An overview of trials of cholesterol lowering and risk of stroke

BACKGROUND

While blood cholesterol level predicts coronary heart disease, whether there is any association with the risk of stroke is unclear. Some, but not all, observational studies suggest that cholesterol level predicts risk of stroke, particularly ischemic stroke. This hypothesis is attractive because ischemic events constitute the vast majority of all strokes and, like coronary heart disease, involve atherogenic processes.

METHODS

To investigate whether lipid lowering reduces the risk of stroke, we performed an overview of randomized trials that included more than 36,000 individuals.

RESULTS

The mean reduction in cholesterol level in the treated as compared with the control subjects ranged from 6% to 23%. Those assigned to treatment experienced no significant reduction in all (fatal plus nonfatal) stroke (relative risk, 1.0; 95% confidence interval, 0.8 to 1.2) or fatal stroke (1.1; 0.8 to 1.6).

CONCLUSIONS

The confidence interval for fatal stroke is wide, and alternative hypotheses, including either a small protective or harmful effect, cannot be excluded; however, the point estimates are compatible with no benefit of cholesterol lowering on the risk of stroke. Additional large-scale randomized trials assessing total mortality would more definitively address any benefits on stroke, as well as any excess nonvascular causes of mortality, for which risks of cholesterol lowering also remain uncertain.

Intermittent claudication and subsequent cardiovascular disease in the elderly

BACKGROUND

This study reports the prevalence of intermittent claudication (IC) in ambulatory community-resident adults age 65 years or older, compares cardiovascular risk factors and comorbidity of persons with and without IC, and examines the independent association of IC in predicting all cause and cardiovascular mortality, myocardial infarction, stroke, and disability.

METHODS

Data are from a pooled sample of 8996 older adults from the East Boston, New Haven, and Iowa sites of the Established Populations for Epidemiologic Studies of the Elderly, conducted between 1982 and 1988.

RESULTS

2.4% and 1.5% of men and women, respectively, reported IC. Persons with IC had significantly higher rates of diabetes and cardiovascular comorbidity than persons without IC, and they were more likely to smoke. Claudication predicted higher rates of mortality, myocardial infarction, stroke, and disability independent of associated cardiovascular conditions and risk factors. Among persons with a history of angina, myocardial infarction, and/or stroke, those who reported IC had a twofold greater risk of cardiovascular mortality.

CONCLUSION

The study demonstrated that IC is an important predictor of mortality and cardiovascular morbidity in ambulatory older adults independent of associated coronary ischemia and cardiovascular disease risk factors. Results suggest that inclusion of a measure of IC improves the prediction of cardiovascular morbidity and mortality in older adults.

beta-carotene and cancer chemoprevention

Evidence supports the potential role of beta-carotene in cancer prevention. Basic research has demonstrated that beta-carotene can trap organic free radicals and/or deactivate excited oxygen molecules which may have an anticancer effect by preventing tissue damage. Although observational epidemiologic studies are not entirely consistent, many show an inverse association between dietary intake or blood levels of beta-carotene and subsequent cancer risk. Two large-scale randomized trials of beta-carotene have been completed. A Finnish trial demonstrated no benefit of beta-carotene among middle-aged male smokers, with those assigned to this supplement in fact experiencing an increased risk of lung cancer. However, because of the long latency period for cancer, which may be a decade or more, the six-year duration of treatment in this trial may have been inadequate to detect an anticancer effect. A Chinese trial demonstrated a modest reduction in cancer mortality from a combined regimen of beta-carotene, vitamin E, and selenium. The effect of the individual agents could not be assessed, and because the trial was carried out among a nutritionally deficient population, its results may not have direct relevance to well-nourished individuals. Several additional large-scale trials of beta-carotene are ongoing. The Physicians' Health Study, which is testing beta-carotene among 22,071 US male physicians, will have an average duration of treatment of 12.5 years at its scheduled termination in late 1995. Data in women will be available from the Women's Health Study, which began in 1992, and will randomize approximately 40,000 US female health professionals.(ABSTRACT TRUNCATED AT 250 WORDS)

A prospective study of plasma hormone levels, nonhormonal factors, and development of benign prostatic hyperplasia

We assessed the relation of plasma hormone levels and nonhormonal factors with subsequent occurrence of surgical treatment for benign prostatic hyperplasia (BPH) among participants in the Physicians' Health Study. Frozen plasma samples, collected at the study onset, were available for 320 men who developed surgically treated BPH up to 9 years later and for 320 age-matched controls. Plasma testosterone (T), dihydrotestosterone (DHT), androstenedione, estradiol (E2), and estrone (E1) were measured for each case-control pair. In unadjusted analyses, none of the hormones or hormone ratios were associated with BPH; for example, for T and E2 the odds ratios (OR) comparing the highest quintile (Q5) with the lowest (Q1) were 0.74 (95% CI = 0.42, 1.30) and 1.07 (95% CI = 0.51, 2.22), respectively. However, in multivariate analyses controlling diastolic blood pressure, exercise, alcohol, E1, and DHT:T ratio, we observed a strong trend for increasing risk across quintiles for E2 (Q5 vs. Q1 OR = 3.56, P trend = 0.009), and a weak inverse trend for E1 (Q5 vs Q1 OR = 0.51, P trend = 0.07). The excess risk associated with E2 was confined to men with relatively low androgen levels. Three nonhormonal factors previously suspected as risk factors were independently associated with surgical BPH in these data. The OR for a 1-mm Hg difference in diastolic blood pressure was 1.04 (95% CI = 1.01, 1.07). Alcohol use and infrequent exercise were inversely associated with risk of BPH surgery; however, risk estimates were not consistent across categories of exercise and alcohol frequency. Our results indicate that normal variation in circulating androgen levels does not influence development of BPH, but that variation in estrogen levels might be important.

Adherence to aspirin in the prevention of myocardial infarction. The Physicians' Health Study

BACKGROUND

The primary aim of this article was to explore, in subgroup analyses, whether participants with differing frequencies of aspirin consumption in a randomized, double-blind, placebo-controlled, primary prevention trial had different magnitudes of benefit in the prevention of myocardial infarction. Secondary aims were to identify factors associated with adherence and to examine the relationship of adherence with cardiovascular outcomes in the placebo group.

METHODS

The Physicians' Health Study randomized 22071 US male physicians who were free of myocardial infarction and cerebrovascular disease at baseline. The average follow-up during the aspirin component of the trial was 60.2 months. Baseline cardiovascular risk factors and adherence to therapy during the trial were assessed by questionnaire; cardiovascular outcomes were reported by questionnaire and confirmed by record review by an Endpoints Committee.

RESULTS

Several cardiovascular disease risk factors assessed at baseline were related to poor adherence (taking < 50% of study tablets): cigarette smoking, obesity, lack of exercise, and history of angina. After adjusting for baseline differences in risk factors, participants in the aspirin group with excellent adherence (taking at least 95% of study tablets) had a statistically significant 51% reduction in myocardial infarction compared with those with excellent adherence in the placebo group. Those in the aspirin group with poor adherence had a smaller, non-significant reduction in risk of myocardial infarction (a 17% reduction associated with taking < 50% of study tablets). In the placebo group better adherence was not associated with decreased risk of myocardial infarction, but was strongly associated with decreased risk of death.

CONCLUSIONS

These subgroup data raise the possibility that a less than alternate day aspirin regimen may yield lower benefits in the prevention of myocardial infarction. Alternate explanations are that these analyses reflect either the play of chance or effects of uncontrolled confounding since comparisons were no longer randomized. Randomized trials are necessary to address the question of frequency of administration of aspirin to achieve optimal benefits in primary prevention of myocardial infarction.

Platelet inhibitors and antioxidant vitamins in cardiovascular disease

Considerable research attention has focused on the possible roles of platelet inhibition, principally using aspirin, and antioxidant vitamins in reducing the risks of cardiovascular disease. Data from large-scale randomized trials indicate that aspirin reduces subsequent vascular events among patients with prior myocardial infarction, stroke, transient ischemic attacks, or unstable angina, as well as among patients with acute evolving myocardial infarction. In primary prevention trials, the Physicians' Health Study showed a clear benefit in decreasing risk of a first myocardial infarction in men; the data on stroke and total number of deaths from vascular causes are inadequate. The Women's Health Study, a trial now under way among apparently healthy women, will provide direct evidence on the balance of risks and benefits of aspirin in primary prevention. Antioxidant vitamins are hypothesized to decrease cardiovascular disease risk by several mechanisms, including inhibition of oxidation of low-density lipoprotein cholesterol and decreasing uptake into the coronary endothelium. Promising results have emerged from observational studies, which show that people with high intakes of antioxidant vitamins through diet or supplements have lowered risks of cardiovascular disease; however, unknown or unmeasured factors associated with high antioxidant vitamin intake may explain all or part of the observed associations. Randomized trials to provide reliable data are now ongoing among apparently healthy men and women, as well as among survivors of prior cardiovascular disease events.

Risk factors for non-insulin-dependent diabetes mellitus requiring treatment in the elderly

OBJECTIVE

To examine the relationship of possible modifiable risk factors, including obesity, physical activity level, alcohol consumption, blood pressure, and thiazide diuretic use with the development of non-insulin-dependent diabetes mellitus (NIDDM) requiring treatment among a large cohort of community-dwelling elderly.

SETTING

The East Boston Senior Health Project, one of four components of the National Institute on Aging-sponsored Established Populations for the Epidemiologic Study of the Elderly (EPESE).

PARTICIPANTS

Residents of East Boston who were 65 years of age or older.

MEASUREMENTS

We performed a prospective cohort study with follow-up over two consecutive 3-year time periods beginning in 1982-1983. The main outcome measure was the occurrence of NIDDM, defined as new treatment with a hypoglycemic agent. A total of 2737 study participants contributed 4682 3-year intervals for analysis.

MAIN RESULTS

NIDDM requiring hypoglycemic therapy occurred in 185 participants over the duration of the study. High body mass index (> 26 kg/m2) (adjusted odds ratio 2.4, 95% confidence interval 1.3-4.4) and low physical activity level (adjusted odds ratio 1.5, 95% confidence interval 1.0-2.1) were significant predictors of NIDDM in a multiple logistic regression model adjusting for age, sex, blood pressure, and self-report of "high blood sugar" moderate alcohol consumption (0.5-<1 ounce per day) had an inverse relation to NIDDM of borderline significance (adjusted odds ratio 0.4, 95% confidence interval 0.2-1.0). Those receiving one or more non-thiazide antihypertensive agents had a higher risk of developing NIDDM in a model including age, sex, body mass index, various antihypertensive regimens, physical activity level, alcohol consumption, blood pressure, and self-report of "high blood sugar." Thiazide diuretic therapy alone or in combination with another antihypertensive was not associated with NIDDM.

CONCLUSIONS

Our findings suggest a positive relationship of obesity and low physical activity level with the development of NIDDM requiring treatment in elderly persons. The inverse association of borderline significance between moderate alcohol use and NIDDM deserves further study. Thiazide diuretic therapy conferred no excess risk for developing NIDDM in this older population although selection factors in the choice of antihypertensive therapy may partially explain the absence of a thiazide effect.

Dehydroepiandrosterone sulfate and the risk of myocardial infarction in US male physicians: a prospective study

High levels of dehydroepiandrosterone sulfate (DHEAS) have been associated with decreased risks of cardiovascular disease. The authors analyzed DHEAS in plasma collected at baseline among 169 participants in the Physicians' Health Study who subsequently had a myocardial infarction and 169 matched controls. The mean prediagnostic plasma DHEAS levels between cases (p = 0.33) (mean, 3.54 mumol/liter; standard deviation, 2.30) and controls (mean, 3.61 mumol/liter; standard deviation, 2.16) did not differ significantly. The relative risk was 1.04 (95 percent confidence interval 0.42-2.60) comparing extreme quintiles after adjustment for several coronary risk factors. In conclusion, these findings do not support the hypothesis that elevated plasma DHEAS is associated with a decreased risk of coronary disease in men, but a small to moderate association cannot be excluded.

Plasma concentration of cross-linked fibrin degradation product (D-dimer) and the risk of future myocardial infarction among apparently healthy men

BACKGROUND

Plasma levels of D-dimer, the primary degradation product of cross-linked fibrin, are elevated in several acute thrombotic disorders. However, whether elevated D-dimer levels among healthy individuals are associated with future coronary thrombosis is unknown.

METHODS AND RESULTS

To evaluate whether levels of D-dimer are associated with the occurrence of future myocardial infarction (MI) among apparently healthy men, levels were measured in plasma samples collected at baseline from 296 participants in the Physicians' Health Study who later developed a first MI and from an equal number of age- and smoking status-matched control subjects who remained free of vascular disease during a mean follow-up period of 60.2 months. In univariate analyses, baseline plasma concentrations of D-dimer in the upper ranges of normal were associated with elevated risks of MI. Specifically, the relative risk of future MI for individuals with baseline D-dimer concentration exceeding the 95th percentile of the control distribution was two times higher than that of individuals with lower levels (relative risk [RR], 2.02; 95% confidence interval [CI], 1.04 to 4.02; P = .04). This association persisted in multivariate analyses controlling for nonlipid cardiovascular risk factors (RR, 2.12; 95% CI, 1.05 to 4.28; P = .04) and for lipoprotein(a) (RR, 2.02; 95% CI, 1.04 to 3.94; P = .03). In contrast, this association was attenuated and no longer statistically significant in analyses that controlled for total and high-density lipoprotein cholesterol (RR, 1.74; 95% CI, 0.78 to 3.91; P = .2) or for endogenous tissue-type plasminogen activator and its primary inhibitor, plasminogen activator inhibitor type 1 (RR, 1.58; 95% CI, 0.67 to 3.77; P = .3).

CONCLUSIONS

Elevated levels of D-dimer are associated with increased risks of future MI, although they do not appear to be an independent predictor when other risk factors are considered. As the presence of D-dimer in plasma reflects ongoing fibrin degradation, these data support the hypothesis that activation of the endogenous fibrinolytic system occurs many years in advance of coronary arterial occlusion.

Contributions of observational evidence and clinical trials in cancer prevention

Observational studies and randomized trials provide relevant and complementary information to a totality of evidence on which to base rational clinical decision making for patients and overall health policy for the general population. Observational studies are particularly useful for detecting moderate to large effects. The 15- to 20-fold greater risk of lung cancer among long term cigarette smokers was established by case-control and prospective cohort studies. The approximate 80% increased risk of coronary heart disease associated with current smoking also has been reliably demonstrated in observational studies. However, as the relative risk gets smaller, there is increasing concern that unmeasured or unknown confounding variables may account for all or part of any observed association. For these reasons, reliable inferences about interventions likely to confer small to moderate benefits will emerge only from randomized trials of sufficient sample size and duration of treatment and follow-up. Dietary variables have been postulated to account for as much as 35% of all human cancers. However, the hypothesized benefit of any specific dietary constituent, such as the antioxidant beta-carotene, is likely to be modest in size, on the order of a 20-30% reduction in risk. Therefore, although a large number of observational studies have demonstrated that individuals with higher dietary intakes or blood levels of beta-carotene have lower risks of cancer, only randomized trials can address this hypothesis definitively. Such trials, however, must be of sufficient duration to allow for the development of an anticancer effect. This may mean a decade or more based on the analogy with smoking cessation and decreased risks of lung cancer. Several ongoing large-scale trials are testing beta-carotene and other promising cancer chemoprevention agents, and their results will provide clear evidence on the balance of benefits and risks of these interventions.

A prospective study of permanent hair dye use and hematopoietic cancer

BACKGROUND

Use of permanent hair dye has been suggested as a risk factor for several types of cancer, although epidemiologic data have not generally supported this hypothesis. Retrospective studies have reported a possible association between hair dyes and hematopoietic cancers.

PURPOSE

Our purpose was to investigate if permanent hair dye was associated with risks of incident lymphoma, leukemia, and multiple myeloma in the Nurses' Health Study, a prospective cohort study of 99,067 women aged 30-55 years in 1976.

METHODS

Questionnaires regarding medical history and other health-related variables were sent to Nurses' Health Study participants every 2 years from 1976 to 1990. The follow-up for mortality in this cohort exceeds 98%. We identified 244 newly diagnosed cases of hematopoietic cancers, confirmed by pathology reports. Permanent hair dye use was ascertained over four cycles of questionnaires from 1976-1982; status of hair dye use established in 1982 was then used for the remainder of the follow-up time (through 1990). Age-specific incidence rates were calculated and used to compute relative risks (RRs) with 95% confidence intervals (CIs).

RESULTS

We found no evidence of a positive association between ever use of permanent hair dye and all hematopoietic cancers (age-adjusted RR = 0.9; 95% CI = 0.7-1.2) or specific types (Hodgkin's lymphoma [RR = 0.9; 95% CI = 0.4-2.1], non-Hodgkin's lymphoma [RR = 1.1; 95% CI = 0.8-1.6], multiple myeloma [RR = 0.4; 95% CI = 0.2-0.9], chronic lymphocytic leukemia [RR = 0.6; 95% CI = 0.3-1.5], and other leukemias [RR = 0.8; 95% CI = 0.3-1.9]). Further examination of age at first use, duration, frequency, and time since first use and risk of all hematopoietic cancers or non-Hodgkin's lymphoma (the largest diagnostic group), indicated no material associations.

CONCLUSION

In this prospective cohort study, permanent hair dye use is not adversely related to risks of hematopoietic cancers.

A prospective study of plasma homocyst(e)ine and risk of ischemic stroke

BACKGROUND AND PURPOSE

Several studies have reported elevated circulating homocyst(e)ine levels in subjects with cerebral atherosclerosis. We assessed prospectively whether high plasma levels of homocyst(e)ine affect risk of ischemic stroke and evaluated whether high blood pressure modifies any such effect.

METHODS

The study sample was drawn from the Physicians' Health Study, a randomized, double-blind, placebo-controlled trial of aspirin and beta-carotene in 22,071 US male physicians. A total of 14,916 subjects 40 to 84 years old with no prior history of stroke, transient ischemic attack, or myocardial infarction provided blood samples at baseline and were followed for 5 years, with 99.7% morbidity and 100% mortality follow-up. Using a nested case-control design, we assayed homocyst(e)ine in samples from 109 subjects who subsequently developed ischemic stroke and 427 control subjects.

RESULTS

The mean plasma concentration of homocyst(e)ine was slightly higher in subjects with stroke (11.1 +/- 4.0 [+/- SD] nmol/mL) than in control subjects (10.6 +/- 3.4 nmol/mL), but the difference was not statistically significant (P = .12). The crude odds ratio of ischemic stroke for subjects in the upper 20% (> 12.7 nmol/mL) compared with those in the bottom 80% of homocyst(e)ine levels was 1.4 (95% confidence interval, 0.8 to 2.2). The odds ratio was 1.2 (95% confidence interval, 0.7 to 2.0) after controlling for several risk factors and other potential confounders. In subgroup analyses, elevated homocyst(e)ine levels appeared to be more strongly predictive of ischemic stroke in normotensive subjects and in men 60 years or younger. Although not statistically significant, in these subgroups increases in risks of 100% and 70%, respectively, were observed for men in the upper 20% of homocyst(e)ine values.

CONCLUSIONS

In this study, the data were compatible with a small but nonsignificant association between elevated plasma homocyst(e)ine and risk of ischemic stroke. However, since the sample size is small and the confidence intervals are wide, either no association or a moderate increase in risk cannot be excluded, particularly in subgroups otherwise at low risk, eg, younger men and those with normal blood pressure.

Association of moderate alcohol consumption and plasma concentration of endogenous tissue-type plasminogen activator

OBJECTIVE

To assess whether an association exists between moderate alcohol consumption and plasma concentration of endogenous tissue-type plasminogen activator (t-PA), a serine protease that plays a central role in the regulation of intravascular fibrinolysis.

DESIGN

Survey of self-reported alcohol consumption and plasma fibrinolytic capacity, controlled for lipid and nonlipid cardiac risk factors.

SETTING

Participants in the Physicians' Health Study.

PARTICIPANTS

A total of 631 apparently healthy male physicians aged 40 to 84 years with no history of myocardial infarction, stroke, or transient cerebral ischemia.

MAIN OUTCOME MEASURE

Plasma concentration of t-PA antigen.

RESULTS

A direct association was found between alcohol consumption and plasma level of t-PA antigen, such that mean plasma levels of t-PA antigen for daily, weekly, monthly, and rare or never drinkers were 10.9, 9.7, 9.1, and 8.1 ng/mL, respectively (P trend = .0002). The relation between alcohol consumption and t-PA antigen level was not materially changed in analyses that adjusted for total cholesterol and high-density lipoprotein cholesterol or nonlipid cardiovascular risk factors including age, body mass index, parental history of coronary heart disease, exercise frequency, and systolic and diastolic blood pressure.

CONCLUSIONS

These data indicate a positive association between moderate alcohol intake and plasma level of endogenous t-PA antigen that is independent of high-density lipoprotein cholesterol. This finding supports the hypothesis that changes in fibrinolytic potential may be an important mechanism whereby moderate alcohol consumption decreases risk of heart disease.

Antioxidant vitamins and cancer

Cancer is the second leading cause of death in the United States as well as most developed countries. Although advances in treatment have afforded tremendous benefits to large numbers of patients, effective primary prevention measures could, in theory, have an even greater public health impact. Diet has been postulated to account for as much as 35% of all human cancers, raising the possibility that specific constituents of diet with cancer chemopreventive effects could be identified. Risk reductions of even 20-30%, which is the most likely magnitude of any protective effect of nutritional agents, could have a significant public health impact on so common and serious a disease. Antioxidant vitamins are among the constituents of diet hypothesized to exert chemopreventive effects. Antioxidant vitamins, which include beta-carotene (pro-vitamin A), vitamin E, and vitamin C, are hypothesized to decrease cancer risk by preventing tissue damage by trapping organic free radicals and/or deactivating excited oxygen molecules, a by-product of many metabolic functions. Over 100 observational epidemiologic studies have assessed the relationship of dietary antioxidant intake or blood nutrient levels with cancer risk. Such studies are not entirely consistent but provide support for the hypothesis that antioxidant vitamin intake may decrease cancer risk. However, the chief limitation of such observational studies is their inability to control for all factors associated with vitamin intake that might independently affect cancer risk. Such unknown or unmeasured confounding variables could account for all or part of any observed associations. For this reason, definitive data on the role of antioxidant vitamins and cancer can derive only from properly conducted large-scale randomized trials of sufficient sample size, dose, and duration of treatment and follow-up. Two large-scale trials have been completed, and several are currently ongoing among well-nourished populations at high risk for cancer as well as among individuals at usual risk. Over the next several years, these trials should provide clear evidence concerning the role of antioxidant vitamins in the prevention of cancer. Such data are crucial for both individual clinical decision making as well as rational public health policy.

Aspirin in the primary prevention of cardiovascular disease

The ability of aspirin to reduce cardiovascular disease risks has been tested in randomized trials in a wide range of patient categories. There are clear benefits of aspirin on nonfatal myocardial infarction, nonfatal stroke, and vascular death among patients with prior manifestations of cardiovascular disease, such as myocardial infarction, unstable angina, and stroke. Aspirin is also beneficial to those in the acute phase of evolving myocardial infarction. In primary prevention, there is a clear reduction in myocardial infarction in men. A large-scale primary prevention trial in women is presently underway.

Confirmation of self-reported cataract in the Physicians' Health Study

Large-scale prospective studies of disease development often rely on self-reported data. To assess the accuracy of self-reports of cataract, we compared the self-reports with medical record data obtained from diagnosing ophthalmologists and optometrists for participants in the Physicians' Health Study, a randomized trial of aspirin and beta-carotene among 22,071 male U.S. physicians aged 40-84 years. A report of cataract, defined as a positive response to a question about whether cataract had ever been diagnosed in either eye and the date of diagnosis, was found to be a very good indicator of lens opacification but was not a good indicator of an incident, age-related opacity that reduced visual acuity. These results indicate that in large prospective studies of clinically significant cataract, where examination of all study participants is not feasible and self-reported data are used, additional documentation to supplement the self-reports should be obtained and strict diagnostic criteria applied to minimize the likely effects of misclassification.

Height, lung function, and mortality from cardiovascular disease among the elderly

The relation between height and death from cardiovascular disease was studied in a cohort of 3,809 persons aged 65 years or older (85% of eligible individuals) enrolled in a population survey in 1982-1983 in East Boston, Massachusetts. Self-reported height and weight were obtained, and peak expiratory flow rate (PEFR) was measured using a mini-Wright peak flow meter (Armstrong Industries, North Brook, Illinois). Vital status and cause of death were obtained through 1988. The median height was 62 inches in women and 66 inches in men. After adjustment for age, body mass index, and cigarette smoking, the risk of cardiovascular death decreased with quintile of height in women, with relative risks of 1.65, 1.16, 1.15, 0.76, and 1.00 over successive quintiles, with the tallest as the referent (p trend = 0.015). The trend in men was not as strong, with relative risks of 1.22, 0.77, 0.90, 0.98, and 1.00 from the shortest to the tallest quintiles (not significant). In both men and women, the strongest association was found with height and height squared, indicating a curvilinear relation. Height remained a predictor in women after adjustment for PEFR and other risk factors. These data suggest that a relation between height and cardiovascular death that is not mediated by lung function exists in the elderly, at least among women.

The use of vitamin supplements and the risk of cataract among US male physicians

OBJECTIVES

The purpose of this study was to examine prospectively the association between reported use of vitamin supplements and risk of cataract and cataract extraction.

METHODS

The study population consisted of 17,744 participants in the Physicians' Health Study, a randomized trial of aspirin therapy and beta-carotene among US male physicians 40 to 84 years of age in 1982 who did not report cataract at baseline and provided complete information about vitamin supplementation and other risk factors for cataract. Self-reports of cataract and cataract extraction were confirmed by medical record review.

RESULTS

During 60 months of follow-up, there were 370 incident cataracts and 109 cataract extractions. In comparison with physicians who did not use any supplements, those who took only multivitamins had a relative risk of cataract of 0.73 after adjustment for other risk factors. For cataract extraction, the corresponding relative risk was 0.79. Use of vitamin C and/or E supplements alone was not associated with a reduced risk of cataract, but the size of this subgroup was small.

CONCLUSIONS

These data suggest that men who took multivitamin supplements tended to experience a decreased risk of cataract and support the need for rigorous testing of this hypothesis in large-scale randomized trials in men and women.

A prospective study of alcohol consumption and risk of cataract

The role of alcohol as a determinant of age-related cataract is largely unexplored, although a possible influence has been suggested by previous retrospective and cross-sectional studies. We used the prospective data base of the Physicians' Health Study to examine the association between alcohol consumption and incidence of cataract as well as cataract extraction among U.S. male physicians. Participants in the Physicians' Health Study, a randomized trial of aspirin and beta-carotene among 22,071 male physicians 40-84 years of age at entry in 1982, were included in these analyses if they did not report cataract at baseline and if they provided information about alcohol consumption and other cataract risk factors. A total of 17,824 physicians satisfied these criteria. An incident cataract was defined as a self-report confirmed by medical record review, first diagnosed after randomization, with an age-related cause, and responsible for a reduction in best corrected visual acuity to 20/30 or worse. During 88,565 person-years of follow-up, 371 participants had a confirmed incident cataract and 110 underwent cataract extraction. Compared to physicians consuming alcohol less than once per month, daily consumers of alcohol had an age-adjusted relative risk (RR) of cataract of 1.31 (95% confidence interval [CI] = 0.95, 1.81). For posterior subcapsular (PSC) cataract, the most disabling subtype in terms of vision loss, the RR was 1.38 (95% CI = 0.84, 2.27); for PSC cataract extraction, the RR was 1.43 (95% CI = 0.71, 2.88).(ABSTRACT TRUNCATED AT 250 WORDS)

Prospective study of endogenous tissue plasminogen activator and risk of stroke

Few haematological or lipid risk factors have been identified for stroke, by contrast with coronary heart disease. To find out whether a marker of endogenous fibrinolytic function might be associated with stroke risk, we measured tissue plasminogen activator (tPA) antigen concentrations in baseline plasma samples from 88 healthy participants in the Physicians' Health Study who subsequently had first-ever strokes (71 thromboembolic, 12 haemorrhagic, 5 indeterminate) and from 471 participants who remained free of cardiovascular disease during 5 years of follow-up (controls). Mean baseline tPA concentrations were significantly higher among men who later had strokes than in the controls (11.14 [SE 0.80] vs 9.59 [0.27] ng/mL, p = 0.03). The difference was largely due to an excess of abnormally high values among stroke cases. The age-adjusted relative risk for stroke among men with baseline tPA concentrations above the 95th percentile of the control distribution was 3.51 (95% CI 1.72-7.17, p = 0.0006) for total stroke and 3.89 (1.83-8.26, p = 0.0004) for thromboembolic stroke. These findings did not change substantially in analyses that also controlled for stroke risk factors (high blood pressure, body-mass index, smoking, presence of diabetes, and parental history of myocardial infarction) or the plasma lipid profile. This prospective study shows that high concentrations of tPA antigen among apparently healthy men are independently associated with high risks of future stroke, especially thromboembolic stroke. This finding is consistent with the hypothesis that activation of the endogenous fibrinolytic system occurs years in advance of arterial vascular occlusion.