Published papers

Hypertension and smoking are major risk factors for death to cardiovascular disease (CVD). These attributions for CVD mortality should be higher in the countries where obesity-related conditions are uncommon. However, the joint effect of these risk factors on CVD and all-cause mortality have not been described. We followed a representative 8,912 Japanese men and women without a history of stroke and heart disease. Participants were categorized into 4 groups as follows: a group of individuals who neither smoked nor had hypertension (HT), a group of current smokers, a group with HT, and a group of current smokers with HT. We further calculated population-attributable fractions (PAF) of CVD and all-cause mortality based on relative hazards assessed by proportional hazard regression models. After 19 years of follow-up, we observed 313 and 291 CVID and 948 and 766 all-cause deaths for men and women, respectively. The PAF of CVD mortality due to smoking or HT were 35.1 % for men and 22.1 % for women. The PAF of CVID mortality was higher in participants <60 years of age (57.4% for men and 40.7% for women) vs. those who were older (26.3% for men and 18.1% for women). Aggressive aftempts to discourage smoking and to curb HT could yield large health benefits in Japan and throughout Asia, particularly for those aged <60 years. Efforts to warn about the adverse consequence of HT and smoking during adolescence and youth could yield the greatest health benefits, since positive behaviors adopted early are more easily continued into middle adulthood and later life.

Explanations for the low prevalence of atherosclerosis in Japan versus the United States are often confounded with genetic variation. To help remove such confounding, the authors compared coronary artery calcification (CAC), a marker of subclinical atherosclerosis, between Japanese men in Japan and Japanese men in Hawaii. Findings were based on risk factors and CAC measured from 2001 to 2005 in 311 men in Japan and 300 men in Hawaii. Men were aged 40-50 years and without cardiovascular disease. After age adjustment, there was a threefold excess in the odds of prevalent CAC scores of ≥10 in Hawaii versus Japan (relative odds = 3.2, 95% confidence interval: 2.1, 4.9). Whereas men in Hawaii had a generally poorer risk factor profile, men in Japan were four 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 explained the low amounts of CAC in Japan. After risk factor adjustment, the relative odds of CAC scores of ≥10 in Hawaii versus Japan was 4.0 (95% confidence interval: 2.2, 7.4). Further studies are needed to identify factors that protect against atherosclerosis in Japanese men in Japan. © The Author 2007. Published by the Johns Hopkins Bloomberg School of Public Health. All rights reserved.

Background: We have previously reported that the prevalence of coronary artery calcification (CAC) was substantially lower among Japanese than American men despite a less favorable profile of many traditional risk factors in Japanese men. To determine whether lipoprotein-associated phospholipase A2 (Lp-PLA2) levels are related to the difference in the prevalence of CAC between the two populations. Methods: A total of 200 men aged 40-49 years were examined: 100 residents in Allegheny County, Pennsylvania, United States, and 100 residents in Kusatsu City, Shiga, Japan. Coronary calcium score (CCS) was evaluated by electron-beam tomography, Lp-PLA2 levels, nuclear magnetic resonance (NMR) lipoprotein subclasses, and other factors were assessed in 2001-2002. Results: Lp-PLA2 levels were higher among American than Japanese men (Mean ± standard deviation 301.7 ± 82.6 versus 275.9 ± 104.7 ng/mL, respectively, p=0.06). Among all Japanese men and those with low density lipoprotein (LDL) cholesterol ≥130 mg/dL, there was an inverse association of the prevalence of CCS>0 with the tertile groups of Lp-PLA2 levels (p=0.08 and p=0.03, respectively). American men did not have any association between CCS>0 with the tertile groups of Lp-PLA2 (p=0.62). Although Lp-PLA2 among both populations correlated positively with LDL and total cholesterol, American and Japanese men had different correlations with NMR lipoprotein subclasses. Reported high odds ratio for CCS>0 among American compared to Japanese men was not reduced after adjusting for Lp-PLA2 levels. Conclusion: Lp-PLA2 may have different mechanisms of action among American and Japanese men. Lp-PLA2 levels can not explain the observed CAC differences between the two populations. Copyright © 2007 by the Japan Epidemiological Association.

Background: An international comparison has indicated that the association between smoking and cardiovascular disease (CVD) differs according to total cholesterol (TC) levels. However, little has been published about the relationship between smoking and CVD mortality among populations with various cholesterol levels. Methods: We calculated the adjusted relative hazard (RH) of smoking for CVD mortality among 8912 Japanese individuals without a history of stroke or heart disease, who were separated according to TC levels of ≥5.40, 4.81-5.39, 4.26-4.80 and <4.25 mmol/l into groups Q4, Q3, Q2 and Q1, respectively. The P-values for multiple interactions between TC and smoking status for CVD mortality were calculated using TC as a continuous variable, dichotomized smoking status (never vs current), and by including cross-product terms in the regression models. Results: After 19 years of follow-up, 313 men and 291 women died of CVD. The RH of CVD mortality among men who currently smoked compared with those who never smoked was increased with higher TC (RH = 2.36 in Q4) and decreased in those with lower TC (RH = 0.85 in Q1) (interaction, P = 0.01). The profiles for coronary heart disease (CHD) mortality and ischaemic CVD (composite endpoint of CHD and ischaemic stroke) in men and for ischaemic CVD mortality in women were identical. The interaction might be explained by a biological mechanism and by frailty of those who have never smoked with lower TC. Conclusions: Counteractive measures should be implemented against smoking targeted towards Japanese with elevated TC levels. © The Author 2007; all rights reserved.

The clinical importance of γ-glutamyltransferase (GGT) has recently been debated. Although some studies have suggested that the relationship between GGT and cardiovascular disease (CVD) mortality is independent of alcohol consumption, to our knowledge no studies have reported the relationship between GGT and CVD mortality in never-drinker subgroups. Since Japanese women are known to have a lower prevalence of alcohol consumption, we examined whether GGT predicts CVD mortality in never-drinkers. We followed 2724 Japanese men and 4122 Japanese women without prior CVD or liver dysfunction for 9.6 years and observed 83 and 82 CVD deaths, respectively. Current alcohol drinkers comprised 59% of men and 7% of women. Among women, the multiple adjusted hazard ratio (HR) for CVD mortality compared with the reference group (GGT: 1-12 U/L) was 2.88 (95% confidence interval (CI), 1.14-7.28) for the elevated group (GGT ≥ 50 U/L). This positive relationship was unchanged in the never-drinkers subgroup (HR for log-transformed continuous GGT, 1.62 (95% CI, 1.11-2.37)). No significant relationships were observed in men. GGT displays a strong positive association with CVD mortality among Japanese women, for whom the prevalence of ever-drinkers is very low. Exploring the significance and biological mechanisms of GGT might provide useful insights into CVD prevention. © 2006 Elsevier Ireland Ltd. All rights reserved.

The objective of this study was to clarify the awareness and treatment status of hypertension in Japanese workers, comparing with those of hypercholesterolemia. The subjects were 6,186 workers (age 19 to 69 years), who worked at 12 workplaces in Japan and who continuously participated in the High-risk and Population Strategy for Occupational Health Promotion (HIPOP-OHP) study at the baseline (1999 or 2000) and the subsequent year (2000 or 2001). No Intervention was performed during the study period. Blood pressure and total cholesterol level were measured at baseline. Awareness and treatment status were examined by a self-administered questionnaire in the following year. Untreated subjects were defined as those were not aware of hypertension (or hypercholesterolemia) or those not taking medication and not undergoing lifestyle modification. The percentages of subjects with hypertension defined as aware were 65.7% in men and 72.7% in women, respectively. The percentages of subjects with hypercholesterolemia defined as aware were 55.7% in men and 58.6% in women, respectively. In subjects 40 years and older, the awareness rate for hypertension was significantly higher than the awareness rate for hypercholesterolemia (75.0% vs. 59.2%, p<0.001). The percentage of untreated subjects with hypertension was significantly lower than the percentage of untreated subjects with hypercholesterolemia (41.0% vs. 57.1%, p<0.001). Although Japanese workers ≥40 years old have a higher awareness of and are more often treated for hypertension than for hypercholesterolernia, there are still many unaware and untreated individuals with either disease in all age groups. It is important to increase the awareness of risk factors in individuals with positive findings for these conditions.

Background: For the Japanese population, a body mass index (BMI) of 25.0-29.9 is classified as obesity and is a risk factor for cardiovascular disorders such as hypertension. Methods: A cohort study to clarify obesity costs for a Japanese population was conducted utilizing baseline BMI and medical costs over a 10-year follow-up period. The participants were 4502 community dwelling Japanese National Health Insurance (NHI) beneficiaries aged 40-69 years. According to their baseline BMI values (kg/m2), participants were classified into the following three categories: BMI < 18.5, 18.5 ≤ BMI < 25.0 and 25.0 ≤ BMI. Medical costs per person per month were compared among the three categories. Excess medical costs attributable to the 25.0 ≤ BMI category compared to the 18.5 ≤ BMI < 25.0 category were estimated. Results: Approximately 20% of the Japanese population studied had a BMI of 25.0 or over. A J-shaped relationship between BMI and personal total medical costs was observed. Personal total medical costs per month determined from the 10-year follow-up in each category were 189 Euros (BMI < 18.5), 134 Euros (18.5 ≤ BMI < 25.0) and 155 Euros (25.0 ≤ BMI). A J-shaped pattern was observed after adjusting for age, sex, smoking and drinking habits, and excluding early deceased participants. Furthermore, smoking habit did not modify the J-shaped pattern of total medical costs. The estimated excess medical costs for the 25.0 ≤ BMI category represented 3.1% of the total medical costs for the entire study population (634 105 Euros). Conclusion: The Japanese NHI beneficiaries with a BMI of 25.0 or over showed increased medical costs compared to those with a BMI of 18.5-24.9. © The Author 2007. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved.

The increased prevalence of obesity in Japan may contribute to the high prevalence of hypertension in Japan. In the present study, we calculated the odds ratio for hypertension in obesity (body mass index (BMI) ≥ 25.0 kg/m2) using data from independent nationwide surveys conducted in 1980 and 1990. We estimated the percentage of hypertensives whose condition was due to obesity among total hypertensives in the general Japanese population. In the 1980 survey, 18.8% of 4,623 male participants were obese and 50.4% were hypertensive, whereas 22.6% of 5,893 female participants were obese and 41.1% were hypertensive. For both sexes, obese participants had a higher odds ratio for hypertension than non-obese participants (BMI < 25.0 kg/m2), and there was a significant dose-response relationship between BMI and the odds ratio for hypertension. Among all hypertensives, the percentage whose hypertension was due to obesity in 1980 and 1990 was 11.4% (95% confidence interval (CI): 4.7-17.7%) and 15.3% (95% CI: 6.8-23.1%) for men and 19.3% (95% CI: 12.1-25.9%) and 22.3% (95% CI: 14.6-29.3%) for women, respectively. Approximately 80-90% of individuals with obesity-induced hypertension were in the 25.0 ≤ BMI < 30.0 kg/m2 category for both sexes in each year. In conclusion, we found that obesity-induced hypertension as a proportion of total hypertension increased between 1980 and 1990 for both sexes. Obesity now is playing a more important role in the high prevalence of hypertension in Japan than it was before. © 2007 Springer Science+Business Media B.V.

The objective of this study is to ascertain if higher normal fasting glucose levels are also an independent risk of developing diabetes in an Asian population, and we thus analysed data from a cohort of healthy Japanese workers. We used data from the non-randomised trial on health promotion intervention, High-risk and Population Strategy for Occupational Health Promotion (HIPOP-OHP) Study. Diabetes cases and those who had fasting blood glucose levels equal to or greater than 100 mg/dl at baseline were excluded, and the Cox proportional-hazards model was used for the analysis. During the four-year follow-up of 2212 participants, we found 37 diabetes cases. In the multivariable model, people with blood glucose levels in the 4th quartile had a higher risk of diabetes than those in the bottom quartile; the multivariable-adjusted odds ratio was 2.52. The risk of diabetes abruptly rose in persons with blood glucose levels higher than 94 mg/dl (fourth quartile). A significant linear trend was not observed in the 1st to 3rd quartiles (p=0.726). In conclusion, higher fasting glucose level was associated with the risk of diabetes, and we found a threshold in the association between fasting blood glucose levels and risk of diabetes in an Asian population. © 2007 Springer-Verlag.

To investigate the association between fish consumption and early atherosclerosis, we analyzed the relationship between fish consumption and average intima-media thickness (AveIMT) by carotid ultrasound in middle-aged Japanese men. Participants were 250 randomly selected, community-based Japanese men aged 40 to 49 years without a prior history of cardiovascular disease. AveIMT was calculated from the mean of 1-cm lengths of both the right and the left carotid arteries at 8 locations. A lifestyle survey was carried out using a self-administered questionnaire including the frequency of fish intake. There were 147 men in the fewer than 4 times per week fish consumption group and 103men in the 4 or more times per week group. The mean AveIMT was significantly higher in the low fish consumption group than in the high fish consumption group (0.623 ± 0.068 vs 0.605 ± 0.065 mm, P = .03). After adjustment for age, waist circumference, pack-years of smoking, alcohol consumption, diabetes, and lipid-lowering medications, the significant difference in the AveIMT between the 2 groups remained. However, after further adjustment for low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, and C-reactive protein in the model, the significant difference disappeared. Fish consumption may be protective against early atherosclerosis in middle-aged men, probably through its beneficial effects on inflammation. © 2007 Elsevier Inc. All rights reserved.

Japanese men consume more alcoholic beverages than men in many other developed countries. The high consumption rate of alcoholic beverages among Japanese men may contribute to the high prevalence of hypertension in Japan. In the present study, we calculated the odds ratio for hypertension in alcohol drinkers based on recent criteria using data from a nationwide survey conducted in Japan in 1990, and estimated, among total hypertensives in a general Japanese population, the percentage of hypertensives whose condition was due to alcohol consumption. Of 3,454 male participants, 64.8% were drinkers (1 gou/day, 28.9%; 2 gou/day, 20.1%; 3 gou/day or more, 8.7%; ex-drinkers, 7.0%) and 49.8% were hypertensive, whereas 7.6% of 4,808 female participants were drinkers (1 gou/day, 5.2%; 2 gou/day or more, 1.3%; ex-drinkers, 1.1%) and 43.1% were hypertensive (1 gou=23.0 g of alcohol). In both sexes, drinkers had a higher odds ratio for hypertension than never drinkers, and there was a significant dose-response relationship between the amount of alcohol consumed and the odds ratio for hypertension. Among all hypertensives, the percentage whose hypertension was due to alcohol consumption was 34.5% (95% confidence interval, 10.9%,51.9%) for men and 2.6% (0.8%,5.8%) for women. The corresponding proportion based on daily alcohol intake was 12.7% for 1 gou/day, 11.1 % for 2 gou/day, 5.8% for 3 gou/day or more, and 4.8% for ex-drinkers in men, and 1.8% for 1 gou/ day, 0.7% for 2 gou/day or more, and -0.1% for ex-drinkers in women. In conclusion, we found that a large percentage of the hypertensives in a general Japanese male population had alcohol-induced hypertension.

Background: Ghrelin, a 28-amino-acid gastric peptide hormone, has an appetite-stimulating effect and controls the energy balance. Serum ghrelin levels inversely correlate with body mass index. Recently, several papers reported the ethnic difference in the ghrelin levels. To our knowledge, however, no studies have compared the serum ghrelin levels between Caucasians in the USA and the Japanese in Japan. Methods: We conducted a cross-sectional study of 189 men 40-49 years of age (91 US Caucasians in the U.S. and 98 Japanese in Japan) to examine serum ghrelin levels and metabolic and other factors. Results: Serum ghrelin levels correlated with waist circumferences and lipid profiles among Caucasian Americans and the Japanese. Serum ghrelin levels were significantly higher among Caucasian Americans than among the Japanese (904.5 (632.0, 1132.0) pg/mL, 508.0 (399.0, 1378.3) pg/mL (median and 95% confidence interval), respectively, P < 0.01), although Caucasian Americans were much more obese (BMI: 26.9 ± 3.3 kg/m2 versus 23.3 ± 3.1 kg/m2 respectively, P < 0.01). The ethnic difference remained after adjusting for metabolic factors, smoking status, and other factors (P < 0.01). Conclusions: We have shown in our population-based study that serum ghrelin levels among men aged 40-49 are significantly higher in Caucasian Americans than in the Japanese in Japan. Reasons for the ethnic difference in the ghrelin levels are largely unknown and warrant further investigation. © 2006 Blackwell Publishing Ltd.

Background: Fatigue is one of the most common symptoms encountered in medical practice. However, little is known about the causal relationship between change in lifestyle and fatigue. Aim: To help prevent fatigue-related disorders, we investigated the association between changes in lifestyle and fatigue among employees. Methods: We studied data sets from the High-risk and Population Strategy for Occupational Health Promotion study for employees at 10 workplaces in Japan. The baseline survey was done in 1999 and the follow-up survey in 2003 via a questionnaire which examined lifestyle and fatigue variables using the vitality domain scale of the SF-36 Health Survey. The lifestyle factors focused on were diet, smoking and alcohol habits and working conditions. Four-year changes in lifestyle that predicted the vitality domain score in the follow-up survey were examined by analysis of covariance. Results: Of the 6284 participants in the baseline survey, 4507 replied to the follow-up survey, of whom 3498, with a mean age of 37 (SD 18) years, returned valid responses. A low vitality score at follow-up was predicted by a change in lifestyle factors such as an increase in overtime work, change to non-sedentary work and increased frequency of eating between meals (P < 0.01, P < 0.01 and P = 0.02, respectively). Conclusion: Fatigue in salaried workers as measured by the vitality domain of the SF-36 is predicted by an increase in overtime work, change to non-sedentary work and an increase in the frequency of eating between meals. © The Author 2007. Published by Oxford University Press on behalf of the Society of Occupational Medicine. All rights reserved.

OBJECTIVES: Hypertension as assessed by blood pressure (BP) measurement is one of the most important risk factors for cardiovascular diseases (CVD). Self-reported history of hypertension (self-reported HT) is an easy way to obtain information on BP and is known to have a certain sensitivity and high specificity for hypertension confirmed by BP measurement (confirmative HT). Thus, it might predict CVD mortality, but few studies have reported on this relationship. METHODS: We followed 6427 participants aged 30-59 years without a history of CVD for 19 years. The multivariate-adjusted hazard ratio (HR) of CVD mortality was estimated by the Cox proportional hazard model. RESULTS: The sensitivity and specificity of self-reported HT for confirmative HT were 52-65% and 95%, respectively. The multivariate-adjusted HR of self-reported HT for CVD death was 2.49 [95% confidence interval (CI) = 1.72-3.61]. Compared to participants with neither self-reported HT nor confirmative HT, those with confirmative HT showed a consistently higher HR for CVD mortality. Self-reported HT without confirmative HT was also significantly related to CVD mortality (HR = 2.10, 95% CI = 1.04-4.26). These tendencies were unchanged when we further adjusted for systolic BP (SBP) level. The age-adjusted mortality rate of individuals with self-reported HT corresponded to the age-adjusted mortality rate of individuals whose SBP was 160-179 mmHg. CONCLUSION: Self-reported HT could screen one-half of the participants for confirmative HT and was significantly associated with CVD mortality. These results indicate that self-reported HT can be a useful tool to screen for individuals with high BP if it is difficult to perform BP measurements continuously among all members of a community. © 2007 Lippincott Williams & Wilkins, Inc.

OBJECTIVE - Metabolic syndrome is diagnosed according to several criteria. Of these, some require glucose intolerance and others require obesity for the diagnosis. We investigated the relationship between metabolic risk factor clustering and cardiovascular disease (CVD) mortality stratified by high blood glucose or obesity. RESEARCH DESIGN AND METHODS - We followed 7,219 Japanese men and women without a history of CVD for 9.6 years. We defined high blood pressure, high blood glucose, high triglycerides, low HDL cholesterol, and obesity as metabolic factors. The multivariate adjusted hazard ratio (HR) for CVD mortality according to the number of clustering metabolic factors was calculated using the Cox proportional hazards model. RESULTS - During follow-up, 173 participants died of CVD. The numbers of metabolic risk factors and CVD mortality were positively correlated (Ptrend = 0.07). The HR was obviously higher among participants with than among those without high blood glucose and clustering of ≥2 other metabolic risk factors (HR 3.67 [95% CI 1.49-9.03]). However, the risk increase was only modest in participants without high blood glucose even if they had ≥2 other metabolic risk factors (1.99 [0.93-4.28]). Conversely, metabolic risk factor clustering was related to CVD mortality irrespective of obesity. CONCLUSIONS - Our findings suggest that glucose tolerance plays an important role in CVD mortality. Because the prevalence of nonobese participants with several metabolic risk factors was quite high and their CVD risk was high, excluding them from the diagnosis of metabolic syndrome because of the absence of obesity might overlook their risk. © 2007 by the American Diabetes Association.

Background: The life expectancy is an important measure for describing health status among population. Several studies from the United States and Europe showed the harm of smoking by describing the life expectancies with different smoking status. No such study is examined in Japan, the country with the world's highest life expectancy irrespective of high smoking rate among men. Methods: The abridged life table method was applied to calculate the life expectancies of men and women among different smoking status from age 40 until age 85. Age-specific mortality rates stratified by different smoking status were obtained from follow-up data from random sample in Japanese population (NIPPON DATA80). Results: Proportion of current smokers was 62.9% in men and 8.8% in women at the baseline survey in 1980. The life expectancies of 40-year-old never smokers, ex-smokers and current smokers were 42.1, 40.4, and 38.6 years in men and 45.6, 45.9, and 43.4 years in women. The life expectancy of 40-year-old men who smoked less than one pack per day was 39.0 and was longer than that of those who smoked one or two packs (38.8) and more than two packs (38.1). Conclusion: Life expectancy decreased gradually as the grade of smoking increased in the Japanese population. Copyright © 2007 by Japan Epidemiological Association.