Background
Throughout the world, the prevalence of the metabolic syndrome (MetS) is rising [
1,
2]. MetS is set as a group of several characteristics including a broadened waist circumference (WC) caused by hypernutrition, a sedentary lifestyle, and sequential abdominal fat accumulation [
1,
2]. Dr. Reaven first described the term “syndrome X” in 1988 with simultaneously grouped risk factors [
3]. Syndrome X, later termed MetS, is recognized as a multiplex risk factor for type 2 diabetes and atherosclerotic cardiovascular disease (ASCVD) due to high correlations with insulin resistance. MetS is currently described as the grouping of abdominal obesity, insulin resistance, elevated blood pressure (BP), and obesity-related dyslipidemia, while also highly correlated with other clinical disorders such as nonalcoholic fatty liver disease, a prothrombotic or proinflammatory state, and reproductive problems [
4‐
6]. Previous studies have revealed that MetS is correlated with unfavorable clinical outcomes, including a fivefold increased risk of new diagnosed type 2 diabetes, and an approximate twofold increased risk of ASCVD [
5].
Currently, there are many different MetS definitions, as the MetS is a combination of different unfavorable conditions with the joint pathophysiology of abdominal obesity and insulin resistance, and not just one single disease [
5]. Moreover, the prevalence and outcome data of the MetS depend on these criteria used and vary [
7,
8]. Hence, it is necessary to find a champion criterion for MetS evaluation. Even though some different MetS criteria were modulated to the one in 2009 [
9], the National Cholesterol Education Program-Adult Treatment Panel III (ATP III), adjusted ATP III by the American Heart Association and the National Heart, Lung, and Blood Institute (ATP III/AHA/NHLBI) [
10] and the International Diabetes Federation (IDF) [
11] are still the most commonly used criteria throughout the world [
12].
Atherosclerosis-related arterial stiffness is examined by pulse wave velocity (PWV), which is highly associated with ASCVD and has been well-studied as a powerful risk factor for coronary artery disease [
13] and all-cause death [
14] in general populations. Moreover, the PWV is also a strong risk factor for death, beyond systolic blood pressure (SBP), among patients with glucose intolerance or diabetes [
15]. However, the associations between the different MetS criteria and clinical outcomes have not yet been elucidated. The purpose of this current study was to objectively compare the performance of ATP III/AHA/NHLBI- and IDF-MetS scores for the risks of atherosclerosis and diabetes through a holistic approach in a general population.
Discussion
To the best of our knowledge, this is the first study to compare, through AUC–ROC, the performance of ATP III/AHA/NHLBI- and IDF-MetS scores to assess the risks of atherosclerosis-related arterial stiffness and diabetes in a general population. In the 26,735 enrolled subjects, 6633 (24.8%) were classified as having MetS on the basis of the ATP III/AHA/NHLBI criteria, compared to 7388 (27.6%) on the basis of the IDF criteria. The stepwise raises in the baPWV and prevalence of diabetic-level hyperglycemia corresponded significantly to both MetS scores, except for a score of 0 in the IDF-MetS, indicating no abdominal obesity. The performance of the ATP III/AHA/NHLBI-MetS score for a higher baPWV and diabetic-level hyperglycemia was superior to the IDF-MetS score. The optimal cutoff points of the ATP III/AHA/NHLBI-MetS score were more than or equal to two, with a batter sensitivity from 32.7 to 62.2% to correlate with a higher baPWV and from 62.6 to 89.4% to correlate with diabetic-level hyperglycemia.
In the enrolled subjects, 24.8 and 27.6% were classified as having MetS on the basis of the ATP III/AHA/NHLBI and IDF criteria, respectively. Retrospectively, a study from Taiwan in 2002 reported a value of 14.3% with the IDF-MetS [
22], almost half of this study. Compared with other Asian countries, the prevalence of the ATP III/AHA/NHLBI-MetS in this study was even more than 13.9% reported in a study from China in 2000–01 [
23] and 16.7% in a study from Hong Kong in 1994–96 [
24], although the unadjusted ATP III criteria of a fasting glucose level more than or equal to 6.1 mmol/L was used. Accordingly, it was believed that the raising prevalence of the MetS, in particular its approximate twofold increase in Taiwan, mirrors a broadening WC worldwide as the consequence of hypernutrition, a sedentary lifestyle, and sequential abdominal fat accumulation [
2,
5]. In addition, a study conducted in the United States from 1999 to 2002 reported that the ATP III/AHA/NHLBI and IDF criteria identified 34.5 and 39.0% [
25], respectively, and the lower rate with the ATP III/AHA/NHLBI-MetS score is similar to our results. The lower WC cutoff value of the IDF-MetS may have contributed to this result.
We also showed that stepwise raises in both MetS scores corresponded to the baPWV and prevalence of diabetic-level hyperglycemia, as they were highly correlated with the joint pathophysiology of insulin resistance and represent groups of risk predictors of diabetes and ASCVD [
5]. In our previous study, stepwise raises in the MetS-associated risk score, containing a more component of high-sensitivity C-reactive protein, also corresponded to the baPWV in a population without or with chronic kidney disease, and it persisted as a strong determinant of baPWV in the presence of chronic kidney disease or not [
19]. In another of our previous studies, stepwise raises in the MetS-associated risk score still corresponded to baPWV before and after fifty in both genders, while it even persisted as a powerful determinant of baPWV only after 50 years old [
20]. The conclusions of the above-mentioned studies are compatible with the current study with regards to the significance of the MetS score.
Since these different MetS definitions were established for more than 10 years, differences in sex-specific characteristics and prevalence of the ATP III/AHA/NHLBI- and IDF-MetS among the same population have been frequently published [
8,
25]. However, objective comparisons of both MetS definitions to predict clinical outcomes are less common. In addition, for the MetS patients, the prediction of the risk of outcomes according to the definition of MetS is one of most important.
In this study, we showed that through a holistic approach, the performance of the ATP III/AHA/NHLBI-MetS score for the risks of atherosclerosis-related arterial stiffness and diabetes was superior to the IDF-MetS score for Asians. Considerable individual variation has been reported in the clinical pattern of metabolic factors in patients with insulin resistance or obesity [
26]. This means that the various expressions of metabolic factors originate from the interactions of genetic factors and environmental exposure [
9].
In a cross-sectional survey, Hwang et al. reported that since the IDF-MetS criteria failed to identify non-obese subjects with hypertension, diabetes and dyslipidemia, the ATP III-MetS definition was better than the IDF-MetS definition to detect subjects at a higher risk of ASCVD [
22]. This is similar to our finding, however their study lacked an objective and direct comparison of the two MetS definitions. Another previous study used the Framingham score calculation and found a higher risk of ASCVD with the ATP III/AHA/NHLBI-MetS than the IDF-MetS (11.3% vs. 6.1%) [
27]. This finding is also similar to ours, however the previous study included further information such as smoking habit beyond MetS itself and lacked a comparison of the risk of diabetes. A recent cross-sectional study enrolling a Chinese elderly population also reported that the updated ATPIII-MetS criteria was a better definition than the IDF-MetS criteria for screening high-risk subjects of ischemic stroke (OR 1.64 vs. 1.55) [
28].
By contrast, in the prospective San Antonio Heart Study with 7.4 years of follow-up, the ATP III/AHA/NHLBI and IDF criteria were both associated with an increased but similar ratio of incident CVD and diabetes [
29]. In the Framingham Offspring Study with 11 years of follow-up, the ATP III/AHA/NHLBI and IDF criteria both identified a similar increase in the risk of diabetes, while the former may have required insulin resistance to identify an increased risk of ASCVD [
30]. In fact, if patients continue to visit an outpatient clinic, longitudinal studies cannot exclude major interference of any additional interventions such as medications, weight reduction or lifestyle modifications during long-term follow-up so that compared to cross-sectional studies, longitudinal studies cannot always detect the natural course and actual clinical outcomes of MetS. This may explain the significant differences in the risk of ASCVD between both MetS definitions in cross-sectional studies [
22,
27,
28] but not in longitudinal studies [
29,
30].
Interestingly, we also showed that for Asians, the presence of more than or equal to 2 of 5 components of the ATP III/AHA/NHLBI-MetS criteria yielded a better sensitivity than the current criteria. This is a novel finding and very important because increasing obesity caused by a hypernutrition and a sedentary lifestyle [
1,
5] is predicted to result in an increase in the global prevalence of diabetes from 6.4 to 7.7%, affecting 439 million adults by 2030 [
31]. The MetS is an evolving concept that continues to be data driven and evidence based. For Chinese populations, lowering the threshold of the MetS criteria may remind us of the importance of earlier lifestyle modifications and weight reduction to reduce the further burden of diabetes and ASCVD. However, there is still a need for more evidence.
The study has some limitations. First, some data containing smoking habits, comorbidities, and medications were missed. The information of the questionnaires including family and personal history of chronic diseases and lifestyle optionally are routinely obtained for patients seeking health check. The information may be incomplete, and has not been verified. To avoid misunderstanding we didn’t include the data in our manuscript. However, subjects were mostly asymptomatic and relatively healthy people, who needed to have their health status checked in a self-paid Health Care Center, but different from the National Health Insurance Program in Taiwan [
16,
19,
20]. Hence, we believed that the subjects from the Health Care Center are relatively healthy populations. Moreover, the examinations were done after an overnight fast, indicating no consumption of any medications or food for 12 h. At last, samples may have originated from a group of certain wealth and lifestyle that do not exactly represent the whole population of Taiwan. However, due to high accessibility of our Heath Care Center service and a large sample size, the research is believed to be adequate for general representations.
Second, we were unsure as to the appropriate WC cutoff value of the ATP III/AHA/NHLBI-MetS in our Chinese population. However, in the published literature, the same WC cutoff value has been applied in other Asian nations including China [
23] and Hong Kong [
24]. Third, it was hard to make clear for the associations of the multiple MetS definitions and their outcomes behind our novel conclusions through a holistic approach since this was not a prospective and longitudinal randomized controlled trial. However, this cross-sectional study may identify more objectively the risks of further unfavorable MetS outcomes just now, compared to longitudinal trials, since we minimized disturbances of any supplementary treatments or unexpected changes of physical conditions during long-term follow up.