Mental Stress Induced-Myocardial Ischemia in Young Patients with Recent Myocardial Infarction: Sex Differences and Mechanisms

Study Sample and Design

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Study participants for the Myocardial Infarction and Mental stress 2 (MIMS2) study included early-onset MI cases and community controls without a history of CAD. The MI cases were recruited from the pool of patients who were admitted with a documented history of MI in the previous 8 months at Emory-affiliated hospitals in Atlanta, Georgia, and who were 18 to 60 years of age at the time of screening. The diagnosis of MI (Type 1) was verified by medical record review based on standard criteria of troponin level increase together with symptoms of ischemia and/or ECG changes or other evidence of myocardial necrosis; presence of obstructive CAD was not a criterion for inclusion.²⁰ Controls were recruited in the Atlanta area from a community-based study of individuals without established CAD.²¹ Inclusion criteria for controls were aged between 18 and 60, and no past history of MI, unstable or stable angina pectoris, congestive heart failure, or stroke. Controls were frequency matched for age and sex to the MI cases, with the goal of achieving approximately 50% women and a similar mean age in both samples.

Subjects were excluded if they had a severe comorbid medical or psychiatric disorder that could interfere with study results, such as cancer, renal failure, severe uncontrolled hypertension, current alcohol or substance abuse or schizophrenia, if they were pregnant or breastfeeding, or if they were currently using immunosuppressant or psychotropic medications other than antidepressants. MI patients were also excluded if they had unstable angina, acute MI or decompensated heart failure within the past week; if they weighed over 450 lbs (due to limits on the weight bearing of the nuclear stress test equipment); and if it was deemed to be unsafe by study cardiologists to hold anti-ischemic medications for 24 hours before the testing.

Both MI cases and healthy controls underwent a standardized mental stress test and vascular testing as described below and following published methodology.²² The MI patients, but not the controls, also underwent single-photon emission computed tomography (SPECT) imaging studies with mental stress and with exercise or pharmacological stress. Sociodemographic and psychosocial data were collected for all participants prior to all testing. At the end of the study protocol, medical records for the MI patients were abstracted for clinical information, including catheterization data. The Emory University Institutional Review Board approved the protocol and all participants provided written informed consent.

Mental Stress Procedure

After resting for 30 minutes in a quiet, dimly lit, temperature-controlled room, mental stress was induced in both MI patients and controls by a standardized public speaking task as previously described.^{4, 18, 22} Participants were asked to imagine a real-life stressful situation, in which a close relative was been mistreated in a nursing home, and asked to make up a realistic story around this scenario. They were given two minutes to prepare a statement and then three minutes to present it in front of a video camera and an audience wearing white coats. Subjects were told that their speech would be evaluated by the laboratory staff for content, quality and duration. We recorded blood pressure and heart rate at five-minute intervals during the resting phase and at one-minute intervals during the mental stress task, and calculated the rate-pressure product as peak systolic blood pressure times peak heart rate. We also obtained subjective ratings of distress with the Subjective Units of Distress Scale ²³ on a linear scale of 0 to 100, with 100 being the highest level of distress.

Vascular Testing

Among both MI patients and controls, we assessed peripheral vasoconstriction and microvascular function using the Endo-PAT2000 (Itamar Medical, Caesarea, Israel), which measures finger pulse volume amplitude (PVA), reflecting peripheral blood volume changes using volume plethysmography technology.^{24, 25} PVA signals detected by a finger probe were filtered, amplified, and analyzed in an operator-independent manner. The baseline PVA was determined by averaging the last 3 minutes of recording that preceded the mental stress test. The stress amplitude was determined as the lowest PVA during the speaking period. The peripheral arterial tonometry (PAT) ratio was then calculated by the software as the ratio of PVA during the speaking task over the resting baseline, with a ratio <1 signifying a vasoconstrictive response. After removing participants with excessive artifacts, the PAT ratio was available in 76% of MI cases and 72% of controls.

Before and 30 minutes after the mental stress test we also used the EndoPAT device to assess the digital reactive hyperemia index (RHI), a measurement of peripheral microvascular function. For this test, PVA is obtained in resting condition and during reactive hyperemia, which is elicited by the release of an upper arm blood pressure cuff inflated to suprasystolic pressure for 5 minutes. The RHI is then calculated as the ratio of post-deflation to baseline pulse amplitude in the hyperemic finger divided by the ratio in the contra-lateral finger.²⁶ This metric is calculated by a computer algorithm and has been associated with cardiovascular outcomes.²⁷ Before and 30 minutes after the mental stress test MI patients and controls also underwent measurement of flow-mediated vasodilation (FMD) of the brachial artery via bi-mode ultrasound to assess endothelial function using standard methodology.²⁸

Myocardial Perfusion Imaging

MI subjects underwent three SPECT myocardial perfusion imaging scans following injection of sestamibi radiolabeled with Technetium-99m ([^99mTc]), at rest, during mental stress, and during conventional stress, at the dose of 10–14 mCi of ^99mTc for rest imaging and 30–40 mCi for stress imaging, based on weight. Testing was done in two separate days up to one week apart on a dedicated SPECT camera (Philips Cardio MD) without attenuation correction. We withheld anti-ischemic medications for 24 hours prior to testing following standard nuclear cardiology protocols. On the mental stress day, [^99mTc]sestamibi was injected one minute after the onset of the public speech task. On the conventional stress day, subjects underwent a standard Bruce protocol, or (if unable to exercise) a pharmacological stress test with regadenoson (Abbott, Chicago, IL). The radioisotope injection was given at peak exertion or immediately after the regadenoson injection.

An experienced nuclear cardiologist (PR) performed visual interpretation blinded to stress test type and clinical data. Each myocardial segment was scored from 0 to 4, with 0 being normal, 1 possibly normal, 2 definitely abnormal, 3 severely abnormal and 4 no perfusion. We calculated summed scores in a conventional fashion, including a summed stress score (SSS), a summed rest score (SRS), and a summed difference score (SDS).²⁹ For exercise and pharmacological stress, presence of ischemia was defined as a SDS ≥4.³⁰ A SDS ≥3 is typically used as evidence of MSIMI.³¹ To check the robustness of the results, we also calculated percent ischemic myocardium as (SDS/68) × 100 as a semiquantitative measure of inducible ischemia for both conditions. Moreover, we computed the number of ischemic segments as an additional measure of ischemic burden. Within each segment, we defined myocardial ischemia as a new perfusion abnormality (a score of ≥2), or worsening of pre-existing abnormality by at least 2 points in a single segment, or by at least 1 point in two or more contiguous segments.

Other Measurements

A research nurse obtained detailed sociodemographic, medical history and medication information, and measured weight and height to calculate the BMI. Trained personnel abstracted medical records for the index MI including clinical and angiographic data. Obstructive CAD was defined as ≥ 70% lumen stenosis. CAD severity was quantified with the Gensini scoring method, which uses a nonlinear point system for degree of luminal narrowing that weighs lesions according to specific coronary tree locations to take into account prognostic significance.³²

We used validated instruments to assess behavioral, social, and health status information for both MI patients and controls. Depressive symptoms were assessed with the Beck Depression Inventory-II (BDI-II),³³ a reliable and valid self-report measure which has been widely used in cardiac as well as non-cardiac populations. In addition, we administered the Structured Clinical Interview for DSM IV (SCID)³⁴ to derive a lifetime diagnosis of major depression and posttraumatic stress disorder (PTSD). The SCID was administered by a trained research nurse under the supervision of the study psychiatrist (JDB), who reviewed psychiatric diagnoses. We also administered the Cohen’s Perceived Stress Scale,³⁵ a 10-item survey of general stress validated in multi-ethnic populations, and the Spielberger’s State-Trait Anxiety Inventory, a 40-item questionnaire to measure anxiety as an emotional state or a personality trait.³⁶

Statistical Analysis

We compared women and men within each group (MI patients and controls) for demographic, behavioral and clinical characteristics, as well as for hemodynamic and vascular changes during mental stress (and, among MI subjects, conventional stress), using t tests for continuous variables and chi-squared tests for categorical variables. Similar testing was done to compare MI patients with controls; furthermore, we tested for sex by group interactions using mixed regression models for continuous variables and Breslow-Day tests for categorical variables. Since the distribution of the PAT ratio was highly skewed, it was log transformed in the analysis, and results were presented as geometric means.

For the MI group, we fitted logistic models to assess sex differences in ischemia as a binary variable for both mental stress and conventional stress. Using nested models, we sequentially adjusted for sociodemographic and lifestyle characteristics, clinical risk variables and CAD severity indicators, and, lastly, psychosocial factors. In a final model, we evaluated whether the results were independent of presence of ischemia with the alternative stress modality.

To assess potential vascular mechanisms for sex differences in MSIMI, we included resting hemodynamics and vascular function measures, and their change with mental stress, to a model with MSIMI as a dependent variable. Additionally, we evaluated whether women and men differed in hemodynamic and vascular predictors of MSIMI, by constructing sex-stratified models and testing interactions by sex. For these analyses, we only considered measures obtained either at rest or during the stress procedure. Measurements of vascular function obtained 30 minutes after stress (FMD and RHI) were not considered relevant for this analysis. All these models were adjusted for body surface area and CAD severity indicators, including Gensini CAD severity score and resting perfusion defect severity (summed rest score).

Missing data were minimal or absent except for angiographic data, which were missing in 7% of the sample, and vascular function data, which were missing between 5% and 25% of the sample, depending on the variable. To avoid loss of information and possible bias due to missing covariate values, we performed multiple imputations for statistical models that included angiographic or vascular function data. Data were imputed 50 times using the Markov chain Monte Carlo method of multiple imputation, implemented using SAS PROC MI, and imputed regression estimates were combined via SAS PROC MIANYLYZE.³⁷ We also assessed the missing at random (MAR) assumption by differential weighting of variable values (SAS MNAR statement) and found results to be nearly identical, thus justifying reliance on the MAR assumption. The estimates in the imputed analysis matched very closely the not-imputed results, suggesting that bias was minimal or absent. Nonetheless, we report results based on imputed datasets for models that included angiographic or vascular function data. All analyses were conducted using SAS statistical software version 9.4 (SAS Inc., Cary, NC).

Study Sample and Baseline Characteristics

Between August 2012 and March 2016, we enrolled 313 MI cases and 112 controls in the MIMS2 study. We excluded 7 MI cases with missing nuclear imaging results, leaving 306 MI subjects (150 women and 156 men) and 112 controls (58 women and 54 men) in the analytical sample. The mean age was 50.5 years overall, and it was 50.8 years in the MI group (range, 25 to 61 years) and 49.5 in the control group (range, 22 to 61 years). All subjects were younger than 61 years, except for three whose birthday occurred after screening but before study visits. The mean and median time from the date of the index MI to the study visit was 5 months for both women and men.

Compared to men with MI, women with MI had a more adverse sociodemographic and psychosocial profile, including lower income and education and higher levels of depression, PTSD, and perceived stress, although there were no differences in anxiety (Table 1). Women with MI were also more likely to be Black/African American than men with MI. No such differences were found among controls. In both cases and controls, there were no sex differences in most cardiovascular risk factors and medications, except that women with MI were more likely to be diabetic than men with MI and women in both groups were more often taking antidepressants. When compared with their MI counterparts, both female and male controls had a better psychosocial risk profile, including higher income and education, lower rates of depression and PTSD, and lower symptoms of perceived stress and anxiety (all p <0.01). Controls also exhibited a more favorable cardiovascular risk profile than MI patients, including lower levels of BMI, smoking, dyslipidemia, diabetes, and resting blood pressure (all p<0.01). This was true for both women and men, with no significant sex-by group interactions. Women in the control group were also more likely to self-describe themselves as premenopausal than women with MI (35% vs 23%, p=0.08).

Despite similarities in CAD risk factors, women with MI had less severe obstructive CAD than men, as evidenced by a lower Gensini score, a lower frequency of three-vessel disease, and a lower summed rest score, a measure of severity of fixed perfusion defects. Women also had a higher left ventricular ejection fraction. There were no sex differences in the proportions of ST-segment elevation MI, history of congestive heart failure, and previous revascularization.

A comparison of resting hemodynamics and vascular function between women and men with MI showed that women had a higher resting systolic blood pressure and heart rate, but vascular parameters (FMD and RHI) were similar at rest (Table 1). There were no sex differences in any of the resting hemodynamic and vascular parameters among controls. As expected, both female and male controls had a more preserved vascular function than their MI counterparts.

Hemodynamic, Vascular and Subjective Responses to Stress

Overall, hemodynamic responses to mental stress were similar in MI cases and controls, except for a blunted heart rate response among patients. Women in both groups exhibited a slightly higher rate pressure product with mental stress than their male counterparts (Table 2). Furthermore, women in both groups showed a lower RHI post-mental stress and a larger post-stress RHI decline compared with men, indicating enhanced microvascular dysfunction with mental stress. In contrast, there were no sex differences in either group in FMD changes with mental stress. In both groups the PAT ratio was higher in women than men (denoting less vasoconstriction), but women with MI, and not men with MI, showed more enhanced vasoconstriction than their respective controls (p=0.036 for the interaction between group and sex). There were no sex differences in any of the hemodynamic and vascular responses to conventional stress.

Myocardial Perfusion

Overall, 50 (16%) MI subjects developed myocardial ischemia with mental stress, and 69 (23%) with conventional stress. Women showed two times more myocardial ischemia than men, with both mental stress and conventional stress (Table 3). This was true irrespective of whether ischemia was analyzed as a dichotomous variable (presence/absence of ischemia), as a quantitative score (percent of ischemic myocardium), or as a count of ischemic myocardial segments. Despite the similarity of results for ischemia with mental and conventional stress, the correlation between the two types of ischemia was modest (Spearman correlation coefficient=0.20, p=0.002), and of the 119 MI cases who developed ischemia with either condition, only 20 (17%) were positive for both mental stress and conventional stress ischemia. This proportion was similar in women (19%) and in men (12%).

Adjustment for sociodemographic and lifestyle factors, CAD severity, and psychosocial factors did not substantially affect sex differences in ischemia with either condition, and the alternative stress test result also did not explain MSIMI or CSIMI, even though they were mildly associated (Table 4).

Vascular Mechanisms for Sex Differences in MSIMI

To assess potential vascular mechanisms of sex differences in MSIMI, we examined whether resting hemodynamic and vascular measures, and their changes with mental stress, explained sex differences in MSIMI. These analyses were adjusted for body surface area and CAD severity (Gensini CAD severity score and SPECT summed rest score). After including resting RPP and resting RHI to the model, the odds ratio (OR) of MSIMI for women compared with men was 2.6 (95% CI, 1.3–5.4). Adding resting FMD to the model did not change the results (OR, 2.6; 95% CI, 1.3–5.3). Adjustment for change in RPP with mental stress (OR, 2.7; 95% CI, 1.3–5.4), and further adjustment for PAT ratio with mental stress (OR, 3.0; 95% CI, 1.5–6.4) also did not weaken the association between sex and MSIMI.

Of the hemodynamic and vascular measures evaluated in these models, only a lower PAT ratio, denoting greater vasoconstriction with mental stress, was significantly associated with MSIMI (p=0.01), while a lower resting RHI, denoting lower baseline microvascular function, was borderline (p=0.06). However, these results were driven by women, as a lower resting RHI and a lower PAT ratio were related to MSIMI in women only (Table 5). The interaction effect with sex was borderline significant for PAT ratio (p=0.058). There was no significant association between rate pressure product (either the resting value or the change with mental stress) and MSIMI, in either women or men. None of these vascular measurements were associated with ischemia with conventional stress (data not shown).