- Email: [email protected]
Impact of Metabolic Syndrome on the Risk of Cardiovascular Disease Mortality in the United States and in Japan
Impact of Metabolic Syndrome on the Risk of Cardiovascular Disease Mortality in the United States and in Japan Longjian Liu, MD, PhD, MSca,*, Katsuyuki Miura, MD, PhDb, Akira Fujiyoshi, MD, PhDb, Aya Kadota, MD, PhDb,c, Naoko Miyagawa, MScb, Yasuyuki Nakamura, MD, PhDd, Takayoshi Ohkubo, MD, PhDe, Akira Okayama, MD, PhDf, Tomonori Okamura, MD, PhDg, and Hirotsugu Ueshima, MD, PhDb,h The United States has a higher prevalence of metabolic syndrome (MS) and cardiovascular disease (CVD) mortality than Japan, but it is unknown how much of the difference in MS accounts for the mortality difference. The aim of this study was to examine the impact of MS on the excess CVD mortality in the United States compared with that in Japan. Data from the United States Third National Health and Nutrition Examination Survey (NHANES III; n [ 12,561) and the Japanese National Integrated Project for Prospective Observation of Noncommunicable Disease and Its Trends in Aged (NIPPON DATA; n [ 7,453) were analyzed. MS was defined as ‡3 of 5 risk factors (obesity, high blood pressure, decreased high-density lipoprotein cholesterol, elevated glycosylated hemoglobin, and elevated triglycerides). The results show that after a median of 13.8 years of follow-up in the United States, 1,683 patients died from CVD (11.75 per 1,000 person-years), and after a median of 15 years of follow-up in Japan, 369 patients died from CVD (3.56 per 1,000 person-years). The ageadjusted prevalence of MS was 26.7% in the United States and 19.3% in Japan. Of 5 MS factors, obesity, high blood pressure, elevated triglycerides, and glycosylated hemoglobin in the United States, and high blood pressure and elevated glycosylated hemoglobin in Japan were significant risk factors for CVD mortality. Estimates of 13.3% and 44% of the excess CVD mortality for the United States could be explained by the higher prevalence of MS and MS plus baseline CVD history than in Japan. In conclusion, the present study is the first to quantitatively demonstrate that MS and MS plus baseline CVD history may significantly contribute to the explanation of excess CVD mortality in the United States compared with Japan. Ó 2014 Elsevier Inc. All rights reserved. (Am J Cardiol 2014;113:84e89) Cardiovascular disease (CVD) has been established as a clear threat to human health. A number of studies have reported that metabolic risk factors, a group of cofactors, increase the risk for CVD.1e6 The United States has higher CVD mortality than most developed countries, including Japan. However, few internationally comparative studies have been conducted to examine the impact of metabolic factors on the risk for CVD mortality across countries.7 Given the differences in the burden of CVD and potential differences in the risk factors, a comparative study of the outcome and risk factors, using nationally representative
samples, may add new insights into the studies of CVD. In the present study, we hypothesized that the prevalence of metabolic risk factors was significantly higher in the United States and that the difference is a significant contributor to the excess CVD mortality for the United States compared with Japan. To test these hypotheses, we used data from the United States Third National Health and Nutrition Examination Survey (NHANES III)8 and the Japanese National Integrated Project for Prospective Observation of Noncommunicable Disease and Its Trends in Aged (NIPPON DATA).4,9e12
a Department of Epidemiology and Biostatistics, Drexel University School of Public Health, Philadelphia, Pennsylvania; bDepartment of Health Science and hLifestyle-Related Disease Prevention Center, Shiga University of Medical Science, Otsu, Japan; cDepartment of School Nursing and Health Education, Osaka Kyoiku University, Kashiwara, Japan; dCardiovascular Epidemiology, Kyoto Women’s University, Kyoto, Japan; eDepartment of Hygiene and Public Health, Teikyo University School of Medicine, Tokyo, Japan; fThe First Institute for Health Promotion and Health Care, Japanese AntiTuberculosis Association, Tokyo, Japan; and gDepartment of Preventive Medicine and Public Health, School of Medicine, Keio University, Tokyo, Japan. Manuscript received June 12, 2013; revised manuscript received and accepted August 20, 2013. The National Integrated Project for Prospective Observation of Noncommunicable Disease and Its Trends in Aged (NIPPON DATA) was
supported by a grant-in-aid from the Ministry of Health and Welfare under the auspices of the Japanese Association for Cerebrocardiovascular Disease Control; a research grant for cardiovascular diseases (7A-2) from the Ministry of Health, Labor, and Welfare; and a Health and Labor Sciences Research Grant, Japan (Comprehensive Research on Aging and Health: H11-Chouju-046, H14-Chouju-003, H17-Chouju-012; Comprehensive Research on Cardiovascular Disease and Diabetes: H22-Seisyu-017). Dr. Liu was named a U.S.-Japan Bridge Fellow of the Japan Society for the Promotion of Science for the U.S.-Japan CVD Comparison Study. See page 89 for disclosure information. *Corresponding author: Tel: (215) 762-1370; fax: (215) 762-1174. E-mail address: [email protected] (L. Liu).
0002-9149/13/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjcard.2013.08.042
www.ajconline.org
Preventive Cardiology/MS and CVD in the U.S. and Japan
85
Table 1 Baseline characteristics of participants aged 30 years in the Third National Health and Nutrition Examination Survey (1988 to 1994) and in the National Integrated Project for Prospective Observation of Noncommunicable Disease and Its Trends in Aged (1990) Variable
Men United States (n ¼ 5,896)
Continuous variables Age (yrs) BMI (kg/m2) Systolic BP (mm Hg) Diastolic BP (mm Hg) Total cholesterol (mg/dl) HDL cholesterol (mg/dl) TG (mg/dl)† HbA1c (%) Categorical variables Smoking status Never Former Current Current alcohol usez Hypertensionx Hypercholesterolemiajj Diabetes mellitus{ Coronary heart disease# Stroke#
49.48 27.00 126.83 78.01 208.83 45.32 165.73 5.49
0.36 0.11 0.43 0.22 1.03 0.38 3.10 0.02
Women
Japan (n ¼ 3,129)
p Value*
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
53.36 22.96 138.01 83.78 198.58 50.37 147.62 5.37
0.24 0.05 0.36 0.21 0.66 0.27 1.88 0.01
United States (n ¼ 6,665) 51.32 26.98 123.86 73.27 212.64 55.34 139.33 5.45
0.52 0.17 0.54 0.22 0.94 0.42 3.09 0.03
Japan (n ¼ 4,324)
p Value*
0.03 <0.001 <0.001 <0.001 <0.001 0.002 <0.001 <0.001
52.48 22.86 133.72 79.62 206.82 56.73 121.12 5.25
0.21 0.05 0.32 0.18 0.59 0.23 1.20 0.01
<0.001 53.77% 8.76% 37.47% 62.00% 37.47% 35.31% 9.14% 6.13% 2.49%
20.29% 24.32% 55.39% 65.10% 53.563% 17.034% 8.69% 3.10% 2.52%
<0.001 <0.001 <0.001 0.35 <0.001 0.13
<0.001 71.63% 4.77% 23.60% 39.16% 37.15% 36.92% 9.57% 3.27% 2.49%
88.32% 2.52% 9.16% 7.54% 46.21% 22.94% 5.20% 2.61% 1.36%
<0.001 <0.001 <0.001 <0.001 0.10 <0.001
Data are expressed as mean SEM or percentages. To convert total and HDL cholesterol from milligrams per deciliter to millimoles per liter, divide by 38.61. To convert TG from milligrams per deciliter to millimoles per liter, divide by 89. * For age-adjusted tests using analysis of covariance for continuous variables and chi-square tests for categorical variables, expect for testing difference in age between countries. † Nonfasting in Japanese data. z Defined by self-report, if a subject had consumed 1 drink of any type (beer, wine, or liquor) per month. x Defined as self-report of physician diagnosis of hypertension, systolic BP 140 mm Hg or diastolic BP 90 mm Hg, or use of antihypertensive medication. jj Defined as self-report of physician diagnosis of hyperlipidemia or total cholesterol 240 mg/dl (6.2 mmol/L). { Defined as self-report of physician diagnosis of diabetes mellitus or serum HbA1c 6.5% or use of medication to treat diabetes. # Defined as self-report of physician diagnosis of each disease.
Methods NHANES III (1988 to 1994) is a nationwide survey conducted by the National Center for Health Statistics of the Centers for Disease Control and Prevention in the United States, which gathers information representing the health and nutritional status of the noninstitutionalized civilian United States population aged 2 months.8 The study consists of interviews, physical examinations, and data from blood sample analyses. A detailed description of the survey and its sampling procedures is available at the NHANES III Web site.8 The NHANES III mortality follow-up study was conducted and linked with the National Death Index.13 This linked study provides mortality follow-up from the date of baseline NHANES III participants (1988 to 1994) through December 31, 2006.13 In the study, we included participants aged 30 years because CVD mortality is substantially lower in those aged <30 years. Of 15,042 participants aged 30 years, we excluded 22 who had ineligible measurements on their vital statistics and 2,459 who did not complete all measurements of 5 metabolic factors (body mass index [BMI], blood pressure [BP], serum high-density lipoprotein cholesterol, triglyceride [TG], and/or glycemia). The remaining 12,561 subjects (83.5% of 15,042) were included in the study sample (5,896 male, 6,665 female). Data from NIPPON DATA90 used in the present study were
approved by the NIPPON DATA steering committee and the institutional review board of Shiga University of Medical Science. NIPPON DATA90, supported by the Ministry of Health and Welfare of Japan, is a cohort study of representative Japanese subjects aged 30 years at baseline surveys (1990). Participants in the study were randomly selected from 300 districts throughout Japan. Data from physical examinations, blood tests, and a self-administered questionnaire on lifestyle were collected in person at individual districts’ health care centers. Mortality follow-up was conducted for all participants from baseline to November 2005.9,14,15 For the purpose of the comparison, we used NIPPON DATA90 because it was conducted approximately in the same time period (1990 to 2005) and had the same measures and standardizations as NHANES III (1988 to 2006).9,11,12,16,17 Of 8,383 baseline participants aged 30 years, we excluded 284 who had invalid follow-up data and 646 who had no measurements of 5 metabolic risk factors. The remaining 7,453 patients (88.9% of 8,383) were included in the study. NHANES III was approved by the institutional review board of the National Center for Health Statistics.8,13 Several criteria for the definition of metabolic syndrome (MS) have been proposed.12,18e20 In the present study, we
86
The American Journal of Cardiology (www.ajconline.org)
Figure 1. Age-specific mortality rate (per 1,000 person-years) from CVD for adult men and women in the United States and in Japan.
Table 2 Multivariate-adjusted hazard ratios of individual metabolic syndrome factors and metabolic syndrome for cardiovascular mortality in the United States and Japan MS Relation to CVD Mortality
United States Rate*
In total participants 1a BMI 30 kg/m2 (U.S.), BMI 25 kg/m2 (Japan)† 1b BMI 30 kg/m2 (both countries)† 2 High BPz 3 Low HDL cholesterolx 4 Elevated triglyceridesjj 5 Elevated HbA1c{ MS (Japan, 3 of 1a to 5 components)# MS (U.S., 3 of 1b to 5 components)# In participants without baseline CVD histories 1a BMI 30 kg/m2 (U.S.), BMI 25 kg/m2 (Japan)† 1b BMI 30 kg/m2 (both countries)† 2 High BPz 3 Low HDL cholesterolx 4 Elevated triglyceridesjj 5 Elevated HbA1c{ MS (Japan, 3 of 1a to 5 components)# MS (U.S., 3 of 1b to 5 components)#
HR (95% CI)
Japan p Value
Rate*
HR (95% CI)
p Value
25.3% 25.3% 43.9% 38.0% 35.8% 25.7% 26.7% 26.7%
1.27 1.27 1.51 1.18 1.17 1.45 1.43 1.43
(1.05e1.55) (1.05e1.55) (1.18e1.93) (1.00e1.40) (1.02e1.35) (1.25e1.68) (1.24e1.64) (1.24e1.64)
0.017 0.017 <0.001 0.06 0.028 <0.001 <0.001 <0.001
23.7% 2.4% 60.5% 29.1% 19.1% 12.4% 19.3% 12.6%
0.80 0.89 2.62 1.06 1.21 1.64 1.35 1.53
(0.62e1.04) (0.39e2.06) (1.74e3.96) (0.85e1.33) (0.95e1.54) (1.31e2.07) (1.08e1.69) (1.20e1.94)
0.09 0.79 <0.001 0.59 0.13 <0.001 0.027 0.001
25.0% 25.0% 42.5% 37.2% 34.8% 24.9% 25.7% 25.7%
1.37 1.37 1.66 1.06 1.09 1.38 1.39 1.39
(1.10e1.69) (1.10e1.69) (1.27e2.16) (0.86e1.31) (0.92e1.28) (1.10e1.73) (1.17e1.66) (1.17e1.66)
0.01 0.01 <0.001 0.58 0.30 0.01 <0.001 <0.001
23.6% 2.4% 60.0% 29.0% 18.8% 12.0% 18.9% 12.2%
0.83 1.04 2.58 0.93 1.16 1.62 1.30 1.48
(0.62e1.10) (0.45e2.39) (1.68e3.97) (0.72e1.20) (0.88e1.52) (1.25e2.09) (1.01e1.67) (1.13e1.93)
0.18 0.94 <0.001 0.57 0.29 <0.001 0.039 <0.01
Baseline CVD history includes coronary heart disease and stroke. HRs were adjusted for age, gender, smoking status, alcohol consumption, and total cholesterol. * Age-adjusted rate using the world population as a standard across the 2 countries. † Obesity was defined per the American Heart Association definition (BMI 30 kg/m2) for the United States population and per the Japan Society for the Study of Obesity (BMI 25 kg/m2) for the Japanese population. z Defined as self-report of physician diagnosis of hypertension, systolic BP 130 mm Hg or diastolic BP 85 mmHg, or medication use. x Defined as HDL cholesterol <40 mg/dl (<1.0 mmol/L) for male participants and <50 mg/dl (<1.3 mmol/L) for female participants. jj Defined as nonfasting TG 200 mg/dl (2.3 mmol/L) in Japan and as fasting TG 150 mg/dl (1.7 mmol/L) in the United States. { Defined as HbA1c 5.7% or self-report of physician diagnosis of diabetes. We did not use glucose level, because fasting blood samples were not available for most participants in the Japanese data. # Defined as having 3 of the 5 factors (obesity, high BP, low HDL cholesterol, elevated TG, and elevated HbA1c).
used the World Health Organization and the American Heart Association criteria by including 5 factors: (1) Obesity was defined using the World Health Organization criteria of BMI 30 kg/m2. When doing data analysis for Japanese patients only, we also defined obesity using a cut-off point of BMI 25 kg/m2, according to the criteria of the Japan Society for
the Study of Obesity.21 (2) High BP was defined as systolic BP 130 mm Hg or diastolic BP 85 mm Hg or current use of antihypertensive medication. (3) Decreased high-density lipoprotein (HDL) cholesterol was defined as HDL cholesterol <40 mg/dl (<1.0 mmol/L) for male subjects and HDL cholesterol <50 mg/dl (<1.3 mmol/L) for female subjects.
Preventive Cardiology/MS and CVD in the U.S. and Japan
87
Table 3 Hazard ratios and 95% confidence intervals for the likelihood of cardiovascular disease mortality between the United States and Japan Model
Covariates in the Model
United States vs Japan HR (95% CI)
In total participants M1† M2 M3 M4 M5 M6 M7 M8 In participants without baseline CVD histories M1† M2 M3 M4 M5 M6 M7
p Value
% of Excess Mortality Accounted For*
Basic model Adjusted for Adjusted for Adjusted for Adjusted for Adjusted for Adjusted for Adjusted for
M1 M1 M1 M1 M1 M1 M7
þ þ þ þ þ þ þ
obesity (BMI 30 kg/m2) high BP low HDL cholesterol elevated TG elevated HbA1c MSz (3 of 5 factors) baseline CVD history
2.01 1.90 2.12 2.00 1.96 1.85 1.88 1.57
(1.84e2.21) (1.71e2.11) (1.93e2.32) (1.82e2.19) (1.79e2.15) (1.67e2.06) (1.71e2.06) (1.22e2.01)
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
Base model 11.26% 10.28% 1.48% 5.04% 15.81% 13.34% 44.17%
Basic model Adjusted for Adjusted for Adjusted for Adjusted for Adjusted for Adjusted for
M1 M1 M1 M1 M1 M1
þ þ þ þ þ þ
obesity (BMI 30 kg/m2) high BP low HDL cholesterol elevated TG elevated HbA1c MSz (3 of 5 factors)
1.81 1.69 1.92 1.81 1.79 1.69 1.70
(1.62e2.03) (1.49e1.92) (1.72e2.15) (1.62e2.03) (1.60e2.00) (1.48e1.92) (1.52e1.91)
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
Base model 15.04% 13.69% 0.25% 2.71% 15.17% 13.19%
Percentage of change ¼ (HR1 HR2)/(HR1 1) 100 (keeping model 1 as HR1). * Percentage of the excess CVD and all-cause mortality in the United States that was accounted for by adding the covariates from model 2 to model 8 (in addition to covariates adjusted in model 1). † Model 1, the basic model, was adjusted for age, gender, smoking, alcohol consumption, and total cholesterol level. z Defined as having 3 of the 5 factors (obesity, high BP, low HDL cholesterol, elevated TG, and elevated HbA1c).
(4) Elevated glucose was defined as serum glycosylated hemoglobin (HbA1c) 5.7%. We used HbA1c because it does not require a fasting blood sample. HbA1c level has been shown to be a highly reliable and correlated marker of fasting glucose in several studies when fasting sample is not available.22e25 Participants with previous diagnoses of diabetes or those using antidiabetic medications (insulin or oral agents) were classified as having elevated glucose.22 (5) Elevated serum TG was defined as serum TG 150 mg/dl (1.7 mmol/L) for a fasting sample or TG 200 mg/dl (2.3 mmol/L) for a nonfasting sample.3 Subjects with 3 of 5 MS components were classified having MS.6,24 The 2 countries’ CVD deaths were classified using the International Classification of Diseases, 10th Revision as CVD (codes I00 to I99), coronary heart disease (codes I20 to I25), and cerebrovascular disease (codes I60 to I69). A serial analysis was conducted. First, we described baseline characteristics of participants by gender and country. Analysis of covariance and chi-square tests were used in the descriptive analysis. The age-adjusted prevalence of MS was estimated by the direct method using the World Standard Population.26,27 Second, we used a Cox proportional-hazards regression model to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) of individual MS factors and MS for the risk for CVD mortality. To verify the Cox proportional-hazards assumptions, we used plots of the log (log) survival curves and Schoenfeld residuals. Third, we examined the impact of MS on the excess CVD mortality of the United States compared with Japan by conducting 8 multivariate models. Model 1 estimated the HR of the United States versus Japan for CVD mortality and was adjusted for 5 covariates (age, gender,
smoking status, alcohol consumption, and total cholesterol). Model 1 served as the base model. Models 2 to 6 adjusted for the same 5 covariates used in model 1 along with each of 5 MS components separately in a step-by-step manner. Model 7 examined the total impact of MS on CVD mortality, and model 8 included adjustment for baseline CVD conditions (coronary heart disease and stroke). The magnitudes of the impact of each MS component and MS on the excess CVD mortality of the United States compared with Japan were expressed using the formula (HR1 HR2)/ (HR1 1.0) 100%, where HR1 represents then HR derived from model 1, HR2 represents the HR after adjusting for additional covariates (i.e., models 2 to 8), and 1.0 represents the HR when there was no excess risk.28,29 Finally, survival functions of participants who had MS for the risk for CVD mortality were estimated and depicted by country and by the number of exposures to MS components. A repeated data analysis was conducted for participants who were free of baseline CVD conditions to examine whether the MS-CVD mortality associations remained observed. All data analyses were conducted using SAS version 9.2 (SAS Institute Inc., Cary, North Carolina). We used sampling weights and accounted for the complex sampling design using “SAS Procedures for Analysis of Sample Survey Data.”26,27 Two-sided p values 0.05 were considered as having statistical significance. Results Table 1 shows that the United States participants had significantly higher means of BMI, total cholesterol, TG and HbA1c and lower HDL cholesterol than the Japanese
88
The American Journal of Cardiology (www.ajconline.org)
participants. Differences in smoking and other risk factors between male and female participants were also observed in the U.S. and Japan. The mean follow-up period was 12.7 years (median 13.8) for the NHANES III participants and 13.9 years (median 15) for the Japanese participants. Of 12,561 United States participants, 1,683 died from CVD. Of CVD deaths, 1,058 patients (62.9%) died from coronary heart disease, 315 (18.7%) from cerebrovascular disease, and 310 (18.4%) from other CVD subtypes. Of 7,453 Japanese participants, 369 died from CVD. Of CVD deaths, 81 patients (22.0%) died from coronary heart disease, 160 (43.4%) from cerebrovascular disease, and 128 (34.6%) from other CVD subtypes. Figure 1 shows that the United States had significantly higher CVD mortality than Japan (11.75% vs 3.56%, p <0.001). Table 2 shows that the United States had significantly higher age-adjusted prevalence rates for individual MS factors, except for having a significantly lower prevalence of high BP than Japan (p <0.01 or p <0.05). The United States had significantly higher prevalence of MS than Japan (26.7% vs 19.3%, p <0.01), when obesity was defined as BMI 30 kg/m2 and BMI 25 kg/m2 for the United States and Japanese samples, respectively. Obesity, high BP, elevated TG, and elevated HbA1c significantly predicted CVD mortality in the United States population, and high BP and elevated HbA1c predicted CVD mortality in Japan (p <0.001 for all). The HR of MS for CVD mortality was 22.9% higher ([1.43 1.35]/[1.35 1]) in the United States than in Japan (p <0.001) when obesity was defined as BMI 30 kg/m2 for the United States and BMI 25 kg/m2 for Japan. However, the HR was 23.3% higher in Japan than in the United States (p <0.001) when obesity was classified as BMI 30 kg/m2 for the 2 populations. Similar associations were observed for participants without baseline CVD conditions. Table 3 shows that the HR of the United States versus Japan for CVD mortality was 2.01 (95% CI 1.84 to 2.21; model 1). After additionally adjusting for obesity (model 2), the HR was reduced to 1.90 (955 CI 1.71 to 2.11). However, the HR increased to 2.12 (95% CI 1.93 to 2.32) when adjusting for high BP (model 3). The HR was largely reduced (15.81%) when adjusting for elevated HbA1c (model 6). An overall 13.34% reduction was observed when adjusting for MS (model 7). The HR was further reduced to 44.17% when adjusting for MS plus baseline CVD conditions (model 8). Similar findings were observed in a subsample excluding those who had CVD conditions at baseline. No significant interaction effects of MS and country on CVD mortality were observed in the total study sample (HR 0.99, 95% CI 0.94 to 1.05, p ¼ 0.84) and in the subsample (HR 0.98, 95% CI 0.92 to 1.06, p ¼ 0.66). Discussion The present study, using nationally comparable databases, is the first to document and compare the differences in MS and the impact of MS on CVD mortality between the United States and Japan. The main findings suggest that the United States had a significantly higher prevalence of MS than Japan. An estimated 13% of the excess CVD mortality
could be explained by the differences in MS and 44% by MS plus baseline CVD conditions in the United States compared with Japan. Of the 5 MS components, obesity significantly predicted CVD mortality in the United States but not in Japan. The reason for the nonsignificant association in Japan is still unknown. Two conditions may partly explain this difference. First, increased BMI as a measure of obesity may be less sensitive than waist circumference. In the present study, we did not use waist circumference as the measure of obesity, because waist circumference was not measured in the Japanese survey. Second, because increased BMI has a stronger association with coronary heart disease than with stroke, the nonsignificant association between obesity and CVD might be due to a significantly different distribution of the subtypes of CVD, in which coronary heart disease is dominant in the United States, while stroke occurs more frequently in Japan.10 It should be noted that when obesity is defined as BMI 30 kg/m2 for the Japanese sample, the HR of MS for CVD mortality was higher in Japan than in the U.S. (1.53 vs 1.43; Table 2). We conducted further sensitivity analyses by comparing the prevalence of MS components in those with MS and with BMIs 30 kg/m2 between the United States and Japanese samples. The results showed that the prevalence rates of high BP, decreased HDL cholesterol, and elevated TG were significantly higher in the Japanese compared with Americans (97.5% vs 82.0% for high BP, 82.5% vs 71.7% for decreased HDL cholesterol, and 80.8% vs 74.7% for elevated TG). This difference may partly contribute to a higher HR of MS for CVD mortality in Japan when obesity is defined as BMI 30 kg/m2. However, it does not mean that Japanese have a higher risk for CVD, because the Japanese have a much lower prevalence of obesity (BMI 30 kg/m2) than Americas (2.4% vs 25.3%, p <0.001). Certainly the magnitudes of individual MS components using different cut-off points on the association between MS and CVD risk in different populations need to be further studied. In the study, we also observed that the HR of high BP for CVD mortality in the Japanese participants was higher than in the American participants. These findings may suggest a different impact of risk factors on CVD mortality across different countries and racial and ethnic populations. The present study had several strengths. First, its findings are derived from nationally representative population samples. The measurements of exposures and outcomes were conducted using standardized approaches. Therefore, it offers a unique opportunity for the study results to be generalized. Second, the association between MS and CVD mortality was analyzed prospectively, which makes it possible to interpret a potentially temporal association. Third, the findings of the study, by addressing the impacts of individual components of MS on the risk for CVD mortality between countries, are informative for health policy decisions and health practice in controlling CVD mortality. There were also limitations that should be kept in mind. First, individual MS components that are continuous measures (i.e., BMI, BP, TG, HDL, and HbA1c) were dichotomized on the basis of the current definition of MS. This dichotomization approach would lead a reduction in statistical power and an underestimation of the strengths of
Preventive Cardiology/MS and CVD in the U.S. and Japan
associations between these factors and CVD mortality. Second, all risk factors had single measurements at baseline. Therefore, we are unable to test any time-varying effects of these variables on CVD mortality. Third, because BMI may have less sensitivity for the measure of obesity, new biomarkers should be included in further studies. Despite these limitations, 2 clear and important conclusions follow from our present study. First, obesity, high BP, elevated TG, and elevated HbA1c in the United States and high BP and elevated HbA1c in Japan were the most significant individual predictors for CVD mortality. Second, MS and MS plus baseline CVD history are among the important risk factors that may significantly contribute to the explanation of excess CVD mortality in the United States versus Japan. Acknowledgment: We thank the members of the NIPPON DATA80 and NIPPON DATA90 Research Group for their important contributions. A list of the members is provided in Nakamura et al.4
12.
13.
14. 15.
16.
17.
Disclosures The authors have no conflicts of interest to disclose. 1. Lakka HM, Laaksonen DE, Lakka TA, Niskanen LK, Kumpusalo E, Tuomilehto J, Salonen JT. The metabolic syndrome and total and cardiovascular disease mortality in middle-aged men. JAMA 2002;288: 2709e2716. 2. Beckman JA, Creager MA, Libby P. Diabetes and atherosclerosis: epidemiology, pathophysiology, and management. JAMA 2002;287: 2570e2581. 3. Kadota A, Hozawa A, Okamura T, Kadowak T, Nakmaura K, Murakami Y, Hayakawa T, Kita Y, Okayama A, Nakamura Y, Kashiwagi A, Ueshima H; NIPPON DATA Research Group. Relationship between metabolic risk factor clustering and cardiovascular mortality stratified by high blood glucose and obesity: NIPPON DATA90 1990-2000. Diabetes Care 2007;30:1533e1538. 4. Nakamura Y, Okamura T, Higashiyama A, Watanabe M, Kadota A, Ohkubo T, Miura K, Kasagi F, Kodama K, Okayama A, Ueshima H; NIPPON DATA80 Research Group. Prognostic values of clockwise and counterclockwise rotation for cardiovascular mortality in Japanese subjects: a 24-year follow-up of the National Integrated Project for Prospective Observation of Noncommunicable Disease and Its Trends in the Aged 1980-2004 (NIPPON DATA80). Circulation 2012;125: 1226e1233. 5. Liu L, Nettleton JA, Bertoni AG, Bluemke DA, Lima JA, Szklo M. Dietary pattern, the metabolic syndrome, and left ventricular mass and systolic function: the Multi-Ethnic Study of Atherosclerosis. Am J Clin Nutr 2009;90:362e368. 6. Liu L, Liu Z, Ma M, Xue F, Sorel E. The cardiometabolic syndrome and risk of mortality from cardiovascular diseases and all causes among African Americans and white Americans. JCMD 2012;3:1e9. 7. Callow AD. Cardiovascular disease 2005—the global picture. Vascul Pharmacol 2006;45:302e307. 8. National Center for Health Statistics. National Health and Nutrition Examination Survey Data. Available at: http://www.cdc.gov/nchs/surveys. htm. Accessed April 16, 2010. 9. Ueshima H. NIPPON DATA. Nippon Rinsho 2006;64(suppl):108e111. 10. Ueshima H, Sekikawa A, Miura K, Turin TC, Takashima N, Kita Y, Watanabe M, Kadota A, Okuda N, Kadowaki T, Nakamura Y, Okamura T. Cardiovascular disease and risk factors in Asia: a selected review. Circulation 2008;118:2702e2709. 11. Nakamura K, Okamura T, Hayakawa T, Hozawa A, Kadowaki T, Murakami Y, Kita Y, Okayama A, Ueshima H; NIPPON DATA80 9RG. The proportion of individuals with obesity-induced hypertension
18.
19. 20.
21. 22. 23. 24.
25.
26. 27. 28. 29.
89
among total hypertensives in a general Japanese population: NIPPON DATA80, 90. Eur J Epidemiol 2007;22:691e698. Liu L, Okamura T, Kadowaki T, Murakami Y, Hozawa A, Kita Y, Takashima N, Okuda N, Okayama A, Ueshima H. Bundle branch block and other cardiovascular disease risk factors: US-Japan comparison. Int J Cardiol 2010;143:432e440. Data Linkage Team. Comparative analysis of the NHANES III publicuse and restricted-use linked mortality files: 2010 public-use data release—National Center for Health Statistics. Available at: http:// www.cdc.gov/nchs/data/datalinkage/nhanes3_mort_compare_2010_final. pdf. Accessed July 18, 2010. NIPPON DATA80 Research Group. Risk assessment chart for death from cardiovascular disease based on a 19-year follow-up study of a Japanese representative population. Circ J 2006;70:1249e1255. Liu L, Choudhury SR, Okayama A, Hayakawa T, Kita Y, Ueshima H. Changes in body mass index and its relationships to other cardiovascular risk factors among Japanese population: results from the 1980 and 1990 National Cardiovascular Surveys in Japan. J Epidemiol 1999;9: 163e174. Okamura T, Hayakawa T, Kadowaki T, Kita Y, Okayama A, Ueshima H; NIPPON DATA90 Research Group. The inverse relationship between serum high-density lipoprotein cholesterol level and all-cause mortality in a 9.6-year follow-up study in the Japanese general population. Atherosclerosis 2006;184:143e150. Ueshima H, Choudhury SR, Okayama A, Hayakawa T, Kita Y, Kadowaki T, Okamura T, Minowa M, Iimura O; NIPPON DATA80 Research Group. Cigarette smoking as a risk factor for stroke death in Japan NIPPON DATA80. Stroke 2004;35:1836e1841. Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, Gordon DJ, Krauss RM, Savage PJ, Smith SC Jr, Spertus JA, Costa F, American Heart Association, National Heart, Lung, and Blood Institute. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute scientific statement. Circulation 2005;112:2735e2752. Batsis JA, Nieto-Martinez RE, Lopez-Jimenez F. Metabolic syndrome: from global epidemiology to individualized medicine. Clin Pharmacol Ther 2007;82:509e524. Liu L, Nunez AE. Cardiometabolic syndrome and its association with education, smoking, diet, physical activity, and social support: findings from the Pennsylvania 2007 BRFSS survey. J Clin Hypertens (Greenwich) 2010;12:556e564. Kanazawa M, Yoshiike N, Osaka T, Numba Y, Zimmet P, Inoue S. Criteria and classification of obesity in Japan and Asia-Oceania. World Rev Nutr Diet 2005;94:1e12. American Diabetes Association. Executive summary: standards of medical care in diabetes—2010. Diabetes Care 2010;33(suppl): S4eS10. Selvin E, Steffes MW, Gregg E, Brancati FL, Coresh J. Performance of A1c for the classification and prediction of diabetes. Diabetes Care 2011;34:84e89. Grundy SM, Brewer HB Jr, Cleeman JI, Smith SC Jr, Lenfant C; American Heart Association, National Heart, Lung, and Blood Institute. Definition of metabolic syndrome: report of the National Heart, Lung, and Blood Institute/American Heart Association Conference on Scientific Issues Related to Definition. Circulation 2004;109:433e438. Ferreira I, Beijers HJ, Schouten F, Smulders YM, Twisk JW, Stehouwer CD. Clustering of metabolic syndrome traits is associated with maladaptive carotid remodeling and stiffening: a 6-year longitudinal study. Hypertension 2012;60:542e549. SAS Institute Inc. SAS/STAT User’s Guide, Version 9.1. Cary, North Carolina: SAS Institute Inc., 2005. Szklo M, Nieto FJ. Epidemiology: Beyond the Basics. Sudbury, Massachusetts: Jones & Bartlett, 2007. Liao Y, Greenlund KJ, Croft JB, Keenan NL, Giles WH. Factors explaining excess stroke prevalence in the US stroke belt. Stroke 2009;40:3336e3341. Cheng CY, Reich D, Haiman CA, Tandon A, Patterson N, Selvin E, Akylbekova EL, Brancati FL, Coresh J, Boerwinkle E, Altshuler D, Taylor HA, Henderson BE, Wilson JG, Kao WH. African ancestry and its correlation to type 2 diabetes in African Americans: a genetic admixture analysis in three u.s. population cohorts. PLoS One 2012;7: e32840.
samples, may add new insights into the studies of CVD. In the present study, we hypothesized that the prevalence of metabolic risk factors was significantly higher in the United States and that the difference is a significant contributor to the excess CVD mortality for the United States compared with Japan. To test these hypotheses, we used data from the United States Third National Health and Nutrition Examination Survey (NHANES III)8 and the Japanese National Integrated Project for Prospective Observation of Noncommunicable Disease and Its Trends in Aged (NIPPON DATA).4,9e12
a Department of Epidemiology and Biostatistics, Drexel University School of Public Health, Philadelphia, Pennsylvania; bDepartment of Health Science and hLifestyle-Related Disease Prevention Center, Shiga University of Medical Science, Otsu, Japan; cDepartment of School Nursing and Health Education, Osaka Kyoiku University, Kashiwara, Japan; dCardiovascular Epidemiology, Kyoto Women’s University, Kyoto, Japan; eDepartment of Hygiene and Public Health, Teikyo University School of Medicine, Tokyo, Japan; fThe First Institute for Health Promotion and Health Care, Japanese AntiTuberculosis Association, Tokyo, Japan; and gDepartment of Preventive Medicine and Public Health, School of Medicine, Keio University, Tokyo, Japan. Manuscript received June 12, 2013; revised manuscript received and accepted August 20, 2013. The National Integrated Project for Prospective Observation of Noncommunicable Disease and Its Trends in Aged (NIPPON DATA) was
supported by a grant-in-aid from the Ministry of Health and Welfare under the auspices of the Japanese Association for Cerebrocardiovascular Disease Control; a research grant for cardiovascular diseases (7A-2) from the Ministry of Health, Labor, and Welfare; and a Health and Labor Sciences Research Grant, Japan (Comprehensive Research on Aging and Health: H11-Chouju-046, H14-Chouju-003, H17-Chouju-012; Comprehensive Research on Cardiovascular Disease and Diabetes: H22-Seisyu-017). Dr. Liu was named a U.S.-Japan Bridge Fellow of the Japan Society for the Promotion of Science for the U.S.-Japan CVD Comparison Study. See page 89 for disclosure information. *Corresponding author: Tel: (215) 762-1370; fax: (215) 762-1174. E-mail address: [email protected] (L. Liu).
0002-9149/13/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjcard.2013.08.042
www.ajconline.org
Preventive Cardiology/MS and CVD in the U.S. and Japan
85
Table 1 Baseline characteristics of participants aged 30 years in the Third National Health and Nutrition Examination Survey (1988 to 1994) and in the National Integrated Project for Prospective Observation of Noncommunicable Disease and Its Trends in Aged (1990) Variable
Men United States (n ¼ 5,896)
Continuous variables Age (yrs) BMI (kg/m2) Systolic BP (mm Hg) Diastolic BP (mm Hg) Total cholesterol (mg/dl) HDL cholesterol (mg/dl) TG (mg/dl)† HbA1c (%) Categorical variables Smoking status Never Former Current Current alcohol usez Hypertensionx Hypercholesterolemiajj Diabetes mellitus{ Coronary heart disease# Stroke#
49.48 27.00 126.83 78.01 208.83 45.32 165.73 5.49
0.36 0.11 0.43 0.22 1.03 0.38 3.10 0.02
Women
Japan (n ¼ 3,129)
p Value*
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
53.36 22.96 138.01 83.78 198.58 50.37 147.62 5.37
0.24 0.05 0.36 0.21 0.66 0.27 1.88 0.01
United States (n ¼ 6,665) 51.32 26.98 123.86 73.27 212.64 55.34 139.33 5.45
0.52 0.17 0.54 0.22 0.94 0.42 3.09 0.03
Japan (n ¼ 4,324)
p Value*
0.03 <0.001 <0.001 <0.001 <0.001 0.002 <0.001 <0.001
52.48 22.86 133.72 79.62 206.82 56.73 121.12 5.25
0.21 0.05 0.32 0.18 0.59 0.23 1.20 0.01
<0.001 53.77% 8.76% 37.47% 62.00% 37.47% 35.31% 9.14% 6.13% 2.49%
20.29% 24.32% 55.39% 65.10% 53.563% 17.034% 8.69% 3.10% 2.52%
<0.001 <0.001 <0.001 0.35 <0.001 0.13
<0.001 71.63% 4.77% 23.60% 39.16% 37.15% 36.92% 9.57% 3.27% 2.49%
88.32% 2.52% 9.16% 7.54% 46.21% 22.94% 5.20% 2.61% 1.36%
<0.001 <0.001 <0.001 <0.001 0.10 <0.001
Data are expressed as mean SEM or percentages. To convert total and HDL cholesterol from milligrams per deciliter to millimoles per liter, divide by 38.61. To convert TG from milligrams per deciliter to millimoles per liter, divide by 89. * For age-adjusted tests using analysis of covariance for continuous variables and chi-square tests for categorical variables, expect for testing difference in age between countries. † Nonfasting in Japanese data. z Defined by self-report, if a subject had consumed 1 drink of any type (beer, wine, or liquor) per month. x Defined as self-report of physician diagnosis of hypertension, systolic BP 140 mm Hg or diastolic BP 90 mm Hg, or use of antihypertensive medication. jj Defined as self-report of physician diagnosis of hyperlipidemia or total cholesterol 240 mg/dl (6.2 mmol/L). { Defined as self-report of physician diagnosis of diabetes mellitus or serum HbA1c 6.5% or use of medication to treat diabetes. # Defined as self-report of physician diagnosis of each disease.
Methods NHANES III (1988 to 1994) is a nationwide survey conducted by the National Center for Health Statistics of the Centers for Disease Control and Prevention in the United States, which gathers information representing the health and nutritional status of the noninstitutionalized civilian United States population aged 2 months.8 The study consists of interviews, physical examinations, and data from blood sample analyses. A detailed description of the survey and its sampling procedures is available at the NHANES III Web site.8 The NHANES III mortality follow-up study was conducted and linked with the National Death Index.13 This linked study provides mortality follow-up from the date of baseline NHANES III participants (1988 to 1994) through December 31, 2006.13 In the study, we included participants aged 30 years because CVD mortality is substantially lower in those aged <30 years. Of 15,042 participants aged 30 years, we excluded 22 who had ineligible measurements on their vital statistics and 2,459 who did not complete all measurements of 5 metabolic factors (body mass index [BMI], blood pressure [BP], serum high-density lipoprotein cholesterol, triglyceride [TG], and/or glycemia). The remaining 12,561 subjects (83.5% of 15,042) were included in the study sample (5,896 male, 6,665 female). Data from NIPPON DATA90 used in the present study were
approved by the NIPPON DATA steering committee and the institutional review board of Shiga University of Medical Science. NIPPON DATA90, supported by the Ministry of Health and Welfare of Japan, is a cohort study of representative Japanese subjects aged 30 years at baseline surveys (1990). Participants in the study were randomly selected from 300 districts throughout Japan. Data from physical examinations, blood tests, and a self-administered questionnaire on lifestyle were collected in person at individual districts’ health care centers. Mortality follow-up was conducted for all participants from baseline to November 2005.9,14,15 For the purpose of the comparison, we used NIPPON DATA90 because it was conducted approximately in the same time period (1990 to 2005) and had the same measures and standardizations as NHANES III (1988 to 2006).9,11,12,16,17 Of 8,383 baseline participants aged 30 years, we excluded 284 who had invalid follow-up data and 646 who had no measurements of 5 metabolic risk factors. The remaining 7,453 patients (88.9% of 8,383) were included in the study. NHANES III was approved by the institutional review board of the National Center for Health Statistics.8,13 Several criteria for the definition of metabolic syndrome (MS) have been proposed.12,18e20 In the present study, we
86
The American Journal of Cardiology (www.ajconline.org)
Figure 1. Age-specific mortality rate (per 1,000 person-years) from CVD for adult men and women in the United States and in Japan.
Table 2 Multivariate-adjusted hazard ratios of individual metabolic syndrome factors and metabolic syndrome for cardiovascular mortality in the United States and Japan MS Relation to CVD Mortality
United States Rate*
In total participants 1a BMI 30 kg/m2 (U.S.), BMI 25 kg/m2 (Japan)† 1b BMI 30 kg/m2 (both countries)† 2 High BPz 3 Low HDL cholesterolx 4 Elevated triglyceridesjj 5 Elevated HbA1c{ MS (Japan, 3 of 1a to 5 components)# MS (U.S., 3 of 1b to 5 components)# In participants without baseline CVD histories 1a BMI 30 kg/m2 (U.S.), BMI 25 kg/m2 (Japan)† 1b BMI 30 kg/m2 (both countries)† 2 High BPz 3 Low HDL cholesterolx 4 Elevated triglyceridesjj 5 Elevated HbA1c{ MS (Japan, 3 of 1a to 5 components)# MS (U.S., 3 of 1b to 5 components)#
HR (95% CI)
Japan p Value
Rate*
HR (95% CI)
p Value
25.3% 25.3% 43.9% 38.0% 35.8% 25.7% 26.7% 26.7%
1.27 1.27 1.51 1.18 1.17 1.45 1.43 1.43
(1.05e1.55) (1.05e1.55) (1.18e1.93) (1.00e1.40) (1.02e1.35) (1.25e1.68) (1.24e1.64) (1.24e1.64)
0.017 0.017 <0.001 0.06 0.028 <0.001 <0.001 <0.001
23.7% 2.4% 60.5% 29.1% 19.1% 12.4% 19.3% 12.6%
0.80 0.89 2.62 1.06 1.21 1.64 1.35 1.53
(0.62e1.04) (0.39e2.06) (1.74e3.96) (0.85e1.33) (0.95e1.54) (1.31e2.07) (1.08e1.69) (1.20e1.94)
0.09 0.79 <0.001 0.59 0.13 <0.001 0.027 0.001
25.0% 25.0% 42.5% 37.2% 34.8% 24.9% 25.7% 25.7%
1.37 1.37 1.66 1.06 1.09 1.38 1.39 1.39
(1.10e1.69) (1.10e1.69) (1.27e2.16) (0.86e1.31) (0.92e1.28) (1.10e1.73) (1.17e1.66) (1.17e1.66)
0.01 0.01 <0.001 0.58 0.30 0.01 <0.001 <0.001
23.6% 2.4% 60.0% 29.0% 18.8% 12.0% 18.9% 12.2%
0.83 1.04 2.58 0.93 1.16 1.62 1.30 1.48
(0.62e1.10) (0.45e2.39) (1.68e3.97) (0.72e1.20) (0.88e1.52) (1.25e2.09) (1.01e1.67) (1.13e1.93)
0.18 0.94 <0.001 0.57 0.29 <0.001 0.039 <0.01
Baseline CVD history includes coronary heart disease and stroke. HRs were adjusted for age, gender, smoking status, alcohol consumption, and total cholesterol. * Age-adjusted rate using the world population as a standard across the 2 countries. † Obesity was defined per the American Heart Association definition (BMI 30 kg/m2) for the United States population and per the Japan Society for the Study of Obesity (BMI 25 kg/m2) for the Japanese population. z Defined as self-report of physician diagnosis of hypertension, systolic BP 130 mm Hg or diastolic BP 85 mmHg, or medication use. x Defined as HDL cholesterol <40 mg/dl (<1.0 mmol/L) for male participants and <50 mg/dl (<1.3 mmol/L) for female participants. jj Defined as nonfasting TG 200 mg/dl (2.3 mmol/L) in Japan and as fasting TG 150 mg/dl (1.7 mmol/L) in the United States. { Defined as HbA1c 5.7% or self-report of physician diagnosis of diabetes. We did not use glucose level, because fasting blood samples were not available for most participants in the Japanese data. # Defined as having 3 of the 5 factors (obesity, high BP, low HDL cholesterol, elevated TG, and elevated HbA1c).
used the World Health Organization and the American Heart Association criteria by including 5 factors: (1) Obesity was defined using the World Health Organization criteria of BMI 30 kg/m2. When doing data analysis for Japanese patients only, we also defined obesity using a cut-off point of BMI 25 kg/m2, according to the criteria of the Japan Society for
the Study of Obesity.21 (2) High BP was defined as systolic BP 130 mm Hg or diastolic BP 85 mm Hg or current use of antihypertensive medication. (3) Decreased high-density lipoprotein (HDL) cholesterol was defined as HDL cholesterol <40 mg/dl (<1.0 mmol/L) for male subjects and HDL cholesterol <50 mg/dl (<1.3 mmol/L) for female subjects.
Preventive Cardiology/MS and CVD in the U.S. and Japan
87
Table 3 Hazard ratios and 95% confidence intervals for the likelihood of cardiovascular disease mortality between the United States and Japan Model
Covariates in the Model
United States vs Japan HR (95% CI)
In total participants M1† M2 M3 M4 M5 M6 M7 M8 In participants without baseline CVD histories M1† M2 M3 M4 M5 M6 M7
p Value
% of Excess Mortality Accounted For*
Basic model Adjusted for Adjusted for Adjusted for Adjusted for Adjusted for Adjusted for Adjusted for
M1 M1 M1 M1 M1 M1 M7
þ þ þ þ þ þ þ
obesity (BMI 30 kg/m2) high BP low HDL cholesterol elevated TG elevated HbA1c MSz (3 of 5 factors) baseline CVD history
2.01 1.90 2.12 2.00 1.96 1.85 1.88 1.57
(1.84e2.21) (1.71e2.11) (1.93e2.32) (1.82e2.19) (1.79e2.15) (1.67e2.06) (1.71e2.06) (1.22e2.01)
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
Base model 11.26% 10.28% 1.48% 5.04% 15.81% 13.34% 44.17%
Basic model Adjusted for Adjusted for Adjusted for Adjusted for Adjusted for Adjusted for
M1 M1 M1 M1 M1 M1
þ þ þ þ þ þ
obesity (BMI 30 kg/m2) high BP low HDL cholesterol elevated TG elevated HbA1c MSz (3 of 5 factors)
1.81 1.69 1.92 1.81 1.79 1.69 1.70
(1.62e2.03) (1.49e1.92) (1.72e2.15) (1.62e2.03) (1.60e2.00) (1.48e1.92) (1.52e1.91)
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
Base model 15.04% 13.69% 0.25% 2.71% 15.17% 13.19%
Percentage of change ¼ (HR1 HR2)/(HR1 1) 100 (keeping model 1 as HR1). * Percentage of the excess CVD and all-cause mortality in the United States that was accounted for by adding the covariates from model 2 to model 8 (in addition to covariates adjusted in model 1). † Model 1, the basic model, was adjusted for age, gender, smoking, alcohol consumption, and total cholesterol level. z Defined as having 3 of the 5 factors (obesity, high BP, low HDL cholesterol, elevated TG, and elevated HbA1c).
(4) Elevated glucose was defined as serum glycosylated hemoglobin (HbA1c) 5.7%. We used HbA1c because it does not require a fasting blood sample. HbA1c level has been shown to be a highly reliable and correlated marker of fasting glucose in several studies when fasting sample is not available.22e25 Participants with previous diagnoses of diabetes or those using antidiabetic medications (insulin or oral agents) were classified as having elevated glucose.22 (5) Elevated serum TG was defined as serum TG 150 mg/dl (1.7 mmol/L) for a fasting sample or TG 200 mg/dl (2.3 mmol/L) for a nonfasting sample.3 Subjects with 3 of 5 MS components were classified having MS.6,24 The 2 countries’ CVD deaths were classified using the International Classification of Diseases, 10th Revision as CVD (codes I00 to I99), coronary heart disease (codes I20 to I25), and cerebrovascular disease (codes I60 to I69). A serial analysis was conducted. First, we described baseline characteristics of participants by gender and country. Analysis of covariance and chi-square tests were used in the descriptive analysis. The age-adjusted prevalence of MS was estimated by the direct method using the World Standard Population.26,27 Second, we used a Cox proportional-hazards regression model to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) of individual MS factors and MS for the risk for CVD mortality. To verify the Cox proportional-hazards assumptions, we used plots of the log (log) survival curves and Schoenfeld residuals. Third, we examined the impact of MS on the excess CVD mortality of the United States compared with Japan by conducting 8 multivariate models. Model 1 estimated the HR of the United States versus Japan for CVD mortality and was adjusted for 5 covariates (age, gender,
smoking status, alcohol consumption, and total cholesterol). Model 1 served as the base model. Models 2 to 6 adjusted for the same 5 covariates used in model 1 along with each of 5 MS components separately in a step-by-step manner. Model 7 examined the total impact of MS on CVD mortality, and model 8 included adjustment for baseline CVD conditions (coronary heart disease and stroke). The magnitudes of the impact of each MS component and MS on the excess CVD mortality of the United States compared with Japan were expressed using the formula (HR1 HR2)/ (HR1 1.0) 100%, where HR1 represents then HR derived from model 1, HR2 represents the HR after adjusting for additional covariates (i.e., models 2 to 8), and 1.0 represents the HR when there was no excess risk.28,29 Finally, survival functions of participants who had MS for the risk for CVD mortality were estimated and depicted by country and by the number of exposures to MS components. A repeated data analysis was conducted for participants who were free of baseline CVD conditions to examine whether the MS-CVD mortality associations remained observed. All data analyses were conducted using SAS version 9.2 (SAS Institute Inc., Cary, North Carolina). We used sampling weights and accounted for the complex sampling design using “SAS Procedures for Analysis of Sample Survey Data.”26,27 Two-sided p values 0.05 were considered as having statistical significance. Results Table 1 shows that the United States participants had significantly higher means of BMI, total cholesterol, TG and HbA1c and lower HDL cholesterol than the Japanese
88
The American Journal of Cardiology (www.ajconline.org)
participants. Differences in smoking and other risk factors between male and female participants were also observed in the U.S. and Japan. The mean follow-up period was 12.7 years (median 13.8) for the NHANES III participants and 13.9 years (median 15) for the Japanese participants. Of 12,561 United States participants, 1,683 died from CVD. Of CVD deaths, 1,058 patients (62.9%) died from coronary heart disease, 315 (18.7%) from cerebrovascular disease, and 310 (18.4%) from other CVD subtypes. Of 7,453 Japanese participants, 369 died from CVD. Of CVD deaths, 81 patients (22.0%) died from coronary heart disease, 160 (43.4%) from cerebrovascular disease, and 128 (34.6%) from other CVD subtypes. Figure 1 shows that the United States had significantly higher CVD mortality than Japan (11.75% vs 3.56%, p <0.001). Table 2 shows that the United States had significantly higher age-adjusted prevalence rates for individual MS factors, except for having a significantly lower prevalence of high BP than Japan (p <0.01 or p <0.05). The United States had significantly higher prevalence of MS than Japan (26.7% vs 19.3%, p <0.01), when obesity was defined as BMI 30 kg/m2 and BMI 25 kg/m2 for the United States and Japanese samples, respectively. Obesity, high BP, elevated TG, and elevated HbA1c significantly predicted CVD mortality in the United States population, and high BP and elevated HbA1c predicted CVD mortality in Japan (p <0.001 for all). The HR of MS for CVD mortality was 22.9% higher ([1.43 1.35]/[1.35 1]) in the United States than in Japan (p <0.001) when obesity was defined as BMI 30 kg/m2 for the United States and BMI 25 kg/m2 for Japan. However, the HR was 23.3% higher in Japan than in the United States (p <0.001) when obesity was classified as BMI 30 kg/m2 for the 2 populations. Similar associations were observed for participants without baseline CVD conditions. Table 3 shows that the HR of the United States versus Japan for CVD mortality was 2.01 (95% CI 1.84 to 2.21; model 1). After additionally adjusting for obesity (model 2), the HR was reduced to 1.90 (955 CI 1.71 to 2.11). However, the HR increased to 2.12 (95% CI 1.93 to 2.32) when adjusting for high BP (model 3). The HR was largely reduced (15.81%) when adjusting for elevated HbA1c (model 6). An overall 13.34% reduction was observed when adjusting for MS (model 7). The HR was further reduced to 44.17% when adjusting for MS plus baseline CVD conditions (model 8). Similar findings were observed in a subsample excluding those who had CVD conditions at baseline. No significant interaction effects of MS and country on CVD mortality were observed in the total study sample (HR 0.99, 95% CI 0.94 to 1.05, p ¼ 0.84) and in the subsample (HR 0.98, 95% CI 0.92 to 1.06, p ¼ 0.66). Discussion The present study, using nationally comparable databases, is the first to document and compare the differences in MS and the impact of MS on CVD mortality between the United States and Japan. The main findings suggest that the United States had a significantly higher prevalence of MS than Japan. An estimated 13% of the excess CVD mortality
could be explained by the differences in MS and 44% by MS plus baseline CVD conditions in the United States compared with Japan. Of the 5 MS components, obesity significantly predicted CVD mortality in the United States but not in Japan. The reason for the nonsignificant association in Japan is still unknown. Two conditions may partly explain this difference. First, increased BMI as a measure of obesity may be less sensitive than waist circumference. In the present study, we did not use waist circumference as the measure of obesity, because waist circumference was not measured in the Japanese survey. Second, because increased BMI has a stronger association with coronary heart disease than with stroke, the nonsignificant association between obesity and CVD might be due to a significantly different distribution of the subtypes of CVD, in which coronary heart disease is dominant in the United States, while stroke occurs more frequently in Japan.10 It should be noted that when obesity is defined as BMI 30 kg/m2 for the Japanese sample, the HR of MS for CVD mortality was higher in Japan than in the U.S. (1.53 vs 1.43; Table 2). We conducted further sensitivity analyses by comparing the prevalence of MS components in those with MS and with BMIs 30 kg/m2 between the United States and Japanese samples. The results showed that the prevalence rates of high BP, decreased HDL cholesterol, and elevated TG were significantly higher in the Japanese compared with Americans (97.5% vs 82.0% for high BP, 82.5% vs 71.7% for decreased HDL cholesterol, and 80.8% vs 74.7% for elevated TG). This difference may partly contribute to a higher HR of MS for CVD mortality in Japan when obesity is defined as BMI 30 kg/m2. However, it does not mean that Japanese have a higher risk for CVD, because the Japanese have a much lower prevalence of obesity (BMI 30 kg/m2) than Americas (2.4% vs 25.3%, p <0.001). Certainly the magnitudes of individual MS components using different cut-off points on the association between MS and CVD risk in different populations need to be further studied. In the study, we also observed that the HR of high BP for CVD mortality in the Japanese participants was higher than in the American participants. These findings may suggest a different impact of risk factors on CVD mortality across different countries and racial and ethnic populations. The present study had several strengths. First, its findings are derived from nationally representative population samples. The measurements of exposures and outcomes were conducted using standardized approaches. Therefore, it offers a unique opportunity for the study results to be generalized. Second, the association between MS and CVD mortality was analyzed prospectively, which makes it possible to interpret a potentially temporal association. Third, the findings of the study, by addressing the impacts of individual components of MS on the risk for CVD mortality between countries, are informative for health policy decisions and health practice in controlling CVD mortality. There were also limitations that should be kept in mind. First, individual MS components that are continuous measures (i.e., BMI, BP, TG, HDL, and HbA1c) were dichotomized on the basis of the current definition of MS. This dichotomization approach would lead a reduction in statistical power and an underestimation of the strengths of
Preventive Cardiology/MS and CVD in the U.S. and Japan
associations between these factors and CVD mortality. Second, all risk factors had single measurements at baseline. Therefore, we are unable to test any time-varying effects of these variables on CVD mortality. Third, because BMI may have less sensitivity for the measure of obesity, new biomarkers should be included in further studies. Despite these limitations, 2 clear and important conclusions follow from our present study. First, obesity, high BP, elevated TG, and elevated HbA1c in the United States and high BP and elevated HbA1c in Japan were the most significant individual predictors for CVD mortality. Second, MS and MS plus baseline CVD history are among the important risk factors that may significantly contribute to the explanation of excess CVD mortality in the United States versus Japan. Acknowledgment: We thank the members of the NIPPON DATA80 and NIPPON DATA90 Research Group for their important contributions. A list of the members is provided in Nakamura et al.4
12.
13.
14. 15.
16.
17.
Disclosures The authors have no conflicts of interest to disclose. 1. Lakka HM, Laaksonen DE, Lakka TA, Niskanen LK, Kumpusalo E, Tuomilehto J, Salonen JT. The metabolic syndrome and total and cardiovascular disease mortality in middle-aged men. JAMA 2002;288: 2709e2716. 2. Beckman JA, Creager MA, Libby P. Diabetes and atherosclerosis: epidemiology, pathophysiology, and management. JAMA 2002;287: 2570e2581. 3. Kadota A, Hozawa A, Okamura T, Kadowak T, Nakmaura K, Murakami Y, Hayakawa T, Kita Y, Okayama A, Nakamura Y, Kashiwagi A, Ueshima H; NIPPON DATA Research Group. Relationship between metabolic risk factor clustering and cardiovascular mortality stratified by high blood glucose and obesity: NIPPON DATA90 1990-2000. Diabetes Care 2007;30:1533e1538. 4. Nakamura Y, Okamura T, Higashiyama A, Watanabe M, Kadota A, Ohkubo T, Miura K, Kasagi F, Kodama K, Okayama A, Ueshima H; NIPPON DATA80 Research Group. Prognostic values of clockwise and counterclockwise rotation for cardiovascular mortality in Japanese subjects: a 24-year follow-up of the National Integrated Project for Prospective Observation of Noncommunicable Disease and Its Trends in the Aged 1980-2004 (NIPPON DATA80). Circulation 2012;125: 1226e1233. 5. Liu L, Nettleton JA, Bertoni AG, Bluemke DA, Lima JA, Szklo M. Dietary pattern, the metabolic syndrome, and left ventricular mass and systolic function: the Multi-Ethnic Study of Atherosclerosis. Am J Clin Nutr 2009;90:362e368. 6. Liu L, Liu Z, Ma M, Xue F, Sorel E. The cardiometabolic syndrome and risk of mortality from cardiovascular diseases and all causes among African Americans and white Americans. JCMD 2012;3:1e9. 7. Callow AD. Cardiovascular disease 2005—the global picture. Vascul Pharmacol 2006;45:302e307. 8. National Center for Health Statistics. National Health and Nutrition Examination Survey Data. Available at: http://www.cdc.gov/nchs/surveys. htm. Accessed April 16, 2010. 9. Ueshima H. NIPPON DATA. Nippon Rinsho 2006;64(suppl):108e111. 10. Ueshima H, Sekikawa A, Miura K, Turin TC, Takashima N, Kita Y, Watanabe M, Kadota A, Okuda N, Kadowaki T, Nakamura Y, Okamura T. Cardiovascular disease and risk factors in Asia: a selected review. Circulation 2008;118:2702e2709. 11. Nakamura K, Okamura T, Hayakawa T, Hozawa A, Kadowaki T, Murakami Y, Kita Y, Okayama A, Ueshima H; NIPPON DATA80 9RG. The proportion of individuals with obesity-induced hypertension
18.
19. 20.
21. 22. 23. 24.
25.
26. 27. 28. 29.
89
among total hypertensives in a general Japanese population: NIPPON DATA80, 90. Eur J Epidemiol 2007;22:691e698. Liu L, Okamura T, Kadowaki T, Murakami Y, Hozawa A, Kita Y, Takashima N, Okuda N, Okayama A, Ueshima H. Bundle branch block and other cardiovascular disease risk factors: US-Japan comparison. Int J Cardiol 2010;143:432e440. Data Linkage Team. Comparative analysis of the NHANES III publicuse and restricted-use linked mortality files: 2010 public-use data release—National Center for Health Statistics. Available at: http:// www.cdc.gov/nchs/data/datalinkage/nhanes3_mort_compare_2010_final. pdf. Accessed July 18, 2010. NIPPON DATA80 Research Group. Risk assessment chart for death from cardiovascular disease based on a 19-year follow-up study of a Japanese representative population. Circ J 2006;70:1249e1255. Liu L, Choudhury SR, Okayama A, Hayakawa T, Kita Y, Ueshima H. Changes in body mass index and its relationships to other cardiovascular risk factors among Japanese population: results from the 1980 and 1990 National Cardiovascular Surveys in Japan. J Epidemiol 1999;9: 163e174. Okamura T, Hayakawa T, Kadowaki T, Kita Y, Okayama A, Ueshima H; NIPPON DATA90 Research Group. The inverse relationship between serum high-density lipoprotein cholesterol level and all-cause mortality in a 9.6-year follow-up study in the Japanese general population. Atherosclerosis 2006;184:143e150. Ueshima H, Choudhury SR, Okayama A, Hayakawa T, Kita Y, Kadowaki T, Okamura T, Minowa M, Iimura O; NIPPON DATA80 Research Group. Cigarette smoking as a risk factor for stroke death in Japan NIPPON DATA80. Stroke 2004;35:1836e1841. Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, Gordon DJ, Krauss RM, Savage PJ, Smith SC Jr, Spertus JA, Costa F, American Heart Association, National Heart, Lung, and Blood Institute. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute scientific statement. Circulation 2005;112:2735e2752. Batsis JA, Nieto-Martinez RE, Lopez-Jimenez F. Metabolic syndrome: from global epidemiology to individualized medicine. Clin Pharmacol Ther 2007;82:509e524. Liu L, Nunez AE. Cardiometabolic syndrome and its association with education, smoking, diet, physical activity, and social support: findings from the Pennsylvania 2007 BRFSS survey. J Clin Hypertens (Greenwich) 2010;12:556e564. Kanazawa M, Yoshiike N, Osaka T, Numba Y, Zimmet P, Inoue S. Criteria and classification of obesity in Japan and Asia-Oceania. World Rev Nutr Diet 2005;94:1e12. American Diabetes Association. Executive summary: standards of medical care in diabetes—2010. Diabetes Care 2010;33(suppl): S4eS10. Selvin E, Steffes MW, Gregg E, Brancati FL, Coresh J. Performance of A1c for the classification and prediction of diabetes. Diabetes Care 2011;34:84e89. Grundy SM, Brewer HB Jr, Cleeman JI, Smith SC Jr, Lenfant C; American Heart Association, National Heart, Lung, and Blood Institute. Definition of metabolic syndrome: report of the National Heart, Lung, and Blood Institute/American Heart Association Conference on Scientific Issues Related to Definition. Circulation 2004;109:433e438. Ferreira I, Beijers HJ, Schouten F, Smulders YM, Twisk JW, Stehouwer CD. Clustering of metabolic syndrome traits is associated with maladaptive carotid remodeling and stiffening: a 6-year longitudinal study. Hypertension 2012;60:542e549. SAS Institute Inc. SAS/STAT User’s Guide, Version 9.1. Cary, North Carolina: SAS Institute Inc., 2005. Szklo M, Nieto FJ. Epidemiology: Beyond the Basics. Sudbury, Massachusetts: Jones & Bartlett, 2007. Liao Y, Greenlund KJ, Croft JB, Keenan NL, Giles WH. Factors explaining excess stroke prevalence in the US stroke belt. Stroke 2009;40:3336e3341. Cheng CY, Reich D, Haiman CA, Tandon A, Patterson N, Selvin E, Akylbekova EL, Brancati FL, Coresh J, Boerwinkle E, Altshuler D, Taylor HA, Henderson BE, Wilson JG, Kao WH. African ancestry and its correlation to type 2 diabetes in African Americans: a genetic admixture analysis in three u.s. population cohorts. PLoS One 2012;7: e32840.