Anger is associated with cardiovascular stress reactivity; however, little is known about the effect of suppressed anger in patients with coronary artery disease (CAD). We examined whether patients with CAD who suppress their anger are at risk of adverse events. At baseline, 644 patients with CAD completed measures of anger, anger-in (reluctance to express anger), and Type D personality (tendency to experience distress and to be inhibited). The combination of high anger and anger-in scores was used to identify the presence of suppressed anger. The end points were major adverse cardiac events (a composite of death, myocardial infarction, and revascularization) and cardiac death/myocardial infarction. After an average follow-up of 6.3 years (range 5 to 10), 126 patients (20%) had experienced a major adverse cardiac event, and 59 (9%) had experienced cardiac death or myocardial infarction. Anger (p = 0.009) and suppressed anger (p = 0.011) were associated with future major adverse cardiac events, but these associations were no longer significant after adjustment for clinical characteristics. However, suppressed anger remained associated with the more rigorous end point of cardiac death or myocardial infarction (odds ratio 2.87, 95% confidence interval 1.15 to 7.15, p = 0.024) after controlling for decreased systolic function, poor exercise tolerance, extent of CAD, and revascularization. Anger alone was not independently associated with this end point. Patients with a Type D personality had a fourfold rate of suppressed anger, and an adjustment for a Type D personality attenuated the observed association between suppressed anger and adverse cardiac events. In conclusion, patients with CAD who suppress their anger were at increased risk of adverse cardiac events, and this was accounted for by individual differences in Type D personality.
Anger has been related to a poor prognosis in patients with coronary artery disease (CAD) and should be considered a potential risk factor for CAD. Anger can induce myocardial ischemia and ventricular arrhythmias and predicts adverse cardiac events. However, negative findings have also been shown for anger and cardiac prognosis, suggesting that individual differences exist in anger-related risk, possibly related to emotion regulation. Anger suppression is a form of emotion regulation that involves inhibiting the expression of angry feelings and has been related to an increased risk of CAD, increased cardiovascular reactivity, decreased heart rate variability, and cardiac mortality. Psychological risk is not uniform across all patients with CAD, and some patients might be particularly vulnerable to suppressing anger. “Negative affectivity” refers to the tendency to experience negative emotions and “social inhibition” to the tendency to inhibit self-expression. Patients with CAD who simultaneously tend to experience negative emotions and inhibit self-expression have a “distressed” or Type D personality and are at an increased risk of adverse events. Hence, Type D patients might be a vulnerable group in terms of suppressed anger, and the Type D personality might explain the associations between suppressed anger and cardiac events. However, this has never been tested. The aims of the present study were (1) to investigate the association between suppressed anger and future cardiac events in patients with CAD, and (2) to test the hypothesis that a Type D personality might account for any associations between suppressed anger and prognosis in patients with CAD.
Methods
The present study included 644 patients with CAD (581 men and 63 women, mean age 55.8 ± 7.9 years) from the University Hospital of Antwerp, Belgium. Patients were recruited to 2 studies; the design of both studies was similar and has been previously described. The first study included 303 patients with CAD, who had undergone screening for decreased left ventricular ejection fraction using ventricular angiography and 19 patients who had undergone echocardiographic screening of left ventricular ejection fraction. The second study included a new sample of 322 patients with CAD. The present study included 347 patients (54%) after myocardial infarction (MI) and 475 patients (74%) after coronary artery bypass graft surgery (CABG) or percutaneous coronary intervention (PCI). Of these 475 patients, 178 had undergone CABG/PCI after MI. Patients with other life-threatening diseases, such as cancer, were excluded. At baseline, all patients underwent an exercise stress test, completed measures of anger and personality, and provided informed consent. The local hospital ethics committee approved the study.
The Spielberger State-Trait Anger Scale is a widely used inventory that assesses the frequency and intensity of anger. This scale has also been used to examine the role of anger in the cardiac prognosis of patients with CAD. In the present study, tertiles were used to classify patients as being low, intermediate, or high on trait-anger; thus, patients with a score of ≥20 were likely to respond with feelings of anger to negative or frustrating situations.
Patients who frequently experienced anger but, at the same time, tended to inhibit the expression of these angry feelings, were considered to suppress their anger. Anger-in was defined as an inability or unwillingness to express one’s anger against people held responsible for one’s frustration, to avoid any interpersonal conflict. As previously described, we used a 4-item scale (“When you get angry, do people around you know about it?,” “If you are angry, do you generally hold it in?,” “When you are being kept waiting for an appointment, would you say anything about it?,” “Do you let people know if you are angry?”). To assess the tendency to inhibit the expression of anger, a 4-point rating scale (from 0 to 3) was used, resulting in an anger-in score of 0 to 12. This scale was internally consistent (Cronbach’s α = 0.73; mean inter-item correlation 0.40). Tertiles were used to classify patients as being low, intermediate, or high in anger-in. Thus, patients with CAD were considered to suppress their anger if they had a score of ≥20 on trait-anger (upper tertile) combined with a score of ≥7 on anger-in (upper tertile).
Patients with a Type D personality simultaneously tend to experience anger and other negative emotions across time and situations (negative affectivity) and to inhibit self-expression in social interaction (social inhibition). According to the initial publications on this personality construct, the cutoff scores on the Trait Anxiety and Heart Patients’ Psychological Questionnaire Social Inhibition scales were used to assess negative affectivity and social inhibition. In the present study, 174 patients (27%) had a Type D personality (i.e., trait-anxiety ≥43 and social inhibition ≥12). We have previously shown that Type D patients from the present sample and other CAD samples are at an increased risk of adverse events.
The clinical risk factors included severity of cardiac disorder, functional disability, and CAD as assessed using 3 different markers. Decreased systolic function was defined as a left ventricular ejection fraction of ≤50%, and decreased functional status as poor exercise tolerance, defined by the peak workload on a symptom-limited exercise stress test (ie, ≤140 and ≤120 W for younger and older men and ≤100 and ≤80 W for younger and older women, respectively). Severe CAD was defined as 3-vessel disease (≥70% reduction in internal diameter of the coronary arteries). Other clinical covariates included anterior MI at baseline, a history of MI, invasive treatment with CABG/PCI, β-blocker therapy, angiotensin-converting enzyme inhibitor therapy, smoking, hyperlipidemia (total cholesterol level >240 mg/dl or taking lipid-lowering medication), and hypertension treatment. Age and gender were entered as covariates in all multivariate models.
The first end point was major adverse cardiac events (MACE), defined as a composite of death, MI, and revascularization (CABG/PCI) after an average follow-up period of 6.3 ± 1.5 years (range 5 to 10). The second end point was a composite of cardiac death and MI as a more rigorous measure of clinical outcome during follow-up. As previously described, follow-up data were derived from the hospital records, and the patient’s attending physician was involved in the classification of the cause of death.
Multivariate logistic regression analyses were used to examine the independent associations of age, gender, disease severity, and other clinical variables with future MACE and cardiac death or MI. Cross tabulation and univariate regression analyses were used to analyze the associations between anger and suppressed anger and future MACE or cardiac death/MI. Multivariate logistic regression analyses were used to examine whether the adjustment for demographic and clinical variables would attenuate the effect of anger or suppressed anger on the occurrence adverse events. Next, cross tabulation and univariate regression analyses were used to test the hypothesis that patients with a Type D personality might be at particularly high risk of suppressing their anger. Finally, multivariate regression models were constructed to investigate whether the adjustment for Type D personality would explain the association between suppressed anger and the risk of future adverse events. All variables were entered simultaneously in the multivariate regression models of MACE and cardiac death or MI. The analyses were performed using the Statistical Package for Social Sciences for Windows, version 17.0 (SPSS, Chicago, Illinois).
Results
During follow-up, 126 patients (20%) experienced MACE and 59 (9%) a fatal or nonfatal cardiac event (29 cardiac deaths and 30 MIs). All deaths were attributable to natural causes. A left ventricular ejection fraction of ≤50% and poor exercise tolerance are markers of disease severity that were independently associated with future MACE ( Table 1 ) and cardiac death or MI ( Table 1 ). No revascularization at baseline was also independently associated with both end points. Age was associated with future MACE, and a trend was seen for 3-vessel disease to be associated with cardiac death/MI. The other clinical variables were not associated with the outcome.
| Baseline Characteristic | Odds Ratio | 95% Confidence Interval | p Value |
|---|---|---|---|
| Major adverse cardiac events (n = 126) | |||
| Demographic variables | |||
| Age | 0.95 | 0.93–0.98 | 0.001 † |
| Gender (men) | 0.99 | 0.49–1.98 | 0.97 |
| Disease severity | |||
| Left ventricular ejection fraction ≤50% | 1.84 | 1.07–3.18 | 0.029 † |
| Poor exercise tolerance ‡ | 1.86 | 1.22–2.83 | 0.004 † |
| Three-vessel disease | 1.34 | 0.84–2.13 | 0.22 |
| Anterior myocardial infarction at baseline | 1.01 | 0.61–1.66 | 0.99 |
| Previous myocardial infarction | 1.07 | 0.54–2.11 | 0.85 |
| Clinical variables | |||
| No coronary artery bypass grafting/percutaneous coronary intervention at baseline | 1.64 | 1.03–2.61 | 0.039 † |
| β-Blocker therapy | 1.35 | 0.89–2.04 | 0.15 |
| Angiotensin-converting enzyme inhibitors | 0.77 | 0.29–1.97 | 0.58 |
| Hyperlipidemia § | 1.20 | 0.77–1.86 | 0.42 |
| Hypertension ¶ | 0.87 | 0.53–1.43 | 0.58 |
| Smoking | 0.98 | 0.59–1.63 | 0.45 |
| Cardiac death/myocardial infarction (n = 59) | |||
| Demographic variables | |||
| Age | 0.98 | 0.94–1.01 | 0.22 |
| Gender (men) | 1.67 | 0.55–5.03 | 0.37 |
| Disease severity | |||
| Left ventricular ejection fraction ≤50% | 2.53 | 1.29–4.96 | 0.007 † |
| Poor exercise tolerance † | 2.80 | 1.56–5.00 | 0.001 † |
| Three-vessel disease | 1.82 | 0.96–3.42 | 0.065 |
| Anterior myocardial infarction at baseline | 1.19 | 0.61–2.30 | 0.62 |
| Previous myocardial infarction | 1.44 | 0.63–3.29 | 0.39 |
| Clinical variables | |||
| No coronary artery bypass grafting/percutaneous coronary intervention at baseline | 2.05 | 1.08–3.90 | 0.029 † |
| β-Blocker therapy | 1.14 | 0.64–2.02 | 0.66 |
| Angiotensin-converting enzyme inhibitors | 0.40 | 0.10–1.59 | 0.19 |
| Hyperlipidemia ‡ | 0.71 | 0.37–1.35 | 0.29 |
| Hypertension § | 1.23 | 0.44–1.79 | 0.74 |
| Smoking | 1.26 | 0.64–2.45 | 0.51 |
‡ For younger and older men, ≤140 and ≤120 W; for younger and older women, ≤100 and ≤80 W, respectively.
§ Total cholesterol level >240 mg/dl or taking lipid-lowering medication.
On univariate analysis, both anger and suppressed anger were associated with future MACE ( Figure 1 ) and with the more rigorous end point of cardiac death and MI ( Figure 1 ). No significant associations between anger or suppressed anger and MACE remained after adjustment for the clinical and demographic characteristics ( Table 2 ). However, suppressed anger remained associated with cardiac death or MI after controlling for decreased systolic function, poor exercise tolerance, extent of CAD, and revascularization ( Table 2 ). In contrast, anger was not independently associated with this end point.
| Baseline Characteristic | Odds Ratio | 95% Confidence Interval | p Value |
|---|---|---|---|
| Major adverse cardiac events (n = 126) | |||
| Anger variables | |||
| Anger | 1.28 | 0.80–2.05 | 0.31 |
| Suppressed anger | 1.47 | 0.71–3.05 | 0.30 |
| Covariates | |||
| Age | 0.95 | 0.93–0.98 | 0.0001 † |
| Gender (men) | 0.96 | 0.48–1.90 | 0.90 |
| Left ventricular ejection fraction ≤50% | 1.63 | 0.98–2.72 | 0.060 |
| Poor exercise tolerance ‡ | 1.86 | 1.23–2.83 | 0.004 † |
| Three-vessel disease | 1.39 | 0.89–2.19 | 0.15 |
| No coronary artery bypass grafting/percutaneous coronary intervention at baseline | 1.55 | 0.98–2.45 | 0.059 |
| Cardiac death/myocardial infarction (n = 59) | |||
| Anger variables | |||
| Anger | 0.98 | 0.49–1.95 | 0.96 |
| Suppressed anger | 2.87 | 1.15–7.15 | 0.024 † |
| Covariates | |||
| Age | 0.97 | 0.94–1.01 | 0.17 |
| Gender (men) | 1.58 | 0.52–4.75 | 0.42 |
| Left ventricular ejection fraction ≤50% | 2.35 | 1.25–4.43 | 0.008 † |
| Poor exercise tolerance ‡ | 2.76 | 1.55–4.93 | 0.001 † |
| Three-vessel disease | 1.91 | 1.02–3.55 | 0.042 † |
| No coronary artery bypass grafting/percutaneous coronary intervention at baseline | 2.08 | 1.11–3.89 | 0.022 † |
‡ For younger and older men, ≤140 and ≤120 W; for younger and older women, ≤100 and ≤80 W, respectively.
Analyses that included the Type D personality indicated that appreciable interindividual variability was present for suppressed anger. Type D patients had a fourfold rate of suppressed anger (31 [18%] of 174) compared to non–Type D patients (20 [4%] of 470; p <0.0001; Figure 2 ). We have previously reported in 2 separate studies that Type D patients from the present sample were at increased risk of adverse events. Accordingly, a Type D personality was also significantly associated with future MACE ( Figure 2 ) and cardiac death or MI ( Figure 2 ) in the present study, with a univariate odds ratio of 2.59 and 3.99, respectively (p <0.0001).
