Mortality from coronary heart disease (CHD), cardiovascular

Epidemiology Impact of the Metabolic Syndrome on Mortality From Coronary Heart Disease, Cardiovascular Disease, and All Causes in United States Adults Shaista Malik, MD, MPH; Nathan D. Wong, PhD, MPH; Stanley S. Franklin, MD; Tripthi V. Kamath, PhD; Gilbert J. L Italien, PhD; Jose R. Pio, BS; G. Rhys Williams, ScD Background Mortality resulting from coronary heart disease (CHD), cardiovascular disease (CVD), and all causes in persons with diabetes and pre-existing CVD is high; however, these risks compared with those with metabolic syndrome (MetS) are unclear. We examined the impact of MetS on CHD, CVD, and overall mortality among US adults. Methods and Results In a prospective cohort study, 6255 subjects 30 to 75 years of age (54% female) (representative of 64 million adults in the United States) from the Second National Health and Nutrition Examination Survey were followed for a mean SD of 13.3 3.8 years. MetS was defined by modified National Cholesterol Education Program criteria. From sample-weighted multivariable Cox proportional-hazards regression, compared with those with neither MetS nor prior CVD, age-, gender-, and risk factor adjusted hazard ratios (HRs) for CHD mortality were 2.02 (95% CI, 1.42 to 2.89) for those with MetS and 4.19 (95% CI, 3.04 to 5.79) for those with pre-existing CVD. For CVD mortality, HRs were 1.82 (95% CI, 1.40 to 2.37) and 3.14 (95% CI, 2.49 to 3.96), respectively; for overall mortality, HRs were 1.40 (95% CI, 1.19 to 1.66) and 1.87 (95% CI, 1.60 to 2.17), respectively. In persons with MetS but without diabetes, risks of CHD and CVD mortality remained elevated. Diabetes predicted all mortality end points. Those with even 1 to 2 MetS risk factors were at increased risk for mortality from CHD and CVD. Moreover, MetS more strongly predicts CHD, CVD, and total mortality than its individual components. Conclusions CHD, CVD, and total mortality are significantly higher in US adults with than in those without MetS. (Circulation. 2004;110:1245-1250.) Key Words: diabetes mellitus risk factors metabolic syndrome X mortality Mortality from coronary heart disease (CHD), cardiovascular disease (CVD), and all causes is greater in persons with diabetes and pre-existing CVD. 1 4 Metabolic syndrome (MetS) comprises insulin resistance, abdominal fat distribution, dyslipidemia, and hypertension. First characterized by Reaven 5 as syndrome X, this condition has had evolving definitions. The World Health Organization (WHO) 6 and the National Cholesterol Education Program (NCEP) Adult Treatment Panel 7 have provided definitions for MetS. The WHO definition includes microalbuminuria and allows the use of plasma insulin or plasma glucose levels to measure insulin resistance. The prevalence of MetS has been reported to be 24% in the US adult population. 8 Most studies have examined how MetS affects mortality in non-us populations. 9 11 Using both the WHO and NCEP definitions, Lakka et al 12 showed in middle-aged men in Finland approximate greater risks for CHD death, CVD death, and total mortality for those with MetS. Isomaa et al 13 used the WHO definition to show higher CVD and overall mortality in adults with MetS in Finland and Sweden. We examined the impact of MetS in a large sample of adults representative of the US population on CHD, CVD, and overall mortality, as well as how these risks compare to those with diabetes or prior CVD. Methods Study Sample and Measurements We used data on 6255 subjects (weighted to 63.9 million in the United States) 30 to 74 years of age who were participants in the Second National Health and Nutrition Examination Survey (NHANES II) who had mortality information with a mean SD follow-up of 13 4 years. NHANES II surveyed 28 000 persons nationwide from 1976 through 1980. The mortality of those with a physical examination (n 9252) was ascertained by computerized matching to 2 national databases, the National Death Index and the Social Security Administration s Death Master File for 1976 to 1992. 14 CHD mortality was defined by the International Classifica- Received February 8, 2004; de novo received April 2, 2004; revision received May 24, 2004; accepted May 27, 2004. From the Heart Disease Prevention Program, University of California, Irvine (S.M., N.D.W., S.S.F., J.R.P.), and Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ (T.V.K., G.J.L., G.R.W.). Presented in part at the American Heart Association Scientific Sessions, Orlando, Fla, November 2003. Reprint requests to Nathan D. Wong, PhD, MPH, Heart Disease Prevention Program, Department of Medicine, C240 Medical Sciences, University of California, Irvine, CA 92697. E-mail ndwong@uci.edu 2004 American Heart Association, Inc. Circulation is available at http://www.circulationaha.org DOI: 10.1161/01.CIR.0000140677.20606.0E 1245

1246 Circulation September 7, 2004 tion of Diseases, ninth revision (codes 410 through 414). CVD mortality was defined by codes 390 through 459. The study was approved by the Institutional Review Board of the University of California, Irvine. Approximately 23% (n 2145) of the cohort was deceased as of December 31, 1992. Death certificates were available for 98% of the decedents (n 2104) and missing for 41 decedents who were included in the total mortality analysis and censored in the CVD and CHD analyses. Two subjects had incomplete identifying data and were excluded. Because of missing triglycerides, HDL cholesterol (HDL-C), and/or glucose, the absence of MetS could not be confirmed for 2995 participants, leaving 6255 subjects for analysis. In these subjects, the presence of 3 MetS risk factors, diabetes, prior CVD, or normal levels of 3 MetS risk factors (confirming the absence of MetS) was confirmed to classify individuals. Subjects not included compared with those included in the analysis were younger (48.4 versus 49.7 years of age), were more likely to be male (50.6% versus 45.6%), had lower body mass index (BMI; 25 versus 26 kg/m 2 ), had higher blood pressure (132/84 mm Hg versus 128/ 81 mm Hg), and were more likely to smoke (38.9% versus 33.9%) (all P 0.0001). Serum total cholesterol and triglycerides were measured with a Technicon Auto-Analyzer II, which uses a Lieberman-Burchard reagent for cholesterol and a fluorimetric measurement of triglycerides. 15 HDL-C was determined by the quantification method, which uses ultracentrifugation and heparin-manganese precipitation. The mean of 2 sitting blood pressure readings was used. Physical activity was assessed by self-reported recreation exercise and dichotomized into much or moderate exercise versus little or no exercise. MetS was defined by modified NCEP criteria if 3 of the following were present: (1) BMI 30 kg/m 2, (2) HDL-C 1.04 mmol/l (40 mg/dl) if male or 1.29 mmol/l (50 mg/dl) if female, (3) triglycerides 1.69 mmol/l (150 mg/dl) if fasting or 4.52 mmol/l (400 mg/dl) if not fasting, (4) blood pressure 130/85 mm Hg or use of antihypertension medication, or (5) glucose 6.1 mmol/l (110 mg/dl) if fasting or 2-hour postload glucose 7.77 mmol/l (140 mg/dl). Of the 6255 included subjects, 73 did not have fasting triglycerides, so a value of 4.52 mmol/l (400 mg/dl) was accepted as abnormal. Diabetes was defined by a physician informing the subject of having diabetes, a fasting glucose 6.99 mmol/l (126 mg/dl), 2-hour postload glucose 11.1 mmol/l (200 mg/dl), or use of dietary treatment. Diabetes was included within those defined with MetS and as a separate category in which we examined mortality risk in those with MetS who did not have diabetes. Also, we examined those with 1 and 2 combined and 3 to 5 combined MetS risk factors (defining MetS), with and without diabetes, as separate groups using an optimal risk reference group consisting of no MetS risk factors. Pre-existing CVD was defined if participants self-reported that a physician had diagnosed them with CHD, heart failure, other cardiac disease, or stroke. We also examined the relation of individual MetS risk factors compared with MetS as a syndrome to mortality. Statistical Analyses The 2 test of proportions or ANOVA was used to compare the prevalence or mean levels of the individual MetS components and other baseline characteristics across clinical conditions. Multivariable Cox proportional-hazards regression, adjusted for age, gender, total cholesterol, physical activity, and cigarette smoking (risk factors not comprising the definition for MetS), examined the risk for CHD, CVD, and overall mortality in those with MetS or CVD compared with those free of these conditions (reference group). To examine for gender differences, interaction terms of the disease group and gender were used, as were analyses stratified by gender. Age- and gender-adjusted rates for each mortality end point were also determined for each clinical condition. We also examined risks in those defined with MetS (but without diabetes) and diabetes in a separate Cox regression model to determine whether the observed findings with MetS were dependent on diabetes being included in the definition. Also, we evaluated risks compared with an optimal risk group (no MetS risk factors) for those with 1 to 2 combined and 3 to 5 combined MetS risk factors, in addition to those groups defined above. To compare the impact of MetS (including those with diabetes) on each mortality end point compared with the impact of individual MetS risk factors, we also conducted multivariable Cox regression with each mortality outcome in 2 separate models: MetS or its individual components. Those with pre-existing CVD were excluded to prevent confounding with the individual component risk factors. Statistical procedures were done with SAS version 8.1 (SAS Institute), and procedures using sample weights for projection to the US population distribution used SUDAAN version 8.02 (Research Triangle Institute). Results Overall, 26.0% of subjects (weighted to a population of 16.6 million) were defined as having MetS, and 19.8% had pre-existing CVD (12.6 million). When those with diabetes were considered separately, further subdividing the MetS category, 19.2% (12.3 million) were defined as having MetS only, 6.8% (4.3 million) as having diabetes, 16.9% (10.7 million) with pre-existing CVD, and 2.9% (1.9 million) with diabetes and pre-existing CVD. Table 1 shows the prevalence of individual MetS risk factors by clinical condition and in our overall study sample. Table 2 shows how age, gender, other risk factors, and follow-up times vary between clinical conditions. The Figure shows age- and gender-adjusted CHD, CVD, and total mortality rates per 1000 person-years by disease group. After adjustment for age, gender, smoking status, physical activity, and total serum cholesterol levels and weighting to the US population by use of the NHANES II sample weights, MetS was associated with a hazard ratio (HR) of CHD mortality of 2.02 (95% CI, 1.42 to 2.89). Individuals with baseline CVD had a higher risk (HR, 4.19; 95% CI, 3.04 to 5.79). When CVD mortality was the outcome, HRs were increased in a similar pattern as for CHD mortality. MetS was also associated with an increase in overall mortality (HR, 1.40; 95% CI, 1.19 to 1.66). Pre-existing CVD led to an even higher risk (HR, 1.87; 95% CI, 1.60 to 2.17) (Table 3). Those with MetS but without diabetes had an HR of CHD mortality of 1.65 (95% CI, 1.10 to 2.47), those with diabetes had an HR of 2.87 (95% CI, 1.84 to 4.47), those with baseline CVD but no diabetes had an HR of 3.89 (95% CI, 2.79 to 5.43), and those with combined diabetes and CVD had the highest risk (HR, 6.45; 95% CI, 4.24 to 9.79) (Table 3, indented categories). Similar trends for increased risks of CVD mortality also were seen. For total mortality, increased risks were seen for those with diabetes and in those with pre-existing CVD, with the highest risk among those with both diabetes and CVD (Table 3). When those with no MetS risk factors were used as the reference group, HRs for CHD mortality were 2.10 (95% CI, 1.05 to 4.19) for those with 1 to 2 MetS risk factors and 3.51 (95% CI, 1.81 to 6.81) for those with MetS (including those with diabetes) (Table 4). Among those with MetS but without diabetes (3 to 5 MetS risk factors), the HR was 2.87 (95% CI, 1.44 to 5.73), and for those with diabetes, the HR was 5.02 (95% CI, 2.47 to 10.23) (Table 4, indented categories). Risks were greatest for those with pre-existing CVD alone and for those with combined diabetes and CVD. For CVD mortality, similar increases in risk across disease and risk factor

Malik et al Impact of Metabolic Syndrome on Mortality 1247 TABLE 1. Prevalence of Individual MetS Abnormalities Among US Adults by Disease Category Weighted No. of Subjects, millions (%) Disease Condition Categories* Elevated Blood Pressure Disease Subjects, n Impaired Glucose Tolerance Low HDL-C High Triglycerides Obesity All groups 6255 63.9 (100) 9.0 46.9 21.8 54.8 20.1 No MetS, diabetes, or CVD 2878 34.6 (54.2) 4.6 25.4 9.6 31.5 3.5 MetS (all) 1698 16.6 (26.0) 18.5 85.0 48.2 90.5 56.2 MetS (no diabetes) 1178 12.3 (19.2) 21.0 92.6 52.6 94.9 63.9 Diabetes 520 4.3 (6.8) 100.0# 63.4 35.9 78.0 34.2 CVD (all) 1679 12.6 (19.8) 8.6 59.3 23.3 77.1 20.7 Pre-existing CVD 1398 10.7 (16.9) 8.6 57.3 20.4 74.8 17.8 Diabetes and CVD 281 1.9 (2.9) 100.0# 70.9 40.1 90.0 37.4 *P 0.0001 across disease condition categories. Glucose 6.1 to 6.94 mmol/l (110 to 125 mg/dl) if fasting or 2-hour postload glucose 7.77 to 11.04 mmol/l (140 to 199 mg/dl). HDL-C 1.04 mmol/l (40 mg/dl) if male or 1.29 mmol/l (50 mg/dl) if female. Triglycerides 1.69 mmol/l (150 mg/dl) if fasting or 4.52 mmol/l if nonfasting (400 mg/dl). Blood pressure 130/85 mm Hg or on antihypertensive medication. BMI 30 kg/m 2. #By definition, all subjects with diabetes mellitus have impaired glucose tolerance, even if on treatment and with normal glucose (only 27.3% of those with diabetes and 19.9% of those with CVD and diabetes had fasting glucose levels of 110 mg/dl or 2-hour postload glucose levels of 140 mg/dl). categories were noted (all P 0.0001 to P 0.01). Although those with 1 to 2 risk factors did not have an increased risk of total mortality, those with MetS (including diabetes) did (HR, 1.47; 95% CI, 1.15 to 1.87). Those with MetS but without diabetes had no statistically significant risk of overall mortality, whereas those with diabetes and/or CVD had significant 2- to 3-fold increases in risk (Table 4). We found no significant differences in mortality prediction between men and women (P 0.56, 0.35, and 0.83 for interactions of MetS with gender for CHD, CVD, and total mortality, respectively). In analyses stratified by gender, women and men had similar HRs for CHD, CVD, and total mortality. MetS carried a 2-fold risk of CHD mortality in women (HR, 2.17; 95% CI, 1.23 to 3.81), whereas men had only a slightly lower risk (HR, 1.92; 95% CI, 1.22 to 3.05). TABLE 2. Other Characteristics at Baseline Women with MetS had a similar risk of CVD death (HR, 2.02; 95% CI, 1.36 to 2.99), whereas men with MetS had a slightly lower risk (HR, 1.71; 95% CI, 1.21 to 2.43). Overall mortality associated with MetS was similar between men (HR, 1.41; 95% CI, 1.13 to 1.76) and women (HR, 1.41; 95% CI, 1.09 to 1.81). In multivariable models examining MetS risk factors in predicting mortality, only single components predicted risk: low HDL-C resulted in a greater risk only for CHD mortality (HR, 1.92; 95% CI, 1.31 to 2.82); hypertension resulted in an increased risk (HR, 1.70; 95% CI, 1.20 to 2.43) only for CVD mortality; and impaired fasting or impaired 2-hour glucose resulted in a higher risk only for overall mortality (HR, 1.31; 95% CI, 1.06 to 1.62). These risks can be compared with slightly greater risks seen for MetS (including diabetes) for Disease All Groups No MetS, Diabetes, or CVD MetS (All) MetS (No Diabetes) Diabetes CVD (All) Pre-Existing CVD Diabetes and CVD Age, y* 49.7 45.6 51.4 50.0 55.4 58.5 57.9 61.7 Gender,* % Male 45.6 43.4 46.4 48.7 39.8 50.6 50.8 49.1 Female 54.4 56.6 53.6 51.3 60.2 49.5 49.2 51.0 Exercise,* % Much/moderate 83.7 86.9 83.5 85.2 78.7 75.22 77.0 65.1 Little/none 16.3 13.1 16.5 14.8 21.3 24.8 23.1 35.0 Smoking,* % Current smoker 33.9 36.4 31.5 33.8 24.8 30.5 31.5 24.3 Not current 66.1 63.6 68.6 66.2 75.2 69.5 68.5 75.7 Total cholesterol, mg/dl* 222.6 213.5 233.6 234.3 231.7 232.9 233.2 231.4 Length of follow-up, y* 14.0 14.5 13.9 14.2 13.0 13.4 12.7 11.0 *P 0.0001 for differences between disease groups.

1248 Circulation September 7, 2004 Age- and gender-adjusted CHD, CVD, and total mortality rates in US adults with MetS with and without diabetes and pre-existing CVD in the NHANES II Follow-Up Study (n 6255; mean followup, 13.3 years). DM indicates diabetes mellitus. CHD, CVD, and total mortality (HR, 1.98, 1.87, and 1.44, respectively; all P 0.0001). Discussion MetS is associated with an increased risk of death from CHD, CVD, or all causes in US adults. Although diabetes is defined as a CHD risk equivalent, 1,7 persons with MetS but without diabetes have a wide spectrum of risk. 16 In persons with MetS who do not have diabetes, increased risks of CVD and CHD mortality remain. CHD, CVD, and total mortality risks are as high or nearly as high as risks in those with diabetes compared with those with pre-existing CVD, providing support to previous studies 1 4 and guidelines. 7 Those with diabetes and CVD are at highest risk. Even those with 1 or 2 MetS risk factors are at a 2-fold-greater risk of CHD and CVD mortality, suggesting that risk is not optimal unless all MetS risk factors are absent. Finally, MetS predicts CHD, CVD, and total mortality more strongly than individual MetS risk factors, consistent with previous reports. 13 Our findings suggest the need for intensified treatment recommendations in persons with MetS and even in those with only 1 or 2 MetS risk factors. Intensive treatment of diabetics involving control of blood pressure, blood sugar, and blood lipids can result in reducing macrovascular and microvascular complications by 50%. 17 In persons with MetS, up to 80% of CHD events may be preventable from optimal control of LDL-C, HDL-C, and blood pressure. 16 Previous studies have not examined mortality risks in persons with MetS after separating out those with diabetes or those with known CVD. 12,13,18 In those with MetS, 40% are at intermediate risk (10% to 20% 10-year risk) of CHD; 20% of men are at higher risk. 16 Moreover, 40% of persons with MetS either are at high risk or have significant coronary calcium ( 75th percentile for age and sex); in these individuals, more aggressive risk factor management may be warranted. 19 Strengths of our study include its prospective design, reliable assessment of causes of death, generalizability from the use of a sample representative of the US population, and assessment of the impact of MetS alone relative to diabetes and pre-existing CVD on mortality end points. Our study excluded 2995 participants because of missing risk factor information on glucose and lipids. From a sensitivity analyses including these subjects, our main effects remained significant and in the same direction, despite some reduction in our HRs, probably as a result of some misclassification in the no-disease category. We classified all included subjects in our main analyses by confirming that those included in the TABLE 3. Multivariable Cox Proportional-Hazards Regression Relating Each Disease/Condition Compared With Neither Condition to CHD, CVD, and Overall Mortality in US Adults CHD Mortality CVD Mortality Overall Mortality Condition n HR 95% CI P HR 95% CI P HR 95% CI P No MetS, diabetes, or CVD 2878 1.0 1.0 1.0 MetS (all) 1698 2.02 1.42 2.89 0.0001 1.82 1.40 2.37 0.0001 1.40 1.19 1.66 0.0001 MetS (no diabetes)* 1178 1.65 1.10 2.47 0.02 1.56 1.15 2.12 0.005 1.17 0.96 1.42 0.11 Diabetes* 520 2.87 1.84 4.47 0.0001 2.42 1.75 3.35 0.0001 1.97 1.59 2.43 0.0001 Pre-existing CVD (all) 1679 4.19 3.04 5.79 0.0001 3.14 2.49 3.96 0.0001 1.87 1.60 2.17 0.0001 Pre-existing CVD (no diabetes)* 1398 3.89 2.79 5.43 0.0001 2.83 2.23 3.61 0.0001 1.74 1.48 2.03 0.0001 Diabetes and CVD* 281 6.45 4.24 9.79 0.0001 5.26 3.82 7.23 0.0001 2.80 2.21 3.54 0.0001 n Indicates unweighted sample sizes. HRs and 95% CIs are weighted to US population and adjusted for gender, age, smoking, physical activity, and total cholesterol. *Categories represented in lieu of MetS (all) and pre-existing CVD (all) in a separate regression model.

Malik et al Impact of Metabolic Syndrome on Mortality 1249 TABLE 4. Multivariable Cox Proportional-Hazards Regression Examining Relation of Each MetS Risk Factor and Disease Category Compared With Those With No MetS Risk Factors to CHD, CVD, and Overall Mortality in US Adults CHD Mortality CVD Mortality Overall Mortality Condition n HR 95% CI P HR 95% CI P HR 95% CI P No MetS RF 1228 1.0 1.0 1.0 1 2 MetS RF 1650 2.10 1.05 4.19 0.04 1.73 1.10 2.70 0.02 1.07 0.82 1.39 0.61 MetS (all) 1698 3.51 1.81 6.81 0.0002 2.71 1.76 4.17 0.0001 1.47 1.15 1.87 0.002 MetS (no diabetes)* 1178 2.87 1.44 5.73 0.003 2.32 1.47 3.67 0.0003 1.23 0.94 1.60 0.12 Diabetes* 520 5.02 2.47 10.23 0.0001 3.63 2.26 5.82 0.0001 2.07 1.57 2.73 0.0001 CVD (all) 1679 7.30 3.83 13.89 0.0001 4.68 3.09 7.09 0.0001 1.95 1.54 2.46 0.0001 Pre-existing CVD* 1398 6.81 3.55 13.06 0.0001 4.24 2.79 6.46 0.0001 1.82 1.44 2.32 0.0001 Diabetes and CVD* 281 11.30 5.66 22.56 0.0001 7.88 4.93 12.60 0.0001 2.94 2.19 3.95 0.0001 RF indicates risk factors; n, unweighted sample sizes. HRs and 95% CIs are weighted to US population and adjusted for gender, age, smoking, physical activity, and total cholesterol. *Categories represented in lieu of MetS (all) and pre-existing CVD (all) in a separate regression model. reference group were negative for 3 MetS risk factors and that those with MetS were confirmed to have 3 MetS risk factors. Because waist circumference measures were not available, we used BMI to classify individuals with obesity. Recent studies have used BMI successfully to classify those with MetS because MetS and diabetes are strongly correlated. 20 Using NHANES III, 97% of those identified as having MetS from BMI would have been identified if waist circumference cut points had been substituted. 19 Despite the difference in definitions used, we note a prevalence of MetS in our study population (26%) that is similar to published estimates. 8,21 From NHANES III, the odds for CVD from definitions using waist circumference compared with BMI are nearly identical, suggesting that risk prediction is comparable using either definition. Moreover, Sattar et al 22 defined MetS by incorporating BMI cut points and reported a similar increased risk of CHD. Including those with a 2-hour postload glucose of 140 to 199 mg/dl in our MetS definition resulted in more individuals being classified with insulin resistance than when fasting glucose alone was used. We found that 68% of these subjects would not have met the criterion of 110 to 125 mg/dl for impaired fasting glucose. Others also show the oral glucose tolerance test to be a more sensitive method. 23,24 But because other factors are involved in the definition of MetS, only 8.3% of our subjects with MetS would not have been identified if the oral glucose tolerance test was not included. Our results relating MetS to each mortality end point remained robust when we ran our analyses without the oral glucose tolerance test. In addition, among the 520 subjects with diabetes only, 29 (5.6%) were identified solely on the basis of dietary treatment; however, our HRs in relation to the mortality end points remained nearly identical when we included only those defined by physician report and/or required glucose cut points. In US adults 30 to 74 years of age, MetS confers an increased risk of CHD, CVD, and total mortality, and persons with diabetes and/or pre-existing CVD are at even higher risk. Even 1 or 2 MetS risk factors confer increased risks of CHD and CVD mortality. MetS is a serious clinical condition associated with a worse prognosis than its individual risk factors. Further investigation is needed to address which treatment strategies are most appropriate given an individual s risk profile. Disclosure Dr Wong has received grant support from Bristol-Meyers Squibb. Acknowledgment This work was supported in part by a research grant from the Bristol Myers-Squibb Pharmaceutical Research Institute. References 1. Haffner SM, Lehto S, Ronnemaa T, et al. Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med. 1998;339: 229 234. 2. Saydah SH, Eberhardt MS, Loria CM, et al. Age and the burden of death attributable to diabetes in the United States. Am J Epidemiol. 2002;156: 714 719. 3. Miettinen H, Lehto S, Salomaa V, et al. Impact of diabetes on mortality after the first myocardial infarction: the FINMONICA Myocardial Infarction Register Study Group. Diabetes Care. 1998;21:69 75. 4. Cho E, Rimm EB, Stampfer MJ, et al. The impact of diabetes mellitus and prior myocardial infarction on mortality from all causes and from coronary heart disease in men. J Am Coll Cardiol. 2002;40:954 960. 5. Reaven GM. Role of insulin resistance in human disease. Diabetes. 1988;37:1595 1607. 6. Albertti KG, Zimmet PZ. Definition, diagnosis, and classification of diabetes mellitus and its complications 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabetes Med. 1998;15:539 553. 7. Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA. 2001;285:2486 2497. 8. Ford ES, Giles WH, Dietz WH. Prevalence of the metabolic syndrome among US adults. JAMA. 2002;287:356 359. 9. Trevisan M, Liu J, Bahsas FB, et al. Syndrome X and mortality: a population-based study. Am J Epidemiol. 1998;148:958 966. 10. Burchfiel CM, Sharp DS, Curb DJ, et al. Hyperinsulinemia and cardiovascular disease in elderly men. Arterioscler Thromb Vasc Biol. 1998; 18:450 457. 11. Solymoss BC, Marcil M, Cahour M, et al. Fasting hyperinsulinism, insulin resistance syndrome, and coronary artery disease in men and women. Am J Cardiol. 1995;76:1152 1156. 12. Lakka HM, Laaksonen DE, Lakka TA, et al. The metabolic syndrome and total and cardiovascular disease mortality in middle-aged men. JAMA. 2002;288:2709 2716.

1250 Circulation September 7, 2004 13. Isomaa B, Almgren P, Tuomi T, et al. Cardiovascular morbidity and mortality associated with the metabolic syndrome. Diabetes Care. 2001; 24:683 689. 14. Loria CM, Sempos CT, Vuong C. Plan and operation of the NHANES II Mortality Study, 1992. Vital Health Stat. 1999;38:1 16. 15. McDowell A, Engel A, Masey J, et al. National Center for Health Statistics: Plan and Operation of the Second National Health and Nutrition Examination Survey, United States, 1976 1980. Washington, DC: US Government Printing Office; 1981. DHHS publication No. PHS81 1317, Vital and Health Statistics, series 1, No. 15. 16. Wong ND, Pio JR, Franklin SS, et al. Preventing coronary events by optimal control of blood pressure and lipids in patients with the metabolic syndrome. Am J Cardiol. 2003;91:1421 1426. 17. Gaede P, Vedel P, Larsen N, et al. Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes. N Engl J Med. 2003;348:383 393. 18. Ninomiya JK L Italien G, Criqui MH, et al. Association of the metabolic syndrome with history of myocardial infarction and stroke in the Third National Health and Nutrition Examination Survey. Circulation. 2004; 109:42 46. 19. Wong ND, Sciammarella M, Miranda-Peats R, et al. The metabolic syndrome, diabetes, and subclinical atherosclerosis assessed by coronary calcium. J Am Coll Cardiol. 2003;41:1547 1553. 20. Janssen I, Katzmarzyk PT, Ross R. Body mass index, waist circumference, and health risk. Arch Intern Med. 2002;162:2074 2079. 21. Park YW, Zhu S, Palaniappan L, et al. The metabolic syndrome: prevalence and associated risk factor findings in the US population from the Third National Health and Nutrition Examination Survey, 1988 1994. Arch Intern Med. 2003;163:427 436. 22. Sattar N, Gaw A, Scherbakova O, et al. Metabolic syndrome with and without C-reactive protein as a predictor of coronary heart disease and diabetes in the West of Scotland Coronary Prevention Study. Circulation. 2003;108:414 419. 23. Carnevale Schianca GP, Rossi A, Sainaghi PP, et al. The significance of impaired fasting glucose versus impaired glucose tolerance: importance of insulin secretion and resistance. Diabetes Care. 2003;26:1333 1337. 24. Hanefeld M, Koehler C, Fuecker K, et al. Insulin secretion and insulin sensitivity pattern is different in isolated impaired glucose tolerance and impaired fasting glucose: the Risk Factor in Impaired Glucose Tolerance for Atherosclerosis and Diabetes Study. Diabetes Care. 2003;26: 868 874.