|Home | About | Journals | Submit | Contact Us | Français|
The goal of this study was to evaluate the prevalence and clinical characteristics of mental stress–induced myocardial ischemia.
Mental stress–induced myocardial ischemia is prevalent and a risk factor for poor prognosis in patients with coronary heart disease, but past studies mainly studied patients with exercise-induced myocardial ischemia.
Eligible patients with clinically stable coronary heart disease, regardless of exercise stress testing status, underwent a battery of 3 mental stress tests followed by a treadmill test. Stress-induced ischemia, assessed by echocardiography and electrocardiography, was defined as: 1) development or worsening of regional wall motion abnormality; 2) left ventricular ejection fraction reduction ≥8%; and/or 3) horizontal or downsloping ST-segment depression ≥1 mm in 2 or more leads lasting for ≥3 consecutive beats during at least 1 mental test or during the exercise test.
Mental stress–induced ischemia occurred in 43.45%, whereas exercise-induced ischemia occurred in 33.79% (p = 0.002) of the study population (N = 310). Women (odds ratio [OR]: 1.88), patients who were not married (OR: 1.99), and patients who lived alone (OR: 2.24) were more likely to have mental stress–induced ischemia (all p < 0.05). Multivariate analysis showed that compared with married men or men living with someone, unmarried men (OR: 2.57) and married women (OR: 3.18), or living alone (male OR: 2.25 and female OR: 2.72, respectively) had higher risk for mental stress-induced ischemia (all p < 0.05).
Mental stress-induced ischemia is more common than exercise-induced ischemia in patients with clinically stable coronary heart disease. Women, unmarried men, and individuals living alone are at higher risk for mental stress-induced ischemia. (Responses of Myocardial Ischemia to Escitalopram Treatment [REMIT]; NCT00574847)
Many studies have demonstrated that mental stress–induced myocardial ischemia (MSIMI) is prevalent among patients with coronary heart disease (CHD) who exhibit exercise stress–induced myocardial ischemia (ESIMI) (1–3). MSIMI is clinically important for this population, as those who exhibit MSIMI have a poorer cardiovascular (CV) prognosis (4–7). MSIMI is generally believed to be a phenomenon limited to CHD patients with comorbid ESIMI (1–3,8–10); however, it may also be present in patients who show no signs of ischemia during exercise (11). In a study conducted by Ramachandruni et al. (11), 6 (28.6%) of the 21 participants with a negative exercise stress test exhibited a mental stress–induced perfusion defect, suggesting that MSIMI may be present in a significant number of CHD patients without ESMI. Nevertheless, studies incorporating larger sample sizes are needed to provide a better estimate of the prevalence of MSIMI, relative to ESIMI, in the CHD population.
In this study, we present data collected from 310 patients with clinically stable CHD who underwent screening for MSIMI as a part of the REMIT (Responses of Myocardial Ischemia to Escitalopram Treatment) study (NCT00574847) (12). The specific focus of the present study was to examine the prevalence and demographic/clinical characteristics of MSIMI, as well as compare left ventricular responses to mental and exercise stress tests.
Both male and female adult patients, age 21 years or older, with documented CHD (by angiographic finding of coronary artery stenosis ≥70%, history of myocardial infarction, and/or revascularization procedures, such as coronary artery bypass graft surgery or percutaneous coronary intervention), were recruited for screening for the REMIT study (12). The REMIT study required that CHD patients demonstrate MSIMI during the baseline screening to be eligible for participation in the trial intervention. The protocol was reviewed and approved by the Duke University Health System Institutional Review Board. All participants voluntarily provided written informed consent.
Patients with CHD who visited the cardiology outpatient clinics of the Duke University Health System between July 2007 and September 2011 were systematically screened for eligibility for the present study. The detailed inclusion and exclusion criteria of REMIT have been previously described and are available in the Online Appendix (12). Once an eligible patient was identified, permission to participate in the study from his/her cardiologist was obtained, and the patient was approached for study consent followed by further assessment.
The REMIT study procedures have been previously published (12). In brief, participants were administered an interview designed for the collection of demographic and clinical characteristics data, a structured psychiatric assessment, and a series of psychometric tests, followed by measurement of resting vital signs (12). Participants completed the stress testing on a separate day.
The stress testing was conducted at the Duke Cardiac Diagnostic Unit between 8 AM and 11 AM. Beta-blockers were withheld for 24 to 48 h, depending on the half-life of these medications, before the stress testing. Following a 20-min rest period, participants underwent 3 mental stress tasks in sequence: 1) mental arithmetic (MS1); 2) mirror trace (MS2); and 3) anger recall public speech (MS3). A rest period of 6 min followed every stress test. Following the completion of MS3 and a resting period, patients performed an exercise treadmill test (ES) using the standard Bruce protocol (13). ES was terminated according to the guidelines of the American College of Sports Medicine. Emotions (sadness, frustration, tension, calm, and in-control) and physical symptoms (chest pain, discomfort, shortness of breath, and others) were collected at rest and after each stress test via a self-rated visual log. One significant difference of this study from previous studies was that except for beta-blockers, other cardiac medications were continued during the stress testing (1,2,9).
Echocardiography and electrocardiography were used to assess for myocardial ischemia (14). Digital acquisition of echo images was obtained during the last 3 min of the baseline resting period, for 3 min during each mental stress test, and at the peak of the ES (12). Blood pressure, heart rate, and standard 12-lead electrocardiography were recorded simultaneously during the acquisition of the echo images. Images (parasternal long- and short-axis views and apical 4- and 2-chamber views) were acquired using a 3-MHz transducer while in the harmonic imaging mode of the Philips iE33 system (Philips Ultrasound, Bothell, Washington). Left ventricular wall motion was assessed using the American Society of Echocardiography’s recommended 16-segment model and determined from 30 to 40 frames of systole from a cardiac cycle. Each segment was graded and scored as normal (normal or hyperdynamic; score = 1) or abnormal (hypokinetic, akinetic, dyskinetic, or aneurysmal; scores = 2, 3, 4, or 5, respectively) wall motion. The kappa value for the intra-/ intervariability of wall motion analysis in this study ranges from 0.80 to 0.87. Wall motion score index, the sum of wall motion scores divided by the total number of segments scored, was calculated at rest, during each mental stress test, and during ES. In addition, we calculated the wall motion score index for the basal, middle, and apical regions of the left ventricle. The purpose of calculating these scores was to document the effects of stress on specific regions of the left ventricle. Spearman coefficients of correlation for the intravariability/intervariability of the wall motion score index of the study range from 0.89 to 0.94. Left ventricular ejection fraction (LVEF) was calculated by measuring the images of the 2 apical windows (parasternal long-axis, apical 4-chamber, and apical 2-chamber) from a 3- to 5-beat loop via the biplane Simpson’s method (15).
Stress-induced myocardial ischemia was defined by 1 or more of the following: 1) development of a new or a worse wall motion abnormality (WMA); 2) reduction of LVEF ≥8%; and/or 3) horizontal or downsloping ST-segment depression ≥1 mm or ST-segment elevation ≥1 mm in 2 or more leads lasting for ≥3 consecutive beats. MSIMI was defined by the aforementioned ischemic changes during 1 or more of the 3 mental stress tasks.
Descriptive statistics and plots were used to assess the demographic and clinical characteristic differences of the following: 1) patients with and without MSIMI; 2) patients with and without ESIMI; and 3) patients with MSIMI only versus patients with ESIMI only. Chi-square or Fisher exact tests were used for categorical variables, and Student t tests for continuous variables. Repeated measures analysis of variance was used for within subjects comparisons of continuous variables (e.g., comparing LVEF changes across tasks), and the McNemar test was used to test for within-subjects differences of categorical variables (e.g., comparing rates of MSIMI and ESIMI). Spearman correlation coefficients were used to examine interrelationships among the various wall motion score index and LVEF changes during the stress tests. Logistic regression models were used to test predictors of MSIMI and ESIMI. SAS version 9.1 (SAS, Cary North Carolina) was utilized for the analysis. A p value of <0.05 was considered statistically significant.
A total of 400 clinically stable CHD patients provided consent for the baseline REMIT assessments, and 310 underwent the stress tests. Study enrollment is summarized in Figure 1. Of those 310 patients, 290 completed all mental and exercise tests.
MSIMI was more frequent than ESIMI, occurring in 43.5% of patients compared with 33.8% for ESIMI (McNemar chi-square  = 9.12, p = 0.003) (Table 1). Of the sample, 46.6% had no ischemia during any of the stress tests, 23.8% showed both MSIMI and ESIMI, 19.7% had MSIMI only, and 10.0% had ESIMI only. Of the patients with ESIMI, 70.4% also had MISMI, and 54.76% of patients with MSIMI demonstrated ESIMI. Relative to ES, mental stress induced greater rates of WMA (36.21% vs. 21.60%, McNemar chi-square  = 31.50, p = 0.0001) and LVEF reduction ≥8% (18.05% vs. 4.96%, McNemar chi-square  = 27.92, p < 0.0001) (Table 1). Similar to previous studies, mental stress testing did not induce electrocardiographic ischemic changes (Table 1) (4). Exercise capacity, indicated by the exercise duration, peak heart rate, and ability to achieve target maximal heart rate, was not statistically different between patients with MSIMI and patients without MSIMI (Table 2).
The characteristics of the overall population and differences between patients with and without MSIMI and between patients with and without ESIMI are summarized in Table 3. Univariate analysis demonstrated that women (odds ratio [OR]: 1.88, 95% confidence interval [CI]: 1.04 to 3.42, p = 0.04), patients who were not married (OR: 1.99, 95% CI: 1.19 to 3.36, p = 0.009), and patients who lived alone (OR: 2.24, 95% CI: 1.19 to 4.20, p = 0.01) were more likely to exhibit MSIMI. In multivariate analysis, however, none of these variables emerged as significantly and independently associated with MSIMI. This is likely due to the intercorrelations of the variables. For example, living arrangement and marital status were highly correlated (r = 0.69), whereas the correlations between sex and marital status and sex and living arrangement were r = 0.39 and r = 0.24, respectively (all p values <0.0001). To further explore the relationships of these variables to MSIMI, we created new variables comprised of different sex and marital status groups (i.e., female/not married, female/married, male/ not married, and male/married), and different sex and living arrangement groups (i.e. female/living alone, female/living with someone, male/living alone, and male/living with someone). Logistic regression showed a significant effect for the sex/marital status variable (p = 0.005). Compared with married men, unmarried men (OR: 2.57, 95% CI: 1.33 to 4.97, p = 0.005) and married women (OR: 3.18, 95% CI: 1.22 to 8.32, p = 0.02) were more likely to show MSIMI. Unmarried women tended to have more MSIMI (OR: 1.82, 95% CI: 0.98 to 4.31, p = 0.11). Analyses of the sex and living arrangement variable yielded a similar pattern of results (overall model p = 0.03): compared with men living with someone, men and women living alone (OR: 2.25, 95% CI: 1.02 to 4.93, p = 0.04, and OR: 2.72, 95% CI: 1.03 to 7.17, p = 0.04, respectively) were more likely to show MSIMI. Women living with someone (OR: 1.78, 95% CI: 0.86 to 3.68, p = 0.12) tended to have more MSIMI.
None of those factors, however, were associated with ESIMI. There were 2 characteristics that significantly separated patients who developed ESIMI from those without ESIMI. Compared with patients without ESIMI, patients who exhibited ESIMI were more likely to have a history of coronary artery bypass graft surgery (52.04% vs. 38.46%, p < 0.03) and had a lower body mass index (28.1 ± 4.3 kg/m2 vs. 29.5 ± 4.6 kg/m2, p = 0.02).
For the purpose of further exploring differences between MSIMI and ESIMI, we compared patients who exhibited MSIMI only and patients who exhibited ESIMI only. Compared with patients with ESIMI only, patients who had isolated MSIMI were younger (age 61.64 ± 9.31 years vs. 64.59 ± 8.63 years), more likely to be women (16.95% vs. 11.11%), not be married (32.20% vs. 18.52%), live alone (15.25% vs. 11.11%), and had lower resting LVEF (54.47 ± 11.23 vs. 59.16 ± 6.26), though none of the differences were statistically significant.
Mean basal, middle, and apical wall motion changes during each mental and exercise task are depicted in Figure 2. Within-task analysis revealed that all mental stress tests and ES induced greater WMA in the mid-region of the left ventricle compared with the basal region, and the differences with MS2, MS3, and ES were statistically significant. All stress tests also induced greater WMA in the apical region than those in the basal region, but the differences were not significant. Compared with ES, MS3 demonstrated greater WMA in the basal and mid-regions, but not the apical region.
A comparison of LVEF change scores revealed that each mental stress task induced greater LVEF reduction than the ES (all p values <0.0001) (Table 1). LVEF reduction occurred in 46.8% with MS1, 52.4% MS2, 49.3% MS3, and 24.1% with ES. Figure 3 depicts the distributions of LVEF change among MS1, MS2, MS3, and ES. Spearman correlation coefficients analysis revealed that all 4 stressor-induced LVEF changes were significantly intercorrelated, but the correlations were stronger among the mental stress tasks (0.540 to 0.585) than the correlation between exercise-induced and mental stress–induced LVEF changes (0.392 to 0.441). Baseline LVEF was weakly associated with the LVEF changes induced by MS1 (r = −0.16, p = 0.006), but not to MS2, MS3, or ES (all p > 0.20).
Stress–induced LVEF change to each mental stressor was weakly, but significantly, correlated with concomitant stress-induced WMA (MS1: r = −0.169, p = 0.005; MS2: r = −0.26, p < 0.0001; MS3: r = −0.22, p = 0.002; and ES: r = −30, p = 0.0001). None of the mental stress-induced LVEF changes correlated to exercise-induced WMA (MS1: r = −0.05, p = 0.41; MS2: r = −0.091, p = 0.15; and MS3: r = −0.071, p = 0.25), whereas exercise-induced LVEF changes were weakly, but significantly, correlated with WMA induced by all 3 mental tasks (MS1: r = −0.21, p = 0.0008; MS2: r = −0.19, p = 0.002; and MS3: r = −0.27, p < 0.0001).
Similar to previous studies, mental stress induced much smaller heart rate, systolic blood pressure, and rate-pressure product changes than those induced by exercise, but a higher proportional diastolic elevation (Table 4) (1,2). However, these CV reactivity measurements were not significantly associated with ischemia status. Nevertheless, chest pain is more common in patients with ESIMI, and MSIMI is relatively painless (Table 4).
This is the largest study to evaluate mental stress–induced WMA and LVEF reduction in patients with CHD that was not confined to patients showing signs of ischemia to exercise or pharmacological stress test before the mental stress testing. Major unique findings from this study are that MSIMI is more common than ESIMI, and the rate MSIMI occurrence was different in men and women, especially when marital and living status were considered.
Using the ischemia criteria of an LVEF reduction ≥8 and/or WMA, approximately one-half of the patients with clinically stable CHD developed MSIMI, which was notably higher than the rate of ESIMI. Previous studies have demonstrated that the prevalence of MSIMI varies depending on the type and duration of mental tasks being performed, susceptibility of individuals being tested, and methods being utilized to detect the myocardial ischemia (3). MSIMI occurred in 73.0% of the patients who had ESIMI in our study, indicating that the potency of the mental tasks and the methods of ischemic assessment in our study are comparable to previous studies (1–3). The percentage of patients who had MSIMI without having ESIMI (21.6%) in our study was somewhat less than the 28.6% reported by Ramachandruni et al. (11), the only previous existing study that tested MSIMI prevalence in CHD patients with a negative exercise stress test.
Female CHD patients were more likely to have MSIMI than men. Furthermore, the results suggest that being married and living with someone may protect men more than women from having MSIMI. Such findings suggest that these social situations may have different biological impacts on men and women. Given that these variables were not associated with ESIMI, the underlying psychosocial and biological interaction resulting in MSIMI must be mediated through the neurocardiovascular interplay. The increased risk of developing MSIMI in female patients may be partially explained by women having smaller coronary arteries and coronary microvascular disease (16). Other studies have reported sex differences in left ventricular function at rest and during physical challenges such as exercise and orthostatic stress (17–19). Findings of these studies indicate that women have a parasympathetic predominance at rest, whereas men have a dominant sympathetic regulation both at rest and during exercise (19–22). The results of the present study may also reflect differences in how men and women respond to social support as well as differences in how they respond to mental stress (23–27). Evidence indicates marriage may affect men and women differently (28–31). Although reports of biological and psychosocial differences of sex in relationship with CHD have recently increased, the understanding of the role of sex in CHD, especially its interaction with psychosocial factors is not well understood (32–34). Previous studies of MSIMI have included too few women to allow for meaningful conclusions to be drawn regarding the prevalence and significance of MSIMI in women versus men (3). Although our study included a larger sample of female patients than other studies of MSIMI, the number of women in our study was still rather small, especially for the analyses exploring sex differences within the context of living arrangement and marital status. Future studies with large samples of both sexes will be required to ascertain the reliability of these findings and to understand the underlying mechanisms.
The present study allowed for a more thorough assessment of the left ventricular responses to various mental tasks and exercise. Although mental stress, especially MS2 and MS3, induced greater increase of WMA than ES induced, mental and physical stress caused similar regional segments of the left ventricle, that is, greater abnormal wall motion in mid- and apical segments than in basal segments (Fig. 2), a commonly observed phenomena in Takotsubo, or stress-induced cardiomyopathy (35–37). In contrast, mental stress, on average, caused LVEF reduction and the exercise test, on average, resulted in LVEF elevation (Table 1, Fig. 3). Exercise-induced LVEF changes were correlated with mental stress–induced wall motion changes, but mental stress– induced LVEF changes were not correlated with the exercise stress–induced wall motion changes. The lack of correlation between mental stress–induced LVEF change and exercise-induced WMA suggests the mechanisms or conditions underlying mental- and exercise-induced ischemia are probably different. The etiology of mental stress– induced LVEF changes has been studied, as it has a significant role in predicting poor CV outcomes (4,6,38). Goldberg et al. (9) demonstrated that both mental and physical stressors induced LVEF changes that were strongly and negatively correlated with the same stressor-induced systemic vascular resistance changes (r = −0.41, −0.53, and −0.44 for the exercise bicycle, speech, and Stroop tests, respectively; all p values <0.001), but not with the stressor-induced heart rate or blood pressure changes. Changes in cardiac output, a major contributor to LVEF, induced by the stress testing, however, could not be fully explained by the stress–induced systemic vascular resistance changes (9). Other factors that contribute to LVEF changes need to be further investigated. The features of mental stress–induced and exercise stress–induced segmental abnormal wall motion changes are particularly interesting in consideration of the Takotsubo, or stress-induced cardiomyopathy (also known as apical ballooning syndrome) (35–37). The stress–induced greater increase of WMA in the mid- and apical segments than in the basal segments of the left ventricle may reflect a combination of myocardial necrosis and regionally decreased beta-adrenoceptor responsiveness with high local catecholamine concentrations, that is, “neurogenic stunned myocardium” that was suggested by findings from animal studies (39,40). However, our study was not designed to evaluate whether the experimental mental and physical stress tests produce Takotsubo cardiomyopathy–like manifestations, and further discussion of its relevance to the present study would be highly speculative.
In summary, MSIMI is a more common condition than previously recognized, occurs in patients with and without ESMI, and is independent of conventional CV risk factors and currently utilized CV protective medications. Further studies are needed to fully characterize the underlying mechanisms and the role of sex, marriage, and living arrangements in MSIMI. A greater appreciation of the high prevalence of MSIMI in stable CHD patients is needed so further research can be directed towards delineating effective management of this condition.
The REMIT study was funded by the National Heart, Lung, and Blood Institute (NHLBI, R01HL085704), Bethesda, Maryland. Dr. Samad is a subinvestigator for the REALISM (Everest II Real World ExpAnded MuLtIcenter Study of the MitraClip System) funded by Abbott Vascular. Dr. Becker has received research support from Bayer and acted as a scientific advisor for Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Johnson & Johnson, and Regado Biosciences. Dr. Williams holds a U.S. patent on 5HTTLPR L allele as risk marker for CVD in persons exposed to chronic stress; In addition, he is a founder and major stockholder in Williams LifeSkills, Inc., a company that develops, tests and markets behavioral products for stress and anger management. Dr. Ortel has received grant support from Daiichi Sankyo, Eisai, GlaxoSmithKline, Instrumentation Laboratory, and Pfizer; and has done consulting work for Bayer, Boehringer Ingelheim, and Instrumentation Laboratory. Dr. O’Connor is a co-owner of Biscardia, a stockholder in Neurotronik/Interventional Autonomics Corporation (Stockholder), and has received financial support from Actelion Pharmaceuticals, Amgen, Astellas Pharma, BG Medicine, Critical Diagnostics, GE Healthcare, Gilead Sciences, HeartWare, Ikaria, Johnson & Johnson, Novartis, Otsuka Pharmaceutical Company, Pfizer, Pozen, ResMed, and Roche Diagnostics. Dr. Velazquez is a consultant for Novartis, and has received research grants from Abbott Vascular and Ikaria Pharmaceuticals. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Peter H. Stone, MD, acted as Guest Editor for this paper.
The authors would like to thank Dr. Kirkwood Adams, Professor of Medicine at Medical School of UNC, Dr. Alan Miller, Professor of Medicine at UF College of Medicine– Jacksonville, and Ms. Zhen Huang, statistician at Duke Clinical Research Institute, for their contributions as the Data Safety Monitoring Board of the REMIT study. The authors also thank Kevin Prybol, MPH, for his contribution in database management and kappa calculations.
For supplementary REMIT inclusion/exclusion criteria, please see the online version of this paper.