Honolulu Heart Program Cohort Research Articles

Kuakini HHP Center for Translational Research on Aging: Aims and Preliminary Findings

Kuakini Medical Center (Kuakini) is creating an interdisciplinary Hawai’i-based Center for translational research on aging. This Center will build upon Kuakini’s five-decades of NIH-funded research, its 420,000-specimen biorepository, and existing strengths in aging research, notably, the 56-year ongoing Kuakini Honolulu Heart Program cohort study (Kuakini HHP), Kuakini Honolulu-Asia Aging Study (Kuakini HAAS), and Kuakini HHP Offspring Study. The overall goal is to find practical means to enhance healthy human lifespan (healthspan). Four research project leaders (RPLs) have been selected from various disciplines for mentorship in translational aging research. The first RPL presentation will introduce a novel mouse model, enabling controlled expression of the pro-longevity gene FoxO3, and assess the impact on lifespan and healthspan phenotypes in mice. These phenotypes will be compared to similar phenotypes in humans with/without the FOXO3 longevity genotype. The second RPL presentation will assess the relation between leukocyte telomere attrition rates (from banked blood collected at three time points over 20-plus years) in older Kuakini HHP men with/without the FOXO3 longevity genotype. The third RPL presentation will assess whether FOXO3 genotype, peripheral leukocyte telomere dynamics (attrition rate, telomerase activity) and inflammatory cytokines mediate the human brain integrity and function with age. This project will utilize structural and functional MRI data from male and female Kuakini HHP Offspring Study participants. The fourth RPL presentation will assess whether APOE e2, e4, and FOXO3 longevity-associated alleles impact 34-year incidence of intracerebral hemorrhage. We will summarize the findings, address the healthspan implications and provide future directions. Supported by NIH 5P20GM125526.

Longevity-Associated FOXO3 Genotype and its Impact on Coronary Artery Disease Mortality in Japanese, Whites, and Blacks: A Prospective Study of Three American Populations.

We recently reported that protection against coronary artery disease (CAD) mortality is the major contributor to longer life associated with FOXO3 genotype. The present study examined this relation in more detail. We performed a 15-year observational study of 3,584 older American men of Japanese ancestry from the Kuakini Honolulu Heart Program cohort and 1,595 White and 1,067 Black elderly individuals from the Health Aging and Body Composition study. Multivariate Cox regression models demonstrated that carriage of the longevity-associated G allele of FOXO3 single nucleotide polymorphisms rs2802292 was a protective factor against CAD mortality in all three populations. In Japanese and Whites, but not in Blacks, the protective effect of the G allele was little changed in models adjusted for other major risk factors. Population-attributable risk (PAR) models found that the nonprotective TT genotype contributed 15%, 9%, and 3% to CAD mortality risk in Japanese, White, and Black Americans, respectively, and was one of the top three contributing factors to CAD mortality. In Japanese, this effect size was comparable with hypertension (15%), but in Whites and Blacks PAR for hypertension was higher (29% and 26%, respectively). G-allele carriers had lower plasma TNF-α than noncarriers, suggesting inflammation as a potential mediating factor for CAD mortality risk. FOXO3 genotype is an important risk factor for CAD mortality in older populations. More research is needed to identify potential mechanisms and targets for intervention.

The FoxO3 gene and cause-specific mortality.

SummaryThe G allele of the FOXO3 single nucleotide polymorphism (SNP) rs2802292 exhibits a consistently replicated genetic association with longevity in multiple populations worldwide. The aims of this study were to quantify the mortality risk for the longevity‐associated genotype and to discover the particular cause(s) of death associated with this allele in older Americans of diverse ancestry. It involved a 17‐year prospective cohort study of 3584 older American men of Japanese ancestry from the Honolulu Heart Program cohort, followed by a 17‐year prospective replication study of 1595 white and 1056 black elderly individuals from the Health Aging and Body Composition cohort. The relation between FOXO3 genotype and cause‐specific mortality was ascertained for major causes of death including coronary heart disease (CHD), cancer, and stroke. Age‐adjusted and multivariable Cox proportional hazards models were used to compute hazard ratios (HRs) for all‐cause and cause‐specific mortality. We found G allele carriers had a combined (Japanese, white, and black populations) risk reduction of 10% for total (all‐cause) mortality (HR = 0.90; 95% CI, 0.84–0.95; P = 0.001). This effect size was consistent across populations and mostly contributed by 26% lower risk for CHD death (HR = 0.74; 95% CI, 0.64–0.86; P = 0.00004). No other causes of death made a significant contribution to the survival advantage for G allele carriers. In conclusion, at older age, there is a large risk reduction in mortality for G allele carriers, mostly due to lower CHD mortality. The findings support further research on FOXO3 and FoxO3 protein as potential targets for therapeutic intervention in aging‐related diseases, particularly cardiovascular disease.

The Relation between Apolipoprotein A-I and Dementia: The Honolulu-Asia Aging Study

The association between apolipoproteins and neurodegeneration is unclear. The authors examined the association of dementia with serum levels of apolipoprotein A-I (ApoA-I) alone and in combination with the apolipoprotein E genotype (ApoE). Subjects were Japanese-American men in Hawaii followed since 1965 in the Honolulu Heart Program cohort and the Honolulu-Asia Aging Study. Lipid levels were assessed in 1980-1982. Dementia was diagnosed in 1991-1993, 1994-1996, and 1997-1999 by using a multistep procedure and international guidelines. The sample consisted of 929 men (107 dementia cases). The relation between ApoA-I and dementia was examined by using Cox proportional hazards models adjusted for age, education, and cardiovascular risk factors. Compared with men in the lowest quartile, men in the highest quartile of ApoA-I concentration had a significantly lower risk of dementia (hazard ratio = 0.25, 95% confidence interval: 0.08, 0.78). Compared with men with both risk factors, those with a high ApoA-I concentration and no ApoE epsilon4 had a significantly lower risk of dementia (hazard ratio = 0.21, 95% confidence interval: 0.08, 0.52). Previous work has demonstrated an inverse relation between ApoA-I and cardiovascular disease, and the authors extended these findings to the risk of dementia. These results raise the possibility that different lipoprotein components of cholesterol may be differentially associated with dementia.

A Nonlinear Association of Serum Cholesterol Level with Incident Coronary Heart Disease in Elderly Men

P62 Increased serum cholesterol has been found to be a direct risk factor for coronary heart disease (CHD) in middle aged individuals in epidemiologic follow-up studies. However, this relationship has not been consistently reproduced in elderly populations. From 1991-93, 2424 elderly (71-93 years old) Japanese American men from the Honolulu Heart Program cohort, who did not have prevalent CHD and who were not on cholesterol lowering drugs, were examined. The subsequent 6-year follow-up provided an opportunity to closely examine the relationship between their baseline cholesterol and their incidence of CHD as determined by surveillance and a physician panel review. The study revealed a significant non-linear association of the baseline cholesterol with the incidence of CHD (p=0.033). Among the 1524 elderly men with cholesterol values below 200mg/dl the age adjusted CHD rate decreased as the mean cholesterol level increased. In those with cholesterol levels of greater than or equal to 200mg/dl (900) the age adjusted CHD rates increased with increasing level of cholesterol. The lowest CHD rates were seen in men with cholesterol levels between 200 and 219 mg/dl. Multivariate adjustment for other know risk factors (BMI, smoking, hypertension, diabetes, and HDL), decreased the strength of this non-linear relationship but it remained significant (p=0.049). A measure of frailty (10% weight loss since an exam 20 years before) further reduced the strength of the cholesterol relationship slightly but did so by reducing the magnitude of the cholesterol/CHD relationship in both those higher cholesterol (positive relationship) and those with lower cholesterol (negative relationship). Thus it is unlikely that frailty can fully explain the reverse cholesterol/CHD relationship seen in those with lower cholesterol levels. In this study the majority of men already had cholesterol levels below the standard cutpoint of 200mg/dl and had the opposite pattern of risk as elderly men with cholesterols above that cutpoint. This would indicate that further data is needed before recommending lowering of cholesterol to levels below 200mg/dl in older individuals.

Apolipoprotein A-1 predicts coronary heart disease only at low concentrations of high-density lipoprotein cholesterol: an epidemiological study of Japanese-Americans.

Conventional epidemiological and clinical studies of apolipoprotein A-1 and high-density lipoprotein-cholesterol have demonstrated, when examined jointly, that high-density lipoprotein is a better predictor of coronary heart disease. This strategy does not take into account known lipid metabolic relationships. A statistical approach that takes into account apolipoprotein A-1 being a constituent of the high-density lipoprotein particle is more appropriate. Among 1,177 Japanese-American men of the Honolulu Heart Program cohort free of disease at baseline (1980-1982), 182 new coronary heart disease cases developed over a 12-year follow-up period. After removing the linear relationship with high-density lipoprotein-cholesterol, a relative measure of apoliprotein A-1 concentration was derived. Based on joint conditions of "low" and "high" relative apoliprotein A-1 concentration and < or =40 and >40 mg/dl for the high-density lipoprotein-cholesterol distribution, four groupings were created. Among relative joint groupings of high/< or =40, low/< or =40, high/>40, and low/>40, respectively, the 12-year coronary heart disease incidence varied from 28.6, 18.2, 8.3, to 11.7 cases per 1,000 person-years. A test of statistical interaction was significant (P=0.028). Additional analyses revealed coronary heart disease cases were more likely among men with triglycerides > 190 mg/dl. Observed patterns of relationships among relative apoliprotein A-1 level, high-density lipoprotein cholesterol, and triglycerides with incident coronary heart disease are consistent with patterns noted in clinical, laboratory, and transgenic animal research more capable of elucidating mechanisms of disease causation. This epidemiological study suggests similar mechanisms may be operating at a population level, and may contribute to the public health burden of coronary heart disease.

Associations of symptoms of sleep apnea with cardiovascular disease, cognitive impairment, and mortality among older Japanese-American men.

To examine the association between symptoms of sleep apnea and prevalent cardiovascular disease, cognitive impairment, and subsequent 3-year mortality. A longitudinal study. Participants lived in the community on Oahu, Hawaii. A total of 2905 older Japanese-American men participating in the fourth examination of the Honolulu Heart Program cohort study from 1991-1993, which is the baseline for the Honolulu-Asia Aging Study of dementia. Self-reported snoring, daytime sleepiness, and breathing pauses; diagnosed cardiovascular disease and dementia; cognitive functioning and vital status approximately 3 years later. More than 12% of the participants reported that they often or always snored loudly, and 8% reported being sleepy most of the day. Fewer than 2% reported that they stop breathing when sleeping, and this was found more frequently among habitual snorers (7%, P < .001) and those sleepy during the day (5%, P < .001). The prevalence of habitual snoring declined in the older age groups, was higher among those with greater Body Mass Index scores, and was not associated with the reporting of daytime sleepiness, diagnosis of heart disease, stroke, dementia, or cognitive impairment. Daytime sleepiness was more prevalent at older ages and was associated with a higher prevalence of heart disease and with cognitive impairment and dementia, chronic obstructive pulmonary disease, and diabetes. Self-reported apnea was associated only with a history of pneumonia. Three-year mortality was not associated with these symptoms after adjusting for prevalent heart disease and cognitive impairment. Symptoms of sleep apnea are reported less frequently in older Japanese-American men. Excessive daytime sleepiness is associated with poor cognition and dementia, but whether it also is an indicator for sleep apnea in this age group remains unclear. Epidemiologic studies of sleep apnea in older adults will require polysomnography to determine accurately the correlates and consequences of this condition.

Influence of Myocardial Infarction, Coronary Artery Bypass Surgery, and Stroke on Cognitive Impairment in Late Life

Relations between cognitive test scores in later life and prior myocardial infarction (MI), coronary artery bypass graft surgery (CABG), and stroke were examined for this study. Subjects were 3,734 Japanese-American men (80% of surviving Honolulu Heart Program cohort) aged 71 to 93 years at the time of cognitive testing. Impairment was defined as scoring below the 16th percentile on a validated cognitive assessment scale. Prior MI, stroke, and CABG were established using hospital surveillance, history, and record review. After adjustment for age, years of education, and years of childhood spent in Japan, men with prior stroke were significantly more likely than others to have poor cognitive performance (odds ratio 4.4, 95% confidence limits 3.0 to 6.7). History of >1 stroke was associated with an odds ratio of 50 (95% confidence limits 10.5 to 238.3). There was no significant association between cognitive performance and ≥1 prior MI or history of CABG. Time between events and cognitive function testing did not affect results. Analyses support a significant association between clinical stroke and persistent cognitive impairment, but fail to implicate CABG or MI.

Prolonged QTc Interval, Impaired Pulmonary Function, and a Very Lean Body Mass Jointly Predict All-Cause Mortality in Elderly Men

PURPOSE: Assess the joint impact of prolonged QTc interval (QT interval corrected for heart rate), impaired lung function, and low body weight on all-cause mortality.METHODS: This is a population-based, prospective study of the Honolulu Heart Program cohort, performed in Oahu Island, Hawaii, during the 1991-1993 fourth examination of cohort survivors. The participants were 3056 Japanese-American men, 71 to 93 years of age. The measurement consisted of: 1) instrument calculated, heart rate corrected QT interval; 2) one second forced expiratory volume (FEV1) as a percentage of age- and height-predicted FEV1; and 3) body mass index (BMI, kg/m2). Relations of subsequent 3 1/2 year, on average, mortality rates with high risk states of these variables are determined. High risk states are QTc > 440 msec, percent predicted FEV1 ⩽ 80%, and BMI ⩽ 21 kg/m2.RESULTS: Mortality rates synergistically increase among groups with one, two, or three high risk states. Men having all three high risk states are seven times more likely to die in the follow-up period than men with no high risk conditions. Very thin men having one other high risk state, pulmonary impairment or prolonged QTc, are four times more likely to die. Excluding diabetics, active smokers, or men taking drugs affecting QT interval does not alter findings. Excluding prevalent coronary heart disease decreases mortality rates among joint high risk groups.CONCLUSION: Results are consistent with clinical studies identifying an autonomic neuropathy associated with wasting chronic lung disease, prolonged QTc, and mortality. Aging populations in developed nations will increase the prevalence of diseases associated with these conditions in decades to come.

The synergy of low lung function and low body mass index predicting all-cause mortality among older Japanese-American men.

To assess the joint characteristics of low standardized weight and compromised pulmonary function in predicting all-cause mortality. A population-based, prospective cohort study. Oahu Island, Hawaii. Surviving Japanese-American men of the Honolulu Heart Program cohort, 71 to 93 years of age (N = 3059). Body mass index (BMI-weight in kilograms/square of height in meters) and 1-second forced expiratory volume (FEV1) as a percentage of age- and height-predicted FEV1 from the 1991 to 1993 examination of the cohort. Mortality data derived from the ongoing tracking of deaths of the cohort. Relations of selected risk factors among joint levels of BMI (< or = 21, > 21 to < 25, > or = 25 kg/m2) and percent predicted FEV1 (< or = 70%, > 70%) were determined. The impact of these covariates on relations between joint BMI/percent predicted FEV1 levels and subsequent all-cause mortality was assessed. The highest age-adjusted mortality rate (91.9 deaths per 1000 person-years) was noted among men characterized by the joint conditions of percent predicted FEV1 < or = 70% and BMI < or = 21 kg/m2. This rate was 4.0 times the mortality rate of a "healthy" reference group characterized by percent predicted FEV1 > 70% and 21 < BMI < 25 kg/m2. This rate ratio is attenuated to 3.2 upon statistical control for measures of current and past smoking behavior. Among the three strata of BMI, statistical interaction is reflected in a heterogeneity of mortality rate differences (49.7, 21.8, -9.6 deaths/person-year, respectively) and rate ratios (2.18, 1.98, .66, respectively) comparing men with percent predicted FEV1 < or = 70% to > 70%. Joint loss of pulmonary function and relative weight is predictive of subsequent all-cause mortality in excess of additive or multiplicative effects of each condition separately. Smoking behavior may contribute to this observation.

Reference values for pulmonary function tests of Japanese-American men aged 71 to 90 years.

Pulmonary function was assessed by spirometry in 3,076 elderly Japanese-American men of the Honolulu Heart Program (HHP) cohort. The assessment was done with a stringent quality assurance program that adhered to American Thoracic Society (ATS) recommendations for spirometry. Less than 6% of the participants were unable to perform three acceptable spirometry maneuvers. A "healthy" subgroup of 528 men between the ages of 71 and 90 yr was identified by excluding almost all smokers and subjects with lung disease and other factors negatively influencing FEV1. Reference equations and normal ranges for FEV1, FVC, and the FEV1/FVC ratio were derived from the healthy group. Use of prediction equations from the Cardiovascular Health Study (CHS) of elderly European-American men consistently overpredicted FVC by 0.3 to 0.4 L and FEV1 by 0.15 L. Men in the HHP were on average 11 cm shorter than those in the CHS. Use of a prediction equation derived from the HHP cohort when the men in the cohort were on average 22.6 yr younger consistently overpredicted FEV1 by 0.2 to 0.3 L. These results underscore the importance of using prediction equations appropriate to the ethnicity, age, and height characteristics of the subjects being studied.

Plasma Fibrinogen and Coronary Heart Disease in Elderly Japanese-American Men

Clinical and epidemiological studies consistently indicate that elevations in plasma fibrinogen concentration are associated with the presence and development of coronary heart disease (CHD). These elevations are strongly correlated with smoking behavior and may play a significant role in mediating a relation of smoking to CHD. This cross-sectional survey of 3571 elderly Japanese-American men, aged 71 through 93 years, represents survivors of the Honolulu Heart Program cohort. Active smokers are almost twice as likely to be represented in the highest quintile of the fibrinogen distribution compared with the lowest quintile (9.8% versus 5.3%, respectively). The highest prevalence of CHD (34%) was noted in past and current smokers who were in the highest quintile of fibrinogen. The age-adjusted relative odds of prevalent CHD comparing the average fibrinogen levels in the first and fifth quintiles were 1.36 (95% confidence interval, 1.13 to 1.64). After adjustment for smoking status, blood pressure, total and HDL cholesterol, diabetes status, hematocrit, and white cell count, the association between fibrinogen and CHD was changed slightly and remained statistically significant (P < .05). These findings in an elderly cohort of Japanese-American men are consistent with previous studies among middle-aged adults demonstrating fibrinogen to be associated with indicators of clinical CHD and CHD risk factors. Because of the cross-sectional nature of this study, it is not possible to distinguish whether the observed relation of fibrinogen to prevalent CHD is causal or whether it represents a marker of active and progressive disease.

To What Extent Do Cardiovascular Risk Factor Values Measured in Elderly Men Represent Their Midlife Values Measured 25 Years Earlier? A Preliminary Report and Commentary From the Honolulu Heart Program

In a Honolulu Heart Program cohort of 1,604 elderly men aged 70-90 years who were sampled from the most recent follow-up examination (1991-1992), current risk factor values were compared with those obtained 25 years earlier when the same men were between ages 45-64 and free of clinically diagnosed cardiovascular disease. Cardiovascular disease risk factors studied included systolic blood pressure, diastolic blood pressure, serum cholesterol, cigarette smoking, body mass index, and alcohol intake. For systolic pressure, 65% of the men had moved into a different quartile by old age, with 25% changing by more than one quartile. For diastolic pressure, 68% had moved into another quartile, with 28% moving more than one quartile, and for body mass index, 53% had moved into another quartile, with 14% moving more than a quartile. Less than 1% started to smoke, while 27% were reclassified from smokers to nonsmokers. Only 4% started to drink alcohol, while 30% were reclassified from drinkers to nondrinkers. When the men were stratified into cardiovascular disease, noncardiovascular disease, and healthy follow-up groups, modest deviations from the overall pattern were observed, with morbidity groups showing a greater tendency to reduction in risk factor levels. The results show that there is a substantial redistribution of major cardiovascular disease risk factor values between midlife and old age. Since midlife values are more likely to represent lifelong exposure values that, in turn, make the main contribution to the development of atherosclerosis, investigators and clinicians may need to be cautious in using risk factor values measured late in life as the only means of assessing risk for subsequent disease.

Elevated serum cholesterol is a risk factor for both coronary heart disease and thromboembolic stroke in Hawaiian Japanese men. Implications of shared risk.

The relation between total serum cholesterol level and thromboembolic or nonhemorrhagic stroke is controversial. The Honolulu Heart Program cohort of Japanese-American men provides data which show that elevated serum cholesterol is an independent predictor of thromboembolic stroke as well as coronary heart disease (CHD). The data are presented to suggest that the association of elevated cholesterol with stroke is sometimes underestimated or underreported partly because of competing or shared risk with CHD, the other major atherosclerotic end point. The data are based on 6352 men (aged 51 to 74 years) at baseline examination (1971 to 1974) who were free of clinical CHD and stroke and were followed an average of 15 years for new cases of both end points. Relative risks of serum cholesterol for CHD and thromboembolic stroke were calculated, controlling for other major cardiovascular covariates. There was a continuous and progressive increase in both CHD and thromboembolic stroke rates with increasing levels of serum cholesterol. The relative risk between the highest and lowest quartiles of serum cholesterol was 1.7 (95% confidence interval, 1.4 to 2.0) for CHD and 1.4 (95% confidence interval, 1.1 to 1.9) for thromboembolic stroke. There was a decline in the difference in relative risks between CHD and thromboembolic stroke in older men (aged 60 years and older) compared with younger men (aged younger than 60 years). These data provide additional evidence that elevated serum cholesterol should be considered a primary risk factor for thromboembolic stroke, presumably through its effect on both coronary and cerebrovascular atherosclerosis. It is suggested that this association is sometimes underestimated or underreported partly because of shared or competing risk with CHD, the clinical manifestation of atherosclerosis that generally occurs earlier in life and with greater frequency than thromboembolic stroke.

Non-recognition of dementia by caregivers; frequency and correlates in the Honolulu heart program cohort : G.W. Ross, L.R. White, C. Trockman, H. Petrovitch, K.H. Masaki, D. Chiu, C. Murdaugh. Honolulu Veterans Affairs; NIA; Honolulu Heart Program (347 N. Kuakini Street, Honolulu, HI.); NINR

Non-recognition of dementia by caregivers; frequency and correlates in the Honolulu heart program cohort : G.W. Ross, L.R. White, C. Trockman, H. Petrovitch, K.H. Masaki, D. Chiu, C. Murdaugh. Honolulu Veterans Affairs; NIA; Honolulu Heart Program (347 N. Kuakini Street, Honolulu, HI.); NINR

Mortality and morbidity among blue and white collar workers in the Honolulu Heart Program cohort.

There has been a controversy regarding occupation, as defined by blue and white collar work, and coronary heart disease (CHD). In a large long-term prospective study of the Honolulu Heart Program cohort, there were no significant differences between these two occupational classifications and the incidence rates of fatal and non-fatal CHD. Nor were there significant differences in total mortality, stroke or total cancer although there were very small but significant differences in some CHD risk factors. Unlike the previous studies, this investigation was free of prevalent cases of CHD at baseline and controlled for the major risk factors of CHD. We question the utility of using this occupational classification to investigate the causes of CHD.

Lipid and lipoprotein predictors of coronary heart disease in elderly men in the Honolulu Heart Program.

Data from the baseline and follow-up examinations of the Honolulu Heart Program (HHP) cohort of 8006 men of Japanese ancestry were used to examine several questions concerning the predictive role of lipids and lipoproteins for incident CHD. For the question "Do serum cholesterol levels measured in middle age predict incident CHD in elderly men 65 years old or older?" the answer was clearly yes. Multivariate relative risks and attributable risks for early and late onset of CHD were similar and statistically significant. For the question "Do serum cholesterol levels measured in the elderly predict subsequent CHD?" the answer again was clearly yes. Multivariate relative risks for elderly men were similar to those for middle-aged men, and attributable risks were consistently higher for elderly men. For the question "Do other lipid or lipoprotein levels measured in the elderly predict incident CHD better than serum cholesterol level?" the answer was no. Multivariate relative risks for low-density-lipoprotein (LDL) and non-high-density-lipoprotein cholesterol were similar to those for total cholesterol. HDL cholesterol was protective for incident CHD, but the patterns were not significant for the elderly. Serum triglyceride level was not a significant predictor of CHD for the elderly. The conclusion was that no matter at what age serum cholesterol was measured, it predicted subsequent CHD in the elderly men in this cohort.

The Nineteen-Year Trends in CHD in the Honolulu Heart Program

Accumulated data from the Honolulu Heart Program cohort of men of Japanese ancestry were analysed for 19-year trends in coronary heart disease (CHD) mortality and incidence rates, case-fatality ratios, and changes in risk factor levels. From 1966 to 1984, there was an estimated average annual decline of 2.7% for CHD mortality rates for the population from which the cohort was drawn. This decline was similar to that of US white males of the same age distribution. Within the examined cohort of men who were free of prevalent CHD at entry into the study, the temporal pattern of incidence rates of definite CHD was quite similar to the pattern of CHD mortality rates. Thus, for the examined cohort, mortality rates were an accurate indication of incidence rates. There was no evidence that the decline in CHD mortality rates was due to reductions in case-fatality ratios. There were, however, significant decreases in the percentages of men who smoked cigarettes, in blood pressure level among treated hypertensives and in the intake of dietary cholesterol. These changes, and a drop in coronary artery atherosclerosis measured during the same time period, are all compatible with the concept that the decline of CHD mortality among Japanese men in Hawaii reflects primary prevention of new disease.

The Nineteen-Year Trends in CHD in the Honolulu Heart Program

Accumulated data from the Honolulu Heart Program cohort of men of Japanese ancestry were analysed for 19-year trends in coronary heart disease (CHD) mortality and incidence rates, case-fatality ratios, and changes in risk factor levels. From 1966 to 1984, there was an estimated average annual decline of 2.7% for CHD mortality rates for the population from which the cohort was drawn. This decline was similar to that of US white males of the same age distribution. Within the examined cohort of men who were free of prevalent CHD at entry into the study, the temporal pattern of incidence rates of definite CHD was quite similar to the pattern of CHD mortality rates. Thus, for the examined cohort, mortality rates were an accurate indication of incidence rates. There was no evidence that the decline in CHD mortality rates was due to reductions in case-fatality ratios. There were, however, significant decreases in the percentages of men who smoked cigarettes, in blood pressure level among treated hypertensives and in the intake of dietary cholestrol. These changes, and a drop in coronary artery atherosclerosis measured during the same time period, are all compatible with the concept that the decline of CHD mortality among Japanese men in Hawaii reflects primary prevention of new disease.

A prospective study of mortality and morbidity among carpenters in the Honolulu Heart Program Cohort.

The relationship of occupation as a carpenter to the development of definite coronary heart disease, stroke, cancer and total mortality was examined in men of Japanese ancestry participating in the Honolulu Heart Program. After 18 years of follow-up those men who indicated that their present and usual occupation was carpentry had a significantly lower age-adjusted rate of definite coronary heart disease and a significantly lower mortality rate compared to participants who were never occupied as carpenters. There were no significant differences for age-adjusted rates for stroke and cancer nor any differences for those whose usual occupation was carpentry but present occupation was not carpentry. These results were unchanged when controlling for several cardiovascular risk factors and potentially confounding variables. Unlike earlier observations, these findings are not affected by the "healthy worker bias" and support the relationship that carpentry is associated with lower rates of cardiovascular disease.