The Effect of Orthonasal and Retronasal Odorant Delivery on Multitasking Stress

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy

in the Graduate School of The Ohio State University

By

Alexandra M. Pierce-Feldmeyer, B. S.

Graduate Program in Food Science and Technology

The Ohio State University

2017

Dissertation Committee

Christopher T. Simons, Ph.D., Co-Advisor

Ken Lee, Ph.D., Co-Advisor

Yael Vodovotz, Ph.D.

Earl Harrison, Ph.D.

1

Copyrighted by

Alexandra Marie Pierce-Feldmeyer

2017

2

Abstract

On a daily basis Americans experience too much stress, often stemming from the multitude of tasks needed to be accomplished in the workplace. Drugs or psychotherapy is often prescribed for chronic stress, but acute stress, which is equally harmful has few effective relief mechanisms.

Aromatherapy has been suggested to elicit stress reducing properties but previous work produces variable results, causing studies to lack reputability. Robust literature is lacking likely because study methodologies vary heavily. Often, delivery method, odorant flow rate, aroma composition and concentration, and exposure time all differ, making studies difficult to compare. One of the variables that has not been formally assessed is the potential for odorant pathway to affect aroma stress reduction. Odorants can be sensed and perceived orthonasally (nostrils) or retronasally (oral cavity). Prior research indicates route of delivery impacts odorant perception, pleasantness, and directed behaviors thus suggesting differential processing of olfactory information. Therefore, differences in aroma delivery may dictate aromatherapy effectiveness. This important concept has not been studied previously. Lavender has a history of use in aromatherapy and thus we studied its impact compound, linalool delivered orthonasally and retronasally. Identical procedures assessed a control condition (ambient air inhalation) and vanillin, a common component of vanilla flavored foods. Adaptation, or the diminution in sensitivity to a stimulus following prolonged and constant exposure was first characterized for both the orthonasal and retronasal pathways. Linalool (12%) or vanillin (25%) were delivered orthonasally (6 LPM) and retronasally (8 LPM) in air phase through

ii a custom built olfactometer. Once the perception of aroma compounds was better understood, the same delivery took place when assessing aroma effects on stress. Thirty-one subjects underwent a baseline period (10 min) where they were instructed to "sit quietly and try to relax" followed by a multitasking computerized stressor (10 min) and finally a recovery period (10 min). For each condition (orthonasal or retronasal) 30 minutes of aroma (linalool or vanillin) or ambient air inhalation occurred. Objective measures (α -amylase, heart rate variability, and mean heart rate), and subjective measures (Bond-Lader Mood rating scale, Nasa Task Load Index and overall stress perception) were recorded during each condition. Linalool was found to have significant effects on physiological markers of stress, whereas air and vanillin showed less support for stress reduction.

Retronasal linalool exhibited the highest effect when compared to orthonasal administration. This study suggests linalool reduces stress in physiological ways, and is more effective when delivered retronasally, likely due to the higher concentration in the bloodstream.

Key words: olfaction, orthonasal, retronasal, stress, linalool, vanillin

iii

Dedication

To Robert Frost who helped inspire me to take the path less traveled.

To my mother who knows what I need before I do.

To my father who taught me how to work hard.

To my brother who isn’t afraid to say he likes Nickelback just as much as I do.

To my husband whom without I would be certifiably insane.

To my advisor because there won’t be a day in my professional life I won’t ask myself:

What would Dr. Simons do?

* clink *

iv

Acknowledgments

I would like to acknowledge my advisor, Dr. Simons and my committee members for guiding me through this process. I also want to acknowledge past and current Simons lab

members that kept me afloat and for always helping me kill the cockroaches in the basement of Howlett Hall. I also want to acknowledge the Ironman brand for teaching me

that Anything is Possible.

v

Vita

Educational Record:

Ph.D candidate Food Science and Technology Ohio State University, January 2014-present Columbus, OH

B.S. Food Science University of Illinois, August 2011-December 2013 Champaign-Urbana, IL

Publications

1. Bangcuyo, R.G., Smith, K.J., Zumach, J.L., Pierce, A.M., Guttman, G.A., Simons, C.T. The use of immersive technologies to improve consumer testing: the role of ecological validity, context and engagement in evaluating coffee. Food Quality and Preference. 41: 84-95, 2015.

Fields of Study

Major Field: Food Science and Technology

vi

Table of Contents

Abstract ...... ii Dedication ...... iv Acknowledgments ...... v Vita ...... vi List of Tables ...... ix List of Figures ...... x Chapter 1. Introduction ...... 1 Chapter 2: Literature Review ...... 6 Stress is a problem ...... 6 Defining Stress ...... 7 Measures of Stress ...... 11 Aromatherapy for stress mitigation ...... 15 Odorants with stress reduction properties ...... 16 Odor effect mechanisms ...... 20 Chiral odorant molecules support the pharmacological hypothesis ...... 21 Psychological theories: the role of expectation ...... 22 Variables of aromatherapy studies ...... 24 Odorant delivery route ...... 24 Exposure time ...... 26 Adaptation ...... 27 Compounds of Interest ...... 30 Lavender ...... 30 Vanillin ...... 40 Inducing cognitive stress ...... 40 Objective ...... 42 Chapter 3: Olfactory adaptation is dependent on route of delivery ...... 43 Chapter 4: Validation of an ecological laboratory induced stressor to assess amelioration techniques ...... 59 Introduction ...... 59 Stress is a problem ...... 59

vii

Types of Stress ...... 59 Laboratory measures ...... 60 Materials and Methods ...... 63 Subjects ...... 63 Procedure ...... 63 Statistical Analysis ...... 66 Results ...... 68 Subjective results ...... 68 Physiological ...... 69 Discussion ...... 74 Chapter 5: The effect of orthonasal and retronasal odorant administration on multitasking stress reduction ...... 77 Introduction ...... 77 Materials and Methods ...... 80 Statistical Analysis ...... 85 Results ...... 86 Discussion ...... 103 α -Amylase ...... 103 Linalool ...... 103 Vanillin ...... 107 Conclusion ...... 111 Future studies ...... 112 Bibliography ...... 114 Appendix A: NASA Task Load Index ...... 138

viii

List of Tables

Table 1. Average HR (beats per minute) reported as M± SE during baseline, stress and recovery for orthonasal air (OA), orthonasal linalool (OL), orthonasal vanillin (OV), retronasal air (RA), retronasal linalool (RL) and retronasal vanillin (RV). Superscripts indicate significant differences between average heart rates within each condition...... 91

Table 2. Orthonasal heart rate variability (M± SE) indexed by RMSSD (ms) and RSA (ms2) during baseline, stress and recovery for OA, OL, and OV. Superscripts indicate significant differences (α =0.05) between HRV within each condition ...... 93

Table 3. Retronasal heart rate variability (M± SE) indexed by RMSSD (ms) and RSA (ms2) during baseline, stress and recovery for RA, RL, and RV. Superscripts indicate significant differences (α =0.05) between HRV within each condition ...... 94

Table 4. Overall stress VAS ratings (M± SE) during baseline, stress and recovery (arbitrary units) for OA, OL, OV, RA, RL and RV. Superscripts indicate significant (α=0.05) differences between perceived stress within each condition ...... 95

Table 5. NASA Task Load Index orthonasal condition averages (M±SE) at baseline, pre- stress, post-stress and following recovery (arbitrary units) for OA, OL, and OV...... 96

Table 6. NASA Task Load Index retronasal condition averages (M±SE) at baseline, pre- stress, post-stress and following recovery (arbitrary units) for RA, RL and RV...... 97

Table 7. α -amylase activity (U/mL) changes from baseline reported as M± SE collapsed over orthonasal and retronasal inhalation conditions for pre-stress, post-stress, recovery (5 minutes) and recovery (10 minutes). Superscripts indicate significant (α=0.05) differences between α –amylase activity within each inhalation condition...... 102

ix

List of Figures

Figure 1. Orthonasal and Retronasal Olfaction (Liu 2013) is depicted in this figure. Orthonasal olfaction occurs as a result of odor molecules traveling via the external nares and being delivered to the olfactory mucosa to be further process, depicted by the triangular-dot shaped receptors. Retronasal olfaction is depicted by arrows directing the movement of odor molecules through the mouth, the back of the throat and finally ascending towards the nostrils for exhalation. Retronasal odors follow this route to culminate at the same olfactory mucosa location as orthonasal volatiles for processing of sensation and perception...... 25

Figure 2. Retronasal (A) and orthonasal (B) aroma delivery systems. Note, the retronasal system was disassembled to show component parts (mouthpiece, glass manifold and connecting tube). Each system was connected to an aroma source (12% or 1.5% linalool and 25% or 6.25% vanillin) that delivered the odorant at 8 LPM (retronasal) or 6 LPM (orthonasal) using deodorized, humidified breathing air...... 47

Figure 3. Orthonasal (A) and retronasal (B) perceived intensity before, during, and after 10-mins of continuous linalool (12%; hollow circles) or vanillin (25%; black circles) delivery. Orthonasal aroma delivery resulted in both linalool (hollow circles) and vanillin (black circles) adaptation whereas the intensity of retronasal stimuli (linalool and vanillin) increased with continued exposure. For each panel, different letters above or below each circle denote significant differences in perceived linalool or vanillin intensities...... 50

Figure 4. Perceived intensity of a weak linalool (1.5%) and vanillin (6.25%) odorant stimulus before, during, and after 10-mins of continuous retronasal (hollow circles) or orthonasal (black circles) delivery. Note for each odorant, orthonasal delivery elicits a response that tends to adapt over time whereas perceived intensity of the same stimulus increases when delivered retronasally. For each panel, different letters denote significant increases or decreases in perceived retronasal or orthonasal intensities. The time- condition interaction term is significant for both (A) linalool and (B) vanillin for a split- plot analysis (linalool: p=0.006, vanillin: p<0.001), suggesting delivery route produces significantly different intensity perceptions over time...... 52

Figure 5. Cross-adaptation. (A) Orthonasal to retronasal cross-adaptation. Evidence of a retronasal stimulus adapting following 10-min of continuous orthonasal exposure was not x observed for linalool (hollow circles) or vanillin (black circles). (B) Retronasal to orthonasal cross-adaptation. Evidence of an orthonasal stimulus adapting following 10- min of continuous retronasal exposure was not observed for linalool (hollow circles) or vanillin (black circles). For each panel, different letters above or below the circles denote significant differences in perceived vanillin or linalool intensities...... 53

Figure 6. Multitasking stressor paradigm with subjective and objective data collection . 66

Figure 7. NTLX perceived changes in mental, physical, and temporal demand, as well as subjects' perceived level of performance, effort and frustration during baseline, pre-stress, post-stress and recovery. When asked to assess their subjective task load index post- stress, subjects reported significant increases in perceived mental (p<0.001), physica l(p<0.001) and temporal demand (p<0.001), as well as perceived level of performance(p<0.001), effort (p<0.001) and frustration(p<0.001)...... 68

Figure 8. Perceived stress level during baseline, pre-stress, post-stress and recovery. Post stressor, subjects perceived a significantly higher (denoted by the *) level of stress (p<0.001) when compared to other session conditions...... 69

Figure 9. Mean heart rate change from baseline in the multitasking stress (MT) at 10 and 20 minutes, and during recovery phases at 5 and 10 minutes. * indicate significant changes in heart beats per minute compared to baseline heart rate (p=0.010)...... 70

Figure 10. Heart rate variability (RMSSD) change from baseline reported as mean plus/minus SE (ms). This change from baseline is the differences in time between successive heart rate peaks compared to baseline time between heart rate peaks. During multitasking (MT) at 10 and 20 minutes there is a significant decrease in HRV (p<0.001) showing HRV was affected in response to the stressor, and that during recovery HRV returned to baseline levels...... 71

Figure 11. Heart rate variability (RSA) change from baseline reported as mean plus/minus SE (ms2). RSA is the natural log transformation of high frequency power measured in ms2. The change from baseline represents changes in RSA during multitasking (MT) and recovery. During MT at 10 and 20 minutes there is a significant decrease in HRV (p<0.001) showing HRV was affected in response to the stressor, and that during recovery HRV returned to baseline levels...... 72

Figure 12. GSC change from baseline reported as mean plus/minus SE (microsiemens). The change from baseline represents changes in skin conductance during multitasking (MT) and recovery. There are no significant differences between MT and recovery conductance levels (p=0.111) showing GSC was not affected or potentially confounded in response to the stressor...... 73

xi

Figure 13. Change from baseline in α-amylase activity (U/mL) pre-stress, post-stress, and in recovery at 5 minutes and 10 minutes. There are no significant differences for α- amylase levels between pre-, post-stress, and recovery, however the trend is a pattern that reflects higher levels in response to a stressor and lower levels at times of rest and recovery...... 74

Figure 14. Procedure for multitasking aroma session ...... 84

Figure 15. Mean HR changes reported in mean plus/minus SE beats per minute in (A.) orthonasal odor conditions and (B) retronasal odor conditions collapse over session time. These values are the average heart rate over baseline, stress and recovery combined. In the orthonasal conditions, both linalool and vanillin average heart rates were significantly (linalool: p; vanillin: p) lower than their control condition. The same pattern was not seen for retronasal conditions; retronasal linalool and vanillin did not differ from the retronasal air condition in average heart rate...... 90

Figure 16. HRV (RMSSD) changes from baseline over time for (A) OA, (B) OL, (C) OV, (D) RA, (E) RL and (F) RV reported as mean±SE (ms). This change from baseline is the differences in time between successive heart beats compared to baseline time between heart beats. Superscripts indicate significant (α=0.05) differences between average heart rates within each condition...... 92

Figure 17. HRV (RSA) changes from baseline over time for (A) OA, (B) OL, (C) OV, (D) RA, (E) RL and (F) RV reported as mean±SE (ms2). RSA is the natural logarithm of high frequency power spectrum reported in ms2. Superscripts indicate significant differences (α =0.05) between average heart rates within each condition...... 93

Figure 18. Overall stress VAS changes over time for (A) OA, (B) OL*, (C) OV, (D) RA, (E) RL and (F) RV reported as mean ± SE (arbitrary units). Superscripts indicate significant (α=0.05) differences between perceived stress within each condition...... 95

Figure 19. α -amylase activity (U/mL) changes from baseline for orthonasal and retronasal A. air, B. linalool, and C. vanillin conditions. Letters depict significant (α = 0.05) differences within that condition...... 102

xii

List of Abbreviations

• ANOVA: analysis of variance

• ANS: autonomic nervous system

• BP: blood pressure

• CgA: chromogranin A

• EEG: electroencephalography

• fMRI: functional magnetic resonance imaging

• GSC: galvanic skin conductance

• HR: heart rate

• HRV: heart rate variability

• IgA: Immunoglobulin A

• NTLX: national aeronautics and space administration task load index (NASA

TLX)

• OA: orthonasal air

• OL: orthonasal linalool

• OV: orthonasal vanillin

• PEBL: psychology experiment building language

• RR: respiration rate

• RSA: Respiratory sinus arrhythmia

• RA: retronasal air

• RL: retronasal linalool

• RV: retronasal vanillin xiii

• RMSSD: root mean square of successive differences

• STAI: state trait anxiety inventory

xiv

Chapter 1. Introduction

In the 2014 Stress in America survey, adults and children claimed they experienced stress levels higher than what they believe to be healthy and feel ill equipped to reduce stress (American Psychological Association 2014). Stress is a problem because it is associated with increased risk for cardiovascular disease (Vrijkotte et al. 200), immune suppression and cancer (Dienstbier 1989; Fraser et al. 1999), along with enhanced feelings of anxiety, worry (Hunter and Thatcher 2007) and nervousness (Cooke and Rousseau, 1984). Medications and psychotherapy are often used to treat chronic stress leaving many suffering from acute stress without appropriate treatment options.

Acute stress can stem from the multitude of responsibilities that people manage on a daily basis. Attempting to juggle multiple responsibilities is often managed using technology.

It was found that current technology use contributes to daily stress as well with little chance of decreasing (American Psychological Association 2017). Managing constant obligations forces people to multitask, especially in today’s fast paced, tech driven society.

Aromatherapy has been suggested to reduce stress, but its benefits are inconclusive, likely due to its complex nature. Multiple variables are associated with odorant delivery, stress induction, and measures used to indicate changes in stress reduction. Aromatherapy has often been studied in ambient environments and nonspecific 1 olfactory delivery, that is orthonasal delivery (odorants reaching the olfactory epithelium via the nostrils) and retronasal delivery (odorants reaching the olfactory epithelium via the oral cavity) have not previously been distinguished when assessing aromatherapy effectiveness. Reasons to separate these pathways are driven by their ability to evoke different responses (Heilmann and Hummel 2004; Small et al. 2005; Hummel et al.

2006), thresholds (Pierce and Halpern 1996; Hummel et al. 2006; Visschers et al. 2006) perceptual qualities (Pierce and Halpern 1996; Hummel et al. 2006), hedonic responses

(Small et al. 2005), and behaviors (Rozin 1982; although see Burdach and Doty 1987 and

Heilmann and Hummel 2004). These differences suggest that aromatherapy effectiveness may also vary if odorants are delivered via different aroma routes. Because previous studies are so variable, to effectively study aromatherapy amelioration, controls need to be put into place to assure results are reproducible.

We aim to study stress reduction effects over a 30 minute exposure to select aroma compounds. Following prolonged and continuous exposure of odorants, a phenomenon known as adaptation occurs. Adaptation is a form of neural processing that results in decreased perceived intensity of a stimulus, which chapter 3 characterizes.

Additionally, prior research indicates route of delivery impacts odorant perception, pleasantness, and directed behaviors thus suggesting differential processing of olfactory information. Chapter 3 also determines whether route of delivery differentially impacts olfactory adaptation and whether cross-adaptation occurs between orthonasal and retronasal pathways. Chapter 3 represents objective 1. This is critical to aromatherapy assessment because of the possibility that (a) pathway intensity perceptions differ over

2 time and (b) that aromatherapy assessment methodology will include extended aroma exposure time. Extended time and intensity perception are important notions for aromatherapy assessment due to previous studies touting effects of expectation during aroma exposure. Aromatherapy assessment takes place over time, and therefore changes in perceptual intensity perception may manipulate subjects’ expectation of stress relieving properties. If adaptation occurs, a lack of perceptual odor inhalation may prompt them to feel less affected by the odorants. If expectation plays a role in aromatherapy effectiveness, it will help researchers better characterize stress reduction effects specific to each pathway based on adaptation perceptions over time of exposure.

Once adaptation is better characterized and therefore controlled for during aromatherapy assessment, it is important to induce a controlled and ecologically valid stressor which is the second objective. Chapter 4 validates a simple, open-source stressor that mimics a real world type of multitasking stress. Multitasking computerized tasks are an effective stressor for both objective and subjective measures of stress. Studying stress is complicated as indicated previously due to its multifaceted nature. Other validated stressors used commonly in stress studies induce maximal levels of stress that are not usually experienced on a daily basis. Chapter 4 validates an acute, real-world, multitasking stress paradigm that is easily replicated in other laboratories. We hypothesized that a low intensity stressor may be more amenable to reduction by way of aroma. Additionally, with an effective stressor, mitigation techniques, such as aromatherapy, may be investigated in a more controlled environment. Often authors try to combine and correlate subjective methods (self-report) with objective ones (physiological

3 biomarkers). Usually a study is stronger by using both types of measures since using one or the other may miss aspects by which stress manifests. Because stress affects human physiology as well as our psychological well-being, it is important to include indices of both objective and subjective changes. Additionally, leaving out subjective reports disregards a person’s perception of stress. It is important to study both manifestations of stress, in order to form a more comprehensive understanding of how stress has been affected due to intervention techniques. For this purpose, physiological measures such as salivary α-amylase, heart rate variability (HRV), mean HR and subjective stress ratings are used simultaneously to assess the impact of odorants on stress.

With a thorough understanding of the potential for adaptation to affect aromatherapy results and a validated controlled stressor, Chapter 5 characterizes stress reduction with two odorant compounds delivered on separate occasions orthonasally and retronasally. Odorants of interest include linalool, the impact compound associated with lavender, as well as a commonly liked flavor compound, vanillin.

Lavender exhibiting stress relief is a more substantiated claim for aromatherapy stress mitigation. Vanillin is also of interest due to its similar hedonic quality because it is often used as a flavoring agent. We hypothesized that linalool would elicit effects on stress measures due to prior studies (Motomura et al. 2001; Bradley et al. 2009; MPham et al. 2012; Bakhsha et al. 2016). Vanillin was assessed to control for pleasantness.

An additional control for this study was the use of aroma compounds instead of essential oil blends. Effects can then be attributed specifically to that compound rather than essential oil blends which have variable ingredients (Robu et al. 2011).

4

With a validated stressor and a thorough understanding of perceptual odorant changes over time, Objective 3 is to elucidate the potential for stress mitigation via linalool and vanillin compounds either orthonasally and retronasally during a multitasking acute stress induction period. With this design I assess linalool and vanillin multitasking stress reduction and if differences between orthonasal and retronasal delivery exist.

5

Chapter 2: Literature Review

Stress is a problem

The Stress in America survey revealed adults and children experience stress levels higher than what they believe to be healthy and are ill equipped to effectively reduce stress in their daily lives (American Psychological Association 2014). In the 2017 survey,

31% of Americans wished they spent less time worried or stressed in the past 10 years of their life, with 36% reporting that reducing stress is a priority for them (American

Psychological Association 2017). Additionally a majority of Americans continue to report that work is a significant source of stress and that they struggle with stress management techniques. The 2017 report found nearly 99% of American adults have an electronic device connected to the internet, and that with more connectivity, stress rose (Stress in American Survey Part 2 2017). Younger Americans have also reported in surveys higher than average levels of stress, which may be due to enhanced technology use. Inevitably, tech use will likely not diminish and stress will continue to rise.

Stress is a problem because it increases risk for cardiovascular disease

(Vrijkotte et al. 2000), immune suppression and cancer (Dienstbier 1989; Fraser et al.

1999), and enhanced feelings of anxiety, worry (Hunter and Thatcher 2007) and nervousness (Cooke and Rousseau 1984). Cohen and Williamson (1991) further support 6 the idea that stress of multiple types increases risk for infectious disease by influencing pathogenesis. Medications and psychotherapy are often used to treat chronic stress leaving many suffering from acute stress without appropriate treatment options. Acute stress can stem from the multitude of responsibilities that people are expected to manage on a daily basis, forcing them to multitask, especially in today’s fast paced, technology directed society. Clearly, methods for mitigating multitasking stress must be investigated.

The difficulty with studying stress arises from its general definition

(Kurniawan and others 2013). Selye (1950) defines it as a disruption in one’s homeostasis. Given this general view of stress, there are potentially many different sources and manifestations of stress in humans, making potential mitigation techniques abundant as well.

Defining Stress

With an adequate definition we can begin to understand the components of stress, and how it might be combatted in alternative ways. Stress has been defined as hardship or adversity (Lazarus 1998). Given this general view of stress, there are potentially many manifestations of stress, especially in humans. When studying stress, often anxiety level is measured to figure out if an intervention has reduced stress. For example, studies have used anxiety perceptions when attempting to understand stress reduction during test taking (Brewer 2002).

7

If anxiety is considered an entity of stress, or at the least a significant component, there are arguably other facets of stress that are measureable. For example some studies assess stress reduction by measuring a variety of perceived moods (Diego and others

1998; Kuroda et al. 2005; MPham and Siripornpanich 2012). Some studies have also correlated lower anxiety to increased alertness, suggesting enhanced alertness indicates reduced stress (Raudenbush et al. 2002; Raudenbush et al. 2009). Pain reduction as an indication of decreased stress has also been used to promote methods of stress reduction, particularly regarding patients undergoing surgical procedures such as needle insertion

(Kim et al. 2011), anesthesia administration (Struys and others 2002), painful electrical stimulation to the forefinger (Masaoka et al. 2013), and infant heel pricks

(Goubet 2003; Rattaz et al. 2005). Stress might be considered chronic, which is defined as persistent stress induction, such as that brought on by chronic illnesses. Other stress may be more acute or induced in the short term, such as stress brought on by cognitively demanding tasks. With different ways stress manifest, there is a need to treat them differently.

Mental stress falls into three categories that include major life events (divorce, death, moving), daily activities (losing things, altercations, responsibilities) and psychological distress (anxiety, depression) (Cohen and Williamson 1991). Major life events, while certainly producing stress, are not a daily occurrence for a majority of people. Often this stress will be ameliorated with time and adjustment to the new life event. Without proper mitigation, stress induced by major life events may manifest into chronic distress or ongoing dysphoria. The remaining type of stress consisting of daily

8 activities such as losing things, social obligations or complications is more common but has perhaps been less studied since it is perceived to be less threatening. Studies have shown support, however that all three types of stress influence risk for infectious diseases, which makes studying daily sources of stress, and ways to combat them, critical to human health (Cohen and Williamson 1991).

In order to study mitigation methods, laboratories must find ways to induce realistic, but controlled, stress. Dickerson and Kemeny (2004) reviewed several studies and organized laboratory induced stressors into four groups: public speaking/verbal interaction, cognitive tasks, emotion induction procedures, and noise exposure. Dickerson and Kemeny’s (1991) review found significant effects of cognitive tasks, public speaking/verbal interaction and a combination to increase stress. This study also found that a combination of public speaking and cognitive tasks induced the highest cortisol response. This suggests this type of stress may elicit larger responses creating a model to study stress mitigation techniques (Dickerson and Kemeny 2004). Prior studies of stress often utilize the Trier’s Social Stress test to study psychobiological responses

(Kirschbaum and others 1993; Balodis and others 2010). Trier’s test consists of anticipation of procedure for 10 minutes, a test period when subjects deliver an original speech, followed by mental arithmetic in front of an audience (Kirschbaum and others

1993). In this study, multiple hormone levels and heart rate differed significantly from baseline, proving efficacy to stimulate stress. Other studies interested in stress mitigation have used Trier’s since it produces fairly consistent stress responses (Brody and others

9

2002; Creswell and others 2005; Kiecolt-Glaser and others 2009; Routledge and others

2011).

Although these studies adequately induce stress, allowing for the study of amelioration techniques, the stress being managed resulting from Trier’s test is of a maximal level and therefore the amelioration methods proposed by those studies might not be effective for a more common, daily multitasking induced stress. Multitasking induced stress has been less frequently studied. Scholey and others (2009) induced acute psychological stress by having subjects attend to four computerized tasks simultaneously while assessing the ability of chewing gum to alleviate stress. The study found chewing gum reduced stress significantly as reported by Bond Lader mood scales and state trait anxiety scores, while also significantly lowering salivary cortisol (Scholey and others

2009). This is one of the only studies that has created a multitasking stress paradigm to assess amelioration techniques via subjective feelings of stress paired with a physiological response. However, Johnson and others (2011) repeated that test design and did not find significantly reduced stress, just heightened alertness from the chewing gum

(Raudenbush and others 2002; Raudenbush and others 2009). Another study analyzed galvanic skin conductance (GSC) during arithmetic and reading tasks (Nourbakhsh and others 2012), while another assessed fragrance as a stress reliever after subjects attended to the Stroop color word task and mental arithmetic (Mezzacappa and others 2010), but both studies had subjects attending to tasks singularly. Kennedy and others (2004) used multitasking stress simulation where subjects attended to four tasks in order to assess a dosage of lemon balm’s effect on stress. The stress reported was only subjectively rated

10 without inclusion of physiological effects of the multitasking. Other studies have used multitasking as a way to assess physiological stress responses but have not studied amelioration techniques (Hjortskov and others 2004; Wetherell and others 2004). A pertinent multitasking stress simulation should affect subjective feelings related to stress while also affecting physiology compared to a baseline measure at rest. With validated multitasking stress induction, appropriate mitigation methods can be studied.

Measures of Stress

In addition to different types of stress, another complication related to studying stress and reduction techniques are the gamut of subjective and physiological measures available. There is a need for adequate stress measures in order to validate specific stress induction and reduction. Dickerson and Kemeny (2004) reviewed 208 laboratory studies and found that cortisol is not responsive to all types of stress, responding in the largest manner to uncontrollable and social-evaluative stressors. When studying stress, anxiety level is a measure used often to assess if an intervention has reduced stress. For example, studies have used anxiety perceptions when attempting to understand stress reduction during test taking (Brewer 2002) and career oriented interventions of odorants (Crow and others 2008).

Multiple stress measures have been found to help characterize different parameters of stress responses (Gordis and others 2006; Balodis and others 2010). With the multifaceted nature of stress responses, a strong study will measure both subjective and objective measures, using appropriate indices dependent on study methodology. 11

One measure of interest for the body’s physiological response to stress is galvanic skin conductance (GSC) (Setz and others 2010; Sun and others 2010; Nourbakhsh and others

2012). When under stress, the body’s sweat response increases, changing the skin’s conductivity (Gruber and Moore 1997). When skin conductance is measured, the amplitude of waves produced in the output is measured and compared to different conditions. This measure indicates sympathetic nervous system arousal (Boucsein et al.

2012). Studies have shown that different events may affect GSC like social engagements, a new environment, or stress and fear (Boucsein et al. 2012). Thus it is critical to note every event during a recording to correct for protocol independent events affecting GSC unnecessarily. GSC may also be affected by relative humidity, age of subjects and ethnicity which may depend on active sweat gland density (Boucsein et al. 2012).

Another measure of interest is heart rate variability (HRV). At its core, heart rate variability is the beat-to-beat changes in heart rate. High HRV indicates good adaptability, whereas low HRV indicates abnormal physiological condition or mental stress (Pumpria et al. 2002). Heart rate variability has been found to indicate vagal sinoatrial control which correlates with parasympathetic activity (Berntson et al.

2007). There are several indices of HRV. Time domain methods are often used to measure HRV and include but are not limited to the root of the mean squared differences of successive NN intervals (RMSSD). These measures also correlate with the parasympathetic nervous system, for example, respiratory sinus arrhythmia (RSA), which

12 is indicative of vagal modulation and thus correlates with parasympathetic vagal activity

(Berntson and others 1993; Task Force of The European Society of Cardiology 1996).

RSA is useful in that it is sensitive not only to physiological variables, but behavioral as well. Decreased RSA has been found to correlate with lower functioning autonomic systems (Feldman and others 2013). With mental and cognitive stressors, RSA is expected to decrease (Grossman et al. 1990). One study found that subjects playing a video game with a known risk of shock produced higher heart rate and decreased RSA values (Grossman and Svebak 1987).

Mean heart rate (mean HR) is the average heart rate recorded during a time period of a given stimulus. Mean heart rate has been reported to elevate when subjects are aroused or stressed and thus may be useful as a measure of multitasking stress induction periods (Connor-Smith et al. 2000, Lang et al. 1993).

Studies have frequently used GSC (Setz and others 2010; Nourbakhsh and others

2012) and heart rate variability, or both (Sun and others 2010) as indicators of stress.

Studies have also used salivary components as indicators of stress such as cortisol and α- amylase. Elevated cortisol occurs during stress, which can inhibit proteins that play a role in the inflammation response, thus affecting a person’s health (Dickerson and Kemeny 2004). Inflammation is reportedly a component of cardiovascular disease, diabetes, dementia, certain cancers and possibly depression (McDade et al.

2010). α-Amylase is an enzyme found in the oral cavity that digests carbohydrates and also responds to stress, reflecting sympathetic nervous system activity

(Bohan and Maye 2010). Cortisol and α-amylase respond differently to stress, with

13 cortisol peaking in response to a stressor and recovering slower than amylase

(Balodis and others 2010). This discrepancy suggests the critical need to understand the characteristics of measures chosen to study laboratory induced stress mechanisms and adjust methodology accordingly. Because the autonomic system regulates salivation and its components such as α-amylase, it reflects sympathetic nervous activity (Chiappelli et al. 2006). The benefit to using α-amylase as a measure of the nervous system response is its reported greater sensitivity and specificity than blood pressure or heart rate (Chiappelli et al. 2006). Also α-amylase has been found to have a much faster recovery time than cortisol (Obayashi 2013).

Effects of psychological stress on physiological systems are variable (Dickerson and Kemeny 2004), thus stressor type influences the pertinent physiological measures recorded. Effective delivery of stress and adequate measures are needed in order to create a reliable and meaningful laboratory induced stressor. To determine relevant measures of multitasking induced stress, studies regarding physiological parameters and subjective assessments in the presence of related stressors were reviewed (Vrijkotte et al. 2000;

Hjortskov et al. 2004; Li et al. 2009; Filaire et al. 2010; Kim et al. 2012; Thayer et al.

2012; O’Donnell et al. 2015).

Filaire et al. (2010) studied gender effects in teachers with stress induced by lecturing, using salivary α -amylase, HRV and subjective ratings. This study found increased stress levels indicated by reduced HRV, elevated salivary amylase and subjective responses from the State Trait Anxiety Inventory and the Perceived Stress

Scale, but did not find a gender effect, making these stress measures conducive to a male

14 and female sample study population (Filaire et al. 2010). Another notable facet of this study was the effect of time when salivary amylase was collected, necessitating change of baseline measures for analysis (Filaire et al. 2010). Hjortskov et al. (2004) indicated that

HRV was more sensitive than blood pressure as a physiological measurement responsive to mental stress. Subjective measures were, however, not different, arguing for utility of both stress measures (objective and subjective) in order to better illuminate stress response. Statistically significant changes in HRV were also found in nursing students

(Kim et al. 2012), subjects attending to video games (Li et al. 2009) and subjects performing a job-demand-control paradigm that also elevated α –amylase (O’Donnell et al. 2015). Work-style stress similarly elevated blood pressure and detrimentally affected

HRV (Vrijkotte et al. 2000).

Prior studies of stress often generate maximal levels of stress and are, therefore, less ecologically valid than the multitasking stress including job related tasks and mental loads often experienced daily. A validated laboratory-based model of multitasking stress induction is therefore needed to investigate mitigation methods for this common stressor.

Aromatherapy for stress mitigation

Aroma inhalation in humans has been suggested as a way to potentially mitigate forms of stress. However, results are variable and inconclusive. Odorant molecules have shown contradicting effects, which is perhaps not surprising since stress is such a multifaceted construct. Although complex, several studies have studied aroma inhalation interventions and their respective stress measures.

15

Aromachology is defined as the scientifically documented study of olfactory effects in humans (Herz 2009). Aromachology, or aromatherapy, has historically been used for several reasons, including healing purposes (Tillett 2010), and emotional health benefits (Edwards 2015). More recently aromas have been assessed for other cognitive benefits such as mood and behavior (Herz 2009). Specifically, aromas have been used for stress reduction. Although aromachology has been in practice for a long time, most of the applications are based on anecdotal evidence and lack scientific support. Without reputable research findings, mechanistic clarity is lacking and aroma effects are inconclusive. Among past studies, aromas with reported effects on stress include lavender, orange, bergamot, ylang-ylang and others, but their effects remain elusive and difficult to define (Herz 2009). A better understanding of aromachology is difficult because of a lack of standard methodologies of odorant administration and/or stress induction (if induced at all), as well as the spectrum of indices that measure stress. In the past, the breadth of uncontrolled variables and different methodologies make overall comparisons and conclusions difficult to obtain.

Odorants with stress reduction properties

Studies include odorants such as floral and musk aromas (Fukui et al. 2007) limonene and carvone, which are associated with citrus (Heuberger et al. 2001), orange oil (Lehrner et al. 2000) isovaleric acid, a cheesy, sweaty smell, thiophenol, a foul sulphurous smell, pyridine, a nauseating, fish-like aroma, L-menthol, a minty cooling sensation, isoamyl acetate, a primary component of a banana aroma, and 1-8 cineole, a

16 eucalyptus smell (Bensafi et al. 2002). Additionally many researchers have looked at effects of bergamot (Chang and Shen 2011; Liu et al. 2013; Peng et al. 2009), ylang- ylang (Cheng et al. 2003), peppermint oil (Goel and Lao 2006), vanillin

(Goubet 2003; Rattaz et al. 2005), neroli (Holm and Fitzmaurice 2008), geranium oil and rosemary (Morris 1995), and coconut extract (Mezzacappa et al. 2010).

By presentation of aromas (isovaleric acid, thiophenol, pyridine, L- menthol, isoamyl acetate and 1-8 cineole) via flasks below the right nostril, Bensafi et al.

(2002) aimed to correlate odor intensity, arousal, pleasantness and familiarity ratings with activation of the autonomic nervous system (ANS). This study was performed on healthy individuals without the induction of any type of stress. ANS measures included GSC, and

HRV. The study reported pleasantness correlated with heart rate variations while arousal correlated with GSC, indicating olfactory stimulation with variable odorants affects ANS parameters (Bensafi et al. 2002).

Bergamot essential oil has also garnered attention for its potential effects on stress. In a study looking at workplace stress-related illness in elementary school teachers, it was found bergamot administered as an aromatic spray decreased BP, HR and elevated HRV (Chang and Shen 2011). Again in Taiwanese elementary school teachers, natural bergamot essential oil was compared to a synthesized version, finding the natural oil to be the only condition exhibiting an effect on HRV (Liu et al. 2013). This again supports the idea that some methods of stress reduction may be fairly specific to the compounds being tested.

17

Another study analyzed bergamot’s effect on ANS upon inhalation via an ultrasonic atomizer. HRV, mean HR, and systolic and diastolic BP were measured.

Authors found significant increases in relaxation based on physiological HR measures induced by bergamot (Peng et al. 2009). Inhalation of steam with either no aroma, geranium oil or rosemary oil was measured with STAI reports. Geranium was found to reduce state and trait anxiety measures (Morris et al. 1995).

Inhalation of coconut extract by nasal cannula in response to laboratory stress via Stroop color-word and arithmetic tasks was measured using HR, HRV and BP measures. HR responses were muted in the arithmetic task when exposed to coconut, but

HRV did not change significantly (Mezzacappa et al. 2010).

Additionally, essential oil of orange was found to reduce anxiety levels, enhance feelings of calmness and positive moods in women in a dental office waiting room

(Lehrner et al. 2000). Essential oil of neroli, produced from blossom of a bitter orange tree, however, did not induce changes in the STAI after exposure in an emergency department waiting room (Holm and Fitzmaurice 2008). This is yet another example of potentially how specific aromatherapy effects can be when controlling for differences among odorant molecules and types of stress being assessed.

One laboratory looked at different scenarios related to stress and how peppermint might be reducing perceived stress by lessening anxiety. This study looked at athletic performance in the presence of different aromas (peppermint, jasmine or dimethylsulfide) aerated and delivered by nasal cannula directed to subject nostrils. Athletes were put through physical stress by a treadmill workout. Oxygen saturation, pulse, and BP were

18 measured, but no effect was found for the odor condition, and additionally no effect on anxiety was found in this study (Raudenbush et al. 2002). This same group looked at peppermint regarding stress and cognitive benefits during a driving simulation task.

Participants wore nasal cannulas through which either low flow oxygen (control condition) or oxygen plus odorant was administered (Raudenbush et al. 2009).

Participants were told that they may or may not perceive a scent through the cannula, but were not informed as to what that scent might be. During driving simulation, participants assessed workload demands of prolonged driving, and completed the NASA Task Load

Index (NTLX) and the Profile of Mood States. They also rated levels of alertness before and after the intervention. There was only a significant effect found in the scent condition. Anxiety ratings were lower in the peppermint condition as compared to both the control and cinnamon condition. One weakness of this study was their lack of biological markers of stress. Correlating those measure may provide more information regarding changes in stress.

Studies assessing vanillin’s analgesic effect have been conducted on infants during routine heel prick tests or blood draws. Goubet et al. (2003) looked at the effect of familiarizing infants with vanillin and found less distress during heel prick tests.

However, vanillin administered without previous exposure did not exhibit these changes, which may support the idea that familiarity was the reason for reduced stress and not the inherent compound. The same findings were reported again by Goubet et al. (2007) during routine infant blood draws. This finding was again reinforced after Rattaz et al.

(2005) found familiarizing infants with vanillin, as well as with mothers’ milk, showing

19 the infants in familiarizing conditions exhibited less crying and grimacing during heel prick tests. These studies suggest that vanillin, as a compound, may not elicit pain or stress relieving effects, unless familiarization takes place beforehand. A limitation of this study is that the effects were only studied on infants, making it impossible to generalize to other age groups.

Odor effect mechanisms

If stress reduction occurs with specific odorants, understanding their mechanisms may generate insight into how other molecules could have similar properties. With mechanistic knowledge, it becomes possible to predict effective stress reduction odor molecules. Herz (2009) makes a point to discuss the potential mechanisms by which odorants change feelings of stress. The ways aroma compounds work primarily fall into two categories (or likely both), pharmacological and psychological. A pharmacological mechanism is described as direct interaction with the autonomic or central nervous system, and possibly the endocrine system (Herz 2009). For a pharmacological mechanisms to take place the effective compound must enter the bloodstream via lung or nasal mucosa (Herz 2009). Another hypothesis with support is the psychological driven effects aromas may have on stress reduction. For this mechanism to take place, emotional learning, expectations or beliefs, as well as conscious perception are what stimulate stress reduction (Herz 2009). A recent study using lavender to assess performance and stress response found both mechanisms were involved (Chamine and Oken 2016). Mechanistic

20 clarity is critical for determining how best to apply research findings when attempting to reduce stress.

Chiral odorant molecules support the pharmacological hypothesis

Chirality of odorant molecules has been studied and offers insight into potential mechanisms of aromatherapy (Heuberger et al. 2001; Heuberger et al. 2004; Kuroda et al.

2005). Chiral molecules exist when the composition of two molecules is the same, but the arrangement of one molecule is non-superimposable on the other. Previously discussed, it was found that only (R)-(-)-linalool and not (S)-(+)-linalool reduced HR and affected mood (Kuroda et al. 2005) and that during transdermal application (R)-(-) linalool decreased BP, with a moderate decrease in skin temperature when compared to the control group receiving a placebo. Interestingly, here subjective effects on well- being were no different between groups (Heuberger et al. 2004). Other studies have shown that chiral molecules may exhibit different effects, further supporting molecular specificity regarding stress reduction mechanism. Chiral molecules of limonene and carvone were studied by administering the aromas through a breathing mask

(Heuberger et al. 2001). Authors aimed to understand if chirality of a molecule had an influence on the way this molecule affected autonomic parameters and subjective mood states. The study found (+)-limonene increased systolic BP, subjective alertness and restlessness whereas (-)-limonene increased systolic BP without effects on psychological reports. (-)-carvone increased pulse rate, diastolic BP and subjective restlessness, while

(+)-carvone increased systolic and diastolic BP, suggesting chirality influenced the differentiated effects associated with inhaled odorants. 21

This is intriguing as it applies to the mechanistic knowledge of aromatherapy. To understand how specific molecules affect humans will potentially enable us to better characterize aroma stress reduction. Enhancing our understanding would also help us define more compounds and their effective mechanisms, especially if similarly shaped molecules elicit alike stress effects. A more complete understanding of the molecular effects would position us to understand optimal ways to apply them and actually utilize their characteristics to help people deal with stress.

Psychological theories: the role of expectation

Another study looked at effects of expectation when using lavender oil after inducing stress through an arousing cognitive task, the Wechsler Adult Intelligence Scale

(Wechsler 1997; Howard and Hughes 2008). Ninety-six healthy undergraduate women were assessed during the task after a priming procedure to manipulate expectancies about the aroma’s likely impact on the ability to relax. Aromas were administered via glass jars containing essential oil saturated cotton wool. Galvanic skin response (GSR) was recorded concurrently with self reports of relaxation using the State Train Anxiety

Inventory (STAI). This study found effects of expectancy associated with relaxation through GSR measures. If subjects were told about lavender’s relaxation effects, they felt them more and when expecting stimulation they relaxed less, but these effects were not indicated in the self-reports, again showing perhaps why multiple types of measures are useful to better understand stress reduction. Different measures may reflect different

22 mechanisms of essential oils in response to particular stress models, since it may not be that all odorants elicit the same mechanism of reducing stress of a certain kind.

Another study looked at psychological priming as an aspect of aromatherapy effectiveness in stress reduction. The authors studied expectations and how they affected

HR, GSC and self-reports of stress (Campenni et al. 2004). Lavender essential oil as well as neroli essential oil (bitter orange) was administered using an essential oil diffuser.

Researchers also included a placebo condition. Authors found that when suggesting either of the aromas or the placebo to be stimulating, mean HR was elevated and vice versa when suggesting relaxation, demonstrating expectation’s potentially large role in aromatherapy stress reduction effectiveness (Campenni et al. 2004). Mood effects were also seen, with increased levels of perceived pleasantness, among lessened tension, depression, fatigue and confusion (Campenni et al. 2004).

Several studies found that expectation played a role in relaxation in the presence of lavender (Diego et al. 1998; Campenni et al. 2004; Howard and Hughes 2008) and other odorants like neroli (Campenni et al. 2004) and rosemary (Diego et al. 1998) all reduced stress objectively as well as subjectively when expecting these results. Also Masaoka et al. (2013) found that subjects reported they felt less pain when it was suggested to them in both lavender and no-aroma conditions. Additionally, pain reports were significantly lower in the presence of lavender than in the no-aroma condition (Masaoka et al. 2013).

23

Variables of aromatherapy studies

Odorant delivery route

Another variable when studying olfaction is the dual nature of volatile pathways to the olfactory epithelium. Before processing volatile molecules, odorants reach the olfactory epithelium in one of two routes: retronasal or orthonasal. Retronasal delivery occurs when odorants reach the olfactory epithelium through the mouth (figure 1).

Studies have used the ascent of odorants through the posterior nares of the nasopharynx as a defining feature of retronasal olfaction (DeWeese and Saunders 1968; Roberts and Acree 1995; Voirol and Daget 1986), where odor identification was more likely for odorants when subjects exhaled solely through their nose after oral cavity inhalation

(Chen and Halpern 2008). Additionally, because retronasal delivery occurs through the oral cavity, it is associated with flavor. The opposing route, termed orthonasal, occurs when aromas are delivered through the nostrils, and is more associated with the external environment. Although odorants ultimately end up in the same place, sensation and perception have reportedly differed (Rozin 1982; Pierce and Halpern 1996; Espinosa-

Diaz 2004; Small et al. 2005; Hummel et al. 2006; Pickering et al. 2007; Hummel and Heilmann 2008; Bender et al. 2009; Welge-Lüssen et al. 2009; Lee and Halpern

2013) although other studies found similar intensity perceptions when odorants were delivered orthonasally and retronasally in the air phase, (Heilmann and Hummel 2004), as well as indirectly as foods or beverages (Burdach and Doty 1987; Sakai et al. 2001).

24

Figure 1. Orthonasal and Retronasal Olfaction (Liu 2013) is depicted in this figure.

Orthonasal olfaction occurs as a result of odor molecules traveling via the external nares and being delivered to the olfactory mucosa to be further processed by olfactory receptors, depicted by the black dots. Retronasal olfaction is a result of odor molecules traveling through the mouth, to the back of the throat and ascending towards the nostrils.

Retronasal odors volatiles culminate in the same olfactory mucosa location as orthonasal volatiles for further processing.

Specifically, retronasal and orthonasal aroma differently affected hedonic responses (Small et al. 2005) and behaviors, such as enjoying a food’s aroma but disliking the flavor (Rozin 1982), a phenomenon that has led to the Duality of Smell hypothesis (Rozin 1982). This hypothesis suggests environmental smells elicit different sensations than the same aroma presented in the mouth. Additionally, retronasal detection thresholds are higher than orthonasal detection thresholds (Cometto-Muñiz 1981; Garcia-

Medina 1981; Pierce and Halpern 1996; Hummel et al. 2006; Visschers et al. 2006).

Orthonasal and retronasal stimuli were further distinguished based upon 25 sensitivity at different concentrations, with orthonasal more readily identified at low concentrations, but at higher concentrations stimuli were identified equally well (Halpern

2008). Additionally, Small et al. (2005) found neural responses to odors were influenced by route of administration.

Exposure time

Another facet of aroma inhalation that could have notable effects on aromatherapy mechanisms and effectiveness is inhalation exposure time. Studies have varied greatly, creating yet another variable that makes comparing findings difficult. For example, studies have used 1 second to 5 minutes of exposure and found effects (Morris et al. 1995; Diego et al. 1998; Bensafi et al. 2002; Kim et al. 2011; Masaoka et al. 2013).

Other studies have found effects after roughly 10 minutes of inhalation (Burnett et al.

2004; Norrish and Dwyer 2004; Howard and Hughes 2008; Toda and Morimoto 2008;

Chang and Shen 2011; Toda and Morimoto 2011). Further, longer exposures from about

15 - 30 minutes have also been applied with notable results (Lehrner et al. 2000;

Heuberger et al. 2001; Motomura et al. 2001; Raudenbush et al. 2002; Campenni et al.

2004; Rho and others 2006; Hoya et al. 2008; Peng et al. 2009; MPham and

Siripornpanich 2012; Liu et al. 2013). Additionally more extensive exposure of about 45 minutes up to about 2 hours have been studied, especially regarding test taking anxiety

(Kutlu et al. 2008; McCaffrey et al. 2009) or during simulated driving conditions

(Raudenbush et al. 2009). Long exposures (45 to 60 minutes) have also been used during massage oil treatment (Takeda et al. 2008), and during physiological and psychological 26 measurements of healthy individuals (Kuroda et al. 2005). Longer inhalation phases assessed in this review included exposure during nurses’ work shifts of about 9 hours

(Pemberton et al. 2008). The longest exposure times were found when assessing infants pain response after overnight exposures from anywhere between 6-23 hours of exposure

(Goubet et al. 2003; Rattaz et al. 2005; Goubet et al. 2007). The studies using different exposure times all show stress reduction in one or multiple measures. However, the drastic differences in time among several other variables make definitive comparisons regarding odorant effects on stress difficult.

Adaptation

One reason exposure time is important from a methodological perspective is that perceptions change after extensive time exposed to an odorant. Adaptation, defined as the diminution in sensitivity to a stimulus following prolonged and constant exposure

(Dalton 2000), occurs among all human sensory systems (Koshland et al. 1982). It is thought that adaptation provides adaptive value by limiting over-stimulation (Ferguson and Caron 1998), and filtering insignificant sensory cues to highlight those distinct from the background (Kadohisa and Wilson 2006). Adaption to an odorant stimulus has been proposed to occur at the peripheral level following the phosphorylation of g-protein coupled receptors in olfactory receptor cells (Ferguson and Caron 1998; Mashukova et al.

2006) and at the cortical level through homosynaptic depression of afferents to the pirform cortex (Kadohisa and Wilson 2006). Olfactory adaptation has been reported in several studies (Cain 1974; Berglund et al. 1986; Dalton and Wysocki 1996; Best and

Wilson 2004), and manifests as an increase of threshold odorant concentration (Berglund 27 et al. 1986). Olfactory adaptation appears to be concentration dependent (Stone et al.

1972; Wuttke and Tompkins 2000) as higher concentrations make for more efficient adaptation (Stone et al. 1972; Wuttke and Tompkins 2000; Jacob et al. 2003).

Results suggest odorant concentration reaching the sensory system is critical, but does not address if stimulus perception is necessary for adaptation. One study tested subthreshold odor levels and found significant decreases in odorant sensitivity

(Keith and Smith 2012). These differences were smaller than those found during supra- threshold odorant administration which are supported by Weber’s Law, but nonetheless suggest that without odor perception adaptation still occurs. Weber’s law states that the size of a just noticeable difference is a constant proportion of the original stimulus perceptual report (Fechner 1860). However, the change in threshold was not significantly proportional to subthreshold adapting stimuli concentrations, although significant threshold changes still occurred.

Multiple odorants have been used in olfactory adaptation studies. Linalool adaptation has been found to occur when delivered via consistent odor pulses (Jacob et al.

2003). Vanillin has also been seen to facilitate adaptation (Lawless 1987; Savic et al.

2002; Goyert et al. 2007; Lee and Halpern 2013). These studies, however, have failed to assess separate orthonasal and retronasal administration over continuous exposure thus creating an absence of knowledge regarding how odorant pathway may contribute to different adaptation and/or cross-adaptation patterns. Vanillin is often experienced with food, whereas linalool has a more floral aroma character associated with the external environment. Previous studies have also compared food and non-food aromas

28 delivered orthonasally and retronasally, finding that the odor category affected salivary flow rate (Bender et al. 2009), fMRI neural recordings (Small et al. 2005) and event related potentials (Hummel et al. 2006; Hummel and Heilmann 2008). These qualities, however, were not assessed over extended exposure periods.

Retronasal odor perception has not been widely investigated, but one study assessed repeated presentation of a food odor through separate odorant routes, resulting in declination of salivary flow (Bender et al. 2009). After switching odor presentation route (i.e., orthonasal to retronasal or vice versa) a recovery of salivary flow rate resulted suggesting no cross-adaptation between odorant delivery routes (Bender et al. 2009). This study found this result only with food related odors (e.g. chocolate, pineapple), and not when a floral aroma was the stimulus.

The delivery of stress reducing compounds over time is an interesting phenomenon. Because there is evidence for psychological methods of aromatherapy, a noticeable aroma dwindling over time may influence feelings of stress reduction effectiveness. It is important to understand if adaptation is taking place in either orthonasal or retronasal routes of odorant delivery and to control for it. Potential aroma route differences may also provide insight into how orthonasal and retronasal sensations are processed.

29

Compounds of Interest

Lavender

A more obvious variable associated with studying aromatherapy are the countless compounds to be studied. One of the most frequently used aromas for stress reduction is lavender. However, studies attributing stress reduction to lavender have administered this odorant in different forms, the most common being diffused lavender essential oil.

Lavender oil contains a blend of several compounds, the main ones being linalool and linalyl acetate (Kuroda et al. 2004; MPham and Siripornpanich 2012). Lavender essential oil has been studied extensively, perhaps due to anecdotal support and persistent historical use. Therefore, studies most often use lavender essential oil as opposed to its individual compounds like linalool (Diego et al. 1998; Motomura et al. 2001; Burnett et al. 2004; Campenni et al. 2004; Kuriyama et al. 2005; Kuroda et al. 2005; Howard and

Hughes 2008; Hoya et al. 2008; Takeda et al. 2008; Toda and Morimoto 2008; Braden

2009; McCaffrey et al. 2009; Toda and Morimoto 2011; MPham and Siripornpanish

2012; Kim et al. 2015; Bakhsha et al. 2016; Venkataramana et al. 2016).

While many studies use essential oil, studies have differed based on types of stress studied. For example one study determined effects of ambient inhalation of lavender incense on test anxiety among students (Kutlu et al. 2008). Researchers found a mean anxiety score significantly lower than that of the control group resulting in a perceived decrease in anxiety during test taking. Another study reporting perceived anxiety reduction through visual analog scales (VAS) in response to a taped cotton ball

30 carrying lavender essential oil on the gowns of pre-operative hospital patients. During transfer to the operation room, the group exposed to the aroma claimed significantly lessened feelings of anxiety, suggesting aromatherapy as an effective pre-operative stress reliever (Braden 2009). These studies show that stress inherent to a specific group was alleviated without the use of stress induction.

Another important aspect to stress reduction is the way it is measured. The above studies touted effects of stress and anxiety reduction via subjective reports alone. While subjective reports are telling of stressful feelings, objective measures, including biological markers such as hormones or proteins related to stress may also indicate stress level changes. One study chose to quantify cortisol and chromagin A (CgA), a protein present in saliva indicative of sympathetic activity (Obayashi et al. 2008) with respect to exposure to lavender and peppermint airborne essential oil (Toda and Morimoto

2011). After exposure to peppermint, cortisol was significantly lower, suggesting less stress, while CgA was higher, which has been found when experiencing feelings of leisure and contentment (Toda and Morimoto 2011). While peppermint results suggested less stress, lavender did not have these effects on subjects, one of the few studies indicating peppermint as relaxing and lavender having no effect. This is questionable mostly due to several other studies showing peppermint’s stimulating qualities (Norrish and Dwyer 2004) as well as its effect on reducing sleepiness (Raudenbush et al. 2009).

However, stimulation and reduction in sleepiness may not be mutually exclusive with stress and anxiety, which is one potential reason these findings occurred. The same researchers studied lavender previously without comparing to a peppermint condition,

31 again measuring cortisol and CgA, with the addition of arithmetic tasks used to induce stress (Toda and Morimoto 2008). Using the same aroma exposure technique, this study found elevated CgA levels in the lavender group, but cortisol was not significantly different (Toda and Morimoto 2008). CgA levels indicated lavender exposure as a potential stress reducer during an arithmetic task (Toda and Morimoto 2008). Different findings even within the same laboratory are potentially due to methodology differences, further showing aromatherapy as complex and thus difficult to characterize.

Differences may be due to methodology, but also may be due to researchers’ lack of adequate stress measures. Weaknesses of the above studies due to methodology may be lessened if more measures of stress are taken. For example, including objective biomarkers in addition to subjective reports of stress may enable a larger perspective of how stress reduction due to lavender essential oil exposure is defined. Perceptions of stress and changes in physiology may not directly correlate, but an effect in either dimension is important. If patients feel and report less stress, then they perceive themselves as ‘better’. Alternatively if physiology causes nervous system changes in line with stress reduction, this may also be beneficial when biological homeostasis is desired.

Due to the different ways stress reduction manifests, measuring both aspects strengthens a study, possibly making results more illuminating.

Another study compared lavender and rosemary essential oils’ effects on reducing anxiety associated with test taking in nursing graduate students (McCaffrey et al. 2009).

This study included both subjective and objective measures. BP and radial pulse were recorded, as well as self-reports on the Test Anxiety Inventory. Aromas were

32 administered via inhalers containing a piece of cotton that was saturated with essential oil. Debriefing meetings regarding aroma effects were also held with subjects, another form of subjective reports. Significant changes in Test Anxiety Inventory scores were reported after both aroma conditions, suggesting lessened anxiety from both rosemary and lavender essential oil. BP did not differ significantly for any condition, but radial pulse significantly lowered in both aroma conditions. Authors make sense of this effect by claiming pulse reacts faster to stimuli, whereas BP reacts slower to immediate changes, suggesting a reason for the seemingly inconsistent physiological effects.

Debriefing conversations from subjects included thoughts that lavender was too relaxing, making concentration difficult, but that rosemary helped them focus (McCaffrey et al.

2009). These reports are potentially confounded due to group discussions, since humans are affected when discussing opinions in the presence of others (Wilder 1978).

There are several more studies that have measured biological stress markers in addition to self-reports of mood, feelings and emotions (Diego et al. 1998; Motomura et al. 2001; Burnett et al. 2004; Campenni et al. 2004; Kuroda et al. 2005; Howard and

Hughes 2008; Hoya et al. 2008; McCaffrey et al. 2009; Kim et al. 2011; MPham and

Siripornpanich 2012; Masaoka et al. 2013). Of these studies, many contained autonomic measures HR, HRV, GSC, BP, respiration rate (RR) and electroencephalography (EEG), which measures electrical activity in the brain. These measures are many times recorded as a potential correlate to mood responses after inhalation of lavender oil. One study used these measures during lavender oil inhalation via respiratory masks. Researchers found decreased BP, HR and GSC, while mood ratings reported subjects feeling more

33 active, refreshed and relaxed (MPham and Siripornpanich 2012). EEG recordings found increased alpha and theta activities, which have been found to occur in states of relaxation (Lagopoulos et al. 2009). Another study developed an odor delivery system that volatilized odorants near the subject’s nostrils. The study administered jasmine tea, in which linalool is naturally present, resulting in significant decreases in HR (Kuroda et al. 2005). The Profile of Mood States was used to measure mood in this study, which historically was developed to assess affective states in clinical and academic environments (King and Meiselman 2009). Results indicated negative mood scores were reduced, and positive ones were increased, albeit insignificantly. The study also found that different enantiomers of linalool produced different stress reduction effects. Only

(R)-(-)-linalool reduced HR and affected mood (Kuroda et al. 2005). These effects are interesting especially when attempting to understand the active components of essential oils of lavender. However the study fails to test effects following acute stress induction, since the subjects were healthy individuals inhaling volatiles at baseline, rather than following an acute stressor or being a part of chronically stressed population. This is another reason results are difficult to compare to studies that have induced stress before measuring changes in stress. Measuring effects in healthy individuals begs the question of how stress might be mitigated when objectively stressed subjects are tested.

Another type of stress measured included that of pre-operative patients. One study using lavender essential oil examined changes in anxiety of gastroscopy patients by the

Face Scale score, in addition to blood pressure changes (Hoya et al. 2008). The Face

Scale score was previously used to measure pain levels in the elderly (Herr et al. 1998),

34 but was used here to measure anxiety levels, potentially introducing a confound, since this was not the intended use of the scale. Lavender was administered in the waiting room via an essential oil burner. Perceived anxiety grew larger from arrival up until the procedure for the control group, but did not significantly increase in the aroma group. BP was significantly lower in the experimental group, suggesting these interventions could be used to reduce pre-gastroscopy anxiety. However, results may not necessarily be attributed to volatiles alone as a video program was also employed to promote pre- gastroscopy relaxation. Also, the authors note this effect as non-pharmacological, which is difficult to determine since it is unknown what mechanism created changes in BP and psychological feelings (Hoya et al. 2008).

Another study measured EEG, and levels of alertness and mood changes using lavender and rosemary essential oil (Diego et al. 1998). Aromas were administered with a vial containing diluted essential oil held near the nose of subjects for 3 minutes. The lavender group showed increased drowsiness due to enhanced beta power of the EEG measure. This condition also lessened depressed moods and enhanced relaxation. They also performed math computations faster and more accurately following the aroma administration. Rosemary showed decreased frontal alpha and beta power, suggesting increased alertness, experienced lower state anxiety scores and reported feeling more relaxed and alert. These subjects were also faster but less accurate when completing math computations, suggesting that essential oils elicit specific effects that are not always exclusively relaxing or sedating, but may also affect alertness and cognitive efficiency.

There may be a differentiation between reduced feelings of stress due to relaxation and

35 feelings of sedation. It might be argued as well that alertness could be stress relieving since becoming more alert could affect the ability of humans to be more efficient, lessening their workload. This, perhaps, helps explain why essential oils can be reported to enhance feelings of alertness and relaxation simultaneously. The multifaceted nature of stress reduction demonstrated by this study’s findings provides insight as to why aromachology effects may be inconsistent, even when effects on stress are repeatedly found. It also underpins the need for several measures of subjective moods and feelings, accompanied by physiological measures that record changes in different study conditions. With multiple measures, a more complete story of aroma effects is attained.

Transdermal Application

Lavender oil has been administered in several ways: coated facial masks (Kim et al. 2011; MPham and Siripornpanich 2012), diffusers (Motomura et al. 2001; Campenni et al. 2004), incense (Kutlu et al. 2008), oil application to subjects’ wrists (Burnett et al.

2004), saturated cotton balls attached to hospital gowns (Braden 2009), cotton dental swabs (Diego et al. 1998), wool (Howard and Hughes 2008), filter paper placed near the nostrils (Toda and Morimoto 2008; Toda and Morimoto 2011), via inhalers (McCaffrey et al. 2009), bubbled through an aroma device near the nostrils (Kuroda et al. 2005), essential oil burners (Hoya et al. 2008), and via litmus strips placed in front of an inspiratory valve connected to a transducer measuring RR (Masaoka et al. 2013).

However, lavender has also been administered transdermally, that is through the skin. This method is often employed for stress reduction effects, but is discussed separately here since it is theoretically not inhalation. However, depending on study 36 controls, inhalation may or may not contribute to transdermal lavender application effects and thus warrants attention. Some studies claim the use of lavender as a massage oil exhibits effects when lavender components transfer through the skin and enter the bloodstream (Jager et al. 1992). Like the previously discussed studies, massage studies also differ by whether they used objective or subjective measures to assess stress reduction.

One study tested several aromas, orange, lavender, marjoram and peppermint, via massage therapy (Takeda et al. 2008). To induce stress, the Advanced Trail Making Test was used in a computerized version that has subjects attend to visual searching designed to test selective attention and spatial working memory (Kajimoto et al. 2007) that may induce mental fatigue (Mizuno and Watanabe 2008). The STAI, the Face scale and VAS were used to assess anxiety, feelings and mood. Cortisol and Immunoglobulin A (IgA), an antibody that’s production correlates with stress, were also measured as autonomic stress indicators. During the massage, the blend of oils consisted of all oils at once. This condition was then compared to a carrier oil (macadamia nut oil) and a rest condition with neither the carrier oil nor massage with the essential oil blend. The study found that subjects massaged with the essential oil blend experienced less fatigue, lower STAI scores, and increased positive feelings and comfort. However, objective measures of stress were unaffected, suggesting massage therapy changed perceived levels of anxiety and mental fatigue but without affecting physiological differences, specifically salivary cortisol and IgA, significantly (Takeda et al. 2008).

37

Another massage oil study looked at several biomarkers such as a collection of peripheral blood cell counts, lymphocytes, hematocrit (volume of red blood cells in blood), and immune related parameters such as IgA in response to massage therapy

(Kuriyama et al. 2005). Psychological evaluations were also collected which included

STAI and a self-rating depression scale. Authors compared carrier oil massage with a blended aromatherapy massage (almond, lavender, cypress and marjoram) following a subtraction task. The control and massage oil conditions were separated by at least two weeks. The authors found no significant differences between the carrier oil massage and the massage used with blended essential oils. The aromatherapy group did find a significant change in partial immune response measures, suggesting it may be beneficial to the immune system. Both the aromatherapy and control massage groups reported reduced STAI scores, however, feelings of depression did not differ before or after either condition (Kuriyama et al. 2005).

Another group analyzed massage therapy using clary sage and lavender oil

(Pemberton et al. 2008) in a pilot study. They used topical application to study potential effects on work-related stress of nurses in an intensive care unit setting. The researchers found decreased perceptions of stress, suggesting a psychological benefit. Another study also found a psychological effect, measuring anxiety and self-esteem in elderly Korean women after using blended 4:3:2:1 ratio of lavender, chamomile, rosemary and lemon for 20 minute sessions 3 times a week for two 3-week periods (Rho et al. 2006). There was a one-week break between. Significant differences in anxiety through STAI and self- esteem (Rosenberg self-esteem scale) were found, but without corresponding BP or pulse

38 rate changes between the control and massage therapy groups, again suggesting a psychological effect without changes to physiology (Rho et al. 2006).

A potential confound of the above studies is that authors attribute effects entirely to massage oil application but if the inhalation of the aroma during application is not controlled, results may not be due solely to percutaneous absorption. Another transdermal application study of aromatherapy used essential oil of lavender to study relaxing effects of just one of the main constituents of the oil, (R)- (-) linalool specifically. (R) – (-

) linalool was applied to the skin of subjects with simultaneous prevention of inhalation using a breathing mask. Researchers measured physiological parameters including blood oxygen saturation, RR, eye-blink rate, pulse, GSC, skin temperature, surface electromyogram and BP, in addition to subjective assessments of well being (Heuberger et al. 2004). In the condition using (R)-(-) Linalool BP and skin temperature decreased when compared to the control group. However, subjective reports of well-being did not differ between groups (Heuberger et al. 2004).

The notable aspect of the above study (Heuberger et al. 2004) was the control and prevention of inhalation. Previous transdermal application studies have attributed effects to massage of essential oils into the skin; however, inhalation could also very well be occurring throughout many of the transdermal application procedures. By controlling the potential for inhalation to confound results, authors position themselves better to attribute findings to transdermal application or inhalation alone. Additionally this study looks at a specific component of lavender essential oil, helping to clarify what component(s) of

39 lavender oil may potentially be affecting subjective feelings or objective physiological changes regarding stress.

Vanillin

Vanillin has less of a history in aromatherapy but is widely used in industry as a major component of vanilla flavoring. Vanilla legally contains no less than 35% alcohol, with at least 13.35 oz. vanilla beans per fold per gallon of extract, vanillin being the main flavor component (Furia 1973).It has recently gained interest from a clinical standpoint, producing anti-inflammatory (Khuda-Bukhsh et al. 2004) and cancer cell apoptosis effects (Ho et al. 2009). Previous studies, however have not assessed vanillin's ability to relieve acute stress. More relevant studies associated with stress in the form of pain found vanillin reduced distress during heel prick tests in infants, however studies further found familiarity to odorants was likely why vanillin was an effective analgesic (Goubet et al.

2003; Rattaz et al. 2005).

Inducing cognitive stress

Another variable needing control is stress induction. A primary difficulty in resolving odorant effects on stress is that many studies employ vastly different stress paradigms or none at all, making comparisons difficult. An important indication of odor stress relief induction is to effectively use a controlled stressor. With effective and controlled stress induction, amelioration can be more easily assessed. Studies often study

40 stress without inducing it. This is another reason potentially that studies are inconclusive. Some studies have induced stress in ways such as demanding excess cognitive or mental work. One study utilizing both subjective and objective measures of stress reduction in the presence of lavender, rosemary or control conditions were tested when completing a crossword puzzle for 10 minutes (Burnett et al. 2004). The Profile of

Mood States and scent pleasantness ratings in conjunction with temperature and HR were measured. Subjects in an aroma condition received 3 drops of diluted rosemary or lavender essential oil on their wrist while working on the puzzle. Authors found that when perception of scent pleasantness is controlled, the aroma can moderate mood states in this “anxiety-provoking task” (Burnett et al. 2004). Authors did not, however, find a difference in physiological states of HR or temperature based on scent condition, but mood ratings differed in aroma conditions and within rosemary and lavender administration. The rosemary condition induced higher reports of tension- anxiety and confusion-bewilderment relative to lavender and control conditions.

Lavender and control conditions elicited higher average vigor-activity ratings relative to the rosemary condition, whereas both rosemary and lavender conditions initiated lower average ratings of fatigue-inertia (Burnett et al. 2004).

In response to acute stress induction, lavender essential oil was studied. During 20 minutes in an empty room, subjects were to sit and wait. Authors considered the waiting experience as stressful. Essential oil of lavender was diffused during one group’s waiting time. BP and HR did not differ significantly, but self-reports found lavender odor was associated with reduced mental stress and increased arousal (Motomura et al. 2001).

41

Objective

There are three main objectives of this study.

1. The first objective is to characterize and better understand olfactory adaptation as it applies to different routes of delivery of linalool and vanillin compounds.

2. The second objective is to develop and validate an open-source, multitasking stressor that has increased ecological validity.

3. Upon validating a laboratory-induced stressor, odor compounds, linalool and vanillin are assessed for stress reduction properties. Additionally, aroma delivery pathway is assessed as to whether orthonasal or retronasal administration is more or less effective.

42

Chapter 3: Olfactory adaptation is dependent on route of delivery

Introduction

Odorants are delivered to the olfactory epithelium via retronasal or orthonasal routes. Retronasal delivery occurs when odorants reach the olfactory epithelium through the mouth. Studies have used the ascent of odorants through the posterior nares of the nasopharynx as a defining feature of retronasal olfaction (DeWeese and Saunders 1968;

Voirol and Daget 1986; Roberts and Acree 1995), where odor identification was more likely for odorants when subjects exhaled solely through their nose after oral cavity inhalation (Chen and Halpern 2008). Additionally, because retronasal delivery occurs through the oral cavity, associated sensations are referred to the mouth and contribute to our perceptions of flavor. Orthonasal olfaction occurs when odorants are inhaled through the nostrils and these stimuli are typically associated with the external environment.

Although both routes deliver the same odorants to the same receptor fields in the olfactory epithelium, sensation and perception have reportedly differed (Rozin 1982;

Pierce and Halpern 1996; Espinosa-Diaz 2004; Small et al. 2005; Hummel et al. 2006;

Pickering et al. 2007; Hummel and Heilmann 2008; Bender et al. 2009; Welge-Lüssen et al. 2009; Lee and Halpern 2013). Indeed, when delivered orthonasally and retronasally, the same odorants can evoke different physiological responses (Heilmann and Hummel

2004; Small et al. et al. 2005; Hummel et al. 2006), thresholds (Pierce and Halpern 1996;

43

Hummel et al. 2006; Visschers et al. 2006), perceptual qualities (Pierce and Halpern

1996; Hummel et al. 2006), hedonic responses (Small et al. 2005), and behaviors (Rozin

1982; Burdach and Doty 1987; Heilmann and Hummel 2004; although see Sakai et al.

2001 for contradictory results). These differences have led to the Duality of Smell hypothesis (Rozin 1982) that suggests olfaction operates as a dual system in which orthonasal and retronasal stimuli are processed differently although the specific mechanism enabling this distinction is unknown.

Adaptation may provide insight into how orthonasal and retronasal sensations are processed. Adaptation, defined as the diminution in sensitivity to a stimulus following prolonged and constant exposure (Dalton 2000), occurs among all human sensory systems (Koshland et al. 1982). It is thought that adaptation provides adaptive value by limiting over-stimulation (Ferguson and Caron 1998), and filtering insignificant sensory cues to highlight those distinct from the background (Kadohisa and Wilson 2006).

Adaption to an odorant stimulus has been proposed to occur at the peripheral level once phosphorylation of g-protein coupled receptors in olfactory receptor cells occurs

(Ferguson and Caron 1998; Mashukova et al. 2006). At the cortical level, homosynaptic depression of afferents to the piriform cortex has been reported (Kadohisa and Wilson

2006). Olfactory adaptation has been reported in multiple studies (e.g. Cain 1974;

Berglund et al. 1986; Dalton and Wysocki 1996; Best and Wilson 2004), and manifests as an increase of threshold odorant concentration (Berglund et al. 1986). Olfactory adaptation has shown concentration dependence (Stone et al. 1972; Wuttke and

Tompkins 2000) with higher concentrations decreasing the time needed to adapt (Stone et

44 al. 1972; Wuttke and Tompkins 2000; Jacob et al. 2003). Historically, adaptation to orthonasal stimuli is well characterized, whereas retronasal adaptation is not. Similarly, cross-adaptation, whereby extended exposure to a stimulus delivered via one pathway

(i.e. ortho- or retronasal) induces adaptation of the same stimulus when delivered via the alternative pathway (retro- or orthonasal), has not been investigated. One study, however, found salivary flow rates to decline with prolonged delivery of retronasal food (but not non-food) odors that rebounded when the same odor was subsequently presented orthonasally (Bender et al. 2009). The results are consistent with a lack of cross- adaptation between odorant pathways.

The objective of the present study was to assess olfactory adaptation following extended exposure to odorant stimuli delivered orthonasally and retronasally. This study assessed the potential for cross-adaptation by evaluating orthonasal odorant intensity after extended exposure to the same stimulus delivered retronasally, and vice versa.

Materials and Methods

Subjects. Sixty subjects were recruited via the Ohio State University’s sensory testing database and enrolled in the study under informed consent (approved IRB protocol 2014B0597). Given the expected effect size, power analysis indicated a need for

30 subjects per condition. Thus, thirty panelists (19 female, 11 male) performed this procedure for vanillin and thirty (13 female, 17 male) panelists performed this procedure for linalool. Recruits were tested if they were over 18 years of age, reported no known smell deficits, no history of heart disease, pulmonary issues (e.g. asthma, fragrance allergies), or dry mouth (xerostomia). They could not smoke and not currently be using

45 any psychoactive (e.g. antidepressants or antipsychotics), or cardiopulmonary (e.g. hypertensive) drugs, statins, beta-blockers, neurologic medications, motion sickness, or smoking/alcohol cessation medications. Participants were also excluded if they were undergoing cancer therapy (chemotherapy, radiation therapy). They were told in the recruitment email they would receive $20 for approximately forty minutes of testing and to refrain from wearing perfume or cologne. The recruitment email indicated they would be evaluating aromas delivered to their nose or oral cavity by a customized aroma delivery device and that they would be asked to assess various aspects of the aromas.

Olfactory stimulation. Linalool (12% w/w; Sigma-Aldrich, St. Louis, MO) or vanillin (25% w/w; Sigma-Aldrich, St. Louis, MO) were dissolved in miglyol (Nature’s

Oil, Streetsboro, OH) or propylene glycol (Fischer Scientific, Fair Lawn, NJ), respectively and delivered in the air phase at 8 liters per min (retronasal) or 6 liters per min (orthonasal) with deodorized, humidified breathing air through a customized aroma delivery device. Linalool and vanillin concentrations were selected to be approximately equally intense. For retronasal delivery, subjects were fitted with a silicone mouthpiece

(Figure 2A) attached to a glass manifold that was connected via silicone tubing to the aroma source. Subjects were instructed to inhale through their mouth and exhale through their nose. Accumulation of saliva was common during retronasal evaluations. To minimize this, saliva was suctioned out using a dental suction device (Henry Schein,

Melville, NY) inserted around the mouthpiece and placed bilaterally between the cheek and gum. For orthonasal evaluations, subjects were fitted with a glass conical nose piece

(Figure 2B) and instructed to inhale through their nose and exhale through their mouth.

46

For cross-adaptation studies, subjects were fitted with both retro- and orthonasal delivery devices simultaneously.

Figure 2. Retronasal (A) and orthonasal (B) aroma delivery systems. Note, the retronasal system was disassembled to show component parts (mouthpiece, glass manifold and connecting tube). Each system was connected to an aroma source (12% or 1.5% linalool and 25% or 6.25% vanillin) that delivered the odorant at 8 LPM (retronasal) or 6 LPM

(orthonasal) using deodorized, humidified breathing air.

Panelists were told the goal of the study was to understand intensity perception of certain odors better and that researchers needed them to rate perceived intensity at different time points and in different smelling conditions. They were then prompted to practice the breathing patterns without being fitted to the device to familiarize themselves with the study conditions. After practicing breathing patterns, panelists smelled a reference solution of the same concentration they would be assessing of either vanillin or linalool to provide familiarity with the odorant character and intensity. Subjects were

47 then fitted with the aroma delivery device and rated perceived odorant intensity during

10-min of continuous delivery at 0-, 5- and 10-min using the generalized labeled magnitude scale (Bartoshuk et al. 2004). Retronasal exposure occurred first, followed after a short recovery period (~5 min), by orthonasal delivery. During the extended exposure, they performed arithmetic problems as a distracter, except when prompted to provide their present perceived intensity ratings. The intensity ratings were collected on an iPad tablet device (Apple Inc., Cupertino, CA) using Compusense software (Guelph,

Ontario, CAN). For cross-adaptation evaluations, subjects were fitted with both orthonasal and retronasal delivery devices. Panelists first provided an initial (pre- adaptation) rating to the stimulus. The route of delivery was then switched after which the subject continued to inhale the odorant for 10-min. Following the 10-min exposure period, the route of odorant delivery was switched back to its original setting and subjects immediately inhaled the stimulus again to provide a cross-adaptation rating. Next, subjects inhaled clean deodorized air for a period of 2-mins to cleanse any residual effects from the persistent exposure and the odorant was delivered again to assess recovery from cross-adaptation. Whether the initial pre-adaptation stimulus was delivered orthonasally or retronasally was counterbalanced across subjects.

A second experiment was conducted to control for initial intensity differences between orthonasal and retronasal stimuli due to the differences in concentrations.

Preliminary experiments were conducted to identify orthonasal concentrations of stimuli that better matched the perceived intensity of the retronasal odorants in the previous

48 experiment. An additional 53 subjects were recruited and underwent the orthonasal adaptation protocol using 1.5% (w/w) linalool (n=26) or 6.25% (w/w) vanillin (n=27).

Statistical Analysis. For each treatment, intensity ratings were subjected to a repeated measures Analysis of Variance (ANOVA) with time and subject as independent factors. Post-hoc Tukey’s test was used to assess significant differences in time. To compare the responses of intensity matched orthonasal stimuli from the second experiment and retronasal stimuli from the first experiment, the intensity ratings were subjected to a mixed-model repeated measures ANOVA with time as a within-subjects factor and route of delivery as a between-subjects factor. Greenhouse-Geisser correction was used when the test of sphericity was violated. Post-hoc Least Significant Difference

(LSD) tests were performed to determine which time points differed between the breathing conditions. All data are reported as means ± SE.

Results

Adaptation

High-Intensity Orthonasal Delivery. When delivered orthonasally, prolonged exposure to the odorants linalool and vanillin resulted in progressively weaker intensity ratings consistent with adaptation (Figure 3A). For linalool, the aroma intensity was perceived as significantly (p=0.010) lower following the 10-min exposure period compared to the initial, pre-exposure rating (16.5 ± 1.2 vs. 23.7 ± 1.1, respectively).

Subjects perceived the linalool intensity to be weaker at 5-min post exposure (19.1 ± 1.1), although it did not reach statistical significance (p=0.169). A similar pattern of response was noted following prolonged vanillin delivery, however, compared to the pre-exposure

49 rating (26.5 ± 1.1), the perceived intensity was significantly lower at both 5-min (19.2 ±

1.2; p=0.005) and 10-min (17.7 ± 1.2; p<0.001).

Retronasal Delivery. When delivered retronasally, neither linalool nor vanillin intensities adapted following a 10-min exposure period (Figure 3B). This contrasts with what was observed following orthonasal delivery of the same aroma compounds. In fact, following 5-min continuous exposure to the retronasal odorants, the perceived intensities of both linalool (11.0 ± 1.2) and vanillin (9.4 ± 1.2) were significantly (p=0.011 and p=0.032, respectively) higher compared to the pre-adaptation intensities (linalool: 6.4 ±

1.2; vanillin: 6.4 ± 1.1). At 10-mins, the perceived vanillin intensity continued to increase

(10.7 ± 1.2; p=0.003), whereas linalool intensity plateaued or slightly decreased (8.1 ±

1.2; p=0.375).

Figure 3. Orthonasal (A) and retronasal (B) perceived intensity before, during, and after

10-mins of continuous linalool (12%; hollow circles) or vanillin (25%; black circles) delivery. Orthonasal aroma delivery resulted in both linalool (hollow circles) and vanillin

(black circles) adaptation whereas the intensity of retronasal stimuli (linalool and 50 vanillin) increased with continued exposure. For each panel, different letters above or below each circle denote significant differences in perceived linalool or vanillin intensities.

Low-Intensity Orthonasal Delivery. To determine if the lack of adaptation observed following retronasal delivery was a consequence of initial perceived intensity, a control treatment adjusted the concentrations of the linalool and vanillin stimuli so that when delivered orthonasally, they more closely matched the retronasal intensities obtained previously. When comparing the responses obtained retronasally to those obtained orthonasally with lowered odorant concentrations, a significant time-condition interaction effect was observed for both linalool (p=0.006) and vanillin (p<0.001) indicating that the pattern of responses obtained were not consistent across delivery routes. Indeed, for linalool, orthonasal delivery resulted in a slight decrease in perceived intensity over the 10-min exposure consistent with adaptation (although the decrease was not significant; p=0.179) compared to the significant (p=0.015) increase following retronasal delivery (Figure 4A). At 5-min, the perceived linalool intensity was significantly greater when delivered retronasally compared to orthonasally (p<0.001;

Figure 3A). Similarly, for vanillin, orthonasal delivery resulted in a significant (p=0.016) decrease in perceived intensity after 5-min of constant exposure that subsequently plateaued (Figure 4B). This is in contrast to retronasal delivery that showed a progressive increase in perceived intensity. For vanillin, there was also a significant (p=0.006) difference in the mean intensity observed for the initial orthonasal and retronasal rating indicating that the orthonasal vanillin was perceived as more intense even though it was 51

1/8th the concentration of the retronasal stimulus (Figure 3B). Results indicate adaptation is occurring at a lower intensity for orthonasally delivered volatiles, but not for retronasal delivery.

Figure 4. Perceived intensity of a weak linalool (1.5%) and vanillin (6.25%) odorant stimulus before, during, and after 10-mins of continuous retronasal (hollow circles) or orthonasal (black circles) delivery. Note for each odorant, orthonasal delivery elicits a response that tends to adapt over time whereas perceived intensity of the same stimulus increases when delivered retronasally. For each panel, different letters denote significant increases or decreases in perceived retronasal or orthonasal intensities. The significant time-condition interaction term for linalool (p=0.006) and vanillin (p<0.001) suggests delivery route produces significantly different intensity perceptions over time.

Cross Adaptation.

Orthonasal to Retronasal. Evidence of a retronasal stimulus adapting following

10-min of continuous orthonasal exposure was not observed. For linalool, the perceived intensities of the retronasal stimuli immediately preceding (6.4 ± 1.2) and following (4.1

± 1.2) the 10-min orthonasal stimulus did not significantly (p=0.057) differ (Figure 5A).

However, after inhaling clean air, the perceived intensity of retronasal linalool rebounded

52 to a significantly (p=0.009) higher level (7.3 ± 1.2). For vanillin, similar results were observed. The perceived intensities of the retronasal vanillin stimuli immediately preceding (6.4 ± 1.2) and following (5.3 ± 1.2) the 10-min orthonasal stimulus did not significantly (p=0.259) differ (Figure 5A).

Retronasal to Orthonasal. Evidence of an orthonasal stimulus adapting following

10-min of continuous retronasal exposure to the same odorant was not observed. Rather, instead of cross-adaptation, the orthonasal stimulus intensity tended to be higher following prolonged retronasal delivery (Figure 5B). This was true for both linalool and vanillin for which the intensity of the orthonasal stimulus following retronasal exposure

(18.9 ± 1.2 and 22.5 ± 1.1, respectively) was greater than that observed before retronasal exposure (14.9 ± 1.1 and 19.8 ± 1.2, respectively), although this difference was not significant (linalool: p=0.211 and vanillin: p=0.463). Following the inhalation of clean air, the perceived intensity of both linalool and vanillin increased further and was found to be significantly (p=0.004 and p=0.024, respectively) greater than pre-adapting levels

(linalool: 23.7 ± 1.0 and vanillin: 26.5 ± 1.1).

Figure 5. Cross-adaptation. (A) Orthonasal to retronasal cross-adaptation. Evidence of a retronasal stimulus adapting following 10-min of continuous orthonasal exposure was not 53 observed for linalool (hollow circles) or vanillin (black circles). (B) Retronasal to orthonasal cross-adaptation. Evidence of an orthonasal stimulus adapting following 10- min of continuous retronasal exposure was not observed for linalool (hollow circles) or vanillin (black circles). For each panel, different letters above or below the circles denote significant differences in perceived vanillin or linalool intensities.

Discussion

We show that olfactory adaptation is dependent upon whether the odorant is delivered orthonasally or retronasally. Whereas orthonasal stimuli adapt readily, retronasally delivered odorants do not. This is the first report that we know of specifically addressing retronasal adaptation and showing a lack thereof. In addition to the lack of retronasal adaptation, we saw no evidence of cross-adaptation whereby extended exposure to a stimulus delivered via one pathway (i.e. ortho- or retronasal) induces adaptation of the same stimulus when delivered via the alternative pathway (retro- or orthonasal). These findings were consistent across two odorants—linalool, a non-food odorant, and vanillin, a food-related odorant—and suggest this lack of adaptation may be a specific property of retronasal olfaction.

One reason contributing to the dearth of retronasal adaptation studies is the logistical complications associated with the continuous delivery of a retronasal stimulus.

In most prior studies investigating retronasal olfaction, the odorant stimulus was dissolved in water and delivered in solution where subjects are either asked to swallow the stimulus bolus (Cerf-Ducastel and Murphy 2001; Espinosa-Diaz 2004) or it was placed in a container on the back of the tongue where panelists were expected to use 54 throat contractions to pump the aroma past the velum (Pierce and Halpern 1996; Sun and

Halpern 2006; Lee and Halpern 2013) and into the nasal sinus. In either case, prolonged stimulus delivery is hindered by the complications associated with consuming large volumes of stimulus or placing an aroma bolus on the back of the tongue. We overcame these issues by delivering volatilized aroma compounds in the air phase directly into the oral cavity through a mouthpiece and training subjects to inhale through the mouth and exhale through the nose. Such a procedure enables the study of purely retronasal stimuli without the confounding issues associated with other delivery methods.

Using this methodology, we observed higher concentrations of linalool and vanillin, when delivered retronasally, elicited perceived intensity ratings that were significantly lower than when the compounds were delivered at lower concentrations orthonasally. This finding is consistent with previous reports that retronasal olfaction is less sensitive than orthonasal olfaction (Espinosa-Diaz 2004; Heilmann and Hummel

2004; Hummel et al. 2006; Furudono et al. 2013; Hummel et al. 2016). This differential sensitivity may reflect differences in air-flow patterns (Zhao et al. 2006) as well as non- uniform receptor distributions across the olfactory epithelium (Schoenfeld and Cleland,

2005). In addition, recent evidence suggests that odorant molecules can be retained in the lung leading to the possible reduction of retronasal odor concentrations and alterations to the odor mixture makeup (Verhagen 2015). In the present experimental design, subjects were instructed to inhale through their mouth and exhale through their nose for retronasal stimulus delivery. As such, odorants necessarily entered the lung prior to being exhaled through the nasal cavity which may have resulted in lower retronasal odorant

55 concentrations. Nevertheless, higher detection thresholds and decreased perceived intensities have also been observed when the same odorant concentration was delivered orthonasally or retronasally through a nasal cannula via the nares to the rostral and caudal aspects of the nasal sinus (Heilmann and Hummel 2004).

In the present experiment, we also observed adaptation to high and low (intensity matched to a retronasal stimulus) concentrations of linalool and vanillin—both pleasant aromas—when delivered orthonasally. In humans, adaptation to orthonasal aroma stimuli has been reported previously (Cain 1974; Berglund et al. 1986; Dalton and Wysocki

1996; Best and Wilson 2004) and the extent to which it occurs depends on the odorant concentration and duration of exposure (Jacob et al. 2003) as well as the stimulus relevance (Dalton 2000, Kobyashi et al. 2007) and pleasantness (Jacob et al. 2003).

Surprisingly, when even higher concentrations of linalool or vanillin were delivered retronasally, no evidence of adaptation was seen. Given the intensity matched orthonasal stimuli adapted whereas the same stimuli delivered retronasally did not suggests that perceived intensity per se is not responsible for the observed effect.

This lack of adaptation to a retronasal olfactory stimulus was unexpected. Prior studies have noted adaptation across all sensory systems, with the possible exception of pain (Sessle 2006) that exhibits sensitization, hyperalgesia and allodynia.

Mechanistically, it is unclear why adaptation following prolonged retronasal stimulation was not observed. One possibility is that adaptation to a retronasal stimulus requires a longer duration of exposure. In this experiment, subjects were exposed to the stimulus for 10 minutes. Exposure for periods longer than this may be required to observe

56 retronasal adaptation. Alternatively, these results are consistent with the Duality of Smell hypothesis (Rozin 1982) and suggest that the same stimulus is processed differently whether it is delivered orthonasally or retronasally. Differential processing of orthonasal and retronasal odors has been suggested to underpin the intensity and quality differences observed between the orthonasal and retronasal delivery of the same odorants (Heilmann and Hummel 2004; Small et al. 2005; Hummel et al. 2006; Verhagen 2015). This differential processing may also impart decreased susceptibility to adaptation. The lack of retronasal adaptation could have adaptive value in that the continued presence of a pleasant retronasal sensation could motivate continued feeding behavior.

Additional support for the Duality of Smell hypothesis comes from the lack of cross- adaptation observed in our present experiments. As retronasal stimuli did not adapt, a lack of cross-adaption to orthonasal stimuli may be expected. However, for both linalool and vanillin, orthonasal adaptation was observed. Interestingly, during the period of orthonasal adaptation, panelists were still sensitive to the same odorant when delivered retronasally. This lack of cross-adaptation suggests that orthonasal and retronasal information is processed differently and, like differential sensitivity, could result from differential air flow (Zhao et al. 2006) or non-uniform receptor distribution in the olfactory epithelium (Schoenfeld and Cleland 2005).

Acknowledgements

Research support provided by state and federal funds appropriated to The Ohio

57

State University, Ohio Agricultural Research and Development Center (Award #

2017007). This study served in partial fulfillment of the dissertation requirements for

Alex Pierce-Feldmeyer.

58

Chapter 4: Validation of an ecological laboratory induced stressor to assess amelioration

techniques

Introduction

Stress is a problem

Adults and children experience stress levels higher than what they believe to be healthy and are ill equipped to effectively reduce it (American Psychological Association

2014). Stress increases risk for cardiovascular disease (Vrijkotte et al. 2000), immune suppression and cancer (Dienstbier 1989; Fraser et al. 1999), among enhanced feelings of anxiety, worry (Hunter and Thatcher 2007) and nervousness (Cooke and Rousseau,

1984). Medications and psychotherapy are often used to treat chronic stress leaving many suffering from acute stress without appropriate treatment options.

Types of Stress

Cohen and Williamson (1991) separate types of mental stress into (1) major life events like moving, divorce and death, (2) daily life events like losing things, arguments with friends and social obligations, and (3) psychological distress like anxiety, dysphoria, and other negative affect states. Stress from daily activities likes losing things, social

59 obligations or arguments are more common but are perhaps less studied since this stress creates a perception of being less threatening. Studies have shown support, however that all three types of stress influence risk for infectious diseases, which makes studying daily sources of stress critical (Cohen and Williamson 1991).

Prior studies of stress have utilized stress paradigms to induce a controlled mental stress, however it is of a maximal level and does not reflect an acute and more common multitasking stress (Kirschbaum et al. 1993; Balodis et al. 2010). Multitasking induced stress has been less studied. Scholey et al. (2009) had subjects attend to 4 computerized tasks simultaneously to assess if chewing gum alleviated stress. The study found reduced stress, but after repeating the design Johnson et al. (2011) found no effect on stress, but instead reported increased alertness. Other studies utilized computerized tasks, but had subjects attending to tasks singularly (Mezzacappa et al. 2010; Nourbakhsh et al.

2012).

Laboratory measures

There are several components of stress that can be measured. Studies often measure a variety of moods or feelings (Diego et al. 1998; Kuroda et al. 2005; MPham and Siripornpanich 2012). Some studies have correlated lower anxiety to increased alertness, suggesting higher alertness reduces stress (Raudenbush et al. 2002;

Raudenbush et al. 2009). Pain reduction has also been used to indicate decreased stress

(Struy and others 2002; Goubet 2004; Rattaz et al. 2005; Kim and others 2011;

Masaoka et al. 2013). Kennedy et al. (2004) assessed the impact of lemon balm on stress using subjective ratings without inclusion of physiological measures (Kennedy et al. 60

2004). Physiological measures have also been used for more objective ways to record stress (Kurniawan et al. 2013). Other studies have measured physiological responses to multitasking but have not studied mitigation techniques (Hjortskov et al. 2004; Wetherell et al. 2004).

Due to the complexity of stress, a well designed study will utilize both subjective and objective measures of stress. Multiple indices have helped characterize different parameters of stress responses (Gordis et al. 2006; Balodis et al. 2010). Hjortskov et al.

(2004) found significant objective changes but insignificant subjective ratings from stress, arguing for the utility of both types of stress measures. For this study, subjective evaluation of the NASA Task Load Index (NTLX) was used for baseline, pre-stress, post- stress, and recovery assessments of mood and stress levels. The NTLX measured perceived levels of mental, physical and temporal load as well as perceived levels of effort, performance and an overall rating of stressfulness (Hart and Staveland 1988).

Another measure indicative of stress is GSC, which measures the electrical conductance of the skin, or the skin’s sweat response, indexing sympathetic nervous system arousal (Fowles et al. 1981; Peres 2012). When sweating, the moisture decreases the resistance of the electrical current through the skin attached to the electrodes, increasing conductance. Stress has been found to affect GSC (Boucsein et al. 2012).

Another measure of interest is heart rate variability (HRV). HRV is the beat-to- beat time changes in heart rate. High HRV indicates good adaptability, whereas low HRV indicates abnormal physiological condition or mental stress (Pumprla et al. 2002). HRV has been found to indicate vagal sinoatrial control which correlates with parasympathetic

61 activity (Berntson et al. 2007). Time domain methods are often used to measure HRV, which includes the root of the mean squared differences of successive heart beat intervals

(RMSSD). Another HRV index is respiratory sinus arrhythmia (RSA). RSA is indicative of vagal modulation and therefore parasympathetic vagal activity (Berntson et al. 1993;

Task Force of The European Society of Cardiology 1996). With mental and cognitive stressors, RSA is expected to decrease (Grossman and Svebak 1987; Grossman et al.

1990). Mean heart rate (mean HR) is also recorded and is the average number of times the heart beats per minute. Mean HR has been reported to elevate in times of arousal or stress (Lang et al. 1993; Connor-Smith et al. 2000).

Studies have successfully used GSC (Setz et al. 2010; Sun and others 2010;

Nourbakhsh et al. 2012) and heart rate variability as indices of stress (Hjortskov et al.

2004; Filaire et al. 2010; Sun et al. 2010). Studies have also used salivary components as indicators of stress such as α-amylase. (Bohan and Maye 2010; Filaire et al. 2010).

Significant differences using HRV were also recorded in nursing students (Kim et al.

2012), from video games (Li et al. 2009), from work stress (Vrijkotte et al. 2000) and in response to a job-demand-control (JDC) model that also elevated salivary amylase levels

(O’Donnell et al. 2015). With adequate evidence of objective stress measures, it is clear that HRV and salivary enzymes will help illuminate the body’s response to stress and potential mitigation techniques.

Objective:

The objective of this study is to develop a computer based multitasking stressor using the open-source code Psychology Experiment Building Language (PEBL) that will increase

62 subjective and physiological indicators of stress. We hypothesized that subjects would report higher levels of stress and that physiological changes representative of stress would occur as a result of the multitasking. With an appropriate model of multitasking stress, potential solutions for acute stress mitigation may be studied.

Materials and Methods

Subjects

Thirty-one (20 female, 11 male) panelists between 19-46 years old were recruited using the Ohio State University’s sensory database and enrolled in the study under informed consent (approved IRB protocol 2014B0597). In order to participate panelists had to be over 18 years of age, have no history of heart disease, pulmonary issues, or dry mouth (xerostomia). They could not smoke and not currently be using any psychoactive or cardiopulmonary drugs, statins, beta blockers, neurologic medications or smoking/alcohol cessation medications. They also could not be undergoing cancer therapy. They were told the study would last one hour, that they would receive $20 compensation and that they would have electrodes attached to their body for data acquisition. Each panelist participated in the study only once.

Procedure

Upon arriving for a test session, panelists were briefed on the study and then prompted to read carefully without time limitations the consent form. After consent, an

63 explanation of the physiological recording procedure was given and panelists were outfitted with ECG electrodes, one ground electrode placed on the skin over the twelfth rib bone, a positive electrode placed on the skin over the twelfth, left rib bone and the negative electrode placed on the skin over the right clavicle. Two galvanic skin reponse (GSR) electrodes were affixed to the medial plantar suface of the subject’s left foot, because hand movements attributed to the stress tasks would confound electrodes placed on the palm. Electrodes were placed early in the study to facilitate a good signal. The electrodes used were 1.5 inches in diameter, containing 7% chloride wet gel (Mindware Technologies LTD., Gahanna, OH).

Next, panelists received information on the tasks they would perform. Tasks were presented on a computer screen running PEBL software. The tasks were described to panelists, beginning with the manual dexterity task. Panelists were to keep a jittering cursor on a target. Each time the cursor was successfully placed on the target they clicked the mouse to restart the next repetition, with the target at a new location on the screen.

For the Stroop task, numbers and letters appeared on the screen. Depending on the amount of letters or numbers, panelists pressed either ‘1,’ ‘2,’ or ‘3,’ at the top left of the keyboard. They were given an example of seeing “MMM” appear and told they would press ‘3,’ for that repetition since three M’s appeared. Manual dexterity and Stroop number tasks were performed simultaneously. Once the panelists indicated their comprehension of the tasks, they were told compensation was based on correct responses. Compensation was falsely based on performance because it was found that

64 motivation paired with a stressor enhances the stress response (Dickerson and Kemeny 2004).

Panelists were told saliva samples would be collected five times. At each time point, saliva was collected for 2 minutes. Subjects spit into a corning polypropylene 50 mL centrifuge tube (Thermo Fischer Scientific, Waltham, MA). Saliva was collected at baseline, pre-stress, post-stress, recovery (5minutes) and at the end of recovery (10 minutes). At these time points (with the exception of recovery at 5 minutes) they would also rate subjective feelings using the NTLX. The NTLX measured perceived levels of the mental, physical and temporal load (Hart and Lowell 1988), as well as perceived levels of effort, performance and an overall rating of stressfulness.

Additionally a visual analog scale (VAS) was used for an overall stress assessment.

VAS’s are psychometric response scales for attitudes or feelings. The VAS measures were recorded using Compusense software (Guelph, Ontario, CA). Before starting the procedure, panelists removed cell phones and other recording devices to reduce signal interference with the recording electrodes.

The procedure is summarized in figure 6. Before beginning recording, the first saliva sample and subjective stress assessments (SA) were collected. Then physiological data acquisition started and was collected continuously over the duration of the study. The baseline recording began and continued for 10 minutes. The panelists were told to sit quietly and relax, and to try not to move. At the end of baseline, saliva and subjective measures were collected again. Next, the multitasking (MT) stressor was activated and continued for 20 minutes. After stress, saliva and subjective measures were

65 collected again. Lastly, 10 minutes of recovery were recorded with the same instructions as the baseline recording. Five minutes into recovery another saliva sample was taken and a final time 5 minutes later along with a subjective measurement. After completing the final questions the panelists were given $20 regardless of their performance.

Figure 6. Multitasking stressor paradigm with subjective and objective data collection

1Subjective Assessment

2Saliva collection

Statistical Analysis

NTLX items and the stress VAS were analyzed using ANOVA for orthonasal and retronasal conditions with time and subject as main effects. Subjective stress ratings were compared with physiological markers obtained at the same time points.

HRV analysis

Heart rate variability was recorded using BioLab acquisition software (Mindware

Technologies Ltd., Gahanna, OH). HRV data was cleared of artifacts using Mindware’s

HRV analysis software or Kubios HRV software (Biosignal Analysis and Medical

66

Imaging Group, Kupio, Finland). Mean HR was recorded and collected without any need to correct for artifacts.

GSC analysis

The amplitude of the pre- and post-stress skin conductance waveform was measured using the Mindware GSC analysis package (Mindware Technologies Ltd., Gahanna, OH) and corrected for artifacts. The change from baseline was log normalized because GSC data were skewed.

Mean HR

Mean HR was recorded using BioLab acquisition software (Mindware

Technologies Ltd., Gahanna, OH).

For physiological stress analysis, each recording was analyzed individually to remove artifacts. The change from baseline was calculated for incorporation into analysis. Subjective and objective measures were quantified and analyzed using ANOVA with time and subject as main effects.

Salivary amylase quantification

Salivary α-amylase activity (U/mL) was quantified using the Salimetrics kit (Salimetrics,

LLC, State College, PA). The kit utilized an α-amylase substrate that produced a color maximally absorbed at 405 nm, in a plate reader over 3 minutes, using a kinetic method.

The change in absorbance from minute 1 and minute 3 was used to quantify U/mL

67 amylase per sample. The amylase changes from baseline were analyzed using ANOVA with time and subject as main effects. Gender and time of sample were also assessed.

Results

Subjective results

Stress was perceived to increase rapidly (within 10 minutes) and sustained over the course of 20 minutes of testing as measured by the NTLX. Panelists reported significantly higher feelings of mental (p<0.001), physical (p<0.001), and temporal (p<0.001) stress, as well as performance level, effort, and frustration (p<0.001)

(Figure 7).

Figure 7. NTLX perceived changes reported as M±SE (arbitrary units) in mental, physical, and temporal demand, as well as subjects' perceived level of performance, effort and frustration during baseline, pre-stress, post-stress and recovery. When asked to assess their subjective task load index post-stress, subjects reported significant increases in 68 perceived mental (p<0.001), physical (p<0.001) and temporal demand (p<0.001), as well as perceived level of performance (p<0.001), effort (p<0.001) and frustration (p<0.001).

Panelists also rated their perceived stress level at each condition and ratings were significantly (p<0.001) greater post-stressor (3.77±0.42) compared to pre-stressor (2.01±

0.38) or recovery (2.06 ±0.34) (Figure 8).

Figure 8. Perceived stress level reported as M±SE (arbitrary units) during baseline, pre- stress, post-stress and recovery. Post stressor, subjects perceived a significantly higher level of stress (p<0.001) when compared to other session conditions. Superscripts depict significant differences (α=0.05).

Physiological

Similar to the subjective ratings, physiological endpoints also indicated the multitasking stressor induced a significant stress responses. Mean HR change from baseline

69 elevated significantly (p=0.010) by 10 minutes (3.50±1.14 bpm) and was sustained for 20 minutes (3.50±1.14 bpm) of multitasking (Figure 9).

Figure 9. Mean HR change from baseline reported as M±SE (beats per minute) in the multitasking stressor (MT) at 10 and 20 minutes, and during recovery phases at 5 and 10 minutes. Superscripts indicate significant changes in mean HR compared to baseline HR

(p=0.010).

Similarly HRV as indexed by RMSSD (ms) and RSA (ms2) was significantly lower within 10 minutes of multitasking (p<0.001) (Figure 10 and 11). RMSSD significantly decreased by 10 minutes (-9.04±2.84 ms) and was sustained for 20 minutes

(-9.02±2.82 ms) of multitasking stress, compared to significant recovery by 5 minutes

(0.32±2.24 ms) and further recovered by 10 minutes (5.78±3.03 ms) (Figure 10).

Similarly, RSA change from baseline significantly decreased by 10 minutes (-1.10±0.22

70 ms2) and 20 minutes (-1.05±0.22 ms2), rebounding significantly by 5 minutes (-0.01±0.16 ms2) and maintaining recovery at 10 minutes (0.28±0.19 ms2) (Figure 11).

Figure 10. Heart rate variability (RMSSD) change from baseline reported as mean±SE

(ms). This change from baseline is the differences in time between successive heart rate peaks compared to baseline time between heart rate peaks. During multitasking (MT) at

10 and 20 minutes there is a significant decrease in HRV (p<0.001) showing HRV was affected in response to the stressor, and that during recovery HRV returned to baseline levels.

71

Figure 11. Heart rate variability (RSA) change from baseline reported as mean±SE (ms2).

RSA is the natural log transformation of high frequency power measured in ms2. The change from baseline represents changes in RSA during multitasking (MT) and recovery.

During MT at 10 and 20 minutes there is a significant decrease in HRV (p<0.001) showing HRV was affected in response to the stressor, and that during recovery HRV returned to baseline levels.

GSC change from baseline was not significant following the multitasking stressor

(p=0.111) (Figure 12). Stress measures at 10 minutes (0.46±0.23 uS) and 20 minutes

(0.35±0.23 uS) were not significantly lower than recovery measures at 5 minutes (0±0.31 uS) or 10 minutes (0.28±0.21 uS). However GSC has been found to differ among people of different cultures, and due to external factors including temperature and humidity

(Boucsein et al. 2012).

72

Figure 12. GSC change from baseline reported as mean±SE (microsiemens, µS). The change from baseline represents changes in skin conductance during multitasking (MT) and recovery. There are no significant differences between MT and recovery conductance levels (p=0.111) showing GSC was not affected or potentially confounded in response to the stressor.

Salivary α-amylase was collected at 5 time points to indicate changes in amylase levels as a result of the stressor, as well as to indicate recovery. Although time points are not significant (p=0.256), the change from baseline post stress indicates a trend of elevated α-amylase post-stress (15.91± 6.31 U/mL) when compared to pre stress (-

9.57±0.65 U/mL) and 5 minutes (-16.86±1.42 U/mL) or 10 minutes (-25.44±1.70 U/mL) of recovery (Figure 13). Recovery change from baseline values also show that rapid recovery occurred at the conclusion of the stressor in just 5 minutes and even more so by 10 minutes.

73

Figure 13. Change from baseline reported as M±SE in α-amylase activity (U/mL) pre- stress, post-stress, and in recovery at 5 minutes and 10 minutes. There are no significant differences for α-amylase levels between pre-, post-stress, and recovery, however the trend is a pattern that reflects higher levels in response to a stressor and lower levels at times of rest and recovery.

Discussion

Using short-term perceived and physiological endpoints of stress, we validated the utility of the multitasking stressor as a way to effectively induce multitasking stress in a laboratory setting.

The multitasking stressor consisted of two computerized tasks, utilized from an open source program, the Psychology Experiment Building Language (PEBL). The

Stroop number task and a manual dexterity task were attended to simultaneously. Due to

74 the open source nature of PEBL, this stressor is easily implemented in any environment with computer access as opposed to other studies using tasks that require commercial software (e.g. Scholey et al. 2009).

In order to validate the stressor we utilized subjective and objective measures of stress. Subjective measures included the NTLX and an overall stress VAS. The NTLX and overall stress VAS were both significantly affected by the multitasking stressor.

These findings indicate subjects perceived themselves to be significantly more stressed.

Without physiological data to support these findings, it could be argued that this is enough to validate a stressor. If subjects perceive themselves to feel stress, then they are in fact feeling stressed. However, perceptions can be biased, especially when referring to amelioration methods. It is helpful to understand if subjects perceive more or less stress, while also understanding how their physiological measures are responding to stress and a lack there of.

In order to strengthen the support for a validated laboratory-induced multitasking stressor, physiological measures including HR, HRV, GSC and salivary α -amylase were recorded or collected, processed and analyzed. With these measures we found significant support for increased stress response in the form of increased mean HR and α-amylase levels, and decreased HRV measures including RMSSD and RSA. GSC did not produce significant results, however this may have been due to the susceptibility of this measure to the influence of multiple, uncontrolled variables. With evidence of physiological biomarkers significantly changing in the presence of the stressor, there is confidence in the validation of this methodology, which represents an easy to emulate stressor. A

75 simple way to induce and study reduction of an ecologically valid stressor such as multitasking is valuable since this type of stress is rarely studied but arguably affects much of the population.

This type of multitasking often causes people to feel stressed which has been found to be detrimental to health, even with its acute nature. Multitasking is often attempted daily by people, but managing this type of stress too often leads to health problems including unwelcome emotions, feelings and moods. In order to study amelioration techniques, validating a laboratory-induced stressor that mimics this acute, daily stress is needed. With this easily emulated stress induction, numerous mitigation techniques can be studied.

Page Break

76

Chapter 5: The effect of orthonasal and retronasal odorant administration on multitasking

stress reduction

Introduction

Americans have high levels of stress, and agree that reducing stress is a top priority (American Psychological Association 2017). Additionally, a majority of

Americans continue to report that work is a significant source of stress and consistently struggle with stress management techniques. Often acute, daily stress originates from a multitasking demand in workplaces. Chronically stressed patients are prescribed drugs or psychotherapy, but acute stress reduction methods are scarce. This is problematic because like chronic stress, acute stress increases risk for disease (Dienstbier 1989; Fraser et al.

1999; Vrijkotte et al. 2000), and debilitating anxious feelings (Cooke and Rousseau,

1984; Hunter and Thatcher 2007) Possible relief techniques need to be investigated.

Aromatherapy has historically been used to mitigate acute stress, often using lavender essential oil (Diego et al. 1998; Motomura and others 2001; Burnett et al.

2004; Campenni et al. 2004; Kuriyama et al. 2005; Kuroda et al. 2005; Howard and

Hughes 2008; Hoya 2008; Takeda et al. 2008; Toda and Morimoto 2008; Braden 2009;

McCaffrey et al. 2009; Toda and Morimoto 2011; MPham and Siripornpanich 2012; Kim et al. 2015; Bakhsha et al. 2016; Venkataramana et al. 2016), and other essential oils that

77 reportedly reduce stress (Morris 1995; Peng et al. 2009; Lehrner and others 2000;

Heuberger 2001; Heuberger et al. 2001; Bensafi and others 2002; Cheng et al. 2003;

Goel and Lao 2006; Fukui et al. 2007; Holm and Fitzmaurice 2008; Chang and Shen 2011; Liu et al. 2013). However, scientific support for a stress-reducing effect is inconclusive, likely due to the vast array of methodologies enlisted to evoke stress or measure subsequent responses.

Odorants often used in aromatherapy are compounds commonly found in nature.

Plant derived compounds with medicinal qualities is not a novel concept. Lycopene, a pigment in tomatoes acts as an effective antioxidant, generating a reputation as a cardiovascular disease inhibitor (Burton-Freeman et al. 2012). Polyphenols of fruits, vegetables (Kaur and Kapoor 2012) and even wine (Frankel et al. 1995) have inherent antioxidant properties that show beverages and foods can elicit benefits beyond basic satiety. It might, therefore, be reasonable to expect aromatic compounds possess beneficial biological activities as well.

In order to assess the activity and potential benefits of aroma compounds on stress reduction, a controlled and ecologically valid stressor was developed. With validated multitasking stress induction, changes in heart rate variability, salivary α-amylase, as well as subjective moods and feelings can more accurately be attributed to a testable stress reduction stimulus. Utilizing both subjective (conscious) and objective (unconscious) measures of stress enables better characterization of stress and effects of potential amelioration tools.

78

In order to further characterize the potential aromatic effects on stress reduction, separate aroma inhalation routes were utilized for aroma administration—orthonasal and retronasal. Orthonasal (via nostrils) olfaction is more associated with environmental odor perception, whereas retronasal (via oral cavity) olfaction is usually associated with flavor perception. Although both routes deliver odorants to the same receptor fields in the olfactory epithelium, sensation and perception have reportedly differed (Rozin 1982;

Pierce and Halpern 1996; Espinosa-Diaz 2004; Small et al. 2005; Hummel et al. 2006;

Pickering et al. 2007; Hummel and Heilmann 2008; Bender et al. 2009; Welge-Lüssen et al. 2009; Lee and Halpern 2013). Indeed, when delivered orthonasally and retronasally, the same odorants can evoke different physiological responses (Heilmann and Hummel

2004; Small et al. 2005; Hummel et al. 2006), thresholds (Pierce and Halpern 1996;

Hummel et al. 2006; Visschers et al. 2006) perceptual qualities (Pierce and Halpern 1996;

Hummel et al. 2006), hedonic responses (Small et al. 2005), and behaviors (Rozin 1982;

Heilmann and Hummel 2004; Burdach and Doty 1987; although see Sakai et al. 2001 for contradictory results).These differences have led to the Duality of Smell hypothesis

(Rozin 1982) that suggests olfaction operates as a dual system in which orthonasal and retronasal stimuli are processed differently although the specific mechanism enabling this distinction is unknown.

The hypothesis of the present study is that odorants will reduce stress when delivered ortho or retronasally. Linalool was hypothesized to reduce stress most effectively physiologically and subjectively due to its known pharmacological effects.

Contrary to linalool, vanillin, which elicits similar pleasantness, was hypothesized to

79 reduce stress in a psychological manner, due to its lack of data as a bioactive stress- relieving compound. Vanillin serves to control for non-specific pleasantness effects. Air was hypothesized to have no effect on stress but served as a control for expectation because panelists were told relaxation would occur regardless of aroma perception.

The objective of this study is to assess if odor compounds, linalool and vanillin, reduce biometric indicators and subjective measures of stress and additionally, if orthonasal and retronasal delivery of odorants differently impact the effectiveness.

Materials and Methods

Subjects

Forty panelists were recruited using the Ohio State University’s sensory database and were enrolled in the study under informed consent (approved IRB protocol

2014B0597). Thirty-one (17 females, 14 males; average age: 27) subjects successfully completed the study. In order to participate panelists had to be over 18 years of age, have no known smell deficits, no history of heart disease, pulmonary issues or dry mouth

(xerostomia). They had to be a non-smoker and not currently be on psychoactive, cardiopulmonary, neurologic, and/or smoking or alcohol cessation medications. They also could not be undergoing cancer therapy.

Subjects were asked to come to the testing facilities for approximately 1 hour over six sessions that occurred on different days. During the initial sign-up, the study subjects were made aware of these expectations. For each completed session, subjects were

80 compensated $20, and received a total of $120 at the culmination of six attended sessions.

Subjects were made aware they would be evaluating aromas and/or flavors delivered to their nose or oral cavity, respectively, by a customized aroma delivery device and they would also be fitted with small adhesive electrodes to measure heart rate and skin conductance.

Procedure

Upon arriving for a test session, panelists were briefed on the study and then prompted to read the consent form carefully without time limitations. After consent, an explanation of the physiological recording procedure was given and panelists were outfitted with ECG electrodes. Electrodes were placed on the skin over the twelfth, right

(ground) and left rib bones and the right clavicle. Electrodes were placed early in the study to facilitate a good signal. The electrodes used were 1.5 inches in diameter, containing 7% chloride wet gel for ECG electrodes (Mindware Technologies LTD.,

Gahanna, OH).

Following electrode placement, panelists received information on the tasks they would be attending to. Tasks were presented on a computer screen running PEBL software. For the manual dexterity task, panelists were to keep a jittering cursor on a target. Each time the cursor was successfully placed on the target, the panelist had to click the left mouse button to restart the next repetition, with the target at a new location on the screen. For the Stroop task, numbers and letters appeared on the screen.

81

Depending on the amount of letters or numbers, panelists pressed either ‘1,’ ‘2,’ or ‘3,’ at the top left of the keyboard. They were given an example of seeing “MMM” appear and told they would press ‘3,’ for that repetition since three M’s appeared. During testing, the manual dexterity and Stroop number tasks were performed simultaneously. Once the panelists indicated their comprehension of the tasks, they were told they would be compensated based on the number of their correct responses, when in fact all subjects received the same $20 per session. Compensation was falsely based on performance because it was found that motivation paired with a stressor enhances the stress response

(Dickerson and Kemeny 2004).

Panelists were told saliva samples would be collected five times. At each time point, saliva was collected for 2 minutes. Subjects spit into a polypropylene 50 mL centrifuge tube (Thermo Fischer Scientific, Waltham, MA). Saliva was collected at baseline, pre-stress, post-stress, recovery (5minutes) and the end of recovery (10 minutes). At these time points (with the exception of recovery at 5 minutes) panelists also rated subjective feelings using the NTLX. Additionally a visual analog scale (VAS) was used for an overall stress assessment. VAS’s are psychometric response scales for attitudes or feelings. Subjective assessments were recorded using Compusense software

(Guelph, Ontario, CA). The NTLX measured perceived levels of different types of stress, the mental, physical and temporal load (Hart and Lowell 1988), as well as perceived levels of effort, performance and an overall rating of stressfulness. The VAS stress scale told subjects to indicate their level of stress at that moment using a line scale anchored,

“no stress,” and “worst possible stress,” on opposing sides.

82

The procedure is summarized in figure 14. Before beginning recording and aroma delivery, the first saliva sample and subjective stress assessments were collected. Then physiological data acquisition started and was collected continuously over the duration of the study. The baseline recording began and continued for 10 minutes. The panelists were told to sit quietly, relax, and to try not to move. At the end of baseline, saliva and subjective measures were collected again. Next, the multitasking (MT) stressor was activated and continued for 10 minutes. After stress, saliva and subjective measures were collected again. Lastly, 10 minutes of recovery was recorded with the same instructions as the baseline recording. Five minutes into recovery another saliva sample was taken and a final time 5 minutes later along with a final subjective measurement. After completing the final questions the subjects were given $20 regardless of their performance. Before starting the procedure, panelists removed cell phones and other recording devices to reduce signal interference with the recording electrodes. After attaching electrodes, the signal was previewed and if acceptable, the test session could begin.

Aroma delivery

12% linalool (Sigma-Aldrich, St. Louis, MO) or 25% vanillin (Sigma-Aldrich, St.

Louis, MO) were dissolved in miglyol (Nature’s Oil, Streetsboro, OH) or propylene glycol (Fischer Scientific, Fair Lawn, NJ), respectively and delivered in the air phase at 8 liters per minute (retronasal) or 6 liters per minute (orthonasal) with deodorized, humidified breathing air through a customized aroma delivery device. For retronasal delivery, subjects were fitted with a silicone mouthpiece (Figure

83

2A) attached to a glass manifold that was connected via silicone tubing to the aroma source. Subjects were instructed to inhale through their mouth and exhale through their nose. Accumulation of saliva was common during retronasal evaluations. To minimize this, saliva was suctioned out using a dental suction device (Henry Schein,

Melville, NY) inserted around the mouthpiece and placed bilaterally between the cheek and gum. For orthonasal evaluations, subjects were fitted with a glass conical nose piece (Figure 2B) and instructed to inhale through their nose and exhale through their mouth.

Figure 14. Procedure for multitasking aroma session

1Subjective Assessment

2 Saliva collection

84

Statistical Analysis

Subjective and objective measures were quantified and analyzed using ANOVA with time and subject as main effects. Post Hoc analyses were performed using Fisher’s

Least Significant Difference test (LSD). Subjective stress ratings were compared with physiological markers obtained at the same time points.

HRV analysis

Heart rate variability was recorded using BioLab acquisition software (Mindware

Technologies Ltd., Gahanna, OH). HRV data was cleared of artifacts using Mindware’s

HRV analysis software or Kubios HRV software (Biosignal Analysis and Medical

Imaging Group, Kupio, Finland). Mean HR was recorded and collected without any need to correct for artifacts.

Salivary α -amylase quantification

Salivary amylase was quantified using the Salimetrics kit (Salimetrics, LLC, State

College, PA). The kit utilized an α-amylase substrate that produced a color maximally absorbed at 405 nm, in a plate reader over 3 minutes, using a kinetic method. The change in absorbance from minute 1 and minute 3 was used to quantify U/mL α-amylase activity per sample. The amylase changes from baseline were analyzed using ANOVA with time and subject as main effects. Gender and time of sample were also assessed.

85

Results

LINALOOL

Baseline and pre-stress

Orthonasal

Before the introduction of stress, linalool administration did little to affect objective indices of stress, except for mean HR in the orthonasal condition, which was significantly lower for orthonasal linalool (OL) than orthonasal air (OA) (p=0.004)

(Figure 15A; Table 1). RMSSD (p=0.402; Figure 16B; Table 2) and RSA (p=0.346;

Figure 17B; Table 2) raw values were not significantly different as a result of OL inhalation when compared to the OA condition during baseline. Subjective stress ratings were significantly (p=0.005) less during the OL condition compared to the OA condition although this rating was acquired prior to aroma inhalation (Figure 18B; Table 4).

Following 10 minutes of baseline recording, stress ratings taken after this time showed no significant differences following OL administration compared to levels obtained during air inhalation (p=0.795) (Figure 18B; Table 4). Subjective ratings via NTLX (p=0.852) did not significantly differ between orthonasal conditions during baseline (Table 5).

Retronasal

Retronasally, there were no significant reported physiological differences in stress indices as a result of either retronasal linalool (RL) or retronasal air (RA). That is, mean

HR (p=0.379; Figure 15E; Table 1), RMSSD (0.958; Figure 16E; Table 3), and RSA

(p=0.917; Figure 16E; Table 3) measures did not change during the RL condition when compared to RA (Figure 17E; Table 3). Subjective measures also did not differ 86 significantly during baseline retronasal sessions via NTLX (p=0.740; Table 6) and the overall stress VAS (p=0.748; Figure 18E; Table 4).

Stress

Orthonasal

Linalool had a significant effect on moderating stress. Orthonasally, mean HR was lower (p=0.007) during the OL condition compared to OA (Table 1; Figure 15A).

However, HRV indices, RMSSD (p=0.608) and RSA (p=0.346) did not reflect significant changes as a result of the condition when compared to OA (Table 2; Figure 16B). During the stressor, orthonasal conditions did not produce significant subjective measure differences in either NTLX (p=0.919; Table 5) or in the overall stress rating (p=0.606;

Figure 18B; Table 4).

Retronasal

Retronasally, linalool affected the body's response to stress as indicated by heart rate variability differences between RL and RA conditions. Compared to baseline,

RMSSD values obtained in the RL condition were not significantly different (p=0.176) indicating a resiliency to the stressor (Figure 16E; Table 3). The same was true when assessing change from baseline. During RL there was no significant depression in

RMSSD (p=0.176) but in RA there was (p=0.006). Also, during RA, RSA scores were significantly lower (p=0.007) but during the RL condition, the indices did not change significantly (p=0.790; Figure 17E; Table 3). The same pattern was shown for RSA change from baseline in the presence of stress. During the RA condition, RSA’s change from baseline was significantly lower (p=0.006) whereas during RL RSA change

87 from baseline was not significantly lower (0.576). Mean HR values (p=0.444) within the

RL condition did not differ significantly, which was also reflected in the air condition, showing no effect of stress relief indexed by HR (Figure 15E; Table 1). Within the RL condition, subjects did not report significant differences on the NTLX (p=0.547; Table

6) or on their overall stress rating (p=0.730; Figure 18E; Table 4).

Recovery

Recovery was the first time point HRV measures reflected benefits from linalool in both orthonasal and retronasal linalool conditions.

Orthonasal

During OL, mean HR decreased at 5 minutes (p=0.020) and further still at 10 minutes (p=0.003; Figure 15A; Table 1). RMSSD levels were also significantly higher by the end of recovery when compared to baseline for OL (p=0.024), whereas OA was not

(0.354) (Figure 16B; Table 2). Similar patterns were reflected for RMSSD values’ change from baseline within the OL condition (p=0.024) and OA (p=0.310). Within the OL condition, RSA (p=0.297) did not reflect the same changes (Figure 17B; Table 2).

At the end of recovery, subjects reported no significant changes in feelings related to stress during orthonasal conditions via NTLX (p=0.175; Table 5) and the overall stress rating (p=0.615; Figure 18B; Table 4).

Retronasal

RMSSD levels within the RL condition were significantly higher than baseline

(p=0.002), but recovery during the RA condition did not reflect a significant recovery by

5 minutes (p=0.259) or 10 minutes (p=0.384; Figure 16E; Table 3). RSA values during

88 the RL condition were not significantly different in recovery or at any time point suggesting that retronasally delivered linalool enabled RSA values to resist the stressor, thus disabling a need for recovery back to baseline (Figure 17E; Table 3).

In recovery, subjects reported no significant changes in feelings related to stress via NTLX (p=0.992; Table 6) or the overall stress rating (p=0.840; Figure 18E; Table 4).

89

Figure 15. Mean HR (beats per minute) reported in mean±SE bpm in (A.) orthonasal odor conditions and (B) retronasal odor conditions collapsed across session time (baseline, stress and recovery). In the orthonasal conditions, both linalool and vanillin mean HR were significantly (linalool: p=0.004; vanillin: p=0.028) lower than the air condition. The same pattern was not seen for retronasal conditions; retronasal linalool (p=0.136) and vanillin (p=0.982) did not significantly differ from the retronasal air condition in average heart rate.

90

Table 1. Mean HR (beats per minute) reported as M±SE during baseline, stress and recovery for orthonasal air (OA), orthonasal linalool (OL), orthonasal vanillin (OV), retronasal air (RA), retronasal linalool (RL) and retronasal vanillin (RV). Superscripts indicate significant differences between mean HR within each condition.

Time Condition Recovery Recovery Baseline Stress (5 minutes) (10 minutes) 77.75 ± 79.17 ± 75.16 ± 76.02 ± OA 1.96a 1.89b 1.67a 1.67a 74.02 75.69 ± 72.49 ± 73.35 OL* ±1.89a* 1.67b* 1.63a* ±1.72a* 74.58 75.28 72.02 72.93 ± OV* ±1.92a* ±1.61b* ±1.70a* 1.60a* 74.34 ± 74.44 ± 71.19 ± 71.31 ± RA 1.93a 1.69a 1.65b 1.57b 75.22 ± 75.40 ± 72.06 ± 72.37 ± RL 2.06a 1.74a 1.85b 1.81b 73.76 ± 74.48 ± 71.62 ± 71.36 ± RV 1.62a 1.54a 1.51b 1.50b * denotes condition was significantly different (α =0.05) from OA when measures are analyzed at specific time points

91

Figure 16. HRV (RMSSD; ms) reported as M±SE over time for (A) OA, (B) OL, (C) OV,

(D) RA, (E) RL and (F) RV. HRV is the differences in time between successive heart beats. Superscripts indicate significant (α=0.05) differences between HRV within each condition.

92

Figure 17. HRV (RSA; ms2) reported as mean±SE over time for (A) OA, (B) OL, (C)

OV, (D) RA, (E) RL and (F) RV. RSA is the natural logarithm of high frequency power spectrum from the ECG recording reported in ms2. Superscripts indicate significant differences (α =0.05) between HRV within each condition.

Table 2. Orthonasal HRV (M± SE) indexed by RMSSD (ms) and RSA (ms2) during baseline, stress and recovery for OA, OL, and OV. Superscripts indicate significant differences (α =0.05) between HRV within each condition.

Orthonasal Time Recovery Recovery Biomarker Condition Baseline Stress (5 min) (10 min) 37.56 48.81 ± 52.60±4.91 OA ±3 52.31±4.43a 5.18a a .06b RMSSD 41.14 (ms) 51.22 ± 55.43 ± 59.38 ± OL ± 4.62a 5.26ab 5.74 b 3.19c OV 54.50 ± 43.32 57.55 ± 58.01 ± 93

5.99a ± 6.12a 6.14 a 4.00b 7.06 ± 6.36 ± 7.17 ± 0.18 OA 7.22 ± 0.19a 0.20a 0.18b a RSA 7.24 ± 6.52 ± 7.37 ± 0.23 OL 7.26 ± 0.22a (ms2) 0.17a 0.18b a 7.05 ± 6.48 ± 7.27 ± OV 7.26 ± 0.20a 0.17a 0.20b 0.21a

Table 3. Retronasal HRV (M± SE) indexed by RMSSD (ms) and RSA (ms2) during baseline, stress and recovery for RA, RL, and RV. Superscripts indicate significant differences (α =0.05) between HRV within each condition.

Retronasal Time Recovery Recovery Biomarker Condition Baseline Stress (5 min) (10 min) 40.32 48.91 ± 52.41 ± 51.61 ± RA ± 6.69a 4.49a 4.75a 3.13b 45.33 RMSSD 48.48 ± 55.81 ± 54.87 ± RL ± (ms) 5.37a 6.15b 5.80b 4.25a 40.61 48.01 ± 53.18 ± 51.90 ± RV ± 4.88a 4.89b 5.24ab 3.90c 6.80 ± 6.44 6.87 ± 6.93 ± RA 0.23a ±0.16b 0.19a 0.17a RSA 6.79 ± 6.75 ± 7.01 ± 7.05 ± RL (ms2) 0.27 0.29 0.26 0.22 6.85 ± 6.26 ± 7.02 ± 6.87 ± RV 0.22a 0.19b 0.20a 0.20a

94

Figure 18. Stress VAS scores reported as mean ± SE (arbitrary units) over time for (A)

OA, (B) OL*, (C) OV, (D) RA, (E) RL and (F) RV. Superscripts indicate significant

(α=0.05) differences between perceived stress within each condition.

* indicates condition was significantly different than OA when measures are collapsed across time points

Table 4. Stress VAS scores (M± SE) during baseline, stress and recovery (arbitrary units) for OA, OL, OV, RA, RL and RV. Superscripts indicate significant (α=0.05) differences between perceived stress within each condition.

Overall Stress Time rating

Recovery Condition Baseline Pre-stress Post-stress (10 min)

95

OA 2.70 ± 0.44a 2.01 ± 0.34a 4.26 ± 0.35b 2.20 ± 0.34a 1.84 ± OL* 1.93 ± 0.28a 3.97± 0.31b 1.91 ± 0.31a 0.33*a OV 2.46 ± 0.40a 2.10 ± 0.34a 3.94 ± 0.33b 2.04 ± 0.32a RA 2.26 ± 0.41a 2.42 ± 0.37a 4.13 ± 0.31b 2.15 ± 0.35a RL 2.59 ± 0.40a 2.46 ± 0.34a 3.77 ± 0.36b 1.97 ± 0.29a RV 2.71 ± 0.45a 2.70 ± 0.38a 4.00 ± 0.38b 2.18 ± 0.35a *significant difference between orthonasal condition and air condition (α =0.05)

Table 5. NASA Task Load Index item average (M±SE) ratings at baseline, pre-stress, post-stress and following recovery (arbitrary units) for OA, OL, and OV.

NTLX Time Subjective Pre- Post- Recovery Condition Baseline item stress stress (10 min) 0.97 ± 2.43 ± 11.66 ± 1.25 ± OA 0.22 0.55 0.92 0.25 Mental 0.68 ± 2.44 ± 11.08 ± 1.33 ± OL demand 0.13 0.49 0.94 0.26 0.81 ± 2.18 ± 11.24 ± 1.50 ± OV 0.24 0.48 0.92 0.37 1.32 ± 3.02 ± 5.11 ± 1.50 ± OA 0.30 0.52 0.88 0.23 physical 1.47 ± 3.52 ± 6.38 ± 1.70 ± OL demand 0.44 0.69 0.91 0.27 1.62 ± 3.13 ± 6.04 ± 1.64 ± OV 0.48 0.57 0.85 0.39 1.47 ± 1.70 ± 12.08 ± 1.23 ± OA 0.38 0.44 0.93 0.23 Temporal 1.21 ± 1.61 ± 12.56 ± 1.63 ± OL demand 0.37 0.34 0.95 0.35 1.38 ± 2.04 ± 11.69 ± 1.35 ± OV 0.34 0.50 1.06 0.37 1.78 ± 2.84 ± 7.28 ± 1.90 ± OA 0.42 0.46 0.65 0.35 2.09 ± 2.68 ± 7.52 ± 1.95 ± Performance OL 0.64 0.47 0.73 0.31 2.48 ± 3.60 ± 6.61 ± 2.01 ± OV 0.61 0.69 0.71 0.37 1.55 ± 3.02 ± 11.24 ± 1.97 ± Effort OA 0.43 0.57 0.82 0.51

96

2.55 ± 3.09 ± 10.97 ± 1.97 ± OL 0.80 0.57 0.89 0.35 2.00 ± 3.80 ± 11.05 ± 2.11 ± OV 0.48 0.79 0.86 0.42 1.64 ± 3.63 ± 7.71 ± 2.27 ± OA 0.42 0.74 0.93 0.39 1.94 ± 3.55 ± 8.42 ± 2.22 ± Frustration OL 0.54 0.80 1.09 0.54 2.12 ± 4.15 ± 7.18 ± 2.35 ± OV 0.57 0.82 0.88 0.41

Table 6. NASA Task Load Index item average (M±SE) ratings at baseline, pre-stress, post-stress and following recovery (arbitrary units) for RA, RL and RV.

NTLX Time Subjective Pre- Post- Recovery Condition Baseline item stress stress (10 min) 1.13 ± 1.96 ± 11.72 ± 1.58 ± RA 0.27 0.48 0.86 0.33 Mental 1.12 ± 1.53 ± 11.82 ± 1.07 ± RL demand 0.23 0.42 0.93 0.21 1.00 ± 1.58 ± 12.19 ± 1.53 ± RV 0.24 0.38 0.87 0.34 1.12 ± 2.06 ± 5.58 ± 1.46 ± RA 0.28 0.38 0.99 0.28 physical 1.84 ± 2.15 ± 5.82 ± 1.15 ± RL demand 0.47 0.45 0.86 0.23 1.45 ± 2.36 ± 6.40 ± 1.27 ± RV 0.41 0.59 0.97 0.27 1.21 ± 1.65 ± 12.72 ± 1.37 ± RA 0.28 0.41 0.86 0.27 Temporal 2.46 ± 1.63 ± 12.26 ± 1.40 ± RL demand 0.61 0.38 0.97 0.30 1.40 ± 1.27 ± 12.75 ± 1.33 ± RV 0.40 0.25 0.82 0.31 1.90 ± 2.27 ± 7.61 ± 1.91 ± RA 0.39 0.39 0.59 0.32 2.56 ± 2.16 ± 8.01 ± 1.99 ± Performance RL 0.47 0.37 0.80 0.34 2.17 v 2.28 ± 7.60 ± 1.99 ± RV 0.53 0.47 0.74 0.39

97

1.63 ± 2.54 ± 12.48 ± 1.89 ± RA 0.43 0.51 0.80 0.44 2.36 ± 2.29 ± 12.12 ± 1.85 ± Effort RL 0.58 0.54 0.90 0.40 1.87 ± 2.16 ± 12.52 ± 1.41 ± RV 0.46 0.49 0.67 0.26 1.51 ± 2.25 ± 7.95 ± 2.16 ± RA 0.39 0.54 0.82 0.54 1.94 ± 2.31 ± 9.26 ± 1.86 ± Frustration RL 0.49 0.48 1.02 0.38 1.59 ± 1.96 ± 8.80 ± 2.15 ± RV 0.39 0.42 1.02 0.53

VANILLIN

Baseline

Orthonasal

At baseline RMSSD (p=0.190; Figure 16C; Table 3) and RSA (p=0.936; Figure

17C; Table 3) HRV values did not significantly differ between orthonasal air (OA) and orthonasal vanillin (OV) conditions. However, mean HR was significantly lower during the OV condition than during the OA condition (p=0.028) (Figure 15A; Table 1).

Subjective ratings via NTLX (p=0.852; Table 5) did not significantly differ between orthonasal conditions during baseline, but the overall stress rating was significantly lower than OA (p=0.039), albeit this was before inhalation began (Figure 18C; Table 4).

Retronasal

During the RV condition, RMSSD (p=0.666; Figure 16F; Table 3), RSA

(p=0.768; Figure 17F, Table 3) and mean HR (p=0.919; Figure 15B; Table 1) values were not significantly different than values recorded during the RA session. Subjects also

98 did not significantly report feeling different during baseline of retronasal conditions via

NTLX (p=0.740; Table 6) or overall stress rating (p=0.0.748; Figure 18F; Table 4).

Stress

Stress reducing effects of vanillin were not indicated in either the orthonasal or retronasal delivery conditions.

Orthonasal

During stress, the only implication of stress reduction was indexed by a significantly lower mean HR for OV compared to OA in the presence of stress (p=0.024;

Figure 15A; Table 1). Within the conditions, OV treatment produced a mean HR that did not significantly elevate, whereas during RA, mean HR marginally increased (p=0.061;

Figure 15A; Table 1). Among orthonasal conditions HRV measures including RMSSD

(p=0.093; Figure 16C; Table 2) and RSA (p=0.462; Figure 17C; Table 3) did not significantly differ. Subjectively, panelists reported no differences among orthonasal conditions during stress via the NTLX (p=0.919; Table 5) or overall stress rating

(p=0.606; Figure 18C; Table 4).

Retronasal

During retronasal conditions, HRV varied in results. During the RV condition

RSA measures were lowest when compared to other retronasal conditions, albeit insignificantly (p=0.188), contradicting vanillin as a relaxing compound (Figure 17F;

Table 3). RMSSD (p=0.203; Figure 16F; Table 4), and mean HR (p=0.421; Figure 15B;

Table 1) values were not significantly different when comparing recordings from RV and

RA.

99

Subjectively, panelists reported no differences among orthonasal conditions during stress via NTLX (p=0.547; Table 6) or overall stress (p=0.730) (Figure 18F; Table 4).

Recovery

Orthonasal

During the OV condition, mean HR values were significantly lower during recovery (p=0.030; Figure 15A; Table 1). However, RMSSD (p=0.135; Figure 16C;

Table 2) and RSA (p=0.582; Figure 17C; Table 2) values did not reflect any differences among orthonasal conditions during 10 minutes of recovery. Subjects reported no significant changes in feelings related to stress during orthonasal conditions via NTLX

(p=0.175; Table 5) and overall stress measure (p=0.615; Figure 18C; Table 4).

Retronasal

During RV administration, RMSSD values reflected stress reduction, but RSA values produced the opposite outcome. RSA change from baseline values were significantly lower during the RV condition when compared to RA (p=0.048; Figure 17F;

Table 3). RMSSD values were significantly higher than baseline levels during the RV condition by 5 minutes (p=0.047), but did not maintain significantly higher values until the end of recovery (p=0.134; Figure 16F; Table 4). Mean HR values did not reflect significant changes during retronasal conditions at the end of recovery (p=0.534; Figure

15B; Table 1). Subjectively, retronasal conditions did not have any significant effects on reported feelings or stress as indicated via NTLX (p=0.992; Table 6) and stress ratings

(p=0.840; Figure 18F; Table 4).

α-Amylase results

100

The diurnal course of saliva α-amylase activity has been reported to show a pronounced decrease within 60 minutes after awakening, followed by stable increases throughout the day (Rohleder et al. 2007). We assessed for a diurnal effect and found that sessions in the morning had significantly less α-amylase activity than in the afternoon for both air (orthonasal: p=0.001; retronasal: p<0.001) and vanillin (orthonasal: p=0.009; retronasal: p=0.004) conditions, however during the linalool conditions there was not a marked diurnal effect on amylase activity (orthonasal: p=0.062; retronasal: p=0.172). After controlling for session time, α-amylase showed a relationship similar to those reflected in the physiological results. Linalool conditions actually showed decreased α-amylase activity over the session in both ortho and retro conditions (Figure

19B; Table 7) although the decreases were not significant. The air condition showed α- amylase increases orthonasally and retronasally post-stressor with a significant increased change from baseline (p=0.001; Figure 19A; Table 7). Vanillin mimicked its contradicting physiological results as well, in that α-amylase activity increased post-stress but did not reach significance (p=0.348; Figure 19C; Table 7).

101

Figure 19. α -Amylase activity (U/mL) changes from baseline reported as mean±SE for orthonasal and retronasal A. air, B. linalool, and C. vanillin conditions. Superscripts depict significant (α = 0.05) differences within that condition.

Table 7. α -Amylase activity (U/mL) changes from baseline reported as M±SE collapsed across orthonasal and retronasal inhalation conditions for pre-stress, post-stress, recovery at 5 minutes and recovery at 10 minutes. Superscripts indicate significant (α=0.05) differences between α –amylase activity within each inhalation condition.

α-amylase average change Time from baseline Recovery Recovery condition Pre-stress Post-stress (5min) (10 min) -22.74 ± 8.17 ± -21.14 ± -21.83 ± Air 12.77a 12.36b 11.26a 12.13a 4.34 ± -7.51 ± -17.06 ± -16.75 ± Linalool 12.28a 12.05a 11.31a 15.49a 4.24 ± -15.68 ± 17.33 ± Vanillin -23 ± 14.39a 14.03a 12.23a 15.83a

102

Discussion

α -Amylase

The lower α-amylase activity produced during the linalool sessions suggests linalool was improving the body’s response to a multitasking stressor. Whereas α- amylase increased in response to stress during the air and vanillin trials, linalool inhalation prevented this effect. This finding is consistent with the stress-reducing properties attributed to linalool or lavender treatments.

Linalool

Linalool has historically been used to relieve stress, often delivered using a lavender essential oil (Diego et al. 1998; Motomura et al. 2001; Burnett et al.

2004; Campenni et al. 2004; Kuriyama et al. 2005; Kuroda et al. 2005; Howard and

Hughes 2008; Hoya et al. 2008; Takeda et al. 2008; Toda and Morimoto 2008; Braden

2009; McCaffrey et al. 2009; Toda and Morimoto 2011; MPham and Siripornpanich

2012; Kim et al. 2015). Prior studies have suggested that the impact compound of lavender oil is linalool (Kuroda et al. 2004; MPham and Siripornpanich 2012).

Several studies support our findings that linalool can affect physiological biomarkers exhibiting changes that correlate with stress relief. One study found lavender exposure reduced stress from arithmetic indexed by salivary chromagranin A (CgA), an indicator of psychological stress (Toda and Morimoto 2008), while radial pulse differed significantly in another study (McCaffrey et al. 2009). We found mean HR significantly

103 affected which was echoed by Kuroda et al. (2005) and Chang and Shen (2011).

While we did not test the above physiological measures (CgA levels and radial pulse), other studies have shown effects of linalool using with the same measures we employed (Burnett et al. 2004; Campenni 2004; Kuroda et al. 2005; Howard and

Hughes 2008; MPham and Siripornpanich 2012). However, many of these effects are without controlled stress induction (Diego et al. 1998; Campenni 2004; Kuroda et al.

2005; Toda and Morimoto 2008) and thus may reflect confounded responses to stress. Studies inducing stress found physiological effects (McCaffrey et al. 2009;

Masaoka et al. 2013), and some did not (Burnett et al. 2004).

Contradictory results are likely due to methodology differences or a lack of enough measurements to provide comprehensive insights, thus underscoring the importance for both controlled methodology and multiple measurements for both objective and subjective stress indices.

Previously studies have also shown subjective feelings of stress relief when assessing effects on test anxiety (Kutlu et al. 2008; McCaffrey et al. 2009) and pre- operative patients (Braden 2009). We did not find similar reported results from the NTLX but did see reported lower overall stress ratings during OL when compared to OA within our stress paradigm. Others have found subjective and objective measures supporting linalool’s effects (Diego et al. 1998; Kuroda et al. 2005; Lagopoulos et al.

2009; MPham and Siripornpanich 2012). One possibility for the lack of reported stress relief via NTLX following orthonasal administration of linalool might be our control for expectation. Expectation has been found to affect stress reports (Campenni et

104 al. 2004; Diego et al. 2008; Howard and Hughes 2008; Masaoka et al. 2013) Presently, panelists were told that they should expect to feel relaxed while being evaluated on task performance and that aromas may be delivered in ways such that perceiving them may not be possible. This frame of mind would have panelists, regardless of condition, expecting to feel relaxed and realize they may not need to smell anything (air control) to feel this way. One reason this may have caused a lack of differentiable subjective reports in the NTLX is that panelists might then report feeling relaxed in all conditions since they expected this to occur, leaving subjective reports indistinguishable. Although expectation may be a reason that subjective assessments were not as illuminating as biometric measures, it is an important control. The controlled nature of this methodology ensures that subject’s need not think they should perceive an aroma to feel relaxed, thus strengthening our control session, and thereby improving the reliability of results.

The NTLX also may carry a demand characteristic. The contrast of doing nothing during baseline and recovery compared to a demanding multitasking paradigm was a stronger difference than the potential to feel more or less relaxed in any one condition.

The stronger difference between baseline and recovery, and the fact that the stressor was present in each inhalation condition, likely overshadowed any smaller changes in mental, physical or temporal demands, or in their perceived level of performance, effort and frustration. It may have been that the context of the stress paradigm made subjective changes more difficult to determine and report.

The one significant difference in subjective indices was the overall stress VAS ratings. The OL condition caused more subjects to rate an overall lower perceived level

105 of stress during the session. The overall stress rating may have been more sensitive to changes in reported stress levels since it was asking subjects verbatim about their stress level. It was a more direct question and it also did not probe the subjects as much as the items on the NTLX did, perhaps allowing for a more visceral reaction reflective of the physiological changes indicative of less stress.

However, this measure did not reflect subjects’ reports of less perceived stress during the retronasal linalool condition, which was the condition that affected biometric indices of stress the strongest. One possible reason for the lack of self-reported stress relief in the RL condition was because the delivery method for retronasal odors was more discomforting and unusual than the method used for orthonasal odor delivery, and was potentially more distracting making it difficult to feel and report less stress.

The RL condition produced stronger effects on stress indices than measures collected during the OL condition. Baseline recordings reflected no immediate changes.

However, upon stress induction, heart rate variability measures were changed during the

RL condition. RMSSD and RSA were significantly affected, with measures recorded during RL exhibiting significantly higher values than the RA condition during stress. Upon the recovery period, RL values were significantly higher for RMSSD measures indicating a more relaxed or less stressed state. Subjectively, however recovery did not show differences between measures reported after RA and RL conditions.

Overall, during RL there are greater reported changes in objective stress reduction indices than in the OL condition over the course of the sessions. This may have occurred because retronasal delivery allowed for a higher blood plasma level of linalool

106 during this condition (Kerns 2017), supporting the idea that linalool concentration in blood contributes to its relaxation effects.

Another reason for RL's stronger effects than OL could be due to the phenomenon of adaptation. We previously found retronasal and orthonasal delivery over time only adapted orthonasally and did not adapt retronasally (Pierce and Simons 2017). This could support the idea that with continued psychological awareness of an effective odorant, stress relief could be strengthened, since adaptation may have taken place with persistent orthonasal delivery of the compound. In other words, panelists could have begun an orthonasal session aware of an aroma, but as it dwindled in intensity, the expectation of stress relieving effects may have also lessened. Theoretically this should not have occurred because panelists were told they need not be aware of an aroma to feel relaxed. However, beginning with a stronger intensity of an expected relaxing aroma and having it decrease in intensity would suggest to subjects that perhaps its relaxing properties are also decreasing, as opposed to beginning with an unperceivable aroma and it never changing.

Vanillin

Vanillin is a familiar food flavor and has less literature support for stress relieving properties although it has been used to mediate newborn pain responses (Goubet et al.

2003). The mechanism underpinning analgesic effects has been related to its familiarity as an aroma (Rattaz et al. 2005). Rattaz et al. (2005) offers support for vanillin, a familiar flavor, to elicit stress reduction, however the results were contradicting, with authors 107 concluding that the familiarization to the odor was the reason for its relaxing effects and not due to the compound itself. Another study reported ameliorating depressive like behaviors in rats due to elevated serotonin and dopamine (Xu et al. 2015) while also exhibiting support for promoting cancer cell apoptosis (Ho et al. 2009) and reduced activity of inflammatory pathways (Khuda-Bukhsh et al. 204). These effects, however do not support vanillin's ability to relieve acute stress incidences in humans at least in the short term, which aligns with the present findings.

Vanillin, although pleasant, did not show the same effects that the linalool treatments exhibited. It was suggested a pleasant aroma may show a stress reduction effect because studies have found stimulus pleasantness and expectation to effect stress responses (Chamine et al. 2016), however our data did not support this hypothesis.

Results for vanillin show that pleasantness alone was not enough to affect objective or subjective stress reduction consistently. During orthonasal administration of vanillin, mean HR was lower than the air condition, but other physiological and subjective indices did not reflect consistent effects. Like that of linalool, perhaps vanillin could have been more effective when administered retronasally due to a potentially higher concentration received. However, retronasally our results do not support vanillin as a stress relieving compound via subjective or objective measures. Baseline measures did not differentiate between RA and RV. Additionally, during stress HRV indices contradicted one another.

RSA values were higher for air than vanillin, supporting air as a better aid in stress relief than vanillin. However, during recovery within the RV condition, RMSSD values reflected positive effects on stress, but RSA again reflected an opposing effect.

108

As our results indicated, vanillin showed little promise as an effective acute stress relief agent. Linalool did not produce any contradictory evidence which suggests the idea that linalool may be a more effective stress relief odorant and potentially one with specific mechanistic qualities compared to vanillin. Vanillin did little to resist physiological changes in stress as indexed by objective measures of HR, HRV and salivary α-amylase. Vanillin also did not affect subjective measures as hypothesized.

One aspect that provides insight into pharmacologcal mechanisms of odorant compounds lies in chirality. For instance, (R)-(-)-linalool was found to reduce HR and affect mood (Kuroda et al. 2005). During transdermal application (R)-(-) linalool decreased BP, with a moderate decrease in skin temperature when compared to the control group receiving a placebo, but found no difference of subjective effects on well- being (Heuberger et al. 2004). Other studies have shown that chiral molecules may exhibit different effects, further supporting molecular specificity regarding stress reduction mechanisms of particular aromatic compounds. This is intriguing from a mechanistic perspective because it shows the possibility of how distinct pharmacologically-active molecules must be to produce an effect. This is not to say it is impossible to feel relaxed with vanillin or other compounds that are common in aromatherapy, but perhaps relaxation mechanisms differ for different aromatherapy compounds, some with larger bioactivities.

Mechanisms of action

109

The idea that psychological awareness of an aroma compound could be needed for stress relieving properties might be just that—psychological. With awareness of a pleasant aroma compound, increased contentment and therefore less stress could result.

However, the present study's results suggest physiological changes may indicate stress still exists. Psychological awareness may enhance stress relieving properties if such odorants being used also carry pharmacological effects simultaneously, like that of linalool in a lavender essential oil. Thus expectation and persistent awareness of a pleasant aroma may support pharmacological effects, but may not solely affect biological stress relief.

A lack of subjective stress reducing effects as measured by NTLX may suggest that perhaps this scale is not as illuminating as a more encompassing overall stress measure, like the stress VAS. The inconsistency between subjective results shows the need for inclusion of multiple measures that respond to different qualities of stress on the body. Previously, some studies have just used subjective reports to assess stress mitigation, but the present study's results show how objective and subjective measures may lead to different and potentially misleading conclusions.

There are reported benefits to physiological HRV resilience because low HRV is associated with inflammation (Sajadieh et al. 2004). With an effective method of keeping

HRV elevated or resilient to effects of everyday types of stress, these methods help fight inflammation and thus many debilitating diseases, such as cancer. With these findings it becomes clear how flavor and fragrance industries might leverage such findings for new product development. Fragrances have often been used as marketable ways to become

110 more relaxed but with our findings, scientific support could push aromatherapy methods from mythic to scientific, on a compound by compound basis. Product developers using impact levels of an aromatic compound could create a functional food product that is more marketable due to its benefits beyond its caloric value. Functional foods are becoming increasingly important to consumers, because health and wellness are major drivers of grocery purchases (Mintel 2017).

Conclusion

It is clear that adaptation differs depending on aroma delivery route, providing more evidence for olfaction being a dual sensory system. A better understanding of adaptation patterns has positioned us to develop stronger methodologies to study aroma stress mitigation techniques.

Additionally, prior studies often used specific stressed populations or induced stress to a maximal degree. Results from these approaches are difficult to generalize to acute stress experienced on a daily basis. Upon validating an ecologically relevant stressor, it becomes possible to assess acute stress reduction and because the stressor is easy to replicate, assessing different aromatic stimuli make stimuli effects more comparable.

The hypothesis that linalool was an effective stress relief aid was supported. Air and vanillin conditions did not produce consistent differentiable effects on stress reduction either subjectively or objectively. These conditions were effective controls for expectation and pleasantness, which appear to play a lesser role on stress relief indices in this study. However, linalool not only exhibited physiological changes reflecting stress 111 reduction when delivered via both aroma pathways, but linalool delivered retronasally produced the strongest heart rate variability and salivary α -amylase stress relief evidence when compared to its orthonasal counterpart. This difference is possibly due to higher concentrations reaching the blood stream.

Linalool showed effects that support its historical use as a component of therapeutic essential oil blends meant for relaxing. The present study elucidated its stronger effect when delivered retronasally, allowing for the opportunity to create products capitalizing on retronasal delivery of linalool for stress relief.

This study also further emphasizes the power of naturally occurring compounds to reduce biological changes brought on by stress. Reducing HRV and potentially inflammation may have important clinical implications. Combatting the body's physiological stress responses may help promote health by inhibiting diseases related to increased inflammation.

This methodology will be simple to replicate in order to further characterize other aroma compounds with potential stress reduction properties. The methodology is already validated, making potential amelioration techniques easier to assess. With an easily emulated acute stressor, studying acute stress reduction techniques becomes more likely.

Future studies

Future studies should aim to quantify the delivered volatile concentration in order to better understand the response to specific concentrations of odorants. Assessing the volatile concentration of odorants is important for our understanding of what level of odor molecules are needed to produce a stress relief response and for comparisons with 112 other aromatic molecules. Additionally developing a dose-response curve would enable thorough characterization of the amount of volatile compound needed for a stress reduction response. Further, the dose-response relationship would allow us to thoroughly assess at what level the compound becomes bioactive, and at what level that bioactivity begins to plateau. This will allow for clarification of how much of the molecule is needed, how much can be added to enhance the effect and to understand at what level does adding more of the compound no longer enhance the effect.

An additional future study would include developing a linalool beverage that could be assessed for stress reduction properties. The benefits to developing a beverage or food product associated with stress reduction are similar to reasons other functional foods are of interest. Food is a part of daily life, making consumption an easy way to help prevent future diseases that stem from stress. Consumption is also an easier lifestyle change to adopt than therapy or other stress reducing mitigation techniques.

Bioavailability of functional compounds is also often enhanced when delivered in food matrices as opposed to capsules like many supplements. Kopec et al. 2014 found carotenoids are better absorbed when consumed with avocados, suggesting compounds that are more stable in hydrophilic or hydrophobic matrices would likely be better absorbed when delivered as such. Functional foods are often more enjoyable to consume than medicines which often increases reported compliance (Lesinki et al. 2015), enabling ease of healthy lifestyle changes.

113

Bibliography

American Psychological Association. (2014). Stress in america: Paying with our

health (Survey

American Psychological Association. (2017). Stress in america: Coping with change,

part 2 (technology and social media).Stress in AmericaTM Survey.

Bakhsha, F., Yousefi, Z., Aryaee, M., Jafari, S. Y., & Derakhshanpoor, F. (2016).

Comparison effect of lavender and citrus aurantium aroma on anxiety in female

students at golestan university of medical sciences.

Balodis, I. M., Wynne-Edwards, K. E., & Olmstead, M. C. (2010). The other side of the

curve: Examining the relationship between pre-stressor physiological responses and

stress reactivity.Psychoneuroendocrinology, 35(9), 1363-1373.

Bender, G., Hummel, T., Negoias, S., & Small, D. M. (2009). Separate signals for

orthonasal vs. retronasal perception of food but not nonfood odors. Behavioral

Neuroscience, 123(3), 481.

Bensafi, M., Rouby, C., Farget, V., Bertrand, B., Vigouroux, M., & Holley, A. (2002).

Autonomic nervous system responses to odours: The role of pleasantness and

arousal. Chemical Senses, 27(8), 703-709.

114

Berglund, B., Berglund, U., & Lindvall, T. (1986). Theory and methods for odor

evaluation. Experientia, 42(3), 280-287.

Berntson, G. G., Cacioppo, J. T., & Quigley, K. S. (1993). Respiratory sinus arrhythmia:

Autonomic origins, physiological mechanisms, and psychophysiological

implications. Psychophysiology, 30(2), 183-196.

Berntson, G. G., Quigley, K. S., & Lozano, D. (2007). Cardiovascular

psychophysiology. Handbook of Psychophysiology, 3, 182-210.

Best, A. R., & Wilson, D. A. (2004). Coordinate synaptic mechanisms contributing to

olfactory cortical adaptation. The Journal of Neuroscience : The Official Journal of

the Society for Neuroscience, 24(3), 652-660.

Bohan, Lt Col Kevin J, & Maye, C. J. P. (2010). Determining the relationship of acute

stress, anxiety, and salivary α-amylase level with performance of student nurse

anesthetists during human-based anesthesia simulator training. AANA Journal, 78(4),

301.

Boucsein, W., Fowles, D. C., Grimnes, S., Ben-Shakhar, G., roth, W. T., Dawson, M. E.,

Filion, D. L., & Society for Psychophysiological Research Ad Hoc Committee on

Electrodermal Measures. (2012). Publication recommendations for electrodermal

measurements. Psychophysiology, 49(8), 1017-1034.

115

Braden, R., Reichow, S., & Halm, M. A. (2009). The use of the essential oil lavandin to

reduce preoperative anxiety in surgical patients. Journal of Perianesthesia

Nursing, 24(6), 348-355.

Bradley, B. F., Brown, S. L., Chu, S., & Lea, R. W. (2009). Effects of orally administered

lavender essential oil on responses to anxiety-provoking film clips. Human

Psychopharmacology: Clinical and Experimental, 24(4), 319-330.

Brewer, T. (2002). Test-taking anxiety among nursing & general college

students. Journal of Psychosocial Nursing and Mental Health Services, 40(11), 22-

29.

Brody, S., Preut, R., Schommer, K., & Schürmeyer, T. H. (2002). A randomized

controlled trial of high dose ascorbic acid for reduction of blood pressure, cortisol,

and subjective responses to psychological stress. Psychopharmacology, 159(3), 319-

324.

Burdach, K. J., & Doty, R. L. (1987). The effects of mouth movements, swallowing, and

spitting on retronasal odor perception. Physiology & Behavior, 41(4), 353-356.

BURNETT, K. M., SOLTERBECK, L. A., & STRAPP, C. M. (2004). Scent and mood

state following an anxiety-provoking task 1. Psychological Reports, 95(2), 707-722.

Burton‐Freeman, B., Talbot, J., Park, E., Krishnankutty, S., & Edirisinghe, I. (2012).

Protective activity of processed tomato products on postprandial oxidation and

116

inflammation: A clinical trial in healthy weight men and women. Molecular

Nutrition & Food Research, 56(4), 622-631.

Cain, W. S. (1974). Perception of odor intensity and the time-course of olfactory

adaptation. ASHRAE Transactions, 80, 53-75.

CAMPENNI, C. E., CRAWLEY, E. J., & MEIER, M. E. (2004). Role of suggestion in

odor-induced mood change 1. Psychological Reports, 94(3c), 1127-1136.

Cerf-Ducastel, B., & Murphy, C. (2001). fMRI activation in response to odorants orally

delivered in aqueous solutions. Chemical Senses, 26(6), 625-637.

Chamine, I., & Oken, B. S. (2016). Aroma effects on physiologic and cognitive function

following acute stress: A mechanism investigation. The Journal of Alternative and

Complementary Medicine, 22(9), 713-721.

Chang, K. M., & Shen, C. W. (2011). Aromatherapy benefits autonomic nervous system

regulation for elementary school faculty in taiwan. Evidence-Based Complementary

and Alternative Medicine : ECAM, 2011, 946537.

Chen, V., & Halpern, B. P. (2008). Retronasal but not oral-cavity-only identification of

"purely olfactory" odorants. Chemical Senses, 33(2), 107-118.

Cheng, A., Chang, J., Kida, E., & Monteath, N. (2003). Evaluating the anxiety-reducing

effects of aromatherapy using cognitive and memory tests. Journal of Young

Investigators, 9(10)

117

Chiappelli, F., Iribarren, F. J., & Prolo, P. (2006). Salivary biomarkers in

psychobiological medicine. Bioinformation, 1(8), 331-334.

Cohen, S., & Williamson, G. M. (1991). Stress and infectious disease in

humans. Psychological Bulletin, 109(1), 5.

Cometto-Muñiz, J. E. (1981). Odor, taste, and flavor perception of some flavoring

agents. Chemical Senses, 6(3), 215-223.

Connor-Smith, J. K., Compas, B. E., Wadsworth, M. E., Thomsen, A. H., & Saltzman, H.

(2000). Responses to stress in adolescence: Measurement of coping and involuntary

stress responses. Journal of Consulting and Clinical Psychology, 68(6), 976-992.

Cooke, R. A., & Rousseau, D. M. (1984). Stress and strain from family roles and work-

role expectations. Journal of Applied Psychology, 69(2), 252.

Creswell, J. D., Welch, W. T., Taylor, S. E., Sherman, D. K., Gruenewald, T. L., &

Mann, T. (2005). Affirmation of personal values buffers neuroendocrine and

psychological stress responses. Psychological Science, 16(11), 846-851.

Crow, S. M., Hartman, S. J., Mahesh, S., McLendon, C. L., Henson, S. W., & Jacques, P.

(2008). Strategic analyses in nursing schools: Attracting, educating, and graduating

more nursing students: Part I--strengths, weaknesses, opportunities, and threats

analysis. The Health Care Manager, 27(3), 234-244.

Dalton, P., & Wysocki, C. J. (1996). The nature and duration of adaptation following

long-term odor exposure. Perception & Psychophysics, 58(5), 781-792.

118

Dalton, P. (2000). Psychophysical and behavioral characteristics of olfactory

adaptation. Chemical Senses, 25(4), 487-492.

DeWeese, David D., Saunders,William H.,,. (1968). Textbook of otolaryngology. St.

Louis: Mosby, 1964.

Dickerson, S. S., & Kemeny, M. E. (2004). Acute stressors and cortisol responses: A

theoretical integration and synthesis of laboratory research. Psychological

Bulletin, 130(3), 355.

Diego, M. A., Jones, N. A., Field, T., Hernandez-Reif, M., Schanberg, S., Kuhn, C.,

Galamaga, M., McAdam, V., & Galamaga, R. (1998). Aromatherapy positively

affects mood, EEG patterns of alertness and math computations. International

Journal of Neuroscience, 96(3-4), 217-224.

Dienstbier, R. A. (1989). Arousal and physiological toughness: Implications for mental

and physical health. Psychological Review, 96(1), 84.

Edwards, V. H. (2015). The aromatherapy companion: Medicinal Uses/Ayurvedic

Healing/Body-care Blends/Perfumes & Scents/Emotional health & well-being Storey

Publishing.

Espinosa Diaz, M. (2004). Comparison between orthonasal and retronasal flavour

perception at different concentrations. Flavour and Fragrance Journal, 19(6), 499-

504.

Fechner, G. T. (1860). Elements of psychophysics.

119

Feldman, R., Rosenthal, Z., & Eidelman, A. I. (2014). Maternal-preterm skin-to-skin

contact enhances child physiologic organization and cognitive control across the first

10 years of life. Biological Psychiatry, 75(1), 56-64.

Ferguson, S. S., & Caron, M. G. (1998). G protein-coupled receptor adaptation

mechanisms. Seminars in Cell & Developmental Biology, , 9(2) 119-127.

Filaire, E., Portier, H., Massart, A., Ramat, L., & Teixeira, A. (2010). Effect of lecturing

to 200 students on heart rate variability and alpha-amylase activity. European

Journal of Applied Physiology, 108(5), 1035-1043.

Fowles, D. C., Christie, M. J., Edelberg, R., Grings, W. W., Lykken, D. T., & Venables,

P. H. (1981). Publication recommendations for electrodermal

measurements. Psychophysiology, 18(3), 232-239.

Frankel, E. N., Waterhouse, A. L., & Teissedre, P. L. (1995). Principal phenolic

phytochemicals in selected california wines and their antioxidant activity in

inhibiting oxidation of human low-density lipoproteins. Journal of Agricultural and

Food Chemistry, 43(4), 890-894.

Fraser, R., Ingram, M. C., Anderson, N. H., Morrison, C., Davies, E., & Connell, J. M.

(1999). Cortisol effects on body mass, blood pressure, and cholesterol in the general

population. Hypertension, 33(6), 1364-1368.

120

Fukui, H., Komaki, R., Okui, M., Toyoshima, K., & Kuda, K. (2007). The effects of odor

on cortisol and testosterone in healthy adults. Neuroendocrinology Letters, 28(4),

433-437.

Furia, T. E. (1973). CRC handbook of food additives CRC press.

Furudono, Y., Cruz, G., & Lowe, G. (2013). Glomerular input patterns in the mouse

olfactory bulb evoked by retronasal odor stimuli. BMC Neuroscience, 14(1), 45.

Garcia-Medina, M. R. (1981). Flavor-odor taste interactions in solutions of acetic acid

and coffee. Chemical Senses, 6(1), 13-22.

Goel, N., & Raymund P. Lao. (2006). Sleep changes vary by odor perception in young

adults. Biological Psychology, 71.3, 341-349.

Gordis, E. B., Granger, D. A., Susman, E. J., & Trickett, P. K. (2006). Asymmetry

between salivary cortisol and α-amylase reactivity to stress: Relation to aggressive

behavior in adolescents.Psychoneuroendocrinology, 31(8), 976-987.

Goubet, N., Rattaz, C., Pierrat, V., Bullinger, A., & Lequien, P. (2003). Olfactory

experience mediates response to pain in preterm newborns. Developmental

Psychobiology, 42(2), 171-180.

Goubet, N., Strasbaugh, K., & Chesney, J. (2007). Familiarity breeds content? soothing

effect of a familiar odor on full-term newborns. Journal of Developmental and

Behavioral Pediatrics : JDBP, 28(3), 189-194.

121

Goyert, H. F., Frank, M. E., Gent, J. F., & Hettinger, T. P. (2007). Characteristic

component odors emerge from mixtures after selective adaptation. Brain Research

Bulletin, 72(1), 1-9.

Grossman, P., Stemmler, G., & Meinhardt, E. (1990). Paced respiratory sinus arrhythmia

as an index of cardiac parasympathetic tone during varying behavioral

tasks. Psychophysiology, 27(4), 404-416.

Grossman, P., & Svebak, S. (1987). Respiratory sinus arrhythmia as an index of

parasympathetic cardiac control during active coping. Psychophysiology, 24(2), 228-

235.

Gruber, M., & Moore, P. (1997). Galvanic skin response. The Science Teacher, 64(9), 52.

Halpern, B. (2008). Retronasal olfaction. Encyclopedia of Neuroscience Oxford

Academic Press (on-Line), , 297-304.

Hart, S. G., & Staveland, L. E. (1988). Development of NASA-TLX (task load index):

Results of empirical and theoretical research. In Peter A. Hancock and Najmedin

Meshkati (Ed.), Advances in psychology (pp. 139-183) North-Holland.

Heilmann, S., & Hummel, T. (2004). A new method for comparing orthonasal and

retronasal olfaction. Behavioral Neuroscience, 118(2), 412-419.

Herr, K. A., Mobily, P. R., Kohout, F. J., & Wagenaar, D. (1998). Evaluation of the faces

pain scale for use with the elderly. The Clinical Journal of Pain, 14(1), 29-38.

122

Herz, R. S. (2009). Aromatherapy facts and fictions: A scientific analysis of olfactory

effects on mood, physiology and behavior. International Journal of

Neuroscience, 119(2), 263-290.

Heuberger, E., Redhammer, S., & Buchbauer, G. (2004). Transdermal absorption of ()-

linalool induces autonomic deactivation but has no impact on ratings of well-being

in humans.Neuropsychopharmacology, 29(10)

Heuberger, E., Hongratanaworakit, T., Bohm, C., Weber, R., & Buchbauer, G. (2001).

Effects of chiral fragrances on human autonomic nervous system parameters and

self-evaluation. Chemical Senses, 26(3), 281-292.

Hjortskov, N., Rissén, D., Blangsted, A. K., Fallentin, N., Lundberg, U., & Søgaard, K.

(2004). The effect of mental stress on heart rate variability and blood pressure during

computer work. European Journal of Applied Physiology, 92(1-2), 84-89.

Ho, K., Yazan, L. S., Ismail, N., & Ismail, M. (2009). Apoptosis and cell cycle arrest of

human colorectal cancer cell line HT-29 induced by vanillin. Cancer

Epidemiology, 33(2), 155-160.

Holm, L., & Fitzmaurice, L. (2008). Emergency department waiting room stress: Can

music or aromatherapy improve anxiety scores? Pediatric Emergency Care, 24(12),

836-838.

123

Howard, S., & Hughes, B. M. (2008). Expectancies, not aroma, explain impact of

lavender aromatherapy on psychophysiological indices of relaxation in young

healthy women. British Journal of Health Psychology, 13(4), 603-617.

Hoya, Y., Matsumura, I., Fujita, T., & Yanaga, K. (2008). The use of

nonpharmacological interventions to reduce anxiety in patients undergoing

gastroscopy in a setting with an optimal soothing environment. Gastroenterology

Nursing : The Official Journal of the Society of Gastroenterology Nurses and

Associates, 31(6), 395-399.

Hummel, T., & Heilmann, S. (2008). Olfactory event-related potentials in response to

ortho-and retronasal stimulation with odors related or unrelated to

foods. International Dairy Journal, 18(8), 874-878.

Hummel, T., Heilmann, S., Landis, B. N., Reden, J., Frasnelli, J., Small, D. M., &

Gerber, J. (2006). Perceptual differences between chemical stimuli presented

through the ortho‐or retronasal route.Flavour and Fragrance Journal, 21(1), 42-47.

Hunter, L. W., & Thatcher, S. M. (2007). Feeling the heat: Effects of stress, commitment,

and job experience on job performance. Academy of Management Journal, 50(4),

953-968.

Jacob, T. J., Fraser, C., Wang, L., Walker, V., & O'Connor, S. (2003). Psychophysical

evaluation of responses to pleasant and mal-odour stimulation in human subjects;

adaptation, dose response and gender differences. International Journal of

Psychophysiology, 48(1), 67-80.

124

Johnson, A. J., Jenks, R., Miles, C., Albert, M., & Cox, M. (2011). Chewing gum

moderates multi-task induced shifts in stress, mood, and alertness. A re-

examination. Appetite, 56(2), 408-411.

Kadohisa, M., & Wilson, D. A. (2006). Olfactory cortical adaptation facilitates detection

of odors against background. Journal of Neurophysiology, 95(3), 1888-1896.

Kaur, C., & Kapoor, H. C. (2001). Antioxidants in fruits and vegetables–the

millennium’s health. International Journal of Food Science & Technology, 36(7),

703-725.

Keith, R. R., & Smith, D. W. (2012). Rapid olfactory adaptation induced by perithreshold

odorant concentrations in human observers. Chemosensory Perception, 5(2), 172-

178.

Kennedy, D. O., Little, W., & Scholey, A. B. (2004). Attenuation of laboratory-induced

stress in humans after acute administration of melissa officinalis (lemon

balm). Psychosomatic Medicine, 66(4), 607-613.

Kerns, A., Elucidating the absorption and metabolism of linalool to understand its

potential health effects. 2017. Ohio State U, Masters thesis.

Khuda-Bukhsh, A. R., Das, S., & Saha, S. K. (2014). Molecular approaches toward

targeted cancer prevention with some food plants and their products: Inflammatory

and other signal pathways. Nutrition and Cancer, 66(2), 194-205.

125

Kiecolt-Glaser, J. K., Heffner, K. L., Glaser, R., Malarkey, W. B., Porter, K., Atkinson,

C., Laskowski, B., Lemeshow, S., & Marshall, G. D. (2009). How stress and anxiety

can alter immediate and late phase skin test responses in allergic

rhinitis. Psychoneuroendocrinology, 34(5), 670-680.

Kim, S., Kim, H., Yeo, J., Hong, S., Lee, J., & Jeon, Y. (2011). The effect of lavender oil

on stress, bispectral index values, and needle insertion pain in volunteers. The

Journal of Alternative and Complementary Medicine, 17(9), 823-826.

King, S. C., & Meiselman, H. L. (2010). Development of a method to measure consumer

emotions associated with foods. Food Quality and Preference, 21(2), 168-177.

Kirschbaum, C., Pirke, K. M., & Hellhammer, D. H. (1993). The 'trier social stress test'--

a tool for investigating psychobiological stress responses in a laboratory

setting. Neuropsychobiology, 28(1-2), 76-81.

Koshland Jr, D. E., Goldbeter, A., & Stock, J. B. (1982). Amplification and adaptation in

regulatory and sensory systems. Science, 2, 16.

Kopec, R. E., Cooperstone, J. L., Schweiggert, R. M., Young, G. S., Harrison, E. H.,

Francis, D. M., Clinton, S. K., & Schwartz, S. J. (2014). Avocado consumption

enhances human postprandial provitamin A absorption and conversion from a novel

high-beta-carotene tomato sauce and from carrots. The Journal of Nutrition, 144(8),

1158-1166.

126

Kuriyama, H., Watanabe, S., Nakaya, T., Shigemori, I., Kita, M., Yoshida, N., Masaki,

D., Tadai, T., Ozasa, K., Fukui, K., & Imanishi, J. (2005). Immunological and

psychological benefits of aromatherapy massage. Evidence-Based Complementary

and Alternative Medicine : ECAM, 2(2), 179-184.

Kurniawan, H., Maslov, A. V., & Pechenizkiy, M. (2013). Stress detection from speech

and galvanic skin response signals. Proceedings of the 26th IEEE International

Symposium on Computer-Based Medical Systems, 209-214.

Kuroda, K., Inoue, N., Ito, Y., Kubota, K., Sugimoto, A., Kakuda, T., & Fushiki, T.

(2005). Sedative effects of the jasmine tea odor and (R)-(−)-linalool, one of its major

odor components, on autonomic nerve activity and mood states. European Journal

of Applied Physiology, 95(2-3), 107-114.

Kutlu, A. K., Yılmaz, E., & Çeçen, D. (2008). Effects of aroma inhalation on

examination anxiety. Teaching and Learning in Nursing, 3(4), 125-130.

Lagopoulos, J., Xu, J., Rasmussen, I., Vik, A., Malhi, G. S., Eliassen, C. F., Arntsen, I.

E., Sæther, J. G., Hollup, S., & Holen, A. (2009). Increased theta and alpha EEG

activity during nondirective meditation. The Journal of Alternative and

Complementary Medicine, 15(11), 1187-1192.

Lang, P. J., Greenwald, M. K., Bradley, M. M., & Hamm, A. O. (1993). Looking at

pictures: Affective, facial, visceral, and behavioral

reactions. Psychophysiology, 30(3), 261-273.

127

Lawless, H. T. (1987). An olfactory analogy to release from mixture suppression in

taste. Bulletin of the Psychonomic Society, 25(4), 266-268.

Lazarus, R. (1998). From psychological stress to the emotions: A history of changing

outlooks. Personality: Critical Concepts in Psychology, 4, 179.

Lee, J., & Halpern, B. P. (2013). High-resolution Time–Intensity tracking of sustained

human orthonasal and retronasal smelling during natural breathing. Chemosensory

Perception, 6(1), 20-35.

Lehrner, J., Eckersberger, C., Walla, P., Pötsch, G., & Deecke, L. (2000). Ambient odor

of orange in a dental office reduces anxiety and improves mood in female

patients. Physiology & Behavior, 71(1), 83-86.

Lesinski, G. B., Reville, P. K., Mace, T. A., Young, G. S., Ahn-Jarvis, J., Thomas-Ahner,

J., Vodovotz, Y., Ameen, Z., Grainger, E., Riedl, K., Schwartz, S., & Clinton, S. K.

(2015). Consumption of soy isoflavone enriched bread in men with prostate cancer is

associated with reduced proinflammatory cytokines and immunosuppressive

cells. Cancer Prevention Research (Philadelphia, Pa.), 8(11), 1036-1044.

Li, Z., Snieder, H., Su, S., Ding, X., Thayer, J. F., Treiber, F. A., & Wang, X. (2009). A

longitudinal study in youth of heart rate variability at rest and in response to

stress. International Journal of Psychophysiology, 73(3), 212-217.

Li, Z., Snieder, H., Su, S., Ding, X., Thayer, J. F., Treiber, F. A., & Wang, X. (2009). A

longitudinal study in youth of heart rate variability at rest and in response to

128

stress. International Journal of Psychophysiology : Official Journal of the

International Organization of Psychophysiology, 73(3), 212-217.

Liu, S., Lin, T., & Chang, K. (2013). The physical effects of aromatherapy in alleviating

work-related stress on elementary school teachers in taiwan. Evidence-Based

Complementary and Alternative Medicine, 2013

Masaoka, Y., Takayama, M., Yajima, H., Kawase, A., Takakura, N., & Homma, I.

(2013). Analgesia is enhanced by providing information regarding good outcomes

associated with an odor: Placebo effects in aromatherapy? Evidence-Based

Complementary and Alternative Medicine, 2013

Mashukova, A., Spehr, M., Hatt, H., & Neuhaus, E. M. (2006). Beta-arrestin2-mediated

internalization of mammalian odorant receptors. The Journal of Neuroscience : The

Official Journal of the Society for Neuroscience, 26(39), 9902-9912.

McCaffrey, R., Thomas, D. J., & Kinzelman, A. O. (2009). The effects of lavender and

rosemary essential oils on Test‐Taking anxiety among graduate nursing

students. Holistic Nursing Practice: The Science of Health and Healing, 23:2, 88-

93.

McDade, T. W., Rutherford, J., Adair, L., & Kuzawa, C. W. (2010). Early origins of

inflammation: Microbial exposures in infancy predict lower levels of C-reactive

protein in adulthood.Proceedings.Biological Sciences, 277(1684), 1129-1137.

129

Mezzacappa, E. S., Arumugam, U., Chen, S. Y., Stein, T. R., Oz, M., & Buckle, J.

(2010). Coconut fragrance and cardiovascular response to laboratory stress: Results

of pilot testing. Holistic Nursing Practice, 24(6), 322-332.

Mizuno, K., & Watanabe, Y. (2008). Utility of an advanced trail making test as a

neuropsychological tool for an objective evaluation of work efficiency during mental

fatigue. Fatigue science for human health (pp. 47-54) Springer.

Morris, N., Birtwistle, S., & Toms, M. (1995). Anxiety reduction by aromatherapy:

Anxiolytic effects of inhalation of geranium and rosemary. International Journal of

Aromatherapy, 7(2), 33-39.

Motomura, N., Sakurai, A., & Yotsuya, Y. (2001). Reduction of mental stress with

lavender odorant. Perceptual and Motor Skills, 93(3), 713-718.

MPham, W. S., & Siripornpanich, V. (2012). The effects of lavender oil inhalation on

emotional states, autonomic nervous system, and brain electrical activity. J Med

Assoc Thai, 95(4), 598-606.

Norrish, M. I. K., & Dwyer, K. L. (2005). Preliminary investigation of the effect of

peppermint oil on an objective measure of daytime sleepiness. International Journal

of Psychophysiology, 55(3), 291-298.

Nourbakhsh, N., Wang, Y., Chen, F., & Calvo, R. A. (2012). Using galvanic skin

response for cognitive load measurement in arithmetic and reading

130

tasks. Proceedings of the 24th Australian Computer-Human Interaction

Conference, 420-423.

O’Donnell, E., Landolt, K., Hazi, A., Dragano, N., & Wright, B. J. (2015). An

experimental study of the job demand–control model with measures of heart rate

variability and salivary alpha-amylase: Evidence of increased stress responses to

increased break autonomy. Psychoneuroendocrinology, 51, 24-34.

Obayashi, K. (2013). Salivary mental stress proteins. Clinica Chimica Acta, 425, 196-

201.

Obayashi, K., Sato, K., Shimazaki, R., Ishikawa, T., Goto, K., Ueyama, H., Mori, T.,

Ando, Y., & Kumamoto, T. (2008). Salivary chromogranin A: Useful and

quantitative biochemical marker of affective state in patients with amyotrophic

lateral sclerosis. Internal Medicine, 47(21), 1875-1879.

Pemberton, E., & Turpin, P. G. (2008). The effect of essential oils on work-related stress

in intensive care unit nurses. Holistic Nursing Practice, 22(2), 97-102.

Peng, S., Koo, M., & Yu, Z. (2009). Effects of music and essential oil inhalation on

cardiac autonomic balance in healthy individuals. The Journal of Alternative and

Complementary Medicine, 15(1), 53-57.

Peres, J. C. (2012). Autonomic and behavioral reactivity to an acute laboratory stressor.

Pierce, J., & Halpern, B. P. (1996). Orthonasal and retronasal odorant identification

based upon vapor phase input from common substances

131

Pierce-Feldmeyer, Alex M., Simons. “Olfactory adaptation is dependent on route of

delivery.” 2017. Unpublished data.

Pumprla, J., Howorka, K., Groves, D., Chester, M., & Nolan, J. (2002). Functional

assessment of heart rate variability: Physiological basis and practical

applications doi:https://doi.org/10.1016/S0167-5273(02)00057-8

Rattaz, C., Goubet, N., & Bullinger, A. (2005). The calming effect of a familiar odor on

full-term newborns. Journal of Developmental & Behavioral Pediatrics, 26(2), 86-

92.

Raudenbush, B., Meyer, B., & Eppich, B. (2002). The effects of odors on objective and

subjective measures of athletic performance. International Sports Journal, 6(1), 14-

27.

Raudenbush, B., Grayhem, R., Sears, T., & Wilson, I. (2009). Effects of peppermint and

cinnamon odor administration on simulated driving alertness, mood and

workload. North American Journal of Psychology, 11(2), 245-256.

Rho, K., Han, S., Kim, K., & Lee, M. S. (2006). Effects of aromatherapy massage on

anxiety and self-esteem in korean elderly women: A pilot study. International

Journal of Neuroscience, 116(12), 1447-1455.

Roberts, D. D., & Acree, T. E. (1995). Simulation of retronasal aroma using a modified

headspace technique: Investigating the effects of saliva, temperature, shearing, and

oil on flavor release. Journal of Agricultural and Food Chemistry, 43(8), 2179-2186.

132

Robu, S., Aprotosoaie, A. C., Spac, A., Cioanca, O., Hancianu, M., & Stanescu, U.

(2011). Studies regarding chemical composition of lavender volatile oils. [Cercetari

privind compozitia chimica a uleiurilor volatile extrase din flori de levantica] Revista

Medico-Chirurgicala a Societatii De Medici Si Naturalisti Din Iasi, 115(2), 584-

589.

Routledge, C., Arndt, J., Wildschut, T., Sedikides, C., Hart, C. M., Juhl, J., Vingerhoets,

A. J., & Schlotz, W. (2011). The past makes the present meaningful: Nostalgia as an

existential resource. Journal of Personality and Social Psychology, 101(3), 638.

Rozin, P. (1982). “Taste-smell confusions” and the duality of the olfactory

sense. Perception & Psychophysics, 31(4), 397-401.

Sajadieh, A., Nielsen, O. W., Rasmussen, V., Hein, H. O., Abedini, S., & Hansen, J. F.

(2004). Increased heart rate and reduced heart-rate variability are associated with

subclinical inflammation in middle-aged and elderly subjects with no apparent heart

disease. European Heart Journal, 25(5), 363-370.

Sakai, N., Kobayakawa, T., Gotow, N., Saito, S., & Imada, S. (2001). Enhancement of

sweetness ratings of aspartame by a vanilla odor presented either by orthonasal or

retronasal routes. Perceptual and Motor Skills, 92(3c), 1002-1008.

Savic, I., Gulyás, B., & Berglund, H. (2002). Odorant differentiated pattern of cerebral

activation: Comparison of acetone and vanillin. Human Brain Mapping, 17(1), 17-

27.

133

Schoenfeld, T. A., & Cleland, T. A. (2005). Anatomical contributions to odorant

sampling and representation in rodents: Zoning in on sniffing behavior. Chemical

Senses, 31(2), 131-144.

Scholey, A., Haskell, C., Robertson, B., Kennedy, D., Milne, A., & Wetherell, M. (2009).

Chewing gum alleviates negative mood and reduces cortisol during acute laboratory

psychological stress.Physiology & Behavior, 97(3), 304-312.

SELYE, H. (1950). Stress and the general adaptation syndrome. British Medical

Journal, 1(4667), 1383-1392.

Setz, C., Arnrich, B., Schumm, J., La Marca, R., Tröster, G., & Ehlert, U. (2010).

Discriminating stress from cognitive load using a wearable EDA device. IEEE

Transactions on Information Technology in Biomedicine, 14(2), 410-417.

Small, D. M., Gerber, J. C., Mak, Y. E., & Hummel, T. (2005). Differential neural

responses evoked by orthonasal versus retronasal odorant perception in

humans. Neuron, 47(4), 593-605.

Stone, H., Pryor, G. T., & Steinmetz, G. (1972). A comparison of olfactory adaptation

among seven odorants and their relationship with several physicochemical

properties. Perception & Psychophysics, 12(6), 501-504.

Struys, M. M., Jensen, E. W., Smith, W., Smith, N. T., Rampil, I., Dumortier, F. J.,

Mestach, C., & Mortier, E. P. (2002). Performance of the ARX-derived auditory

evoked potential index as an indicator of anesthetic DepthA comparison with

134

bispectral index and hemodynamic measures during propofol administration. The

Journal of the American Society of Anesthesiologists, 96(4), 803-816.

Sun, F., Kuo, C., Cheng, H., Buthpitiya, S., Collins, P., & Griss, M. (2010). Activity-

aware mental stress detection using physiological sensors. International Conference

on Mobile Computing, Applications, and Services, 211-230.

Takeda, H., Tsujita, J., Kaya, M., Takemura, M., & Oku, Y. (2008). Differences between

the physiologic and psychologic effects of aromatherapy body treatment. The

Journal of Alternative and Complementary Medicine, 14(6), 655-661.

Task Force of The European Society of Cardiology and The North American Society of

Pacing and Electrophysiology. (1996). Heart rate variability: Standards of

measurement, physiological interpretation, and clinical use. European Heart

Journal, 17, 354–381.

Thayer, J. F., Åhs, F., Fredrikson, M., Sollers, J. J., & Wager, T. D. (2012). A meta-

analysis of heart rate variability and neuroimaging studies: Implications for heart

rate variability as a marker of stress and health. Neuroscience & Biobehavioral

Reviews, 36(2), 747-756.

Tillett, J., & Ames, D. (2010). The uses of aromatherapy in women's health. The Journal

of Perinatal & Neonatal Nursing, 24(3), 238-245.

Toda, M., & Morimoto, K. (2008). Effect of lavender aroma on salivary endocrinological

stress markers. Archives of Oral Biology, 53(10), 964-968.

135

Toda, M., & Morimoto, K. (2011). Evaluation of effects of lavender and peppermint

aromatherapy using sensitive salivary endocrinological stress markers. Stress and

Health, 27(5), 430-435.

Venkataramana, M., Pratap, K., Padma, M., Kalyan, S., Reddy, A. A., & Sandhya, P.

(2016). Effect of aromatherapy on dental patient anxiety: A randomized controlled

trial. Journal of Indian Association of Public Health Dentistry, 14(2), 131.

Verhagen, J. V. (2015). A role for lung retention in the sense of retronasal

smell. Chemosensory Perception, 8(2), 78-84.

Visschers, R. W., Jacobs, M. A., Frasnelli, J., Hummel, T., Burgering, M., & Boelrijk, A.

E. M. (2006). Cross-modality of texture and aroma perception is independent of

orthonasal or retronasal stimulation. Journal of Agricultural and Food

Chemistry, 54(15), 5509-5515.

Voirol, E., & Daget, N. (1986). Comparative study of nasal and retronasal olfactory

perception. Lebensmittel-Wissenschaft Technologie, 19(4), 316-319.

Vrijkotte, T. G., van Doornen, L. J., & de Geus, E. J. (2000). Effects of work stress on

ambulatory blood pressure, heart rate, and heart rate variability. Hypertension, 35(4),

880-886.

Wechsler, D. (1997). WAIS-III, wechsler adult intelligence scale: Administration and

scoring manual Psychological Corporation.

136

Welge-Lüssen, A., Husner, A., Wolfensberger, M., & Hummel, T. (2009). Influence of

simultaneous gustatory stimuli on orthonasal and retronasal olfaction. Neuroscience

Letters, 454(2), 124-128.

Wetherell, M. A., Hyland, M. E., & Harris, J. E. (2004). Secretory immunoglobulin A

reactivity to acute and cumulative acute multi-tasking stress: Relationships between

reactivity and perceived workload.Biological Psychology, 66(3), 257-270.

Wilder, D. A. (1978). Perceiving persons as a group: Effects on attributions of causality

and beliefs. Social Psychology, , 13-23.

Wuttke, M. S., & Tompkins, L. (2000). Olfactory adaptation in drosophila

larvae. Journal of Neurogenetics, 14(1), 43-62.

Xu, J., Xu, H., Liu, Y., He, H., & Li, G. (2015). Vanillin-induced amelioration of

depression-like behaviors in rats by modulating monoamine neurotransmitters in the

brain. Psychiatry Research, 225(3), 509-514.

Zhao, K., Pribitkin, E. A., Cowart, B. J., Rosen, D., Scherer, P. W., & Dalton, P. (2006).

Numerical modeling of nasal obstruction and endoscopic surgical intervention:

Outcome to airflow and olfaction.American Journal of Rhinology, 20(3), 308-316.

137

Appendix A: NASA Task Load Index

138