Published online before print April 28, 2008, doi: 10.1161/HYPERTENSIONAHA.107.102459
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Hypertension. 2008;51:1483.)
© 2008 American Heart Association, Inc.
Original Articles |
Relation of Blood Pressure and All-Cause Mortality in 180 000 Japanese Participants
Pooled Analysis of 13 Cohort Studies
From the Department of Health Science (Y.M., A.H., H.U.), Shiga University of Medical Science, Otsu, Japan; and the Department of Preventive Cardiology (T.O.), National Cardiovascular Center, Suita, Japan.
Correspondence to Yoshitaka Murakami, Department of Health Science, Shiga University of Medical Science. Seta Tsukinowa-cho, Otsu City, Shiga, 520-2192, Japan. E-mail ymura{at}belle.shiga-med.ac.jp
 |
Abstract |
|---|
 Top Abstract IntroductionStudy Participants and MethodsResultsDiscussionAppendixReferences |
|---|
Hypertension is a leading cause of death because of cardiovascular disease and predominantly affects total mortality. To reduce avoidable deaths from hypertension, we need to collect blood pressure data and assess their impact on total mortality. To examine this issue, a meta-analysis of 13 cohort studies was conducted in Japan. Poisson regression was used for estimating all-cause mortality rates and ratios. In the model, blood pressure data were treated as continuous (10-mm Hg increase) and categorical (every 10 mm Hg) according to recommendations of the Seventh Joint National Committee on Prevention, Detection, Evaluation, and Treatment of Hypertension. Potential confounders included body mass index, smoking, drinking, and cohort. The impact of hypertension was measured by the population-attributable fraction. After excluding participants with cardiovascular disease history, 176 389 participants were examined in the analysis. Adjusted mortality rates became larger as the blood pressure increased, and these were more distinct in younger men and women. Hazard ratios also showed the same trends, and these trends were more apparent in younger men (hazard ratio [unit: 10-mm Hg increase] aged 40 to 49 years: systolic blood pressure 1.37 (range: 1.15 to 1.62); diastolic blood pressure 1.46 [range: 1.05 to 2.03]) than older ones (hazard ratio: aged 80 to 89 years: systolic blood pressure 1.09 [range: 1.05 to 1.13]and diastolic blood pressure 1.12 [range: 1.03 to 1.22]). Population-attributable fraction of hypertension was
20% when the normal category was used as a reference level and was 10% when we included the prehypertension group in the reference level. In conclusion, high blood pressure raised the risk of total mortality, and this trend was higher in the younger Japanese population.
Key Words: pooled analysis • total mortality • cohort study • blood pressure • population attributable fraction
|
Introduction |
|---|
|
TopAbstractIntroductionStudy Participants and MethodsResultsDiscussionAppendixReferences |
|---|
High blood pressure is a well-established leading cause of cardiovascular disease mortality. The contribution of hypertension to total mortality is also large,1 and the importance of management of hypertension is widely accepted not only in clinical practice but also in public health practice. Before measures for reducing hypertension are implemented, more information about the relationship between hypertension and total mortality is needed. For example, one study found that the hazard ratio of cardiovascular disease in those with high blood pressure was larger in younger than in older participants,2 suggesting that the contribution of high blood pressure to mortality differed at different ages. The examination of this important issue requires a large number of participants, and a single cohort study estimating this contribution is limited by small sample size. The Joint National Committee of Hypertension 7 (JNC-7) also mentioned the importance of the risk of mortality and the contribution of prehypertension to total mortality.3 The relation between blood pressure and total mortality is, therefore, of great interest. It is difficult to address this issue in a single cohort study, because few events can be observed in cohorts subgrouped, eg, by age or prehypertension status. A large-scale cohort study could answer these questions, but the huge amount of cost and effort involved represent serious obstacles. Meta-analysis using data on individual participants4 is an efficient way to deal with this issue, and the approach has been used in studies of cardiovascular disease epidemiology.2,5,6
In Japan, meta-analysis of individual participants’ data for cardiovascular disease was conducted in the Japanese population. The study, called Evidence for Cardiovascular Prevention From Observational Cohorts in Japan (EPOCH-JAPAN), included 13 cohort studies of existing Japanese cohorts. The total number of EPOCH-JAPAN participants was 188 321, with 10 years of follow-up. The purpose of this study was to examine sex- and age-specific hazard ratios and the effect of blood pressure on total mortality and to estimate the contribution of high blood pressure to all-cause death by performing a meta-analysis of the data from 13 population-based cohort studies conducted in Japan.
|
Study Participants and Methods |
|---|
|
TopAbstractIntroductionStudy Participants and MethodsResultsDiscussionAppendixReferences |
|---|
Study Cohorts
The EPOCH-JAPAN Study is the pooled analysis of 13 cohort studies examining the relation between health measures (laboratory measures and lifestyle and behavioral factors) and disease (mortality and incidence) in the Japanese population. The criteria for inclusion of meta-analysis were as follows: collection of health examination measures, follow-up of
10 years, and a number of participants >1000 persons. Both nationwide and single-site cohort studies were included. The name of each cohort study7–22 is listed in Table 1
. Inclusion criteria for participants were age at entry (40 to 90 years old) and availability of information about sex, age at entry, systolic blood pressure, and diastolic blood pressure. Since the end of follow-up varied between cohorts, we limited age ranges of follow-up from 40 to 90 years, and the end of the observation period was set at age 90 years.
|
|
Statistical Methods
Hazard ratios for total mortality were estimated in men and women separately. Participants were stratified into 10-year age groups from 40 to 80 years, and a statistical model was made to analyze the data of each age group separately. They were also divided on the basis of systolic blood pressure (SBP) into 10-mm Hg groups from <120 mm Hg to 
160 mm Hg and on the basis of diastolic blood pressure (DBP) into 10-mm Hg groups from <70 mm Hg to 100 mm Hg. The lowest blood pressure group (<120 mm Hg for SBP and <70 mm Hg for DBP) served as the reference group.
A Poisson regression model was constructed for each sex and age group. When we analyzed the sex-combined results, we included sex in the model. In the model, we analyzed continuous and categorical blood pressure data to estimate the hazard ratio. Body mass index, smoking status (smokers, ex-smokers, or never-smokers), drinking status (drinkers, ex-drinkers or never-drinkers), and cohort were included in the model as confounders. Person-years of observation was separated into 5 categories (1 for every 10 years from age 40 to age 80 years) and used as an offset variable in Poisson regression. Multivariate adjusted mortality rates were estimated from the Poisson regression. Mortality rates among groups were calculated after adjusting for the population-averaged effects of confounders (eg, smoking, drinking, and mean body mass index).
The MacMahon and Peto method was used to correct for regression dilution bias of blood pressure.4,23 The dilution factors of blood pressure were derived from longitudinal blood pressure measurements made by the Ohasama Study that followed 1900 participants from 1999 to 2002. From the data, the calculated regression dilution ratio was 0.59 for SBP and 0.48 for DBP. Hazard ratio for total mortality according to the JNC-7 definition was also estimated in the analysis. The prehypertension (120
SBP<140 mm Hg or 80DBP<90 mm Hg), hypertension stage 1 (140SBP<160 mm Hg or 90DBP<100 mm Hg), and hypertension stage 2 (160 mm HgSBP or 100 mm HgDBP) groups were compared with the normal reference groups (<120 mm Hg for SBP and <80 mm Hg for DBP).
The population-attributable fraction of high blood pressure according to the JNC 7 classification was calculated from the hazard ratio. Two reference levels (the normal and below prehypertension [normal plus prehypertension]) and the excess hazard ratio were used to calculate the population-attributable fraction. The prevalence of hypertension was set as the total number of participants in this study.24 All of the statistical analysis was performed using SAS release 9.13 (SAS Institute Inc).
|
Results |
|---|
|
TopAbstractIntroductionStudy Participants and MethodsResultsDiscussionAppendixReferences |
|---|
Table 1 shows baseline characteristics of participants in each cohort. There were 188 141 participants (men: 70 558; women: 117 583). Average age at study entry was 59.6 years in men and 58.4 years in women, and the average follow-up period was 9.6 years in men and 9.9 years in women. The total number of deaths (all-cause) was 17 757 (men: 9880; women: 7877). The characteristics of the cohorts were similar, and no apparent differences in hazard ratio were found at each blood pressure level. In our study, we analyzed participants without cardiovascular disease history. Because there was no information of cardiovascular disease history in the Tanno-Sobetsu cohort, we excluded these participants, and, therefore, 176 389 (male: 65 463; female: 110 926) participants were investigated in our study.
The Figure shows the relation between blood pressure and multivariate-adjusted hazard rate for total mortality. In almost every age group, mortality rate was higher at a higher SBP level and higher DBP levels. The difference in absolute mortality rate between the highest and lowest blood pressure groups was larger in older than in younger participants. The results did not change even when we excluded the participants who did not take antihypertensive medication.
|
Table 2 shows the multivariate-adjusted mortality ratio for all-cause mortality for each 10-mm Hg blood pressure increase. In all of the age groups and categories, the hazard ratios were statistically significant. The hazard ratio was larger in the younger group than in the older group. The effect of modifying age was apparent when we included the interaction terms of age and blood pressure into the model (men: P<0.01 in both SBP and DBP; women: P=0.02 in SBP and P=0.05 in DBP). Although we excluded body mass index or alcohol drinking from confounding factors, our result did not changed largely.
|
Table 3 shows the multivariate-adjusted hazard ratio for all-cause mortality according to JNC-7 criteria. Although they fluctuated in some categories to some extent, hazard ratio gradually increased in all of the sex and age categories. With some exceptions, the hazard ratio in groups with hypertension was consistently higher for almost every age range in both men and women.
|
|
The population-attributable fraction for each age range was similar in men and women. In men, except for age 80 to 89 years, 10.5%, the population-attributable fraction was 20% to 30%. Except for the lowest population-attributable fraction in women aged 40 to 49 and aged 80 to 89 years, the population-attributable fraction was 10% to 20%. For the overall population, the population-attributable fraction of nonnormal blood pressure was 22.7% in men and 17.9% in women. The population-attributable fraction became smaller when the reference was the combination of normal and prehypertension groups as compared with the normal group alone.
|
Discussion |
|---|
|
TopAbstractIntroductionStudy Participants and MethodsResultsDiscussionAppendixReferences |
|---|
On the basis of a meta-analysis of individual data from 176 389 Japanese participants, we confirmed that high blood pressure affects total mortality in all age categories. We found that there was an apparent effect modification by age and blood pressure and that hazard ratio was higher in younger than in older groups. We also examined the impact of primary prevention of high blood pressure on total mortality by calculating the population-attributable fraction and found that it was considerable in both men and women.
Blood pressure values are known to be relatively higher in Japan than other developed countries. This situation still exists although values have dropped dramatically.25 One third of Japanese men and women died from cardiovascular diseases,26 and Japan has one of the highest stroke mortalities among developed countries.27 Thus, the contribution of high blood pressure to all-cause mortality should be higher in Japan than other countries. Thus, it should be important to know how many all-cause deaths are attributable to high blood pressure in different age groups. Our huge data set from 180 000 participants can be used to show the health consequence of high blood pressure in Japan. Each of our cohort studies was well administered and provided reliable data sets. The cohort studies were conducted all over Japan, confirming that our results are applicable to the general population in Japan.
Prospective studies collaboration showed that the relation of blood pressure to cardiovascular disease mortality was stronger in younger than in older subjects.2 Recent studies in Japan showed that risk (all causes and cardiovascular disease) of high blood pressure in the Japanese population has increased.21,28,29 The increasing trends in age-specific28 and age-adjusted29 hazard ratios were observed in nationwide cohort studies, and sex- and age- (aged 40 to 59 years and aged 60 to 79 years) specific hazard ratios showed increasing trends in the large-scale cohort study of the Ibaraki prefecture.21 These studies raised the possibility that the association of blood pressure with all-cause mortality was stronger in the younger than in the older Japanese population. A data set from a huge population after fine age stratification could be used to prove findings that were never shown in Asian (Japanese) populations. We found significant interaction between age and blood pressure for all-cause mortality. This suggests that aggressive primary prevention of high blood pressure by all-cause mortality reduction benefits younger more than older people, although the absolute level of all-cause mortality was lower in younger than in older participants. As a consequence, absolute risk reduction should be lower in younger than in older people. Conversely, although the hazard ratio of high blood pressure was relatively lower in the elderly, absolute risk reduction should be higher in older than in younger people. Thus, the risk of hypertension should be age dependent. Because risk reduction is smaller in younger people, the high-risk approach might not be cost-effective. Therefore, the population approach is more suitable in younger than in older individuals. Conversely, for older individuals, the high-risk approach might be more beneficial.
The impact of lower blood pressure on the total mortality was determined by the population-attributable fraction. The study by Sairenchi et al21 showed that the population-attributable fraction is 10% in men and 3% in women. In our study, the EPOCH-JAPAN estimated an age-specific population-attributable fraction for the contribution of blood pressure. In younger age ranges, a large number of deaths can be avoided by lowering blood pressure. This impact of blood pressure decreased with increasing age. When the prehypertension group was used for reference, the population-attributable fraction was 10% in each group. This proportion showed that lowering blood pressure is still effective even when all of the participants have prehypertension, which is a level of blood pressure that is achievable in practice.
We confirmed that all-cause mortality is consistently higher in individuals with prehypertension than in individuals with normal blood pressure. Thus, recommending lifestyle modification for them might reduce their all-cause mortality. However, because the absolute risk difference between normal blood pressure and prehypertension was not large, and population-attributable fraction was small, antihypertensive medication for them might not be recommended. Thus, we considered that the JNC-7 recommendation to modify lifestyle is the appropriate measure for prehypertension participants.
Hypertension is a leading cause of cardiovascular disease mortality and, thus, is a main contributor to total mortality. One of the advantages of selecting total mortality as an end point is that there is no misclassification issue in all-cause deaths compared with disease-specific ones. Although the interpretation of hypertension effect on total mortality was not intuitive, our examination of total mortality provided substantial information for public health purposes.
There were limitations in this study. First, the pooled data of most of the cohort studies were from baseline surveys performed during community health examinations. Participants in the cohort study volunteered to receive their health examinations, and for that reason their characteristics might be somewhat different from those of nonparticipants. This would influence the absolute measure of effect (mortality rate) and might underestimate the risk. However, these differences have little effect on relative measures of effect (such as hazard ratio). Thus, we considered that comparing hazard ratios between age groups or population-attributable fractions might be largely unaffected. Second, we did not adjust for diabetes in this study. Because diabetes is an obesity-related risk factor for hypertension, we might have overestimated the risk posed by hypertension, per se. However, because prevalence of diabetes was very low during the baseline period in Japan, we believe that not adjusting for diabetes should have no substantial effect on our result.
In conclusion, high blood pressure raised the risk of total mortality, and this increase was higher relatively, but not absolutely, in younger than in older individuals. A relatively large amount of the population-attributable fraction was observed in the younger age group. Blood pressure management is an important preventive measure for the Japanese population regardless of age.
Perspectives
The present study showed the relation between blood pressure and total mortality in the Japanese population in detail. The results showed that apparent relation between blood pressure and total mortality was present not only in the elderly but also in a younger age group for both men and women. The result encourages us that blood pressure management was an important preventive measure for Japanese participants regardless of age. Furthermore, the people with prehypertension showed high hazard ratios in most age groups. This detailed information would provide effective public health policy and clinical practice not only in the Japanese but also in the Asian population.
|
Appendix |
|---|
|
TopAbstractIntroductionStudy Participants and MethodsResultsDiscussionAppendixReferences |
|---|
Evidence for Cardiovascular Prevention From Observational Cohorts in Japan Research Group is composed of the following individuals: chairperson: Hirotsugu Ueshima (Shiga University of Medical Science); writing committee: Yoshitaka Murakami, Atsushi Hozawa, Tomonori Okamura, and Hirotsugu Ueshima; statistical analysis: Yoshitaka Murakami; secretariat: Yoshitaka Murakami; and the executive committee: Hirotsugu Ueshima (Shiga University of Medical Science), Yutaka Imai (Tohoku University Graduate School of Medicine), Hiroyasu Iso (Osaka University Graduate School of Medicine), Yutaka Kiyohara (Kyushu University Graduate School of Medicine), Kazunori Kodama (Radiation Effects Research Foundation), Hideaki Nakagawa (Kanazawa Medical University), Takeo Nakayama (Kyoto University School of Public Health), Tomonori Okamura (National Cardiovascular Center), Akira Okayama (Japan Anti-Tuberculosis Association), Shigeyuki Saitoh (Sapporo Medical University), Akiko Tamakoshi (National Center for Geriatrics and Gerontology), Ichiro Tsuji (Tohoku University Graduate School of Medicine), and Yoko Izumi (Ibaraki Prefecture).
|
Acknowledgments |
|---|
Sources of Funding
This research was supported by Grant-in-Aid from the Ministry of Health, Labor, and Welfare, Health and Labor Sciences research grants, Japan (Research on Health Services: H17-Kenkou-007, Comprehensive Research on Cardiovascular Disease and Life-Related Disease: H18-Junkankitou[Seishuu]-Ippan-012 and Comprehensive Research on Cardiovascular Disease and Life-Related Disease: H19-Junkankitou[Seishuu]-Ippan-012).
Disclosures
None.
Received October 5, 2007; first decision October 19, 2007; accepted April 9, 2008.
|
References |
|---|
|
TopAbstractIntroductionStudy Participants and MethodsResultsDiscussionAppendixReferences |
|---|
1. Marmot M, Elliott P, eds. Coronary Heart Disease Epidemiology. 2nd ed. New York, NY: Oxford University Press; 2005.
2. Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 perspective studies. Lancet. 2002; 360: 1903–1913.[CrossRef][Medline] [Order article via Infotrieve]
3. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL Jr, Jones DW, Materson BJ, Oparil S, Wright JT Jr, Roccella EJ; National Heart, Lung, and Blood Institute Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension. 2003; 42: 1206–1252.
4. Woodward M. Epidemiology–Study Design and Data Analysis. 2nd ed. Boca Raton, FL: Chapman & Hall/CRC; 2002.
5. MacMahon S, Peto R, Cutler J, Collins R, Sorlie P, Neaton J, Abbott R, Godwin J, Dyer A, Stamler J. Blood pressure, stroke, and coronary heart disease. Part 1, prolonged difference in blood pressure: prospective observational studies corrected for the regression dilution bias. Lancet. 1990; 335: 765–774.[CrossRef][Medline] [Order article via Infotrieve]
6. Asia Pacific Cohort Studies Collaboration. Blood pressure and cardiovascular disease in the Asia Pacific region. J Hypertension. 2003; 21: 707–716.[CrossRef][Medline] [Order article via Infotrieve]
7. Fujiwara T, Saitoh S, Takagi S, Ohnishi H, Ohata J, Takeuchi H, Isobe T, Chiba Y, Katoh N, Akasaka H, Shimamoto K. Prevalence of asymptomatic arteriosclerosis obliterans and its relationship with risk factors in inhabitants of rural communities in Japan: Tanno-Sobetsu study. Atherosclerosis. 2004; 177: 83–88.[Medline] [Order article via Infotrieve]
8. Ohnishi H, Saitoh S, Takagi S, Katoh N, Chiba Y, Akasaka H, Nakamura Y, Shimamoto K. Incidence of type 2 diabetes in individuals with central obesity in a rural Japanese population: the Tanno and Sobetsu study. Diabetes Care. 2006; 29: 1128–1129.
9. Kuriyama S, Shimazu T, Ohmori K, Kikuchi N, Nakaya N, Nishino Y, Tsubono Y, Tsuji I. Green tea consumption and mortality due to cardiovascular disease, cancer, and all causes in Japan: the Ohsaki study. JAMA. 2006; 296: 1255–1265.
10. Ohkubo T, Kikuya M, Metoki H, Asayama K, Obara T, Hashimoto J, Totsune K, Hoshi H, Satoh H, Imai Y. Prognosis of "masked hypertension" and "white-coat hypertension" detected by 24-hour ambulatory blood pressure monitoring: 10-year follow-up from the Ohasama study. J Am Coll Cardiol. 2005; 46: 508–515.
11. Soyama Y, Miura K, Morikawa Y, Nishijo M, Nakanishi Y, Naruse Y, Kagamimori S, Nakagawa H. High-density lipoprotein cholesterol and the risk of stroke in Japanese men and women: the Oyabe Study. Stroke. 2003; 34: 863–868.
12. Yoshita K, Miura K, Morikawa Y, Ishizaki M, Kido T, Naruse Y, Soyama Y, Suwazono Y, Nogawa K, Nakagawa H. Relationship of alcohol consumption to 7 year blood pressure change in Japanese men. J Hypertens. 2005; 23: 1485–1490.[Medline] [Order article via Infotrieve]
13. Nakamura K, Okamura T, Kanda H, Hayakawa T, Okayama A, Ueshima H; Health Promotion Research Committee of the Shiga National Health Insurance Organizations. Medical costs of patients with hypertension and/or diabetes: a 10-year follow-up study of National Health Insurance in Shiga, Japan. J Hypertens. 2006; 24: 2305–2309.[Medline] [Order article via Infotrieve]
14. Mannami T, Konishi M, Baba S, Nishi N, Terao A. Prevalence of asymptomatic carotid atherosclerotic lesions detected by high-resolution ultrasonography and its relation to cardiovascular risk factors in the general population of a Japanese city: the Suita Study. Stroke. 1997; 28: 518–525.
15. Mannami T, Baba S, Ogata J. Strong and significant relationships between aggregation of major coronary risk factors and the acceleration of carotid atherosclerosis in the general population of a Japanese city: the Suita Study. Arch Intern Med. 2000; 160: 2297–2303.
16. Nakanishi S, Yamada N, Hattori N, Suzuki G. Relationship between HbA1c and mortality in a Japanese population. Diabetologia. 2005; 48: 230–234.[CrossRef][Medline] [Order article via Infotrieve]
17. Ohmura T, Ueda K, Kiyohara Y, Kato I, Iwamoto H, Nakayama K, Nomiyama K, Ohmori S, Yoshitake T, Shinkawa A, Hasuo Y, Fujishima M. Prevalence of type 2 (non-insulin-dependent) diabetes mellitus and impaired glucose tolerance in the Japanese general population: the Hisayama Study. Diabetologia. 1993; 36: 1198–1203.[CrossRef][Medline] [Order article via Infotrieve]
18. Iso H, Date C, Yamamoto A, Toyoshima H, Tanabe N, Kikuchi S, Kondo T, Watanabe Y, Wada Y, Ishibashi T, Suzuki H, Koizumi A, Inaba Y, Tamakoshi A, Ohno Y, and J Am Coll Cardiol Study Group. Perceived mental stress and mortality from cardiovascular disease among Japanese men and women. The Japan Collaborative Cohort Study for Evaluation of Cancer Risk Sponsored by Monbusho (J Am Coll Cardiol Study). Circulation. 2002; 106: 1229–1236.
19. Ueshima H, Choudhury SR, Okayama A, Hayakawa T, Kita Y, Kadowaki T, Okamura T, Minowa M, Iimura O. Cigarette smoking as a risk factor for stroke death in Japan NIPPON DATA80. Stroke. 2004; 35: 1836–1841.
20. 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: 143–150.[CrossRef][Medline] [Order article via Infotrieve]
21. Sairenchi T, Iso H, Irie F, Fukasawa N, Yamagishi K, Kanashiki M, Saito Y, Ota H, Nose T. Age-specific relationship between blood pressure and the risk of total and cardiovascular mortality in Japanese men and women. Hypertens Res. 2005; 28: 901–909.[CrossRef][Medline] [Order article via Infotrieve]
22. Irie F, Iso H, Sairenchi T, Fukasawa N, Yamagishi K, Ikehara S, Kanashiki M, Saito Y, Ota H, Nose T. The relationships of proteinuria, serum creatinine, glomerular filtration rate with cardiovascular disease mortality in Japanese general population. Kidney Int. 2006; 69: 1264–1271.[CrossRef][Medline] [Order article via Infotrieve]
23. Frost C, Thompson SG. Correcting for regression dilution bias: comparison of methods for a single predictor variables. J R Statist Soc A. 2000; 163: 173–189.[CrossRef]
24. Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic Research–Principles and Quantitative Methods. New York, NY: Van Nostrand Reinhold; 1982.
25. Ueshima H, Zhang XH, Choudhury SR. Epidemiology of hypertension in China and Japan. J Hum Hypertens. 2000; 14: 765–769.[CrossRef][Medline] [Order article via Infotrieve]
26. Statistics and Information, Department Minister’s Secretariat Ministry of Health, Labour and Welfare. Statistical abstracts on health and welfare in Japan 2005. Tokyo, Japan: Japan Health and Welfare Statistics Association; 2005.
27. Ueshima H. Changes in dietary habits, cardiovascular risk factors and mortality in Japan. Acta Cardiol. 1990; 45: 311–327.[Medline] [Order article via Infotrieve]
28. NIPPON DATA80 Research Group. Impact of elevated blood pressure on mortality from all causes, cardiovascular diseases, heart disease and stroke among Japanese: 14 year follow-up of randomly selected population from Japanese–Nippon data 80. J Hum Hypertens. 2003; 17: 851–857.[CrossRef][Medline] [Order article via Infotrieve]
29. Okayama A, Kadowaki T, Okamura T, Hayaakawa T, Ueshima H, on behalf of the NIPPON DATA80 Research Group. Age-specific effects of systolic and diastolic blood pressures on mortality due to cardiovascular diseases among Japanese men (NIPPON DATA 80). J Hypertens. 2006: 24; 459–462.[Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  Y. Kokubo, K. Kamide, T. Okamura, M. Watanabe, A. Higashiyama, K. Kawanishi, A. Okayama, and Y. Kawano Impact of High-Normal Blood Pressure on the Risk of Cardiovascular Disease in a Japanese Urban Cohort: The Suita Study Hypertension, October 1, 2008; 52(4): 652 - 659. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||