Psychological stress is frequently assumed to contribute to cardiovascular disease, but empirical support for this belief is just beginning to emerge.1 One approach to stress examines the cardiovascular responses to challenging tasks presented in laboratory settings. Individual differences in such cardiovascular reactivity are correlated with the current and future occurrence of a number of cardiovascular diseases.1 “Preclinical”2 indices of cardiovascular disease have typically been employed because such indices can be examined in community samples, permitting identification of disease in its early stages. Carotid intima-media thickness (IMT), for example, is related to systemic and coronary atherosclerosis as well as incidence and prevalence of myocardial infarction and stroke.3,4
We previously observed that the amplitude of blood pressure reactions to laboratory challenges were associated with carotid IMT measurements collected concurrently in the population-based Kuopio Ischemic Heart Disease (KIHD) study.5 We now report on the same measures of blood pressure reactivity as predictors of 7-year progression of carotid IMT.6
Participants and methods
Participants. KIHD is an epidemiologic investigation sampling the community of Kuopio, Finland—a relatively high-risk geographical region.7 KIHD is composed of four cohorts of men aged 42, 48, 54, and 60 years at the time of study initiation. Between 1991 and 1993, we tested 901 men from this study with both a standardized laboratory stress reactivity battery and with carotid ultrasound. Seven years later, between 1999 and 2001, 756 of these men participated in a carotid artery ultrasound follow-up.
Carotid atherosclerosis. High-resolution B-mode ultrasonography was used to assess right and left common carotid arteries (CCAs) in a 1.0- to 1.5-cm section at the distal end of the CCA, proximal to the carotid bulb. A Biosound Phase 2 scanner equipped with a 10-MHz annular array probe was employed in both initial and follow-up examination.8 Prosound software (University of Southern California, Los Angeles, CA) provided a computerized analysis of the videotaped ultrasound images, using an edge-detection algorithm.9 This permitted assessment of IMT at approximately 100 points in both the right and left CCAs. In addition, lipoproteins and glucose were assessed with standard techniques.
Cardiovascular reactivity testing. An automated test battery assessed individual differences in cardiovascular reactivity. This consisted of four standardized computer-based tasks, each 9 minutes long (memory task, reaction time task, tracing task, and computerized version of the Stroop Color-Word task).10 Tasks were separated by a 9-minute baseline (rest and recovery) period. The computer maintained difficulty of the tasks at approximately 60% success. Maintaining a consistent challenge contributed to the ability of this battery to show consistent individual differences in cardiovascular reactivity over repeated assessments.10
A two-lead electrocardiogram was continuously assessed and blood pressure measurements (Dinamap Vital Signs Monitor) were taken every 90 seconds during the baseline and task periods. Heart rate, systolic, and diastolic blood pressures were averaged separately across the 9-minute period for each rest period and each task. Reactivity was then scored by subtracting the average rest period values from each averaged task score and averaging standardized measures across tasks.
Three measures of IMT were used: (1) mean IMT, the mean of all IMT estimates from the right and left CCAs; (2) maximum IMT, the mean of the points of maximum thickness from
the right and left CCAs; and (3) plaque height, the mean of right and left CCA measurements of plaque height, measured by the difference between maximum and minimum thickness.
Analyses. Blood pressure and heart rate responses were used in regression analyses to predict IMT. Age and educational level (factors known to influence cardiovascular responses) were entered with the physiological responses initially, and then IMT from the initial assessment was added. Finally, analyses were repeated including standard risk factors (smoking status, low-density lipoprotein, high-density lipoprotein, serum triglyceride, fasting serum glucose, and resting systolic and dia-stolic blood pressures). Separate analyses were conducted for younger and older age groups and for participants reporting no cardiovascular disease
or cardiovascular medication usage at the initial assessment.
Seven-year prospective relations between reactivity and carotid measures. Enhanced blood pressure reactions to mental stress at the initial testing were related both to greater atherosclerosis 7 years later and to the increase in atherosclerosis over these 7 years. Controlling only for age and education, systolic blood pressure reactions were significantly related to degree of mean and maximal IMT as well as the plaque score (P < .001). This relationship remained after controlling for the carotid IMT at initial testing, thus assessing progression of IMT thickening (P < .008). Diastolic blood pressure responses to the mental stress challenges showed the same pattern of results, albeit with less statistical significance (P < .02 with age and education covaried; P < .08 with initial IMT added).
Heart rate responses were unrelated to IMT measures. Figure 1 presents the relationship between systolic blood pressure reactivity at initial testing and carotid IMT assessed 7 years later. For illustrative purposes, values of systolic blood pressure reactivity are separated into quintiles, and the younger and older two cohorts are presented separately to illustrate the change in IMT with age and the applicability of our finding to both age groups. Figure 2 presents the relationship between quintiles of systolic blood pressure reactivity and progression of thickness of carotid IMT over 7 years. The Figures further illustrate the possibility, which we cannot yet interpret, that the relationship between systolic blood pressure reactivity and carotid IMT may not be strictly linear.
An important issue is whether traditional risk factors may explain the relationship between blood pressure reactivity to laboratory challenge and progression of carotid artery disease. In fact, covarying known risk factors did not significantly modify the positive relationship between systolic blood pressure reactions and carotid IMT or its progression. The Table shows the results of this analysis. Diastolic blood pressure relationships were eliminated when risk factors were covaried. The Table presents standardized beta values in order to permit direct comparison of the strength of relationship between risk factors.
Analyses of healthy individuals not taking medications at the time of initial testing (n = 195) continued to support the relationship between systolic blood pressure reactivity and subsequent carotid IMT (P < .02), but covarying the initial carotid IMT reduced the significance to P < .10. Analyses covarying and blocking on age did not demonstrate any age differences in the effects of blood pressure reactivity on IMT (despite such a finding in our initial cross-sectional report5).
Systolic blood pressure reactivity to mental stress was observed to be positively and prospectively related to carotid artery IMT measures taken
7 years later. Mean carotid IMT was 0.035 mm thicker for each standardized unit of mental-stress induced change of systolic blood pressure (approximately 7 mm Hg of change in blood pressure). This prospective relationship was not altered when both the preexisting carotid artery values and recognized risk factors were taken into account. A change of approximately 7 mm Hg of blood pressure reactivity to stress was related to a progression over 7 years of 0.02 mm of thickening of the carotid IMT. Based on previous work, this would translate into an increase of approximately 1% every 3 years in risk for myocardial infarction.4,11 This magnitude of association may have public health significance and possibly clinical significance when considered in concert with other independent risk factors.
The results reported here provide substantial support for the hypothesis that cardiovascular reactivity is re-
lated to the progression of the atherosclerosis. The current results are based on a larger sample and more thorough measurements than previous work.12-15 The current results are consistent with these studies and, taken together, suggest the predictive validity of magnitude of systolic blood pressure responses to laboratory challenges regardless of gender and nature of the sample, eg, community versus clini-cal sample.
A number of limitations of this study should be noted. The sample was all men and from a high-risk area of eastern Finland. This limits generalization. We did not study mortality or morbidity of myocardial infarction or stroke. The methods are correlational and unknown/unmeasured factors related to blood pressure reactivity might account for the results. Finally, the results do not illuminate any mechanism for the relationship of reactivity and atherosclerosis (although a number of mechanisms have been proposed1). We do claim that relative to our prior cross-sectional findings,5 these prospective results more strongly support a role for cardiovascular reactivity in the atherosclerotic process.