Systolic Blood Pressure Predicts Cardiovascular Mortality in a Farming but Not in a Fishing Community

Transcription

1 Circulation Journal Official Journal of the Japanese Circulation Society ORIGINAL ARTICLE Epidemiology Systolic Blood Pressure Predicts Cardiovascular Mortality in a Farming but Not in a Fishing Community A 40-Year Follow up of the Japanese Cohorts of the Seven Countries Study Yuji Hirai, MD, PhD; Johanna M. Geleijnse, PhD; Hisashi Adachi, MD, PhD; Tsutomu Imaizumi, MD, PhD; Daan Kromhout, PhD Background: Blood pressure (BP) is a strong determinant of cardiovascular diseases (CVD). The strength of this association in 2 Japanese communities with different intakes of fish was investigated. Methods and Results: The analysis was carried out in the Japanese cohorts of the Seven Countries Study (Tanushimaru and Ushibuka), which were followed for 40 years. We included 1,006 subjects for whom data on baseline BP and relevant potential confounders were available. Data were analysed using multivariable Cox proportional hazard models. In Tanushimaru men, the systolic BP level was strongly directly related to risk of stroke and, with hazard ratios (HR) of 4.42 ( ) for stroke and 3.05 ( ) for CVD for BP levels 140 mmhg compared to <120 mmhg. In Ushibuka, the HR were 1.74 ( ) for stroke mortality and 1.66 ( ) for for high vs. low systolic BP. With regard to diastolic BP, the associations with stroke and were similar in Tanushimaru and Ushibuka subjects. Conclusions: This study showed that the well-known relationship of systolic BP with stroke and was more pronounced in the Japanese farming community than in the fishing community. This brings up the hypothesis that the detrimental effect of raised systolic BP could be attenuated by a high intake of fish. (Circ J 2011; 75: ) Key Words: Blood pressure; Cardiovascular diseases; Epidemiology; Nutrition The mortality pattern of Japan is characterized by a low mortality rate from coronary heart disease (CHD) and a high rate from stroke. 1 In spite of a large decrease in the past decades, stroke is still the major contributor to cardiovascular mortality in the Japanese population. 2 4 Elevated blood pressure (BP) is a major contributor to mortality and incident cardiovascular diseases (CVD), especially stroke, in Japan. 5,6 In spite of an increasing trend in serum cholesterol in more recent years in Japan, the CHD mortality rates have remained relatively low possibly because the average serum cholesterol level stayed below 5 mmol/l. 7 Also the high level of fish intake and consequently the high content of the very long-chain n-3 fatty acids, eicosapentanoic acid (EPA) and docosahexaenoic acid (DHA), in the Japanese diet, were put forward as an explanation for the low CHD mortality rate. 8 Meta-analyses of prospective cohort studies have shown that a small amount of fish lowers the risk of fatal CHD and the incidence of fatal and non-fatal stroke. 9,10 Also a modest intake of marine-derived n-3 fatty acids ( mg/day) was associated with a low risk of CHD mortality in a metaanalysis that included both prospective cohort studies and clinical trials. 11 This low level of n-3 fatty acids does not lower BP but a high intake (>3 g/day) does. 12,13 In Japan several studies have been carried out comparing farming and fishing villages These studies showed that the average level of fish consumption in farming villages was approximately 100 g/day, whereas it was g/day in fishing villages. The levels of n-3 fatty acids in blood were higher and platelet aggregation, pulse wave velocity, an indicator of atherosclerosis, and Intima Media Thickness were more favorable in people from the fishing villages In the present study, we investigated whether the strength of the well-known graded association between BP and longterm risk of stroke and total differed between a farming and a fishing community. We examined this re- Received October 5, 2010; revised manuscript received March 23, 2011; accepted March 31, 2011; released online June 15, 2011 Time for primary review: 51 days Department of Internal Medicine, Division of Cardiovascular Medicine, Kurume University School of Medicine, Kurume (Y.H., H.A., T.I.), Japan; Division of Human Nutrition, Wageningen University, Wageningen (Y.H., J.M.G., D.K.), The Netherlands Mailing address: Johanna M. Geleijnse, PhD, Division of Human Nutrition, Wageningen University, P.O. Box 8129, 6700 EV Wageningen, The Netherlands. marianne.geleijnse@wur.nl ISSN doi: /circj.CJ All rights are reserved to the Japanese Circulation Society. For permissions, please cj@j-circ.or.jp

2 BP and CVD in Japanese Farmers and Fishermen 1891 Table 1. Baseline Characteristics of Middle-Aged Men in Tanushimaru in 1958 and Ushibuka in 1960 Tanushimaru (n=508) Ushibuka (n=498) P value* Age, years 50.1 (5.7) 49.5 (5.6) 0.07 Height, cm (5.4) (5.9) 0.08 Weight, kg 56.0 (6.9) 56.4 (7.2) 0.46 Body mass index, kg/m (2.3) 22.1 (2.5) 0.05 Serum cholesterol, mmol/l** 3.9 (0.9) 3.9 (0.8) 0.84 Systolic BP, mmhg (24.4) (25.5) 0.02 Diastolic BP, mmhg 73.7 (14.1) 78.4 (13.0) <0.01 Hypertension, % Resting heart rate, beats/min 60.6 (10.6) 65.2 (10.9) <0.01 Smoking status, % Never Former Current Job-related physical activity, % Hard physical work Moderate physical work <0.01 Sedentary Occupation Farmers, % Fishermen, % Professionals, % <0.01 Manual workers, % Prevalence of CVD, % Values are presented as mean (SD) or percentages. BP, blood pressure; CVD, cardiovascular diseases. *P value obtained by unpaired t-test (continuous variables) or chi-squared test (categorical variables). **Multiply by to convert to mg/dl. Systolic BP > 140 mmhg or diastolic BP > 90 mmhg. Clinically diagnosed myocardial infarction or stroke. search question in 2 Japanese cohorts of the Seven Countries Study. Methods Seven Countries Study The Seven Countries Study is an epidemiologic populationbased study that started in the late 1950 s. In 16 cohorts including 2 Japanese ones, men aged years were enrolled between 1958 and The cohort of Tanushimaru (n= 508, examined in 1958) consisted predominantly of farmers, whereas the cohort of Ushibuka (n=502, examined in 1960) consisted mainly of fishermen. 19,20 The present analysis includes a total of 1,006 men for whom complete data on baseline BP and major confounders was available. BP BP was measured from the right arm by trained physicians at the end of the physical examination using a calibrated mercury sphygmomanometer, with the subject in the supine position, as described in the World Health Organization (WHO) manual, Cardiovascular Survey Methods. 21 The mean of 2 measurements, taken 1 min apart, was computed for both the systolic and the diastolic (phase V) BP. Hypertension was defined as a systolic BP of 140 mmhg or higher, a diastolic BP of 90 mmhg or higher, or both. During the baseline surveys, medication to lower BP was rarely prescribed in Japan and therefore medication use was not included in the definition of hypertension. 22 Other Baseline Risk Factors Height and weight were measured once when subjects were in light underwear and were without shoes, and the body mass index (BMI) was calculated. 21 Resting heart rate was calculated from the resting ECG as the average of the rates in lead I and V6. Non-fasting blood samples were drawn and cholesterol was determined in fresh serum after saponification and by color development with ferric ion. 23,24 A few dried serum samples were sent from Japan to Minneapolis, USA, the coordinating center of the Seven Countries Study. Minneapolis serum cholesterol analyses were conducted according to the Abell Kendall method, 25 modified by Anderson and Keys. 26 The average serum cholesterol value was higher in Tanushimaru samples than in Minneapolis samples (4.34 vs mmol/l) in Therefore, the cholesterol values in Tanushimaru, were adjusted to the Minneapolis values by multiplying the result with In 1960 in Ushibuka, the data were in good agreement with those from Minneapolis. Information about smoking habits and occupation was collected by means of standardized questionnaires. Smoking status was coded into 3 categories: current, former and never cigarette smokers. Occupation was coded into 4 categories: farmers, fishermen, manual workers and professionals including business men and governmental officials. 19 The men were classified according to their habitual job-related physical activity into 3 categories: sedentary, moderate physical work and hard physical work. 19 Mortality During 40 years of follow up, the vital status of the men was

3 1892 HIRAI Y et al. Table 2. Rates for 40-Year Mortality (Expressed per 10,000 Person-Years) in Middle-Aged Men in Tanushimaru and Ushibuka Tanushimaru (n=508) Ushibuka (n=498) Person-years 13,116 12,003 Total cardiovascular diseases (n) 90.0 (118) (130) Stroke (n) 57.2 (75) 65.0 (78) Coronary heart disease (n) 22.9 (30) 19.2 (23) Other cardiovascular diseases (n)* 9.9 (13) 24.2 (29)** Total mortality (n) (418) (413) *Comprised mainly symptomatic heart disease (ICD-8 code 427), that is, 9 cases in Tanishimaru and 19 cases in Ushibuka; other ICD-8 codes were 395, 398, 400, 402, 403, 424, 444 and 450. **P<0.05; other differences between the 2 communities were not statistically significant. Table 3. Systolic BP and 40-Year Mortality From Stroke and Total CVD in Tanushimaru and Ushibuka Baseline systolic BP <120 mmhg mmhg > 140 mmhg P for trend Tanushimaru n=158 n=188 n=162 No. of events HR, model 1* ( ) 4.22 ( ) <0.001 HR, model 2** ( ) 4.42 ( ) <0.001 No. of events HR, model ( ) 3.09 ( ) <0.001 HR, model ( ) 3.05 ( ) <0.001 Ushibuka n=122 n=186 n=190 No. of events HR, model ( ) 1.91 ( ) <0.001 HR, model ( ) 1.74 ( ) <0.001 No. of events HR, model ( ) 1.53 ( ) HR, model ( ) 1.66 ( ) <0.001 HR, hazard ratio, obtained by Cox proportional hazard analysis, with 95% confidence interval. Other abbreviations see in Table 1. *Adjusted for age. **Adjusted for age, height, weight, smoking status, physical activity and occupation. continuously monitored and no one was lost to follow up. Information was collected on the different causes of death. In Tanushimaru, the 40-year follow-up period was completed. In Ushibuka, the censor date was September 1999 and the baseline examinations took place between May and July So, the follow-up period in Ushibuka was 39.3 years. During the first 25 years of follow up, information about the causes of death was obtained from official death certificates, medical and hospital records, and relatives of the person who died or other witnesses. After 25 years, only death certificates were available. In cases in which multiple causes of death were possible, priority was given to violent death, followed by cancer in an advanced stage, CHD, and stroke, respectively. The final adjudication of the underlying cause of death and the contributory causes was made according to the criteria of the WHO International Classification of Diseases, 8 th Revision (ICD-8), 27 by an experienced clinical epidemiologist. Stroke was defined as ICD-8 codes 430 through 438, and CHD as ICD-8 codes 410 through 414 and 795. Total CVD were defined as diseases of the circulatory system, ICD-8 codes 390 through 458, and 795 when sudden death of cardiac origin was mentioned. Statistical Methods We computed means and standard deviations for continuous variables and relative frequencies (percentages) for discrete ones. We compared means of both cohorts using the t- or χ2 test and a 2-sided P-value of <0.05 was considered statistically significant. BP was classified into 3 categories: <120 mmhg, mmhg and 140 mmhg for systolic BP and <70, and 85 mmhg for diastolic BP. Cox proportional hazard analysis was performed for the 2 different cohorts to estimate the strength of associations between BP and 40-year mortality from stroke and total CVD. Hazard ratios were obtained with 95% confidence intervals. In multivariable analysis, we adjusted for age, height, weight, smoking status (current or former/never), physical activity (sedentary, moderate physical work or hard physical work), and occupation (farmers, fishermen, manual workers or professionals). We repeated the analyses in a subpopulation of farmers from Tanushimaru (n=393) and fishermen from Ushibuka (n=324). Statistical interaction between area of residence (Tanushimaru vs.

4 BP and CVD in Japanese Farmers and Fishermen 1893 Hazard ratio 6.0 A Tanushimaru Ushibuka P for interaction = 0.09 < 120 mmhg mmhg 140 mmhg Systolic BP Hazard ratio B Tanushimaru farmers Ushibuka fishermen P for interaction = 0.05 < 120 mmhg mmhg Systolic BP 140 mmhg Figure. Hazard ratios with 95% confidence intervals for 40-year mortality from cardiovascular diseases in categories of systolic blood pressure (BP) in 2 Japanese communities. Hazard ratios have been adjusted for height, weight, smoking status, physical activity and occupation. Findings for Tanushimaru are depicted in blue, and findings for Ushibuka are depicted in red. (A) Total sample of Tanushimaru men (n=508) and Ushibuka men (n= 498). (B) Selection of farmers from Tanushimaru (n=393) and fishermen from Ushibuka (n=324). Ushibuka) and systolic BP (as a continuous variable) was tested by adding the product term to the multivariable models. We did not analyze the association of BP with CHD mortality or mortality from other CVD because of the limited number of cases. All statistical analyses were conducted using the SPSS system statistical software (version 17.0). Results The baseline characteristics of the 2 cohorts are shown in Table 1. Average BMI (22 kg/m 2 ) and serum cholesterol (4 mmol/l) were low in both communities. The prevalence of hypertension was 7% lower in Tanushimaru than in Ushibuka (P=0.02). Most men smoked; 71% in Tanushimaru and 78% in Ushibuka. In Ushibuka, more men had a sedentary job (7% vs. 4% in Tanushimaru; P<0.01). Men from Tanushimaru were mainly farmers (78%) whereas men from Ushibuka were mainly fishermen (65%). The prevalence of CVD (defined as a clinically diagnosed myocardial infarction or stroke) was below 1% in both communities. Long-term mortality rates for different causes of death in Tanushimaru and Ushibuka cohorts are presented in Table 2. The total CVD and overall mortality were lower in the Tanushimaru than in the Ushibuka cohort, but these differences were not statistically significant. Within the category of, Tanushimaru men had a significantly lower risk of symptomatic heart disease (ICD-8 code 427) than Ushibuka men. In the Tanushimaru cohort, the risk of stroke and showed a strong and graded relationship with systolic BP (Table 3; Figure A), which remained after adjustment for confounders. Men with a systolic BP of 140 mmhg or higher had a more than 4-fold higher risk of stroke mortality and a 3-fold higher risk of compared with men who had a BP below 120 mmhg. Ushibuka men with a systolic BP of 140 mmhg or higher (but not those with a BP of mmhg) had a 1.7-fold higher risk of stroke and (Table 3; Figure A). The interaction between area of residence and systolic BP was statistically significant for stroke (P=0.03) and borderline significant for CVD mortality (P=0.09). In the subpopulation of Tanushimaru farmers, the 40-year

5 1894 HIRAI Y et al. Table 4. Diastolic BP and 40-Year Mortality From Stroke and Total CVD in Tanushimaru and Ushibuka Baseline diastolic BP <70 mmhg mmhg > 85 mmhg P for trend Tanushimaru n=184 n=219 n=105 No. of events HR, model 1* ( ) 2.47 ( ) <0.001 HR, model 2** ( ) 2.81 ( ) <0.001 No. of events HR, model ( ) 2.18 ( ) <0.001 HR, model ( ) 2.32 ( ) <0.001 Ushibuka n=92 n=266 n=140 No. of events HR, model ( ) 2.48 ( ) <0.001 HR, model ( ) 2.71 ( ) <0.001 No. of events HR, model ( ) 2.05 ( ) <0.001 HR, model ( ) 2.58 ( ) <0.001 Abbreviations see in Tables 1,2. *Adjusted for age. **Adjusted for age, height, weight, smoking status, physical activity and occupation. mortality rate from stroke (per 10,000 person-years) was 66.2, whereas it was lower (62.5 per 10,000 person-years) in the subpopulation of Ushibuka fishermen. The rates for CVD were 99.8 and 100.8, respectively (data not presented in Tables). In Tanushimaru farmers, the relative risks for stroke were 2.13 ( ) for systolic BP levels of mmhg, and 4.62 ( ) for levels >140 mmhg compared to low BP levels, which was comparable to relative risks in the total Tanushimaru cohort. For, stronger associations were observed in farmers only (Figure B), with relative risks of 2.01 ( ) for systolic BP levels of mmhg and 3.92 ( ) for systolic BP 140 mmhg (P for trend <0.001). When examining only fishermen in the Ushibuka cohort, systolic BP was no longer significantly related to stroke mortality, with relative risks of 0.92 ( ) for levels of mmhg and 1.36 ( ) for levels >140 mmhg. Also in fishermen, was no longer related to systolic BP (Figure B). The relative risks in this subpopulation were 0.76 ( ) for levels of mmhg and 1.05 ( ) for levels >140 mmhg (P for trend: 0.70). The statistical interaction terms between area of residence and systolic BP in this restricted analysis of farmers and fishermen were borderline statistically significant both for stroke mortality (P=0.06) and (P=0.05). For diastolic BP, no differences in the association with CVD or stroke was observed between Tanushimaru and Ushibuka cohorts (Table 4). In both communities, the risk of mortality from stroke and CVD was non-significantly increased by ~30% for diastolic BP between mmhg, and significantly increased by ~150% in the category of 85 mmhg or higher, compared to levels <70 mmhg (Table 4). Discussion The associations of systolic BP with 40-year mortality from stroke and CVD differed between the two communities. In the Tanushimaru cohort, a graded relationship was observed and in the Ushibuka cohort, only men with a systolic BP of 140 mmhg or higher had an elevated risk. When the analysis was restricted to the subpopulation of farmers in Tanushimaru, the strong associations of systolic BP with stroke and CVD mortality remained. However, in Ushibuka fishermen, the associations of BP with stroke and disappeared. Remarkably, this phenomenon was not observed for diastolic BP for which the risk of stroke and was doubled at high levels (ie, 85 mmhg) in both communities. The present study has several limitations. The number of men examined at baseline in the Tanushimaru and Ushibuka cohorts was relatively low. Also, we analysed baseline BP data only because repeated BP measurements during follow up were only available for part of the cohort. If more BP measurements had been taken into account during follow up, this would have increased the precision and probably yielded stronger associations with stroke and. We examined only men and findings might not apply to women. The study also has major strengths, including the large difference in fish consumption between the 2 communities within 1 culture with the same genetic background. Furthermore, no one was lost to follow up. Dietary surveys were carried out in Tanushimaru in 1964 and in Ushibuka in The results of these surveys and the chemical analysis of food composites representing the average food consumption pattern of these men have been published previously The average fish consumption was more than twice as high in the Ushibuka cohort (207 g/day) than in the Tanushimaru cohort (93 g/day). 28 The average intake of the n-3 fatty acids EPA and DHA was 1.9 g/day in Ushibuka and 1.2 g/day in Tanushimaru. 29 Saturated fat intake was low in both communities; 14 and 10 g/day, respectively. 30 The average intake of other macronutrients, flavonoids and total energy showed little difference An overview of dietary intakes and CVD risk factors in both communities, including changes during follow up, is provided as a Web

6 BP and CVD in Japanese Farmers and Fishermen 1895 appendix (Tables S1 and S2). The main finding of the present study is that fishermen from Ushibuka, who were probably the largest fish consumers in that community, showed no relationship between systolic BP and long-term risk of mortality from stroke and total CVD. This is in contrast with the well-known graded relationship of systolic BP with the endpoints, as confirmed in the farming community of Tanushimaru, where farmers with a systolic BP >140 mmhg were at a 4-fold risk of dying from stroke and CVD compared to men with levels <120 mmhg. An intriguing question is how to explain this difference in terms of the association of systolic BP with stroke and CVD mortality between Japanese farmers and fishermen. In 1994, Yamada et al examined Japanese middle-aged farmers from the village, Haze, and fishermen from the village, Kamishima. 18 The average fish consumption was 100 g/day in the farming village and almost twice as high in the fishing village. In that study, the Intima Media Thickness and the number of plaques were much lower in the fishermen compared with the farmers. 18 The most likely explanation for the differences in these measures of atherosclerosis is the difference in fish consumption between farmers and fishermen. A comparative study of Japanese and American whites aged years showed that Japanese men had twice the level of marine-derived fatty acids in their blood and the lowest level of atherosclerosis. 32 Significant differences between Japanese and American whites in Intima Media Thickness and Coronary Artery Calcification prevalence became nonsignificant after adjustment for n-3 fatty acids. This suggests that n-3 fatty acids might reduce the burden of atherosclerosis. The effect of very long-chain n-3 fatty acids (1.4 g/day) on plaque stability was examined in elderly patients waiting to undergo carotid endarterectomy. 33 This study showed that n-3 fatty acids were rapidly incorporated into the plaque where they contributed to plaque stability. The authors concluded that the increased plaque stability might reduce the risk of cardiovascular events. Remarkably, we found no difference in the association of diastolic BP with mortality from stroke and CVD between the Tanushimaru and Ushibuka cohorts. We have no explanation why only the association with systolic BP was blunted in Ushibuka men. However, in line with our hypothesis, one may speculate that n-3 fatty acids prevents rupture of atherosclerotic plaques, especially at peak BP levels. In summary, the present study showed that the established graded relationship of systolic BP with stroke and was not present in the fishing community of Ushibuka. These results suggest that the detrimental effect of systolic hypertension on could possibly be attenuated by very-long chain n-3 fatty acids. In the Honolulu Heart Program that started in 1965, CHD mortality rates were not related to the number of daily cigarettes smoked by year-old Japanese-American men with a high fish intake ( 2 times/week). 34 These data confirm the idea that a high fish intake could blunt the adverse effects of classical risk factors on CVD. This hypothesis, however, needs to be confirmed in other long-term, population-based studies with subjects who have a wide range of fish intake. Acknowledgments We are grateful to members of the Japan Medical Association of Ukiha and Amakusa, the elected officials and residents of Tanushimaru and Ushibuka, and the team of cooperating physicians for their help in performing the health examinations. This study was supported, in part, by the Kimura Memorial Heart Foundation, Fukuoka, Japan. Disclosure Authors conflicts of interest: None. References 1. Feigin VL, Lawes CMM, Bennett DA, Anderson CS. Stroke epidemiology: A review of population-based studies of incidence, prevalence, and case-fatality in late 20th century. Lancet Neurology 2003; 2: Ma E, Iso H, Takahashi H, Yamagishi K, Tanigawa T. Age-Period- Cohort analysis of mortality due to ischemic heart disease in Japan, 1955 to Circ J 2008; 72: Ma E, Takahashi H, Mizuno A, Okada M, Yamagishi K, Iso H. Stratified Age-Period-Cohort analysis of stroke mortality in Japan, 1960 to J Stroke Cerebrovasc Dis 2007; 16: Kitamura A, Sato S, Kiyama M, Imano H, Iso H, Okada T, et al. Trends in the incidence of coronary heart disease and stroke and their risk factors in Japan, 1964 to 2003: The Akita-Osaka study. J Am Coll Cardiol 2008; 52: Kang G, Guo L, Guo Z, Hu X, Wu M, Zhou Z, et al. Impact of blood pressure and other components of the metabolic syndrome on the development of cardiovascular disease. Circ J 2010; 74: Mori I, Ishizuka T, Morita H, Matsumoto M, Uno Y, Kajita K, et al. Comparison of biochemical data, blood pressure and physical activity between longevity and non-longevity districts in Japan. Circ J 2008; 72: Blackburn H, Jacobs DR. The ongoing natural experiment of cardiovascular diseases in Japan. Circulation 1989; 79: Harris W. Omega-3 fatty acids: The Japanese factor? J Am Coll Cardiol 2008; 52: He K, Song Y, Daviglus ML, Liu K, Van Horn L, Dyer AR, et al. Accumulated evidence on fish consumption and coronary heart disease: A meta-analysis of cohort studies. Circulation 2004; 109: He K, Song Y, Daviglus ML, Liu K, Van Horn L, Dyer AR, et al. Fish consumption and the incidence of stroke: A meta-analysis of cohort studies. Stroke 2004; 35: Mozaffarian D, Rimm EB. Fish intake contaminants, and human health: Evaluating the risks and the benefits. JAMA 2006; 296: Morris MC, Sacks F, Rosner B. Does fish oil lower BP? A metaanalysis of controlled trials. Circulation 1993; 88: Geleijnse JM, Giltay EJ, Grobbee DE, Donders ART, Kok FJ. BP response to fish oil supplementation: Metaregression analysis of randomized trials. J Hypertens 2002; 20: Kagawa Y, Nishizawa M, Suzuki M, Miyatake T, Hamamoto T, Goto K, et al. Eicosapolyenoic acids of serum lipids of Japanese islanders with low incidence of cardiovascular diseases. J Nutr Sci Vitaminol 1982; 28: Yamori Y, Nara Y, Iritani N, Workman RJ, Inagami T. Comparison of serum phospholipid fatty acids among fishing and farming Japanese populations and American inlanders. J Nutr Sci Vitaminol 1985; 31: Hamazaki T, Urakaza M, Sawazaki S, Yamazaki K, Taki H, Yano S. Comparison of pulse wave velocity of the aorta between inhabitants of fishing and farming villages in Japan. Atherosclerosis 1988; 73: Hirai A, Terano T, Tamura Y, Yoshida S. Eicosapentaenoic acid and adult diseases in Japan: Epidemiological and clinical aspects. J Intern Med Suppl 1989; 731: Yamada T, Strong J, Ishii T, Ueno T, Koyama M, Wagayama H, et al. Atherosclerosis and omega-3 fatty acids in the populations of a fishing village and a farming village in Japan. Atherosclerosis 2000; 153: Keys A, Aravanis C, Blackburn H, Van Buchem FS, Buzina R, Djordjevic BS, et al. Epidemiological studies related to coronary heart disease: Characteristics in men aged in seven countries. Acta Med Scand Suppl 1966; 460: Toshima H, Koga Y, Blackburn H, Keys A. Lessons for science from the Seven Countries Study. Tokyo: Springer-Verlag, 1994; Rose GA, Blackburn H. Cardiovascular survey methods. World Health Organization monograph series no. 56. Geneva: World health Organization, Van den Hoogen P, Feskens E, Nagelkerke N, Menotti A, Nissinen A, Kromhout D, for the Seven Countries Study Research Group. The relation between BP and mortality due to coronary heart disease among men in different parts of the world. N Engl J Med 2000;

7 1896 HIRAI Y et al. 342: Zlatkis A, Zak B, Boyle AJ. A new method for the direct determination of serum cholesterol. J Lab Clin Med 1953; 41: Henley AA. The determination of serum cholesterol. Analyst 1957; 82: Abell LL, Levy BB, Brodie BB, Kendall FE. A simplified method for the estimation of total cholesterol in serum and the demonstration of its specificity. J Biol Chem 1952; 195: Anderson JT, Keys A. Cholesterol in serum and lipoprotein fractions: Its measurement and stability. Clin Chem 1956; 2: World Health Organization. Manual of the international statistical classification of diseases, injuries, and causes of death: Based on the recommendations of the Eighth Revision Conference, 1965, and adopted by the Nineteenth World Health Assembly. Vol. 1. International classification of diseases. Geneva: World Health Organization, Kromhout D, Keys A, Aravanis C, Buzina R, Fidanza F, Giampaoli S, et al. Food consumption patterns in the 1960s in seven countries. Am J Clin Nutr 1989; 49: De Vries J, Jansen A, Kromhout D, Van de Bovenkamp P, Van Staveren W, Mensink R, et al. The fatty acids and sterol content of food composites of middle-aged men in seven countries. J Food Comp Anal 1997; 10: Kromhout D, Menotti A, Bloemberg B, Aravanis C, Blackburn H, Buzina R, et al. Dietary saturated and trans fatty acids and cholesterol and 25-year mortality from coronary heart disease: the Seven Countries Study. Prev Med 1995; 24: Hertog MG, Kromhout D, Aravanis C, Blackburn H, Buzina R, Fidanza F, et al. Flavonoid intake and long-term risk of coronary heart disease and cancer in the seven countries study. Arch Intern Med 1995; 155: Sekikawa A, Curb JD, Ueshima H, El-Saed A, Kadowaki T, Abbott RD, et al. Marine-derived n-3 fatty acids and atherosclerosis in Japanese, Japanese-Americans, and white men: A cross-sectional study. J Am Coll Cardiol 2008; 52: Thies F, Garry JMC, Yaqoob P, Rerkasem K, Williams J, Shearman CP, et al. Association of n-3 polyunsaturated fatty acids with stability of atherosclerotic plaques: A randomized controlled trial. Lancet 2003; 361: Rodriguez BL, Sharp DS, Abbott RD, Burchfiel CM, Masaki K, Chyou PH, 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: Supplemental Files Supplemental File 1 Table S1. Dietary Intakes of Middle-Aged Men in Tanushimaru in 1964 and Ushibuka in 1971 (Data Derived From References i iv) Table S2. Trends in Risk Factors in Farmers and Fishermen s Villages in Japan Please find supplemental file(s);