Old ideas, new directions: re-examining the predictive utility of the hemodynamic profile of the stress response in healthy populations

ABSTRACT

The ‘reactivity hypothesis’ has a long and fruitful history in health psychology and behavioural medicine, with elements of its thesis taken as core and others lost in the plethora of research on its utility as a theory of psychosomatic disease. One such thesis is that the underlying hemodynamic profile of the stress response may be particularly revealing when detailing the impact of psychological stress on the development of cardiovascular disease. This paper re-examines old ideas surrounding the hemodynamic profile of the stress response, asking why its health-predictive properties were never fully explored. Further, this paper reviews the evidence that a vascular profile of stress responding may be especially predictive of disease development, particularly in the case of hypertension. In addition, measurement of hemodynamic profile, as well as its known psychosocial moderators, is reviewed including how examination of patterns of cardiovascular stress response adaptation may extend the field. This paper highlights that the extension of the reactivity hypothesis to include both hemodynamic profile and patterns of cardiovascular stress response adaptation may hold much explanatory power in detailing the impact of how stress responding and stress tolerance promotes disease development.

KEYWORDS:

• Cardiovascular reactivity
• stress
• hemodynamic profile
• stress response adaptation

The ‘reactivity hypothesis’, that exaggerated and/or sustained increases in blood pressure and heart rate in response to stress is either a causal or contributory factor in the development of cardiovascular disease is neither new nor revolutionary; but it is arguably health psychology’s greatest contribution to medicine. The notion that psychological stress leads to biological changes that promote disease development has been powerful in the medical literature, particularly in relation to cardiovascular diseases. In psychology, it is basic knowledge now, that how we think, feel, and behave has important and long-lasting implications on our health. Known to affect health in a myriad of ways, psychological stress and its physiological consequences continue to be at the core of health psychology and behavioural medicine.

The reactivity hypothesis has a long history in health psychology and behavioural medicine. Since its detailed explanation in the 1980s (Obrist, 1981), across 40 years of research, many studies have confirmed its prospective value in predicting a range of indicators of cardiovascular disease and disease endpoints, including future hypertension (e.g., Carroll, Ginty, Painter, et al., 2012), atherosclerosis (e.g., Barnett et al., 1997), and cardiovascular disease mortality (Carroll, Ginty, Der, et al., 2012), although research is not always supportive of its prognostic role in cardiovascular disease development (Bourassa et al., 2021). In fact, while it appears there exists a reliable association between exaggerated reactivity to stress and disease, the strength of the association is often low (Chida & Steptoe, 2010) and suggests exaggerated reactivity has the greatest impact on progression of disease rather than initiation (Alderman et al., 1990; Lovallo, 2010; Manuck et al., 1992). Further, in the last 10 years, a large literature has emerged identifying low cardiovascular reactivity (or ‘blunted reactivity’) to be prospectively associated with a range of negative outcomes (e.g., Phillips et al., 2013; Whittaker et al., 2021). Today, a ‘cardiovascular reactivity’ search in APA PsycInfo returns nearly 5000 ‘hits’, speaking to its immense popularity in psychology, with a number of position and review articles appearing over the years; from Manuck’s (1994) excellent overview of the status quo in the 1990s, the more recent Chida and Steptoe’s (2010) review highlighting the predictive utility of the reactivity hypothesis on disease endpoints, as well as Phillips et al.’s (2013) exploration of the health-relevant outcomes of blunted reactivity.

However, amongst the plethora of research exploring the many facets of the reactivity hypothesis, some elements although highlighted at an early stage, have all but disappeared from view. While these elements occasionally appeared on the periphery of the literature, despite the promise of much explanatory power, they remained largely unexplored. One such feature is the hemodynamic profile of the stress response.

Hemodynamic profile of the stress response

Hemodynamic profile describes the dynamic relationship that exists between cardiac output (CO) and total peripheral resistance (TPR). Blood pressure change can be a function of a change in CO, a change in TPR, or a change in both (Turner et al., 1994). While the reactivity hypothesis has mainly detailed the predictive utility of blood pressure and heart rate responses to stress on disease endpoints (for a review see Chida & Steptoe, 2010), examination of the hemodynamic profile of the stress response can be particularly revealing. Equivalent blood pressure change can be driven by vastly different movements on CO and TPR. In fact, CO and TPR share a reciprocal relationship where an increase in one parameter should be matched by a decrease in the other, thereby ensuring homeostatic control of blood pressure. Therefore, blood pressure elevation in response to stress is driven by one of three potential mechanisms: (1) an increase in CO not offset by a compensatory decrease in TPR; (2) an increase in TPR not offset by a compensatory decrease in CO; or (3) a (often, smaller) increase in both parameters. These hemodynamic patterns of responding reveal vastly different profiles in the stress response that may have implications for disease processes. All three patterns, though different, can lead to equivalent blood pressure change.

An early indication of the value in examining the underlying hemodynamic determinants of blood pressure change was demonstrated by Sherwood and Turner (1993). While equivalent changes in blood pressure were observed in response to a stress-task in participants when both standing and seated, the underlying hemodynamic variables of CO and TPR were significantly altered. While the effects of the alterations in these hemodynamic variables were explained by gravity acting upon the columns of blood in major vessels (Sherwood & Turner, 1993), the results highlighted an important point for consideration in reactivity research at the time; unchanged or equivalent blood pressure responses to a task may not necessarily indicate participants are not reacting. Rather, their reactivity may be revealed by examination of the underlying hemodynamic variables of CO and TPR, an assertion previously supported by others (e.g., Light et al., 1993; Light & Sherwood, 1989; Manuck et al., 1990), and shown to underlie blunted cardiovascular reactions to stress in some samples (e.g., O’Leary et al., 2013).

The fact that examination of the hemodynamic profile of the stress response may hold explanatory power in profiling how stress responsivity contributes to disease risk is not new. In fact, it was 1955 when Wolf et al. (1955) reported two different hemodynamic patterns evident in response to stressful interviews; high output (marked by high CO) and high resistance (marked by high TPR). Wolf and Wolff (1951) had previously claimed that normotensives were more likely to exhibit high output while hypertensive were more likely to exhibit high resistance. Hejl (1957) continued this work and identified a third pattern; increases on both parameters. These patterns of the hemodynamic profile are now known as myocardial (blood pressure change driven by the increase on CO with little or no decreases on TPR), vascular (blood pressure change driven by increases on TPR with little or no decreases on CO), and mixed (small increases on both CO and TPR). A brief summary of these profiles is shown in Table 1.

Table 1. Summary of myocardial, vascular, and mixed (dual) hemodynamic profiles in response to stress.

Profile	Myocardial	Vascular	Mixed
Cardiac output	Increase	Small decrease	Small increase
Total peripheral resistance	Small decrease	Increase	Small increase

What is important to consider at this point, is that examination of CO or TPR alone as an outcome variable cannot reveal the precise nature of the hemodynamic profile of the stress response. The terms myocardial and vascular denote the profile as driven by the reciprocal relationship between CO and TPR, with a myocardial profile identified by reactions mainly due to increases on CO that are not entirely offset by a decrease in TPR, and a vascular profile identified by reactions mainly due to increases on TPR not offset by a decrease in CO. Therefore, it is important to consider in evaluating research whether the true nature of this reciprocal relationship is being described, or rather, data are reporting cardiac or vascular reactivity (as opposed to hemodynamic profiles of reactivity). Further, an increase on either CO or TPR alone does not tell us if that profile is myocardial or vascular; rather, it could reflect a mixed pattern of response (i.e., both variables are increasing). Only when both are examined in the one model can the true nature of the hemodynamic profile be revealed.

Hemodynamic profile and health outcomes: direct evidence

Why the hemodynamic profile of the stress response might be particularly important in the biology of stress is that it may hold stronger explanatory power when tracking the impact of psychosocial moderators of the stress response on disease processes. A vascular pattern of responding may be particularly important in tracking the development and maintenance of essential hypertension. Again, this is not a new thought or observation; but within the reactivity literature, this observation is rarely made explicit today. The notion that a vascular profile of reactivity may be particularly predictive of the development of hypertension specifically arises from early research on essential hypertension; in almost all cases of essential hypertension (that is hypertension with no underlying cause), elevated blood pressure is characterised by elevated TPR (Beevers et al., 2001).

However, perhaps more convincing is Julius and colleagues’ work on the ‘hyperkinetic’ model of hypertension (Julius, 1988; Julius et al., 1971; Julius & Nesbitt, 1996). Across over three decades of research, Julius and colleagues identified a number of important features distinguishing young people with borderline hypertension from those with established hypertension. In the early stages of hypertension and in those with borderline hypertension, elevated blood pressure is characterised by elevated CO (e.g., Eich et al., 1962; Eich et al., 1996; Lund-Johansen, 1983). However, as borderline hypertension is sustained and develops into established hypertension, the hemodynamic profile of the elevated blood pressure is altered; established hypertension is now characterised by elevated TPR. This suggests that vascular patterns of reactivity to stress may be the most damaging pattern in terms of hypertension risk.

A further illustration that vascular patterns of reactivity, in particular, may be most damaging arises from the observation that elevated blood pressure driven by elevated TPR rather than CO has been linked to increased cardiac events and mortality (Julius, 1988; Mayet & Hughes, 2003; Sung et al., 1993). Therefore, it may be a powerful risk marker for the development of hypertension if young adults, free from existing disease display a vascular profile of reactivity in response to psychological stress. Although proposed in the past, it is surprising that there exist no studies with a prospective research design assessing if vascular patterns of reactivity predict the development of hypertension. Of course, perhaps such studies have been confined to the file drawer due to null effects. Which raises the first question to be highlighted in this paper; does there exist data where vascular profiles of stress responsivity can be examined in order to assess if they prospectively predict future disease outcomes?

Matthews et al. (2003) appear to be the only study to attempt such a feat. They followed 100 children, initially aged 8–10 years old, and found that TPR reactivity to a range of stressors did not predict future blood pressure levels three years later. This could of course be because TPR reactivity is not a preclinical marker of hypertension risk. Alternatively, given the young age of the sample and the short follow-up period, perhaps not enough time had elapsed to track meaningful changes on blood pressure in such a young sample. In addition, it is perhaps elevated TPR in the absence of a significant decrease in CO (i.e., a vascular profile rather than TPR reactivity) that is damaging, a feature not identifiable in Matthews et al. data who did not examine the hemodynamic profile as a health-relevant predictor. Apart from this one study, it appears that we have no data with a prospective follow-up of participants, either young or old, tracking the predictive validity of myocardial, vascular, or mixed patterns of reactivity on future disease endpoints.

One study which may give some sign that vascular profiles of reactivity are predictive of hypertension development is Brindle et al. (2016). In a middle-aged sample of over 600 individuals, despite equivalent (exaggerated) blood pressure responses to the psychological stressor at baseline, it was only those who also showed moderate heart rate responses (compared to those who showed elevated or blunted heart rate responses) that were at greater risk of hypertension development five years later. This cluster, identified by Brindle et al., given it was characterised by a moderate heart rate response paired with an exaggerated blood pressures response, suggests a vascular hemodynamic profile of reactivity, identifying that in the case of hypertension development, this profile of reactivity is more damaging. However, this is speculation, as Brindle et al. did not have data examining the specific hemodynamic profile of the stress response in this sample. Unpublished data from our lab, however, confirms that this cluster of reactivity in young healthy adults (high blood pressure responses paired with a moderate heart rate response), is characterised by a vascular profile of response to psychological stress. Prospective studies are needed, however, to establish if vascular patterns of reactivity in young, healthy, adults are predictive of future hypertension risk.

Hemodynamic profile and health outcomes: indirect evidence

Indirect strands of research that suggest that vascular profiles of reactivity and mechanisms as being particularly health-damaging span the medical literature but rarely overlap with the psychological literature. For example, patients with evidence of ischaemia and no obstructive narrowing of the coronary arteries (INOCA) have adverse cardiovascular morbidity and mortality outcomes. While the pathophysiology of INOCA is not clear, abnormal coronary vascular reactivity has been implicated (Pasupathy et al., 2016).

Likewise, arterial stiffness, which refers to the relative rigidity of the arteries, is associated with both hypertension and cardiovascular disease in adults (e.g., Hansen, Staessen, et al., 2006); in fact, it appears to precede the development of hypertension (Dernellis & Panaretou, 2005) and is a key biomarker of vascular health (Segers et al., 2020). Given that increased arterial stiffness will elevate vascular resistance, this points towards arterial stiffness measurement as an exciting and well-justified parameter that has shown strong predictive utility over and above blood pressure on cardiovascular outcomes. However, despite its emerging and exciting role in cardiovascular disease identification of at-risk populations, it has yet to be incorporated in the cardiovascular reactivity literature and offers an exciting area of future research in cardiovascular reactivity to stress and links with future disease. Likewise, non-invasive measurement of diastolic dysfunction may also be particularly revealing in developing the reactivity hypothesis with a focus on the mediators of hemodynamic profile as a potential health-relevant measure in healthy samples.

Another indirect strand that straddles both medicine and psychology, relates to patient samples and mental stress-induced myocardial ischaemia (MSIMI). Myocardial ischaemia occurs when myocardial oxygen demand exceeds supply; MSIMI is when myocardial ischaemia occurs while under mental stress. In the clinic or research lab, it is assessed through a variety of methods, including electrocardiogram, echocardiography, radionuclide ventriculography, positron emission tomography, and quantitate coronary angiography (Zhang et al., 2020). Research has shown that in patients with coronary artery disease, MSIMI is a predictor of poor prognosis and mortality (e.g., Sheps et al., 2002) and so there exists interest in MSIMI as a metric that may reveal an individual’s vulnerability to psychological stressors (Wokhlu & Pepine, 2016). It appears that a pattern of vascular reactivity underlies MSIMI (Goldberg et al., 1996; Jain et al., 1998), which is different from that which underlies exercise-induced myocardial ischaemia (Strike & Steptoe, 2003). In fact, recent research has identified that women with MSIMI show vascular responses to mental stress rather than an increased hemodynamic workload that is usually associated with MSIMI in men (Sullivan et al., 2018; Vaccarino et al., 2018). This indirect evidence highlights vascular reactivity as being potentially health damaging, particularly in women who are more likely to suffer from MSIMI than men (van Loo et al., 2014). Indirect evidence linking CO with negative outcomes arises from Jefferson et al.’s (2010) examination of the Framingham data, identifying that cardiac index (CO divided by body surface area) was associated with accelerated brain aging. Together, these studies all indirectly identify that differentiated patterns of hemodynamic response are linked to disease outcomes, with vascular mechanisms in particular, indicative of negative health outcomes and future disease states.

Hemodynamic profile as a situational or individual response stereotypy?

Over the years, examination of hemodynamic profile as task-specific or an individual response stereotypy has been the subject of some investigation. It appears that tasks characterised by different coping demands placed on the person are associated with different hemodynamic response profiles. Psychological stressors that involve participant engagement and control over the outcome are often considered ‘active’ tasks and are associated with myocardial hemodynamic profiles (Obrist, 1981). Such psychological challenges, including mental arithmetic, speech, reaction time tasks, or any other cognitive task where the participant can actively control the outcome, are assumed to be elicited by central command centres (e.g., frontal lobe, hypothalamic, and brainstem influences), preparing motor output systems in preparation for fight and flight (see Kamarck & Lovallo, 2003). These tasks have been shown to have consistent associations with increases in CO rather than TPR (Kasprowicz et al., 1990). For example, Sherwood et al. (1990) requested male participants to complete a series of tasks that required both active engagement (competitive reaction time) and passive coping (viewing of film clips). They demonstrated that the tasks requiring ‘active’ coping were characterised by increases on CO paired with small decreases in TPR.

Passive tasks, meanwhile, are those tasks where participants cannot influence the outcome; they must be endured rather than performed. Passive tasks used throughout the literature have included viewing negative emotional images or videos, mirror tracing, and the cold pressor test. The cold pressor test is one of the original ‘passive’ stressors, first described by Hines and Brown (1932) as a stimulus for measuring vasomotor reactions (for a brief history of the cold pressor test and its use, see Lamotte et al., 2021). However, its use as a passive psychological stressor is problematic; as well as requiring passive psychological endurance, it also contains a large physical element whereby the cold temperature leads to vasoconstriction, which in turn raises TPR. In fact, the effects of the cold pressor test on coronary vasculature are complex and are not fully understood (Lamotte et al., 2021). The cold pressor test, although established as predictive of cardiovascular disease progression (Pouwels et al., 2019), is not a true psychological stressor, where activation of the autonomic system occurs due to forebrain processing and appraisal of the stressor. However, other passive tasks have been developed that ensure this physical element is absent, including viewing negative emotional images (e.g., Kim & Hamann, 2012), audiovisual clips (e.g., Nyklícek et al., 2005), watching another participant engage in a speech task (Park et al., in press), or watching oneself engage in a speech task on playback (e.g., Griffin & Howard, 2020; Soye & O'Súilleabháin, 2019). These passive tasks have shown associations with vascular patterns of reactivity or at the very least, a mixed pattern of response; stressors that mimic the cold pressor test in physiological reaction, but do not contain physiological confounds of the vasoconstriction induced by the cold temperature.

Others, however, have proposed that individuals are themselves particular types of reactors and this individual difference variable is what predicts disease development. Physiological response specificity (Engel, 1972), that individuals will show a similar response to a range of stimuli, has been demonstrated in a number of studies with regards to hemodynamic profile. Kasprowicz et al. (1990) examined the pattern of hemodynamic response to three different types of tasks in 48 men; mental arithmetic, mirror tracing, and bicycle ergometry. They characterised participants into myocardial, vascular, or mixed reactors in response to the mental arithmetic task. They found substantial generalisation of the categories to the different psychological tasks, evidence of physiological response specificity. This was not evident in response to the physical exercise task, suggesting that hemodynamic profile in response to psychological stress alone is an individual difference variable. However, they also found that certain tasks were associated with particular patterns of reactivity; providing further evidence that the hemodynamic response profile is an individual difference variable, but is affected by task-specific variables. Other work has confirmed that the hemodynamic basis of reactivity is an individual characteristic with task demands only partially explaining differing hemodynamic profiles of the stress response (e.g., Lawler et al., 1995; Sherwood et al., 1990)

A further theoretical framework, the biopsychosocial (BPS) model of challenge and threat (Blascovich & Tomaka, 1996; Seery, 2013; drawing on the work from Dienstbier, 1989), offers further food-for-thought in the consideration of the hemodynamic profile of the stress response, particularly in relation to whether it represents a situational or individual response stereotypy. According to this highly influential model, stress provokes a challenge or a threat response in the individual, resulting from the individual’s evaluation of the situational demands and their personal resources to meet these demands. Where a person evaluates they have the resources to meet the demands of the situation, they exhibit a challenge response; where they evaluate they do not have the resources to meet demands, they exhibit a threat response. This challenge or threat response can be identified by the nature of movement on CO and TPR, with challenge identified by predominantly CO reactivity and threat by TPR reactivity. While originally identified as relevant to motivated performance situations (Blascovich & Tomaka, 1996; Hase et al., 2020; Seery, 2013), further work showed that situations do not necessarily have to be performance-based, with interactions with a stranger shown to engender differential response patterns (e.g., Blascovich et al., 2001; Mendes et al., 2007; Mendes et al., 2008). What this model aimed to identify was the psychological meaning of physiological responses (Blascovich & Kelsey, 1990), which differs from the reactivity hypothesis in the behavioural medicine literature, which aimed to outline the health consequences of physiological reactions to stress.

The BPS model of challenge and threat and the reactivity hypothesis grew organically, but separately from each other. It is now timely to expand both areas of research with due regard for the theoretical and practical implications of both expansive and productive literatures. Locating the BPS model within the hemodynamic profile literature it might be argued that, on its simplest level, a challenge response to an active stressor is the expected and healthy response; a threat state is potentially more damaging as it reflects a dysregulated stress response and reveals internal stress where the person has evaluated that they do not have the resources to meet the demands of the stressor. This reflects the hypothesised effect of a vascular versus a myocardial hemodynamic profile in the face of active stress, with a vascular profile of reactivity potentially predictive of future hypertension risk.

Measurement of hemodynamic profile

Early investigations into hemodynamic profiles relied on categorical identification of myocardial reactors, vascular reactors, or mixed (aka dual) reactors (e.g., Kasprowicz et al., 1990; Lawler et al., 2001). This usually involved identifying individuals as myocardial, vascular, or mixed reactors based on their CO or TPR reactivity in response to an active task; usually a mental arithmetic task. If changes were driven by CO, these individuals were categorised as myocardial reactors; if changes were driven by TPR, these individuals were categorised as vascular reactors; and finally, if changes were driven by largely equivalent increases on both CO and TPR, these reactors were termed ‘mixed’. However, the quantification of what constituted a ‘change’ varied across studies, with residualised change scores and responses outside of one standard deviation from the sample mean (Kasprowicz et al., 1990; Lawler et al., 1995; Lawler et al., 2001) or reactions in the upper tertiles of the distribution (Sherwood et al., 1990) used to categorise individuals as myocardial, vascular, or mixed reactors. Using this categorisation, early research confirmed that the hemodynamic pattern of the stress response was reflective of particular tasks as well as representing an individual difference variable (e.g., Kasprowicz et al., 1990). However, the research was limited in that the criteria used often only allowed categorisation of a subset of the sample and indeed forced discontinuity on what is essentially continuous measurement, resulting in a lack of power in statistical analyses.

In an attempt to address the limitations of this categorisation of a continuous variable, Gregg et al. (2002) proposed a new model of blood pressure regulation; the hemodynamic profile – compensation deficit model of blood pressure regulation (HP/CD model). Given that mean arterial pressure is the product of CO and TPR, a log transform of CO and TPR allows these to be expressed as an additive function to represent mean arterial pressure. Then, taking the log of the ratio of the task to baseline values for both CO and TPR (log CO_r and log TPR_r), mean arterial pressure reactivity can be expressed as an additive function (log CO_r + log TPR_r). If the two-dimensional space represented by logCO and logTPR is rotated 45 degrees, the orthogonal relationship between hemodynamic profile and compensation deficit is formed. This model then allows a description of both the magnitude of the blood pressure response, identified by compensation deficit (CD; which will indicate the degree to which CO compensated for an increase in TPR), as well as hemodynamic profile (HP; whether CO or TPR dominated the change in blood pressure). (For a full description of the model and early studies with its use, see; Gregg et al., 2002; James et al., 2012).

The HP/CD model has been used in a handful of studies investigating various psychosocial moderators of the cardiovascular stress response as well as indicators of cardiovascular disease risk. Gregg et al. (2005) identified in a sample of 30 young men that a mixed or vascular hemodynamic profile was associated with elevated ambulatory 24 h ambulatory pulse pressure. Ottaviani et al. (2006) also reported that HP in response to a passive task (mirror tracing) was associated with ambulatory blood pressure in a mixed-gender sample of participants with a mean age of approximately 33 years. Together these studies identified that a vascular or mixed hemodynamic response to stress in the lab is associated with elevated ambulatory blood pressure over 24 h, a known marker of cardiovascular disease risk (Hansen, Jeppesen, et al., 2006). In a further use of the model, Carnevali et al. (2019) compared blood pressure responses to a physical task (orthostasis) and a emotional task (anger recall), in normotensive African American and European American college students. While blood pressure stress reactivity did not differ between the groups, African Americans showed a prominent vascular response to both the physical and emotional tasks, whereas European Americans showed no hemodynamic response to the physical task and a mixed profile in response to the anger recall task. These findings further illustrate that equivalent blood pressure response to stress-tasks can be characterised by vastly different hemodynamic profiles.

With regards to psychosocial moderators of hemodynamic profile, both neuroticism (Hughes et al., 2011) and Type D personality (Howard et al., 2011) are associated with weaker myocardial reactions to an active task as identified by the HP/CD model, while perseverative cognition has been associated with a vascular profile of reaction (Ottaviani et al., 2017). Age has been shown to be associated with vascular profile to a cognitive recall task (Hogan et al., 2012), while a behavioural index of repressive behaviour showed no associations with differing hemodynamic profiles (Howard et al., 2017). Together these studies point towards hemodynamic profile of the stress response as being particularly elucidating when examining potential moderators of the cardiovascular stress response. In fact, an interesting feature reported by Gregg et al. (1999) identified that regardless of whether a task was active or passive, recovery from such a task was markedly vascular in nature. Ottaviani et al. (2007) confirmed this observation, showing that the hemodynamic profile of recovery patterns from both active and passive tasks are characterised by a vascular hemodynamic profile. In addition, they showed that this vascular profile of recovery was associated with both higher ambulatory blood pressure and higher levels of soluble intercellular adhesion molecule-1, an inflammatory marker associated with heart disease.

However, perhaps the most interesting findings arising from the HP/CD model in cardiovascular reactivity research relate to initial studies on the impact of caffeine and sleep restriction on stress reactivity using the HP/CD model. The proposers of the model originally pursued research into the impact of caffeine on stress reactivity and confirmed previous research that caffeine intake resulted in increases in TPR (James & Gregg, 2004), even in adolescents (James et al., 2018). More interesting though, was the finding that acute sleep restriction, despite having no impact on blood pressure reactivity, was characterised by a markedly vascular hemodynamic profile (James & Gregg, 2004). This vascular profile in response to sleep restriction was later confirmed in an independent sample (O’Leary et al., 2013). This is particularly interesting given that short sleep duration is associated with an increased risk of cardiovascular diseases (Hoevenaar-Blom et al., 2011) and mortality (Kojima et al., 2000).

But the HP/CD model is not without its critics, and indeed, has never been adopted as a core measurement within reactivity hypothesis. Despite the statistical benefit in being able to represent both CO and TPR as a unitary construct, as opposed to the categorical approaches used in the past, use of this model has mainly arisen from two independent research labs; those originating from the original proposer of the model and their collaborators (e.g., Howard et al., 2011; Hughes et al., 2011); and those arising from research by Ottaviani and colleagues (e.g., Ottaviani et al., 2006; Ottaviani et al., 2017). In fact, two separate papers have directly targeted the HP/CD model of blood pressure regulation with criticisms (Sawada, 2006; Why & Chen, 2013), with one proposing an alternative model (Sawada, 2006) and the other highlighting that baseline differences may be overly influential on its computation (Why & Chen, 2013); although this latter critique only rings true if multiple tasks with differing baselines are used in the calculation of the HP of each task. One aspect of the HP/CD model sometimes under question is whether HP and CD are truly statistically independent, with some studies showing a lack of orthogonality (e.g., Ottaviani et al., 2006).

With regards to the calculation of HP specifically, there is substantial statistical overlap between the HP/CD model of blood pressure regulation and the challenge-threat index used sometimes within the BPS model. In fact, recent research has identified that on a numeric level, the HP index from the HP/CD model and the challenge-threat index used in the BPS model (Blascovich et al., 2004) are mathematically measuring the same construct (Griffin & Howard, in press). In the BPS model, a challenge-threat index is computed by standardising the CO and TPR reactivity scores and then subtracting the standardised TPR reactivity score from the standardised CO reactivity score (e.g., Blascovich et al., 2004; Seery et al., 2004). In this computation, higher scores indicate greater challenge. This challenge-threat index, directly, but inversely, mirrors that of the HP score from the HP/CD model where a log transform of the reactivity scores for TPR and CO are computed and these log-transformed scores are subtracted (logTPRr – logCOr) to reflect hemodynamic profile (e.g., Gregg et al., 2002; James et al., 2012). Greater (negative) scores on this index represent a myocardial profile. Unsurprisingly, recent work has shown that these two constructs share considerable variance, with correlations between the challenge-threat index and HP of -.96 and above reported in a sample of young adults in response to a speech task (Griffin & Howard, in press).

However, theoretically speaking, these two overlapping constructs are proposed to measure different underlying continuums; with the HP/CD model measuring a type of response from myocardial to a vascular response, while the challenge-threat index identifies the psychological meaning behind the physiological response; a challenge (myocardial) or threat (vascular) response. While the vascular response appears to reflect passive coping in the health psychology literature, within the BPS domain, both the challenge and threat response are relevant to the motivated performance situations (i.e., active tasks). This overlap in statistical computation, but departure on the meaning of the underlying constructs, has the potential to further research in both literatures. By adopting both the use of physiological responses to stress as reflective of the underlying meaning of the psychological experience (as proposed in the BPS model) and by acknowledging and exploring the health-relevant impact of differentiated hemodynamic responses to stress (the core of the reactivity hypothesis), significant advancements could be made in our understanding of the biology of the psychological stress response and its relevance for future health.

Recent moves to data-driven approaches to hemodynamic profiling of the cardiovascular stress response have also identified some interesting avenues for future research that may help to marry and extend these two areas of research. As described earlier, Brindle et al. (2016) showed a cluster of blood pressure and heart rate reactivity that implied more vascular responding was associated with hypertension risk on five-year follow-up. More recently, Wormwood et al. (2019) identified two clusters in two large independent datasets in response to an active task, with one cluster identified by an increase in CO paired with decreases in TPR, pre-ejection period, and inter-beat interval (i.e., myocardial reactors) and the other identified by more modest changes, akin to a threat response (that is, a more vascular response). These data-driven approaches may offer strengths in identifying the disease risk associated with different hemodynamic profiles of the stress response, mapping onto differences in psychological experience (e.g., challenge versus threat). It is certainly not a new idea that cardiovascular reactivity to stress is a multivariate structure, but it is most commonly examined as a univariate construct. Data-driven approaches and examination of the hemodynamic profile of the stress response acknowledge the multivariate construct that is cardiovascular reactivity to stress.

Another interesting outcome of Wormwood et al.’s (2019) data showed that, when a three-cluster solution was allowed, a clear pattern of non-responders emerged. This is interesting in the context of the emerging body of literature identifying the ‘blunted’ cardiovascular response profile as an important predictor of future negative events (see Whittaker et al., 2021 and Phillips et al., 2013, for discussions on blunted cardiovascular reactivity), but also further highlights the need to adopt multivariate approaches in reactivity measurement.

How do different hemodynamic profiles of stress response promote or progress disease?

The potential mechanisms and pathways to cardiovascular disease within the reactivity hypothesis have previously been excellently laid out by Lovallo and Gerin (2003) and Lovallo (2005), involving three different systemic levels (exaggerated cognitive-emotional response, heightened hypothalamic and brainstem responsivity, and peripherally altered tissues). However, it still remains to be established whether these differential patterns are the mechanism linking psychological states to disease, part of the pathway, or are indirectly related to disease outcomes by altering health behaviours (see Cross et al., 2020). Myocardial reactions, where individuals typically respond with high CO that is not compensated by decreased vascular tone may be at risk of cardiovascular disease due to tissue overperfusion and endothelial damage due to shear stress. Vascular reactors then, where elevated TPR is not adequately compensated by decreased myocardial activity might be at risk of suffering impaired vascular contractility due to repeated ongoing episodes of increased vascular resistance. In fact, the vascular profile of reactivity associated with the cold pressor test, which shows reliable associations with cardiovascular health, has been shown to influence downstream metabolic vasodilatation, which impacts on endothelial function, release of hormones that active β2-adrenergic receptors located on the coronary vascular smooth muscle cells, as well as stimulating nitric oxide synthesis via stimulation of α2-adrenergic receptors located on the coronary artery endothelial cells (for a review see Pouwels et al., 2019). These downstream effects, on inflammatory factors, vascular tone, and vascular blood flow highlight why vascular profiles of reactivity to psychological stress may be of particular interest in the prediction of disease development in healthy populations.

Hemodynamic profile and stress tolerance

One feature of reactivity identified by examination of hemodynamic profile not often acknowledged is that patterns of responding become more vascular as the duration of the stressor continues (Carroll et al., 1990). This was demonstrated clearly by Ring et al. (2002), who showed in a 28-min laboratory protocol comprising of baseline, task and recovery phases, mean arterial pressure showed a sustained increase, with CO increasing early in task but returning to baseline by the end of the task. Conversely, TPR continued to increase as the task progressed. This is interesting as it suggests that while participants may show the expected myocardial reaction to an active stressor, as the stressor continues, the pattern of responding will swing from myocardial to vascular patterns of responding; essentially a swing in active coping to passive coping occurs.

Within the BPS model, this dynamic change in reactivity would look like a swing from a challenge to a threat response. The magnitude of this swing, and whether it might reveal important facets of how stress tolerance influences disease processes, has not yet been examined in any reactivity research. This is unfortunate as it may represent a bridge between the behavioural medicine/health psychology research where the reactivity hypothesis arose and the social and personality psychology literature which is the home of the BPS model of challenge and threat.

While measurement of reactivity in the health psychology literature was often on a 2-min or more ‘bin’, within the emotion and social literatures, physiologic response is often in a 10-s or 15-s bins. Of course, in health psychology, sometimes other physiological parameters such as hormones or immune responses are also measured, necessitating longer measurement periods. These temporal differences in measurement time periods between studies arising from the different literatures could in themselves offer fruitful extensions to the health literature. For example, the nature of the swing, or how long it takes the response to move from active to passive (or from challenge to threat), could have important health-relevant implications and may reveal further insight into the nature of the psychological experience in ongoing or chronic stress.

There exists a paradigm within which this notion could easily be investigated and together would advance what we know about the impact of both stress responsivity and stress tolerance on health. Patterns of cardiovascular stress response habituation-sensitization have previously been identified as holding much promise in the elucidation of how psychosocial moderators influence stress responding and therefore disease processes. Work began by Kelsey and colleagues in the 1990s (e.g., Kelsey, 1993; Kelsey et al., 2000) and later fine-tuned and developed by Hughes and colleagues (e.g., Hughes et al., 2018), examine how cardiovascular stress response adaptation patterns are influenced by various environmental and psychosocial factors. Kelsey et al. (2004), for example, demonstrated that repeated exposure to an active stressor attenuated cardiac responding, with evaluative observation disrupting this habituation. Other work has shown various moderators of cardiovascular stress response adaptation patterns, such as personality (Howard & Hughes, 2013; Hughes, 2007a; Hughes et al., 2011; Lü et al., 2016; Lü et al., 2018), smoking (Hughes & Higgins, 2010), social support (Howard & Hughes, 2012; Hughes, 2007b), and performance feedback (Brown & Creaven, 2017).

What this theoretical and data-driven development of the reactivity hypothesis adds to the literature is a focused examination of how individual stress tolerance may be reflective of disease processes and factors that may moderate such processes. Such paradigms usually look at how individuals show cardiovascular stress response habituation (or lack thereof), by examining patterns of response adaptation either within stressors (e.g., O’Súilleabháin et al., 2018), between repeated exposures to the same stressor within the same session (e.g., Howard & Hughes, 2013), or between repeated exposures to the same stressor between sessions (e.g., Johnson et al., 2012). These studies have revealed important differences in cardiovascular stress response adaptation as a result of pertinent stress-related personality traits and health-relevant factors, particularly in relation to cardiovascular diseases. For example, rumination (Johnson et al., 2012), neuroticism (Hughes et al., 2011), and Type D personality (Howard & Hughes, 2013) have shown maladaptive patterns of cardiovascular stress response adaptation to recurrent stress, highlighting potential mechanisms by which these traits are associated with increased disease risk.

Examination of the hemodynamic profile of these patterns of cardiovascular stress response adaptation may further extend the utility of the reactivity hypothesis by revealing important disease processes associated with stress tolerance in the individual. While previous research suggests that most participants will swing from a myocardial profile of response to a vascular profile of response (Carroll et al., 1990), such studies have mainly focused on the examination of just cardiac or vascular responding. Such research has not focused on the dynamic relationship between CO and TPR and whether a swing occurs from myocardial profiles to vascular profiles, rather than just an attenuation of cardiac in favour of an exaggeration in TPR responding. In fact, such physiological response patterns may indicate if certain individuals are more likely to swing from active coping to passive coping when exposed to ongoing, or chronic, stress. Indeed, such analyses would also suggest that initial challenge responses within the BPS model become threat-like as the duration of the stressor continues or as repeated exposure to the stressor occurs. Do initial challenge-like responses move to a more threat-based experience as the duration of the stressor continues, not only due to temporal differences in how vascular and cardiac variables respond but also because the individual revaluates their resources as not being able to meet demands? It is here, the interchange between the psychological meaning of the stress response and health-relevant physiological outcomes offered by both models that have the potential to significantly expand what we know about the biology of the stress experience.

As such, the hemodynamic profile of the stress response, as well as the profile of patterns of stress response adaptation, is particularly relevant when tracking the impact of stress tolerance on disease development. It may be the case that patterns of healthy stress response adaptation (i.e., habituation) are underpinned by a smaller swing from myocardial to vascular in the face of active stress, thereby showing a pattern of a healthy response to recurrent, ongoing, or chronic stress. Furthermore, this would also place active response profiles (or challenge-like in the BPS model) as the healthy, expected cardiovascular response to active tasks. Deviations from this expected response to initial active stress might signal dysfunction in the regulatory mechanisms of blood pressure control, with more vascular profiles of reactivity an early indicator of something amiss in the stress response of the healthy individual. This could therefore mark a potential early identifier of future disease risk, particularly in the case of hypertension. Further, patterns of stress response adaptation would highlight whether such dysregulated stress responses are maintained in terms of stress tolerance and ongoing stress reactions.

Conclusions

While the predictive utility of the hemodynamic profile of the stress response has yet to be prospectively established, there appears much promise in existing research that it holds some explanatory power in mediating the link between cardiovascular reactivity to stress and the development of disease. This is revealed when examining the different response profiles associated with various psychosocial factors, particularly stress-relevant personality traits such as neuroticism. Therefore, one aim of the present paper is to seek researchers to examine data they may have in existing prospective research designs that measured the hemodynamic profile of the stress response in healthy, disease-free adults, in order to see if hemodynamic profiles of the stress response predict health-relevant outcomes. Disease endpoints such as hypertension development would be particularly welcome here. While there is some evidence to suggest healthy individuals with a vascular hemodynamic profile show elevated ambulatory blood pressure, these were cross-sectional in design and therefore lacking in conviction that hemodynamic profile is an important mechanism through which reactivity is a contributory factor in the development of cardiovascular disease. Therefore, the first question that remains to be answered and should motivate future research is; are vascular patterns of reactivity predictive of future cardiovascular disease risk?

The adoption of the BPS model within investigations within the reactivity paradigm, with a particular focus on the hemodynamic profile, may offer significant developments on what we know about how the psychological experience of stress is associated with the physiological response to stress and potential consequent disease initiation or promotion. Adoption of the hemodynamic profile of the stress response, either as an indicator of psychological experience or health-informative response profile, could help further both overlapping, but as yet, independent areas of research.

Further development of the reactivity paradigm by the introduction of stress phases to detect patterns of cardiovascular stress response habituation-sensitization (see Hughes et al., 2011 and subsequently Manigault et al., 2021), may also be a particularly fruitful extension of both the reactivity hypothesis and the BPS model of challenge and threat. In collaboration with the examination of the hemodynamic profile, physiological indices of stress tolerance (i.e., habituation patterns) may be particularly revealing on detailing the health-relevant impact of personality and social variables on stress-relevant disease processes, as well as how the psychological experience of stress may change (moving from challenge to threat) as the duration of a stressor continues. This may have important implications for literatures examining both acute and chronic stress and its impact on health.

Therefore, the purpose of this paper is to highlight and review what are essentially old ideas, unfinished in their exploration of the health consequence of the biology of the stress response. There may be existing data in which we can examine these questions; or alternatively, future research may wish to include such guiding hypotheses in their design. The relevance of hemodynamic profile as a health-relevant consequence of the biological stress response, as well as the potential moderators of it, may highlight in young healthy individuals those most at risk for hypertension development. If upheld, the hemodynamic profile of the stress response would act as a powerful preclinical marker of those individuals most in need of health promotion interventions to reduce other, more amenable, risk factors for disease.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.