In this review, we find evidence that work stressors are associated with increased levels of catecholamines (particularly from urinary samples), lower levels of heart rate variability, a greater cortisol awakening response, higher levels of prolactin and lower levels of testosterone (see ). Thus there is evidence that work stressors are related to elevated stress responses in terms of sympatho-adrenal and HPA axis biomarkers.
Summary figure of the association between work stressors and neuroendocrine biomarkers
Our review of the studies available suggests that measurement of biomarkers of work stress tends to be limited to small focussed studies rather than large scale epidemiological cohorts. Studies are also largely cross sectional, with few studies on longitudinal measurements of work based stress and autonomic or neuroendocrine function. Despite measuring work stress only cross-sectionally, a number of studies discuss how chronic work stress leads to blunted stress reactivity profiles. This had led to the interpretation of reduced levels of stress biomarkers with work stress as evidence of the effect of chronic work stress on physiological stress responses. However, only longitudinal studies would be able to examine the suggested bi-modal response of these outcomes to work based stressors; whether work stress leads to an initial activation followed by exhaustion of the pathways.(Appels, 1997
) Furthermore, even longitudinal assessment of work stressors may not necessarily reveal exhaustion if the assessment periods of work stress has not been long enough.
We did not conduct a formal meta-analysis of the literature because of the wide heterogeneity in the measurement of the work stress exposure and stress response. For example, although heart rate variability had the most consistent associations in the literature, the methodology varied considerably between studies. For example, Chandola et al. (Chandola et al. 2008
) measured heart rate variability for 15 minutes during a clinical visit and found a positive association between job strain and lower HRV while Riese et al., (Riese et al. 2004
) measured heart rate variability for 24 hours on two occasions and failed to find an association. Measurement decisions are based on many factors including participant burden, size of study, but harmonisation of methodologies across a number of studies may serve to reduce imprecision in findings.
We were careful to distinguish the association of the stressor, in this case work based stress from the stress response in our review. This review did not consider burnout, anxiety(Kawachi et al. 1995
) and panic disorder(Yeragani et al. 1993
) as within the remit as these measures do not distinguish the stress generating working conditions from the psychological stress response. We also did not consider inflammatory markers as within the remit of this review even though they may be important mediators between stressors and health. This is because we aimed to examine ‘upstream’ markers that may mediate the stressor-health relationship. In the allostatic load paradigm, these may be considered ‘primary mediators’ in the cascade. (McEwen and Seeman, 1999
) Inflammatory markers represent secondary mediators in the cascade. Additionally, inflammatory marker levels are substantially influenced by adiposity as adipocytes release inflammatory markers(Mohamed-Ali et al. 1997
) contributing up to one third of circulating levels.
Most of the studies reviewed used either the job strain or the effort-reward imbalance models to measure work place stressors. Although these models differ in their theoretical basis, they also measure overlapping concepts (especially in relation to job demands and efforts at work). The review did not find any systematic association of either model with specific biomarkers, suggesting that physiological stress reactions to work stress are basically the same irrespective of the source of stressors.
Our review suggests that work based stress measures are consistently associated with alterations in the autonomic nervous system such that there is an apparent alteration in sympathetic parasympathetic balance. These changes in autonomic function have been associated with adverse health outcomes. Delayed heart rate recovery and lower cardiac autonomic reactivity (heart rate, preejection period, and heart rate variability) were prospectively associated with higher carotid atherosclerosis over a 2-year period.(Heponiemi et al. 2007
) In 1999 participants of the Whitehall II study, measures of autonomic function are associated with some measures of work based stess and in prospective analyses with measures of health, in particular with obesity, blood pressure and cholesterol.(Britton et al. 2007
) This adds to the plausibility that work stress leads to coronary heart disease partly through activation of the physiological stress responses reviewed in this study.
The association of cortisol secretion with health outcomes is currently unclear with no consistent evidence from large scale prospective surveys of an independent association with health outcomes. Currently there is poor empirical evidence that the cortisol awakening response is associated with clinical end points while there is evidence that slope in diurnal cortisol secretion is predictive of increased mortality in patient populations.(Sephton et al. 2000