Association Between Anger and Mental Stress–Induced Myocardial Ischemia

Pratik Pimple, MBBS, MPH; Amit Shah, MD, MSCR; Cherie Rooks, PhD; J. Douglas Bremner, MD; Jonathon Nye, PhD; Ijeoma Ibeanu, MPH; Nancy Murrah, RN; Lucy Shallenberger, MPH; Mary Kelley, PhD; Paolo Raggi, MD, FACC; Viola Vaccarino, MD, PhD

Disclosures

Am Heart J. 2015;169(1):115-121.e2.

Abstract and Introduction

Abstract

Background Mental stress–induced myocardial ischemia is associated with adverse prognosis in coronary artery disease patients. Anger is thought to be a trigger of acute coronary syndromes and is associated with increased cardiovascular risk; however, little direct evidence exists for a link between anger and myocardial ischemia.

Methods [^99mTc]-sestamibi single-photon emission tomography was performed at rest, after mental stress (a social stressor with a speech task) and after exercise/pharmacologic stress. Summed scores of perfusion abnormalities were obtained by observer-independent software. A summed-difference score, the difference between stress and rest scores, was used to quantify myocardial ischemia under both stress conditions. The Spielberger's State-Trait Anger Expression Inventory was used to assess different anger dimensions.

Results The mean age was 50 years, 50% were female, and 60% were non-white. After adjusting for demographic factors, smoking, coronary artery disease severity, depressive, and anxiety symptoms, each IQR increment in state-anger score was associated with 0.36 U–adjusted increase in ischemia as measured by the summed-difference score (95% CI 0.14–0.59); the corresponding association for trait anger was 0.95 (95% CI 0.21–1.69). Anger expression scales were not associated with ischemia. None of the anger dimensions was related to ischemia during exercise/pharmacologic stress.

Conclusion Anger, both as an emotional state and as a personality trait, is significantly associated with propensity to develop myocardial ischemia during mental stress but not during exercise/pharmacologic stress. Patients with this psychologic profile may be at increased risk for silent ischemia induced by emotional stress, and this may translate into worse prognosis.

Introduction

Psychologic stress has long been suspected to be a risk factor for coronary heart disease (CHD), but its exact role as a trigger of acute ischemia is unclear.^[1] Mental stress–induced myocardial ischemia is a transient myocardial ischemic response to a standardized mental stress challenge,^[2] which can be induced in approximately one-third to one-half of patients with CHD.^[2] Mental stress ischemia is analogous to exercise or pharmacologically induced myocardial ischemia during standard cardiac testing (here referred together as physical stress–induced myocardial ischemia), except that the stressor used is psychologic instead of physical.^[2] Mental stress ischemia has similar prognostic value to physical stress ischemia but is typically painless, occurs at lower levels of oxygen demand, and is not related to severity of coronary artery disease (CAD)^[2,3] or previous revascularization.^[2,3] Mental stress, but not physical stress–induced myocardial ischemia, correlates with myocardial ischemia measured in daily life ambulatory monitoring^[4] and is associated with worse prognosis in subjects with stable CHD, with a 2-fold increased risk of future cardiac events independent of physical stress–induced ischemia.^[5]

Anger has long been considered a potential precipitant of acute myocardial infarction (MI) and a significant risk factor for CHD. A recent meta-analysis of 44 prospective studies found that anger and hostility were significantly associated with increased CHD risk in both healthy (19% increase) and preexisting CHD populations (24% increase).^[6] Substantial research also suggests that acute anger is a potential trigger of acute coronary syndromes.^[1,7] These previous data suggest that anger could play a role in the development of acute myocardial ischemia. However, direct evidence of a link between anger and myocardial ischemia is scarce.^[8]

To clarify this issue, we examine whether anger as an acute state or as a personality trait or specific anger expression patterns is positively associated with the occurrence of mental stress but not physical stress–induced ischemia. Given that emotional distress appears to play a larger role in early-onset CHD than in older age groups,^[9] we elected to study young and middle-aged (≤60 years) men and women who survived a recent MI.

1 2 3 4 5

Next Section

Abstract and Introduction
Methods
Results
Discussion
Conclusion
References

Table I. Characteristics of the study population
Variables	Characteristics (n = 98)
Demographics
Age in years, mean (SD)	50 (6)
Age≤50 y (%)	49 (50%)
Female (%)	49 (50%)
Non-white (%)	59 (60%)
Income below poverty line (%)*	31 (32%)
Current smokers (%)	28 (29%)
Medical history and CHD risk factors
ST-elevation MI (%)	44 (45%)
Time since MI (m), mean (SD)	4.8 (1.3)
Hypertension (%)†	67 (69%)
Hyperlipidemia (%)†	71 (73%)
Diabetes (%)†	20 (21%)
BMI (kilograms per square meter), mean (SD)†	31 (6)
Obese (BMI ≥30 kg/m²) (%)†	45 (46%)
Gensini score, mean (SD)	17 (32)
> 0 (%)	65 (66%)
≥1 angina episode in past month (%)	41 (42%)
Treatment history and current medications
Percutaneous coronary intervention (%)	73 (75%)
Coronary artery bypass graft (%)	11 (11%)
Aspirin (%)†	85 (88%)
β-blockers (%)†	85 (88%)
ACE inhibitors (%)†	53 (55%)
Statins (%)†	85 (88%)
Antidepressants (%)†	13 (13%)
Psychologic factors
BDI-II score, mean (SD)	11 (9)
>13	36 (37%)
Anxiety state, mean (SD)†	38 (11)
Anxiety trait, mean (SD)†	39 (11)

Abbreviations: BMI, Body mass index; ACE, angiotensin-converting enzyme.
*Two observations missing.
†One observation missing.

Table II. Characteristics of anger subscales
Anger subscales	No of items	Scale score mean (SD)	Scale score range	Scale item mean (SD)*
Anger state	15	18.0 (7.5)	15–60	1.20 (0.50)
Anger trait	10	15.8 (4.9)	10–36	1.58 (0.49)
Anger expression
Anger out	8	13.9 (3.5)	8–27	1.73 (0.44)
Anger in	8	15.8 (4.3)	8–28	1.98 (0.54)
Anger control (out)	8	24.1 (5.3)	14–32	3.01 (0.66)
Anger control (in)	8	23.1 (5.2)	10–32	2.89 (0.65)

*Scale items ranged between 1 and 4 on a Likert scale from 1, almost never, to 4, almost always.

Table III. Association between anger subscale scores and myocardial ischemia severity, as quantified by the SDS, during mental stress
Anger subscales	Model 1		Model 2		Model 3
Anger subscales	Δ (95% CI)	P	Δ (95% CI)	P	Δ (95% CI)	P
Anger state	0.40 (0.20–0.60)	<.001	0.41 (0.20–0.62)	<.001	0.36 (0.14–0.59)	.002
Anger trait	1.02 (0.39–1.65)	.002	1.12 (0.46–1.78)	.001	0.95 (0.21–1.69)	.01
Anger expression
Anger out	0.68 (0.08–1.29)	.03	0.68 (0.06–1.31)	.03	0.52 (−0.15 to 1.18)	.13
Anger in	0.67 (−0.09 to 1.42)	.08	0.68 (−0.10 to 1.46)	.09	0.31 (−0.62 to 1.23)	.52
Anger control (out)	−0.69 (−1.51 to 0.13)	.10	−0.72 (−1.54 to 0.10)	.08	−0.52 (−1.39 to 0.35)	.24
Anger control (in)	−0.27 (−1.22 to 0.67)	.57	−0.33 (−1.27 to 0.61)	.49	0.01 (−0.99 to 1.02)	.98

Abbreviation: Δ, Unit change in SDS per increase in IQR of corresponding anger subscale.
Model 1: unadjusted model.
Model 2: adjusted for age, gender, race, smoking status, and Gensini score.
Model 3: adjusted for all the covariates in model 2 + BDI-II score and trait anxiety.

Supplementary Table I. Association between anger subscale scores and other study factors
Covariates	Anger-state, mean (SD) or r	P	Anger-trait, mean (SD) or r	P	Anger expression
Covariates	Anger-state, mean (SD) or r	P	Anger-trait, mean (SD) or r	P	Anger-out, mean (SD) or r	P	Anger-in, mean (SD) or r	P	Anger-control (out), mean (SD) or r	P	Anger-control (in), mean (SD) or r	P
Demographic and lifestyle covariates
Age	−0.13	.05	−0.26	.01	−0.18	.08	0.007	.95	0.05	.60	−0.04	.73
Gender
Female	19.5 (9.8)	.36	15.8 (5.2)	.94	14.1 (3.8)	.51	16.3 (4.5)	.33	24.1 (4.8)	.94	23.4 (5.2)	.63
Male	16.6 (3.5)	.36	15.9 (4.7)	.94	13.6 (3.2)	.51	15.4 (4.1)	.33	24.1 (5.8)	.94	22.9 (5.4)	.63
Race
White	17.9 (7.6)	.61	16.2 (3.9)	.57	13.9 (3.1)	.87	16.4 (3.8)	.29	23.6 (5.0)	.47	22.4 (5.1)	.24
Non-white	18.1 (7.5)	.61	15.6 (5.6)	.57	13.8 (3.8)	.87	15.5 (4.6)	.29	24.4 (5.5)	.47	23.6 (5.3)	.24
Income status
Above poverty	17.0 (6.1)	.15	15.5 (4.0)	.53	13.3 (3.0)	.06	15.5 (3.8)	.47	24.5 (5.1)	.37	23.2 (5.2)	.82
Below poverty	20.0 (9.8)	.15	16.3 (6.5)	.53	15.0 (4.1)	.06	16.3 (5.2)	.47	23.4 (5.7)	.37	22.9 (5.4)	.82
Smoking status
Non-smokers	17.1 (6.0)	.07	15.3 (4.3)	.18	13.4 (3.3)	.04	15.1 (3.9)	.02	24.4 (5.4)	.32	23.2 (5.7)	.67
Current smokers	20.5 (10.1)	.07	17.1 (6.2)	.18	15.1 (3.8)	.04	17.6 (4.9)	.02	23.3 (5.0)	.32	22.8 (3.9)	.67
Medical history and coronary angiographic data
Hypertension
Absent	17.1 (3.0)	.54	14.8 (3.7)	.09	12.9 (2.5)	.02	15.4 (4.5)	.47	24.5 (5.5)	.60	22.5 (4.6)	.48
Present	18.5 (8.8)	.54	16.4 (5.4)	.09	14.4 (3.8)	.02	16.1 (4.3)	.47	23.9 (5.3)	.60	23.3 (5.5)	.48
Gensini score	−0.03	.76	−0.09	.37	0.05	.62	0.04	.69	−0.07	.50	−0.07	.52
Angina frequency
None	17.6 (6.5)	.57	15.4 (4.3)	.37	13.0 (3.0)	.01	15.7 (4.3)	.66	24.3 (5.2)	.62	23.6 (5.2)	.27
At least 1/month	18.7 (8.8)	.57	16.4 (5.8)	.37	15.1 (3.8)	.01	16.1 (4.5)	.66	23.8 (5.5)	.62	22.4 (5.2)	.27
Psychosocial factors
BDI-II score	0.48	<.001	0.45	<.001	0.40	<.001	0.57	<.001	−0.28	.005	−0.29	.004
Anxiety-state	0.47	.005	0.31	.002	0.23	.02	0.43	<.001	−0.26	.01	−0.29	.003
Anxiety-trait	0.39	.03	0.35	<.001	0.23	.02	0.54	<.001	−0.30	.002	−0.37	<.001

SD: standard deviation; r: Pearson/Spearman correlation coefficients; BDI: Beck Depression Inventory.
The table shows means and standard deviation of anger subscales according to levels of categorical variables, or regression coefficient for continuous variables.

Supplementary Table II. Hemodynamic and subjective distress rating changes with mental stress
Variables	Change, post test vs. pre test	P value
Hemodynamic measures
Systolic blood pressure (mmHg), mean change (SD)	48 (23)	<.001
Diastolic blood pressure (mmHg), mean change (SD)	30 (13)	<.001
Heart rate (beats/min), mean change (SD)	28 (18)	<.001
Rate-pressure product, mean change (SD)	7850 (4519)	<.001
Subjective ratings of distress
Subjective Units of Distress Scale, mean change (SD)	15 (34)	<.001
Visual analog scale of nervousness, mean change (SD)	0.5 (1.4)	.001
Visual analog scale of anxiety, mean change (SD)	0.3 (1.6)	.08
Visual analog scale of fear, mean change (SD)	0.4 (1.1)	.002
Visual analog scale of anger, mean change (SD)	0.4 (1.1)	.002

*P value for one sample t-test with null hypothesis being mean = 0.

Supplementary Table III. Association between anger subscale scores and myocardial ischemia severity, as quantified by the SDS, during physical stress
Anger subscales	Model 1		Model 2		Model 3
Anger subscales	Δ (95% CI)	P	Δ (95% CI)	P	Δ (95% CI)	P
Anger-state	−0.06 (−0.32–0.20)	.63	−0.07 (−0.34–0.20)	.61	−0.10 (−0.39–0.19)	.51
Anger-trait	−0.32 (−1.16–0.47)	.42	−0.34 (−1.17–0.48)	.41	−0.37 (−1.29–0.55)	.43
Anger expression
Anger-out	−0.22 (−0.98–0.53)	.56	−0.28 (−1.05–0.49)	.47	−0.29 (−1.10–0.52)	.48
Anger-in	0.41 (−0.48–1.32)	.36	0.36 (−0.57–1.29)	.45	0.85 (−0.24–1.96)	.13
Anger-control (out)	0.64 (−0.33–1.61)	.20	0.61 (−0.36–1.59)	.21	0.45 (−0.57–1.47)	.39
Anger-control (in)	0.39 (−0.73–1.50)	.50	0.36 (−0.74–1.47)	.52	0.09 (−1.09–1.26)	.88

Δ: unit change in SDS per increase in IQR of corresponding anger subscale.
Model 1: unadjusted model.
Model 2: adjusted for age, gender, race, smoking status, and Gensini score.
Model 3: adjusted for all the covariates in model 2 + BDI-II score and anxiety-trait.

Supplementary Table IV. Sensitivity analysis for the association between anger subscale scores and myocardial ischemia severity (quantified by the SDS), during both mental and physical stress, after excluding 9 subjects with SPECT imaging artifacts
Anger subscales	SDS during mental stress (unadjusted)*		SDS during physical stress (unadjusted)*
Anger subscales	Δ (95% CI)	P	Δ (95% CI)	P
Anger-state	0.24 (0.12 - 0.36)	.0001	−0.02 (−0.18 - 0.15)	.86
Anger-trait	0.83 (0.23 - 1.42)	.006	−0.43 (−1.19 – 0.33)	.27
Anger expression
Anger-out	0.75 (0.05 - 1.45)	.04	−0.28 (−1.18 – 0.62)	.54
Anger-in	0.41 (−0.34 - 1.17)	.28	0.62 (−0.30 - 1.54)	.19
Anger-control (out)	−0.30 (−1.07 - 0.49)	.46	0.88 (−0.05 – 1.82)	.07
Anger-control (in)	−0.03 (−0.90 - 0.85)	.95	0.63 (−0.43 – 1.69)	.25