Abstract

Psychosocial characteristics may be associated with an increased risk of coronary heart disease (CHD). Whether hostility predicts recurrent coronary events is unknown. A total of 792 women in the Heart and Estrogen/progestin Replacement Study (HERS) were evaluated prospectively to determine the role of hostility as a risk factor for secondary CHD events (nonfatal myocardial infarction and CHD death). The mean age of study participants was 67 years, and the average length of follow-up was 4.1 years. The study was conducted between 1993 and 1998, and all study sites were in the United States. High Cook-Medley hostility scores were associated with greater body mass index (p = 0.01) and higher levels of serum triglycerides (p = 0.05), and they were inversely associated with high density lipoprotein cholesterol (p = 0.04), self-rated general health (p < 0.001), age (p = 0.05), and education (p = 0.001). Compared with women in the lowest hostility score quartile, women in the highest quartile were twice as likely to have had a myocardial infarction (relative hazard = 2.03, 95% confidence interval: 1.02, 4.01). The relation between hostility and CHD events was not mediated or confounded by the biologic, behavioral, and social risk factors studied. In this study, hostility was found to be an independent risk factor for recurrent CHD events in postmenopausal women.

Received for publication January 22, 2002; accepted for publication May 29, 2002.

Epidemiologic evidence suggests that psychological behaviors may be associated with coronary heart disease (CHD) risk. Hostility, a construct that includes cynicism, anger, mistrust, and aggression (1), has been correlated with carotid atherosclerosis (2, 3), angiographic coronary artery disease (4, 5), exercise-induced ischemia (6),and restenosis after mechanical revascularization in women (7). High hostility scores have also been associated with an increased risk of nonfatal myocardial infarction among older women (8, 9) but not with fatal events in patients with documented CHD (10).While these observations suggest a link between hostility and CHD, the association between hostility and recurrent CHD outcomes has not been established.

To examine the relation between hostility and CHD risk in a clinical population, we analyzed data collected from a cohort of postmenopausal women with CHD enrolled in the Heart and Estrogen/progestin Replacement Study (HERS). Although prior research suggests that the risk associated with psychosocial factors occurs early in the disease process (4, 1113), several of the proposed mechanisms linking hostility and CHD could also contribute to recurrent events. Therefore, our objective was to determine whether hostility affects cardiovascular health once advanced atherosclerosis is present. We evaluated the effect of hostility on CHD outcomes in older women, the effect of postmenopausal hormone therapy on hostility score, and the interaction between hormone therapy and level of hostility.

MATERIALS AND METHODS

Study design

To determine the association between hostility and CHD outcomes in postmenopausal women with known coronary disease, we conducted an ancillary prospective cohort study (HERS Hostility Ancillary Study) within HERS. HERS was a randomized, blinded, placebo-controlled trial to test the effect of postmenopausal hormone therapy on CHD events in women with established CHD. Women were randomly assigned to receive 0.625 mg of conjugated equine estrogen plus 2.5 mg of medroxyprogesterone in one capsule or an identical placebo. The methods and primary results from HERS have been published previously (14, 15).

Participants and setting

Entry into HERS required that participants be younger than age 80 years and have had documented coronary disease, defined by prior myocardial infarction, coronary artery bypass graft surgery, mechanical coronary revascularization, or angiographic evidence of at least a 50 percent occlusion of one or more major coronary arteries. Exclusion criteria included a clinical cardiac event (myocardial infarction or coronary revascularization procedure) within 6 months of randomization, a contraindication to hormone therapy or having received hormone therapy within the 3 months prior to the screening visit, uncontrolled hypertension (systolic blood pressure ≥200 mmHg or diastolic blood pressure ≥105 mmHg), uncontrolled diabetes mellitus (fasting blood sugar ≥300 mg/dl), elevated triglycerides (≥300 mg/dl), New York Heart Association class IV or severe class III congestive heart failure, or a disease judged to be fatal within 4 years.

We enrolled 792 of the 2,763 HERS participants from eight of the HERS clinical centers: Hartford, Connecticut; Baltimore, Maryland; Pittsburgh, Pennsylvania (Oakland and West Penn); Chicago, Illinois; Iowa City, Iowa; La Jolla, California; and Stanford, California. Geographic selections were based on the interest of the principal investigator at each clinical site. All study sites were in the United States. The HERS Hostility Ancillary Study was conducted between 1993 and 1998.

Measurements

The Cook-Medley hostility scale (16), a 50-item component of the Minnesota Multiphasic Personality Inventory (MMPI), was used to measure hostility at the baseline, year 1, and closeout visits. To increase variance in the scores, the original yes/no format was revised to a four-item response scale: always true (three points), usually true (two points), usually false (one point), always false (zero points). Hence, the Cook-Medley hostility scale ranged from zero (nonhostile) to 150 (very hostile). Several studies have used the Cook-Medley cynicism subscale to evaluate the association between hostility and CHD events (9, 17, 18); thus, we also included the 13-item cynicism subscale in our analysis. When converted to the four-item response format, the cynicism scale ranges from zero (noncynical) to 39 (highly cynical).

Follow-up visits were conducted by each of the clinic sites and included an assessment of compliance with study medications and the recording of outcome and adverse events. Primary outcomes were nonfatal myocardial infarction (symptomatic or silent), CHD death (fatal myocardial infarction, sudden death within 1 hour of symptom onset, unobserved death in the absence of other known cause, and death during sleep), and CHD event (nonfatal myocardial infarction and CHD death). This definition of CHD death was narrower than the one used in the main HERS report (15). In the Hostility Ancillary Study, deaths due to coronary revascularization procedure or congestive heart failure were analyzed separately. Secondary cardiovascular outcomes included hospitalization for unstable angina, coronary artery bypass graft surgery, percutaneous transluminal coronary angioplasty, stroke, transient ischemic attack, and peripheral arterial disease. Data pertaining to outcome measures were collected by the Coordinating Center at the University of California San Francisco. Two physicians at the HERS Coordinating Center, who were blinded to treatment assignment, independently reviewed all outcome events. An independent committee of cardiologists, also blinded to treatment assignment, adjudicated all primary outcomes. Participants were followed for an average of 4.1 years (range, 3.6–5.3 years).

Data analysis

Enrollment into HERS (February 1993–September 1994) was initiated before the Hostility Ancillary Study began; therefore, not all participants were available for a baseline evaluation. A total of 425 women (54 percent of ancillary study participants) completed the Cook-Medley hostility questionnaire at baseline, and an additional 367 women (46 percent) were evaluated for the first time at the year 1 visit. Among the 792 participants, 630 (80 percent) completed a closeout questionnaire.

To evaluate the effect of postmenopausal hormone therapy, hostility scores at baseline, year 1, and closeout were compared by using linear mixed models. To analyze hostility as a risk factor for cardiovascular events, participants were classified by quartile of hostility score. Among women who had baseline and year 1 evaluations, this classification was made by using the baseline score. We used univariate and multivariate proportional odds models to identify the independent predictors of hostility quartile among baseline biologic, behavioral, and social risk factors for cardiovascular disease.

The incidence of CHD events was initially compared by quartile of hostility score using Kaplan-Meier survival curves. Independent associations between hostility quartile and CHD outcomes were assessed by using multivariable Cox proportional hazards regression models. Covariate selection was based on published data and known CHD risk factors in HERS.

Biologic covariates included body mass index (kg/m2); serum low density lipoprotein cholesterol, high density lipoprotein cholesterol, and triglyceride levels (mg/dl); lipoprotein(a) >25.3 mg/dl; creatinine clearance <40 ml/minute (Cockcroft-Gault equation (19)); diabetes (use of diabetic medications or self-reported history of diabetes); hypertension (systolic blood pressure >140 mmHg or diastolic blood pressure >90 mmHg, use of antihypertensive medication, or self-reported history of hypertension); prior history of myocardial infarction (≥2); and fair-to-poor self-rated general health. Demographic variables included age (years), race (White, non-White), education (years), and marital status (married, not married). Behavioral measures included current tobacco use, any current alcohol consumption, exercise (≥3 times per week), and medication use (beta-blocker, aspirin, or 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitor). The significant risk factors for nonfatal myocardial infarction or CHD death in HERS include non-White race, hypertension, low density lipoprotein cholesterol >130 mg/dl, high density lipoprotein cholesterol <35 mg/dl, diabetes, renal insufficiency (creatinine clearance <40 ml/minute), two or more prior myocardial infarctions, heart failure symptoms, and nonuse of HMG-CoA reductase inhibitors.

In a supplementary analysis, adjusted Cox model relative hazards were compared with unadjusted estimates to assess potential mediation of the relation between hostility and study outcomes. We also tested for interactions between medication (i.e., 0.625 mg of conjugated equine estrogen plus 2.5 mg of medroxyprogesterone, beta-blocker, aspirin, and HMG-CoA reductase inhibitor) and hostility score to determine whether treatment modified the effect of hostility or whether hostility modified the association between use of beta-blockers or HMG-CoA reductase inhibitors and cardiovascular outcomes. Finally, we examined interactions between hostility score and time of entry into the Hostility Ancillary Study.

In all survival analyses, follow-up was initiated, and covariates assessed, at the time of entry into the Hostility Ancillary Study. The proportional hazards assumption was verified by using log-log survival plots and by testing for interactions between hostility and time since randomization. Two-tailed tests in which p < 0.05 were considered statistically significant.

All analyses were performed by using SAS software, version 6.12 (SAS Institute Inc., Cary, North Carolina).

RESULTS

Treatment assignment was not associated with mean hostility scores, and hostility scores remained constant over time (table 1). We also found no evidence for an interaction between treatment assignment and hostility quartile. Thus, we assessed the effect of hostility on cardiovascular events in both groups combined.

The median Cook-Medley hostility score for the 792 HERS Hostility Ancillary Study participants was 51 (interquartile range, 41–60). High hostility scores were associated with the interrelated variables of increased body mass index (p = 0.01), lower high density lipoprotein cholesterol level (p = 0.04), and higher serum triglyceride level (p = 0.05) and were inversely associated with self-rated general health (p < 0.001), age (p = 0.05), and education (p = 0.001) (table 2). Behavioral risk factors for CHD, including tobacco use, alcohol consumption, and low levels of physical activity, were not significantly associated with hostility score. Among the variables studied, increased body mass index, fair-to-poor general health, lower levels of education, and current tobacco use appeared to be the strongest independent correlates of hostility.

Kaplan-Meier survival estimates for time to CHD events were determined for each hostility quartile (figure 1). Compared with women whose hostility scores were in the lowest quartile (score <41), women whose scores were in the highest quartile (score >60) were at an increased risk of CHD events (nonfatal myocardial infarction and CHD death; relative hazard (RH) = 1.88, 95 percent confidence interval (CI): 1.01, 3.53) and nonfatal myocardial infarction (RH = 2.03, 95 percent CI: 1.02, 4.01) (table 3). Risks for either endpoint among women in the second and third quartiles were not significantly different from those for women in the first quartile, suggesting a threshold effect near the upper quartile of the hostility score distribution. However, a linear trend in risk across quartiles could not be rejected in favor of a threshold model (p = 0.33). As a continuous variable, hostility scores predicted CHD events (RH per 10-point increase in hostility score = 1.15, 95 percent CI: 1.0, 1.3) but not other CHD outcomes.

In contrast to the results for CHD events and nonfatal myocardial infarction, risk of a CHD death among women in the uppermost quartile was not clearly increased (RH =1.34, 95 percent CI: 0.28, 6.47). Similarly, other cardiovascular outcomes in HERS, including hospitalization for unstable angina, revascularization procedures (i.e., coronary artery bypass graft surgery, percutaneous transluminal coronary angioplasty), congestive heart failure, stroke, transient ischemic attack, and peripheral arterial disease, were not significantly associated with high hostility scores. The results for adjusted and unadjusted models were similar, suggesting that the relation between hostility and CHD was not mediated or confounded by the biologic, behavioral, and social factors that we examined. In addition, we found no evidence of an interaction between hostility quartile and use of beta-blockers, use of HMG-CoA reductase inhibitors, or time of entry into the Hostility Ancillary Study.

The median cynical hostility subscale score was seven (interquartile range, six–nine). Results from analyses including cynicism subscores were similar to the hostility score findings. Compared with women whose cynicism scores were in the first quartile, women whose scores were in the fourth quartile were at greater risk of CHD events (RH = 2.12, 95 percent CI: 1.09, 4.12) and nonfatal myocardial infarction (RH = 2.31, 95 percent CI: 1.08, 4.96), independent of other CHD risk factors. Cynicism scores in the fourth quartile were not significantly associated with CHD death (RH = 1.42, 95 percent CI: 0.39, 5.13) or with the secondary CHD outcomes of unstable angina (RH = 0.60, 95 percent CI: 0.26, 1.37) and coronary artery bypass graft surgery/ percutaneous transluminal coronary angioplasty (RH = 1.11, 95 percent CI: 0.71, 1.74). The correlation between hostility scores and cynicism subscores was high (r = 0.84).

DISCUSSION

Prior reports indicate that psychosocial characteristics may be a risk factor for CHD in younger adults but that older populations might fail to demonstrate this association (4, 1113). However, we have shown that postmenopausal women with high Cook-Medley hostility scores are at a twofold increased risk for recurrent CHD events. This association was statistically significant and independent of other cardiovascular risk factors. Among HERS women, the risk associated with high levels of hostility was greater than that associated with elevated serum low density lipoprotein cholesterol levels, tobacco use, hypertension, and diabetes (data not shown).

Several mechanisms have been proposed to explain the association between hostility and CHD. The health behaviors model suggests that hostility serves as a marker for behaviors that increase risk for disease (20). In one study of middle-aged men (21), associations between cynicism and cardiovascular mortality were mediated by adverse health practices related to CHD outcomes (e.g., tobacco use, alcohol consumption, and low activity levels). However, in the HERS Hostility Ancillary Study, we found no evidence that the relation between hostility and CHD events was mediated or confounded by the behavioral and social risk factors that were measured.

An alternative mechanism linking hostility and CHD involves the sympathetic nervous system. Exaggerated adrenergic responses to stressful psychological stimuli can increase heart rate and blood pressure (2224), induce coronary vasoconstriction in atherosclerotic arteries (25, 26), and activate platelets (27). A triggering event, such as an outburst of anger, may precipitate a response that promotes the acute rupture of vulnerable plaque and thrombosis, increasing the risk of nonfatal myocardial infarction or CHD death (2830). Additionally, persons who have high levels of hostility tend to have more negative world views and feelings of chronic cynicism and mistrust, which may result in continual sympathetic nervous system arousal (31). Although the proximity to a stressful event was not examined in HERS, our finding that hostility is associated with nonfatal myocardial infarction and CHD death, but not with angina or revascularization procedure, is consistent with the cardiovascular reactivity hypothesis. Revascularization procedures are often elective and not representative of an acute pathophysiologic process. This finding also suggests that the results are not due to differential patient or physician behavior.

In the Canadian Amlodipine/Atenolol in Silent Ischemia Study (CASIS), high levels of hostility diminished the benefits of anti-ischemic medications (i.e., amlodipine, atenolol) (32). We did not find evidence of an interaction between hostility and medication use (i.e., beta-blockers and HMG-CoA reductase inhibitors). In addition, beta-blockers may also modify the sympathetic response to a triggering event, such as an episode of anger, thereby reducing the risk of plaque rupture (28, 30). In the HERS Hostility Ancillary Study, beta-blockers did not appear to influence the association between hostility and CHD events. However, use of medications other than estrogen and progestin (0.625 mg of conjugated equine estrogen plus 2.5 mg of medroxyprogesterone) was not randomized in HERS, so our ability to detect interactions may have been limited and subject to confounding by indication.

Specific traits in the hostility construct may act independently on CHD risk. We found an equally strong association with CHD events by using the cynicism subscore of the Cook-Medley hostility scale. This finding is in contrast to several prior studies that failed to find a significant relation between cynicism scores and CHD outcomes (9, 17, 18). Discrepancies between our research findings and those of other investigators may be due to differences in the populations studied (e.g., age, gender, CHD risk). However, when compared with the global Cook-Medley hostility score, cynicism subscores were only marginally less predictive of acute CHD events in older women (9). The HERS Hostility Ancillary Study findings indicate that, among postmenopausal women with CHD, the cynicism subscale and the full Cook-Medley hostility scale may be measuring similar psychosocial characteristics.

The Hostility Ancillary Study included a subgroup of women from selected HERS sites. Thus, although the women in HERS were comparable to women with CHD in the Third National Health and Nutrition Examination Survey (NHANES III) (33), a national probability sample of persons in the United States,our subsample may be less representative. Furthermore, our study findings may not apply to women without established CHD or to men.

Since the mid-1960s, mortality rates from cardiovascular disease have been declining, with the greatest decreases attributed to changes in risk factors (e.g., lifestyle modifications, tobacco use) (34). However, the variance in CHD is only partially explained by traditional risk factors, and, despite favorable trends, cardiovascular disease remains a leading cause of morbidity and mortality in women. As the population ages and the prevalence of women with CHD rises, efforts to identify modifiable CHD risk factors will become increasingly important. Prior research suggests that interventions during cardiac rehabilitation, such as counseling sessions, breathing-relaxation therapy, health education, and exercise, may favorably affect coronary-prone psychosocial characteristics and health outcomes (3537).

In summary, we have shown that hostility is a significant independent predictor of CHD events in postmenopausal women with known CHD. Further research is needed to confirm these findings and to assess the possible role of interventions directed at behavioral risk factors.

ACKNOWLEDGMENTS

The Heart and Estrogen/progestin Replacement Study (HERS) was supported by a contract with Wyeth-Ayerst Research.

Correspondence to Dr. Lily A. Chaput, UCSF/Mt. Zion Women’s Health Clinical Research Center, University of California San Francisco, 1635 Divisadero Street, Suite 600, San Francisco, CA 94115 (e-mail: lily.chaput@ucsfmedctr.org).

FIGURE 1. Kaplan-Meier estimates of the cumulative incidence of coronary heart disease events (nonfatal myocardial infarction and coronary heart disease death) by hostility score quartile, Hostility Ancillary Study, Heart and Estrogen/progestin Replacement Study, United States, 1993–1998.

FIGURE 1. Kaplan-Meier estimates of the cumulative incidence of coronary heart disease events (nonfatal myocardial infarction and coronary heart disease death) by hostility score quartile, Hostility Ancillary Study, Heart and Estrogen/progestin Replacement Study, United States, 1993–1998.

TABLE 1.

Mean hostility score over time, stratified by treatment group, Hostility Ancillary Study, Heart and Estrogen/progestin Replacement Study, United States, 1993–1998

Visit E + P  Placebo p value* 
No. Mean (SE†)  No. Mean (SE) 
Baseline 214 49.8 (1.0)  211 49.7 (1.0) 0.91 
Year 1 339 50.9 (0.8)  342 50.8 (0.8) 0.87 
Closeout 311 51.1 (0.9)  319 50.9 (0.9) 0.92 
Visit E + P  Placebo p value* 
No. Mean (SE†)  No. Mean (SE) 
Baseline 214 49.8 (1.0)  211 49.7 (1.0) 0.91 
Year 1 339 50.9 (0.8)  342 50.8 (0.8) 0.87 
Closeout 311 51.1 (0.9)  319 50.9 (0.9) 0.92 

* p values are from linear mixed models for the treatment effect; p values for tests of homogeneity within treatment groups = 0.38 (0.625 mg conjugated equine estrogen plus 2.5 mg medroxyprogesterone (E + P)) and 0.37 (placebo).

† SE, standard error.

TABLE 2.

Baseline cardiovascular risk factors, by hostility score quartile (n = 792), Hostility Ancillary Study, Heart and Estrogen/progestin Replacement Study, United States, 1993–1998

Variables Quartile 1 (4–40) (n = 201) Quartile 2 (41–51) (n = 202) Quartile 3 (52–60) (n = 191) Quartile 4 (61–113) (n = 198) p value 
Biologic      
Body mass index (mean kg/m2 28.1  29.1  29.4  29.5  0.01 
Low density lipoprotein cholesterol (mean mg/dl)* 135 136 136 140  0.24 
High density lipoprotein cholesterol (mean mg/dl)*  53  52  51  50  0.04 
Triglycerides (mean mg/dl)† 168 163  178 178  0.05 
Lipoprotein(a) >25.3 mg/dl (%)‡  45  44  49  41  0.68 
Diabetes (%)  23  22  30  27  0.17 
Hypertension (%)  63  69  73  67  0.25 
Prior myocardial infarction ≥2 (%)   3   4   8   4  0.47 
Creatinine clearance <40 ml/minute (%)  12   9  14  12  0.73 
Fair-poor self-rated health (%)  16  14  23  29 <0.001 
Demographic      
Age (mean years)   68  67  66  66  0.05 
White race (%)  96  91  93  93  0.49 
Education (mean years)  13.0  13.4  13.0  12.3  0.001 
Married (%)  59  61  62  51  0.13 
Behavioral      
Current tobacco use (%)   9  10  10  15  0.12 
Any alcohol consumption (%)  46  54  45  44  0.37 
Exercise ≥3 times/week (%)  44  35  38  34  0.07 
Medication use      
Beta-blocker (%)  36  38  40  31  0.46 
Aspirin (%)  81  80  82  73  0.09 
HMG-CoA§ reductase inhibitors (%)  41  41  39  42  0.95 
Variables Quartile 1 (4–40) (n = 201) Quartile 2 (41–51) (n = 202) Quartile 3 (52–60) (n = 191) Quartile 4 (61–113) (n = 198) p value 
Biologic      
Body mass index (mean kg/m2 28.1  29.1  29.4  29.5  0.01 
Low density lipoprotein cholesterol (mean mg/dl)* 135 136 136 140  0.24 
High density lipoprotein cholesterol (mean mg/dl)*  53  52  51  50  0.04 
Triglycerides (mean mg/dl)† 168 163  178 178  0.05 
Lipoprotein(a) >25.3 mg/dl (%)‡  45  44  49  41  0.68 
Diabetes (%)  23  22  30  27  0.17 
Hypertension (%)  63  69  73  67  0.25 
Prior myocardial infarction ≥2 (%)   3   4   8   4  0.47 
Creatinine clearance <40 ml/minute (%)  12   9  14  12  0.73 
Fair-poor self-rated health (%)  16  14  23  29 <0.001 
Demographic      
Age (mean years)   68  67  66  66  0.05 
White race (%)  96  91  93  93  0.49 
Education (mean years)  13.0  13.4  13.0  12.3  0.001 
Married (%)  59  61  62  51  0.13 
Behavioral      
Current tobacco use (%)   9  10  10  15  0.12 
Any alcohol consumption (%)  46  54  45  44  0.37 
Exercise ≥3 times/week (%)  44  35  38  34  0.07 
Medication use      
Beta-blocker (%)  36  38  40  31  0.46 
Aspirin (%)  81  80  82  73  0.09 
HMG-CoA§ reductase inhibitors (%)  41  41  39  42  0.95 

* To convert to mmol/liter, multiply values by 0.0259.

† To convert to mmol/liter, multiply values by 0.0113.

‡ To convert to mmol/liter, multiply values by 0.0357.

§ HMG-CoA, 3-hydroxy-3-methylglutaryl coenzyme A.

TABLE 3.

Risk of recurrent coronary heart disease events, by hostility score quartile (n = 792), Hostility Ancillary Study, Heart and Estrogen/progestin Replacement Study, United States, 1993–1998

Variables Quartile 1 Quartile 2 Quartile 3 Quartile 4 
  (4–40) (41–51)  (52–60)  (61–113) 
 (n = 201)  (n = 202) (n = 191) (n = 198) 
CHD* events (nonfatal MI* and CHD death)     
No. 16 15 19 29 
Unadjusted RH* (95% CI*)  1.0 0.95 (0.47, 1.92) 1.28 (0.66, 2.48) 1.95 (1.06, 3.59) 
Adjusted RH† (95% CI)  1.0 1.09 (0.53, 2.24) 1.05 (0.53, 2.09) 1.88 (1.01, 3.53) 
Nonfatal MI     
No. 13 13 13 26 
Unadjusted RH (95% CI)  1.0 1.01 (0.47, 2.18) 1.07 (0.50, 2.31) 2.14 (1.10, 4.17) 
Adjusted RH† (95% CI)  1.0 1.20 (0.55, 2.63) 0.91 (0.41, 1.99) 2.03 (1.02, 4.01) 
CHD death     
No.  3 
Unadjusted RH (95% CI)  1.0 2.03 (0.51, 8.12) 2.12 (0.53, 8.50) 1.37 (0.31, 6.13) 
Adjusted RH† (95% CI)  1.0  2.38 (0.52, 10.82) 1.96 (0.43, 8.96) 1.34 (0.28, 6.47) 
Unstable angina     
No. 18 14 11 
Unadjusted RH (95% CI)  1.0 0.78 (0.39, 1.57) 0.34 (0.14, 0.86) 0.63 (0.30, 1.34) 
Adjusted RH† (95% CI)  1.0 0.81 (0.39, 1.68) 0.33 (0.13, 0.86) 0.76 (0.35, 1.64) 
CABG*/PTCA*     
No. 43 33 27 39 
Unadjusted RH (95% CI)  1.0 0.74 (0.47, 1.17) 0.64 (0.39, 1.03) 0.94 (0.61, 1.45) 
Adjusted RH† (95% CI)  1.0 0.72 (0.45, 1.14) 0.59 (0.36, 0.96) 0.93 (0.60, 1.45) 
Stroke/TIA*     
No. 16 15 14 
Unadjusted RH (95% CI)  1.0 0.93 (0.46, 1.88) 0.59 (0.26, 1.32) 0.89 (0.44, 1.83) 
Adjusted RH† (95% CI)  1.0 1.07 (0.49, 2.30) 0.55 (0.23, 1.30) 0.83 (0.39, 1.78) 
Peripheral arterial disease     
No. 15 18 13 13 
Unadjusted RH (95% CI)  1.0 1.21 (0.61, 2.40) 0.92 (0.44, 1.93) 0.90 (0.43, 1.89) 
Adjusted RH† (95% CI)  1.0 1.17 (0.57, 2.40) 0.84 (0.39, 1.79) 0.85 (0.40, 1.83) 
Variables Quartile 1 Quartile 2 Quartile 3 Quartile 4 
  (4–40) (41–51)  (52–60)  (61–113) 
 (n = 201)  (n = 202) (n = 191) (n = 198) 
CHD* events (nonfatal MI* and CHD death)     
No. 16 15 19 29 
Unadjusted RH* (95% CI*)  1.0 0.95 (0.47, 1.92) 1.28 (0.66, 2.48) 1.95 (1.06, 3.59) 
Adjusted RH† (95% CI)  1.0 1.09 (0.53, 2.24) 1.05 (0.53, 2.09) 1.88 (1.01, 3.53) 
Nonfatal MI     
No. 13 13 13 26 
Unadjusted RH (95% CI)  1.0 1.01 (0.47, 2.18) 1.07 (0.50, 2.31) 2.14 (1.10, 4.17) 
Adjusted RH† (95% CI)  1.0 1.20 (0.55, 2.63) 0.91 (0.41, 1.99) 2.03 (1.02, 4.01) 
CHD death     
No.  3 
Unadjusted RH (95% CI)  1.0 2.03 (0.51, 8.12) 2.12 (0.53, 8.50) 1.37 (0.31, 6.13) 
Adjusted RH† (95% CI)  1.0  2.38 (0.52, 10.82) 1.96 (0.43, 8.96) 1.34 (0.28, 6.47) 
Unstable angina     
No. 18 14 11 
Unadjusted RH (95% CI)  1.0 0.78 (0.39, 1.57) 0.34 (0.14, 0.86) 0.63 (0.30, 1.34) 
Adjusted RH† (95% CI)  1.0 0.81 (0.39, 1.68) 0.33 (0.13, 0.86) 0.76 (0.35, 1.64) 
CABG*/PTCA*     
No. 43 33 27 39 
Unadjusted RH (95% CI)  1.0 0.74 (0.47, 1.17) 0.64 (0.39, 1.03) 0.94 (0.61, 1.45) 
Adjusted RH† (95% CI)  1.0 0.72 (0.45, 1.14) 0.59 (0.36, 0.96) 0.93 (0.60, 1.45) 
Stroke/TIA*     
No. 16 15 14 
Unadjusted RH (95% CI)  1.0 0.93 (0.46, 1.88) 0.59 (0.26, 1.32) 0.89 (0.44, 1.83) 
Adjusted RH† (95% CI)  1.0 1.07 (0.49, 2.30) 0.55 (0.23, 1.30) 0.83 (0.39, 1.78) 
Peripheral arterial disease     
No. 15 18 13 13 
Unadjusted RH (95% CI)  1.0 1.21 (0.61, 2.40) 0.92 (0.44, 1.93) 0.90 (0.43, 1.89) 
Adjusted RH† (95% CI)  1.0 1.17 (0.57, 2.40) 0.84 (0.39, 1.79) 0.85 (0.40, 1.83) 

* CHD, coronary heart disease; MI, myocardial infarction; RH, relative hazard; CI, confidence interval; CABG, coronary artery bypass graft surgery; PTCA, percutaneous transluminal coronary angioplasty; TIA, transient ischemic attack.

† Multivariate models were adjusted for body mass index, low density lipoprotein cholesterol, high density lipoprotein cholesterol, triglycerides, lipoprotein(a), diabetes, hypertension, prior MI ≥2, creatinine clearance <40 ml/minute, self-rated general health, age, race, education, marital status, tobacco use, alcohol consumption, exercise ≥3 times/week, and use of beta-blocker, aspirin, and 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors.

References

1.
Barefoot JC, Dodge KA, Peterson BL, et al. The Cook-Medley hostility scale: item content and ability to predict survival.
Psychosom Med
 
1989
;
51
:
46
–57.
2.
Knox SS, Adelman A, Ellison RC, et al. Hostility, social support, and carotid artery atherosclerosis in the National Heart, Lung, and Blood Institute Family Heart Study.
Am J Cardiol
 
2000
;
86
:
1086
–9.
3.
Matthews KA, Owens JF, Kuller LH, et al. Are hostility and anxiety associated with carotid atherosclerosis in healthy postmenopausal women?
Psychom Med
 
1998
;
60
:
633
–8.
4.
Low KG, Fleisher C, Colman R, et al. Psychosocial variables, age, and angiographically-determined coronary artery disease in women.
Ann Behav Med
 
1998
;
20
:
221
–6.
5.
Williams RB, Haney TL, Lee KL, et al. Type A behavior, hostility, and coronary atherosclerosis.
Psychosom Med
 
1980
;
42
:
539
–49.
6.
Helmers KF, Krantz DS, Howell RH, et al. Hostility and myocardial ischemia in coronary artery disease patients: evaluation by gender and ischemic index.
Psychosom Med
 
1993
;
55
:
29
–36.
7.
Goodman M, Quigley J, Moran G, et al. Hostility predicts re-stenosis after percutaneous transluminal coronary angioplasty.
Mayo Clin Proc
 
1996
;
71
:
729
–34.
8.
Lahad A, Heckbert SR, Koepsell TD, et al. Hostility, aggression and the risk of nonfatal myocardial infarction in postmenopausal women.
J Psychosom Res
 
1997
;
43
:
183
–95.
9.
Barefoot JC, Larsen S, von der Lieth L, et al. Hostility, incidence of acute myocardial infarction, and mortality in a sample of older Danish men and women.
Am J Epidemiol
 
1995
;
142
:
477
–84.
10.
Barefoot JC, Peterson BL, Harrell FE, et al. Type A behavior and survival: a follow-up study of 1,467 patients with coronary artery disease.
Am J Cardiol
 
1989
;
64
:
427
–32.
11.
Kop W. Acute and chronic psychological risk factors for coronary syndromes: moderating effects of coronary artery disease severity.
J Psychosom Res
 
1997
;
43
:
167
–81.
12.
Williams RB, Barefoot JC, Haney TL, et al. Type A behavior and angiographically documented coronary atherosclerosis in a sample of 2,289 patients.
Psychosom Med
 
1988
;
50
:
139
–52.
13.
Siegman AW, Dembroski TM, Ringel N. Components of hostility and the severity of coronary artery disease.
Psychosom Med
 
1987
;
49
:
127
–35.
14.
Grady D, Applegate W, Bush T, et al. Heart and Estrogen/progestin Replacement Study (HERS): design, methods, and baseline characteristics.
Control Clin Trials
 
1998
;
19
:
314
–35.
15.
Hulley S, Grady D, Bush T, et al. Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women.
JAMA
 
1998
;
280
:
605
–13.
16.
Cook W, Medley D. Proposed hostility and pharasaic-virtue scales for the MMPI.
J Appl Psychol
 
1954
;
38
:
414
–18.
17.
Siegman AW, Townsend ST, Blumenthal RS, et al. Dimensions of anger and CHD in men and women: self-ratings versus spouse ratings.
J Behav Med
 
1998
;
21
:
315
–36.
18.
Kawachi I, Sparrow D, Spiro A, et al. A prospective study of anger and coronary heart disease. The Normative Aging Study.
Circulation
 
1996
;
94
:
2090
–5.
19.
Cockcroft DW, Gault MN. Prediction of creatinine clearance from serum creatinine.
Nephron
 
1976
;
16
:
31
–41.
20.
Leiker M, Hailey BJ. A link between hostility and disease: poor health habits?
Behav Med
 
1988
;
14
:
129
–33.
21.
Everson SA, Kauhanen J, Kaplan GA, et al. Hostility and increased risk of mortality and acute myocardial infarction: the mediating role of behavioral risk factors.
Am J Epidemiol
 
1997
;
146
:
142
–52.
22.
Powch IG, Houston BK. Hostility, anger-in, and cardiovascular reactivity in white women.
Health Psychol
 
1996
;
15
:
200
–8.
23.
Spicer J, Chamberlain K. Cynical hostility, anger, and resting blood pressure.
J Psychosom Res
 
1996
;
40
:
359
–68.
24.
Vitaliano PP, Russo J, Bailey SL, et al. Psychosocial factors associated with cardiovascular reactivity in older adults.
Psychosom Med
 
1993
;
55
:
164
–77.
25.
Boltwood MD, Taylor CB, Burke MB, et al. Anger report predicts coronary artery vasomotor response to mental stress in atherosclerotic segments.
Am J Cardiol
 
1993
;
72
:
1361
–5.
26.
Yeung AC, Vekshtein VI, Krantz DS, et al. The effect of atherosclerosis on the vasomotor response of the coronary arteries to stress.
N Engl J Med
 
1991
;
325
:
1551
–6.
27.
Jern C, Eriksson E, Tengborn L, et al. Changes in plasma coagulation and fibrinolysis in response to mental stress.
Thromb Haemost
 
1989
;
62
:
767
–71.
28.
Moller J, Hallqvist J, Diderichsen F, et al. Do episodes of anger trigger myocardial infarction? A case-crossover analysis in the Stockholm Heart Epidemiology Program (SHEEP).
Psychosom Med
 
1999
;
61
:
842
–9.
29.
Tofler GH. Triggering and the pathophysiology of acute coronary syndrome.
Am Heart J
 
1997
;
134
:
S55
–61.
30.
Mittleman MA, Maclure M, Sherwood JB, et al. Triggering of acute myocardial infarction onset by episodes of anger. Determinants of Myocardial Infarction Onset Investigators.
Circulation
 
1995
;
92
:
1720
–5.
31.
Adams SH, John OP. A hostility scale for the California psychological inventory: MMPI, observer Q-sort, and big-five correlates.
J Pers Assess
 
1997
;
69
:
408
–24.
32.
Rutledge T, Linden W, Davies RF. Psychological risk factors may moderate pharmacological treatment effects among ischemic heart disease patients. Canadian Amlodipine/Atenolol in Silent Ischemia Study (CASIS) Investigators.
Psychosom Med
 
1999
;
61
:
834
–41.
33.
Herrington DM, Fong J, Sempos CT, et al. Comparison of the Heart and Estrogen/progestin Replacement Study (HERS) cohort with women with coronary disease from the National Health and Nutrition Examination Survey III (NHANES III).
Am Heart J
 
1998
;
136
:
115
–24.
34.
Hu FB, Stampfer MJ, Manson JE, et al. Trends in the incidence of coronary heart disease and changes in diet and lifestyle in women.
N Engl J Med
 
2000
;
343
:
530
–7.
35.
Lavie CJ, Milani RV. Effects of cardiac rehabilitation and exercise training programs on coronary patients with high levels of hostility.
Mayo Clin Proc
 
1999
;
74
:
959
–66.
36.
Gidron Y, Davidson K. Development and preliminary testing of a brief intervention for modifying CHD-predictive hostility components. J
Behav Med
 
1996
;
19
:
203
–20.
37.
Linden W, Stossel C, Maurice J. Psychosocial interventions for patients with coronary artery disease: a meta-analysis.
Arch Intern Med
 
1996
;
156
:
745
–52.