Abstract
Explanations for the low prevalence of atherosclerosis in Japan versus United States are often confounded with genetic variation. To help remove such confounding, coronary artery calcification (CAC), a marker of subclinical atherosclerosis, was compared between Japanese men in Japan and Japanese men in Hawaii. Findings are based on risk factor and CAC measurements that were made from 2001 to 2005 in 311 men in Japan and 300 men in Hawaii. Men were aged 40 to 50 years and without cardiovascular disease. After age-adjustment, there was a 3-fold excess in the odds of prevalent CAC scores ≥10 in Hawaii versus Japan (relative odds [RO] = 3.2; 95% confidence interval [CI] = 2.1,4.9). While men in Hawaii had a generally poorer risk factor profile, men in Japan were 4-times more likely to smoke cigarettes (49.5 vs. 12.7%, p<0.001). In spite of marked risk factor differences between the samples, none of the risk factors provided an explanation for the low amounts of CAC in Japan. After risk factor adjustment, the RO of CAC scores ≥10 in Hawaii versus Japan was 4.0 (95% CI = 2.2,7.4). Further studies are needed to identify factors that offer protection against atherosclerosis in Japanese men in Japan.
Keywords: Atherosclerosis, cohort studies, coronary disease, Japan, men, risk factors
Compared to men in the United States, Japanese men in Japan have a low risk of coronary heart disease. Increasing evidence further suggests that the Japanese are less prone to subclinical atherosclerosis [1,2]. In a recent study of coronary artery calcification (CAC), a marker of subclinical atherosclerosis, Japanese men in Japan had higher blood pressure levels, higher concentrations of low-density lipoprotein cholesterol (LDL-C), higher levels of fasting blood glucose, and were more likely to have diabetes than Caucasian men in the United States [1,2]. Men in Japan were also 3-times more likely to smoke cigarettes. In spite of having a less favorable cardiovascular risk profile for several important risk factors, the men in Japan had half the prevalence of CAC than the men in the United States [1].
Unfortunately, it could not be determined if the low levels of atherosclerosis in Japan versus the United States could be explained by genetic or environmental factors. In an attempt to remove potential explanations due to genetic differences, this report compares the prevalence and correlates of CAC between men in Japan and Hawaii who are entirely Japanese. For both samples, measurement of CAC and concomitant risk factors followed identical protocols.
MATERIALS AND METHODS
Study samples
From 2001 to 2005, 311 Japanese men from Japan and 300 Japanese men from Hawaii were enrolled in a study of CAC and its correlated risk factors. Men were aged 40 to 50 years. In Japan, men were residents of Kusatsu City in Shiga prefecture and were descendents of Japanese parents [1,2]. The largely westernized men from Hawaii were second and third generation offspring of Japanese who migrated to Hawaii near the beginning of the 20th century. All participants were without clinical cardiovascular disease, type I diabetes, cancer (except for skin cancer in the past two years), renal failure, and genetic familial hyperlipidemias.
In Kusatsu City, subjects were randomly selected from a registry of residents containing information on name, birth date, and address [1,2]. In Hawaii, men were randomly selected from lists of offspring of fathers who were participants in the Honolulu Heart Program [3,4]. The latter is a long-term follow-up study of coronary heart disease and stroke in Japanese-American men that began in 1965. For both samples, procedures were in accordance with institutional guidelines and approved by an institutional review committee. Informed consent was obtained from the study participants.
CAC screening
At both sites, the same study protocol was followed. Screening for CAC was based on the use of a GE-Imatron C150 Electron Beam Tomography scanner (GE Medical Systems, South San Francisco, CA) with frequent calibration by centrally trained technicians. For both scanners, measures of water density (0 Hounsfield Unit [HU]), air (-1000 HU), and calcification (≥130 HU) were the same. Single scans were completed following a rigid protocol to obtain 30 to 40 contiguous 3 mm thick transverse images from the level of the aortic root to the apex of the heart. Images were recorded during a maximal breath hold using ECG guided triggering of 100 m second exposures during the same phase of the cardiac cycle. Scans were read centrally at the Cardiovascular Institute in Pittsburgh, Pennsylvania by a trained radiology technician using a DICOM (Digital Imaging and Communications in Medicine) workstation with software from the AccuImage Diagnostic Corporation, San Francisco, CA. Scoring adhered to the method of Agatston [5]. The reproducibility of the scans had an intraclass correlation of 0.99 [6]. Further description of the scanning procedure is provided elsewhere [1,2,5,6].
Concomitant risk factors
Concomitant risk factors measured at the time of CAC screening included age, body mass index, systolic blood pressure, hypertension status, the use of medications to treat hypertension, diabetes, fasting blood glucose, insulin, LDL-C, hypercholesterolemia, the use of medications to treat hypercholesterolemia, high-density lipoprotein cholesterol (HDL-C), triglycerides, C - reactive protein, fibrinogen, homocysteine, and parental history of coronary heart disease. Lifestyle factors included current and past cigarette smoking, alcohol intake, being sedentary at work, and the frequent intake of fish, beef or pork, and soy products (≥4 times/week). Body mass index was calculated as weight divided by height squared (kg/m2). A diagnosis of hypertension was made when a systolic blood pressure was ≥140 mm Hg, when a diastolic blood pressure was ≥90 mm Hg, or when a subject was being treated for high blood pressure. Diabetes was diagnosed when a fasting serum glucose level was ≥7 mmol/L or a subject was receiving insulin or oral hypoglycemic therapy. Hypercholesterolemia was defined when LDL-C levels were ≥4.14 mmol/L or when medications were being taken for the treatment of an adverse lipid profile. A description of the methods used to measure the other risk factors is given elsewhere [1,2,7].
Statistical methods
The percent of men with CAC scores above common CAC cut-points was derived for each sample and compared through the use of exact testing methods. Comparisons were also made between the median and 20th and 80th percentiles. Because the distribution of CAC scores is skewed (with many scores =0), the latter comparisons were based on quantile regression models [8]. Use of standard regression techniques was not considered in this report since they are often limited to modeling average values where inference and prediction can be adversely influenced by skewness and unstable variances. While log and power transformations often provide a means for removing such influences, they are also limited to applications with nonzero CAC scores. For comparison of concomitant risk factors between the samples, t-tests and standard procedures for comparing binomial proportions were used.
In this report, the association between a risk factor and the prevalence of a CAC score ≥10 was examined through the use of logistic regression models. Here, a CAC cut point of 10 was selected because of its clinical importance [9] and because of the possibility that scores ranging from >0 to <10 could be an imaging artifact from spurious noise [10]. Through the use of the logistic models, estimates of the relative odds (and 95 percent confidence interval) of having a CAC score ≥10 within each sample were derived by comparing men with and without a risk factor condition. For continuous risk factors, relative odds were calculated for high versus low risk factor levels, where high versus low tended to correspond with a comparison between the 75th and 25th percentiles. The relative odds of having a CAC score ≥10 was also adjusted for age and several concomitant risk factors. In addition, tests were made to determine if the relative odds associated with a risk factor differed between the study samples.
To determine if differences in the prevalence of a CAC score ≥10 between Japan and Hawaii could be explained by confounding risk factors, the percent prevalence of men with scores ≥10 in the two cohorts were calculated and compared after adjusting for the effects of age and other risk factors using logistic regression models and standard analysis of covariance methods [11]. Estimated regression coefficients were used to derive the relative odds (and 95 percent confidence interval) of having a CAC score ≥10 in Hawaii versus Japan. All reported p-values were based on two-sided tests of significance.
RESULTS
The distribution of CAC scores in the Japanese men in Japan and Hawaii are shown in table 1. The percent prevalence of scores >0, ≥10, and ≥100 for men in Hawaii were significantly higher than they were in Japan (p<0.001). Compared to men in Japan, men in Hawaii had a near 3-fold excess in the prevalence of CAC scores ≥10 (32.0 vs. 11.6 percent) and a near 6-fold excess of scores ≥100 (13.3 vs. 2.3 percent).
Table 1.
Distribution, range, and percentiles of CAC scores in 311 Japanese men in Japan and 300 Japanese men in Hawaii aged 40-50 years from 2001 to 2005.
| Japan | Hawaii | |||
|---|---|---|---|---|
| No. | % | No. | % | |
|
| ||||
| CAC range | Distribution | |||
| 0 | 214 | 68.8 | 152 | 50.7 |
| >0* | 97 | 31.2 | 148 | 49.3 |
| ≥10* | 36 | 11.6 | 96 | 32.0 |
| ≥100* | 7 | 2.3 | 40 | 13.3 |
| ≥400† | 2 | 0.6 | 10 | 3.3 |
| ≥1000 | 0 | 0.0 | 3 | 1.0 |
| For all CAC scores | ||
| Range | 0 - 538 | 0 – 3170 |
| Median | 0 | 0 |
| 80th percentile‡ | 3.4 | 51.4 |
|
| ||
| For CAC scores >0 | ||
| 20th percentile | 1.5 | 3.1 |
| Median‡ | 5.7 | 34.2 |
| 80th percentile‡ | 36.6 | 141.4 |
|
| ||
| For CAC scores ≥10 | ||
| 20th percentile | 20.8 | 32.5 |
| Median§ | 37.2 | 79.3 |
| 80th percentile | 129.0 | 211.3 |
Significant excess percent in Hawaii versus Japan (p<0.001).
Significant excess percent in Hawaii versus Japan (p=0.019).
Significantly higher in Hawaii versus Japan (p<0.001).
Significantly higher in Hawaii versus Japan (p=0.007).
CAC: Coronary artery calcification.
In Japan, the maximum CAC score was 538, and in Hawaii, the maximum was 3170. While the median CAC scores were identical, the 80th percentile in Hawaii was significantly higher than the 80th percentile in Japan (51.4 vs. 3.4, p<0.001). The excess 80th percentile in Hawaii remained significant after adjustment for age, body mass index, systolic blood pressure, treatment for high blood pressure, diabetes, current and past cigarette smoking, low and high-density lipoprotein cholesterol, treatment for hypercholesterolemia, and C-reactive protein (p=0.007). For men with a CAC score >0, the median and 80th percentile were each significantly higher in Hawaii versus Japan (p<0.001). Findings remained significant after adjusting for age and the other risk factors (p=0.007 for the median and p=0.001 for the 80th percentile). For men with a CAC score ≥10, only the median was significantly higher in Hawaii than in Japan (p=0.007). After adjustment for age, however, the latter became insignificant (p=0.195).
Table 2 provides a comparison between the study risk factors in the two samples. In Japan, two of the study participants were missing data on body mass index and 12 were missing data on homocysteine. In Hawaii, one man has missing data for HDL-C. Except for cigarette smoking, LDL-C, homocysteine, and alcohol intake, men in Hawaii had a poorer risk factor profile. Compared to men in Japan, the average body mass index was 4.3 kg/m2 higher for men in Hawaii (p<0.001). For the average height of 170 cm (which was nearly identical in the two samples), this corresponds to a difference of 12.4 kg. The average weight in Japan was 68.5 kg, and in Hawaii, it was 80.2 kg.
Table 2.
Percents and average risk factor levels in Japanese men in Japan and Hawaii aged 40-50 years from 2001 to 2005.
| Risk factor | Japan | Hawaii | p-value |
|---|---|---|---|
| Age (y) | 45.1 (2.8)* | 46.1 (2.8) | <0.001 |
| Body mass index (kg/m2) | 23.6 (2.9) | 27.9 (4.3) | <0.001 |
| Systolic blood pressure (mm Hg) | 125 (16) | 128 (13) | 0.033 |
| Hypertension (%) | 26.7 | 32.7 | 0.111 |
| Treatment for high blood pressure (%) | 5.5 | 20.3 | <0.001 |
| Diabetes (%) | 6.1 | 13.3 | 0.003 |
| Fasting blood glucose (mmol/L) | 5.93 (1.04) | 6.22 (1.17) | 0.001 |
| Insulin (pmol/L) | 71 (30) | 104 (62) | <0.001 |
| Current cigarette smoker (%) | 49.5 | 12.7 | <0.001 |
| Past cigarette smoker (%) | 33.8 | 21.7 | 0.001 |
| LDL-C (mmol/L) | 3.43 (0.93) | 3.15 (0.85) | <0.001 |
| Hypercholesterolemia (%) | 22.2 | 33.7 | 0.002 |
| Treatment for hypercholesterolemia (%) | 3.5 | 23.0 | <0.001 |
| HDL-C (mmol/L) | 1.40 (0.35) | 1.31 (0.32) | 0.002 |
| Triglycerides (mmol/L) | 1.76 (0.91) | 2.10 (1.61) | 0.002 |
| C-reactive protein (mg/L) | 0.7 (1.8) | 1.3 (2.2) | 0.001 |
| Fibrinogen (μmol/L) | 7.52 (1.93) | 9.28 (2.14) | <0.001 |
| Homocysteine (μmol/L) | 98 (49) | 60 (23) | <0.001 |
| Parental history of CHD (%) | 13.2 | 28.7 | <0.001 |
| Alcohol intake (g/day) | 27.0 (28.7) | 18.0 (33.0) | <0.001 |
| Sedentary at work (%) | 5.5 | 31.3 | <0.001 |
| Fish intake ≥4 times/week (%) | 41.8 | 9.7 | <0.001 |
| Beef or pork intake ≥4 times/week (%) | 21.5 | 51.0 | <0.001 |
| Soy product intake ≥4 times/week (%) | 26.7 | 10.7 | <0.001 |
Mean (SD).
LDL-C: Low-density lipoprotein cholesterol. HDL-C: High-density lipoprotein cholesterol. CHD: Coronary heart disease.
Men in Hawaii also had a higher average systolic blood pressure than men in Japan (p=0.033) in spite of a greater frequency of treatment for hypertension in Hawaii (p<0.001). The frequency of diabetes in Hawaii was more than double the frequency in Japan (13.3 vs. 6.1 percent, p=0.003). Men in Hawaii also had higher levels of fasting blood glucose and insulin. While average levels of LDL-C were lower in Hawaii (3.15 vs. 3.43 mmol/L, p<0.001), men in Hawaii were 6-times more likely to be treated for an adverse lipid profile (23.0 vs. 3.5 percent, p<0.001). The excess in treatment for an adverse lipid profile in Hawaii further explains the significant excess of hypercholesterolemia in Hawaii. In addition, compared to Japan, levels of HDL-C were lower in Hawaii (p=0.002), while levels of C-reactive protein were higher (p=0.001). Compared to men in Hawaii, Japanese men had lower levels of fibrinogen but higher levels of homocysteine. Men in Hawaii were more than twice as likely to have had a parental history of coronary heart disease as men in Japan (28.7 vs. 13.2 percent, p<0.001).
Among the lifestyle factors, men in Hawaii were 6 times more likely to be sedentary at work, more than twice as likely to be frequent consumers of beef or pork (≥4 times/week), and far less likely to be frequent consumers of fish or soy products (≥4 times/week). In contrast to the excess in several adverse risk factors in Hawaii, men in Japan consumed more alcohol and were 4-times more likely to smoke cigarettes than men in Hawaii (49.5 vs. 12.7 percent, p<0.001). Men in Hawaii were also less likely to have been former smokers (p=0.001).
The relationship between the risk factors in table 2 and prevalent CAC scores ≥10 is described in table 3 within each sample after risk factor adjustment. Here, the risk factor adjusted relative odds of CAC scores ≥10 associated with specific risk factor comparisons are shown. Compared to unadjusted observations, findings are largely unchanged. Although the age range is limited to men aged 40 to 50 years, it seems noteworthy that in each sample, a 5 year difference in age is associated with a 2 to 3-fold excess in the odds of a CAC score ≥10. The prevalence of CAC scores ≥10 was also higher in men who had greater body mass index, although it was significant only in Japan. In the latter, a difference of 6 kg/m2 is associated with a 2.8-fold excess in the odds of a CAC score ≥10. In both samples, the odds of an elevated CAC score is increased by approximately 3-fold in current versus nonsmokers, although it was significant only in Hawaii. Among the remaining risk factors, high LDL-C levels in Japan were related to elevated levels of CAC. Here, an LDL-C difference of 1.30 mmol/L is associated with a 2-fold excess in the prevalence of a CAC score ≥10. In Japan, as described in an earlier report [12], excessive intake of alcohol was also associated with a greater prevalence of elevated CAC scores, while a similar finding in Hawaii was absent. In Japan, although diabetes and the treatment for hypertension and hypercholesterolemia seemed to also be associated with higher amounts of CAC, excesses were not statistically significant. The latter, in part, could be due to a limited sample size. Removing men who were being treated for hypercholesterolemia failed to markedly change the observations contained in table 3. While some of the differences in tables 3 in the relative odds of a prevalent CAC score ≥10 between the two samples appear large (e.g., body mass index and treatment for high blood pressure) none were statistically significant.
Table 3.
Risk factor adjusted relative odds of a CAC score ≥10 associated with selected risk factor comparisons for Japanese men in Japan and Hawaii aged 40-50 years from 2001 to 2005.
| Risk factor | Risk factor comparison** | Japan†† | Hawaii†† | ||
|---|---|---|---|---|---|
| RO | 95% CI | RO | 95% CI | ||
| Age | 5 years | 2.9* | 1.3,6.6 | 2.0† | 1.2,3.4 |
| Body mass index | 6 kg/m2 | 2.8‡ | 1.1,7.0 | 1.3 | 0.8,2.1 |
| Systolic blood pressure | 20 mm Hg | 0.8 | 0.5,1.4 | 1.3 | 0.9,2.1 |
| Hypertension | Yes vs. no | 0.5 | 0.2,1.5 | 1.3 | 0.6,2.9 |
| Treatment for high blood pressure | Yes vs. no | 3.4 | 0.8,13.6 | 1.1 | 0.5,2.3 |
| Diabetes | Yes vs. no | 1.9 | 0.5,7.1 | 1.5 | 0.7,3.4 |
| Fasting blood glucose | 0.7 mmol/L | 1.0 | 0.8,1.3 | 0.9 | 0.8,1.1 |
| Insulin | 49 pmol/L | 1.1 | 0.5,2.1 | 0.9 | 0.7,1.2 |
| Current cigarette smoker | Yes vs. no | 2.7 | 0.7,10.6 | 3.2§ | 1.5,6.9 |
| Past cigarette smoker | Yes vs. no | 1.3 | 0.3,5.5 | 1.2 | 0.6,2.2 |
| LDL-C | 1.3 mmol/L | 2.1¶ | 1.2,3.8 | 1.3 | 0.8,2.0 |
| Hypercholesterolemia | Yes vs. no | 1.6 | 0.6,3.9 | 1.1 | 0.5,2.6 |
| Treatment for hypercholesterolemia | Yes vs. no | 1.8 | 0.3,9.4 | 1.6 | 0.8,3.3 |
| HDL-C | 0.4 mmol/L | 1.3 | 0.8,2.1 | 1.4 | 1.0,1.9 |
| Triglycerides | 1.1 mmol/L | 0.9 | 0.6,1.5 | 1.0 | 0.8,1.3 |
| C-reactive protein | 1 mg/L | 1.0 | 0.7,1.3 | 0.9 | 0.7,1.0 |
| Fibrinogen | 3.0 μmol/L | 1.1 | 0.6,2.0 | 1.5 | 0.9,2.4 |
| Homocysteine | 37 μmol/L | 1.1 | 0.9,1.4 | 1.1 | 0.7,1.7 |
| Parental history of CHD | Yes vs. no | 1.8 | 0.6,5.3 | 1.3 | 0.8,2.4 |
| Alcohol intake | 36 g/day | 1.7# | 1.1,2.7 | 1.1 | 0.8,1.6 |
| Sedentary at work | Yes vs. no | 1.0 | 0.2,5.0 | 1.0 | 0.5,1.7 |
| Fish intake ≥4 times/week | Yes vs. no | 0.7 | 0.3,1.5 | 1.0 | 0.4,2.4 |
| Beef or pork intake ≥4 times/week | Yes vs. no | 1.2 | 0.5,3.2 | 0.7 | 0.4,1.3 |
| Soy product intake ≥4 times/week | Yes vs. no | 1.3 | 0.5,3.7 | 0.6 | 0.1,3.2 |
Significant relative odds:
p=0.012,
p=0.007,
p=0.025,
p=0.002,
p=0.011,
p=0.028.
For continuous risk factors, the estimated relative odds compare the risk of a CAC score ≥10 in men with a high versus low risk factor level where risk factors differ by the amount specified.
Relative odds are based on models that include age, body mass index, systolic blood pressure, treatment for high blood pressure, diabetes, current and past cigarette smoking, LDL-C, treatment for hypercholesterolemia, HDL-C, and C-reactive protein as covariates. For models with fasting blood glucose and insulin, diabetes is removed as a covariate. When modeling hypertension status, systolic blood pressure is removed as a covariate, and when modeling hypercholesterolemia, LDL-C is removed.
CAC: Coronary artery calcification. RO: Relative odds.
CI: Confidence interval. LDL-C: Low-density lipoprotein cholesterol.
HDL-C: High-density lipoprotein cholesterol. CHD: Coronary heart disease.
In table 4, the possibility that the risk factors considered in this report could explain the low prevalence of CAC scores ≥10 in Japan versus Hawaii is explored. After age and risk factor adjustment (including adjustment for lifestyle factors), the difference in the prevalence of scores ≥10 remained unchanged (p<0.001). In attempting to avoid over parameterizing the statistical models used in table 4, effects of smaller subsets of risk factors and other characteristics in table 2 were also examined. Results failed to alter the findings that appear in table 4. An excess of CAC scores ≥10 in the sample from Hawaii versus Japan also persisted in current and non-cigarette smokers and in men who were not being treated for hypercholesterolemia.
Table 4.
Adjusted and risk factor adjusted percent prevalence of men with CAC scores ≥10 in Japanese men in Japan and Hawaii aged 40-50 years from 2001 to 2005.
| Study sample | % prevalence | RO | 95% CI |
|---|---|---|---|
| Age-adjusted | |||
| Japan | 12.4 | reference | |
| Hawaii | 30.3 | 3.2 | 2.1,4.9 |
| p-value | <0.001 | ||
|
| |||
| Model A† | |||
| Japan | 11.5 | reference | |
| Hawaii | 31.4 | 4.0 | 2.2,7.4 |
| p-value | <0.001 | ||
|
| |||
| Model B‡ | |||
| Japan | 11.4 | reference | |
| Hawaii | 31.3 | 4.1 | 2.2,7.8 |
| p-value | <0.001 | ||
Model A: Adjusted for age, body mass index, systolic blood pressure, treatment for high blood pressure, diabetes, current and past cigarette smoking, low and high-density lipoprotein cholesterol, treatment for hypercholesterolemia, and C-reactive protein.
Model B: Adjusted for age, alcohol intake, frequent intake of fish, beef or pork, and soy products (≥4 times/week), body mass index, systolic blood pressure, diabetes, current cigarette smoking, low and high-density lipoprotein cholesterol, and parental history of coronary heart disease.
CAC: Coronary artery calcification. RO: Relative odds. CI: Confidence interval.
DISCUSSION
In this comparison of Japanese men in Japan and Hawaii, the prevalence of CAC scores ≥10 was nearly 3 times higher in Hawaii than in Japan. Although relatively young (aged 40 to 50 years), there were substantial levels of CAC in Hawaii that were well over the threshold for significant obstructive atherosclerosis [13]. Over 13 percent of the men in Hawaii had CAC scores ≥100 as compared to 2.3 percent in Japan. With increased exposure to Western lifestyles, these findings suggest that Japanese in Japan are capable of developing the same excessively high levels of atherosclerosis as in Hawaii. They are also consistent with observations from the Multi-Ethnic Study of Atherosclerosis (MESA) where increased acculturation into the United States has been associated with CAC within other ethnic samples [14,15]. In contrast to the current comparison, MESA includes a broad spectrum of acculturation in the United States. Here, comparisons are made between men in Japan who have been acculturated to Japanese lifestyles in Japan and Japanese men in Hawaii who have been acculturated to western exposures in the United States through multiple generations.
No single risk factor considered in the current report (or set of risk factors) provided explanation for the low amounts of CAC in Japan versus Hawaii, and it could not be determined if the adverse cardiovascular consequence from the generally poorer risk factor profile in Hawaii was worse than the excessive rates of smoking in Japan. When replacing the CAC cut point of 10 with scores >0, it is interesting to note that each of the risk factors, except for body mass index, persisted in failing to explain the excess of CAC scores >0 in Hawaii versus Japan (49.3 vs. 31.2 percent, p<0.001). When adjustments for body mass index were made, the prevalence of CAC scores >0 became similar (38.9 percent in Japan and 41.1 percent in Hawaii, p=0.605). The latter could partly be due to an exceptionally high frequency of obesity (body mass index ≥30 kg/m2) in the men from Hawaii with CAC scores ranging from >0 to <10. Here, 42.3 percent were obese as compared to 18.4 percent of the men with CAC scores =0 and 33.3 percent in those with CAC scores ≥10. In contrast, only 4.9 percent of the men in Japan with CAC scores >0 to <10 were obese. For those with CAC scores =0 and ≥10, 1.4 and 8.6 percent were obese, respectively. It may be that the 1.6-fold excess of CAC scores >0 in Hawaii versus Japan is more easily explained by a difference in body mass index as compared to the 3 and near 6-fold excesses of scores ≥10 and ≥100 that were observed in Hawaii.
The prevalence of high CAC scores in Japan also remained low in spite of a 4-fold excess in the use of cigarettes in Japan versus Hawaii (49.5 vs. 12.7 percent, p<0.001). Whether there are risk factor associations with prevalent CAC that differ between nonsmokers in Japan and those who smoke excessively warrants consideration. One possible factor might include fish intake. In the Honolulu Heart Program, risk of death was halved in heavy smokers who consumed high amounts of fish as compared to those whose consumption was low [16]. In the current study, high fish intake could not be shown to be associated with low amounts of CAC for either smokers or nonsmokers, although the latter could be due to limited statistical power.
Differences in the intake of other dietary factors such as beef or pork and soy products also failed to explain the high prevalence of CAC scores in Hawaii versus Japan. In an earlier report involving the current sample from Japan, alcohol intake was shown to have a J-shaped relationship with CAC [12]. In Hawaii, a similar finding was absent. Differences in the intake of ethanol and its relationships to CAC between the two samples also failed to explain the excess of CAC that was observed in Hawaii.
Cardiovascular risk factors also failed to explain the 2-fold excess of CAC in a sample of Caucasian men in the United States in a comparison with the current sample in Japan [1,2]. This occurred in spite of the Japanese men having a more adverse cardiovascular risk profile for several risk factors including higher blood pressures, higher concentrations of LDL-C, and higher levels of fasting blood glucose. Compared to the Caucasian men, the Japanese in Japan were also more likely to have diabetes and to smoke cigarettes. While genetic factors unique to the Japanese may offer some protection against this adverse risk factor profile, the results of the current study imply that it is unlikely that genetic variation can provide more than a partial explanation for the low prevalence of CAC in Japanese men in Japan.
Of course, this assumes that gene distributions and genetic susceptibilities are similar between the Japanese samples being compared. In addition to being entirely Japanese, the two samples were also all men, of the same age, and with identical periods of study execution. As in all populations, while genetic variation is present in Japan, the extent of this variation is likely to be considerably less than in Caucasian men in the United States where ancestral diversity is high. As a result, it is likely that the Japanese samples in the current comparison are genetically quiet similar. Even with equal frequencies of genetic susceptibility, however, genetics may still be important. An excess in non-genetic risk factors in one sample could alter disease risk by providing an environment for genetic effects to have a role in disease processes.
In addition to a comparison between samples of men who are all Japanese, there are other strengths in the current report worth noting. Most important is that the study samples were examined through a common research protocol. All phases of the study were coordinated by principal investigators in Japan, Hawaii, and at the Cardiovascular Institute in Pittsburgh, Pennsylvania. Technicians received identical training, and all CAC scans were centrally scored. The excess prevalence of CAC in the Japanese men in Hawaii versus Japan is also consistent with the excess in coronary heart disease that has been observed to occur in Japanese who migrate to the United States [3,17-22].
While these observations are specific to men of Japanese ancestry, it needs to be determined if they can be generalized to include women. In addition, failure for the risk factors considered in this report to explain the excess CAC in Hawaii versus Japan could be due to the single measurement of the risk factors at the time of CAC determination. It may be that risk factors that coexist with CAC levels are poor measures of long-term exposure to adverse risk factor profiles with origins in young adulthood. Whether this long-term exposure offers insight into the difference in CAC development between Japanese in Japan and Hawaii warrants consideration. Findings for the age range of 40 to 50 years may also not apply to other age groups, although having a narrow age range can increase the homogeneity of the samples being compared. It seems clear, however, that there are significant implications for the prevention of cardiovascular disease in all ethnic groups if the factors which protect against subclinical atherosclerosis can be identified, particular among Japanese men in Japan who smoke excessively. Further studies of CAC and its progression in larger samples are needed in order to increase the capacity to identify such factors that promote atherosclerosis or offer protection against its development.
Acknowledgments
Supported by a contract (N01-AG-4-2149) and grant (1-R01-AG17155) from the National Institute on Aging; a contract (N01-HC-05102) and grants (R01-HL068200 and R01-HL071561) from the National Heart, Lung, and Blood Institute; a grant (1-R01-NS41265-01) from the National Institute of Neurological Disorders and Stroke; a grant from the United States Department of the Army (DAMD17-98-1-8621); a grant (0160512U0) from the American Heart Association; a grant-in aid for scientific research (A: 13307016) from the Japanese Ministry of Education, Culture, Sports, Science and Technology; the Japan Society for the Promotion of Science. The authors wish to thank Ms. Lori Givens and Ms. Susan Simmons for their expert technical role in the scoring of coronary artery calcification.
Abbreviations
- CAC
Coronary artery calcification
- CHD
Coronary heart disease
- LDL-C
Low-density lipoprotein cholesterol
- HDL-C
High-density lipoprotein cholesterol
- RO
Relative odds
- CI
Confidence interval
References
- 1.Sekikawa A, Ueshima H, Kadowaki T, et al. Less subclinical atherosclerosis in Japanese men in Japan than in white men in the United States in the post World-War-II birth cohort. Am J Epidemiol. 2007;165:617–624. doi: 10.1093/aje/kwk053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sekikawa A, Ueshima H, Zaky WR, et al. Much lower prevalence of coronary calcium detected by electron-beam computed tomography among men aged 40–49 in Japan than in the US, despite a less favorable profile of major risk factors. Int J Epidemiol. 2004;34:173–179. doi: 10.1093/ije/dyh285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kagan A, Harris BR, Winkelstein W, Jr, et al. Epidemiologic studies of coronary heart disease and stroke in Japanese men living in Japan, Hawaii and California: Demographic, physical, dietary, and biochemical characteristics. J Chron Dis. 1974;27:345–364. doi: 10.1016/0021-9681(74)90014-9. [DOI] [PubMed] [Google Scholar]
- 4.Yano K, Reed DM, McGee DL. Ten-year incidence of coronary heart disease in the Honolulu Heart Program: Relationship to biologic and lifestyle characteristics. Am J Epidemiol. 1984;119:653–666. doi: 10.1093/oxfordjournals.aje.a113787. [DOI] [PubMed] [Google Scholar]
- 5.Agatston AS, Janowitz WR, Hildner FJ, et al. Quantification of coronary artery calcification using ultrafast computed tomography. J Am Coll Cardiol. 1990;15:827–832. doi: 10.1016/0735-1097(90)90282-t. [DOI] [PubMed] [Google Scholar]
- 6.Sutton-Tyrrell K, Kuller LH, Edmundowicz D, et al. Usefulness of electron beam tomography to detect progression of coronary and aortic calcium in middle-aged women. Am J Cardiol. 2001;87:560–564. doi: 10.1016/s0002-9149(00)01431-4. [DOI] [PubMed] [Google Scholar]
- 7.Moriyama Y, Okamura T, Kajinami K, et al. Effects of serum B vitamins on elevated plasma homocysteine levels associated with the mutation of methylenetetrahydrofolate reductase gene in Japanese. Atherosclerosis. 2002;164:321–328. doi: 10.1016/s0021-9150(02)00105-3. [DOI] [PubMed] [Google Scholar]
- 8.Yu K, Lu Z, Stander J. Quantile regression: Applications and current research areas. The Statistician. 2003;52:331–350. [Google Scholar]
- 9.Budoff MJ, Shaw LJ, Liu ST, et al. Long-term prognosis associated with coronary calcification: Observations from a registry of 25,253 patients. J Am Coll Cardiol. 2007;49:1860–1870. doi: 10.1016/j.jacc.2006.10.079. [DOI] [PubMed] [Google Scholar]
- 10.Jain T, Peshock R, McGuire DK, et al. African Americans and Caucasians have a similar prevalence of coronary calcium in the Dallas Heart Study. J Am Coll Cardiol. 2004;44:1011–1017. doi: 10.1016/j.jacc.2004.05.069. [DOI] [PubMed] [Google Scholar]
- 11.Lane PW, Nelder JA. Analysis of covariance and standardization as instances of prediction. Biometrics. 1982;38:613–621. [PubMed] [Google Scholar]
- 12.Okamura T, Kadowaki T, Sekikawa A, et al. Alcohol consumption and coronary artery calcium in middle-aged Japanese men. Am J Cardiol. 2006;98:141–144. doi: 10.1016/j.amjcard.2006.01.095. [DOI] [PubMed] [Google Scholar]
- 13.O’Rourke RA, Brundage BH, Froelicher VF, et al. American College of Cardiology/American Heart Association expert consensus document on electron-beam computed tomography for the diagnosis and prognosis of coronary artery disease. J Am Coll Cardiol. 2000;36:326–340. doi: 10.1016/s0735-1097(00)00831-7. [DOI] [PubMed] [Google Scholar]
- 14.Diez Roux AV, Detrano R, Jackson S, et al. Acculturation and socioeconomic position as predictors of coronary calcification in a multiethnic sample. Circulation. 2005;112:1557–1565. doi: 10.1161/CIRCULATIONAHA.104.530147. [DOI] [PubMed] [Google Scholar]
- 15.Bild DE, Detrano R, Peterson D, et al. Ethnic differences in coronary calcification: The Multi-Ethnic Study of Atherosclerosis (MESA) Circulation. 2005;111:1313–1320. doi: 10.1161/01.CIR.0000157730.94423.4B. [DOI] [PubMed] [Google Scholar]
- 16.Rodriguez BL, Sharp DS, Abbott RD, et al. Fish intake may limit the increase in risk of coronary heart disease morbidity and mortality among heavy smokers: The Honolulu Heart Program. Circulation. 1996;94:952–956. doi: 10.1161/01.cir.94.5.952. [DOI] [PubMed] [Google Scholar]
- 17.Gordon T. Mortality experience among the Japanese in the United States, Hawaii, and Japan. Pub Health Reports. 1957;72:543–553. [PMC free article] [PubMed] [Google Scholar]
- 18.Kagan A, Marmot MG, Kato H. The Ni-Hon-San Study of cardiovascular disease epidemiology: Population characteristics and epidemiology of stroke. In: Kesteloot H, Joossens JV, editors. Epidemiology of Arterial Blood Pressure. Vol. 1980. The Hague/ Boston/ London: Martinus Nijhoff Publishers; pp. 423–436. [Google Scholar]
- 19.Takeya Y, Popper JS, Shimizu Y, et al. Epidemiologic studies of coronary heart disease and stroke in Japanese men living in Japan, Hawaii and California: Incidence of Stroke in Japan and Hawaii. Stroke. 1984;15:15–28. doi: 10.1161/01.str.15.1.15. [DOI] [PubMed] [Google Scholar]
- 20.Worth RM, Kato H, Rhoads GG, et al. Epidemiologic studies of coronary heart disease and stroke in Japanese men living in Japan, Hawaii and California: Mortality. Am J Epidemiol. 1975;102:481–490. doi: 10.1093/oxfordjournals.aje.a112186. [DOI] [PubMed] [Google Scholar]
- 21.Yano K, Reed DM, Kagan A. Coronary heart disease, hypertension and stroke among Japanese-American men in Hawaii: The Honolulu Heart Program. Hawaii Med J. 1985;44:297–312. [PubMed] [Google Scholar]
- 22.Robertson TL, Kato H, Rhoads GG, et al. Epidemiologic studies of coronary heart disease and stroke in Japanese men living in Japan, Hawaii and California. Incidence of myocardial infarction and death from coronary heart disease. Am J Cardiol. 1977;39:239–243. doi: 10.1016/s0002-9149(77)80197-5. [DOI] [PubMed] [Google Scholar]