Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Emotion regulation and heart rate variability among smokers
(USC Thesis Other)
Emotion regulation and heart rate variability among smokers
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
EMOTION REGULATION AND HEART RATE VARIABILITY AMONG SMOKERS
by
Georgia Christodoulou
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
Preventive Medicine (Health Behavior)
August 2021
Copyright 2021 Georgia Christodoulou
ii
DEDICATION
To my Mom
For her Love and Dedication
iii
ACKNOWLEDGEMENTS
Thank you to my dissertation committee for the innumerable hours of meetings,
discussions, and deliberations. Your guidance and support these past five years have been
substantial to say the least. I will always be appreciative for all I have learned from you.
To Garen, my partner and best friend –Your ambition, confidence, and determination
have taught me that anything is possible. Thank you for your unwavering enthusiasm,
encouragement, and support.
To my dear friends that I met through the program: Anuja–You have been a pillar of
solace and comfort for me. Thank you for your grounding presence, your positive affirmations,
and your infinite patience when answering all my incessant questions. Tobin – You continue to
amaze me with all that you accomplish and your perseverance. Thank you for the countless
snacks and for reminding me to always keep things in perspective.
Lastly, I would like to acknowledge the University of Southern California for offering me
an outstanding education and home for almost a decade of my life. I will be forever grateful.
iv
TABLE OF CONTENTS
DEDICATION ............................................................................................................................... ii
ACKNOWLEDGEMENTS .......................................................................................................... iii
LIST OF TABLES ......................................................................................................................... vi
LIST OF FIGURES ...................................................................................................................... vii
ABSTRACT ................................................................................................................................... ix
CHAPTER 1. INTRODUCTION .................................................................................................... 1
BACKGROUND AND SIGNIFICANCE .................................................................................. 1
Overview of Tobacco Use Disorder ........................................................................................ 1
Using the Transdiagnostic Approach to Target Smoking ....................................................... 1
Emotion Regulation as a Treatment Indicator ......................................................................... 2
Vagally-Mediated Heart Rate Variability (vmHRV) .............................................................. 7
HRV Metrics and Assessment ................................................................................................. 8
Relevant Theory on vmHRV and Emotion Regulation ........................................................... 9
vmHRV: A Transdiagnostic Biomarker ................................................................................ 11
vmHRV in Tobacco Research ............................................................................................... 12
OVERVIEW OF DISSERTATION STUDIES ............................................................................. 13
CHAPTER 2: EMOTION REGULATION AND HEART RATE VARIABILITY AMONG
SMOKERS WITH DEPRESSION (STUDY 1) ............................................................................ 16
INTRODUCTION ..................................................................................................................... 16
METHODS ................................................................................................................................ 20
Participants ............................................................................................................................ 20
Procedure ............................................................................................................................... 21
Experimental Tasks ............................................................................................................... 22
Self-Report Measures ............................................................................................................ 23
Data Capture and Preprocessing ............................................................................................ 26
RESULTS .................................................................................................................................. 29
DISCUSSION ............................................................................................................................ 33
CHAPTER 3: EMOTION REGULATION, HEART RATE VARIABILITY, AND
MINDFULNESS TRAINING AMONG SMOKERS (STUDY 2) ............................................... 47
INTRODUCTION ..................................................................................................................... 47
METHODS ................................................................................................................................ 52
Participants ............................................................................................................................ 52
Procedure ............................................................................................................................... 53
Interventions .......................................................................................................................... 54
Experimental Tasks ............................................................................................................... 55
Self-Report Measures ............................................................................................................ 57
Data Capture and Preprocessing ............................................................................................ 59
RESULTS .................................................................................................................................. 63
v
DISCUSSION ............................................................................................................................ 65
CHAPTER 4: EMOTION REGULATION, HEART RATE VARIABILITY (HRV), AND
RELATIONSHIP SATISFACTION AMONG SMOKERS (STUDY 3) ..................................... 76
INTRODUCTION ..................................................................................................................... 76
METHODS ................................................................................................................................ 80
Participants ............................................................................................................................ 80
Procedure ............................................................................................................................... 81
Self-Report Measures ............................................................................................................ 82
Data Capture and Preprocessing ............................................................................................ 84
RESULTS .................................................................................................................................. 87
DISCUSSION ............................................................................................................................ 89
CHAPTER 5. DISCUSSION AND CONCLUSIONS .................................................................. 98
Limitations ................................................................................................................................. 98
Implications ............................................................................................................................. 100
Future Research Directions ..................................................................................................... 107
Concluding Remarks ............................................................................................................... 111
REFERENCES ............................................................................................................................ 113
vi
LIST OF TABLES
Table 1. Demographics and Sample Characteristics in Study 1………………………...………42
Table 2. vmHRV, DERS, Smoking Characteristics and Smoking Outcomes between Deprived
and Non-Deprived Sessions in Study 1…………….………………………………………...….43
Table 3. Demographics, DERS, vmHRV, and Smoking Characteristics and Outcomes by
Depression Symptom Status in Study 1……………………….……………………………...…44
Table 4. Generalized Estimating Equations (GEE) of HRV Metrics on Smoking Outcomes
[Minutes to Smoke, Number of Cigarettes Smoked] in Study 1………………………………...45
Table 5. Sample Demographics, DERS, Smoking Characteristics, and vmHRV between
Intervention Groups in Study 2………………………………...………………………………...72
Table 6. Paired Samples Analyses of DERS, Smoking Characteristics, and Resting vmHRV in
Study 2……………………………………………………………………………………….…..73
Table 7. vmHRV Metrics and Smoking Outcomes between Intervention Groups during Smoking
Abstinence in Study 2……………………………………………………….…………….…......74
Table 8. Sample Demographics, DERS, Smoking Characteristics, vmHRV, and Smoking
Outcomes by Relationship Satisfaction Status in Study 3……………………………………….95
Table 9. Pairwise Associations between Baseline DERS and vmHRV Metrics at Baseline in
Study 3………...…………………………………………………………………………….….. 96
Table 10. Predictive Modeling of DERS and vmHRV on Cigarette Use Variability in Study
3…….……………………………………………………………………………….……………97
vii
LIST OF FIGURES
Figure 1. Conceptual Model of Dissertation Studies 1-3………………………………...……...15
Figure 2. Procedure for Experimental Sessions (Deprived and Non-Deprived of Smoking) in
Study 1…………………………………………………………………………………………...41
Figure 3. Interaction of DERS x Session Type on Number of Cigarettes Smoked in Study 1
………………………………………………………………………………………………........46
Figure 4. Procedure for Baseline and Experimental Sessions in Study 2……………………….71
Figure 5. vmHRV Reactivity to TSST between Intervention Groups in Study 2……………….75
viii
ABSTRACT
Tobacco use disorder remains a major public health issue within the United States.
Cessation efforts that incorporate transdiagnostic indicators, such as difficulties in emotion
regulation, can help identify particularly vulnerable smokers. The purpose of this dissertation
was to determine whether vagally-mediated heart rate variability (vmHRV) could serve as a
reliable, transdiagnostic biomarker of emotion regulation, capable of predicting smoking
behavior outcomes. This dissertation aimed to (1) test the linear association between difficulties
in emotion regulation (DERS) and vmHRV (Study 1-3) during various experimental paradigms
(induction of emotion, induction of acute social stress, and at rest) to validate vmHRV as an
objective marker of emotion regulation and (2) to test whether DERS and vmHRV could predict
smoking behavior outcomes in laboratory (Studies 1 & 2) and naturalistic settings (Study 3). For
Aim 1, all studies revealed non-significant findings regarding the main effect of DERS on
vmHRV (Studies 1-3). For Aim 2, generalized estimating equation (GEE) modeling in Study 1
revealed a significant interaction effect of DERS and experimental session type (deprived of
smoking versus not deprived of smoking) on number of cigarettes smoked during a smoking
reinstatement task. Additionally, GEE results in Study 1 demonstrated that the odds of smoking
during a smoking reinstatement task versus waiting to smoke until the task was complete was
less for every unit increase in vmHRV during rest and during the experimental film task. Studies
2 and 3 produced null findings surrounding main effects of DERS and vmHRV on smoking
behavior outcomes, most likely due to a small sample sizes (N~20) and a lack of power to detect
statistically significant effects. Nonetheless, preliminary results from Studies 2 and 3 point to the
utility of vmHRV reactivity for future investigations and offer a foundation of knowledge
surrounding methodology and best practices in vmHRV assessment. The results of this
ix
dissertation offer great insight to the current literature and emphasize the potential predictive
value of vmHRV on smoking behavior outcomes, increasing its applicability and utility in
research studies that target tobacco addiction.
1
CHAPTER 1. INTRODUCTION
BACKGROUND AND SIGNIFICANCE
Overview of Tobacco Use Disorder
Tobacco use disorder is a major public health issue in the United States (Hatsukami,
Stead, & Gupta, 2008). Tobacco use disorder has been clinically defined by the Diagnostic
Statistical Manuel of Mental Disorders (DSM-5) as problematic tobacco use (presence of at least
two listed symptoms) that leads to clinical implications and distress. Rates of smoking decline
and cessation attempts among those diagnosed with psychiatric disorders are notably lower than
non-clinical populations (Benowitz, 2010; Centers for Disease & Prevention, 2013; Farris,
Aston, Zvolensky, Abrantes, & Metrik, 2017; Kinnunen, Doherty, Militello, & Garvey, 1996;
Weinberger et al., 2017). The presence of psychopathology and related symptomology, such as
anxiety and depression, may worsen smoking behavior. Reports of increased smoking demand
intensity and more severe withdrawal symptoms are prevalent among this subgroup (Farris et al.,
2017; Weinberger et al., 2017). Current health promotion efforts have focused not only on
preventing future use of tobacco products but also on cessation efforts among current users.
Identifying underlying factors that drive tobacco use is an essential step to creating interventions
that can counteract maladaptive smoking behaviors. Furthermore, identifying these factors
among individuals with psychopathology and related symptomology is a priority, as they
represent a vulnerable population that is more susceptible to poor health outcomes.
Using the Transdiagnostic Approach to Target Smoking
Targeting smoking behaviors in the context of psychiatric conditions and symptomology
has inherent challenges due to heterogeneity in symptom experience, even among the same
2
diagnostic category (Leventhal & Zvolensky, 2015). Research is needed to reveal factors that
may reinforce smoking that move beyond traditional syndrome-based classification systems,
such as the Diagnostic and Statistical Manual of Mental Disorders (DSM), in which individuals
are diagnosed based on categorical designation that can narrow their treatment options (Garland,
2014). Differences between normality and psychopathology are not clearly defined within these
systems, and rigid definitions of the latter may exclude individuals who meet subthreshold levels
of these disorders but may be suffering from debilitating symptoms (Cross & Hickie, 2017). The
transdiagnostic movement in mental health has garnered increasing attention due to its feasibility
within the clinical field, in which mental health providers can personalize their interventions by
treating an array of transdiagnostic symptoms versus adhering to treatment plans dictated solely
by clinical diagnosis (Garland, 2014). This approach, endorsed by the National Institutes of
Mental Health (NIMH), not only offers individualized treatment designed to target comorbid
symptomology across disorders but also provides the foundation for identifying at-risk
individuals (Fernandez, Jazaieri, & Gross, 2016). Researchers have recently begun to focus on
underlying symptomology of comorbid disorders in an effort to broaden the way these conditions
are defined and thus treated (Leventhal & Zvolensky, 2015). Thus, adopting a transdiagnostic
framework can provide an innovative opportunity to better understand the underlying
mechanisms driving smoking behaviors among individuals with and without a psychiatric
diagnosis, who may experience a wide range of pathological symptoms.
Emotion Regulation as a Treatment Indicator
Both the affective model of drug motivation and incentive learning theory emphasize the
important role of negative affect and the ability to regulate emotional states in the initiation and
3
maintenance of tobacco use disorder (Baker, Piper, McCarthy, Majeskie, & Fiore, 2004;
Mathew, Hogarth, Leventhal, Cook, & Hitsman, 2017). Negative affect (defined as negative
emotional states such as anxiety, irritability, sadness, etc.) is considered an important precursor
to smoking behavior among individuals who suffer from tobacco use disorder and depression.
Each of these theories offers insight on how negative affect may interact with other static (low
positive affect in depression) and temporal variables (deprived versus non-deprived conditions)
to produce poor smoking outcomes. Furthermore, these indicators become particularly important
among those who suffer from comorbid psychopathology, such as depression and tobacco use
disorder, as these individuals may have inherent deficits (such as cognitive impairment) that
affect their ability to tolerate perceived distressful states.
The affective model of drug motivation supports the reinforcing role of negative affect in
the context of smoking abstinence as a major motivator in the maintenance of tobacco use.
Although individuals typically initiate smoking for the positive and rewarding effects of nicotine,
sustained use and threat of relapse primarily occurs in an effort to reduce the negative symptoms
associated with withdrawal (Baker et al., 2004). According to the affective model of drug
motivation, negative affect can augment preconscious attentional bias to smoking cues. More
specifically, the presence of negative affect can lead to three types of attentional bias: (1)
positively reinforced behaviors have greater reward-producing effects (such as smoking) and the
behavior (smoking) is repeated; (2) unconscious bias to negative cues trigger autonomic behavior
responses and downregulate cognitive processes that may help deter problematic behaviors
(smoking); and (3) reduced value of alternative reinforcers in the presence of the drug (Baker et
al., 2004). The effects of attentional biases may become even more pronounced during smoking
abstinence and when an individual is experiencing negative affect associated with withdrawal
4
(anxiety, irritability, etc.). The presence of comorbid conditions, such as depression, may further
augment these overall attentional biases and the subjective experience of withdrawal symptoms.
In support of this theory, smokers with history of depression show more attentional bias and
negative reinforcement expectancies related to smoking and its ability to reduce negative
symptomology (Weinberger, George, & McKee, 2011). Thus, depressed smokers represent a
particularly vulnerable population that may be more sensitive to the effects of negative affect in
the context of withdrawal.
The very recently published incentive learning theory offers a more nuanced explanation
of the association between depression and smoking behavior by expanding on the importance of
negative affect regulation in the context of depression and environmental circumstances. Moving
away from traditional negative reinforcement models of behavior, the incentive learning theory
proposes that both effects of internal (automatic, subconscious bias) and external motivational
states combine to produce heightened risk for smoking behaviors. These motivational states
include low positive affect (low energy, reduced pleasure), high negative affect (irritability,
anxiety, sadness), and cognitive impairment [deficits in attention, memory, and executive
functioning (encompassing emotion regulation)] (Mathew et al., 2017). In tandem with goal-
directed knowledge of the rewarding effects of nicotine in various environmental contexts, these
motivational states form a conscious inference that reinforces the predicted positive effects of
smoking (augments smoking desire) and increases the likelihood of repetitive use and addiction
(smoking behavior). Low positive affect and increased cognitive deficits (characteristics of
depression) are introduced as important indicators for smoking behavior along with negative
affect. Increased negative affect and decreased positive affect are theorized to be two distinct
biological constructs that collectively worsen in the context of abstinence among smokers with
5
depression versus smokers without depression (Audrain-McGovern, Wileyto, Ashare, Cuevas, &
Strasser, 2014). The role of cognitive impairment among individuals who suffer from depression
becomes an especially prominent indicator and can resemble the cognitive deficits demonstrated
by non-psychiatric smokers during acute withdrawal. Furthermore, depressed smokers
demonstrate worse cognitive performance during withdrawal compared to non-depressed
smokers, indicating that the combination of these motivational states and environmental context
(smoking deprivation) may worsen smoking outcomes for smokers with depression compared to
smokers without depression.
Emotion regulation has become an important transdiagnostic feature across various
psychopathologies and represents a broad psychological construct, typically defined as the ability
to modulate emotions, and regulate reactivity in response to emotional experiences (Gross &
Thompson, 2007(Fernandez et al., 2016; Sloan et al., 2017). Difficulties in emotion regulation
(defined as the application of maladaptive emotion regulation strategies in the face of negative
affect) can be prevalent among both clinical and non-clinical populations (Aldao, Nolen-
Hoeksema, & Schweizer, 2010); however, more severe deficits are typically reported among
those suffering from psychiatric conditions, such as depression. Difficulties in emotion
regulation have been conceptualized into six distinct factors, including: (1) inattention to or lack
of emotional awareness of negative emotions (AWARE); (2) non-acceptance of negative affect
or non-willingness to accept negative affect (NON-ACCEPTANCE); (3) difficulty initiating
goal-directed behavior to modulate the intensity or duration of negative affect (GOALS); (4)
reduced capability to initiate strategies to counteract negative affect (STRATEGIES); (5)
difficulties in controlling impulsive behavior when experiencing negative affect (IMPULSE);
and lack of clarity or understanding of negative affective states (CLARITY)(Gratz, 2004).
6
The difficulties in emotion regulation scale (DERS), which assesses the aforementioned
six distinct factors with 36 self-report items, has become a broadly used, self-report measure
within research and has shown associations with biological neural correlates of emotion
regulation (Li et al., 2008) and laboratory tasks designed to test distress tolerance (Tull, 2010).
The DERS has also been translated to multiple languages, demonstrates good test re-test
reliability at two-week follow-up, and shows high internal consistency and construct validity
across various studies (Moreira, 2020). Due the length of the originally published DERS (36-
item measure), a shortened version of the scale (an 18-item measure) has since been published
and validated for use in research studies that seek to reduce participant burden (Moreira, 2020).
Additionally, a state-based version of the DERS (21 items) was recently published to assess
state-based difficulties in emotion regulation as opposed to the trait-based items of the original
DERS (Lavender, Tull, DiLillo, Messman-Moore, & Gratz, 2017). While changes have been
made to the original 36-item factor structure (including elimination of the AWARE items in the
DERS-SF due to poor factor loading as well as several examinations of various alternative factor
structures to the original six-factor scale), researchers recommend using the total score on the
DERS as an index of overall emotion regulation difficulties (Osborne, Michonski, Sayrs, Welch,
& Anderson, 2017) .
While there has been a wide range of studies that have focused on the associations
between various psychiatric conditions and difficulties in emotion regulation (Mennin & Fresco,
2015), there have only been a few investigations that have assessed theses associations in the
context of tobacco use disorder. Among smokers with depression, lack of emotional acceptance
was significantly associated with recent smoking (Adams, Tull, & Gratz, 2012). Moreover, the
interaction of difficulties in emotion regulation and the presence of psychopathology can
7
increase odds of smoking relapse during cessation attempts (Farris, Zvolensky, & Schmidt,
2016). Among a non-clinical sample, individuals who reported difficulties in emotion regulation
showed increased craving and attentional bias to smoking cues (Szasz, Szentagotai, & Hofmann,
2012). All of the aforementioned studies used self-report measures and did not incorporate
objective measures of emotion regulation to corroborate their findings. A general limitation
found across these studies is an overreliance on self-report measures that are subject to bias. This
bias is especially prominent among psychiatric populations, who are prone to alexithymia
(inability to identify and label feelings) and demonstrate limited self-awareness (Grabe, Spitzer,
& Freyberger, 2004; Moeller & Goldstein, 2014). Thus, the incorporation of objective
biomarkers of emotion regulation in tobacco research is integral to strengthening the field and
identifying appropriate treatment targets.
Vagally-Mediated Heart Rate Variability (vmHRV)
Vagally-mediated heart rate variability (vmHRV) may represent a promising, inexpensive
biomarker of overall health that is linked to self-regulatory capacity, including emotion
regulation. HRV denotes the variability in time between heartbeats and corresponds to activity of
the vagus nerve (McCraty & Shaffer, 2015; Shaffer & Ginsberg). The vagus nerve is largely
responsible for the parasympathetic nervous system and its influence over the heart (McCraty &
Shaffer, 2015; Shaffer & Ginsberg). As opposed to the sympathetic nervous system, which
controls organs and bodily functions when an individual is facing a task or stressor, vagal tone
(referring to the functioning and activity of the vagus nerve) is activated when an individual is
calm and at rest. vmHRV represents an objective marker of vagal tone with higher variability
typically representing efficient autonomic regulation; the system effectively responds to the
8
changing environment and maintains a state of homeostasis (marked by overall parasympathetic
nervous system dominance). Lower vmHRV, in turn, represents sympathetic nervous system
dominance and disruption in homeostasis caused by a variety of individual and environmental
influences (McCraty & Shaffer, 2015).
HRV Metrics and Assessment
There are approximately 70 HRV available metrics and the decision of choosing which
metric to report in research largely resides on the theoretical understanding of research constructs
and proposed hypotheses (Laborde, Mosley, & Thayer, 2017). Recommendations from the
European Task Force of Cardiology and the North American Society of Pacing and
Electrophysiology published in 1996 provide guidelines for the assessment and analysis of HRV
in research (Task Force, 1996). Laborde and colleagues (2017) recently outlined guidelines for
planning, conducting, and reporting HRV results, offering clear and updated recommendations
for HRV assessment (Laborde et al., 2017). Short-term measurements of HRV can be divided
into three categories of metrics: time-domain, frequency-domain, and non-linear variables. While
time-domain metrics are used to measure variability between heart beats, frequency-domain
metrics denote frequency bands that comprise the HRV signal and represent activity of various
systems of the ANS (vagal tone versus SNS dominance). While there is still some controversy
surrounding metrics of non-linear indices, time-domain and frequency-domain variables have
gained validity within research, particularly vmHRV metrics that correspond to vagal tone and
PNS dominance, including the root mean square of successive differences (RMSSD) and the
high-frequency HRV band (HF-HRV)(0.15-0.4Hz)(Laborde et al., 2017). In addition to
representing vagal tone, lower RMSSD and HF-HRV values are highly associated with
9
decreased cognitive functioning and difficulties in emotion regulation (Segerstrom & Nes, 2007;
Thayer, Hansen, Saus-Rose, & Johnsen, 2009; Thayer & Lane, 2000; Williams et al., 2015).
These metrics and their unique associations have thus broadened the utility of vmHRV as a
biomarker for emotion regulatory capacity in psychophysiological research (Laborde et al.,
2017).
Relevant Theory on vmHRV and Emotion Regulation
There have been many theoretical models introduced to explain the association between
emotion regulation and vmHRV, most notably the neurovisceral integration model, the vagal
tank theory, and the polyvagal theory. The neurovisceral integration model proposed by Thayer
and Lane (2000) explains the intricate neurobiological underpinnings linking the brain to the
heart, labeled the central autonomic network (CAN)(Thayer & Lane, 2000). The CAN represents
a network of executive neural structures (including the insula, anterior cingulate cortex, and
prefrontal cortices) that connect to the heart and control goal-directed behavior, necessary for
adaptation and autonomic regulatory control (Thayer & Lane, 2000). The systems involved in
the CAN are complex and interconnected, allowing for reciprocal feedback loops necessary to
inhibit and disinhibit bodily processes in response to environmental demands. The output of the
CAN and its activities are reflected by vmHRV; higher vmHRV is associated with a more
efficient network, improved emotion regulation, and effective autonomic control in the face of
environmental demands (Thayer, Ahs, Fredrikson, Sollers, & Wager, 2012; Thayer et al., 2009;
Thayer & Lane, 2000). Conversely, lower vmHRV is associated with decreased autonomic
flexibility among these systems, indicating poor adaptability to the environment and worse
overall health outcomes (Thayer & Lane, 2000).
10
The vagal tank theory builds upon the neurovisceral integration model by broadening the
context of how vmHRV changes depending on the environmental context (Laborde, Mosley, &
Mertgen, 2018) 2018). Laborde and colleagues (2018) describe the vagus nerve as a “tank” full
of self-regulatory resources controlled by the CAN and represented by vmHRV. The fullness of
the tank, as a representation of autonomic adaptability, largely depends on the environmental
context (specifically during rest, reactivity to a stressor or environmental demand, and recovery).
For example, higher vmHRV values during rest usually denote better self-regulation. During
tasks that only necessitate physical exertion (such as exercise), an observed decrease in vmHRV
from baseline (sympathetic nervous system dominance) represents an appropriate response to
environmental demands and thus represents efficient adaptation (Laborde et al., 2018). During
tasks that necessitate a low level of physical activity and engagement of executive functioning
(such as a mental arithmetic task and/or social stress task), an increase in vmHRV from baseline
(remaining calm while performing the task) represents a more efficient system, sustained
activation of higher order neural networks of the CAN, and autonomic control (Laborde et al.,
2018). Thus, the vagal tank theory provides a foundation for more informed hypotheses within
studies that incorporate vmHRV assessment in the context of adaptation to stress and emotion
regulation.
The polyvagal theory by Porges (2007) discusses the evolutionary value of the vagus
nerve toward social behavior and communication (Porges, 2007). Porges discusses vmHRV as a
proxy for the vagus nerve and its functioning. Efficient self-regulation allows the vagal nerve to
“break” when confronted with environmental demands that necessitate physical
resources/activation of the sympathetic nervous system and trigger a “fight or flight” response.
During moments of rest and relaxation (parasympathetic nervous system dominance), self-
11
regulatory resources can be redirected toward social mobilization and adaptive behaviors
necessary to survive in a social environment (Porges, 2007). In situations that do not require
physical resources and sympathetic nervous system activation, the vagus nerve is dominant and
reflected by higher vmHRV values. This “mode” enables prosocial behavior and spontaneous
social engagement, in which individuals are able to shift the focus of their autonomic resources
toward navigating, building, and reinforcing social bonds (Porges, 2007). The polyvagal theory
emphasizes the evolutionary role of vmHRV in a social context, emphasizing both its
intrapersonal and interpersonal value in maintaining relationships that are necessary for the
survival of the species.
vmHRV: A Transdiagnostic Biomarker
Thayer and Beauchaine (2015) have recently introduced the applicability of high
frequency HRV (HF-HRV) as a transdiagnostic biomarker of psychopathology due to its
association with emotion regulation across multiple disorders, including depression, anxiety
disorders, post-traumatic stress disorder, substance use disorders, etc. (Beauchaine & Thayer,
2015). The utility of HF-HRV in the context of psychopathology has been broad and research
has focused on its assessment in relation to indicators of psychopathology, including both
internalizing (feelings of anxiety, depression, and social withdrawal) and externalizing symptoms
(bullying, aggressive behavior)(Beauchaine & Thayer, 2015). During moments of stress, HF-
HRV is more affected by the presence of both types of symptoms. Thus, HF-HRV, as an
objective indicator, can adequately represent the severity and range of pathological symptoms
(Beauchaine & Thayer, 2015). During rest and in response to tasks that induce stress, HF-HRV is
able to demonstrate difficulties in emotion regulation and maladaptive strategies in the face of
12
negative affect and distress that characterize these conditions, making it a useful, objective
biomarker for studies that investigate these deficits among vulnerable populations.
vmHRV in Tobacco Research
A limited number of studies have assessed the association between vmHRV metrics and
smoking (Ashare et al., 2012; Bodin et al., 2017; Ferdous M, 2014; Libby, Worhunsky, Pilver, &
Brewer, 2012; Murgia et al., 2019), and hardly any investigations have examined vmHRV in the
context of psychopathology and related symptoms (Taylor et al., 2011)(Harte et al., 2013).
Generally, smoking leads to reduction in HRV with smokers demonstrating lower values of
several HRV metrics (SDNN, RMSSD, HF-HRV, etc.) compared to non-smokers (Murgia et al.,
2019). Harte and colleagues found that smokers with clinical depression had significantly lower
HRV (HF-HRV, Low Frequency HRV, etc.) compared to non-depressed smokers (Harte et al.,
2013), which represents one of the only studies that has systematically looked at the association
between smoking and HRV in the context of depression. Clearly more studies are needed to
understand the association between specific vmHRV metrics and smoking outcomes. Drawing
from the neurovisceral integration model and vagal tank theory, investigating the association
between vmHRV and smoking behavior outcomes during stress or induction of emotion may
better elucidate the predictive properties of vmHRV on smoking behavior. Additionally, drawing
from the polyvagal theory, examining the association of difficulties in emotion regulation and
vmHRV in the context of social relationships and social functioning may reveal its potential as
an objective biomarker of both intrapersonal and interpersonal well-being, offering a great
contribution to the field of tobacco use disorder.
13
OVERVIEW OF DISSERTATION STUDIES
Using the transdiagnostic framework and aforementioned vmHRV theories of emotion
regulation, this dissertation aimed to: (1) test the association between difficulties in emotion
regulation (DERS) and vmHRV (during induction of emotion, induction of acute social stress,
and at rest) to validate vmHRV as an objective marker of emotion regulation and (2) test whether
DERS and vmHRV could predict smoking behavior outcomes within laboratory (Study 1 & 2)
and naturalistic settings (Study 3). The dissertation also aimed to elucidate whether these
associations were moderated by predefined groups [smokers with high depression symptoms
versus low depression symptoms (Study 1); smoking deprivation status (Study 1); smokers
undergoing mindfulness training versus an active control condition (Study 2); and smokers in
satisfactory versus unsatisfactory romantic relationships (Study 3)]. Findings would reveal
whether vmHRV can serve as a reliable, transdiagnostic, objective biomarker of emotion
regulation, capable of predicting smoking behavior outcomes and identifying at-risk smokers.
Study 1 aimed to determine if the DERS and vmHRV during emotionally valanced film
clips could predict smoking behavior outcomes during a smoking reinstatement task (number of
cigarettes smoked and minutes to smoke) and whether these associations were moderated by
depression symptom status and smoking deprivation status (participants underwent two
experimental sessions: one in which they were deprived of smoking and another session in which
they were not deprived). Smokers with high depression symptoms were expected to demonstrate
lower vmHRV during the film clip task and worse smoking behavior outcomes during the
smoking reinstatement task under smoking deprivation compared to smokers with low
depression symptoms. Study 2 aimed to test the association between DERS and vmHRV during
an acute social stressor task among smokers randomly assigned to mindfulness training versus an
14
active control condition. This study also aimed to test whether vmHRV and DERS could predict
smoking behavior outcomes during a smoking reinstatement task following induction of acute
social stress (number of cigarettes smoked, minutes to smoke, and craving) and whether these
associations were moderated by intervention group status (mindfulness training versus the active
control condition). Study 3 aimed to determine if the DERS and vmHRV at rest could predict
smoking behavior outcomes in real-world settings using ecological momentary assessment
(EMA) and whether these associations were moderated by relationship satisfaction status
(smokers that in satisfactory versus unsatisfactory romantic relationships). Smokers with
increased DERS scores and lower vmHRV at baseline were expected to demonstrate worse
smoking behavior outcomes (cigarette use variability) at follow-up (28 days) using EMA
methodology.
No studies to date have assessed the association between the DERS and vmHRV and
their respective main effects on smoking behavior outcomes under a single hypothesis. These
dissertation studies will demonstrate whether vmHRV can be used as a biomarker of emotion
regulation, both as a stable indicator during rest and/or representative of emotion regulation
capability during emotion or stress-based tasks. This dissertation will also reveal whether
vmHRV can predict smoking behavior outcomes in laboratory (smoking reinstatement task) and
naturalistic conditions (via EMA methodology), offering internal and external validity to the
study findings. The main objective of this dissertation is to validate the utility of vmHRV in
future tobacco research studies that aim to incorporate viable biomarkers and target vulnerable
individuals who may be at elevated risk of smoking relapse given their emotion regulation
deficits.
15
Figure 1. Conceptual Model of Dissertation Studies 1-3
16
CHAPTER 2: EMOTION REGULATION AND HEART RATE VARIABILITY AMONG
SMOKERS WITH DEPRESSION (STUDY 1)
INTRODUCTION
Tobacco use disorder among individuals with depression is a major public health concern
(Mathew et al., 2017). Smokers with depression represent a vulnerable population, who report
more adverse health outcomes compared to psychiatrically healthy smokers, including increased
mortality and morbidity (Mathew et al., 2017). Past research has shown that the presence of
psychopathology can exacerbate smoking behaviors, such as augmented smoking demand,
withdrawal symptoms, and incidents of smoking relapse (Farris et al., 2017; Weinberger et al.,
2017). Cessation attempts among smokers with depression are notably lower than smokers
without depression, and those who attempt to quit smoking fail more readily (Centers for Disease
& Prevention, 2013; Kinnunen et al., 1996). Research efforts to assess the underlying
mechanisms driving the association between depression and smoking behaviors is integral to
targeting poor health outcomes and developing successful cessation programs for smokers with
depression.
There has been limited theory to describe the comorbidity between depression and
tobacco use disorder within the literature, which may be a contributing factor as to why so few
interventions have targeted smoking behaviors among smokers with depression. The very
recently published incentive learning theory offers a specific and detailed explanation for the
association between depression and smoking behavior. The incentive learning theory highlights
greater expectations regarding the value of smoking among smokers with depression, particularly
in the context of three internal states that act as motivators of behavior, including low positive
affect (low energy, reduced pleasure), high negative affect (irritability, anxiety, sadness), and
cognitive impairment (deficits in attention, memory, and executive functioning)(Mathew et al.,
17
2017). In the face of negative life circumstances, these internal states, coupled with goal directed
knowledge of nicotine reinforcement in various external contexts, form a conscious inference
that reinforces the predicted, positive effects of smoking (augments smoking desire) and
increases the likelihood of repetitive use and addiction. The incentive learning theory emphasizes
the importance of emotion regulation (the ability to monitor and modulate emotional experiences
such as negative affect) as an underlying mechanism of the association between depression and
tobacco use disorder.
Driven by these theories, researchers have begun to investigate difficulties in emotion
regulation (referring to maladaptive emotion regulation strategies) among smokers with
psychopathology as a prominent factor in sustained smoking behaviors (Adams et al., 2012;
Farris et al., 2016; Fucito, Juliano, & Toll, 2010; Gehricke et al., 2007). Adams and colleagues
found that non-acceptance of negative emotions mediated the association between depression
and recent smoking among a group of smokers in a residential substance use treatment program
(Adams et al., 2012). Additionally, adult smokers with mild or greater depressive symptoms
showed augmented attentional bias to smoking cues (Fucito et al., 2010). Researchers found that
smokers with lower self-reported difficulties in emotion regulation were less likely to relapse
early when trying to quit smoking; however, this effect was not observed among smokers with
past-year psychopathology (Fucito et al., 2010). This finding may be indicative of possible bias
related to self-report measures of difficulties in emotion regulation. Researchers posit that
perceived difficulties in emotion regulation may differ from actual difficulties in monitoring and
regulating distress and negative affect (Fucito at al., 2010). Thus, the use of objective
biomarkers that are associated with emotion regulation and smoking behavior outcomes may be
integral to better understanding the interrelationships of these research constructs.
18
A potential objective biomarker of emotion regulation is vagally-mediated heart rate
variability (vmHRV), representing the variability between heart beats and corresponding to the
activity of the vagus nerve (Laborde et al., 2017; Thayer et al., 2009). The vagus nerve is largely
responsible for the parasympathetic nervous system, which is activated when an individual is not
facing an environmental stressor or physical demand (Laborde et al., 2017; Thayer et al., 2009).
Higher vmHRV values represent a more efficient autonomic nervous system in which the body is
able to efficiently respond to environmental stressors and maintain overall parasympathetic
nervous system dominance. The polyvagal theory by Porges (2007) discusses the evolutionary
value of the vagus nerve, represented by vmHRV, toward emotion regulation and adaptive social
behavior (Porges, 2007). Efficient self-regulation allows the vagal nerve to “break” when
confronted with environmental demands that necessitate physical resources/activation of the
sympathetic nervous system and trigger a “fight or flight” response. During moments of rest and
relaxation (parasympathetic nervous system dominance), self-regulatory resources can be
redirected toward adaptive emotion regulation and social mobilization necessary to survive in a
social environment (Porges, 2007). In situations that do not require physical resources and
sympathetic nervous system activation, the vagus nerve is dominant and reflected by higher
vmHRV values. This “mode” enables prosocial behavior and spontaneous social engagement, in
which individuals are able to shift the focus of their autonomic resources toward navigating,
building, and reinforcing social bonds (Porges, 2007).Higher vmHRV values at rest have been
correlated with greater cognitive capacity, self-regulatory strength, and improved emotion
regulation (including reduced negative emotional arousal)(Thayer et al., 2009; Williams et al.,
2015). Conversely, lower vmHRV values have been associated with various psychiatric
conditions, including depression, anxiety disorders, and tobacco use disorder (Beauchaine &
19
Thayer, 2015; Murgia et al., 2019). Thus, vmHRV (representing the vagus nerve and
parasympathetic nervous system dominance) may be an informative and viable objective
biomarker, capable of demonstrating the interrelationships between depression, emotion
regulation, and smoking behaviors.
There have been a limited number of studies that have assessed the association between
vmHRV metrics and smoking outcomes, particularly in the context of tobacco abstinence
(Ashare et al., 2012; Bodin et al., 2017; Ferdous M, 2014; Libby et al., 2012; Murgia et al.,
2019). Researchers have generally found that smokers demonstrate lower vmHRV compared to
non-smokers. Hardly any investigations have assessed this association in the context of
psychopathology or related symptomology (Taylor et al., 2011)(Harte et al., 2013). In a cross-
sectional study, Harte and colleagues (2003) found that smokers with clinical depression had
significantly lower vmHRV compared to non-depressed smokers with effect sizes (eta-squared)
ranging from .11 to .17, which represents one of the only studies that has systematically looked
at the association between smoking and vmHRV in the context of depression. Taylor and
colleagues (2011) found that vmHRV weakly mediated the association between depression and
smoking; however, results were based on cross-sectional data and so linear associations must be
interpreted with caution. No other longitudinal studies to date have assessed the predictive
properties of vmHRV on smoking behavior outcomes among smokers with depression in the
context of smoking deprivation.
The current study. vmHRV assessment was added to the parent study to (1) to test the
main effect of difficulties in emotion regulation at baseline on vmHRV during an emotionally-
valanced film clip task to validate vmHRV as an objective biomarker of emotion regulation and
(2) to test the respective main effects of difficulties in emotion regulation and vmHRV on
20
smoking behavior outcomes during a smoking reinstatement task. Moderation effects of
depression symptom level and smoking abstinence on these aims were also examined. The
following hypotheses were tested:
1. Smokers with higher self-reported difficulties in emotion regulation are expected to
demonstrate lower vmHRV in response to an emotionally-valanced film clip task
2. Smokers with higher self-reported difficulties in emotion regulation and lower
vmHRV are expected to demonstrate poorer smoking behavior outcomes (including
less time to initiate smoking and an increased number of smoked cigarettes) during a
smoking reinstatement task
3. These hypothesized linear associations are expected to be more pronounced among
smokers with high depression symptoms in smoking abstinence, indicative of
moderation effects by depression symptom status and experimental session type
(deprived versus non-deprived of smoking cigarettes)
Study findings may validate vmHRV as an objective biomarker of emotion regulation that can
help differentiate particularly vulnerable smokers. The results will also demonstrate the
predictive properties of difficulties in emotion regulation and vmHRV metrics on smoking
behavior outcomes in the context of psychopathology and abstinence, offering great insight to
the current literature and the development of smoking cessation programs.
METHODS
Participants
Community-based cisgender cigarette smokers with and without depression (N=76, 38%
female) were recruited via locally distributed advertisements in Los Angeles for the parent study.
21
Fifty-six participants of the parent study provided vmHRV data. Of those, three participants’
data had to be excluded from the analyses (due to noisiness (too many artifacts) and a device
malfunction), which resulted in a total sample of N=53 participants. Inclusion criteria included
the following: (1) greater than or equal to 21 years of age; (2) smoked regularly for the past year;
(3) smoke 5+ cigarettes/day; and (4) fluent in English. Exclusion criteria included: (1) current
Diagnostic and Statistical Manual (DSM) 5 dependence for substances other than nicotine; (2)
bipolar disorder in the lifetime; (3) current PTSD or psychotic disorder; (4) breath Carbon
Monoxide (CO) levels < 8ppm at screening session (to confirm smoking status); (5) current use
of nicotine replacement or medications for smoking cessation; (6) regular use of non-cigarette
tobacco and nicotine products; (7) current pregnancy or intent to be pregnant; (8) current or
planned cessation attempt.
Procedure
Participants were screened by research personnel via telephone. Following this initial
telephone screening, participants came to the lab to complete a baseline visit that included
informed consent procedures, a psychiatric interview, alcohol and CO breath testing, and
baseline survey measures. Each participant was then scheduled for two counterbalanced
experimental sessions: (1) the deprived session: participants remained abstinent from smoking 16
hours prior to the session and (2) the non-deprived session: participants were instructed to smoke
normally, as shown in Figure 2. The experimental sessions were identical except participants
were instructed to smoke a cigarette in the lab at the beginning of the deprived session for
smoking recency standardization. Participants’ CO and alcohol levels were tested at the
beginning of the sessions using Vitalgraph and an alcohol test. Participants, who showed CO
22
levels greater than 8 ppm (during the deprived session) or with a positive alcohol test (i.e., BrAC
> 0.00 g/dl), were rescheduled, since they had not abstained from smoking or drinking alcohol.
Participants then engaged in the following laboratory tasks: 1) instructed to smoke their own
cigarettes (in the non-deprived session) or rest for ten minutes (deprived session); 2) instructed to
complete self-report measure assessments; 3) vmHRV assessment; 4) a film clip task (including
vmHRV assessment); and 5) a smoking reinstatement task.
Experimental Tasks
Film Clip Task. The Film Clip Task was initially designed to measure affect variability
in response to emotionally valanced film clips in a laboratory setting (Koval, Pe, Meers, &
Kuppens, 2013). During each experimental session (total of two sessions per participant),
participants watched five clips spanning under three minutes in duration (two negative, two
positive, and one neutral) in fixed order, counterbalanced across experimental sessions. All film
clips were derived from a validated database of emotional clips (Schaefer, Nils, Sanchez, &
Philippot, 2010).
Smoking Reinstatement Task. This task is designed to measure the ability to resist
smoking temptation under preset conditions that make it advantageous to abstain from smoking
(McKee, 2009; McKee, Weinberger, Shi, Tetrault, & Coppola, 2012). Participants were
presented with their own cigarette box with 8 cigarettes, a lighter, and an ashtray. There was a
delay period at the beginning of the task which lasted 50 minutes, for which participants were
told they could smoke at any point. However, for each 5 minutes of delay, the participant was
told they would receive $0.50 for smoking abstinence (a maximum of $5.00 can be earned).
Next, the participant entered the self-administration period. During this period, participants could
23
smoke as little or as much as they wanted for the next 50 minutes. Participants were told they
had a $4.00 credit and each cigarette they smoked would cost them $0.50 from this credit. At the
end of self-administration, participants entered a rest period (with no smoking). This rest period
was necessary to prevent an opportunity to smoke once they left the laboratory, which could
reduce smoking during the delay and self-administration periods (DeGrandpre, Bickel, Higgins,
& Hughes, 1994). Participants were monetarily rewarded to 1) delay the opportunity to smoke
(outcome: minutes to smoke, range: 0-50 min), and 2) smoke fewer cigarettes once given the
opportunity to smoke (outcome: number of cigarettes smoked, range: 0-8).
Self-Report Measures
Center for Epidemiologic Studies Depression Scale (CES-D). The Center for
Epidemiologic Studies Depression Scale (CES-D) is a 20-item self-report measure that assesses
the frequency of depression symptoms within the past week (e.g., depressed mood, feelings of
guilt and worthlessness, loss of interest, psychomotor retardation, sleep disturbance) on a 4-point
Likert from 0 (Rarely or none of the time, 0-1 days) to 3 (Most or all of the time, 5-7 days)
(Radloff, 1977). Total scores range from 0 to 60 with higher scores indicating greater depression
symptoms. Participants with a CES-D score ≥20 were categorized as having an elevated
depression symptom level (=1), and those with CES-D score <20 were categorized as having a
low depression symptom level (=0). This cut-off score was determined by referencing a meta-
analysis that included studies that screened for major depressive disorder by using the CES-D.
Researchers determined that this cut-off score yielded the best trade-off between sensitivity
(83%) and specificity (78%)(Vilagut, Forero, Barbaglia, & Alonso, 2016).
24
Difficulties in Emotion Regulation Scale (DERS) - Short Form. The DERS-SF is a
multidimensional, 16-item self-report measure (a validated short-form version of the original 36-
item self-report measure) that assesses trait difficulties in emotion regulation on a 5-point Likert
from 1 (Almost never (0-10%) to 5 (Almost always (91-100%))(Bjureberg et al., 2016;
Neumann, van Lier, Gratz, & Koot, 2010). There are five subscales for the DERS-SF, including
(1) Clarity (e.g., “I have difficulty making sense of my feelings”); (2) Goals (e.g., “When I’m
upset, I have difficulty getting work done”); (3) Impulse (e.g., “When I’m upset, I become out of
control”); (4) Strategies (e.g., “When I’m upset, I believe that I will remain that way for a long
time”); (5) Non-acceptance (e.g., “When I’m upset, I feel that I am weak)”. This measure
produces a total score (indicating higher levels of emotion regulation difficulties) and five
subscale scores. Internal consistency reliability for this measure was high: lack of emotional
clarity. (a = .90), difficulty engaging in goal-directed behavior (a = .90), impulse control
difficulties (a = .88), limited emotion regulation strategies (a = .94), and non-acceptance of
emotional responses (a = .92). This measure was administered at the baseline session.
State Difficulties in Emotion Regulation Scale (S-DERS). The S-DERS is a validated,
multidimensional, 21-item self-report measure that assesses state-based (versus trait-based)
difficulties in emotion regulation in the present moment on a 5-point Likert from 1 (Not at all) to
5 (Completely) (Lavender et al., 2017). There are four subscales for the S-DERS, including (1)
Clarity (e.g., “I am confused about how I feel”); (2) Modulate (e.g., “I believe I will end up
feeling very depressed); (3) Awareness (e.g., “I care about what I am feeling” (Reverse-Coded));
and (4) Non-acceptance (e.g., “I feel ashamed with myself for feeling this way”). This measure
produces a total score (indicating higher levels of state-based emotion regulation difficulties) and
four subscale scores. Internal consistency reliability for this measure ranged from acceptable to
25
high (lack of emotional clarity (a = .73), limited modulation strategies (a = .92), lack of
emotional awareness (a = .81), non-acceptance of emotional responses (a = .92) for the
deprived session and (lack of emotional clarity (a = .86), limited modulation strategies (a = .91),
lack of emotional awareness (a = .73), non-acceptance of emotional responses (a = .95) for the
non-deprived session. This measure was administered during each experimental session.
Fagerström Test for Cigarette Dependence (FTCD). The FTCD is a 6-item self-report
measure that assesses the severity of cigarette dependence (Heatherton, Kozlowski, Frecker, &
Fagerstrom, 1991). Higher summed scores (max=10) indicate greater cigarette dependence. This
scale is widely used in tobacco research and has demonstrated good reliability and validity
among smokers (Pomerleau, Carton, Lutzke, Flessland, & Pomerleau, 1994). Smoking
dependence has been shown to reduce vmHRV values and negatively affect smoking outcomes
and tasks designed to prompt motivation to smoke, such as cigarette abstinence (Bold, Yoon,
Chapman, & McCarthy, 2013; Hayano et al., 1990). Thus, mean FTCD score at baseline was
included as an a priori covariate to adjust for baseline differences in tobacco dependence.
Minnesota Nicotine Withdrawal Scale (MNWS). The MNWS assesses 11 DSM
withdrawal symptoms, including smoking urge (craving), depressed mood, irritability,
frustration, or anger, anxiety, difficulty concentrating, restlessness, increased appetite, on a 4-
point Likert from 0 (None) to 5 (Severe) (Cappelleri et al., 2005; Hughes & Hatsukami, 1986;
Toll, O'Malley, McKee, Salovey, & Krishnan-Sarin, 2007). The MNWS has been well-validated
and frequently used within the literature to assess withdrawal symptoms and severity among
smokers (Cappelleri et al., 2005; Toll et al., 2007; West, Ussher, Evans, & Rashid, 2006). The
MNWS mean score was used to assess withdrawal symptoms during the deprived and non-
deprived experimental sessions.
26
10-item Brief Questionnaire of Smoking Urges (QSU). This 10-item measure assesses
desire, intention, urge, and need to smoke on a 7-point Likert from 0 (Strongly Disagree to 5
(Strongly Agree) (Cox, Tiffany, & Christen, 2001). Items from the scale include statements such
as, “I have an urgent desire to smoke,” “I could control things better right now if I could smoke,”
“Smoking would make me less depressed,” and “I am going to smoke as soon as possible.” The
QSU mean score was used to assess urge to smoke among participants during the deprived and
non-deprived experimental sessions.
Data Capture and Preprocessing
vmHRV assessment. vmHRV data were assessed and analyzed under the HRV
assessment guidelines set forth by the Task Force of The European Society of Cardiology and
The North American Society of Pacing and Electrophysiology (Task Force, 1996). RR intervals
were assessed at 1000Hz frequency via the FirstBeat Body Guard 2 (Firstbeat Technologies Ltd,
Jyväskylä, Finland), which is a reliable, portable, and lightweight R-R recording device for short
or long-term recordings. Participants could move freely while wearing this device. The device
was attached to participants via two electrodes (55 mm wide) with solid gel (to reduce any
possible skin irritation): one electrode was placed above the sternum on the right side of the body
and the other electrode was placed on the ribcage on the left side of the body. vmHRV was
assessed for 5 minutes during baseline (or rest) and during the entire film clip task (total of 15
minutes). Recordings from negative and positive film clips from each experimental session were
combined respectively and averaged. There was a total 6-minute recordings for positive film
clips, 6-minute recordings for negative film clips, and 3-minute recordings for neutral film clips
for each experimental session.
27
vmHRV preprocessing. R-R interval data were first uploaded to the Firstbeat Analysis
Server (Parak J, 2013) and then transferred to the Kubios software program 3.3.0 (University of
Kuopio, Finland). Built-in preprocessing and an automatic artifact correction tool offered by
Kubios software was applied to preprocess the data for statistical analyses. This automatic
correction algorithm detects abnormal beats with high accuracy. In a recent study that assessed
its performance against a state-of-the-art algorithm (using the MIT-BIH arrhythmia database),
sensitivity in detecting ectopic beats was 100%; sensitivity in detecting real atrial and ventricular
ectopic beats was 96.96%; and the corresponding specificity was 99.94%)(Lipponen &
Tarvainen, 2019). Two vmHRV metrics were used for this study: RMSSD (square root of the
mean squared differences between successive RR intervals) and HF-HRV (representing the
frequency range of 0.15 Hz to 0.4 Hz)("Heart rate variability: standards of measurement,
physiological interpretation and clinical use. Task Force of the European Society of Cardiology
and the North American Society of Pacing and Electrophysiology," 1996). These specific
validated metrics of vmHRV are highly associated with vagal tone (parasympathetic nervous
system dominance) and emotion regulation in the literature (Segerstrom & Nes, 2007; Thayer et
al., 2009; Thayer & Lane, 2000; Williams et al., 2015) and correspond to recommendations
offered by Laborde and colleagues on the assessment of vmHRV in psychophysiological
research (Laborde et al., 2017).
Statistical Analyses
Distributions of variables were assessed for satisfaction of linear model assumptions.
Non-normal continuous variables were log-transformed to create normal distributions of scores
for further analyses. All vmHRV metrics used as dependent variables within the models that did
not show normal distributions were log-transformed (Laborde et al., 2017). Preliminary analyses
28
contrasting variables of interest between depression symptom status groups (high vs. low
determined by the CES-D>20) and experimental session type (deprived versus non-deprived)
were conducted to determine if moderation for depression symptom status and/or experimental
session type on main effects was warranted. Additionally, contrasts of vmHRV during the film
clip task (positive, negative, neutral) and experimental session were also conducted to determine
if vmHRV reactivity metrics (in response to induction of emotion via the film clip tasks) would
be included in subsequent linear modeling.
Stata IC version 15 was used to construct generalized estimating equations (GEE), a
methodology that accounts for repeated (non-independent) data to test study aims (1) the main
effect of DERS on vmHRV at rest and during the film clips; (2) the main effect of DERS and
vmHRV during rest and during the film clip task on smoking behavior outcomes (minutes to
smoke and number of cigarettes smoked) during the smoking reinstatement task.
Multiple studies that have incorporated the smoking reinstatement task have used GEE
modeling to evaluate their outcomes (Day et al., 2015; Kahler et al., 2014). GEE takes into
account the dependency of observations by specifying a working correlation structure, and no
assumptions on the distribution of data are made (Cui, 2007). Since AIC criterion is no longer
applicable to determine model fit (GEE is not based on a maximum likelihood theory), a GEE
data-driven approach was used in the analyses to select the correlation structure called the quasi-
likelihood criterion (QIC) ((Cui, 2007). The correlation structure that resulted in the lowest QIC
value was selected and applied.
GEE models included two levels (level 1- repeated within-person data, level 2- between-
person data). All models included the primary predictors (DERS mean score; vmHRV) and
relevant covariates that can affect vmHRV, including baseline tobacco dependence (mean FTCD
29
score), sex, body mass index (BMI) (calculated as (weight (lbs) / [height (in)])
2
* 703 and age
(Laborde et al., 2017). Separate models were constructed for each outcome of the smoking
reinstatement task [minutes to smoke (range 0-50) and number of cigarettes smoked following
smoking initiation during the self-administration period (range: 0-8)]. Since minutes to smoke
showed a bimodal distribution between those who waited until the task was complete to smoke
and those who smoked within a few minutes, this variable was dichotomized accordingly
(1=smoked during the task; 0=waited to smoke until task was complete). GEE modeling with a
binomial distribution and logit link were thus specified.
RESULTS
Participants
Fifty-six participants of the parent study provided vmHRV data. Of those participants,
three participants’ data had to be excluded from the analyses [due to noisiness (too many
artifacts) and a device malfunction], which resulted in a sample of N=53 participants for the
analyses. Demographics of this subset are provided in Table 1. Participants reported a mean
FTCD score of 4.79, representing low to medium cigarette dependence. Twenty-three
participants (representing approximately 43% of the sample) reported elevated depression
symptoms according to the CES-D. vmHRV values were within the normative ranges with a
mean lnRMSSD value of 3.538 and a HF-HRV value of 6.088 (represented in log units).
Preliminary Analyses
Effects of experimental session type. Paired t-tests were conducted to contrast DERS,
vmHRV, and smoking outcomes between experimental sessions (deprived versus non-deprived)
among participants (Table 2). There was a significant difference in resting vmHRV variables
30
between the deprived and non-deprived experimental sessions (p<.01). Generally, participants
showed higher vmHRV in the deprived session versus the non-deprived session. There were also
significant differences in lnRMSSD score during the film clip task between sessions (p<.01), as
participants showed lower scores in the non-deprived session versus deprived session.
Additionally, there were significant differences found in mean DERS score (state-version) and
the aware subscale between experimental sessions. Participants in the deprived session showed
significantly greater difficulty in awareness of their emotions compared to the non-deprived
sessions (p=.002). There were significant differences between both smoking characteristics
(withdrawal symptoms, craving) and smoking outcomes (minutes to smoke, number of cigarettes
smoked) between experimental sessions. Participants showed greater withdrawal symptoms
(p<.001) and craving (p<.001) during the deprived session versus the non-deprived session.
Participants took less time to delay their smoking (Difference= -13.792, p<.001) and smoked
more cigarettes overall (Difference=+0.453, p=.011) during the deprived session versus the non-
deprived session.
Effects of elevated depression symptom status. Independent samples t-tests showed
that participants with elevated depression symptoms were significantly younger (p<.001) and had
a significantly lower mean BMI score (p=.017) than participants with lower depression
symptoms, as shown in Table 3. There were significant differences between participants with
high depression symptoms versus low depression symptoms on total DERS score at baseline and
every subset (p<.001); those with elevated depression symptoms self-reported more difficulties
in emotion regulation. There were no significant differences in vmHRV at rest and vmHRV
during the film clip tasks between depression symptom groups (p>.05) in both the deprived and
non-deprived experimental sessions. Participants with higher depression symptoms showed
31
increased (yet non-significant) vmHRV values compared to smokers with low depression
symptoms. Expectedly, participants with elevated depression symptoms reported significantly
higher withdrawal symptoms during both the deprived and non-deprived experimental sessions.
However, there were no significant differences between groups on any of the smoking outcomes
(minutes to smoke and total cigarettes smoked between session type) during the smoking
reinstatement task. Therefore, elevated depression symptoms status was not included in further
analyses as a moderator of interest.
Contrasts of vmHRV between film clips & experimental sessions. A two-way
repeated measures ANOVA was run on participants to determine if there were differences in
vmHRV during the film clips (positive versus negative versus neutral) by experimental session
type (deprived versus non-deprived). The results showed no significant differences in mean
vmHRV values (RMSSD) across film clip types during both the deprived and non-deprived
sessions (all p>.10). Therefore, vmHRV reactivity variables (differences in vmHRV between
resting and film clip tasks) were not included in subsequent modeling of smoking outcomes.
Main Outcomes Analyses
DERS as a predictor of vmHRV. Results of GEE modeling indicated that DERS mean
score at baseline was not associated with vmHRV metrics [lnRMSSD during rest (standardized
beta = 0.072, p=.496); High Frequency (HF) in log units during rest (standardized beta = 0.102,
p=.603); and lnRMSSD during film clips overall (standardized beta = 0.010, p=.923)], after
adjusting for sex, age, BMI, and cigarette dependence. The state-based DERS was also not a
significant predictor of vmHRV metrics among the deprived [lnRMSSD during rest (beta = -
0.037, p=.692), HF-log during rest (beta = -0.099, p=.562), lnRMSSD during film clips overall
32
(beta = -0.099, p=.309)] and non-deprived experimental sessions [lnRMSSD during rest (beta =
0.068, p=.422), HF-log during rest (beta = 0.106, p=.494), lnRMSSD during film clips overall
(beta = 0.062, p=.477)].
DERS as a predictor of smoking behavior outcomes. There was no significant effect of
DERS mean score at baseline on minutes to smoke (beta = -0.190, OR= 0.826, p=.143) adjusting
for age, sex, and cigarette dependence; however, there was a significant interaction effect of
DERS x experimental session type on number of cigarettes smoked (standardized beta = 0.374,
p=.021)(Figure 3).
vmHRV as a predictor of smoking behavior outcomes. Results of GEE modeling are
depicted in Table 4. The odds of smoking during the task versus waiting to smoke was less for
every unit increase in lnRMSSD during rest (OR=0.402, p=.018) and for every unit increase in
lnRMSSD during the film clip task (OR=0.466, p=.043)(Table 5). An alternative explanation
due to the log transformation of the predictor variable on a dichotomous outcome is as follows:
The odds of smoking during the task versus waiting to smoke decreased by 1.62% for every 10%
increase in resting vmHRV, and the odds of smoking during the task versus waiting to smoke
decreased by 1.36% for every 10% increase in vmHRV while watching the film clips [derived
from the following calculation: resting vmHRV e^(log(1.1)*log(0.402)) = 0.983751; (0.983751 -
1) * 100 = -1.62 %; vmHRV during film clip task: e^(log(1.1)*log(0.466)) = 0.986367;
(0.986367 - 1) * 100 = -1.36 % ]. The inclusion of an interaction term (lnRMMSD x
experimental session type) was not significant across both models (p=.796 and p=.791,
respectively). Results of GEE modeling indicated that there was no association between vmHRV
and number of cigarettes smoked during the smoking reinstatement task, holding all other
variables constant.
33
DISCUSSION
In this study, we sought to determine whether (1) difficulties in emotion regulation could
predict vmHRV at rest and during emotion induction via film clips and whether (2) difficulties in
emotion regulation and vmHRV could predict smoking behavior outcomes during a smoking
reinstatement task. Although we did not find evidence of a main effect of DERS on vmHRV,
study findings demonstrated that both the DERS at baseline and vmHRV (at rest and during the
film clip tasks) were significant predictors of smoking behavior, respectively. Additionally,
results from preliminary analyses offer some insight on the vmHRV methodology in future
tobacco studies, specifically the effects of abstinence on the assessment of vmHRV. This chapter
discussion will focus on the impact of abstinence on vmHRV metrics; potential explanations for
the non-significant findings (including the lack moderation effects by depression symptom status
and the lack of a linear association between the DERS and vmHRV); the predictive properties of
the DERS and vmHRV on smoking behavior outcomes, respectively; and the utility of vmHRV
as a viable biomarker of emotion regulation in future tobacco research.
vmHRV and smoking abstinence. The assessment of vmHRV during smoking
abstinence has led to several revelations concerning both short-term and long-term effects of
smoking exposure on the autonomic nervous system. Research has shown that smoking causes
an acute decrease in vmHRV (following 3 minutes of smoking) and subsequently increases
sympathetically-mediated HRV indices (after 10 to 17 minutes)(Hayano et al., 1990). Acute
changes related to smoking a cigarette in vmHRV typically dissipate after 5-10 minutes of
smoking (Karakaya et al., 2007). In the context smoking abstinence, a period of one day can
lead to an increase in resting HRV indices (based on 24-hour measurements) among heavy
smokers (one or more packs per day for 2-year period). In this study, participants remained
34
abstinent from smoking for a mean period of 18.21 hours prior to their participation in the
deprived experimental session. Study results indicate that participants’ short-term recordings of
vmHRV values may demonstrate an increase during a period of abstinence that is shorter than 24
hours. Additionally, mean minutes between smoking administration and vmHRV assessment
during the non-deprived experimental session averaged to 55.75 minutes. A previous study has
shown that smoking a cigarette can reduce vmHRV values for up to 30 minutes before beginning
to return to baseline with the greatest effects seen within 5-10 minutes of smoking. Therefore,
increased vmHRV values seen during the deprived session likely demonstrate the acute effects of
smoking abstinence on short-term vmHRV metrics. Although vmHRV can increase following a
short period of smoking deprivation, these initial increased levels tend to decrease following
seven days of smoking cessation (Yotsukura et al., 1998). This indicates that smoking has long-
term, blunting effects on vmHRV that are not subject to change due to sustained abstinence
behavior. In fact, a longitudinal study that assessed HRV among a large group of heavy smokers,
light smokers, and non-smokers (N=1485) for 11 years showed that a 15 to 25 year abstinent
period may be required for heavy former smokers ( > 20 pack-years) to reach the vmHRV levels
of lifelong non-smokers (Girard et al., 2015). Although abstinence from smoking can increase
vmHRV in a period shorter than 24 hours, whether or not the cardiovascular system can recover
from prolonged smoking exposure (related to tobacco use disorder) in the long-term is unsettled.
Despite this gap in knowledge, this study provides data that support an observed increase in
vmHRV within an abstinence period shorter than 24 hours, offering important insight into future
studies that assess vmHRV in the context of smoking deprivation. Researchers can take this
important methodological factor into account when implementing short-term vmHRV
assessment during smoking abstinence.
35
Depression symptom status, DERS, and vmHRV. Our results showed that vmHRV
and smoking outcomes did not differ between depression symptom level status (high versus low
depression symptoms based on the CES-D). As expected, participants in the high depression
symptom group reported greater difficulties in emotion regulation overall and across all
subscales. However, smokers with high depression symptoms did not show significantly lower
vmHRV values compared to smokers with low depression symptoms. This finding contradicts
past studies have shown that both smoking and depression can disrupt functioning of the
autonomic nervous system and lead to reductions in vmHRV (Wang et al., 2013; Harte et al.,
2013; Taylor et al., 2011). Ehrenthal and colleagues (2010) found that participants with
depression compared to psychiatrically healthy participants showed reduced vmHRV reactivity
indices and poor cardiovascular outcomes (including measures of heart rate, blood pressure,
cardiac output, and peripheral resistance) in response to two stress tasks (anger recall and mental
arithmetic tasks) (Ehrenthal et al., 2010). Additionally, Agelink et al. (2002) found that vmHRV
indices during a deep breathing task were negatively associated with both the status and severity
of depressive symptoms adjusting for age, gender, and smoking (Agelink, Boz, Ullrich, &
Andrich, 2002). Within a very recently published systematic review paper of 10 original articles
that assessed vmHRV reactivity to stress among participants with and without depression,
researchers did not find significant differences in resting vmHRV among participants across
studies; however, vmHRV reactivity in response to acute social stress tasks was significantly
blunted among individuals suffering from depression compared to psychiatrically healthy
controls across a majority of studies (Schiweck, Piette, Berckmans, Claes, & Vrieze, 2019). This
indicates that individuals with depression symptoms may demonstrate blunted reactivity to stress
as a result of chronic overactivation of the autonomic nervous system associated with their
36
condition (Schiweck et al., 2019). Therefore, likely reasons for the discrepancy between these
study findings and past studies may be due to a smaller sample size and lack of a stress-based
task to elicit vmHRV changes. Due to the overall observed blunting effect of both depression
symptoms and smoking on vmHRV, future studies that utilize much larger sample sizes (N>90)
and measure vmHRV reactivity in response to stress may be warranted when investigating the
effects of depression on vmHRV among smokers.
Main effect of DERS on vmHRV. The results of the study showed non-significant main
effects of DERS at baseline and the state-based DERS on vmHRV during rest and during the
film clip tasks. The lack of significant main effects may have been due to a small sample size,
lack of variance within the distribution of scores, and response bias. Our findings contradict past
studies that have shown significant correlations (albeit low effect sizes) between the DERS and
vmHRV at rest (Visted et al., 2017; Williams et al., 2015). These studies had much larger sample
sizes (N=60-183) (Visted et al., 2017; Williams et al., 2015) and used 24-hour recorded vmHRV
measurements (Visted et al., 2017), which may be more sensitive in assessing nuanced
differences in emotion regulation difficulties compared to short-term vmHRV measurements
(Visted et al., 2017). Additionally, these studies used college student populations, who were
much younger than our current sample and may not demonstrate the same blunted vmHRV
indices compared to older participants who smoke and suffer from depression. The lack of a
significant association between the DERS at baseline and the S-DERS during the experimental
sessions with vmHRV may also be due to floor effects from positively skewed data distributions
of the measures. Additionally, only the S-DERS demonstrated a negative association with
vmHRV (albeit non-significant), which was in the expected direction. This suggests that
participants may have been more accurate in their self-assessment when using a more proximal,
37
state-based measure that assesses maladaptive emotion regulation strategies in the present
moment. As expected, participants reported more state-based difficulties in emotion regulation
during the deprived session versus the non-deprived session, most likely due to the experience of
withdrawal-associated negative affect. It is possible that participants were more likely to
accurately self-report their difficulties in emotion regulation when they were able to directly
assess their ability to manage negative affect in the present moment versus reflecting on overall
trait-like qualities. Response bias may also be especially prominent among population samples
with psychiatric conditions, such as depression, who demonstrate limited self-awareness and
who are prone to alexithymia (inability to identify and label feelings)(Grabe et al.,2004; Moeller
& Goldstein, 2014). Thus, participants suffering from symptoms of depression may be prone to
inaccuracies when completing self-report measures that assess for (trait-based and state-based)
difficulties in monitoring and modulating emotional experiences.
Main effect of DERS on smoking behavior outcomes. The association between DERS
at baseline and number of cigarettes smoked significantly differed between the deprived and
non-deprived experimental sessions. There was a significant interaction effect of the DERS at
baseline and experimental session type (deprived versus non-deprived) on number of cigarettes
smoked. These findings indicate that the DERS, as a self-report trait-based measure, may have
more predictive value over smoking behavior outcomes when participant smokers are deprived
of smoking cigarettes. In other words, the DERS may demonstrate more predictive sensitivity
during smoking deprivation when smokers’ ability to manage negative affect related to
withdrawal (such as anxiousness, irritability, etc.) is tested and possibly compromised.
Withdrawal from smoking during abstinence is theorized to be a major motivator in the
maintenance of tobacco use. Although individuals typically initiate smoking for the positive and
38
rewarding effects of nicotine, sustained use and threat of relapse primarily occurs in an effort to
reduce negative affect associated with withdrawal (Baker et al., 2004). Thus, the DERS, during
smoking abstinence specifically, may serve an important indicator of relapse susceptibility
among smokers.
Main effect of vmHRV on smoking behavior outcomes. The most notable finding of
this study was the significant main effect of vmHRV on smoking delay during the smoking
reinstatement task. Greater vmHRV scores at rest and during the film clips reduced the odds of
smoking during the task versus waiting to smoke until the task was complete, irrespective of
smoking deprivation. This indicates that higher vmHRV at rest may represent greater self-
regulatory strength employed during a smoking reinstatement task that is designed to test
capacity to abstain from smoking. This study corroborates past research that has observed similar
linear associations of vmHRV on smoking outcomes. Ashare and colleagues (2012) found that
reduced vmHRV reactivity during a stressful imagery task was significantly associated with less
time to initiate smoking and increased craving among smokers. Libby et al. (2012) tested the
effects of mindfulness training on smoking and found that vmHRV reactivity (increase versus
decrease between baseline and during meditation tasks) predicted fewer cigarettes at 17-week
follow-up, irrespective of mindfulness training. In their limitations, researchers noted that a
stressful versus relaxing tasks may have elicited significant and more meaningful changes in
vmHRV reactivity (Libby et al., 2012). This study noted a lack of significant differences in
vmHRV reactivity in response to film clips (positive, negative, neutral) across both experimental
sessions. Induction of emotion via the film clip tasks within this study may not have been strong
or potent enough to elicit a physiological response, reflected by a substantial change in vmHRV
among participants. As noted by previous researchers, future studies may benefit from using a
39
stressful task or stress-based paradigm to elicit vmHRV changes among participant smokers that
are prominent for detection and sensitive enough to predict smoking behavior outcomes (Libby
et al., 2012).
Limitations. There are several inherent limitations to note. Current smoking and the
presence of depression symptoms can both function to reduce vmHRV values; therefore, there
were challenges in detecting differences between depression symptom level and vmHRV among
smokers (Harte et al., 2013). Additionally, the film task parameters (positive versus negative)
may not have induced sufficient emotional arousal and elicited emotion-driven regulation
changes in vmHRV metrics, particularly among participants who were already subject to blunted
vmHRV due to their comorbid conditions. Due to COVID-related impacts on participant
numbers, this study suffered from a small sample size and may have lacked necessary power to
detect between-group differences among variables of interest. Another limitation of the study
was the inability to control for respiration. Differences in respiration can impact vmHRV metrics
if participants are breathing slower than a specific frequency range (e.g., at a frequency less than
0.15-0.4 Hz)(Laborde et al., 2017). Current recommendations surrounding the assessment of
vmHRV call for adjustment of respiration, if indicated (Laborde et al., 2017). However, some
researchers argue that adjustment for respiration removes crucial variability associated with
neural regulation over the heart and thus takes away a crucial component of vmHRV (Larsen,
Tzeng, Sin, & Galletly, 2010). Nonetheless, the lack of respiration assessment represents a
limitation of our study and therefore we chose to focus on vmHRV indices that are not as
affected by respiration, namely RMSSD (Hill & Siebenbrock, 2009).
Conclusions. There have been limited tobacco studies that have assessed biological
markers of emotion regulation related in the context of withdrawal. Overreliance on self-report
40
measures can lead to bias (including reduced validity related to the repeated administration of
emotion-based question items). Studies that have tested objective biomarkers of emotion
regulation in smoking abstinence are scarce. One such example, Engelmann and colleagues
(2011) demonstrated that tobacco abstinence increased emotional physiological reactivity to
negative stimuli. Among non-abstinent and abstinent adult smokers, researchers measured the
level of potentiation and suppression of the startle reflex (a physiological marker associated with
emotion modulation) in response to emotionally-valanced images; however, researchers did not
assess smoking behaviors (Engelmann, Gewirtz, & Cuthbert, 2011). Our study represents one of
few studies that have incorporated a biological marker of emotion regulation (vmHRV) in the
context of smoking abstinence. The implementation of an experimental smoking task designed to
measure the ability to resist smoking temptation under preset conditions that make it
advantageous to abstain from smoking may not translate to real-world settings; however, this
nuanced investigation with high internal validity points to the potential predictive value of
vmHRV on smoking behavior outcomes, thus increasing its applicability and utility in
interventions that target and treat tobacco use disorder.
41
Figure 2. Procedure for Experimental Sessions (Deprived and Non-Deprived of Smoking) in
Study 1
42
Table 1. Demographics and Sample Characteristics in Study 1 (N=53)
Variables M (SD) or N(%)
Demographics
Age (range 21-65 yrs)
39.925(12.00)
Females
21 (38.89)
BMI
28.415 (6.51)
Baseline Measures
Cigarette Dependence (FTCD)(range 0-10)
4.79 (1.14)
DERS Total (range 16-80)
DERS Item Mean (range 1-5)
34.962 (17.598)
2.185 (1.110)
Elevated Depression
(CES-D cut-off>20)
23 (43.40)
vmHRV during Rest (N=44)
RMSSD overall 43.022 (27.990)
lnRMSSD 3.538 (0.708)
High-Frequency (HF) 6.088 (1.338)
vmHRV during Film Task (N=40)
RMSSD overall 41.592 (29.224)
lnRMSSD 3.490 (0.719)
RMSSD positive clips 41.087 (29.101)
lnRMSSD 3.469 (0.738)
RMSSD negative clips 42.241 (21.056)
lnRMSSD 3.487 (0.7440)
RMSSD neutral clips 43.446 (29.244)
lnRMSSD 3.535 (0.734)
Smoking Characteristics
Age first smoked (range 12-30 yrs) 16.55 (4.27)
Age smoked regularly (range 8-29 yrs) 18.62 (4.37)
Cigarettes smoked daily (range 3-30) 11.68 (4.95)
Smoking Behavior Outcomes
Total Time Delayed to Smoke (min) 32.48 (21.68)
Total Cigarettes Smoked (#) 1.15 (1.02)
RMSSD = root mean square of successive differences in milliseconds (ms); lnRMSSD = natural log-
transformed RMSSD values; HF: Reflection of auto-regressive modeling in log units
43
Table 2. vmHRV, State-based DERS, Smoking Characteristics and Smoking Outcomes between
Deprived and Non-Deprived Sessions in Study 1[Mean, (SD)]
Variables Deprived Non-Deprived Difference P value
vmHRV at Rest
RMSSD (ms) 49.206 (31.343) 36.396 (22.755) +12.810 .006**
lnRMSSD 3.689 (0.682) 3.389 (0.683) + 0.300 .002**
High-Frequency (HF) 6.379 (1.236) 5.847 (1.346) + 0.531 .005**
vmHRV during Film Clips
RMSSD (ms) 48.289 (32.552) 33.816 (23.275) +14.473
lnRMSSD 3.634 (0.760) 3.314 (0.664) + 0.321 .002**
vmHRV during Positive Clips
RMSSD (ms) 48.132 (33.308) 34.474 (23.996) +13.658
lnRMSSD 3.612 (0.801) 3.325 (0.678) + 0.287 .010**
vmHRV during Negative
Clips
RMSSD (ms) 49.876 (34.043) 33.852 (24.800) +16.024
lnRMSSD 3.665 (0.763) 3.296 (0.691) + 0.369 .001**
Difficulties in Emotion
Regulation (DERS) (State
Version) (N=53)
Overall
1.768 (0.588) 1.662 (0.627) +0.136 .071
Non-acceptance
1.512 (0.812) 1.474 (0.860) +0.038 .719
Modulate 1.582 (0.889) 1.499 (0.806) +0.084 .410
Aware
2.581 (1.014) 2.189 (0.800) +0.392 .002**
Clarity 1.594 (0.995) 1.575 (1.080) +0.019 .902
Smoking Characteristics
(N=53)
Withdrawal Symptoms
(MNWS)
1.840 (1.104) 1.212 (0.893) +0.627 <.001**
Craving Symptoms (QSU) 2.949 (1.349) 1.177 (1.347) +1.772 <.001**
Smoking Outcomes (N=53)
Total minutes to Smoke (min) 25.585 (23.193) 39.277 (17.734) -13.792 <.001**
Total Cigarettes Smoked (#) 1.377 (1.130) 0.925 (1.035) +0.453 .011**
Note: Conducted via paired t-tests; RMSSD = root mean square of successive differences in milliseconds
(ms); lnRMSSD = natural log-transformed RMSSD values; HF: Reflection of auto-regressive modeling in
log units
**= Significant at 0.05 level
44
Table 3. Demographics, DERS at Baseline, vmHRV, Smoking Characteristics, and Smoking
Outcomes by Depression Symptom Status in Study 1 [Mean (SD) or N(%)
b
]
Variables
High
Depression
Symptoms
(N=23)
Low Depression
Symptoms
(N=30)
Difference
P
value
Effect Size
(Cohen’s D
or OR
b
)
Age 31.26 (6.98) 46.57 (10.89) -15.31 <.001** -1.628
Females
b
8 (34.78) 13 (43.33) .528 1.434
BMI 25.97 (6.88) 30.55 (5.53) -4.58 .017** -0.740
Cigarette Dependence
(FTCD)
4.826 (1.337 4.767 (1.006) +0.059 .854 0.051
Difficulties in Emotion Regulation (DERS)
Overall 3.122 (0.892) 1.467 (0.597) +1.655 <.001** 2.242
Clarity 2.761 (0.864) 1.167 (0.356) +1.594 <.001** 2.359
Goals 3.536 (0.968) 1.678 (0.805) +1.858 <.001** 2.115
Impulse 2.522 (1.082) 1.378 (0.654) +1.144 <.001** 1.323
Strategy 3.235 (1.092) 1.360 (0.557) +1.875 <.001** 2.257
Non-acceptance 3.363 (1.306) 1.722 (1.029) +1.641 <.001** 1.418
Deprived Experimental Session
vmHRV Metrics
lnRMSSD at rest 3.826 (0.791) 3.571 (0.615) +0.255 .236 0.359
High-Frequency (HF) at rest 6.115 (1.331) 5.676 (1.377) +0.439 .298 0.379
lnRMSSD during film clips 3.726 (0.215) 3.525 (0.692) +0.201 .354 0.370
Smoking Characteristics
(N=53)
Withdrawal Symptoms
(MNWS)
2.266 (1.197) 1.513 (0.918) +0.753 .012** 0.720
Craving Symptoms (QSU) 3.013 (1.213) 2.900 (1.462) +0.113 .766 0.083
Smoking Outcomes
Minutes to Smoke 30.130 (22.219) 22.100 (23.687) +8.030 .215 0.348
Total Cigarettes Smoked (#) 1.652 (1.191) 1.167 (1.052) +0.485 .112 0.436
Non-Deprived Experimental Session
vmHRV Metrics
lnRMSSD at rest 3.547 (0.610) 3.298 (0.754) +0.249 .252 0.359
High-Frequency (HF) at rest 6.653 (1.468) 6.089 (1.120) +0.564 .156 0.379
lnRMSSD during film clips 3.537 (0.522) 3.248 (0.717) +0.289 .186 0.370
Smoking Characteristics
Withdrawal Symptoms
(MNWS)
1.565 (0.832) 0.942 (0.856) -0.623 .010** 0.757
Craving Symptoms (QSU) 0.983 (1.073) 1.327 (0.278) -0.344 .362 -0.255
Smoking Outcomes
Minutes to Smoke 40.652 (17.012) 38.400 (3.380) +2.252 .651 0.126
Total Cigarettes Smoked (#) 1.043 (1.224) 0.833 (0.160) +0.210 .469 0.202
Note: Conducted via independent t-tests and Chi-squared test
b
; RMSSD = root mean square of successive
differences in milliseconds (ms); lnRMSSD = natural log-transformed RMSSD values; HF: Reflection of
auto-regressive modeling in log units; **= Significant at 0.05 level
45
Table 4. Generalized Estimating Equations (GEE) of HRV Metrics on Smoking Outcomes
[Minutes to Smoke, Number of Cigarettes Smoked] in Study 1
Minutes to Smoke B
Robust
SE (B) 95% CI
Odds
Ratio
P value
vmHRV at Rest
lnRMSSD overall -1.105 0.455 -1.996, -0.213 0.331 .015**
vmHRV during Film
Clips
lnRMSSD overall -1.081 0.511 -2.082, -0.081 0.339 .034**
Number of Cigarettes
Smoked B
Robust
SE (B) 95% CI
P value
vmHRV at Rest
lnRMSSD overall -0.344 0.209 -0.753, 0.066 .100
vmHRV during film
clips
lnRMSSD overall -0.230 0.307 -0.831, 0.371 .454
Note: All models adjusted for baseline cigarette dependence, experimental session type, BMI, sex, and
age
**Significant at 0.05 level
Minutes to Smoke: 1 = smoked during the task, 0 = waited to smoke until task was complete
46
Figure 3. Interaction of DERS x Session Type on Number of Cigarettes Smoked in Study 1
Note: The effect of DERS mean score on number of cigarettes smoked differed between the deprived and
non-deprived experimental sessions. The significant interaction of DERS mean score and experimental
session type indicates that the DERS may have more predictive value over number of cigarette smoked
when participants are deprived from smoking.
47
CHAPTER 3: EMOTION REGULATION, HEART RATE VARIABILITY, AND
MINDFULNESS TRAINING AMONG SMOKERS (STUDY 2)
INTRODUCTION
Emotion regulation refers to the ability to monitor and modulate emotional experiences
(Gross & Thompson, 2007). Difficulties in emotion regulation pertain to maladaptive strategies
that are employed in the presence of distressful circumstances and can be prevalent among both
clinical and non-clinical populations (Aldao et al., 2010). Thus, difficulties in emotion regulation
represents a psychological construct applicable to a variety of populations and has become a
target of tobacco cessation research (Kauffman, Farris, Alfano, & Zvolensky, 2017; Rogers et
al., 2018). In the context of tobacco use disorder, difficulties regulating negative emotions (such
as anger, shame, humiliation, disgust, nervousness, etc.) have been associated with increased
rates of smoking relapse, reduced cessation attempts, and maintenance of addiction (Adams et
al., 2012; Farris et al., 2016; Fucito et al., 2010; Gehricke et al., 2007). Increased negative affect
(pertaining to the general experience of negative emotions) has also been identified as an
important factor in the transition from experimental smoking to habit-forming smoking
(Carmody, Vieten, & Astin, 2007). Moreover, individuals who demonstrate increased difficulties
in emotion regulation show augmented craving and attentional bias to smoking cues (Szasz et al.,
2012).
Although the construct of mindfulness has been operationalized in a variety of ways
within the literature (referring to a trait-like quality, general practice of meditation, or a transient
mental state), mindfulness is typically defined as the heightened focus on experiences in the
present moment, leading to a greater capacity for self-regulation that encompasses emotion
regulation and improved self-awareness of internal states (Holzel et al., 2011; Keng, Smoski, &
Robins, 2011). Researchers have recently proposed that emotion regulation may be the core
48
mechanism that underlies the beneficial effects of mindfulness, which has been corroborated by
both empirical and theoretical models (Guendelman, Medeiros, & Rampes, 2017). These
researchers posit that mindfulness offers a distinct benefit from other psychological interventions
in that it helps individuals expand their awareness and acceptance of bodily sensations, leading
to a greater overall tolerance of negative emotional states. This is referred to as implicit emotion
regulation that operates outside of subjective control to reduce emotion reactivity and arousal as
opposed to explicit emotion regulation that encompasses conscious strategies, such as cognitive
control, reappraisal, and affect labeling (Guendelman et al., 2017). While there is still debate on
exactly how mindfulness leads to benefits in emotion regulation, researchers posit that using an
integrative and holistic approach that considers mindfulness effects on both implicit and explicit
systems of emotion regulation may be warranted (Guendelman et al., 2017).
Mindfulness training which aims to increase mindfulness (as a disposition and mental
state) has shown promise in reducing cigarette smoking among smokers, derived from both
clinical and non-clinical populations (e.g., patients, prisoners, students, community volunteers,
etc.)(de Souza et al., 2015; Spears et al., 2017). A recent systematic review on mindfulness
interventions for the treatment of smoking behaviors identified 13 randomized controlled trials
that tested the effects of mindfulness training on smoking outcomes (craving, smoking cessation,
and relapse prevention)(de Souza et al., 2015). Researchers identified the beneficial effects of
mindfulness on cessation maintenance across studies but pointed out the need for more
rigorously conducted experiments and homogeneity across trials (de Souza et al., 2015).
Specifically, the majority of the reviewed studies relied on subjective measures, used non-
standardized mindfulness training protocols, did not incorporate active control groups, and did
not assess for underlying mechanisms of action driving the beneficial effects of mindfulness on
49
smoking outcomes, such as emotion regulation. Vagally-mediated heart rate variability
(vmHRV) may represent a promising biomarker of emotion regulation that can be used in
mindfulness research.
The most notable theoretical models that explain the association between emotion
regulation, stress regulation, and vmHRV include the neurovisceral integration model and the
vagal tank theory. The neurovisceral integration model proposed by Thayer and Lane (2000)
elucidates the neurobiological network between the central nervous system and the heart labeled
as the central autonomic network (CAN) (Thayer & Lane, 2000). The CAN represents a network
of executive neural structures (including the insula, anterior cingulate cortex, and prefrontal
cortices) that connect to the heart and control goal-directed behavior that is necessary for
autonomic regulatory control and stress regulation (Thayer & Lane, 2000). The output of the
CAN is reflected by vmHRV; higher vmHRV is associated with a more efficient network,
including improved emotion regulation capabilities and more effective autonomic control in the
face of perceived environmental stressors (Thayer et al., 2012; Thayer et al., 2009; Thayer &
Lane, 2000). The vagal tank theory builds upon the neurovisceral integration model by
broadening the context of how vmHRV changes depending on the environmental context
(Laborde et al., 2018). Laborde and colleagues (2018) describe the vagus nerve as a “tank” full
of self-regulatory resources. The fullness of the tank largely depends on the environmental
context. For example, higher vmHRV values during restful states usually denote better self-
regulation. During tasks that only necessitate physical activity or a physical stressor (such as
exercise), a decrease in vmHRV from rest represents more efficient adaptation, represented by
sympathetic nervous system dominance (Laborde et al., 2018). During tasks that necessitate a
low level of physical activity yet high level of executive functioning (such as a working memory
50
task with social evaluative threat), an increase in vmHRV from rest represents a more efficient
system, sustained engagement of executive neural structures, and better autonomic control
(Laborde et al., 2018). Thus, the vagal tank theory provides a theoretical foundation for which
researchers can base their hypotheses surrounding vmHRV reactivity in the context of emotion
and/or stress regulation task paradigms.
There have been a few studies that have tested the effects of vmHRV metrics,
mindfulness, and smoking under a single hypothesis (Ashare et al., 2012; Bodin et al., 2017;
Ferdous M, 2014; Harte et al., 2013; Libby et al., 2012; Murgia et al., 2019). Generally, smoking
leads to reduction in resting HRV with smokers demonstrating lower values on several HRV
metrics (SDNN, RMSSD, HF-HRV, etc.) compared to non-smokers during rest (Murgia et al.,
2019). Only a couple of studies have assessed these constructs using laboratory-based tasks
among smokers (Ashare et al., 2012; Libby et al., 2012). Ashare and colleagues (2012) found
that lower within-subject vmHRV values during a stressful imagery task was significantly
associated with less time to initiate smoking and increased craving among smokers. Libby and
colleagues (2012) tested the effects of mindfulness training on smoking and found that vmHRV
reactivity (increase versus decrease between baseline and during meditation tasks) predicted
fewer cigarettes at a 17-week follow-up, irrespective of mindfulness training. In their limitations,
researchers noted that a stressful versus relaxing task (meditation) may have elicited significant
and more meaningful changes in HF-HRV reactivity following implementation of mindfulness
training that targeted smoking behaviors (Libby et al., 2012). Our recent review on the utility of
vmHRV among randomized control studies that incorporate mindfulness training showed an
increase in vmHRV reactivity (corresponding to adaptive changes of vmHRV in response to a
task/stressor) following implementation of a mindfulness-based intervention (MBI) among non-
51
clinical samples across studies (Christodoulou et al., 2019). Thus, investigating vmHRV among
smokers when confronted with an acute social stressor may better elucidate the unique predictive
properties of vmHRV on smoking behavior outcomes.
The current study. vmHRV assessment was added to an existing parent study that
focused on testing the effects of mindfulness training on smoking behavior outcomes during a
smoking reinstatement task following exposure to an acute social stressor. The study aimed to
(1) test the main effect of difficulties in emotion regulation at baseline on vmHRV during an
acute social stressor task to validate vmHRV as an objective biomarker of emotion regulation
and (2) to test the main effects of difficulties in emotion regulation and vmHRV respectively on
smoking behavior outcomes during a smoking reinstatement task. Moderation effects of
mindfulness training versus an active control condition on these aims were also examined. The
following hypotheses were tested:
1. Smokers with greater self-reported difficulties in emotion regulation at baseline are
expected to demonstrate lower vmHRV values at rest and when exposed to an acute
social stressor during smoking abstinence
2. Smokers with greater self-reported difficulties in emotion regulation at baseline and
lower vmHRV values (in reaction to an acute social stressor) are expected to
demonstrate worse smoking behavior outcomes (craving, minutes to smoking, and
number of cigarettes smoked) during a smoking reinstatement task
3. Smokers in the mindfulness training group are expected to show lower difficulties in
emotion regulation, greater vmHRV metrics during exposure to an acute social
stressor, and better smoking behavior outcomes (craving, minutes to smoke, and
52
number of cigarettes smoked) post-intervention compared to smokers in the active
control condition
Addressing these hypotheses could add predictive value to vmHRV reactivity as a biomarker of
emotion regulation. Study findings could also demonstrate if difficulties in emotion regulation
and vmHRV reactivity metrics can predict smoking behavior outcomes, respectively. The
examination of moderation effects by intervention group status (mindfulness training versus an
active control condition) on these pathways would demonstrate the beneficial effects of
mindfulness training using an objective biomarker of emotion regulation, offering innovative
methodology and insight to future mindfulness-based studies and smoking cessation programs.
METHODS
Participants
Cigarette smokers (N=36) in the local geographic vicinity of Los Angeles were recruited
for the parent study, which was a randomized control trial (RCT) that examined the effects of
intranasal oxytocin and mindfulness training on smoking outcomes following stress induction in
a laboratory setting. Participants of the study were recruited via flyers, direct referrals, and
advertisements on Craigslist, Twitter, Facebook, and Instagram. Participants were eligible for the
study if they were 18-40 years of age, smoked at least 10 cigarettes per day, and were fluent in
English. Exclusion criteria for the study included the following: (1) current DSM-5 substance use
disorder, excluding tobacco use disorder (to minimize alcohol or drug withdrawal symptoms
during the study sessions); 2) any medical condition that increased risk for study participation
(such as sinus infection or other condition blocking access to the olfactory epithelium); 3)
women who were pregnant or nursing; 4) Current use of psychiatric medication for anxiety or
53
mood disorders; 5) breath Carbon Monoxide (CO) levels of at least 10ppm measured during
study intake (to confirm smoking status); 6) planning to quit or reduce smoking in the next 30
days (as the focus of the study was among smokers without an active plan to quit); 7) current
regular use of other nicotine-containing products; and 8) current mindfulness or meditation
practice of greater than 5 minutes per day. Additional vmHRV exclusion criteria included 1) use
of beta-blocker medications and 2) use of arrhythmia medications. Participants were asked not to
drink alcohol, energy drinks, or engage in extensive exercise the day of testing. Participants were
compensated for each study session (a total of two sessions) to help incentivize participation
throughout the intervention and to avoid study attrition.
Procedure
For the parent study, participants were required to attend two in-person sessions at the
research facility and undertake a 14-day intervention (mindfulness training or active control
condition) self-administered outside of the facility between their in-person sessions
(administered through auditory recordings via their cell phones based on group assignment).
Study procedures for the baseline and experimental sessions are depicted in Figure 4.
During the baseline session, participants read the consent form and discussed the study procedure
with research personnel. Participants, who agreed to participate, signed the informed consent and
then completed the screening procedures to establish their eligibility for the study. Following this
screening, participants’ medical results were sent to a study physician who reviewed the results
and cleared the participant for study enrollment. Enrolled participants were randomized into four
study conditions for a 2 (mindfulness training versus TED Talks) x 2 (oxytocin spray versus
placebo) factorial design: (1) mindfulness training + oxytocin spray, (2) mindfulness training +
54
placebo, (3) TED Talks+ oxytocin spray, and (4) TED Talks + placebo. Participants were asked
to engage in their assigned intervention (mindfulness training or TED Talks) for 20 minutes per
day (two 10-minute sessions) for a total of 14 days. Intervention adherence was monitored in
percent of audio completed by the participant for each audio session. This site recorded the
percentage of listening completion and time completed for each assigned session.
For the post-intervention experimental session (approximately 14 days following the
baseline session and intervention assignment), participants were asked to refrain from smoking
(for a minimum of 12 hours prior to the experimental session). Participants’ CO levels were
tested at the beginning of the experimental session using the Bedfont Scientific Smokelyzer®.
Participants who showed CO levels >10 ppm or with a positive alcohol test (i.e., BrAC > 0.00
g/dl) were rescheduled as they were not in a state of withdrawal. Participants completed post-
intervention self-report and physiological assessments (cortisol assessment, cardiovascular
assessment, vmHRV assessment, etc.). Participants then participated in the Trier Social Stress
Test (TSST) followed by the smoking reinstatement task thereafter. After task completion,
participants rested for 60 minutes and then completed the final measures and assessments.
Participants were then debriefed following completion of the experimental session. A
mindfulness training app was made available to each participant when their participation in the
study ended after the final follow up assessment.
Interventions
App-based mindfulness training. Participants assigned to mindfulness training listened
and practiced 20-minute guided meditations each day (in the morning and evening) for a total of
14 days minimum between the baseline and experimental sessions. Participants downloaded the
55
Headspace app and activated an account using a free code provided by the Headspace
administrative team. Headspace is an empirically tested app that incorporates mindfulness
practice and is designed to be completed by novices to mindfulness; thus, the instructions are
easier to follow. Participants were then trained on how to access their practice sessions in the
basic series provided (including body scans and seated stillness meditation exercises for novice
meditators). Participants were encouraged to complete 20 minutes of mindfulness training (two
10-minute sessions) a day at two separate intervals for a total of 14 days.
App-based psychoeducation control group (TED Talks). Participants were instructed
to listen to 20 minutes of TED Talk podcasts (two 10-minute talks per day) each day for a total
of 14 days to match the attention, time, and education features of the mindfulness training.
Additionally, participants were instructed to listen to the podcasts and focus their complete and
mindful attention to the audio, similar to instructions offered in the mindfulness training group.
None of the talks incorporated topics on mindfulness, meditation, or related themes/concepts.
Experimental Tasks
Trier Social Stress Test (TSST). The TSST is a standardized social stressor task
developed for use in a laboratory setting to induce acute social stress among participants
(Kirschbaum, Pirke, & Hellhammer, 1993). The TSST has been validated for use in smoking
studies and has been shown to successfully increase subjective stress states and physiological
stress among smokers (Childs & de Wit, 2009). For this task, participants were led to an
interview room that contained a false mirror and an iPad that was set up for recording.
Participants were instructed by research personnel to prepare (5 minutes) and deliver a 5-minute
speech on why they were qualified for their dream job position to a panel of experts behind the
56
false mirror. Participants were told that these experts were trained to detect verbal and non-
verbal signs of stress and would be monitoring and evaluating their speech (based on the content
and delivery) during the test and afterward on the iPad recording. Participants were able to see
themselves delivering the speech on the iPad during the test to increase self-evaluative stress and
self-consciousness during speech. After delivering the 5-minute speech, participants were then
asked to perform a 5-minute verbal arithmetic task (serial subtraction) in which they were
instructed to count backwards from 1022 in steps of 13. If the participant relayed the incorrect
number, they were corrected and asked by research personnel to start from the beginning.
Smoking reinstatement task. This task is designed to measure the ability to resist
smoking temptation under preset conditions that make it advantageous to abstain from smoking
(McKee, 2009; McKee et al., 2012). Participants were presented with a tray of 8 of their own
cigarettes, a lighter, and ashtray. There was a delay period at the beginning of the task which
lasted 50 minutes. During this delay period, the participant was told they could smoke at any
point, as they preferred. For each five minutes of delay, the participant was told they would
receive $0.20 for refraining from smoking (a maximum of $3.60 could be earned). Next, the
participant entered the self-administration period. During this period, the participant could smoke
as little or as much as they wanted for the next 50 minutes. They were told that they have a $1.60
credit and each cigarette they smoked would cost them $0.20 from this credit. At the end of self-
administration period, participants entered a rest period (with no smoking). This rest period was
necessary to prevent an opportunity to smoke once they left the laboratory, which could reduce
smoking during the delay and self-administration periods (DeGrandpre et al., 1994). Two
variables were computed from behavioral observation made by study staff during this task
57
including: time to first smoke (range of 0-50 minutes) and number of cigarettes smoked during
the self-administration period (range of 0-8 cigarettes).
Self-Report Measures
Difficulties in Emotion Regulation Scale (DERS). The DERS is multidimensional 36-
item measure of emotional regulation that is self-administered. Items include statements such as,
“I pay attention to how I feel,” “When I’m upset, I feel out of control,” and “When I’m upset, I
believe there is nothing I can do to make myself feel better” (Neumann et al., 2010). Items are
rated on a 5-point scale ranging from 1 or almost never (0-10%) to 5 or almost always (91-
100%). This measure produces a total score (indicating higher levels of emotion dysregulation)
and relevant sub-scores, including the following categories: nonacceptance of emotional
responses, difficulty engaging in goal-directed behavior, impulse control difficulties, lack of
emotional awareness, limited access to emotion regulation strategies, and lack of emotional
clarity. Internal consistency reliability for this measure ranged from acceptable to high: non-
acceptance of emotional responses. (a = .93), difficulty engaging in goal-directed behavior (a =
.87), impulse control difficulties (a = .86), lack of emotional awareness (a = .70), limited
emotion regulation strategies (a = .88), and lack of emotional clarity (a = .74). Mean total DERS
score upon baseline assessment was used as the primary predictor for all models as
recommended by previous research (Osborne et al., 2017).
Fagerström Test for Cigarette Dependence (FTCD). The FTCD scale measures the
severity of cigarette dependence (Heatherton et al., 1991). Higher summed scores (max=10)
indicate greater dependence. Items on this scale include: (1) How soon after you wake up do you
smoke your first cigarette, (2) Do you find it difficult to refrain from smoking in places where it
is forbidden, (3) Which cigarette would you hate most to give up, (4) How many cigarettes per
58
day do you smoke, (5) Do you smoke more frequently during the first hours after waking than
during the rest of the day, and (6) Do you smoke when you are so ill that you are in bed most of
the day. This scale is widely used in tobacco research and has demonstrated good reliability and
validity among smokers (Pomerleau et al., 1994). Mean FTCD score at baseline was used as a
covariate to adjust for baseline differences in tobacco use disorder severity.
Minnesota Nicotine Withdrawal Scale (MNWS). The MNWS assesses 11 DSM
withdrawal symptoms (including smoking urge (craving), depressed mood, irritability,
frustration, or anger, anxiety, difficulty concentrating, restlessness, increased appetite) on a 6-
point Likert-type scale (Cappelleri et al., 2005; Hughes & Hatsukami, 1986; Toll et al., 2007).
The MNWS total score will serve as the primary measure of subjective withdrawal. The MNWS
has been well-validated and frequently used within the literature to assess withdrawal symptoms
and severity among smokers (Cappelleri et al., 2005; Toll et al., 2007; West et al., 2006). Mean
MNWS score was used to assess differences in smoking characteristics between the baseline and
experimental sessions.
10-item Brief Questionnaire of Smoking Urges (QSU). The QSU is a 10-item measure
assesses desire, intention, urge, and need to smoke using a 7-point likert scale (Cox et al., 2001).
Items from the scale include statements, such as “I have an urgent desire to smoke,” “I could
control things better right now if I could smoke,” “Smoking would make me less depressed,” and
“I am going to smoke as soon as possible.” This measure was used to assess differences in
smoking characteristics between the baseline and experimental sessions. Urge to smoke as a
single item, “I have an urgent desire to smoke,” was assessed as an outcome variable following
the TSST administration.
59
Data Capture and Preprocessing
vmHRV Assessment. vmHRV data were assessed and analyzed under
the HRV assessment guidelines set forth by the Task Force of The European Society of
Cardiology and The North American Society of Pacing and Electrophysiology (Task Force,
1996). vmHRV was assessed for 5 minutes during rest at the pre-intervention baseline session, 5
minutes during rest during the post-intervention experimental session, and during the entire
TSST administration (15-minute TSST session) which started at approximately 11:00 AM for all
participants. R-R recordings from each session (resting pre-intervention, resting post-
intervention, speech preparation, speech task, mental arithmetic task) were collected. RR
intervals were assessed at 1000Hz frequency via the FirstBeat Body Guard 2 (Firstbeat
Technologies Ltd, Jyväskylä, Finland), which is a reliable, portable, and lightweight R-R
recording device for short or long-term recordings. Participants could move freely while wearing
this device. The device was attached to participants via two electrodes (55 mm wide) with solid
gel (to reduce any possible skin irritation): one electrode was placed above the sternum on the
right side of the body and the other electrode was placed on the ribcage on the left side of the
body.
vmHRV Preprocessing. R-R interval data were first uploaded to the Firstbeat Analysis
Server (Parak J, 2013) and then transferred to the Kubios software program 3.3.0 (University of
Kuopio, Finland). Built-in preprocessing and an automatic artifact correction tool offered by
Kubios software was applied to preprocess the data for statistical analyses. In a recent study that
assessed its performance against a state-of-the-art algorithm (using the MIT-BIH arrhythmia
database), sensitivity in detecting ectopic beats was 100%; sensitivity in detecting real atrial and
ventricular ectopic beats was 96.96%; and the corresponding specificity was 99.94%)(Lipponen
60
& Tarvainen, 2019). Two vmHRV were used: RMSSD (square root of the mean squared
differences between successive RR intervals) and HF-HRV (representing the frequency range of
0.15 Hz to 0.4 Hz)("Heart rate variability: standards of measurement, physiological
interpretation and clinical use. Task Force of the European Society of Cardiology and the North
American Society of Pacing and Electrophysiology," 1996). These specific validated metrics of
vmHRV are highly associated with PNS dominance and emotion regulation in the literature
(Segerstrom & Nes, 2007; Thayer et al., 2009; Thayer & Lane, 2000; Williams et al., 2015) and
correspond to recommendations offered by Laborde and colleagues on the assessment of HRV in
psychophysiological research (Laborde et al., 2017). vmHRV reactivity was calculated by
subtracting RMSSD values during the TSST (overall, math, speech) from baseline vmHRV
values during the experimental session.
Statistical Analyses
Stata IC version 15 was used to conduct all statistical analyses. Distributions of variables
were assessed for satisfaction of linear model assumptions. Non-normal variables were log-
transformed to create normal distributions of scores for further analyses. All vmHRV metrics
used as dependent variables within the models that did not show normal distributions were log-
transformed as recommended (Laborde et al., 2017). There were no missing data since research
personnel monitored participants’ responses throughout the baseline and experimental
procedures.
Among the total sample, data from 12 participants were drawn from the pilot study. The
pilot study included the same study protocol but excluded the nasal spray administration.
Comparisons of participants in the pilot study and main trial were conducted to see if we could
combine the study samples for greater statistical power. Comparisons of demographic variables,
61
DERS scores at baseline, smoking characteristics at baseline, vmHRV metrics at baseline and
during the experimental sessions, and smoking behavior outcomes revealed no significant
differences between participants that received a nasal spray (N=14) and participants that did not
receive a nasal spray (N=12) during the experimental session; therefore, study samples from the
respective studies were combined for further analyses.
Due to the COVID-19 pandemic, the parent study was forced to prematurely end all
procedures. Thus, much fewer participants participated in the study than originally anticipated,
and the majority of the proposed linear models between of predictors of interest on smoking
behavior outcomes were underpowered. A preliminary power analysis using G*Power (v3.1)
revealed that a sample size of N=55 for each linear model was required for 80% power to detect
a small to medium effect (f2 =0.15) at an alpha of .05. The sample size of N=20 participants
prevented execution of our initially proposed analyses, and most intervention group comparisons
were limited to non-regression analyses.
Our final analyses included partial correlation analyses of the DERS, vmHRV, and
smoking behavior outcomes during the smoking reinstatement task including minutes to smoke
(minutes lapsed before participants began smoking (range 0-50 minutes)), number of cigarettes
smoked (number of cigarettes smoked following smoking initiation during the self-
administration period (range: 0-8)), and urge to smoke (single item from the QSU), that were
adjusted for baseline cigarette dependence and vmHRV. These tests were run to assess if linear
modeling would be warranted in a future study with a larger sample size. Bivariate correlations
could show an indication of a linear pathway between aforementioned hypotheses including:
62
(1) Expected negative association between DERS at baseline and vmHRV metrics pre-
intervention (at rest) and post-intervention [greater DERS scores correspond to lower
vmHRV metrics at rest and during the TSST]
(2) Expected positive and negative associations between DERS, vmHRV post-
intervention, and smoking behavior outcomes (greater DERS scores and lower
vmHRV during rest and TSST correspond to less minutes to initiate smoking, greater
number of cigarettes smoked, and greater self-reported urgency to smoke,
respectively)
Preliminary analyses to test for intervention effects were undertaken. Successful
randomization of the study groups was confirmed using independent samples testing to contrast
demographic variables and baseline variables of interest (DERS, vmHRV, and smoking
characteristics) between participants in mindfulness training and the TED Talks group.
Independent samples tests were also run to contrast vmHRV reactivity metrics (during rest and
the TSST) and smoking behavior outcomes (minutes to smoke, number of cigarettes smoked,
and urge to smoke). Significant differences in vmHRV and smoking behavior outcomes between
groups would point to potential intervention effects and address our hypotheses:
(3) Smokers in the mindfulness training group were expected to show significantly
greater vmHRV at rest and during the TSST as well as better smoking behavior
outcomes (longer delay to initiate smoking, less number of cigarettes smoked, and
less self-reported urgency to smoke) compared to the active control group
As an additional exploratory analysis, within-subject contrasts of the DERS overall mean score
and subscales, smoking characteristics, and vmHRV at rest between the baseline and
experimental sessions were also conducted.
63
RESULTS
There were 36 total participants that provided vmHRV data at baseline and 26 of those
participants provided vmHRV data during the experimental session. Six out of the 26
participants’ vmHRV data from the experimental session had to be excluded due to too many
artifacts/noise [above the predetermined threshold of the Kubios automatic correction algorithm
(>2 artifacts for each 5-minute segment and/or more than 5% corrected beats)] and device
malfunctions (R-R intervals were not readable). Out of the 20 participants that provided
experimental session data, 11 participants were assigned to mindfulness training while 9
participants were assigned to the control group.
Partial spearman correlations were conducted on predictors of interest (DERS, vmHRV)
and smoking outcomes, adjusted for cigarette dependence and/or baseline vmHRV. Spearman
correlations revealed significant positive associations between baseline DERS subscales and
craving during the experimental session (non-acceptance of emotions rs = .444, p = .03; impulse
rs = .414, p = .04; strategy rs = .499, p = .01). None of the other correlations revealed significant
results except for mean vmHRV during the TSST (RMSSD) and minutes to smoke during the
smoking reinstatement task. vmHRV during the TSST (RMSSD) showed a positive association
with longer duration to smoke (rs = .486, p = .04)(i.e., higher vmHRV during acute social stress
was positively associated with delayed minutes to smoke during the smoking reinstatement task).
Independent samples t-tests and chi-squared tests demonstrated randomization to study
groups was successful, as there were no significant differences between groups on demographics
(age, sex, and BMI), DERS overall mean score and subscales, smoking characteristics, and
resting vmHRV at the baseline session (Table 5). Due to the small sample sizes for the outcome
64
variables (N=20), independent samples t-tests, chi-square tests, and Mann-Whitney tests (for
non-parametric variables) were used to assess differences in variables of interest between
intervention groups. Since minutes to smoke showed a bimodal distribution between those who
waited the full 50 minutes to smoke and those who smoked within a few minutes, this variable
was dichotomized accordingly (1=smoked during the task; 0=waited to smoke until after the task
ended). Bootstrapping was also conducted for parametric tests to re-estimate standard errors of
mean differences and acquire more robust confidence intervals.
Paired samples t-test via and Wilcoxin signed rank tests revealed no significant
differences between participants’ DERS scores at baseline and the experimental sessions except
for the clarity subscale (p=.018) and smoking characteristics, as shown in Table 6. Although
participants demonstrated greater vmHRV metrics during the experimental session versus the
baseline session, these differences did not reach statistical significance.
Results of all parametric tests and bootstrapping revealed no significant differences
between the two study groups on resting vmHRV, vmHRV during the TSST, and smoking
outcomes (Table 7). However, there were significant and marginally significant differences in
vmHRV (RMSSD) reactivity to the TSST between by study group. Chi-squared testing did not
reveal significant differences between study groups on positive HRV change; however, the effect
size (or odds ratio) was relatively large at OR=6.125. A Mann-Whitney test for vmHRV
reactivity overall revealed marginally significant differences between groups (p=.063), and the
effect size was considered medium (rho=.416). As expected, participants in mindfulness training
showed marginally significant overall increase in vmHRV in response to the TSST compared to
the control group.
65
The Mann-Whitney test for vmHRV reactivity during the math portion of the TSST
revealed a significant difference between intervention groups (p=0.044) with participants in
mindfulness training demonstrating more positive overall vmHRV (RMSSD) change values in
reaction to the math portion of the TSST compared to the control group. No correction was made
for multiple comparisons as this was an exploratory preliminary analysis. Visual comparison of
participants that showed a positive vmHRV change to the TSST (vmHRV increased in response
to TSST) versus negative vmHRV change (vmHRV decreased in response to TSST) versus no
change (less than a 1.0 unit change in RMSSD values) are depicted in Figure 5. Results show
that participants in the mindfulness training group demonstrated greater positive vmHRV change
(less reactivity to stress) compared to the control group.
DISCUSSION
The aims of this study centered on investigating the linear associations of the DERS,
vmHRV, and smoking behavior outcomes, while also testing the moderation effects of
mindfulness training versus an active control condition on these pathways. Despite overall null
findings most likely due to a small sample size, the results of the preliminary analyses reveal
promising avenues for future research. Specifically, the significant findings surrounding vmHRV
reactivity during the TSST (including overall medium to large effect sizes) between groups point
to its potential utility as a biomarker of stress regulation following mindfulness training. While
there are multiple factors to consider before making conclusive inferences (including study
limitations discussed in detail below), preliminary findings point to the possible salutary effects
of mindfulness training on acute social stress reactivity among smokers. Despite our limited
sample size, interpretation of the preliminary findings can offer informative guidance for future
66
research. The discussion will center on interpreting analyses that revealed larger effect sizes
between groups in an effort to guide forthcoming studies that seek to understand and test the
interrelationships between the DERS, vmHRV, and smoking behavior outcomes in the context of
mindfulness training.
In our recently published review on the utility of vmHRV in mindfulness intervention
research (Christodoulou, Salami, & Black, 2020), we assessed the current state of the literature
and examined 17 original studies that have incorporated both a mindfulness-based intervention
(MBIs) and vmHRV assessment (as both a dependent variable and predictor of study outcomes).
Although our scoping review focused on highlighting the variability found across vmHRV
methodology (including the duration of vmHRV recording, reported vmHRV metrics, covariates
in vmHRV analyses, and exclusion criteria) and offering recommendations for future studies, we
found that the effects of mindfulness on vmHRV were better detected in trials that incorporated a
stressor during vmHRV assessment (Christodoulou et al., 2020). In other words, mindfulness-
related salutary effects on vmHRV may be more apparent during acute stress (assessed via
vmHRV reactivity). This corroborates both the theoretical and empirical evidence on the role of
mindfulness training and its association with improved emotion regulatory capacity and stress
adaptation across clinical and non-clinical populations. Thus, as demonstrated within this study
and in our previous work, the salutary effects of mindfulness on HRV may be more apparent
through a stress induction paradigm (Christodoulou et al., 2020). Future research would also
benefit from incorporating a measurement of magnitude of vmHRV change as well (such as
percent change) when assessing the impact of mindfulness training on vmHRV. Due to the
design of the parent study (posttest-only group design), our results were limited to a binary
67
designation of an increase or decrease in vmHRV in response to acute social stress thus
representing a limitation in our findings.
As a result of increasing evidence of the association between emotion regulation and
vmHRV, vmHRV has become not only an indicator of static, trait-based emotion regulation
capacity but can also demonstrate successful adaptation to stress through task-based HRV acute
changes (vmHRV reactivity) in response to stress induction (Thayer et al., 2009). The effects of
mindfulness training have shown to increase functionality in the same executive regions of the
brain that comprise the central autonomic network, including augmented activity in the
prefrontal cortex and downregulation of the amygdala (Chiesa, Serretti, & Jakobsen, 2013; Tang,
Holzel, & Posner, 2015; Tang, Yang, Leve, & Harold, 2012). These neuroimaging findings
corroborate the utility of vmHRV as a biomarker of mindfulness-related effects on these neural
networks (specifically downregulation of sympathetic response to stressful stimuli (amygdala)
and increased activation of executive regions of the brain that support parasympathetic
dominance through activation of the vagus nerve). A very recent met-analysis on vmHRV and
mindfulness training that focused on randomized controlled trials indicated that there is limited
evidence to demonstrate that mindfulness has beneficial effects on vmHRV; however, the
majority of studies that showed null findings focused on vmHRV at rest and did not assess
vmHRV reactivity in response to acute social stress (Brown et al., 2020). In partial agreement
with these authors, we urge researchers to continue to conduct rigorously designed studies to test
the interrelationships between mindfulness training and vmHRV reactivity using stress-induction
paradigms (Christodoulou et al., 2020).
Although significant associations between vmHRV metrics and smoking behavior
outcomes were not found, results from the TSST reveal insights on how smokers handle stress
68
during smoking abstinence. Differences in vmHRV reactivity in response to the TSST may
demonstrate how smokers are able to regulate their stress and associated negative affect during
withdrawal from smoking. Studies in the past that have assessed the predictive properties of
vmHRV on smoking indicate that vmHRV reactivity may be a reliable predictor for smoking
behavior (Ashare et al., 2012; Libby et al., 2012). Libby et al. (2012) tested the effects of
mindfulness training on smoking and found that HRV reactivity (increase versus decrease
between baseline and during meditation tasks) predicted fewer cigarettes at follow-up,
irrespective of mindfulness training. In their limitations, researchers noted that a stressful versus
relaxing task (meditation) could have elicited significant and more meaningful changes in HF-
HRV reactivity following implementation of mindfulness training (Libby et al., 2012). Although
findings between vmHRV metrics and our primary smoking behavior outcomes were not
significant, the current study, in corroboration with past research, points to the applicability of
mindfulness training on stress reduction and improved smoking behavior outcomes. Although
non-significant, participants in mindfulness training reported less craving compared to the
control group during the experimental session (during smoking abstinence). Additionally,
participants in mindfulness training demonstrated more positive vmHRV change in reaction to
the TSST (acute social stress during smoking abstinence). A recent systematic review on
mindfulness interventions for the treatment of smoking behaviors pointed out the need for more
rigorously executed experiments to test the effects of mindfulness training on smoking behavior
outcomes (de Souza et al., 2015). Specifically, the majority of the reviewed studies relied on
subjective measures and did not assess for underlying mechanisms of action driving the
beneficial effects of mindfulness on smoking behavior. Therefore, the incorporation of objective
69
biomarkers, such as vmHRV reactivity, to studies that have larger sample sizes and are well-
powered represents a promising direction for future research.
Limitations. There were several limitations to note within the study aside from a small
sample. We lacked the research personnel to conduct a full in-person panel evaluation for the
TSST, changes to the standard TSST protocol were necessary (e.g., replacing a full panel of
evaluators with a double mirror). Despite this protocol adaptation, studies have shown that
modifications to the TSST that retain social-evaluative threat are robust, and participants
demonstrate an appropriate stress response (via both biological and self-report measures)
(Dickerson & Kemeny, 2004; Liu et al., 2017). The effects of stress on vmHRV among smokers
may have been attenuated since smokers demonstrate generally lower vmHRV compared to non-
smokers. We were also not able to adjust for influential covariates within our preliminary
analyses of intervention group differences that can affect vmHRV, including age, sex, BMI,
baseline vmHRV values, etc. (Laborde et al., 2017). Therefore, conclusive inferences regarding
intervention-specific effects on vmHRV reactivity cannot be made. Due to the design of the
parent study (posttest-only group design), study results were limited to a binary designation of an
increase or decrease in vmHRV in response to acute social stress (versus magnitude of vmHRV
change) thus representing a limitation in our findings. Additionally, the inability to control for
respiration which may affect some vmHRV metrics is also a limitation (Laborde et al., 2017).
Participants remained seated during the entire TSST in an effort to reduce variability due to the
effects of postural changes on respiration (H. G. Kim, Cheon, Bai, Lee, & Koo, 2018).
Additionally, the exact same procedures to measure vmHRV at rest, vmHRV during the TSST,
and vmHRV during recovery were employed across participants.
70
Conclusions. Although this study suffered from a sample size, it managed to address
several limitations of past studies, including the integration of an objective measure of emotion
and stress regulation; mindfulness training derived from a standardized program that is available
to the public; and the inclusion of an active control group (de Souza et al., 2015).
Generalizability of the study results to naturalistic stressors may be limited due to the
experimental nature of the data; however, controlled laboratory settings offer inherent internal
validity that may better elucidate the associations between our constructs of interest and increase
the predictive value of vmHRV within tobacco research. Large effect sizes and marginally
significant results point to the utility of vmHRV reactivity as a promising biomarker for future
investigations. Studies that incorporate mindfulness training in an attempt to improve smoking
behavior outcomes would benefit from incorporating biological indicators of emotion regulation
and stress adaptation, such as vmHRV reactivity, to add predictive validity to their findings.
71
Figure 4. Procedure for Baseline and Experimental Sessions in Study 2
72
Table 5. Sample Demographics, DERS, Smoking Characteristics, and vmHRV between
Intervention Groups in Study 2 (N=36) [Mean(SD) or N(%)
b
]
Mindfulness
Training (N=17)
TED Talks
(N=19) p-value
Effect Size
Demographics
Age 32.235 (5.298) 29.632 (6.865) .215
0.422
Females
b
7 (41.18) 5 (26.32) .345 0.510
b
BMI 26.027 (4.886) 25.221 (4.263) .358
0.374
Difficulties in
Emotion Regulation
(DERS) (Range 1-5)
Item Overall
1.938 (1.578) 2.012 (0.662) .747
-0.108
Acceptance 1.794 (0.910) 2.079 (1.022) .386
-0.386
Goals 2.224 (1.095) 2.284 (0.939) .859
-0.060
Impulse 1.814 (0.803) 1.816 (0.880) .949
-0.002
Aware 2.149 (0.538) 2.235 (0.888) .724
-0.119
Strategy 1.961 (0.868) 1.926 (0.918) .910
0.038
Clarity 1.635 (0.725) 1.811 (0.770) .489
-0.233
Smoking
Characteristics
Age first smoked
(range 8-34 yrs) 17.412 (5.444) 18.526 (5.551) .548
-0.203
Age smoked
regularly (range 9-34
yrs) 18.764 (5.761) 19.947 (5.759) .543
-0.205
Cigarettes smoked
daily (range 4-40) 14.765 (9.208) 14.263 (3.572) .827
0.073
FTCD (range 0-10)
5.176 (2.038) 5.105 (1.370) .902
0.041
MNWS (range 0-5) 1.005 (1.007) 0.880 (0.630) .655
0.151
QSU (range 0-5) 2.741 (1.433) 2.611 (1.262) .773
0.097
HRV at Rest
lnRMSSD 3.479 (0.763) 3.598 (0.647) .624
-0.167
High-Frequency (HF)
5.979 (1.562) 6.165 (1.286) .702
-0.131
a
Independent samples two-sided t-test;
b
Chi-squared test; Effect size reported as Cohen’s D or OR
b
RMSSD: Root Mean Square of Successive Differences in milliseconds; lnRMSSD = natural log-
transformed RMSSD values; HF: Reflection of auto-regressive modeling in log units
73
Table 6. Paired Samples Analyses of DERS, Smoking Characteristics, and vmHRV at Rest in
Study 2
Note:
a
Paired samples t-test via, Wilcoxon signed rank test; **significant at p<.05
RMSSD: Root Mean Square of Successive Differences in milliseconds; lnRMSSD: natural log-
transformed RMSSD values; HF: Reflection of auto-regressive modeling in log units
Measures (# of Participants)
Baseline
Mean (SD) or
Median
Experimental
Mean (SD) or
Median Difference p-value
Difficulties in Emotion
Regulation (DERS) (N=12) b
Item Overall 1.736 1.889 .092
Acceptance 1.750 1.583 .125
Goals 2.000 2.100 .191
Impulse 1.667 1.833 .115
Aware 2.080 2.250 .408
Strategies 1.750 1.750 .080
Clarity 1.500 2.200 .018**
Smoking Characteristics
(N=26)
Withdrawal (MNWS) a 1.017(0.879) 1.126 (0.872) -0.109 .469
Craving (QSU) a 2.581(1.254) 2.688 (1.325) -0.107 .688
HRV at Rest (N=22)
RMSSD(ms) 38.200 42.153
lnRMSSD a 3.536 (0.780) 3.725 (0.736) -0.189 .091
High-Frequency (HF)
a
6.110 (1.574) 6.445 (1.439) -0.335 .111
74
Table 7. vmHRV Metrics and Smoking Outcomes between Intervention Groups during Smoking
Abstinence in Study 2 [Mean (SD), N(%)
b
, or Median
c
]
Variable Name
Mindfulness
Training
(N=11)
TED Talks
(N=9)
p-
value
Effect
Size
Bootstrap Results
Diff SE 95% CI
vmHRV at Rest (lnRMSSD)
a
lnRMSSD 3.451 (0.779) 3.968 (0.588) .085 -0.295 -0.517 0.284 -1.081 0.001
High-Frequency
(HF)
5.938 (1.638)
6.872 (1.036) .114 -0.689 -0.934 0.585 -2.082 0.214
vmHRV during TSST(lnRMSSD)
a
Overall 3.545 (0.713) 3.728 (0.599) .541 -0.280 -0.183 0.290 -0.753 0.387
Prep 3.517 (0.669) 3.873 (0.642) .232 -0.544 -0.356 0.285 -0.914 0.203
Speech 3.529 (0.705) 3.796 (0.634) .375 -0.400 -0.267 0.297 -0.850 0.318
Math 3.564 (0.759) 3.570 (0.691) .986 -0.008 -0.006 0.329 -0.651 0.639
vmHRV Reactivity to TSST
Overall
Reactivity
c
4.670 -6.542 .063 0.416
- - - -
Speech
Reactivity
c
5.597 -5.769 .118 0.341
- - - -
Math
Reactivity
c
4.973 -8.036 .044* 0.450
- - - -
Positive HRV
change
b
7 (77.78) 4 (36.36) .064 6.125
b
N/A 5.753 0.972 38.599
Smoking Outcomes
Withdrawal
(MNWS)
a
1.076 (0.989) 1.169 (0.793) .792 0.105 -0.093 0.343 -0.766 0.580
Craving (QSU)
a
2.317 (1.460) 3.007 (1.155) .191 0.530 -0.690 0.513 -1.749 0.368
Urge to Smoke
c
70.000 76.500 .273 0.219
c
-
Minutes to
Smoke
b
10 (83.33) 11(78.57) .759 1.364
b
N/A 0.151 -0.247 0.342
# of Cigarettes
Smoked
c
1.000 1.500 .810 -0.047
c
- - - -
Note: *Significant at p<.05; Conducted via Independent samples t-test
a
; Chi-squared test
b
; Mann-
Whitney test
c
; Effect size reported as Cohen’s D
a
, Odds Ratio
b
, Rho
c
Bootstrap Methodology: Based on 1000
replications
HRV Reactivity: HRV during TSST (overall, speech, math) – HRV at rest;
Positive HRV Change: 1 = HRV increased, 0= HRV decreased
Minutes to Smoke: 1=smoked during the task, 0=waited to smoke until task was complete
75
Figure 5. vmHRV Reactivity to TSST between Intervention Groups in Study 2
Note: vmHRV Reactivity was calculated by taking RMSSD values during TSST (overall, speech, math)
and subtracting RMSSD values at rest during the experimental session
76
CHAPTER 4: EMOTION REGULATION, HEART RATE VARIABILITY (HRV), AND
RELATIONSHIP SATISFACTION AMONG SMOKERS (STUDY 3)
INTRODUCTION
Conflict in romantic relationships represents a strong interpersonal stressor that has acute
and long-term effects on health, particularly among smokers who are more susceptible to
increased morbidity and mortality (Coppotelli & Orleans, 1985; Kiecolt-Glaser & Newton,
2001). Abstinent smokers are more likely to relapse in the face of interpersonal conflict and
associated stress (Coppotelli & Orleans, 1985). Additionally, smokers who are attempting to quit
tobacco products report better smoking outcomes when they feel increased social and emotional
support from their partners (Coppotelli & Orleans, 1985; Mermelstein, Lichtenstein, & McIntyre,
1983). Thus, smokers in unsatisfactory romantic relationships marked by interpersonal conflict
represent a particularly vulnerable population, who may be more susceptible to relapse and failed
smoking cessation attempts.
Emotion regulation, defined as the ability to monitor and regulate emotional experiences,
can be a strong indicator of the quality of intimate relationships and self-reported relationship
satisfaction. Studies report that emotion regulation is associated with improved intrapersonal
relationship satisfaction and adaptive attachment styles (Bloch, Haase, & Levenson, 2014; John
& Gross, 2004; Salvatore, Kuo, Steele, Simpson, & Collins, 2011). Additionally, positive
changes related to physiological and emotional reactivity in response to interpersonal conflict in
one partner can transfer to the other partner (Diamond, 2003). This indicates that changes in
emotion regulation are thus not only expressed on an individual level but are dyadic in nature
and often experienced interpersonally. Specifically, one such benefit of improved individual
emotion regulation is associated reductions in physiological and emotional arousal among both
partners. Researchers theorize that the reason for this observation is that improved emotion
77
regulation capabilities in one partner can produce a non-reactive space during interpersonal
conflicts for engagement in healthier communication patterns toward reconciliation (Bloch et al.,
2014). In a 13-year, three-wave longitudinal study of middle-aged married couples (N=82),
greater regulation of negative emotion during discussion of marital conflict (assessed via
observed reduction in emotional experience, behavior, and physiological arousal) predicted
greater marriage satisfaction for both partners between Waves 1 and 2 (spanning approximately
six years)(Bloch et al., 2014). On the contrary, difficulties in emotion regulation can be
predictive of long-term relationship conflict throughout the lifespan. In a longitudinal study
spanning 21 years among 190 men, individual emotion dysregulation mediated the
intergenerational transmission of relationship conflict between parent and child participants and
between participants and their romantic partners (H. K. Kim, Pears, Capaldi, & Owen, 2009).
Thus, difficulties in emotion regulation on an individual level can have widespread and long-
lasting repercussions on social relationships.
Despite the established association between difficulties in emotion regulation and
smoking outcomes, including increased rates of smoking relapse and reduced cessation attempts
(Carmody et al., 2007), very few studies to date have assessed the association between emotion
regulation and smoking in the context of relationship satisfaction. Shoham and colleagues (2007)
found that partners in dual-smoker couples reported increased positive affect and relationship
unity while jointly lighting a cigarette, as opposed to single-smoker couples. Moreover, a model
on the social dynamics of smoking emphasizes the social cohesion and sense of solidarity that
accompanies smoking in dual smoker relationships, for which the activity of concurrent smoking
provides a sense of interconnection, emotional stability, and perceived emotional support
(Doherty & Whitehead, 1986). The majority of these studies have focused on interpersonal
78
conflicts in the context of a single smoker versus dual smoker relationships and associated
smoking outcomes as opposed to explicitly focusing on the predictive properties of individual
emotion regulation. Additionally, the majority of these studies have incorporated only self-report
measures of emotion regulation, which are subject to bias. Studies that use laboratory-based
emotion regulation tasks typically request participants to regulate their emotions under very
controlled conditions, which increases internal validity but may not generalize to real-world
conditions and interactions. Thus, the need for biological indicators of these constructs in real-
world settings is imperative to elucidate underlying mechanisms that contribute to smoking
behaviors among smokers in romantic relationships.
Vagally-mediated HRV (vmHRV) may represent an important biological indicator given
its association with both emotion regulation and relationship satisfaction. vmHRV denotes the
variability in time between heartbeats and corresponds to activity of the vagus nerve (McCraty &
Shaffer, 2015; Shaffer & Ginsberg), largely responsible for activities of the parasympathetic
nervous system and its influence over the heart (McCraty & Shaffer, 2015; Shaffer & Ginsberg).
The polyvagal theory by Porges (2007) discusses the evolutionary value of the vagal nerve in
promoting social behavior and interpersonal connections (Porges, 2007). Efficient self-regulation
allows the vagal nerve to “break” when confronted with environmental demands that necessitate
physical resources and trigger activation of the sympathetic nervous system. During moments of
rest and relaxation, emotion regulatory resources can be redirected toward social mobilization
and adaptive behaviors necessary to survive in a social environment (Porges, 2007). In situations
that do not require a “fight or flight” response, the vagus nerve is activated, represented by
parasympathetic dominance and reflected by increased vmHRV metrics. This shift in the
autonomic system enables prosocial behavior and spontaneous social engagement, in which
79
individuals are able to redirect the focus of their autonomic resources toward navigating and
building social relationships (Porges, 2007). The polyvagal theory thus demonstrates the
evolutionary role of vmHRV in a social context, emphasizing both its intrapersonal and
interpersonal value.
Only a couple of studies to date have assessed the association of vmHRV and marital
satisfaction. These studies found that participants who report more satisfaction in their
relationships have better cardiovascular health and demonstrate increased vmHRV at rest
compared to participants who report lower relationship and marital satisfaction (Donoho,
Seeman, Sloan, & Crimmins, 2015; Smith et al., 2011). No other studies have assessed the
association between difficulties in emotion regulation and vmHRV at rest specifically among
smokers in satisfactory romantic relationships versus unsatisfactory relationships. This study
aims to address this gap in the literature by investigating whether vmHRV during rest at baseline
can predict smoking behaviors between these groups at long-term, 28-day follow-up, using
ecological momentary assessment (EMA) procedures.
The Current Study. Two measures of emotion regulation were added to the parent
study, including a self-report measure (the Difficulties in Emotion Regulation Scale (DERS))
and an objective biomarker (vmHRV). This study aimed (1) to assess the correlation between the
DERS and vmHRV at rest to validate vmHRV as an objective biomarker of emotion regulation
and (2) to test the respective main effects of DERS and vmHRV on smoking behavior outcomes
(smoking variability and cigarette use percentage change) following a 28-day observation period.
Moderation effects of relationship satisfaction on these linear associations were also examined.
The following hypotheses were tested:
80
1. An inverse association between self-reported difficulties in emotion regulation and
vmHRV is expected. Smokers who report more difficulties in emotion regulation are
expected to demonstrate lower vmHRV at rest
2. Smokers with higher self-reported difficulties in emotion regulation and lower
vmHRV at rest are expected to demonstrate poorer smoking behavior outcomes
(greater smoking variability and greater cigarette use percentage change)
2a. These hypothesized linear associations will be especially prominent if participants
report lower relationship satisfaction. In other words, smokers with low relationship
satisfaction will report more difficulties in emotion regulation, lower vmHRV at rest,
and worse smoking behavior outcomes compared to smokers with high relationship
satisfaction.
Study findings would validate vmHRV as an objective biomarker of emotion regulation and as
an independent predictor of smoking behavior outcomes within real-world conditions. The
results of the study could also point to interpersonal factors that may be influencing the
association between emotion regulation, vmHRV, and smoking behaviors among couples,
particularly relationship satisfaction. Rather than relying on self-report measures, vmHRV may
be used in future studies to reliably identify vulnerable individuals, who are more susceptible to
smoking relapse due to their interpersonal environment.
METHODS
Participants
Participants included community-based (from the Los Angeles area) cohabitating couples
aged 18-65 that were recruited from August 2018 to December 2019 via flyers, direct referrals,
81
and advertisements on Craigslist, Twitter, Facebook, and Instagram. Inclusion criteria included:
(1) adult individuals in a minimum of a one-year-long relationship with no plans to separate or
move in the next 30 days and (2) participant smokers must have smoked at least 5 cigarettes per
day for at least the past 2 years. Exclusion criteria included: (1) non-fluency in English; (2)
breath carbon monoxide (CO) levels < 5ppm measured during study intake to confirm smoking
status; (2) current pregnancy or intention to become pregnant within 6 months; (3) and currently
nursing.
Procedure
Participants completed an intake session in which their eligibility was assessed.
Participants, who were enrolled and completed their informed consent, were administered
baseline measures, EMA surveys/smartphone training, CO verification procedural training (to
confirm smoking status), and HRV assessment simultaneously. For the parent study, participants
were trained on the use of EMA equipment, which included a 28-day EMA procedure.
EMA data assessment was based on a smartphone application that provided survey items
to participants (smokers): self-initiated morning survey (upon awakening) and evening survey
(bedtime) each day. Morning surveys included question items on negative affect, a single rating
item assessing urge to smoke, and abstinence plan (“Do you plan to quit smoking today”) with a
binary response (“yes”/“no”). Nightly surveys included an assessment of number of daily
cigarettes smoked. For participants that were attempting to quit smoking, this EMA procedure
began 7 days prior to their selected quit date. The percentage of days that participants did not
plan to quit across the 28 days was assessed to be used as a covariate in subsequent linear
modeling. For EMA data, plan to quit smoking that day was included as a potential covariate as
well. After 28 days, participants came back to the research facility to return the EMA equipment.
82
Due to missing data associated with EMA procedures, contingencies in the parent study had been
placed to encourage participants to respond to EMA prompts (including daily reminders, ability
to review their own progress, and bonus compensation for 80% completion).
Self-Report Measures
Difficulties in Emotion Regulation Scale (DERS). The DERS is a multidimensional 36-
item measure of emotional regulation that is self-administered. Items include statements such as
“I pay attention to how I feel,” “When I’m upset, I feel out of control,” and “When I’m upset, I
believe there is nothing I can do to make myself feel better” (Neumann et al., 2010). Items are
rated on a 5-point scale ranging from 1 or almost never (0-10%) to 5 or almost always (91-
100%). This measure produces a total score (indicating higher levels of difficulties in emotion
regulation) and relevant subscales, including the following categories: nonacceptance of
emotional responses, difficulty engaging in goal-directed behavior, impulse control difficulties,
lack of emotional awareness, limited access to emotion regulation strategies, and lack of
emotional clarity. Internal consistency reliability for this measure ranged from acceptable to high
(a = .72 clarity, a = .83 goals, a = .89 impulse, strategies a = .92, non-acceptance a = .87,
emotional awareness a = .70). Mean DERS score upon baseline assessment was used as the
primary predictor for all models, as recommended by previous research (Osborne et al., 2017).
Relationship Assessment Scale (RAS). The RAS is a seven-item measure of
relationship satisfaction, including items regarding the provision and receipt of interpersonal
support and relationship conflict using a 5-point likert scale (Hendrick, Dicke, & Hendrick,
1998). Items in this questionnaire include (1) How well does your partner meet your needs; (2)
In general, how satisfied are you with your relationship; (3) How good is your relationship
83
compared to most; (4) How often do you wish you hadn’t gotten in this relationship; (5) To what
extent has your relationship met your original expectations; (6) How much do you love your
partner; and (7) How many problems are there in your relationship. Higher summed scores
denote greater relationship satisfaction after reverse scoring of items 4 and 7. The RAS has
demonstrated good test-retest reliability across samples of ethnically and racially diverse healthy
couples (Renshaw KD, 2011; Vaughn, 1999). Internal consistency reliability for this measure
was acceptable (a = .78).
Fagerström Test for Cigarette Dependence (FTCD). The FTCD scale measures the
severity of cigarette dependence (Heatherton et al., 1991). Higher summed scores (max=10)
indicate greater dependence. Items on this scale include: (1) How soon after you wake up do you
smoke your first cigarette, (2) Do you find it difficult to refrain from smoking in places where it
is forbidden, (3) Which cigarette would you hate most to give up, (4) How many cigarettes per
day do you smoke, (5) Do you smoke more frequently during the first hours after waking than
during the rest of the day, and (6) Do you smoke when you are so ill that you are in bed most of
the day. This scale is widely used in research and has demonstrated good reliability and validity
among smokers (Pomerleau et al., 1994). Mean FTCD score at baseline was used as a covariate
in linear modeling.
Smoking Behavior Outcomes
Cigarette Use Variability. Cigarette use variability was calculated as the mean square of
successive differences (MSSD) in cigarette use across 28 days of assessment. N time
measurements were calculated as follows:
84
MSSD represents an aggregate measure of temporal dependency and variability linked to within-
person variance. This variable has been shown to be consistent with a “prn” smoking style (i.e.,
smoking “as needed” in response to stressors in the environment) versus smoking to relieve
perceived fears of withdrawal symptoms associated with habitual smoking (Powers et al., 2016).
High smoking variability could be indicative of smoking behaviors aimed to relieve negative
affect due to stressors within the environment, which can ultimately lead to augmented
difficulties during smoking cessation attempts (Powers et al., 2016).
Data Capture and Preprocessing
vmHRV assessment. vmHRV data were assessed and analyzed under the HRV
assessment guidelines set forth by the Task Force of The European Society of Cardiology and
The North American Society of Pacing and Electrophysiology (Task Force, 1996). RR intervals
were assessed at 1000Hz frequency via the FirstBeat Body Guard 2 (Firstbeat Technologies Ltd,
Jyväskylä, Finland), which is a reliable, portable, and lightweight R-R recording device for short
or long-term recordings. Participants could move freely while wearing this device. The device
was attached to participants via two electrodes (55 mm wide) with solid gel (to reduce any
possible skin irritation): one electrode was placed above the sternum on the right side of the body
and the other electrode was placed on the ribcage on the left side of the body. vmHRV at rest
was assessed for five minutes during the baseline session prior to participants completing
subjective measures. All participants were instructed to breathe normally in a seated position for
five minutes without speaking.
vmHRV preprocessing. R-R interval data were first uploaded to the Firstbeat Analysis
Server (Parak J, 2013) and then transferred to the Kubios software program 3.3.0 (University of
Kuopio, Finland). Built-in preprocessing and an automatic artifact correction tool offered by
85
Kubios software was applied to preprocess the data for statistical analyses. In a recent study that
assessed its performance against a state-of-the-art algorithm (using the MIT-BIH arrhythmia
database), sensitivity in detecting ectopic beats was 100%; sensitivity in detecting real atrial and
ventricular ectopic beats was 96.96%; and the corresponding specificity was 99.94%)(Lipponen
& Tarvainen, 2019). Two vmHRV were used for analyses: RMSSD (square root of the mean
squared differences between successive RR intervals) and HF-HRV (representing the frequency
range of 0.15 Hz to 0.4 Hz)("Heart rate variability: standards of measurement, physiological
interpretation and clinical use. Task Force of the European Society of Cardiology and the North
American Society of Pacing and Electrophysiology," 1996). These specific validated metrics of
vmHRV are highly associated with parasympathetic nervous system dominance and emotion
regulation in the literature (Segerstrom & Nes, 2007; Thayer et al., 2009; Thayer & Lane, 2000;
Williams et al., 2015) and correspond to recommendations offered by Laborde and colleagues on
the assessment of HRV in psychophysiological research (Laborde et al., 2017).
Statistical Analyses
Stata IC version 15 was used to conduct all statistical analyses. Distributions of variables
were assessed for satisfaction of linear model assumptions. Non-normal continuous variables
were log-transformed to create normal distributions of scores for further analyses. All vmHRV
metrics used as dependent variables within the models that did not show normal distributions
were log-transformed as recommended (Laborde et al., 2017).
Due to the COVID-19 pandemic, the parent study was forced to prematurely end all
procedures. A preliminary power analysis using G*Power (v3.1) revealed that a sample size of
N=68 for each linear model was required for 80% power to detect a small to medium effect (f2
86
=0.15) at an alpha of .05. Fewer participants participated in the study than originally anticipated,
and the majority of the proposed linear models between of predictors of interest on smoking
outcomes were underpowered. Specifically, there were low sample sizes for the outcome
variable: cigarette use variability (N=25). This was most likely due to the fact that these data
originated from bedtime reports on successive days. Overall compliance was generally low
although similar to compliance rates in other EMA studies among study samples with tobacco
use disorder (Jones et al., 2019)(Pang et al., 2020). As previously reported, participants
completed a mean 24.5 days (SD=7.5) with a range of 1-28 days (Loftus et al., 2021).
Percentages for compliance was 71.9 ± 29.6% (M ± SD) for morning surveys and 58.3 ± 29.2%
for evening surveys (Loftus et al., 2021). Bedtime compliance was the lowest; therefore, there
were a number of participants out of the total sub-sample (N=30) that did not have enough
consistent data to calculate these outcomes (e.g., day-to-day variation could not be calculated).
Multiple imputation (MI) for models predicting cigarette use variability was considered but
deemed inappropriate. Data were only missing for the dependent/outcome variable, indicating
that MI would not provide notable advantages over full case analysis (GD, 2015; Lee &
Simpson, 2014).
The sample size also prevented examination of moderation effects of relationship
satisfaction on linear pathways. Therefore, preliminary analyses included group mean contrasts
of demographic variables and variables of interest (DERS, vmHRV, and smoking behavior
outcomes) between individuals who reported high relationship satisfaction versus low
relationship satisfaction to help guide future research with larger sample sizes.
General/generalized linear models (GLM) were constructed to run preliminary tests on
the main effects of DERS mean score and vmHRV on cigarette use variability, respectively.
87
Percentage of days not dedicated to quitting and cigarette use variability were not significantly
correlated (p=.429). To keep the models parsimonious, percent of days not dedicated to quitting
was not included as a covariate in GLM modeling. Age, sex, and cigarette dependence at
baseline (FTCD) were included in the models as covariates to account for potential effects on the
outcome variables. Cigarette use variability was also dichotomized due to a non-normal
distribution via median split (median = 3.2) with 1=high cigarette use variability across 28 days;
0 = low cigarette use variability across 28 days.
RESULTS
There were 38 total participants that provided vmHRV data. Two participants’ data had
to be discarded due to a device malfunction (e.g., device slipped off during assessment and
clothing interference), resulting in a total sample size of N=36. Six out of the 36 participants
were non-smokers; therefore, they were excluded from the main analyses, leading to total sample
size of N=30.
Association of DERS and vmHRV. As shown in Table 8, there were no significant
pairwise correlations between the DERS and any of its subscales with vmHRV at rest during the
baseline session. Additionally, the observed correlation coefficients were generally low and in
the opposite hypothesized direction (positive versus negative correlation coefficients). Higher
reported difficulties in emotion regulation were positively associated with vmHRV values at rest;
however, these correlations did not reach statistical significance.
Effects of relationship satisfaction status. Sample demographics stratified by
relationship satisfaction status (high versus low determined by median split) are shown in Table
9. A significantly greater number of women were in the low relationship satisfaction group
88
compared to the high relationship satisfaction group (p=.003). Participants in the low relationship
satisfaction group self-reported more difficulties in emotion regulation overall and within each
subscale (except for awareness); however, these differences did not reach statistical significance
with low to medium effect sizes. Participants did not demonstrate significantly different smoking
characteristics or smoking behavior outcomes between relationship satisfaction groups
(including age first smoked, age first smoked regularly, cigarettes smoked daily within the past
month, cigarette dependence (FTCD), mean number of cigarettes smoked, and cigarette use
variability). Participants in the high relationship satisfaction group demonstrated significantly
higher HF-HRV scores (p=.039); however, this difference was most likely attributed to sex-
related differences of vmHRV, as there were more women within this subgroup. An analysis of
covariance (ANCOVA) was run to assess the difference in HF-HRV between groups while
controlling for sex. The results indicated that the difference between HF-HRV between groups
was no longer significant after adjustment for sex (F=3.13 (2,32), p=.06). Due to the non-
significant findings of the preliminary analyses among variables of interest (DERS, vmHRV, and
smoking behavior outcomes) between groups, relationship satisfaction was no longer considered
a moderator of interest for further linear modeling of smoking behavior outcomes.
DERS on smoking behavior outcomes. Results of GLM are provided in Table 10 and
revealed that DERS mean score at baseline was not a significant predictor of cigarette use
variability group status (high versus low) adjusted for age, sex, and cigarette dependence at
baseline (beta = 0.243, OR=1.275, p=0.852).
vmHRV on smoking behavior outcomes. Results of GLM showed that vmHRV metrics
(lnRMSSD and HF-HRV) at baseline were not significant predictors of cigarette use variability
89
status (high versus low), adjusted for age, sex, and cigarette dependence at baseline [beta = -
0.400, OR= 0.670, p=.232 (lnRMSSD); beta = -1.422, OR = 0.241, p=.284 (HF-HRV)].
DISCUSSION
Although the majority of the hypothesized linear associations produced null findings,
important insights drawn from the data can help guide future investigations that wish to test the
effects of difficulties in emotion regulation and vmHRV on smoking behavior outcomes in the
context of relationship satisfaction. In the next section, implications of the study findings are
discussed, including the lack of association between difficulties in emotion regulation and
vmHRV at rest; lack of associations between relationship satisfaction, difficulties in emotion
regulation, and vmHRV; and vmHRV and DERS as predictors of smoking behavior outcomes.
The limitations of the study are also provided, along with recommendations for future research
studies that implement vmHRV assessment.
Association of DERS and vmHRV. DERS mean score at baseline was not strongly
correlated with vmHRV indices with overall low correlations (rs =.153 to .336))(p>.05). These
results indicate that vmHRV at rest may not be a strong indicator of self-reported difficulties of
emotion regulation. This result contradicts findings from past studies that have shown negative
associations (albeit low effect sizes) between the DERS and vmHRV at rest controlling for age,
sex, and BMI (Visted et al., 2017; Williams et al., 2015). These studies had much larger sample
sizes (N=60-183) (Visted et al., 2017; Williams et al., 2015), conducted hierarchal regression
analyses, and used 24-hour recorded vmHRV measurements (Visted et al., 2017), which may be
more sensitive in assessing nuanced differences in emotion regulation difficulties compared to
short-term vmHRV measurements (Visted et al., 2017). Additionally, these studies used
undergraduate student populations, who were younger than our current sample (mean age ranged
90
from 19-23). Although researchers did not account for smoking behaviors (other than requesting
for participants not to smoke 24 hours prior to assessment), these younger participants may not
have demonstrated the same blunted vmHRV indices compared to older participants.
Additionally, previous studies have shown that self-reported difficulties in emotion regulation
decrease with age (Giromini, Ales, de Campora, Zennaro, & Pignolo, 2017), which may have
been a contributing factor to the observed positively skewed distribution of DERS scores among
our sample of middle-aged participants. Therefore, potential floor effects and associated lack of
variance among DERS scores may have been contributing factors to the null findings.
DERS and vmHRV by relationship satisfaction status. As expected, smokers in the
low relationship satisfaction group self-reported more difficulties in emotion regulation overall
and within each subscale (except for awareness); however, these differences did not reach
statistical significance with observed low to medium effect sizes. The inclusion of the awareness
subscale on the DERS has been debated among studies that have conducted factor analyses on
the scale. In revised versions of the DERS (including shorter forms to eliminate participant
burden), the awareness subscale has been removed due to repeated poor factor loading across
studies (Osborne et al., 2017). Additionally, researchers posit that the awareness subscale may
assess a different construct than the other scale items and propose that awareness may actually
precede the employment of emotion regulation strategies (Osborne et al., 2017). The current
study also did not find significant associations between vmHRV at rest and self-reported
relationship satisfaction status (high versus low). Correlations between vmHRV metrics and
relationship satisfaction were generally low and in the opposite hypothesized direction (r=-.142
to -.229), meaning those who reported less relationship satisfaction showed higher vmHRV at
rest. This observed inverse correlation was likely driven by sex and age effects on vmHRV
91
(more women who were significantly younger were in the low relationship satisfaction group
than men). Women have higher vmHRV values than men, and these sex-driven differences in
short-term vmHRV metrics tend to dissipate among older ages (>55 years)(Voss, Schroeder,
Heitmann, Peters, & Perz, 2015). Additionally, vmHRV sex-driven age differences tend to be
more pronounced between the ages of 35-44 years and 45-54 years (Voss et al., 2015). Lack of
significant associations may also be due to small sample size. Past studies that have found
significant associations between vmHRV and relationship satisfaction had much larger study
samples (N=228-907) (Donoho et al., 2015; Smith et al., 2011). Additionally, one study assessed
vmHRV during a discussion of a disagreement in which participants were first primed with a
negative, positive, or neutral task (Smith et al., 2011). This enabled researchers to directly assess
within-subject vmHRV reactivity. Thus, within-subject variation in vmHRV reactivity to a social
stressor or interpersonal task may be a more sensitive and robust indicator of relationship
satisfaction than vmHRV at rest. Future studies may benefit from incorporating a task-based
paradigm that enables vmHRV reactivity assessment among smoker participants.
Main effect of DERS on smoking behavior outcomes. Past studies have demonstrated
that the DERS can significantly predict smoking outcomes, including past-hour smoking
frequency (Adams et al., 2012); reinforcement smoking outcome expectancies and negative
reinforcement smoking motives (Kauffman et al., 2017); and negative affect reduction
expectancies, coping motives for smoking, perceived barriers for smoking cessation, and severity
of problems experienced during past quit attempts (Rogers et al., 2018). Our findings represent
the first attempt to assess these constructs of interest under a single hypothesis using EMA
methodology, which offers ecological validity and reduced retrospective bias associated with
self-report cigarette use (Shiffman, Stone, & Hufford, 2008). Our findings indicate that the
92
DERS was not a significant predictor of cigarette use variability. Future studies could explore
negative affect variability and smoking outcomes to further elucidate the predictive value of
DERS on smoking outcomes, including whether the DERS can be used as a tool to help
differentiate habitual smokers (smokers that engage in a more consistent manner of smoking)
versus smokers that have a “prn” smoking style (smoke in the face of overwhelming stressors in
their environment). The DERS could be used as potential tool in smoking cessation interventions
and programs to identify and target at-risk individuals, who may be more vulnerable and
susceptible to smoking relapse.
Main effect of vmHRV on smoking behavior. The reason for the null linear association
between vmHRV and smoking behavior was most likely due to low power from a small sample
size and low compliance among participants. The predictive properties of vmHRV on smoking
behavior outcomes may also not be apparent in restful states; thus, this study may have been
limited by the lack of incorporation of a stress-based task. Ashare and colleagues (2012) found
that reduced vmHRV during a stressful imagery task was significantly associated with less time
to initiate smoking and increased craving among smokers. Libby and colleagues (2012) tested
the effects of mindfulness training on smoking and found that vmHRV reactivity (increase
versus decrease between baseline and during meditation tasks) predicted fewer cigarettes at
follow-up. In their limitations, researchers noted that a stressful versus relaxing task (meditation)
may elicit significant and more meaningful changes in vmHRV reactivity (Libby et al., 2012).
As noted by previous researchers, future studies may benefit from using a stressful task or stress-
based paradigm. This might elicit vmHRV changes that are prominent enough for detection and
may be more demonstrative of difficulties managing distressful states. Thus, vmHRV reactivity
93
versus vmHRV at rest may be more sensitive in detecting emotion regulatory changes that may
be predictive of smoking behaviors.
Limitations. There were several limitations to the study. To detect small to medium
effect sizes between constructs of interest, studies may need larger study samples to attain
sufficient statistical power. The long duration of the EMA procedures (28 days) may have been
too taxing on participants, and therefore, the main outcome analyses suffered from small samples
due to low overall compliance among smoker participants. Assessment of emotion regulation
was based on self-report (DERS) and vmHRV metrics at rest. Previous studies have shown that
HF-HRV reactivity or HF-HRV synchrony (the degree of synchrony between couples’ vmHRV
scores) during marital disputes are more predictive of marital satisfaction and health-related
outcomes (Smith et al., 2011; Wilson et al., 2018). Drawing from studies that have shown an
association between vmHRV reactivity and smoking outcomes, the current study may have
benefited from incorporating an emotion-based priming and/or a stress-based task to better elicit
changes to vmHRV metrics (Smith et al., 2011; Wilson et al., 2018). Within-subject comparisons
between vmHRV at rest and vmHRV during the task could objectively identify participants, who
show more difficulties in emotion regulation based on their physiological responses versus self-
report. The parent study precluded stress induction among couples using an interpersonal conflict
task; therefore, detection of significant differences in resting vmHRV among participants in
satisfactory versus unsatisfactory romantic relationships that were not accounted for by age, sex,
or BMI proved to be challenging. Based on past findings and the current study results, the
predictive properties of vmHRV on smoking behavior outcomes may be limited to vmHRV
change or reactivity in response to a stressor or induction of emotion.
94
Conclusions. Despite these limitations, no studies to date have assessed the association
between emotion regulation, vmHRV at rest, and smoking behavior in naturalistic settings in the
context of relationship satisfaction. More studies are needed to elucidate the predictive properties
of vmHRV reactivity in the context of interpersonal relationships and smoking behavior
outcomes. Rather than simply relying on self-report measures (which are subject to bias),
vmHRV reactivity can be used as an objective biomarker that could potentially identify
particularly vulnerable individuals, who may be more susceptible to smoking relapse due to their
interpersonal environment.
95
Table 8. Pairwise Associations between Baseline DERS and vmHRV Metrics at Baseline in
Study 3 (N=35)
Overall
DERS Accept Goals Impulse Aware Strategy Clarity lnRMSSD
High -
Frequency
Overall
DERS
1
Accept .809** 1
Goals .819** .503** 1
Impulse .918** .655** .842** 1
Aware .182 .108 .034 .103 1
Strategy .828** .625** .690** .772** .018 1
Clarity .854** .786** .623** .687** .012 .776** 1
lnRMSSD .210 .195 .316 .215 .289 .316 .168 1
High-
Frequency
.186 .153 .307 .184 .323 .336 .196 .961** 1
** p<0.01, * p<0.05
96
Table 9. Sample Demographics, DERS, Smoking Characteristics, vmHRV, and Smoking
Outcomes by Relationship Satisfaction Status in Study 3 (N=35)
High
Relationship
Satisfaction
(N=17)
Low
Relationship
Satisfaction
(N=18)
Mean (SD) or
N(%)
Mean (SD) or
N(%) p-value
Effect
Size
Demographics
Age 47.852 (10.888) 43.676 (12.284) .296
0.359
Females
b
3 (17.65) 12 (66.67) .003**
9.333
BMI 28.203 (5.825) 27.351 (7.270) .706
0.129
Difficulties in Emotion
Regulation (DERS) (1-5)
Overall
1.936 (0.625) 2.254 (0.850) .223
-0.427
Acceptance 2.029 (0.874) 2.352 (1.136) .356 -0.317
Goals 2.176 (0.886) 2.816 (1.296) .103
-0.576
Impulse 1.706 (0.853) 2.127 (1.091) .218
-0.431
Aware 2.240 (0.723) 2.111 (0.723) .625
0.177
Strategy 1.738 (0.844) 2.263 (1.010) .125
-0.533
Clarity 1.865 (0.809) 2.144 (0.760) .299
-0.357
Smoking Characteristics
Smoker 15 (88.24) 14 (77.78) .939 0.933
Age first smoked
(range 7-45 yrs) 17.286 (4.159) 19.333 (10.118) .488
-0.100
Age smoked regularly
(range 11-46 yrs) 20.143 (6.491) 22.333 (8.723) .452
-0.283
Cigarettes smoked daily in
past month (range 1-50) 14.571 (7.013) 12.800 (11.730) .629
0.182
FTCD
a
(range 0-10)
4.214 (1.968) 4.000 (2.803) .815
-0.027
Resting vmHRV at Baseline
lnRMSSD 3.075 (0.620)
3.387 (0.493)
.108 -.559
High-Frequency (HF)
4.927 (1.457)
5.827 (0.989)
.039** -.727
Smoking Outcomes
Mean number of cigs smoked 6.563 (4.412) 5.559 (4.834) .560 .216
Cigarette use variability 5.939 (7.521) 4.314 (6.836) .630 .230
a
Independent samples t-test;
b
Chi-squared test; **significant at p<.05; Effect size reported as Cohen’s D;
RMSSD = root mean square of successive differences in milliseconds (ms); lnRMSSD = natural log-
transformed RMSSD values HF: Reflection of auto-regressive modeling in log units
97
Table 10. Predictive Modeling of DERS and vmHRV on Cigarette Use Variability in Study 3
Cigarette Use
Variability
a
B
SE (B) 95% CI
Odds
Ratio
P value
DERS 0.018 0.028
-0.035, 0.073
1.018 0.852
vmHRV
lnRMSSD -0.400 0.334 -1.056, 0.256 0.670 0.232
Note: All models adjusted for sex, age, and baseline cigarette dependence (FTCD)
a
Generalized linear modeling (GLM) Cigarette Use Variability: 1= high cigarette use variability; 0=low cigarette use
variability
98
CHAPTER 5. DISCUSSION AND CONCLUSIONS
Limitations
Prominent limitations across studies. There are several limitations among the
dissertation studies that are particularly important to note and may help guide future
investigations that seek to investigate the interrelationships of DERS, vmHRV, and smoking
behavior. Tobacco use disorder and the presence of psychopathology can both function to reduce
vmHRV (Harte et al., 2013); therefore, parsing out the effects of depression symptoms on
vmHRV among smokers in Study 1 was challenging (Chapter 2). Study 1 did not implement
social stress and/or emotion regulation behavioral tasks within the laboratory session. Therefore,
the film task (showing positive negative, and neutral film clips) may not have induced sufficient
emotional arousal and emotion-driven regulation changes in vmHRV metrics compared to tasks
designed to induce acute stress, particularly among participants who were already subject to
blunted vmHRV due to their comorbid conditions (tobacco use disorder and depression)
(Schiweck et al., 2019). Additionally, the only direct measure of emotion regulation in this study
was based on self-report, which is subject to bias. Although Study 2 (Chapter 3) demonstrated
promising preliminary results regarding mindfulness training effects on vmHRV reactivity in
response to acute social stress (TSST), the lack of statistical power from a small sample size
limited the examination of linear associations, adjustment for influential covariates, and the
subsequent assessment of moderation effects of the intervention group. Additionally, due to the
control-group, post-test-only design, assessment of magnitude of change in relation to vmHRV
reactivity that was attributable to the intervention was not possible. Study 3 (Chapter 4) did not
incorporate stress induction among couples through the use of an interpersonal conflict task.
Previous studies have shown that HF-HRV synchrony during marital disputes within the
99
laboratory can predict marital satisfaction (Smith et al., 2011; Wilson et al., 2018). vmHRV at
rest may not be a sensitive enough indicator to differentiate the effects of relationship satisfaction
among smokers. Additionally, due to the nature of the parent study design for Study 3 (Chapter
4), comparisons between task-based vmHRV change and vmHRV at rest on smoking behavior
outcomes among couples were not possible.
Effects of respiration on vmHRV. Differences in respiration can impact vmHRV
metrics if participants are breathing slower than a specific frequency range (e.g., at a frequency
less than 0.15-0.4 Hz)(Laborde et al., 2017). Current recommendations surrounding the
assessment of vmHRV call for adjustment of respiration, if indicated (Laborde et al., 2017).
However, some researchers argue that adjustment for respiration removes variability associated
with neural regulation over the heart (Laborde et al., 2017). Thus, adjusting for respiration may
strip away crucial information that is integral to research questions regarding the influence of
high-order neural structures of the central autonomic network (CAN) on vmHRV (Larsen et al.,
2010). Despite this viewpoint, the lack of assessment of respiration across these dissertation
studies represents a limitation. RMSSD and HF-HRV are both representative of vagal tone or
parasympathetic nervous system dominance, and across all three studies, these metrics were
highly correlated (r>.90, p<.001). This observed high positive correlation between RMSSD and
HF-HRV has been corroborated in previous investigations (Laborde et al., 2017). To address the
lack of assessment of respiration, analyses of the main outcomes focused primarily on RMSSD,
which is not as affected by changes in respiration (Hill & Siebenbrock, 2009). HF-HRV metrics
were reported only when participants were resting and not engaged in a task. Previous studies
have demonstrated that respiration-based changes of vmHRV during rest are minimal, and
spontaneous or “normal” breathing (versus controlled breathing or breathing at a specific rate) is
100
optimal for assessment (Bertsch, Hagemann, Naumann, Schachinger, & Schulz, 2012; Larsen et
al., 2010). Additionally, standardized procedures were implemented in controlled laboratory
settings, thus reducing variability in breathing rates across smoker participants. Moreover, during
the experimental tasks, participants remained in a sedentary, seated position while their vmHRV
was assessed across all three studies, which would have reduced respiratory-induced changes in
vmHRV as a result of postural alterations (H. G. Kim et al., 2018).
Effects of COVID-19. Unfortunately, shortly following the acceptance of this
dissertation proposal, the advent of the COVID-19 pandemic in March 2020 prevented the
enrollment of more participants for an undefined period of time. In the months that followed the
discontinuation of in-person participants, all parent studies and their procedures ended and were
redesigned for virtual administration. These new procedures excluded vmHRV and experimental
smoking behavior assessment, which required in-person participant attendance. Therefore, the
majority of the initially proposed analytical plans that included the examination of moderation
effects on linear pathways were underpowered across all three studies. Despite this disruption,
new analytical plans were undertaken that were amenable to smaller sample sizes and focused on
the main aims of this dissertation, specifically examining 1) the main effect of the DERS on
vmHRV during various experimental paradigms to validate vmHRV as an objective biomarker
of emotion regulation; and 2) the respective main effects of DERS and vmHRV on smoking
behavior outcomes within laboratory and naturalistic settings.
Implications
The contributions of this dissertation include insight on two novel research questions,
including (1) can the DERS at baseline predict vmHRV in various experimental contexts (during
101
rest, during induction of emotion via film clips, and during induction of acute social stress via the
TSST) (2) can the DERS and vmHRV respectively predict smoking behavior outcomes in both
laboratory and real-world settings. These two questions represent major gaps within tobacco
addiction literature that have yet to be answered thoroughly by studies to date. Although
experimental tasks designed to induce emotion and stress may not translate to naturalistic
settings, nuanced investigations with high internal validity may point to the potential predictive
value of the DERS and vmHRV on smoking behavior outcomes, thus increasing their
applicability and utility in interventions that target tobacco use disorder. Additionally, the
incorporation of smoking behavior outcomes derived from ecological momentary assessment
contributes to the external validity of the study findings. Most previous studies that have
investigated emotion regulation and smoking outcomes relied solely on self-report measures.
Under a single hypothesis supported by relevant theory, this dissertation incorporated novel
methodology to existing studies and provided evidence that validates vmHRV as an objective
biomarker, capable of predicting smoking behavior outcomes and potentially identifying
vulnerable smokers who may be more susceptible to smoking relapse.
Association of DERS and vmHRV. The lack of significant findings surrounding the
linear association between the DERS and vmHRV across all three studies was surprising, given
contradictory evidence provided by previous research. Past studies have shown significant
associations (albeit low effect sizes) between the DERS and vmHRV at rest (Visted et al., 2017;
Williams et al., 2015). One possible reason for the discrepancy in study findings between these
dissertations studies and previous research was lack of statistical power. These studies had larger
sample sizes [N=63 (Visted et al., 2017) and N=183 (Williams et al., 2015)] and used 24-hour
recorded vmHRV measurements (Visted et al., 2017), which may be more sensitive in assessing
102
nuanced differences in emotion regulation difficulties compared to short-term vmHRV
measurements (Visted et al., 2017)(Laborde et al., 2017). Additionally, these studies used college
student populations, who were much younger than the dissertation study samples and may not
demonstrate the same blunted vmHRV indices compared to older participants, who smoke and
suffer from symptoms of depression (Visted et al., 2017; Williams et al., 2015). Previous studies
have shown that self-reported difficulties in emotion regulation decrease with age (Giromini et
al., 2017), which may have also been a contributing factor to the observed positively skewed
distribution of DERS scores among the samples of middle-aged participants.
Although non-significant findings between the DERS and vmHRV across studies could
be attributed to low statistical power and the effects of age and psychopathology on vmHRV,
there are several other factors that may explain the lack of observed linear associations,
particularly bias associated with self-report measures. The DERS is a trait-based, self-report
scale that relies on self-awareness of the participant to accurately and honestly report difficulties
in their ability to recognize, manage, and tolerate negative affect. In Study 1, only the (state-
based) S-DERS demonstrated a negative association (in the expected direction) with vmHRV,
meaning those who reported more difficulties in emotion regulation in the present moment had
lower vmHRV values. This suggests that participants may have been more accurate in their self-
assessment of their emotion regulation difficulties when using a more proximal, state-based
measure. A general limitation of self-report measures is response bias and the tendency to alter
answers based on socially desirability. All three study samples showed positively skewed
distributions of DERS scores, irrespective of the presence of pathological symptoms
(depression), pointing to a possible floor effect among the small study samples. Self-report bias
may be especially prominent among population samples with psychiatric conditions, who are
103
prone to alexithymia (inability to identify and label feelings) and who demonstrate limited self-
awareness capacity (Grabe et al., 2004; Moeller & Goldstein, 2014). Participants suffering from
symptoms of depression and tobacco use disorder may be prone to inaccuracies when completing
self-report measures that assess their awareness of internal emotional states and emotion
regulation capabilities. Therefore, potential floor effects (due to age, self-report bias, etc.) and
associated lack of variance among DERS scores may have been contributing factors to the null
findings. Investigations of vmHRV and behavioral measures of emotion regulation (versus self-
report) may be warranted to better detect linear associations between the DERS and vmHRV
among small sample sizes of participant smokers with and without depression symptomology.
vmHRV as a biomarker of emotion regulation. The transdiagnostic movement in
mental health has garnered increasing attention due to its feasibility within the clinical field, in
which mental health providers can personalize their interventions by treating an array of
transdiagnostic symptoms versus adhering to treatment plans dictated solely by clinical diagnosis
(Garland, 2014). This approach, endorsed by the National Institutes of Mental Health (NIMH),
not only offers individualized treatment designed to target comorbid symptomology, such as
difficulties in emotion regulation, across disorders but also provides the foundation for
identifying at-risk individuals (Fernandez et al., 2016). Adopting a transdiagnostic framework
can provide an innovative opportunity to better understand the underlying mechanisms driving
smoking behaviors among individuals with and without a psychiatric diagnosis, who may
experience a wide range of pathological symptoms.
Emotion regulation, defined as the ability to monitor and modulate emotional
experiences, has been identified as a promising transdiagnostic indicator across multiple
conditions, including tobacco use disorder and depression (Gross & Thompson, 2007). The
104
Research Domain Criteria (RDoC) is an organizing framework developed by a working group of
the National Institutes of Mental Health (NIMH) as a way to identify and organize domains of
psychopathology that align with innovative research developments in the fields of genetics and
clinical neuroscience (Insel et al., 2010). Within the originally published framework, Insel and
colleagues (2010) included five transdiagnostic domains including negative valence systems
(constructs such as threat and loss); positive valence systems (reward learning); cognitive
systems (attention, perception, memory); social processes (social communication, understanding
of self and others); and arousal and regulatory systems (sleep-wake, cardiac activity). The
criteria matrix for this framework represents a classification system and a “working target,” in
which investigators are expected to make alterations and additions as a result of new research
findings and developments. Fernandez and colleagues (2016) proposed to make emotion
regulation the 6
th
RDoC domain and provided substantial empirical evidence for its inclusion.
Although this study was published in 2016, vmHRV is notably missing from the physiology
domain. A small number of studies have assessed the predictive role of emotion regulation on
smoking outcomes, and only one out of five studies (Fucito et al., 2010) assessed three
components of the RDoC domain, including behavior, self-report, and paradigm (Adams et al.,
2012; Farris et al., 2016; Fucito et al., 2010; Kauffman et al., 2017; Rogers et al., 2018). The
majority of these studies relied on self-report measures and did not incorporate physiological
biomarkers such as vmHRV, which represents a major limitation in the field.
The utility of vmHRV as a viable and sensitive biomarker of emotion regulation remains
to be established. Findings from this dissertation indicate the necessity of additional research to
evaluate the associations between vmHRV reactivity (versus vmHRV at rest) and emotion
regulation so that it can be incorporated across transdiagnostic research studies that target the
105
comorbid conditions, such as depression and tobacco use disorder. The inclusion of vmHRV
assessment requires certain advances in research, most notably the validation of specific vmHRV
reactivity indices that are grounded in theory and are most sensitive to detecting emotion
regulation difficulties and smoking behavior outcomes (RMSSD, HF-HRV). Thus, future
studies with limited resources and small sample sizes may benefit from incorporating within-
subjects vmHRV assessment in the context of stress reactivity. Troy & Mauss (2011) posit that
emotion regulation strategies (specifically attention control and cognitive reappraisal) moderate
the association between perceived stressful events and resilience. When looking at vmHRV
reactivity in the context of stress, difficulties in emotion regulation may be a valuable indicator
that can differentiate individuals who are resilient to the effects of stress (adaptive emotional
responses) versus those who show impairment (maladaptive emotional responses)(Troy, 2011).
Future studies that are underpowered to detect a small effect between vmHRV at rest and
difficulties in emotion regulation may benefit from investigating within-subject vmHRV
reactivity in response to acute stress, as it may represent a more sensitive, objective metric that is
able to better demonstrate stress regulation capacity (encompassing emotion regulation and the
ability to withstand and regulate negative emotional states). vmHRV reactivity may serve as an
informative biomarker, representative of higher-order stress regulation processes and the
application of adaptive emotion regulation strategies. Additionally, interventions designed to
improve emotion and stress regulation (such as mindfulness training) to curb smoking behaviors
could incorporate vmHRV reactivity assessment, either as an objective indicator of successful
treatment effects or as a predictor of smoking behaviors, offering a great asset and tool for future
studies.
106
DERS as a predictor of smoking behavior outcomes. The predictive effects of DERS
score on smoking outcomes was apparent in Study 1, most likely reflecting an increase in
statistical power due to the repeated-measures, within-subjects design of the respective parent
study. In Study 1, the association between DERS score at baseline and number of cigarettes
smoked significantly differed between the deprived and non-deprived experimental sessions,
demonstrating the DERS may have more predictive value over number of cigarettes smoked
when participants are deprived from smoking. The results of this study indicate that the trait-
based DERS may be a reliable predictor of smoking behavior specifically in the context of
smoking deprivation, contributing to its potential utility in smoking cessation research and
interventions.
vmHRV as a predictor of smoking behavior outcomes. Although Study 2 and 3 did
not show that vmHRV was a significant predictor of smoking behavior outcomes (most likely
due to the parent study design and lack of power from a small sample size), findings from Study
1 (originating from an adequately powered, repeated measures, within-subjects design)
demonstrated that vmHRV at rest and during the film clip task predicted minutes to smoke status
while controlling for important covariates, including age, sex, cigarette dependence, and
smoking abstinence. These findings demonstrate the utility of vmHRV as an objective prognostic
biomarker of smoking behavior irrespective of smoking deprivation, offering a great contribution
to tobacco research and interventions. In Study 1, vmHRV values overall were significantly
higher during the non-deprived experimental session, providing data that support an observed
increase in vmHRV values within an abstinence period even shorter than 24 hours. This finding
offers important insight on best practices in methodology for future studies that include
assessment of vmHRV in the context of smoking deprivation, allowing researchers to improve
107
their study hypotheses and accommodate for an initial rise in vmHRV within 24 hours of
smoking abstinence.
Findings from Studies 2 and 3 point to the utility of vmHRV reactivity in the context of
acute social stress. Results from Study 2 show substantial preliminary differences in vmHRV
reactivity between intervention groups (mindfulness training versus and active control
condition), indicating that vmHRV reactivity to acute stress may provide important insight on
adaptive emotion regulation strategies and their effect on subsequent smoking behaviors. Only a
couple of studies have assessed the predictive properties of vmHRV reactivity on smoking
outcomes in a laboratory setting (Ashare et al., 2012; Libby et al., 2012). In their limitations,
researchers noted that stressful tasks may elicit more significant and meaningful changes in HF-
HRV reactivity that are capable of predicting smoking behavior longitudinally (Libby et al.,
2012). As previously mentioned in our recent review, the sensitivity of vmHRV to predict
relevant smoking behavior outcomes may be more apparent through a stress induction paradigm
(Christodoulou, Salami, & Black, 2020). Therefore, investigating the association between
vmHRV reactivity and smoking behaviors during acute social stress among a larger sample of
participants may better elucidate the unique predictive properties of vmHRV on smoking
behavior outcomes.
Future Research Directions
To validate the utility of vmHRV as a biomarker of emotion regulation on smoking
outcomes, future studies may benefit from the following recommendations: (1) assessment of
vmHRV reactivity in response to acute social stress as a more sensitive indicator of autonomic
regulatory response (and employment of adaptive emotion regulation strategies); (2) the
108
incorporation of behavioral emotion regulation tasks to account for limitations and bias
associated with self-report measures. Additionally, to address statistical power limitations
associated with smaller sample sizes and external factors that can affect vmHRV (such as
medications, alcohol, caffeine, physical fitness), within-subject study designs have been highly
recommended for vmHRV assessment (Laborde et al., 2017) Future investigations that assess
underlying mechanisms that may drive the association between emotion regulation and vmHRV
in the context of depression and tobacco use disorder may also be warranted, including both
neuroimaging studies and assessment of inflammatory markers.
Neuroimaging Studies. Understanding the association between vmHRV and neural
structures and networks associated with emotion regulation can be accomplished by
incorporating vmHRV assessment (vmHRV reactivity) in fMRI studies. Smokers could engage
in emotion regulation and/or social stress-based task(s) within the fMRI scanner and then
undergo a smoking lapse task outside the scanner to test the predictive properties of vmHRV
reactivity on smoking behavioral outcomes. The results of this research could reveal important
underlying mechanisms (regarding specific neural structures and network activation) that
correspond to specific changes in emotion/stress regulation represented by vmHRV reactivity
indices among smokers. By examining neural correlates of vmHRV, researchers can narrow
down which vmHRV metric (reactivity) is most sensitive to emotion regulation, validating
vmHRV as a cost-effective, objective biomarker of autonomic control and higher-order, neural
activity of the central autonomic network.
The Role of Inflammation. Inflammatory marker assessment can also enhance
understanding of the biological mechanisms that link depression, vmHRV, and smoking
behaviors. Both smoking and depression can disrupt functioning of the autonomic nervous
109
system and lead to reductions in vmHRV (Wang et al., 2013). Although most studies that have
examined the underlying associations between depression and vmHRV have been based on
animal models, researchers theorize that the biological mechanism that underlies depression and
vmHRV is rooted in both dysfunction of the autonomic nervous system and increased
inflammation (Kop et al., 2010).
vmHRV is controlled by the vagus nerve, specifically through its parasympathetic
inhibitory influence of the heart. Parasympathetic activation of the autonomic nervous system
through the vagus nerve (represented by vmHRV) is mediated by acetylcholine activity and
neurotransmission through two types of acetylcholine receptors, including muscarine receptors
and nicotine receptors (Huston & Tracey, 2011). The cholinergic anti-inflammatory pathway is
activated through signals from the vagus nerve to nAChRs (mirroring the pathway of
parasympathetic activation represented by vmHRV). Researchers have begun to look at vmHRV
as being a representative biomarker of both parasympathetic dominance and this anti-
inflammatory pathway. It is hypothesized that the reduced functioning of the vagus nerve in
humans and inability to activate the cholinergic anti-inflammatory pathway leads to an increase
in overall cytokine production (an inflammatory marker) and results in a reduction of vmHRV
(Huston & Tracey, 2011).
Raison and colleagues (2006) propose that depression can result from stress-induced
changes in autonomic control that result in increased inflammation, demonstrated by augmented
inflammatory markers among individuals with depression. The experience of a stressor typically
precedes the onset of a depressive episode (observed across cultures) and results in alterations in
the central nervous system, including reduced functionality of the prefrontal cortex; increased
production of stress-related hormones; and disruption in the metabolism of serotonin,
110
norepinephrine, and dopamine (Raison, Capuron, & Miller, 2006). Additionally, an increase in
proinflammatory cytokines can lead to changes in the autonomic nervous system and the
proliferation of “sickness” behaviors (weakness, fatigue, loss of concentration) that overlap with
depression symptomology (Dantzer, 2004). Additionally, patients that receive cytokines for the
treatment of viral diseases (such as Hepatitis C) show an approximately 50% chance of
developing depression and associated symptoms (Musselman et al., 2001). Although more
human studies are needed to explain the role of proinflammatory cytokines in depression, some
researchers postulate that one known pathway might be through the vagus nerve (Musselman et
al., 2001; Raison et al., 2006). Cytokine molecules, which are too large to penetrate the blood-
brain barrier, may access the central nervous system and affect neurotransmitter activity through
vagal afferent fibers, representing a possible pathway from the vagus nerve to increased
systematic inflammation to the eventual onset of depression (Musselman et al., 2001; Raison et
al., 2006).
Nicotine naturally increases activity of the vagus nerve and its activation of the
cholinergic anti-inflammatory pathway. However, increased overexposure to nicotine through
smoking behaviors is hypothesized to dysregulate the normal functioning of the vagus nerve by
desensitizing nAChRs. This dysregulation process may lead to sympathetic bias, defined as a
compensatory sympathetic reflex state triggered by nicotine withdrawal and activated in-between
smoking sessions (Yun, Bazar, Lee, Gerber, & Daniel, 2005). Because the vagus nerve is no
longer able to efficiently and effectively activate the cholinergic anti-inflammatory pathway due
to chronic activation, an increase in inflammation ensues among smokers (Huston & Tracey,
2011; Yun et al., 2005). Examining both the acute and chronic effects of nicotine on the vagus
nerve among humans is integral to understanding its effects. Generally, both acute and long-term
111
smoking leads to reduction in vmHRV, as indicated by blunted levels of resting vmHRV among
smokers versus non-smokers. Although more studies are needed to elucidate the underlying
mechanisms of these constructs, researchers hypothesize that chronic nicotine exposure
desensitizes nAChRs and leads to compensatory changes in the autonomic nervous system (Yun
et al., 2005). Consequently, the sympathetic nervous system dominates and functioning of the
vagus nerve is ultimately disrupted. As a result of continued tobacco use and addiction,
depressed vagal activity leads to reduced vmHRV indices and higher levels of inflammatory
markers among smokers (Yun et al., 2005).
Assessment of relevant inflammatory markers such as cytokines that are sensitive to
vagus nerve dysfunction, coupled with vmHRV reactivity assessment, can reveal important,
underlying systematic effects of smoking and can add to the current lack of human studies that
have assessed these constructs among smokers with and without depression. Therefore,
identifying specific inflammatory markers associated with stress/emotion regulation tasks and
vmHRV reactivity indices, as well as corroboration from neuroimaging studies that focus on
neural functionality, would serve as promising research directions for future studies.
Concluding Remarks
Significant findings support the utility of vmHRV as a viable predictor of smoking
outcomes. Specifically, the predictive properties of vmHRV on smoking behaviors may open up
the utility of vmHRV as a transdiagnostic biomarker in other fields as well, particularly studies
that investigate comorbid conditions characterized by difficulties in emotion regulation.
Furthermore, vmHRV reactivity may be used as an indicator of treatment or intervention effects
in clinical or research environments, specifically programs that focus on improving emotion
112
and/or stress regulation and curbing smoking behaviors. In clinical practice, vmHRV reactivity
can be assessed as an outcome variable (capable of predicting smoking behaviors longitudinally)
or measured routinely throughout smoking prevention efforts to validate treatment effects.
In conclusion, there are broad research and practice implications of this dissertation,
including the examination of vmHRV metrics of parasympathetic dominance that have not been
adequately investigated within tobacco addiction studies. This dissertation also offers important
methodological value and insight on the effects of abstinence on these metrics. Assessment of
vmHRV reactivity during emotionally valanced film clips and an acute social stressor followed
by the implementation of smoking behavioral tasks were particularly unique and represent
innovative research methodology. Preliminary findings contribute to existing vmHRV theory by
elucidating differences in vmHRV reactivity during acute social stress between participants who
underwent mindfulness training versus an active control condition. A common theme shared
across all studies is the determination of whether vmHRV during experimental paradigms (at
rest, during an induction of emotion, or during an induction of acute stress) represents a viable,
transdiagnostic, objective biomarker of emotion regulation. Findings point to the potential utility
of vmHRV reactivity as a representation of higher-order neural processes of the central
autonomic network that are crucial to emotion regulation. Additionally, this dissertation
validated the vmHRV as a prognostic indicator of smoking behavior outcomes, increasing the
internal and external validity of these linear associations and offering a great contribution to the
current literature.
113
REFERENCES
Adams, C. E., Tull, M. T., & Gratz, K. L. (2012). The role of emotional nonacceptance in the
relation between depression and recent cigarette smoking. Am J Addict, 21(4), 293-301.
Aldao, A., Nolen-Hoeksema, S., & Schweizer, S. (2010). Emotion-regulation strategies across
psychopathology: A meta-analytic review. Clin Psychol Rev, 30(2), 217-237.
Ashare, R. L., Sinha, R., Lampert, R., Weinberger, A. H., Anderson, G. M., Lavery, M. E., . . .
McKee, S. A. (2012). Blunted vagal reactivity predicts stress-precipitated tobacco
smoking. Psychopharmacology (Berl), 220(2), 259-268.
Audrain-McGovern, J., Wileyto, E. P., Ashare, R., Cuevas, J., & Strasser, A. A. (2014). Reward
and affective regulation in depression-prone smokers. Biol Psychiatry, 76(9), 689-697.
Baker, T. B., Piper, M. E., McCarthy, D. E., Majeskie, M. R., & Fiore, M. C. (2004). Addiction
motivation reformulated: an affective processing model of negative reinforcement.
Psychol Rev, 111(1), 33-51.
Beauchaine, T. P., & Thayer, J. F. (2015). Heart rate variability as a transdiagnostic biomarker of
psychopathology. Int J Psychophysiol, 98(2 Pt 2), 338-350.
Benowitz, N. L. (2010). Nicotine addiction. N Engl J Med, 362(24), 2295-2303.
Bertsch, K., Hagemann, D., Naumann, E., Schachinger, H., & Schulz, A. (2012). Stability of
heart rate variability indices reflecting parasympathetic activity. Psychophysiology, 49(5),
672-682.
Bjureberg, J., Ljotsson, B., Tull, M. T., Hedman, E., Sahlin, H., Lundh, L. G., . . . Gratz, K. L.
(2016). Development and Validation of a Brief Version of the Difficulties in Emotion
Regulation Scale: The DERS-16. J Psychopathol Behav Assess, 38(2), 284-296.
Bloch, L., Haase, C. M., & Levenson, R. W. (2014). Emotion regulation predicts marital
satisfaction: more than a wives' tale. Emotion, 14(1), 130-144.
Bodin, F., McIntyre, K. M., Schwartz, J. E., McKinley, P. S., Cardetti, C., Shapiro, P. A., . . .
Sloan, R. P. (2017). The Association of Cigarette Smoking With High-Frequency Heart
Rate Variability: An Ecological Momentary Assessment Study. Psychosom Med, 79(9),
1045-1050.
Bold, K. W., Yoon, H., Chapman, G. B., & McCarthy, D. E. (2013). Factors predicting smoking
in a laboratory-based smoking-choice task. Exp Clin Psychopharmacol, 21(2), 133-143.
Brown, L., Rando, A. A., Eichel, K., Van Dam, N. T., Celano, C. M., Huffman, J. C., & Morris,
M. E. (2020). The Effects of Mindfulness and Meditation on Vagally-Mediated Heart
Rate Variability: A Meta-Analysis. Psychosom Med, Publish Ahead of Print.
Cappelleri, J. C., Bushmakin, A. G., Baker, C. L., Merikle, E., Olufade, A. O., & Gilbert, D. G.
(2005). Revealing the multidimensional framework of the Minnesota nicotine withdrawal
scale. Curr Med Res Opin, 21(5), 749-760.
Carmody, T. P., Vieten, C., & Astin, J. A. (2007). Negative affect, emotional acceptance, and
smoking cessation. J Psychoactive Drugs, 39(4), 499-508.
Centers for Disease, C., & Prevention. (2013). Vital signs: current cigarette smoking among
adults aged >/=18 years with mental illness - United States, 2009-2011. MMWR Morb
Mortal Wkly Rep, 62(5), 81-87.
Chiesa, A., Serretti, A., & Jakobsen, J. C. (2013). Mindfulness: top-down or bottom-up emotion
regulation strategy? Clin Psychol Rev, 33(1), 82-96.
Childs, E., & de Wit, H. (2009). Hormonal, cardiovascular, and subjective responses to acute
stress in smokers. Psychopharmacology (Berl), 203(1), 1-12.
114
Christodoulou, G., Salami, N., & Black, D. S. (2020). The Utility of Heart Rate Variability in
Mindfulness Research. Mindfulness, 11, 554–570.
Coppotelli, H. C., & Orleans, C. T. (1985). Partner support and other determinants of smoking
cessation maintenance among women. Journal of consulting and clinical psychology,
53(4), 455-460.
Cox, L. S., Tiffany, S. T., & Christen, A. G. (2001). Evaluation of the brief questionnaire of
smoking urges (QSU-brief) in laboratory and clinical settings. Nicotine Tob Res, 3(1), 7-
16.
Cross, S. P., & Hickie, I. (2017). Transdiagnostic stepped care in mental health. Public Health
Res Pract, 27(2).
Cui, J. (2007). QIC program and model selection in GEE analyses. The Stata Journal, 7(2), 209-
220.
Dantzer, R. (2004). Cytokine-induced sickness behaviour: a neuroimmune response to activation
of innate immunity. Eur J Pharmacol, 500(1-3), 399-411.
Day, A. M., Kahler, C. W., Metrik, J., Spillane, N. S., Tidey, J. W., & Rohsenow, D. J. (2015).
Working Memory Moderates the Association Between Smoking Urge and Smoking
Lapse Behavior After Alcohol Administration in a Laboratory Analogue Task. Nicotine
Tob Res, 17(9), 1173-1177.
de Souza, I. C., de Barros, V. V., Gomide, H. P., Miranda, T. C., Menezes Vde, P., Kozasa, E.
H., & Noto, A. R. (2015). Mindfulness-based interventions for the treatment of smoking:
a systematic literature review. J Altern Complement Med, 21(3), 129-140.
DeGrandpre, R. J., Bickel, W. K., Higgins, S. T., & Hughes, J. R. (1994). A behavioral economic
analysis of concurrently available money and cigarettes. J Exp Anal Behav, 61(2), 191-
201.
Diamond, L., & Aspinwall, LG. (2003). Emotion regulation across the life span: An integrative
perspective emphasizing self-regulation, positive affect, and dyadic processes. Motivation
and Emotion, 27(2), 125–156.
Dickerson, S. S., & Kemeny, M. E. (2004). Acute stressors and cortisol responses: a theoretical
integration and synthesis of laboratory research. Psychol Bull, 130(3), 355-391.
Doherty, W. J., & Whitehead, D. A. (1986). The social dynamics of cigarette smoking: a family
systems perspective. Family Process, 25(3), 453-459.
Donoho, C. J., Seeman, T. E., Sloan, R. P., & Crimmins, E. M. (2015). Marital status, marital
quality, and heart rate variability in the MIDUS cohort. J Fam Psychol, 29(2), 290-295.
Engelmann, J. M., Gewirtz, J. C., & Cuthbert, B. N. (2011). Emotional reactivity to emotional
and smoking cues during smoking abstinence: potentiated startle and P300 suppression.
Psychophysiology, 48(12), 1656-1668.
Farris, S. G., Aston, E. R., Zvolensky, M. J., Abrantes, A. M., & Metrik, J. (2017).
Psychopathology and tobacco demand. Drug Alcohol Depend, 177, 59-66.
Farris, S. G., Zvolensky, M. J., & Schmidt, N. B. (2016). Difficulties with Emotion Regulation
and Psychopathology Interact to Predict Early Smoking Cessation Lapse. Cognit Ther
Res, 40(3), 357-367.
Ferdous M, F. S. (2014). Acute Effect of Cigarette Smoking On HRV in Current Cigarette
Smokers. Journal of Bangladesh Social Physiology, 9(2), 59-64.
Fernandez, K. C., Jazaieri, H., & Gross, J. J. (2016). Emotion Regulation: A Transdiagnostic
Perspective on a New RDoC Domain. Cognit Ther Res, 40(3), 426-440.
115
Fucito, L. M., Juliano, L. M., & Toll, B. A. (2010). Cognitive reappraisal and expressive
suppression emotion regulation strategies in cigarette smokers. Nicotine Tob Res, 12(11),
1156-1161.
Garland, E. L., Howard, M.O. . (2014). A transdiagnostic perspective on cognitive, affective, and
neurobiological processes underlying human suffering. Research on Social Work
Practice, 24(1), 142-151.
GD, G. (2015). Missing Values Analysis and Data Imputation. In S. A. Publishers (Ed.).
Gehricke, J. G., Loughlin, S. E., Whalen, C. K., Potkin, S. G., Fallon, J. H., Jamner, L. D., . . .
Leslie, F. M. (2007). Smoking to self-medicate attentional and emotional dysfunctions.
Nicotine Tob Res, 9 Suppl 4, S523-536.
Girard, D., Delgado-Eckert, E., Schaffner, E., Häcki, C., Adam, M., Stern, G. L., . . . Frey, U.
(2015). Long-term smoking cessation and heart rate dynamics in an aging healthy cohort:
Is it possible to fully recover? Environ Res, 143(Pt A), 39-48.
Giromini, L., Ales, F., de Campora, G., Zennaro, A., & Pignolo, C. (2017). Developing Age and
Gender Adjusted Normative Reference Values for the Difficulties in Emotion Regulation
Scale (DERS. Journal of Psychopathology and Behavioral Assessment, 39, 705-714.
Grabe, H. J., Spitzer, C., & Freyberger, H. J. (2004). Alexithymia and personality in relation to
dimensions of psychopathology. Am J Psychiatry, 161(7), 1299-1301.
Gratz, K. L., & Roemer, L. (2004). Multidimensional assessment of emo- tion regulation and
dysregulation: development, factor structure, and initial validation of the difficulties in
emotion regulation scale. Journal of Psychopathology and Behavioral Assessment, 26,
41-54.
Guendelman, S., Medeiros, S., & Rampes, H. (2017). Mindfulness and Emotion Regulation:
Insights from Neurobiological, Psychological, and Clinical Studies. Front Psychol, 8,
220.
Harte, C. B., Liverant, G. I., Sloan, D. M., Kamholz, B. W., Rosebrock, L. E., Fava, M., &
Kaplan, G. B. (2013). Association between smoking and heart rate variability among
individuals with depression. Ann Behav Med, 46(1), 73-80.
Hatsukami, D. K., Stead, L. F., & Gupta, P. C. (2008). Tobacco addiction. Lancet, 371(9629),
2027-2038.
Hayano, J., Yamada, M., Sakakibara, Y., Fujinami, T., Yokoyama, K., Watanabe, Y., & Takata,
K. (1990). Short- and long-term effects of cigarette smoking on heart rate variability. Am
J Cardiol, 65(1), 84-88.
Heart rate variability: standards of measurement, physiological interpretation and clinical use.
Task Force of the European Society of Cardiology and the North American Society of
Pacing and Electrophysiology. (1996). Circulation, 93(5), 1043-1065.
Heatherton, T. F., Kozlowski, L. T., Frecker, R. C., & Fagerstrom, K. O. (1991). The Fagerstrom
Test for Nicotine Dependence - a Revision of the Fagerstrom Tolerance Questionnaire.
British journal of addiction, 86(9), 1119-1127.
Hendrick, S. S., Dicke, A., & Hendrick, C. (1998). The Relationship Assessment Scale. Journal
of Social and Personal Relationships, 15(1), 137-142.
Hill, L. K., & Siebenbrock, A. (2009). Are all measures created equal? Heart rate variability and
respiration - biomed 2009. Biomed Sci Instrum, 45, 71-76.
Holzel, B. K., Lazar, S. W., Gard, T., Schuman-Olivier, Z., Vago, D. R., & Ott, U. (2011). How
Does Mindfulness Meditation Work? Proposing Mechanisms of Action From a
Conceptual and Neural Perspective. Perspect Psychol Sci, 6(6), 537-559.
116
Hughes, J. R., & Hatsukami, D. (1986). Signs and symptoms of tobacco withdrawal. Arch Gen
Psychiatry, 43(3), 289-294.
Huston, J. M., & Tracey, K. J. (2011). The pulse of inflammation: heart rate variability, the
cholinergic anti-inflammatory pathway and implications for therapy. J Intern Med,
269(1), 45-53.
John, O. P., & Gross, J. J. (2004). Healthy and unhealthy emotion regulation: personality
processes, individual differences, and life span development. J Pers, 72(6), 1301-1333.
Kahler, C. W., Metrik, J., Spillane, N. S., Day, A., Leventhal, A. M., McKee, S. A., . . .
Rohsenow, D. J. (2014). Acute effects of low and high dose alcohol on smoking lapse
behavior in a laboratory analogue task. Psychopharmacology (Berl), 231(24), 4649-4657.
Kauffman, B. Y., Farris, S. G., Alfano, C. A., & Zvolensky, M. J. (2017). Emotion dysregulation
explains the relation between insomnia symptoms and negative reinforcement smoking
cognitions among daily smokers. Addict Behav, 72, 33-40.
Keng, S. L., Smoski, M. J., & Robins, C. J. (2011). Effects of mindfulness on psychological
health: a review of empirical studies. Clin Psychol Rev, 31(6), 1041-1056.
Kiecolt-Glaser, J. K., & Newton, T. L. (2001). Marriage and health: his and hers. Psychol Bull,
127(4), 472-503.
Kim, H. G., Cheon, E. J., Bai, D. S., Lee, Y. H., & Koo, B. H. (2018). Stress and Heart Rate
Variability: A Meta-Analysis and Review of the Literature. Psychiatry Investig, 15(3),
235-245.
Kim, H. K., Pears, K. C., Capaldi, D. M., & Owen, L. D. (2009). Emotion dysregulation in the
intergenerational transmission of romantic relationship conflict. J Fam Psychol, 23(4),
585-595.
Kinnunen, T., Doherty, K., Militello, F. S., & Garvey, A. J. (1996). Depression and smoking
cessation: characteristics of depressed smokers and effects of nicotine replacement. J
Consult Clin Psychol, 64(4), 791-798.
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.
Kop, W. J., Stein, P. K., Tracy, R. P., Barzilay, J. I., Schulz, R., & Gottdiener, J. S. (2010).
Autonomic nervous system dysfunction and inflammation contribute to the increased
cardiovascular mortality risk associated with depression. Psychosom Med, 72(7), 626-
635.
Koval, P., Pe, M. L., Meers, K., & Kuppens, P. (2013). Affect dynamics in relation to depressive
symptoms: variable, unstable or inert? Emotion, 13(6), 1132-1141.
Laborde, S., Mosley, E., & Mertgen, A. (2018). Vagal Tank Theory: The Three Rs of Cardiac
Vagal Control Functioning - Resting, Reactivity, and Recovery. Front Neurosci, 12, 458.
Laborde, S., Mosley, E., & Thayer, J. F. (2017). Heart Rate Variability and Cardiac Vagal Tone
in Psychophysiological Research - Recommendations for Experiment Planning, Data
Analysis, and Data Reporting. Front Psychol, 8, 213.
Larsen, P. D., Tzeng, Y. C., Sin, P. Y., & Galletly, D. C. (2010). Respiratory sinus arrhythmia in
conscious humans during spontaneous respiration. Respir Physiol Neurobiol, 174(1-2),
111-118.
Lavender, J. M., Tull, M. T., DiLillo, D., Messman-Moore, T., & Gratz, K. L. (2017).
Development and Validation of a State-Based Measure of Emotion Dysregulation.
Assessment, 24(2), 197-209.
117
Lee, K. J., & Simpson, J. A. (2014). Introduction to multiple imputation for dealing with missing
data. Respirology, 19(2), 162-167.
Leventhal, A. M., & Zvolensky, M. J. (2015). Anxiety, depression, and cigarette smoking: a
transdiagnostic vulnerability framework to understanding emotion-smoking comorbidity.
Psychol Bull, 141(1), 176-212.
Li, C. S., Huang, C., Yan, P., Bhagwagar, Z., Milivojevic, V., & Sinha, R. (2008). Neural
correlates of impulse control during stop signal inhibition in cocaine-dependent men.
Neuropsychopharmacology, 33(8), 1798-1806.
Libby, D. J., Worhunsky, P. D., Pilver, C. E., & Brewer, J. A. (2012). Meditation-induced
changes in high-frequency heart rate variability predict smoking outcomes. Front Hum
Neurosci, 6, 54.
Lipponen, J. A., & Tarvainen, M. P. (2019). A robust algorithm for heart rate variability time
series artefact correction using novel beat classification. J Med Eng Technol, 43(3), 173-
181.
Liu, J. J. W., Ein, N., Peck, K., Huang, V., Pruessner, J. C., & Vickers, K. (2017). Sex
differences in salivary cortisol reactivity to the Trier Social Stress Test (TSST): A meta-
analysis. Psychoneuroendocrinology, 82, 26-37.
Mathew, A. R., Hogarth, L., Leventhal, A. M., Cook, J. W., & Hitsman, B. (2017). Cigarette
smoking and depression comorbidity: systematic review and proposed theoretical model.
Addiction, 112(3), 401-412.
McCraty, R., & Shaffer, F. (2015). Heart Rate Variability: New Perspectives on Physiological
Mechanisms, Assessment of Self-regulatory Capacity, and Health risk. Glob Adv Health
Med, 4(1), 46-61.
McKee, S. A. (2009). Developing human laboratory models of smoking lapse behavior for
medication screening. Addict Biol, 14(1), 99-107.
McKee, S. A., Weinberger, A. H., Shi, J., Tetrault, J., & Coppola, S. (2012). Developing and
validating a human laboratory model to screen medications for smoking cessation.
Nicotine Tob Res, 14(11), 1362-1371.
Mennin, D. S., & Fresco, D. M. (2015). Advancing Emotion Regulation Perspectives on
Psychopathology: The Challenge of Distress Disorders. Psychol Inq, 26(1), 80-92.
Mermelstein, R., Lichtenstein, E., & McIntyre, K. (1983). Partner support and relapse in
smoking-cessation programs. Journal of consulting and clinical psychology, 51(3), 465-
466.
Moeller, S. J., & Goldstein, R. Z. (2014). Impaired self-awareness in human addiction: deficient
attribution of personal relevance. Trends Cogn Sci, 18(12), 635-641.
Moreira, H., Gouveia, M.J. & Canavarro, M.C. . (2020). A bifactor analysis of the Difficulties in
Emotion Regulation Scale - Short Form (DERS-SF) in a sample of adolescents and
adults. Current Psychology.
Murgia, F., Melotti, R., Foco, L., Gogele, M., Meraviglia, V., Motta, B., . . . Pattaro, C. (2019).
Effects of smoking status, history and intensity on heart rate variability in the general
population: The CHRIS study. PLoS One, 14(4), e0215053.
Musselman, D. L., Lawson, D. H., Gumnick, J. F., Manatunga, A. K., Penna, S., Goodkin, R. S.,
. . . Miller, A. H. (2001). Paroxetine for the prevention of depression induced by high-
dose interferon alfa. N Engl J Med, 344(13), 961-966.
118
Neumann, A., van Lier, P. A., Gratz, K. L., & Koot, H. M. (2010). Multidimensional assessment
of emotion regulation difficulties in adolescents using the Difficulties in Emotion
Regulation Scale. Assessment, 17(1), 138-149.
Osborne, T. L., Michonski, J., Sayrs, J., Welch, S. S., & Anderson, L. K. (2017). Factor structure
of the Difficulties in Emotion Regulation Scale (DERS) in adult outpatients receiving
dialectical behavior therapy (DBT). Journal of Psychopathology and Behavioral
Assessment, 39(2), 355-371.
Parak J, K. I. (2013). Accuracy of Firstbeat Bodyguard 2 beat-to-beat heart rate monitor
Firstbeat Technologies Ltd. 2013. Retrieved from
https://assets.firstbeat.com/firstbeat/uploads/2015/11/white_paper_bodyguard2_final.pdf
Pomerleau, C. S., Carton, S. M., Lutzke, M. L., Flessland, K. A., & Pomerleau, O. F. (1994).
Reliability of the Fagerstrom Tolerance Questionnaire and the Fagerstrom Test for
Nicotine Dependence. Addict Behav, 19(1), 33-39.
Porges, S. W. (2007). The polyvagal perspective. Biol Psychol, 74(2), 116-143.
Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general
population. Applied psychological measurement, 1(3), 385-401.
Raison, C. L., Capuron, L., & Miller, A. H. (2006). Cytokines sing the blues: inflammation and
the pathogenesis of depression. Trends Immunol, 27(1), 24-31.
Renshaw KD, M. P., Caska, CM, Blais RK. (2011). The utility of the Relationship Assessment
Scale in multiple types of relationships. Journal of Social and Personal Relationships,
28(4), 435-447.
Rogers, A. H., Bakhshaie, J., Viana, A. G., Manning, K., Mayorga, N. A., Garey, L., . . .
Zvolensky, M. J. (2018). Emotion dysregulation and smoking among treatment-seeking
smokers. Addict Behav, 79, 124-130.
Salvatore, J. E., Kuo, S. I., Steele, R. D., Simpson, J. A., & Collins, W. A. (2011). Recovering
from conflict in romantic relationships: a developmental perspective. Psychol Sci, 22(3),
376-383.
Schaefer, A., Nils, F., Sanchez, X., & Philippot, P. (2010). Assessing the effectiveness of a large
database of emotion-eliciting films: A new tool for emotion researchers. Cognition and
Emotion, 24(7), 1153-1172.
Schiweck, C., Piette, D., Berckmans, D., Claes, S., & Vrieze, E. (2019). Heart rate and high
frequency heart rate variability during stress as biomarker for clinical depression. A
systematic review. Psychol Med, 49(2), 200-211.
Segerstrom, S. C., & Nes, L. S. (2007). Heart rate variability reflects self-regulatory strength,
effort, and fatigue. Psychol Sci, 18(3), 275-281.
Shaffer, F., & Ginsberg, J. P. (2017). An Overview of Heart Rate Variability Metrics and Norms.
Front Public Health, 5, 258.
Shiffman, S., Stone, A. A., & Hufford, M. R. (2008). Ecological momentary assessment. Annu
Rev Clin Psychol, 4, 1-32.
Sloan, E., Hall, K., Moulding, R., Bryce, S., Mildred, H., & Staiger, P. K. (2017). Emotion
regulation as a transdiagnostic treatment construct across anxiety, depression, substance,
eating and borderline personality disorders: A systematic review. Clin Psychol Rev, 57,
141-163.
Smith, T. W., Cribbet, M. R., Nealey-Moore, J. B., Uchino, B. N., Williams, P. G., Mackenzie,
J., & Thayer, J. F. (2011). Matters of the variable heart: respiratory sinus arrhythmia
119
response to marital interaction and associations with marital quality. J Pers Soc Psychol,
100(1), 103-119.
Spears, C. A., Hedeker, D., Li, L., Wu, C., Anderson, N. K., Houchins, S. C., . . . Wetter, D. W.
(2017). Mechanisms underlying mindfulness-based addiction treatment versus cognitive
behavioral therapy and usual care for smoking cessation. J Consult Clin Psychol, 85(11),
1029-1040.
Szasz, P. L., Szentagotai, A., & Hofmann, S. G. (2012). Effects of emotion regulation strategies
on smoking craving, attentional bias, and task persistence. Behav Res Ther, 50(5), 333-
340.
Tang, Y. Y., Holzel, B. K., & Posner, M. I. (2015). The neuroscience of mindfulness meditation.
Nat Rev Neurosci, 16(4), 213-225.
Tang, Y. Y., Yang, L., Leve, L. D., & Harold, G. T. (2012). Improving Executive Function and
its Neurobiological Mechanisms through a Mindfulness-Based Intervention: Advances
within the Field of Developmental Neuroscience. Child Dev Perspect, 6(4), 361-366.
Thayer, J. F., Ahs, F., Fredrikson, M., Sollers, J. J., 3rd, & 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. Neurosci Biobehav Rev, 36(2), 747-756.
Thayer, J. F., Hansen, A. L., Saus-Rose, E., & Johnsen, B. H. (2009). Heart rate variability,
prefrontal neural function, and cognitive performance: the neurovisceral integration
perspective on self-regulation, adaptation, and health. Ann Behav Med, 37(2), 141-153.
Thayer, J. F., & Lane, R. D. (2000). A model of neurovisceral integration in emotion regulation
and dysregulation. J Affect Disord, 61(3), 201-216.
Toll, B. A., O'Malley, S. S., McKee, S. A., Salovey, P., & Krishnan-Sarin, S. (2007).
Confirmatory factor analysis of the Minnesota Nicotine Withdrawal Scale. Psychol
Addict Behav, 21(2), 216-225.
Troy, A., & Mauss, I. (2011). Resilience in the face of stress: Emotion regulation as a protective
factor. In S. Southwick, Litz, B.,Charney, D., & Friedman, M. (Ed.), Resilience and
Mental Health: Challenges Across the Lifespan (pp. 30-44). Cambridge: Cambridge
University Press.
Tull, M. T., Gratz, K. L., Latzman, R. D., Kimbrel, N. A., & Lejuez, C. W. (2010).
Reinforcement sensitivity theory and emotion regulation difficulties: a multimodal
investigation. Personality and Individual Differences, 49, 989–994.
Vaughn, M., Matyastik Baier, ME. (1999). Reliability and validity of the Relationship
Assessment Scale. Journal of Family Therapy, 27(2), 137-147.
Vilagut, G., Forero, C. G., Barbaglia, G., & Alonso, J. (2016). Screening for Depression in the
General Population with the Center for Epidemiologic Studies Depression (CES-D): A
Systematic Review with Meta-Analysis. PLoS One, 11(5), e0155431.
Visted, E., Sørensen, L., Osnes, B., Svendsen, J. L., Binder, P. E., & Schanche, E. (2017). The
Association between Self-Reported Difficulties in Emotion Regulation and Heart Rate
Variability: The Salient Role of Not Accepting Negative Emotions. Front Psychol, 8,
328.
Voss, A., Schroeder, R., Heitmann, A., Peters, A., & Perz, S. (2015). Short-term heart rate
variability--influence of gender and age in healthy subjects. PLoS One, 10(3), e0118308.
Wang, Y., Zhao, X., O'Neil, A., Turner, A., Liu, X., & Berk, M. (2013). Altered cardiac
autonomic nervous function in depression. BMC Psychiatry, 13, 187.
120
Weinberger, A. H., George, T. P., & McKee, S. A. (2011). Differences in smoking expectancies
in smokers with and without a history of major depression. Addictive Behaviors, 36(4),
434-437.
Weinberger, A. H., Kashan, R. S., Shpigel, D. M., Esan, H., Taha, F., Lee, C. J., . . . Goodwin, R.
D. (2017). Depression and cigarette smoking behavior: A critical review of population-
based studies. Am J Drug Alcohol Abuse, 43(4), 416-431.
West, R., Ussher, M., Evans, M., & Rashid, M. (2006). Assessing DSM-IV nicotine withdrawal
symptoms: a comparison and evaluation of five different scales. Psychopharmacology
(Berl), 184(3-4), 619-627.
Williams, D. P., Cash, C., Rankin, C., Bernardi, A., Koenig, J., & Thayer, J. F. (2015). Resting
heart rate variability predicts self-reported difficulties in emotion regulation: a focus on
different facets of emotion regulation. Front Psychol, 6, 261.
Wilson, S. J., Bailey, B. E., Jaremka, L. M., Fagundes, C. P., Andridge, R., Malarkey, W. B., . . .
Kiecolt-Glaser, J. K. (2018). When couples' hearts beat together: Synchrony in heart rate
variability during conflict predicts heightened inflammation throughout the day.
Psychoneuroendocrinology, 93, 107-116.
Yotsukura, M., Koide, Y., Fujii, K., Tomono, Y., Katayama, A., Ando, H., . . . Ishikawa, K.
(1998). Heart rate variability during the first month of smoking cessation. Am Heart J,
135(6 Pt 1), 1004-1009.
Yun, A. J., Bazar, K. A., Lee, P. Y., Gerber, A., & Daniel, S. M. (2005). The smoking gun: many
conditions associated with tobacco exposure may be attributable to paradoxical
compensatory autonomic responses to nicotine. Med Hypotheses, 64(6), 1073-1079.
Abstract (if available)
Abstract
Tobacco use disorder remains a major public health issue within the United States. Cessation efforts that incorporate transdiagnostic indicators, such as difficulties in emotion regulation, can help identify particularly vulnerable smokers. The purpose of this dissertation was to determine whether vagally-mediated heart rate variability (vmHRV) could serve as a reliable, transdiagnostic biomarker of emotion regulation, capable of predicting smoking behavior outcomes. This dissertation aimed to (1) test the linear association between difficulties in emotion regulation (DERS) and vmHRV (Study 1-3) during various experimental paradigms (induction of emotion, induction of acute social stress, and at rest) to validate vmHRV as an objective marker of emotion regulation and (2) to test whether DERS and vmHRV could predict smoking behavior outcomes in laboratory (Studies 1 & 2) and naturalistic settings (Study 3). For Aim 1, all studies revealed non-significant findings regarding the main effect of DERS on vmHRV (Studies 1-3). For Aim 2, generalized estimating equation (GEE) modeling in Study 1 revealed a significant interaction effect of DERS and experimental session type (deprived of smoking versus not deprived of smoking) on number of cigarettes smoked during a smoking reinstatement task. Additionally, GEE results in Study 1 demonstrated that the odds of smoking during a smoking reinstatement task versus waiting to smoke until the task was complete was less for every unit increase in vmHRV during rest and during the experimental film task. Studies 2 and 3 produced null findings surrounding main effects of DERS and vmHRV on smoking behavior outcomes, most likely due to a small sample sizes (N~20) and a lack of power to detect statistically significant effects. Nonetheless, preliminary results from Studies 2 and 3 point to the utility of vmHRV reactivity for future investigations and offer a foundation of knowledge surrounding methodology and best practices in vmHRV assessment. The results of this dissertation offer great insight to the current literature and emphasize the potential predictive value of vmHRV on smoking behavior outcomes, increasing its applicability and utility in research studies that target tobacco addiction.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Cultural risk and protective factors for tobacco use behaviors and depressive symptoms among American Indian adolescents in California
PDF
The role of depression symptoms on social information processing and tobacco use among adolescents
PDF
Individual differences in heart rate response and expressive behavior during social emotions: effects of resting cardiac vagal tone and culture, and relation to the default mode network
PDF
Tobacco and marijuana surveillance using Twitter data
PDF
Developing and testing novel strategies to detect inattentive responding in ecological momentary assessment studies
PDF
Examining tobacco regulation opinions and policy acceptance among key opinion leaders and tobacco retailers in low socioeconomic status African American, Hispanic, and non-Hispanic White communities
PDF
Energy drink consumption, substance use and attention-deficit/hyperactivity disorder among adolescents
PDF
Applications of mindfulness toward promoting psychosocial health in diverse communities
PDF
A network analysis of online and offline social influence processes in relation to adolescent smoking and alcohol use
PDF
The acute and longitudinal associations between sedentary behaviors, affective states, and emotional disorder symptoms among youth
PDF
Prospective associations of stress, compulsive internet use, and posttraumatic growth among emerging adults
PDF
Anxiety symptoms and nicotine use among adolescents and young adults
PDF
Adolescent social networks, smoking, and loneliness
PDF
Role transitions, past life events, and their associations with multiple categories of substance use among emerging adults
PDF
Exposure to alcohol advertising on television and alcohol use among young adolescents
PDF
Free-living and in-lab effects of sedentary time on cardiac autonomic nervous system function in youth with overweight and obesity
PDF
A sociocultural and developmental approach to intimate partner violence among a sample of Hispanic emerging adults
PDF
Biological and behavioral correlates of emotional flexibility and associations with exposure to family aggression
PDF
Dedifferentiation of emotion regulation strategies in the aging brain: an MVPA investigation
PDF
The role of social support in the relationship between adverse childhood experiences and addictive behaviors across adolescence and young adulthood
Asset Metadata
Creator
Christodoulou, Georgia
(author)
Core Title
Emotion regulation and heart rate variability among smokers
School
Keck School of Medicine
Degree
Doctor of Philosophy
Degree Program
Preventive Medicine (Health Behavior Research)
Degree Conferral Date
2021-08
Publication Date
07/15/2021
Defense Date
03/26/2021
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
emotion regulation,heart rate variability,mindfulness,OAI-PMH Harvest,Smoking,tobacco use disorder,vagal tone
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Black, David (
committee chair
), Kirkpatrick, Matt G. (
committee chair
), Mather, Mara (
committee member
), Pang, Raina (
committee member
), Unger, Jennifer (
committee member
)
Creator Email
georgia.christo@gmail.com,georgiac@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-oUC15582114
Unique identifier
UC15582114
Legacy Identifier
etd-Christodou-9744
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Christodoulou, Georgia
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Repository Email
cisadmin@lib.usc.edu
Tags
emotion regulation
heart rate variability
mindfulness
tobacco use disorder
vagal tone