Psychosocial stress and cardiovascular disease: An exposure science perspective
By Thomas Kamarck
Thomas Kamarck is a professor of psychology and psychiatry at the University of Pittsburgh. His research examines the role of psychosocial factors in cardiovascular disease using ambulatory or ecological momentary methods, as a means of examining the mechanisms by which these factors may contribute to health risk. Kamarck completed his doctoral training in psychology at the University of Oregon and his postdoctoral training at the University of Pittsburgh. He received the Distinguished Early Career Contributions Award from the Society for Psychophysiological Research in 1991 and he is a Fellow in the Academy of Behavioral Medicine Research. Author website.
Chronic Stress and Cardiovascular Disease
Over the past 30 years, a large and varied body of research has examined the role of chronic psychosocial stress in the development of cardiovascular disease (CVD) (Cohen, Janicki-Deverts, & Miller, 2007; Kaplan, Chen, & Manuck, 2009; Rozanski, Blumenthal, & Kaplan, 1999). Psychosocial stress may be defined as aversive or demanding conditions that tax or exceed the behavioral resources of the organism (Lazarus, 1966). Chronic stress involves conditions with sufficient longevity to alter one’s long term risk trajectory, for example, by contributing to atherosclerosis (Strong, 1992). In humans, the largest and most consistent literature in this area is the research on occupational stress, and in the research linking occupational stress with CVD, the dominant paradigm is the “job strain” model (Belkic, Landsbergis, Schnall, & Baker, 2004; Kivimaki, et al., 2006). According to the job strain model, a workplace characterized by high Psychological Demands (high output requirements and multiple responsibilities) and low Decision Latitude or Control (task variety and control over how the work is done) is linked with enhanced CVD risk. Most of this work has relied upon existing survey data or standard questionnaires to measure the presence or absence of occupational stress. Reviews of the existing literature have concluded that such measures show significant associations with CVD, at least in men, but that there is substantial heterogeneity in the magnitude of observed effects (Belkic, Landsbergis, Schnall, & Baker, 2004; Kouvonen, Kivimaki, Virtanen, Pentti, & Vahtera, 2005).
Challenges in the Assessment of Psychosocial Stress Exposure
One important problem in most of the existing literature linking chronic stress and disease in humans involves limitations in the assessment of stress exposure. Most work in this field has relied upon the use of survey data or standard questionnaires to measure the presence or absence of stress, using single point estimates that require the participant to summarize his or her experience over prolonged periods of time. Research in autobiographical memory, however, suggests that we are not very proficient at these types of estimates or summaries (Stone, et al., 1999). In the developing field of “exposure science” (Lioy, 2010), cumulative exposures to environmental toxins are estimated using information from personal monitoring, capturing the temporal (frequency and duration) and spatial (concentration and proximity) features of contact with a toxin rather than relying upon point estimates of exposure. This perspective has been applied successfully to the study of a diverse range of toxins such as airborne cigarette smoke, pesticides, and asbestos exposure (Al-Delaimy, Crane, & Woodward, 2000; Panahi, Kakooei, Marioryad, Mehrdad, & Golhosseini, 2011; see also the Food Quality Protection Act (FQPA) Background website).
Although not usually discussed in this light, the literature on ambulatory blood pressure (ABP) monitoring embodies the principles exposure science. ABP measures, blood pressure measures collected in the natural environment, are more strongly associated with CVD risk than blood pressures (BP) collected in the doctor’s office (Conen & Bamberg, 2008). The superior prognostic value of ABP does not appear to be due to methodological confounds (for example, use of more observations for ABP than for standard clinic BP); our work and that of others suggests that ABP outperforms clinic BP as a correlate of subclinical cardiovascular disease even when the number of readings taken is comparable across settings (3 to 4 clinic BP vs. 3 to 4 ABP assessments; Gerin, et al., 2006; Kamarck, Polk, Sutton-Tyrrell, & Muldoon, 2002). One might conclude then that ABP is a better predictor than clinic BP because it provides a better representation of daily risk exposure, capturing cumulative exposures to heightened blood pressure more effectively than point estimates of BP from the doctor’s office.
This exposure science perspective may be applied to the measurement of behavioral factors as well as to chemical agents and BP dynamics (Schwartz, 2006); this expanded application of exposure science principles is reflected by recent Exposure Biology initiatives by the National Institutes of Health that have funded our work and others. Applying exposure science principles to the measurement of behavior may enhance our ability to demonstrate associations between chronic stress and disease, and may assist us in the identification, management and reduction of psychosocial stress (cf. Lioy, 2010). Here we outline how we have adopted principles of exposure science to the study of occupational stress and cardiovascular disease.
Developing Mobile Measurement Strategies for Capturing Psychosocial Stress
Ecological Momentary Assessment (EMA) methods involve sampling experience or behavior in the natural environment, in real time, and using repeated assessments (A. Stone & Shiffman, 1994). With the assistance of electronic data collection tools, brief self-report “interviews” can be presented at random or fixed intervals throughout the day in order to capture a representative slice of life for each participant. To the extent that they include personal monitoring in the natural environment and a focus on representative sampling, EMA methods are an optimal means of operationalizing the principles of exposure science in the assessment of chronic psychosocial stress (Kamarck, Shiffman, & Wethington, 2011).
In a recent study of CVD risk among initially healthy older adults (the Pittsburgh Healthy Heart Project or PHHP), we assessed ambulatory blood pressure every 45 minutes during waking hours for three 3-day periods; at baseline, 4 months later, and 6 years following the initial assessment (Kamarck, Muldoon, Shiffman, Sutton-Tyrrell, & Gwaltney, 2004). Participants were equipped with an ABP unit, an electronic motor worn on the belt attached to an inflatable arm cuff, and they carried a portable electronic diary throughout the day. During each of these 3-day monitoring periods, the cuff inflated to record participants’ BP every 45 minutes during the waking day. In conjunction with each BP reading, participants were asked to evaluate their current activities using items presented by the electronic diary. Among the constructs we assessed were Demand (“Required working hard?” “Required working fast?” “Juggling several tasks at once?”) and Control (“Could change activity if you chose to?” and “Choice in scheduling this activity?”), each assessed using continuous scales ranging from No! to Yes!. These items were derived from the job strain model of occupational stress described above, but they were modified to apply to one’s current daily activities (“in the past ten minutes”) rather than assessing, in a more general fashion, the characteristics of one’s workplace.
Participants completed about 50 electronic diary interviews over the course of each of the three-day monitoring periods in this study. Because subjects were asked to rate their current activities for each interview, ratings fluctuated over time, as expected, and changes in this “within-person” or state-level variance (accounting for over 2/3 of the variability in ratings) were associated, as expected, with fluctuations in the psychosocial environment. For example, ratings of Demand were significantly higher during periods of the day that participants were at work than when they were at home, and ratings of Control were lower at work compared to home (Kamarck, Shiffman, Smithline, et al., 1998). We have shown longitudinal changes in the experiences of daily life with aging; on average, daily activities are rated as less psychologically demanding and more controllable as we age (over the six year follow-up period, the sample progressed from age 60 to 66 on average), especially among those who have transitioned into retirement. Such changes appear to moderate effects of aging on blood pressure during daily living: Among adults in this age range who are not taking blood pressure medication, mean ambulatory systolic blood pressure increases substantially (about 11 mmHg) over a 6-year period (Kamarck, Shiffman, Sutton-Tyrrell, Muldoon, & Tepper, in press) but this occurs to a significantly smaller degree among those who show larger six-year declines in self-rated daily Demand (Kamarck, Shiffman, Garcia, & Muldoon, 2009).
In addition to examining fluctuations in daily experiences over time, we can use time-averaged EMA ratings to measure individual differences in exposures that are relatively stable. Three-day averaged assessments of Demand and Control show substantial test-retest reliability over a 4-month period (r=.73, r= .70, n= 354; Kamarck, Shiffman, et al., 2002). Even six-year stability of these ratings are quite substantial (r=.56 for Demand, .49 for Control, n=225) and this stability is on the same order of magnitude as averaged measures of ambulatory blood pressure measured at the same time (r= .55 for SBP, r= .61 for DBP; Kamarck, et al., in press). Such stability is important, as it suggests that these measures reflect chronic sources of stress that may exert a cumulative impact on health outcomes. Time-averaged assessments of Demand and Control are associated, as expected, with traditional measures of occupational stress, as well as with other conventional measures of trait negative mood (Kamarck, Muldoon, et al., 2004). Such associations are only modest to moderate in magnitude, however, suggesting that the EMA ratings capture some unique determinants which may not be represented when the assessments are based upon single-point estimates of stress exposure.
Daily Psychosocial Demands Are Associated with Cardiovascular Risk
Consistent with the tenets of exposure science, our data suggest that EMA assessments of psychosocial stress yield useful estimates of cumulative health risk exposures. We have shown that ambulatory blood pressure (ABP) is relatively elevated during periods of the day that are rated as more demanding and less controllable, even after controlling for posture, activity, and substance use (within-subject analyses) (Kamarck, Janicki, Shiffman, Polk, Muldoon, et al., 2002). At the same time, people who rate their daily lives as more demanding and less controllable, in general, have greater mean ABP, even after adjusting for demographic differences (between-subject analyses using data averaged across observations and days). These within-subject and between-subject findings are interrelated. That is, the reason ABP is higher for individuals who report higher demands is that it tends to be higher acutely at the times that higher demands are reported (that is, in multilevel models, mean effects are completely accounted for by momentary effects). In short, measures that focus on daily life experiences may be particularly well suited for capturing determinants of biological exposures, like blood pressure, that also fluctuate over the course of daily living, and their ability to capture the acute determinants of these biological exposures may be one reason that time-averaged EMA measures of psychosocial stress might be expected to be particularly effective in predicting disease risk.
Because they have the capacity to sample biology and behavior simultaneously and in real time, this type of measurement strategy may allow us to examine some of the daily life mechanisms accounting for the association between stress and disease. In our data, those who rate their daily lives as more demanding, on average, not only show greater ABP, but also greater intima-medial thickening (IMT) in the carotid arteries, an association that appears to be independent of demographic characteristics and cardiovascular risk factors (Kamarck, et al., 2004). IMT is an ultrasonographic marker of atherosclerosis that has been shown to be associated with increased risk for future heart attack and stroke (O'Leary & Polak, 2002), so the association between daily demands and IMT can be presumed to have important prognostic significance. Most recently, we have shown that these same ratings of daily Demand even predicted six-year progression of IMT and carotid artery plaque, but only among those who were not medically treated for blood pressure reduction over the follow-up period (Kamarck, et al., in press). These latter findings are consistent with the possibility that ABP may partially mediate the association between daily experiences of Demand and disease progression. Indeed, parallel analyses with our cross sectional as well as our longitudinal data are consistent with such a mediation effect (Kamarck, et al., in press; Kamarck, et al., 2004). A demanding psychosocial environment may elicit recurrent activation of sympathetic nervous system pathways during daily life in a manner that may contribute both to elevated ABP as well as to some of the pathophysiological changes that contribute to early signs of atherosclerosis.
Among those who were employed, EMA measures of Demand taken in the workplace predicted IMT and six-year IMT progression in the PHHP sample, whereas traditional measures of Psychological Demands did not (Kamarck, et al., 2004; Kamarck et al., in press). In short, our EMA-based assessment tool for measuring daily psychosocial stress is not only useful as an index of disease risk, but it may outperform traditional measures as well. We are intrigued by the possibility that assessments of psychosocial stress informed by exposure science may yield findings more consistently linked with human health than the measures that are traditionally employed in these literatures.
There are a number of steps that we are currently taking to replicate and extend our work. First, we are examining the utility of these exposure-based measures of psychosocial stress in a larger independent sample of employees, an approach that will permit us to better interpret our findings in the context of the existing literature. Second, we are expanding the scope of our mobile interviewing tools and methods, in a manner that we hope will ultimately enhance our understanding of psychosocial stress and the mechanisms of its effects. And third, we are investigating methods to make these types of assessment tools more exportable, to facilitate future research on exposures to psychosocial stress in the context of daily living.
The research described here was supported by the National Institutes of Health (HL56345, DA7005, HL40962) and was conducted in close collaboration with Saul Shiffman, Matthew Muldoon, Stephen Manuck, Joseph Schwartz, Dan Siewiorek, Asim Smailagic, Barbara Anderson, Clem Stone, Elaine Wethington, Kim Sutton Tyrrell, and Ping Tepper.
Al-Delaimy, W., Crane, J., & Woodward, A. (2000). Questionnaire and hair measurement of exposure to tobacco smoke. Journal of Exposure Analysis and Environmental Epidemiology, 10, 378-384.
Belkic, K. L., Landsbergis, P. A., Schnall, P. L., & Baker, D. (2004). Is job strain a major source of cardiovascular disease risk? Scandinavian Journal of Work and Environmental Health, 30, 85-128.
Cohen, S., Janicki-Deverts, D., & Miller, G. (2007). Psychological stress and disease. JAMA, 298, 1685-1687.
Conen, D., & Bamberg, K. (2008). Noninvasive 24-hour ambulatory blood pressure and cardiovascular disease: A systematic review and meta-analysis. Journal of Hypertension, 26, 1290-1299.
Gerin, W., Schwartz, J. E., Devereux, R. B., Goyal, T., Shimbo, D., Ogedegbe, G., et al. (2006). Superiority of ambulatory to physician blood pressure is not an artifact of differential measurement reliability. Blood Pressure Monitoring, 11, 297-301.
Kamarck, T. W., Janicki, D.L., Shiffman, S., Polk, D.E., Muldoon, M.F., Liebenauer, L.L. & Schwartz, J.E. (2002). Psychosocial demands and ambulatory blood pressure: A field assessment approach. Physiology and Behavior, 77, 699-704.
Kamarck, T. W., Muldoon, M., Shiffman, S., Sutton-Tyrrell, K., & Gwaltney, C. J. (2004). Experiences of demand and control in daily life as correlates of subclinical carotid atherosclerosis in a healthy older sample: the Pittsburgh Healthy Heart Project. Health Psychology, 23, 24-32.
Kamarck, T. W., Polk, D. E., Sutton-Tyrrell, K., & Muldoon, M. F. (2002). The incremental value of ambulatory BP persists after controlling for methodological confounds: Associations with carotid atherosclerosis in a healthy sample. Journal of Hypertension, 20, 1-7.
Kamarck, T. W., Shiffman, S., Garcia, J., & Muldoon, M. (2009). Daily psychosocial demands are associated with six-year changes in ambulatory blood pressure: The Pittsburgh Healthy Heart Project. Psychosomatic Medicine, 71, A-39.
Kamarck, T. W., Shiffman, S., Janicki, D., Liebenauer, L. L., Glasgow, C., Gwaltney, C., et al. (2002). Evaluating an EMA-based assessment os psychosocial risk: Replication and extension. Annals of Behavioral Medicine, 24, S096.
Kamarck, T., Shiffman, S., Sutton-Tyrrell, K., Muldoon, M., & Tepper, P. (in press). Daily psychological demands are associated with six-year progression of carotid artery atherosclerosis: The Pittsburgh Healthy Heart Project. Psychosomatic Medicine.
Kamarck, T. W., Shiffman, S., & Wethington, E. (2011). Measuring psychosocial stress using ecological momentary assessment methods. In R. J. Contrada & A. Baum (Eds.), The handbook of stress science: Biology, psychology, and heatlh (pp. 597-617). New York: Springer.
Kamarck, T. W., Shiffman, S.M., Smithline, L., Goodie, J.L., Thompson, H.S., Ituarte, P.H.G., Jong, J.Y., Pro, V., Paty, J.A., Kassel, J.D., Gnys, M., et al. (1998). The Diary of Ambulatory Behavioral States: a new approach to the assessment of psychosocial influences on ambulatory cardiovascular activity. In D. S. Krantz, & A. Baum, (Ed.), Perspectives in behavioral medicine: technology and methods in behavioral medicine (pp. 163-194). Mahwah, NJ: Lawrence Erlbaum.
Kaplan, J. R., Chen, H., & Manuck, S. B. (2009). The relationship between social status and atherosclerosis in male and female monkeys as revealed by meta-analysis. American Journal of Primatology, 71, 732-741.
Kivimaki, M., Viortanen, M., Elovainio, M., Louvonen, A., Vaananen, A., & Vahtera, J. (2006). Work stress in the etiology of coronary heart disease: A meta-analysis. Scandinavian Journal of Work Environment and Health, 32, 431-432.
Kouvonen, A., Kivimaki, M., Virtanen, M., Pentti, J., & Vahtera, J. (2005). Work stress, smoking status, and smoking intensity: an observational study of 46,190 employees. Journal of Epidemiology and Community Health, 59, 63-69.
Lazarus, R. (1966). Psychological stress and the coping process. New York: McGraw-Hill.
Lioy, P. J. (2010). Exposure science: A view of the past and milestones for the future. Environmental Health Perspectives, 118, 1081-1090.
O'Leary, D. H., & Polak, J. F. (2002). Intima-medial thickness: A tool for atherosclerosis imaging and event prediction. American Journal of Cardiology, 90, 18L-21L.
Panahi, D., Kakooei, H., Marioryad, H., Mehrdad, R., & Golhosseini, M. (2011). Evaluation of exposure to the airborne asbestos in an asbestos cement sheet manufacturing industry in Iran. Environmental Monitoring and Assessment, 178, 449-454.
Rozanski, A., Blumenthal, J. A., & Kaplan, J. (1999). Impact of psychological factors on the pathogenesis of cardiovascular disease and implications for therapy. Circulation, 99, 2192-2217.
Schwartz, D. A. (2006). The importance of gene-environment interactions and exposure assessment in understanding human diseases. Journal of Exposure Science and Environmental Epidemiology, 16, 474-476.
Stone, A., & Shiffman, S. (1994). Ecological momentary assessment (EMA) in behavioral medicine. Annals of Behavioral Medicine, 16, 199-202.
Stone, A. A., Turkkan, J. S., Bachrach, C. A., Jobe, J. B., Kurtzman, H. S., & Cain, V. S. (Eds.). (1999). The science of self-report: Implications for research and practice. Mahwah, New Jersey: Lawrence Erlbaum Associates.
Strong, P. (1992). Atherosclerotic lesions: Natural history, risk factors, and topography. Archives Pathology Laboratory Medicine, 116, 1268-1275.
The views expressed in Science Briefs are those of the authors and do not reflect the opinions or policies of APA.
PSA is the monthly e-newsletter of the APA Science Directorate. It is read by psychologists, students, academic administrators, journalists, and policymakers in Congress and federal science agencies. Subscribe here.