The lifetime prevalence of cardiovascular disease (CVD) in adults is >50% and has remained the leading cause of death in the United States for a century (1). Whereas a large emphasis has been placed on traditional risk factors for CVD, there is a growing understanding that psychosocial factors such as psychological stress also contribute independently to CVD risk (2–5). For example, a case-control study that included nearly 25,000 adults from 52 countries reported that individuals who experienced an acute myocardial infarction were 1.5 times more likely to report two or more stressful life events related to work, home, family, or finances in the year prior (6). In support of these findings, evidence from a large, prospective cohort study in 118,706 adults living in 21 different low-, middle-, and high-income countries demonstrated that psychosocial stress at work or home, from major life events, or due to finances contributed to a 20–30% increase in CVD, coronary heart disease, and stroke risk even with adjustment for traditional social, behavioral, and clinical risk factors (7). Thus, compelling epidemiological evidence exists to suggest that psychosocial stress exposure independently contributes to CVD risk, and that the magnitude of risk is comparable to that attributed to traditional CVD risk factors (7, 8). There has been an alarming trend of increasing rates of CVD among young adults (9, 10). Given that young adults are currently reporting higher stress levels than older generations (11) and that young adulthood is a period of life where individuals are at heightened risk to develop stress disorders (12, 13), it is plausible that psychosocial stress may be contributing to these increases in CVD risk.

The acute neural response to psychosocial stressors includes activation of the limbic system and subsequently of the hypothalamic-pituitary-adrenal and sympathetic-adrenal-medullary axes (4, 14). The downstream physiological effects include increased sympathetic nervous system activity and parasympathetic withdrawal, the secretion of stress-related hormones, immune system activation and the production of inflammatory cytokines and prothrombotic factors, and an increase in peripheral resistance (4, 15). Collectively, this response also contributes to transient reductions in endothelial vasodilatory function (16–18). Whereas this acute physiologic response is protective in emergent situations, it is thought that chronic, repeated exposure to psychosocial stressors may result in cumulative, maladaptive effects that include detrimental vascular endothelial sequalae, the promotion of atherosclerotic progression, and ultimately increased CVD risk (16, 19, 20).

The vascular endothelium is a regulatory organ that plays a critical role in maintaining cardiovascular homeostasis (21). Consequently, vascular endothelial dysfunction is a pathophysiological mechanism contributing to the development of atherosclerosis and CVD (21, 22), and is characterized by reduced endothelium-dependent vasodilatory function in conduit arteries. Recent reports have indicated that psychosocial stress is associated with reduced vascular endothelial function (VEF) among young adults (19, 20, 23), perhaps implicating vascular endothelial dysfunction as a pathophysiological link between psychosocial stress exposure and the development of CVD.

In examining the health-related effects of psychosocial stress, two prevailing models include the differential exposure and differential vulnerability hypotheses (24, 25). The differential exposure hypothesis proposes that differential exposure to psychosocial stressors explain group differences in health, including cardiovascular risk (26, 27). The differential vulnerability hypothesis posits that stress appraisal dictates the effect of stress exposure on health, whereby individuals who perceive stressors as more stressful (i.e., greater stress sensitization) are more strongly impacted (25, 28–30). Notably, while these hypotheses are often presented as competing models, they are not necessarily mutually exclusive. Accordingly, evidence from the Midlife in the United States (MIDUS) cohort study indicated that both the total number of naturalistic stressors experienced during eight days of daily life and the affective reactivity to these stressors were predictive of mortality over 20 years of subsequent follow-up, although the association of affective reactivity with mortality was only present among those with a chronic illness (31). Therefore, to better understand the effects of stress exposure versus stress appraisal on cardiovascular health in young adults, we examined the associations of chronic stress exposure, chronic stress appraisal, and proximal perceived stress with VEF measured by the brachial artery flow mediated dilation technique. Our hypothesis was that the degree of exposure to chronic stressors, rather than chronic stress appraisal or proximal perceived stress, would be more strongly and inversely associated with VEF in young adults. We further conceptualized that stress appraisal would moderate the association between stress exposure and VEF.

This is among the first studies to assess the effects of chronic, ongoing naturalistic stress on conduit artery VEF, inclusive of cumulative chronic stressor exposure and chronic and proximal stress appraisal in young adults. As hypothesized, those who reported greater cumulative exposure to chronic stressors also exhibited reduced VEF. Interestingly, the present findings showed that neither appraisal of chronic stressors, nor perceived stress in the past month were significantly associated with VEF in our sample. While chronic stress appraisal did not modify the relation between chronic stress exposure and VEF, proximal perceived stress did. Specifically, the relation between chronic stress exposure and VEF was significant at low and average levels of proximal perceived stress, whereas there was no relation at high levels of proximal perceived stress. We then determined that the association of chronic stress exposure on VEF was significant only at PSS scores less than 22. Notably, individuals with PSS scores ≥22 had greater exposure to factors that may sensitize individuals to stress than individuals with PSS scores <22, including greater adverse childhood experiences and depressive symptoms, and lower psychological resilience. Thus, overall, our results demonstrated that exposure to common, chronic psychosocial stressors was most strongly predictive of VEF in young adults, and that this association is dependent upon proximal perceived stress.

Our findings indicate that even among healthy, young adults with no existing cardiovascular or metabolic disease, exposure to chronic stressors is associated with reduced VEF but not brachial blood pressure, which aligns with prior evidence linking chronic naturalistic stressor exposure to the development and progression of disease (48, 49) and portends risk for the future development of CVD. However, neither chronic stress appraisal nor proximal perceived stress were individually associated with VEF in our study. Similarly, Greaney and colleagues (50) reported that the number of exposures to common daily stressors is positively associated with microvascular vasoconstrictor responsiveness to noradrenergic stimulation in young adults. Also consistent with our findings, Greaney et al. (50) reported that perceived stressor severity was not significantly associated with vascular function. These findings also agree with data from the MIDUS study, which indicated that greater cumulative stressor exposure in mid-life was linked with increased mortality risk during 20 years of follow-up, whereas negative affective reactivity to naturalistic stressors only predicted increased mortality among individuals with a chronic health condition (31). Therefore, our findings lend support to reduced VEF as an important pathophysiological mechanism linking chronic stress exposure with future CVD and, potentially, mortality risk. These findings also lend credence to the differential exposure hypothesis, indicating that increasing exposure to chronic stressors is sufficient to promote reduced VEF.

Conceptually, exposure to and appraisal of stressors are not independent phenomena, and instead stressor appraisal may modify the degree to which stressor exposure negatively impacts health. Indeed, whereas repeated activation of the physiological stress systems would be expected to adversely impact cardiovascular health, it is reasonable to suspect that this may be further exacerbated by heightened reactivity among individuals who perceive the stressor(s) as more upsetting. Thus, our study takes an important step by building on prior studies to test the hypothesis that stress appraisal moderates the association of exposure to chronic naturalistic stressors with reduced VEF. In partial accordance with our hypothesis, proximal perceived stress was a significant moderator (Figure 1B), but appraisal of the chronic stressors themselves was not (Figure 1A). Interestingly, our data indicate that while stress exposure magnitude was significantly and inversely associated with VEF at low (−1SD) and moderate (average) levels of proximal perceived stress for our sample, it was not among individuals who reported high (+1SD) levels of proximal perceived stress. Examination of the simple slopes (Figure 2B) further suggests that individuals with high perceived stress have lower VEF even at low chronic stress exposure. Therefore, our findings indicate that the association of chronic stressor exposure with reduced VEF is not dependent on appraisal of the chronic stressors but is dependent on proximal perceived stress.

As stated, when individuals reported greater proximal perceived stress, VEF was less impacted by the magnitude of exposure to common, chronic psychosocial stressors. Although seemingly counterintuitive, these findings could be explained by increased stress sensitization among those with high levels of proximal perceived stress, such that individuals are more impacted by chronic stressors and perceive high levels of stress at low levels of chronic stress exposure (51). For example, for an individual who is sensitized to stress, low exposure to chronic stress will result in higher proximal perceived stress, whereas for someone who is less sensitized to stress, high exposure to chronic stress will be required to result in high proximal perceived stress (52). This model has been used previously to identify groups who may be more prone to experience psychopathology in response to psychosocial stress exposure (53). For example, individuals with a history of adverse childhood experiences have been shown to be sensitized to subsequent proximal stress, increasing their risk of later psychopathology (51). To further test this hypothesis, we performed the Johnson-Neyman technique to determine the level of proximal perceived stress at which the association of chronic stress exposure magnitude with VEF ceased to be significant (46). Our analysis indicated that the critical PSS score in our sample was approximately 22. Thus, we stratified our sample into two groups based on PSS scores below and equal to or above this critical value and performed comparisons examining factors that have been previously linked to stress sensitization or resilience (19, 39–41). Our findings confirmed that individuals with a PSS ≥22 indeed had greater exposure to adverse childhood experiences, greater depressive symptoms, and lower psychological resilience (Figure 3A–C).

While our findings appear to support an effect of stress sensitization in association with prior early life adversity, depressive symptoms, and low psychological resilience, there are several very important considerations that must be taken into account. First, there were no differences observed in chronic stress appraisal among individuals below and equal to the critical PSS score. However, stress sensitization may best be defined by the coupling of stress and affect (i.e., the affective response to stress exposure) (54) which was not assessed in this study, and thus the lack of difference in chronic stress appraisal also does not preclude an effect of stress sensitization. It is also important to consider that proximal stress appraisal as measured in this study was not directly resultant from the chronic stress exposures assessed and, when also considered against the potential temporal discordance of chronic stressor exposure and proximal perceived stress, it is difficult to ascertain with certainty the role of stress sensitization. However, it is also worth considering that the PSS has also been shown to reflect, in part, a stable characteristic describing how an individual perceives stressful life circumstances (38). Regardless, these findings raise several interesting hypotheses that will require future work with careful determination of temporality and more wholistic examination of stress sensitization, including assessments of affective responsiveness.

It should also be considered that our observance of differences in adverse childhood experiences, greater depressive symptoms, and lower psychological resilience are a statistical artifact as a result of expected linear relations among these variables. Thus, an alternative and simpler explanation is that these findings reflect interacting negative impacts of both proximal stress appraisal and chronic stress exposure on VEF, where high perceived stress proximal to the experimental visit superceded, and thus obscured, the effect of chronic stress exposure on VEF (55). Consequently, the effect of chronic stress exposure on VEF was apparent at low, but not high PSS scores. This finding has important implications for rigor, reproducibility, and study design, indicating that careful measurement of acute (proximal) stress is a necessary consideration when attempting to precisely delineate the effect(s) of chronic stress exposure on VEF. Finally, while it is beyond the scope of this study, it will be important for future studies to better understand how and if factors such as adverse childhood experience exposure, depression, and psychological coping interact in these associations and how this translates to cardiovascular risk, with the goal of identifying interventions that ultimately promote cardiovascular resilience among susceptible populations.

There are several strengths, limitations, and considerations to note when considering our findings. First, while our participant sample is representative of the population of the local region in which the data was collected, White/non-Hispanic adults are overrepresented in our sample relative to national ethno-racial demographics. Second, whereas chronic stress exposure and appraisal were the focus of this study, rumination and affective (emotional) responsiveness have been previously associated with reduced endothelial function, such as in those with major depressive disorder (19), and may not necessarily be dependent on stressor severity. Future studies should include measures of affective responsiveness to elucidate whether these modify the effect of stress exposure on VEF. Additionally, both chronic and proximal stress were assessed via retrospective self-report, which increases risk of response bias (i.e., artificially high levels of stress may have been reported if the day questions were answered was particularly stressful). To partially circumvent this possibility, stress was assessed via multiple modalities, such that the measurement of stress and subsequent conclusions were not entirely dependent on one self-report measure. Furthermore, the scales used to assess stress were implemented at least one week prior to the experimental visit where VEF was assessed to better isolate the effect of chronic versus acute stress. Future studies using a prospective design and daily or semi-daily (e.g., ecological momentary assessments) surveys of naturalistic stressors and stress appraisal would be advantageous. Of consideration, the range of reported perceived stress scores in our sample corresponds qualitatively with “moderate” (e.g., PSS scores of 14–26; n=60) and “high” (e.g., PSS scores of 27–40; n=5) perceived stress levels. Thus, in our moderation analysis, the association of chronic stress exposure with VEF, as measured by FMD, at low perceived stress (−1 SD) is in fact at moderate (~18) perceived stress. However, it is important to consider that these results are representative of young adults, and that this may in fact reflect typical stress levels in this population, particularly in contemporary society (56).

Given that young adulthood is a period of life where individuals are at heightened risk to develop stress disorders (12, 13) and the alarming recent increases in CVD risk among young adults (9, 10), our data provide important insight linking chronic stress with CVD risk among this population. More specifically, our findings indicate that the cumulative magnitude of chronic stress exposure is significantly associated with reduced VEF among otherwise healthy young adults. While chronic stressor severity and proximal perceived stress did not directly associate with VEF, proximal perceived stress did moderate the effect of chronic stress exposure on VEF. Therefore, our data highlight reduced VEF as an important pathophysiological mechanism linking chronic stress exposure with CVD risk in young adults, and indicate that the degree to which chronic stress is associated with reduced VEF is dependent on promixal stress appraisal. Accordingly, our findings emphasize that measures of proximal stress are necessary to fully understand the influence of chronic stress exposure on cardiovascular health. Our findings may also suggest that factors such as adverse childhood experiences, depression, and psychological resilience may play a role in stress-exposure related cardiovascular risk, although prospective longitudinal studies will be critical to better understand how these factors interact to ultimately promote cardiovascular risk or resilience.