Key words: anticoagulants; coronary artery bypass; health-related quality of life; lipid-lowering therapy

Transcription

1 Quality of Life After Coronary Artery Bypass Graft* Results From the POST CABG Trial Norma Lynn Fox, PhD, MPH; Byron J. Hoogwerf, MD; Susan Czajkowski, PhD; Ruth Lindquist, PhD; Gilles Dupuis, PhD; J. Alan Herd, MD; Lucien Campeau, MD; Ann Hickey, MD; Franca B. Barton, MS; and Michael L. Terrin, MD, MPH, FCCP; for the POST CABG Study Investigators Objectives: The POST CABG (Post Coronary Artery Bypass Graft) Trial showed that aggressive lowering of low-density lipoprotein (LDL) cholesterol levels reduced the progression of atherosclerosis in saphenous vein grafts. In the extended follow-up phase, aggressive lowering of LDL cholesterol levels was associated with reduced rates of clinical events. Low-dose anticoagulation therapy did not reduce the progression of atherosclerosis. We conducted this analysis to determine the effects of both lipid-lowering and low-dose anticoagulation therapy on healthrelated quality of life (HRQL). Design: Randomized clinical trial, factorial design. Setting: Outpatients in five tertiary care medical centers. Patients: A cohort of 852 patients enrolled in the POST CABG Trial completed an HRQL questionnaire at baseline, and at the year 2 and year 4 follow-up visits. Intervention: Aggressive LDL cholesterol lowering vs moderate LDL cholesterol lowering, and low-dose warfarin vs placebo. Measurements: Domains included emotional status, basic physical and social functioning, perceived health status, symptoms of pain, a variety of physical symptoms, and global life satisfaction. Results: Overall, there were no indications of systematic differences among treatment groups for any of the HRQL parameters at baseline, year 2, or year 4. Conclusions: These data indicate that patients did not experience detrimental or beneficial effects on HRQL parameters while receiving LDL cholesterol-lowering therapy that had demonstrable benefits for treatment of atherosclerosis. (CHEST 2004; 126: ) Key words: anticoagulants; coronary artery bypass; health-related quality of life; lipid-lowering therapy Abbreviations: BAATAF Boston Area Anticoagulation Trial for Atrial Fibrillation; CABG coronary artery bypass graft; CES-D Center for Epidemiologic Studies Screening Test for Depression; HRQL health-related quality of life; INR international normalized ratio; LDL low-density lipoprotein; LRC-CPPT Lipid Research Clinics Coronary Primary Prevention Trial; POST CABG Post Coronary Artery Bypass Graft *From the Maryland Medical Research Institute (Drs. Fox and Terrin, and Ms. Barton), Baltimore, MD; Cleveland Clinic Foundation (Dr. Hoogwerf), Cleveland, OH; National Heart, Lung, and Blood Institute (Dr. Czajkowski), Bethesda, MD; University of Minnesota (Dr. Lindquist), Minneapolis, MN; Montreal Heart Institute (Drs. Dupuis and Campeau), Montreal, PQ, Canada; Cedar Sinai Medical Center (Dr. Hickey), Los Angeles, CA; and Baylor College of Medicine (Dr. Herd), Houston, TX. Dr. Fox is currently at Human Genome Sciences, Inc., Rockville, MD. Dr. Fox owns stock in Merck & Company (the manufacturer of study treatment and a contributor to the Post CABG Trial), which was purchased after the end of Dr. Fox s employment at the Post CABG Studies Coordinating Center. The POST CABG (Post Coronary Artery Bypass Graft) Trial tested two hypotheses: aggressive lowering of low-density lipoprotein (LDL) cholesterol, targeted to 60 to 85 mg/dl (1.6 to 2.2 mmol/l) was more effective in delaying the progression of atherosclerosis in saphenous vein grafts than moderate lowering; and that low-dose anticoagulation, to POST CABG Studies Investigators listed in references 1 and 21. This study was supported by contracts N01-HC-75071, 75072, 75073, 75074, and from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, MD and was partially supported by Merck & Company. Manuscript received February 25, 2003; revision accepted March 30, Reproduction of this article is prohibited without written permission from the American College of Chest Physicians ( permissions@chestnet.org). Correspondence to: Norma Lynn Fox, PhD, MPH, Human Genome Sciences, Inc., Rockville, MD 20850; normalynn_ fox@hgsi.com CHEST / 126 / 2/ AUGUST,

2 maintain an international normalized ratio (INR) 2, was effective, as compared with placebo, in reducing obstruction of bypass grafts. The results showed that aggressive lowering of LDL cholesterol levels reduced the progression of atherosclerosis in grafts. Low-dose anticoagulation therapy did not reduce the progression of atherosclerosis. 1 In the extended follow-up phase, reduced rates of clinical events were observed in the groups assigned to aggressive lowering of LDL cholesterol levels and in the groups assigned to low-dose anticoagulants. 2 The clinical and angiographic benefits of lipid-lowering therapies are now well recognized. 3 8 Furthermore, the efficacy and safety of the statin class of lipidlowering drugs has resulted in widespread clinical use of these agents. The Lipid Research Clinic Coronary Primary Prevention Trial (LRC-CPPT) 9 investigators reported an increased risk for accidental death and suicide in patients randomly assigned to cholestyramine, compared to patients assigned to placebo. The Helsinki Heart Study, 10 a primary prevention trial of lipidlowering therapy using gemfibrozil, also reported that patients randomized to gemfibrozil experienced more deaths due to violence and accidents than patients assigned to placebo. These early observations from the LRC-CPPT and the Helsinki Heart Study were important in the controversy about whether there was a causal relationship between lipid-lowering therapy and noncardiovascular mortality. Muldoon et al 11 published results of a metaanalysis reporting that lipid-lowering interventions, including diet, clofibrate, colestipol, cholestyramine, and gemfibrozil, tended to lower coronary deaths, but not overall mortality, with an apparent excess of deaths not related to disease (accidents, suicide, or violence). Subsequent analyses that included data from the statin class of lipid-lowering therapy suggested that the excess of noncardiovascular mortality observed in some clinical trials may be related to particular classes of therapy, specifically fibrates and hormones, rather than to lipid lowering. 12 Nevertheless, some investigators speculated that the association noted by Muldoon et al 11,13 could be explained by negative effects of lipid lowering on neurochemistry or behavior, for example with depression or violent behavior as mediating conditions. 14,15 There have been some observational and experimental studies that showed correlations between relatively low lipid levels and depression, risk of suicide, and other disorders or adverse outcomes. Results of other studies 19,20 suggest that lipid-lowering interventions do not influence health-related quality of life (HRQL), or even have positive effects on conditions such as depression. These studies have differed widely with regard to baseline characteristics of patients, interventions under evaluation (behavioral or pharmacologic, class of pharmacologic agent), length of follow-up (weeks to years), and outcome measures (sleep disorders to comprehensive qualityof-life questionnaires). There have been limited evaluations of the effects of warfarin therapy on HRQL. The results of the Boston Area Anticoagulation Trial for Atrial Fibrillation (BAATAF) suggested that warfarin therapy that results in an INR range of 1.5 to 2.5 was not generally associated with changes in measures of functional status, well being, and perceived health. However, patients who experienced a bleeding episode showed decreases in perceived health. 21,22 In the POST CABG Trial, we examined the effects of both lipid-lowering and low-dose anticoagulation therapy on HRQL. Materials and Methods Institutional review board approvals were obtained at each participating center, and informed consent was obtained from each patient enrolled. HRQL measures involving psychosocial and physical well being were obtained through a self-administered questionnaire at baseline (randomization visit), and at the year-2 and year-4 follow-up visits. Patients completed the questionnaire during a clinic visit or at home. Domains included emotional status based on assessment of anxiety, scored on a scale of 20 to 80, with higher scores indicating increased anxiety (Spielberger State Anxiety Inventory 23 ), and depression, scored on a scale of 0 to 60, with higher scores indicating more depressive symptoms and a screening score 16 considered to warrant further evaluation (Center for Epidemiologic Studies Screening Test for Depression [CES-D]). 24 Basic physical functioning and social functioning were evaluated with the functional status questionnaire, scored on a scale of 0 to 100%, with 100% indicating that all functions can be performed with no difficulty whatsoever. 25 Additional factors assessed included the patients perceived health status, 26 pain symptom levels (McGill pain questionnaire), 27 a variety of physical symptoms based on patient recall, and global life satisfaction at present and anticipated 5 years hence. 28 Their measurement has been described in detail. 28,29 A summary of the HRQL domains and measures used is provided in Table 1. The details of the design of the POST CABG Trial have been reported previously. 1 Eligibility criteria included men and women aged 21 to 74 years who underwent CABG procedures 1 to 11 years earlier. In a 2 2 factorial design, patients were randomly assigned to aggressive LDL cholesterol lowering using lovastatin (and cholestyramine, as necessary) aimed to achieve a LDL cholesterol goal of 60 to 85 mg/dl (1.6 to 2.2 mmol/l), compared to moderate LDL cholesterol lowering (same medications in lower doses) to achieve a LDL cholesterol goal of 130 to 140 mg/dl (3.4 to 3.6 mmol/l). Patients were also randomly assigned to receive warfarin or placebo; the warfarin dose (1 to 4 mg) was titrated to achieve a mean INR 2. Statistical Analysis Comparisons of proportions of patient groups having given attributes were made with 2 tests, with one degree of freedom for unadjusted two-group comparisons. 30 For continuous data, 488 Clinical Investigations

3 Table 1 HRQL Domains Domain Measure Comment Emotional status Anxiety Spielberger et al 23 Scale 20 80, with higher scores indicating higher anxiety. Depression CES-D 24 Scale 0 60, with higher scores indicating more depressive symptoms. Functional status Basic physical Functional status questionnaire 25 Score of 0 100%, with 100% indicating that all functions can be performed with no difficulty. Social Functional status questionnaire 25 Perceived health status Rand medical outcomes study 26 Based on patients ranking of their perceived health as excellent, very good, good, fair, or poor. Symptoms of pain McGill pain questionnaire short form 27 Scale of 0 45, based on 15 items in the affective and sensory dimension. Each item is scored 0 to 3, with 0 corresponding to the absence of the symptom and 3 corresponding to severe symptoms. Physical symptoms Study-specific questions Life satisfaction Ladder of life 28 Based on patient s assessments of their overall life satisfaction, with a score of 0 corresponding to their perception of the worst possible life and a score of 10 corresponding to their perception of the best possible life. mean values ( SD) are presented, and comparisons were made with two-sample t tests. 31 Tests for homogeneity of the effects among treatment groups were performed for proportions using the Mantel-Haenzel method and for continuous data with interaction terms in analysis of variance models. 30,31 Because no important interactions were detected, the results were pooled for single comparisons of treatment effects, and p values used are two sided without correction for multiple comparisons. Because many comparisons are made in these analyses, some p values 0.05 would be expected due to chance, and can only be interpreted as suggesting associations; p 0.01 provides some evidence, and p provides strong evidence of associations. The number of patients whose data are included in this analysis was determined by the enrollment and timing of the POST CABG Trial. With the number of patients in each treatment group, comparisons of an level of 0.01 were made with 80% power for differences of 0.35 SDs and 90% power for differences of 0.39 SDs. The analyses were performed using the SAS statistical package (SAS Institute; Cary, NC). Results Patient Characteristics The characteristics of the POST CABG Trial patients have been described elsewhere. 1 A total of 852 of the 1,351 patients enrolled in the POST CABG Trial completed the HRQL questionnaire at the randomization visit, year 2, and year 4, and are referred to as the HRQL cohort and are included in the analyses reported in this article. The majority of patients (n 316) who were excluded did not complete the questionnaire at baseline, primarily because the questionnaire was not put into use until some months after the initiation of the POST CABG Trial. Forty-eight patients died prior to the year-4 assessment and are not included in the analysis (13 patients in the aggressive treatment plus warfarin group, 10 patients in the aggressive treatment plus placebo group, 9 patients in the moderate treatment plus warfarin group, and 16 patients in the moderate treatment plus placebo group). The baseline characteristics of 852 patients included in the analysis were compared to those of the 499 patients who were not included (Table 2). A slightly higher percentage of patients who were Table 2 Baseline Characteristics of Patients Included in and Excluded From HRQL Analysis* Characteristics HRQL Cohort (n 852) Non-HRQL Cohort (n 499) p Value Male gender White race CABG 5 yr Family history of coronary heart disease Family history of myocardial infarction Current smoker ST-segment depression mm Aspirin blocker Calcium-channel blocker Insulin or glucoselowering agent Thiazide diuretic Age, yr Years since CABG Cholesterol, mg/dl LDL, mg/dl High-density lipoprotein Triglycerides Ejection fraction *Data are presented as % unless otherwise indicated. Data are presented as mean. CHEST / 126 / 2/ AUGUST,

4 Table 3 Anxiety and Depressive Symptoms According to Aggressive (n 436) Moderate (n 416) (n 431) (n 421) Mean SD Mean SD p Value Mean SD Mean SD p Value Anxiety Baseline Year Year Depressive symptoms Baseline Year Year included in the HRQL cohort had undergone CABG within the past 5 years (57% vs 52%, p 0.07). A lower percentage of patients in the HRQL cohort had a family history of coronary heart disease (68% vs 74%, p 0.02) or a family history of myocardial infarction (47% vs 53%, p 0.02). A higher percentage of patients in the HRQL cohort were receiving aspirin at baseline (80% vs 68%, p 0.001). Patients in the HRQL cohort had a slightly higher mean age (61.9 years vs 60.9 years, p 0.01). HRQL The HRQL results are summarized in Tables 3 7. Overall, there were no indications of systematic differences among treatment groups for any of the HRQL parameters at baseline, year 2, or year 4. Specifically, the mean scores on the Spielberger State Anxiety Inventory were 30 to 32 (Table 3). The average scores for the CES-D scale were 5.8 to 7.0. There is a small difference between the warfarin and placebo groups at year 2 (7.0 vs 5.8, p 0.02). Subscales of the functional status questionnaire for self report of basic physical functioning and social functioning averaged 95%, with the large majority of patients responding that they were able to perform all the basic physical and social activities (Table 4). There was a small difference at baseline between the aggressive treatment group and the moderate treatment group (97.8% vs 98.9%, p 0.02) that was not present at the later observation periods. The patients reports of symptoms are summarized in Table 5. Tiredness was experienced at least once per week for approximately half the patients. Fewer than one fourth of patients experienced chest pain or shortness of breath at least once per week. There were no indications of changes in these symptoms during the 4-year follow-up period. Approximately one third of patients reported trouble sleeping at least once a week. Fewer than 20% of patients reported trouble concentrating at least once a week. Less than 10% had problems eating at least once a week. Edema at least once per week was reported for slightly more patients in the aggressive LDL cholesterol-lowering treatment group at year 4 (19% vs 13%, p 0.02). Approximately one third of patients experienced forgetfulness at least once per week, and approximately 40% experienced muscle aches and pains at least once a week. The McGill pain scale Table 4 Physical and Social Functioning According to Aggressive (n 436) Moderate (n 416) (n 431) (n 421) Mean SD Mean SD p Value Mean SD Mean SD p Value Basic physical functioning Baseline Year Year Social functioning Baseline Year Year Clinical Investigations

5 Table 5 Symptoms Once or More Per Week According to Aggressive (n 436), % Moderate (n 416) (n 421) % p Value (n 431), % % p Value Tiredness Baseline Year Year Chest pain Baseline Year Year Shortness of breath Baseline Year Year Trouble sleeping Baseline Year Year Trouble concentrating Baseline Year Year Irritability Baseline Year Year Problems eating Baseline Year Year Edema Baseline Year Year Forgetfulness Baseline Year Year Muscle aches and pains Baseline Year Year scores averaged 3.5 to 4.7, with no consistent differences among treatment groups or over time (Table 6). Patients were asked to rank their overall global life satisfaction on the ladder of life. The average score at present was 7.9 to 8.2, with similar atpresent scores reported at baseline, year 2, and year 4 (Table 6). The patients anticipated their global life satisfaction 5 years in the future to be somewhat higher, with scores averaging 7.8 to 8.5, with small differences between the warfarin and placebo groups at year 2 (7.9 vs 8.4, p 0.002) and year 4 (7.8 vs 8.1, p 0.02). The majority of patients ranked their health as excellent or very good; 25% ranked their health as fair or poor (Table 7). These responses were consistent across treatment groups and at all three time points. Discussion This study shows that long-term treatment to lower LDL cholesterol levels, effective in reducing progression of atherosclerosis, was not associated with any detrimental or beneficial effects on HRQL parameters, including symptoms of depression and anxiety over a 4-year period. These observations are CHEST / 126 / 2/ AUGUST,

6 Table 6 McGill Pain Scale and Ladder of Life According to Aggressive (n 436) Moderate (n 416) (n 431) (n 421) Mean SD Mean SD p Value Mean SD Mean SD p Value McGill pain scale Baseline Year Year Ladder of life now Baseline Year Year Ladder of life 5 yr from now Baseline Year Year especially noteworthy because of the approach to lipid lowering that was used in the aggressive LDL cholesterol-lowering treatment group. In the aggressive LDL cholesterol-lowering treatment group, the mean LDL cholesterol level at 1 year was 93 mg/dl, and 66% of the participants had LDL cholesterol levels 100 mg/dl. In both the 4-year and longer follow-up assessments, these reductions in LDL cholesterol levels were associated with clinically and statistically significant improvements in the angiographic end points, and favorable trends in clinical events, including a reduced number of revascularization procedures (coronary artery bypass graft [CABG] or percutaneous transluminal coronory angioplasty). 1,2 The absence of any adverse effect on HRQL parameters, including symptoms of depression (as measured on the CES-D, which standardizes the inquiry for symptoms to screen for individuals for whom a proportion will be diagnosed with clinical depression, a proposed precursor of suicide in other writings on noncardiovascular mortality and lipidlowering therapy), is reassuring as increasing numbers of patients are being treated with medications to lower LDL cholesterol levels to ranges that are Table 7 Perceived Health According to Aggressive (n 436), % Moderate (n 416), % p Value (n 431), % (n 421) p Value Perceived health Baseline Excellent Very good Good Fair Poor Perceived health Year Excellent Very good Good Fair Poor Perceived health Year Excellent Very good Good Fair Poor Clinical Investigations

7 comparable to those in the aggressive LDL cholesterol-lowering treatment group of the POST CABG Trial. with warfarin was not associated with an impact on HRQL parameters, nor was it associated with changes in angiographic measures or clinical events in the POST CABG Trial. The BAATAF investigators 21,22 observed that patients who experienced a bleeding event during treatment with warfarin experienced a decrease in their perception of their health status. Patients treated with warfarin in BAATAF had an INR of 1.5 to 2.5 and a slight increase in risk of bleeding events. In contrast, patients randomly assigned to warfarin treatment in the POST CABG Trial had a lower average INR of 1.4, and no excess risk of bleeding that required hospitalization and/or transfusion was detected. Accordingly, the current observations may be attributable to the low level of anticoagulation achieved in the POST CABG Trial. It is noteworthy that the study patient population showed generally very positive scores on all components of the HRQL assessment at study entry and on 2-year and 4-year follow-up. The activities of daily living scales showed high levels of basic physical and social functioning. A minority of patients reported clinical symptoms such as chest pain or shortness of breath once a week or more often. It seems likely that the overall positive HRQL status reflects the selection of patients with the physical, psychological, and social resources needed to participate in this long-term clinical trial. These patients had undergone CABG surgery a minimum of 1 year and an average of 4.3 years prior to study entry. Accordingly, the majority were well beyond the perioperative period, when patients generally have lower HRQL scores. 29 These results are consistent with the findings of the Randomized Intervention of Angina Trial 32 that relief of angina by percutaneous transluminal coronary angioplasty or CABG is associated with positive perceived health status. Also, the absence of any finding of improved HRQL scores in association with aggressive lipid-lowering therapy provides a cautionary note. Hypercholesterolemia is a silent killer like high BP. Not until there are severe abnormalities in the coronary artery anatomy do patients have symptoms or reduced HRQL scores. Patients must be encouraged to comply with aggressive lipid-lowering treatments with the expectation that the therapy will prevent unnecessary losses in quality of life instead of the expectation of rapid and noticeably improved quality of life. The suspicion that reduction in cholesterol might be associated with increased risk of violent death (eg, accidental death, suicide) was initially raised in the context of two randomized, double-blind, placebo controlled clinical trials: the LRC-CPPT 9 and the Helsinki Heart Study. 10 In the LRC-CPPT, cholesterol lowering with cholestyramine was associated with a reduction in coronary heart disease events. Among the patients who died of noncardiovascular events, there were 11 accidental/violent deaths in the cholestyramine group, compared to 4 deaths in the placebo group; 7 of these deaths were suicide/ homicide, and 7 deaths were automobile/motorcycle accidents. Half of these deaths were associated with high alcohol levels. The LRC-CPPT investigators 9 carefully investigated violent deaths to ensure that they were not related to coronary events, but they did not report any association with cholesterol levels, perhaps because of the small number of accidental deaths. The Helsinki Heart Study 10 (a randomized, double-blind controlled trial comparing the effects of gemfibrozil vs placebo) had a slightly higher number of violent deaths in the gemfibrozil group, causing some speculation about a possible relationship between cholesterol level reduction and the risk of violent death. Subsequent to these two clinical trials, several observational studies were summarized by Muldoon et al, 11 showing a relationship between lower serum cholesterol and increased noncardiovascular mortality. In a summary of randomized clinical trial data from dietary and medication lipid-lowering studies, Muldoon et al 11 conclude that there are statistically significant increases in nonillness mortality in the cholesterol-lowering groups. 11 Subsequent to these reports, in addition to the POST CABG Trial, there have been eight large randomized, double-blind clinical trials 3,5 8,33 35 reporting violent deaths including accidental deaths and suicides. 3-hydroxy-3-methylglutaryl coenzyme-a reductase inhibitors (statins) were used in seven trials, 3 8,34 35 and gemfibrozil was used in one study 33 (Table 8). When data are summarized from the 10 randomized clinical trials with violent deaths reported, there are nearly equal numbers of violent deaths in the placebo groups (n 67) and the cholesterol-lowering treatment groups (n 63). When data from only statin trials are considered, there are slightly greater numbers of violent deaths in the placebo groups (n 54) than in the statin treatment groups (n 40). These data from large randomized clinical trials support the concept advanced in the current report that aggressive cholesterol lowering is not associated with adverse psychological changes or an increased risk for violent behavior. Our data are limited to POST CABG Trial participants for whom data were available throughout the follow-up period; exclusion of patients for whom long-term data are not available, for example because of death, may bias the observations to overall CHEST / 126 / 2/ AUGUST,

8 Table 8 Ten -Controlled, Lipid-Lowering Clinical Trials* Participants, No. Clinical Trials Lipid-Lowering Agent LRC-CPPT Cholestyramine 1,900 1,906 Helsinki Heart Study 10 Gemifibrozil 2,030 2,051 Scandinavian Simvastatin Survival Study Simvastatin 2,223 2,221 CARE 5 Pravastatin 2,078 2,081 WOSCOPS 6 Pravastatin 3,293 3,302 AFCAPS/TexCAPS 7 Lovastatin (with or without) cholestyramine 3,301 3,304 LIPID 8 Pravastatin 4,502 4,512 VAHIT 33 Gemifibrozil 1,267 1,264 Heart Protection Study 34 Simvastatin 10,267 10,269 PROSPER 35 Pravastatin 2,913 2,891 Total 33,774 33,801 All statin trials 28,577 28,580 *CARE Cholesterol and Recurrent Events Trial; WOSCOPS West of Scotland Coronary Prevention Study; AFCAPS/TexCAPS Air Force/Texas Coronary Atherosclerosis Prevention Study; LIPID Long-Term Intervention with Pravastatin in Ischaemic Disease; VAHIT Veterans Affairs High Density Cholesterol Intervention Trial Study Group; PROSPER Prospective Study of Pravastatin in the Elderly at Risk. high scores on the HRQL domains. It seems likely that this potential effect would be small, given that the majority of patients not included in these analyses were excluded because data were not collected at baseline, rather than because of loss of patients to follow-up. Further, the low mortality was similar across the treatment groups, and does not change the impression that no differences occurred among treatment groups in HRQL outcomes. In summary, POST CABG Trial participants did not experience detrimental or beneficial effects on HRQL parameters while receiving LDL cholesterollowering therapy that had demonstrable benefits for treatment of their atherosclerosis. These data support the concept that the recommendation of the National Cholesterol Education Program to lower LDL cholesterol level 100 mg/dl in patients with coronary artery disease may be accomplished without apparent adverse effects on HRQL. 36 References 1 The Post Coronary Artery Bypass Graft Trial Investigators. The effect of aggressive lowering of low-density lipoprotein cholesterol levels and low-dose anticoagulation on obstructive changes in saphenous-vein coronary-artery bypass grafts. N Engl J Med 1997; 336: Knatterud GL, Rosenberg Y, Campeau L, et al. Long-term effects on clinical outcomes of aggressive lowering of lowdensity lipoprotein cholesterol levels and low-dose anticoagulation in the Post Coronary Artery Bypass Graft Trial. Circulation 2000; 102: Scandinavian Simvastin Survival Study Group. Randomized trial of cholesterol lowering in 4,444 patients with coronary heart disease: the Scandinavian Simvastin Survival Study (4S). Lancet 1994; 344: Pedersen TR, Wilhelmsen L, Faergeman O, et al, on behalf of the Scandinavian Simvastatin Survival Study Group. Follow-up study of patients randomized in the Scandinavian Simvastatin Survival Study (4S) of cholesterol lowering. Am J Cardiol 2000; 86: Sacks FM, Pfeffer MA, Moye LA, et al. The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. N Engl J Med 1996; 335: Shepherd J, Cobbe SM, Ford I, et al. Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia: West of Scotland Coronary Prevention Study Group. N Engl J Med 1995; 333: Downs JR, Clearfield M, Weis S, et al. Primary prevention of acute coronary events with lovastatin in men and women with average cholesterol levels: results of AFCAPS/TexCAPS. JAMA 1998; 279: The Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) Study Group. Prevention of cardiovascular events and death with pravastatin in patients with coronary heart disease and a broad range of initial cholesterol levels. N Engl J Med 1998; 339: Lipid Research Clinics Program. The Lipid Research Clinics Coronary Primary Prevention Trial results: I. Reduction in incidence of coronary heart disease. JAMA 1984; 251: Frick MH, Elo O, Haapa K, et al. Helsinki Heart Study: primary prevention trial with gemfibrozil in middle-aged men with dyslipidemia. N Engl J Med 1987; 317: Muldoon MF, Manuck SB, Matthews KA. Lowering cholesterol concentrations and mortality: a quantitative review of primary prevention trials. BMJ 1990; 301: Gould AL, Rossouw JE, Santanello NC, et al. Cholesterol reduction yields clinical benefit: impact of statin trials. Circulation 1998; 97: Muldoon M, Rossouw JE, Manuck SB, et al. Low or lowered cholesterol and risk of death from suicide and trauma. Metabolism 1993; 42(9 Suppl 1): Wardle J. Cholesterol and psychological well-being. J Psychosom Res 1995; 39: Golomb BA. Cholesterol and violence: is there a connection? Ann Intern Med 1998; 128: Lindberg G, Rastam L, Gullberg B, et al. Low serum cholesterol concentration and short term mortality from 494 Clinical Investigations

9 injuries in men and women. BMJ 1992; 305: Lindberg G, Larsson G, Setterlind S, et al. Serum lipids and mood in working men and women in Sweden. J Epidemiol Community Health 1994; 48: Benton D. Do low cholesterol levels slow mental processing? Psychosom Med 1995; 57: Weidner G, Connor SL, Hollis JF, et al. Improvements in hostility and depression in relation to dietary change and cholesterol lowering: the Family Heart Study. Ann Intern Med 1992; 117: Wardle J, Armitage J, Collins R, et al. Randomized placebo controlled trial of effect on mood of lowering cholesterol concentration: Oxford Cholesterol Study Group. BMJ 1996; 313: Hirsh J. Influence of low-intensity warfarin treatment on patients perceptions of quality of life. Arch Intern Med 1991; 151: Lancaster TR, Singer DE, Sheehan MA, et al. The impact of long-term warfarin therapy on quality of life: evidence from a randomized trial; Boston Area Anticoagulation Trial for Atrial Fibrillation Investigators. Arch Intern Med 1991; 151: Spielberger CD, Gorsuch RL, Lushene RE. Manual for the state-trait anxiety inventory. Palo Alto, CA. Consulting Psychologists, Radloff L. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas 1977; 1: Jette AM, Davies AR, Cleary PD, et al. The functional status questionnaire: reliability and validity when used in primary care. J Gen Intern Med 1986; 1: Stewart AL, Hays RD, Ware JE Jr. The MOS short-form general health survey: reliability and validity in a patient population. Med Care 1988; 26: Melzack R. The short-form McGill pain questionnaire. Pain 1987; 30: Czajkowski SM, Terrin M, Lindquist R, et al, for the POST CABG Biobehavioral Study Investigators. Comparison of preoperative characteristics of men and women undergoing coronary artery bypass grafting (The Post Coronary Artery Bypass Graft [CABG] Biobehavioral Study). Am J Cardiol 1997; 79: Lindquist R, Dupuis G, Terrin ML, et al, for the POST CABG Biobehavioral Study Investigators. Comparison of health-related quality-of-life outcomes in men and women after coronary artery bypass surgery through 1 year: findings from the POST CABG Biobehavioral Study. Am Heart J 2003; 146: Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959; 22: Graybill PA. Theory and application of the linear model. North Boston, MA: Duxbury Press, 1976; Pocock SJ, Henderson RA, Seed P, et al, for the RITA trial participants. Quality of life, employment status, and anginal symptoms after coronary angioplasty or bypass surgery: 3-year follow-up in the Randomized Intervention of Angina (RITA) trial. Circulation 1996; 94: Rubins HB, Robins SJ, Collins D, et al, for the Veterans Affairs High Density Cholesterol Intervention Trial Study Group. Gemfibrozil of the secondary prevention of coronary heart disease in men with low levels of high-density lipoprotein cholesterol. N Engl J Med 1999; 341: Heart Protection Study Collaborative Group. MRC/BHF heart protection study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomized placebocontrolled trial. Lancet 2002; 360: PROSPER Study Group. Pravastatin in elderly individuals at risk of vascular disease (PROSPER): a randomized controlled trial. Lancet 2002; 360: Executive Summary of the Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and of High Blood Cholesterol in Adults (adult treatment panel III). JAMA 2001; 285: CHEST / 126 / 2/ AUGUST,