Introduction

The metabolic syndrome has been addressed as one of the major public health targets worldwide.1, 2 Some committees have engineered definitions of the metabolic syndrome;3, 4, 5, 6, 7 the Japanese government started a nation-wide intervention strategy for the metabolic syndrome in April 2008.8

Traditionally, the metabolic syndrome involves obesity,9 leads to a cluster of other cardiovascular risk factors and increases the risk of cardiovascular disease.10, 11, 12, 13, 14, 15, 16, 17 The American Heart Association/National Heart, Lung, and Blood Institute (AHA/NHLBI) defined the metabolic syndrome as a constellation of cardiovascular risk factors including obesity,5 whereas the criteria of the International Diabetes Federation (IDF)6 were based on the presence of obesity. The metabolic syndrome, by the IDF criteria, is defined as a condition of obese persons with certain cardiovascular risk factors; thus, non-obese persons are not diagnosed with the metabolic syndrome, even if they have any of these cardiovascular risk factors. Earlier Japanese studies showed that underweight individuals also had a high risk of hemorrhagic stroke,18 due to various pathological mechanisms.19 It is of value to examine the criteria of the metabolic syndrome that effectively predict the risk of cardiovascular disease among less-obese populations, such as the Japanese.20

We examined the contribution of the metabolic syndrome and its components to the incidence of ischemic heart disease and stroke among Japanese men and women, among whom there is a low prevalence of obesity.

Methods

Study cohort

We used two cohorts in this study. The Japan Public Health Center-based Prospective Study (JPHC Study) initiated the first (Cohort I) and the second cohort (Cohort II), registered in 27 administrative districts supervised by nine public health centers.21 The subjects were identified using population registries maintained by the local municipalities. Data regarding lifestyle, serum total and high-density lipoprotein (HDL) cholesterol, serum triglycerides, glucose, blood pressure, height in stocking feet and weight in light clothing were available for 8582 men and 15 521 women aged 40–69 years; these subjects responded to the self-administered questionnaire and undertook health examinations conducted by municipal governments in 1995 (Cohort I) and 1993–1994 (Cohort II). Together, these questionnaires and health examinations represented the baseline survey. We excluded the subjects who reported having cancer, myocardial infarction, angina pectoris or stroke at the baseline survey, leaving 23 313 people (8249 men and 15 064 women) for analysis. We informed all study subjects of the study details. The protocol of the study was approved by the institutional review board of the National Cancer Center, Tokyo, Japan.

The subjects were followed-up from the date of the baseline survey to 31 December 2003. Residence and survival were confirmed annually using the residential registers maintained by each municipality. For subjects who moved out of the study area, we used the data reported by the municipal office within the area to which they had moved. In Japan, residency and death registration are required by law, and the registries are believed to be complete. Information on the cause of death was obtained through the death certificate provided by the Ministry of Health, Labor, and Welfare after the Ministry of Internal Affairs and Communications granted permission. The study protocol was approved by the Human Ethics Review Committees of the National Cancer Center and the University of Tsukuba.

Baseline survey

Serum total and HDL cholesterol and triglycerides were measured in 23 laboratories. The precision and accuracy of lipid measurement in all laboratories were satisfactory, according to the Osaka Medical Center for Health Science and Promotion, a member of the Cholesterol Reference Method Laboratory Network.22 Arterial blood pressure was measured using a standard mercury sphygmomanometer applied to the right arm of the seated participant after a 5-min rest. Height in stocking feet and weight in light clothing were measured. Body mass index (BMI) was calculated as weight (kg) divided by the square of the height in meters (m2). A self-administrated questionnaire was carried out to ascertain the smoking status, the number of cigarettes smoked per day and the usual weekly intake of alcohol in go units (a Japanese traditional unit of volume corresponding to 23 g ethanol).

A modified definition by the AHA/NHLBI5 and the IDF6 was used to categorize the subjects according to the number of components of the metabolic syndrome shown. As waist circumference was not measured in this study, BMI 25.0 kg m−2 was used as the criterion for obesity; this BMI level was reported to correspond well with the Asian criterion for high waist circumference of 85 cm in men and 90 cm in women and 100 cm2 of visceral fat area.23 The components of the metabolic syndrome were defined as follows: (1) high blood pressure: blood pressure 130/85 mm Hg and/or medication use; (2) high glucose: glucose 5.55 mmol l−1 (100 mg per 100 ml) fasting or 7.77 mmol l−1 (140 mg per 100 ml) non-fasting and/or on treatment; (3) low HDL cholesterol: HDL cholesterol <1.03 mmol l−1 (40 mg per 100 ml) for men and <1.29 mmol l−1 (50 mg per 100 ml) for women; (4) high triglycerides: high serum triglycerides 1.69 mmol l−1 (150 mg per 100 ml); and (5) overweight: BMI 25.0 kg m−2. The metabolic syndrome was defined as the presence of three or more of the components (high blood pressure, high glucose, low HDL cholesterol, high triglycerides and overweight) according to the modified criteria of the AHA/NHLBI and the presence of two or more of the cardiovascular risk factors (high blood pressure, high glucose, low HDL cholesterol and high triglycerides) among individuals who were overweight according to the modified criteria of the IDF.

Confirmation of ischemic heart disease and stroke subtypes

For confirmation of cardiovascular events, a total of 78 hospitals were registered in the sampling area of the study cohort.24 All were major hospitals capable of treating patients with acute ischemic heart disease and stroke. To reduce uncaptured events, we also sent follow-up questionnaires to participants and reviewed death certificates for identifying suspected cardiovascular events. Therefore, all suspected events of cardiovascular disease were collected using ascertainment resources, including active patient notification, follow-up questionnaires and death certificate diagnoses. For the final diagnosis of cardiovascular events, physicians blinded to the patients' lifestyle data reviewed the medical records at each hospital.

The details of the surveillance for ischemic heart disease were described earlier.25 Myocardial infarction was confirmed in the medical records according to the criteria of the Multinational MONItoring of trends and determinants in CArdiovascular disease (MONICA) project,26 which requires evidence from electrocardiograms, cardiac enzymes and/or autopsy. When such workup was not performed and typical chest pain was present, a probable diagnosis was made. In the absence of diagnosis for myocardial infarction, deaths that occurred within 1 h from the onset of symptoms were regarded as sudden cardiac deaths.

Strokes were confirmed according to the criteria of the National Survey of Stroke,27, 28 which requires a constellation of neurological deficits of sudden or rapid onset lasting at least 24 h or until death. For each subtype of stroke (namely, hemorrhagic stroke (intraparenchymal hemorrhage and subarachnoid hemorrhage) or ischemic stroke (thrombotic or embolic stroke)), a definite diagnosis was established based on the examination of computer tomographic scan, magnetic resonance images or autopsy.

Statistical analysis

Outcome was defined as the new development of a primary ischemic heart disease or stroke during the study period. For each subject, person-months of follow-up were calculated from the baseline questionnaire collected to the first endpoint, death, emigration or 1 January 2004, whichever came first. Overall, 3% of the subjects had moved out of the communities (740 people) or were lost to follow-up (45 people) and were treated as censored.

Sex-specific hazard ratios of ischemic heart disease, stroke subtypes and the respective 95% confidence intervals (95% CI) were calculated with reference to the risk of individuals without the metabolic syndrome, without each individual component or with none of these components, using the Cox proportional hazards model. The multivariable hazard ratios were adjusted for potential confounding factors, including age at baseline (years), study area (nine public health center areas), time since last meal (<8 h vs. 8 h), total cholesterol levels (mg per 100 ml), smoking status (never smoked, ex-smokers, <20 cigarettes per day, 20–29 cigarettes per day, 30 cigarettes per day) and ethanol intake (non-drinkers, occasional drinkers (<75 or 75–149 g per week), 150–299, 300–449, or 450 g per week). The sex-specific association between the metabolic syndrome and the risk of cardiovascular disease was examined, stratified by the presence or absence of overweight. We also calculated the population-attributable fraction (PAF) to examine the contribution of the metabolic syndrome and its components to the risk of cardiovascular disease, using multivariable hazard ratios and the proportion of cases in each category.29 The PAF was estimated as pd × (HR−1)/HR, where pd is the proportion of cases falling into the category and HR the hazard ratio in the category. The 95% confidence limits (95% CL) for the PAF were calculated using the bootstrap method.30

All statistical tests were two-sided and a P-value <0.05 was regarded as statistically significant. All statistical analyses were conducted using the SAS, version 9.13 (SAS Institute, Inc., Cary, NC, USA).

Results

A total of 23 313 people (8249 men and 15 064 women) were followed for a median of 11. 0 years; 395 men and 298 women presented with cardiovascular disease as either ischemic heart disease (82 men and 40 women) or stroke (314 men and 258 women). Ischemic heart disease included 95 definite myocardial infarctions, 9 probable myocardial infarctions and 18 sudden cardiac deaths. We also identified the stroke subtypes: ischemic stroke (206 men and 142 women) or hemorrhagic stroke (108 men and 116 women). Hemorrhagic stroke was classified into either intraparenchymal hemorrhage (90 men and 70 women) or subarachnoid hemorrhage (18 men and 46 women). Of the total cardiovascular disease events, the percentage of ischemic cardiovascular disease (ischemic heart disease and ischemic stroke combined) was 72.9% in men and 61.1% in women. The incidence of stroke (398 incidents per 100 000 person-years in men and 172 incidents per 100 000 person-years in women) was 3.8-fold higher in men and 6.5-fold higher in women than that of ischemic heart disease (104 incidents per 100 000 person-years in men and 27 incidents per 100 000 person-years in women). With regard to stroke incidence, the percentage of intraparenchymal hemorrhage was 29% in men and 27% in women.

Men were older and had higher mean systolic and diastolic blood pressure, serum glucose and triglycerides compared with women. Moreover, men had lower mean total and HDL cholesterol and BMI compared with women (Table 1). Men were more likely to use medication for diabetes, drink heavily, smoke and have metabolic syndrome; however, they were less likely to use medication for hypertension. The percentage of obesity (BMI30 kg m−2) was 2.3% in men and 4.0% in women.

Table 1 Sex-specific baseline characteristics of cardiovascular risk factors, the metabolic syndrome based on modified criteria of the American Heart Association/National Heart, Lung, and Blood Institute (AHA/NHLBI) and the International Diabetes Federation (IDF) and its components

We showed a significant increased risk of ischemic cardiovascular disease with regard to each component of the metabolic syndrome in men (Table 2). The multivariable hazard ratio of ischemic cardiovascular disease among men was 2.37 (1.72–3.25) for high blood pressure, 1.46 (1.13–1.89) for high glucose, 1.82 (1.36–2.42) for low HDL cholesterol, 1.50 (1.16–1.94) for high triglycerides and 1.44 (1.12–1.85) for being overweight among men. These associations were weaker among women (Table 3); the respective hazard ratios were 2.28 (1.57–3.31), 1.44 (0.99–2.08), 1.40 (1.04–1.90), 1.17 (0.83–1.65) and 1.13 (0.83–1.53) for high blood pressure, high glucose, low HDL cholesterol, high triglycerides and for being overweight.

Table 2 Sex-specific hazard ratios (95% confidence interval) and population-attributable fraction (95% confidence limits) of cardiovascular disease according to the metabolic syndrome based on modified criteria of the American Heart Association/National Heart, Lung, and the Blood Institute (AHA/NHLBI) and international diabetes federation (IDF) and its components among men
Table 3 Sex-specific hazard ratios (95% confidence interval) and population-attributable fraction (95% confidence limits) of cardiovascular disease according to the metabolic syndrome based on modified criteria of the American Heart Association/National Heart, Lung, and Blood Institute (AHA/NHLBI) and the International Diabetes Federation (IDF), and its components among women

There was no significant excess risk of hemorrhagic stroke for any component of the metabolic syndrome, except for high blood pressure. Among men, the multivariable hazard ratio for high blood pressure was 2.75 (1.62–4.67) for hemorrhagic stroke, 2.77 (1.55–4.95) for intraparenchymal hemorrhage and 2.52 (0.70–8.98) for subarachnoid hemorrhage. The respective hazard ratios were 3.98 (2.38–6.64), 4.75 (2.33–9.67) and 3.13 (1.48–6.58) among women.

Men with metabolic syndrome based on the AHA/NHLBI criteria (25.1% of the population) had a 2.3-fold higher risk of ischemic heart disease and a 1.9-fold higher risk of ischemic stroke, although men with metabolic syndrome based on the IDF criteria (18.5% of the population) had a non-significant excess risk of ischemic heart disease and a 1.9-fold higher risk of ischemic stroke compared with men without metabolic syndrome, after adjusting for age, study area and other cardiovascular risk factors. The multivariable hazard ratios for ischemic cardiovascular disease were 1.97 (1.54–2.51) for metabolic syndrome based on the AHA/NHLBI criteria and 1.82 (1.39–2.38) for metabolic syndrome based on the IDF criteria among men. The respective hazard ratios among women were 1.51 (1.10–2.06) and 1.49 (1.06–2.08).

The PAFs of ischemic heart disease, total stroke and stroke subtypes were between 43 and 69% for high blood pressure among men and women; these values were much higher than those of the other components. The PAF (95% CL) of ischemic cardiovascular disease was 19% (11,26) for the metabolic syndrome based on the AHA/NHLBI criteria and 12% (6,18) for the metabolic syndrome based on the IDF criteria among men (P for difference=0.003), whereas the respective PAF among women was 12% (2,21) and 8% (1,16) (P for difference=0.12).

We found no significant excess risk of hemorrhagic stroke for the metabolic syndrome based on either the AHA/NHLBI or the IDF criteria among both men and women.

Increasing the number of metabolic syndrome components increased the risk of ischemic heart disease among men and that of ischemic stroke among men and women (P for trend <0.0001) (Tables 4 and 5). Although women with 1–3 components had a higher risk of hemorrhagic stroke than did women with 0 or 4 components, there was no dose-response association between the number of components of the metabolic syndrome and the risk for either sex (P for trend >0.05). Further, there was no dose-response association of the metabolic syndrome components with either intraparenchymal or subarachnoid hemorrhage (P for trend >0.05).

Table 4 Sex-specific hazard ratios (95% confidence interval) and population-attributable fraction (95% confidence limits) of cardiovascular disease according to the number of components of the metabolic syndrome among men
Table 5 Sex-specific hazard ratios (95% confidence interval) and population-attributable fraction (95% confidence limits) of cardiovascular disease according to the number of components of the metabolic syndrome among women

We also analyzed the association between the metabolic syndrome and the risk of ischemic cardiovascular disease after study participants were stratified into non-overweight and overweight individuals (Table 6). Among men, the multivariable hazard ratio of ischemic cardiovascular disease was 1.90 (1.14–3.14) in non-overweight persons with 1 risk factor (any component of the metabolic syndrome, except overweight), 2.72 (1.65–4.48) in non-overweight persons with 2 risk factors, 2.23 (1.21–4.11) in overweight persons with 1 risk factor and 3.61 (2.17–6.00) in overweight persons with 2 risk factors. The respective hazard ratios were 1.73 (0.99–3.02), 2.08 (1.19–3.64), 1.73(0.91–3.27) and 2.38 (1.36–4.18) among women. The prevalence was 29% in non-overweight persons with 1 risk factor, 24% in non-overweight persons with 2 risk factors, 9% in overweight persons with 2 risk factors and 19% in overweight persons with 2 risk factors among men. The respective prevalence was 26, 19, 12 and 16% among women.

Table 6 Sex-specific hazard ratios (95% confidence interval) and population-attributable fraction (95% confidence limits) of ischemic cardiovascular disease according to the number of components of the metabolic syndrome (except overweight) stratified by overweight (BMI25 kg m−2)

Men with 2 risk factors had 20% (12,28) of the PAF in ischemic cardiovascular disease among non-overweight individuals and 19% (13,26) of the PAF among overweight people compared with non-overweight people with no risk factors (as reference). The respective PAFs were 14% (4,23) and 15% (6,24) among women. Non-overweight persons with 1 risk factors represented 33% (18,47) of the PAF with regard to ischemic cardiovascular disease among men and 25% (8,41) of the PAF among women, whereas overweight persons with 1 risk factors represented 23% (15,32) of the PAF among men and 20% (8,32) of the PAF among women (data not shown).

Discussion

In this large population-based prospective study of Japanese people, we observed significant associations of the metabolic syndrome with the risk of ischemic heart disease and ischemic stroke, but not with hemorrhagic stroke. People with the metabolic syndrome had 1.5- to 2.0-fold higher risks of ischemic cardiovascular disease than people without it, and the contribution of the metabolic syndrome to ischemic cardiovascular disease was between 8 and 19%. However, the excess risks of or contributions to cardiovascular disease did not exceed those of high blood pressure.

The magnitude of the relative risks of the metabolic syndrome for total cardiovascular incidence and stroke incidence was smaller in our study than in earlier meta-analyses. There was no association between the metabolic syndrome and incidence of ischemic heart disease among women; the relative risk ranged from 1.5 to 2.2 for the incidence of cardiovascular disease, was 1.8 for the incidence of stroke and ranged from 1.5 to 1.6 for the incidence of ischemic heart disease.15, 16

The contribution of the metabolic syndrome based on the modified criteria of the IDF was lower than that based on the modified criteria of the AHA/NHLBI. The weaker predictive value was because of the inclusion of a high-risk group among non-overweight persons. An earlier study among the Korean population showed that the IDF criteria failed to detect 44.9% of men and 16.6% of women with the metabolic syndrome identified by the revised NCEP/ATP criteria, because of the mis-identification of non-obese high-risk persons.31 In our Japanese population, non-overweight persons with 2 risk factors constituted a large proportion of the population (24% in men and 19% in women). The PAFs for ischemic cardiovascular disease among non-overweight people with 2 risk factors were 20% in men and 14% in women, similar to or larger than those among overweight people with 2 risk factors.

Moreover, non-overweight men with one risk factor represented a similarly large PAF for cardiovascular disease, whereas overweight men with one risk factor did not. An earlier study also showed that non-overweight persons with one risk factor had a 2.0-fold higher risk of cardiovascular death.14 In our study, high blood pressure exhibited a higher contribution with regard to the incidence of cardiovascular disease compared with other components of the metabolic syndrome. Earlier studies support the importance of hypertension as a component of the metabolic syndrome and as a risk factor for the development of cardiovascular disease.32, 33, 34 These data suggested that an effect of non-clustered risk factors, such as high blood pressure, should also be taken into account for the prevention of cardiovascular disease among Japanese people, characterized by a low prevalence of obesity.

Metabolic syndrome, as defined by the Japanese committee (Committee to Evaluate Diagnostic Standards for Metabolic Syndrome),7, 17 as well as by the IDF, is a condition in obese persons with certain cardiovascular risk factors. Non-obese persons can therefore be considered not to have the metabolic syndrome, even if they possess any of those cardiovascular risk factors. Our finding showed that these criteria, based on the presence of obesity, dismiss high-risk populations that are not overweight.

There was no dose-response association between the metabolic syndrome and the risk of hemorrhagic stroke. This finding was expected because hemorrhagic stroke has different pathological mechanisms,19 than those involved in ischemic heart disease or ischemic stroke. Further, being overweight is not associated with the risk of hemorrhagic stroke.18

The strengths of this study include the large population-based sample of middle-aged men and women and the use of standardized methods for the measurement of serum lipids and risk characteristics. The stroke surveillance was almost complete and a high percentage of stroke events was confirmed using imaging studies (97%).

The limitations of the study were, first, that we did not measure waist circumference at the baseline survey. However, earlier studies have shown that a BMI of 25.0 kg m−2 is equal to the 100 cm2 of visceral fat area that defines central obesity,23 and that an increased risk of cardiovascular disease or that of ischemic heart disease was observed only among people with BMI 25 kg m−2.18 Second, we used non-fasting data, in particular non-fasting serum triglycerides 1.69 mmol l−1 (150 mg per 100 ml), as a component of the metabolic syndrome. Although the justification for the use of the same cutoff point as fasting status is debated, the data represented by non-fasting triglycerides can be used because of its significant predictive ability with regard to ischemic heart disease.35

In conclusion, this study provides epidemiological evidence that the metabolic syndrome was associated with the risks of ischemic heart disease and ischemic stroke, but not with the risk of hemorrhagic stroke. The metabolic syndrome based on the AHA/NHLBI criteria predicts the risk of ischemic cardiovascular disease better than that based on the IDF criteria among less-obese populations. This is because non-overweight people with cardiovascular risk factors are also at a high risk for ischemic heart disease and stroke.