Hypertension is a major risk factor for future atherosclerotic changes in the Japanese population

Abstract

Background

Great differences in age-standardized mortality rates by cardiovascular disease exist among countries. We prospectively assessed determinants of future cardiovascular changes in a Japanese cohort.

Methods

In 1996, 1011 men and 1153 women from a Japanese community participated in a study on cardiovascular risk factors at a local health centre. Of these, the 896 subjects who visited the centre both in 1996 and 2001 were selected for the analysis. The presence of cardiovascular changes was defined as the appearance of one or more of the following in five years: positive electrocardiographic findings, intima–media thickness of the carotid artery ≥0.8 mm and retinal vascular changes ≥Keith–Wegener–Barker classification stage I.

Results

Of the 607 subjects who had no history of cardiovascular disease in 1996, 421 showed changes in 2001. Both the age-adjusted and multivariate models showed that the risk of the cardiovascular changes increased as systolic blood pressure (SBP) increased to ≥135 mmHg (multivariate odds ratio = 1.739, 95% confidence interval = 1.076–2.810, P < 0.05) compared with those with an SBP of 110–134 mmHg. When we made the analyses only for laboratory test results by excluding SBP, body mass index, alcohol intake and current smoking from the regression model, high-density lipoprotein cholesterol and fasting plasma glucose were significant variables.

Conclusion

The risk of future cardiovascular changes is significantly greater with higher SBP in the Japanese population.

Introduction

The existing population-based studies have shown an association between cardiovascular diseases and characteristics such as dyslipidaemia, diabetes mellitus, hypertension, high body mass index (BMI), increased alcohol intake and smoking.^1,2 However, World Health Organization (WHO) Statistics from 2006 show a great difference of age-standardized mortality rate by cardiovascular disease (per 100,000 population) among countries.³ The Japanese population has a significantly lower mortality rate of 106 compared with the national average of other countries, such as the United States of America (mortality rate of 188) and the United Kingdom (mortality rate of 182). Studies based on ethnicity can add new information to the existing knowledge on the pathogenesis and prevention of cardiovascular disease.

We previously reported the characteristics relating to the risk of cardiovascular disease based on a cross-sectional study of a Japanese population from the Niigata study, an ongoing epidemiological study for the prevention of cardiovascular disease.⁴ Briefly, intima–media thickness of the carotid artery (IMT) was independently correlated with total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), systolic blood pressure (SBP) and BMI in men and with TC, triglyceride (TG), LDL-C, HDL-C, haemoglobin type A_1c (HbA_1c), SBP, diastolic blood pressure (DBP) and BMI in women.

In the present study report, we prospectively assessed determinants of future cardiovascular changes in a sub-cohort of 896 subjects from the Niigata study. Measurements of the lipid profile, fasting plasma glucose, blood pressure, alcohol intake and smoking habit, which had been obtained in 1996, were compared with the findings of electrocardiography (ECG), IMT and retinal vascular changes in 2001.

Methods

Study population

The Niigata study is a population-based follow-up trial with the purpose of investigating characteristics relating to the risk of cardiovascular disease in the Japanese population. Full details of the study have been reported previously.⁴ Briefly, in 1996, 2164 subjects (1011 men and 1153 women; 23–89 y of age) were recruited from a Japanese community. The exclusion criteria at the initial entry were as follows: current and previous medication, pregnant women, history of diabetes mellitus, history of cardiovascular disease and fasting plasma glucose (FPG) concentration >0.69 mmol/L (125 mg/dL). All data were collected, analysed and evaluated at a local health centre.

For the present study, the 896 subjects (394 men and 502 women) who visited the centre both in 1996 and in 2001 were selected.

Procedures

The clinical measurements taken in the present study were as follows: IMT measurements, ECG and retinal vascular changes; concentrations of fasting plasma insulin, FPG and HbA_1c; serum concentrations of TC, TG, HDL-C and LDL-C; SBP, DBP and BMI; and habits regarding drinking alcohol and smoking cigarettes.

The assessment of IMT was performed using high-resolution B-mode ultrasonography according to Salonen et al. ⁵ The extracranial carotid arteries were scanned from the lowest portion visible in the supraclavicular fossa to the carotid bifurcation, with each subject in the supine position. Measurements of IMT were based on the most severely affected site in either the right or left side of the neck. The affected sites included homogeneous thickening and localized plaque with or without calcification or ulceration or both.

An FCP-4720 ECG (Fukuda Denshi, Tokyo, Japan) was used to record the 12-lead ECGs. ECG findings were considered as positive if there were abnormalities associated with coronary heart disease, such as abnormal Q wave, ST segment depression, ST segment elevation and/or T wave change according to the Minnesota code.

Retinal examination was performed in one randomly selected eye of each subject using a standard protocol. The fundus was photographed using an FF-1000 fundus camera with a 45° field (Fukuda Denshi). Every photograph was evaluated by an experienced doctor (MO) who was unaware of the clinical data of each subject. The retinal vascular changes sought were light reflex change of arterioles, nipping of the venule at crossings, narrowing of arterioles, haemorrhages, exudates, cotton-wool spots and oedema of the discs. Retinal lesions were classified according to the Keith–Wagener–Barker classification as follows: stage 0, normal retinal finding; stage I, slightly detectable light reflex changes in retinal vessels; stage II, changes in retinal vessels more marked than in stage I; stage III, retinal vessel changes plus haemorrhage and/or cotton–wool spots and stage IV, stage III plus oedema of the discs.⁶

After an overnight fast, blood was drawn from an antecubital vein with each subject seated. All analyses were performed within 3 h after the blood was drawn. The plasma insulin concentration was determined using an enzyme immunoassay (AxSYM Insulin, Dainabot, Tokyo, Japan) Blood glucose was measured using the glucose oxidase–oxygen electrode method on a Glucoreader-C apparatus (Shino-Test, Tokyo, Japan). HbA_1c was measured using the high-performance liquid chromatography method on a model HLC-723GHb chromatography column (Tosoh, Tokyo, Japan). Serum TC, TG and HDL-C were measured enzymatically on a model 736 automatic analyser (Hitachi, Tokyo, Japan). The amount of LDL-C was calculated from the amounts of TC, HDL-C and TG using Friedewald's formula.

Sitting SBP and DBP were measured at the upper arm using a sphygmomanometer. The subjects refrained from physical activity and eating for at least 30 min before measurement. SBP was recorded as the appearance of the Korotkoff sound, and DBP as the disappearance of sound. BMI, calculated as weight (kg) divided by the square of height (m) (kg/m²), was used as an index of overall adiposity. A questionnaire on drinking alcohol was categorized into four categories: never/hardly ever, monthly, weekly or daily. Smoking cigarettes was a dichotomous (yes or no) variable.

Statistical analysis

The presence of cardiovascular changes was defined as the appearance of one or more of the following in five years: positive ECG findings, IMT≥0.8 mm and retinal vascular changes ≥Keith–Wegener–Barker classification stage I.

We used logistic regression modelling to estimate the odds ratio (OR) of variables for the risk of cardiovascular changes in all multivariate analyses. Subjects who were diagnosed as having cardiovascular changes with the data in 2001 were assigned 1 as the dependent variable of the logistic regression analysis; otherwise, 0 was assigned. To analyse the effect of potential confounders on the ORs, we first adjusted multivariate models for age only and next by including sex (male or female), LDL-C, HDL-C, TG, FPG, HbA_1c, fasting insulin, SBP, BMI, alcohol intake (never, monthly, weekly or daily) and current smoking (yes or no). In the former model, age and the parameter concerned were included, while in the latter model, age, sex and all other parameters were included. Selection of a significant subset of independent risk factors was achieved by the backward stepwise elimination method. We also performed analyses stratified by SBP (<110 mmHg, 110–134 mmHg and ≥135 mmHg) and BMI (<22, 22–24 and ≥25 kg/m²).

All statistical analyses were conducted using SPSS software, version 11.0 J for Windows (SPSS Japan Inc, Tokyo, Japan). Values express mean±standard deviation and P < 0.05 was considered statistically significant. For all ORs, we calculated the 95% confidence interval (CI).

The study protocol was approved by the Committee on Human Research, Niigata University School of Medicine, Japan, and informed consent was obtained from all participants.

Results

Of the 896 participants, no cardiovascular changes were found in 607 subjects (251 men and 356 women) in 1996. The baseline characteristics of the 607 subjects both in 1996 and 2001 are shown in Table 1. During the five years of follow-up between 1996 and 2001, the serum concentration of LDL-C increased slightly and the number of smokers decreased. The other variables were similar across the years. Of the subjects, 421 (178 (71%) men and 243 (68%) women) showed cardiovascular changes, and 186 did not, in 2001. Univariate analyses showed that higher FPG and higher SBP were associated with the risk of cardiovascular changes in the following five years (Table 2). No significant association was found for the other variables.

Table 1

Baseline characteristics of the 607 subjects in 1996 and 2001

Characteristic	Year examined
Characteristic	1996	2001
Age (y)	57.1 (9.7)*
Sex (male/female) (no.)	251/356
LDL-C (mmol/L)	3.07 (0.75)	3.27 (0.78)
HDL-C (mmol/L)	1.46 (0.38)	1.53 (0.37)
Triglyceride (mmol/L)	1.18 (0.88)	1.21 (0.82)
Fasting plasma glucose (mmol/L)	5.29 (0.71)	5.32 (0.84)
HbA_1c (%)	5.1 (0.6)	4.9 (0.6)
Fasting insulin (pmol/L)	32.2 (75.9)	34.9 (22.2)
Systolic blood pressure (mmHg)	126 (18)	127 (17)
Body mass index (kg/m²)	22.8 (2.9)	22.8 (3.0)
Alcohol intake (no.)
Never/hardly ever	288	246
Monthly	77	62
Weekly	70	62
Daily	172	181
Current smoking (no.)	125	101

HbA_1c, haemoglobin type A_1c

*Mean (standard deviation)

Table 2

Baseline characteristics in 1996 of subjects with and without the cardiovascular change

Characteristic	Cardiovascular change		P value
Characteristic	Without	With	P value
Subjects (no.)	186	421
Age (y)	52.4 (10.7)	59.2 (8.4)	0.000
Sex (male/female) (no.)	73/113	178/243	0.484
LDL-C (mmol/L)	2.96 (0.78)	3.12 (0.74)	0.183
HDL-C (mmol/L)	1.53 (0.35)	1.43 (0.40)	0.107
Triglyceride (mmol/L)	1.02 (0.67)	1.25 (0.95)	0.536
Fasting plasma glucose (mmol/L)	5.13 (0.51)	5.36 (0.77)	0.025
HbA_1c (%)	5.0 (0.7)	5.2 (0.5)	0.613
Fasting insulin (pmol/L)	28.8 (15.6)	33.6 (90.6)	0.311
Systolic blood pressure (mmHg)	119.6 (15.6)	129.4 (18.6)	0.001
Body mass index (kg/m²)	22.1 (2.6)	23.2 (2.9)	0.100
Alcohol intake (%)			0.360
Never/hardly ever	45.2	48.5
Monthly	12.9	12.6
Weekly	10.8	11.9
Daily	31.2	27.1
Current smoking (%)	24.7	18.8	0.086

HbA_1c, forms of glycosylated haemoglobin

Table 3 provides two logistic regression models for predicting the risk of cardiovascular changes with 10 variables. The age-adjusted model showed that HDL-C, TG, FPG, HbA_1c, SBP and BMI were significant variables. A multivariate-adjusted model showed that the significance of ORs of HDL-C, TG, FPG, HbA_1c and BMI were attenuated when adjusted for each other, while SBP was the only significant variable. In either models, LDL-C, fasting insulin, alcohol intake and current smoking were not significant variables.

Table 3

Odds ratio (OR) of overall abnormality: Niigata follow-up study, 1996–2001

Characteristic*	Age-adjusted model^†		Multivariate model^‡
Characteristic*	OR (95% CI)	P value	OR (95% CI)	P value
LDL-C	1.003 (0.996–1.009)	0.428	1.001 (0.994–1.009)	0.688
HDL-C	0.980 (0.968–0.992)	0.002	0.989 (0.974–1.004)	0.137
Triglyceride	1.004 (1.001–1.008)	0.009	1.001 (0.998–1.004)	0.554
Fasting plasma glucose	1.026 (1.007–1.045)	0.007	1.010 (0.990–1.030)	0.317
HbA_1c	1.482 (1.020–2.154)	0.039	1.333 (0.878–2.024)	0.178
Fasting insulin	1.021 (0.961–1.085)	0.504	1.000 (0.984–1.016)	0.989
Systolic blood pressure	1.025 (1.014–1.036)	0.000	1.019 (1.006–1.031)	0.003
Body mass index	1.150 (1.074–1.230)	0.000	1.069 (0.989–1.155)	0.092
Alcohol intake	1.028 (0.891–1.186)	0.707	0.979 (0.799–1.200)	0.841
Current smoking	0.898 (0.577–1.396)	0.632	0.740 (0.432–1.266)	0.271

CI, confidence interval; HbA_1c, haemoglobin type A_1c

*Examined in 1996

^†Data were adjusted for age

^‡Variables in the model were age, sex, LDL-C, HDL-C, triglyceride, fasting plasma glucose, haemoglobin type A_1c, fasting insulin, systolic blood pressure, body mass index, alcohol intake (never, monthly, weekly, daily) and current smoking

Analysis by the backward stepwise elimination method – in which the criterion for determining variables to be removed from the model was the likelihood ratio – showed that SBP and BMI were significant variables that could independently predict the risk of cardiovascular changes other than age (Table 4). The subset of possible predictors also comprised HDL-C and HbA_1c in combination with SBP and BMI. In order to look at the effect of excluding SBP, BMI, alcohol intake and current smoking from the regression model, we made another analysis by the backward stepwise elimination method and found that HDL-C and FPG were significant variables that could independently predict the risk of cardiovascular changes other than age (Table 5).

Table 4

Odds ratio (OR) of the variables selected by the Backward Stepwise Elimination Method

Characteristic*	Multivariate OR (95% CI)^†	P value
Age	1.066 (1.044–1.088)	0.000
HDL-C	0.987 (0.973–1.000)	0.051
HbA_1c	1.400 (0.949–2.065)	0.090
Systolic blood pressure	1.020 (1.009–1.032)	0.001
BMI	1.080 (1.003–1.164)	0.043

BMI, body mass index; CI, confidence interval; HbA_1c, haemoglobin type A_1c

*Examined in 1996

^† Variables evaluated for the elimination were age, sex, LDL-C, HDL-C, triglyceride, fasting blood sugar, HbA_1c, systolic blood pressure, BMI, alcohol intake and current smoking

Table 5

Odds ratio (OR) of the variables selected by the Backward Stepwise Elimination Method (excluding blood pressure and body mass index)

Characteristic*	Multivariate OR (95% CI)^†	P value
Age	1.075 (1.055–1.097)	0.000
HDL-C	0.981 (0.969–0.994)	0.003
Fasting plasma glucose	1.024 (1.005–1.043)	0.012

CI, confidence interval

*Examined in 1996

^† Variables evaluated were age, sex, LDL-C, HDL-C, triglyceride, fasting blood sugar, haemoglobin type A_1c, alcohol intake and current smoking

The relationships between categories of SBP and the risk of cardiovascular changes are summarized in Table 6. Both the age-adjusted and multivariate models showed that the risk of the cardiovascular changes increased as SBP increased to ≥135 mmHg (multivariate OR = 1.739, 95% CI = 1.076–2.810, P < 0.05), while there was no significant difference in the risk for an SBP<110 mmHg compared with an SBP of 110–134 mmHg. For BMI, age-adjusted models showed that the risk of cardiovascular changes increased as BMI increased to ≥25 kg/m² (OR = 1.911, 95% CI = 1.093–3.340, P < 0.001), while the multivariate OR for an increase in BMI to ≥25 kg/m² and that for a decrease to <22 kg/m² were not significant compared with a BMI of 22–24 kg/m².

Table 6

Odds ratio of overall abnormality according to blood pressure in 1996

	Blood pressure (mmHg)
	<110	110–134	≥135
Number of participants	121	281	205
Age-adjusted odds (95% CI)	0.734 (0.462–1.165)	1	2.108 (1.336–3.29)
Multivariate odds* (95% CI)	0.877 (0.535–1.438)	1	1.739 (1.076–2.810)

CI, confidence interval

*Data were adjusted for age, sex, LDL-C, HDL-C, triglyceride, fasting plasma glucose, haemoglobin type A_1c, fasting insulin, BMI, alcohol intake (never, monthly, weekly, daily) and current smoking

In the analyses, we did not make any normalization of data from the theoretical point of view. However, from a practical point of view, outlying or peripherally located data points in the skewed distribution may affect the analytical results in some situation. Therefore, we transformed TG and FPG logarithmically and repeated the analyses. We found that there was no significant difference between the results obtained from the original data and those from the transformed data.

Discussion

We prospectively assessed the determinants of future cardiovascular changes in a Japanese cohort of 607 subjects with no history of cardiovascular disease. The risk of cardiovascular changes significantly increased with higher SBP and higher BMI. Furthermore, higher HbA_1c and lower HDL-C were possible risk factors. The other variables of sex, LDL-C, TG, FPG, fasting insulin, alcohol intake and smoking did not significantly correlate with the cardiovascular changes.

During the five-year follow-up, 421 among 607 subjects developed cardiovascular changes. This prevalence seems to be very high compared with those of existing studies, of which the primary objective was to assess major cardiac events as cardiovascular deaths, non-fatal myocardial infarction, non-fatal stroke, etc. In our study, the endpoint is surrogate and can occur more often than those major events.

When we made analyses only for laboratory test results by excluding SBP, BMI, alcohol intake and current smoking from the regression model, HDL-C and FPG were significant variables that could independently predict the risk of cardiovascular changes other than age. Although the laboratory test results are known as important contributors of metabolic syndrome, these variables were strongly correlated with SBP and/or BMI, which in turn were the most significant independent variables in the Japanese population. This is the most significant finding of our study.

As the WHO statistics demonstrate, the mortality rate by cardiovascular disease is significantly different from country to country. Disparity in health outcomes across countries or ethnicities is a complex phenomenon with many factors, such as race, culture and socioeconomic condition. A study of two Mexican-American communities, for example, showed that residents in Mexico City consumed more carbohydrates and less fat and had higher TG and lower HDL-C compared with their counterparts living in San Antonio, TX, USA.⁷ However, in a related paper, it was reported that the differences in carotid IMT between the two Mexican communities were inconsistent.⁸ Two studies showed that a high-carbohydrate diet and a resultant dyslipidaemia might be less atherogenic for the Mexican population. In Australia, a steady decline in cardiovascular disease mortality in the general Australian population has been documented, while the death rate attributed to the disease among adult Aboriginal people – i.e. native Australians – is more than twice that of the non-Aboriginal population and continues to rise. One study showed that an excess of obesity-related cardiovascular risk factors such as high FPG and dyslipidaemia, as well as smoking habit, contributed to the risk in the urban Australian Aboriginal population.⁹ These findings are completely consistent with the traditional knowledge that a lifestyle consisting of a high-fat diet, lower physical activity and smoking is the major factor that raises the mortality rate by cardiovascular disease.

In contrast, in Asian countries, the cultures, lifestyles and genetic backgrounds are quite different from those of Western countries. The effects of westernized lifestyles on a Chinese population were studied by Woo et al,¹⁰ who examined carotid IMT and traditional cardiovascular risk factors in rural Chinese, urban Chinese and Chinese-Australian subjects. They found that carotid IMT increased in both urban Chinese subjects and Chinese-Australian subjects compared with rural Chinese subjects, but the overall risk factor profile was rather better in the urban Chinese subjects, who had higher concentrations of HDL-C and lower SBP compared with rural Chinese subjects. They noted that the pathogenesis in the Chinese populations could not be explained by the concentrations of serum lipoproteins or by blood pressure. Thus, the susceptibility of Chinese populations to the atherogenic effects of the above-mentioned risk factors may be different from other ethnicities. A magnetic resonance imaging-based study of the carotid arteries revealed that there was no significant difference in the measurements, except for plaque size between Chinese subjects in Beijing, China and Caucasian Americans in Seattle, WA, USA.¹¹ The average size of the lesion was significantly larger in the Chinese subjects even after adjustment for confounding variables including statin therapy, and their lesions were more homogeneous than those of Caucasian Americans. This indicates that the lesions of the Chinese subjects were less mature, with a shorter history of atherogenesis, which may be due to rapid westernization and consequent changes in the diet and lifestyle in their community. In another study, Wu et al. ¹² reported substantial differences in the ECG findings between Taiwanese Chinese subjects and US Caucasians, which could not be explained by traditional risk factors for cardiovascular disease. It was found that US Caucasians had a higher prevalence of Q wave abnormalities, while the prevalence of ST segment depression and T wave abnormalities was similar between Taiwanese Chinese male subjects and Caucasian male subjects in the USA. Although there were significant differences in the concentrations of many confounding variables such as lipid profile, blood pressure and obesity, one needs to take into account that each risk and/or cut-off level may differ between the two populations.

Similar studies have been performed for Japanese subjects in Japan and Japanese Americans in Hawaii.^13,14 They found that the lifestyles of Japanese Americans, who have an identical genetic background to that of native Japanese subjects, had been rapidly westernized to a greater extent as follows: the Japanese American subjects consumed a total calorie count 1.6 times higher than the native Japanese subjects; the Japanese Americans had significantly higher concentrations of serum TC, LDL-C, TG and higher BMI than the native Japanese subjects and the death rate of the Japanese Americans due to cardiovascular disease was significantly higher than that of the native Japanese subjects. Nevertheless, the death rate of cardiovascular disease in Japanese Americans is still less than half the rates reported for Caucasian men in the continental USA.

One of the physical characteristics of average Japanese subjects is higher blood pressure with lower body weight than Western individuals. The higher rate of hypertension in the Japanese population appears to be due to the higher intake of salt,^15,16 higher rate of cigarette smoking¹⁷ and lower rate of treatment of hypertension.¹⁸ Salt intake is the most significant factor involved in the pathogenesis of hypertension. Karppanen et al. ¹⁶ reported that, in Finland, a progressive and marked decrease in the average blood pressure has taken place during the past three decades, which is possibly explained by the fall in the average intake of salt in the population. There are great differences among ethnicities in salt intake, which ranges from less than 0.1 g/day to 15 g/day throughout the world. The Japanese population previously took in more than 24 g salt/day and this is now about 11 g/day.¹⁵ This may explain why in the present study SBP was the most significant variable to independently predict future cardiovascular changes.

Do these findings represent only lifestyle changes, or do genetic factors also contribute to the differences in the pathogenesis of hypertension among ethnicities? For example, blood pressure reductions in subjects with a T allele at amino acid 235 of the angiotensinogen gene are more sensitive to lowering dietary sodium intake.¹⁹ Asian populations, including the Japanese, have higher frequencies – around 60–80% – of this allele, while Western populations show lower frequencies, around 35–45%.²⁰ The fasting insulin concentration, which is closely related to hypertension, is another candidate phenotype that is less influenced by lifestyle and represents genetic background. The average fasting insulin concentration of non-diabetic men free of ischaemic heart disease in the Japanese population is about 30 pmol/L (5 μU/mL),^4,21 while that in British men is 72 pmol/L (12 μU/mL),²² in Canada 78 pmol/L (13 μU/mL)²³ and in the USA 72 pmol/L (12 μU/mL).²⁴ The rates of insulin resistance or hyperinsulinaemia vary substantially according to ethnicity.^25,26 For example, the A/T54 polymorphism in the intestinal fatty acid binding protein is a possible candidate for the phenotype. However, the epidemiological data linking this polymorphism with insulin resistance is inconsistent.²⁷ Many studies conducted in Western countries have shown that hyperinsulinaemia is related to the incidence of cardiovascular disease.^23,28,29 However, both the present study and a study for Japanese-American men,¹⁴ each of which was conducted prospectively in non-diabetic subjects, showed that fasting insulin concentration was not a significant variable. Burchfiel et al. ¹⁴ noted that several populations, including the Pima Indians, Hispanics and Japanese, had paradoxically low rates of cardiovascular disease despite relatively high proportions of individuals with glucose intolerance. Although potential confounders have not been well analysed to date, these differences may reflect the existence of genetic variants in hyperinsulinaemia and/or glucose intolerance among ethnicities.

Several limitations of the present study should be discussed. First, we defined future cardiovascular events by ECG findings, IMT measurements and retinal vascular changes, and did not include definitive variables such as angiographic findings. The subjects recruited in the Niigata study were non-diabetic healthy men and women without history of cardiovascular disease. However, carotid echosonography is recognized as a valid non-invasive surrogate for atherosclerosis, and a number of variables have been identified as directly implicated in the increase in IMT.^30,31 In addition, it has been proven that an increased IMT is a predictive marker of future cardiovascular disease³² and cerebrovascular events.³³ Given this information, the use of the combination of IMT with ECG and retinal vascular findings was the logical choice for the present study. Second, recent studies have shown that high-sensitivity C-reactive protein,³⁴ uric acid,³⁵ fibrinogen³⁶ and abdominal obesity³⁷ are also possible risk factors. Data of these variables were not included in the Niigata study, which started in 1996. Third, the major cardiac events and total mortality were not examined in the present study. Furthermore, we did not include any accounts on subjects who withdrew from the study. Since our study has been designed so as to continue the follow-up for more than five years, the subjects who did not visit the centre in 2001 are not necessarily censored cases. Further research should include these data.

Conclusion

By prospectively analysing data of a Japanese cohort with no history of cardiovascular disease, we found that the risk of cardiovascular changes was significantly increased with higher SBP and higher BMI, which was 1.739 times higher among subjects whose SBP was higher than or equal to 135 mmHg compared with those whose SBP were 110–134 mmHg. One of the physical characteristics of the Japanese population is higher blood pressure with lower body weight. A variety of genetic variants, which are different from those in Western ethnicities, may play roles in the pathogenesis of cardiovascular disease in ethnicities such as the Japanese.

DECLARATIONS

Competing interests: All authors declare that there are no conflicts of interest.

Funding: None.

Ethical approval: The study protocol was approved by the Committee on Human Research, Niigata University School of Medicine, Japan, and informed consent was obtained from all participants.

Guarantor: MO.

Contributorship: TA acquired and drafted the manuscript, AT acquired data, SY handled supervision, MM performed statistical analysis and MO conceived and designed the research.

References

Bϕnaa

, Thelle

. Association between blood pressure and serum lipids in a population: the Tromsø study. Circulation 1991;83:1305–14

Greenland

, Knoll

, Stamler

, Major risk factors as antecedents of fatal and nonfatal coronary heart disease events. JAMA 2003;290:891–7

Okada

, Miida

, Hama

, Possible risk factors of carotid artery atherosclerosis in the Japanese population: a primary prevention study in non-diabetic subjects. Intern Med 2000;39:362–8

Salonen

, Salonen

. Progression of carotid atherosclerosis and its determinants: a population-based ultrasonography study. Atherosclerosis 1990;80:33–40

Keith

, Wagener

, Barker

. Some different types of essential hypertension: their course and prognosis. Am J Med Sci 1974;268:336–49

Stern

, Gonzalez

, Hernandez

, Performance of semiquantitative food frequency questionnaires in international comparisons: Mexico City versus San Antonio, Texas. Ann Epidemiol 1993;3:300–7

Wei

, Gonzalez

, Haffner

, Ultrasonographically assessed maximum carotid artery wall thickness in Mexico City residents and Mexican Americans living in San Antonio, Texas: association with diabetes and cardiovascular risk factors. Arterioscler Thromb Vasc Biol 1996;16:1388–92

Bradshaw

, Wilkes

, Thompson

. Determinants of carotid intima–medial thickness in an urban Australian Aboriginal population. Atherosclerosis 2007;192:218–23

10.

Woo

, Chook

, Raitakari

, Westernization of Chinese adults and increased subclinical atherosclerosis. Arterioscler Thromb Vasc Biol 1999;19:2487–93

11.

Saam

, Cai

, An

, Carotid plaque composition differs between ethno-racial groups: an MRI pilot study comparing mainland Chinese and American Caucasian patients. Arterioscler Thromb Vasc Biol 2005;25:611–6

12.

C-C

, Yeh

W-T

, Crow

, Comparison of electrocardiographic findings and associated risk factors between Taiwan Chinese and US white adults. Int J Cardiol 2008;128:224–31

13.

Yano

, Reed

, McGee

. Ten-year incidence of coronary heart disease in the Honolulu heart program: relationship to biologic and lifestyle characteristics. Am J Epidemiol 1984;119:653–66

14.

Burchfiel

, Sharp

, Curb

, Hyperinsulinemia and cardiovascular disease in elderly men: the Honolulu heart study. Arterioscler Thromb Vasc Biol 1998;18:450–7

15.

Kawano

, Ando

, Matsuura

, Report on the working group for dietary salt reduction of the Japanese Society of Hypertension: (1) rationale for salt restriction and salt-restriction target level for the management of hypertension. Hypertens Res 2007;30:879–86

16.

Karppanen

, Mervaala

Sodium intake and hypertension. Prog Cardiovasc Dis 2006;49:59–75

17.

Watanabe

, Yamane

, Fujikawa

, Westernization of lifestyle markedly increases carotid intima–media wall thickness (IMT) in Japanese people. Atherosclerosis 2003;166:67–72

18.

Ueshima

, Okayama

, Saitoh

, Difference in cardiovascular disease risk factors between Japanese in Japan and Japanese-Americans in Hawaii: the INTERLIPID study. J Hum Hypertens 2003;17:631–9

19.

Hunt

, Geleijnse

, Wu

, Enhanced blood pressure response to mild sodium reduction in subjects with the 235T variant of the angiotensinogen gene. Am J Hypertens 1999;12:460–6

20.

Nakase

, Mizuno

, Harada

, Angiotensinogen gene polymorphism as a risk factor for ischemic stroke. J Clin Neurosci 2007;14:943–7

21.

Toyoda

, Fukushima

, Mitsui

, Factors responsible for age-related elevation in fasting plasma glucose: a cross-sectional study in Japanese men. Metabolism 2008;57:299–303

22.

Perry

, Wannamethee

, Whincup

, Serum insulin and incident coronary heart disease in middle-aged British men. Am J Epidemiol 1996;144:224–34

23.

Després

J-P

, Lamarche

, Mauriège

, Hyperinsulinemia as an independent risk factor for ischemic heart disease. New Engl J Med 1996;334:952–7

24.

Donahue

, Bean

, Donahue

, Does insulin resistance unite the separate components of the insulin resistance syndrome?: evidence from the Miami community health study. Arterioscler Thromb Vasc Biol 1997;17:2413–7

25.

Reimann

, Schutte

, Schwarz

PEH

. Insulin resistance – the role of ethnicity: evidence from Caucasian and African cohorts. Horm Metab Res 2007;39:853–7

26.

Godsland

, Johnston

, Chaturvedi

. Mechanisms of disease: lessons from ethnicity in the role of triglyceride metabolism in ischemic heart disease. Nat Clin Pract Endocrinol Metab 2007;3:530–8

27.

Koning

, Merchant

, Hegele

, Association of the FABP2 T54 variant with plasma triglycerides and insulin resistance in a multiethnic population. Clin Chem 2008;54:1742–4

28.

Bonora

, Willeit

, Kiechl

, Relationship between insulin and carotid atherosclerosis in the general population: the Bruneck study. Stroke 1997;28:1147–52

29.

Haffner

, D'Agostino

Jr , Mykkänen

, Proinsulin and insulin concentrations in relation to carotid wall thickness: insulin resistance atherosclerosis study. Stroke 1998;29:1498–503

30.

Joakimsen

, Bϕnaa

, Stensland-Bugge

, Age and sex differences in the distribution and ultrasound morphology of carotid atherosclerosis: the Tromsø study. Arterioscler Thromb Vasc Biol 1999;19:3007–13

31.

Schott

, Wildman

, Brockwell

, Segment-specific effects of cardiovascular risk factors on carotid artery intima–medial thickness in women at midlife. Arterioscler Thromb Vasc Biol 2004;24:1951–6

32.

Staub

, Meyerhans

, Bundi

, Prediction of cardiovascular morbidity and mortality: comparison of the internal carotid artery resistive index with the common carotid artery intima–media thickness. Stroke 2006;37:800–5

33.

Mathiesen

, Bϕnaa

, Joakimsen

. Echolucent plaques are associated with high risk of ischemic cerebrovascular events in carotid stenosis: the Tromsø study. Circulation 2001;103:2171–5

34.

Arima

, Kubo

, Yamamoto

, High-sensitivity C-reactive protein and coronary heart disease in a general population of Japanese: the Hisayama study. Arterioscler Thromb Vasc Biol 2008;28:1385–91

35.

Ishizaka

, Ishizaka

, Toda

, Association between serum uric acid, metabolic syndrome, and carotid atherosclerosis in Japanese individuals. Arterioscler Thromb Vasc Biol 2005;25:1038–44

36.

Aono

, Ohkubo

, Kikuya

, Plasma fibrinogen, ambulatory blood pressure, and silent cerebrovascular lesions: the Ohasama study. Arterioscler Thromb Vasc Biol 2007;27:963–8

37.

Litwin

. Which measures of obesity best predict cardiovascular risk? J Am Coll Cardiol 2008;52:616–9