Halit Silbershatz

Prediction of Coronary Heart Disease Using Risk Factor Categories

BACKGROUND The objective of this study was to examine the association of Joint National Committee (JNC-V) blood pressure and National Cholesterol Education Program (NCEP) cholesterol categories with coronary heart disease (CHD) risk, to incorporate them into coronary prediction algorithms, and to compare the discrimination properties of this approach with other noncategorical prediction functions. METHODS AND RESULTS This work was designed as a prospective, single-center study in the setting of a community-based cohort. The patients were 2489 men and 2856 women 30 to 74 years old at baseline with 12 years of follow-up. During the 12 years of follow-up, a total of 383 men and 227 women developed CHD, which was significantly associated with categories of blood pressure, total cholesterol, LDL cholesterol, and HDL cholesterol (all P<.001). Sex-specific prediction equations were formulated to predict CHD risk according to age, diabetes, smoking, JNC-V blood pressure categories, and NCEP total cholesterol and LDL cholesterol categories. The accuracy of this categorical approach was found to be comparable to CHD prediction when the continuous variables themselves were used. After adjustment for other factors, approximately 28% of CHD events in men and 29% in women were attributable to blood pressure levels that exceeded high normal (> or =130/85). The corresponding multivariable-adjusted attributable risk percent associated with elevated total cholesterol (> or =200 mg/dL) was 27% in men and 34% in women. CONCLUSIONS Recommended guidelines of blood pressure, total cholesterol, and LDL cholesterol effectively predict CHD risk in a middle-aged white population sample. A simple coronary disease prediction algorithm was developed using categorical variables, which allows physicians to predict multivariate CHD risk in patients without overt CHD.

Impact of Atrial Fibrillation on the Risk of Death The Framingham Heart Study

BACKGROUND Atrial fibrillation (AF) causes substantial morbidity. It is uncertain whether AF is associated with excess mortality independent of associated cardiac conditions and risk factors. METHODS AND RESULTS We examined the mortality of subjects 55 to 94 years of age who developed AF during 40 years of follow-up of the original Framingham Heart Study cohort. Of the original 5209 subjects, 296 men and 325 women (mean ages, 74 and 76 years, respectively) developed AF and met eligibility criteria. By pooled logistic regression, after adjustment for age, hypertension, smoking, diabetes, left ventricular hypertrophy, myocardial infarction, congestive heart failure, valvular heart disease, and stroke or transient ischemic attack, AF was associated with an OR for death of 1.5 (95% CI, 1.2 to 1.8) in men and 1.9 (95% CI, 1.5 to 2.2) in women. The risk of mortality conferred by AF did not significantly vary by age. However, there was a significant AF-sex interaction: AF diminished the female advantage in survival. In secondary multivariate analyses, in subjects free of valvular heart disease and preexisting cardiovascular disease, AF remained significantly associated with excess mortality, with about a doubling of mortality in both sexes. CONCLUSIONS In subjects from the original cohort of the Framingham Heart Study, AF was associated with a 1.5- to 1.9-fold mortality risk after adjustment for the preexisting cardiovascular conditions with which AF was related. The decreased survival seen with AF was present in men and women and across a wide range of ages.

Intermittent Claudication A Risk Profile From The Framingham Heart Study

BACKGROUND Intermittent claudication identifies persons at increased risk for death and disability. METHODS AND RESULTS Using 38-year follow-up data for the original cohort in the Framingham Heart Study, we developed an intermittent claudication risk profile. Intermittent claudication occurred in a total of 381 men and women. Age, sex, serum cholesterol, hypertension, cigarette smoking, diabetes, and coronary heart disease were associated with an increased risk for claudication and were included in the profile. A pooled logistic regression model was used to compute the probability of intermittent claudication for specified levels of risk factors. CONCLUSIONS The intermittent claudication risk profile allows physicians to identify high-risk individuals during a routine office visit and can be used to educate patients about modifiable risk factors, particularly smoking and blood pressure. Improved compliance with risk factor modification strategies may result in a beneficial impact on survival.

Nonfasting plasma total homocysteine levels and stroke incidence in elderly persons: the Framingham Study.

Meta-analyses (1, 2) strongly suggest that mildly to moderately elevated circulating levels of the sulfur amino acid homocysteine, either fasting or nonfasting, confer an independent risk for clinical arteriosclerotic outcomes, including stroke. However, just four reports (3-6) have provided somewhat limited, conflicting prospective data on the potential relation between total homocysteine levels and stroke incidence. Only one of these studies (3) used a population-based sample. Samples from two studies were limited to middle-aged men (4, 5): One consisted of U.S. physicians responding to a survey (4), and the other was derived from family practice registries in Great Britain (5). The fourth study (6) was a longitudinal investigation that primarily included women with systemic lupus erythematosus (mean age at baseline, 35 years). No prospective studies have specifically evaluated mild hyperhomocysteinemia as a potential risk factor for stroke in men and women 60 years of age or older, an age group that experiences the most pronounced morbidity and mortality from cerebrovascular disease (7). Accordingly, we examined the association between baseline nonfasting plasma total homocysteine levels and incident stroke in a well-characterized, population-based cohort of elderly women and men who at baseline had not had stroke. Methods The study sample consisted of the original Framingham Study cohort (8). Baseline examinations for the current analyses took place between May 1979 and May 1982, with follow-up occurring through May 1992. Of 2351 persons examined during the baseline period, 1947 had not previously had stroke and had specimens available for measurement of plasma total homocysteine levels. Additional baseline covariables assessed for the current analyses were age, sex, cigarette smoking, diabetes, history of atrial fibrillation, history of coronary heart disease, systolic blood pressure, and creatinine levels. Detailed operational definitions for all these covariables are provided elsewhere (8). Stroke outcome ascertainment and definition methods used in the Framingham Study, including subtype classification, have been described in detail previously (8, 9). Total homocysteine levels were determined by high-performance liquid chromatography with fluorescence detection (10). Nonfasting plasma aliquots were stored at 20 C from the baseline examination period until mid-1997. Data from long-term storage studies conducted at 20 C have confirmed both the biochemical stability and long-term within-person reproducibility of total homocysteine determinations (10). Creatinine levels were measured in nonfasting plasma by the Jaffe method, adapted for autoanalyzers. The skewed total homocysteine data were natural log-transformed, and differences in geometric mean total homocysteine levels according to sex, diabetes, history of atrial fibrillation or coronary heart disease, and smoking status were compared by using unpaired t-tests. The Spearman rho was used to assess unadjusted rank-order correlations between untransformed total homocysteine levels and age, creatinine level, and systolic blood pressure. Unadjusted and adjusted (for age, sex, history of atrial fibrillation or coronary heart disease, diabetes, smoking, systolic blood pressure, and creatinine level) relative risk estimates (hazards ratios with 95% CIs) for total stroke, nonhemorrhagic stroke, and atherothrombotic brain infarction were generated by proportional hazards modeling. Total homocysteine level (natural log-transformed or expressed in quartiles) was the independent variable. All statistical analyses were performed by using SAS software, version 6.12 (SAS Institute, Inc., Cary, North Carolina). Results The study sample (n=1947) consisted of 1158 women (59.5%) and 789 men (40.5%). Four hundred thirteen patients (21.2%) were cigarette smokers, 340 (17.5%) had a history of coronary heart disease, 182 (9.3%) were diabetic, and 81 (4.2%) had a history of atrial fibrillation. Arithmetic means (SD) for key continuous variables were as follows: age, 70 7 years (range, 59 to 91 years); systolic blood pressure, 141 21 mm Hg (range, 86 to 225 mm Hg); creatinine level, 97.2 26.5 mol/L (range, 35.3 to 442 mol/L); and total homocysteine level, 12.65 7.19 mol/L (range, 4.13 to 219.84 mol/L). Geometric mean total homocysteine levels were higher in men than in women (12.35 compared with 11.32 mol/L; P<0.001), in patients with a history of atrial fibrillation than in those without (13.17 compared with 11.66 mol/L; P=0.003), and in patients with a history of coronary heart disease than in those without (12.49 compared with 11.57 mol/L; P<0.001). However, these levels did not differ according to the presence or absence of diabetes (11.83 compared with 11.71 mol/L; P>0.2) or between current cigarette smokers and nonsmokers (11.67 compared with 11.74 mol/L; P>0.2). Weak but significant Spearman correlations were observed between total homocysteine levels and age (r=0.212; P<0.001), creatinine level (r=0.178; P<0.001), and systolic blood pressure (r=0.111; P<0.001). Quartiles of total homocysteine were as follows: quartile 1, 4.13 to 9.25 mol/L; quartile 2, 9.26 to 11.43 mol/L; quartile 3, 11.44 to 14.23 mol/L; quartile 4, 14.24 to 219.84 mol/L. During a median follow-up of 9.9 years, 165 incident total strokes occurred; 153 of these were incident nonhemorrhagic strokes, and 100 were incident atherothrombotic brain infarctions. Five hundred twenty-four persons were censored during follow-up because of death from causes other than stroke. Age, systolic blood pressure, current smoking, diabetes, and history of atrial fibrillation or coronary heart disease were independently predictive of total stroke occurrence (Table 1). Levels of total homocysteine (natural log) as a continuous variable (data not shown) and across quartiles (P<0.001 for linear trend) were associated with all stroke outcomes in unadjusted and multivariable-adjusted proportional hazards analyses (Tables 1 and 2). The interaction term between sex and total homocysteine level (quartile analyses) was nonsignificant (P=0.1); thus, the stroke incidence analyses were not stratified by sex. Further adjustment for creatinine level did not change the results of any of these analyses (data not shown). Table 1. Predictors of Incident Total Stroke in Elderly Women and Men in the Framingham Study Cohort Table 2. Relative Risk Estimates in Elderly Women and Men in the Framingham Study Cohort: Comparison of Each of the Upper Three Quartiles to the Lowest Quartile of Nonfasting Plasma Total Homocysteine Level Discussion Our findings are consistent with previously reported data, derived primarily from elderly female and male participants in the Framingham Study (9, 11), indicating that age, systolic blood pressure, diabetes, cigarette smoking, and history of atrial fibrillation or coronary heart disease were independently predictive of stroke incidence. We report population-based evidence that elevated nonfasting total homocysteine levels are also independently associated with stroke incidence among elderly women and men. Four earlier studies (3-6) have examined the relation between total homocysteine levels and stroke incidence. Alfthan and colleagues (3) did not find an association between total homocysteine levels and incident stroke among Finnish men and women 40 to 64 years of age (total events, 76). Similarly, Verhoef and coworkers (4) found only a weak, nonsignificant association between total homocysteine level and stroke incidence (total events, 109) in a cohort of male physicians whose mean age was 59.7 years (upper quintile compared with lower four quintiles: odds ratio, 1.2 [CI, 0.7 to 2.0]). In contrast, Perry and colleagues (5) reported a robust, independent association between total homocysteine levels (across quartiles) and incident stroke (total events, 107) among British men whose mean age was 54.0 years (quartile 3 compared with quartile 1: odds ratio, 3.3 [CI, 0.9 to 11.5]; quartile 4 compared with quartile 1: odds ratio, 7.4 [CI, 1.9 to 29.0]). More recently, elevated total homocysteine levels were independently linked to the development of stroke outcomes in a cohort of predominantly younger women (mean age, 34.9 years) with systemic lupus erythematosus (6). The two negative studies (3, 4) were characterized by modest stroke event rates and overall exposure to lower total homocysteine levels on the basis of sound nutritional (4) or, possibly, favorable genetic (3) influences. More widespread exposure to elevated total homocysteine levels, due perhaps to worse nutritional status, may have accounted for the strong association between total homocysteine level and incident stroke reported by Perry and colleagues (5). The positive relation between total homocysteine level and stroke occurrence described by Petri and associates in the Hopkins Lupus Cohort (6) was confined to patients with the highest total homocysteine levels (>14 mol/L). Our analyses, which revealed an independent association between total homocysteine level and incident stroke (total events, 165), were performed in an elderly sample characterized, as expected, by a relatively higher stroke event rate (7, 9, 11, 12). This higher rate occurred in conjunction with an increased prevalence of mild hyperhomocysteinemia (that is, in approximately 25% of patients with total homocysteine level>14 mol/L) at the baseline examination. Despite the lack of substantiation by either proven or biologically plausible mechanisms, it has nevertheless been proposed that hyperhomocysteinemia is an epiphenomenon of clinical or even subclinical arteriosclerosis (13, 14). This hypothesis appears untenable in view of the following published findings from both human and animal studies. First, despite the absence of any traditional arteriosclerotic risk factors, 50% of untreated children and young adults with homocystinuria due to cystathionine synthase deficiency ex

NIDDM and Blood Pressure as Risk Factors for Poor Cognitive Performance: The Framingham Study

OBJECTIVE To determine if NIDDM and blood pressure are risk factors for poor cognitive performance and if history and duration of NIDDM and blood pressure interact such that the risk of poor performance is greater for subjects with both NIDDM and hypertension. RESEARCH DESIGN AND METHODS We used a large prospective cohort sample with 187 NIDDM subjects and 1,624 nondiabetic subjects who were followed for 28–30 years. Cognitive function was assessed using eight tests of learning, memory, visual organization, verbal fluency attention, concept formation, and abstract reasoning. A composite score was also calculated. Odds ratios were used to estimate the relative risk of performing below the lower 25th percentile of z scores on these tests. RESULTS NIDDM and blood pressure interacted such that diagnosis and duration of NIDDM were associated with greater risk of poor performance on tests of visual memory and on the composite score for hypertensive subjects. Duration of NIDDM was associated with increased risk for poor performance on tests of verbal memory and concept formation. Insulin-treated NIDDM subjects were at higher risk for poor cognitive performance than those NIDDM subjects treated with oral agents or diet. Blood pressure level was associated independently with a measure of verbal fluency. CONCLUSIONS History and duration of NIDDM and high blood pressure are significant risk factors for poor cognitive performance. Hypertensive people with NIDDM are at greatest risk for poor performance on tests measuring visual organization and memory.

Trends in the Prevalence of Hypertension, Antihypertensive Therapy, and Left Ventricular Hypertrophy from 1950 to 1989

BACKGROUND Men and women with hypertension are at increased risk for cardiovascular disease, especially when left ventricular hypertrophy is present. We examined temporal trends in the use of antihypertensive medications and studied the relation between their use, the prevalence of high blood pressure, and the presence of electrocardiographic evidence of left ventricular hypertrophy. METHODS A total of 10,333 participants in the Framingham Heart Study who were 45 to 74 years of age underwent a total of 51,756 examinations from 1950 to 1989. Data were obtained on blood pressure and the use of antihypertensive medications, and electrocardiograms were assessed for left ventricular hypertrophy. The generalized-estimating-equation method was used to test for trends over time. RESULTS From 1950 to 1989, the rate of use of antihypertensive medications increased from 2.3 percent to 24.6 percent among men and from 5.7 percent to 27.7 percent among women. The age-adjusted prevalence of systolic blood pressure of at least 160 mm Hg or diastolic blood pressure of at least 100 mm Hg declined from 18.5 percent to 9.2 percent among men and from 28.0 percent to 7.7 percent among women. This decline was accompanied by age-adjusted reductions in the prevalence of electrocardiographic evidence of left ventricular hypertrophy, from 4.5 percent to 2.5 percent among men and from 3.6 percent to 1.1 percent among women. CONCLUSIONS Our findings support the notion that the increasing use of antihypertensive medication has resulted in a reduced prevalence of high blood pressure and a concomitant decline in left ventricular hypertrophy in the general population. Our observations may in part explain the considerable decline in mortality from cardiovascular disease observed since the late 1960s.

CUMULATIVE EFFECTS OF HIGH CHOLESTEROL LEVELS, HIGH BLOOD PRESSURE, AND CIGARETTE SMOKING ON CAROTID STENOSIS

BACKGROUND Single measurements of cardiovascular risk factors may not accurately reflect a persons past exposure to those risk factors. We therefore studied the long-term associations of cardiovascular risk factors such as high serum cholesterol levels, high blood pressure, and cigarette smoking with the prevalence of carotid stenosis. METHODS We studied cross-sectional and longitudinal information from a sample of 429 men and 661 women in the Framingham Heart Study who underwent B-mode ultrasound measurements of the carotid artery. Their mean age was 75 years, and each had attended most of the biennial clinic examinations over the 34 years before the carotid ultrasound study. We used time-integrated measurements to assess the associations between various cardiovascular risk factors and the degree of carotid stenosis. RESULTS Moderate carotid stenosis (> or =25 percent) was present in 189 men and 226 women. We assessed the odds ratios for this degree of stenosis as compared with minimal stenosis (<25 percent) according to increases in risk factors. In the men, the odds ratio for moderate carotid stenosis associated with an increase of 20 mm Hg in systolic blood pressure was 2.11 (95 percent confidence interval, 1.51 to 2.97). The odds ratio for an increase of 10 mg per deciliter (0.26 mmol per liter) in the cholesterol level was 1.10 (95 percent confidence interval, 1.03 to 1.16), and for an increase of five pack-years of smoking it was 1.08 (95 percent confidence interval, 1.03 to 1.13). The results were similar in the women. Time-integrated measurements of diastolic blood pressure showed significant associations with carotid stenosis in men and insignificant associations in women. CONCLUSIONS Over the long term, high systolic blood pressure, high cholesterol levels, and smoking were associated with an increased risk of carotid stenosis in this elderly population.

Association of Fibrinogen With Cardiovascular Risk Factors and Cardiovascular Disease in the Framingham Offspring Population

BackgroundFibrinogen has been identified as an independent risk factor for cardiovascular disease and associated with traditional cardiovascular risk factors. Also, the role of elevated fibrinogen in thrombosis suggests that it may be on the causal pathway for certain risk factors to exert their effect. These associations remain incompletely characterized. Moreover, the optimal fibrinogen assay for risk stratification is uncertain. Methods and ResultsIn 2632 subjects from cycle 5 of the Framingham Offspring Population, fibrinogen levels were determined with a newly developed immunoprecipitation test (American Biogenetic Sciences) and the functional Clauss method. With the immunoprecipitation method, there were significant linear trends across fibrinogen tertiles (P <0.001) for age, body mass index, smoking, diabetes mellitus, total cholesterol, HDL cholesterol, and triglycerides in men and women. The Clauss method had significant results (P <0.030), except for triglycerides in men. Fibrinogen levels were higher for those with compared with those without cardiovascular disease. After covariate adjustment, fibrinogen remained significantly higher in those with cardiovascular disease with the use of the immunoprecipitation test (P =0.035 and P =0.018 for men and women, respectively) but not with the Clauss method. ConclusionsFibrinogen was associated with traditional cardiovascular risk factors. Elevation of fibrinogen may provide a mechanism for risk factors to exert their effect. Also, fibrinogen levels were higher among subjects with cardiovascular disease compared with those without disease. The immunoprecipitation test showed a stronger association with cardiovascular disease than the Clauss method, suggesting that it may be a useful screening tool to identify individuals at increased thrombotic risk.

Increased Platelet Aggregability Associated With Platelet GPIIIa PlA2 Polymorphism : The Framingham Offspring Study

The platelet glycoprotein IIb/IIIa (GP IIb/IIIa) plays a pivotal role in platelet aggregation. Recent data suggest that the Pl A2 polymorphism of GPIIIa may be associated with an increased risk for cardiovascular disease. However, it is unknown if there is any association between this polymorphism and platelet reactivity. We determined GPIIIa genotype and platelet reactivity phenotype data in 1422 subjects from the Framingham Offspring Study. Genotyping was performed using PCR based restriction fragment length polymorphism analysis. Platelet aggregability was evaluated by the Born method. The threshold concentrations of epinephrine and adenosine diphosphate (ADP) were determined. Allele frequencies of Pl A1 and Pl A2 were 0.84 and 0.16, respectively. The presence of one or two Pl A2 alleles was associated with increased platelet aggregability as indicated by incremen- tally lower threshold concentrations for epinephrine and ADP. For epinephrine, the mean concentrations were 0.9 µmol/L (0.9–1.0) for homozygous Pl A1 , 0.7 µmol/L (0.7–0.9) for the heterozygous Pl A1 / Pl A2 and 0.6 µmol/L (0.4–1.0) for homozygous Pl A2 individuals, p = 0.009. The increase in aggregability induced by epinephrine remained highly significant (p = 0.007) after adjustment for covariates. For ADP-induced aggregation, the respective mean concentrations were 3.1 µmol/L (3.0–3.2), 3.0 µmol/L (2.9–3.2), and 2.8 µmol/L (2.4–3.3), p = 0.19 after adjustment for covariates. Our findings indicate that molecular variants of the gene encoding GPIIIa play a role in platelet reactivity in vitro. Our observations are compatible with and provide an explanation for the reported association of the Pl A2 allotype with increased risk for cardiovascular disease.

Increased Platelet Aggregability Associated With Platelet GPIIIa PlA2 Polymorphism

The platelet glycoprotein IIb/IIIa (GP IIb/IIIa) plays a pivotal role in platelet aggregation. Recent data suggest that the PlA2 polymorphism of GPIIIa may be associated with an increased risk for cardiovascular disease. However, it is unknown if there is any association between this polymorphism and platelet reactivity. We determined GP IIIa genotype and platelet reactivity phenotype data in 1422 subjects from the Framingham Offspring Study. Genotyping was performed using PCR-based restriction fragment length polymorphism analysis. Platelet aggregability was evaluated by the Born method. The threshold concentrations of epinephrine and ADP were determined. Allele frequencies of PlA1 and PlA2 were 0.84 and 0.16, respectively. The presence of 1 or 2 PlA2 alleles was associated with increased platelet aggregability as indicated by incrementally lower threshold concentrations for epinephrine and ADP. For epinephrine, the mean concentrations were 0.9 micromol/L (0.9 to 1.0) for homozygous PlA1, 0.7 mmol/L (0.7 to 0.9) for the heterozygous PlA1/PlA2, and 0.6 micromol/L (0.4 to 1.0) for homozygous PlA2 individuals, P=0.009. The increase in aggregability induced by epinephrine remained highly significant (P=0.007) after adjustment for covariates. For ADP-induced aggregation, the respective mean concentrations were 3.1 micromol/L (3.0 to 3.2), 3.0 micromol/L (2.9 to 3.2), and 2.8 micromol/L (2.4 to 3.3); P=0.19 after adjustment for covariates. Our findings indicate that molecular variants of the gene encoding GP IIIa play a role in platelet reactivity in vitro. Our observations are compatible with and provide an explanation for the reported association of the PlA2 allotype with increased risk for cardiovascular disease.