Comparing the Predictive Abilities of Different Metabolic Syndrome Definitions for Acute Coronary Syndrome: A Case

Abstract

Background:

Different institutions have proposed various definitions for metabolic syndrome, which is a combination of risk factors for cardiovascular diseases (CVD). This study aimed to compare the feasibilities and abilities of different metabolic syndrome definitions in predicting acute coronary syndrome (ACS) in Chinese adults.

Methods:

A case–control study was designed. This study recruited 162 newly diagnosed ACS patients (the case group) and 162 non-ACS patients (the control group) according to the study criteria. Metabolic syndrome definitions proposed by the National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III), International Diabetes Federation (IDF), American Heart Association/National Heart, Lung, and Blood Institute (AHA/NHLBI), Chinese Diabetes Society (CDS), and Joint Committee for Developing Chinese Guidelines on Dyslipidemia in Adults (JCDCG) were studied. After collecting demographic and clinical data, sensitivity, specificity, positive and negative predictive values (PPV, NPV), the likelihood ratio of a positive test and a negative test (LR+, LR−), odds ratios (OR), diagnostic accuracy, and the Youden index (YI) were compared.

Results:

Of the 324 participants, the mean age was 59.1±10.5 years, and 56.8% were males. The AHA/NHLBI and IDF definitions had better sensitivity (53.09%, 48.77%). The CDS definition was more specific (76.54%), but less sensitive (25.93%). The IDF definition performed better in PPV (53.74%), NPV (53.11%), LR+ (1.15) and LR− (0.89), OR (1.32), and diagnostic accuracy (53.4%). The IDF definition also provided optimal cutoff points with the biggest YI.

Conclusion:

The IDF definition performed better in detecting the onsets of nonfatal ACS in the northwestern Chinese population. All studied definitions were feasible in Chinese clinical settings.

Introduction

Metabolic syndrome is recognized as a combination of risk factors for cardiovascular disease (CVD) and diabetes.¹ As a result of epidemic overweight and sedentary lifestyles in modern society, metabolic syndrome has become a worldwide health problem. It was estimated that 20%–25% of adults in the world had metabolic syndrome.² In China, the prevalence of metabolic syndrome in adults was 9.8% in males and 17.8% in females.³

People with metabolic syndrome have increased risks in developing CVD, which is the leading cause of death in the world.^4

–7 Previous studies found that the co-existence of metabolic syndrome in CVD patients worsened the progression of the disease, leading to more complications and higher mortalities.^5,8 In China, there are 110 million patients with coronary heart disease (CHD), which cost 2.45 billion US dollars in health care each year.⁹ Acute coronary syndrome (ACS) is the most acute and fatal type of CHD and requires immediate medical interventions. In China, ACS has caused 1.75 million deaths annually.⁹ Given the adverse health outcomes of ACS patients, the abilities of metabolic syndrome in detecting this high-risk population draw worldwide attention.

The main components of metabolic syndrome have been recognized as raised blood pressure, dyslipidemia, raised fasting glucose, and central obesity. However, different organizations have proposed various definitions. In 1999, the World Health Organization (WHO) established a metabolic syndrome definition that proposed insulin-resistance as a prerequisite component.¹⁰ In 2001, the National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) proposed another definition. In 2005, International Diabetes Federation (IDF) established a central-obesity-centered definition.^2,11 In 2009, the IDF and American Heart Association/National Heart, Lung, and Blood Institute (AHA/NHLBI) published a scientific statement that claimed no obligatory component for metabolic syndrome.¹ Nevertheless, the original IDF definition is still widely used in various studies, especially in China.^12

–16

Although a universally agreed-upon definition is beneficial for comparing studies globally, national definitions would be more specific in setting obesity criteria and more practical for local medical resources.¹ In China, there are two national definitions for metabolic syndrome; one was published by the Chinese Diabetes Society (CDS) in 2004¹⁷ and the other by the Joint Committee for Developing Chinese Guidelines on Prevention and Treatment of Dyslipidemia in Adults (JCDCG) in 2007.¹⁸ Both definitions proposed cutoff points based on population-based studies, and have been widely used in China.^19
–21

Of these definitions, identifying which can more accurately predict ACS in Chinese adults has become a concern. Previous studies focused on comparing the performances of different metabolic syndrome definitions in predicting CVD in various populations.^{20,22

–29} However, no consistent results have been found. Limited studies were designed for comparing the predictive abilities of metabolic syndrome definitions for ACS. Only one study compared NCEP, IDF, and AHA/NHLBI definitions in prediction of ACS in Greece, but the characteristics of age and gender were not matched between two groups (P<0.05). In addition to the ethnic differences, their findings would not be applied in Chinese population directly.²⁵

Previous Chinese studies either only compared international definitions,^28
–30 or compared international definitions with a single Chinese definition.^20,26,27 To the best of our knowledge, none have compared CDS and JCDCG in the same study. Although previous cohort studies reported large numbers of participants with metabolic syndrome, a relatively small number of CVD/CHD incidents were observed,^20,26 which has limited their statistical abilities in predicting CVD/CHD incidents. Moreover, participants' metabolic syndrome status may have changed during years of follow-up, which would confound the real association between metabolic syndrome and CVD/CHD incidents.^20,26,30 To overcome these barriers, this study applied a case–control design, which recruited a greater proportion of CHD patients to compare the two Chinese definitions (CDS and JCDCG) with international ones. To avoid the influences of lifestyle changes on metabolic syndrome, this study selected newly diagnosed ACS patients as cases. The results of this study will provide a suitable metabolic syndrome definition to identify high risk populations for ACS.

Because insulin resistance and microalbuminuria required by the WHO definition are not routinely tested in current clinical practices in China, the WHO definition was excluded in the comparisons. The WHO definition may be more appropriate for research purposes, but it is not feasible for clinical practice, especially in China.

This study aims at: (1) Comparing the overall performances of five metabolic syndrome definitions (NCEP, IDF, AHA/NHLBI, CDS, JCDCG) in predicting ACS in terms of sensitivity, specificity, positive (PPV) and negative predictive (NPV) values, likelihood ratios of a positive and negative test, odds ratios (ORs), diagnostic accuracy, and Youden index (YI) and (2) exploring the feasibilities of the definitions in Chinese clinical settings.

Methods

Study design

A case–control study design was used in this study.

Study settings and participants

This study was conducted in two affiliated hospitals of Medical College of Xian Jiaotong University. Both hospitals are III-A level hospitals, indicating the best public hospitals in China. All participants of the study were recruited from December, 2012, to February, 2013.

The case group was recruited consecutively from hospitalized patients with newly diagnosed ACS in the cardiac units. The criteria for inclusion in the case group were: (1) Being at least 18 years old and (2) being hospitalized for first onsets of ACS. The period of time from the onset of the disease until data collection was within 2 weeks. In this study, the types of ACS exclusively included ST segment elevation myocardial infarction (STEMI), non-ST segment elevation myocardial infarction (NSTEMI), and unstable angina pectoris. Participants were defined as having STEMI if they had continuous chest pain for ≥30 min and with an electrocardiogram (ECG) showing ST-segment elevation ≥1 mm in ≥2 continuous limb leads or ≥2 mm in ≥2 continuous precordial leads or new left bundle branch block, and typical rise and fall of cardiac biomarkers [troponin I and T, or muscle and brain fraction of creatine kinase (CK-MB)].^25,31,32 NSTEMI and unstable angina pectoris were defined as having chest pain at rest or minimal exertion of ≥5 min during previous 24 hr and with ST-segment abnormalities other than elevation; and/or having raised concentration of troponin I and T or CK-MB; and/or with a presence of ≥50% stenosis of the coronary artery in angiography.^25,31
–33 Patients with diagnoses of other atherosclerotic CVD like stroke and peripheral arterial disease were excluded. In addition, the subjects had to be able to communicate in Chinese.

For each case, one control matching the study criteria was recruited from non-ACS patients of the same hospital according to the recommended methods by previous case–control study.³¹ The matching and including criteria for the control group were: (1) Matched age (±5 years) and gender to the cases and (2) being hospitalized within 2 weeks and with no diagnosis or past health history of CHD-related diseases. The controls were recruited from noncardiac units, including physical health check units, ophthalmic units, and orthopedic and general surgical units. Their reasons for hospitalization varied, e.g., receiving physical health checks, myopia corrections, cataracts, or receiving elective surgery. None of their diagnoses were related to ACS or its risk factors. To eliminate the possibility of potential ACS, patients with any symptoms or history of chest pain or with ECG reports indicating any possibility of CHD were excluded from the control group. In addition, those subjects who developed ACS during their hospitalization periods were excluded. These subjects also had to be able to communicate in Chinese.

The sample size was calculated based on McNemar test using PASS 11.0. It was estimated that a total of 316 subjects would give the study 80% power at 5% level of significance to detect a 10% difference in the studied metabolic syndrome definitions, assuming the discordant proportion was at most 30%.

Definitions of metabolic syndrome

The metabolic syndrome definitions studied in this paper are as follows.

NCEP definition

At least three of the following risk determinants: (1) central obesity, waist circumference (WC) >102 cm (male), >88 cm (female); (2) blood pressure (BP) >130/85 mmHg; (3) triglycerides (TGs) ≥1.70 mmol/L; (4) high-density lipoprotein cholesterol (HDL-C) <1.03 mmol/L in males, <1.29 mmol/L in females; (5) fasting plasma glucose (FPG) ≥6.1 mmol/L.¹¹

IDF definition for Chinese

A prerequisite of WC ≥90 cm for men, ≥80 cm for women, or body mass index (BMI) ≥30 kg/m², plus at least two of the following disorders: (1) TGs ≥1.70 mmol/L, or undergoing specific treatment for this lipid abnormality; (2) HDL-C <1.03 mmol/L in men, <1.29 mmol/L in women, or undergoing specific treatment for this lipid abnormality; (3) BP ≥130/85 mmHg, or undergoing antihypertensive treatment; (4) FPG ≥5.6 mmol/L, or undergoing treatment for type 2 diabetes.²

AHA/NHLBI definition

Three or more of the following disorders: (1) Central obesity, WC ≥90 cm (male), ≥80 cm (female); (2) BP ≥130/85 mmHg, or undergoing antihypertensive treatment; (3) TGs ≥1.70 mmol/L, or undergoing treatment for elevated TGs; (4) HDL-C, <1.03 mmol/L in males, <1.29 mmol/L in females, or undergoing treatment for reduced HDL-C; (5) FPG ≥5.6 mmol/L, or undergoing treatment of type 2 diabetes.¹

CDS definition

At least three of the following disorders: (1) BMI ≥25 kg/m²; (2) FPG ≥6.1 mmol/L, or undergoing treatment for type 2 diabetes; (3) BP ≥140/ 90 mmHg, or undergoing treatment for hypertension; (4) TGs ≥1.70 mmol/L, and/ or HDL-C <0.9 mmol/L in males, <1.0 mmol/L in females.¹⁷

JCDCG definition

At least three of the five listed disorders: (1) WC >90 cm (male), >85 cm (female); (2) TGs ≥1.70 mmol/L; (3) HDL-C <1.04 mmol/L; (4) BP ≥130/85 mmHg; (5) FPG ≥6.1 mmol/L, and/ or with a history of type 2 diabetes.¹⁸

Data collection procedures

This study was approved by the Research Ethics Committee of the university and study hospitals. During the study period, all potential participants in study hospitals were screened by the researcher according to medical records once their diagnoses were confirmed. Eligible patients matching the study criteria were invited and received full introduction of the study. Written consents were obtained from all participants before joining the study.

Once the patient was recruited in the study, demographic information including age, gender, marital conditions, and educational levels were obtained by face-to-face interviews. Measurements of body weight, height, WC, and BP were administered by the researcher in the same manner following WHO guidelines.³⁴ BMI was calculated as weight divided by the square of height (kg/m²).¹⁰ Past health history and results of lab examinations in terms of FPG, TGs, and HDL-C were obtained from patients` medical records of this hospitalization. All blood samples for the tests were collected after an 8-hr fast. In this study, laboratory tests were conducted after current hospital admissions, which were all within 2 weeks until the time of data collection.

Statistical analysis

The data were analyzed by SPSS 20.0 version for Windows. Continuous variables were presented as mean±standard deviation for normal distributed ones and median (lower quartile, upper quartile) for skewed distributed ones. Categorical variables were presented as frequency (percentage). Demographic and metabolic syndrome–related characteristics were compared between groups by t-tests, Mann–Whitney U-tests, or chi-squared tests as appropriate.

Sensitivity, specificity, PPV and NPV, likelihood ratios of a positive test and a negative test (LR+, LR−), as well as ORs and diagnostic accuracy were calculated and compared across definitions. LR+ is calculated as sensitivity/(1-specificity), indicating the probabilities of having metabolic syndrome in case group compared with the controls; LR− equals specificity/(1-sensitivity), meaning the chances of not having metabolic syndrome in cases than that in controls.³⁵ Diagnostic accuracy is the proportion of true positive and true negative among all subjects, indicating the definitions' abilities to provide correct results.³⁶ McNemar chi-squared tests were applied to compare the sensitivity and specificity between different definitions.³⁷ The YI was calculated as (sensitivity + specificity −1). One definition with the biggest YI provides the optimal cutoff point in identifying ACS.³⁸

All probability values reported were two sided and P values<0.05 were defined as statistically significant.

Results

In this study, 162 cases and 162 controls were recruited. The mean age of the cases was 59.48±10.17 years, ranging from 33 to 83 years. The control patients were from 33 to 80 years old. In the case group, there were 92 (56.8%) males, who were younger than the females (t=2.051, P=0.042). NCEP, IDF, AHA/NHLBI, CDS, and JCDCG definitions identified 114 (35.2%), 147 (45.4%), 169 (52.2%), 80 (24.7%), and 126 (38.9%) participants with metabolic syndrome, respectively. Table 1 shows the demographic and metabolic syndrome–related characteristics of the participants. There were significant differences between groups in WC, BMI, and systolic blood pressure (SBP) (P values were 0.002, 0.014, 0.005, and 0.013).

Table 1.

Demographic and Metabolic Syndrome–Related Characteristics of the Participants

Characteristics	Total (n=324)	Case group (n=162)	Control group (n=162)	P value
Age (years)	59.1±10.5	59.5±10.2	58.7±10.9	0.528
Males	184 (56.8)	92 (56.8)	92 (56.8)	1.000
Marital status				0.237
Married	305 (94.14)	150 (92.60)	155 (95.68)
Divorced/widowed	19 (5.86)	12 (7.40)	7 (4.32)
Education levels				0.811
0–6 years	72 (22.2)	34 (21.0)	38 (23.5)
7–9 years	102 (31.5)	49 (30.2)	53 (32.7)
10–12 years	87 (26.9)	47 (29.0)	40 (24.7)
>12 years	63 (19.4)	32 (19.8)	31 (19.1)
History of hypertension	148 (45.7)	72 (44.4)	76 (46.9)	0.656
Medication for hypertension	105 (70.9)	54 (75.0)	51 (67.1)	0.290
History of dyslipidemia^a	67 (20.7)	36 (22.2)	31 (19.1)	0.493
Medication for dyslipidemia	16 (23.9)	9 (25.0)	7 (22.6)	0.817
History of diabetes	48 (14.8)	20 (12.3)	28 (17.3)	0.211
Medication for diabetes	37 (77.1)	16 (80.0)	21 (75.0)	0.684
Body height (cm)	166.29±7.40	165.70±7.10	166.89±7.66	0.148
Body weight (kg)	66.34±10.66	67.30±10.67	65.38±10. 60	0.106
WC (cm)	87.48±10.35	89.22±9.87	85.74±10.55	0.002^**
BMI (kg/m²)	23.96±3.30	24.47±3.24	23.44±3.29	0.005^**
SBP (mmHg)	129.60±20.19	126.84±19.07	132.37±20.95	0.013^*
DBP (mmHg)	80.31±12.81	79.72±12.48	80.90±13.14	0.411
FPG^b (mmol/L)	5.20 (4.64, 6.52)	5.44 (4.70, 6.71)	5.05 (4.85, 6.37)	0.394
TGs^b (mmol/L)	1.36 (1.00, 1.89)	1.37 (1.00, 1.97)	1.36 (1.05, 1.74)	0.202
HDL-C^c (mmol/L)	1.02 (0.87, 1.21)	1.01 (0.87, 1.24)	1.02 (0.86, 1.20)	0.710

In this table, categorical variables were presented by n (%), and tested by chi-squared tests; normal distributed variables were described as (mean standard deviation) and tested by t-tests.

Dyslipidemia is diagnosed following the Chinese Guidelines on Prevention and Treatment of Dyslipidemia in Adults (2007). The types of dyslipidemia include: high total cholesterol (≥6.22 mmol/L), high low-density lipoprotein cholesterol (≥4.14 mmol/L), abnormal high-density lipoprotein cholesterol (≤1.04 mmol/L, or ≥1.55 mmol/L) and high triglycerides (≥2.26 mmol/L).

Skewed distribution variable, described as median (upper quartile, lower quartile), and applied t-test for variables transformed by log (10).

Skewed distribution variable, described as median (upper quartile, lower quartile), and applied Mann–Whitney test.

P<0.05.

P<0.01.

WC, waist circumference; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; FPG, fasting plasma glucose; TGs, triglycerides; HDL-C, high-density lipoprotein cholesterol.

Table 2 summarizes the sensitivity, specificity, PPV, NPV, LR+, LR−, and OR of each definition with 95% confidence intervals. The AHA/NHLBI and IDF definitions had better sensitivity (53.1% and 48.8%). Although the CDS definition had the lowest sensitivity (25.93%), it was more specific (76.54%). The IDF definition performed slightly better in predictive values (PVs), LRs, and OR (1.32), whereas the AHA/NHLBI definition had smaller PPV, LR+, and OR values; the CDS definition showed smaller NPV and worse LR− values.

Table 2.

Sensitivity, Specificity, PPV, NPV, LR+, LR−, and OR for Studied Definitions % (95% CI)

	Sensitivity	Specificity	PPV	NPV	LR+	LR−	OR
NCEP	37.04 (29.98–44.69)	66.67 (59.1–73.47)	52.63 (43.53–61.56)	51.43 (44.7–58.1)	1.11 (1.01–1.22)	0.94 (0.92–0.97)	1.18 (0.74–1.86)
IDF	48.77 (41.19–56.4)	58.02 (50.33–65.35)	53.74 (45.69–61.6)	53.11 (45.77–60.32)	1.15 (1.08–1.21)	0.89 (0.86–0.93)	1.32 (0.85–2.04)
AHA/NHLBI	53.09 (45.42–60.61)	48.77 (41.19–56.40)	50.89 (43.41–58.32)	50.97 (43.17–58.72)	1.04 (0.99–1.08)	0.96 (0.91–1.01)	1.08 (0.70–1.67)
CDS	25.93 (19.79–33.18)	76.54 (69.45–82.41)	52.50 (41.70–63.08)	50.82 (44.58–57.03)	1.10 (0.92–1.33)	0.97 (0.95–0.99)	1.14 (0.69–1.89)
JCDCG	40.74 (33.47–48.44)	62.96 (55.31–70.02)	52.38 (43.72–60.90)	51.52 (44.59–58.38)	1.10 (1.02–1.19)	0.94 (0.91–0.97)	1.17 (0.75–1.83)

PPV, positive predictive values; NPV, negative predictive values; LR+, likelihood ratio for a positive test; LR−, likelihood ratio for a negative test; OR, odds ratios; CI, confidence intervals; NCEP ATP III, National Cholesterol Education Program Adult Treatment Panel III; IDF, International Diabetes Federation; AHA/NHLBI, American Heart Association/National Heart, Lung, and Blood Institute; CDS, Chinese Diabetes Society; JCDCG, Joint Committee for Developing Chinese Guidelines on Prevention and Treatment of Dyslipidemia in Adults.

In the McNemar tests, the NCEP and JCDCG definitions performed similarly in sensitivity (P=0.238); two pairs showed no difference in specificity, including IDF versus JCDCG (P=0.229), and NCEP versus JCDCG (P=0.109). Other comparisons in sensitivity and specificity were significantly different (P values<0.05).

The IDF definition had better diagnostic accuracy (53.4%), followed by the NCEP and JCDCG (51.85%), and CDS (51.23%) definitions. The AHA/NHLBI definition had the smallest diagnostic accuracy (50.93%). Similarly, the IDF definition showed the biggest YI (6.79%). The AHA/NHLBI definition had the smallest YI (1.86%).

Discussion

In this study, the two groups had similar demographic characteristics (all P values>0.05). Most cases were middle-aged and elderly people, and women were older than men (t=2.051, P=0.042). These characteristics were consistent with those of Chinese CHD populations.^27,29,31,39

The prevalence of metabolic syndrome in the case group was similar with previous surveys in Chinese CHD patients, but much lower than that in western CHD patients.^22,25 This may be due to ethnic differences, or the older age of their participants. Although the prevalence of metabolic syndrome varied among different definitions, no significant difference was detected between groups in this study (P values>0.05). One possible reason may be the age characteristics of the participants. To match with the cases in age, most participants of the control group (79.6%) were aged 45–64 years. In Chinese adults, people aged 45–64 years had the highest prevalence of metabolic syndrome.³ The high risks for metabolic syndrome in the middle-aged group may conceal the differences in metabolic syndrome prevalence between groups. Another possible reason may be that the controls were recruited from hospitalized patients in which the prevalence was much higher than that in the general population.^3,20,26,28 The differences may have been more significant if the controls were selected with more stringent criteria, e.g., with no past health history of hypertension or diabetes.

The AHA/NHLBI and IDF definitions were more sensitive, but with lower specificity. Conversely, the CDS definition that proposed more stringent thresholds was more specific, but less sensitive. This result echoed that strict diagnostic tests were generally associated with higher specificity and lower sensitivity.⁴⁰ McNemar tests confirmed the different performances of the IDF, AHA/NHLBI, and CDS definitions in sensitivity and specificity. They also indicated that the local definition of JCDCG worked similarly with some international ones. The sequences of definitions in sensitivity, specificity, and predictive values were consistent with previous studies.^20,26 However, this study showed higher sensitivity and PPV, but lower specificity and NPV than previous Chinese studies.^20,26 The differences were reasonable because all these indicators were related with disease prevalence.⁴⁰ Previous studies observed few incidences of CHD (1.39% and 3.01%).^20,26 In our study, half of the participants were ACS patients, which led to higher sensitivity and PPV, but lower specificity and NPV.

As an indicator that combines sensitivity and specificity, LR is more useful in estimating the likelihood of a disease for each patient. A larger LR+ and smaller LR− indicate better performances of a test.³⁵ In this study, the IDF definition had better LRs, indicating its better ability to distinguish ACS. Moreover, the IDF definition had a greater OR (1.32) and diagnostic accuracy (53.4%) and provided optimal cutoff points. The better performances of IDF-metabolic syndrome were also reported in previous studies.^24,25,28 Its advantages may come from the mandatory requirement of central obesity, which has been recognized as an independent indicator for CVD.³¹ In this study, the ACS patients had higher WC and BMI than the controls (P=0.002 and 0.005).

In this study, all definitions performed marginally in PVs, LRs, and OR, which raised a concern that metabolic syndrome may not be enough to predict ACS. This may be caused by the limitations of metabolic syndrome definitions. Current metabolic syndrome definitions only include obesity, raised BP, hyperglycemia, and dyslipidemia. Other well-acknowledged risk factors, such as gender, age, smoking status, total cholesterol, family history of hypercholesterolemia, and CVD, are not mentioned in the definitions.³¹ Moreover, current metabolic syndrome definitions do not include low-density lipoprotein cholesterol (LDL-C), which plays a critical role in the genesis of ACS.⁴¹ LDL-C and family history of hypercholesterolemia may be considered for predicting ACS in future studies. Future studies could explore the performances of more comprehensive methods including more CVD risk factors in predicting ACS.

Current findings were different from some western and Chinese studies, which reported stronger predictive powers.^20,22,25,26 On one hand, the discrepancies may come from different end points. For example, some studies used CVD (including both CHD and stroke) as the end point; ²⁶ some applied fatal and nonfatal CHD,^20,22 whereas our study used newly diagnosed nonfatal ACS. Although the incidents of nonfatal ACS may be observed in previous studies, no data were reported separately for this end point. On the other hand, participants' distinct characteristics in ethnic, age, and gender may contribute to the different results. The prevalence of metabolic syndrome and ACS varied across ethnic groups, which would lead to different results in Chinese and western studies.^42,43 Even in China, the prevalence was different between northern and southern populations.^3,39 Previous Chinese studies recruited more women with younger age from southern China.^20,26,28,29 This study recruited more middle-aged and elderly (45–64 years) men from northwestern China. Therefore, results of previous studies were not comparable with ours. The current study demonstrated the performances of metabolic syndrome in detecting onsets of nonfatal ACS in northwestern Chinese adults.

As to feasibility, a good metabolic syndrome definition should be simple to apply with low cost. All studied definitions require simple assessments in terms of WC, BMI, BP, FPG, TGs, and HDL-C. Therefore, studied definitions were all feasible in Chinese clinical practices.

In summary, the greatest significance of metabolic syndrome is to detect high-risk people for CVD at an early stage.^42,44 From this perspective, a better metabolic syndrome definition should identify more at-risk individuals, which means a higher PPV, sensitivity, and LR+.^35,40 At the same time, a good definition should also be accurate and specific enough to exclude healthy people.^36,40 In this study, the AHA/NHLBI definition had higher sensitivity, but it performed poorly in PPV, specificity, LRs, and diagnostic accuracy. The IDF definition showed balanced sensitivity and diagnostic accuracy and also had better PVs, LRs, and YI. Overall, the IDF definition performed better among all definitions.

There were some limitations in this study. First, the case–control study design had inherent limitations in confirming the cause–effect relationships. But it is more efficient to recruit enough patients, providing more statistical power at relatively lower costs.^25,31 Second, our study was conducted in two university-affiliated hospitals in a big city in northwestern China, and all ACS patients were recruited in the acute phase of the first onsets. Caution is advised when generalizing the results to other Chinese populations. Third, participants of the control group may still have subclinical CHD or other CVD. Although they were screened by symptoms and ECG reports carefully to fit the study criteria, none received exercise stress tests or coronary angiography. Future studies are suggested using more accurate methods to exclude subclinical CHD patients from the controls, e.g., exercise stress tests.

In spite of these limitations, this was the first study to compare both Chinese metabolic syndrome definitions with other definitions in detecting the onset of nonfatal ACS. The results provide valuable evidence for selecting a suitable metabolic syndrome definition to identify high-risk people for onset of ACS. Given the inclusion of three widely used international definitions—NCEP, IDF, and AHA/NHLBI—in this study, the findings should also provide references for studies in other populations. Future studies are suggested that use a multicentered prospective study design in general populations to further explore the performance of metabolic syndrome in predicting ACS. Other risk factors like LDL-C, family history of hypercholesterolemia and CVD-related diseases, age, gender, and smoking should be assessed together with metabolic syndrome in the prediction of ACS.

In conclusion, current metabolic syndrome definitions performed differently in various perspectives. The IDF definition may perform better in detecting the onset of nonfatal ACS in the northwestern Chinese population. All studied definitions were feasible in Chinese clinical practices. This study should provide references for future studies in other populations.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Alberti

, Eckel

, Grundy

, et al. International Diabetes Federation Task Force on Epidemiology and Prevention, Hational Heart, Lung, and Blood Institute, American Heart Association.World Heart Federation, International Atherosclerosis Society, International Association for the Study of Obesity. Harmonizing the metabolic syndrome: A joint interim statement of the International Diabetes Federation task force on epidemiology and prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation, 2009; 120:1640–1645.

The International Diabetes Federation. The IDF consensus worldwide definition of the metabolic syndrome. 2006. Available at www.idf.org/webdata/docs/IDF_Meta_def_final.pdf Accessed December 1, 2012 .

, Reynolds

, Wu

, et al. Prevalence of the metabolic syndrome and overweight among adults in china. Lancet, 2005; 365:1398–1405.

, Qiao

, Tuomilehto

, et al. Prevalence of the metabolic syndrome and its relation to all-cause and cardiovascular mortality in non-diabetic European men and women. Arch Intern Med, 2004; 164:1066–1076.

Gami

, Witt

, Howard

, et al. Metabolic syndrome and risk of incident cardiovascular events and death: A systematic review and meta-analysis of longitudinal studies. J Am Coll Cardiol, 2007; 49:403–414.

Mottillo

, Filion

, Genest

, et al. The metabolic syndrome and cardiovascular risk: A systematic review and meta-analysis. J Am Coll Cardiol, 2010; 56:1113–1132.

Thomas

, Lewis

, Hyde

, et al. “The solution needs to be complex.” Obese adults' attitudes about the effectiveness of individual and population based interventions for obesity. BMC Public Health, 2010; 10:420.

Clavijo

, Pinto

, Kuchulakanti

, et al. Metabolic syndrome in patients with acute myocardial infarction is associated with increased infarct size and in-hospital complications. Cardiovasc Revasc Med, 2006; 7:7–11.

The World Bank. Toward a healthy and harmonious life in China: Stemming the rising tide of non-communicable diseases. 2011; 62318-CN. Available at www.worldbank.org/en/news/feature/2011/07/26/toward-health-harmonious-life-china-stemming-rising-tide-of-non-communicable-diseases Accessed March 5, 2013 .

10.

World Health Organization. Definition, diagnosis and classification of diabetes mellitus and its complications. 1999;WHO/NCD/NCS/99.2.

11.

National Cholesterol Education Programme. 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). 2001; NIH no. 01-3670.

12.

, He

, Hu

, et al. Prevalence of metabolic syndrome and its influencing factors among the Chinese adults: The China Health and Nutrition Survey in 2009. Prev Med, 2013; 57:867–871.

13.

, Jia

, Zheng

, et al. Metabolic syndrome and stroke recurrence in Chinese ischemic stroke patients—the ACROSS-China study. PLoS One, 2012; 7:e51406.

14.

Zeng

, Zhu

, Zhang

, et al. Metabolic syndrome and the early detection of impaired glucose tolerance among professionals living in Beijing, China: A cross sectional study. Diabetol Metab Syndr, 2013; 5:65.

15.

Cai

, Huang

, Xu

, et al. Prevalence and determinants of metabolic syndrome among women in Chinese rural areas. PLoS One, 2012; 7:e36936.

16.

, Yang

, Lv

, et al. Prevalence of the metabolic syndrome in the Yan-an region of northwest China. J Int Med Res, 2012; 40:673–680.

17.

Group of Metabolic Syndrome in Chinese Diabetes Society. Suggestions on definition of metabolic syndrome from Chinese Diabetes Society. Chinese J Diabetes, 2004; 12:156–161.

18.

Joint Committee for Developing Chinese Guidelines on Prevention and Treatment of Dyslipidemia in Adults. Chinese guidelines on prevention and treatment of dyslipidemia in adults. Chinese J Cardiol, 2007; 35:1076–1095.

19.

, Zhao

, Zhou

, et al. Cut offs and risk stratification of dyslipidemia in Chinese adults. Chinese J Cardiol, 2007; 35:6.

20.

Wang

, Hou

, Bao

, et al. The metabolic syndrome increased risk of cardiovascular events in Chinese—a community based study. Int J Cardiol, 2010; 139:159–165.

21.

Jia

, Lu

, Xiang

, et al. Prediction of abdominal visceral obesity from body mass index, waist circumference and waist-hip ratio in Chinese adults: Receiver operating characteristic curves analysis. Biomed Environ Sci, 2003; 16:206–211.

22.

Bataille

, Perret

, Dallongeville

, et al. Metabolic syndrome and coronary heart disease risk in a population-based study of middle-aged men from France and northern Ireland: A nested case-control study from the PRIME cohort. Diabetes Metab, 2006; 32(5 Pt 1):475–479.

23.

Lawlor

, Smith

, Ebrahim

. Does the new International Diabetes Federation definition of the metabolic syndrome predict CHD any more strongly than older definitions? Findings from the British Women's Heart and Health Study. Diabetologia, 2006; 49:41–48.

24.

Lorenzo

, Williams

, Hunt

, et al. The National Cholesterol Education Program-Adult Treatment Panel III, International Diabetes Federation, and World Health Organization definitions of the metabolic syndrome as predictors of incident cardiovascular disease and diabetes. Diabetes Care, 2007; 30:8–13.

25.

Koutsovasilis

, Protopsaltis

, Triposkiadis

, et al. Comparative performance of three metabolic syndrome definitions in the prediction of acute coronary syndrome. Int Med, 2009; 48:179–187.

26.

Zhou

, Guo

, Yu

, et al. Evidence on the applicability of the ATPIII, IDF and CDS metabolic syndrome diagnostic criteria to identify CVD and T2DM in the Chinese population from a 6.3-year cohort study in mid-eastern China. Diabetes Res Clin Pract, 2010; 90:319–325.

27.

Zhu

, Gao

, Huang

, et al. The predictive values of metabolic syndrome to coronary heart disease. Chinese J Cardiovasc Rev, 2004; 2:201–203.

28.

, Li

, Hung

, et al. Metabolic syndrome defined by IDF and AHA/NHLBI correlates better to carotid intima-media thickness than that defined by NCEP ATP III and WHO. Diabetes Res Clin Pract, 2009; 85:335–341.

29.

Chang

, Chu

, Wang

, et al. Differences in prevalence and severity of coronary artery disease by three metabolic syndrome definitions. Can J Cardiol, 2012; 28:208–214.

30.

Wang

, Li

, Kinnunen

, et al. How well does the metabolic syndrome defined by five definitions predict incident diabetes and incident coronary heart disease in a Chinese population?. Atherosclerosis, 2007; 192:161–168.

31.

Yusuf

, Hawken

, Ounpuu

, et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): Case-control study. Lancet, 2004; 364:937–952.

32.

Cannon

, Battler

, Brindis

, et al. American College of Cardiology key data elements and definitions for measuring the clinical management and outcomes of patients with acute coronary syndromes. A report of the American College of Cardiology Task Force on clinical data standards (acute coronary syndromes writing committee). J Am Coll Cardiol, 2001; 38:2114–2130.

33.

Ryan

, Bauman

, Kennedy

, et al. Guidelines for percutaneous transluminal coronary angioplasty. A report of the American Heart Association/American College of Cardiology Task Force on assessment of diagnostic and therapeutic cardiovascular procedures (committee on percutaneous transluminal coronary angioplasty). Circulation, 1993; 88:2987–3007.

34.

World Health Organization. Waist circumference and waist–hip ratio: Report of a WHO expert consultation. Geneva, December 8–11, 2008.

35.

Akobeng

. Understanding diagnostic tests 2: Likelihood ratios, pre- and post-test probabilities and their use in clinical practice. Acta Paediatr, 2007; 96:487–491.

36.

Eusebi

. Diagnostic accuracy measures. Cerebrovas Dis, 2013; 36:267–272.

37.

Trajman

, Luiz

. McNemar chi² test revisited: Comparing sensitivity and specificity of diagnostic examinations. Scand J Clin Lab Invest, 2008; 68:77–80.

38.

Fluss

, Faraggi

, Reiser

. Estimation of the Youden index and its associated cutoff point. Biometr J, 2005; 47:458–472.

39.

Liu

, Yu

, Pan

, et al. Cross-sectional study on the prevalence of metabolic syndrome in patients with coronary heart disease in China. Natl Med J China, 2006; 86:2095–2098.

40.

Akobeng

. Understanding diagnostic tests 1: Sensitivity, specificity and predictive values. Acta Paediatr, 2007; 96:338–341.

41.

Stone

, Robinson

, Lichtenstein

, et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: A report of the American College of Cardiology/American Heart Association task force on practice guidelines. J Am Coll Cardiol, 2013; 11:002–086.

42.

Levesque

, Lamarche

. The metabolic syndrome: Definitions, prevalence and management. J Nutrigenet Nutrigenom, 2008; 1:100–108.

43.

Ford

, Li

, Zhao

. Prevalence and correlates of metabolic syndrome based on a harmonious definition among adults in the US. J Diabetes, 2010; 2:180–193.

44.

Desroches

, Lamarche

. The evolving definitions and increasing prevalence of the metabolic syndrome. Appl Physiol Nutr Metab, 2007; 32:23–32.

Comparing the Predictive Abilities of Different Metabolic Syndrome Definitions for Acute Coronary Syndrome: A Case–Control Study in Chinese Adults

Abstract

Background:

Methods:

Results:

Conclusion:

Introduction

Methods

Study design

Study settings and participants

Definitions of metabolic syndrome

NCEP definition

IDF definition for Chinese

AHA/NHLBI definition

CDS definition

JCDCG definition

Data collection procedures

Statistical analysis

Results

Discussion

Footnotes

Author Disclosure Statement

References