardiovascular disease

Diabetes mellitus is one of the main risk factors for atherosclerosis, as indicated by several studies, including ours 13,14. Accordingly, diabetes is associated with an increased cardiovascular morbidity and mortality. This was clearly documented by the Framingham Study 15, the Multiple Risk Factors Intervention Trial (MRFIT) 16, the Nurses' Health Study 17 and many other longitudinal surveys 18–20. Diabetes increases the risk of myocardial infarction about 2‐fold in men and about 4‐fold in women. It also increases 2‐ to 4‐fold the risk of stroke and 4‐ to 9‐fold the risk of peripheral vascular disease, especially in women. Moreover, after any acute cardiovascular event, the presence of diabetes makes the outcome poorer 21,22. Not surprisingly, CVD is the leading cause of death in diabetes, largely exceeding cancer or other diseases 23,24.

Interestingly, the pre‐diabetic states of IFG and IGT are already featured by an increased cardiovascular risk 25,26. This is the main explanation of why the improvement of glucose control in T2DM did not result in a remarkable reduction of cardiovascular disease 27. In fact, to achieve a substantial reduction of CVD in diabetic individuals their glycaemic control should be brought down to the condition of normoglycaemia. Accordingly, the plasma glucose values representing the cut‐off points of the condition of IFG were recently lowered to 100 mg/dL 28.

Obesity, central fat distribution and cardiovascular disease

Obese subjects experience CVD at a higher rate than non‐obese individuals 29,30. Nevertheless, there has been a great debate on the independent role of obesity in CVD. The debate is more academic than substantial because the true issue is whether obese people have an increased risk rather than whether obesity is a risk factor per se, independently of the disturbances commonly accompanying the excess of body fat. In fact, obesity (and also overweight) is often associated with T2DM, dyslipidaemia and hypertension 31,33 and, through these classic risk factors, it obviously conveys a higher cardiovascular risk. Anyway, there are studies claiming also a role of obesity independently of these factors 34,35, and this further strengthens the message that excess weight is deleterious and, therefore, should be prevented and corrected.

The pioneer studies on central fat distribution and CVD were carried out in Sweden more that 20 years ago. These studies showed that subjects with high waist‐to‐hip ratio (WHR) had an increased incidence of CVD 36,37 and T2DM 38,39, which is regarded as a special type of CVD by some authorities 40. In fact, subjects with a predominantly central fat distribution have a constellation of metabolic, haemodynamic and pro‐coagulant abnormalities 41,42, and are featured by insulin resistance 43,44. Recently, the focus was on waist as a risk factor because a high waist measurement is a good proxy of an excess of visceral adipose tissue 42,45,46, the fat depot which seems to convey the greater risk of developing metabolic disturbances and cardiovascular events 47,48.

Hypertriglyceridaemia and cardiovascular disease

There has been a great debate on the independent role of hypertriglyceridaemia in CVD, but in recent years a consensus was reached on the risk conveyed by this abnormality in the lipid profile 49,50. More recently, it has been reported that even mild hypertriglyceridemia can significantly increase the risk of myocardial infarction and stroke 51,52. The increased cardiovascular risk related to hypertriglyceridaemia is more evident in subjects with a concomitant hypercholesterolaemia 53, but the role of hypertriglyceridaemia in the so‐called atherogenic dyslipidaemia is no longer debated 54. The risk associated with hypertriglyceridaemia is lower than that associated with hypercholesterolaemia, but triglyceride‐rich lipoproteins (very‐low‐density lipoproteins, VLDL) seem to play a direct role in the vascular injury. Moreover, an excess of these particles has a crucial role in the genesis of small dense low‐density lipoproteins (LDL) 55, which are very susceptible to oxidation and are implicated in early stages of the atherosclerotic process 56, and also in the genesis of lower concentrations of high‐density lipoproteins (HDL) 57.

Low HDL cholesterol and cardiovascular risk

Nowadays, low HDL cholesterol stands out among major risk factors for CVD with a status not inferior to high LDL cholesterol. Strong evidence on the contribution of this abnormality in the lipoprotein profile to the risk of CVD came initially from the Framingham Study 58, but other longitudinal surveys confirmed this finding 59,60. Low HDL cholesterol is another feature of atherogenic dyslipidaemia 54, which is typical of subjects with central obesity and/or T2DM 61,62. HDL particles not only drive cholesterol from peripheral tissues (including the vasculature) to the liver 63 but also exert anti‐inflammatory and anti‐oxidant actions in the vascular wall 64.

Hypertension and cardiovascular risk

The Framingham Study 65 as well as the MRFIT 66 and many other studies 67–69 have documented the continuous increase in the risk of myocardial infarction and stroke with increase in blood pressure. More recently, it has been pointed out that even a small increase in blood pressure can predispose to clinical events 70. This is one of the reasons why the cut‐off values for diagnosing hypertension were recently reduced 71. Hypertension is one of the factors which more strongly damage the vascular wall, thus yielding an endothelial dysfunction 72. The latter is the first step in the atherosclerotic process 73 and can predict subsequent clinical vascular events 74,75.

Microalbuminuria and cardiovascular disease

A moderate increase in the urinary albumin excretion rate (30 to 300 mg/day) is considered a biochemical sign of incipient diabetic nephropathy and a risk factor for subsequent development of proteinuria and end‐stage renal disease in diabetes 76. However, it has been repeatedly reported that microalbuminuria can also be considered a marker of endothelial dysfunction 77. Accordingly, in both diabetic and non‐diabetic subjects microalbuminuria is a risk factor of CVD 78,79. Interestingly, an increased risk of atherosclerosis was observed also when albuminuria was in the top part of the range of normality 80.

Insulin resistance and cardiovascular disease

Insulin resistance is associated with T2DM, obesity, dyslipidaemia, hypertension 43,81–84. It is a common finding in several other clinical conditions and can be found also in healthy subjects 85,86. Several cross‐sectional studies reported an independent association between insulin resistance of glucose metabolism, as assessed by various techniques, and coronary, carotid or peripheral vascular disease 87–91. More recently, longitudinal studies suggested that insulin resistance can also predict CVD independently of classic risk factors in both non‐diabetic and diabetic subjects 92–94. In diabetic subjects we used homeostasis model assessment of insulin resistance (HOMA‐IR), a surrogate measure well correlated with the gold standard measure of insulin sensitivity 95, and found that for one standard deviation increase in (log)HOMA‐IR, the risk of developing a cardiovascular event was about 50% higher (OR 1.54, 1.14–2.12, P<0.001) 96. In the general population, subjects in the top quartile of distribution of (log)HOMA‐IR had a ∼80% increase in the risk for CVD (OR 1.77, CI 1.03–3.02, P = 0.038), after adjusting for classic and non‐traditional risk factors (Bonora et al. unpublished data).

Other features of the metabolic syndrome are novel cardiovascular risk factors

Over the last decade many investigators, including ourselves 97, reported that subjects with the metabolic syndrome have a wide spectrum of additional biochemical abnormalities (Table IV). The long list of ancillary features of the metabolic syndrome includes higher levels of PAI‐1, fibrinogen, coagulation factors VII and VIII, von Willebrand factor (vWF), apoprotein B, oxidized LDL, free fatty acids (FFA), urate, leukocytes, CRP, erythrocyte sedimentation rate (ESR), sialic acid, α‐1 acid glycoprotein, ferritin, endothelial adhesion molecules, homocystein, leptin, and lower levels of apoprotein A‐1, and adiponectin 97–114. Overall, these abnormalities document chronic mild inflammation, increased oxidant stress, thrombophilia and endothelial dysfunction. Interestingly, most of these abnormalities are associated with insulin resistance 97,115–120. Moreover, all of these abnormalities represent further risk factors for CVD and many of them are thought to contribute causally to the development of atherosclerosis and the occurrence of clinical CVD in subjects with the metabolic syndrome. In the next paragraphs the main evidence supporting these conclusions are briefly summarized.

Table IV. Ancillary (not diagnostic) features of the metabolic syndrome. These features are often but not necessarily found in subjects with the metabolic syndrome. The list is incomplete.

Raised levels	Reduced levels
PAI‐1
Fibrinogen
Factor VII, factor VIII
vonWillebrand factor
Apoprotein B	Apoprotein A‐1
Small dense LDL
Oxidized LDL
Free fatty acids
Urate
Leukocytes
CRP
ESR
Ferritin
Sialic acid
α‐1 Acid Glycoprotein
Amyloid A
TNF‐α
IL‐6
Homocystein
E‐selectin
P‐selectin
ICAM‐1
VCAM‐1
Leptin	Adiponectin
Resistin
Asymmetric dimethylarginine
GOT (AST)‐ GPT (ALT)
Gamma‐GT

Thrombosis, fibrinolysis and cardiovascular disease

Thrombus formation upon atherosclerotic plaques is often the precipitating factor in the occurrence of clinical cardiovascular events 121. Therefore a pro‐coagulant state and/or impaired fibrinolysis often play a key role in the development of myocardial infarction or stroke, the leading causes of death in western countries. Fibrinogen and PAI‐1 are among the several potential markers of a pro‐coagulant state and an impaired fibrinolysis. Many reports pointed out that higher fibrinogen levels represent a condition of increased cardiovascular risk 122–124. Also high PAI‐1 concentrations turned out to be a risk factor for CVD 125–127. In addition, other factors of coagulation, e.g. factor VII and factor VIII, are able to predict future cardiovascular events 128,129.

Apoprotein B, oxidized LDL, FFA and cardiovascular disease

Several studies clearly demonstrated that high concentrations of apoprotein B can predict CVD 14,58,60,69,130–132. In some studies high apoprotein B represented a risk factor even stronger than total or LDL cholesterol. Each LDL particle contains one molecule of apoprotein B. Therefore, the higher the number of circulating apoprotein B molecules, the higher is the number of LDL circulating particles. In the metabolic syndrome, serum LDL cholesterol is not significantly increased 97, whereas apoprotein B levels are higher 97. This means that the number of LDL particles is increased in subjects with the metabolic syndrome and that these particles are smaller and denser. Experimental data showed that small dense LDL is more susceptible to oxidation 56, and that oxidized LDL particles play a crucial role in the initiation and the progression of the atherosclerotic lesion 56,133. Accordingly, the circulating levels of small dense LDL and oxidized LDL are predictors of cardiovascular events 134,135.

Increased concentrations of FFA are typically observed in conditions of insulin resistance and excess visceral fat 136. The experimental elevation of circulating FFA impairs the endothelial function 137. As previously mentioned, endothelial dysfunction is a predictor of clinical cardiovascular events 74,75. Interestingly, higher FFA were associated with sudden death 138, which is a dramatic manifestation of CVD.

Urate and cardiovascular disease

Hyperuricaemia is a common finding in many clinical conditions, including obesity, dyslipidaemia, hypertension and T2DM 139. High serum urate is an independent risk factor for cardiovascular events 140–142. However, uric acid seems to participate in the anti‐oxidant defence of the body 143. This finding suggested that hyperuricaemia might be just a marker of an increased risk but without any causal role. On the other hand, a pathogenetic role of urate in the atherosclerotic process has been more recently postulated 144.

Inflammatory markers and cardiovascular disease

In the last decade a number of experimental studies documented that atherosclerosis is an inflammatory process 145,146. Accordingly, circulating markers of inflammation were reported to predict atherosclerosis and cardiovascular events. In particular many studies showed that CRP is a biomarker of future myocardial infarction as well as CVD morbidity and mortality 147–149, and that its power of prediction might be even stronger than that of LDL cholesterol 150. Other markers of inflammation were associated with atherosclerosis and CVD. They include leukocytes, ESR, ferritin, and others 151–155. Moreover, inflammatory cytokines like tumour necrosis factor‐α (TNF‐α) and Interleukin‐6 (IL‐6) have been associated with subsequent cardiovascular events 156,157.

Adhesion molecules and cardiovascular disease

It is well known that one of the first steps in the development of atherosclerosis is the adhesion of monocytes to endothelium and the subsequent migration of these cells in the subendothelial layer where they differentiate into macrophages 158. This process requires the interaction of circulating monocytes with endothelial adhesion molecules like E‐selectin, P‐selectin, intercellular adhesion molecule‐1 (ICAM‐1), vascular cell adhesion molecule‐1 (VCAM‐1) and others 159. These molecules are released into the circulating blood where they can be measured. Higher levels of these molecules are thought to represent endothelial dysfunction and progressing atherosclerosis. In recent years a number of longitudinal studies documented that serum levels of these molecules can predict atherosclerosis and CVD 160–163.

Adipokines and cardiovascular disease

Adipose tissue is now regarded as the greatest endocrine gland of the body. In fact, adipocytes secrete several molecules which act with autocrine, paracrine and endocrine mechanisms 164. Among the products of adipose cells are adipokines such as leptin and adiponectin, two molecules which also seem to have a strong impact on the vasculature. Adiponectin, in particular, possesses several anti‐inflammatory effects 165,166. On the contrary leptin seems to exert effects with a pro‐atherosclerotic potential 167,168. Preliminary studies indicated that adiponectin is a marker of cardiovascular disease with a protective meaning (the higher is adiponectin, the lower is the risk) 169, whereas higher leptin levels are associated with an increased risk 170.

Metabolic syndrome and cardiovascular disease

The metabolic syndrome is very common 97,171–174 and its prevalence and incidence are thought likely to increase in the future, paralleling the increase in the prevalence and incidence of obesity 175 and T2DM 176. At present, several hundred million people, and perhaps more than one billion, do have the metabolic syndrome throughout the world 177. A fraction of these people have a full‐blown syndrome, with most or all of the abnormalities featuring the metabolic syndrome well expressed (i.e. diabetes, hypertension, dyslipidaemia and obesity), others have only a cluster of mild abnormalities (e.g. IFG, abdominal adiposity and low HDL cholesterol) with a variety of possible clinical/biochemical phenotypes and with a phenotype in continuous evolution, depending on changes in life‐style and, as a consequence, in insulin sensitivity and body weight, the two major determinants of the metabolic syndrome 97,178.

The majority of subjects with the metabolic syndrome are thought to be at risk for CVD because most of classic and non‐classic clinical and biochemical features of the syndrome are cardiovascular risk factors. However, until an accepted set of diagnostic criteria for the syndrome has been established, it has been quite difficult to assess the true cardiovascular risk in these individuals. After the WHO criteria 6 and the NCEP‐ATP III criteria 7 had been issued, a number of studies examined the incidence of atherosclerosis and CVD in subjects with and without the metabolic syndrome according to these criteria. Quite disappointingly, in most of these studies modified WHO and/or modified NCEP‐ATP III criteria were used. Moreover, the endpoints were often different. Therefore, the results of these studies are not exactly comparable. In the next paragraphs the most relevant findings of these studies are presented, and in Table V data on coronary heart disease (CHD) are briefly summarized. More emphasis is given to the results of the Bruneck study because, within its frame, we have addressed several aspects related to the metabolic syndrome and we examined both atherosclerosis and clinical CVD.

Table V. Risk of CHD morbidity or mortality in subjects with the metabolic syndrome.

First author: (Study)	Country	No. of subjects	Years of follow‐up	WHO criteria OR/HR/RR	NCEP criteria OR/HR/RR
Bonora: (Bruneck study)	Italy	888	5	2.03 (1.2–4.3)	1.50 (NS)
Isooma: (Botnia study)	Finland	3606	7	1.81 (1.24–2.65)
Lakka: (Kuopio study)	Finland	1209	11	Men 2.87 (1.22–6.78)	Men 4.16 (1.60–10.8)
Sattar: (WOSCOPS)	Scotland	6447	5		Men 1.30 (1.00–1.67)
Hu: (DECODE)	Europe	11,512	9	Men 2.26 (1.61–3.17)
				Women 2.78 (1.57–4.94)
McNeill: (ARIC)	USA	12,089	11		Men 1.46 (1.23–1.74)
					Women 2.05 (1.59–2.64)
Hunt: (SAHS)	USA	2815	13	Men 1.15 (NS)	Men 1.82 (1.14–2.91)
				Women 2.83 (1.55–5.17)	Women 4.65 (2.35–9.21)
Malik: (NHANES)	USA	6255	13		2.02 (1.42–2.89)

OR = odds ratio; HR = hazard ratio; RR = relative risk. 95% C.I. are given in parentheses. See text for more details about the studies.

The Botnia study

This study examined Finnish subjects participating in a project focusing on metabolic defects in families with T2DM. A sub‐study on the metabolic syndrome analyzed data from 4483 subjects aged 30–70 years who had T2DM (n = 1697), IFG and/or IGT (n = 798) or normal glucose regulation (NGR) with various degree of insulin sensitivity (n = 1697) 179. Therefore this is not a population‐based study. In these subjects the metabolic syndrome was diagnosed with slightly modified WHO criteria: it was coded as present when insulin resistance (as assessed by HOMA‐IR) or abnormal glucose regulation (IFG, IGT or T2DM) was associated with two further abnormalities among obesity, hypertension, dyslipidaemia or microalbuminuria. During a median follow‐up of 6.9 years, cardiovascular death was assessed in 3,606 subjects. The relative risk (RR) of cardiovascular mortality associated with the metabolic syndrome was 1.81 (C.I. 1.24–2.65, P = 0.002), after adjusting for sex, age, LDL cholesterol and smoking.

The Kuopio ischaemic heart disease risk factor study

In this study 1209 Finnish men aged 42 to 60 years were randomly recruited from the general population and followed up for a median of 11.9 years 180. NCEP‐ATP III or modified WHO criteria (no microalbuminuria assessment; hyperinsulinaemia as a proxy of insulin resistance) were used for diagnosing the metabolic syndrome. Diabetic subjects were excluded from analyses. After adjusting for age, family history of CHD, smoking and LDL cholesterol, the RR of CHD mortality was 4.16 (1.60–10.8) when the metabolic syndrome was diagnosed by NCEP‐ATP III criteria, and 2.87 (1.22–6.78) when it was diagnosed by modified WHO criteria. The corresponding figures for CVD mortality were 2.52 (1.10–5.78) and 2.63 (1.37–5.05), respectively. All these RRs were highly significant.

The Bruneck study

This is a prospective population‐based survey examining 888 subjects aged 40–79 years of whom those fulfilling the WHO criteria or the NCEP‐ATP III criteria for the metabolic syndrome were identified 181. As compared with controls, subjects with the metabolic syndrome by WHO criteria had an increased 5‐year incidence and progression of carotid atherosclerosis, after adjusting for several confounders. Subjects with the metabolic syndrome by these criteria also had an increased incidence of CHD during follow‐up. The multiple‐adjusted odds ratio (OR) for incident carotid plaques and for incident carotid stenosis in subjects with the metabolic syndrome as compared to controls were 1.5 (95% confidence intervals 1.1–2.1, P = 0.02) and 2.4 (C.I. 1.3–4.1, P = 0.01), respectively. The multiple‐adjusted odds ratio for incident CHD in subjects with the metabolic syndrome was 2.3 (C.I. 1.2–4.3, P = 0.01). When NCEP‐ATPIII criteria were used the multiple‐adjusted risk of incident new plaques (OR 1.30, P = NS) and incident carotid stenosis (OR 3.1, P<0.001), as well as the multiple‐adjusted risk of incident CHD (OR 1.5, P = NS) were not superimposable but were similar to those we yielded when using WHO criteria. Interestingly, both subjects positive for the metabolic syndrome only by WHO criteria and those positive only by NCEP‐ATPIII criteria were at greater risk 182. Therefore, none of these sets of criteria seems superior to the other in the definition of the cardiovascular risk. However, NCEP‐ATP III criteria identified only 50% of subjects identified with the WHO criteria. Thus, a greater number of subjects at risk are identified when WHO criteria are used.

In our opinion, a major breakthrough related to the concept of the metabolic syndrome is the recognition of the high cardiovascular risk in subjects with a cluster of mild abnormalities or with a cluster of abnormalities that are not regarded as driving forces in CVD (e.g. overweight or higher plasma triglycerides). Accordingly, when we excluded from the analysis subjects with diabetes and/or definite hypertension (treatment and/or systolic blood pressure >160 and/or diastolic blood pressure >95 mmHg) and focused on subjects with mild abnormalities (IFG/IGT, mild hypertension, along with dyslipidaemia, central obesity, microalbuminuria and insulin resistance), we found that none of the individual components of the metabolic syndrome was an independent predictor of carotid atherosclerosis or CHD but that the metabolic syndrome was associated with a 4‐fold increased risk of atherosclerosis and CHD 182. These findings support the conclusion that a focus on the coexistence of multiple mild abnormalities allows one to identify a large number of subjects at risk of atherosclerosis progression and cardiovascular events who would be missed if the focus were limited to major risk factors (e.g. definite hypertension, diabetes, hypercholesterolaemia). As a consequence, it seems that there is a clear improvement in vascular risk prediction when using the ‘metabolic syndrome approach’.

The West of Scotland coronary prevention study (WOSCOPS)

Within this trial on primary prevention of CHD by pravastatin carried out in 6447 Scottish men, the cardiovascular risk related to the metabolic syndrome was calculated 182. Modified NCEP‐ATP III criteria were used (body mass index (BMI) replaced waist circumference as a marker of obesity). Diabetic subjects were excluded. During a 4.9 year follow‐up subjects with the metabolic syndrome had an unadjusted hazard ratio (HR) for CHD of 1.76 (1.44–2.15). The risk was slightly lower but still significant (1.30, 1.00–1.67, P = 0.045) when the model included as covariates age, smoking, systolic blood pressure, total cholesterol to HDL‐cholesterol ratio and pravastatin treatment. When subjects with no abnormality were the reference category, those with three abnormalities (i.e. with the metabolic syndrome) had an HR for CVD of 3.19 (1.98–5.12) and those with four abnormalities had an HR of 3.65 (2.11–6.33).

The women's health study (WHS)

In this intervention trial on primary prevention of CVD by aspirin and vitamin E, 14,719 apparently healthy women aged 45 years and over were followed‐up for 8 years 183. The metabolic syndrome was diagnosed with modified NCEP‐ATP III criteria (BMI replaced waist circumference) and women were stratified also for CRP (normal, high). As compared with women without the metabolic syndrome and with normal CRP (<3 mg/l), those with the metabolic syndrome had a RR of CVD events of 2.3 (1.6–3.3) when CRP was normal, and 4.0 (3.0–5.4) when CRP was high. The corresponding figures for CHD events were 3.1 (2.0–4.9) and 5.5 (3.8–8.0), respectively. As compared to subjects without any abnormality, the risk in women with three disorders (i.e. the metabolic syndrome) was 4‐fold higher when CRP was normal and 7‐fold higher when CRP was high.

The Verona diabetes complications study

Subjects with the metabolic syndrome, as defined according to WHO criteria, were identified among 946 non‐insulin‐treated type 2 diabetic patients examined within the Verona diabetes complications study 184. Mean age was 64 and mean duration of diabetes 9 years. At baseline and after a mean of 4.5 years follow‐up, CVD was assessed by medical history, physical examination, electrocardiogram (ECG) and echo‐duplex of carotid and lower limb arteries. Death certificates and medical records of subjects who died during the follow‐up were carefully scrutinized in order to identify CVD deaths. In statistical analyses, CVD was considered as an aggregate end‐point, including fatal and non‐fatal coronary, cerebrovascular and peripheral vascular disease as well as ischaemic ECG abnormalities and vascular lesions at the echo‐duplex. The proportion of subjects with the metabolic syndrome was very high (92.3%). Among subjects free of CVD at the baseline (n = 559), CVD events during the follow‐up were significantly increased in patients with the metabolic syndrome as compared with those without it (19.9 % versus 3.9%, P<0.001). Multiple logistic regression analysis showed that, along with sex, age, smoking and glycated hemoglobin (HbA1c), the presence of the metabolic syndrome independently predicted incident CVD and increased the risk of about 5‐fold (OR 4.89, 1.16–20.67, P = 0.031).

The DECODE study

This combined analysis of data from 11 prospective European cohort studies included 6156 men and 5356 women without diabetes and aged 30 to 89 years 185. The median follow‐up was 8.8 years and the metabolic syndrome was diagnosed with modified WHO criteria (microalbuminuria was not assessed, and hyperinsulinaemia was used as a surrogate of insulin resistance). Diabetic subjects were excluded. In subjects with the metabolic syndrome the HR for cardiovascular mortality was 2.26 (1.61–3.17) in men, and 2.78 (1.57–4.94) in women, after adjusting for age, smoking and total cholesterol. When hyperinsulinaemia was not considered and the diagnosis was made with an approach similar to NCEP‐ATP III criteria, the corresponding figures were 1.74 (1.19–2.55) in men, and 2.17 (1.13–4.19), in women, respectively.

The San Antonio heart study (SAHS)

This study was based upon the follow‐up (average 12.7 years) of 2815 Hispanic and non‐Hispanic white subjects from San Antonio, Texas, USA, aged 25 to 64 years 186. The metabolic syndrome was diagnosed by NCEP‐ATP III criteria. The HR for cardiovascular mortality was 1.82 (1.14–2.91) in men and 4.65 (2.35–9.21) in women, after adjusting for ethnicity and age. When modified WHO criteria were used (no microalbuminuria assessment; hyperinsulinaemia as a surrogate for insulin resistance) the corresponding figures were 1.15 (0.72–1.86) in men and 2.83 (1.55–5.17) in women. When subjects with CVD at baseline were excluded from analysis, a significant association of the metabolic syndrome with CVD was still found: NCEP‐ATP III criteria HR 1.81 (0.72–4.57) in men, and 3.93 (1.87–8.28) in women; WHO criteria HR 1.15 (0.65–2.06) in men, and 2.70 (1.36–5.37) in women.

The second national health and nutrition examination survey (NHANES)

Subjects of the Second NHANES (n = 6255; age 30–75 years) representative of 64 million adults in the United States were followed‐up for a mean of about 13 years 187. The metabolic syndrome was diagnosed by modified NCEP‐ATP III criteria (BMI was used instead of waist circumference; two‐hour OGTT along with fasting plasma glucose was used to identify impaired glucose regulation). After adjusting for gender, age, smoking, physical activity and LDL‐cholesterol, the HR for CHD mortality was 2.02 (1.42–2.89) and that for CVD mortality was 1.82 (1.40–2.37) in subjects with the metabolic syndrome as compared to subjects without the syndrome, diabetes or CVD at baseline. In those with the metabolic syndrome but without diabetes the corresponding figures were 1.65 (1.10–2.47) and 1.56 (1.15–2.12). When subjects without any risk factors were used as the reference category, the risk for CHD mortality in those with the metabolic syndrome was higher (HR 3.51, 1.81–6.81). In this analysis also having one to two risk factors increased the risk of CHD mortality (HR 2.10, 1.05–4.19). In subjects with pre‐existing diabetes, the risk associated with the metabolic syndrome was 3‐ to 5‐fold higher, depending on the reference category. The risk was 4‐ to 12‐fold higher in those with the metabolic syndrome and pre‐existing CVD, especially when diabetes was also present.

The 4S and AFCAPS/TexCAPS

Post‐hoc analyses of placebo data from the Scandinavian Simvastatin survival study (4S) and the Air Force/Texas coronary atherosclerosis prevention study (AFCAPS/TexCAPS) were used to estimate the long‐term relative risk of CHD associated with the metabolic syndrome, as assessed by modified NCEP‐ATP III criteria 188. Subjects with diabetes mellitus were excluded. Placebo‐treated patients with the metabolic syndrome had a risk of CHD about 1.5‐fold higher in both 4S (HR 1.5; 1.2–1.8) and AFCAPS/TexCAPS (HR 1.4; 1.04–1.9). Therefore, the metabolic syndrome was associated with increased risk of CHD both in patients with hypercholesterolaemia and pre‐existing CHD (4S) and in subjects with low HDL cholesterol but without CHD (AFCAPS/TexCAPS). Quite surprisingly, the risk was not different in the two categories.

The atherosclerosis risk in communities study (ARIC)

In this multicentre U.S. study 12,089 black and white individuals aged 45–64 years were followed‐up for an average of 11 years 189. The metabolic syndrome was diagnosed with NCEP criteria but subjects with diabetes or CVD at baseline were excluded. The risk of incident CHD was 1.5‐fold higher in men (HR 1.46, 1.23–1.74) and 2‐fold higher in women (HR 2.05, 1.59–2.64) with the metabolic syndrome, after adjustment for centre, race, age, smoking and LDL‐cholesterol. Similar association was found with stroke: HR in men 1.42 (0.96–2.11), HR in women 1.96 (1.28–3.00). As compared to subjects without any abnormality, women and men with three disorders (i.e. the metabolic syndrome) had a 4‐fold and a 2‐fold increase in CHD, respectively. The corresponding increase in the risk in subjects with four disorders was 7‐fold and 2.5‐fold in women and in men, respectively.

The Framingham offspring study

This survey included 3037 men and women aged 26 to 82 years who were followed up for 7 years 190 The metabolic syndrome was diagnosed by NCEP‐ATP III criteria. Subjects with diabetes and/or pre‐existing CVD were excluded. The end‐points were cardiovascular events, including angina, myocardial infarction, stroke, transitory ischaemic attack, heart failure and intermittent claudication. As compared to subjects without the syndrome, those with the metabolic syndrome had a sex‐, age‐ and CRP‐adjusted HR for CVD of 1.8 (1.4–2.5). The risk was higher in women (2.4, 1.1–5.4) than in men (1.8, 1.2–2.6).

The cardiovascular health study (CHS)

A total of 2175 American subjects from the cardiovascular health study who were free of CVD at baseline and were not taking antihypertensive or lipid‐lowering medications were followed up for about 4 years 191. The metabolic syndrome was assessed with modified‐WHO and with NCEP‐ATP III criteria. When the metabolic syndrome was defined with the latter criteria, the HR for CVD (coronary and cerebrovascular events) was 2.04 (1.69–2.46), and when it was diagnosed with WHO criteria the HR was 1.63 (1.33–2.01), after adjusting for sex, age, smoking, LDL cholesterol and family history of myocardial infarction. Interestingly, when individual components of the syndrome were included in the model, the HR still was significant (HR 1.38; 1.06–1.79, P<0.01). This means that classic components of the syndrome do not explain all the risk it conveys and, therefore, that the assessment of its individual components without focusing the presence of the syndrome might underestimate global risk of a given individual.

The Hoorn study

In this population‐based Dutch study a cohort of 615 men and 749 women aged 50 to 75 years were followed up for 10 years to evaluate fatal and non‐fatal CVD 192. The metabolic syndrome was diagnosed by NCEP‐ATP III and modified WHO criteria, and also with those proposed by European Group for the Study of Insulin Resistance (EGIR) 193, and American College of Endocrinology (ACE) 194. Subjects with diabetes and CVD at baseline were excluded. The NCEP‐ATP III definition was associated with about a 2‐fold increase in age‐adjusted risk of fatal CVD in men and non‐fatal CVD in women. For the WHO, EGIR, and ACE definitions, the HR were slightly lower. The risk increased with the number of risk factors.

Studies with factor analysis

A number of studies used factor analysis to evaluate whether the cluster of abnormalities featuring the metabolic syndrome can predict CVD. In these studies plasma insulin was assessed and used as a proxy of insulin resistance. An insulin resistance factor was generated by treating variables with factor analysis. This factors generally included insulin, glucose, triglycerides, BMI, WHR and blood pressure, which are among the variables used to diagnose the metabolic syndrome. Such an insulin resistance factor was an independent predictor of subsequent CHD and/or stroke in both non‐diabetic and diabetic subjects. In elderly non‐diabetic men HR for CHD associated with insulin resistance factor was 1.33 (1.08–1.65) 195. In middle‐age non‐diabetic men HR for CHD associated to insulin resistance factor was 1.28 (1.10–1.50) and HR for stroke was 1.64 (1.29–2.08) 196. In diabetic men HR for CHD associated with insulin resistance factor was 1.71 (1.08–2.71) 197. The results of these studies are important because the components of the metabolic syndrome were modelled as continuous variables and no cut‐offs were arbitrarily imposed. Moreover, these studies pointed out the potential role played by insulin resistance as a central underlying disorder.

Conclusions

The risk of CVD in subjects with the metabolic syndrome is 2‐ to 4‐fold higher than in subjects without the syndrome. The risk associated with the syndrome is higher in women (about 4‐fold) that in men (about 2‐fold). The risk is particularly high when affected subjects have pre‐existing diabetes and/or CVD and/or chronic mild inflammation (high CRP). Of note is that the cluster increases the risk even when major risk factors, such as diabetes and hypertension, are not present. However, it should be underlined that in most studies carried out so far the reference category was subjects without the syndrome and not subjects without any risk factor. In fact, many among subjects without the syndrome have one to two disorders and this misleadingly attenuates the increase in the risk in subjects with the metabolic syndrome. Interestingly, when subjects without any disorders and subjects with multiple disorders (three or more, i.e. with the metabolic syndrome) were compared, the latter had a risk of CVD several‐fold higher. Therefore, the metabolic syndrome is certainly a high risk condition but the true risk of subjects affected is underestimated when a dichotomous categorization is made (metabolic syndrome yes versus no). This concept should be carefully considered when criticisms of the syndrome are made 198. In fact, the use of an appropriate reference category is crucial before stating that the risk of CVD in subjects with the metabolic syndrome is not greater than the sum of the risk related to its individual components. In this regard, the evaluation of the number of risk factors clustering in the given subject, and the comparison of these subjects to those free of any disorder, give a better assessment of true individual risk. Nevertheless, the ‘metabolic syndrome approach’ remains valid because it provides a nosologic dignity for a very common clinical condition which is often neglected when it is composed of multiple minor disorders (central adiposity, impaired fasting glucose, etc.). The clustering of these minor disorders, however, is still a risky condition. Moreover, the metabolic syndrome approach points out the existence of underlying pathogenic disorders of the cluster, i.e. central obesity and insulin resistance. These underlying disorders can be targeted by specific interventions in order to prevent the metabolic syndrome in those not affected yet, and CVD in those already affected. The notion that treating the underlying disorders, i.e. obesity and/or insulin resistance, with lifestyle changes or medication resulting in an amelioration of several classic and ancillary components of the metabolic syndrome 199–211 and, therefore, in a substantial reduction of the cardiovascular risk, deserves great emphasis.