Stress and health: Lifespan stress and cardio-metabolic disease risk pathways through cardiovascular stress reactivity

Background information

An exciting opportunity for a full-time PhD period of study is available in the Faculty of Health Sciences and Sport, University of Stirling.

High and low physiological reactivity to psychological stressors is implicated in the development of cardiovascular (CV) disease, or CV risk factors, but what determines the magnitude of an individual’s reactivity to stress is less well understood. Childhood adversity is a potential key factor in predicting adult stress reactivity, and independently predicts disease risk.  However, what is not known is how childhood adversity and adulthood stressors interact to predict CV risk.  What is even less well understood is how stress across the lifespan might interact with stress reactivity to predict CV and cardio-metabolic disease risk.  This knowledge would help us to identify individuals at greatest risk of future disease through stress screening, and by understanding which type of stress is most important as well as the mechanisms of its impact (i.e. high or low CV reactivity).  This project will address this gap in knowledge by examining the interaction between childhood and adulthood stress and its association with stress reactivity as well as cardio-metabolic risk factors.  It will examine whether stress reactivity mediates any association between lifespan stress and cardio-metabolic risk in a cross-sectional experimental laboratory study and existing cohort datasets.  There is also potential to explore other associations such as the link with social networks, physical fitness and exercise dependence.

The primary supervisor (60%) Whittaker is a Professor of Behavioural Medicine and HCPC Health Psychologist working on interdisciplinary ageing research.  She has considerable experience of psychophysiological and neuroendocrine measurement, and multi-disciplinary research.  She is internationally renowned for research on cardiovascular stress reactivity.  Whittaker has led an EC Marie Curie ITN training PhD students in interdisciplinary research into physical activity in ageing and is Director of Research Development for Sport at Stirling.  Coffee (40%) the second supervisor is an expert on the psychology of sport, specifically group memberships and social networks, and has extensive PhD supervision experience and is a HCPC Sport and Exercise Psychologist.

This interdisciplinary expertise means the student will receive training in psychosocial assessments, psychophysiological testing, stress testing, neuroendocrine sampling and assays, physiological function tests, analysis of multi-disciplinary data, and public participant involvement among many other techniques, resulting in creating a PhD graduate with a broad range of interdisciplinary skills applicable for many settings. They will use Stirling’s Research COMPASS tool to identify training needs annually. 

Hosted within the Faculty of Health Sciences and Sport, the student will be provided with facilities, including their own desk in an office, and full access to all software and equipment as required. The student will participate in the faculty’s internal research group ‘Stirling Physical Activity Research, Knowledge & Learning Exchange - SPARKLE’. The research group has regular meetings, provides a forum for PhD students to discuss their research and gain support and advice from the broader teams. In addition, the student will have access to subject-specific training on a one-to-one basis with the supervisors and with other academics within the Faculty, and through their participation in internal and external subject-specific training events, including those designed by the PhD students in the Faculty.

Project description

There are large physiological variations in how individuals respond to stressful stimuli which have important health and behavioural implications (Carroll et al., 2009a; Chida and Steptoe, 2010; Phillips and Hughes, 2011).  Stress responses include the acute sympathetic nervous system response with adrenaline, measured via cardiovascular changes, as well as the slower cortisol response.  The magnitude of individual differences in CV reactions during acute exposure to psychological stress have been shown to be relatively consistent and stable over time, indicating that the extent of stress reactivity is an individual trait rather than state (Ginty et al., 2013; Sherwood et al., 1990). It is also clear that these differences can have an impact on health and behaviour (Carroll et al., 2017, 2009b; Chida and Steptoe, 2010); individuals who show large magnitude CV reactions to acute psychological stress are at increased risk of developing CV Disease (CVD).  Supportive evidence is provided by a number of large scale cross-sectional and prospective observational studies by ourselves and others that show associations between heightened CV reactions to laboratory stress exposures and hypertension (e.g., Carroll et al., 2011, 2003), atherosclerosis development (e.g., Everson et al., 1997), and left ventricular hypertrophy (e.g., Allen et al., 1997). Both qualitative reviews (e.g., Treiber et al., 2003) and a very elegant meta-analysis (Chida and Steptoe, 2010) of this evidence confirm that exaggerated CV stress reactions signal poor future CV health, and we have even shown a link with CVD mortality (Carroll et al., 2012).  Further, the Principal supervisor has recently demonstrated that individuals most at risk of hypertension at five year follow-up are those characterised by high blood pressure reactivity to acute stress in the presence of a small to moderate heart rate response, rather than those who had exaggerated stress reactivity across blood pressure and heart rate measures, as expected (Brindle et al., 2016), showing that the absence of a robust heart rate response to stress confers hypertension risk.  This association withstood adjustment for standard risk factors such as anti-hypertensive medication, body mass index, smoking status, age, sex, and socio-economic status, and thus perhaps reflects the increased risk associated with vascular as opposed to cardiac reactions determining the magnitude of blood pressure stress responses (Manuck, 1994).

In contrast, low or blunted CV reactivity to acute stress has been regarded as benign or even health protective.  However, more recent evidence suggests that low CV, and also low cortisol reactivity to stress are also correlated with negative health outcomes and behaviours.  For example, low or blunted CV and cortisol reactions to acute psychological stress characterise both smokers (e.g., Al’Absi, 2006; Ginty et al., 2014; Phillips et al., 2009) and those with alcohol (Panknin et al., 2002) and other substance addictions (Lovallo, 2006).  Indeed, blunted CV and/or cortisol reactions to stress are also evident in the adolescent offspring of alcoholic parents (Moss et al., 1999; Sorocco et al., 2006), pathological gamblers (Paris et al., 2010) and those identified as exercise dependent (Heaney et al., 2011).  This suggests that the link between addiction and blunted stress reactivity does not primarily reflect the chronic or acute effects of ingested toxins on autonomic function.   In addition, we and others have shown that blunted CV and cortisol reactivity are characteristic of individuals with obesity (e.g., Carroll et al., 2008; Jones et al., 2012; Phillips et al., 2012), symptoms of depression (e.g., Carroll et al., 2007; de Rooij et al., 2010; York et al., 2007), major depressive disorder (Rottenberg et al., 2007; Salomon et al., 2013, 2009), anxiety (de Rooij et al., 2010; Souza et al., 2015), and poor self-reported health (De Rooij and Roseboom, 2010; Phillips et al., 2009), and indeed predicts the likelihood of becoming obese (Carroll et al., 2008), depressed (Phillips et al., 2011), and subjectively unhealthy (Phillips et al., 2009) over five years.  Differences in samples, stress tasks, and psychological measures used illustrate that these findings are not specific to a particular dataset or laboratory.  These negative health outcomes and behaviours associated with blunted reactivity (obesity, poor self-reported health, depression, smoking, excessive alcohol intake) are, of course, also implicated in the development of CVD (Guh et al., 2009; Van der Kooy et al., 2007)  as well as associated cardio-metabolic risk factors and diseases such as diabetes (DeBoer, 2013; Slagter et al., 2014; Yu et al., 2015).  As such, blunted stress reactivity may indicate an alternative pathway to increased risk of CVD via negative health behaviours (see model, Appendix 1).

Although there is a substantial body of evidence demonstrating the correlates of both exaggerated and blunted CV and cortisol reactivity, the exact origins of reactivity magnitude are unknown (Phillips et al., 2013).  As we have shown that blunted reactors do not differ from exaggerated reactors in terms of task appraisals, effort, or physiological capacity to respond to stress (Brindle et al., 2017), these can be discounted as behavioural or physiological mechanisms.  However, a meta-analysis has provided strong support for a genetic component (Wu et al., 2010) and it has recently been suggested that genetic influences may interact with adverse childhood experiences to influence stress reactivity (Lovallo et al., 2016).  Therefore, early-life adversity has also been implicated as a potential pathway to reactivity differences. Childhood adversity is also associated with poor health; being related to CVD (Appleton et al., 2017) and pre-disease markers such as increased carotid intima-media thickness/plaque (Khan et al., 2011) as well as smoking (Anda et al., 1999), depression (Dahl et al., 2017), anxiety (Li et al., 2016), cardio-metabolic disease (Winning et al., 2015), behavioural impulsivity (Lovallo, 2013) and risky sexual behaviours (Thompson et al., 2017).  It is plausible that childhood adversity may, in part, link to these negative health outcomes via dysregulated stress reactivity.

Research has suggested that early-life trauma influences future stress reactivity magnitude (Carpenter et al., 2011; Heim et al., 2000). It was originally thought that stressful childhoods contribute to the development of exaggerated physiological reactions to stress (Boyce and Ellis, 2005). However, recent research suggests that childhood adversity/stress may instead lead to blunted heart rate (Lovallo et al., 2012) and cortisol reactivity (Goldman-Mellor et al., 2012). This is supported by evidence using functional magnetic resonance imaging; blunted responses in the brain regions that control stress reactions (e.g., amygdala, anterior cingulate cortex, etc.) are found in those reporting traumatic childhoods (Banihashemi et al., 2015). Despite the mixed findings, the bulk of research in healthy humans, implies that childhood adversity is associated with blunted and not exaggerated reactivity to acute psychological stress. However, clear explanations to why other studies have found conflicting evidence are lacking and thus the precise role of childhood adversity in determining reactivity magnitude remains unclear, particularly as studies are, by necessity, observational, and thus causality cannot be inferred. It is possible that childhood adversity might result in either blunted or exaggerated reactivity, depending on a range of intervening factors in adulthood, including stressful life events in adulthood. We have previously shown that high levels of stressful events in adolescence and in older age are associated with blunted CV responses to acute stress (Carroll et al., 2005; Phillips et al., 2005).  However, what is not known is whether early life and recent stressful life events interact to predict blunted or exaggerated stress reactivity, and thus, CVD risk.  Understanding this means we could use stress screening to quickly identify patients most at risk.

There is some evidence that childhood and adult stress interact to influence cortisol reactivity where those with early life adversity and adulthood distress showed the most blunted reactivity to acute stress (Goldman-Mellor et al., 2012). However, it would be important to test whether this extends to effects on cardiovascular reactivity and CVD risk.  This question regarding both childhood and adult stressful life events has also received little attention in the context of cardio-metabolic risk.  Various studies have indicated that childhood adversity as a risk factor for cardio-metabolic issues as an adult (Danese and Tan, 2014; Lehman et al., 2005; Miller et al., 2011; Slopen et al., 2013).  Further, adult has been associated with a range of measures of cardio-metabolic health including high blood pressure (Liu et al., 2017), adiposity (Wardle et al., 2011), and abnormal lipid and glucose levels (Bergmann et al., 2014). However, most studies examine the impact of childhood or adult stress independently, rather than together, despite the contention that early exposure to stress and adult stress experience may additively contribute to later health outcomes (Evans & Kim, 2010) (accumulation model), or alternatively that childhood adversity may amplify the response to future stress exposure in adulthood resulting in pathology (McLaughlin et al., 2010) (sensitisation model).  Further, it has been theorised that the trajectory of exposure to different risk factors across the lifespan can influence the development of future CVD, including risk factors such as depression and poor social support (Hardy et al., 2015).  This highlights the importance of considering the interactions between stress and other psychosocial factors as well as stress at different points in the lifespan.

A very recent study has partially addressed this gap by examining early life adversity and recent life events and their association in the prediction of cardio-metabolic disease risk.  This was measured through resting blood pressure, obesity markers, blood lipid and blood glucose measures, and showed that both types of stress were associated with increased body mass index, waist circumference and cardio-metabolic risk score (an aggregation of body fat and blood markers).  These effects were not mediated by psychosocial factors such as depression, personality or coping styles (Gebreab et al., 2018).  However, no interaction between childhood and adult adversity was shown in this study (suggesting additive not synergistic effects), and measurements did not incorporate stress reactivity, so it was not able to address the issue of the interaction of stress across the lifespan effects on reactivity.  One study has examined the link between childhood adversity, CV reactivity, and adolescent psychopathology, and shown CV reactivity to be a linking mechanism such that individuals subject to early life trauma displayed externalising psychopathology which was mediated by inefficient reactivity, characterised by blunted cardiac output responding (Heleniak et al., 2016).  However, as this focused on adolescent psychopathology, it was not able to examine the link between childhood trauma, CV reactivity and adult psychopathology or indeed CVD risk.  A further study in adults tested the impact of both childhood and adult stress on physiological functioning in adulthood, characterised as allostatic load; both types of stress predicted higher allostatic load, but early life stress potentiated the impact of adult stress on allostatic load in men only (Dich et al., 2015).  Again, this study focused on cardio-metabolic markers of ill health but did not measure CV reactivity.  Given that we know that high CV reactivity relates to CVD risk, but low CV reactivity is associated with many of the risk factors for CVD and cardio-metabolic risk (as outlined above) including depression, obesity, smoking, alcoholism, as well as childhood adversity and adult stress independently, it is crucial to understand how different trajectories of stress link to cardiovascular/cardio-metabolic risk across the life course.  Addressing this question is novel, as many existing datasets, while including cardio-metabolic risk factors, do not assess both childhood adversity and adult life events stress or do not contain CV reactivity data; whereas in cross-sectional student studies often CV and cortisol reactivity are not both measured, and/or childhood adversity or stressful life events are studied separately.  This project will address these gaps in knowledge and use it to understand how different types of stress relate through different pathways to increased CV disease risk.  This could then be applied to stress screening to inform early psychosocial interventions to reduce CV disease risk.

Consequently, the present project aims to assess the impact of trajectories of stress across the lifespan on cardio-metabolic risk as well as stress reactivity.  To address this aim, the project has two specific goals:

  • to examine the interaction between childhood adversity and adult stressful life events on CV and cortisol reactivity to acute psychological stress in a new experimental study
  • to examine whether any relationship between accumulation of stress across the lifespan and markers of cardio-metabolic risk, are linked via the magnitude of CV and cortisol reactivity in existing cohort studies.

Applicants are likely to be from a Psychology, Sports Sciences, Biomedical Sciences or Human Biology background, with experience of either psychological, physiological or lab-based methodologies and measures.  They will be supported in the development of new psychosocial, biological or physiological skills as appropriate including one-to-one training in neuroendocrine assays (Whittaker), social network mapping (Coffee), and psychosocial assessment (Whittaker).  Public engagement training will be administered by the supervisors who have a strong public engagement profile. 

Candidates are welcome to make informal enquiries about the project to Prof Anna C. Whittaker (

The project is self-funded, however the opportunity may occasionally arise for paid tasks relating to teaching and research within the Faculty. Further information relating to fees and funding can be found on our postgraduate tuition fees page.

Entry Requirements

The successful candidate should have:

  • a 1st or 2:1 degree in Psychology or a Health/Natural Sciences subject with a statistical component and experience of working with human participants
  • a Master degree in a relevant topic area (e.g. sports science, sport and exercise psychology, health psychology, public health,)
  • an interest in working with older adults
  • an interest and willingness to work on an interdisciplinary project
  • an interest in the research areas and methodologies involved in the project
  • good command of one package for statistical computing (e.g., SPSS, R, Stata, …)

How to apply

Applicants are asked to send in all the documentation listed below, attached as a single email, to using the subject header ‘Stress and health: Lifespan stress and cardio-metabolic disease risk pathways through cardiovascular stress reactivity’

Documents to attach include:

  • Academic Transcript(s) and Degree Certificate(s): Final degree transcripts including grades and degree certificates (and official translations, if needed) - scanned copy in colour of the original documents.
  • References: Two references on headed paper (academic and/or professional). At least one reference must be academic. The other can be academic or professional. Your references should be on official headed paper. These should also be signed by the referee. If your referees would prefer to provide confidential references direct to the University then we can also accept the reference by email, from the referee’s official university or business email account to clearly labelling the reference e.g. ‘SHAPE reference’
  • Copy of CV: detailing relevant education and work experience.
  • Applicant Statement: a brief 1-page letter of motivation, outlining your research interests and your thoughts on how you could contribute to our research agenda.

Applications will be assessed by the project team and shortlisted applicants may be invited to an interview late July/early August. The successful candidate will be based in the Faculty of Health Sciences and Sport, University of Stirling. It is anticipated the candidate would start in September 2020 (flexible).