**THE ROLE OF PHYSICAL FITNESS ON CARDIOVASCULAR RESPONSES TO STRESS THE ROLE OF PHYSICAL FITNESS ON CARDIOVASCULAR RESPONSES TO STRESS**

**Topic Editors Daniel A. Boullosa, Arto J. Hautala and Anthony S. Leicht Topic Editors Daniel A. Boullosa, J. Hautala and Anthony S. Leicht**

PHYSIOLOGY

#### *FRONTIERS COPYRIGHT STATEMENT*

© Copyright 2007-2015 Frontiers Media SA. All rights reserved.

All content included on this site, such as text, graphics, logos, button icons, images, video/audio clips, downloads, data compilations and software, is the property of or is licensed to Frontiers Media SA ("Frontiers") or its licensees and/or subcontractors. The copyright in the text of individual articles is the property of their respective authors, subject to a license granted to Frontiers.

The compilation of articles constituting this e-book, wherever published, as well as the compilation of all other content on this site, is the exclusive property of Frontiers. For the conditions for downloading and copying of e-books from Frontiers' website, please see the Terms for Website Use. If purchasing Frontiers e-books from other websites or sources, the conditions of the website concerned apply.

Images and graphics not forming part of user-contributed materials may not be downloaded or copied without permission.

Individual articles may be downloaded and reproduced in accordance with the principles of the CC-BY licence subject to any copyright or other notices. They may not be re-sold as an e-book.

As author or other contributor you grant a CC-BY licence to others to reproduce your articles, including any graphics and third-party materials supplied by you, in accordance with the Conditions for Website Use and subject to any copyright notices which you include in connection with your articles and materials.

All copyright, and all rights therein, are protected by national and international copyright laws.

The above represents a summary only. For the full conditions see the Conditions for Authors and the Conditions for Website Use.

**ISSN** 1664-8714 **ISBN** 978-2-88919-463-6 **DOI** 10.3389/978-2-88919-463-6

# *ABOUT FRONTIERS*

Frontiers is more than just an open-access publisher of scholarly articles: it is a pioneering approach to the world of academia, radically improving the way scholarly research is managed. The grand vision of Frontiers is a world where all people have an equal opportunity to seek, share and generate knowledge. Frontiers provides immediate and permanent online open access to all its publications, but this alone is not enough to realize our grand goals.

# *FRONTIERS JOURNAL SERIES*

The Frontiers Journal Series is a multi-tier and interdisciplinary set of open-access, online journals, promising a paradigm shift from the current review, selection and dissemination processes in academic publishing.

All Frontiers journals are driven by researchers for researchers; therefore, they constitute a service to the scholarly community. At the same time, the Frontiers Journal Series operates on a revolutionary invention, the tiered publishing system, initially addressing specific communities of scholars, and gradually climbing up to broader public understanding, thus serving the interests of the lay society, too.

# *DEDICATION TO QUALITY*

Each Frontiers article is a landmark of the highest quality, thanks to genuinely collaborative interactions between authors and review editors, who include some of the world's best academicians. Research must be certified by peers before entering a stream of knowledge that may eventually reach the public - and shape society; therefore, Frontiers only applies the most rigorous and unbiased reviews.

Frontiers revolutionizes research publishing by freely delivering the most outstanding research, evaluated with no bias from both the academic and social point of view.

By applying the most advanced information technologies, Frontiers is catapulting scholarly publishing into a new generation.

# *WHAT ARE FRONTIERS RESEARCH TOPICS?*

Frontiers Research Topics are very popular trademarks of the Frontiers Journals Series: they are collections of at least ten articles, all centered on a particular subject. With their unique mix of varied contributions from Original Research to Review Articles, Frontiers Research Topics unify the most influential researchers, the latest key findings and historical advances in a hot research area!

Find out more on how to host your own Frontiers Research Topic or contribute to one as an author by contacting the Frontiers Editorial Office: researchtopics@frontiersin.org

# **THE ROLE OF PHYSICAL FITNESS ON CARDIOVASCULAR RESPONSES TO STRESS**

Topic Editors:

**Daniel A. Boullosa,** Universidade Católica de Brasília, Brazil **Arto J. Hautala,** Verve Research, Finland **Anthony S. Leicht,** James Cook University, Australia

A greater fitness can help individuals to jump the hurdle of stress. Image created by Pilar Boullosa.

Cardiovascular responses to physical and/ or mental stressors has been a topic of great interest for some time. For example, significant changes of cardiovascular control and reactivity have been highlighted as important mechanisms for the protective effect of exercise as a simple and effective, non medical therapy for many pathologies. However, despite the great number of studies performed to date (e.g. >54,000 entries in Pubmed for "cardiovascular stress"), important questions of the role stress has on cardiovascular function still remain. For instance, What factors account for the different cardiovascular responses between mental and physical stressors? How do these different components of the cardiovascular system interact during stress? Which cardiovascular responses to stress are the most important for identifying normal, depressed, and enhanced cardiovascular function? Can these stress-induced responses assist with patient diagnosis and prognosis? What impact does physical

fitness have on the relationship between cardiovascular function and health? The current topic examined our current understanding of cardiovascular responses to stress and the significant role that physical fitness has on these responses for improved function and health. Manuscripts focusing on heart rate variability (HRV), heart rate recovery, and other novel cardiovascular assessments were especially encouraged.

# Table of Contents


Sarah K. Sasse, Tara J. Nyhuis, Cher V. Masini, Heidi E. W. Day and Serge Campeau

*Wheel Running in Rats*

# Introduction to the research topic: the role of physical fitness on cardiovascular responses to stress

#### *Daniel A. Boullosa1 \*, Arto J. Hautala2 and Anthony S. Leicht <sup>3</sup>*

*<sup>1</sup> Post-Graduate Program in Physical Education, Catholic University of Brasilia, Águas Claras, Brazil*

*<sup>2</sup> Department of Exercise and Medical Physiology, Verve Research, Oulu, Finland*

*<sup>3</sup> College of Healthcare Sciences, James Cook University, Townsville, QLD, Australia*

*\*Correspondence: d\_boullosa@yahoo.es*

#### *Edited by:*

*Mikko Paavo Tulppo, Verve, Finland*

#### *Reviewed by:*

*Harri Lindholm, Finnish Institute of Occupational Health, Finland*

**Keywords: physical activity, stress, exercise, cardiovascular diseases, physical fitness**

This e-book is the culmination of countless hours of meticulous work by global scientists. We would like to thank the researchers for their great contributions to this hot topic. The combination of these studies reflects the importance of the topic amongst researchers and practitioners and the wide interest from numerous laboratories around the world. The contributions include a variety of formats including five original investigations, three review articles, one opinion article and a hypothesis and theory article. Notably, these contributions included both human and animal models that encompassed a range of techniques from molecular mechanisms to real life interventions thus reinforcing the translational approach for the understanding of cardiovascular responses to stress.

The three review articles (Huang et al., 2013; Carnevali and Sgoifo, 2014; Tonello et al., 2014) provided a great insight into the current knowledge of cardiovascular stress and its relationship with physical activity (PA). The first review article by Huang et al. (2013) considered the big picture of the topic by combining the classical perspective–examining how different forms of induced cardiovascular stress can be attenuated in physically trained individuals, with the addition of the interrelationships between obesity, inflammation and oxidative stress with these stress responses and cardiovascular health. The subsequent review from Carnevali and Sgoifo (2014) involved animal studies that focused on a mechanistic approach with resting vagal tone leading to stress resilience, reducing the development of anxiety and depression, and the important role of PA to mediate these relationships. The mini-review from Tonello et al. (2014) extended the topic into one of the most important social stressors in modern life, the work environment. This review revealed that factors related to adverse working conditions such as excessive effort, effort-reward imbalance, over commitment, irregular shift work, and work stress were associated with reduced cardiac autonomic function. Importantly this review identified the need for further studies in this area with heart rate variability (HRV) recognized as an important tool for evaluating both work related stress including harmful physical inactivity, and adaptations to potentially important stress buffers as PA and enhanced physical fitness.

Following the review articles, two interesting papers addressed novel and important aspects for stress management therapies. The opinion article by Stults-Kolehmainen (2013) highlighted an important and frequently forgotten aspect: How does stress negatively influence PA levels in individuals with and without cardiovascular diseases? Additionally, Stults-Kolehmainen (2013) introduced the concept of "mindfulness" for reducing stress and subsequently enhancing patients PA levels. Subsequently, Demarzo et al. (2014) discussed potentially effective means to mediate the role of physical fitness on cardiovascular responses to stress using "mindfulness" interventions that may be applied to this new area of future research.

As mentioned previously, the variety of original research investigations included within this topic reinforced the importance of research translation. The cross-sectional study by Childs and de Wit (2014) reported a significant relationship between positive affect decline after a social stressor and regular exercise in healthy individuals despite no difference between exercisers and non-exercisers for post-stress heart rate, blood pressure and cortisol responses. This study effectively illustrated how regular PA enhances psychological health and strengthens emotional resilience to stressors. Likewise, the study of Hanson et al. (2013) provided additional support for the effectiveness of regular PA in attenuating physiological responses to social stressors. Interestingly, an antidepressant drug attenuated cardiovascular responses (i.e., HR and HRV) to stress only in irregular exercisers that exhibited a comparable stress-induced response similar to that of regular exercisers during placebo (Hanson et al., 2013). These two studies (Hanson et al., 2013; Childs and de Wit, 2014) highlighted the positive impact that regular PA has on mental health, especially for depressed patients, with further studies needed to elaborate on mechanisms, benefits, and potential new therapies.

Other important responses during different sources of stress were also included in this e-book. The study of Rauber et al. (2014) was the first to our knowledge that documented postexercise hypotension (PEH) in children. This study (Rauber et al., 2014) reported an enhanced PEH after traditional games compared to active video game playing and watching TV, and an attenuated blood pressure response during the cold pressor test following traditional games. The greater exercise intensity and metabolic demands of traditional games were highlighted as important factors for these responses with children's playing strategies fundamental for cardiovascular health.

The study of Franklin et al. (2014) documented the important contribution of other physiological responses during stress. These authors (Franklin et al., 2014) noted that endothelial function after resistance exercise (i.e., physical stressor) was impaired for obese women compared to lean women with endothelium independent-vasodilation correlated to body weight for all women. These findings provided important guidance for resistance exercise prescription of obese women to minimize cardiovascular disease risk.

Finally, the study of Sasse et al. (2013) presented a very novel hypothesis suggesting that exercise might facilitate adaptation to repeated stress via both hypothalamic-pituitary-adrenocortical axis and cardiovascular response habituation. In this study (Sasse et al., 2013), the brains of rats that performed voluntary exercise on a wheel and those that were sedentary were analyzed following control, acute and repeated noise exposures. These authors identified that unlike sedentary rats, exercising rats regulated corticotropin-releasing factor and brain derived neurotrophic factors across several brain regions. Subsequently, the hypothesis was supported with habituation to stress using exercise resulting in multiple system responses. Future studies may elaborate on the results of Sasse et al. (2013) to understand the response of different physiological systems and their interactions to stressors and applicable translation.

This e-book has taken the initial action to integrate current research findings to stimulate further discussion and research. We would like to thank the authors for their significant contributions and the many reviewers who critiqued and improved the overall topic. Future studies will clarify the importance of physical fitness, exercise and PA to regulate cardiovascular responses during stress and such benefits for cardiovascular health.

#### **REFERENCES**

Carnevali, L., and Sgoifo, A. (2014). Vagal modulation of resting heart rate in rats: the role of stress, psychosocial factors, and physical exercise. *Front. Physiol*. 5:118. doi: 10.3389/fphys.2014.00118


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Received: 03 September 2014; accepted: 03 November 2014; published online: 19 November 2014.*

*Citation: Boullosa DA, Hautala AJ and Leicht AS (2014) Introduction to the research topic: the role of physical fitness on cardiovascular responses to stress. Front. Physiol. 5:450. doi: 10.3389/fphys.2014.00450*

*This article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology.*

*Copyright © 2014 Boullosa, Hautala and Leicht. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

**REVIEW ARTICLE** published: 07 November 2013 doi: 10.3389/fphys.2013.00314

#### *Chun-Jung Huang1 \*, Heather E. Webb2, Michael C. Zourdos <sup>1</sup> and Edmund O. Acevedo3*

*<sup>1</sup> Department of Exercise Science and Health Promotion, Florida Atlantic University, Boca Raton, FL, USA*

*<sup>2</sup> Department of Kinesiology, Mississippi State University, Starkville, MS, USA*

*<sup>3</sup> Department of Health and Human Performance, Virginia Commonwealth University, Richmond, VA, USA*

#### *Edited by:*

*Daniel Boullosa, Universidade Católica de Brasília, Brazil*

#### *Reviewed by:*

*Arto J. Hautala, Verve Research, Finland Shane A. Phillips, University of Illinois at Chicago, USA*

#### *\*Correspondence:*

*Chun-Jung Huang, Department of Exercise Science and Health Promotion, Florida Atlantic University, 777 Glades Road, FH11A-126B, Boca Raton, FL 33431, USA e-mail: chuang5@fau.edu*

Psychological stress has been proposed as a major contributor to the progression of cardiovascular disease (CVD). Acute mental stress can activate the sympathetic-adrenal-medullary (SAM) axis, eliciting the release of catecholamines (NE and EPI) resulting in the elevation of heart rate (HR) and blood pressure (BP). Combined stress (psychological and physical) can exacerbate these cardiovascular responses, which may partially contribute to the elevated risk of CVD and increased proportionate mortality risks experienced by some occupations (e.g., firefighting and law enforcement). Studies have supported the benefits of physical activity on physiological and psychological health, including the cardiovascular response to acute stress. Aerobically trained individuals exhibit lower sympathetic nervous system (e.g., HR) reactivity and enhanced cardiovascular efficiency (e.g., lower vascular reactivity and decreased recovery time) in response to physical and/or psychological stress. In addition, resistance training has been demonstrated to attenuate cardiovascular responses and improve mental health. This review will examine stress-induced cardiovascular reactivity and plausible explanations for how exercise training and physical fitness (aerobic and resistance exercise) can attenuate cardiovascular responses to stress. This enhanced functionality may facilitate a reduction in the incidence of stroke and myocardial infarction. Finally, this review will also address the interaction of obesity and physical activity on cardiovascular reactivity and CVD.

**Keywords: psychological stress, obesity, physical activity, microvascular reactivity, inflammation, resistance exercise, oxidative stress, stress hormones**

# **INTRODUCTION**

Chronic psychological stress is a risk factor for cardiovascular disease (CVD). In addition, acute psychological stress is asssociated with factors that explain the development of atherosclerosis; endothelial dysfunction, inflammatory reactivity and oxidative stress. The American Psychological Association has provided evidence that 20% of Americans report extreme stress and 80% report that their stress levels have increased over the past year (APA, 2012). Additionally, over the past 5 years, 60% of Americans have attempted to reduce their stress, with just 7% reporting success in reducing stress (APA, 2012). Examinations of acute responses to psychological stress provide insight into the potential mechanisms that may explain the relationship of psychological stress to CVD. Greater understanding can also provide support for considering therapeutic alternatives that may alleviate the ill effects of stress.

This review will examine stress-induced cardiovascular reactivity and plausible explanations for how exercise training and physical fitness (aerobic and resistance exercise) can attenuate cardiovascular responses to stress. Important to our understanding of the development of CVD is how the benefits of physical activity in attenuating the cardiovascular stress response (enhanced functionality) may also support a reduction in the incidence of stroke and myocardial infarction. Finally, in light of the high prevalence of overweight and obesity (68.8% of US adults were categorized as overweight in 2008, and 35.7% were categorized as obese (Flegal et al., 2012) and support for the concept that obesity can be considered a chronic stressor marked by chronic inflammation, oxidative stress, and endothelial dysfunction, this review will also address the interaction of obesity and physical activity on cardiovascular reactivity and CVD (**Figure 1**).

### **SYMPATHOADRENAL RESPONSES TO ACUTE STRESS AND ADAPTATION TO PHYSICAL ACTIVITY**

Initially, the neuroendocrine response to stress was believed to be attributable solely to the release of catecholamines from the adrenal medulla (Cannon and De La Paz, 1911; Cannon, 1914). Cannon and De La Paz (1911) research regarding sympathetic activation in response to threat or danger resulted in the introduction of the concept known as the "fight or flight" response. Furthering Cannon's work, Hans Selye (1936) conceptualized that factors such as heat or cold, forced immobilization or exercise, as well as chemical, biological, and psychological factors will elicit the exact same non-specific response of not only the catecholamines, but also corticosteroids (Cannon and De La Paz, 1911; Cannon, 1914; Selye, 1936, 1950, 1954, 1976).

The relationship between psychological stress and cardiovascular reactivity has long been suggested as an explanation for the association between psychological stress and CVD. It is known that chronic psychological stressors can lead to increased risk of arteriosclerosis, hypertension, and other metabolic disorders (Chrousos, 2000b, 2009; McCrone et al., 2001; Kyrou and Tsigos,

2009), while acute stressors result in acute increases in blood pressure, HR, and decreased metabolic efficiency (Crews and Landers, 1987; Hamer et al., 2006; Webb et al., 2011).

When an individual is psychologically or physiologically stressed in an acute manner, a complex chain of reactions occur, stemming from responses occurring within the sympathoadrenal (SA) and hypothalamic pituitary adrenal (HPA) axes, and the parasympathetic (PNS) and sympathetic nervous system (SNS) pathways in the body. Acute psychological stress has been shown to elicit increases in the secretion of epinephrine (EPI) and norepinephrine (NE) from the SA axis (Frankenhaeuser, 1991; Schoder et al., 2000; Gerra et al., 2001), and cortisol from the HPA axis (Frankenhaeuser, 1991; Chrousos, 2000a; Gerra et al., 2001). It has also been suggested that the increases in HR and BP response are due to a decrease in vagal tone attributed to the PNS as a result of rhythmic shifts in HR mediated by the brainstem medullary mechanisms through the vagus nerve (Hatfield et al., 1998; Spalding et al., 2000; Smeets, 2010) or an increase in afferent sympathetic neuromuscular activation (Kaufman and Hayes, 2002; Smith et al., 2006). Physical stressors, such as exercise, will elicit NE release in a curvilinear manner in response to increasing workloads, while EPI secretion increases at workloads above 60% of an individual's VO2max (Hjemdahl, 1993; Rowell and Shepherd, 1996), and these increases are likely responsible for the concomitant increases in HR and BP.

While these same physiological mechanisms are experienced in response to a stressor of any type, the amount of reactivity and type of response experienced has been suggested to be impacted by multiple factors, including an individual's perception of control over a situation, the combination of multiple stressors, the level of cardiorespiratory fitness, obesity levels, and sex. The greater the amount of control an individual perceives they have over a situation results in lesser catecholamine, HR, and BP responses relative to a situation where an individual feel less control over a situation (Frankenhaeuser et al., 1976; Frankenhaeuser, 1991; Hinton et al., 1991).

While the majority of investigations of stress are often addressed from a unidirectional perspective, it has been suggested that psychological appraisals can interact with and cause alterations in peripheral physiological responses. The proposed transactional psychobiological model of cognitive appraisal during exercise suggests an individual's perception of the demands imposed by psychological and physical stressors is related to an individual's perceived ability to meet these demands Acevedo and Ekkekakis (2001). The combination of mental and physical stress has been shown to result in an exacerbated SA (Roth et al., 1990; Szabo et al., 1994; Rousselle et al., 1995; Acevedo et al., 2006; Huang et al., 2010a; Webb et al., 2010, 2008) and HPA (Webb et al., 2008, 2011, 2013) responses above that of a single stressor alone. Therefore, if pathological events are related to psychological and physical stress independently, then this combination of stressors, resulting in an exacerbated SA and HPA response, are likely be responsible for greater pathophysiological alterations in these systems.

Exercise is also believed to have an attenuating effect on an individual's reactivity level at resting levels, and research has shown this to generally be true (de Geus and van Doornen, 1993; de Geus et al., 1993; Porges, 1995; Sothmann et al., 1996; Schuler and O'Brien, 1997), with individuals of higher fitness levels exhibiting a lesser HR response to psychological stress (Claytor, 1991; Boutcher and Nugent, 1993; Spalding et al., 2000), as well as an attenuated EPI and NE response (Boutcher and Nugent, 1993). It has also been shown that PNS responses increase after exercise and this response may assist in blunting responses that elicit increases in HR and BP (Porges, 1995; Dishman et al., 2002). Exercise may have be beneficial effects on HPA functioning because of lower stress-induced cortisol responses in physically fit compared with unfit subjects (Traustadóttir et al., 2005; Rimmele et al., 2009). However, the actual impact of cardiorespiratory fitness on cardiovascular reactivity is still a topic of debate as multiple meta-analyses and larger clinical studies investigating this topic have differed in their conclusions (Crews and Landers, 1987; Forcier et al., 2006; Hamer et al., 2006; Jackson and Dishman, 2006; Sloan et al., 2011; Alex et al., 2013), with the studies citing the differences in participant demographics, research methodologies, and measured variables adding to the difficulty in coming to definitive conclusion.

Similarly, studies investigating the relationship between adiposity and SA axis responses also lack a clear relationship, with inconsistent relationships found among obesity and catecholamine response (Macdonald, 1995), although there is support for an increase in SNS reactivity in obesity (Grassi et al., 1995; van Baak, 2001; Alvarez et al., 2002). Abdominal obesity has been associated with both exaggerated (Goldbacher et al., 2005; Steptoe and Wardle, 2005) and blunted (Hamer et al., 2007; Carroll et al., 2008; Phillips, 2011; Phillips et al., 2012) cardiovascular reactivity. Thus, these differences in findings may be related to a number of factors, including participant demographic differences and methodological issues. It can be suggested that generally, studies with appropriate statistical adjustments and stringent inclusion criteria have reported negative associations between cardiovascular reactivity and abdominal obesity (Laederach-Hofmann et al., 2000; Hamer et al., 2007; Phillips, 2011).

It is also interesting to note that while both males and females react to stress through the same psychophysiological pathways, they do so with markedly different results. Research demonstrates that males respond with greater diastolic blood pressure and total peripheral resistance changes during acute stressors (Stoney et al., 1987, 1988; Matthews and Stoney, 1988; McAdoo et al., 1990; Lai and Linden, 1992; Matthews et al., 1992; Allen et al., 1993), while females experience greater changes in HR during a psychologically stressful situation (Frankenhaeuser et al., 1976; McAdoo et al., 1990; Frankenhaeuser, 1991; Allen et al., 1993). In addition to cardiorespiratory differences, it has also been shown that when subjected to an acute psychological stress, males had a significantly greater rise in EPI levels in comparison to females', whose EPI rose slightly or not at all. A similar, but less pronounced sex difference was also found for NE, with males again registering a greater change in hormonal levels (Frankenhaeuser et al., 1978; Forsman and Lindblad, 1983; Lundberg, 1983). These differences have led to suggestions (Allen et al., 1993) that male and female responses to mental stress may be attributed to different mechanisms, with males being considered "vascular" reactors and females classified as "cardiac" reactors.

One mechanism that has been suggested to account for differences in male and female vascular response during psychological stress is a greater sensitivity in peripheral alpha- and betaadrenergic receptors in women (Freedman et al., 1987; Girdler et al., 1990; Kajantie and Phillips, 2006). Another proposed mechanism that may account for the larger cardiac response in women may be a greater sensitivity and/or density of adrenergic receptors in the myocardium (Girdler et al., 1990; Kajantie and Phillips, 2006). Both hypotheses have been supported by research showing women respond with lower catecholamine secretions during an acute psychological stress when compared to males (Frankenhaeuser et al., 1978; Rauste-von Wright et al., 1981; Collins et al., 1985). It was also noted that differences in cortisol secretions during psychologically stressful tasks are negligible between males and females (Frankenhaeuser et al., 1978; Forsman and Lindblad, 1983; Lundberg, 1983), further suggesting that the SA axis may play a key role in reactivity differences among males and females during an acute psychological stress.

In addition, it should be noted that a number of studies have shown that men have higher BP levels than women through much of their lifetime regardless of race and ethnicity (Sandberg and Ji, 2012). Particularly, acute stress results in immediate increases in arterial blood pressure (Lutgendorf et al., 2000) to be a result of vasoconstriction prompted by enhanced SA activity (McCarty and Gold, 1996). Furthermore, chronic stress may lead to hypertension and other cardiovascular dysfunction as a result of disturbances in the SA axis and the nitric-oxide pathways (Alvarez et al., 2001; Stefano et al., 2001; Esch et al., 2002).

Interestingly, several studies demonstrate that cardiovascular responses during mental stress are better predictors of future hypertension (Wood et al., 1984; Matthews et al., 1993) than resting BP measurements. Furthermore, because an individual's BP levels respond to many factors, including daily activities and affect (Pickering, 1997), it has been hypothesized that cardiovascular responses during mental challenges may be better predictors of future resting BP levels than BP at rest. Further, it has been reported that BP levels during mental stress are more closely associated with left ventricular mass than are resting BP levels (Georgiades et al., 1996).

These acute responses to mental challenge can elicit increased cardiac output, increased systemic vascular resistance, and thus an elevation of arterial blood pressure. While sympathetic elevations that are transient in nature prepare the body for accommodating to possible physical demands that an individual may encounter, prolonged or frequently occurring elevations of the catecholamines can result in vasoconstriction in most systemic arteries and veins leading to allostatic alterations in cardiovascular responses (Gidron et al., 2002). These alterations provide conditions favorable for the development of hypertension, endothelial dysfunction, and may contribute to the development of arteriosclerosis (Chrousos and Gold, 1998; Gidron et al., 2002; Spieker et al., 2002).

Research has shown psychological stress may be attenuated by a number of factors including exercise (Crews and Landers, 1987; van Doornen and de Geus, 1989; de Geus et al., 1993; Sothmann et al., 1996; Schuler and O'Brien, 1997; Acevedo et al., 1999). Additionally, exercise has been shown to have immediate psychological benefits comparative to those of other traditional therapeutic modalities and is a more effective anxiolytic agent than cognitive therapies when assessing anxiety (Petruzzello et al., 1991). This proposed explanation is also supported by Dienstbier's (1989) physiological toughness hypothesis, in that it is suggested that cognitively based stress reduction strategies only provide a short-term solution to anxiety reduction.

# **MICROVASCULAR REACTIVITY TO OXIDATIVE STRESS AND INFLAMMATION**

Chronic stress has been demonstrated to be a determinant of CVD (Olinski et al., 2002). One of the earliest sub-clinical stages in the atherosclerotic process is an impairment of endotheliumdependent vasodilation, also known as endothelial dysfunction (Singhai, 2005). Acute mental stress is capable of altering physiological homeostasis such as microvascular reactivity (Huang et al., 2013; Szijgyarto et al., 2013). Laboratory-induced psychological stress has been demonstrated to induce transient endothelial dysfunction [impairment of brachial artery flow-mediated dilation (FMD)] (Ghiadoni et al., 2000; Szijgyarto et al., 2013). This stress-induced impaired FMD has been shown to be worsened in high-stress occupations (e.g., firefighting, law enforcement) and patients with stress disorders (e.g., depression) (Violanti et al., 2006; Mausbach et al., 2012; Wagner et al., 2012). Fahs et al. (2009) found increased aortic and carotid artery stiffness in firefighters. This observation has also been discovered following 3 h of firefighting activities (Fahs et al., 2011a,b). Furthermore, Joseph et al. (2010) showed a lower FMD in police officers compared to controls. This impaired FMD in police officers was associated with decreased carotid intima-media thickness, and 16 and 36% of these police officers (*N* = 100) met criteria for depression and reported posttraumatic stress disorder symptoms, respectively (Violanti et al., 2006).

A number of mechanisms have been shown to be involved in endothelial dysfunction that occurs as a result of acute mental stress. For example, Broadley et al. (2005)showed that acute mental stress-induced endothelial dysfunction (lower FMD) was prevented with an oral administration of metyrapone (an inhibitor of cortisol synthesis) in healthy individuals. Furthermore, plasma cortisol levels are higher at rest and in response to acute mental stress in firefighters and police officers compared to controls (Tomei et al., 2003; Rosati et al., 2011; Robinson et al., 2013). These reported elevations in cortisol levels have been associated with impaired FMD in police offices (Violanti et al., 2009). These findings suggest that the challenges of the HPA axis experienced by high-stress occupations may lead to an increased risk of CVD. Although the underlying mechanisms remain to be determined, elevated levels of oxidative stress and inflammation have also been implicated as contributing factors that link acute mental stress to endothelial dysfunction.

A mediator of endothelial dysfunction is shear stress (a dragging frictional force generated by blood flow in the vasculature), leading to oxidative stress (Bagi et al., 2003). Oxidative stress is an imbalance between antioxidants [e.g., nitric oxide (NO)] and reactive oxygen species (ROS) [superoxide (O2−), hydrogen peroxide (H2O2), and hydroxyl radical (OH−)] (Sies, 1997). In healthy vascular cells, ROS is generated during the metabolism of oxygen, with the rate of ROS production being balanced by the rate of oxygen elimination (Vider et al., 2001). When ROS production is elevated, the process of cell damage occurs and can possibly facilitate the development of CVD (Ji et al., 2006). Furthermore, any increase in vascular shear stress from the cardiovascular alterations in response to a mental or physical perturbation can influence the balance of oxidative stress.

Research has previously shown that psychological stress may contribute to the development of atherosclerosis by eliciting an elevation in ROS, which can further induce oxidative DNA damage (Olinski et al., 2002). Subsequently, in a study on medical students, Sivonova et al. (2004) demonstrated greater nuclear DNA damage in lymphocytes on the day of an examination (stress condition) compared with during the time between two examination periods (non-stress condition). To further understand endothelial dysfunction by oxidative stress, Chung et al. (2010) recently examined the effects of immobilization stress (120 min/day for 14 days) in rats and found that arterial endothelial nitric oxide synthase (eNOS) mRNA and NO decreased and plasma malondialdehyde level (a marker of oxidative stress) increased along with decreased acetylcholine-induced relaxation of arteries as compared to controls. Interestingly, Huang et al. (2010a) examined oxidative stress response in healthy individuals who were exposed to acute dual challenge (physical and psychological stress). This study found that the dual challenge condition elicited greater plasma 8-isoprostane levels (a biomarker of oxidative stress) compared to exercise alone condition. This finding may provide the potential explanation that a combined physical and psychological stress often experienced chronically by occupational professionals (e.g., emergency responders, firefighters, and police officers) is a contributing factor to endothelial dysfunction, thereby increasing the risk of CVD.

In addition, vascular inflammation plays a critical role in endothelial dysfunction which can be induced by the production of pro-inflammatory cytokines such as tumor necrosis factoralpha (TNF-α) and interleukin-6 (IL-6) (Ross, 1999; Esteve et al., 2007). Notably, a previous study has suggested that individuals who have a greater laboratory-induced stress response are more likely to experience higher stress in daily life (Wirtz et al., 2008), and this response is associated with increased risk for atherosclerosis (Kamarck et al., 1997). An increase in circulating TNF-a and IL-6 has been observed following acute mental stress (Steptoe et al., 2001; Heinz et al., 2003) via activation of nuclear factor-kappa B (Barnes and Karin, 1997; Bierhaus et al., 2003). Furthermore, Huang et al. (2010b) found a greater plasma IL-2 levels in firefighters who experienced a dual stress (physical and psychological stress) compared to exercise alone group. It is important to note that the magnitude and direction of inflammatory cytokine response to psychological stress vary and are dependent on the acute or chronic nature of the stimuli. For example, Kang and Fox (2001) examined chronic academic stress during examinations and found that decreased IL-2 [in both peripheral blood mononuclear cell (PBMC) and whole blood measures] and IFN-g (only PBMC) levels were observed, whereas an elevation was seen in IL-6 (in both PBMC and whole blood measures). In a recent study, Ramey et al. (2012) examined the inflammatory cytokine response in law enforcement officers and found that the job demand (physical and psychological) assessed by the Job Content Questionnaire was positively associated with resting IL-1 beta and IL-6. This finding is further supported by Groer et al. (2010) who demonstrated a significant increase in salivary IL-6 in police officers following simulated workplace scenario (6 min of tracking a gunman through a building. It is important to note that ∼23% of law enforcement officers who reported metabolic syndrome were physically inactive (Yoo et al., 2009). In a cross-sectional study of 527 firefighters, Durand et al. (2011) demonstrated the CVD risk factors such as triglyceride and low density lipoprotein cholesterol levels are negatively associated with time and frequency of physical activity per week. These elevations in the mediators of endothelial dysfunction may be partially explained by physical inactivity. Thus, it is pivotal to understand how lifestyle changes such as exercise may help alleviate the potential for the negative health outcomes in these occupations.

Epidemiological evidence consistently shows that the benefits of physical activity and fitness on physiological and psychological health. One of the purported benefits associated with aerobic fitness is the attenuation of the cardiovascular response during psychological stress and recovery (Dienstbier, 1989; Sothmann et al., 1996; Spalding et al., 2004). For example, aerobically trained individuals have shown to attenuate reactivity and recovery of HR and blood pressure (Crews and Landers, 1987; McCubbin et al., 1992) and lower cortisol reactivity to acute mental stress (Webb et al., 2013). Furthermore, aerobic exercise training has been shown to defend against ROS-induced lipid peroxidation and to decrease the occurrence of ROS-associated diseases such as CVDs and Alzheimer disease (Mattson and Wan, 2005; Perry et al., 2005; Radak et al., 2005). A number of studies have shown that exercise training can enhance the adaptation of oxidative stress by increasing antioxidant defenses demonstrated by the up-regulation of antioxidant gene expressions such as superoxide dismutase and glutathione peroxidase (Powers and Lennon, 1999; Leeuwenburgh and Heinecke, 2001). Additionally, Nojima et al. (2008) found that an oxidative stress marker (urinary 8-OHdG level) decreased in patient with type 2 diabetes following a 12-month of aerobic exercise training. These findings suggest that regular exercise is beneficial in up-regulating the resistance against oxidative stress.

In addition, physical fitness has been shown to provide a more resilient immune defense and greater stress protection. For example, a reduction in pro-inflammatory levels (TNF-a and IL-6) was found in patients with coronary heart disease following aerobic exercise training (Goldhammer et al., 2005), and lower IL-6 concentrations have been observed in individuals who had selfreported higher physical activity levels (Pischon et al., 2003). In response to acute mental stress, physical fitness is associated with diminished pro-inflammatory cytokine responses (TNF-α and IL-6) (Hamer and Steptoe, 2007). While stress management alone has not demonstrated an improvement in FMD (Blumenthal et al., 2005), Dod et al. (2010) have shown an improvement of FMD along with decreased IL-6 levels in patients with coronary artery disease (CAD) following 12-weeks of lifestyle interventions including exercise and stress management. Therefore, the development of strategies, including exercise training, to address the negative consequences of chronic stress may help alleviate the elevated risk for CVD.

# **RESISTANCE EXERCISE: CARDIOVASCULAR REACTIVITY TO STRESS**

Resistance exercise has long demonstrated positive adaptations in relation to skeletal muscle hypertrophy, muscular strength, and body composition (Wilmore, 1974). However, the effects of resistance training on cardiovascular reactivity and acute and long-term markers of stress are significant and perhaps overlooked. To appropriately understand these adaptations, in-depth insight must be pursued to examine the psychological effects of resistance exercise, and varying hormonal (cortisol, EPI, and NE) and hemodynamic responses (HR and BP). Additionally, knowledge related to the blood flow and inflammatory cytokine response, which may occur as a result of structured and periodized resistance training programs designed to elicit a desired training adaptation.

Pioneer research from Morgan (1969) has demonstrated a greater incidence of depression in unfit individuals compared to those who are more physically fit. Since then a plethora of data has illustrated the benefits of exercise to be associated with lower emotional distress (Steptoe and Butler, 1996) and decreased levels of depression (Steptoe et al., 1997). Additionally, longitudinal data from Paffenbarger et al. (1994) found physical activity to be inversely correlated with depression over the course of 25 years in a sample of over 10,000 men. The confounding variable, however, is that the research related to mental health has mostly examined aerobic exercise activity and is limited regarding resistance exercise. However, chronic resistance training has been shown to cause reductions in resting concentrations of the stress hormone cortisol (Hakkinen et al., 1988; Kraemer et al., 1998). This factor is of great importance as increased resting cortisol levels have been associated with impaired declarative memory in healthy adults (Kirschbaum et al., 1996). Additionally, resistance training may be an attractive strategy over the life span to not only attenuate the onset of sarcopenia (the loss of type II fibers with age), but to prevent elevations in cortisol, which have been related to memory impairments in elderly humans (Lupien et al., 1996) and to predict hippocampal dysfunction in an aging population (Lupien et al., 1998). Therefore, even though the prominent research related to exercise and mental stress focuses on endurance training, a physiological analysis of the benefits associated with resistance training suggests positive outcomes for mental stress when resistance training is included in an individual exercise routine.

Commonly, analyzed hormonal response in regards to resistance training protocols includes the anabolic hormone testosterone and the catabolic hormone cortisol, which are secreted from the HPA axis. Via the endocrine system, secretion of these hormones in response to training occurs to maintain homeostasis in the body (Galbo, 1983). Fluctuations in these hormones can occur acutely in response to exercise, or resting concentrations may be altered as an adaptation to a chronic training stress (Fry et al., 1994; Deschenes et al., 1998). As an acute response, it is typical to for significant increases in testosterone and cortisol to occur, however, smaller changes in resting levels over the long-term may indicate a positive response to stress (Staron et al., 1994). Moreover, the literature has demonstrated that acute hormonal fluctuations are simply due to the metabolic stress response of physical activity and bear little to no impact on long-term muscle performance adaptations (Ahtiainen et al., 2005). Further, these acute changes may even lead to decreased stress and increased exercise performance over time (Hakkinen et al., 1988). Recently, investigators have demonstrated that an intense bout of resistance training (5 sets of 8 repetitions @75% of one repetitions maximum -1RM) significantly increases both the catabolic hormone cortisol and the anabolic hormone testosterone by a similar percentage (McCaulley et al., 2009). Additionally, the acute stress response seems to be augmented by total work performed or training volume (Sets X Repetitions X Weight Lifted). In response to long-term training, data has shown decreased cortisol levels to lead to greater levels of muscle performance (Staron et al., 1994; Kraemer et al., 1999). Furthermore, data exists demonstrating that long-term (2 years) resistance training may increase the testosterone to cortisol ratio (T/C ratio), which also results in an increase in muscular force development (Hakkinen et al., 1988). Therefore, a reduction in cortisol over time may be an important signal of resistance training adaptation and a decline in stress levels leading to increased exercise performance.

Another stress response to consider is the elevation and recovery paradigm of the neurohormones of the adrenal medulla: EPI and NE. A plethora of research has demonstrated that, in response to training, there are acute elevations in levels of EPI and NE similar to cortisol, signaling a stress response (Kraemer et al., 1987). This acute response shows elicited activation of the adrenal medulla, evidenced not only by increases of EPI and NE up to 15 min post-resistance training, but also a decrease in plasma peptide f (P-F) immediately following training (Bush et al., 1999). Data from Bush et al. (1999) also revealed that following the initial decrease of P-F, the hormone then reverses course and increases up to 4 h following a resistance exercise protocol of either 10 or 15 repetitions per set. This observation of elevated P-F for up to 240 min following exercise is likely demonstrating that the adrenal medulla serves a physiological function to maintain appropriate homeostasis. The prolonged increase in P-F is intriguing as other research has shown elevations only up to 15 min post-exercise (Kraemer et al., 1985a,b, 1990), but a prolonged increase may suggest that secretion of P-F from the adrenal medulla may aid in recovery via immune response. The relationship between P-F and immune response is plausible, as data has demonstrated P-F to enhance antibodies as a result of increased T-Cell function (Hiddinga et al., 1994). Ultimately, T-Cells then serve as a mediator for B-Cell stimulation and response capability (Hiddinga et al., 1994). Therefore, the prolonged increase in P-F may signal as a response to combat stress by serving as a signaling mechanism for enhanced immune system activity. Further, these authors reported that a group who performed 10 repetitions per set had greater force output and blood lactate than those who performed 15 repetitions per set, however, there was no difference in the acute hormonal stress response to exercise. This lack of difference is likely because total work was equated.

Further analysis of adrenal medulla secretions in response to resistance exercise reveals that in previously trained lifters EPI and NE seem to increase while P-F seems to decrease (Bush et al., 1999). Additionally, in trained weightlifters, there is a significant increase in P-F up to at least 240 min following exercise (Bush et al., 1999). This demonstrates an important training adaptation acting to increase a trainee's ability to combat stress. In other words, a trained lifter may fight the high physiological demands of resistance training with an increase in P-F to signal immune response leading to recovery. This training adaptation is of importance, as it is well established that the eccentric phase of resistance training leads to significant skeletal muscle myofiber damage and fatigue as evidenced by elevated plasma creatine kinase (CK) levels (Nosaka et al., 2001; Chen and Hsieh, 2005) and soreness (Nosaka et al., 2001). Concomitantly, as myofiber damage occurs an influx of neutrophils and blood monocyte secretion of IL-1 Beta into the muscle, this may last for up to 5 days (Cannon et al., 1989; Fielding et al., 1993). Therefore, elevated P-F for up to 4 h may augment the recovery process to attenuate the muscle damage response. Ultimately, total work seems to be the key factor to elicit neurohomronal response, however, training status may increase secretions of hormones from the adrenal medulla leading to an improved immune response and faster recovery.

Additionally, resistance training has frequently been used as a component of cardiac rehabilitation programs (Pollock et al., 2000) and has shown resistance training status to effect hemodynamic responses (i.e., HR and BP). It is well known that acute increases in HR and BP are significant during and following an intense bout of resistance training (Darr et al., 1988). While parasympathetic activity is responsible for the initial rise in HR during resistance training due to vagal withdrawal, it is the elevation in activity of the SNS during an intense strength training session, which is responsible for increasing HR. Interestingly, individuals with previous training experience have seen HR return to baseline levels sooner following exercise than those of a less-advanced training status (Darr et al., 1988). This response suggests that chronic adaptations to resistance training act to handle increasing levels of stress and cause a more rapid return of HR to baseline levels following an intense bout of resistance training.

Moreover, resistance training has been shown to be a safe mode of exercise in patients with CAD or congestive heart failure CHF (Karlsdottir et al., 2002), this supports the use of chronic resistance training as a means to decrease the risk of stroke and myocardial infarction. In support of this notion, Karlsdottir et al. (2002) reported that subjects with CAD and CHF increased HR and BP to similar levels as cycling at 90% ventilatory threshold when performing one set of 10 repetition at 60–70% of one-repetition maximum on either the biceps curl, leg press, or shoulder press exercises. Additionally, these authors reported the stability of left ventricular function with resistance exercise in cardiac patients. Finally, CAD and CHF patients show a similar pattern of hemodynamic responses following resistance training to that of healthy individuals. Therefore, including resistance training in a rehabilitation program for cardiac patients may be a key component and appropriate stress reducer designed to improve quality of life and increase muscle mass.

Recently, blood-flow restriction training (BFR) at rest and in conjunction with resistance exercise has become an increasingly attractive method in which to examine cardiovascular reactivity. BFR training has been shown to increase muscle strength and hypertrophy by restricting blood flow proximal to the exercising muscle resulting in blood pooling and a restriction of venous return (Iida et al., 2007; Loenneke et al., 2012). It has been demonstrated that even without exercise, BFR training via application of a KAATSU device with a pressure of 200 mmHg has produced HR and BP responses similar to that of an orthostatic impetus (Iida et al., 2007). These intriguing results suggest that BFR may be a plausible method to be utilized as a countermeasure against orthostatic intolerance. Interestingly, recent data suggest that the introduction of BFR with resistance exercise has demonstrated greater HR and BP during exercise in young (30 years) and older (66 years) subjects (Vieira et al., 2013). However, in healthy individuals, BP and arterial compliance did not change following 6 weeks of full body resistance training in a low-intensity BFR resistance exercise group (20% of 1RM), nor did it change in moderate-intensity (70% of 1RM) or low-intensity (50% 1RM) resistance exercise groups (Fahs et al., 2011a,b). Therefore, individuals who have contraindications to performing high-intensity resistance training, such as cardiac patients, may be able to get resistance training benefits while not negatively altering arterial compliance by using blood flow restricted exercise.

In addition to the catabolic hormone cortisol discussed previously, the pro-inflammatory cytokine, IL-6, and antiinflammatory cytokine, IL-10, can be important factors to alert of a stress response and subsequent fatigue and recovery. Because resistance exercise causes myofibrillar disruption and localized edema, an inflammatory response mediated by cytokines will occur (Izquierdo et al., 2009). It is therefore likely that the response of inflammatory cytokines in the presence of skeletal muscle damage serve as a mediator of satellite cell activation and myofiber repair to reduce stress and induce hypertrophic adaptations of the skeletal muscle (Izquierdo et al., 2009). Data has shown IL-6 to be elevated up to 48 h post-resistance training of 5 sets of 10 repetitions (Izquierdo et al., 2009). This elevation in IL-6 is consistent with research showing increased CK levels at 48 h following similar resistance protocols (Chen and Hsieh, 2005) and suggesting that IL-6 is directly correlated with the muscle damage response, further signifying the role of IL-6 in recovery and its importance in alleviating a stress response to training (Smith et al., 2000; Suzuki et al., 2002; Izquierdo et al., 2009).

The anti-inflammatory cytokine IL-10 has been increased by catecholamines (EPI and NE) following exercise (Peake et al., 2005). Further, the responsiveness of IL-10 seems to be augmented in a subsequent training bout 4 weeks later (Hirose et al., 2004). The greater response of IL-10 in a subsequent training bout may be viewed as an adaptation to stress as a result of decreased inflammation. The attenuated inflammatory response to training is a component of the repeated bout effect (RBE), which stipulates that when the same exercise (Nosaka et al., 2001) is performed in a subsequent resistance training bout, the muscle damage response is attenuated. Ultimately, IL-6 and IL-10 not only play important processes in recovery, but their blunted release secondary to chronic training can be a demonstration of the RBE and may signify a reduction in the stress response to resistance training.

Data has demonstrated the ability of resistance-trained individuals to improve adaptations to stress, including; lower cortisol levels (Hakkinen et al., 1988), a more rapid return of HR to baseline levels (Darr et al., 1988), and increases in P-F possibly leading to enhanced immune function (Bush et al., 1999). Moreover, a reduction in cortisol levels that is associated with chronic resistance training is likely an appropriate method to attenuate decreasing hippocampal function with aging. Additionally, resistance training seems to be a safe method for patients with CAF and CAD (Karlsdottir et al., 2002) and its benefits may decrease the risk of CVD. Ultimately, for an individual to reduce chronic stress it seems that resistance training is a key component to optimize this goal.

# **FUTURE RESEARCH DIRECTIONS: ENDOTHELIAL FUNCTION TO STRESS IN OBESITY**

Obesity is associated with job-associated stress (tension and anxiety) and depression, and these stress related disorders have been found to lead to an increased risk of CVD and mortality (Nishitani and Sakakibara, 2006; Valtonen et al., 2012). Recent studies have reported ∼77% overweight and obesity rates in high-stress professionals such as young emergency responders (firefighters and ambulance recruits), police officers, and military personnel (Franke et al., 2002; Hsu et al., 2007; Tsismenakis et al., 2009; Ramey et al., 2009). Importantly, obesity has been shown to disturb cardiovascular responsivity to acute mental stress (Hamer et al., 2007), which may associate with stress-related endothelial dysfunction. Furthermore, in response to acute mental stress, Ghiadoni et al. (2000) have also shown that diabetic patients have a lower FMD compared to the control subjects. In addition, Martin et al. (2013) found that the index of adiposity (e.g., BMI, waist circumference, and waist-to-hip ratio) were negatively correlated with measures of vascular function such as hyperemic velocity time integral and hyperemic shear stress in a total of 1462 male firefighters. Thus, understanding the link between obesity and psychological stress may provide a critical contribution in determining the pathopysiological mechanisms of CVD development.

Although there is limited information investigating the impact of psychological stress on the oxidative stress in obese populations, one mediator of obesity-induced endothelial dysfunction may be the level of oxidative stress (Timimi et al., 1998; Schäfer and Bauersachs, 2008). Flint et al. (2007) has demonstrated that cortisol, norepinephrine, and epinephrine released during psychological stress can induce DNA damage within 10 min. A study by Epel et al. (2000) showed that elevated cortisol responses to acute psychological stress are associated with increased waist-tohip ratio. This increase in oxidative stress during psychological stress could possibly be the negative effect of high circulating levels of stress hormones (e.g., cortisol).

Another potential mechanism to explain stress-induced endothelial dysfunction in obesity is the interactions of leptin with oxidative stress and inflammation. Leptin, an adipocytederived hormone, plays an important role in metabolism, adiposity, and vascular inflammation, and has been implicated in the development of coronary heart disease (Wannamethee et al., 2007). *In vitro* stimulation of cultured human endothelial cells with leptin has induced an increased accumulation of ROS and levels of pro-inflammatory mediator (e.g., monocyte chemotactic protein-1) via activation of nuclear factor-kappa B (Bouloumie et al., 1999). Interestingly, recent research has shown that people who undergo acute psychological stress demonstrate increases in leptin levels, and these increases are positively correlated with waist circumference (Otsuka et al., 2006; Brydon et al., 2008). Brydon et al. (2008) also showed that a positive correlation between basal circulating leptin and IL-6 responses in response to mental stress. Taken together, these findings suggest that stress-induced leptin may partially contribute to the development of endothelial dysfunction. However, although a number of investigators have expressed interest in examining the mechanisms underlying psychological stress-induced endothelial dysfunction, the possible interaction (additive or synergistic) of obesity and psychological stress on the development of endothelial dysfunction are still not fully understood. Future investigation should attempt to expand the understanding of the mechanisms contributing endothelial dysfunction that links between obesity, psychological stress and CVD.

# **CONCLUSIONS**

The studies that have been conducted examining the effect of physical activity on cardiovascular reactivity and CVD generally support Dienstbier's physiological toughness hypothesis (1989). More specifically, physical activity can not only reduce the immediate effects of stress but can also enhance the recovery from stressors (Crews and Landers, 1987; van Doornen and de Geus, 1989; de Geus et al., 1993; Sothmann et al., 1996; Schuler and O'Brien, 1997; Acevedo et al., 1999). Additionally, exercise has been shown to have immediate psychological benefits relative to other therapeutic treatments and can likely serve as a very effective adjuvant therapy (Petruzzello et al., 1991).

Furthermore, aerobic exercise training has been shown to defend against ROS-induced lipid peroxidation and to decrease the occurrence of ROS-associated diseases such as CVDs and Alzheimer disease (Mattson and Wan, 2005; Perry et al., 2005; Radak et al., 2005). These findings suggest that regular exercise is beneficial in up-regulating the resistance against oxidative stress. The immune system has also demonstrated positive adaptations to physical activity. Physical fitness has been shown to elicit a more resilient immune defense and greater stress protection. In response to acute mental stress, physical fitness is associated with diminished pro-inflammatory cytokine responses (TNF-α and IL-6) (Hamer and Steptoe, 2007). This effect is not only seen in aerobic activity, but also with resistance exercise. In particular, the attenuated inflammatory response to training is a component of the repeated bout effect, which stipulates that when the same exercise (Nosaka et al., 2001) is performed in a subsequent resistance training bout, the muscle damage response is attenuated. Ultimately, IL-6 and IL-10 are important in recovery, and their blunted release following chronic training can be a demonstration of the repeated bout effect that signifies a reduction in the stress response to resistance training. Further investigation into the benefits of resistance exercise and stress is warranted, although initial reports are promising and parallel the benefits of aerobic exercise.

Finally, the interaction of obesity and psychological stress on the development of CVD is not fully understood. Future examination of the mechanisms contributing to endothelial dysfunction and the links between obesity, psychological stress and CVD will undoubtedly lead to the development of more specific and effective strategies, such as physical activity training, to address the ill effects of stress on CVD.

# **REFERENCES**


blood flow by KAATSU in healthy subjects. *Eur. J. Appl. Physiol*. 100, 275–285. doi: 10.1007/s00421-007-0430-y


Stress (BCOPS) pilot study: methods and participant characteristics. *Ann. Epidemiol*. 16, 148–156. doi: 10.1016/j.annepidem.2005.07.054


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Received: 01 August 2013; paper pending published: 28 August 2013; accepted: 13 October 2013; published online: 07 November 2013.*

*Citation: Huang C-J, Webb HE, Zourdos MC and Acevedo EO (2013) Cardiovascular reactivity, stress, and physical activity. Front. Physiol. 4:314. doi: 10.3389/fphys. 2013.00314*

*This article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology.*

*Copyright © 2013 Huang, Webb, Zourdos and Acevedo. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Vagal modulation of resting heart rate in rats: the role of stress, psychosocial factors, and physical exercise

# *Luca Carnevali and Andrea Sgoifo\**

*Stress Physiology Laboratory, Department of Neuroscience, University of Parma, Parma, Italy*

#### *Edited by:*

*Daniel Boullosa, Universidade Católica de Brasília, Brazil*

#### *Reviewed by:*

*Arto J. Hautala, Verve Research, Finland Ben N. Greenwood, University of Colorado Boulder, USA*

#### *\*Correspondence:*

*Andrea Sgoifo, Stress Physiology Laboratory, Department of Neuroscience, University of Parma, Parco Area delle Scienze 11/a, Parma 43124, Italy e-mail: andrea.sgoifo@unipr.it*

In humans, there are large individual differences in the levels of vagal modulation of resting heart rate (HR). High levels are a recognized index of cardiac health, whereas low levels are considered an important risk factor for cardiovascular morbidity and mortality. Several factors are thought to contribute significantly to this inter-individual variability. While regular physical exercise seems to induce an increase in resting vagal tone, chronic life stress, and psychosocial factors such as negative moods and personality traits appear associated with vagal withdrawal. Preclinical research has been attempting to clarify such relationships and to provide insights into the neurobiological mechanisms underlying vagal tone impairment/enhancement. This paper focuses on rat studies that have explored the effects of stress, psychosocial factors and physical exercise on vagal modulation of resting HR. Results are discussed with regard to: (i) individual differences in resting vagal tone, cardiac stress reactivity and arrhythmia vulnerability; (ii) elucidation of the neurobiological determinants of resting vagal tone.

**Keywords: autonomic nervous system, heart rate variability, arrhythmias, stress, anxiety, depression, exercise, rat**

#### **INTRODUCTION**

The sinoatrial node has an intrinsic rate of spontaneous automaticity that sets the basic rhythm of the heartbeat (Jose and Collison, 1970). The dynamic balance between the sympathetic and parasympathetic (vagal) influences on sinoatrial node activity primarily determines the actual heart rate (HR) of a given physiological state. In healthy individuals, vagal modulation (or "tone") prevails under resting conditions. The role of resting vagal tone in healthy cardiac function has been increasingly recognized over the past 20 years (Levy and Schwartz, 1994). This was predominantly due to the rise of research that applied heart rate variability (HRV) analysis as a window into cardiac autonomic control. Mounting evidence indicates that those individuals showing higher vagal tone than average at rest tend to be more resilient to stress, adapting well across a number of different situations (El-Sheikh et al., 2001; Kok and Fredrickson, 2010; Smeets et al., 2010; Souza et al., 2013). On this regard, the beneficial effects of regular physical exercise on cardiac health appear to be mediated by an increase in resting vagal outflow (Smith et al., 1989; Rosenwinkel et al., 2001; Rennie et al., 2003; Soares-Miranda et al., 2009; Fu and Levine, 2013). On the contrary, low levels of vagal modulation may not adequately counteract sympathetic stimulation, leaving the heart vulnerable to ventricular tachyarrhythmias and sudden cardiac death (Schwartz et al., 1988; Esler, 1992; Volders, 2010). Several factors, including chronic life stress, negative personality traits, anxiety and mood states have been associated with vagal withdrawal and sympathetic predominance (Sloan et al., 1994; Friedman and Thayer, 1998; Rozanski et al., 1999; Gorman and Sloan, 2000; Lucini et al., 2005). Moreover, depression of vagal modulation characterizes cardiovascular pathology (e.g., heart failure) (Sabbah et al., 2011) and strongly predicts cardiac mortality after myocardial infarction (La Rovere et al., 1998). Interventional therapies for restoring the autonomic balance in such psychological and cardiac conditions include pharmacological, biobehavioral, and exercise strategies (Bryniarski et al., 1997; Iellamo et al., 2000; Tsai et al., 2006; Blumenthal et al., 2007; Nolan et al., 2008). Traditionally, pharmacological approaches have mainly been directed at reducing cardiac sympathetic outflow, while overlooking the possibility of enhancing vagal tone. This was presumably due to a lack of a comprehensive understanding of the central neurobiological determinants of resting vagal tone. Preclinical research is now attempting to fill this knowledge gap.

This paper focuses on current and past HRV research in rats that can potentially increase our understanding of the neurobiological mechanisms underlying vagal tone impairment/enhancement. It describes (i) how resting vagal tone is regulated and measured, (ii) the relations among individual differences in resting vagal tone, cardiac stress reactivity, and arrhythmia vulnerability, and (iii) the role of stress, psychosocial factors and physical exercise on the neural determination of resting vagal tone.

#### **ORIGIN, COMPONENTS, AND MEASUREMENTS OF RESTING VAGAL TONE**

Cardiac vagal activity is generated by cardiac vagal pre-ganglionic neurones mainly located in the nucleus ambiguous (NAmb) of the lower brainstem (**Figure 1**). Their axons innervate

**Abbreviations:** DMH, dorsomedial hypothalamus; HA, high-aggressive; HAB, high-anxiety behavior; HF, high frequency; HR, heart rate; HRV, heart rate variability; LAB, low-anxiety behavior; LF, low frequency; NA, non-aggressive; NAmb, nucleus ambiguus; NTS, nucleus tractus solitarii; RMSSD, root mean square of successive R-R interval differences; RSA, respiratory sinus arrhythmia; VLF, very low frequency; 5-HT, serotonin.

post-ganglionic motoneurones located in a ganglionic plexus on the heart itself (Cheng et al., 1999; Cheng and Powley, 2000) (**Figure 1**). Under resting conditions, this vagal pathway fires a rapid and continuous signal (or "tonus") to the sinoatrial node, slowing HR via acetylcholine release from efferent vagal nerve discharge. Vagally-released acetylcholine counteracts sympathetic effects both post-synaptically (via cAMP) and pre-synaptically (by reducing noradrenaline release from the sympathetic terminals) (Levy, 1971).

Cardiac vagal activity is dependent on excitatory and inhibitory synaptic inputs (Mendelowitz, 1996), including baroreflex [mediated by the nucleus tractus solitarii (NTS)] and respiratory input (Rentero et al., 2002) (**Figure 1**). Breathing modulates cardiac vagal activity through two mechanisms: (i) directly by projections from the central respiratory pattern generator, and (ii) indirectly by ascending afferents from lung stretch receptors that are activated during inspiration (**Figure 1**). These ascending afferents activate GABA-ergic neurons in the NTS that inhibit cardiac vagal motorneurons in the NAmb (**Figure 1**). Collectively, these mechanisms produce respiratory sinus arrhythmia (RSA)—rhythmic oscillations of HR around its mean value—with increases in HR during inspiration as vagal influence is momentarily suppressed, and decreases in HR during the early expiration phase as vagal influence resumes. Although moment-to-moment HR fluctuations can also be generated in response to, for example, physical movement and blood pressure and thermoregulatory changes, respiration is reliably periodic. Therefore, the strength of vagal influence can be estimated by measuring the rhythmic oscillation in the intervals between consecutive heart beats that are due to RSA—this is called HRV (Task Force, 1996).

HRV analysis is an established tool to estimate cardiac autonomic regulation in humans and animal models. Although the suitability of using HRV parameters to estimate sympathetic modulation is highly debated (Reyes Del Paso et al., 2013), this approach produces reliable measures of vagal tone (Berntson et al., 1997; Reyes Del Paso et al., 2013). Traditional HRV methods fall under the broader description of being either "time-domain" or "frequency domain" analyses. Among time domain measures of HRV, which are assessed with calculations based on statistical operations on R-R intervals (Kleiger et al., 1992), the root mean square of successive R-R interval differences (RMSSD) detects high frequency oscillations of HR, and therefore estimates parasympathetic nervous system activity (Stein et al., 1994). Frequency domain measures are based on spectral analysis of a sequence of R-R intervals and provide information on how power (variance) is distributed as a function of frequency. Usually, three oscillatory components are distinguished in the spectral profile: the very low (VLF), the low (LF), and the high (HF) frequency band. The power of HF band includes respiration-linked oscillations of HR and therefore reflects the modulation of vagus nerve discharge caused by respiration (Berntson et al., 1997), whereas the LF and VLF bands are related to a more gradual interplay between sympathetic and parasympathetic influences (Reyes Del Paso et al., 2013). The power of LF and HF bands is often reported in normalized (relative or fractional) units, which correspond to the relative value of each power in proportion to the total power (usually minus the VLF component). In particular, LF to HF ratio is taken as a synthetic measure of sympatho-vagal balance as it estimates the fractional distribution of power (Task Force, 1996).

It is worth mentioning for completeness sake that algorithms based on the chaos theory and nonlinear dynamics have been developed in order to evaluate in greater detail the intrinsic complexity of HRV (for an historical overview of the evolution of the concept of HRV and its application see Billman, 2011). Nonlinear methods are based on the assumption that the mechanisms involved in HR regulation interact with each other in a nonlinear way. The basic concept of nonlinear HRV methods is to try to capture the non-periodic behavior and complexity that exist inside the R-R interval dynamics. Various nonlinear methods have been tested in several sets of R-R interval data (Bigger et al., 1996; Lombardi et al., 1996; Makikallio et al., 1997, 1999; Huikuri et al., 1998, 2000), providing additional prognostic information and complementing traditional time- and frequency-domain analyses.

#### **INDIVIDUAL DIFFERENCES IN RESTING VAGAL TONE: PHYSIOLOGICAL SIGNIFICANCE**

Porges' polyvagal theory (Porges, 1995b) introduces a new perspective relating vagal tone during steady states (i.e., resting vagal tone) and vagal reactivity in response to environmental demands. According to this theory, measures of resting vagal tone are informative of an organism's ability to maintain homeostasis and the potential responsiveness of that organism. In particular, high levels of resting vagal tone can be considered a sign of autonomic flexibility, the capability of the parasympathetic nervous system to generate adequate responses to environmental challenges by modifying HR, respiration and arousal (Porges, 1995a,b, 2007; Beauchaine, 2001). On the contrary, depression of vagal modulation may predict mismatches between environmental demands and cardiac (re)-activity, thus increasing vulnerability to cardiac arrhythmias and sudden cardiac death (Schwartz et al., 1988; Esler, 1992; Volders, 2010).

A deeper insight into the complex relations among individual differences in resting vagal tone, cardiac stress reactivity and arrhythmia vulnerability might be facilitated by HRV studies in rats. Conventional methods for HRV research in rats rely on conscious state electrocardiogram (ECG) recorded by telemetry (Sgoifo et al., 1996). By applying this approach, we obtained a detailed characterization of autonomic regulation of resting HR in two strains of rats, Wild-type and Wistar (**Figure 2**). Results indicate that Wild-type rats were characterized by higher HR than Wistar counterparts both during the dark/active and light/inactive phases of the daily cycle (**Figure 2**). HRV analysis allowed unveiling the autonomic determinants underlying the differences in resting HR between these two rat strains. As suggested by the vagal indexes RMSSD and HF power, Wildtype rats were clearly characterized by lower levels of vagal modulation compared to Wistar rats (**Figure 2**). Consequently,

**FIGURE 2 | Daily rhythms of heart rate, RMSSD, spectral power in HF band (0.75–2.5 Hz) and LF (0.2–0.75 Hz) to HF ratio in 4-month-old Wild-type (***n* **= 10) and Wistar (***n* **= 10) rats.** These values (reported as means ± s.e.m.) were obtained by averagingmultiple 2-min segments acquired every hour during

the 12 hours-light and 12 hours-dark phases, as previously described in detail (Carnevali et al., 2013a). Results of Two-Way ANOVA for repeated measures: group difference for heart rate (*F* = 103*.*7, *p <* 0*.*01), RMSSD (*F* = 6*.*8, *p <* 0*.*05) HF (*F* = 7*.*4, *p <* 0*.*5), and LF to HF ratio (*F* = 4*.*4, *p <* 0*.*05) values. in Wild-type rats the sympatho-vagal balance (indexed by the LF to HF ratio) was shifted toward a sympathetic dominance (**Figure 2**). These results indicate that these two rat strains may be viewed as extremes, in terms of cardiac autonomic modulation, within the range of the natural variation of this species, and therefore may represent a valid model for investigating the neurobiological determinants underlying individual differences in resting vagal tone. As discussed before, this inter-individual variability may influence the individual's response to subsequent challenges. Addressing this issue, the two strains were tested under stress conditions (i.e., restraint test). Even though the peak HR reached in response to restraint stress was higher in Wild-type rats, these animals showed reduced overall HR stress responsiveness compared to Wistar rats, as suggested by the values of the area under the response curve above baseline (AUC) (**Figure 3**). In Wild-type rats, this phenomenon appeared to be related to a reduced vagal flexibility, namely a smaller vagal withdrawal in response to stress, and was coupled with a larger, although modest, vulnerability to arrhythmias compared to Wistar counterparts (**Figure 3**). Interestingly, a previous study demonstrated that arrhythmia vulnerability was remarkably larger in Wild-Type than Wistar rats in response to a social stressor (Sgoifo et al., 1998). Collectively, these findings support the view that low levels of vagal modulation of resting HR may be associated with an inability to flexibly generate adequate cardiac responses to stress, even in the absence of evident cardiovascular pathologies.

# **EFFECTS OF STRESS AND PSYCHOSOCIAL FACTORS ON VAGAL MODULATION OF RESTING HR**

Several humans studies provide clear and convincing evidence that chronic life stress, negative personality traits such us aggressiveness, anger, and hostility, anxiety and mood states contribute significantly to the pathogenesis and progression of cardiac disorders (Sloan et al., 1994; Friedman and Thayer, 1998; Rozanski et al., 1999; Gorman and Sloan, 2000; Lucini et al., 2005; Albus, 2010). Putative underlying mechanisms may include a disruption of the sympatho-vagal balance, through an increase in sympathetic activity and/or a decrease in vagal tone. Rat studies on the integration of stress, psychosocial factors, and cardiac autonomic neural control have just started investigating the underlying mechanistic links.

#### **STRESS**

Several rat studies have investigated the short- and long-term effects of repeated stress exposure on cardiac autonomic regulation. In a study conducted by Trombini et al. (2012), rats

**FIGURE 3 | Cardiac response to a restraint stress in 4-month-old Wild-type (***n* **= 10) and Wistar (***n* **= 10) rats.** Panels **(A–C)** show the time course of changes in heart rate, RMSSD values, and high-frequency (HF, 0.75–2.5 Hz) spectral power, respectively, in baseline conditions (bas, 30 min) and during the restraint test (15 min). Each point during the restraint phase is the mean of 5-min intervals. Inner graphs in **(A–C)** represent the area under

the response time curve above baseline (AUC) of heart rate, RMSSD, and HF values during the restraint phase. Panel **(D)** illustrates the incidence of cardiac arrhythmias during the restraint phase. Data analysis was conducted as described previously in detail (Carnevali et al., 2013a). Values are reported as means <sup>±</sup> s.e.m. <sup>∗</sup> and # indicate a significant difference between HAB and LAB rats (*p <* 0*.*05 and *p <* 0*.*01, respectively; Student's *t*-test).

submitted to intermittent episodes of restraint stress showed a prolonged increase in resting vagal drive in the days that followed the end of the stress period, as indicated by the increase in the vagal index RMSSD. Likewise, in another study, sub-chronically stressed rats (5 days of footsocks) exhibited a substantial increase in resting vagal tone that lasted well beyond the duration of the stressor (Carnevali et al., 2011). This peculiar phenomenon might be called "enduring vagal rebound" or "persistent vagal rebound" (Carnevali et al., 2011). In the literature, the term "vagal rebound" usually refers to a short-term vagal hyperactivity following a stressor, a sympathetic overdrive or reperfusion following acute myocardial infarction (Chiladakis et al., 2001; Mezzacappa et al., 2001). In this study (Carnevali et al., 2011), it was a relatively persistent, long-term consequence of repeated stress exposure, whose underlying neurobiological determinants have yet to be determined. However, it appears that this enduring vagal rebound is corticosterone- and serotonin-independent, as it was not prevented by metyrapone (inhibitor of corticosteroid synthesis) or fluoxetine (serotonin-selective reuptake inhibitor) treatments (Carnevali et al., 2011). Such sustained vagal activation after repeated stress exposure has been interpreted as a sign of adaptation, a transient compensatory phenomenon that initially overcomes the commonly observed stress-induced sympathetic hyperactivity (Carnevali et al., 2011). This adaptation might fail after a more prolonged exposure to stress and/or exposure to a more severe stressor, leading the organism to a maladaptive phase of vagal withdrawal and sympathetic dominance. Indeed, like humans, the most pervasive stressors in rats fall within the social domain (Bjorkqvist, 2001; Sgoifo et al., 2014). On this regard, the resident-intruder stress paradigm (Miczek, 1979) is regarded as an ethologically relevant model of psychosocial stress in rats, which involves subjecting a male intruder rat to aggressive threats and overt attacks by an unfamiliar male rat (resident) in the resident's home cage. This stress paradigm has recently been applied to study the shared pathophysiology that links stress-related psychological disorders such as anxiety and depression with cardiovascular disease (Carnevali et al., 2012b, 2013b; Wood et al., 2012; Sevoz-Couche et al., 2013). This raises another important issue that will be discussed in the next sections: the interplay between stress and related psychological disorders in the modulation of resting vagal tone.

#### **ANXIETY**

Cardiac autonomic function in rats displaying symptoms of stress-evoked anxiety was investigated in a study conducted by Sevoz-Couche et al. (2013). In this study, HRV analysis was conducted in rats submitted to intermittent daily episodes of social defeat, following a classical resident-intruder paradigm. Five days after the last defeat, stressed rats showed symptoms of an anxiety-like behavior that were accompanied by signs of reduced vagal modulation of resting HR, as suggested by the decrease in the vagal indexes RMSSD and HF power (Sevoz-Couche et al., 2013). Interestingly, such reduction in cardiac vagal tone was prevented by inhibition of the dorsomedial hypothalamus (DMH) with muscimol and the blockade of the NTS 5-HT3 receptors with granisetron. During acute stress, the DMH acts on the rostral cuneiform nucleus that, in turn, activates the rostral periacqueductal gray (Netzer et al., 2011). Downstream to this structure, serotonin (5-HT) is released to activate 5-HT3 receptors that are localized presynaptically on NTS vagal afferents (Leslie et al., 1990) (**Figure 1**). When activated, these receptors trigger glutamatergic activation of local GABAergic interneurons that project to the NAmb on pre-ganglionic vagal neurons, thereby inhibiting vagal activity (Loewy, 1990) (**Figure 1**). Therefore, it is reasonable to hypothesize that chronic activation of 5-HT3 receptors in the NTS due to overactivity of 5-HT in this region and/or hypersensitivity of 5-HT3 receptors is at the origin of vagal tone impairment in repeatedly stressed animals. These findings highlight the role of DMH and NTS in vagal tone impairment following repeated social challenge.

The alterations in cardiac vagal tone that have been described in this animal model of stress-evoked anxiety may also characterize other forms of anxiety. Addressing this issue, we have recently investigated cardiac autonomic function in a rat model of trait anxiety, namely Wistar rats selectively bred for either high (HAB) or low (LAB) anxiety-related behavior (Carnevali et al., 2014). The HAB/LAB rats have been proved to display robust and consistent differences in their level of baseline anxiety (Landgraf and Wigger, 2002) and therefore represent a valid and reliable model for investigating the autonomic correlates of extremes in anxietyrelated behavior. HAB rats clearly displayed a lower vagal modulation of resting HR (indexed by RMSSD and HF values) compared to LAB rats during both the light/inactive and dark/active phases of the daily cycle (**Figure 4**). One of the most puzzling findings was that, despite this evident difference in resting vagal tone, HAB and LAB rats had similar HR values (**Figure 4**). This is indicative of the fact that simple measurements of HR do not necessarily provide accurate insights into the functional regulatory characteristics of the autonomic nervous system. Indeed, mean HR values are indicative of the net effects of sympathetic and vagal influences on cardiac activity, which often result from a combination of concurrent changes in activity within both branches (Berntson et al., 1991). Therefore, it is reasonable to hypothesize that the reduced cardiac vagal tone observed in HAB rats was coupled with a decreased cardiac sympathetic influence on the sinoatrial node compared to LAB rats. Supporting this view, HAB and LAB rats showed similar LF to HF ratio (index of sympatho-vagal balance), suggesting that vagal and sympathetic influences on cardiac pacemaker activity were equally balanced in the two groups, leading to similar HR values (**Figure 4**). When tested under stress conditions (i.e., restraint test), HAB rats showed reduced HR stress responsiveness compared to LAB counterparts, as suggested by the values of the area under the response curve above baseline (AUC) (**Figure 5**). In HAB rats, this phenomenon appeared to be related to a reduced vagal flexibility, namely a smaller vagal withdrawal in response to stress, and was coupled with a tendentially larger vulnerability to ventricular arrhythmias compared to LAB counterparts (**Figure 5**). These findings support the view that a low tonic vagal modulation of HR in HAB rats may have determined an inability to flexibly generate adequate cardiac responses to environmental demands. Furthermore, these findings provide a strong basis for future mechanistic investigation aimed at defining, using this rat model, the central neural determinants of vagal

Carnevali and Sgoifo Stress, exercise and heart rate

**FIGURE 4 | Daily rhythms of (A) heart rate, (B) RMSSD, (C) spectral power in HF band (0.75–2.5 Hz) and (D) LF (0.2–0.75 Hz) to HF ratio in 4-month-old Wistar rats selectively bred for either high (HAB,** *n* **= 10) or low (LAB,** *n* **= 10) anxiety-related behavior.** These values (reported as means ± SEM) were obtained by averaging multiple 2-min

control impairment in individuals with high levels of baseline anxiety.

#### **DEPRESSION**

The complex interplay among social stress, depression and autonomic neural modulation of HR was investigated in a rat model of social stress by Wood et al. (2012). Rats exposed to 7 consecutive days of social defeat displayed symptoms of a depressive-like state. Forty eight hours after the final social defeat stress, depressed rats showed reduced HRV and signs of sympathetic predominance (increased LF to HF ratio). It was not determined whether such autonomic imbalance was due to a decrease in vagal tone and/or an increase in sympathetic tone. Similar changes in cardiac autonomic neural outflow were observed in rats submitted to a chronic mild stress model of depression (Grippo et al., 2004, 2006). Importantly, such abnormal modulation of resting HR in depressed rats was partially abolished by fluoxetine treatment (Grippo et al., 2006), suggesting a potential involvement of 5-HT neurotransmission in mediating autonomic changes in depressed individuals. Supporting this view, a study conducted by Hildreth et al. (2008) provided evidence of resting vagal tone impairment in a genetic rat model of depression (i.e., the Flinders-Sensitive

segments acquired every hour during the 12 h-light and 12 h-dark phases, as previously described in detail (Carnevali et al., 2014). Results of two-way ANOVA for repeated measures: group difference for RMSSD (*F* = 16*.*4, *p <* 0*.*01) and HF (*F* = 17*.*48, *p <* 0*.*01) values. Modified from Carnevali et al. (2014).

Line rat) that was related to abnormal serotonergic control of vagal modulation of HR. Human studies indicate that brain 5-HT levels (Mann and Stoff, 1997) and 5-HT receptor function, particularly the 5-HT1A receptors (Parsey et al., 2006), are abnormal in depression. Importantly, it has been demonstrated that chronic social defeat downregulates the 5-HT1A receptor in the rat brain (Kieran et al., 2010). The 5-HT1A receptors play an important role in inhibiting sympathetic neurons located in the medullary raphe area that activates the heart during stress (Nalivaiko, 2006; Nalivaiko and Sgoifo, 2009). In addition to sympathoinhibition, activation of 5-HT1A receptors may have vagomimetic effects (Ngampramuan et al., 2008; Dutschmann et al., 2009; Carnevali et al., 2012a). Indeed, given that vagal preganglionic neurons in the NAmb are under tonic inhibition by GABA-ergic interneurons that express 5-HT1A receptors, activation of such inhibitory 5-HT1A receptors may lead to disinhibition of cardiomotor vagal neurons and consequently increase cardiac vagal tone (Jordan, 2005) (**Figure 1**). Therefore, a decreased density/function of 5- HT1A receptors in these areas may be at the origin of the imbalance in the autonomic neural modulation of resting HR in chronically stressed and depressed individuals. Further work is required in order to determine whether serotonergic mechanisms

do contribute to driving the changes in autonomic modulation of resting HR in rat models of depression.

#### **AGGRESSIVENESS**

The association between personality traits (i.e., aggressiveness) and autonomic neural control of resting HR was the focus of one of our recent investigations (Carnevali et al., 2013a). In this study, high-aggressive (HA) and non-aggressive (NA) rats were selected from a population of adult male Wild-type rats on the basis of their latency time to attack a male intruder in a classical residentintruder test performed on 3 consecutive days (HA: mean attack latency *<*90 s; NA: no overt aggression during the three tests, each lasting 600 s) (Koolhaas et al., 2013). As suggested by RMSSD and HF power indexes, HA rats exhibited lower levels of vagal modulation of resting HR during both the light/inactive and dark/active active phases of the daily cycle (**Figure 6**). However, despite different levels of resting vagal tone, HA and NA rats had similar HR values (**Figure 6**). Similarly to what has been argued above, it has been hypothesized that the reduced cardiac vagal tone observed in HA rats was coupled with a decreased sympathetic influence on sinoatrial node activity compared to NA rats. Consequently, vagal and sympathetic influences were equally balanced in the two groups (LF to HF ratio), leading to similar HR values (**Figure 6**) (Carnevali et al., 2013a). When tested under stress conditions (i.e., restraint and psychosocial stress test), HA and NA rats showed similar tachycardic responses. However, under these conditions HA rats displayed lower vagal antagonism and larger incidence of tachyarrhythmias compared to NA rats (**Figure 7**) (Carnevali et al., 2013a). In addition, injection of beta adrenergic agonist isoproterenol provoked a much larger incidence of ventricular tachyarrhythmias in HA rats compared to NA counterparts (Carnevali et al., 2013a) (**Figure 7**). The results of this study support the view that low levels of vagal modulation of resting HR are associated with arrhythmia susceptibility that may predict vulnerability to cardiac morbidity and mortality. Future mechanistic experiments using this rat model are needed in order to determine the neural determinants of vagal control impairment in aggressive individuals.

**FIGURE 6 | Daily rhythms of heart rate, RMSSD, spectral power in HF band (0.75–2.5 Hz) and LF (0.2–0.75 Hz) to HF ratio in high-aggressive (HA,** *n* **= 10) and non-aggressive (NA,** *n* **= 10) Wild-type rats.** For the 12 hours-light and 12 hours-dark phases, values are reported as means ± s.e.m.

of data obtained by averaging multiple 2-min segments acquired every hour over a period of 6 days. <sup>∗</sup> and # indicate a significant difference between HAB and LAB rats (Student's *t*-test, *p <* 0*.*05 and *p <* 0*.*01, respectively). Modified with permission from Carnevali et al. (2013a).

#### **EFFECTS OF PHYSICAL EXERCISE ON VAGAL MODULATION OF RESTING HR**

Human studies provide evidence that regular physical exercise increases life expectancy in healthy individuals and reduces cardiac-related events in patients with coronary heart disease and heart failure (Rosenwinkel et al., 2001; Buch et al., 2002). The cardioprotective effects of physical training may be mediated, at least in part, by an increase in vagal modulation of resting HR (Crimi et al., 2009). However, central neural mechanisms that underlie these exercise-induced effects on vagal tone are not well understood. Cardiac autonomic adaptation induced by physical training has been subject of research in several carefully designed and rigorously conducted rat studies. These studies reported a clear decrease in resting HR in rats submitted to training with aerobic exercises (i.e., swimming, running) (Medeiros et al., 2004; Rossi et al., 2009; Souza et al., 2009; Sant'Ana et al., 2011; Neto et al., 2013). The effect of the autonomic nervous system in mediating training-induced resting bardycardia was investigated by means of (i) pharmacological approaches (after cardiac muscarinic and adrenergic blockade) and (ii) HRV analysis. Resting bradycardia after exercise training was mainly explained by a vagal effect (Medeiros et al., 2004) or by a reduction in intrinsic HR accompanied by a moderate increase in vagal modulation (indexed by HF spectral power values) (Rossi et al., 2009; Souza et al., 2009; Sant'Ana et al., 2011; Neto et al., 2013).

One possible limitation of these studies is that they applied protocols of forced exercise (treadmill or swimming). In doing so, the effects of physical exercise *per se* on cardiac autonomic function might have been confounded by the effects of accompanying psychological stress (the effects of repeated stress exposure on vagal modulation have been discussed before). On this regard, an elegant study conducted by Beig et al. (2011) investigated the effects of voluntary running on cardiac autonomic function. This study addressed another important issue: that is, whether the effects of voluntary exercise on autonomic function were long-lasting. This question is of major relevance, since loss of exercise-induced cardioprotection has been described soon after exercise cessation (Lin and Horvath, 1972). Rats submitted to a protocol of voluntary exercise exhibited resting bradycardia associated with increased vagal modulation (indexed by HF spectral power values) (Beig et al., 2011). Importantly, these effects persisted 10–12 days after termination of the training protocol, suggesting that voluntary exercise has an enduring effect on resting vagal tone.

Interestingly, voluntary training did not affect stress-induced tachycardia, but augmented resistance to cardiac arrhythmias, as evidenced by the higher doses of the proarrhythmic drug aconitine needed to provoke arrhythmic effects in trained rats (Beig et al., 2011). Such antiarrhythmic effect of exercise training seems to reflect electrophysiological changes (i.e., prolongation of the effective refractory period) induced by long-lasting increase

in vagal outflow in trained rats (Beig et al., 2011). This suggests that the benefits of increased resting vagal tone induced by exercise training may be attributed to (i) an HR-lowering effect and (ii) other parasympathetic effects on the electrophysiological properties of the myocardium. However, the mechanistic links whereby physical training provokes changes in resting vagal outflow are currently unknown. One hypothesis is that exercise might reduce tonic GABAAergic inhibition of neurons in the NTS involved in HR control (**Figure 1**), therefore increasing vagal influences on cardiac peacemaker activity (Mueller and Hasser, 2006). Given that physical exercise increases central 5-HT synthesis (Blomstrand et al., 1989), and that central 5-HT increases vagal modulation in conscious rats (Ngampramuan et al., 2008), it is tempting to speculate that activation of inhibitory 5-HT1A receptors located on GABA-ergic interneurons in the Namb might mediate disinhibition of cardiomotor vagal neurons in that area (**Figure 1**). Further work is required in order to (i) evaluate potential correlations between duration/intensity of voluntary exercise and effects on resting vagal tone and (ii) provide a more comprehensive understanding of central neural mechanisms underlying exercise-induced vagal tone enhancement.

## **CONCLUDING REMARK**

The influence of vagal tone on resting HR is highly variable among individuals. Several psychosocial and physiological conditions are thought to contribute significantly to this large interindividual variability. From the data summarized in this paper, it is quite evident that rat studies can be extremely useful for investigating (i) the effects of stress, psychosocial factors, and physical exercise on vagal modulation of resting HR, and (ii) the consequences of such effects on cardiac stress reactivity and arrhythmia vulnerability. Moreover, the results of these studies have provided preliminary insights into the central neural mechanisms underlying vagal tone impairment/enhancement, which may be exploited by future experiments aimed at developing pharmacological approaches for enhancing vagal activity and cardioprotection.

# **REFERENCES**


as a predictor of mortality in the elderly. *Circulation* 97, 2031–2036. doi: 10.1161/01.CIR.97.20.2031


Wood, S. K., McFadden, K. V., Grigoriadis, D., Bhatnagar, S., and Valentino, R. J. (2012). Depressive and cardiovascular disease comorbidity in a rat model of social stress: a putative role for corticotropin-releasing factor. *Psychopharmacology* 222, 325–336. doi: 10.1007/s00213-012-2648-6

**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Received: 03 February 2014; paper pending published: 01 March 2014; accepted: 10 March 2014; published online: 24 March 2014.*

*Citation: Carnevali L and Sgoifo A (2014) Vagal modulation of resting heart rate in rats: the role of stress, psychosocial factors, and physical exercise. Front. Physiol. 5:118. doi: 10.3389/fphys.2014.00118*

*This article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology.*

*Copyright © 2014 Carnevali and Sgoifo. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# The role of physical activity and heart rate variability for the control of work related stress

#### *Laís Tonello1, Fábio B. Rodrigues 1, Jeniffer W. S. Souza1, Carmen S. G. Campbell 1, Anthony S. Leicht <sup>2</sup> and Daniel A. Boullosa1 \**

*<sup>1</sup> Department of Physical Education, Catholic University of Brasilia, Brasilia, Brazil*

*<sup>2</sup> Institute of Sport and Exercise Science, James Cook University, Queensland, QLD, Australia*

#### *Edited by:*

*Arto J. Hautala, Verve Research, Finland*

#### *Reviewed by:*

*Harri Lindholm, Finnish Institute of Occupational Health, Finland Arja L. T. Uusitalo, Helsinki University Hospital, HUSLAB, Finland*

#### *\*Correspondence:*

*Daniel A. Boullosa, Post-Graduate Program in Physical Education, Catholic University of Brasilia, QS 07, LT1 S/N - Sala 111 - Bloco G, Águas Claras 71966-700, Brasilia, Brazil e-mail: d\_boullosa@yahoo.es*

Physical activity (PA) and exercise are often used as tools to reduce stress and therefore the risk for developing cardiovascular diseases (CVD). Meanwhile, heart rate variability (HRV) has been utilized to assess both stress and PA or exercise influences. The objective of the present review was to examine the current literature in regards to workplace stress, PA/exercise and HRV to encourage further studies. We considered original articles from known databases (PubMed, ISI Web of Knowledge) over the last 10 years that examined these important factors. A total of seven studies were identified with workplace stress strongly associated with reduced HRV in workers. Longitudinal workplace PA interventions may provide a means to improve worker stress levels and potentially cardiovascular risk with mechanisms still to be clarified. Future studies are recommended to identify the impact of PA, exercise, and fitness on stress levels and HRV in workers and their subsequent influence on cardiovascular health.

**Keywords: autonomic nervous system, physical fitness, exercise, allostatic load, employees**

#### **INTRODUCTION**

The adaptive process by which an organism maintains homeostasis is known as allostasis with variable allostatic loads commonly experienced by humans (McEwen and Seeman, 1999; Frodl and O'Keane, 2012). When exposure to chronic stress becomes excessive, the allostatic load experienced may promote important alterations in stress sensitive systems that are intimately linked to the pathophysiology of many diseases (Juster et al., 2010). The most studied of all stress related disorders is cardiovascular disease (CVD), which has been highlighted as the leading cause of mortality worldwide (World Health Organization, 2011). Previously, stress experienced by a person at their place of employment or work has been suggested to substantially increase their CVD risk (Yarnell, 2008; Thayer et al., 2010) with the risk of coronary heart disease being increased by 50% among workers (Kivimaki et al., 2006). As individuals spend most of their daily time at work (e.g., 8–10 h per day), a greater examination of the impact of interventions focusing on managing work-related stress as an important component of allostatic loads is warranted to reduce CVD risk and promote life-long resiliency against abnormal allostatic loading for workers (Juster et al., 2010). Development of healthy workplaces and practices may provide important environments that combat chronic stress and its consequent adverse contribution to the increasing work-related development of disease (Taylor et al., 1997).

Physical activity (PA) and exercise have been extensively recognized as important influences on the relationship between psychosocial stress and CVD (Hamer, 2012) probably because of its influence on physical fitness. Thus, it would be expected that more active individuals, who conversely possess higher physical fitness, would be more resilient to mental stresses (Hamer, 2012). In this regard, interventions that involve PA in conjunction with other beneficial practices (e.g., social support) in the workplace may be very effective for the control of allostatic load at an individual level (Juster et al., 2010). However, the expected greater stress resilience in those individuals with a greater physical fitness has been questioned as the stress-buffering effect of physical fitness has not always been demonstrated (Jackson and Dishman, 2006). This lack of demonstration may be related to methodological constraints of previous studies (Hamer, 2012) with further studies warranted to elucidate the important role that PA and exercise could have on workers cardiovascular health, potentially as an important stress-buffer.

Heart rate variability (HRV) is an easy and non-invasive tool for the assessment of variations in beat-to-beat intervals and autonomic nervous system activity with HRV obtained by linear methods within the domains of time and frequency analyses, and nonlinear methods (Task Force, 1996). HRV has been studied extensively in regards to CVD (Vrijkotte et al., 2000; Kivimaki et al., 2006; Yarnell, 2008; Thayer et al., 2010; Frodl and O'Keane, 2012), exercise (Proper et al., 2002; Kiviniemi et al., 2007; Buchheit et al., 2010; Boullosa et al., 2012), and stress (Hjortskov et al., 2004; Collins et al., 2005; Yarnell, 2008; Loerbroks et al., 2010; Uusitalo et al., 2011). The use of HRV as a practical monitoring tool for allostatic load though has been scarce and may provide a simple instrument for workers in the workplace. Greater HRV has been related to lower cardiovascular risk (Kiviniemi et al., 2010), greater physical fitness and responsiveness to aerobic training (Hautala et al., 2010), greater PA levels in workers (Rennie et al., 2003), and lower work related stress in workers (Uusitalo et al., 2011). Collectively, these and other previous studies (Orsila et al., 2008; Thayer and Lane, 2009; Hynynen et al., 2011) emphasize the importance of HRV for the assessment of cardiovascular stress in the workplace with PA and fitness potentially enhancing HRV control. Therefore, the aim of this mini-review was to review the current literature regarding HRV, PA, fitness, and workplace stress to better delineate the current understanding and potential for future studies.

# **MATERIALS AND METHODS**

An extensive search of relevant studies listed within the National Library of Medicine (PubMed) and ISI Web of Knowledge databases over the past 10 years (2003–2013) was conducted. The following combination of terms was utilized during the search: (exercise OR PA OR physical fitness) AND (workers OR occupational OR work OR job) AND cardiovascular stress (**Figure 1**). Inclusion criteria included: original articles; written in English; study population consisted of workers; included assessments of PA and/or exercise; assessment of work-related stress, and cardiac autonomic control monitored via HRV. The review, evaluation and selection of articles based upon inclusion criteria were carried out independently by three of the authors.

# **RESULTS**

After initial searches based on the combination of specific terms were performed, there were 228 and 383 articles identified within PubMed and ISI Web of Knowledge databases, respectively (**Figure 1**). The titles and abstracts of these 611 articles were reviewed with 593 excluded as not meeting the inclusion criteria. Subsequently, 18 articles were obtained in full and examined further for details according to the inclusion criteria (**Figure 1**). Based upon these criteria, only seven articles were included in this study (see **Table 1**). These studies examined autonomic function via HRV for job stress evaluation, however studies included a variety of analyses, different scheduling of R-R recordings, and various populations that make comparisons difficult. Nevertheless, several studies (Chandola et al., 2008; Clays et al., 2011; Uusitalo et al., 2011) reported a reduced HRV in workers reporting workrelated stress. Interestingly, only Chandola et al. (2008) reported a positive association between work stress and CVD possibly as a consequence of the negative influence of stress on health behaviors (i.e., reduced PA and poor diet). Meanwhile, others (Eller

et al., 2011b; Lindholm et al., 2012) did not find any direct relationship between stress and HRV indices although a reduced HRV was exhibited during an imbalance between effort and reward, and in those workers experiencing irregular shiftwork, important sources of work-related stress. Likewise Melville et al. (2012) and Cheema et al. (2013) reported no relationship between HRV and stress with reductions in stress and anxiety following workplace yoga interventions accompanied by no changes in HRV.

All included studies assessed the level of PA or exercise with most simply reporting the level in a descriptive function. Only one study (Cheema et al., 2013) examined PA or exercise as an independent variable in a longitudinal design of work related stress and autonomic control of HR. Surprisingly, this recent study (Cheema et al., 2013) reported a worsening of autonomic control of HR (i.e., decrements in HRV indices) for the yoga intervention group vs. the control group. Paradoxically, the participants of this study exhibited improvements in some fitness characteristics and a reduced anxiety state after the yoga intervention suggesting an effective intervention but with an unexpected reduced HRV (Cheema et al., 2013). Previously, Melville et al. (2012) examined the acute effects of both yoga and meditation interventions on perceived stress and HRV with no significant changes in HRV during the acute post-intervention period (i.e., 15 min), despite significant reductions in perceived stress after both interventions.

# **DISCUSSION**

This review suggests that autonomic function could be a simple and effective measure to identify workplace stress. Additionally, a single study with longitudinal workplace PA interventions reported improved worker stress levels without significant positive changes in cardiac autonomic activity, a mechanism known to be cardioprotective. Based upon the current few studies of PA and stress, and equivocal relationships between stress and cardiovascular health in workers, we encourage more studies involving varying PA and exercise interventions in the workplace to better examine the benefits of PA and exercise on both stress and HRV.

Overall, the current review provides further support of the applicability of cardiac autonomic function monitoring for work related stress. Factors related to adverse working conditions such as excessive effort (Vrijkotte et al., 2004), effort-reward imbalance (Eller et al., 2011a; Uusitalo et al., 2011), over commitment (Vrijkotte et al., 2004; Lindholm et al., 2012), irregular shift work (Lindholm et al., 2012), and work stress (Chandola et al., 2008) were significantly related to reduced cardiac autonomic function. Therefore, HRV monitoring may provide a simple and non-invasive assessment of stress and allostatic load in working environments that employers could utilize in the efficient management of employees. The recent systematic review of Jarczok et al. (2013) provides further support for HRV use in employee management with adverse psychosocial work conditions reported to be negatively associated with autonomic nervous system function as indexed by HRV.

It should be pointed out that within the current review, a variety of different HRV analyses were employed that limited comparisons between studies. The selection of HRV methodology is an important issue as tool sensitivity may compromise

**Table 1 | Summary** 


*(Continued)*


**Table 1 | Continued**


*Abbreviations: RMSSD, root mean squared differences of successive NN intervals; pNN50, % differences between adjacent RR intervals N50 ms; SDNN, standard deviation of normal-to-normal (NN) intervals;CHD, Coronary Heart Disease; LF, Low Frequency; LF/HF, Low Frequency to High Frequency Ratio; HR, Heart Rate; HF, High Frequency; JSQ, Job Stress Questionnaire; ISW, Irregular shift work; RDW, Regulardaytime work; ERI, Effort-reward imbalance; IQR, interquartile range. HRV, heart rate variability; ECG, electro cardiogram; lnHR, normalized heart rate; ln(LF/HF), normalized Low Frequency to High FrequencyRatio; TP, total power; JDI, Job Descriptive Index; JIG, Job in General scale.*

tended to improve the push-up test

(*p* = 0*.*07) vs. control.

the detection of cardiac autonomic responses (Hynynen et al., 2011). Aspects such as variations in HRV measures examined (e.g., linear, non-linear, etc.), data analysis (e.g., supine, seated, Fast Fourier Transform, Autoregression, etc.) as well as the quality of data (e.g., degree of ectopy/artifact, sampling rate, recording length, etc.) make comparisons between studies challenging (Jarczok et al., 2013). For instance, Hynynen et al. (2011) reported that HRV during an orthostatic test upon wakening may be more useful for the analysis of stress in real life compared to night time recordings. In contrast, HRV measures recorded during both work and at night were reported to be also sensitive markers of mental stress alterations at work (Hjortskov et al., 2004; Yarnell, 2008). Further studies should determine the optimal HRV methods for detecting cardiac autonomic stress related adaptations. Factors such as the HRV measure (i.e., time or frequency domain, nonlinear, etc.), time, and duration of recordings and body posture (Young and Leicht, 2011; Boullosa et al., 2012) may play an important role that requires further clarification. Given the possible influence of factors external to work related stressors during long recordings, the use of laboratory controlled recordings should be included in further studies for a better evaluation of autonomic control of HR (Lombardi and Stein, 2011). Additionally, other cardiac autonomic indices (i.e., heart rate recovery; HRR, post-exercise) could also be employed as a delayed HRR was reported in individuals with high levels of stress and depression (Gordon et al., 2012). Consequently, the simple assessment of HRV and HRR may provide a comprehensive evaluation of the cardiac autonomic health of workers and prognosis for work-related stress and responsiveness to exercise training (Huovinen et al., 2011).

Another important restriction of studies within the current review was the different assessment tools for stress levels or psychosocial characteristics of work. For instance, the following tools were utilized in the current review studies: effort reward imbalance model (ERI-model) (Eller et al., 2011a; Uusitalo et al., 2011), Job Descriptive Index (JDI); Job in General (JIG) scale (Cheema et al., 2013), job stress questionnaire (Clays et al., 2011), stress questionnaire (Lindholm et al., 2012), job-strain questionnaire (Chandola et al., 2008), and a 100 mm visual analog scale (Melville et al., 2012). While each of the aforementioned tools has been used to assess stress at work, differences in specificity, reliability, and validity of these tools and their relationship with cardiac autonomic activity made comparisons difficult. Standardized use of stress assessment tools may help clarify the relationship between stress and cardiac autonomic activity. For example, Jarczok et al. (2013) highlighted that measurements of stress could be grouped when evaluated by ERI and JDC questionnaires with these questionnaires demonstrating a significant association between high strain/stress work and decreased vagal tone. Identification of the ultimate assessment tool may help to better understand the relationship between cardiovascular stress and workers health. Further, this tool may also overcome the confounding variables that affect the measurement of cardiac autonomic control such as age, sex, disease, caffeine intake, PA, smoking, and alcohol consumption (Jarczok et al., 2013).

One disappointing aspect of the current review was the low incidence of exercise and PA with only two studies involving exercise interventions (i.e., yoga or yoga and meditation) (Melville et al., 2012; Cheema et al., 2013) in the workplace. Further, these recent studies (Melville et al., 2012; Cheema et al., 2013) revealed contradictory results regarding the relationship between exercise interventions, stress, and HRV. The remaining studies (Vrijkotte et al., 2004; Clays et al., 2011; Eller et al., 2011a; Uusitalo et al., 2011; Lindholm et al., 2012) simply recorded PA as a factor of consideration in their analyses. Moreover, only one study objectively recorded the levels of PA via accelerometry (Uusitalo et al., 2011) with the remaining studies utilizing questionnaires for this purpose. Further, very few examined physical fitness with those studies assessing physical fitness (Clays et al., 2011; Uusitalo et al., 2011; Cheema et al., 2013) and not the potential role of this factor on stress responses and HRV. Therefore, the results of our review confirmed the variety of factors utilized in past studies with future studies recommended to examine objective measures of exercise and PA as these have been previously associated with HRV levels (Rennie et al., 2003; Hautala et al., 2010; Takada et al., 2010).

Previous studies have observed a weak and inverse association between job strain and leisure time PA indicating the potentially important role of PA for managing work related stress (Loerbroks et al., 2010). Previously, Choi et al. (2010) reported that having on-the-job learning opportunities and decision authority about their tasks may be conducive to increasing active leisure time PA in middle-aged US workers. While these previous studies have analyzed the impact of work conditions on leisure time PA, other studies have analyzed the impact of complex interventions including exercise in the workplace. For instance, Eriksen et al. (2002) previously demonstrated the greater effectiveness of complex interventions including PA than more focused non-PA interventions (e.g., stress management alone). Subsequently, Tveito and Eriksen (2009) reported in a randomized controlled study that a program consisting of physical exercise, stress management training, health information and an examination of the participants' workplace, promoted improvements in health, physical fitness, muscle pain, stress management, maintenance of health, and work situation. Overall, these previous studies highlighted the appropriateness of complex, exercise or PA based interventions at work for stress management and other health related outcomes. Paradoxically, in our focused review we found only two studies with workplace exercise interventions. Melville et al. (2012) evaluated the impact of a brief intervention (e.g., 15 min of yoga) at work on the acute cardiovascular responses (i.e., HR, HRV, and blood pressure) and perceived stress. Interestingly, this study reported a higher HRV during yoga and meditation interventions that returned to baseline immediately after the session with a reduction of perceived stress for 15 min post-intervention (Melville et al., 2012). Although the results of this study were promising, these results, within a small time of evaluation, should be considered with caution, especially due to the low intensity and duration of yoga practice. However, such an exercise intervention could be promising for the control of acute workrelated stress responses (Hamer, 2012). Another study (Cheema et al., 2013) also evaluated the effectiveness of a workplace program based of hatha yoga (10 weeks, 50 min at lunch time, 3 times per week) on physiological stress (HRV) and associated health-related outcomes in a cohort of office workers. Despite the reduction of anxiety and fitness improvements among those with high adherence rates, the decrease in HRV after the intervention period raised concerns about the effectiveness of this exercise modality for the cardiac health of workers. Further studies with other exercise and PA interventions, which are different concepts (Caspersen et al., 1985), may clarify the most appropriate modality and dose-response for the control of stress and associated CVD in workers.

In the current review, most studies were cross-sectional with only two involving exercise interventions (yoga and meditation). Based on this review, there is a notable lack of studies comparing longitudinal interventions and the cause-effect relationships between physical fitness, exercise, and PA, and cardiovascular stress (Hamer, 2012). Moreover, identification of barriers and facilitators for PA or exercise at work (Renton et al., 2011; Leicht et al., 2013) may assist with the development of the most appropriate interventions to assist workers based on their employment demands. For instance, emergency physicians with different shifts (14 or 24-h) exhibited significant differences in stress levels (Dutheil et al., 2012). Such factors may assist in developing the best interventions and any potential changes in working environments (Conn et al., 2009). Although the use of complex interventions including exercise, psychological and nutritional habits have been previously recommended (Carson et al., 2010; Strijk et al., 2012), the inclusion of sole exercise and/or PA interventions may determine the appropriate dose-response (Vanhees et al., 2012). This would be very important as the levels of physical fitness that protect against CVD when considering specific job demands remain undefined (Huang and Acevedo, 2011). Moreover, CVD risk factors and stress exhibit a bidirectional relationship as stress promotes changes in behavior and pathophysiological parameters and vice versa (Huang and Acevedo, 2011) that should be considered in both prevention and treatment interventions (Stults-Kolehmainen, 2013). Thus, stressed workers may be less inclined to perform PA contributing to a greater deterioration of their health and progression of pathophysiological conditions (Azevedo et al., 2012; Silva and Barreto, 2012). Identification of barriers and motivators within workers would be an important aspect for designing effective and individualized intervention programs that also considers PA levels and sitting time derived from specific job activities (Leicht et al., 2013).

In summary, the results of the current review suggested a relationship between work related stress and autonomic control of HR. However, the extent of the positive benefits of both PA and exercise on both stress and HRV remains to be elucidated. Thus, future research, especially longitudinal studies, with different work categories and samples are needed to better understand how different levels of physical fitness, energy expenditure and exercise modes prevent or minimize the allostatic load and subsequent stress on workers and their impact on cardiovascular health.

#### **ACKNOWLEDGMENTS**

Laís Tonello was partially supported by a scholarship from CNPq.

#### **REFERENCES**


*Based Complement. Alternat. Med.* 2012, 501986. doi: 10.1155/2012/ 501986


heart rate variability. *Hypertension* 35, 880–886. doi: 10.1161/01.HYP. 35.4.880


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Received: 30 October 2013; accepted: 04 February 2014; published online: 21 February 2014.*

*Citation: Tonello L, Rodrigues FB, Souza JWS, Campbell CSG, Leicht AS and Boullosa DA (2014) The role of physical activity and heart rate variability for the control of work related stress. Front. Physiol. 5:67. doi: 10.3389/fphys.2014.00067*

*This article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology.*

*Copyright © 2014 Tonello, Rodrigues, Souza, Campbell, Leicht and Boullosa. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# The interplay between stress and physical activity in the prevention and treatment of cardiovascular disease

# *Matthew A. Stults-Kolehmainen\**

*Department of Psychiatry, Yale Stress Center, Yale University Medical School, New Haven, CT, USA \*Correspondence: matthew.stults-kolehmainen@yale.edu*

#### *Edited by:*

*Daniel Boullosa, Universidade Católica de Brasília, Brazil*

**Keywords: exercise, distress, mental health, physical activity, stress, psychological, cardiovascular diseases**

Cardiovascular disease (CVD) continues to menace developed and developing nations alike. However, in the United States alone, 40.5% of the population is projected to have CVD or a closely related disorder by 2030 (Heidenreich et al., 2011). Within the same time frame, costs associated with CVD are expected to balloon up to \$818 billion (Heidenreich et al., 2011). These societal strains are increasing parallel to rising mental health disorders and psychosocial stress (Centers for Disease Control, 2011). In this context, stress refers to both physical and emotional challenges, some of which may be transient and rather innocuous (e.g., final examinations, strenuous exercise) with potential positive adaptations. Other stress may be chronic and uncontrollable (e.g., caregiving for a loved one with a terminal illness), which may produce excessive wear and tear, resulting in lasting and harmful insults to one's physical and mental health (McEwen, 2007). The American Psychological Association (APA) reports that 72% of Americans perceived that their stress increased or held steady from 2007 to 2011. Fourth fifths of this population reported an increase from 2010 to 2011 (American Psychological Association, 2012).

Unsurprisingly, there are wellestablished connections between stress and CVD. Indeed, stress provides an independent contribution to CVD outcomes (Vitaliano et al., 2002). Stress is linked to the pathogenesis of coronary heart disease (CHD) (Rozanski et al., 1999), incidence of acute myocardial infarctions (Rosengren et al., 2004) and worse survival from cardiac events (Kivimaki et al., 2002; Milani and Lavie, 2009). In one eye opening investigation, Milani and Lavie (2009) found that patients with high psychosocial stress in cardiac rehabilitation were almost 4 times as likely to die as those with low stress (22 vs. 5%). The experience of distress, the emotional outcome of stress, also exacerbates the morbidity associated with CVD (Centers for Disease Control, 2011). Those who are objectively stressed, such as caretakers of those with chronic conditions, have higher resting heart rates, blood pressure, and greater incidence of metabolic syndrome (Vitaliano et al., 2002). Individuals reporting higher levels of stressful life events have higher scores on risk factors for CVD, such as smoking (Ansell et al., 2012), systemic inflammation (Puustinen et al., 2011) and obesity (Sinha and Jastreboff, 2013). They are nearly twice as likely to start using anti-hypertensive medication over time compared to less stressed individuals (Rod et al., 2009). Furthermore, those with CVD are often stressed by their condition—adjusting to a life with a long-term hardship (Bodenheimer et al., 2002). This demonstrates that stress both contributes toward disease and, reciprocally, emanates from the experience of disease, which leaves little doubt that mental health interventions are needed throughout the progression and treatment of CVD.

Given these findings, an effective prevention and/or treatment plan for CVD would also target stress, but would minimize additional side effects and healthcare costs. To this end, exercise interventions have proved effective. In terms of CVD pathology, exercise improves the odds of recuperating from stroke, reduces hypertension and diminishes symptoms of heart failure and CHD—in each case with few serious side effects (Kujala, 2009). Recent evidence finds that exercise is just as effective as or even more effective than medications. For instance, Naci and Ioannidis (2013) recently found that exercise had a stronger effect than anticoagulants and antiplatelets in the treatment of stroke. Moreover, there is convincing evidence that those who exercise are much less likely to develop CVD over time (Lee et al., 2012; Matheson et al., 2013). The same is true for depression and post-traumatic stress disorder (PTSD) (Leardmann et al., 2011). In a large military cohort, those who exercised at least 20 min of vigorous activity two times a week were *>*40% less likely to develop new onset PTSD over *>*3 year period (Leardmann et al., 2011). Among those experiencing difficulties with stress, aerobic and resistance exercise has been effective in reducing stress-induced cardiac reactivity (King et al., 2002; Faulk and Bartholomew, 2012) and perceived stress (Wilcox et al., 2008), particularly when paired with healthy dietary changes (Imayama et al., 2011) and behavioral modification programming (Atlantis et al., 2004). Given the strong association between stress and CVD, is it possible that part of the health-enhancing power of exercise stems from its ability to mitigate the effects of stress (Milani and Lavie, 2009)? Extensive data supports the thought that physical activity (PA) buffers the relationship between stress and physical health problems (Gerber and Pühse, 2009; Emeny et al., 2012; Hamer, 2012; Rueggeberg et al., 2012). In other words, at high levels of stress, greater PA is associated with better health (Rueggeberg et al., 2012). Additionally, those who are physically fit are more resilient to the effects of stress, such as high work demands, resulting in less heart disease and associated mortality (Holtermann et al., 2010).

Stults-Kolehmainen Interplay between stress and PA

Data from our own laboratory show that higher levels of moderate to vigorous exercise are associated with fewer complaints of cardiovascular problems particularly when stress is low (Stults-Kolehmainen et al., accepted). Under high levels of stress, however, higher levels of exercise were not related to better health, contradicting previous research (Rueggeberg et al., 2012). While PA—and more specifically, exercise—certainly have a salubrious effect on both psychosocial stress and CVD, and lifestyle interventions increasingly emphasize all forms of PA, I was left to wonder about the possible limitations of these health behaviors. One possible explanation for these discrepancies is that various forms of PA may have differential influence on the stress and CVD relationship (Fredman et al., 2006). PA, by definition, is any movement that results in energy expenditure, including occupational and spontaneous forms of locomotion (Garber et al., 2011). This is important because stressed populations, particularly laborers, may engage in moderate to high levels of occupational activity. Exercise, however, is typically performed with the intent of increasing physical fitness. Aside from athletes, military personnel and physical education students, it is usually considered recreational and is completed during one's leisure time. Therefore, it is possible that quantifications of exercise simply capture the luxury of having more time for rest and relaxation (Iwasaki et al., 2001). A preponderance of evidence would support the observations that those with less leisure are more frequently stressed, but also that those who are stressed typically have less leisure time (Fredman et al., 2006; Lutz et al., 2007).

Consequently, in the pursuit of better health, stress may have the upper hand over one's ability to engage in healthful levels of PA (Salmon, 2001). In fact, Rafer Lutz established that stress has a stronger effect on PA than the reverse order of influence (Lutz et al., 2007). This was the first and only prospective design with the specific aim of untangling these effects. Lutz, John Bartholomew and I later published data that suggested that those who are in earlier stages-of-change for exercise (i.e., pre-contemplators, contemplators) are most vulnerable to this effect (Lutz et al., 2010). This association does not appear to be limited to non-habituated exercisers, however. In our recent systematic review, Rajita Sinha and I found that *>*85% of prospective studies examining the association of stress and PA/exercise reported an inverse relationship between these two constructs. In other words, the experience of high mental stress predicted less PA. Several studies used sophisticated designs to objectively capture periods of greater stress and examine future levels of PA or exercise (Griffin et al., 1993; Steptoe et al., 1996; Vitaliano et al., 1998; Roemmich et al., 2003; Oaten and Cheng, 2005; Smith et al., 2005; Sherman et al., 2009). However, only one investigation employed a true experimental design and manipulated an acute mental stressor (Roemmich et al., 2003). In this study, children participated in two conditions, the order of which was randomly selected. In the stress condition, these subjects were required to give a prepared talk on a social topic. This task induced a moderate to strong stress response, which was followed by the opportunity to be either physically active (riding a bicycle) or remain sedentary. Behavioral responses were compared to changes observed within a passive reading control condition. While responses subjective stress were not related to changes in exercise (*r* = −0*.*19), children were less likely to expend energy after the stress condition compared to the control (a 21% difference in total minutes of activity). Total energy expenditure was also lower in the stress condition. Also examining this outcome, Smith et al. (2005) found approximately a 1000 kcal difference of weekly activity in a cohort of parents with a child who recently received a cancer diagnosis vs. parents of a healthy control child. Effects sizes for this association were large: 1.71 (Cohen's d) at diagnosis and 1.13 at a 3-month follow up. Such differences were due to less PA and more sedentary behavior, such as TV viewing. Given the long waits in doctors' offices to which these parents are subject, this is not surprising. In another longitudinal case-control study, caregivers had more stress and less frequent exercise bouts over two time points compared to matched controls (Vitaliano et al., 1998).

Following students before and during a final examinations period is a highly convenient method to observe potential influences of stress on health behaviors (Griffin et al., 1993; Steptoe et al., 1996; Oaten and Cheng, 2005; Sherman et al., 2009). In two cases, control groups have been utilized to determine whether other factors could explain changes in PA, such as the weather (Steptoe et al., 1996; Oaten and Cheng, 2005). These studies have identified significant declines in PA compared to controls during examinations, and such changes are due to less frequency, duration, and perceived ease of exercise. In an interesting permutation on this design, Sherman (Sherman et al., 2009) found that PA declined across a 14-day period that terminated with students' most stressful examination—a medium effect (*p* = <sup>0</sup>*.*03, *<sup>d</sup>* <sup>=</sup> <sup>0</sup>*.*62; <sup>η</sup><sup>2</sup> <sup>=</sup> <sup>0</sup>*.*26). Unfortunately, none of these studies looked at changes in PA after the completion of a stressful examination period or over multiple cycles of examinations, which would provide an additional insight on the dynamic effects of stress on exercise behavior. A rebound in PA after a stressful frame of time would complement existing data. Nevertheless, when viewed collectively, the published data substantiate the notion that both acute and transient or chronic life stressors have an impact on facets of PA.

These findings have direct implications for clinicians interested in improving CVD outcomes. Exercise is effective for treating both CVD and stress, but the patient would benefit from a treatment plan that goes beyond a prescription of exercise and "usual care." Navigating the complexities of a more active lifestyle is stressful in itself, particularly for the uninitiated (Stults-Kolehmainen and Sinha, 2013). Second, those who are chronically stressed experience both poor recovery from exercise and derive less affective benefits, placing them at risk for dropout early after initiation of an exercise routine (Stults-Kolehmainen and Bartholomew, 2012). Exercise cannot help those who do not adhere to a structured regimen, and stressed populations are at risk for drop out and poor compliance with their doctor's prescriptions. For instance, depressed cardiac rehabilitation patients are less likely to adhere to an exercise program; coming to fewer sessions and dropping out at high rates (Glazer et al., 2002). As a metaphor, the sun-scorched plant benefits from water, but inordinate heat will also evaporate the new moisture. Only when shade is introduced can full restoration occur. In a similar light, the human organism needs relief from stress to effectively take advantage of exercise. The combination of PA and stress management would likely have a synergistic effect. In a randomized controlled trial, Blumenthal et al. (2005) demonstrated that exercise training and stress management groups significantly improved flow-mediated arterial dilation and other clinical outcomes compared to a third group receiving only usual care. They did not test the utility of both exercise and stress management combined, which has been successfully tested by Ornish et al. (1998). Their lifestyle intervention resulted in 50% reductions in the re-occurrence of cardiac events. Given this progression, a logical follow-up study should fully intertwine stress management procedures, such as mindfulnessbased stress reduction (MBSR) (Ludwig and Kabat-Zinn, 2008), as part of the exercise regimen. Mindfulness practices can be conducted during movement (i.e., controlled breathing and contractions with proper alignment), during rest periods (e.g., focusing on changing internal states) and directly following a workout (e.g., sensing stillness and the process of unwinding). Such a combination may enhance physical recovery (Stults-Kolehmainen and Bartholomew, 2012), boost affective valence (increase positive emotions), and reinforce exercise behavior resulting in more effective treatment with time and cost savings.

Stress is a common denominator underlying CVD and the lack of PA. It independently contributes to pathologies of the cardiovascular system, exacerbates distress associated with these problems and impedes one's progress toward a healthier lifestyle. As this is a complicated set of relationships, I would argue that novel therapeutic regimens are needed to address the manifestation of stress at multiple levels. Fully melding exercise interventions with proven stress management techniques, such as MBSR, would be a useful strategy.

### **REFERENCES**


and mood on health practices. *J. Behav. Med.* 16, 643–661. doi: 10.1007/BF00844724


stressed: stages of change and the stress-exercise participation relationship. *Psychol. Sport Exerc.* 11, 560–567. doi: 10.1016/j.psychsport.2010.06.005


*Received: 31 October 2013; accepted: 08 November 2013; published online: 27 November 2013.*

*Citation: Stults-Kolehmainen MA (2013) The interplay between stress and physical activity in the prevention and treatment of cardiovascular disease. Front. Physiol. 4:346. doi: 10.3389/fphys.2013.00346*

*This article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology.*

*Copyright © 2013 Stults-Kolehmainen. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Mindfulness may both moderate and mediate the effect of physical fitness on cardiovascular responses to stress: a speculative hypothesis

#### *Marcelo M. P. Demarzo1 \*, Jesús Montero-Marin2, Phyllis K. Stein3, Ausiàs Cebolla4,5, Jaime G. Provinciale6 and Javier García-Campayo2*

*<sup>1</sup> Department of Preventive Medicine, "Mente Aberta" – Brazilian Center for Mindfulness and Health Promotion, Universidade Federal de São Paulo, São Paulo, Brazil*

*<sup>2</sup> Department of Psychiatry, Instituto Aragonés de Ciencias de la Salud de la Red de Actividades Preventivas y de Promoción de la Salud (REDIAPP), Universidad de Zaragoza, Zaragoza, Spain*

*<sup>3</sup> Heart Rate Variability Laboratory, Division of Cardiology, Internal Medicine Department, Washington University School of Medicine, Saint Louis, MO, USA*

*<sup>4</sup> Department of Basic and Clinical Psychology and Psycholobiology, Universitat Jaume I, Castellón, Spain*

*<sup>5</sup> Centro de Investigación Biomédica en Red - Fisiopatología de la Obesidad y Nutrición (CIBERobn), Madrid, Spain*

*<sup>6</sup> Human Centered Technology Department, Instituto en Bioingeniería y Tecnología Orientada al Ser Humano, Universitat Politècnica de València, Valencia, Spain*

#### *Edited by:*

*Daniel Boullosa, Universidade Católica de Brasília, Brazil*

#### *Reviewed by:*

*Laurinda Abreu, Independent Researcher, Spain Matthew A. Stults-Kolehmainen, Yale University Medical School, USA*

#### *\*Correspondence:*

*Marcelo M. P. Demarzo, Department of Preventive Medicine, "Mente Aberta" – Brazilian Center for Mindfulness and Health Promotion, Universidade Federal de São Paulo, Rua Botucatu, 740. São Paulo-SP 04023-900, Brazil e-mail: demarzo@unifesp.br*

The psychological construct of mindfulness refers to an awareness that emerges by intentionally paying attention to the present experience in a non-judgmental or evaluative way. This particular quality of awareness has been associated to several indicators of physical and psychological health, and can be developed using mindfulness-based interventions (MBIs), and therefore MBIs have been successfully applied as preventive and complementary interventions and therapies in medicine and psychology. Together with quiet sitting and lying meditation practices, mindful physical exercises such as "mindful walking" and "mindful movement" are key elements in MBIs and couple muscular activity with an internally directed focus, improving interoceptive attention to bodily sensations. In addition, MBIs seem to share similar mechanisms with physical fitness (PF) by which they may influence cardiovascular responses to stress. Based on these facts, it is feasible to raise the question of whether physical training itself may induce the development of that particular quality of awareness associated with mindfulness, or if one's dispositional mindfulness (DM) (the tendency to be more mindful in daily life) could moderate the effects of exercise on cardiovascular response to stress. The role of mindfulness as a mediator or moderator of the effect of exercise training on cardiovascular responses to stress has barely been studied. In this study, we have hypothesized pathways (moderation and mediation) by which mindfulness could significantly influence the effects of PF on cardiovascular responses to stress and discussed potential practical ways to test these hypotheses.

**Keywords: mindfulness, moderator effects, mediational model, physical fitness, cardiovascular system, psychological stress, mindful exercise**

# **MINDFULNESS IN MEDICINE AND PSYCHOLOGY**

Interest in mindfulness has increased exponentially in recent decades in academic and clinical contexts (Dimidjian and Kleiber, 2013). The psychological construct of mindfulness refers to an awareness that emerges by intentionally paying attention to the present experience in a non-judgmental way (Kabat-Zinn, 2003). In other words, a "mindful" mind brings together attentional and attitudinal features at the same time, self-regulating attention toward present-moment direct experiences and attitude in a non-judgmental tone toward internal and external phenomena (physical, affective and behavioral).

This particular quality of non-evaluative awareness can improve one's physical and psychological health status, and therefore several approaches generally called "mindfulness-based interventions" (MBIs) have been developed and tested in the last 40 years (Cullen, 2011). Evidence about their efficacy and

effectiveness in improving mindfulness and health is rapidly accumulating (Khoury et al., 2013).

MBIs, e.g., "Mindfulness-based Stress Reduction" (MBSR) and "Mindfulness-based Cognitive Therapy" (MBCT), are generally 8-weeks group-based interventions that mix mindfulness practices (mostly derived from ancient eastern meditative techniques) with contemporary cognitive and behavioral approaches to improve one's mindfulness levels and capabilities to adaptively manage stressful life events (Cullen, 2011). Moreover, MBIs have been successfully applied as preventive and adjuvant therapies in patients with depression, anxiety, chronic pain, cardiovascular disorders, cancer, and other non-communicable diseases (Bonadonna, 2003; Edelman et al., 2006; Kuyken et al., 2008; Fortney and Taylor, 2010; Khoury et al., 2013). In addition, MBIs have been adapted to special targeted population (Carson et al., 2006; Duncan and Bardacke, 2010; Cullen, 2011), including the sports psychology field, in this case aiming to improve athletes' psychological well-being and to enhance their performance (Gardner and Moore, 2012).

It has been suggested that MBIs result in improvements in autonomic and central nervous systems regulation, attention control, emotional and behavioral self-awareness and regulation, self-compassion, resilience, dispositional flow, and body awareness, and these mechanisms may account for their efficacy (for reviews about mindfulness mechanisms see Hölzel et al., 2011; Gardner and Moore, 2012).

#### **MINDFULNESS: TRAIT AND STATE**

Several questionnaires and scales to measure mindfulness trait or state have been developed in the last two decades (Park et al., 2013), which allow systematic investigation in this field. It has been observed that to some degree the particular type of awareness associated to mindfulness may be innate (Garland et al., 2013).

Trait or dispositional mindfulness (DM), the tendency to be more mindful in daily life (Garland et al., 2013), seems to be a sum of individual genetics and life experiences, and DM is not necessarily related to having participated in an MBI or practicing meditative exercises. DM can be measured by scales addressing mindfulness trait, e.g., the Mindful Attention Awareness Scale (MAAS) (Brown and Ryan, 2003) and the Five Facets Mindfulness Questionnaire (FFMQ) (Bohlmeijer et al., 2011), and it has been significantly related to several indicators of psychological and physical health including: higher levels of positive affect, improvement in personal stress management skills, and in adaptive emotional regulation (Hayes-Skelton and Graham, 2013; Khoury et al., 2013; Garcia-Campayo et al., 2014).

As expected, DM can be also modulated by participating in MBIs and practicing specific mindfulness techniques such as several types of meditative practices (Garland et al., 2010; Park et al., 2013). These interventions can induce transient increases in the mindfulness state, and may promote long-term improvements in DM if applied in a regular basis (Soler et al., 2014). The "mindfulness state," i.e., mindfulness addressed as a statelike mental behavior, which is context-dependent, transient and variable, is also measureable, and few scales have been developed for this purpose (Tanay and Bernstein, 2013).

# **MINDFULNESS AND CARDIOVASCULAR RESPONSES TO STRESS**

Chronic stress is a well-known risk factor for cardiovascular diseases (CVD). Physical and psychological stressors can provoke non-adaptive stress-induced cardiovascular responses marked by dysfunctional cortisol and catecholamine releasing, systemic inflammation, oxidative stress, and unbalanced autonomic nervous system, leading to endothelial dysfunction, increased blood pressure, among others factors, which chronically may induce hypertension and CVD (Koolhaas et al., 2011; Huang et al., 2013; Stoner et al., 2013). Current evidence suggests that MBIs may buffer responses to stressors by mechanisms direct or indirectly related to these biomarkers of non-adaptive cardiovascular responses to stress.

A recent study showed that an MBSR program (the original program that have influenced subsequent types of MBIs) had a favorable influence both on biomarkers of stress regulation, such as cortisol secretion and sleep (Brand et al., 2012). In this regard, stress-sleep connection may be an important mechanism influencing responses to stress (Demarzo and Stein, 2012). In the same direction, another study showed that an MBI resulted in increases in both objectively- and subjectively-measured sleep continuity in patients using anti-depressant medication (Britton et al., 2012).

MBIs and other behavioral interventions designed to reduce emotional reactivity may be of therapeutic benefit in chronic inflammatory conditions (Rosenkranz et al., 2013). Even a brief mindfulness intervention in the workplace may be an effective and probably cost-effective way to reduce systemic inflammation (Malarkey et al., 2013). MBSR has also shown promise as a novel treatment approach for reducing loneliness-related proinflammatory gene expression in older adults (Creswell et al., 2012). Moreover, analysis of oxidative stress levels in people who meditate indicated that meditation correlates with lower oxidative stress and higher melatonin levels, a well-known antioxidant agent (Martarelli et al., 2011).

Another mechanism that may link mindfulness to cardiovascular responses to stress is influencing on the autonomic nervous system, the body's most primitive and automatic regulator of the stress responses (Thayer et al., 2010). One way to measure the functioning of the autonomic nervous system is to analyze heart rate variability (HRV). It is well-known that depressed HRV, suggesting a lack of flexibility of autonomic control, has been linked to stress factors (Thayer et al., 2010). Recent studies have shown the potential of mindfulness practices to improve HRV, e.g., increasing the parasympathetic tone (Peressutti et al., 2011; Libby et al., 2012; Krygier et al., 2013).

Furthermore, it was observed in a recent study that MBSR helped reduce blood pressure levels and blood pressure reactivity to stress among healthy community-dwelling individuals reporting elevated stress levels (Nyklícek et al., 2013a ˇ ). In addition, Prakhinkit et al. (2013) observed in a randomized controlled study that both walking meditation (a key mindfulness practice in MBIs) and traditional walking exercise may improve endothelium-dependent vasodilation in elderly with depressive symptoms (Prakhinkit et al., 2013).

Moreover, MBIs may improve both the cardiovascular response to stressors and provide a complementary approach for the prevention and treatment of CVD. A large study showed that a multidimensional intervention based on mindfulness and other integrative medicine principles reduced the 10-year risk of coronary heart disease (CHD) for outpatients with 1 or more known cardiovascular risk factors when compared to usual care, although the effect of the specific components of the intervention were not tested separately (Edelman et al., 2006). Furthermore, a MBSR program may reduce characteristics of the distressed personality ('Type D), and this effect seems to be mediated by increase in self-reported mindfulness (Nyklícek et al., 2013b ˇ ). The "Type D" personality is characterized by a combination of negative affectivity and social inhibition, and has been associated with adverse health outcomes, including CVD (Nyklícek et al., 2013b ˇ ).

# **HYPOTHESIS RATIONALE**

Together with sitting and lying meditation practices, mindfulness exercises such as "mindful walking" and "mindful movements" are key elements in MBIs and couple muscular activity with an internally directed focus (Kabat-Zinn, 2005), improving interoceptive attention to bodily sensations (Ospina et al., 2007; Posadzki and Jacques, 2009; Farb et al., 2013; Gryffin and Chen, 2013; Hagins et al., 2013). This may suggest that a regular physical activity made with internal focus may also improve one's mindfulness levels (Tsang et al., 2008).

In addition, as shown in the previous section, MBIs may buffer cardiovascular responses to stress probably sharing similar mechanisms by which physical fitness (PF) influence these responses (Huang et al., 2013; Stults-Kolehmainen, 2013). **Figure 1** schematically represents potential physiological mechanisms shared by physical activity and mindfulness by which they may attenuate cardiovascular responses to stress (Huang et al., 2013).

Based on these facts, it is feasible to raise the question of whether physical training itself may induce the development of that particular quality of awareness associated with mindfulness (mindfulness as a mediator variable), or if one's DM could moderate the effects of exercise on cardiovascular response to stress. Until now, there has been a lack of information in scientific literature regarding the role of mindfulness as a mediator or moderator of effects of exercise training on stress and health in general, and on cardiovascular responses to stress in particular. Thus, in this study we aimed to develop the hypothesis that mindfulness influences the effects of PF on cardiovascular responses to stress by both moderating or mediating these effects, and discussed potential practical ways to test these hypotheses.

### **MINDFULNESS AS A MODERATOR OF THE EFFECTS OF EXERCISE ON CARDIOVASCULAR RESPONSE TO STRESS**

As represented in the **Figure 2** (line "C"), our first hypothesis is that DM acts as a moderator in the association between PF (independent variable) and cardiovascular responses to stress (dependent variable), by altering the strength of the association. By considering both PF and DM as continuous variables (Baron and Kenny, 1986), probably the moderation occurs in a linear direction (the more DM, the more improvement on dysfunctional stress-induced cardiovascular reactivity caused by PF). Thus, we may theorize that exercise training affects differently people with higher or lower DM.

Although there is no published data about this potential moderator role of DM, it is feasible to speculate on it as mindfulness may be seen as an "independent variable" (see **Figure 2**, line "B") in the relationship with the "dependent variable" cardiovascular response to stress, an important pre-requisite to be considered as a moderator factor (Baron and Kenny, 1986). In addition, it has been suggested that sex, obesity levels, and individual's perception of control over a situation may be moderators of the effects of PF (Huang et al., 2013), and so DM could be another factor

to be taken in account in future studies. Moreover, it is interesting to note that some authors have discussed the idea that only less stressed people could benefit from exercise effects (Stults-Kolehmainen, 2013), and so DM could be a potential confounder factor in this phenomena, as it is well-known that DM is related to less perceived stress and stress-induced biomarkers (Ciesla et al., 2012; Murphy et al., 2012).

In addition, recent findings have identified variables potentially associated with DM that may be confounding variables for the effects of PF on cardiovascular responses to stress. Lower or higher DM may mean lower or higher levels of psychological well-being, healthy behaviors, adherence to an exercise program, emotional reactivity, social anxiety, and co-morbidity, and so may alter the strength of the association between exercise training and cardiovascular responses to stress (Ulmer et al., 2010; Bränström et al., 2011; Salmoirago-Blotcher et al., 2011; Ciesla et al., 2012; Brown et al., 2013; Garcia et al., 2014). It matters mainly if investigators are not using a randomized controlled design to study that association. Even in this case, stratified randomization for DM may be of interest in order to prevent this potential bias.

In order to test the hypothesis that DM is a moderator of the effects of PF on cardiovascular responses to stress, an investigator should address this fundamental research question: does an exercise-training program produce more benefits for people with higher levels of DM but fewer for those with lower levels? If it was the case, it would be expected to find an interaction as hypothesized in **Figure 3**, where is shown potential additive or synergistic effects between PF and DM. To test such a question, the investigator should pre-specify DM as a moderator variable of interest before the intervention is delivered, and should observe whether DM influences the strength of the association between the exercise training intervention and outcomes related to cardiovascular responses to stress. In other words, to test whether the causal relation between PF and cardiovascular responses to stress changes as a function of DM (Baron and Kenny, 1986). In statistical terms, the slope of the interaction between DM and

the exercise-training program improving PF, in a multivariate regression model, should be significant.

### **MINDFULNESS AS A MEDIATOR OF THE EFFECTS OF EXERCISE ON CARDIOVASCULAR RESPONSE TO STRESS**

Our second hypothesis is that mindfulness may be useful to understand the processes through which PF influences changes in cardiovascular responses to stress (i.e., mindfulness as a mediator mechanism of PF effects). In this case, rather than hypothesizing a direct causal relationship between the exercise training and outcomes, our mediational model hypothesizes that exercise training causes changes on mindfulness levels, which in turn helps to account for outcomes related to cardiovascular responses to stress (**Figure 4**, lines "B" plus "C").

Two recent studies give support to this mediational hypothesis. Goldin and colleagues (Goldin et al., 2012) using a randomized controlled design observed, among other findings, that both MBI and an 8-week aerobic exercise (AE) program were with reductions in self-reported negative emotional reactivity to negative self-beliefs. Although it was not their primary outcome, interestingly, they also observed that AE increased participants' mindfulness levels measured by the Kentucky Inventory of Mindfulness Skills (Baum et al., 2010). This increase was statistically significant but less intense than that induced by the MBI, although there were no significant differences between groups in pre-to-post change (Goldin et al., 2012). Another study, led by Zgierska et al. (2013), interestingly showed that increases in DM measured by MAAS seems to be a better predictor than exercise intensity in preventing cold illness severity when comparing a MBI to an AE program to prevent acute respiratory infection, using a randomized controlled design (Zgierska et al., 2013). Together, these findings suggest that a "classical" AE program also has the potential to promote mindful awareness, and the hypothesis that mindfulness is a mediator of the exercise training on cardiovascular responses to stress may be feasible.

A potential mechanism that could explain this finding is that during and just after (recovery period) a physical exercise, mainly of moderate to vigorous intensity, it is common the awareness of our present moment body sensations such as our breathing rhythm, body temperature, or some kind of body transient momentary discomfort, or even some changing in our internal affective state. Such a combination may enhance one's acute

mindfulness state, and chronically one's DM. This has resemblance with the cross-stressor adaptation hypothesis that postulates that there is not only a direct adaptation to the exercise load but also a transfer to everyday life stress situations (Sothmann et al., 1996).

To confirm the mediational hypothesis presented herein, an investigator should test whether an exercise program produces the observed outcomes in cardiovascular responses to stress partially via changes in mindfulness levels (mediator variable). In general, to test mediation hypotheses, it is common to use statistical methods such as regression analysis in order to verify whether both the intervention (exercise program) and the outcome (cardiovascular responses to stress) covary as expected with the hypothesized mediating variable (mindfulness levels), and whether controlling for mindfulness levels explains part of the exercise program effects (Baron and Kenny, 1986). In other words, it is necessary to demonstrate that the relationship between the exercise program and the cardiovascular responses to stress is significantly lower when the mindfulness level is included in the equation. Using the representation in **Figure 4**, when paths "A" and "B" are controlled, a previously significant relation between the independent (PF) and dependent variable (cardiovascular responses to stress) is less or no longer significant (Baron and Kenny, 1986).

# **MINDFULNESS AS BOTH A MODERATOR AND MEDIATOR OF THE EFFECTS OF EXERCISE ON CARDIOVASCULAR RESPONSE TO STRESS**

Our third hypothesis is a combined model with mindfulness having both mediator and moderator status (**Figure 5**). Baron and Kenny coined the term "mediated moderation" for this potential kind of relation among variables (Baron and Kenny, 1986). In **Figure 5**, mediated moderation would be indicated by "PF × DM1" affecting "CRS" in line "D," and "PF × DM1" affecting "DM2" and "DM2" affecting "CRS" in lines "D" plus "B". So it is

**effects on cardiovascular responses to stress ("D" represents "mediated moderation").** Adapted from Baron and Kenny (1986). PF: physical fitness. DM: dispositional mindfulness. CRS: cardiovascular responses to stress. DM 1: DM pre-intervention. DM 2: DM post-intervention.

possible for "DM2" (post-intervention DM) to mediate both the effect of PF on cardiovascular responses to stress, and the effect of "PF × DM1" (pre-intervention DM moderating PF) on that outcome.

It may be explained by the fact that pre-intervention DM levels (DM1) may differentially affect post-intervention DM (DM2) induced by PF ("mediated moderation"). In other words, one's pre-intervention DM moderates the effects of PF on postintervention DM (DM2) levels, and then DM2 mediates the effects of PF on cardiovascular responses to stress (dependent variable). Once more, there is no direct evidence about this hypothesis, but it has been suggested that individuals with higher pre-intervention DM have greater benefits (e.g., higher postintervention DM) from participating in a MBI (Shapiro et al., 2011), and the same process may occur regarding the effects of an exercise intervention. To test this hypothetical model, researchers might use modeling of covariance structures.

#### **IMPLICATIONS FOR FUTURE STUDIES**

Based on our hypotheses, some further studies could be performed. One possibility would be to compare results from AE programs done with additional mindfulness training to usual training (non-mindful component) using the mediational hypothesis analysis. Another study, maybe the most innovative type, would be to compare an exercise session performed with additional instruction to enhance mindfulness, i.e., with full attention to movements and bodily sensations, to a usual training session without this kind of instruction. This last possibility was also suggested by Stults-Kolehmainen (2013, p. 3). In this same study, another possibility for control group would be individuals that intentionally do not stay aware of the exercise, as it has been suggested that multitasking during exercise results in less beneficial stress adaptations (Breus and O'Connor, 1998).

With regard to this last potential study, there are few welldesigned studies that have already compared mindful modes of exercise to non-mindful ones but not specifically related to cardiovascular responses to stress (Netz and Lidor, 2003; Tsang et al., 2008). Although an initial study had ambiguous results (Brown et al., 1995), one interesting study examined the effect of a single session of mindful exercise on mood. The intervention compared four physical exercise modes: yoga, the Feldenkrais technique (awareness through movement), aerobic dance, and swimming, with a cognitive exercise delivered by computer lessons serving as control. Measures of mood improved following Feldenkrais, swimming, and yoga, but did not improve following aerobic dance and computer lessons (Netz and Lidor, 2003), supporting the idea that mindful exercise may have additional benefits compared to non-mindful modes of exercises.

In addition, future studies should examine some competing theories as alternatives to our hypotheses. One possibility is that higher scores in DM could help persons to better perform or tolerate exercise training (e.g., through better training psychological and physiological adaptation), consequently improving adherence and compliance to exercise protocols. Another competing argument would be that exercise training might not result in an increase in DM, but that exercise training could improve certain factors such as perceived stress (Starkweather, 2007) which are also improved by the practice of mindfulness, and so both would have positive effects, but exercise would not necessarily lead to improvements in DM.

As expected, our theoretical model has limitations. A main limitation is that scales and questionnaires to measure mindfulness are still object of scientific debate, and some authors have argued that they lack construct validity in measuring the complexity of mindfulness (Grossman et al., 2011; Park et al., 2013). In this way, future situational measurements of mindfulness (Mitchell et al., 2013) could be more effective to support or reject our hypotheses. Another main limitation is that mindfulness seems to be a multidimensional construct (Grossman et al., 2011; Park et al., 2013) and we have not addressed this fact herein in order to simplify the model. In this direction, further research should address the differential moderator and mediator roles of these mindfulness sub-dimensions, e.g., those addressed by the FFMQ ("observing," "describing," "acting with awareness," "non-judging of inner experience," and "non-reactivity to inner experience") (Bohlmeijer et al., 2011).

# **CONCLUSION**

In conclusion, we have developed the hypothesis that mindfulness may both moderate and mediate the effects of PF on cardiovascular response to stress. This innovative structural model may oxygenate the debate in this research field and help researchers better understand the effects of PF on cardiovascular responses to stress. In addition, general research on both fields of exercise and mindfulness may benefit from this kind of theoretical approach as well.

Conceivably, a "mindful" exercise may provide benefits that are not available in "non-mindful" regular exercise or in opposition to situations where exercise is performed in a multitasking context (Breus and O'Connor, 1998). If so, a mixture of aerobic or resistance physical training with mindfulness would have the potential to improve cardiovascular response to stress in a more effective direction.

Based on these feasible hypotheses, exercise researchers should consider mindfulness in their future research protocols in order to address questions such as whether a higher DM is associated with better health outcomes in an exercise protocol, or if a mindfulness training session mixed into an exercise program could improve health and physical performance among patients, general population, and athletes.

#### **AUTHOR CONTRIBUTIONS**

Marcelo M. P. Demarzo presented the initial concept of the hypothesis, drafted the first version, and organized subsequent versions until the final format of the manuscript. Jesús Montero-Marin, Phyllis K. Stein, Ausiàs Cebolla, Jaime G. Provinciale, and Javier García-Campayo contributed to the development and improvement of the hypothesis and manuscript until their final version.

#### **ACKNOWLEDGMENTS**

Marcelo M. P. Demarzo is grateful to the CNPq - Brazilian National Council for Research and Technology Development for a postdoctoral fellowship under supervision of Professor Javier García-Campayo ("Science without Borders Program"). The authors would like to thank the reviewers for the constructive environment for debating hypotheses and concepts presented herein, which effectively allowed the manuscript's development and improvement.

# **REFERENCES**


the prevention of acute respiratory infection: possible mechanisms of action. *Evid. Based Complement. Altern. Med.* 2013:952716. doi: 10.1155/2013/952716

**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Received: 07 August 2013; accepted: 02 March 2014; published online: 25 March 2014. Citation: Demarzo MMP, Montero-Marin J, Stein PK, Cebolla A, Provinciale JG and García-Campayo J (2014) Mindfulness may both moderate and mediate the effect of physical fitness on cardiovascular responses to stress: a speculative hypothesis. Front. Physiol. 5:105. doi: 10.3389/fphys.2014.00105*

*This article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology.*

*Copyright © 2014 Demarzo, Montero-Marin, Stein, Cebolla, Provinciale and García-Campayo. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Regular exercise is associated with emotional resilience to acute stress in healthy adults

# *Emma Childs\* and Harriet de Wit*

*Human Behavioral Pharmacology Laboratory, Department of Psychiatry and Behavioral Neuroscience, The University of Chicago, Chicago, IL, USA*

#### *Edited by:*

*Anthony Leicht, James Cook University, Australia*

#### *Reviewed by:*

*Antti M. Kiviniemi, Verve, Finland Daniel Boullosa, Universidade Católica de Brasília, Brazil*

#### *\*Correspondence:*

*Emma Childs, Human Behavioral Pharmacology Laboratory, Department of Psychiatry and Behavioral Neuroscience, The University of Chicago, 5841 S Maryland Ave., MC3077, Chicago, IL 60637, USA e-mail: echilds@uchicago.edu*

Physical activity has long been considered beneficial to health and regular exercise is purported to relieve stress. However empirical evidence demonstrating these effects is limited. In this study, we compared psychophysiological responses to an acute psychosocial stressor between individuals who did, or did not, report regular physical exercise. Healthy men and women (*N* = 111) participated in two experimental sessions, one with the Trier Social Stress Test (TSST) and one with a non-stressful control task. We measured heart rate, blood pressure, cortisol, and self-reported mood before and at repeated times after the tasks. Individuals who reported physical exercise at least once per week exhibited lower heart rate at rest than non-exercisers, but the groups did not differ in their cardiovascular responses to the TSST. Level of habitual exercise did not influence self-reported mood before the tasks, but non-exercisers reported a greater decline in positive affect after the TSST in comparison to exercisers. These findings provide modest support for claims that regular exercise protects against the negative emotional consequences of stress, and suggest that exercise has beneficial effects in healthy individuals. These findings are limited by their correlational nature, and future prospective controlled studies on the effects of regular exercise on response to acute stress are needed.

**Keywords: physical activity, stress, TSST, heart rate, blood pressure, cortisol, mood**

### **INTRODUCTION**

The Centers for Disease Control and Prevention (CDC) and the American College of Sports Medicine (ACSM) recommend that American adults exercise for at least 30 min on most days to improve their health and quality of life (Pate et al., 1995; Haskell et al., 2007). Indeed, clinical trials have shown that regular exercise is an effective treatment for disease, including physical ailments e.g., cardiovascular disease (Elrick, 1996), and psychiatric disorders e.g., depression (Dinas et al., 2011). Further, regular exercise is frequently associated with general well-being and lower rates of mood and anxiety disorders in cross-sectional studies (Dua and Hargreaves, 1992; Slaven and Lee, 1997; Goodwin, 2003) and with improved longevity and decreased mortality in prospective studies (Kujala et al., 1998; Samitz et al., 2011). However, the mechanisms underlying the benefits of exercise are not clear. One way in which exercise may promote health is via enhanced resilience to stress, since stress exposure and chronic stress burden has been associated with physical and mental illness (McEwen, 2007).

Acute stress produces a cascade of physiological and psychological effects that are coordinated by the sympathetic nervous system (SNS) and the hypothalamic pituitary adrenal axis (HPAA, López et al., 1999; Chrousos, 2009). Soon after encountering a stressor, defined as any emotional, physical or psychological threat that perturbs homeostasis, heart rate and blood pressure increase, along with mental alertness and tension, and cortisol is released into the blood from the adrenals (Habib et al., 2001). This multidimensional stress response is extremely beneficial and serves to ready the organism to deal with the imminent threat, however when it is improperly activated it can have deleterious effects and contribute to diseases such as atherosclerosis, obesity and depression (McEwen, 2006). Intense physical activity can also be considered a stressor since it activates the same systems involved in responding to an external threat (Hackney, 2006); bouts of exercise increase heart rate, blood pressure and levels of cortisol. Thus, regular activation of stress systems by physical exercise may produce beneficial adaptations such that these systems are able to respond to acute stress more effectively, for example with reduced vigor or shorter duration. This idea has been termed the cross-stressor adaptation hypothesis (Sothmann, 2006).

Previous studies have investigated psychophysiological responses to mentally challenging or psychologically stressful laboratory tasks among healthy individuals with respect to levels of physical fitness. Most of these studies have focused upon cardiovascular reactivity to psychologically stressful tasks with mixed findings. For example, more physically fit individuals have exhibited enhanced reactivity (de Geus et al., 1993), blunted reactivity (Crews and Landers, 1987) or no difference (Steptoe et al., 1990; Blumenthal et al., 1991; Summers et al., 1999; Spalding et al., 2004; Poole et al., 2011) in comparison to their less physically fit counterparts. Further, two recent meta-analyses of studies reached different conclusions regarding cardiovascular reactivity; Jackson and Dishman (2006) reported that physical fitness was associated with a greater cardiovascular response yet quicker recovery, while Forcier et al. (2006) reported that cardiovascular responses were blunted. A limited number of studies have examined cortisol responses to psychological stress. In these studies, physically fit men, older women, and highly active children exhibited blunted cortisol responses to psychological stress in comparison to their less physically active counterparts (Traustadottir et al., 2004; Rimmele et al., 2007; Martikainen et al., 2013). Finally, few studies have reported on emotional responses to stress with respect to levels of physical exercise, and again the findings have been mixed (e.g., Sinyor et al., 1983; Choi and Salmon, 1995; Summers et al., 1999). The discrepancies in findings between studies may be due to methodological differences, including stress induction methods (e.g., speech tasks tend to produce greater psychophysiological responses than mental challenge tasks, Dickerson and Kemeny, 2004) and subject samples studied (e.g., previous studies show evidence of sex differences in responses to psychological stress, Kirschbaum et al., 1999; Kelly et al., 2008; Childs et al., 2010).

In this study we aimed to assess multidimensional responses i.e., physiological (heart rate, blood pressure, salivary cortisol) and psychological (anxiety, positive mood) to a standardized acute psychosocial stressor among young healthy men and women, and to compare stress responses between individuals who reported regular physical exercise and those who did not. To our knowledge no studies have measured multidimensional aspects of stress responses i.e., cardiovascular, hormonal and emotional, in the same individuals which is important to show the existence of relationships between the different modalities. Furthermore, very few studies have examined stress-induced changes in positive and negative mood states with respect to levels of habitual physical exercise. We hypothesized, based on previous reports, that nonexercisers participants would exhibit greater cardiovascular and emotional reactivity, and dampened cortisol responses to stress in comparison to regular exercisers.

# **MATERIALS AND METHODS**

#### **SUBJECTS**

Participants (*n* = 111) were recruited from the University and surrounding area by flyers and advertisements. They attended the laboratory for an in-person medical screening that included a health and drug use questionnaire and an ECG. Participants were healthy adults, aged 18–32, with body mass index 19– 29 kg/m2. Exclusion criteria included a current or past year diagnosis of a Major Axis I psychiatric disorder (American Psychiatric Association, 1994), an abnormal electrocardiogram, use of prescription medications including, in women, oral contraceptives (Kirschbaum et al., 1995, 1999), or night shift work. Individuals who smoked *>*5 cigarettes/week were also excluded as smoking has been shown to alter responses to the TSST (Kirschbaum et al., 1993b; al'Absi et al., 2003; Childs and de Wit, 2009). Information on physical activity was obtained from a questionnaire administered at screening; participants indicated whether they exercised on a regular basis (outside of normal activities including commuting), and how many times per week they exercised. Participants were then classified as sedentary (*n* = 30) or regular exercisers (i.e., ≥1 occasion per week, *n* = 81). Participants were told that the study aim was to examine the effects of verbal tasks on mood and physiology. At the end of the study, they were fully debriefed about the study aims and paid for their participation.

#### **PROCEDURE**

The University of Chicago Hospital's Institutional Review Committee for the use of human subjects approved the study protocols. All participants provided informed consent at a separate orientation session conducted before the study began. Participants completed two sessions at least 48 h apart, one with a stressful task and another with a non-stressful control task, in randomized order. The stressor that we used was the Trier Social Stress Test (TSST; Kirschbaum et al., 1993a). This is a standardized and widely used psychosocial stressor that reliably induces changes in physiological and psychological dimensions (Dickerson and Kemeny, 2004). Participants also completed a non-stressful control task on a separate day to account for diurnal rhythms in mood and physiology (Childs and de Wit, 2009; Het et al., 2009; Childs et al., 2010; Lovallo et al., 2010).

All study procedures were conducted at the Human Behavioral Pharmacology Laboratory at the University of Chicago. Experimental sessions were conducted in testing rooms furnished as a comfortable living area, with an easy chair for relaxing (when participants were not completing study measures), a television and video player, and a desk with a computer for completing study questionnaires. On arrival, participants provided breath and urine samples to detect recent drug or alcohol use (no one tested positive) and then relaxed for 30 min to acclimatize to the laboratory. Baseline measures were obtained and then participants were read instructions for the task to be performed that day. They were allowed 10 min to prepare for each task, at the end of which they were escorted to an adjacent room to perform the task. The TSST consisted of a 5 min speech and 5 min mental arithmetic (serial subtraction) performed before two interviewers who were unknown to the participant and who provided no feedback. There was also a video camera present which projected the participants' image onto a television screen throughout the task. The control task was performed in the absence of a video camera, and involved the participant talking to the research assistant for 5 min about a favorite book, movie or television program, followed by playing a computer game (Solitaire) for 5 min. Before, and at repeated times after the tasks, participants rated their mood, saliva samples were obtained for cortisol analysis, and vital signs were obtained.

#### **DEPENDENT MEASURES**

Saliva samples were collected using Salivette® cotton wads (Sarstedt Inc., Newton, NC) at −30, 10, 20, and 60 min after the tasks and were analyzed by the Core Laboratory at the University of Chicago Hospitals General Clinical Research Center for levels of cortisol (Salimetrics LLC, State College, PA, sensitivity = 0.003 ug/dL). Heart rate (HR) was measured continuously throughout the experimental session (one reading per minute) using a Polar chest band and monitor (Mini-Logger, Mini Mitter/Respironics, Bend, OR). Scores were averaged over consecutive 10 min periods. Blood pressure was measured using a monitor (Critikon Dinamap Plus Vital Signs Monitor, GE Healthcare Technologies, Waukesha, WI) at −30, 0, 10, 20, and 30 min after the tasks. Self-reported mood was measured using the Profile of Mood States questionnaire (POMS, McNair and Droppleman, 1971) at −30 and 0 min after the task.

Personality traits were assessed using the Multidimensional Personality Questionnaire Brief Form (MPQ-BF, Patrick et al., 2002). This questionnaire is an empirically-derived personality instrument with an orthogonal factor structure that yields 11 well-defined primary trait scores and three superfactors termed Positive Emotionality (PEM, extraversion), Negative Emotionality (NEM, neuroticism) and Constraint (CON, behavioral spontaneity).

Some data points were missing due to sample loss or equipment failure, thus there were minor variations in samples sizes between the separate analyses (heart rate *n* = 77; blood pressure *n* = 111; mood *n* = 96; cortisol *n* = 99; personality *n* = 102).

#### **STATISTICAL ANALYSES**

Two markers of stress reactivity were calculated for the outcome measures; (1) peak change from pre-task baseline, which provides a measure of the intensity of the response, and (2) area under the curve relative to the pre-task baseline (AUC, Altman, 1991; Pruessner et al., 2003), which provides information about response duration and recovery of homeostasis after stress exposure. We first confirmed the efficacy of the stress task by comparing responses in the outcome measures between the two tasks using one factor (Task) repeated measures analysis of variance (ANOVA). We also compared responses to the tasks between men and women using two factor (Task∗Sex) repeated measures ANOVA, since others have previously reported sex differences in responses to the TSST (Kirschbaum et al., 1999; Kelly et al., 2008; Childs et al., 2010). Sex was included as an additional factor in later analyses for any outcome measures that were significantly influenced by Sex.

We then compared demographic and personality characteristics between the groups using independent samples *t*-test (for continuous variables) and chi-squared analysis (for categorical variables). We also compared baseline measures (average of pre-task scores from each session) between the groups using independent samples *t*-test. We compared responses to the tasks between groups using two-factor repeated measures (Task∗Group) ANOVA. All analyses were conducted using SPSS v19 for windows. Finally, we assessed relationships between selfreported frequency of exercise per week and net responses to stress (i.e., response after TSST minus response after control task) using Pearson's correlation coefficient. Effect sizes are reported using partial eta squared (η<sup>2</sup> <sup>ρ</sup>) for analyses of variance; 0.01, 0.06, and 0.14 are considered, respectively, small, medium, and large effect sizes.

# **RESULTS**

#### **DEMOGRAPHICS**

Most participants were of European descent (53%) and in their early twenties (22*.*1 ± 0*.*4 years, **Table 1**). The groups did not differ on any demographic or personality characteristics.

#### **Table 1 | Demographic characteristics of study participants.**


*Values indicate mean* ± *s.e.m. Asterisks indicate a significant difference between the samples (\*p < 0.05, Chi-squared analysis).*

#### **BASELINE MEASURES**

Before the tasks began, heart rate was significantly lower among individuals who reported regular exercise [*t*(80) = 2*.*2; *p <* 0*.*05, mean difference = 6.3 ± 2.9 bpm] and baseline heart rate was significantly negatively correlated with the frequency of exercise per week (*r* = −0*.*24; *p <* 0*.*05) in the whole group. Blood pressure, cortisol and mood did not differ significantly between the groups at baseline (**Table 2**).

#### **STRESS REACTIVITY AND SEX DIFFERENCES**

In comparison to the control task, the TSST significantly increased heart rate, blood pressure, and cortisol among all participants (**Table 3**). The TSST also significantly increased negative affect (Anxiety, Depression, Anger, Confusion) and decreased positive mood states (Friendly, Elation, Positive Mood).

Overall, men exhibited greater cortisol responses to the tasks than women [Sex effect: Peak change *F*(1*,* 96) = 16*.*7; *p <* 0*.*001; η<sup>2</sup> <sup>ρ</sup> <sup>=</sup> <sup>0</sup>*.*15; AUC *<sup>F</sup>*(1*,* 95) <sup>=</sup> <sup>18</sup>*.*1; *<sup>p</sup> <sup>&</sup>lt;* <sup>0</sup>*.*001; <sup>η</sup><sup>2</sup> <sup>ρ</sup> = 0*.*16] and greater stress-induced increases in systolic blood pressure than women [Task∗Sex effect: *<sup>F</sup>*(1*,* 106) <sup>=</sup> <sup>4</sup>*.*2; *<sup>p</sup> <sup>&</sup>lt;* <sup>0</sup>*.*05; <sup>η</sup><sup>2</sup> <sup>ρ</sup> = 0*.*04]. Therefore, Sex was included as a factor in later analyses of these measures. There were no other sex differences in other cardiovascular or emotional responses to the tasks.

#### **INFLUENCE OF EXERCISE ON STRESS REACTIVITY**

Heart rate, blood pressure and cortisol reactivity to the stress procedure did not differ between the groups. Individuals who did not regularly exercise exhibited significantly greater decreases in positive affect after stress [Task∗Group effect: Elation *F*(1*,* 94) = 8*.*38; *p <* 0*.*01; η<sup>2</sup> <sup>ρ</sup> = 0*.*08, Positive Mood *F*(1*,* 94) = 3*.*06; *p* = 0*.*08; η2 <sup>ρ</sup> = 0*.*03, **Figure 1**]. Overall, regular exercisers also felt more "friendly" after both tasks [Group effect: Friendliness *F*(1*,* 94) = 4*.*39; *p <* 0*.*05; η<sup>2</sup> <sup>ρ</sup> = 0*.*05, **Figure 1**]. Correlation analyses did not show any evidence of significant relationships between the

**Table 2 | Baseline values of physiological measures.**


*Values indicate mean* ± *s.e.m. Asterisks indicate a significant difference between the samples (\*p < 0.05, Independent Samples t-test).*

**Table 3 | Responses to the control task and TSST among all participants.**


*Values indicate mean* ± *s.e.m. Asterisks indicate a significant difference between the samples (\*p < 0.05, \*\*p < 0.01, \*\*\*p < 0.001, Independent Samples t-test).*

frequency of exercise per week and psychophysiological responses to stress.

### **DISCUSSION**

This study aimed to compare reactivity to acute stress between healthy individuals who exercise regularly and those who do not. We examined cardiovascular, cortisol and emotional responses to a standardized psychosocial stressor in comparison to a non-stressful control task among healthy male and female participants and in relation to their self-reported levels of regular physical exercise. There were several interesting findings. Overall, heart rate was significantly lower among regular exercisers than sedentary participants, yet cardiovascular reactivity to stress or the tasks overall did not differ between the groups. The groups did not significantly differ in levels of cortisol at baseline or after stress. Also, while subjective mood states did not differ between the groups at baseline, emotional responses to the tasks did differ between the groups; participants who regularly exercised exhibited less of a decline in positive affect after stress than sedentary participants. These findings suggest that regular exercisers may be more resistant to acute stress, which may protect them against future poor health.

The finding that regular exercisers exhibited a smaller decline in positive affect during a stressful situation provides some of the first direct evidence to support that habitual physical activity is associated with stress resilience in healthy individuals. Exercise

has been associated with greater well-being in cross-sectional studies of healthy adults but empirical evidence of its beneficial effects is limited. Both groups showed similar levels of positive and negative affect at baseline, and the groups did not differ in personality measures of positivity or negativity. Thus, regular exercise was not associated with higher baseline levels of positive mood, but instead selectively influenced the ability of a stressful situation to diminish positive affect. Interestingly, stress-induced increases in negative affect were similar between the groups. Recent theories have begun to place more importance on the role of positive emotions during stress independent of negative affect (Folkman, 2008). Moreover, positive, but not negative affect, has been linked to a decreased risk of mortality (Moskowitz et al., 2008; Davis, 2009). Thus, an ability to maintain greater positive mood during stress exposure among regular exercisers may serve a protective function, minimizing the accumulation of stress burden with repeated exposures that is linked with the development of disease. Possible explanations for a resistance to stress-induced decreases in positive affect include that individual appraisals of the situation, self-resources, or coping strategies may be more positive among regular exercisers. For example, positive coping strategies have been linked to greater positive affect during stress (Folkman and Moskowitz, 2000, 2004; Lazarus, 2000). Thus, future studies should also look to assess primary and secondary appraisals of stressful situations and the various coping strategies utilized by exercisers and non-exercisers.

Regular exercisers did not exhibit altered reactivity in other components of responses to acute stress which echoes the findings of some previous reports (Blumenthal et al., 1991; Summers et al., 1999; Spalding et al., 2004; Poole et al., 2011). Also, in line with others' findings, heart rate was significantly lower overall among exercisers in comparison to nonexercisers (e.g., de Geus et al., 1990, 1993; Summers et al., 1999), and baseline heart rate was correlated with the frequency of self-reported exercise per week. Thus, although our analysis depended upon self-reported levels of exercise which are more unreliable than objective measures of physical activity, the significant relationship between baseline heart and frequency of self-reported exercise reinforces the validity of our approach.

There were several limitations to the present study. First, as mentioned, we relied upon self-reports of regular physical exercise which can be unreliable. A better method for future studies would be to obtain an objective measure of physical activity, such as that provided by an activity monitor or accelerometer which participants could wear for a week before testing. A second limitation was that the groups were self-selecting, that is there may be a bias introduced when subjects who spontaneously engage in exercise are compared to non-exercising persons. Although there was no difference between regular exercisers and sedentary participants in baseline mood or the personality traits of extraversion and neuroticism, there may be an underlying factor associated with both exercise and stress resilience that is not accounted for in this study. Finally, the proportion of European Americans who reported regular exercise was significantly greater than that reported by other races, and since race has also been shown to influence stress responses (Shen et al., 2004; Chong et al., 2008; Fauvel and Ducher, 2009; Li et al., 2009; Christian et al., 2013), we cannot negate the influence of race upon our findings. Nevertheless, studies of differences in stress responses between races have mainly reported differences in physiological responses to stress, and in this study we did not find any group differences in this measure (Murphy et al., 1992; Saab et al., 1992; Kelsey et al., 2000; Wilcox et al., 2005; Kim, 2008). Thus, in order to conclusively disentangle the effects of race from physical exercise, our findings should be replicated in samples with similar distributions of European Americans.

#### **CONCLUDING REMARKS**

In conclusion, in this study we assessed multidimensional responses to acute stress in healthy participants who differed in levels of regular exercise. The stress procedure produced a smaller decline in positive mood among the regular exercisers, compared to the sedentary individuals. Responses to the psychosocial stressor used in this study may reflect the way individuals typically respond to daily stressors, suggesting that regular exercisers are more resistant to the emotional effects of acute stress, which in turn, may protect them against diseases related to chronic stress burden.

#### **ACKNOWLEDGMENTS**

Emma Childs and Harriet de Wit wrote the protocols. Emma Childs conducted the studies, data analyses and wrote the first draft of the manuscript. Funding for this study was provided by a grant awarded to Harriet de Wit (NIDA DA02812). The authors thank Les Sidney, Heather Longstreth, Michael Helzer, Nicole Noga, and Celina Joos for their technical assistance.

#### **REFERENCES**


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Received: 18 February 2014; accepted: 05 April 2014; published online: 01 May 2014. Citation: Childs E and de Wit H (2014) Regular exercise is associated with emotional resilience to acute stress in healthy adults. Front. Physiol. 5:161. doi: 10.3389/fphys. 2014.00161*

*This article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology.*

*Copyright © 2014 Childs and de Wit. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# The impact of escitalopram on vagally mediated cardiovascular function to stress and the moderating effects of vigorous physical activity: a randomized controlled treatment study in healthy participants

*Camilla S. Hanson1, Tim Outhred1,2,3, Andre R. Brunoni 4, Gin S. Malhi 2,3 and Andrew H. Kemp1,2,3,4\**

*<sup>1</sup> SCAN Research and Teaching Unit, School of Psychology, University of Sydney, Sydney, NSW, Australia*

*<sup>2</sup> Discipline of Psychiatry, Sydney Medical School, Northern Clinical School, University of Sydney, Sydney, NSW, Australia*

*<sup>3</sup> CADE Clinic, Royal North Shore Hospital, St. Leonards, NSW, Australia*

*<sup>4</sup> Center for Clinical and Epidemiologic Research, University Hospital, University of São Paulo, São Paulo, Brazil*

#### *Edited by:*

*Arto J. Hautala, Verve Research, Finland*

#### *Reviewed by:*

*Arto J. Hautala, Verve Research, Finland Daniel Boullosa, Universidade Católica de Brasília, Brazil*

#### *\*Correspondence:*

*Andrew H. Kemp, Center for Clinical and Epidemiologic Research, University Hospital, University of São Paulo, Av. Prof. Lineu Prestes 2565, CEP 05508-000 Sao Paulo, Brazil*

*e-mail: andrew.kemp@hu.usp.br; andrew.kemp@sydney.edu.au*

Recent concerns over the impact of antidepressant medications, including the selective serotonin reuptake inhibitors (SSRIs), on cardiovascular function highlight the importance of research on the moderating effects of specific lifestyle factors such as physical activity. Studies in affective neuroscience have demonstrated robust acute effects of SSRIs, yet the impact of SSRIs on cardiovascular stress responses and the moderating effects of physical activity remain to be determined. This was the goal of the present study, which involved a double-blind, randomized, placebo-controlled, cross-over trial of a single-dose of escitalopram (20 mg) in 44 healthy females; outcomes were heart rate (HR) and its variability. Participants engaging in at least 30 min of vigorous physical activity at least 3 times per week (regular exercisers) showed a more resilient cardiovascular stress response than irregular vigorous exercisers, a finding associated with a moderate effect size (Cohen's *d* = 0*.*48). Escitalopram attenuated the cardiovascular stress response in irregular exercisers only (HR decreased: Cohen's *d* = 0*.*80; HR variability increased: Cohen's *d* = 0*.*33). HR during stress under escitalopram in the irregular exercisers was similar to that during stress under placebo in regular exercisers. These findings highlight that the effects of regular vigorous exercise during stress are comparable to the effects of an acute dose of escitalopram, highlighting the beneficial effects of this particular antidepressant in irregular exercisers. Given that antidepressant drugs alone do not seem to protect patients from cardiovascular disease (CVD), longitudinal studies are needed to evaluate the impact of exercise on cardiovascular stress responses in patients receiving long-term antidepressant treatment.

**Keywords: selective serotonin reuptake inhibitors (SSRIs), escitalopram, exercise, physical activity, cardiovascular stress response, heart rate, heart rate variability, HRV**

#### **INTRODUCTION**

Depression and cardiovascular disease (CVD) are leading burdens of disease and this burden is projected to worsen up to 2030 and beyond with ageing of the population and increasing prevalence of multi-morbidity (Mathers and Loncar, 2006; Langan et al., 2013). Critically, depression increases risk for the development of CVD 1.5-fold, while patients with CVD and depression have a 2- to 3-fold increased risk of future cardiac events compared to those cardiac patients who do not have depression (Rudisch and Nemeroff, 2003). Psychological stress plays a key role in the development of depression (Cohen et al., 2007), and increases risk of mortality from a number of causes including CVD, cancer and external causes over an 8 year followup period (Russ et al., 2012). While many biological factors, including the hypothalamic-pituitary-adrenal (HPA) axis dysfunction and altered inflammatory processes, contribute to the relationship between stress, psychiatric illness and CVD, vagally mediated cardiovascular function—indexed by increases in heart rate (HR) and decreases in heart rate variability (HRV)—may underlie a substantial part of this risk (Thayer et al., 2010a; Nemeroff and Goldschmidt-Clermont, 2012). Importantly, the inhibitory function of the vagus nerve regulates both the HPA axis and inflammatory processes (Tracey, 2002; Thayer and Sternberg, 2006; Huston and Tracey, 2010; Kemp and Quintana, 2013), leading to proposals that chronic impairment in vagal function is an early marker of future morbidity and mortality (Kemp et al., 2010; Thayer et al., 2010b; Åberg et al., 2013; Kemp and Quintana, 2013).

Acute psychological stress is associated with parasympathetic withdrawal, sympathetic activation, an increase in HR and reductions in its variability (Madden and Savard, 1995; Porges, 1995; Steptoe and Kivimäki, 2012), while chronic stress is associated with persistent cardiovascular stress responses, which may then contribute to psychiatric illness, physical ill-health and all-cause mortality (Thayer and Brosschot, 2005; Thayer et al., 2010b; Lemogne et al., 2011; Åberg et al., 2013; Kemp and Quintana, 2013). Epidemiological research has even demonstrated that an increased resting HR increases the risk of suicide by 19–37% over a follow-up period of 9 years even after accounting for covariates including depressed mood (Lemogne et al., 2011). In another study lower cardiovascular fitness determined using the cycle ergonometric test at 18 years of age predicts an increased risk of suicide attempt and death by suicide over a follow-up period of 42 years (Åberg et al., 2013). As lower cardiovascular fitness is associated with increased HR and reduced HR variability (Rennie, 2003), these findings highlight an important role of vagally mediated cardiovascular function in regards to psychological wellbeing as well as physical health. Importantly, increased parasympathetic control is associated with positive emotions (Kok and Fredrickson, 2010), resilience (Kashdan and Rottenberg, 2010) and improved regulation over the HPA axis and inflammatory processes (Tracey, 2002; Thayer and Sternberg, 2006; Huston and Tracey, 2010; Kemp and Quintana, 2013), thereby contributing to psychological wellbeing and physical health. National guidelines (Australian Government Department of Health and Ageing: http://www*.*health*.*gov*.*au/internet/main/publishing*.*nsf/ content/health-pubhlth-strateg-phys-act-guidelines) distinguish between regular moderate-intensity physical activity and regular vigorous exercise, recommending at least 30 min of vigorous exercise on top of regular activity for health and fitness benefits.

The selective serotonin reuptake inhibitors (SSRIs) are a firstline treatment option (Kemp et al., 2008) for both mood and anxiety disorders and admixtures of the two, and are also used widely in bipolar depression (Malhi, 2012). SSRIS have important acute neurophysiological effects (Kemp and Nathan, 2004; Kemp et al., 2004) providing the foundation upon which subsequent therapeutic response is based (Harmer et al., 2009). Acute effects are often studied in healthy samples to avoid confounds of illness and psychopathology (Harmer, 2010). Interestingly, acute SSRI treatment has been shown to attenuate cardiovascular responses to stress (Straneva-Meuse, 2004a; Golding et al., 2005; Jiang et al., 2013), However, longitudinal research indicates that they may also have adverse cardiovascular effects, including reductions in vagally mediated cardiovascular function (Licht et al., 2010b) and sudden cardiac death (Whang et al., 2009), highlighting the importance of research on antidepressant actions and the moderating effects of specific lifestyle factors such as physical activity. Physical activity is associated with improved psychological health (Powell et al., 2011; Åberg et al., 2013) and lower cardiovascular (and proinflammatory) responses to stress (Hamer and Steptoe, 2007) lowering the risk of future CVD (Mora et al., 2007; Hamer and Stamatakis, 2009; Hamer et al., 2009) and mortality (Leitzmann et al., 2007). The higher the parasympathetic activity at rest—indicated by lower HR and higher HRV—the more autonomic resources available to tackle subsequent stressors; a proposal known as the "autonomic resource hypothesis" (Hynynen et al., 2008; Boullosa et al., 2012). Thus, research on the impact of antidepressants and the moderating effects of physical activity has important implications for not only better understanding therapeutic response to antidepressant medications, but also for the long-term health of the cardiovascular system. Recent research in outpatients with coronary heart disease and depression (Blumenthal et al., 2012) demonstrates that exercise and sertraline (a SSRI) are equally effective at reducing depressive symptoms after 16 weeks of treatment, but that exercise led to greater improvements in HRV collected using 24-h Holter recordings compared with sertraline. However, it remains unclear, to what extent physical activity and SSRIs interact in moderating the effects of SSRIs on psychological stress.

Here we report on a proof-of-concept, laboratory-based, experimental study to examine the acute effects of a commonly prescribed SSRI, escitalopram, on HR and its variability under stress, with the aim of determining whether individual differences in levels of physical activity moderate these effects. Better understanding the impact of the SSRIs and the moderating effects that specific lifestyle factors exert has important implications for therapeutic responsiveness, psychological wellbeing, and physical health. We focus specifically on acute treatment effects in healthy female volunteers. We hypothesized that acute administration of escitalopram would attenuate increases in HR and decreases in HRV during stress, relative to placebo, and that regular vigorous exercise would facilitate this effect.

# **METHODS**

#### **PARTICIPANTS**

Forty-four healthy female volunteers (aged 18–47, *M* = 23.70, *SD* = 5*.*89) completed the study, and gave written informed consent in accordance with National Health Medical Research Council guidelines. Participants were recruited using universitywide staff and student newsletters. Participants were free from medication (other than hormonal contraceptives), physical and psychiatric illness, symptoms of depression and anxiety [PHQ-9 (Kroenke et al., 2001) and GAD-7 (Spitzer et al., 2006) assessment], illicit drug use, alcoholism, smoking, brain injury, neurological disorders, and sustained loss of consciousness. Only female participants were recruited for this study as females display higher rates of mental disorders (Nolen-Hoeksema, 2001; Kessler, 2003; Slade et al., 2009); focusing on females also allowed us to avoid known gender differences in baseline HRV (Rajendra Acharya et al., 2006), HRV responses to cognitive stress (Li et al., 2009), and responsiveness to SSRI treatment (Khan et al., 2005). Finally, participants abstained from caffeine on the morning of the experiment and no participant tested positive on pregnancy tests, which were conducted at each session. Ethics approval for the trial was secured from the University of Sydney's Human Research Ethics Committee (ref. 13901) and the Northern Sydney Central Coast Area Health Service Human Research Ethics Committee (ref. 1105-178 M), and it was registered with the Australian New Zealand Clinical Trials Registry (ANZCTR; ACTRN126111000719932).

#### **EXPERIMENTAL DESIGN**

Participants were randomly assigned to receive escitalopram (20 mg) or placebo *per os*, tested in a double-blind crossover design, with two sessions per participant separated by at least 1 week to ensure a sufficient drug washout of at least 5 half-lives (*t*1*/*<sup>2</sup> = 26*.*7 h; Sogaard, 2005). Forty-five percent of participants received placebo at the first session. Relative to other commonly prescribed SSRIs, escitalopram has improved efficacy, a lower side effect profile, and greater selectivity for serotonergic transport proteins (Fernandez et al., 2007; Cipriani et al., 2009a). The maximum recommended dosage of 20 mg was used to maximize receptor occupancy (Kasper et al., 2009) and subsequent physiological effects from a single dose.

#### **MEASUREMENT OF PHYSICAL ACTIVITY**

Participants reported the frequency in which they engaged in vigorous physical activity using the International Physical Activity Questionnaire (IPAQ; Craig et al., 2003), a widely used, reliable and valid self-report assessment. Physical activity data is transformed into energy expenditure estimates of metabolic equivalent tasks (MET), such that one MET is equivalent to the energy cost of sitting quietly for an hour (1 kcal/kg/h). Participants were categorized into low or high vigorous activity groups—according to national guidelines (Australian Government Department of Health and Ageing: http://www*.*health*.*gov*.*au/internet/main/publishing*.*nsf/content/ health-pubhlth-strateg-phys-act-guidelines). These guidelines recommend at least 30 min of vigorous physical activity 3 days a week on top of regular moderate-intensity activity for health and fitness benefits. Participants categorized in the low activity grouping did not meet the criteria of at least 30 min of vigorous physical activity 3 days a week, while participants in the high activity grouping met or exceeded this criterion. Groups did not differ on time spent sitting, walking or performing moderate physical activity (all *p >* 0*.*1). Groups were compared on resting state HR, providing a (partial) validation of questionnaire-based categorization (see participant characteristics).

#### **STRESS MANIPULATION**

The mental arithmetic task component of the Trier Social Stress Test (TSST; Kirschbaum et al., 1993) was used to elicit stress (Jönsson et al., 2010) and associated physiological correlates, including HR, blood pressure, catecholamines and cortisol (Straneva-Meuse, 2004a; Vermetten, 2006). Participants were instructed to count backwards subtracting thirteens, beginning at either 2083 or 2027 for 5-min. Participants were given one of these two alternate versions of the arithmetic task across sessions. No feedback was given for correct responses, and the experimenter vocalized the word "error" when a mistake was made, instructing the participant to start counting from the beginning, further increasing stress and social pressure as per prior studies (Hjortskov et al., 2004).

#### **PROCEDURE**

Testing was conducted in a psychophysiology laboratory at the Clinical Assessment and Diagnostic Evaluation (CADE) Clinic (www.cadeclinic.com), Royal North Shore Hospital. Participants arrived in the early morning, having consumed breakfast, and abstained from caffeine, as caffeine increases sympathetic nervous system (SNS) activity (Sondermeijer et al., 2002). Participants completed a consent form and pregnancy test. Height and weight were measured at the first session to calculate BMI. Testing commenced 3 h after administration of either placebo or 20 mg of escitalopram, so as to coincide with the time of the expected peak plasma concentration (*T*max = 3*.*0 ± 1*.*5 h; Sogaard, 2005; Rao, 2007) During the waiting period, the participant was provided with a standardized snack and in the first session only, completed the IPAQ. The snack helped to minimize autonomic and mood changes associated with hunger. In order to determine the presence of side effects, subjective wellness was recorded during the waiting period and after psychophysiological testing. Participants reported whether they experienced side effects and were asked to guess their treatment condition in order to examine the potential impact of treatment unblinding at each session.

All participants were tested in the late morning, and at the same time on each session to control for potential changes in HRV with circadian rhythms (Kleiger et al., 2005). Participants sat in a reclined chair with their legs raised and were instructed to breathe normally, remain still and awake for 5-min of data collection in the resting state. Following this, 5-min recordings were made during the stress condition in which participants were required to perform the mental arithmetic task. The experimenter was seated in close proximity to the participant and acted in a formal manner, in order to further increase the stressful nature of the procedure through social pressure (Hjortskov et al., 2004). After testing, subjective ratings of relaxation and stress for each condition were recorded on a five-point Likert scale (from low to high) to determine the efficacy of the experimental manipulations.

Participants provided saliva samples (1 mL) for estradiol and progesterone analysis before treatment administration at each session. The hormonal saliva samples were stored frozen until assay. On the day of assay, samples were thawed for determination of salivary progesterone and estradiol using commercially available kits (Salimetrics, USA) according to the manufacturers instructions. Thawed samples were centrifuged at 1500 × *g* for 15 min to collect clear saliva which was used without further processing for all assays. All samples were brought to room temperature before adding to assay wells and all samples were analyzed in duplicate. Hormonal menstrual phase was determined in accordance with previous research (Lu et al., 1999; Gandara et al., 2007).

# **DATA COLLECTION AND ANALYSIS**

HR and R-R interval recordings—on which analyses of HRV were conducted—were made during two conditions, consistent with Task Force guidelines (Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, 1996): (1) a 5-min resting state and (2) a 5-min social stress task, using a Polar RS800CX training device at a sampling rate of 1000 Hz. These devices have been validated against the electrocardiogram (ECG) and have excellent reliability, especially in young, healthy individuals in a supine position and when analysis is conducted on normalized values (Gamelin et al., 2006; Weippert et al., 2010) (see also Quintana et al., 2012), as was the case in the current study. This device is wirelessly connected to an electrode strap worn just beneath the chest, that has been wetted with saline solution (0.9% NaCl) to simulate sweat and ensure conductivity. R-R intervals were extracted from text files and analyzed in Kubios HRV analysis software (version 2.0, 2008, biosignal analysis and medical imaging group, University of Kuopio, Finland, MATLAB). Each file was visually inspected for artifacts (ectopic and missing beats), and medium automatic filter corrections were applied to each data set. Measures of HRV included HF HRV (normalized units, n.u.) (0.15–0.40 Hz) and the Root Mean Square of the Successive Differences (RMSSD).

#### **STATISTICAL ANALYSIS**

All statistical analyses were conducted using IBM SPSS Statistics (Version 20) for Windows 7. The impact of escitalopram and the moderating effects of physical activity levels on HR and heart rate variability (HRV) were examined using mixed withinand between-subjects analysis of variance (ANOVA). Within subject factors were treatment (escitalopram vs. placebo) and task (stress vs. rest) and the between subjects factor was physical activity (high vs. low). A factor of "treatment order" was not included in statistical analysis for parsimony; order effects are rare, are generally underpowered even when an appreciable effect is present, and sufficient drug washout ameliorates such concerns (Senn, 1994; Senn et al., 2004; Mills et al., 2009). Independent samples *t*-tests were conducted to determine whether low and high activity groupings differed across demographic variables (**Table 1**). ANOVA was also employed to determine whether subjective ratings of relaxation and stress differed across treatment sessions and physical activity groupings. Cohen's *d* effect size statistics were calculated for each pairwise comparison. Cohen's guidelines (Cohen, 1988a,b) identify 0.2, 0.5, and 0.8 as small, medium, and large effects, respectively. Partial eta-squared (η<sup>2</sup> *<sup>p</sup>*) was reported for ANOVA effects as an indicator of effect size (small = 0.01 medium = 0.06, large = 0.14; (Cohen, 1988a,b). The statistical threshold of 0.05 (two-tailed) was set for all analyses, except where directional effects were expected, in which one-tailed tests were employed. Manipulation checks included a focus on whether hormone or menstrual status, or treatment blinding differentially impacted on the results reported below. With respects to hormone and menstrual status, χ<sup>2</sup> tests were conducted to check for equal proportions of participant menstrual phase within treatment sessions and equal proportions of participants using hormonal contraceptives in each order of treatment. Regression analyses were also conducted on hormone concentrations in order to determine the relationships between hormone levels and the HR and HRV responses under each treatment and task condition. With respects to treatment blinding, χ<sup>2</sup> tests were performed to examine the relationship between guessing treatment condition and experiencing side effects. The effect of guessing and side effects on HR and HRV under each treatment and task condition was determined using repeated-measures ANOVAs on HR and HRV separately.

# **RESULTS**

#### **PARTICIPANT CHARACTERISTICS**

Participant characteristics are presented in **Table 1** and flow of participants through the different stages of the study is presented in **Figure 1**. The two physical activity groupings were matched in terms of age, BMI, years of education, time spent sitting, walking or performing moderate physical activity (all *p >* 0*.*1). The high activity group spent more minutes in the past week in vigorous activity [*t(*38*)* = 6*.*93, *p <* 0*.*001, *d* = 2*.*26], and had a higher weekly energy expenditure [*t(*38*)* = 3*.*50, *p* = 0*.*001, *d* = 1*.*12] relative to the low activity group. Regardless of treatment, resting HR was significantly lower in the high activity group in comparison to the low activity group [*t(*38*)* = 1*.*71, *p* = 0*.*048 (one-tailed), *d* = 0*.*55], suggesting cardiovascular adaptations in the high activity group that may be explained by the link between bradycardia, regular exercise and physical fitness. Four participants were identified as multivariate outliers on HR variables using Tabachnick and Fidell's (2007) multivariate outliers identification procedure, and were subsequently excluded from further analysis. Forty participants aged between 18 and 47 (*M* = 23.70, *SD* = 5*.*89) were included in final analyses.

#### **HORMONAL RESULTS**

The χ<sup>2</sup> tests demonstrated that participants were equally distributed between menstrual phases on placebo, χ<sup>2</sup> *(*2*)* = 0*.*950, p = 0.622, and escitalopram, χ<sup>2</sup> *(*2*)* = 0*.*950, *p* = 0*.*622, treatment sessions. There was an equal distribution of participants who took hormonal contraceptives to those who did not, χ<sup>2</sup> *(*2*)* = 0*.*100, *p* = 0*.*752 and hormonal contraceptive status was equally distributed between those who received placebo or escitalopram at their first session, χ<sup>2</sup> *(*1*)* = 0*.*002, *p* = 0*.*962. Regression analyses on HR and HRV data for each task and treatment condition showed no significant associations with progesterone or estradiol concentration [all *p >* 0*.*05 (Bonferroni corrected for multiple analyses)]. Additionally, there were no differences between those who did and did not take hormonal contraceptives on each task and treatment condition [HR, *F(*1*,* <sup>38</sup>*)* = 0*.*544, *p* = 0*.*465; HRV, *F(*2*,* <sup>37</sup>*)* = 0*.*733, *p* = 0*.*487].

#### **TREATMENT BLINDING MANIPULATION CHECK**

A greater proportion of participants (70%) were found to correctly guess treatment condition than that which was expected by chance, χ<sup>2</sup> *(*1*)* = 6*.*400, *p* = 0*.*011. Side effects also occurred in 48% of the sample, although these did not occur in a greater proportion than that expected by chance, χ<sup>2</sup> *(*1*)* = 0*.*100, *p* = 0*.*752. There was, however, a significant association between the subjective experience of side effects and correct treatment guessing, χ2 *(*1*)* = 6*.*535, *p* = 0*.*011, indicating that participants who correctly guessed treatment condition were likely to have made this correct guess based on their experience of side effects. Repeatedmeasures ANOVAs testing the interaction between presence of side effects and correct guessing of treatment were performed on the HR and HF HRV measures. Importantly, no significant interaction effects of correct guessing × side effects × treatment × task on HR [*F(*1*,* <sup>36</sup>*)* = 0*.*428, *p* = 0*.*517] or HRV [*F(*1*,* <sup>35</sup>*)* = 0*.*995, *p* = 0*.*380] were observed.

#### **SUBJECTIVE RATINGS OF TASK**

A significant interaction between task and physical activity groupings on subjective ratings was observed, [*F(*1*,* <sup>32</sup>*)* = 7*.*17, *<sup>p</sup>* <sup>=</sup> <sup>0</sup>*.*012, <sup>η</sup><sup>2</sup> *<sup>p</sup>* = 0*.*18]. Follow-up pairwise comparisons demonstrated that participants in the high physical activity group reported significantly lower mental stress during the stress task, than the low physical activity group, *t(*32*)* = 2*.*84, *p* = 0*.*008, d = 1.00. Therefore, the less physically active group found the mental stress task more stressful than the more physically active

#### **Table 1 | Participant demographic information.**


*Abbreviations: PA, physical activity; BMI, body mass index; PHQ-9, patient health questionnaire; GAD-7, generalised anxiety disorder screener; IPAQ, international physical activity questionnaire; MET, metabolic equivalent task.*

group, regardless of treatment. Importantly, however, a main effect of task was also observed indicating that participants were more stressed in the arithmetic task than during the rest condition [*F(*1*,* <sup>32</sup>*)* <sup>=</sup> <sup>102</sup>*.*77, *<sup>p</sup> <sup>&</sup>lt;* <sup>0</sup>*.*001, <sup>η</sup><sup>2</sup> *<sup>p</sup>* = 0*.*76], confirming the validity of the stress condition in our sample, regardless of physical activity groupings. No main effects of physical activity [*F(*1*,* <sup>32</sup>*)* = 2*.*49, *p* = 0*.*124] or treatment [*F(*1*,* <sup>32</sup>*)* = 2*.*35, *p* = 0*.*135], or interactions between treatment and task [*F(*1*,* <sup>32</sup>*)* = 2*.*26, *p* = 0*.*142] or treatment, task, and physical activity [*F(*1*,* <sup>32</sup>*)* = 0*.*228, *p* = 0*.*637] were observed on subjective ratings.

# **IMPACT OF ESCITALOPRAM AND THE MODERATING EFFECTS OF PHYSICAL ACTIVITY**

Descriptive statistics for HR and HRV in high and low activity groupings for task and treatment are presented in **Figure 2** and **Table 2**.

#### *Heart rate*

A three-way interaction between treatment, task and activity level was significant, [*F(*1*,* <sup>38</sup>*)* <sup>=</sup> <sup>7</sup>*.*18, *<sup>p</sup>* <sup>=</sup> <sup>0</sup>*.*<sup>011</sup> <sup>η</sup><sup>2</sup> *<sup>p</sup>* = 0*.*16]. Relative to placebo, escitalopram reduced HR in both the low activity [*t(*38*)* = 2*.*80, *p* = 0*.*008, *d* = 0*.*45] and high activity [*t(*38*)* = 2*.*24, *p* = 0*.*031, *d* = 0*.*36] groups during rest. Escitalopram also reduced HR in the low activity group [*t(*38*)* = 4*.*97, *p <* 0*.*001, *d* = 0*.*80] during stress. By contrast, escitalopram did not reduce HR in the high activity group during stress (**Figure 2**). Notably, HR for the high activity group during stress under placebo, was markedly reduced relative to those in the low activity group [*t(*38*)* = 2*.*13, *p* = 0*.*040, *d* = 0*.*48], and equivalent to HR for the low activity group under drug (**Figure 2**), indicating that those in the high activity group display a more resilient cardiovascular response to stress than those in the low activity group. This interpretation is further supported by the observation of an interaction between task and physical activity grouping on subjective ratings of task (reported above). These findings were observed on a background of a significant treatment × task interaction [*F(*1*,* <sup>38</sup>*)* <sup>=</sup> <sup>6</sup>*.*61, *<sup>p</sup>* <sup>=</sup> <sup>0</sup>*.*014, <sup>η</sup><sup>2</sup> *<sup>p</sup>* = 0*.*15], a main effect of treatment [*F(*1*,* <sup>38</sup>*)* <sup>=</sup> <sup>22</sup>*.*18, *<sup>p</sup> <sup>&</sup>lt;* <sup>0</sup>*.*001, <sup>η</sup><sup>2</sup> *<sup>p</sup>* = 0*.*37], and a main effect of task [*F(*1*,* <sup>38</sup>*)* <sup>=</sup> <sup>210</sup>*.*43, *<sup>p</sup> <sup>&</sup>lt;* <sup>0</sup>*.*001, <sup>η</sup><sup>2</sup> *<sup>p</sup>* = 0*.*85] collapsed across physical activity groupings. These findings indicate that the increase in HR under stress (relative to rest) was decreased by escitalopram (a treatment × task interaction), that HR was decreased by escitalopram (a main effect of treatment) and that HR was increased under stress (a main effect of task).

#### *Heart rate variability*

No significant three-way interaction was found for HRV [*F(*2*,* <sup>37</sup>*)* = 0*.*23, *p* = 0*.*80]. Univariate tests for the three-way interaction between treatment, task and vigorous activity were not significant for HF *F(*1*,* <sup>38</sup>*)* = 0*.*06, *p* = 0*.*804, or RMSSD, *F(*1*,* <sup>38</sup>*)* = 0*.*47, *p* = 0*.*496. However, a main effect of task was observed indicating that HRV was decreased by stress, relative to the rest, regardless of treatment [RMSSD, *F(*1*,* <sup>38</sup>*)* = 35*.*99, *p <* 0*.*001, *d* = 0*.*96; and HF, *F(*1*,* <sup>38</sup>*)* = 84*.*60, *p <* 0*.*001, *d* = 1*.*47]. Interestingly, *post-hoc*, pairwise comparisons indicated that escitalopram—relative to placebo—increased RMSSD (but not HF) during stress in the low activity group [*t(*37*)* = 2*.*07, *p* = 0*.*046, *d* = 0*.*33], mirroring the findings of HR. No other pairwise comparisons were significant.

#### **DISCUSSION**

Recent reports on the adverse cardiovascular effects of the secondgeneration antidepressants (Whang et al., 2009; Licht et al.,

**Table 2 | HR and RMSSD means and standard deviations for vigorous exercise groups under each treatment for rest and stress conditions.**


2010a), highlight the importance of research on specific lifestyle factors such as physical activity. Major findings observed here were that (a) participants in the high physical activity group (regular vigorous exercisers) reported feeling less stressed than those in the low physical activity grouping (irregular exercisers) during the stress task (i.e., an interaction between task and physical activity groupings on subjective ratings on task); this finding was associated with a large effect size (Cohen's *d* = 1*.*00), (b) all participants reported feeling more stressed in the stress vs. rest condition (i.e., a main effect of task), a finding associated with a large effect size (η<sup>2</sup> *<sup>p</sup>* = 0*.*76), providing an important validation of our task in the context of an interaction between group and physical activity grouping, (c) irregular—relative to regular—exercisers display attenuated cardiovascular responses to stress following treatment with escitalopram (i.e., a three-way interaction between treatment, task and activity for HR, also indicated in a *posthoc* test on HRV), a finding associated with a moderate to large effect size (HR: Cohen's *d* = 0*.*80; RMSSD HRV: Cohen's *d* = 0*.*33), and (d) regular—relative to irregular—exercisers displayed a markedly lower HR under placebo (i.e., a three-way interaction between treatment, task and activity for HR), a finding associated with a moderate effect size (Cohen's *d* = 0*.*48).

We show here that irregular exercisers—those participants that did not engage in 30 min of vigorous physical activity at least 3 days a week—reported feeling more stressed after the stress task and following an acute dose of escitalopram treatment, they displayed improvements in vagally mediated cardiovascular function during stress. Importantly, these subjective and objective findings relating to the impact of acutely administered escitalopram during stress were also observed regardless of physical activity grouping. The interactions with physical activity simply highlight that the beneficial effects of escitalopram were greatest in irregular exercisers. This effect was most prominent for HR, but was also observed in planned comparisons for HR variability. The impact of escitalopram on HRV under stress was not as robust as the observed findings for HR, however, this observation is understandable because under conditions of stress, parasympathetic activity, which is the primary driver of changes in HRV, is withdrawn and this reduces the capacity of other factors to moderate reductions in HRV during stress (observed as a main effect of task).

While animal and human studies (Babyak et al., 2000; Engesser-Cesar et al., 2007; Arunrut et al., 2009) have indicated that the combination of exercise and antidepressant medication may confer no advantage over either treatment alone, there is another issue at stake here. Although antidepressant medication—including the tricyclic medications, the serotonin and noradrenaline reuptake inhibitor and the selective serotonin reuptake inhibitors—may have short- to medium-term benefits including the amelioration of depressive symptoms and increased resilience to stress, research has begun to highlight the longerterm adverse cardiovascular effects of antidepressants (Whang et al., 2009; Licht et al., 2010a). Reductions in HRV have been attributable to the specific effects of antidepressants (Licht et al., 2008, 2010a) while increases in HRV are associated with their cessation (Licht et al., 2010a). An epidemiological study (Whang et al., 2009) on 63,469 women aged 30–55 years without baseline coronary heart disease (CHD) reported that while depressive symptoms were associated with fatal CHD, antidepressant use (61% of participants were using an SSRI) was specifically associated with a 3.34 increased risk for sudden cardiac death even after controlling for a variety of confounds. A more recent study has reported that reduced cardiovascular fitness is associated with an increased risk of suicide over a 42 year follow-up period (Åberg et al., 2013) highlighting the importance of cardiovascular health over the lifespan. Together, these findings highlight the importance of research on the cardiovascular effects of antidepressants and specific lifestyle factors such as physical activity. The present study makes an important contribution to this effort highlighting that the cardiovascular effects of regular vigorous exercise and of an acute dose of escitalopram during stress are comparable, i.e., both are associated with a decrease in HR and an increase in HRV.

Here we propose a simplified model for the beneficial effects of regular vigorous physical activity and SSRI treatment on vagal tone, initiating a path toward mental and physical wellbeing (see **Figure 3**). Increased vagal function is associated with increased psychophysiological flexibility (Thayer et al., 2009; Kashdan and Rottenberg, 2010) and stress resilience, which in turn promotes wellbeing (Oveis et al., 2009; Geisler et al., 2010, 2013; Kashdan and Rottenberg, 2010; Kok and Fredrickson, 2010; Thayer et al., 2010b; Kemp and Quintana, 2013). Increased vagal function also initiates downstream neurophysiological changes such as enhanced growth factors (Follesa et al., 2007), improved regulation of HPA axis function and inflammatory processes (Tracey, 2002; Thayer and Sternberg, 2006; Huston and Tracey, 2010; Pavlov and Tracey, 2012), leading to improved wellbeing (Thayer et al., 2010b; Kemp and Quintana, 2013). The model also highlights the direct links between regular vigorous physical activity and wellbeing (discussed above). However, it is noted that overtrained athletes also display a decease in HRV, which limits the autonomic resources available to respond to emotional and physical stress (Hynynen et al., 2008; Boullosa et al., 2012).

In the present study we demonstrated that acute administration of escitalopram is associated with increased vagal function under both resting and task conditions; a finding that was particularly robust for HR. This finding is consistent with reported short- to medium-term beneficial effects of SSRIs on cardiovascular and neuroendocrine responses to stress in depressed patients (e.g., Straneva-Meuse, 2004b), but contrast against longer-term research outcomes (Whang et al., 2009; Licht et al., 2010a) highlighting the adverse effects of the SSRI class of antidepressants. Here, an explanation for this apparent discrepancy is the complexity of central and autonomic 5-HT effects on cardiovascular function, which include bradycardia, associated with activation of 5-HT1A receptors, as well as tachycardia, associated with activation of 5HT2 receptors. It is possible that the effects of SSRIs shift from parasympathetic—as shown here to sympathetic activation with increasing length of use. Further research is urgently needed on the combination of antidepressants and physical activity over longer-time frames than typical clinical trials (i.e., years rather than months).

A number of limitations are worth noting. Participants correctly guessed treatment condition beyond chance due to side effects, consistent with those typically experienced with escitalopram (Cipriani et al., 2009b). However, neither experience of side effects, nor accuracy of treatment guess impacted on HR or HRV. Experimenters were also blinded to individual participants' physical activity levels and participants were not aware that physical activity was of primary interest. It is therefore unlikely that the experience of side effects and correctly guessing treatment would have confounded the obtained findings. Another limitation is the use of a questionnaire based-measure of physical activity. It is possible that our measure reflects recent regular physical activity, rather than fitness *per se*. Future studies could consider use of objective measures of fitness such as VO2 max to confirm that those engaging in regular activity are indeed characterized by higher levels of fitness. Future studies with clinical samples and chronic treatment are also needed to further examine the utility of combining antidepressant medication and vigorous physical activity.

In summary, results from this experimental study highlight the adverse impact of stress on vagally mediated cardiovascular function and the beneficial moderating effects of escitalopram as well as regular vigorous physical exercise. Escitalopram was observed to increase vagal function under both rest and stress conditions, and under stress, these effects are specific to irregular exercisers, but not regular exercisers. We propose a simplified model for understanding the beneficial effects of physical activity and antidepressants on cardiovascular function that may subsequently lead to improved mental and physical activity. Our study and model provide a foundation on which future research could be based. Increasing concerns about the bidirectional relationship between depression and CVD, reinforce the need for further research on the impact of antidepressants, physical activity and their combination, that is not only focused on amelioration of depressive symptoms, but also on vagally mediated cardiovascular function.

# **DISCLOSURE**

This research was supported by an Australian Research Council Discovery Project Grant (DP0987332), a National

# **REFERENCES**


Health and Medical Research Council (NHMRC) Project Grant (464863) and a NHMRC Career Development Award (571101) awarded to Andrew H. Kemp. The authors Andrew H. Kemp and Tim Outhred are currently supported by an International Visiting Research Professorship from the Universidade de São Paulo and an Australian Postgraduate Award, respectively. Camilla S Hanson was supported by a Charles Perkins Centre Summer Research Scholarship.

# **ACKNOWLEDGMENTS**

This research was supported by an Australian Research Council Discovery Project Grant (DP0987332) and a National Health and Medical Research Council Project Grant (464863) awarded to Andrew H. Kemp. The authors would like to acknowledge support and contributions of the following people at different stages of the project: Prof Richard Bryant, Dr Kim Felmingham, Kristi Griffiths, Sasha Saunders, Kristy-Lee Feilds, Mimi Leith, Daniel Quintana, James Heathers, Caroline Fields, Jonathan Kreiger, Sara Shahrestani, Julie Ji, and Matthew Beauregard.

treatment have differential effects in the hippocampus and the spinal cord. *Neuroscience* 144, 1033–1044. doi: 10.1016/j.neuroscience.2006. 10.016


mediated by emotion regulation. *Pers. Individ. Dif.* 49, 723–728. doi: 10.1016/j.paid.2010.06.015


of antidepressant drug action. *Br. J. Psychiatry* 195, 102–108. doi: 10.1192/bjp.bp.108.051193


*N. Z. J. Psychiatry*. 46, 289–292. doi: 10.1177/0004867412445131


H., Glynn, R. J., et al. (2009). Depression and risk of sudden cardiac death and coronary heart disease in women. *J. Am. Coll. Cardiol.* 53, 950–958. doi: 10.1016/j.jacc.2008.10.060

**Conflict of Interest Statement:** Gin S. Malhi has received research support from AstraZeneca, Eli Lilly, Organon, Pfizer, Servier, and Wyeth. He has been a speaker for AstraZeneca, Eli Lilly, Janssen Cilag, Lundbeck, Pfizer, Ranbaxy, Servier, and Wyeth. He has been a consultant for AstraZeneca, Eli Lilly, Janssen Cilag, Lundbeck, and Servier. Andre R. Brunoni declares no potential conflict of interest.

*Received: 24 June 2013; accepted: 03 September 2013; published online: 24 September 2013.*

*Citation: Hanson CS, Outhred T, Brunoni AR, Malhi GS and Kemp AH (2013) The impact of escitalopram on vagally mediated cardiovascular function to stress and the moderating effects of vigorous physical activity: a randomized controlled treatment study in healthy participants. Front. Physiol. 4:259. doi: 10.3389/fphys.2013.00259*

*This article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology.*

*Copyright © 2013 Hanson, Outhred, Brunoni, Malhi and Kemp. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Traditional games resulted in post-exercise hypotension and a lower cardiovascular response to the cold pressor test in healthy children

*Suliane B. Rauber <sup>1</sup> \*, Daniel A. Boullosa1,2, Ferdinando O. Carvalho1,3, José F. V. N. de Moraes 1,3, Ioranny R. C. de Sousa1, Herbert G. Simões <sup>2</sup> and Carmen S. G. Campbell 1,2*

*<sup>1</sup> Graduate Program on Physical Education and Health, Laboratory Study of Physical Activity and Health, Catholic University of Brasilia, Brasilia, Brazil <sup>2</sup> Undergraduate Program on Physical Education, Study Group of the Benefits of Physical Activity for Children's Health (GEBEXFISI-Grupo de Estudos dos*

*Benefícios do Exercício Físico para Saúde Infantil), Catholic University of Brasilia, Brasilia, Brazil*

*<sup>3</sup> Department of Physical Education, College of Physical Education, Federal University of Vale do São Francisco, Petrolina, Brazil*

#### *Edited by:*

*Arto J. Hautala, Verve Research, Finland*

#### *Reviewed by:*

*Douglas L. Jones, The University of Western Ontario, Canada Silja Angervo Martikainen, University of Helsinki, Finland*

#### *\*Correspondence:*

*Suliane B. Rauber, Laboratory Study of Physical Activity and Health, Catholic University of Brasilia, QS 07, Lote 01, Águas Claras, CEP 71966-700, Brasilia, Brazil e-mail: suliane.edfisica@gmail.com*

The present study aimed to verify if blood pressure (BP) reactivity could be reduced through a previous single session of active playing when compared to sedentary leisure. Sixteen pre-pubertal healthy children participated in this study. After familiarization with procedures and anthropometric evaluation, participants performed three sessions in randomized order: (1) 30 min of traditional Brazilian games (PLAY); (2) 30 min of video game playing (DDR); and (3) 30 min of watching TV (TV). Each session lasted 80 min, being 10 min of rest; 30 min of intervention activity; and 40 min of recovery. After recovery, the Cold Pressor Test (CPT) was used for the assessment of acute cardiovascular reactivity. BP was recorded at 30 s and 1 min during the CPT. Analysis of variance showed post-exercise hypotension (PEH) only after PLAY, and that systolic and diastolic BP were significantly increased in all conditions during CPT. However, the magnitude of the CPT-induced BP response was significantly less in PLAY compared to DDR and TV. The PEH observed during recovery and the reduced BP response to CPT following playing traditional games may be due its higher cardiovascular and metabolic demand as was indicated by the increased heart rate, oxygen consumption, and BP. It was concluded that BP reactivity to stress may be reduced through a previous single session of traditional games and that PEH was recorded only after this exercise form. This benefit indicates a potential role of playing strategies for cardiovascular health in childhood.

**Keywords: cardiovascular system, blood pressure, heart rate, children, active playing**

## **INTRODUCTION**

Cardiovascular disease (CVD) is the primary cause of death in western countries (Ergin et al., 2004). More recently, researchers have provided evidence that CVD is becoming more frequent among pediatric populations (Bigi et al., 2011) and the increase of blood pressure (BP) during infancy and puberty is becoming more premature (Kark et al., 2009; Maximova et al., 2009), and has been identified as the first etiologic factor for the development of hypertension in adulthood (Chiolero et al., 2008; Tsioufis et al., 2011). Therefore, it is important to identify factors that could influence cardiovascular health during childhood, specifically those lifestyle-related that are potentially modifiable (Bell and Belsky, 2008; Daniels, 2011; Moraes et al., 2014).

Physical inactivity is considered to be one of the main causes of primary hypertension, independently of weight status (Martinez-Gomez et al., 2009; Daniels, 2011). Physical activity (PA) levels are becoming lower among children and adolescents worldwide (Rossow et al., 2010; Nettle and Sprogis, 2011), demonstrating that new generations are more sedentary than the previous ones (Chinapaw et al., 2011). This raises concerns about which strategies could help children and adolescents to increase their PA levels, as PA in childhood has been demonstrated to be an effective intervention to combat obesity and hypertension in adult life (Guy et al., 2011; Siegrist et al., 2013).

Among several mechanisms related to the effectiveness of PA on BP control, there is the so-called post-exercise hypotension (PEH) (MacDonald, 2002). The effect of a single exercise bout induced BP reduction in relation to a non-exercise control session has been systematically observed after different exercise modes can last several hours post-exercise (Ciolac et al., 2008; Mota et al., 2009; Rabelo et al., 2009; Morais et al., 2011), thus helping for control BP in both hypertensive (Mota et al., 2009) and healthy people (Richter et al., 2010). However, the occurrence of PEH after exercise in children was not investigated yet to our knowledge.

An important application of PEH is that, during its occurrence, individuals are supposed to present lowered BP reactivity to different stress situations (MacDonald et al., 2001; Brownley et al., 2003). Stress can be considered an important factor that typically promotes a rise in BP (Lehman et al., 2009), which is dependent on the nature of the stimulus and the individual characteristics of the subject (Probst et al., 1997). Previous studies have demonstrated that an exercise bout of moderate intensity could attenuate the increase of BP and HR after a stressful event (Probst et al., 1997; Bond et al., 2002) in adults. More recently, the studies of Roemmich et al. (2009) and Lambiase et al. (2010) have showed a dampened cardiovascular reactivity to speech and stroop tasks after interval cycling and a self-paced walk respectively, in children. These important studies (Roemmich et al., 2009; Lambiase et al., 2010) highlighted the positive impact of different forms of exercise performed in laboratory conditions on acute stress reactivity. However, the short recovery interval used (20 min) in these previous studies did not allow a precise evaluation of the PEH phenomenon. Moreover, it would be interesting to evaluate also the impact of other activities more enjoyable to children, such as street games and active video game playing, that have been recently proposed to increase energy expenditure (Lanningham-Foster et al., 2006; Wang and Perry, 2006; Warburton et al., 2009; Rauber et al., 2013). This information would help to know the impact of other forms of exercise on both PEH and stress reactivity in children; and therefore to better understand the role of different exercise strategies for cardiovascular health during childhood.

On the other hand, the cold pressure test (CPT) is a simple and time-saving protocol proposed for the evaluation of cardiovascular reactivity to acute thermal stress (Hines and Brown, 1936; Wood et al., 1984) that requires passive coping in contrast to the active coping required in previous studies with children (Roemmich et al., 2009; Lambiase et al., 2010). Briefly, this test consists of the immersion of the left hand in cold water for the assessment of the BP responses (Hines and Brown, 1936). The validity of this test has been previously demonstrated as it is a good predictor of hypertension in adulthood from the recorded values during childhood (Wood et al., 1984; Menkes et al., 1989).

Thus, the purpose of present study was to analyze the occurrence of PEH after different games (i.e., video games and traditional Brazilian children's games), and to assess the potential protective effect of these games on cardiovascular reactivity to subsequent stress in children. The hypotheses were that both games would induce PEH and that cardiovascular reactivity would be lower after active games in comparison to the sedentary activity (TV).

# **METHODS**

#### **SAMPLE AND ETHIC PROCEDURES**

Sixteen children aged 9–10 years (8 boys and 8 girls) participated in this study. After authorization from the school, the children's parents received an informed consent letter describing all procedures, risks and benefits of the study and signed the consent for children's participation.

The exclusion criteria were: any physical impairment that could prevent the participant from performing the programmed activities; having any previous cardiometabolic disease (e.g., diabetes, hypertension); and any previous experience with the video game "Dance Dance Revolution" (DDR). The study was approved by the local ethics committee for human research (protocol 147/2009).

### **GENERAL PROCEDURES**

Four visits were necessary in order to perform the present study. The first one was for familiarization with the experimental procedures and for anthropometric evaluation. The other three visits were randomly performed between 8 and 11 a.m. Each visit lasted 80 min, divided in: rest (10 min); programmed activity (30 min); and recovery (40 min). Immediately after the 40th min of the post-activity recovery period, the CPT was applied over 1 min.

# *Familiarization and anthropometric evaluation*

Evaluation of body weight (Electronic Scale Tech 05®, China), height (Portable Stadiometer, Sanny®, Brazil), and subscapular and triceps skinfolds (Adipometer, Lange Skinfold Caliper, USA) were performed for body composition estimation. Body fat percentage was estimated using the skinfolds method according to Slaughter et al. (1988). The CPT and BP measurements were also performed so the children were familiarized with these procedures. Subsequently, the children received orientations and practiced the DDR video game 10–15 times.

# *Traditional games (PLAY)*

Volunteers participated of a physical education class together with other 10 children so the activity could be the most realistic possible. A physical education teacher guided an activity composed by three traditional games usually applied in physical education classes in Brazil, and organized as follows: 10 min of "run and catch;" 10 min of "dodge ball;" and 10 min of "capture the flag." These activities were performed in the same order in a gymnasium close to university's laboratory. The intensity of the games was monitored by HR (Polar Electro Oy FS1, Finland), as well as through the responses of oxygen uptake (VO2), and respiratory exchange ratio (RER) (Cortex Metamax 3B, Germany).

# *"Dance Dance Revolution" video game (DDR)*

Volunteers played an interactive video game (DDR) installed in a laptop computer. They selected the songs from a low-level difficulty list. Authors decided to use a basic level for standardization purposes, since their first contact with this video game was during the familiarization session.

# *Watching television (TV)*

Volunteers watched two popular cartoons ("Sponge Bob" and "Ben 10") for 30 min (15 min per cartoon) while seated in a quiet room.

# *BP and HR measurements*

Systolic (SBP), diastolic (DBP) BP and HR were collected during each session at the 5th and 10th min of rest before sessions; at the 10th, 20th, and 30th min during sessions; and at the 2nd, 10th, 20th, 30th, and 40th min after all sessions (see **Figure 1**). Mean arterial pressure (MAP) was calculated using the following equation: MAP = DBP + [(SBP − DBP)/3]. BP was measured, in the right arm, with the volunteer seated, through the auscultatory method using a pediatric sphygmomanometer with an appropriate cuff for children (Becton Dickinson, Brazil) and a children's stethoscope Duo Sonic (Missouri, Brazil). Heart rate was measured using a HR monitor (FS1, Polar Electro Oy, Finland).

# *Cold Pressor Test (CPT)*

This thermal stress test was performed after the completion of 40 min of recovery after each session. The volunteers immersed the left hand of the child in cold water (∼4-5◦C) for 60 s. During this period, BP was measured at the 30th and 60th s, and subsequently compared with BP at rest. This test was performed in order to verify the BP response after the different sessions (PLAY, DDR, and TV). BP was measured using auscultatory method by the same evaluator, on the arm opposite to that immersed one, and HR was also measured using the same procedures and equipment cited above.

#### **STATISTICAL ANALYSIS**

Shapiro-Wilk's test was used to confirm the normality of data. One-Way ANOVA with Scheffé as a *post hoc* were applied to compare variables between sessions (PLAY, DDR, and TV). ANCOVA for repeated measures adjusted for exercise intensity (i.e., VO2) was applied at different time points during recovery and CPT after TV, DDR, and PLAY sessions. The level of significance was set at *p* ≤ 0*.*05. The data was analyzed using the Statistical Package for the Social Sciences (SPPS), version 15.0 for windows.

# **RESULTS**

The characteristics of the children were: mean age 9.6 ± 0.5 years, 133.8 ± 9.9 cm of height, weight 32.4 ± 4.0 kg with a mean body mass index (BMI) of 18.4 <sup>±</sup> 3.7 kg· <sup>m</sup>−2, and body fat of 17.3 <sup>±</sup> 3.8 %. All children were characterized as eutrophic.

**Table 1** shows the cardiovascular and ventilatory responses at rest and during each experimental session. Heart rate, SBP, DBP, and MAP were significantly higher during PLAY and DDR when compared to TV. Also, PLAY elicited a significantly higher VO2 and RER when compared to TV. Furthermore, all cardiovascular and ventilatory responses were significantly higher during PLAY in comparison to DDR session.

**Figures 2**–**4** illustrate HR and BP responses during CPT. One of the volunteers was considered hyper-reactive to the test, having an increase in SBP of ≥25 mmHg during the test (Wood et al., 1984). Heart rate increased 4, 7.2, and 7.5 bpm at the 30th s, and 4.4, 8.2, and 8 bpm at the 60th s during the CPT after TV, DDR and PLAY, respectively. However, when adjusted for exercise intensity, only PLAY elicited a significant HR increase at both moments when compared to pre-exercise rest.

**Figures 3**, **4** show the responses of SBP and DBP, respectively, at the 30th and 60th s during CPT after each session. PLAY elicited a significantly smaller increase in SBP and DBP when compared to TV and in SBP when compared to DDR at both moments. When adjusted for exercise intensity, all sessions elicited a significant SBP increase at both moments when compared to pre-exercise rest with the extent of this change in SBP being greater in TV with respect to DDR and PLAY, and lower in PLAY with respect to TV and DDR. Meanwhile, only PLAY exhibited a lower DBP at both moments when compared to TV. Additionally, SBP at rest was significantly lower in PLAY when compared to TV.

**Tables 2**, **3** show the mean values of SBP and DBP over each session and subsequent recovery. With or without adjustment for exercise intensity, SBP and DBP in DDR and PLAY were significantly higher at the 10th, 20th, and 30th min of intervention when compared to TV. PLAY also showed significantly higher values when compared to DDR at all times for SBP and DBP, but only at the 10th and 20th for DBP when controlled for exercise intensity. During the recovery period, SBP and DBP were significantly reduced after adjustment for exercise intensity at the 30th min, and SBP at the 40th min when compared to rest only in PLAY. Interestingly, SBP at 2nd min of recovery in DDR was different from rest in the same session before adjustment for exercise intensity. Moreover, SBP at the 40th min of recovery and DBP at the 2nd of recovery in PLAY were significantly different from the values recorded after TV session before adjustment for exercise intensity. In contrast, DBP was not different from values recorded after TV at the 30th and 40th min of recovery before adjustment for exercise intensity.

# **DISCUSSION**

The main findings of present study were that children experienced PEH after traditional games and that BP reactivity to stress induced by CPT was also lower after traditional games when compared to active video game playing and TV sessions. These responses may be related to exercise intensity once the more intense session (PLAY) also induced a more significant hypotensive effect during the post-exercise recovery period. Moreover, this lowered BP reactivity was observed 40 min after the end of the experimental session, which is longer than that observed in previous studies (Roemmich et al., 2009; Lambiase et al., 2010).


**Table 1 | Cardiovascular and ventilatory responses at rest and during each session.**

*Values expressed as mean* ± *standard deviation. TV, television; DDR, dance dance revolution; PLAY, traditional children's game; HR, heart rate; SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; VO*2*, oxygen uptake; RER, respiratory exchange ratio. \*p* <sup>≤</sup> *0.01 in relation to TV; ‡p* <sup>≤</sup> *0.01 in relation to DDR.*

Overall, these findings suggest that traditional games are better suited for cardiovascular health in children when compared to other activities of lower cardiovascular and metabolic stimuli like active video game playing or watching TV.

To the best of our knowledge, this is the first study reporting PEH in children. Thus, significant reductions in SBP and DBP were observed in children at the 30th and 40th min of recovery when compared to rest only after PLAY. Interestingly, DBP was significantly reduced when compared to TV session after adjustment for exercise intensity. In contrast, SBP at the 40th min of recovery was not different from SBP values recorded

**Test.** TV, television; DDR, dance dance revolution; PLAY, traditional games; <sup>∗</sup>*<sup>p</sup>* <sup>≤</sup> <sup>0</sup>*.*05 vs. TV in the same moment; †*<sup>p</sup>* <sup>≤</sup> <sup>0</sup>*.*05 vs. rest in the same session.

after TV session when controlled for exercise intensity. From a methodological point of view, these findings are interesting and suggest the necessity in further studies of comparing the effect of different activities matched for exercise intensity on cardiovascular responses. Additionally, as the recovery period lasted only 40 min in the current study, further studies should evaluate if the PEH exerts its positive influence over longer periods in children.

Previously, Roemmich et al. (2009) reported that previous interval exercise protocol in cycloergometer also dampened the subsequent cardiovascular response to a speech stressor test when compared to watching TV. Similarly, Lambiase et al. (2010) have reported that the cardiovascular reactivity during cognitive stress was dampened after a treadmill walking when compared to a simulated sedentary drive to school. These results are consistent with the results of the current study, and also with previous studies in adults (Probst et al., 1997; Bond et al., 2002), thus confirming the positive impact that acute exercise has on stress tolerance. However, this is the first study that reproduces traditional games which is a natural condition that imitates the habitual patterns of PA in children. Furthermore, the greater positive impact of traditional games (PLAY) could be due to its higher intensity as inferred from the higher exercising HR, BP, and VO2 recorded when compared to active video game playing, which have exhibited similar values as in previous reports (Wang and Perry, 2006). Therefore, this suggests that street playing in a non-controlled



*TV, television; DDR, dance dance revolution; PLAY, traditional games; \*p* <sup>≤</sup> *0.05 vs. TV in the same moment; †p* <sup>≤</sup> *0.05 vs. rest in the same session; ‡p* <sup>≤</sup> *0.05 vs. DDR in the same moment; §p* ≤ *0.05 vs. 2 min recovery in the same session;* #*<sup>p</sup>* <sup>≤</sup> *0.05 vs. 10 min recovery in the same session.*

**Table 3 | Diastolic blood pressure mean values (±***SD***) of all sessions (TV, DDR, and PLAY) (***n* **= 18).**


*TV, television; DDR, dance dance revolution; PLAY, traditional games; \*p* <sup>≤</sup> *0.05 vs. TV in the same moment; †p* <sup>≤</sup> *0.05 vs. rest in the same session; ‡p* <sup>≤</sup> *0.05 vs. DDR in the same moment; §p* ≤ *0.05 in relation to 2 min recovery in the same session.*

environment is an effective way for reducing the cardiovascular reactivity in children. PLAY seems to be more motivating, thus naturally increasing the levels of PA in an easy manner.

While we did not evaluate the underlying mechanisms that accounts for these differences, it may be suggested that the greater intensity achieved by children during street playing, as reflected in a greater HR increase (i.e., 72% for HRmax during PLAY vs. 54% HRmax during DDR; Machado and Denadai, 2011), and thus the occurrence of PEH after this more intense intervention, may have a role in a lowered post-exercise BP reactivity. Previous studies have reported that a greater sympathetic activation during exercise could mediate a lower post-exercise reactivity to stress (Lovallo, 2005). In this regard, Brownley et al. (2003) have demonstrated that a single bout of aerobic cycle exercise in adults reduces the post-exercise norepinephrine production in response to stress, favoring a lower cardiovascular reactivity that could also be related to a greater vasodilatory effect of exercise. Additionally, Negrão et al. (2003) have also suggested that exercise could induce a greater nitric oxide release after exercise that could elicit the vasodilatation response, and consequently favor both the PEH and the lower BP reactivity to post-exercise stress. As we did not evaluate possible mechanisms in the current study, it may be suggested that exercise intensity and subsequent metabolic demands could be the major factor accounting to these differences between conditions, as recently reported (Eicher et al., 2010; Simões et al., 2010). Additionally, the significant increase in HR during the CPT after PLAY that was only evident when cardiovascular responses were controlled for exercise intensity, may suggest the existence of other mechanisms (e.g., nitric oxide) different from autonomic control that may account for both the reduced BP reactivity and PEH as suggested in a previous study with women exhibiting PEH after resistance exercises (Tibana et al., 2013). Therefore, further studies are needed in order to determine the exact physiological mechanisms that account to these differences in cardiovascular reactivity in children after exercise, considering exercise mode and intensity.

The current study has some limitations. First, the number of boys and girls is limited and do not allow comparisons between genders. Second, the CPT does not reproduce the habitual source of stress in children's daily activities although is a valid, simple and time saving tool for cardiovascular stress reactivity evaluation. Therefore, further studies should be conducted for testing the potential differences between boys and girls regarding other stressors, such as those experienced at school, in a more realistic environment. Third, we only recorded cardiovascular responses within 40 min of recovery. Therefore, it is unknown how these responses could be maintained in the longterm (e.g., 24 h). Lastly, it is well known that genetic factors and ethnicity could be related to hypertension in adulthood (Douglas et al., 2003), but no evidence is available regarding such differences in children. In this scenario, as our sample was composed of blacks and whites, it may be interesting to look for such potentially genetic differences in further studies.

### **CONCLUSION**

The most relevant finding of the present study was that the 30-min exercise session based on traditional games ("run and catch," "dodge ball," and "capture the flag") resulted both in PEH and lower BP reactivity to stress in children. These results could be due to greater exercise intensity and metabolic demands recorded during traditional games. Additional studies should be conduct to elucidate the influence of intensity, mode and duration of activities usually performed by children, the impact of these variables, and the possible mechanisms underlying the protective cardiovascular effect during subsequent stress-induced BP reactivity.

# **ACKNOWLEDGMENTS**

To CNPq (National Conselho Nacional de Desenvolvimento Científico e Tecnológico - Council for Scientific and Technological Development) for their support and scholarship for scientific research.

#### **REFERENCES**


diabetes. *J. Strength Cond. Res.* 24, 1277–1284. doi: 10.1519/JSC.0b013e3181 d67488


Wood, D. L., Sheps, S. G., Elveback, L. R., and Schirger, A. (1984). Cold pressor test as a predictor of hypertension. *Hypertension* 6, 301–306.

**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Received: 07 March 2014; accepted: 05 June 2014; published online: 25 June 2014.*

*Citation: Rauber SB, Boullosa DA, Carvalho FO, de Moraes JFVN, de Sousa IRC, Simões HG and Campbell CSG (2014) Traditional games resulted in post-exercise hypotension and a lower cardiovascular response to the cold pressor test in healthy children. Front. Physiol. 5:235. doi: 10.3389/fphys.2014.00235*

*This article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology.*

*Copyright © 2014 Rauber, Boullosa, Carvalho, de Moraes, de Sousa, Simões and Campbell. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Reduced vasodilator function following acute resistance exercise in obese women

*Nina C. Franklin1,2, Mohamed Ali 1,2, Melissa Goslawski 1, Edward Wang3 and Shane A. Phillips 1,2,4\**

*<sup>1</sup> Department of Physical Therapy, University of Illinois at Chicago, Chicago, IL, USA*

*<sup>2</sup> Integrative Physiology Laboratory, College of Applied Health Sciences, University of Illinois at Chicago, Chicago, IL, USA*

*<sup>3</sup> Department of Biomedical and Health Information Sciences, University of Illinois at Chicago, Chicago, IL, USA*

*<sup>4</sup> Department of Medicine, University of Illinois at Chicago, Chicago, IL, USA*

#### *Edited by:*

*Daniel Boullosa, Universidade Católica de Brasília, Brazil*

#### *Reviewed by:*

*Jonato Prestes, Catholic University of Brasilia, Brazil Eliseo Iglesias-Soler, Department of Physical Education and Sports-University of A Coruña, Spain*

#### *\*Correspondence:*

*Shane A. Phillips, Department of Physical Therapy, University of Illinois at Chicago, 1919 W. Taylor St. (M/C 898), Chicago, IL 60612, USA e-mail: shanep@uic.edu*

Obesity contributes to stress induced impairments in endothelium-dependent vasodilation (EDV), a precursor to atherosclerosis. Since obesity is associated with inflammation and oxidative stress, we sought to determine if a single bout of strenuous weight lifting (SWL) reduces EDV among sedentary obese adults. Participants included 9 obese (OB) (BMI 30.0–40.0 kg/m2) and 8 lean (LN) (BMI 18.5–24.9 kg/m2) sedentary young women. All participants underwent a single bout of SWL using a progressive leg-press protocol. Brachial artery flow-mediated dilation (FMD) (an index of EDV) was determined using ultrasonography before and after SWL. Sublingual nitroglycerin (NTG) was used to determine brachial artery endothelium-independent vasodilation following SWL. Brachial artery FMD was significantly reduced in OB and LN women (LN: 6.4 ± 1.6%, *p* = 0*.*22) after SWL. There was no difference in the magnitude of change pre- and post-SWL between groups (OB: −2*.*4 ± 0*.*6% and LN: −2*.*2 ± 1*.*6%, *p* = 0*.*84). Dilation to NTG was lower in OB (21.6 ± 1.3%) compared to LN women (27.6 ± 2.1%, *p* = 0*.*02) and associated with body weight (*r* = −0*.*70, *p* = 0*.*01). These data suggest that EDV is reduced in woman after acute resistance exercise. Dilations to NTG were lower in obese compared to lean woman and associated with body weight suggesting that changes in sensitivity of blood vessels to NO occurs during obesity. These findings may be important for understanding vascular risk following acute exercise in obesity.

**Keywords: endothelium, vasodilation, acute exercise, obesity, women**

# **INTRODUCTION**

Rates of overweight and obesity have reached epidemic proportions (Flegal et al., 2012) with currently 2 in 3 adults either overweight or obese. In fact, atherosclerotic cardiovascular disease (CVD) is the third leading cause of death among women between the ages of 25 and 44 and obesity independently contributes to increased risk in this particular demographic (Mosca et al., 1997). Vascular dysfunction may play a pivotal role in the development of atherosclerosis in obesity (Sturm et al., 2009). Increased body fat is associated with increased reactive oxygen species production (ROS) (Anfossi et al., 2010) and impaired nitric oxide (NO) mediated endothelium-dependent vasodilation (EDV) (Al et al., 2001). While impaired endothelium-independent dilation (EID), indicative of smooth muscle dysfunction, is also linked to the development of atherosclerosis, most research suggests that it is preserved (Arkin et al., 2008).

Resistance exercise training is recommended for preventing weight gain and reducing the risk of obesity associated cardiovascular risk (Mason et al., 2007) and may reduce the ill effects of chronic stress on cardiovascular function (i.e., psychological stressors) (Paine et al., 2013). A previous study found that resistance exercise training (2 days per week) for 18 weeks had no effect on endothelial function in healthy post-menopausal women (Casey et al., 2007), while others have found improvements in endothelial function after 16 weeks of resistance exercise when combined with other lifestyle interventions of aerobic exercise and calorie reduction (Cotie et al., 2014). Acute resistance exercise increases blood flow intermittently, yields increased shear stress and improves nitric oxide (NO)-mediated vasodilation (Tinken et al., 2010). However, acute resistance exercise is associated with elevated systolic and diastolic blood pressures (Sale et al., 1994) to levels known to impair vascular function (Jurva et al., 2006).

Previous studies on stress-induced cardiovascular dysfunction have focused on the neural and hormonal regulation of cardiovascular function following acute psychological challenges (Chrousos, 2000, 2009). Research on the direct effects of acute physical stressors on vasodilator function and the contribution of obesity to these responses has been limited. In previous studies from our laboratory brachial artery flow mediated dilation (FMD) was reduced following acute weight lifting in sedentary compared to exercise-trained men, despite similar elevations in blood pressure during exercise (Phillips et al., 2011). Resistance exercise is an important component of exercise prescription in overweight and obesity to promote lean body mass and insulin sensitivity (Laskowski, 2012). Therefore, there is a need to determine the clinical impact of different modes of exercise (including resistance exercise) on vascular function in woman since (1) young women suffer a disproportionate burden of morbidity and mortality attributable to obesity and have higher morbidity and mortality rates after myocardial infarction compared to men (Maas et al., 2011) and (2) intense exercise which may be experienced particularly in the initial stages of an exercise program may be harmful to blood vessel function. For example, acute high intensity resistance exercise (*>*80% 1RM) can reduce brachial FMD (Phillips et al., 2011; Choi et al., 2014). In contrast, moderate intensity, aerobic exercise (50–65% VO2max) has been shown to improve FMD in lean individuals (Rooks et al., 2011) or reduce FMD in overweight individuals (Harris et al., 2008). The purpose of this study was to determine whether a single bout of weight lifting reduces FMD in sedentary obese women compared to lean woman in order to determine if obesity (in the absence of other risk factors) translates to more severe reductions in FMD following a strenuous exercise session. In addition, we sought to determine if changes in FMD associated with SWL and are greater in obese compared to lean women. We hypothesized that acute physical stress (in the form of a single session of weight lifting) will reduce vascular function and that this change is augmented in obese compared to sedentary lean women.

### **MATERIALS AND METHODS**

#### **STUDY POPULATION**

Seventeen women were studied and were recruited in the community. Initially, each participant was screened by telephone and if inclusion criteria were met, they were invited for an in-person screening at which time eligibility was confirmed upon completion of a medical and exercise history questionnaire and physical examination. We included woman who were 18–40 years of age, obese (OB) (BMI 30.0–40.0 kg/m2) or lean (LN) (BMI 18.5– 24.9 kg/m2), and sedentary (less than 150 min of moderate physical activity/week).We excluded individuals with a history of CVD, diabetes mellitus or thyroid dysfunction, woman who were pregnant or recently pregnant and lactating, woman with a history of cancer, history of smoking (for at least 6 months prior to participation), amenorrhea or irregular menses, and use of vasoactive medications. Written informed consent was obtained prior to participation. The study protocol was approved by the Office for the Protection of Research Subjects and the Institutional Review Board of the University of Illinois at Chicago.

#### **STUDY DESIGN, BIOCHEMICAL, BLOOD PRESSURE, AND BODY COMPOSITION MEASUREMENTS**

All individuals were evaluated at the University of Illinois at Chicago, Clinical Research Center after a 12-h overnight fast between 0700 and 1100 in a temperature controlled study suite. **Figure 1** is a schematic summarizing the study design. Subjects were screened for eligibility at Visit 1. Approximately 1 week later (Visit 2) brachial artery flow-mediated dilation (FMD) was assessed in all subjects prior to and 30 min after performance of a single bout of bilateral lower-body strenuous weight lifting (SWL). Venous blood samples were drawn before and immediately following resistance exercise from an antecubital vein.

Plasma was separated by centrifugation for off-site laboratory analysis of total cholesterol, high-density lipoproteins (HDLs), low-density lipoproteins (LDLs), and glucose (Alverno Clinical Laboratories, LLC; Hammond, IN). Total cholesterol, HDLs, LDLs, and triglycerides were measured using spectrophotometric assays with intra-assay variances of 1.4, 3.0, 2.5, and 2.0%, respectively. Glucose concentration was measured using the glucose oxidase procedure (Beckman Autoanalyser II; Beckman Coulter Inc.; Fullerton, CA) with intra-assay variance of 1.5%. Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured after at least 10 min of rest using a standard manual mercury sphygmomanometer with an appropriate cuff size and during the last repetition of each set of exercise. Waist circumference was measured at the narrowest part of the waist (above the umbilicus and below the xiphoid process) (Ohrvall et al., 2000). Body composition was determined by dual energy xray absorptiometry (DEXA) using established methods (Mattila et al., 2007).

#### **DETERMINATION OF VASODILATOR FUNCTION**

Brachial artery FMD was used as a measure of endothelium dependent vasodilation (EDV) using techniques previously described (Corretti et al., 2002; Phillips et al., 2011). In premenopausal women, FMD may vary during the menstrual cycle (Hashimoto et al., 1995); therefore, subjects were recruited during the early follicular phase of menses. In the supine state, ultrasound imaging (MicroMaxx; Sonosite; Seattle, WA) of the brachial artery was performed in a longitudinal plane at a site 1–3 cm proximal to the antecubital fossa of the arm (Phillips et al., 2011). Baseline images were recorded and a blood pressure cuff was placed on the forearm and inflated to 50 mmHg above SBP for 5 min. Arterial diameter was determined during peak hyperemia after release of the blood pressure cuff from the forearm. To assess dilation, 10 s of images were captured 1–3 min after cuff release. Flow velocity was recorded at baseline and just after cuff release where maximal velocity was observed. Brachial artery reactivity to FMD was assessed before and 30 min following a single bout of SWL. The original position of the ultrasound probe was marked and measured according to the distance from the antecubital crease and the post-exercise examination was performed in the same position. Ten minutes after recording the last brachial artery diameter measurement following exercise, EID was determined with 0.4 mg of sublingual nitroglycerin (NTG) for 5 min. Since there is a blood pressure lowering effect of NTG making it unsafe to administer immediately prior to maximal exercise, we analyzed NTG dilations in a separate group of age and BMI matched adults who did not perform exercise. All ultrasonographic images were recorded and transferred to a PC for offline analysis of FMD and NTG responses using edge-detection software (Medical Imaging; Iowa City, IA). The coefficient of variation for FMD% was 6.3 and 3.2% for NTG-induced dilation. To estimate brachial artery wall shear stress, peak shear rates (SR) were calculated during FMD using the following equation: peak SR = maximal flow velocity (mm/s) ÷ diameter (mm). FMD was normalized for the peak SR using the following equation: normalized FMD = FMD ÷ peak SR (Phillips et al., 2011).

#### **EXERCISE PROTOCOL**

The exercise protocol involved performance of a single bout of bilateral lower-body weight lifting using a variable-resistance leg press machine (Hoist HD-1610 Selectorized Leg Press; Hoist Fitness Systems; San Diego, CA). After becoming familiarized with the leg press machine, participants performed a warm up period of 1–2 sets of 10 repetitions at a perceived capacity of approximately 30–40% of 1-repetition maximum (RM; calculated with a prediction equation) (Kemmler et al., 2006). Then 3–4 sets of 10 repetitions at a perceived capacity of approximately 80– 90% of 1-RM performed with a final isometric hold. A 2-min rest interval was allotted between each set. Exercise heart rate and the 10-point Borg rating of perceived exertion (RPE) scale were used as indices of intensity after each set. However, all volunteers completed a minimum of 4 sets of 10 repetitions or to failure on the final set. The RPE scale was not used as an intensity threshold since (1) the perceived capacity was 80–90% 1RM and (2) the goal was to elicit similar increases in BP between participants (Jurva et al., 2006).

#### **STATISTICS**

Results are expressed as mean ± standard deviation (SD), unless otherwise stated. Normality of the distribution was confirmed for all data using the Shapiro-Wilk test. Differences in physiological and physical characteristics (lipid panels, glucose measurements, vital signs, and anthropometric measurements) as well as dietary characteristics between OB and LN subjects were compared using student *t*-test. Pre-post differences in hemodynamic and vascular characteristics were compared using paired *t*-test within OB and LN groups. The mixed effects model was performed to examine the effects of acute resistance exercise between OB and LN subjects on hemodynamic and vascular variables including brachial artery diameter, maximum percentage change in diameter (FMD), and normalized FMD. For each outcome variable, group (OB vs. LN), time (pre vs. post) and group∗time interaction were modeled as fixed effects and subject intercept was modeled as random effect. Using only baseline data, we also conducted linear regression analyses to investigate the association of obesity with brachial artery FMD, independent of potential confounding effects. Pearson's correlations were used to evaluate how brachial artery reactivity to FMD and NTG responses relate to physiological and physical characteristics. Throughout the paper the effect size was computed using Cohen (1988). Cohen's *d* is an appropriate effect size for the comparison between two means. It indicates the standardized difference between two means, and expresses this difference in SD units ( mean/SD). Due to its pilot nature, we did not perform a formal power analysis for the current study. With sample size of 8 and 9 in each group, the study will have approximately 80% power to detect a large effect size of 1.5 standard unit mean difference between OB and LN using a two-sided alpha of 0.05. The results of our data could provide a more precise power estimate for future studies. Throughout the paper, the level of statistical significance for all analyses was set at two-sided *P <* 0*.*05. Data were analyzed using SPSS software (Version 19.0; IBM Corp., Armonk, NY).

#### **RESULTS**

#### **BASELINE AND VASCULAR CHARACTERISTICS**

Baseline characteristics for the OB (*n* = 9) and LN (*n* = 8) women who underwent a single bout of SWL are presented in **Table 1**. As was expected, body weight was significantly different between sedentary OB and LN women. In addition, all body weight-related characteristics including BMI, body fat percentage, and waist circumference were significantly different between OB and LN women. Total cholesterol and LDL cholesterol were also higher in OB compared to LN women. **Table 2** shows hemodynamic and vascular characteristics before and after



*All values expressed as mean (*<sup>±</sup> *SD). \*P <sup>&</sup>lt; 0.05, Statistically significant. BMI, body mass index; LDL, low density lipoprotein; HDL, high density lipoprotein; SBP, systolic blood pressure; DBP, diastolic blood pressure.*



*All values expressed as mean (*±*SD). \*P <sup>&</sup>lt; 0.05, Statistically significant compared to Before SWL (P <sup>&</sup>lt; 0.05). BP, blood pressure; FMD, flow-mediated dilation; SR, shear rate; BFV, blood flow velocity; FMD -, absolute change in FMD; NTG, nitroglycerin; ND, not determined.*

Peak BFV, cm/s 136*.*7 ± 40 140*.*7 ± 40 0.10 0.779 135*.*1 ± 37*.*8 143*.*4 ± 43 0.21 0*.*415 Baseline SR, s−<sup>1</sup> <sup>267</sup>*.*<sup>2</sup> <sup>±</sup> 47 287*.*<sup>8</sup> <sup>±</sup> 112 0.24 0.676 193*.*<sup>9</sup> <sup>±</sup> 63 218*.*<sup>8</sup> <sup>±</sup> 91 0.32 0*.*<sup>561</sup> Peak SR, s−<sup>1</sup> <sup>385</sup>*.*<sup>6</sup> <sup>±</sup> 130 421*.*<sup>1</sup> <sup>±</sup> 81 0.33 0.468 375*.*<sup>0</sup> <sup>±</sup> 123 404*.*<sup>9</sup> <sup>±</sup> 189 0.19 0*.*<sup>365</sup> Normalized FMD 0*.*018 ± 0*.*011 0*.*011 ± 0*.*01 0.67 0.102 0*.*032 ± 0*.*009 0*.*024 ± 0*.*009\* 0.89 0*.*008

SWL. Although the participants were not hypertensive (SBP *<* 140 mmHg), between-group comparisons showed significantly increased SBP and DBP levels in the OB compared to LN women. In addition, FMD and normalized FMD was reduced after compared to before SWL in OB but not in LN women (**Table 2**). None of the interactions between group and time for hemodynamic and vascular variables listed in **Table 2** were found significant in the mixed effects ANOVA models.

#### **EFFECTS OF ACUTE EXERCISE ON BRACHIAL FLOW AND NTG DILATIONS**

The RPE during the last repetition was 9.4 ± 0.7 in obese and 9.5 ± 0.8 in lean women. Brachial artery FMD was significantly reduced after SWL in OB group but not in LN group (**Figure 2**). In the mixed models, however, the Group∗SWL interaction was not significant, indicating that the slop of change in brachial artery FMD was not difference between LN and OB groups after SWL. There was a significant SWL effect (*b* = −0*.*024, *p* = 0*.*038) suggesting that the overall brachial artery FMD was significantly reduced after the SWL exercise (**Table 3**). NTG-dilation was lower in OB (21.6 ± 1.3%) compared to LN women (27.6 ± 2.1%, *p* = 0*.*02) (**Figure 3A**) and this response was negatively correlated with body weight (*r* = −0*.*70, *p* = 0*.*01) (**Figure 3B**), BMI (*r* = 0*.*62, *p* = 0*.*02), and WC (*r* = 0*.*67, *p* = 0*.*03) but not completely with %BF (*r* = 0*.*41, *p* = 0*.*10). Furthermore, correlations were independent of other physiological and physical characteristics. In separate studies (*n* = 12) obese women recruited to perform NTG dilations before exercise had similar NTG dilation (21.9 ± 1.9%) compared to after SWL.

#### **DISCUSSION**

The major findings of this study are that: (1) a single bout of strenuous physical stress (induced by weight lifting) reduces EDV in obese women who are sedentary, and (2) nitroglycerininduced endothelium-independent dilation are reduced in sedentary obese compared to sedentary lean women measured after a single bout of SWL.

**FIGURE 2 | The effect of a single bout of strenuous weight lifting (SWL) on brachial artery flow-mediated dilation (FMD) in lean (LN) and obese (OB) sedentary women.** ∗Significant difference observed after SWL vs. before SWL (*p <* 0*.*05).

**Table 3 | Mixed effects model on brachial artery FMD (%).**


*\*Mixed effects model was performed using maximum likelihood method with unstructured covariance.*

#### **VASCULAR REACTIVITY AND ACUTE EXERTION IN OBESITY**

Under normal physiologic conditions, NO is the chief vasodilator released by the endothelium and functions in modulating smooth muscle tone and inflammation (Vanhoutte et al., 2009). In addition, NO has anti-proliferative effects on the vascular wall and plays a critical role in vasodilation under conditions of increased blood flow (Landmesser et al., 2004). During increased flow, ROS are also produced by endothelial nitric oxide synthase (eNOS) but are usually scavenged by vascular antioxidant enzymes and, subsequently, NO diffuses into adjacent smooth muscle cells and mediates vasodilation (Wolin, 2000).

Repetitive increases in blood flow during chronic exercise training have been shown to improve NO-mediated vasodilation as assessed by FMD (Walsh et al., 2003; Olson et al., 2006) and may protect against vascular dysfunction through regulation of eNOS and reduction of oxidative stress (Hambrecht et al., 2003). However, acute exercise induces oxidative stress in individuals who are unaccustomed to exercise training by enhancing vascular ROS production, which contributes to a decrease in NO bioavailability and impairs FMD. Similar to previous studies (Phillips et al., 2011), brachial artery FMD after a single bout of weight lifting was reduced in sedentary adults (**Table 2**). Since a sedentary lifestyle is closely linked to obesity and both are associated with vascular dysfunction and increased risk of CVD, we hypothesized the degree of FMD impairment to be augmented in OB compared to LN adults. Although FMD was reduced in sedentary OB adults after SWL, the absolute difference (pre- vs. post-SWL) was similar between groups. Other studies have found impaired FMD responses after acute aerobic exercise in sedentary overweight and obese men (Harris et al., 2008). Other studies found reduced FMD following hand grip exercise in individuals at risk of CVD (McGowan et al., 2006). To our knowledge, this is the first study to determine the effects of acute resistance exercise on FMD responses in obese women.

Previous studies showed that vascular smooth muscle responses to NTG are impaired in the presence of obesity (Ayer et al., 2011) and other CVD risk factors (Raitakari et al., 2004; Adams et al., 2005); however, other studies examining NTGdilations in obesity independent of other risk factors suggest that it is preserved. Of considerable interest were the findings in this study that brachial artery EID responses to NTG were lower in OB compared to LN women (**Figure 3**). This result may represent a mechanistically different effect of exercise on FMD whereby vascular sensitivity to NO may contribute to reduce FMD in obesity compared to lean adults. Moreover, NTGinduced dilation after acute resistance exercise was associated with body weight, BMI, and WC with a tendency to correlate with %BF but not the other metabolic risk factors assessed (i.e., total cholesterol and LDL cholesterol). Although SBP and DBP were higher in OB women, there was no relationship between baseline values and NTG-induced vasodilation. Taken together, these data suggest an independent link between obesity and reduced vascular smooth muscle function after a single bout of SWL.

Both inflammation and oxidative stress could influence EID by reducing bioconversion of NTG to or scavenging of NO in smooth muscle (Schulz et al., 2002). In fact, obesity is associated with a pro-inflammatory milieu characterized by increased production of pro-atherogenic adipokines that enhance production of proinflammatory cytokines and increase ROS generation promoting oxidative stress (Tilg and Moschen, 2006). Acute resistance and aerobic exercise has been shown to increase plasma markers of oxidative stress in obesity (Vincent et al., 2004). Future studies will focus on the specific vascular effects of acute resistance exercise and other physical stressors on smooth muscle sensitivity to NO during obesity.

#### **STUDY LIMITATIONS**

There are some limitations of this study. First, the generalizability of the study is a limited to flow mediated dilation responses in women up to an hour following acute exertion. However, given that the burden of CVD is increasing in women, the results of this study contribute to a better understanding of the mechanisms by which exercise influences vascular health in women. Sex specific influences of physical stress on endothelium-dependent and endothelium-independent vasodilator responses may be important considerations in the future. Second, we were unable to evaluate the effects of NTG in sedentary OB and LN adults before a single bout of SWL due to the residual vasodilator effects of NTG on blood pressure during exercise. However, we

**FIGURE 3 | (A)** Nitroglycerin (NTG)-induced dilation of the brachial artery in lean (LN) and obese (OB) sedentary women after a single bout of strenuous weightlifting (SWL). \*Significant difference observed in OB vs. LN (*p <* 0*.*05). **(B)** Correlation between nitroglycerin (NTG)-induced

dilation of the brachial artery and body weight in sedentary women after a single bout of strenuous weight lifting (SWL). There was a negative correlation between brachial artery NTG-induced dilation and body weight (*r* = −0*.*70, *p <* 0*.*01).

have found that NTG dilations in a similar cohort of obese subjects were similar before and after exercise (Mean: 21% ± 1.4) suggesting that lower NTG dilation in obese women is not mediated by weight lifting. Third, CVD risk factors may have influenced NTG-dilations after SWL in OB adults since baseline blood pressure, total cholesterol, and LDL levels were higher compared to LN adults. However, in our analysis there were no relationships between these CVD risk factors and EID after acute resistance exercise. Finally, our results may have been confounded by nutritional variation and that the % body fat in the lean group was 27% a level that may be associated with metabolic dysfunction. There was no relationship between % body fat and FMD.

#### **CLINICAL INSIGHTS AND CONCLUSIONS**

In conclusion, our results suggest that acute physical stress induced by resistance exercise reduces FMD in obese women who are sedentary. Atherosclerotic CVD is a leading cause of preventable death among women between the ages of 25 and 44 and a large majority of women with CVD are asymptomatic (Maas et al., 2011) which makes early detection difficult. Since FMD correlates well with coronary artery function (Hashimoto et al., 2000) the results of our studies may extend to predict CVD risk among young women since acute exertion often triggers myocardial infarction in individuals who are sedentary (Mittleman et al., 1993).

In addition, endothelium-independent dilation to nitroglycerin was reduced in sedentary obese compared to lean women suggesting changes in vasodilator sensitivity of blood vessels to NO in obese adults. Reduced arterial function (FMD) and reduced sensitivity to NO donors (NTG) after acute physical exertion may be an important composite marker for future CV risk in young obese women without baseline arterial dysfunction or overt disease. Furthermore, these findings may be important in understanding the link between obesity and vascular dysfunction and may have important implications for how vascular and hemodynamic responses to other acute physiological perturbations such as acute hypertension, systemic hypoxemia, and hyperglycemia are altered during obesity.

#### **ACKNOWLEDGMENTS**

We would like to thank Weihua Gao and Hajwa Kim (UIC) for their statistical assistance. This work was supported by the National Heart, Lung, and Blood Institute grants K23HL85614, RO1HL095701, and HL095701-01A2S to Shane A. Phillips. This project was also supported by the University of Illinois at Chicago, Center for Clinical and Translational Science (CCTS), Award Number UL1RR029879 from the National Center for Research Resources. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

#### **REFERENCES**

Adams, V., Linke, A., Krankel, N., Erbs, S., Gielen, S., Mobius-winkler, S., et al. (2005). Impact of regular physical activity on the NAD(P)H oxidase and angiotensin receptor system in patients with coronary artery disease. *Circulation* 111, 555–562. doi: 10.1161/01.CIR.0000154560.88933.7E


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Received: 15 February 2014; accepted: 17 June 2014; published online: 07 July 2014. Citation: Franklin NC, Ali M, Goslawski M, Wang E and Phillips SA (2014) Reduced vasodilator function following acute resistance exercise in obese women. Front. Physiol. 5:253. doi: 10.3389/fphys.2014.00253*

*This article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology.*

*Copyright © 2014 Franklin, Ali, Goslawski, Wang and Phillips. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*

# Central gene expression changes associated with enhanced neuroendocrine and autonomic response habituation to repeated noise stress after voluntary wheel running in rats

# *Sarah K. Sasse , Tara J. Nyhuis , Cher V. Masini , Heidi E. W. Day and Serge Campeau\**

*Department of Psychology and Neuroscience, University of Colorado, Boulder, CO, USA*

#### *Edited by:*

*Anthony Leicht, James Cook University, Australia*

#### *Reviewed by:*

*Anastassios Philippou, National and Kapodistrian University of Athens, Greece Sladjana Dronjak, University of Belgrade, Serbia*

#### *\*Correspondence:*

*Serge Campeau, Department of Psychology and Neuroscience, University of Colorado, Muenzinger D244, 345 UCB, Boulder, CO 80309, USA e-mail: serge.campeau@*

*colorado.edu*

Accumulating evidence indicates that regular physical exercise benefits health in part by counteracting some of the negative physiological impacts of stress. While some studies identified reductions in some measures of acute stress responses with prior exercise, limited data were available concerning effects on cardiovascular function, and reported effects on hypothalamic-pituitary-adrenocortical (HPA) axis responses were largely inconsistent. Given that exposure to repeated or prolonged stress is strongly implicated in the precipitation and exacerbation of illness, we proposed the novel hypothesis that physical exercise might facilitate *adaptation* to repeated stress, and subsequently demonstrated significant enhancement of both HPA axis (glucocorticoid) and cardiovascular (tachycardia) response habituation to repeated noise stress in rats with long-term access to running wheels compared to sedentary controls. Stress habituation has been attributed to modifications of brain circuits, but the specific sites of adaptation and the molecular changes driving its expression remain unclear. Here, *in situ* hybridization histochemistry was used to examine regulation of select stress-associated signaling systems in brain regions representing likely candidates to underlie exercise-enhanced stress habituation. Analyzed brains were collected from active (6 weeks of wheel running) and sedentary rats following control, acute, or repeated noise exposures that induced a significantly faster rate of glucocorticoid response habituation in active animals but preserved acute noise responsiveness. Nearly identical experimental manipulations also induce a faster rate of cardiovascular response habituation in exercised, repeatedly stressed rats. The observed regulation of the corticotropin-releasing factor and brain-derived neurotrophic factor systems across several brain regions suggests widespread effects of voluntary exercise on central functions and related adaptations to stress across multiple response modalities.

**Keywords: audiogenic stress, exercise, habituation, glucocorticoid, cardiovascular,** *in situ* **hybridization**

# **INTRODUCTION**

Stress is a significant risk factor for numerous physical (Brindley and Rolland, 1989; Khansari et al., 1990; Forsen, 1991; Pasternac and Talajic, 1991; Vanitallie, 2002; Kalantaridou et al., 2004) and psychological (Dunner et al., 1979; Brown et al., 1987; Hammen et al., 1992; Arborelius et al., 1999; Vanitallie, 2002; Swaab et al., 2005) disorders. Remarkably, many of the same disorders precipitated or exacerbated by stress can be prevented or improved by regular physical exercise (Manson et al., 1992; Paffenbarger et al., 1992; Chodzko-Zajko and Moore, 1994; Bérard et al., 1997; Wannamethee et al., 1998; Dunn et al., 2001; Goodwin, 2003), suggesting that routine physical activity may benefit health in part through a stress-mitigating effect (Roth and Holmes, 1985; Brown and Lawton, 1986; Moraska and Fleshner, 2001; Fleshner, 2005). Indeed, exercise can reduce some of the physiological consequences of acute stress under certain experimental conditions (Brown and Siegel, 1988; Dishman et al., 1995, 1997; Moraska and Fleshner, 2001; Greenwood et al., 2003a,b, 2007a,b; Adlard and Cotman, 2004; Campeau et al., 2010; Masini et al., 2011). However, results are inconsistent across stress stimuli when hypothalamic-pituitary-adrenocortical (HPA) axis-mediated glucocorticoid responses are assessed (Dishman et al., 1995, 1997; Fleshner, 2000; Campisi and Fleshner, 2003; Droste et al., 2003, 2006, 2007; Fediuc et al., 2006; Sasse et al., 2008; Nyhuis et al., 2010), and only limited data are available with regard to cardiovascular reactivity (Morimoto et al., 2000; Masini et al., 2011). Collectively, these findings argue against generalized reductions in acute stress sensitivity by prior exercise, and rather point to complex regulation involving selective modulation of responses to some, but not all, acute stress stimuli. How such a high degree of regulatory specificity is conferred by regular exercise remains to be understood.

While many stress-responsive systems regulate vital physiological functions under both normal and acute stress conditions, it is their sustained, excessive or dysregulated activation by repeated stress that is most strongly associated with pathogenesis (Chrousos and Gold, 1992; Tsigos and Chrousos, 1994; Charmandari et al., 2005). Repeated exposure to the same stressor is often accompanied by a progressive decrease in response amplitude, or habituation, which is considered an adaptation that lessens the physiological toll of repeatedly activating stressresponsive systems (Armario et al., 1984, 1986; Kant et al., 1985; Melia et al., 1994; Campeau et al., 2002). Dysfunctional response habituation could contribute to increased susceptibility to stress-related pathology. Prior work in our laboratory indicated that habituation of glucocorticoid (Sasse et al., 2008; Nyhuis et al., 2010) and tachycardia (Masini et al., 2011) responses to repeated audiogenic stress is facilitated after chronic voluntary wheel running in rats, lending support to the hypothesis that the health benefits of regular exercise include enhanced adaptive mechanisms that reduce the cumulative impact of repeated or prolonged stress. Habituation of stress responses involves active plastic processes mediated by changes within the central nervous system (Armario et al., 1988; Melia et al., 1994; Akana and Dallman, 1997; Ezzeddine and Glanzman, 2003; Esdin et al., 2010). Enhanced glucocorticoid and cardiovascular response habituation to repeated stress in exercised animals is thus likely to result from regulation at a central level by physical activity. To explore this possibility, *in situ* hybridization histochemistry was performed on brains collected from animals that displayed exercise-induced facilitation of glucocorticoid response habituation to repeated audiogenic stress (Sasse et al., 2008; Experiment 2). Similar experimental manipulations also produce enhanced cardiovascular response habituation in exercised, repeatedly stressed rats (Masini et al., 2011). The primary goal was to define neuroanatomical regions and associated neurochemical substrates potentially underlying facilitation of response habituation to repeated audiogenic stress by prior exercise. Although correlational, any reliable changes in central gene expression could subsequently be tested for their specific relevance to the regulation of stress adaptation by physical activity.

First, messenger RNA (mRNA) expression of the immediateearly gene FBJ osteosarcoma oncogene (*Fos*, also known as c-*fos*) was characterized in select regions based on their prior association with audiogenic stress activation. These included the central control station of HPA axis activation, the paraventricular nucleus of the hypothalamus (PVN), in addition to various cortical (cingulate [CG], infralimbic [IL], orbitofrontal/claustrum [OFC/CL], and prelimbic [PL] regions of the medial prefrontal cortex [mPFC]) and forebrain (septohypothalamic nucleus [SHy] and closely associated ventrolateral septum [LSv], anterior bed nucleus of the stria terminalis [BNST]) areas (Campeau and Watson, 1997; Campeau et al., 2002; Burow et al., 2005). Regions differentially activated by repeated audiogenic stress in exercised vs. sedentary brains could reflect important sites of exerciseinduced regulation associated with stress response habituation. Within many of these regions, additional molecules reported to mediate glucocorticoid or cardiovascular responses to stress and/or to be regulated by stress or regular exercise were also examined for putative transcriptional regulation. These included transcripts of corticotropin-releasing hormone (*Crh*) and arginine vasopression (*Avp*) at the level of the PVN (Timofeeva et al., 2003; Kawashima et al., 2004; Park et al., 2005), *Crh* in stressrelated extra-hypothalamic regions (e.g., BNST, central [CeA] and basolateral [BLA] amygdaloid nuclei—Antoni, 1993; Gu et al., 2003; Dong and Swanson, 2006; Hauger et al., 2006; Radley et al., 2009; Ulrich-Lai and Herman, 2009), *Crh* receptor subtypes *Crhr1* and *Crhr2* (LSv, PVN, BLA and medial amgydaloid nucleus [MeA], and ventromedial hypothalamic nucleus [VMH]—Luo et al., 1994; Chalmers et al., 1995; Makino et al., 1995; Imaki et al., 1996; Van Pett et al., 2000), brain-derived neurotrophic factor (*Bdnf* ; hippocampus, PVN—Smith et al., 1995a,b; Neeper et al., 1996; Oliff et al., 1998; Nibuya et al., 1999; Adlard and Cotman, 2004; Farmer et al., 2004; Gomez-Pinella et al., 2008; Nyhuis et al., 2010), and tyrosine receptor kinase B subtype (*Trkb*), the main receptor for *Bdnf*, in the hippocampus (Nibuya et al., 1999). Our group previously examined neurotrophin regulation by prior exercise and repeated stress (Nyhuis et al., 2010), but this analysis did not include comparisons in non-stressed and acutely stressed animals, which were available in the current study.

#### **MATERIALS AND METHODS**

#### **SUBJECTS AND EXPERIMENTAL DESIGN**

Readers are referred to Sasse et al. (2008) (Experiment 2) for explicit details regarding manipulation of animal subjects and experimental design. All animal procedures were reviewed and approved by the Institutional Animal Care and Use Committee of the University of Colorado and conformed to the National Research Council's *Guide for the Care and Use of Laboratory Animals* (8th Edn., 2011). All efforts were made to minimize animal suffering and the number of animals used. Briefly, following an acclimation period to our colony facility, the 48 young adult (∼2 months of age upon arrival), male Sprague–Dawley rats (Harlan, Indianapolis, IN) were divided into six groups matched for body weight (*n* = 8/group) and individually housed. Three groups were assigned to the Exercise (Ex) condition and given 24 h unlimited access to a stainless steel running wheel (Nalge Nunc International) attached to the wire lid of their home cages for the 6 weeks prior to and the 11 days of the experimental testing phase. During the same time period, the remaining 3 groups were individually housed in similar cages in the same colony room, but under Sedentary (Sed) conditions, without running wheels in their home cages.

**Figure 1** provides a graphic summary of the experimental design. On each of the 3 days prior to the onset of testing, all Ex and Sed animals were pre-exposed to the testing conditions, which involved transporting rats to a separate testing room, placing them within their home cages into assigned acoustic chambers, and exposing them only to the background noise (∼60 decibel-A scale [dB]) generated by the ventilation fans for 30 min. Importantly, Ex rats would be denied running wheel access during all 30 min stress (or no stress) exposures, so they were habituated to this additional experimental manipulation during these preexposures by placing a long metal rod through both ends of each wheel that completely blocked rotation. Then, on each of 11 consecutive days, rats were transported to the testing room, placed within their home cages into their assigned acoustic chambers, and exposed to one of the following *Stress Treatment* conditions for 30 min: 98 dB noise stress (Ex- or Sed-Repeated Noise, *n* = 8/group), 60 dB background noise (Ex- or Sed-No Noise, *n* = 8/group), or 60 dB background noise exposures on the first 10 days followed by a single 98 dB noise stress presentation on Day

11 (Ex- or Sed-Acute Noise, *n* = 8/group). Running wheels of all Ex rats were locked during each noise (or no noise) exposure, after which wheels were immediately unlocked and all animals were returned to the colony. Immediately following the final 30 min 98 dB or background noise exposure on Day 11, all rats were killed by decapitation (without anesthesia), and brains were removed immediately and frozen in isopentane chilled to −30 to −40◦C prior to storage at −80◦C until further processing.

# *In situ* **HYBRIDIZATION HISTOCHEMISTRY**

Ten micron coronal sections of collected brains were cut on a cryostat (Leica model 1850, Wetzlar, Germany), thaw-mounted onto poly-L-lysine-coated slides, and stored at −80◦C until processed. Slides were first fixed in a phosphate-buffered 4% paraformaldehyde solution for 1 h, and then rinsed 3 times in 2X sodium saline citrate (SSC), acetylated for 10 min in 0.1 M triethanolamine containing 0.25% acetic anhydride, rinsed in distilled water, and dehydrated in graded ethyl alcohol concentrations. 35S-labeled cRNA riboprobes were generated for *Fos*, *Crh*, *Crhr1*, *Crhr2*, *Avp*, *Bdnf*, and *Trkb* from cDNA subclones in transcription vectors using standard *in vitro* transcription methodology. The rat *Fos* (courtesy of Dr. T. Curran, St. Jude Children's Research Hospital, Memphis, TN) and *Crh* (courtesy of Dr. R. T. Thompson, University of Michigan) cDNA clones were subcloned in pGem 3Z and cut with HindIII to yield 680 nucleotide (nt) and 770 nt cDNA templates, respectively. The rat *Crhr1* cDNA clone (Dr. J. P. Herman, University of Cincinnati) was subcloned in pCR-Blunt II-Topo and cut with BamHI to yield a 345 nt cDNA template. The rat *Crhr2* (courtesy of Dr. R. T. Thompson, University of Michigan) and *Trkb* (courtesy of Dr. D. McKinnon, SUNY, Stony Brook) cDNA clones were subcloned in pBluescript SK and cut with HindIII to yield 899 nt and 306 nt cDNA templates, respectively. The rat *Avp* cDNA clone (courtesy of Dr. T. G. Sherman, Georgetown University Medical Center) was subcloned into pGem 3 and cut with EcoRI to yield a 235 nt cDNA template. The rat *Bdnf* cDNA clone (courtesy of Dr. J. P. Herman, University of Cincinnati) was subcloned in pBluescript and cut with PvuII to yield a 759 nt cDNA template.

Copy riboprobes were labeled in a reaction mixture consisting of 1µg of the appropriate linearized plasmid, 1X T3, T7, or SP6 transcription buffer (Promega), 125µCi 35S-UTP, 150µM NTP's (ATP, CTP, and GTP), 12.5 mM dithiothreitol, 20 U RNase inhibitor, and 6 U of T3 (for *Crhr2*, *Bdnf*, and *Trkb*), T7 (for *Fos*, *Crh*, and *Crhr1*), or SP6 (for *Avp*) RNA polymerase. The reaction was allowed to proceed for 2 h at 37◦C, after which probe was separated from free nucleotides over a Sephadex G50–50 column. Riboprobes were diluted in hybridization buffer consisting of 50% formamide, 10% dextran sulfate, 2X SSC, 50 mM sodium phosphate buffer (pH 7.4), 1X Denhardt's solution, and 0.1 mg/ml yeast tRNA, to yield <sup>∼</sup>1–2 <sup>×</sup> <sup>10</sup><sup>6</sup> dpm/70µl buffer. Diluted probe (70µl) was applied to tissue sections on each slide, which were then coverslipped, placed in sealed plastic boxes lined with filter paper moistened with 60% formamide in distilled water, and incubated overnight (12–18 h) at 55◦C. Coverslips were then removed, and slides were rinsed 3 times in 2X SSC before being incubated in a solution containing RNase A (2.0µg/ml) for 1 h at 37◦C. Slides were next washed successively in 2X, 1X, 0.5X, and 0.1X SSC for 5 min each, and then incubated in 0.1X SSC for 1 h at 65◦C. They were subsequently returned to room temperature in 2 rinses of 0.1X SSC for 5–10 min each prior to being dehydrated in graded ethyl alcohols and air-dried. Slides were then exposed to Kodak MR X-ray film for optimized exposure times.

Control experiments were performed on tissue sections pretreated with RNase A (2.0µg/ml at 37◦C for 1 h) prior to hybridization; this treatment prevented labeling. Additionally, some control sections were hybridized with the sense cRNA strands, which in all cases did not lead to significant hybridization to tissue sections (data not shown). For each riboprobe, two to three slides (4–6 sections/slide) representing a given brain region from each rat included in the study were processed simultaneously to allow for direct comparisons in the same regions. In more cases than not, multiple *in situ* hybridizations were performed for the same target mRNA at different levels of the brain with all animals represented, which reduced the effects of technical variation within regions, but limited the ability to make direct comparisons between different regions.

#### **IMAGE ANALYSIS**

Semi-quantitative densitometric analyses were performed on digitized images from X-ray films in the linear range of the gray values obtained with our acquisition system (Northern Light lightbox model B95 [Imaging Res. Inc., St. Catharines, Ontario], a SONY TV camera model XC-ST70 fitted with a Navitar 7000 zoom lens [Rochester, NY], connected to an LG3-01 frame grabber [Scion Corp., Frederick, MD] inside a Dell Dimension 500, captured with Scion Image beta rel. 4.02). Signal pixels in a region of interest were defined as being 3.5 standard deviations above the mean gray value of the representative background, which was set by a cell- and/or signal- poor area close to the region of interest. Mean integrated gray values were computed from the product of the number of pixels comprising the positive signal and their average gray level within the region of interest.

The panels in the left column of **Figures 2A,E,I,M** provide schematic representations of the brain that were adapted from *The Rat Brain in Stereotaxic Coordinates* (CD-ROM; Paxinos and Watson, 2005) atlas, with permission from Elsevier. The right sides of these coronal sections are labeled for the regions of interest in which semi-quantitative analyses were performed, while the left sides of the same sections illustrate the templates that were consistently used to define these regions. For example, panel **A** represents a section at the level of the prefrontal cortex (PFC; 3.24 mm anterior to bregma, as designated in the lower right corner), in which *Fos* mRNA expression was analyzed in: the prelimbic (PL) PFC, designated by a 50 × 60 pixel rectangle template; IL PFC, designated by a 40 × 50 pixel ellipse; the region encompassing the orbitofrontal cortex and claustrum (OFC/CL), designated by a 110 × 60 pixel ellipse; and the cingulate (CG) PFC, for which a trapezoidal template was designed using the Paxinos and Watson atlas for guidance. For each riboprobe, 4–12 bilateral measurements were made within each relevant region of interest (as defined by the template boundaries illustrated in **Figure 2**) on appropriate sections from each animal. These values were averaged to obtain the mean integrated gray value per region for each rat, which provided the relative mRNA expression level. This analytic method gives relative semi-quantitative results that are comparable to performing a quantitative grain analysis on photographic emulsion-dipped sections (Day et al., 2005). Occasionally, sections or slides for a specific animal were missing, damaged, or otherwise inappropriate for analysis following these multiple levels of processing and were therefore excluded. These instances will be explicitly stated in the appropriate Results section.

#### **STATISTICAL ANALYSES**

All statistical analyses were performed using the *Statistical Analysis Software (SAS)* program. Two-way (*Activity Status* × *Stress Treatment*) ANOVAs were used to analyze the various target mRNA expression levels in the brain regions of interest. These were followed by Scheffé's *post-hoc* multiple means comparisons to determine the source of reliable effects when present. Statistical significance for all analyses was set at *p* ≤ 0*.*05.

#### **RESULTS**

#### **REGIONAL Fos mRNA INDUCTION**

Expression levels and basic statistical comparisons of *Fos* mRNA measured in several brain regions of exercised and sedentary rats immediately following the last of 11 consecutive daily noise (Repeated Noise) or background noise (No Noise) exposures, or an acute noise stress presentation on Day 11 (Acute Noise), are presented in **Table 1**. Of the eight regions analyzed, six displayed a similar pattern of *Fos* mRNA induction, in which only rats from the acutely exposed groups demonstrated reliable induction that was greater than that of the repeatedly exposed and no noise control groups, which did not differ from each other (see panels **B–D** of **Figure 2**). These regions included the PVN [main effect of *Stress Treatment*: *F(*2*,* <sup>38</sup>*)* = 35*.*73, *p <* 0*.*0001], LSv [*Stress Treatment*: *F(*2*,* <sup>42</sup>*)* = 55*.*77, *p <* 0*.*0001], SHy [*Stress Treatment*: *F(*2*,* <sup>42</sup>*)* = 82*.*27, *p <* 0*.*0001], ventrolateral aspect of the anterior BNST [vBNST; *Stress Treatment*: *F(*2*,* <sup>42</sup>*)* = 23*.*64, *p <* 0*.*0001], and the PL [*Stress Treatment*: *F(*2*,* <sup>38</sup>*)* = 6*.*93, *p* = 0*.*0027] and IL [*Stress Treatment*: *F(*2*,* <sup>38</sup>*)* = 17*.*15, *p <* 0*.*0001] regions of the mPFC, as evidenced by Two-Way ANOVAs and confirmed and further characterized by Scheffé's *post-hoc* analyses (all *p*'s ≤ 0.05). Additional Two-Way ANOVAs showed that group differences in *Fos* mRNA expression in the OFC/CL region did not reach statistical significance, and while there was a significant *Activity Status* × *Stress Treatment* interaction effect on *Fos* mRNA expression in the CG [*F(*2*,* <sup>38</sup>*)* = 3*.*28, *p* = 0*.*0487], Scheffé's *post-hoc* analyses did not reveal any reliable differences among pair-wise comparisons. Of the four mPFC regions examined for *Fos* mRNA induction, the following groups had at least one animal excluded from the analysis: Sed-Repeated Noise (*n* = 7), Ex-No Noise (*n* = 7), and Ex-Repeated Noise (*n* = 6). The following groups had at least one animal excluded from the analysis in the PVN: Sed-Repeated Noise (*n* = 7) and Ex-Repeated Noise (*n* = 7).

#### **FIGURE 2 | Diagram specifying brain regions analyzed by** *in situ* **hybridization and corresponding representative autoradiograms depicting important groups differences.** Panels in the left column provide images adapted from the Paxinos and Watson (2005) atlas (with

permission to reprint from Elsevier) of coronal brain sections at the level of the **(A)** medial prefrontal cortex (mPFC), **(E)** ventrolateral septum (LSv) and anterior bed nucleus of the stria terminalis (BNST), **(I)** paraventricular nucleus of the hypothalamus (PVN), and **(M)** hippocampus and amygdala. All regions of interest analyzed in this study are labeled on the right sides of these sections while the left sides depict the templates used to consistently define their boundaries. Remaining panels in each row show representative autoradiogram images of specified mRNAs in brain sections corresponding to these atlas diagrams. Panels **(B–D)** *Fos* mRNA expression illustrating robust induction of *Fos* mRNA by acute noise stress and nearly complete habituation of *Fos* responses to levels similar

#### to No Noise controls following repeated noise exposures in some (prelimbic [PL] and infralimbic [IL]), but not all (orbitofrontal cortex/claustrum [OFC/CL] and cingulate [CG]) mPFC regions. Panels **(F–H)** *Crh* mRNA in the dorsal (d) and ventral (v) aspects of the BNST demonstrating repeated noise-induced reduction of *Crh* mRNA expression in sedentary but not exercised rats, which exhibited levels similar to No Noise controls. Panels **(J–L)** *Crhr1* mRNA expression in the PVN illustrating the trend for mRNA induction that was specific to exercised, repeatedly stressed rats. Panels **(N–P)** *Bdnf* mRNA expression showing the trend for voluntary exercise to prevent the repeated stress-induced reduction in hippocampal *Bdnf* mRNA expression observed in sedentary rats. Other abbreviations: BLA, basolateral nucleus of amygdala; CeA, central nucleus of amygdala; dDG, dorsal dentate gyrus; MeA, medial nucleus of amygdala; SHy, septohypothalamic nucleus; vDG, ventral dentate gyrus; VMH, ventromedial nucleus of hypothalamus.

#### **HYPOTHALAMIC Crh AND Avp mRNA EXPRESSION**

*Crh* mRNA expression levels in the PVN are presented in **Table 2**. A Two-Way ANOVA revealed significant *Activity Status* [*F(*1*,* <sup>42</sup>*)* = 6*.*07, *p* = 0*.*0180], *Stress Treatment* [*F(*2*,* <sup>42</sup>*)* = 7*.*28, *p* = 0*.*0019], and *Activity Status* × *Stress Treatment* [*F(*2*,* <sup>42</sup>*)* = 5*.*17, *p* = 0*.*0098] effects on *Crh* mRNA expression in this key locus of stress response integration. Specifically, when controlling for *Stress Treatment*, Ex rats exhibited reliably lower PVN-*Crh* mRNA expression than Sed rats, and while holding *Activity Status* constant, animals in the Repeated Noise groups had significantly greater mRNA expression than those exposed to Acute Noise, although neither of these two groups differed reliably from the No Noise controls (Scheffé, *p*'s ≤ 0.05). The significant interaction effect reflects that, while acute and repeated noise stress exposure had little effect on *Crh* mRNA expression in the PVN of Ex rats, the Sed rats acutely exposed to

**Table 2 | Effects of stress and wheel running on central CRH signaling.**

#### **Table 1 | Stress and voluntary exercise effects on central** *Fos* **mRNA expression.**


*Mean relative Fos mRNA expression in indicated brain regions of exercised (Ex) and sedentary (Sed) rats exposed to no, acute, or repeated noise stress. Group means are mean integrated gray values/100 (*±*1 s.e.m.), presented as arbitrary units. CG, cingulate cortex; IL, infralimbic cortex; OFC/CL, orbitofrontal cortex/claustrum; PL, prelimbic cortex; vBNST, ventrolateral bed nucleus of stria terminalis; SHy, septohypothalamic nucleus; LSv,ventrolateral septum; PVN, paraventricular nucleus of hypothalamus.*

*\*Omnibus ANOVA, p* <sup>≤</sup> *0.05.*

*†Two-Way ANOVA: significant main effect of Stress Treatment (p* <sup>≤</sup> *0.05); <sup>a</sup>,bacross each row, Stress Treatment group means with same (or no) letter symbol were not statistically different (Scheffé, p* ≤ *0.05).*

noise showed a reliable reduction in mRNA expression as compared to the other two Sed groups, which did not differ (Scheffé, *p* ≤ 0*.*05). A graphic representation of *Crh* mRNA expression in the PVN is depicted in **Figure 3A** to help illustrate this interaction. *Avp* mRNA expression was abundant in both the paraventricular and supraoptic nuclei of the hypothalamus in all groups. However, the extent of this expression was not reliably affected by *Activity Status* or *Stress Treatment* in either structure, as evidenced by non-significant Two-Way ANOVAs (data not shown).

#### **EXTRA-HYPOTHALAMIC Crh mRNA EXPRESSION**

**Table 2** also depicts relative *Crh* mRNA expression levels in specified extra-hypothalamic regions. In the dorsal aspect of the anterior BNST (dBNST), a significant *Activity Status* × *Stress Treatment* interaction effect was found on *Crh* mRNA expression [*F(*2*,* <sup>42</sup>*)* = 8*.*42, *p* = 0*.*0008]. In general, while expression was decreased by the repeated noise stress in Sed rats, it was relatively unaffected, and even somewhat increased, in Ex animals exposed to the same repeated stress paradigm (see panels **F–H** of **Figure 2**). The only statistically significant group difference revealed by Sheffé's *post-hoc* analyses, however, was that *Crh* mRNA expression was reliably lower in the Sed-Repeated Noise group than all other groups (*p* ≤ 0*.*05), which did not differ amongst each other. *Crh* mRNA levels in the dBNST are depicted in **Figure 3B** to aid in the visualization of this interaction. A similar expression pattern was observed in the vBNST and


*Mean relative expression of Crh, Crhr1, and Crhr2 mRNAs in indicated brain regions of Ex and Sed rats following no, acute, or repeated audiogenic stress. Group means are mean integrated gray values/100 (*±*1 s.e.m.), presented as arbitrary units. dBNST, dorsal BNST; Hypothal., hypothalamus; CeA, central nucleus of amygdala; MeA and BLA, medial and basolateral nuclei of amygdala; VMH, ventromedial nucleus of hypothalamus.*

*\*Omnibus ANOVA, p* <sup>≤</sup> *0.05.*

*Two-Way ANOVA: ‡significant main effect of Activity Status, †significant main effect of Stress Treatment (p* <sup>≤</sup> *0.05); <sup>a</sup>,bacross each row, Stress Treatment group means with same (or no) letter symbol(s) were not statistically different (Scheffé, p* ≤ *0.05). ¶significant Activity Status x Stress Treatment interaction (p* ≤ *0.05); group differences are not indicated by letter symbols here but rather presented in the Results section and illustrated in Figure 3.*

the CeA, and while significant *Activity Status* × *Stress Treatment* interaction effects were uncovered for both regions by Two-Way ANOVAs [vBNST: *F(*2*,* <sup>42</sup>*)* = 9*.*27, *p* = 0*.*0005; CeA: *F(*2*,* <sup>41</sup>*)* = 6*.*62, *p* = 0*.*0032], none of the *post-hoc* pair-wise comparisons revealed reliable group differences (Scheffé, all *p*'s *>* 0.05). One animal in the Ex-Acute Noise group was excluded from the analysis of *Crh* mRNA expression in the CeA (*n* = 7).

#### **Crh RECEPTOR mRNAs**

Two-Way ANOVAs revealed no significant differences in *Crhr1* mRNA levels in the MeA or BLA. Similarly, no reliable effects of *Activity Status* or *Stress Treatment* were found with respect to *Crhr1* mRNA expression in the PVN when analyzing all six experimental groups (see **Table 2**). Interestingly, when comparing only the Ex groups using a One-Way ANOVA for *Stress Treatment*,

**FIGURE 3 | Stress-induced** *Crh* **regulation in the PVN and dBNST is different in exercised compared to sedentary rats. (A)** Relative *Crh* mRNA expression levels in the paraventricular nucleus of the hypothalamus (PVN). Significant main effects of *Activity Status* and *Stress Treatment* were observed, with reliably lower PVN-*Crh* mRNA expression levels in Ex as compared to Sed rats, and significantly greater expression levels in Repeated as compared to Acute Noise groups, although neither of these two groups differed reliably from the No Noise controls. The interaction effect was also significant, such that, while acute and repeated noise stress exposure had little effect on *Crh* mRNA expression in the PVN of Ex rats, the Sed rats acutely exposed to noise showed a reliable reduction in mRNA expression as compared to the other two Sed groups, which did not differ. Group means are mean integrated gray values/100 (±1 s.e.m.), presented as arbitrary units. <sup>∗</sup>*p* ≤ 0*.*05 vs. Sed-No Noise and Sed-Repeated Noise. **(B)** Relative *Crh* mRNA expression levels in the dorsal aspect of the anterior bed nucleus of the stria terminalis (dBNST). A significant *Activity Status* × *Stress Treatment* interaction effect was found such that, while expression levels were decreased by the repeated noise stress in Sed rats, they were relatively unaffected, and even slightly increased, in Ex animals exposed to the same repeated stress paradigm. A similar pattern in CRH mRNA expression was also observed in the anteroventral portion of the BNST (vBNST), as well as the central nucleus of the amygdala (CeA; see **Table 2**). Group means are mean integrated gray values/100 (±1 S.E.M.), presented as arbitrary units. <sup>∗</sup>*p* ≤ 0*.*05 vs. all other experimental groups, which were not reliably different amongst each other.

a significant main effect was uncovered [*F(*2*,* <sup>21</sup>*)* = 4*.*67, *p* = 0*.*0210], such that *Crhr1* mRNA expression was significantly greater in the PVN of Ex-Repeated Noise as compared to Ex-Acute Noise animals (Scheffé, *p* ≤ 0*.*05), although neither of these groups differed reliably from the Ex-No Noise group (see panels **J–L** of **Figure 2**). A Two-Way ANOVA revealed a significant main effect of *Activity Status* [*F(*1*,* <sup>40</sup>*)* = 13*.*88, *p* = 0*.*0006] on *Crhr2* mRNA expression in the VMH, with reliably lower expression in brains of Ex compared to Sed animals. The following groups had at least one animal excluded from the analysis in the VMH:

**Table 3 | Stress and wheel running regulation of central** *Bdnf* **and** *Trkb* **mRNA expression.**


*Mean relative Bdnf and Trkb mRNA expression levels in indicated brain regions of Ex and Sed rats following no, acute, or repeated stress. Group means are mean integrated gray values/100 (*±*1 s.e.m.), presented as arbitrary units. dDG and vDG, dorsal and ventral blades of dentate gyrus, respectively. \*Omnibus ANOVA, p* <sup>≤</sup> *0.05.*

*Two-Way ANOVA: ‡significant main effect of Activity Status (p* <sup>≤</sup> *0.05).*

Sed-Acute Noise (*n* = 7) and Sed-Repeated Noise (*n* = 7). *Crhr2* mRNA levels were not reliably different between treatment groups in the LSv or MeA (**Table 2**).

#### **Bdnf AND Trkb mRNA EXPRESSION**

Expression levels and basic statistical comparisons of *Bdnf* and *Trkb* transcripts are presented in **Table 3**. In three of the four hippocampal sub-regions examined for *Bdnf* mRNA expression, Ex rats displayed significantly greater expression than Sed animals when controlling for *Stress Treatment*, as shown by Two-Way (*Activity Status* × *Stress Treatment)* ANOVAs. These included the CA3 region [main effect of *Activity Status*: *F(*1*,* <sup>42</sup>*)* = 17*.*22, *p* = 0*.*0002] and the dorsal [*Activity Status*: *F(*1*,* <sup>40</sup>*)* = 33*.*04, *p <* 0*.*0001] and ventral [*Activity Status*: *F(*1*,* <sup>42</sup>*)* = 27*.*70, *p <* 0*.*0001] blades of the dentate gyrus. Relatively low levels of *Bdnf* mRNA were expressed in the CA1 region that did not differ reliably between the experimental groups. Although not statistically significant, a general pattern in hippocampal *Bdnf* mRNA expression emerged such that, in Sed rats, acute and repeated noise had a tendency to decrease expression relative to non-stressed controls, while Ex animals appeared to be buffered from this stress-induced reduction in hippocampal *Bdnf* transcription (see panels **N–P** of **Figure 2**). **Figure 4** depicts *Bdnf* mRNA expression levels in the CA1 region, which serves to clearly illustrate this pattern of results. Expression of *Bdnf* mRNA was also examined in the PVN, in which a Two-Way ANOVA uncovered no reliable group differences. A significant main effect of *Activity Status* was revealed by Two-Way ANOVAs on *Trkb* mRNA expression in the CA1 [*F(*1*,* <sup>42</sup>*)* = 8*.*63, *p* = 0*.*0053] and CA3 [*F(*1*,* <sup>42</sup>*)* = 5*.*37, *p* = 0*.*0254] hippocampal sub-regions, such that mRNA levels

were significantly greater in Ex compared to Sed rats (Scheffé, *p*'s ≤ 0.05). No reliable group differences in *Trkb* mRNA expression were observed in either the dorsal or ventral blades of the dentate gyrus.

# **DISCUSSION**

The results of this study indicate that physical activity can significantly alter central nervous system expression patterns of several genes strongly associated with stress responsiveness. As summarized below, a number of studies have already identified various genes that are differentially regulated in the brains of physically active compared to sedentary rats under control (no stress) conditions. Importantly, some of this regulation can be further differentiated in exercised and sedentary brains in response to acute or repeated stress exposures. In view of our recent findings that prior physical activity enhances the rate of neuroendocrine and autonomic response habituation to repeated stress (Sasse et al., 2008; Nyhuis et al., 2010; Masini et al., 2011), we reasoned that these previously reported sites and molecular targets of regulation could also underlie this particular mode of stress adaptation. While the brains analyzed in the current study were obtained following measurements of HPA axis activation (Sasse et al., 2008), a nearly identical experimental protocol led to significantly facilitated habituation of heart rate and core body temperature responses in exercised rats (Masini et al., 2011). It is therefore highly likely that the central gene regulation observed in the current study could be associated with modulation of either or both the neuroendocrine and autonomic response adaptations observed in our prior studies.

#### **REGIONAL** *Fos* **mRNA INDUCTION AND HABITUATION**

As reported previously (Campeau et al., 1997, 2002; Campeau and Watson, 1997; Burow et al., 2005), acute loud noise exposure induced widespread *Fos* mRNA induction, but importantly, the amplitude of this expression did not differ between exercised and sedentary rats. This finding supports a number of prior studies reporting that acute HPA axis (Dishman et al., 1995, 1997, 1998; Fleshner, 2000; Fediuc et al., 2006; Sasse et al., 2008; Campeau et al., 2010; Nyhuis et al., 2010) and autonomic (Morimoto et al., 2000; Salam et al., 2009; Masini et al., 2011) responses to stress do not always differ between sedentary and more active animals. These indices of neuronal activation provide further evidence that general central responsiveness to acute stress can be preserved in exercising animals, which is important given the critical role of stress responses in adaptation and survival (Munck et al., 1984; Levine and Ursin, 1991; Chrousos and Gold, 1992; Akil and Morano, 1995). Notably, whereas our previous work demonstrated enhanced HPA axis (Sasse et al., 2008; Nyhuis et al., 2010) and autonomic (Masini et al., 2011) response habituation to repeated stress exposures in exercised animals, the quantified brain regions of Ex and Sed rats did not exhibit significant differences in *Fos* mRNA induction immediately following the final (11th) repeated noise exposure. This general result may reflect the fact that brains were collected at a point during repeated stress exposures at which the neuroendocrine and autonomic response differences between exercised and sedentary rats have mostly dissipated (Sasse et al., 2008; Nyhuis et al., 2010; Masini et al., 2011). Future studies should include earlier time-points during repeated stress exposures (3rd–8th exposures) when HPA axis and autonomic response differences are observed. In addition, although non-significant, an interesting pattern emerged in the brains of exercised animals, such that acute and especially repeated noise stress appeared to *decrease Fos* mRNA induction in prefrontal cortical regions relative to that of No Noise controls (see **Table 1**). Prior associations with motivated behaviors (Cardinal et al., 2002) and stress adaptation (Campeau et al., 2002; Weinberg et al., 2009) make these regions important targets for future studies of the impact of physical activity on stress adaptation.

### **Crh AND Avp EXPRESSION IN THE PARAVENTRICULAR HYPOTHALAMIC NUCLEUS**

In the PVN, stress-induced release of CRH and AVP peptides are typically accompanied by increases in their gene transcription (Herman et al., 1992; Kovács and Sawchenko, 1996; Ma et al., 1997; Girotti et al., 2006) and subsequent elevations of steady state mRNA levels (Herman et al., 1992; Luo et al., 1994; Makino et al., 1995). It was therefore surprising to observe the lowest PVN-*Crh* mRNA levels in the Sed-Acute Noise group relative to that of the other Sed groups. However, PVN-*Crh* mRNA increases are typically observed 1–4 h after stress onset (Herman et al., 1992; Luo et al., 1994; Ma et al., 1997) and thus may have been missed given the shorter 30 min interval employed in the current study. This may also explain the lack of a *Stress Treatment* effect on the expression of *Avp* mRNA in the PVN, as elevation in *Avp* transcription is further delayed compared to that of *Crh* (Herman et al., 1992; Kovács and Sawchenko, 1996; Ma et al., 1997). An alternative explanation might be provided by unusually elevated *Crh* mRNA expression levels in the Sed-No Noise rats. Importantly, the lack of differences in plasma corticosterone concentrations between exercised and sedentary rats under basal conditions, after acute stress exposure, and following 11 days of repeated stress exposures (Sasse et al., 2008; Nyhuis et al., 2010) suggests that observed differences in PVN-*Crh* mRNA exerted little influence on HPA axis tone or sensitivity. Repeated stress is also reported to elevate *Crh* mRNA expression in the PVN (Aguilera, 1994; Herman et al., 1995; Gómez et al., 1996), yet the comparatively milder repeated noise stress paradigm used here did not produce significant increases in PVN-*Crh* or *Avp* mRNA expression in sedentary or exercised rats. The similar expression levels of these principal hypothalamic regulators of stress responses between exercised and sedentary rats therefore argues against an important role for these hypophysiotropic signals in the differential rates of stress adaptation we have observed (Sasse et al., 2008; Nyhuis et al., 2010; Masini et al., 2011).

#### **EXTRA-HYPOTHALAMIC** *Crh* **REGULATION**

*Crh* mRNA expression was examined in three additional, extrahypothalamic regions, including the central amygdala and the dorsal and ventral BNST. Previous studies have demonstrated that both acute and repeated stress activate the CRH systems in these regions (Chappell et al., 1986; Merali et al., 1998; Makino et al., 1999; Figueiredo et al., 2003) and have established important roles for their recruitment in mediating the expression and integration of behavioral and autonomic components of the stress response (Fisher et al., 1982; Koob et al., 1993). Each of these three regions exhibited a similar and interesting pattern of *Crh* regulation, in which expression was generally decreased by repeated noise stress in sedentary rats but was relatively unaffected, if not increased (although non-significantly), in voluntarily exercising animals exposed to the same repeated stress paradigm. The relatively short, 30 min interval between stress onset and euthanasia may again account for the lack of clear acute stress effects in these regions when compared to prior studies (Chappell et al., 1986; Merali et al., 1998; Makino et al., 1999; Figueiredo et al., 2003). Based on the results obtained in the Sed groups, one might hypothesize that a decrease in extra-hypothalamic CRH contributes to response habituation to repeated stress exposures. If this were the case, an even greater reduction in extra-hypothalamic *Crh* transcript would be expected in the Ex-Repeated Noise group, since response habituation to repeated noise was enhanced in these animals relative to their sedentary counterparts (Sasse et al., 2008; Nyhuis et al., 2010; Masini et al., 2011), but this was not observed. Whether and how differential regulation of these extra-hypothalamic influences contributes to the different rates of response habituation exhibited by exercised and sedentary rats remains an open question.

#### **REGULATION OF** *Crh* **RECEPTORS**

Expression levels of *Crhr1* mRNA in the medial and basolateral amygdala were similar between Ex and Sed groups, and were relatively insensitive to the different *Stress Treatment* conditions of this experiment, which is consistent with previous observations (Van Pett et al., 2000). In the PVN, *Crhr1* mRNA is typically expressed at very low levels basally, although several groups have reported inducible expression of PVN-*Crhr1* following exposure to a variety of different stressors (Luo et al., 1994; Makino et al., 1995; Rivest et al., 1995; Imaki et al., 1996; Van Pett et al., 2000). The only reliable difference in *Crhr1* expression observed in this study was exhibited in the PVN specifically of Ex rats, with significantly greater expression in the Repeated Noise compared to the Acute Noise group (although neither group differed significantly from No Noise controls). While others have demonstrated *Crhr1* mRNA induction in the PVN several hours following stressor onset (e.g., Van Pett et al., 2000), its expression was below the threshold of detection in the PVN of both Ex and Sed rats 30 min after acute noise onset. It thus remains possible that differential regulation of PVN-*Crhr1* might be revealed at later time-points following acute stress exposure.

Exploration of *Crhr2* mRNA expression, which is regionally more limited than that of *Crhr1* (Chalmers et al., 1995; Van Pett et al., 2000), did not reveal any differences in the lateral septum or medial amygdala. However, *Crhr2* mRNA expression was significantly reduced in the ventromedial hypothalamic nucleus of exercised compared to sedentary groups. Since *Crhr2* signaling in this region has previously been implicated in the regulation of food intake and energy balance (Steller, 1954; Spina et al., 1996), and significant alterations in these same measures are consistently reported in voluntarily exercising rats (Tokuyama et al., 1982; Rodnick et al., 1989; Afonso and Eikelboom, 2003), this finding may be important for future studies aimed at unraveling the complex metabolic adaptations associated with regular exercise.

#### **EFFECTS WITHIN THE BDNF-TRKB SIGNALING SYSTEM**

In three of the four hippocampal sub-regions in which *Bdnf* mRNA levels were assessed, expression was significantly greater in Ex compared to Sed rats. This finding is consistent with several other reports of increased hippocampal *Bdnf* expression in voluntarily exercising animals, both at the mRNA (Neeper et al., 1996; Oliff et al., 1998; Farmer et al., 2004; Gomez-Pinella et al., 2008; Nyhuis et al., 2010) and protein levels (Adlard and Cotman, 2004). Further, although stress has been demonstrated to reduce *Bdnf* expression in the hippocampus and PVN (Smith et al., 1995a,b; Nibuya et al., 1999; Adlard and Cotman, 2004), only non-significant reductions were observed here, particularly in the Sed-Repeated Noise rats. Importantly, the general trends we observed in the hippocampus were consistent with previously published results, in that voluntary exercise appeared to prevent the non-significant stress-induced decreases in hippocampal *Bdnf* mRNA expression observed in sedentary animals. Similarly, greater *Trkb* mRNA levels were expressed in the CA1 and CA3 pyramidal cell layers of Ex compared to Sed rats. Taken together, the exercise-induced increases in both ligand (*Bdnf*) and receptor (*Trkb*) transcripts in the present study are suggestive of an enhanced capacity for expression of the neuroprotective and synapse-strengthening functions of this signaling system. However, the overall lack of significant stress or interaction effects implies that the observed differences in hippocampal *Bdnf* and *Trkb* mRNA expression are not likely to provide the key neural substrate(s) directly mediating differential stress response habituation in exercised and sedentary rats.

#### **CONCLUSIONS**

This study provided a starting point for the elucidation of neurochemical mechanisms potentially underlying exercise-induced facilitation of glucocorticoid and cardiovascular response habituation to repeated audiogenic stress exposures that were recently reported (Sasse et al., 2008; Nyhuis et al., 2010; Masini et al., 2011). The primary conclusion that can be drawn from these results is that the brains of sedentary and physically active animals respond in multiple different ways to acute and repeated stress exposures. It is likely that some of the central changes reported here mediate various exercise-induced physiological adaptations (i.e., metabolic regulation). Additional direct intervention studies will be required to determine whether or how these observed brain modifications contribute to the enhanced rate of stress response habituation observed in exercised animals. Because it is the situations involving prolonged or repeated stress that are most strongly associated with illness, the possibility that interventions as cost-effective as regular exercise could enhance the rate of adaptation of potentially deleterious physiological responses to such situations strongly indicates the need for further investigation.

### **AUTHOR CONTRIBUTIONS**

Sarah K. Sasse performed the animal experiment with significant technical contributions from Tara J. Nyhuis, Cher V. Masini, Heidi E. W. Day and Serge Campeau. All *in situ* hybridization histochemistry was conducted by Sarah K. Sasse and Tara J. Nyhuis. Data analysis was performed by Sarah K. Sasse and interpreted by Sarah K. Sasse and Serge Campeau. Sarah K. Sasse and Serge Campeau wrote the manuscript.

### **ACKNOWLEDGMENTS**

This study was supported by NINDS R03NS054358 (Serge Campeau).

#### **REFERENCES**


integration of neuroendocrine, autonomic, and drinking responses. *J. Comp. Neurol.* 494, 75–107. doi: 10.1002/cne.20790


helplessness/behavioral depression: role of dorsal raphe serotonergic neurons. *J. Neurosci.* 23, 2889–2898.


hypothalamic paraventricular nucleus in the rat. *Brain Res.* 850, 136–143. doi: 10.1016/S0006-8993(99)02114-9


**Conflict of Interest Statement:** The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

*Received: 10 September 2013; accepted: 05 November 2013; published online: 25 November 2013.*

*Citation: Sasse SK, Nyhuis TJ, Masini CV, Day HEW and Campeau S (2013) Central gene expression changes associated with enhanced neuroendocrine and autonomic response habituation to repeated noise stress after voluntary wheel running in rats. Front. Physiol. 4:341. doi: 10.3389/fphys.2013.00341*

*This article was submitted to Clinical and Translational Physiology, a section of the journal Frontiers in Physiology.*

*Copyright © 2013 Sasse, Nyhuis, Masini, Day and Campeau. This is an openaccess article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.*