JRSM >
Volume 99, Number 9 >
Pp. 457-462
Review |
Professor of Epidemiology, School of Population and Health Sciences, University of Newcastle upon Tyne, Medical School, Newcastle NE2 4HH, UK
E-mail: N.C.Unwin{at}newcastle.ac.uk
It has long been known that certain cardiovascular and metabolic risk factors tend to cluster together. For example, in the 1940s the French clinician, Vague, described the diabetic and atherogenic risk associated with android (upper body) obesity.
The relatively recent concept of a metabolic syndrome suggested that the clustering represents a discrete entity due to a single underlying aetiology (insulin resistance). Yet, despite several definitions and its own International Classification of Diseases code its validity and utility remains unclear.
The evidence does not fully support the existence of a single underlying aetiology, and most, but not all, evidence suggests that the syndrome as currently defined appears to have little value in predicting cardiovascular disease risk over other approaches that consider individual risk factors. At present there is no specific treatment for the syndrome other than targeting the individual risk factors.
It is concluded that at this time its main utility is as a reminder to the clinician that certain risk factors tend to cluster, and as an indicator of the public health importance of the risk factor clustering associated with obesity and physical inactivity.
INTRODUCTION
The metabolic syndrome (sometimes also known as syndrome X or insulin resistance syndrome) has been recognized since the late 1980s, arguably long before, and at the time of writing, there are at least five published definitions of it. Most recently the International Diabetes Federation has proposed ‘... a new worldwide definition’.1 Prospective data evaluating its prognostic importance for the development of both diabetes and cardiovascular disease have shown that it identifies high risk individuals.2,3 The metabolic syndrome even has its own ICD (International Classification of Diseases) code. Yet the apparent consensus on the existence and importance of the syndrome belies the considerable debate and uncertainty that remains.4-6 Questions remain on the underlying pathogenesis of the syndrome:
The aim of this paper is to provide a critical overview of the concept of the metabolic syndrome, and in particular to reach a conclusion, based on current evidence, on its clinical and public health utility.
METHODS
The paper starts by tracing some of the origins of the concept of the metabolic syndrome, as these are informative for today's debate. It then provides a description of some of the definitions of the syndrome, and reviews some of the analyses that have addressed the issue of clustering of proposed components of the syndrome. Finally, studies that have investigated the prognostic implications of some of the definitions are discussed, before attempting to reach a conclusion on the current utility of the concept of the syndrome. The article is based on a detailed Medline review that aimed to identify all articles concerned with the clustering of cardiovascular disease risk factors between 1966 and 2004, with other articles identified from the references of those reviewed. This literature search was updated to June 2005. In addition, the author also benefited from being part of the International Diabetes Federation (IDF) meeting in 2004 that reviewed the evidence on which the recent IDF definition is based.1
ORIGINS OF THE CONCEPT OF THE METABOLIC SYNDROME
Most authors on the metabolic syndrome relate their work to description by Reaven in 1988 of syndrome X or ‘the insulin resistance syndrome’.7 However, the roots of the concept run deeper and broader than what is covered in this description, and can be seen as the culmination of several strands of observational and experimental work over the second half of the last century. The common thread to this work was the observation that certain metabolic and biological characteristics, associated with an increased risk of diabetes and atherosclerotic disease, tend to cluster (occur together greater than predicted by chance) within individuals.
Broadly, two lines of work can be identified: one based on the epidemiological studies established to identify risk factors for cardiovascular disease; and the other based on clinical and experimental studies concerning the pathogenesis of diabetes and atherosclerosis.
Epidemiological studies
The early epidemiological studies established to identify risk factors for
cardiovascular disease initially described the importance of risk factor
concurrence. For example, the first major publication from the Framingham
study based on 6 years follow up (in which the term ‘risk factor’
was coined) identified raised blood pressure, raised serum cholesterol and
electrocardiograph evidence of left ventricular hypertrophy as risk
factors.8 It also
demonstrated that the concurrence of these risk factors tended to increase
risk in a multiplicative rather than an additive fashion. The multiplicative
nature of concurrent cardiovascular disease risk factors was subsequently
described in many other studies and is now considered established
fact.9-11
Explicit recognition and analysis of cardiovascular disease risk factor clustering, the concurrence of risk factors in individuals greater than predicted by chance (based on their prevalence in the underlying population), in published epidemiological studies is not found until the 1970s. Descriptions included the clustering of socio-economic, lifestyle (such as smoking and aspects of diet) and biological factors,12-14 as well as the clustering of biological factors alone, such as glucose, lipids, blood pressure and anthropometry.15,16 It was noted that these clusters of risk factors predispose individuals to the development of both diabetes and cardiovascular disease, and, based largely on epidemiological evidence, Jarrett proposed that ‘... atherosclerosis and diabetes share a number of antecedents, the balance between them determining the clinical outcome’.17
Clinical studies
Of the early clinical and experimental work it is that of Vague, initially
published in the late 1940s and 1950s, that is most clearly the forerunner of
today's concept of the metabolic
syndrome.18 His
work built on observations made several decades
earlier.19-21
Vague et al. described the strong relationship between android (upper
body) fat distribution and adult onset (type 2) diabetes, atherosclerosis and
gout. In contrast to android obesity, they concluded that although gynoid
(predominantly lower body) obesity is associated with mechanical complications
it ‘... does not exercise any influence on the metabolic disorders, such
as diabetes, gout... and atherosclerosis’. One of the potential
mechanisms discussed by Vague was that android obesity is associated with an
absolute or relative over activity of the hypothalamic pituitary adrenal axis,
one of the effects of which is to oppose the peripheral action of insulin.
Recognition of the potential pathophysiological importance of insulin resistance can be traced to Himsworth in the 1930s. He first recognized its potential importance in the pathogenesis of diabetes.22 Thirty to forty years later Reaven et al. confirmed the insulin resistant nature of clinical non-insulin dependent or maturity onset diabetes (i.e. type 2) and lesser forms of glucose intolerance.23-25 Reaven's group also described the fact that insulin resistance, and associated hyperinsulinaemia, is common in individuals with normal glucose tolerance.7 This observation led directly to a consideration of whether despite normal or at least non-diabetic glucose levels insulin resistance is related to other adverse consequences, and consideration of this question to the formulation of syndrome X or the insulin resistance syndrome.7
DEFINITIONS AND DESCRIPTIONS OF THE METABOLIC SYNDROME
Reaven's original description of ‘syndrome X’
In the 1988 Banting Lecture, Reaven used the term syndrome X to refer to
the tendency of glucose intolerance, hypertension, low high density
lipoprotein (HDL) cholesterol and raised triglycerides, and hyperinsulinaemia
to occur in the same
individual7
(Table 1). He proposed that the
common feature of the syndrome is insulin resistance and that ‘... all
other changes are likely to be secondary to this basic abnormality’.
Neither obesity nor abdominal obesity was included in Reaven's original
description, although secondary sources often state that they were included.
However, Reaven did suggest that avoiding obesity and remaining physically
active were measures that would protect against insulin resistance.
|
Recent descriptions and definitions
The term syndrome X failed to gain widespread acceptance. ‘Insulin
resistance syndrome’ is still used, but suggests a single underlying
mechanism. The term ‘the metabolic syndrome’ is more neutral, and
used to indicate the tendency of certain risk factors to cluster without tying
it to a single underlying mechanism. It is now the most widely used.
Until earlier this year there were two widely used definitions of the metabolic syndrome: that of the World Health Organization26 and that of the National Cholesterol Education Program, Adult Treatment Panel III.27 The components of each definition and criteria for making the diagnosis of the metabolic syndrome are summarized in Table 1. In addition, definitions were proposed by the European Group for the Study of Insulin Resistance28 and the American Association of Clinical Endocrinologists.29 Essentially, these are modifications of the WHO and NCEP definitions, respectively.
Recently the International Diabetes Federation has proposed a new definition (see Table 1) that it hopes will become the international standard.1 This definition is similar to the NCEP definition, being based on relatively simple measures, applicable in a clinical or epidemiological setting, but differs in three important respects. Central obesity, as determined by waist circumference, is mandatory, and different waist cut points for different ethnic groups are given based on available data linking waist circumference to other components of the syndrome. Finally, the IDF definition uses a lower fasting glucose level than the original NCEP definition, using the American Diabetes Association 2003 cut point for impaired fasting glucose.30
The prevalence of the metabolic syndrome
The metabolic syndrome by any of the definitions so far examined is very
common in adults in many parts of the world, typically being found from one in
six to one in three adults (sometimes more), rising with age and being higher
in men than
women.31
Figure 1 shows the prevalence
in adults (25-64 years) from three ethnic groups resident in Newcastle upon
Tyne, UK,32 with
differences between the ethnic groups mainly reflecting differences in the
prevalence of diabetes and obesity. Although common by both definitions,
agreement between the definitions is not particularly good. Thus, even in the
Europeans (see Figure 1) where
the overall NCEP and WHO prevalence was similar, between 20-30% of individuals
categorized as having the syndrome on one definition were not classified as
metabolic syndrome on the other. Studies reporting agreement between the new
IDF definition and others are awaited.
|
Multivariate analyses of risk factor clustering—one or more clusters?
Multivariate methods of analysis have been widely used to investigate the
clustering of defined and suggested components of the metabolic syndrome. The
most widely used technique has been factor analysis, although its value is not
without controversy and its interpretation often subjective. Nonetheless, it
can provide some insight into the underlying relationships between a set of
interrelated variables. A basic assumption of factor analysis is that it is
possible to explain the correlation between two or more variables by a common
underlying, but unmeasured, factor. It has been argued, for example, that if
the proposed components of metabolic syndrome tend to cluster because of a
single underlying cause, such as insulin resistance, they should all group
into one factor. In the vast majority of populations studied using factor
analysis (there are over 20 published studies) more than one factor has been
extracted. At a broad level the components of the factors were similar, with
for example, blood pressure and lipids in separate factors.
Figure 2 illustrates this by
summarizing the results from two of the earliest studies to apply factor
analysis to the metabolic
syndrome.33,34
Both studies found three, to some extent overlapping, factors. One
interpretation of the factors is that they represent underlying physiological
domains. The names given to each of the factors in
Figure 2 are those given by the
authors and reflect this interpretation. Overall the results of the factor
analyses are against the unity hypothesis that a single underlying determinant
(such as insulin resistance) is responsible for the
clustering.4 The
results suggest that more than one, possibly several, determinants contribute
to the interrelationships of glucose, insulin, lipids, blood pressure and
other related variables.
|
METABOLIC SYNDROME AND FUTURE DISEASE RISK
The current definitions of the metabolic syndrome are somewhat disjunctive, in the sense that it is possible for two individuals to be labelled as having the syndrome but to have different risk factor profiles. In the case of the NCEP definition two individuals labelled as having the metabolic syndrome need share only one of the five features in common (see Table 1). In the case of the WHO definition it is possible (although unlikely) that two individuals could have none of the different components in common. Thus, within these definitions there is considerable scope for heterogeneity and it is highly likely this will be associated with heterogeneity in the prognostic implications of the diagnostic label of metabolic syndrome. Certainly, applying the definitions to populations with different prevalences and patterns of the components of metabolic syndrome, such as is often the case between different populations, whether defined geographically or by ethnic group, is likely to result in different phenotypes.32
The weight of evidence strongly suggests that individuals with the metabolic syndrome (WHO or NCEP definitions, publications on the IDF definition are awaited) have a much higher risk for cardiovascular disease and diabetes than individuals without the syndrome. This is hardly surprising given that the components of the syndrome are well known risk factors. The estimates of the increased risk vary from 30-400%,5 probably mainly representing differences in the underlying populations studied and the length of follow up. In such studies it has been found that excess risk remains after adjustment for other conventional cardiovascular risk factors.5
Of importance to the clinical value of the metabolic syndrome is whether it provides more predictive information than considering its components individually, i.e. is it greater than the sum of its parts? Five studies that address this issue were recently reviewed,5 and none found that considering the components together (as a syndrome) had any greater predictive value than considering them separately. However, one recent study has suggested that the syndrome does provide predictive power for cardiovascular disease (CVD) events even after adjusting for its individual components.35
Also relevant to its clinical value is whether it provides a better alternative to other approaches to predicting CVD risk, such as those based on the Framingham risk equation and made widely available for use in clinical practice. Studies addressing this issue are limited and somewhat inconsistent in their findings, but overall suggest that the metabolic syndrome performs no better or worse than the Framingham risk equation. For example, in the San Antonio Heart Study, using data on subjects free of diabetes and CVD at baseline and followed for 7-8 years, the Framingham risk equation had a higher sensitivity for predicting cardiovascular disease events than the presence of the syndrome, and adding metabolic syndrome did not improve the prediction of events.36 Similar comparisons with alternative risk prediction scores, such as the European version (SCORE)37 would be worthwhile.
Finally, comparison to the Framingham, or other risk equations, highlights another potential shortcoming of the metabolic syndrome in any of its current definitions. In the Framingham and European equations different risk factors are given different weights - some contribute more to risk than others. In the metabolic syndrome each component is weighted equally.
CONCLUSIONS—SYNDROME OR TENDENCY, AND IS IT USEFUL?
The concept of the metabolic syndrome grew out of the recognition that certain risk factors tend to cluster together greater than predicted by chance. The clearest forerunner of the concept, particularly as recently defined by the IDF, is the description by Vague of diabetic and atherogenic risk associated with android obesity.18 The attempt to link the syndrome to a single underlying aetiology—insulin resistance—has in large part failed. The model of single cluster linked together by a single underlying mechanism (insulin resistance) does not match the empirical data. This is not to deny, of course, that insulin resistance is often associated with many of the proposed components of the syndrome. Thus current evidence does not support the use of the term ‘syndrome’ in the sense of a distinct clinical entity with a single underlying pathology. However, it may be justified in the way it is often used in epidemiology;38 and that is to indicate the more frequent coexistence of certain characteristics, in this case cardiovascular and metabolic risk factors, than would be expected by chance. More work is needed to fully understand the physiological basis of the tendency for certain risk factors to occur together.
The available studies do not suggest that the metabolic syndrome provides greater predictive value for cardiovascular disease than the sum of its parts, or that it performs better than other approaches to its risk prediction. In addition, it is worth noting that, at the time of writing, there are no specific interventions for the syndrome. Treatment of the syndrome entails treating the individual components, along with the type of lifestyle advice (healthy diet and regular physical activity) that is common to the prevention of many non-communicable diseases.
Given all of the above, does the concept of the metabolic syndrome as currently defined and understood have any utility, whether in clinical or public health terms? Perhaps its main value in clinical terms is to remind the practitioner that certain risk factors tend to cluster within individuals. Its main value in public health terms is to highlight the clustering of risk factors associated with obesity, particularly abdominal obesity, and physical inactivity. The prevalence of the metabolic syndrome within a population may provide a simple summary measure of the level of cardiovascular disease risk related to obesity and physical inactivity that can be used to compare different populations and to track changes over time. What the best definition of the syndrome would be to be used in this way requires further evaluation.
Footnotes
Competing interests None declared.
REFERENCES
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
