Cholesterol-Lowering Therapy in Women and Elderly Patients With Myocardial Infarction or Angina Pectoris
Findings From the Scandinavian Simvastatin Survival Study (4S)
Background The Scandinavian Simvastatin Survival Study (4S) demonstrated pronounced reductions in mortality and major coronary events in a cohort of patients with established coronary heart disease (CHD). The present study provides a detailed, post hoc assessment of the efficacy and safety of simvastatin therapy in the following subgroups of 4S patients: those ≥65 years of age, those <65 years of age, women, and men.
Methods and Results The 4S cohort of 4444 CHD patients included 827 women and 1021 patients ≥65 years of age. Total cholesterol at baseline was 5.5 to 8.0 mmol/L with triglycerides ≤2.5 mmol/L. Patients were randomized to therapy with simvastatin 20 to 40 mg daily or placebo for a median follow-up period of 5.4 years. End points consisted of all-cause and CHD mortality, major coronary events (primarily CHD death and nonfatal myocardial infarction), other acute CHD and atherosclerotic events, hospitalizations for CHD and cardiovascular events, and coronary revascularization procedures. Mean changes in serum lipids were similar in the different subgroups. In patients ≥65 years of age in the simvastatin group, relative risks (95% confidence intervals) for clinical events were as follows: all-cause mortality, 0.66 (0.48 to 0.90); CHD mortality, 0.57 (0.39 to 0.83); major coronary events, 0.66 (0.52 to 0.84); any atherosclerosis-related event, 0.67 (0.56 to 0.81); and revascularization procedures, 0.59 (0.41 to 0.84). In women, the corresponding figures were 1.16 (0.68 to 1.99); 0.86 (0.42 to 1.74), 0.66 (0.48 to 0.91), 0.71 (0.56 to 0.91), and 0.51 (0.30 to 0.86), respectively.
Conclusions Cholesterol lowering with simvastatin produced similar reductions in relative risk for major coronary events in women compared with men and in elderly (≥65 years of age) compared with younger patients. There were too few female deaths to assess the effects on mortality in women. Because mortality rates increased substantially with age, the absolute risk reduction for both all-cause and CHD mortality in simvastatin-treated subjects was approximately twice as great in the older patients.
A direct relationship between serum cholesterol levels and risk for developing CHD has been most clearly demonstrated in studies of middle-aged men. Although a similar relationship has been observed in middle-aged women and in some older populations (≥65 years of age), the strength of the relationship is reported to be weaker in the elderly and is less convincingly established for elderly women compared with elderly men.1 Women and elderly patients have been poorly represented in prior cholesterol-lowering clinical trials. Consequently, the value of lipid screening and cholesterol-lowering therapy in these populations has been subject to question.2
The 4S randomized 4444 moderately hypercholesterolemic patients with a history of myocardial infarction or angina pectoris to double-blind treatment with simvastatin or placebo. Patients 35 to 70 years of age were entered regardless of sex. Treatment with simvastatin over a median follow-up period of 5.4 years resulted in a 30% reduction in risk of death from all causes (P<.0003), attributable to a 42% reduction in risk of CHD death (P<.00001). Of the total patient population, 19% were female and 23% were ≥65 years of age at the start of the study. The initial 4S publication3 provided limited results for women and for patients ≥60 years of age. In the present article, the efficacy and safety of simvastatin therapy in elderly (age ≥65) and female 4S patients are reported in depth.
Study Design and Conduct
Details of the 4S design, organization, conduct, and baseline characteristics of the entire study cohort have been published previously.3,4 In brief, 4444 patients (827 female) 35 to 70 years of age (1021 were ≥65 years) were enrolled at 94 clinical sites within the Nordic countries. To qualify for entry, patients were required to have total serum cholesterol between 5.5 and 8.0 mmol/L (213 to 309 mg/dL), a triglyceride level ≤2.5 mmol/L (220 mg/dL) on a lipid-lowering diet, and either a documented prior acute myocardial infarction or active, stable angina pectoris. Premenopausal women of childbearing potential were excluded. Other exclusions have been described previously.3,4 Qualifying patients were randomized to receive simvastatin 20 mg daily or placebo. The dosage was increased to 40 mg daily in patients whose cholesterol remained >5.2 mmol/L (201 mg/dL) at the 6- or 18-week measurements. All-cause mortality and major coronary events (CHD death, nonfatal myocardial infarction, or resuscitated cardiac arrest) were the primary and secondary end points of the study, respectively. Any CHD-related event, any atherosclerosis-related event, and coronary revascularization procedures were tertiary end points.
Safety and Tolerability Assessment
Patients were evaluated for adverse events at clinical visits that were scheduled after the initiation of therapy at 6 weeks, 12 weeks, and 6 months and semiannually thereafter. Measurement of serum AST, ALT, and creatine kinase were performed at each of these visits; ECG and routine physical examination were performed annually. An adverse event is defined as any new or worsening symptom or any unfavorable change in the structure, function, or chemistry of the body or the worsening of a preexisting condition. The study investigators at each of the participating clinics classified each adverse event as to relationship to study therapy using the categories of definitely, probably, possibly, probably not, or definitely not drug related. They also recorded whether the adverse event was “serious,” ie, one that resulted in death, was immediately life threatening, resulted in permanent or substantial disability, was a malignant neoplasm, resulted in hospitalization, or prolonged an existing hospitalization.
Treatment group differences were assessed by the log-rank test. Relative risk and 95% confidence intervals were calculated with the Cox regression model.5 Mortality, major coronary events, and revascularization data were also analyzed with baseline variables that were significantly related to outcome (two-sided probability values of P<.05) included in the model. Time to first hospitalization and time to first revascularization were determined by use of the Kaplan-Meier procedure. Treatment differences in the rate of acute CHD hospitalizations, revascularization procedures, and total acute cardiovascular hospitalizations were assessed with a Cox regression method for recurrent events.6 ANOVA was used to test treatment differences in the average length of stay per hospitalization and total bed days per patient. All data were analyzed by intention to treat.
The distribution of male and female patients by age (5-year intervals) in the simvastatin and placebo treatment groups is shown in Fig 1⇓. Mean age of the subgroup of patients ≥65 years of age was ≈67 years compared with 56 years for the <65 subgroup. Baseline characteristics (including lipids) for the four subpopulations of interest (women, men, patients ≥65 years of age, and patients <65 years of age) are provided in Table 1⇓. In general, the two treatment groups within each age or sex subgroup were well matched at baseline. However, there were clear differences in baseline characteristics between women and men and between patients ≥65 and <65 years of age (Table 1⇓). Female participants were slightly older at entry (on average, 2.3 years), and more female patients were entered on the basis of documented angina pectoris without a history of prior myocardial infarction. There were fewer prior revascularization procedures, fewer patients with major Q waves on baseline ECGs, more nonsmokers, more patients with a diagnosis of hypertension, and greater use of nitrates and, to a lesser extent, calcium channel blockers among women compared with men. Women had higher baseline HDL cholesterol on average compared with men. The elderly (≥65 years of age) subgroup included a higher proportion of women, more patients entered on the basis of prior myocardial infarction, and more nonsmokers compared with the younger (<65 years of age) subgroup.
The potential importance of chance differences in baseline characteristics between the two treatment groups for any of these four subpopulations was evaluated by assessment of the relationship of all of the listed baseline variables to total mortality and to major coronary events, as described in “Methods.” The inclusion of those baseline characteristics that were significantly related to outcome in the Cox regression model for the analysis of treatment group differences had virtually no impact on the results. Consequently, only results from analyses unadjusted for baseline variables are reported in the following sections.
Use of estrogens was very low among women participating in 4S and was equally distributed between the two treatment groups (31 patients at baseline: 15 of 407 (3.7%) in the simvastatin group and 16 of 420 (3.8%) in the placebo group).
Serum Lipids During Treatment
Table 2⇓ lists the (intention-to-treat) mean between-treatment group differences for percent change from baseline in total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides. Values are presented from measurements taken 6 weeks after initiation of therapy with the study drug (at which point all patients allocated to simvastatin were still taking 20 mg/d), and values representing the mean between-group differences over the full duration of the trial are given. In general, the observed changes in each of the four subpopulations of interest were similar, and the changes measured at 6 weeks were remarkably stable over the entire 5- to 6-year study period.
Among patients in the simvastatin group, the proportion of women whose dose required titration from 20 to 40 mg was 39% compared with 37% for men. Corresponding figures for patients ≥65 and <65 years of age were 31% and 39%, respectively. The percentages of patients achieving total cholesterol levels <5.2 mmol/L (201 mg/dL) at the 1-year time point were 68.6% for women, 73.1% for men, 75.8% for patients ≥65 years of age, and 71.2% for patients <65 years of age.
Total and CHD Mortality
Total and CHD mortality for the four subpopulations of interest are provided in Table 3⇓. There were 53 deaths among the 827 female patients (6.4%), 28 in the simvastatin group, and 25 in the placebo group. (This includes one female CHD death in the simvastatin group that had been incorrectly reported as a male death in the first 4S publication, an error discovered after publication.) There were 17 deaths attributed to CHD among the female patients in the placebo group (a rate of 4.0%, less than half the 9.5% CHD mortality rate observed in men) compared with 14 CHD deaths in the simvastatin group. Thus, the overall numbers of deaths were too low to permit meaningful assessment of either total or CHD mortality. There were no meaningful differences between the treatment groups in the individual categories of non-CHD death classified by the blinded End Point Classification Committee; the non-CHD deaths for women in the simvastatin:placebo groups were distributed in the following categories: other cardiac, 1:1; nonembolic cerebral infarction, 1:0; intracerebral hemorrhage, 2:0; other arteriosclerotic, 1:0, cancer, 6:6; suicide/violence/accident, 1:1; and other noncardiovascular, 2:1.
All-cause mortality was significantly reduced in patients ≥65 years of age in the simvastatin group compared with the group receiving placebo (relative risk, 0.66; P=.009). The reduction in relative risk was slightly greater than that observed for patients <65 years of age (0.72, P=.007), with overlapping 95% confidence intervals. Relative risk for CHD mortality in the simvastatin group was similarly reduced in the two age groups, by 43% and 42% for patients ≥65 and <65 years of age, respectively, and these results were highly statistically significant for both groups (Table 3⇑). All-cause and CHD mortality rates in the placebo group over the duration of the study were more than twice as high for patients ≥65 years of age (19.1% and 14.5%, respectively) compared with patients <65 years of age (9.3% and 6.7%, respectively). Consequently, the absolute risk reduction in the ≥65 years of age group was more than twice that in the <65 years of age group for both total and CHD mortality (ie, the numbers of deaths prevented by simvastatin therapy in the ≥65- and <65-year-old 4S subpopulations were 61.5 versus 25.5 per 1000 patients [all cause] and 60.2 versus 28.1 [CHD], respectively; see Table 3⇑).
The results of an analysis of the proportion of patient deaths (CHD and all cause) as a function of age with a logistic regression model (with treatment group, age, and treatment by age interaction in the model) are shown in Fig 2⇓. The curves estimate the proportion of patients who died in each group as a function of patient age at entry into the study from the 5th to 95th age percentiles. Total and CHD mortality increased with age in both groups, with a beneficial simvastatin treatment effect seen across the full range of patient ages shown in the figures.
Kaplan-Meier curves for all-cause mortality in patients ≥65 and <65 years of age are shown in Fig 3⇓. The curves begin to separate at <1 to 2 years, as was the case for the entire study population,4 and continue to diverge for the duration of follow-up in the study.
Major Coronary Events and Other Event Categories
Analysis of major coronary events in female 4S patients provides considerably greater statistical power compared with CHD mortality because there were approximately five times the number of such events observed. The relative risk for major coronary events for women in the simvastatin group (Table 3⇑) was 0.66 (P=.012), identical to the relative risk observed in men. The related event category of nonfatal myocardial infarctions showed a similar degree of risk reduction and statistical significance. Significant risk reductions in women were seen in all the other prespecified tertiary end-point categories for which significant results were observed in the entire study cohort, including any acute CHD-related event, any acute atherosclerosis-related event, and need for coronary revascularization procedure. The magnitude of risk reductions in these categories was quantitatively similar to that observed in men (Table 3⇑) and in the entire study population.3
Highly significant risk reductions were also observed for these secondary and tertiary end-point categories in patients ≥65 years of age in the simvastatin group (Table 3⇑), with 95% confidence intervals overlapping those observed for the patients <65 years of age and the entire 4S study cohort.3 The magnitude of the relative risk reductions in each case was as great as or greater for patients ≥65 years of age compared with patients <65 years of age.
Kaplan-Meier curves for major coronary events and for revascularization procedures are shown for the subpopulations of interest in Fig 4⇓. The curves are similar to those previously published for the entire study population,3 separating at <1 to 2 years and continuing to diverge during the remainder of the follow-up period.
The numbers of hospitalizations and total hospital bed days for acute CHD and acute cardiovascular events in patients in the simvastatin and placebo groups within each of the four subpopulations are shown in Table 4⇓. Statistically significant reductions in both numbers of hospitalizations and total hospital bed days for both categories were observed in each subgroup. There were no consistent differences in the magnitude of these reductions in the different subgroups. The frequency of all coronary artery revascularization procedures (as opposed to only the first procedure, captured as a tertiary end point) is also listed Table 4⇓. Significant reductions were seen in simvastatin- compared with placebo-treated patients in both age subgroups and for men as well as women.
Safety and Tolerability
A detailed analysis of adverse experiences, safety, and tolerability for the entire study cohort has been published separately,7 including an assessment of cancers, which were approximately equally distributed between simvastatin and placebo patients. There was no evidence for increased prevalence of any sex-specific cancer in patients treated with simvastatin (breast, 3 simvastatin and 6 placebo; ovary, 2 simvastatin and 1 placebo; penis, 1 simvastatin and 0 placebo; prostate, 17 simvastatin and 20 placebo; and uterus, 1 simvastatin and 3 placebo).
The numbers of patients within each of the age and sex subpopulations who permanently discontinued study drug for any reason other than death are listed in Table 5⇓. Discontinuations were ≈20% lower for simvastatin-treated patients compared with placebo-treated patients in each subpopulation. Discontinuation rates were similar for patients ≥65 and <65 years of age. Adverse clinical experiences were the reason for discontinuation in similar proportions of patients in the simvastatin and placebo groups within each of the age and sex subpopulations. Discontinuations because of adverse laboratory experiences were uncommon (none at all in the age ≥65 subgroup) and were distributed nearly equally between simvastatin- and placebo-treated patients in each subpopulation. They consisted primarily of elevations in liver enzymes.
A summary of the numbers/percentages of patients in the different subgroups who experienced adverse clinical or laboratory events is also provided in Table 5⇑. The lower occurrence of adverse clinical events within the simvastatin subgroups in each of the four subpopulations (particularly those rated as serious) was a consequence of the larger numbers of cardiovascular adverse events in placebo-treated patients, consistent with the end-point findings. Review of specific clinical adverse event categories did not reveal any significant age- or sex-specific toxicities of simvastatin within the subpopulations. Among women, the only adverse clinical experience category not related to end point with a statistically significant (P<.05) difference between the treatment groups was new or increased hypertension (simvastatin, 18; placebo, 38, P=.009). The latter difference was also seen in the entire 4S cohort (simvastatin, 133; placebo, 168; P=.042) but not in men (simvastatin, 115; placebo, 130; P=.32). In patients ≥65 years of age, laboratory adverse experiences of increases in ALT or AST occurred significantly more frequently in simvastatin- versus placebo-treated patients (for ALT, 40 patients (7.7%) versus 22 patients (4.4%), P=.026; for AST, 26 patients (5.0%) versus 10 patients (2.0%), P=.01). The minor and generally transient nature of these elevations in the elderly is reflected both by the fact that no patient ≥65 years of age discontinued the study drug permanently because of liver enzyme elevations and by the fact that elevations to levels three times the upper limit of normal among these patients were infrequent and comparably distributed between the treatment groups (for ALT, 5 simvastatin and 3 placebo; for AST, 3 simvastatin and 5 placebo).
Patients in 4S were randomized regardless of sex provided that they met the entry criteria specified in the protocol.4 The percentage of women in 4S (19%) was similar to percentages randomized in smaller secondary prevention trials that entered CHD patients of both sexes within a similar age range.8–10
Differences in baseline characteristics for women compared with men in 4S included a lower proportion of prior myocardial infarction, more nonsmokers, more baseline diagnoses of hypertension, greater use of nitrates and calcium antagonists, and fewer prior coronary revascularization procedures. Female participants were slightly older than male participants and had higher HDL cholesterol levels with comparable LDL cholesterol levels. Some of these differences undoubtedly contributed to the lower incidence of CHD mortality and nonfatal CHD events in women compared with men. Estrogen use among women was very low and was equally distributed between treatment groups; it is unlikely, therefore, to have had any impact on results other than possibly contributing slightly to the lower mortality and morbidity seen in women. The most notable baseline differences between the age subgroups were fewer smokers and greater numbers of patients entered on the basis of myocardial infarction in older versus younger patients. Because of the randomization process, however, between-treatment group differences for the age subpopulations were minor and did not significantly affect the results.
The mean changes in serum lipids produced by simvastatin were similar in each of the examined subpopulations and were similar to the changes observed in the entire study cohort.3 For the subjects ≥65 versus <60 years of age, lipid changes of a similar magnitude were achieved despite titration of fewer patients from 20 to 40 mg (31% compared with 39%). This is consistent with other data suggesting that increasing age enhances cholesterol-lowering responsiveness to HMG CoA reductase inhibitors.11 Baseline total, LDL, and HDL cholesterol levels showed no significant relationship to treatment outcome (reduction in relative risk) for any of the examined subpopulations (data not shown), as was also the case for the entire study cohort.12
The 4S is the first clinical trial to convincingly demonstrate that cholesterol-lowering therapy reduces the risk of major coronary events in women with preexisting CHD. The magnitude of the reduction (34%) was similar to that observed in men, as were the magnitudes of risk reductions for the tertiary end points of any CHD event, any atherosclerotic event, and need for coronary revascularization. Decreased rates of cardiovascular hospitalizations and hospital bed days further support the benefit of treatment. Small numbers of CHD and total deaths in women precluded meaningful assessment of simvastatin effects on all-cause or CHD mortality in women. However, the finding that all CHD end-point categories other than mortality were similarly positive for men and women (Table 3⇑) makes it highly likely that a study with adequate statistical power to address mortality in women (such as the ongoing Heart Protection Study in the United Kingdom) will yield a positive outcome, particularly because there is no suggestion in the 4S data of any adverse influence of simvastatin on non-CHD mortality in women or in the general 4S population.3,7
The 4S is also the first clinical trial to clearly demonstrate that cholesterol-lowering therapy reduces the risk of all-cause mortality and major coronary events in CHD patients ≥65 years of age. Simvastatin treatment produced a highly significant 34% reduction in risk for all-cause mortality (P=.009), attributable to a 43% reduction in risk for CHD mortality (P=.003). The observed relative risk reductions in both mortality categories, as well as for other secondary and tertiary CHD end points (Table 3⇑), were as great as or greater than those seen in younger patients. Because mortality and CHD event rates increased with age (Fig 2⇑), the absolute reductions in deaths and events were substantially greater for patients ≥65 compared with those <65 years of age. It should be remembered that at the conclusion of the study, those patients who were ≥65 years old at entry ranged from 70 to 76 years of age.
Because the strength of the relationship between serum cholesterol and development of CHD observed in some epidemiological studies is reported to weaken in elderly compared with middle-aged subjects,1,13–16 the above findings may at first appear surprising. However, data on the predictive importance of cholesterol in elderly patients with established CHD are limited. Moreover, the mechanisms underlying the relationship between cholesterol and the development of CHD may differ from the mechanisms contributing to therapeutic benefit from cholesterol lowering in the setting of advanced atherosclerosis. There is evidence that “lipid-rich” atherosclerotic plaques are particularly susceptible to rupture, leading to clinical events of myocardial infarction or unstable angina. Lipid depletion of rupture-prone plaques may stabilize such lesions, resulting in fewer clinical events.17 A mechanism of this nature would be expected to reduce risk in patients having vulnerable plaques, regardless of age.
In the CARE trial, 4159 patients with prior myocardial infarction and total cholesterol level <6.2 mmol/L (240 mg/dL) were randomized to treatment with pravastatin 40 mg/d or placebo for ≈5 years.18 Women (n=576) and patients ≥60 years of age (n=2129) had statistically significant reductions in major coronary events. It is difficult to compare specific event rates meaningfully in 4S and CARE because there were major differences in the definitions of end points and baseline characteristics of the patients in the two studies. For example, the end-point category of major coronary events in CARE, unlike 4S, included revascularization procedures but excluded silent myocardial infarctions. All women studied in CARE were entered on the basis of a prior myocardial infarction, while only 63% of 4S women had a history of prior definite infarction. Pertinent characteristics of the female cohort in CARE included significantly more current smokers and higher prevalences at baseline of diabetes, hypertension, family history of CHD, and multiple risk factors compared with the male CARE cohort.19 Larger risk reductions for major coronary events were observed in women compared with men with pravastatin treatment in CARE (46% versus 20%, respectively). In contrast, the observed relative risk reductions for major coronary events in 4S were identical for the two sexes. Whether the apparent male-female treatment effect difference in CARE reflects a chance finding (the reported 95% confidence intervals for men and women overlapped), a consequence of differences in patient baseline characteristics, or some other explanation cannot be determined from the available data. Neither 4S nor CARE had sufficient power to address total or CHD mortality in women. Total and CHD mortality data for women and for patients ≥65 have not yet been reported for CARE.
In summary, cholesterol lowering with simvastatin in 4S produced statistically significant risk reductions in women, men, and patients ≥65 or <65 years of age for major coronary events and all tertiary CHD and atherosclerosis-related study end points that were positive for the entire 4S cohort. The magnitudes of the observed risk reductions in these subgroups were remarkably similar to those reported for the entire study cohort and for other clinically relevant subgroups that have been analyzed.3,12,20 Although 4S was not specifically designed to address mortality benefit in the elderly, high event rates combined with the substantial portion of patients who fell into this subgroup provided the power to demonstrate highly significant risk reductions for both all-cause and CHD mortality. Safety and tolerability findings for the age and sex subpopulations showed no important differences and were generally consistent with the findings for the entire study cohort.7 We conclude that beneficial effects of cholesterol lowering with simvastatin in CHD populations such as that studied in 4S occur independent of age and sex.
Selected Abbreviations and Acronyms
|4S||=||Scandinavian Simvastatin Survival Study|
|CARE||=||Cholesterol and Recurrent Events|
|CHD||=||coronary heart disease|
This study was supported by a grant from Merck Research Laboratories, Rahway, NJ.
- Received July 11, 1997.
- Revision received September 15, 1997.
- Accepted September 23, 1997.
- Copyright © 1997 by American Heart Association
Manolio AT, Pearson AT, Wenger NK, Barrett-Connor E, Payne GH, Harlan WR. Cholesterol and heart disease in older persons and women: review of an NHLBI workshop. Ann Epidemiol. 1992;3:2161–2176.
Garber AM, Littenberg B, Sox HC Jr, Gluck, ME, Wagner JL, Duffy BM. Costs and Effectiveness of Cholesterol Screening in the Elderly. Washington DC: Health Program, Office of Technology Assessment, Congress of the United States; April. 1989. GPO publication No:052–003-01151–4.
Cox DR. Regression methods of life tables (with discussion). J R Stat Soc. 1972;B34:187–220.
Pedersen TR, Berg K, Cook TJ, Færgeman O, Haghfelt T, Kjekshus J, Miettinen T, Musliner TA, Olsson AG, Pyörälä K, Thorgeirsson G, Tobert JA, Wedel H, Wilhelmsen L. Safety and tolerability of cholesterol lowering with simvastatin during 5 years in the Scandinavian Simvastatin Survival Study. Arch Intern Med. 1996;156:2085–2092.
Research Committee of the Scottish Society of Physicians. Ischaemic heart disease: a secondary prevention trial using clofibrate. BMJ. 1971;4:775–784.
Physicians of the Newcastle Upon Tyne Region. Trial of clofibrate in the treatment of ischaemic heart disease. BMJ. 1971;4:767–775.
Shear CL, Franklin FA, Stinnett S, Hurley DP, Bradford RH, Chremos AN, Nash DT, Langendorfer A. Expanded Clinical Evaluation of Lovastatin (EXCEL) study results. Circulation. 1992;85:1293–1303.
Dawber TR. The Framingham Study: the Epidemiology of Atherosclerotic Disease. Cambridge, Mass: Harvard University Press; 1980.
Fuster V. Mechanisms leading to myocardial infarction: insights from studies of vascular biology. Circulation. 1994;90:2126–2146.
Sacks FM, Pfeffer MA, Moye LA, Rouleau JL, Rutherford JD, Cole TG, Brown L, Warnica JW, Arnold JMO, Wun C-C, Davis BR, Braunwald E, for the Cholesterol and Recurrent Events Trial Investigators. The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. N Engl J Med. 1996;335:1001–1009.
Lewis SJ, Mitchell JS, East C, Glasser P, Grant J, Johnstone DE, Kell S, Letterer R, Limacher M, Timmis G, Wun C-C, Rouleau JL, Pfeffer MA, Sacks F, Braunwald E. Women in CARE have earlier and greater response to pravastatin post myocardial infarction. Circulation. 1996;94(suppl I):I-12. Abstract.