Self-Rated Stress Is Noncontributory to Coronary Artery Disease in Higher Socioeconomic Strata

Abstract

Stress and its attendant psychosocial and lifestyle variables have been associated with coronary artery disease (CAD), yet the contribution of socioeconomic status (SES) has not been addressed. The aim of this study is to determine if stress assessment is associated with CAD independent of SES, and is incremental to the Framingham Score. The study group consisted of 325 executive patients undergoing comprehensive health assessment. Stress was assessed utilizing the validated “Self-Rated Stress” (SRS) instrument. Coronary artery calcification (CAC) served to assess the degree of atherosclerosis, a CAD equivalent and risk assessment tool. The relationship between SRS and CAC was assessed, with adjustment by potential confounders. CAC was modeled by a variety of cut points (>0, ≥5, ≥100, ≥200) for the test of trend across stress levels per Mantel-Haenszel chi-square (1 df) with nonsignificant P values of 0.9960, 0.5242, 0.1692, 0.3233, respectively. A logistic regression model with SRS as a categorically ranked and continuous variable to predict binary outcome of calcification yielded P values of 0.2366 and 0.9644; this relationship, further adjusted by age, fruit and vegetable consumption, exercise, and education, yielded no statistically significant association. No improvement of fit was observed for the established Framingham Score to CAC relation utilizing SRS. The study concluded that SRS did not play a role in early CAD when focusing on a population in higher socioeconomic strata, and SRS did not add predictive value beyond patient age or calculated Framingham risk. Future studies should focus on additional validated instruments of stress to differentiate between subtypes of stress for varying SES strata. (Population Health Management 2013;16:332–340)

Introduction

S tress contributes to major adverse cardiovascular events, particularly acute myocardial infarction and sudden death,^1,2 often related to asymptomatic preclinical measures of the disease,^{3

–12} if not also to its established risk factors.^13

–17 What is not entirely certain is whether stress contributes to the atherosclerotic burden of the disease with plaque formation and ultimately obstructive coronary artery disease (CAD), or instead incites an often terminal event by an altogether different mechanism. It would be important to know whether stress accelerates the formation of coronary plaques beyond known risk factors, thus necessitating inclusion into the Framingham Score (FS) and other noninvasive risk assessment scores. Public health initiatives could have an overall larger impact as those risk factors operating early represent a more upstream approach.

There is good reason to look beyond the conventional established risk factors. Framingham risk factors have not been able to predict cardiovascular disease outcomes in individuals younger than 40 years of age,¹⁸ and it is well known that the process of atherosclerosis starts much earlier, generally in the late teens to early 20s.^19

–23 The traditional Framingham risk factors also fail to account for the 33%–50% of risk associated with the varying strata of socioeconomic status (SES).^24
–26

Stress has a demonstrated impact on human physiology lending biological plausibility to its hypothesized association with cardiovascular disease. In particular, it has been associated with hypertension and cardiovascular reactivity, endothelial dysfunction, increased levels of inflammatory markers, enhanced immune response, platelet reactivity, levels of coagulation factors, hemostasis, fibrinogen, lipid levels, glucose metabolism, and heart rate variability, all processes intimately linked to the cardiovascular system and the disease process. The process is thought to be mediated by the hypothalamic-pituitary adrenal axis and the catecholamine response to stress.^27,28

A review highlighted evidence to support independent causal relationships between chronic stress, depression, SES, social support, and development of CAD.²⁹ However, this has not been uniformly established in all cases.³⁰

Two major difficulties arise in establishing the association between stress and CAD. The first is that stress is a highly subjective and individually perceived response, and well-developed coping strategies might mitigate this risk. The second major difficulty is that it also is associated with myriad psychosocial and lifestyle factors, including SES, known to be linked to many adverse behavioral risk factors. For example, CAD is more prevalent in people from low SES backgrounds, and low SES is associated with an increased exposure to psychological stress.³¹ However, low SES also could be a surrogate marker for inadequate health care access, poor diet, sedentary lifestyle, and inadequate education, any one of which is perhaps a more plausible reason for adverse cardiovascular outcomes in this population than stress. Given that socioeconomic difference tendencies were found to exist in coronary artery calcification (CAC) measures in both men and women in their 30s, a time in life when traditional risk factors are generally nonexistent or small,³² it would be important to elucidate the actual role of stress in the process without the confounding element of low SES and its associated adverse behavioral risk factors.

The authors propose that stress contributes to the atherosclerotic burden via elevated CAC as measured by spiral computed tomography, and further propose that measures of stress add additional predictive information beyond that provided by the FS, which does not take into consideration SES and stress factors in its calculation. These associations will be examined in a cross-sectional study of executive patients from Mayo Clinic. In this population, perceived stress has been assessed by a simple validated instrument along with a composite FS. A variety of psychosocial and lifestyle predictors will be added to the models to determine if these associations are modified or attenuated. To isolate the effects of stress on an intermediate and early marker of CAD as determined by CAC, this study focuses on a study population of higher socioeconomic strata and a variety of psychosocial and lifestyle factors in an attempt to control for the confounding elements of low SES and health care access.

Methods

This study utilizes data from the year 2002 executive patient population from Mayo Clinic in Rochester, Minnesota. All patients seen in the Preventive Executive Health Program during the period from January 1, 2002 to December 31, 2002 were selected. These are patients who electively, or as part of their executive benefit package through their respective company employers, undergo a structured comprehensive preventive health assessment. These executives are largely from the Midwest, centered in Minnesota, but include patients from the entire United States and a small number of executives from abroad.

This executive study population numbered 1730 patients, although 58 refused authorization to release their private medical information, leaving 1672 eligible in the patient population. Of the 1672, a subpopulation of 329 individuals underwent coronary artery calcium testing, although 4 of these were excluded secondary to procedure limitations and invalid assessments, leaving 325 individuals remaining for the subgroup analysis. Twelve of these 325 were missing data for the self-rated stress score (SRSS), leaving 313 individuals for this study's consideration and the basis for the results presented in this paper.

For this cross-sectional study, multiple database sources were utilized for the data collection process to avoid misclassification. Data from the electronic medical record (EMR) were abstracted utilizing International Classification of Diseases coding information. A further enhancement for accuracy was accomplished by a manual review of each EMR. Behavioral risk factors were collected by a standardized survey of “Patient Provided Information” (PPI), a form filled out by all patients prior to their medical episode of care.

Information abstracted from the comprehensive EMR included sex, age, body mass index (weight in kilograms divided by height in meters squared), blood pressure (mmHg systolic and diastolic), diabetes as defined by American Diabetes Association 2003 criteria, and laboratory data including portions of a lipid profile (low-density lipoprotein [LDL] mg/dL, high-density lipoprotein [HDL] mg/dL). Information from the PPI included smoking status, highest educational level attained, physical activity level, consumption of fruits/vegetables, and perceived stress.

The level of stress was assessed using the SRSS.³³ This instrument is a single-item rating of present stress level. The question asks: “How would you rate your stress level?” and uses a Likert scale from 1 (low stress) to 5 (high stress). This measure has shown good validity when compared to the Global Severity Index (GSI) of the Symptom Checklist-90-Revised (SCL-90-R; r=0.40, P<0.001) in a large sample of outpatients (n=266).³³

CAC scanning was performed according to clinical protocol. The outcome of CAC was utilized as a surrogate marker for atherosclerotic burden; these scores were determined by electron beam computed tomography. This procedure is noninvasive, and detects and quantifies epicardial CAC using serial and contiguous ECG-triggered, 100 millisecond, 3 mm thin sections from the aorta through the heart apex. An Agatston score was developed to quantify CAC.³⁴ Direct relationships have been established between these scores and histologic,^35,36 ultrasonic,³⁷ and angiographic^38,39 measures of CAD on a vessel-by-vessel, even segmental basis. Although calcification is more frequently observed in advanced lesions, it also can occur in small amounts in earlier stages of atherosclerosis, particularly in high-risk patients and among patients as young as 11 to 24 years of age with familial hypercholesterolemia,⁴⁰ making this the assessment of choice to capture asymptomatic preclinical disease.

A composite calculation was performed utilizing the FS in order to obtain a 10-year risk assessment. This is a validated tool used in clinical practice to stratify individual risk, reported as the percent likelihood for a coronary event to occur within a 10-year period of time.⁴¹ Parameters included in this FS calculation are sex, age, diagnosis of diabetes, LDL cholesterol, HDL cholesterol, smoking status, and systolic/diastolic blood pressure measurements. Other known risk factors (eg, remote smoking history, secondary smoking exposure, family history of CAD), and the novel risk factors of lipoprotein (a), homocysteine, and C-reactive protein levels are not included in this calculation.

Potential psychosocial exposure variables from the PPI were abstracted for use in the FS (current smoking status) as well as in the assessment for potential confounding factors. These exposure variables were represented as ordinal scores and included highest educational level attained (number of years of education: range 4 through 17+ years), physical activity level (number of minutes of exercise per week: 0, 15, 45, 90, 150, 210, 270, 330+), and consumption of fruits/vegetables (number of servings per day: 0.5, 2, 3, 4, 5).

SAS statistical software, version 9.3 (SAS Institute Inc., Cary, NC) was used in all analyses. Multiple models were considered and focused on self-rated stress (SRS) as a potential predictor of CAC measures. The authors also assessed models that contained both SRSS and FS as covariates to be able to determine if the SRS association was confounded by the FS, or alternatively, if the SRSS provided additional predictive information beyond that already contained within the FS by assessing improvement of fit. All models were adjusted for age when utilizing the SRSS unless the FS had been included in the model, as age had already been factored into the score. The SRSS had been modeled as a continuous variable and as a class variable given its ordinal scoring system.

The authors modeled the outcome variable of the CAC score as a continuous variable and as a dichotomized variable, using a series of different cut points for high and low CAC score.

Results

The study population consisted predominately of highly-educated, middle-aged men of overweight body habitus with excellent access to health care. Baseline characteristics are presented in Table 1.

Table 1.

Baseline Characteristics: Individuals Undergoing EBCT-CAC Testing (n=329)

	Count	Mean or %	Std Dev	Range
Demographics
Age, years	325	51.0	(7.1)	31–70
BMI	324	28.4	(3.9)	19.5–41.1
Sex, male	325	87%
Height (cm)	325	176.5	(8.2)	146.3–195.6
Weight (kg)	324	88.6	(14.8)	49.8–142.7
Coronary artery calcium, by EBCT (4 procedure limitations)
Score=0	325	37%
Score 1–100	325	39%
Score >100	325	23%
Score average	325	108.1	(294.5)	0–3219
Score average, where CAC score >0	202	173.9	(358.2)	1–3219
Psychosocial Factors
Dietary intake, #fruits/veg servings per day (7 missing)
≥3	318	18%
≥1 and ≥2	318	52%
None	318	28%
Average	318	1.4	(1.2)	1–5
Average, where CAC score >0	197	1.4	(1.2)	1–5
Education, #years (4 missing)
<12	321	2%
≥12 and <17	321	47%
≥17	321	51%
Average	321	15.9	(1.9)	4–17+
Exercise, minutes per week
None	325	20%
≥1 hour	325	37%
>1 hour and <2.5 hours	325	26%
≥2.5 hours to 5.5 hours	325	18%
Stress, perceived (ordinal 1 low to 5 high) (12 missing)
1 to 2	313	15%
3	313	32%
4	313	41%
5	313	12%
Average	313	3.5	(1.0)	1–5
Average, where CAC score >0	195	3.5	(1.0)	1–5
Risk factors for coronary artery disease
Blood pressure parameters
Diastolic blood pressure, mmHg	325	81.4	(9.3)	59–120
Systolic blood pressure, mmHg	325	129.0	(15.9)	98.0–186.0
Diagnoses
Diabetes mellitus	325	13%
Hyperlipidemia, on Rx	325	27%
Hypertension, on Rx	325	19%
Impaired fasting glucose, >100 mg/dL	325	20%
Family history of coronary artery disease (58 missing)
Very premature, <50 in first-degree relative	267	13%
Premature, <55 Male/<65 Female in first-degree relative	325	25%
Any first-degree relative	325	63%
Framingham score	325	13.1	(3.3)	0–22
Laboratory values
HDL mg/dL	320	51.5	(15.0)	19–122
HgA1C %	49	6.1	(1.6)	4.0–12.6
LDL mg/dL	317	119.6	(31.5)	48.0–264.0
Total Cholesterol, mg/dL	320	196.9	(37.0)	106–377
Total Cholesterol/HDL ratio	320	4.1	(1.3)	2.0–9.2
Triglycerides mg/dL	320	131.0	(80)	37–689
Tobacco use
Current cigarette smoker	315	6%
Current cigar smoker	310	13%
Current pipe smoker	310	3%
Current chew tobacco user	310	3%
Current heavy use any tobacco(>2pk/day or >5 years)	310	6%
Current pack-years	310	0.8	(4.7)	0–46
Remote use any tobacco	302	32%
Remote heavy use any tobacco(>2pk/day or >5 years)	315	25%
Remote pack-years	302	4.3	(9.6)	0–57.5

BMI, body mass index; CAC, coronary artery calcification; EBCT, electron-beam computed tomography; HDL, high-density lipoprotein; LDL, low- density lipoprotein; Rx, prescription.

Mantel-Haenszel chi-square test with 1 degree of freedom (df) was used to assess the crude model of stress utilizing 4 cutoff points for the CAC. No statistical significance was observed for this linear test for trend (Table 2).

Table 2.

Linear Test for Trend Across Levels of Stress, Crude Model: Mantel-Haenszel Chi-Square Test (1df) Tabular Method (N=313)

N, %	Stress Level 1 12, 4	Stress Level 2 35, 11	Stress Level 3 100, 32	Stress Level 4 127, 41	Stress Level 5 39, 13	P value
CAC >0195, 62	7, 58	27, 77	56, 56	78, 61	27, 69	0.9960
CAC ≥5164, 52	7, 58	22, 63	51, 51	61, 48	23, 59	0.5242
CAC ≥10073, 23	4, 33	12, 34	23, 23	24, 19	10, 26	0.1692
CAC ≥20045, 14	1, 8	10, 29	13, 13	15, 12	6, 15	0.3233

CAC, coronary artery calcification.

Logistic regression was performed modeling the SRSS with these same outcome cutoff points while adjusting for age and assessing for confounding by the potential contributory or explanatory psychosocial factors of fruits and vegetable consumption, exercise level, years of education, and FS (Table 3). No statistically significant relationship was observed for the SRSS to CAC relation, nor was this primary association consistently modified or attenuated by adding any of the potential confounders. Age and FS, when added to the model, were independent significant predictor variables of the CAC outcome with the exception of some inconsistency with the CAC outcome score of ≥500 where a positive outcome had been observed in only 5.8% of cases. This latter cutoff point had been excluded in all reports. Also, no confounding was suggested by any of the aforementioned psychosocial variables in the age to CAC score relation or FS to CAC score relation when adjusted for the SRSS.

Table 3.

Logistic Regression Model to Predict CAC>0

		95% CI					95% CI
Parameter	Odds ratio	Lower	Upper	P value	Parameter	Odds ratio	Lower	Upper	P value
Model 1
stress2	2.321	0.576	9.363	0.2366	Stress (as Continuous)	1.005	0.795	1.271	0.9644
stress3	0.893	0.265	3.007	0.8548
stress4	1.108	0.333	3.687	0.8674
stress5	1.607	0.423	6.101	0.4857
Model 2
stress2	2.761	0.634	12.019	0.1760	Stress (as Continuous)	1.165	0.902	1.506	0.2426
stress3	1.364	0.378	4.925	0.6353
stress4	1.921	0.534	6.911	0.3177
stress5	2.795	0.678	11.527	0.1551
Age	1.100	1.059	1.143	<.0001	Age	1.102	1.061	1.145	<0.0001
Model 3
stress2	2.328	0.563	9.626	0.2432	Stress (as Continuous)	0.976	0.766	1.243	0.8446
stress3	0.821	0.241	2.798	0.7531
stress4	1.027	0.305	3.455	0.9656
stress5	1.363	0.350	5.303	0.6554
Fruit/veg	0.954	0.806	1.130	0.5847	Fruit/veg	0.947	0.802	1.117	0.5172
Exercise	1.002	1.000	1.005	0.1041	Exercise	1.002	1.000	1.005	0.0632
EdYr	1.083	0.956	1.227	0.2119	EdYr	1.055	0.936	1.189	0.3804
Model 4
stress2	2.535	0.595	10.807	0.2087	Stress (as Continuous)	1.107	0.866	1.416	0.4167
stress3	1.021	0.286	3.639	0.9746
stress4	1.457	0.412	5.154	0.5588
stress5	2.294	0.567	9.283	0.2445
FS	1.196	1.098	1.301	<.0001	FS	1.194	1.097	1.299	<0.0001
Model 5
stress2	2.496	0.561	11.111	0.2299	Stress (as Continuous)	1.066	0.827	1.373	0.6222
stress3	0.906	0.246	3.339	0.8822
stress4	1.322	0.362	4.828	0.6732
stress5	1.822	0.432	7.679	0.4138
Fruit/veg	0.905	0.759	1.078	0.2638	Fruit/veg	0.901	0.758	1.071	0.2381
Exercise	1.002	1.000	1.005	0.0964	Exercise	1.002	1.000	1.005	0.0638
EdYr	1.084	0.951	1.236	0.2249	EdYr	1.055	0.930	1.196	0.4056
FS	1.202	1.103	1.309	<.0001	FS	1.198	1.101	1.305	<.0001

CAC, coronary artery calcification; EdYr, years of education; FS, Framingham Score.

Improvement of fit was undertaken for the previously established FS to CAC relation (Table 4). Adding the SRSS or any of the psychosocial factors did not demonstrate a consistent incremental improvement of fit for the entire CAC cutoff points. Improvement of fit also was undertaken for the SRSS, modeled as a continuous and ordinal variable. Although the SRSS was not statistically significant as a predictor of CAC, a variety of potential explanatory variables were found to significantly add to its predictive ability, although there was inconsistency in this finding for the varying cutoff points of the CAC score.

Table 4.

Improvement of Fit Using Logistic Model Analyses

	FS
Improvement of Fit Models	Coefficient	Std Error	P value	-2 Log L Intercept+Covariates	LR Chi Square	df	Significance P<0.05
Part1: Start with FS
CAC >0
FS	0.1714	0.0425	<0.0001	393.123	17.832	1	<0.0001
FS+Stress II, III, IV, V (category)	0.1786	0.0433	<0.0001	386.307	6.816	4	0.1459
FS+Stress (continuous)	0.1772	0.0431	<0.0001	392.463	0.660	1	0.4166
FS+Fruit/vegetable	0.1767	0.0428	<0.0001	392.163	0.960	1	0.3272
FS+Exercise	0.1706	0.0425	<0.0001	389.628	3.495	1	0.0616
FS+Yrs of education	0.1715	0.0425	<0.0001	392.155	0.968	1	0.3252
FS+Stress (continuous)+Exercise	0.1758	0.0431	<0.0001	389.071	4.052	2	0.1319
CAC ≥5
FS	0.5290	0.0404	0.0002	412.397	15.421	1	<0.0001
FS+Stress II, III, IV, V (category)	0.1544	0.0411	0.0002	409.511	2.886	4	0.5771
FS+Stress (continuous)	0.1541	0.041	0.0002	412.366	0.031	1	0.8602
FS+Fruit/vegetable	0.1580	0.0407	0.0001	411.470	0.927	1	0.3356
FS+Exercise	0.1523	0.0405	0.0002	405.435	6.962	1	0.0083
FS+Yrs of education	0.1531	0.0404	0.0002	412.059	0.338	1	0.5610
FS+Stress (continuous)+Exercise	0.1541	0.0411	0.0002	405.423	6.974	2	0.0306
CAC ≥100
FS	0.1473	0.0456	0.0012	324.599	10.906	1	0.0010
FS+Stress II, III, IV, V (category)	0.1433	0.0466	0.0021	322.244	2.355	4	0.6708
FS+Stress (continuous)	0.1421	0.0463	0.0022	324.250	0.349	1	0.5547
FS+Fruit/vegetable	0.1620	0.0467	0.0005	321.230	3.369	1	0.0664
FS+Exercise	0.1494	0.0459	0.0011	320.277	4.322	1	0.0376
FS+Yrs of education	0.1497	0.0460	0.0011	323.298	1.301	1	0.2540
FS+Stress (continuous)+Exercise	0.1437	0.0469	0.0022	319.934	4.665	2	0.0971
CAC ≥200
FS	0.1198	0.0531	0.0241	247.817	5.123	1	0.0236
FS+Stress II, III, IV, V (category)	0.1204	0.0555	0.0299	243.629	4.188	4	0.3812
FS+Stress (continuous)	0.1161	0.0542	0.0321	247.703	0.114	1	0.7356
FS+Fruit/vegetable	0.1292	0.0545	0.0177	246.879	0.938	1	0.3328
FS+Exercise	0.1240	0.0544	0.0227	242.076	5.741	1	0.0166
FS+Yrs of education	0.1227	0.0538	0.0225	246.267	1.550	1	0.2131
FS+Stress (continuous)+exercise	0.1195	0.0559	0.0324	241.965	5.852	2	0.0536

	SRSS
	Coefficient	Std Error	P value	-2 Log L Intercept+Covariates	LR Chi Square	df	Significance P<0.05
Part 2: Start with Stress and Age
CAC >0
SRSS	0.0054	0.1198	0.9644	410.953	0.002	1	0.9644
SRSS+Age	0.1529	0.1308	0.2426	382.475	28.478	1	<0.0001
SRSS+FS	0.1020	0.1256	0.4167	392.463	18.490	1	<0.0001
SRSS+Age+Exercise	0.1406	0.1314	0.2846	381.061	29.892	2	<0.0001
SRSS+FS+Exercise	0.0944	0.1265	0.4555	389.071	21.882	2	<0.0001
CAC ≥5
SRSS	-0.0618	0.1168	0.5966	427.536	0.282	1	0.5957
SRSS+Age	0.1037	0.1288	0.4208	389.153	38.383	1	<0.0001
SRSS+FS	0.0214	0.1215	0.8604	412.366	15.170	1	0.0001
SRSS+Age+Exercise	0.0872	0.1297	0.5013	385.391	42.145	2	<0.0001
SRSS+FS+Exercise	0.0136	0.123	0.9123	405.423	22.113	2	<0.0001
CAC ≥100
SRSS	−0.1641	0.1359	0.2274	334.057	1.448	1	0.2289
SRSS+Age	0.0498	0.473	0.7353	294.216	39.841	1	<0.0001
SRSS+FS	−0.0833	0.1409	0.5542	324.250	9.807	1	0.0017
SRSS+Age+Exercise	0.0529	0.1483	0.7213	291.805	42.252	2	<0.0001
SRSS+FS+Exercise	−0.0834	0.1422	0.5573	319.934	14.123	2	0.0009
CAC ≥200
SRSS	−0.1660	0.1636	0.4393	252.348	0.592	1	0.4418
SRSS+Age	0.1248	0.1742	0.4737	219.117	33.231	1	<0.0001
SRSS+FS	−0.0570	0.1688	0.7358	247.703	4.645	1	0.0311
SRSS+Age+Exercise	0.1473	0.1774	0.4063	214.794	37.554	2	<0.0001
SRSS+FS+Exercise	−0.0571	0.1708	0.7381	241.965	10.383	2	0.0056

CAC, coronary artery calcification; FS, Framingham Score; LR, logistic regression; SRSS, self-rated stress score.

Discussion

Self-rated stress was not found to be a significant predictor of preclinical CAD (or atherosclerotic burden) after adjusting for age and FS in a study population of higher SES. No statistically significant relation was observed in these analyses when a variety of psychosocial and lifestyle variables were assessed as possible confounders to ensure that these had not, instead, masked an underlying relationship. These findings raise the question that SRS may not be linearly applicable across population strata.

No predictive enhancement of the FS was accomplished by adding the SRSS or any of the available psychosocial variables to the model. In the future, more emphasis should be placed on the use of the FS alone rather than considering stress as a contributory component to the clinical risk assessment for atherosclerotic disease plaque burden.

Many models were considered primarily to ensure that the atherosclerotic burden of disease was captured, given that the distribution of CAC was less than ideal. The observation that no trend was observed in all of these analyses lends credence to the authors' conclusion.

This study suffered some limitations. Utilizing prevalent point-in-time data often fails to establish a direct cause-and-effect relationship between predictor variables and the outcome of interest, given that no temporality component exists. With respect to CAC, although this may represent a particular point in time, given its inherent increase in score over time and its previously demonstrated strong correlation with atherosclerotic burden, it remains a valid measure. It could be argued that preclinical measures of CAD per these scores ultimately avoid disease misclassification, whereas attributing CAD outcomes only to terminal events may miss those individuals who succumb to another demise or disease process. However, disease misclassification still may be problematic, as it remains to be determined whether atherosclerotic plaque rupture, the critical event in the disease process, occurs in the calcified or noncalcified portions of the atheroma, as perhaps calcium, although correlating with atheromatous burden, confers local protection rather than instability. Similarly, there may be other benign processes that concurrently cause CAC that would confound the diagnosis and cause misclassification. Confounding by indication likely existed as one third of the study population had been maintained on an antihyperlipidemic agent and most would have received, at a minimum, counseling about therapeutic lifestyle choices known to treat an adverse lipid profile. Mixing of effects likely has occurred between the disease process and the treatment itself, which may impact the outcome of the study and obscure the risk potentially attributable to stress.

The study population focused only on those with a higher SES who chose to participate in an Executive Health program; the majority comprised individuals who have these expenses paid for by a corporate benefit plan, although those individuals personally willing and able to pay for preventive services and enhanced access also are included. A full accounting of SES has not been undertaken because data such as personal income are not collected.

Although the stress assessment per the SRSS has shown good validity, especially at higher levels of SRS when compared to SCL-90-R GSI scores,³³ the use of a single-item stress measure poses several challenges. The reliability of the SRSS has not been adequately tested via a comprehensive psychometric assessment. However, the SRSS continues to be widely used as a screen at Mayo Clinic and elsewhere because of its simplicity and brevity as part of a comprehensive evaluation. The SRSS is a single-item measure and therefore limited in variability or complexity of responses when compared to multi-item, self-reported stress measures (eg, the Weekly Stress Inventory,⁴² the Reeder Stress Inventory,⁴³ the Perceived Stress Scale⁴⁴); these latter self-report assessments have demonstrated good psychometric characteristics with various populations. The SRSS also is not temporally anchored (eg, over the past week, month), thus current situations and events may be overemphasized when the patient responds to SRS.

Similarly, adaptive coping mechanisms and resources may exist that might mitigate perceived stress and would be difficult to capture as data. This study population likely would have had access to these resources and also to adaptive coping mechanisms as one might imagine. SRS also encompasses a wide variety of stress subtypes, some which could be implicated instead. Perhaps a high SRS level is viewed as energizing in this population (eustress), but conversely associated with a sense of defeat (distress) in others. As an example, high job strain and demand with low decision latitude⁵ are stress subparameters linked to low SES, yet not represented in this particular study population and precluded by study design. As such, future studies should perhaps focus on other study populations to elucidate the role of stress in preclinical CAD, utilizing this same SRS instrument or others. One health risk assessment survey accomplished under the Health Enhancement Research Organization had utilized a binary measure of stress where high risk had been determined if they were “almost always” troubled by stress and, further, did not handle stress well.^45,46 It also may be beneficial to investigate alternative validated tools of stress that might better capture duration and intensity in these populations, as well as stress subtype.

The SRSS, although not found to be associated with atherosclerotic burden, could instead be involved by inciting a terminal event by another mechanism. This possibility should be given due consideration, as it may be well to target these individuals for therapeutic intervention despite low CAC or even low FS. SRS still may be associated with risk factors for the disease, such as hypertension, although this was not overtly discerned in these analyses.

Although generalizability of these results would be strictly limited to those individuals who would come to the Mayo Clinic Preventive Executive Health Program for a comprehensive general medical evaluation, the intent of this study was to elucidate the etiologic role of SRS on the development of preclinical CAD in order to identify or refute SRS as a potential novel risk factor that could be acted upon in early disease identification with appropriately targeted primary and secondary prevention interventions. SRS was not suggestive of playing a role in preclinical measures of CAD in this study population utilizing this particular instrument for stress detection, nor did it add any contributory predictive capability beyond patient age or calculated FS.

Footnotes

Author Disclosure Statement

Drs. Kermott, Hagen, and Behrenbeck, and Mr. Cha declared no conflicts of interest with respect to the research, authorship, and/or publication of this article. The authors received no financial support for research, authorship, and/or publication of this article.

References

Kopp

, Falger

, Appels

, Szedmak

. Depressive symptomatology and vital exhaustion are differentially related to behavioral risk factors for coronary artery disease. Psychosom Med, 1998; 60:752–758.

Mittleman

, Maclure

, Sherwood

et al. Triggering of acute myocardial infarction onset by episodes of anger. Determinants of Myocardial Infarction Onset Study Investigators. Circulation, 1995; 92:1720–1725.

Koertge

, Al-Khalili

, Ahnve

, Janszky

, Svane

, Schenck-Gustafsson

. Cortisol and vital exhaustion in relation to significant coronary artery stenosis in middle-aged women with acute coronary syndrome. Psychoneuroendocrinology, 2002; 27:893–906.

Pedersen

, Denollet

. Type D personality, cardiac events, and impaired quality of life: A review. Eur J Cardiovasc Prev Rehabil, 2003; 10:241–248.

Karasek

, Theorell

, Schwartz

, Schnall

, Pieper

, Michela

. Job characteristics in relation to the prevalence of myocardial infarction in the US Health Examination Survey (HES) and the Health and Nutrition Examination Survey (HANES) Am J Public Health, 1988; 78:910–918.

Alfredsson

, Karasek

, Theorell

. Myocardial infarction risk and psychosocial work environment: An analysis of the male Swedish working force. Soc Sci Med, 1982; 16:463–467.

Kuper

, Marmot

. Job strain, job demands, decision latitude, and risk of coronary heart disease within the Whitehall II study. J Epidemiol Community Health, 2003; 57:147–153.

Johnson

, Hall

, Theorell

. Combined effects of job strain and social isolation on cardiovascular disease morbidity and mortality in a random sample of the Swedish male working population. Scand J Work Environ Health, 1989; 15:271–279.

Eriksen

. The role of social support in the pathogenesis of coronary heart disease. A literature review. Fam Pract, 1994; 11:201–209.

10.

Bunker

, Colquhoun

, Esler

et al. “Stress” and coronary heart disease: Psychosocial risk factors. Med J Aust, 2003; 178:272–276.

11.

Koh

, Choe

, An

. Anger and coronary calcification in individuals with and without risk factors of coronary artery disease. Yonsei Med J, 2003; 44:793–799.

12.

Iribarren

, Sidney

, Bild

et al. Association of hostility with coronary artery calcification in young adults: The CARDIA study. Coronary Artery Risk Development in Young Adults. JAMA, 2000; 283:2546–2551.

13.

Yan

, Liu

, Matthews

, Daviglus

, Ferguson

, Kiefe

. Psychosocial factors and risk of hypertension: The Coronary Artery Risk Development in Young Adults (CARDIA) study. JAMA, 2003; 290:2138–2148.

14.

Friedman

, Rosenman

. Association of specific overt behavior pattern with blood and cardiovascular findings; blood cholesterol level, blood clotting time, incidence of arcus senilis, and clinical coronary artery disease. JAMA, 1959; 169:1286–1296.

15.

Richards

, Hof

, Alvarenga

. Serum lipids and their relationships with hostility and angry affect and behaviors in men. Health Psychol, 2000; 19:393–398.

16.

Niaura

, Stoney

, Herbert

. Lipids in psychological research: The last decade. Biol Psychol, 1992; 34:1–43.

17.

Brunner

, Marmot

, White

et al. Gender and employment grade differences in blood cholesterol, apolipoproteins and haemostatic factors in the Whitehall II study. Atherosclerosis, 1993; 102:195–207.

18.

Greenland

, Knoll

, Stamler

et al. Major risk factors as antecedents of fatal and nonfatal coronary heart disease events. JAMA, 2003; 290:891–897.

19.

Enos

, Holmes

, Beyer

. Coronary disease among United States soldiers killed in action in Korea; Preliminary report. JAMA, 1953; 152:1090–1093.

20.

McNamara

, Molot

, Stremple

, Cutting

. Coronary artery disease in combat casualties in Vietnam. JAMA, 1971; 216:1185–1187.

21.

Tejada

, Strong

, Montenegro

, Restrepo

, Solberg

. Distribution of coronary and aortic atherosclerosis by geographic location, race, and sex. Lab Invest, 1968; 18:509–526.

22.

World Health Organization. Report of a WHO Expert Committee: Prevention of Coronary Heart Disease. Geneva, Switzerland: World Health Organization, 1982.

23.

Van Horn

, Greenland

. Prevention of coronary artery disease is a pediatric problem. JAMA, 1997; 278:1779–1780.

24.

Marmot

, Theorell

. Social class and cardiovascular disease: The contribution of work. Int J Health Serv, 1988; 18:659–674.

25.

Stansfeld

, Fuhrer

, Shipley

, Marmot

. Psychological distress as a risk factor for coronary heart disease in the Whitehall II Study. Int J Epidemiol, 2002; 31:248–255.

26.

Pekkanen

, Tuomilehto

, Uutela

, Vartiainen

, Nissinen

. Social class, health behaviour, and mortality among men and women in eastern Finland. BMJ, 1995; 311:589–593.

27.

Goodkin

, Appels

. Behavioral–neuroendocrine–immunologic interactions in myocardial infarction. Med Hypotheses, 1997; 48:209–214.

28.

Brunner

. Stress mechanisms in coronary heart diesease. Stansfeld

, Marmot

. Stress and the Heart: Psychosocial Pathways to Coronary Heart Disease. London: BMJ Books, 2002; 181–199.

29.

Strike

, Steptoe

. Psychosocial factors in the development of coronary artery disease. Prog Cardiovasc Dis, 2004; 46:337–347.

30.

O'Malley

, Jones

, Feuerstein

, Taylor

. Lack of correlation between psychological factors and subclinical coronary artery disease. N Engl J Med, 2000; 343:1298–1304.

31.

Brydon

, Edwards

, Mohamed-Ali

, Steptoe

. Socioeconomic status and stress-induced increases in interleukin-6. Brain Behav Immun, 2004; 18:281–290.

32.

Colhoun

, Rubens

, Underwood

, Fuller

. Cross sectional study of differences in coronary artery calcification by socioeconomic status. BMJ, 2000; 321:1262–1263.

33.

Lim

, Williams

, Hagen

. Validation of a five-point self-rated stress score. Am J Health Promotion, 2005; 19:438–441.

34.

Agatston

, Janowitz

, Hildner

, Zusmer

, Viamonte

Jr , Detrano

. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol, 1990; 15:827–832.

35.

Mautner

, Mautner

, Froehlich

et al. Coronary artery calcification: Assessment with electron beam CT and histomorphometric correlation. Radiology, 1994; 192:619–623.

36.

Rumberger

, Simons

, Fitzpatrick

, Sheedy

, Schwartz

. Coronary artery calcium area by electron-beam computed tomography and coronary atherosclerotic plaque area. A histopathologic correlative study. Circulation, 1995; 92:2157–2162.

37.

Baumgart

, Schmermund

, Goerge

et al. Comparison of electron beam computed tomography with intracoronary ultrasound and coronary angiography for detection of coronary atherosclerosis. J Am Coll Cardiol, 1997; 30:57–64.

38.

Budoff

, Georgiou

, Brody

et al. Ultrafast computed tomography as a diagnostic modality in the detection of coronary artery disease: A multicenter study. Circulation, 1996; 93:898–904.

39.

Kajinami

, Seki

, Takekoshi

, Mabuchi

. Coronary calcification and coronary atherosclerosis: site by site comparative morphologic study of electron beam computed tomography and coronary angiography. J Am Coll Cardiol, 1997; 29:1549–1556.

40.

Gidding

, Bookstein

, Chomka

. Usefulness of electron beam tomography in adolescents and young adults with heterozygous familial hypercholesterolemia. Circulation, 1998; 98:2580–2583.

41.

Wilson

, D'Agostino

, Levy

, Belanger

, Silbershatz

, Kannel

. Prediction of coronary heart disease using risk factor categories. Circulation, 1998; 97:1837–1847.

42.

Brantley

, Bodenlos

, Cowles

, Whitehead

, Ancona

, Jones

. Development and validation of the Weekly Stress Inventory-Short Form. J Psychopathol Behav Assess, 2007; 29:55–60.

43.

Metcalfe

, Smith

, Wadsworth

et al. A contemporary validation of the Reeder Stress Inventory. Br J Health Psychol, 2003; 8:83–94.

44.

Hewitt

, Flett

, Mosher

. The Perceived Stress Scale: Factor structure and relation to depression symptoms in a psychiatric sample. J Psychopathol Behav Assess, 1992; 14:247–257.

45.

Goetzel

, Anderson

, Whitmer

, Ozminkowski

, Dunn

, Wasserman

. The relationship between modifiable health risks and health care expenditures. An analysis of the multi-employer HERO health risk and cost database. J Occup Environ Med, 1998; 40:843–854.

46.

Goetzel

, Pei

, Tabrizi

et al. Ten modifiable health risk factors are linked to more than one-fifth of employer-employee health care spending. Health Aff (Millwood), 2012; 31:2474–2484.