Adult Treatment Panel III 2001 but Not International Diabetes Federation 2005 Criteria of the Metabolic Syndrome Predict Clinical Cardiovascular Events in Subjects Who Underwent Coronary Angiography

The design of this prospective cohort study has been described in detail earlier (19,20). In brief, we enrolled 756 consecutive Caucasian patients referred to coronary angiography for routine evaluation of established or suspected coronary artery disease (CAD). Six patients with type 1 diabetes (C-peptide negative) were excluded from the analyses. The ethics committee of the University of Innsbruck approved the present study, and all participants gave written informed consent.

At baseline, coronary angiography was performed with the Judkins technique. Stenoses ≥50% were considered significant, and coronary arteries were defined as angiographically normal in the absence of any visible lumen narrowing at angiography, as described previously (21,22).

From the 750 patients included in the analyses, 164 had diabetes according to World Health Organization criteria (23) and 586 did not have diabetes. Among patients with diabetes, 42.7% were not receiving any antidiabetic medication and 34.8, 32.3, 24.4, and 1.2% were receiving—alone or in combination—sulfonylurea, biguanides, insulin, and α-glucosidase inhibitors, respectively. Overall, 64.7% of our patients were on aspirin, 31.3% on statins, 2.9% on fibrates, 12.1% on calcium antagonists, 47.1% on β-blocking agents, 36.0% on ACE inhibitors, and 3.9% on angiotensin II blocking agents.

According to ATPIII criteria (4), the metabolic syndrome was diagnosed in the presence of any three of the following: waist circumference >102 cm in men and >88 cm in women, triglycerides ≥150 mg/dl (1.7 mmol/l), HDL cholesterol <40 mg/dl (1.0 mmol/l) in men and <50 mg/dl (1.3 mmol/l) in women, blood pressure ≥130/≥85 mmHg, or fasting glucose ≥110 mg/dl (6.1 mmol/l). Using IDF criteria (10), the metabolic syndrome was diagnosed in patients who had a high waist circumference (≥94 cm in men and ≥80 cm in women) plus any two of the following: triglycerides ≥150 mg/dl (1.7 mmol/l) or specific treatment for this lipid abnormality, HDL cholesterol <40 mg/day (1.0 mmol/l) in men and <50 mg/dl (1.3 mmol/l) in women or specific treatment for this lipid abnormality, systolic blood pressure ≥130 or diastolic blood pressure ≥85 mg/dl or treatment of previously diagnosed hypertension, and fasting plasma glucose ≥100 mg/dl (5.6 mmol/l) or previously diagnosed type 2 diabetes.

During a follow-up period of 3.9 ± 0.8 years, we recorded fatal and nonfatal cardiovascular events, including coronary death (fatal myocardial infarction, sudden cardiac death, and mortality from congestive heart failure due to CAD), fatal ischemic stroke, nonfatal myocardial infarction, nonfatal ischemic stroke, and need for coronary artery bypass grafting, percutaneous coronary intervention, or noncoronary revascularization. Angina pectoris per se was not included among the end points.

Differences in baseline characteristics were tested for statistical significance with the χ² test and the Mann-Whitney U test for categorical and continuous variables, respectively. The Wilcoxon-Gehan statistic was used to compare differences in the cumulative incidence rates of vascular events. Adjusted hazard ratios (HRs) for the incidence of vascular events were derived from Cox proportional hazards models; for these calculations, continuous variables were z transformed. Results are given as means ± SD if not denoted otherwise. All statistical analyses were performed with the software package SPSS 11.0 for Windows.

From the total study population, 280 patients (37.3%) met the ATPIII criteria and 341 patients (45.5%) the IDF criteria for the metabolic syndrome. Baseline characteristics of patients with respect to the metabolic syndrome status according to the ATPIII and IDF definitions are summarized in Table 1.

From the individual ATPIII metabolic syndrome criteria, the blood pressure criterion was most frequently (in 85.4%) met in patients with the metabolic syndrome according to the ATPIII definition, followed by the criteria for triglycerides (75.0%), fasting plasma glucose (70.4%), HDL cholesterol (68.9%), and waist circumference (59.3%). From the patients with the metabolic syndrome according to the IDF definition, 100, 89.4, 83.9, 59.2, and 52.2% fulfilled the IDF criteria for waist circumference, blood pressure, fasting plasma glucose, triglycerides, and HDL cholesterol, respectively.

Concordance between the ATPIII and IDF definitions was moderate (Cohen-κ coefficient = 0.522; P < 0.001). From our 280 patients (37.3% of the total study population) with the metabolic syndrome according to the ATPIII definition, 223 (79.6%) also had the metabolic syndrome according to the IDF definition; from the 470 patients who did not have the ATPIII metabolic syndrome (62.7% of the total study population), 352 (74.9%) also did not fulfil the criteria of the IDF definition (online appendix [available from http://care.diabetesjournals.org]). Thus, 57 patients (7.6% of the total study cohort) fulfilled the ATPIII criteria but not the IDF criteria (ATPIII+/IDF− group), and 118 patients (15.7% of the total study cohort) fulfilled the IDF criteria but not the ATPIII criteria (ATPIII−/IDF+ group).

By definition, all ATPIII−/IDF+ patients met the IDF waist criterion. No ATPIII+/IDF− patient fulfilled the ATPIII waist criterion. ATPIII−/IDF+ patients less frequently than ATPIII+/IDF− patients met the respective triglycerides (30.5 vs. 86.0%; P < 0.001) and HDL (24.6 vs. 77.2%; P < 0.001) criteria. There were no significant differences in the prevalence of the respective blood pressure (83.9 vs. 84.2%; P = 0.958) and glucose (78.0 vs. 68.4%; P = 0.173) criteria between these subgroups.

During a mean (±SD) follow-up time of 3.9 ± 0.8 years, we recorded 160 vascular end points, encompassing 42 coronary deaths, 12 fatal ischemic strokes, 13 nonfatal myocardial infarctions, 14 nonfatal ischemic strokes, 23 coronary artery bypass graftings, 33 percutaneous coronary interventions, and 23 revascularizations at the carotid (n = 6) or peripheral (n = 17) arteries. Thus, the overall incidence rate was 21.3%.

Event-free survival was significantly lower (P < 0.001) in patients with the metabolic syndrome according to the ATPIII definition than in patients who did not meet the ATPIII criteria for the diagnosis of the metabolic syndrome (Fig. 1A). In Cox regression analyses adjusting for age, sex, smoking, LDL cholesterol, and baseline CAD, the ATPIII metabolic syndrome proved strongly and independently predictive of vascular events, with an adjusted HR of 1.745 (95% CI 1.255–2.427; P = 0.001). After additional adjustment for type 2 diabetes, this HR decreased, but the metabolic syndrome (ATPIII definition) remained predictive of vascular events (1.578 [1.114–2.234]; P = 0.010). Further adjustment for treatment with medications did not substantially change the power of the metabolic syndrome (ATPIII definition) as a predictor of vascular events (Table 2).

When patients receiving treatment for individual metabolic syndrome traits were considered to meet the respective criteria for these traits, the thus-modified ATPIII definition of the metabolic syndrome significantly predicted vascular events after adjustment for age, sex, smoking, LDL cholesterol, and baseline CAD (adjusted HR 1.810 [95% CI 1.299–2.520]; P < 0.001). Also, a modified ATPIII definition that was recently proposed by the American Heart Association and the National Heart, Lung, and Blood Institute (7), which considers treatment and uses a fasting glucose of 100 mg/dl as a cutoff for the high-glucose criterion, significantly predicted vascular events in our patient cohort (1.634 [1.173–2.276]; P = 0.004). Further, the metabolic syndrome as defined by World Health Organization criteria (3) proved significantly predictive of vascular events (1.405 [1.011–1.953]; P = 0.043).

In patients with the metabolic syndrome according to the IDF definition, event-free survival was not significantly different (P = 0.283) from patients who did not fulfil the IDF criteria for the diagnosis of the metabolic syndrome (Fig. 1B). Also, in Cox regression analysis adjusting for age, sex, smoking, LDL cholesterol, and baseline CAD, the IDF metabolic syndrome was not significantly associated with vascular events, with an adjusted HR of 1.189 (95% CI 0.859–1.646; P = 0.297).

The incidence of vascular events was significantly lower (P = 0.012) among ATPIII−/IDF+ patients than among ATPIII+/IDF− patients (Fig. 1C). Cox regression analysis adjusting for age, sex, smoking, LDL cholesterol, diabetes status, and baseline CAD confirmed that ATPIII+/IDF− patients had a significantly worse prognosis than ATPIII−/IDF+ patients, with an adjusted HR of 2.073 (95% CI 1.012–4.246; P = 0.046).

Table 2 summarizes the cardiovascular risks related to the dichotomous diagnosis of the metabolic syndrome and to its single components. In Cox regression analyses adjusting for age, sex, smoking, LDL cholesterol, diabetes, baseline CAD, and for the presence of individual metabolic syndrome component features, the metabolic syndrome as defined by ATPIII criteria remained significantly predictive of vascular events after adjustment for the ATPIII metabolic syndrome traits high fasting glucose (HR 1.530 [95% CI 1.045–2.239]; P = 0.029), high blood pressure (1.500 [1.041–2.163]; P = 0.030), and high waist circumference (1.865 [1.265–2.747]; P = 0.002). However, after adjustment for the metabolic syndrome traits high triglycerides or low HDL cholesterol, the ATPIII metabolic syndrome no longer was predictive of vascular events (1.251 [0.825–1.899]; P = 0.292 and 1.205 [0.806–1.801], respectively; P = 0.363). In contrast, the individual metabolic syndrome traits high triglycerides and low HDL cholesterol significantly predicted vascular events even after adjustment for the overall diagnosis of the ATPIII metabolic syndrome, with adjusted HRs of 1.502 (1.006–2.244); P = 0.047 and 1.686 (1.130–2.514); P = 0.011, respectively. In a multivariate analysis including both the clinical category of the metabolic syndrome (ATPIII definition) and at the same time all component features from the ATPIII definition, no significant association remained between the metabolic syndrome and vascular events (adjusted HR 1.167 [95% CI 0.626–2.178]; P = 0.627). From the individual metabolic syndrome criteria, only the low HDL feature independently predicted vascular events in this model (1.755 [1.156–2.664]; P = 0.008).

In analyses focusing on specific secondary study end points, the ATPIII metabolic syndrome after adjustment for age, sex, smoking, LDL cholesterol, and baseline CAD proved significantly predictive of end points related to CAD (n = 111), with an adjusted HR of 1.708 (1.158–2.519; P = 0.007). The IDF metabolic syndrome in contrast was not associated with CAD-related end points (0.945 [0.642–1.389]; P = 0.772). We did not observe significant associations between either metabolic syndrome category and the smaller number (n = 32) of end points related to cerebrovascular disease (1.333 [0.553–3.213]; P = 0.521 and 0.758 [0.311–1.848]; P = 0.542 for the ATP and IDF definitions, respectively).

Among the 509 men included in our analyses, 196 (38.5%) fulfilled the ATPIII criteria for the definition of the metabolic syndrome and 226 (44.4%) fulfilled the IDF criteria. Among women (n = 241), 84 (34.9%) and 115 (47.7%) met the ATPIII and IDF metabolic syndrome definitions, respectively. The metabolic syndrome as defined by ATPIII criteria after adjustment for age, smoking, LDL cholesterol, and baseline CAD proved significantly predictive of vascular events among both men (adjusted HR 1.625 [95% CI 1.111–2.377]; P = 0.012) and women (2.321 [1.179–4.567]; P = 0.015). In contrast, the metabolic syndrome as defined by the IDF criteria predicted vascular events neither in men nor in women (1.204 [0.829–1.748], P = 0.330, and 1.276 [0.648–2.513], P = 0.481, respectively). There was no significant difference in the impact of the metabolic syndrome on vascular events between men and women, irrespective of whether the ATPIII criteria or the IDF criteria were applied (P_interaction = 0.313 and 0.903, respectively).

When patients with diabetes (n = 164) were excluded from the analyses, of the remaining 586 nondiabetic patients, 168 (28.7%) had the metabolic syndrome according to ATPIII criteria and 229 (39.1%) met the IDF criteria for the diagnosis of the metabolic syndrome. Consistent with the total study cohort, the metabolic syndrome as defined by ATPIII criteria significantly predicted vascular events in nondiabetic patients (HR 1.531 [95% CI 1.007–2.328]; P = 0.046), whereas the metabolic syndrome as defined by IDF criteria did not significantly predict vascular events in these patients (1.139 [0.765–1.695]; P = 0.522).

From our results, we conclude that among angiographied coronary patients the ATPIII definition and the IDF definition of the metabolic syndrome do not identify the same patients. The metabolic syndrome as defined by ATPIII criteria confers a higher risk of vascular events than the metabolic syndrome as defined by IDF criteria.

The prevalence of the metabolic syndrome defined according to the IDF definition in our investigation was higher than the metabolic syndrome defined according to the ATPIII definition. Concordantly, Athyros et al. (24) have found a higher prevalence of the metabolic syndrome diagnosed according to IDF criteria in a Greek population–based study, and Ford (25) described a higher prevalence of the IDF metabolic syndrome among adults in the U.S. A recent investigation from northern Mexico reported a high concordance between the ATPIII and IDF definitions of the metabolic syndrome (26). However, in 23.3% of our Caucasian patients the two metabolic syndrome definitions led to discordant diagnoses; thus, concordance was moderate.

The IDF wanted its new definition to encourage the identification of patients at increased risk of cardiovascular events (10). However, in our study population, we did not find a significant association between vascular events and the IDF metabolic syndrome, whereas the metabolic syndrome as defined according to ATPIII criteria was strongly predictive of future vascular events. The incidence of vascular events was significantly higher in patients who met the ATPIII criteria but not the IDF criteria for the diagnosis of the metabolic syndrome than in patients who met the IDF criteria but not the ATPIII criteria.

There are important differences between the ATPIII and IDF definitions of the metabolic syndrome that may explain the weaker association of the IDF metabolic syndrome with vascular events. First, the lowering of the cutoff value for waist circumference leads to the inclusion of patients with a relatively lower level of this risk factor in the IDF category of the metabolic syndrome. Second, the mandatory status of the waist criterion in the new definition results in a relatively lower prevalence of other (potentially stronger) metabolic syndrome risk factors in patients with the metabolic syndrome. In particular, the high triglycerides and low HDL cholesterol metabolic syndrome components are relatively underrepresented in patients with the IDF metabolic syndrome. However, the greatest vascular risk in our investigation was conferred by these lipid traits. Concordant with this finding, high triglycerides and low HDL cholesterol were the only metabolic syndrome components that were independently associated with myocardial infarction and stroke in an evaluation of data from the third National Health and Nutrition Examination Survey (27). Consistent with their less pronounced dyslipidemia, patients with the metabolic syndrome according to the IDF definition had a lower degree of homeostasis model assessment of insulin resistance (Table 1), which is a strong cardiovascular risk factor among coronary patients (19). Third, in the IDF definition, patients receiving treatment for metabolic syndrome components are considered to meet the respective criteria, whereas therapy was not explicitly considered in the ATPIII definition. Treatment of risk factors possibly could dilute their effects on prognosis and thus weaken their prognostic power. However, in our investigation, a modification of the ATPIII category in which treatment was considered still proved strongly predictive for vascular events. Thus, the consideration of treatment in the IDF definition of the metabolic syndrome does not appear to account for its weak association with vascular events.

A recent report from a population with a high proportion of Mexican Americans suggested that the increased vascular risk of patients with the metabolic syndrome is entirely due to the high prevalence of patients with diabetes among subjects with the metabolic syndrome (28). However, in our population of Caucasian coronary patients, the metabolic syndrome defined according to ATPIII criteria conferred an increased vascular risk independent from the diabetic state.

Importantly, the cardiovascular risk associated with the ATPIII metabolic syndrome in our patients resided basically in its lipid components. Moreover, our data suggest that these lipid traits are better predictors of vascular events than the clinical category of the metabolic syndrome itself and thus lend support to recent criticism of the use of the metabolic syndrome as a clinical category in cardiovascular risk stratification (8,9). Of note, the decisive role of the dyslipidemic metabolic syndrome features in cardiovascular risk prediction is well in line with our recent report on the paramount impact of the low HDL/high triglyceride pattern on the short-term incidence of vascular events in patients with type 2 diabetes (20).

In conclusion, patients with the metabolic syndrome according to the ATPIII definition are at a significantly higher risk of vascular events than patients with the metabolic syndrome diagnosed according to IDF criteria. However, as even the ATPIII definition of the metabolic syndrome did not provide prognostic information beyond its dyslipidemic features, physicians in the care of coronary patients should pay more attention to the high triglyceride/low HDL pattern of dyslipidemia. Evaluations of new and old criteria for the metabolic syndrome in other populations as well as interventional data on the potential of cardiovascular risk reduction residing in an improvement of the high triglycerides and low HDL traits in patients with the metabolic syndrome are urgently needed.

The Vorarlberg Institute for Vascular Investigation and Treatment thanks Dr. Egmond Frommelt and the Innovationsstiftung of the Liechtenstein Global Trust Bank for providing us with a generous research grant.

We are grateful to Franz Rauch and the Vorarlberger Industriellenvereinigung; to Dr. Peter Woess and the Vorarlberger Aerztekammer; and to Luis Patsch, Director, Vorarlberger Landeskrankenhaus-Betriebsgesellschaft for continuously supporting our research institute.