Search

Introduction: Over the past several decades, Asian countries including Japan have experienced westernization of their lifestyles, leading to changes in prevalence of major risk factors for cardiovascular disease (CVD). The estimation of changed in trends of population attributable fraction (PAF) of major CVD risk factors would be useful for planning of CVD prevention strategies. Hypothesis: We assessed the hypothesis that PAFs of major CVD risk factors have changed among a Japanese population over the past several decades. Methods: We conducted four 10-year cohorts from 1975 to 2014: the baseline of the first cohort as 1975-1977 (n=4,415), that of the second as 1985-1986 (n=7,155), that of the third as 1995-1997 (n=6,892), and that of the fourth as 2005-2007 (n=5,067), consisting of Japanese men and women aged 40-79 years, initially free of CVD, and with valid information on major CVD risk factors (hypertension, diabetes mellitus, hypercholesterolemia, obesity, smoking, alcohol drinking, and atrial fibrilization). Hazard ratios (HRs) and 95 % confidence intervals (95 % CIs) of CVD (incident stroke and ischemic heart disease) risk factors were calculated by Cox proportional hazard model and their PAFs (95 %CIs) were estimated. Results: During the median follow-up of 8.1 years, we documented 227 incident CVDs in the first, 251 in the second, 223 in the third, and 132 in the fourth cohort. The age-adjusted incidence rates per 1,000 person-years of CVD decreased over time; 6.4 in the first, 4.3 in the second, 3.4 in the third, 2.3 in the fourth cohort. Through the four cohorts, hypertension was the leading attributable risk factor for CVD, but its contributions decreased from the first to the fourth cohorts: PAF (95% CI)= 51 (33-64)% in the first; 42 (29-53)% in the second; 47 (32-58)% in the third; and 27 (4-45)% in the fourth cohorts. In contrast, the contributions of diabetes mellitus increased between the third and the fourth cohorts and then diabetes mellitus was promoted to the second leading risk factor: PAF (95% CI)= 1 (-4-6)% in the first; 3 (-1-8)% in the second; 3 (-1-7)% in the third; and 17 (8-25)% in the fourth cohorts. The contributions of hypercholesterolemia and other risk factors were small and did not change over time materially. Similar trends were observed for stroke and coronary heart disease. Conclusions: The contribution of hypertension for CVD risk decreased, while that of diabetes mellitus increased among Japanese over the past 40 years. Taken together with the steep decrease in stroke incidence, decrease in hypertension has contributed to reducing PAF of hypertension, and also resulted in increase in PAF of diabetes. Although hypertension is still the leading attributable risk factor for CVD, diabetes could be another target of CVD prevention strategy among Japanese.

Although case series reporting pulmonary embolism or deep vein thrombosis after prolonged television watching have been published,1 no prospective study has examined the association between time spent watching television and the risk of mortality from pulmonary embolism. We examined this association in a large cohort study of Japanese men and women. The Japanese Collaborative Cohort Study is a population-based cohort study that started between 1988 and 1990 in 45 regions of Japan involving 110 585 participants 40 to 79 years of age.2 Written or verbal informed consent was obtained according to the guidelines of the Council of International Organizations of Medical Science.3 This study was approved by the ethics committees of the Nagoya University and Osaka University. After the exclusion of those who did not provide information about time spent watching television and those who reported a history of cancer, stroke, myocardial infarction, or pulmonary embolism, 86 024 participants (36 006 men and 50 018 women) were included in the analysis. Baseline information was collected by a self-administered questionnaire that included items about demographic characteristics, medical history, and lifestyle factors. Participants were asked for their …

Several studies report that daily walking reduces the risk of pneumonia. However, the elderly often experience underlying chronic diseases such as myocardial infarction (MI) or stroke which also increase the risk of pneumonia; thus, the association of walking with pneumonia may merely be a result of confounding of the underlying conditions. Therefore, we explored the links between daily walking and pneumonia mortality stratified by the presence of medical histories in 22,280 Japanese elderly (9,067 men and 13,213 women) aged 65 to 79 years from 1988-1990. The Cox proportional hazards model was used to calculate hazard ratio (HR) and 95% confidence interval (CI) for pneumonia mortality (ICD10: J9-18, J69) adjusted for age, sex, and possible confounders. P for trend was calculated across the categories of walking time. After a median of 11.9 years’ follow-up, 1,203 participants died of pneumonia. In participants without a medical history of MI or stroke, the HR of pneumonia mortality in participants who had walked at baseline for ≥1 h/day was 0.73 (95% CI, 0.62-0.85; P for trend

Background: Limited evidence is available on the association of urinary sodium and potassium excretion and risk of stroke in population samples of Japanese. Method: A total of 1,403 Japanese men and women aged 40 to 79 years with no history of cardiovascular disease at baseline survey, whose was measured 24-hour urinary excretion of sodium and potassium in the Circulatory Risk in Communities Study (CIRCS). Results: During the median 11.1-years of follow-up, we documented 62 incident strokes and 87 incident cardiovascular disease (stroke and coronary heart disease). The age-adjusted risks of stroke and total cardiovascular disease were higher among persons in the highest tertile of urinary sodium excretion compared with those in the lowest tertile (9.0g/day). These associations did not change substantially after adjusting for cardiovascular risk factors. There were no associations of the urinary potassium excretion with risk of stroke and cardiovascular disease. The respective multivariable hazard ratios (HRs, 95%CI) of stroke and total cardiovascular disease for highest tertile of urinary sodium excretion were 2.06(1.08-3.96) and 1.90(1.09-3.24). For urinary potassium excretion, the corresponding HRs was 1.09(0.51-2.34) and 1.45(0.85-2.49), respectively. Conclusions: Higher excretion of urinary sodium was associated with higher risk of stroke and total cardiovascular disease in general Japanese men and women.

Background— It is unclear whether mild hypertensive retinopathy is a risk factor for mortality. This study examined whether mild hypertensive retinopathy could be a risk factor for cardiovascular mortality in subjects with and without hypertension. Methods and Results— In this cohort study, 87 890 individuals (29 917 men and 57 973 women) 40 to 79 years of age in 1993 were followed up until 2008. Retinal photography was classified as normal, grade 1, or grade 2 based on the Keith-Wagener-Barker system. Risk ratios for all-cause and cause-specific mortality for each classification were calculated with Cox proportional hazards regression models. Covariates included age, systolic blood pressure, antihypertensive medication use, and other cardiovascular risk factors. Multivariable hazard ratios for total cardiovascular disease mortality were 1.24 (95% confidence interval [CI], 1.12–1.38) and 1.23 (95% CI, 1.03–1.47) for grades 1 and 2 among men and 1.12 (95% CI, 1.01–1.24) and 1.44 (95% CI, 1.24–1.68) for grades 1 and 2 among women, respectively. Hazard ratios for total stroke mortality were 1.31 (95% CI, 1.13–1.53) and 1.38 (95% CI, 1.08–1.77) for grades 1 and 2 among men and 1.30 (95% CI, 1.12–1.50) and 1.70 (95% CI, 1.36–2.11) for grades 1 and 2 among women, respectively. For both hypertensive and normotensive subjects of each sex, multivariable hazard ratios for all-cause mortality, total cardiovascular mortality, and total stroke mortality were significantly higher for grade 1 or 2 compared with normal. Conclusions— Mild hypertensive retinopathy is a risk factor for cardiovascular mortality independently of cardiovascular risk factors among men and women with and without hypertension.

Background: Both high insulin resistance and low insulin secretion are the causes of diabetes mellitus. Japanese level of insulin secretion said to be lower than Western people. However, the association between insulin resistance, secretion and risk of coronary heart disease (CHD) or stroke in non-diabetic population remains unknown. Hypothesis: We assessed the hypothesis that each higher insulin resistance and lower insulin secretion is associated with increased risk of CHD even in non-diabetic population. Methods: A prospective study of Japanese subjects aged 40-79 years was undertaken using data collected from 5,639 participants (1,971 men and 3,668 women) who had no past history of cardiovascular disease or diabetes mellitus in cardiovascular risk surveys conducted in three communities of the Circulatory Risk in Communities Study (CIRCS) in 2000 and 2001. Insulin resistance was calculated on the basis of fasting glucose (mg/dL) and insulin levels (μU/mL) according to the homeostasis model assessment (HOMA-r) method: glucose x insulin / 405. The HOMA-beta cell function (HOMA-b) was calculated by using the following formula: 360 x fasting insulin (μU/mL) / (fasting glucose (mg/dL) - 63). Each incidence of CHD and stroke was ascertained by systematic surveillance. The hazard ratios (HRs) of incidence of CHD and 95% confidence intervals (CI) associated with a 1 standard deviation (SD) increase in log(HOMA-r) and in log(HOMA-b) were calculated with adjustment for age and other potential confounding factors using the Cox proportional hazards model. Results: During an average follow up of 9.8-years, 39 incident CHD (including 23 myocardial infarction) and 123 incident stroke occurred. Higher insulin resistance tended to increase risk of CHD or myocardial infarction, but all HRs did not reach statistical significance. The multivariable HRs in log(HOMA-r) for risk of CHD were 1.30(95% CI: 0.23 to 7.42) for men, 2.84(0.60 to 13.56) for women, and 1.56(0.40 to 6.04) for total subjects. The corresponding HRs of myocardial infarction were 3.73(0.31 to 44.87) for men, 2.45(0.46 to 12.98) for women, and 2.71(0.48 to 15.36) for total subjects. Lower insulin secretion associated with risk of CHD or myocardial infarction, and some of HRs reached statistical significance. The multivariable HRs in log(HOMA-b) for risk of CHD were 1.17(0.56 to 2.45) for men, 0.31(0.10 to0.96, p=0.042) for women, and 0.75(0.40 to1.42) for total subjects. The corresponding HRs of myocardial infarction were 0.53(0.18 to1.53) for men, 0.31(0.07 to 1.35) for women, and 0.42(0.18 to1.00, p=0.049) for total subjects. There was no interaction between HOMA-r and HOMA-b. HOMA-r and HOMA-b did not associated with incidence of stroke. Conclusions: Lower insulin secretion may be associated with increased risk of coronary heart disease among non-diabetic Japanese population.

Introduction: Despite the high risk of cardiovascular disease (CVD) in those with chronic kidney disease (CKD), there are conflicting data as to whether two key kidney measures, estimated glomerular filtration rate (eGFR) and albuminuria, contribute to better CVD prediction, beyond conventional risk factors, warranting a more comprehensive investigation over a broad range of populations. Methods: We studied 127,825 participants without history of CVD from 12 general population, 3 high risk and 1 CKD cohorts with data on eGFR (based on the CKD-EPI creatinine equation) and urinary albumin-creatinine ratio (ACR) and at least 4 years of median follow-up for CVD mortality (4,133 deaths from 15 cohorts), coronary heart disease (CHD) (5,420 events from 9 cohorts), stroke (2,651 events from 9 cohorts), or heart failure (2,507 events from 8 cohorts). To compare eGFR and ACR with conventional predictors independently of the order of modeling, we examined the worsening of 5-year prediction of CVD outcomes by omitting each predictor in turn compared to a full model with all kidney and conventional predictors. Results: C-statistics for full models ranged from 0.759-0.836 in general population and high risk cohorts and 0.712-0.796 in the CKD population (Table). All the conventional and kidney measures contributed to better prediction of CVD outcomes. The contribution of ACR was greater than that of any conventional modifiable risk factors except in predicting CHD in both general/high-risk cohorts and CKD population. Although weaker than ACR, eGFR also contributed significantly to better prediction of CVD mortality (especially in CKD populations) and CHD. Largely similar results were observed for categorical net reclassification index. Conclusion: The two key kidney measures (particularly albuminuria) contribute as much as some or all of the conventional risk factors to CVD prediction, supporting their use for CVD risk classification in certain circumstances.

Background: To evaluate risk factors for intraparenchymal hemorrhage (IPH) by its subtypes may contribute to more effective prevention of IPH, but few prospective studies have characterized risk factors for specific subtypes of IPH. Hypothesis: We assessed the hypothesis that the associations between traditional risk factors such as blood pressures and risk of IPH vary by IPH subtype. Methods: A prospective study of Japanese subjects aged 40-69 years was undertaken using the data collected from 9,086 participants (3,560 men and 5,526 women) in cardiovascular risk surveys of three communities between 1985 and 1994. The incidence of IPH between 1985 and 2008 was ascertained by systematic surveillance. The location of hemorrhage was assigned based on admission CT scan or MRI by trained physicians. IPH selectively involving the thalamus, basal ganglia or brainstem was defined as deep IPH, while IPH isolated to the cortex was defined as lobar. The hazard ratios (HRs) of incidence of IPH and its subtypes (deep IPH and lobar or cerebellar IPH) and 95% confidence intervals (CI) relative to the baseline variables were calculated with adjustment for age and other potential confounding factors using the Cox proportional hazards model. Results: During an average follow up of 19.2-years, 139 incident IPHs (103 deep, 31 lobar or cerebellar, and 5 mixed or unclassified IPHs) occurred. Systolic blood pressure, major ST-T abnormality on electrocardiogram, and excess ethanol intake (≥46g/day) were associated with increased risk of deep IPH, while serum total cholesterol levels was inversely associated with incidence of lobar and cerebellar IPH, even after adjustment for potential confounding variables. Serum total cholesterol levels also tended to be inversely associated with incidence of deep IPH, although it did not reach statistically significance. The multivariate-adjusted HRs (95%CI) of deep IPH were 1.14 (1.05-1.23) for systolic blood pressure (10mmHg), 2.09 (1.19- 3.66) for major ST-T abnormality, and 1.86 (1.02- 3.38) for excess ethanol intake. The adjusted HR (95%CI) of lobar and cerebellar IPH was 0.22 (0.07- 0.76) for the participants with higher total cholesterol levels (≥220 mg/dL), compared with those with lower total cholesterol levels ( Conclusions: Impact of risk factors on incidence of IPH varied according to its subtype.

Background— Evidence on the relationship of a wide range of alcohol consumption with risk of incident atrial fibrillation has been limited. Methods— Between 1991 and 1995, 8602 Japanese men and women aged 30 to 80 years and free of clinical atrial fibrillation took part in the first examination of the Circulatory Risk in Communities Study(CIRCS)- a population based cohort study of cardiovascular risk factors, cardiovascular disease incidence, and their trends in Japanese communities. In the first examination, we checked a detailed medical history, physical examination, blood and urine examination, and electrocardiogram (ECG). An interviewer obtained histories in detail for weekly alcohol intake. In the follow-up period, incident atrial fibrillations were ascertained by annual ECG record and medical history of treatment of atrial fibrillation. ECGs were coded with the Minnesota Code by trained physician-epidemiologists. Differences in baseline characteristics between atrial fibrillation cases and controls were compared using Student t-tests or chi-squared tests. The hazard ratios (HRs) of incidence of atrial fibrillation and 95% confidence interval (CI) relative to the never-drinking group were calculated with adjustment for age and other potential confounding factors using the Cox proportional hazard model. Results— During an average follow-up of 6.4 years, 290 incident atrial fibrillation occurred. The higher incidence rate of atrial fibrillation was observed among participants with more than 69 g of ethanol drinking per week, compared with less than 69 g of ethanol drinking per week. On the other hand, light to moderate alcohol consumption was not associated with risk of atrial fibrillation. Compared with the never drinking group, the multivariable-adjusted HRs of past, light (69 g) drinking groups were 1.20 (95% CI, 0.61-2.35), 0.85 (95% CI, 0.57-1.27), 1.05 (95% CI, 0.63-1.75), 1.34 (95% CI, 0.78-2.32), and 2.92 (95% CI, 1.61-5.28), respectively. Conclusions— Heavy alcohol consumption was associated with the higher risk of atrial fibrillation, whereas there was no association of less than moderate alcohol consumption and atrial fibrillation.

Background: The incidence of hemorrhagic stroke has substantially decreased nearly to a quarter over years from 1960s to 1980s in Japan. However, there is little evidence on the trends after 1980, or whether subtypes of hemorrhagic stroke have decreased. We examined trends in the incidence of hemorrhagic stroke and its subtypes. Hypothesis: Hemorrhagic stroke have decreased after 1980s. For intraparenchymal hemorrhage (IPH), lobar and cerebellar intraparenchymal hemorrhage has not decreased. Methods: All residents aged 30 to 84 years in two communities (Akita and Osaka, census population: 18,998 in 2000) in Japan was surveyed. The incidence of IPH and subarachnoid hemorrhage (SAH) from 1986 to 2010 was ascertained by systematic surveillance of hospital records and confirmed by physicians with CT or MRI images. The implementation rate of CT or MRI has not changed materially over years (78.9~97.4%). IPH was further categorized to lobar and cerebellar IPH, and deep IPH (thalamus, basal ganglia or brainstem). The respective age-adjusted incidence rates per 100,000 person-years were calculated by each 5 years of time periods: 1986-1990, 1991-1995, 1996-2000, 2001-2005, and 2006-2010. Results: Age-adjusted incidence of SAH has decreased in 1996-2000 but increased after 2000. Respective age-adjusted incidence rates (per 100,000 person-years) of SAH for 5 time periods were 46.2, 46.5, 14.5, 23.0, and 23.7 (p for trend=0.23). Age-adjusted incidence of all IPH has decreased from 1986-1990 to 2006-2010. Respective incidence rates of IPH were 56.8, 53.7, 48.7, 29.0, and 26.5 (p for trend=0.01). For the subtypes of IPH, incidence of lobar and cerebellar IPH has not decreased from 1986-1990 to 2006-2010, while it have decreased for deep IPH. Respective incidence rates for lobar and cerebellar IPH were 13.1, 7.6, 16.6, 10.9, and 10.3 (p for trend=0.85) and those of deep IPH were 52.0, 53.1, 49.8, 40.8, and 29.5 (p for trend=0.03). Conclusion: The age-adjusted incidence of IPH after 1980s has decreased over years, although such trend was not clear for SAH. Among IPH, the incidence of lobar and cerebellar IPH has not decreased, while the incidence of deep IPH has decreased.

We appreciate the interest of Triantafyllou et al and Cheung et al in our recent article on retinopathy and cardiovascular mortality.1 They pointed out several issues with the article. Triantafyllou et al discussed 3 major points. First, they stated that grade 1 retinopathy did not possess a predictive role for cardiovascular mortality in normotensive individuals. However, the probability value for the trend showed significance in nonhypertensive men and women (Table 3); in addition, the multivariable (including blood pressure levels and antihypertensive medication use) adjusted hazard ratio for grade 1 retinopathy was significantly higher in all (ie, nonhypertensive and hypertensive) men and women (Table 2). These results appear to suggest the predictive role of grade 1 retinopathy. Second, the lack of information on antihypertensive drugs during the follow-up period, especially in nonhypertensive subjects, was pointed out. We excluded subjects who used antihypertensive drugs …

Introduction: Evidence for the longitudinal impact of snoring on incident hypertension in Asians is still limited. The objective of this study was to examine whether the impact of snoring with and without overweight (body mass index ≥25.0 kg/m 2 ) on the incidence of hypertension among Japanese general population. Hypothesis: We assessed the hypothesis that the impact of snoring on incident hypertension among Japanese general population was more evident for non-overweight (body mass index 2 ) subjects probably because overweight is a primary determinant of both snoring and hypertension. Method: A total of 7,482 non-hypertensive persons (2,891 men and 4,591 women) aged 35-79 years from three communities in Japan who participated in the Circulatory Risk in Communities Study (CIRCS) were followed up for 7.9 years. Hazard ratios for the incidence of hypertension were calculated according to snoring (self-administered questionnaire) stratified by body mass index. Results: At baseline, 74 % of men and 59 % of women had sometimes or everyday snoring and 31 % of men and 28 % of women had being overweight. During the 7.9-year follow-up period, 1087 individuals (451 men and 636 women) developed hypertension. The multivariable-adjusted odds ratio (95%confidence interval) of incident hypertension for snoring compared with non-snoring was 1.37 (1.06 to 1.76) for men and 1.32 (1.10 to 1.58) for women. When stratified by body mass index at baseline, the significant association was observed only in non-overweight; the multivariable adjusted odds ratio of incident hypertension with non-overweight was 1.52 (1.12 to 2.04) for men and 1.39 (1.11 to 1.73) for women, and respective odds ratio for snoring with overweight was 1.11 (0.68 to 1.81) for men and 1.14 (0.83 to 1.53) for women compared with non-snoring categories although the interaction was not statistically significant. The population-attributable fraction (PAF) of snoring with non-overweight for incident hypertension was 26 % for men and 17% for women. Conclusions: Snoring may have larger impact on development hypertension among non-overweight Japanese men and women.

Background: Previous studies have reported that anger expression style is associated with cardiovascular disease and its risk factors. Epidemiologic data regarding the association between anger expression and risk of stroke have been inconsistent, and few studies have reported the association of anger expression with stroke subtypes. Hypothesis: We assessed the hypothesis that suppressed anger is associated with increased risk of stroke, and the associations vary by stroke subtype. Methods: A prospective study of Japanese subjects aged 30–74 years was undertaken using data collected from 6,276 participants (2,285 men and 3,991 women) in cardiovascular risk surveys conducted in four communities between 1995 and 1998. We used the Spielberger Anger Expression Scale to measure self-reported levels of anger expression style; anger expressed outwardly is regarded as “anger-out” and anger held in or suppressed is classified as “anger-in.” The incidence of stroke was ascertained by systematic surveillance. The hazard ratios (HRs) of incidence of stroke and its subtypes and 95% confidence intervals (CI) relative to the lowest tertile of anger expression were calculated with adjustment for age and other potential confounding factors using the Cox proportional hazards model. Results: During an average follow up of 9.1-years, 129 incident strokes (50 hemorrhagic, 77 ischemic, and 2 unclassified strokes) occurred. Men with “anger-in” scores in the highest tertile had a 1.7-fold multivariable-adjusted relative risk of stroke as compared with those in the lowest tertile, but it did not reach statistical significance (HR; 1.73, 95% CI; 0.94–3.19, p=0.08). When stratifying for stroke subtypes, the associations between “anger-in” and stroke tended to be stronger for hemorrhagic stroke than ischemic stroke; HRs (95% CI) were 3.82 (1.03–14.1) for hemorrhagic stroke and 1.32 (0.64–2.71) for ischemic stroke. Anger-in score was not associated with incidence of either hemorrhagic or ischemic stroke in women. In men and women, there were no significant associations between “anger-out” and incidence of stroke. As compared with the lowest tertile of “anger-out” scores, the multivariable-adjusted HRs of stroke for the highest tertile of “anger-out” scores were 1.39 (0.74–2.59) for men and 1.48 (0.78–2.78) for women, respectively. Conclusions: Suppressed anger may be associated with increased risk of hemorrhagic stroke among middle-aged Japanese men, but not women.

Introduction: The evidence for the relation between diabetes mellitus and risk of dementia in Asian population has been limited. This study investigated the association between diabetes mellitus and risk of dementia in community-based samples of Japanese. Hypothesis: There is an association between diabetes mellitus and risk of disabling dementia Methods: We performed a nested case-control study based on a cohort of about 6,000 Japanese aged 45-69 at baseline between 1984 and 1994. The information of the status of diabetes at baseline was collected for 174 disabling dementia cases incident between 1999 and 2004 and 348 controls matched with cases by age±3, sex and baseline-year. Incident disabling dementia was defined as dependent individuals who had moderate to severe dementia-related behavioral disturbance or cognitive impairment. This criterion was previously validated with 5-cog test. Dementia cases were further classified according to presence of history of stroke. The conditional odds ratio (OR) for dementia was calculated according to the status of (1) glucose intolerance; fasting serum glucose 110-125mg/dl or non-fasting serum glucose 140-199 mg/dl; and (2) diabetes; fasting serum glucose ≥126 mg/dl, non-fasting serum glucose ≥200 mg/dl, and/or the use of glucose-lowering medication or insulin therapy. Variables for multivariable adjustment were body mass index, smoking status, total cholesterol and hypertension status (140≤SBP Results and Conclusions: Mean follow-up year was 11.4 years. The proportion of men was 32.8% and the prevalence of glucose intolerance and diabetes among controls were 16.7% and 5.8%, respectively. Of dementia cases, 44.8% had history of stroke. Compared with persons with normal glucose level, ORs (95%CI) for glucose intolerance and diabetes were 1.12 (0.68-1.84) and 2.18 (1.13-4.22), and multivariable ORs were 0.98 (0.58-1.65) and 2.04 (1.03-4.03), respectively. For sex stratified analysis, ORs for diabetes were 1.39 (0.45-4.31) for men, and 2.79 (1.20-6.50) for women. The association of diabetes was primarily observed in dementia cases with stroke history [OR=3.19 (1.04-9.82)], but not in those without it [OR=1.78 (0.78-4.07)]. In conclusion, we found an association between diabetes mellitus and risk of disabling dementia. The association was confined to women, and dementia with stroke history.

BACKGROUND A previous prospective study showed that dietary intakes of fish or ω-3 polyunsaturated fatty acids (PUFA) were associated with reduced risk of heart failure (HF), but no study has examined the association between plasma ω-3 PUFA and HF. Our a priori hypothesis was that plasma ω-3 PUFA would be inversely associated with incident HF. METHODS We included 3,592 white participants from the Minneapolis field center of the ARIC Study, aged 45– 64 at baseline (1987– 89), initially free of coronary heart disease, stroke and HF. Fractions of plasma cholesterol ester (CE) and phospholipid (PL) ω-3 PUFA of marine origin (eicosapentaenoic acid [EPA], ω-3 docosapentaenoic acid [DPA] and docosahexaenoic acid [DHA]) were measured by gas-liquid chromatography. The subjects were followed through 2003, and incident HF was defined by a hospital discharge or death including a HF ICD code. The hazard ratio and 95% confidence interval (HR [95%CI]) for HF was calculated according to quintiles of CE and PL fractions of ω-3 PUFA. RESULTS During the 14.3-year follow-up, we identified 197 cases of CHF (110 for men and 87 for women). After adjustment for age and potential risk factors (smoking, pack-years, alcohol consumption, body mass index, systolic blood pressure, antihypertensive medication use, plasma total cholesterol, diabetes and total energy intake), the incidence of HF for women was associated inversely and linearly with ω-3 PUFA in the PL fraction (HR for highest quintile vs lowest quintile was 0.27 [0.12– 0.63], p for trend CONCLUSIONS We found a significant inverse association between plasma ω-3 PUFA and incident HF among women, consistent with a prior study using a food frequency questionnaire. Contrary to our hypothesis, however, no such association was observed for men.

Background— To date, no prospective studies have examined the association between serum homocysteine levels and the risk of stroke and stroke subtypes in Asian populations. Methods and Results— A prospective, nested, case-control study of Japanese subjects 40 to 85 years of age was conducted by using frozen serum samples from 11 846 participants in cardiovascular risk surveys collected from 1984 to 1995 for one community and 1989 to 1995 for the other two communities. By the end of 2000, we identified 150 incident strokes, the subtypes of which were confirmed by imaging studies. Three control subjects per case were selected by matching for sex, age, community, year of serum storage, and fasting status. Serum total homocysteine levels were measured by high-performance liquid chromatography. Compared with control subjects, total (n=150), hemorrhagic (n=52), and ischemic (n=98) strokes had higher geometric mean values of total homocysteine and higher proportions of homocysteine ≥11.0 μmol/L. The multivariate odds ratios (95% CI) for highest (≥11.0 μmol/L) versus lowest quartiles ( Conclusions— High total homocysteine concentrations were associated with the increased risk of total stroke, more specifically ischemic stroke and lacunar infarction, among Japanese men and women.