University of Southern California Dissertations and Theses

Distress tolerance and mindfulness disposition: associations with substance use during adolescence and emerging adulthood

(USC Thesis Other)

Distress tolerance and mindfulness disposition: associations with substance use during adolescence and emerging adulthood

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Copyright 2020 Afton Victoria Kechter

DISTRESS TOLERANCE AND MINDFULNESS DISPOSITION: ASSOCIATIONS WITH
SUBSTANCE USE DURING ADOLESCENCE AND EMERGING ADULTHOOD

by

Afton Victoria Kechter

_________________________________________________________

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PREVENTIVE MEDICINE/HEALTH BEHAVIOR RESEARCH)

May 2020

ii
Table of Contents
Acknowledgments ........................................................................................................................................ iii
List of Tables ................................................................................................................................................ iv
List of Figures............................................................................................................................................... vi
Abstract........................................................................................................................................................ vii
Chapter 1. Introduction .................................................................................................................................. 1
Background and Significance ........................................................................................................... 1
Malleable Traits/Skills for Stress Reactivity .................................................................................... 2
Specific Aims ................................................................................................................................... 8

Chapter 2. Distress Tolerance and Subsequent Substance Use Throughout High School .......................... 10
A.1 Introduction ............................................................................................................................. 12
A.2 Methods ................................................................................................................................... 15
A.3 Results ..................................................................................................................................... 20
A.4 Discussion................................................................................................................................ 23
A.5 Figures and Tables ................................................................................................................... 26

Chapter 3. Longitudinal Associations Between Substance Use and Distress Tolerance Trajectory Groups
Throughout High School ............................................................................................................................. 34
B.1 Introduction.............................................................................................................................. 36
B.2 Methods ................................................................................................................................... 39
B.3 Results...................................................................................................................................... 44
B.4 Discussion ................................................................................................................................ 47
B.5 Figures and Tables ................................................................................................................... 50

Chapter 4. Facets of Mindfulness Disposition and Frequency of Substance Use After High School ........ 62
C.1 Introduction.............................................................................................................................. 64
C.2 Methods ................................................................................................................................... 67
C.3 Results...................................................................................................................................... 71
C.4 Discussion ................................................................................................................................ 73
C.5 Figures and Tables ................................................................................................................... 76

Chapter 5. Discussion .................................................................................................................................. 87
Overview ........................................................................................................................................ 87
Key Findings .................................................................................................................................. 87
Implications and Conclusions ........................................................................................................ 96
Bibliography ................................................................................................................................................ 98
Appendix ................................................................................................................................................... 110

iii
Acknowledgments

I would like to thank my committee for their guidance and support. Each member played a
significant and complementary role to my growth and achievements that I am deeply grateful
for. A special thank you to my parents, partner, and puppy for their unwavering love.

iv
List of Tables

Table A1. Baseline Participant Descriptive Characteristics.......................................................28

Table A2. Baseline Distress Tolerance Predicting Past 30-Day Use of Substances Across
High School...............................................................................................................29

Table A3. Moderators of Baseline Distress Tolerance Predicting Past 30-Day Use of
Substances Across High School................................................................................30

Supplemental Table A1. Baseline Distress Tolerance Mean and Subscales Predicting Past
30-Day Substance Use by Baseline Never vs. Ever User................. 31

Supplemental Table A2. Baseline Distress Tolerance Predicting Past 30-Day Use of
Substances Across High School Among Baseline Never-Users.......32

Supplemental Table A3. Baseline Distress Tolerance Predicting Past 30-Day Use of
Substances Across High School Among Baseline Ever-Users.........33

Table B1. Distress Tolerance Scale Across 4-Years of High School........................................53

Table B2. Linear versus Quadratic Growth Model Comparison for Distress Tolerance Scale
Across High School...................................................................................................54

Table B3. Quadratic Growth Mixture Model Fit Statistics for the Distress Tolerance Scale
Across High School...................................................................................................55

Table B4. Participant Characteristics by Distress Tolerance Classes........................................56

Table B5. Past 30-Day Use Frequency Baseline Intercept and Linear Slope Across High
School on Distress Tolerance Class...........................................................................57

Supplemental Table B1. Model Fit Statistics For Alcohol Use Growth Model Comparisons
Across High School...........................................................................60

Supplemental Table B2. Past 30-Day Substance Use Linear Growth Models Across High
School................................................................................................61

Table C1. Baseline Participant Descriptive Characteristics.......................................................77

Table C2. Correlation Matrix of Five-Facet Mindfulness Questionnaire Subscales, Substance
Use, and Mental Health Variables.............................................................................78

Table C3. Effect of Mindfulness Disposition Skills on Frequency of Past 30-Day Substance
Use Adjusted for Sociodemographics and Mental Health.........................................79

v
Table C4. Interactive Effects Of Mindfulness Disposition Skills on Frequency of Past
30-Day Substance Use Adjusted for Sociodemographics and Mental Health..........80

Supplemental Table C1. Acting with Awareness and Non-Reactivity on Past 30-Day Use
Frequency of Alcohol and E-Cigarettes............................................84

Supplemental Table C2. Observing and Non-Reactivity on Past 30-Day Use Frequency of
Cannabis............................................................................................85

Supplemental Table C3. Acting with Awareness and Non-Judging on Past 30-Day Use
Frequency of Alcohol........................................................................86

Table D1. Protective Subscales Across the DTS and FFMQ-SF...............................................95

Appendix 1. Distress Tolerance Scale..........................................................................................110

Appendix 2. Five-Facet Mindfulness Questionnaire-Short Form................................................111

Appendix 3. Average Past 30-Day Substance Use by Distress Tolerance Class.........................112

vi
List of Figures

Figure A1. Analytic Sample Based on Study Accrual and Baseline Distress Tolerance Scale
and Substance Use Respondents................................................................................26

Figure A2. Baseline Distress Tolerance Predicting Past 30-Day Use of Substances Across
High School...............................................................................................................27

Figure B1. Analytic Sample Based on Study Accrual and Distress Tolerance Scale
Respondents...............................................................................................................50

Figure B2. Conceptual Model of Latent Substance Use Intercept and Slope on Distress
Tolerance Scale Classes Across High School...........................................................51

Figure B3. Distress Tolerance Scale Sample Means of Three Class Trajectories Across High
School........................................................................................................................52

Supplemental Figure B1. Distress Tolerance Scale Sample Means for Two Class Trajectories
Across High School...........................................................................58

Supplemental Figure B2. Distress Tolerance Scale Sample Means for Four Class Trajectories
Across High School...........................................................................59

Figure C1. Acting with Awareness and Non-Reactivity on Past 30-Day Alcohol Use..............76

Supplemental Figure C1. Acting with Awareness and Non-Reactivity on Past 30-Day
E-Cigarette Use..................................................................................81

Supplemental Figure C2. Observing and Non-Reactivity on Cannabis Use on Past 30-Day
Cannabis Use.....................................................................................82

Supplemental Figure C3. Acting with Awareness and Non-Judging on Past 30-Day Alcohol
Use.....................................................................................................83

Appendix 4. Alcohol Use by Distress Tolerance Class Across High School..............................113

Appendix 5. Cigarette Use by Distress Tolerance Class Across High School.............................114

Appendix 6. E-Cigarette Use by Distress Tolerance Class Across High School.........................115

Appendix 7. Cannabis Use by Distress Tolerance Class Across High School............................116

vii
Abstract
The goal of the current dissertation was to deepen the scientific understanding of adolescent substance
use etiology, particularly in regard to malleable individual traits/skills related to stress reactivity (i.e.,
distress tolerance and mindfulness disposition). In a large, diverse, cohort of adolescents and emerging
adults from the Happiness & Health Study, the specific research objectives were to: (1) examine the
association of distress tolerance with subsequent substance use throughout high school; (2) characterize
developmental growth trajectories of the Distress Tolerance Scale (DTS) across high school and
examine associations between trajectories of substance use across high school with DTS trajectory
membership; and (3) examine the association of mindfulness disposition by mechanisms (subscales of
attention monitoring and acceptance in the Five-Facet Mindfulness Questionnaire [FFMQ]) with
frequency of substance use after high school. Findings suggest (1) prior to substance use, greater
distress tolerance may be a protective factor against substance use frequency, (2) after initial substance
use and with greater rate of use (specifically alcohol), optimal distress tolerance development may
decline, and (3) the association between mindfulness facets and substance use is multifaceted and only
one combination of attention monitoring and acceptance facets emerged as protective against substance
use frequency. Collectively, four skills captured by the DTS and FFMQ subscales emerged as protective
against greater substance use frequency in our diverse sample of adolescents and emerging adults: DTS
absorption, DTS appraisal, FFMQ acting with awareness, and FFMQ non-reactivity. Future research
should examine the factor structure of these subscales to evaluate the viability of developing a
condensed and integrative scale. The overarching themes of these subscales appear to be (1) intentional
awareness in the present moment, (2) accepting and tolerating stressors, and (3) choosing an adaptive
response. Given the transdiagnostic processes of these skills, such an integrative scale has the potential
to more precisely test associations between protective stress reactivity traits/skills and transdiagnostic
health outcomes to inform larger public health prevention program initiatives.

1
Chapter 1. Introduction

Background and Significance

Adolescence is a pivotal period in human development between childhood and adulthood
spanning from ages 10-24 (Sawyer, Azzopardi, Wickremarathne, & Patton, 2018). Specifically
during middle to late adolescence, corresponding to ages 14-19 or roughly the high school years,
the adolescent grows and learns rapidly by engaging in behaviors that promote skill
development, including high peer socialization and risk taking (Dahl, Allen, Wilbrecht, &
Suleiman, 2018). Due to neurobiological and hormonal changes, this period is also commonly
characterized by heightened emotional states whereby developing adaptive coping capacities,
such as emotion regulation skills, becomes important for successfully navigating the
socioemotional demands (Skinner & Zimmer-Gembeck, 2007). One longstanding concern during
this developmental time period relates to the adolescents increased experimentation with health -
risk behaviors, such as substance use. Previous research indicates early substance use has
detrimental effects on the adolescent brain development and remodeling (Ewing, Sakhardande,
& Blakemore, 2014), which may affect their ability to navigate socioemotional demands in an
adaptive, and healthy manner.
Efforts to minimize substance use during adolescence and prevent the development of
misuse patterns is a longstanding public health priority. Addiction and commonly co-occurring
mental and physical health ailments is a preventable public health issue (Sandler, 2014).
However, despite alcohol and drug prevention efforts, short- and long-term program effects
remain mixed (Skara & Sussman, 2003). Risks leading to adolescent substance use
experimentation and long-term consequences from subsequent misuse/addiction vary based on a

2
number of neurocognitive predispositions and neural exposures during development (Whyte,
2018) as well as individual traits/skills and receptivity leading to successful prevention and
treatment programs (Davis, Kendall, & Suveg, 2019). Thus, deepening the scientific
understanding of adolescent substance use etiology, particularly in regard to malleable individual
traits/skills related to stress reactivity, temporal nature of substance use exposure, and new
products and trends (e.g., vaping nicotine and cannabis), is imperative in order to continue to
update and align prevention efforts suitably.

Malleable Traits/Skills for Stress Reactivity
While associations between psychopathology symptomology (i.e., anxiety and depression
disorders) and substance use problems have been well-documented in the literature (Miettunen,
Murray, Jones, Maki, Ebeling, Taanila, & Moilanen, 2014; Sandler, 2014; Zubrick, 2012), there
are also naturally occurring stress reactivity traits/skills that may have incremental predictive
value beyond this commonly developed comorbidity. Identifying the degree to which these
traits/skills act as risk factors beyond mental health symptoms could hold important prevention
implications (Krank & Robinson, 2017); particularly, for traits that may be poor even prior to
mental health symptom development and are capable of being enhanced skills through training,
such as distress tolerance and mindfulness disposition (Creswell, 2017; Veilleux, 2019). Both of
these traits/skills may directly reflect the propensity one has to act on aversive experiences by
using substances and low trait/skill levels may partially explain why some youth are at greater
risk for both initiating and/or maintaining substance use compared to their peers.
In adult substance use literature, it is well-established that substance use disorders are
maintained through the alleviation of negative emotional states and a hijacked reward system.
However, this process of addiction takes time before a habitual appetitive intake response is

3
established. For example, when an individual experiences a challenge or stressor beyond their
perceived capacity to navigate or cope, a common stress-initiated risk chain of
neurophysiological arousal, perseverative cognition, and negative affect follows (Garland,
Boettiger, & Howard, 2011). This becomes a learned response of the sympathetic nervous
system (i.e., fight or flight) and interrupts the parasympathetic nervous system (i.e., rest and
digest) such that when substance intake is repeated to alleviate distress (vs. more adaptive coping
behaviors such as exercise), then a maladaptive behavior is paired through negative
reinforcement (Baker, Piper, McCarthy, Majeskie, & Fiore, 2004). During adult substance use
treatment, a key focus is learning how to manage stress and negative emotional states.
Adolescents with increased stress-reactivity (or poorer abilities to respond to stress
adaptively) are especially vulnerable to maladaptive behaviors such as being more likely to
initiate substance use and progress to addiction or relapse (Chaplin et al., 2010; Liu et al., 2012).
In order to maximize the efficacy of behavioral interventions, a better understanding of common
contributing and underlying problems such as stress reactivity and other malleable markers that
cut across physical and mental health conditions have the potential to enhance transdiagnostic
interventions (Stein, Freeman, & Smits, 2019). There are a number of constructs that aim to
capture stress reactivity. Two mechanisms to be focused on in this dissertation are distress
tolerance and mindfulness disposition. Previous research has reported that lower levels of both
distress tolerance and mindfulness disposition cluster together and associate with greater risk
health behaviors including greater substance use and obesogenic behaviors (Warren, Kechter,
Christodoulou, Cappelli, & Pentz, 2020). Therefore, it is important to deepen our scientific
understanding of the role that these malleable risk factors (i.e., distress tolerance and mindfulness
disposition) may have on substance use frequency, particularly during the critical developmental

4
period of adolescence. Findings have the potential to inform ways to update prevention
intervention efforts by considering the most salient mechanisms to target and ultimately reduce
the unfolding of risk during development.
Distress tolerance is a construct reflecting ones ability to withstand aversive
psychological and physical experiences (e.g., negative affect, disturbing thoughts, physical
discomfort; Leyro, Zvolensky, & Bernstein, 2010). Distress tolerance has been conceptualized as
a core component of emotion regulation and plays a role in the link between emotional
psychopathology and substance use (Leventhal & Zvolensky, 2015). Negative reinforcement and
experiential/avoidant coping theories indicate people with low distress tolerance are more likely
to engage in maladaptive coping strategies (such as substance use) to alleviate unpleasant
experiences, through motivation of managing negative affect and cravings (Baker et al., 2004;
Trafton, 2011). This framework suggests that people with low distress tolerance initiate
substance use as a low-effort regulation method and fail at cessation as low-effort relief method
from distress-eliciting experiences (Leventhal & Zvolensky, 2015; Leyro et al., 2010; Veilleux,
2019). Alternatively, it is possible that maladaptive coping strategies (such as substance use)
alters reward responsivity (Trafton, 2011) and may change the developmental course of or lower
ones ability to tolerate distress. Research is needed to further test these opposing hypotheses of
the role distress tolerance plays particularly during early (versus later) stages of substance use.
The current state of the field for distress tolerance research is emphasizing the importance
of examining contextual and temporal factors during early stages of substance use to better
inform theory and practice (Leventhal & Zvolensky, 2015; Veilleux, 2019). At large, distress
tolerance has been implicated as a relatively stable, multifaceted, individual trait and risk factor
for substance use (i.e., lower levels are associated with smoking status and behavior) and target

5
in cessation outcomes (Cummings et al., 2013; Veilleux, 2019). However, varying context,
sample characteristics, measurements, and definitions of the distress tolerance construct may
partially explain inconsistent findings and more work is needed to refine the current evidence-
base and conceptualization of distress tolerance in substance use, particularly during adolescence
(Glassman et al., 2016). In our current conceptualization, low distress tolerance is likely both a
vulnerability factor by which some youth are at greater risk for escalating and consequence from
maintaining substance use.
Mindfulness disposition is another trait and psychological capacity, capable of being
enhanced (E. Garland, Gaylord, & Park, 2009). Mindfulness disposition refers specifically to the
degree to which one is purposefully attentive to and aware of the present moment, and ability to
accept all experiences (positive, negative, and neutral internal and external) moment-by-moment
(Brown & Ryan, 2003; Crane et al., 2017; Kabat-Zinn, 1982). Greater levels of mindfulness
disposition have been positively associated with psychological health, negatively associated with
substance use, and associated with improved outcomes across domains of mental health and
addiction following mindfulness interventions (Bowen & Enkema, 2014; Creswell, 2017;
Robinson, Ladd, & Anderson, 2014; Tomlinson, Yousaf, Vittersø, & Jones, 2017). Altogether,
there is a general consensus that mindfulness disposition is an intrapersonal characteristic that
alters the degree to which individuals perceive events as threatening (Kechter et al., 2019) and
permits adaptive (vs. maladaptive) coping strategies (e.g., positive reappraisal and solution-
focus) that interposes previously discussed cycles in the addiction process (Black, 2012, 2014).
Currently, the field of mindfulness research is prioritizing the examination of the
mechanisms (i.e., why, how, and for whom) underlying these beneficial health or treatment
outcomes (Garland & Howard, 2018). Psychometric research has established mindfulness

6
disposition is a multidimensional construct with five distinct facets: observing, acting with
awareness, non-reactivity, non-judging, and describing (Baer, Hopkins, Krietemeyer, Smith, &
Toney, 2006). However, it is unclear whether all five facets correlate with more adaptive
responses or if select facets/skills are more protective. Monitor and Acceptance Theory (MAT),
proposes that not all of these mindfulness facets are equally important; only the attention
monitoring (the what of mindfulness captured by the observing facet) and the acceptance (the
how of mindfulness captured by the non-reactivity and non-judging facets) are necessary,
sufficient, and together explain how mindfulness improves negative affect, stress, and health
outcomes (Lindsay & Creswell, 2017). Future research is needed to directly test this theory.

Gaps in Literature
Based on the existing theories and research just discussed, specifically in regard to distress
tolerance and mindfulness disposition during early stages of substance use, the following gaps
will be examined in this dissertation.
1. Is distress tolerance associated with substance use frequency in adolescents? Empirical
data on perceived distress tolerance in relation to progression across the early stages of
the substance use trajectory as well as the temporal nature between distress tolerance and
substance use across adolescence is lacking (Leventhal & Zvolensky, 2015). Specifically,
the field currently lacks developmental work evaluating distress tolerance as a
prospective predictor of substance use (Veilleux, 2019). Addressing the temporal
association is needed to help ascertain whether low distress tolerance is a factor, beyond
mental health, by which some youth are at a greater or lower risk for escalating substance
use frequency.

7
2. Does substance use during adolescence lower or change the course of distress tolerance
development? Next, the field also lacks work addressing the competing explanation that
substance use lowers or changes the course of distress tolerance (Veilleux, 2019).
Specifically, no published research has examined the growth trajectories of distress
tolerance during high school in association with real outcomes such as substance use
frequency. Addressing this alternative temporal association is needed to help ascertain
whether low distress tolerance is a consequence from substance use.
3. Which mechanisms of mindfulness are protective against substance use frequency in
emerging adults? Limited literature on substance use was used to develop the
aforementioned mechanisms of mindfulness theory (MAT) and no published literature
has since examined the interaction of mindfulness facets and associations with substance
use in a large and diverse sample of emerging adults (Eisenlohr-Moul, Walsh, Charnigo,
Lynam, & Baer, 2012; Tomlinson et al., 2017). Given the multifaceted nature of
mindfulness disposition and complexity of interventions aimed at enhancing related skills
(i.e., mindfulness-based interventions), addressing which combinations of mindfulness
skills associate with lower substance use hold potential for future prevention programs to
consider targeting those skills in precision interventions.

Across these three dissertation studies, traditional and novel substances are tested. There has
been a recent evolution in substance use products affecting adolescents, such as the legalization
of cannabis and popularization of e-cigarettes. Such products have raised urgent concerns for
adolescent health due to the high risk addiction potential associated with many public health
concerns including transition to combustible products, decline in respiratory health, as well as

8
distress impairment and mood problems (Barrington-Trimis & Leventhal, 2018; Gotts, Jordt,
Mcconnell, & Tarran, 2019; Layden et al., 2019). Examining these new nicotine and products in
association with previously discussed traits/skills may provide insight into theoretical, clinical,
and policy targets for personalized prevention of these highly addictive products and reduce the
later public health burden of addiction.

Specific Aims
To address these gaps in literature, I leverage data from a longitudinal prospective cohort study
called the Happiness and Health Study (H&H). Participants are ages 14-19 with diverse
backgrounds. H&H measures past 30-day frequency of alcohol, cigarette, e-cigarette, cannabis,
and opioids across a total of 9 timepoints. This dissertation investigates high levels of distress
tolerance and mindfulness disposition as protective factors and low levels as consequences of
substance use over a 4.5-year follow-up.
Aim 1. To examine the association of distress tolerance (Distress Tolerance Scale [DTS]) at the
first semester of 9th grade with subsequent substance use (past 30 days) across 42 months-follow-
ups for alcohol, cigarettes, e-cigarettes, cannabis, and opioids (in separate models).
1a. To test baseline substance use status (users vs. non-user) as a moderator of the
association between DTS score and subsequent frequency of substance use.
1b. To test baseline mental health status (clinical or subclinical vs. non-clinical) as a
moderator of the association between DTS score and subsequent frequency of substance
use.

9
Aim 2. To characterize developmental growth trajectories of DTS across high school and
examine associations between growth trajectories of substance use across high school with DTS
trajectory membership.
2a. To examine trajectories of DTS from growth mixture modeling across high school,
average age range 14-18 years.
2b. To examine intercept and slope of alcohol, cigarette, e-cigarette, and cannabis
frequency of use (in separate models) in association with DTS group belonging.
Aim 3. To examine the association of mindfulness disposition (Five-Facet Mindfulness
Questionnaire-Short Form [FFMQ-SF]) with frequency of substance use after high school.
3a. To examine the association between FFMQ-SF sum and subscale scores with
substance use frequency (past 30 days) for alcohol, cigarette, e-cigarette, and cannabis
use (in separate models).
3b. To examine interactions between FFMQ-SF attention monitoring subscales
(observing, acting with awareness) with FFMQ-SF acceptance subscales (non-reactivity,
non-judging, describing) on frequency of alcohol, cigarette, e-cigarette, and cannabis use
(in separate models).

10
Chapter 2. Distress Tolerance and Subsequent Substance Use
Throughout High School

Abstract
Background: Distress tolerance (DT), ones ability to endure aversive experiences, has been
implicated in substance use problems and treatment outcomes. However, empirical data on DT in
relation to progression across the early stages of the substance use is lacking.

Methods: Data were from high school students (N=3,203) surveyed twice per year from 2013-
2017 (8 waves). A series of generalized estimating equation (GEE) models assessed prospective
associations of distress tolerance (Distress Tolerance Scale [DTS], range 1=strongly agree to
5=strongly disagree, where higher values reflect greater ability to withstand distress) at baseline
and use frequency (number of days used in past 30 days) of alcohol, cigarettes, e-cigarettes,
cannabis, and opioids across all available follow-up timepoints. Baseline use status (lifetime
never-user vs. user) and mental health status (clinical or sub-clinical vs. non-clinical) were tested
as moderators. The negative binomial regression coefficients from the GEE models were
exponentiated and are presented as Incident Rate Ratios (IRR).

Results: Baseline use status (lifetime never-user vs. user) was a significant moderator (p=.03)
and mental health status (clinical or sub-clinical vs. non-clinical) was not a significant moderator
(p>.05) of the association between baseline mean DTS score and past-30 day substance use
composite score averaged across available follow-up timepoints. Among baseline never users of
each respective substance, a greater baseline mean DTS score was associated with fewer days of
alcohol, cigarette, e-cigarette, and opioid use in past-30 days averaged across follow-up

11
timepoints (IRR range=0.60-0.84). For example, for each unit increase in baseline mean DTS
participants reported 16% fewer alcohol using days (IRR [95% CI]= 0.84 [0.74, 0.97]) averaged
across the 7 follow-ups among baseline never-users. In models among baseline ever-users with
the same covariates tested among baseline never-users, all associations between baseline mean
DTS and substance use across follow-up timepoints were not significant.

Conclusions: This study suggests greater perceived distress tolerance at the start of high school
may be protective against substance use frequency among adolescents who have never used
substances (but not among those who have ever used substances). Future research is needed to
replicate findings in the current study and extend by examining distress tolerance trajectories
before and across substance use stages.

12
A.1 Introduction

Distress tolerance, ones ability to endure aversive experiences and a core component of emotion
regulation, has been implicated in substance use problems and treatment outcomes (Leyro et al.,
2010). However, empirical data on distress tolerance in relation to progression across the early
stages of the substance use trajectory is lacking (Leventhal & Zvolensky, 2015). Distress
tolerance may be one mechanism by which some youth are at greater risk for both initiating
and/or maintaining substance use compared to their peers and holds important implications for
screening and interventions. The current study aims to address the association between perceived
distress tolerance and substance use frequency in adolescence.
Examining use frequency by specific substances may offer additional insight in the
relation between distress tolerance and substance use. Substances are categorized into
pharmacological profiles or classes (e.g., depressant, stimulant, opiate, hallucinogen) based on
the neurotransmitter systems (e.g., GABA, dopamine) they target (Tomkins & Sellers, 2001).
These different psychoactive properties lead to various effects on mood, emotion, and behavior,
leading to individual preference or outcome expectancies for specific substances following use
(Winters & Arria, 2011). For example, the primary neurotransmitter that alcohol targets is
glutamate while the primary neurotransmitter that nicotine targets is acetylcholine, which
contributes to depressant or anxiolytic effects of alcohol and stimulant effects of nicotine on
cognitive processes (Banerjee, 2014). Based on negative reinforcement and coping theories, this
becomes a key link in substance use disorders specifically for alleviation of aversive stimuli and
stress. However, using alcohol and drugs to achieve such desired effects of specific substances
follows initiation of that substance.

13
The motivation to use alcohol and other drugs is different among youth who have never
used and youth who use for the first time to experiment versus youth who have used and
continue to use in order to achieve a desired effect. Previous research has shown that distress
tolerance was lower in youth who used cigarettes (vs. e-cigarettes and no use) and lower in youth
who used e-cigarettes (vs. no use; Leventhal et al., 2016). Therefore, it is likely that those who
have previously used a substance (vs. never-used) and report lower (vs. higher) perceived
distress tolerance may have already progressed to a later stage of substance use or addiction
processes and it is critical to examine such differences separately by prior-use status.
Furthermore, while previous research suggests use status is associated with lower distress
tolerance, it is unclear whether lower distress tolerance also corresponds with greater frequency
of use during early stages. A recent review of the literature found evidence suggestive that while
cigarette smokers had lower distress tolerance compared to non-smokers, there was no indication
of a dose -dependent relation such that greater frequency nor dependence were consistently
associated with lower distress tolerance in adults, regardless of self-report or behavioral
measurement (Veilleux, 2019).
Distress tolerance is highly associated with internalizing symptoms (i.e., anxiety and
mood disorders) such that those with lower distress tolerance are more likely to also have higher
internalizing symptoms (Cummings et al., 2013; Daughters et al., 2009; Felton et al., 2017;
Felton et al., 2018; Juarascio et al., 2016). Therefore, the degree to which ones distress tolerance
predicts substance use frequency may differ by expression of anxiety and mood disorders (i.e.,
generalized anxiety disorder, obsessive compulsive disorder, panic disorder, social phobia, and
major depressive disorder). It is likely that those with low distress tolerance use substances more
frequently when they also meet clinical threshold for anxiety and mood disorders.

14
In the current study, we examined whether the previously documented null-association
between lower distress tolerance and greater smoking frequency (Veilleux, 2019) varies based on
substance (nicotine vs. alcohol, cannabis, opioids), product administration (smoking vs. vaping),
prior use status (lifetime user vs. never-user), and clinical (vs. non-clinical or sub-clinical)
expression of anxiety and mood disorders in order to elucidate the magnitude to which distress
tolerance prospectively predicts past 30-day substance use among a large, diverse sample of high
school adolescents. We used the Distress Tolerance Scale (DTS; Simons & Gaher, 2005), a 15-
item self-report survey, which has the potential for easy screening efforts. We hypothesized that
DTS at the beginning of high school would inversely associate with all subsequent substance use
across high school, and the magnitude of this association would be greater for lifetime users (vs.
lifetime never-users) for each product and participants meeting clinical (vs. non-clinical or sub-
clinical) cut points for anxiety and mood disorders. Findings offer important temporal insights
for distress tolerance, substance use, and mental health.

15
A.2 Methods
Participants
Data were collected via pencil-and-paper surveys from students at 10 high schools in the Los
Angeles Metropolitan area, who participated in the Happiness & Health (H&H) prospective
cohort study (R01-DA033296). To enroll, students were required to provide written or verbal
assent and their parents were required to provide written or verbal consent (N=3,396).
Assessment of distress tolerance was collected in the supplemental (vs. core) section of the
survey at baseline in Fall 2013. Assessment of substance use was collected at all eight
timepoints, or every 6-months across four years of high school. See Figure 1 for study accrual.
The analytic sample was generated from all participants with complete mean score of baseline
distress tolerance scale (DTS) as well as at least one follow-up substance use outcome (timepoint
2-8). One-hundred and ninety-two participants had missing data for baseline DTS and two
participants that had complete DTS data were missing at least one substance use outcome leaving
the analytic sample of N=3,202. See Table A1 for baseline sociodemographic characteristics by
baseline ever vs. never users of any substance. The University of Southern California
institutional review board approved the study.

Measures
Distress Tolerance Scale (DTS; Simons & Gaher, 2005) is a validated, 15-item self-report survey
with 4 subscales: absorption (3-items), appraisal (6-items), regulation (3-items), and tolerance (3-
items) that measures distress tolerance. Items are rated on a 5-point Likert scale (1 = Strongly
Agree to 5 = Strongly Disagree), where a higher composite score reflects a higher level of
distress tolerance. The four subscale scores were calculated by taking the mean of the
corresponding items. Example items include, When I feel distressed or upset, I cannot help b ut

16
concentrate on how bad the distress actually feels (absorption= feeling much attention is
absorbed by the negative emotion and how much it interferes with functioning), My feelings of
distress or being upset are not acceptable (appraisal=experiencing emotional distress as
unacceptable), Ill do anything to stop feeling distressed or upset (regulation=engaging in
behaviors to immediately terminate distress), and Theres nothing worse than feeling distressed
or upset (tolerance=low perceived ability to tolerate affective distress). See Appendix 1 for full
scale and subscale breakdown.

Substance Use: At each timepoint, participants reported lifetime ever use (no / yes) and past 30-
day frequency (0 days, 1 to 2 days, 3 to 5 days, 6 to 9 days, 10 to 14 days, 15 to 19 days, 20 to 24
days, 25 to 29 days, all 30 days) of alcohol, cigarettes, e-cigarettes, cannabis, and opioids.
Specific substances were collected at varying timepoints due to the timepoint they were added to
the survey. Reports on past 30-day use of alcohol, cigarettes, cannabis, and opioids were
collected at all 8 timepoints while past 30-day e-cigarette use was added at time 3 and collected
the remaining timepoints. Repeated substance use data was restructured into longitudinal cases
(using all available within-subject repeated measures) for past 30-day frequency for each
substance and used as the outcome variables. Lifetime ever vs. never use was tested as a
moderator and used as the model stratification variable for each product.

The Revised Child Anxiety and Depression Scale (RCADS) (Chorpita, Yim, Moffitt, Umemoto,
& Francis, 2000) is a measure of DSM-IV anxiety and depression disorders and includes
subscales for Generalized Anxiety Disorder (GAD) with 6-items, Major Depressive Disorder
(MDD) with 10-items, Panic Disorder (PD) with 9-items, Social Phobia (SP) with 9-items, and

17
Obsessive Compulsive Disorder (OCD) with 6-items. Items are rated on a 4-point Likert scale
(never, sometimes, often, and always) where a higher score indicates higher levels of anxiety and
depressive symptoms. Example items include, I feel sad or empty (MDD scale) and When I
have a problem, I get a funny feeling in my stomach (PD scale). To examine whether clinical
manifestations of anxiety and depression disorders moderates the association between distress
tolerance and substance use, a dichotomous variable was created to indicate whether a participant
reported symptomology at the subclinical or clinical cut points (vs. nonclinical; 1 vs. 0) for one
or more of the five disorders.

Covariates: Measures of age, gender (Female, Male), race/ethnicity (Hispanic, White, Asian,
Black, Multiracial, Other), highest parental education (8th or less, Some high school, High
school graduate, Some college, College graduate, Advanced degree, or Don't know), and school
(district 1-3) were collected at baseline using the self-report response to investigator-defined
forced-choice items. To adjust for independent measurement violations from participant repeated
measures, subject and within-subject variables included student identification (SID) and
timepoint (1-8), respectively.

Statistical Analyses
Descriptive Analyses
Prevalence of baseline ever use of alcohol, cigarettes, e-cigarettes, cannabis, and opioids was
examined. Then participant characteristics by ever use of any substance (for table simplification)
was examined.

18
Primary Analyses: Test Association Between Baseline DTS and Substance Use Frequency
Across All Available Timepoints And Model Specifications
A series of generalized estimating equation (GEE) regression models were conducted to
assess the prospective associations of baseline distress tolerance (DTS in 9th grade) and past 30-
day use of specific substances (alcohol, cigarettes, e-cigarettes, cannabis, and opioids) across all
available timepoints in separate models. All substance use outcomes were 0 to 30 (past 30 days)
count variables and showed overdispersion so negative binomial model distribution with a log
link function were specified. Due to the within and across subject correlation of substance use
variables, the exchangeable correlation structure was specified (Ballinger, 2004).
All models were adjusted for mental health status (RCADS, 0/1), sociodemographics
(gender, race/ethnicity, parental education), and school district (1-3). Student ID was specified in
the model as repeated subject effects and time was specified in the model as within-subject
effects. Missing baseline sociodemographic covariates were handled with multiple imputation
(Rubin, 1987) to avoid listwise deletion of missing data. The negative binomial regression
coefficients from the GEE models were exponentiated and are presented as Incident Rate Ratios
(IRR). Significance was set at .05. All analyses were conducted in SPSS 25.

Secondary Analyses: Test Effects of Moderating Variables
Baseline DTS Lifetime use status for each product at baseline (0/1) interaction terms
were added in subsequent models to test for moderating effect on substance use frequency. All
models were stratified by lifetime use status and separately for each product (operationalized as
lifetime never-user [0] vs. ever-user [1] for alcohol, cigarette, e-cigarette, cannabis, and opioid)
at baseline. In other words, for each product two nearly identical models were tested. The first

19
was among the selected sample of those who reported never having used the respective substance
(0) and the second among those who reported having used the respective substance (1). All
primary analysis model specifications in the previous sub-section were applied in all 10 models
with the addition that in baseline ever-user models past 30-day use at timepoint 1 of respective
substance was included.
Baseline DTS mental health status at baseline (RCADS [0/1] operationalized as
meeting subclinical or clinical level of symptomology for one or more of the five disorders:
MDD, GAD, PD, SP, OCD) was added in subsequent models to test for moderating effect on
substance use frequency.

Supplemental Analyses: Composite Substance Use and DTS Subscale Models
The following supplemental analyses were included. First, a negative binomial random
effect repeated measure GEE regression model (same specifications as primary analyses)
including a composite substance use outcome (i.e., past 30-day alcohol, cigarette, e-cigarette,
cannabis, and opioid use) across timepoints 4-8 (since e-cigarette use was added at timepoint 3
and available timepoints need to match for variable to case transformation) estimated the
averaged association between baseline mean DTS and past 30-day substance use across five
follow-up waves. Second, the four mean DTS subscales (DTS absorption mean, DTS appraisal
mean, DTS regulation mean, and DTS tolerance mean) were entered separately in subsequent
models as predictors of the composite substance use outcome model.

20
A.3 Results

Descriptive Data
Baseline prevalence ever versus never use of each substance was examined and showed the most
ever use of alcohol (27.9%) and least for opioids (3.8%). See Table A1 for baseline
sociodemographic characteristics by baseline ever vs. never users of any substance. Baseline
never users (M[SD]=3.51 [0.83]) reported significantly greater levels of DTS at baseline
compared to baseline ever users (M[SD]=3.32 [0.87]) of any substance (p<.001). The interaction
between DTS lifetime use of any substance (in models with all substance use frequency
outcome) was significant (p=.029; Table A3) so models were stratified by ever/never use for
each substance.

Baseline DTS and Substance Use Frequency Across High School by Ever Use Status
Baseline never users. Among baseline never users of each respective substance, a greater
baseline mean DTS score was associated with fewer days of alcohol, cigarette, e-cigarette, and
opioid use in past-30 days averaged across follow-up timepoints (Table A2 and Figure A2). For
each unit increase in baseline mean DTS participants reported 16% fewer alcohol using days
(IRR [95% CI]= 0.84 [0.74, 0.97]) averaged across the 7 follow-ups among baseline never-
users. For each unit increase in baseline mean DTS participants reported 40% fewer cigarette
using days (IRR [95% CI]= 0.60 [0.46, 0.79]) averaged across the 7 follow-ups among baseline
never-users. For each unit increase in baseline mean DTS participants reported 26% fewer e-
cigarette using days (IRR [95% CI]= 0.74 [0.57, 0.96]) averaged across the 5 follow-ups among
baseline never-users. For each unit increase in baseline mean DTS participants reported 33%
fewer opioid using days (IRR [95% CI]= 0.67 [0.50, 0.88]) averaged across the 7 follow-ups

21
among baseline never-users. Mean DTS did not significantly predict past 30-day use of cannabis
averaged across the 7 follow-ups among baseline never-users.
Baseline ever-users. In models among baseline ever-users with the same covariates tested
among baseline never-users, all associations between baseline mean DTS and substance use
across follow-up timepoints were not significant (Table A2 and Figure A2). Supplemental Tables
A2 and A3 show covariate estimates of models in Table A2/Figure A2.

Mental Health Status Tested as Moderator Models
Additional models including the mental health status (using 0/1 RCADS) DTS interaction
terms found all terms were not significant (all ps>.05). Since there was no evidence of mental
health status moderating the association of mean DTS and past 30-day substance use frequency
across follow-up timepoints no models were not stratified by mental health status.

Composite Substance Use and DTS Subscale Models
Overall, for each unit increase of mean DTS, baseline never-users reported 27% fewer days (or
lower rate) of past-30 day substance use (alcohol, cigarettes, e-cigarettes, cannabis, and opioids)
across follow-up timepoints, in fully adjusted models (IRR [95% CI]= 0.73 [0.59, 0.91];
Supplemental Table A1). However, in parallel models for baseline ever-users, the association
was not significant (IRR [95% CI]= 0.97 [0.84, 1.12]). Furthermore, mean DTS subscales were
differentially associated with lower rates of composite substance use across timepoints 4 through
8 among baseline never-users, but not baseline ever-users. A greater baseline mean DTS
Absorption and Appraisal scores (but not DTS Regulation and Tolerance) was associated with
significantly fewer days (lower rate) of past 30-day composite substance use (alcohol, cigarettes,

22
e-cigarettes, cannabis, and opioids) across follow-up timepoints, in fully adjusted models
(Absorption: IRR [95% CI]=0.78 [0.68, 0.89)]; Appraisal: IRR [95% CI]=0.74 [0.59, 0.94)];
Supplemental Table A1).

23
A.4 Discussion

This is the first study to comprehensively examine adolescent self-reported distress tolerance as a
predictor of substance use frequency across high school. The current study found that (1) distress
tolerance was inversely associated with substance use frequency across high school among
baseline never users, but not among baseline ever users of any substance and that (2) the
association between distress tolerance and substance use frequency was not moderated by the
manifestation of clinical/subclinical (vs. nonclinical) anxiety and depression symptomology.
While previous research has established lower distress tolerance as a risk factor in substance
maintenance and cessation stages of use (Leventhal & Zvolensky, 2015; Veilleux, 2019), this is
the first study to establish higher distress tolerance as a protective factor against substance use
prior to initial use.
Lifetime use status moderated the association between distress tolerance and substance
use frequency such that greater distress tolerance was associated with lower substance use
frequency across high school among baseline never users (and was not significantly associated
among ever users). Notably, skills captured in the DTS absorption and DTS appraisal subscales
seemed to drive this effect among the never-users. This finding of which specific DTS subscales
show the strongest magnitude of association with health behavior not only corresponds to
previous work in this sample (Kechter & Leventhal, 2018), but also extends to other samples
where appraisal predicted substance use above and beyond overall distress tolerance and the
three other DTS facets (Greenberg, Martindale, Fils-Aimé, & Dolan, 2016). Our findings
contribute evidence for the theoretical conceptualization that distress tolerance is comprised of
many facets rather than a hierarchical construct with many related constructs of coping with
stressors (Evanovich, Marshall, David, & Mumma, 2019). Altogether, findings suggest that a

24
worthwhile substance use prevention effort may be to target the improvement of DTS absorption
and DTS appraisal skills in youth prior to substance use initiation.
Lower distress tolerance did not associate with greater substance use frequency
regardless of substance or device. This finding aligns with previous literature among cigarette
smokers (Veilleux, 2019). Given that distress tolerance did not significantly associate with
substance use frequency among those who had used before the hypothesis that the magnitudes of
association for this group would differ for alcohol, cigarettes, e-cigarettes, cannabis, and opioids
due to the pharmacological targets of the drug was also not found. Previous work on adolescents
has reported that lower distress tolerance (as measured by behavioral vs. self-report indicators)
was associated with greater risk taking behaviors and internalizing symptomology (Daughters et
al., 2009; Felton et al., 2017; Felton et al., 2018; Juarascio et al., 2016). Incidentally, the ever-
users showed significantly lower mean baseline distress tolerance scores and a greater proportion
of poor mental health status compared to never-users. Thus, one plausible explanation may be
that participants with lower distress tolerance were inherently more vulnerable than participants
with greater distress tolerance to initiate early substance use and experience mental health
symptomology (i.e., ever users in our sample).
Clinical and subclinical (vs. nonclinical) anxiety and depression disorder symptomology
did not moderate the association between DTS and subsequent substance use. This finding holds
important clinical relevance as it suggests DTS seems to be a distinguishable marker from
internalizing symptomology. In other words, based on findings in the current study that mental
health status does not moderate the association between distress tolerance and substance use
frequency, it can be inferred that adolescents do not need different distress tolerance enhancing
interventions based on their mental health. For example, middle school substance use prevention

25
programs could aim to train distress tolerance skills in early adolescence – ideally before
substance use experimentation or initiation stages so that adolescents are all never users.
The main limitation of the current study is that observational, survey-based data do not
allow inference of causality due to correlational nature. Additionally, the current study could not
ascertain whether lower distress tolerance among ever users was due to innately poor coping
skills or whether they had progressed to later stages of substance use. Future research is needed
to replicate findings in the current study and extend by examining distress tolerance trajectories
before and across substance use stages. Despite these limitations, findings in this study suggests
greater perceived distress tolerance, specifically absorption and appraisal facets, at the start of
high school may be protective against greater risk of substance use frequency among adolescents
who have never used substances (vs. those who have) and thus DTS may be a worthwhile
screening target. Prevention programs targeting acceptance of emotions and coping techniques
(Carpenter, Sanford, & Hofmann, 2019; Greenberg et al., 2016) may be most beneficial in
training youth to effectively navigate stressors without the use of substances.

26
A.5 Figures and Tables

Figure A1. Analytic Sample Based on Study Accrual and Baseline Distress Tolerance Scale and
Substance Use Respondents

Note. Analytic sample consists of 3,203 high school students surveyed twice per year from 2013-2017 (8
time points) with complete data for DTS use at baseline and at least one substance outcome at any follow-
up timepoint (timepoints 2-8).

27
Figure A2. Baseline Distress Tolerance Predicting Past 30-Day Use of Substances Across High School

Note. Graphical representation of estimates in Table A2 - fully adjusted models stratified by use status. *Significant group differences of p<.05

28
Table A1. Baseline Participant Descriptive Characteristics (N=3,203)

Study Variable Never
Substance User
(N=1,893)
Ever
Substance User
(N=1,310)
N(%) or M(SD)
Gender
Male 890 (47.0%) 589 (45.0%)
Female 1003 (53.0%) 721 (55.0%)
Age* 14.54 (0.39) 14.62 (0.42)
Highest Parental Education Level*
<8th grade 50 (2.6%) 56 (4.3%)
Some high school 119 (6.3%) 132 (10.1%)
High school graduate 223 (11.8%) 246 (18.8%)
Some college 313 (16.5%) 224 (17.1%)
College graduate 573 (30.3%) 312 (23.8%)
Advanced degree 359 (19.0%) 173 (13.2%)
Dont know 256 (13.5%) 167 (12.7%)
Race/Ethnicity*
Hispanic 792 (41.8%) 737 (56.3%)
Asian 415 (21.9%) 136 (10.4%)
Black 91 (4.8%) 60 (4.6%)
Multiracial 126 (6.7%) 82 (6.3%)
Other 129 (6.8%) 94 (7.2%)
White 340 (18.0%) 201 (15.3%)
Distress Tolerance Scale* 3.51 (0.83) 3.32 (0.87)
Mental Health Status*
Clinical or Subclinical 250 (13.2%) 234 (17.9%)
Nonclinical 1643 (86.8%) 1072 (81.8%)

Note. Mean Distress Tolerance Scale (DTS) baseline scores ranged from 1.00-5.00.
Mental Health Status = Revised Child Anxiety and Depression Scale (RCADS), coded as 0=without
clinical anxiety and mood disorder symptomology and 1=with subclinical or clinical anxiety and mood
disorder symptomology.
This table is stratified by ever vs. never use of any substance for simplified table visual. However, each
model was stratified by each respective substance.
*P-value <.01 group difference from t-tests and ANOVA tests for equality of means.

29
Table A2. Baseline Distress Tolerance Predicting Past 30-day Use of Substances Across High School

Incidence Rate Ratio (95% CI)
Alcohol Cigarette E-Cigarette Cannabis Opioid
Baseline Never Users
DTS 0.84 (0.74, 0.97)** 0.60 (0.46, 0.79)*** 0.74 (0.57, 0.96)* 0.84 (0.69, 1.04) 0.67 (0.50, 0.88)**
Baseline Ever Users
DTS 0.98 (0.88, 1.11) 1.47 (0.98, 2.21) 0.77 (0.57, 1.04) 0.95 (0.80, 1.14) 0.96 (0.60, 1.54)

Note. DTS= Distress Tolerance Scale, range 1=strongly agree to 5=strongly disagree, where higher values reflect greater ability to withstand distress. All
models control for gender, race/ethnicity, parental education, and school district. Ever-user models were additionally controlled for self-reported
past 30-day use first available timepoint of corresponding substance. Student ID was specified as repeated subject effects and Time was specified
as within-subject effect. Alcohol, cigarettes, cannabis, and opioid use data are from timepoints 2-8 and stratified by timepoint 1 use status. E-
cigarette data are from timepoints 4-8 and stratified by timepoint 1 use status. Substance use count variables from 0-30. Corresponds to Figure A2.
*P-value < .05 **P-value < .01 ***P-value < .001

30
Table A3. Moderators of Baseline Distress Tolerance Predicting Past 30-day Use of Substances Across High School

Incidence Rate Ratio (95% CI)
Composite Use Alcohol Cigarette E-Cigarette Cannabis Opioid
Lifetime Use Status as Moderator
DTS 0.74 (0.60, 0.90)** 0.85 (0.74, 0.97)* 0.62 (0.45, 0.86)** 0.75 (0.54, 1.05) 0.92 (0.85, 1.00)* 0.91 (0.86,0.97)**
Use Status 1.53 (0.67, 3.49) 1.82 (0.97, 3.43) 0.49 (0.74, 3.16) 2.66 (0.59, 11.96) 1.60 (0.87, 2.95) 0.94 (0.39, 2.31)
DTS Use Status 1.30 (1.03, 1.64)* 1.17 (0.97, 1.41) 2.25 (1.28, 3.97)** 1.15 (0.72, 1.82) 1.10 (0.92, 1.32) 1.27 (0.96, 1.68)
Mental Health Status as Moderator
DTS 0.88 (0.76, 1.01) 0.92 (0.83, 1.03) 0.72 (0.51, 1.01) 0.82 (0.60, 1.13) 0.95 (0.87, 1.03) 0.90 (0.58, 1.41)
Mental Health Status 1.96 (0.71, 5.40) 1.87 (0.79, 4.40) 1.08 (0.13, 9.07) 2.87 (0.45, 18.19) 1.07 (0.58, 1.99) 2.27 (0.16, 32.98)
DTS Mental Health Status 0.91 (0.64, 1.30) 0.85 (0.62, 1.15) 1.21 (0.57, 2.56) 0.80 (0.41, 1.55) 1.04 (0.83, 1.29) 1.09 (0.46, 2.58)

Note. DTS Use status, where 1 (vs. 0)=ever-use (vs. never-use) of respective substance.
DTS Mental Health status, where 1 (vs. 0)=clinical or subclinical (vs. none) anxiety and mood disorders.
Composite Use: Past 30-day use from timepoints 4-8 for alcohol, cigarettes, e-cigarettes, cannabis, and opioid were used to create a sum score for
substance use stratified by those who reported never vs. ever using any substance at baseline. Substance use variables are 0-30 count. In all models
testing mental health status as moderator, lifetime use for each substance was a significant covariate (ps<.001).
*P-value < .05 **P-value < .01 ***P-value < .001
Exact P-Values for Significant Interactions:
DTS Use Status Composite Use (p=.029)
DTS Use Status Cigarette (p=.005)

31
Supplemental Table A1. Baseline Distress Tolerance Mean and Subscales Predicting Past 30-Day Substance Use by Baseline Never
vs. Ever User

Incidence Rate Ratio (95% CI)
Baseline Never User Baseline Ever User
DTS Overall Mean 0.73 (0.59, 0.91)** 0.97 (0.84, 1.12)
DTS Absorption 0.78 (0.68, 0.89)*** 0.95 (0.87, 1.04)
DTS Appraisal 0.74 (0.59, 0.94)** 0.95 (0.82, 1.09)
DTS Regulation 0.90 (0.80, 1.01) 0.99 (0.89, 1.10)
DTS Tolerance 0.93 (0.80, 1.08) 1.05 (0.95, 1.16)

Note. Past 30-day use from timepoints 4-8 for alcohol, cigarettes, e-cigarettes, cannabis, and opioid were used to create a sum score for substance
use stratified by those who reported never vs. ever using any substance at baseline. Substance use variables are 0-30 count.
Stratification Variable: Baseline Never User (N=1893) Ever User (N=1310).
*P-value < .05 **P-value < .01 ***P-value < .001

32
Supplemental Table A2. Baseline Distress Tolerance Predicting Past 30-Day Use of Substances across High School among Baseline
Never-Users

Incidence Rate Ratio (95% CI)
Alcohol Cigarette E-Cigarette Cannabis Opioid
DTS 0.84 (0.74, 0.97)** 0.60 (0.46, 0.79)*** 0.74 (0.57, 0.96)* 0.84 (0.69, 1.04) 0.67 (0.50, 0.88)**
RCADS (1 vs. 0) 1.20 (0.87, 1.67) 1.54 (0.74, 3.19) 1.59 (0.99, 2.55) 1.52 (0.98, 2.37)* 1.80 (1.00, 3.26)*
Gender (M v. F) 1.02 (0.80, 1.31) 3.09 (1.74, 5.49)*** 3.07 (2.15, 4.38)*** 1.40 (1.04, 1.87)* 1.27 (0.74, 2.18)
Parent Education Level (v. College graduate reference group)
Advanced degree 1.01 (0.71, 1.42) 1.04 (0.39, 2.83) 0.65 (0.36, 1.20) 0.70 (0.45, 1.08) 0.77 (0.36, 1.64)
Some college 1.00 (0.73, 1.38) 1.15 (0.41, 3.23) 0.70 (0.39, 1.24) 1.01 (0.66, 1.55) 0.42 (0.18, 0.98)*
High school grad 1.32 (0.96, 1.81) 0.71 (0.30, 1.67) 1.44 (0.71, 2.90) 1.11 (0.70, 1.77) 0.91 (0.41, 1.99)
Some high school 1.74 1.14, 2.67)* 2.36 (0.85, 6.54) 1.13 (0.50, 2.53) 1.09 (0.59, 2.00) 1.41 (0.52, 3.77)
< 8th 0.65 (0.36, 1.20) 0.68 (0.18, 2.52) 1.00 (0.23, 4.25) 1.06 (0.39, 2.91) 0.88 (0.23, 3.42)
Dont know 1.12 (0.68, 1.84) 1.39 (0.56, 3.46) 1.20 (0.64, 2.24) 0.80 (0.47, 1.38) 0.85 (0.31, 2.36)
Race/Ethnicity (v. Hispanic reference group)
Asian 0.62 (0.40, 0.94)* 1.76 (0.58, 5.33) 1.42 (0.75, 2.68) 0.60 (0.37, 0.98)* 0.54 (0.22, 1.30)
Black 1.01 (0.50, 2.05) 1.38 (0.41, 4.60) 1.13 (0.34, 3.78) 0.84 (0.34, 2.06) 0.70 (0.23, 2.14)
Multiracial 2.02 (0.93, 4.37) 7.79 (1.97, 30.75)** 3.74 (1.02, 13.77)* 1.10 (0.35, 3.41) 1.54 (0.36, 6.56)
Other 1.07 (0.76, 1.52) 1.35 (0.60, 3.04) 1.63 (0.93, 2.86) 0.57 (0.37, 0.88)* 0.62 (0.23, 1.67)
White 1.01 (0.78, 1.31) 2.05 (0.94, 4.45) 2.62 (1.27, 4.03)* 1.16 (0.80, 1.68) 0.89 (0.46, 1.71)
School District (v. District 1 reference group)
2 1.05 (0.74, 1.48) 0.62 (0.28, 1.41) 0.91 (0.56, 1.47) 0.96 (0.63, 1.46) 0.73 (0.37, 1.43)
3 1.36 (0.98, 1.88) 0.89 (0.28, 2.76) 1.21 (0.78, 1.86) 1.48 (0.97, 2.26) 0.91 (0.51, 1.62)

Note. All models control for gender, race/ethnicity, parental education level, and school district. Student ID modeled as repeated subject effects;
Time modeled as within-subject effect. Alcohol, cigarettes, cannabis, and opioid use data are from timepoints 2-8 and stratified by timepoint 1 use
status. E-cigarette data are from timepoints 4-8 and stratified by timepoint 1 use status. Substance use variables coded as past 30-day 0-8 count
scale. Stratification Variables: W1 Never Substance Use (N=1893) Ever User (N=1310); W1 Alcohol Never User (N=1869) Ever User (N=584);
W1 Cigarette Never User (N=2261) Ever User (N=200); W1 E-Cigarette Never User (N=2325) Ever User (N=350); W1 Cannabis Never User
(N=2177) Ever User (N=285); W1 Opioid Never User (N=2383) Ever User (N=76).
Parent Education Level: 0=College graduate;1=Advanced degree; 2=Some college; 3=High school graduate; 4=Some high school; 5=Less than
8th grade; 6=Dont know
Race/Ethnicity: 0=Hispanic; 1=Asian; 2=Black; 3=Multiracial; 4=Other; 5=White
*P-value < .05 **P-value < .01 ***P-value < .001

33
Supplemental Table A3. Baseline Distress Tolerance Predicting Past 30-Day Use of Substances across High School among Baseline
Ever-Users

Incidence Rate Ratio (95% CI)
Alcohol Cigarette E-Cigarette Cannabis Opioid
DTS 0.98 (0.88, 1.11) 1.47 (0.98, 2.21) 0.77 (0.57, 1.04) 0.95 (0.80, 1.14) 0.96 (0.60, 1.54)
Baseline Past 30 Use 1.38 (1.28, 1.48)*** 1.17 (1.02, 1.35)* 1.52 (1.37, 1.69)*** 1.27 (1.21, 1.34)*** 1.17 (1.03, 1.34)*
RCADS (1 v. 0) 1.05 (0.83, 1.32) 3.39 (1.34, 8.58)* 1.61 (0.89, 2.93) 0.83 (0.55, 1.26) 2.16 (1.12, 4.18)*
Gender (M v. F) 1.11 (0.90, 1.36) 1.46 (0.82, 2.59) 3.56 (2.08, 6.07)*** 1.37 (0.96, 1.95)*** 1.63 (0.64, 4.13)
Parent Education Level v. College graduate reference group)
Advanced degree 1.09 (0.80, 1.48) 3.43 (1.17, 10.05)* 0.67 (0.28, 1.60) 1.49 (0.78, 2.84) 1.46 (0.29, 7.46)
Some college 1.06 (0.78, 1.44) 3.19 (0.89, 11.48) 0.53 (0.25, 1.14) 1.48 (0.77, 2.84) 1.11 (0.25, 4.94)
High school grad 0.93 (0.70, 1.24) 1.21 (0.36, 4.01) 1.11 (0.49, 2.50) 1.23 (0.67, 2.25) 3.84 (1.34, 11.02)*
Some high school 0.83 (0.59, 1.16) 1.72 (0.45, 6.62) 1.32 (0.46, 3.79) 2.02 (1.01, 4.03)* 1.22 (0.35, 4.27)
< 8th 0.97 (0.62, 1.51) 2.22 (0.53, 9.28) 0.97 (0.41, 2.28) 2.11 (0.92, 4.84) 1.42 (0.27, 7.35)
Dont know 0.84 (0.59, 1.20) 0.71 (0.24, 2.09) 0.60 (0.26, 1.38) 1.67 (0.84, 3.30) 0.96 (0.28, 3.32)
Race/Ethnicity (v. Hispanic reference group)
Asian 0.71 (0.49, 1.04) 1.23 (0.23, 6.65) 2.22 (0.91, 5.38) 1.20 (0.76, 1.90) 3.93 (1.20, 12.91)*
Black 0.84 (0.48, 1.49) 0.54 (0.16, 1.91) 0.55 (0.19, 1.56) 0.92 (0.52, 1.63) 3.41 (0.65, 17.92)
Multi 0.90 (0.52, 1.55) 0.51 (0.12, 2.28) 1.54 (0.26, 9.05) 1.77 (0.83, 3.79) 0.54 (0.27, 1.07)
Other 0.78 (0.57, 1.05) 1.60 (0.71, 3.63) 1.57 (0.77, 3.21) 1.62 (0.98, 2.69) 1.49 (0.31, 7.04)
White 1.36 (1.02, 1.81)* 1.39 (0.61, 3.15) 2.91 (1.35, 6.26)** 2.52 (1.61, 4.03)*** 3.76 (0.76, 18.54)
School District (v. District 1 reference group)
2 1.10 (0.87, 1.40) 0.98 (0.50, 1.94) 0.59 (0.35, 0.99)* 1.20 (0.78,1.84) 2.37 (0.68, 8.23)
3 1.22 (0.96, 1.55) 1.23 (0.57, 2.64) 0.80 (0.45, 1.42) 1.07 (0.71, 1.61) 0.79 (0.27, 2.30)
Note. All models control for gender, race/ethnicity, parental education level, and school district. Student ID modeled as repeated subject effects;
Time modeled as within-subject effect. Alcohol, cigarettes, cannabis, and opioid use data are from timepoints 2-8 and stratified by timepoint 1
use status. Stratification Variables: W1 Never Substance Use (N=1893) Ever User (N=1310); W1 Alcohol Never User (N=1869) Ever User
(N=584); W1 Cigarette Never User (N=2261) Ever User (N=200); W1 E-Cigarette Never User (N=2325) Ever User (N=350); W1 Cannabis Never
User (N=2177) Ever User (N=285); W1 Opioid Never User (N=2383) Ever User (N=76). E-cigarette data are from timepoints 4-8 and stratified by
timepoint 1 use status. Substance use coded as 0-30 count.
*P-value < .05 **P-value < .01 ***P-value < .001

34
Chapter 3. Longitudinal Associations Between Substance Use and
Distress Tolerance Trajectory Groups Throughout High School

Abstract
Background: Previous research among adolescents has not fully addressed the temporal nature
between distress tolerance and substance use and it is unclear whether perceived distress
tolerance is a risk factor or consequence of risky substance use.

Methods: The current study leveraged self-report data from a prospective cohort study that
measured perceived distress tolerance (Distress Tolerance Scale, DTS; Simons & Gaher, 2005)
and substance use frequency (past 30-day alcohol, cigarette, e-cigarette, and cannabis)
throughout high school in a sample of 14 to 18-year olds. A series of latent growth mixture
models were estimated to identify the best fitting number of heterogeneous growth trajectories
from the DTS. Latent growth curve modeling was then used to estimate intercept (baseline level)
and linear slope (rate of change across assessment periods) for each substance use frequency
(alcohol, cigarette, e-cigarette, and cannabis). Results of the association between changes in
substance use and DTS class are reported as odds ratios (OR) with 95% confidence intervals for
class comparisons.

Results: Finding suggest there are three qualitatively distinct trajectories of perceived distress
tolerance based on this diverse sample of high school adolescents and were named High
Increasing, Moderate Stable, and Poor Declining. Greater baseline levels of alcohol, e-
cigarettes, and cannabis use at the start of high school was associated with belonging to the
Poor Declining v ersus High Increasing distress tolerance trajectories (OR range=1.08-1.20).

35
Greater change in alcohol use at each timepoint was associated with belonging to the Poor
Declining versus High Increasing (OR [95% CI]=10.05 [2.42, 41.75]) and Moderate Stable
(OR [95% CI]=4.77 [1.14, 20.05]) distress tolerance trajectories.

Conclusions: While additional research is warranted for replication and temporal sequence,
these findings suggest greater use of commonly used substances at the start of high school could
potentially derail optimal development of distress tolerance.

36
B.1 Introduction

Limited literature has examined longitudinal changes in distress tolerance during adolescence.
While there is some evidence that distress tolerance may be stable when measured by Behavioral
Indicator of Resiliency to Distress across 4-years of follow-up among a sample of baseline 11-
year olds (Cummings et al., 2013); weve recently shown that during a 1 -year period in high
school, stability of perceived distress tolerance measured by the Distress Tolerance Scale (DTS)
was fairly low in this sample of baseline 15-year olds (range for DTS subscales=.39-.48; Kechter
& Leventhal, 2018). Previous developmental research suggests the greatest degree of change in
psychological traits occurs during later adolescence and early adulthood (Compas et al., 2014).
Therefore, the lack of concordance across studies may stem from variability in measurement
(i.e., behavioral vs. self-report) and age (i.e., early adolescence vs. middle adolescence). It is
unclear if and how perceived distress tolerance changes throughout high school. Given the stable
yet malleable trait-like conceptualization of distress tolerance during this critical developmental
period, it is likely that individual perceptions of distress tolerance during high school begin and
change at different levels and rates.
Previous research among adolescents has not fully addressed the temporal nature between
distress tolerance and substance use. One study examined how behavioral distress tolerance
during abstinence among heavy drinking adolescents compared to demographically matched
non-drinking adolescents (Winward, Bekman, Hanson, Lejuez, & Brown, 2014). As expected,
heavy drinking adolescents demonstrated poorer distress tolerance during early abstinence, but
group differences diminished with continued abstinence indicative that initial group differences
in distress tolerance may be an acute state from chemical withdrawal. Yet, in this same sample,
adolescents who were exposed to alcohol use at an earlier age and had greater lifetime

37
consumption showed worse distress tolerance and greater emotional reactivity during abstinence
(Winward et al., 2014) indicative that group differences in distress tolerance may be an
enduring trait, possibly from structural changes in the brain. Miettuen et al. (2014) found
externalizing symptoms often precede substance use while internalizing symptoms (which
distress tolerance seems to underlie) often follow initiation of substance use, especially among
females (Miettunen, Murray, Jones, Maki, Ebeling, Taanila, Joukamaa, et al., 2014). Combining
these studies with negative reinforcement theory, this may be suggestive that distress tolerance
development follows substance initiation, which perpetuates substance maintenance.
Collectively, it is unclear whether perceived distress tolerance is a risk factor or consequence of
risky substance use. Given the neuroadaptations from substance use on the adolescents
developing brain as well as immature prefrontal cortex during this time-period often linked with
low emotion regulation skills predisposing the adolescent to engage in riskier behaviors (Jordan
& Andersen, 2017), it is plausible that perceived distress tolerance may both increase risk of
substance use and be altered by substance use.
To address these gaps in literature and elucidate the temporal nature between distress
tolerance and substance use, the current study leveraged data from a prospective cohort study
that measured adolescent distress tolerance and substance use throughout high school. The first
objective in the current study was to evaluate time-dependent patterns of distress tolerance
development across high school. To examine the malleable nature of the perceived ability to
tolerate distress, we use repeated measures of the Distress Tolerance Scale (DTS; Simons &
Gaher, 2005) at four time points spanning four years. The second objective was to test the
hypothesis that participants with (1) greater past 30-day substance use frequency at baseline
and/or (2) greater average increase in 30-day substance use at each follow-up would predict

38
belonging to the poorest DTS group trajectory. Findings offer important temporal insights
between substance use trajectories and distress tolerance development.

39
B.2 Methods
Participants
Data were collected from high school students in a longitudinal survey-based study called
Happiness and Health (R01-DA033296). To participate, students from 10 high schools in Los
Angeles provided written or verbal assent and their parents provided written or verbal consent
(N=3,396). Assessment of distress tolerance was in the supplemental survey, which was
collected once a year in the Fall from 2013-2017 or four timepoints total (DTS at time 1
N=3,204; time 3 N= 2,946; time 5 N= 2,671; and time 7 N=2,490). Participants with complete
DTS data across timepoints were included in the current study analytic sample (Figure 1).
Assessment of substance use exposure variables was collected at timepoints 1-8. Paper-and-
pencil surveys were administered onsite in students classrooms. The authors university
institutional review board approved the study.

Measures
Distress Tolerance Scale (DTS; Simons & Gaher, 2005) is a validated, 15-item self-report survey
with 4 subscales: absorption (3-items), appraisal (6-items), regulation (3-items), and tolerance (3-
items) that measures distress tolerance ones ability to withstand negative psychological states.
Items are rated on a 5-point Likert scale (1 = Strongly Agree to 5 = Strongly Disagree), where a
higher composite score reflects a higher level of distress tolerance. The four subscale scores were
calculated by taking the mean of the corresponding items. Example items include When I feel
distressed or upset, I cannot help but concentrate on how bad the distress actually feels
(absorption=feeling much attention is absorbed by the negative emotion and how much it
interferes with functioning), My feelings of distress or being upset are not acceptable
(appraisal=experiencing emotional distress as unacceptable), Ill do anything to stop feeling

40
distressed or upset (regulation=engaging in behaviors to immediately terminate distress), and
Theres nothing worse than feeling distressed or upset (tolerance=low perceived ability to
tolerate affective distress). See Appendix for full scale and subscale breakdown.

Substance Use
At each timepoint, past 30-day frequency (0 days, 1 to 2 days, 3 to 5 days, 6 to 9 days, 10 to 14
days, 15 to 19 days, 20 to 24 days, 25 to 29 days, all 30 days) of alcohol, cigarette, e-cigarette
and cannabis use was collected. Specific substances were collected at varying timepoints due to
the timepoint they were added to the survey. Reports on past 30-day use of alcohol, cigarettes,
and cannabis were collected at all 8 timepoints. Past 30-day e-cigarette use was added at time 3.
All variables were transformed from 0-8 count to 0-30 count (0, 2, 4, 8, 12, 18, 22, 27, 30) for
ease of interpretation.

Baseline Mental Health and Sociodemographic Covariates
Revised Child Anxiety and Depression Scale (RCADS; Chorpita et al., 2000) is a measure of
DSM-IV anxiety and depression disorders and includes subscales for Generalized Anxiety
Disorder (GAD) with 6-items, Major Depressive Disorder (MDD) with 10-items, Panic Disorder
(PD) with 9-items, Social Phobia (SP) with 9-items, and Obsessive Compulsive Disorder (OCD)
with 6-items. Items are rated on a 4-point Likert scale (never, sometimes, often, and always)
where a higher score indicates higher levels of anxiety and depressive symptoms. To control for
clinical manifestations of anxiety and depression disorders in the association between distress
tolerance and substance use without using too many degrees of freedom with all five subscales, a

41
dichotomous variable was created to indicate whether a participant reported symptomology at the
subclinical or clinical cut points (vs. nonclinical; 1 vs. 0) for one or more of the five disorders.
Measures of age, gender (Female, Male), and race/ethnicity (American Indian or Alaska
Native, Asian, Black or African American, Native Hawaiian or Pacific Islander, White, More
than one race, Unknown or Not Reported) were collected at baseline using the self-report
response to investigator-defined forced-choice items.

Statistical Analyses
Descriptive Analyses
Means, internal consistency, and inter-correlation of the Distress Tolerance Scale were examined
at each available timepoint (fall semester of each year in high school corresponding to
measurement timepoints 1, 3, 5, and 7).

Primary Analyses: DTS Class Selection
A growth model for overall DTS was tested as linear versus quadratic to determine best fit for
latent growth mixture modeling (LGMM). To do this, we used negative two loglikelihood ratio
test. LGMM is a person-centered (vs. variable-centered) statistical approach that identifies and
models heterogenous patterns in longitudinal data by grouping participants with more
homogenous patterns into latent classes (Berlin, Parra, & Williams, 2014; Jung, 2008). A series
of LGMM were estimated to identify heterogeneity in growth trajectories of DTS. The number
of classes that best fit the available DTS data was selected based on model fit comparisons using
a series of standard fit indices. Specifically, lower values on Akaikes Information Criterion
(AIC), Bayesian Information Criterion (BIC), and Sample-Size Adjusted BIC (aBIC) and non-

42
significant values for Vuong-Lo-Mendell-Rubin Likelihood Ratio Test (VLRT), Lo-Mendell-
Rubin Adjusted Test (LRT), and Bootstrapped Parametric Likelihood Ratio Test (BLRT). After
DTS class trajectories were identified, we used the manual three step approach (Muthén &
Muthén, 1998-2017) where logits are set for each class to prevent movement of classes in
subsequent models.

Secondary Analyses: Changes in Substance Use (Intercept and Slope) Effect on Emergent DTS
Classes
Associations between substance use trajectories for alcohol, cigarette, e-cigarette, and cannabis
with DTS classes were examined using a series of multi-process models. See Figure 2 for
conceptual model (this was performed separately for each substance). Latent growth curve
modeling (LGCM) was used to estimate intercept (baseline level) and linear slope (rate of
change across assessment periods) for each substance use frequency (alcohol, cigarette, e-
cigarette, and cannabis). A quadratic function was tested for each substance use growth model
but did not show significantly better fit statistics than linear. To estimate the most appropriate
growth models of substance use, the substance use variables were specified as count data in the
model. Negative binomial distribution for substance use count data showed best fit statistics
(Supplemental Table B2). The default estimator in Mplus for negative binomial distribution is
maximum likelihood robust (MLR), which outputs robust standard errors under assumptions of
non-normality (Muthén & Muthén, 1998-2017). The specification of maximum likelihood (ML)
versus MLR did not substantially change model fit or estimates so MLR was maintained to be
more conservative. All models were adjusted for baseline mental health symptoms (RCADS) and
sociodemographics (age, gender, race/ethnicity) by including paths from them to DTS latent

43
class variable. The estimated intercept and linear slope of substance use frequency, in a separate
model for each substance, were examined as predictors of distress tolerance class identified in
the primary analyses. Results of the association between changes in substance use and DTS class
are reported as odds ratios with 95% confidence intervals for class comparisons. Significance
was set to .05. All descriptive analyses were conducted in SPSS 25 and all growth modeling was
conducted in Mplus 8.

Supplementary Analyses: Examination of DTS Classes
To better describe the emergent distress tolerance classes, participant demographic
characteristics (e.g., age, gender, race/ethnicity) were assessed by class belonging. Additionally,
after DTS class trajectories were identified, mean substance use frequencies at each timepoint
were examined by distress tolerance classes to examine group characteristics. See Appendix 3-7.

44
B.3 Results
Descriptive Data
Table B1 reports mean, standard deviations, bivariate correlations, and Cronbachs alpha for
self-reported distress tolerance across 4-years of high school. Internal consistency of the DTS
was excellent with Cronbach alphas ranging from .91 -.94 at each time point.

DTS Class Selection and Description
From LGMM and class-solution comparison, a 3-class quadratic model was determined to fit
the data best (see Table B2 and Table B3 for fit statistics from models with 1-5 classes and
Supplemental Figures 1 and 2 for classes 2 and 4, respectively). The largest class named
Moderate Stable represented 47.6% of the sample and showed an intercept at 3.3 of 5 on the
DTS and remained relatively stable (slope= 0.10, quadratic= -0.05) across timepoints (Class 1 in
Figure B3). The second class named Poor Declining represented 10.1% of the sample and
showed an intercept at 2.9 of 5 on the DTS and moderately decreased (slope= 0.03, quadratic= -
0.11) across the timepoints (Class 2 in Figure B3). Finally, the third class named High
Increasing represe nted 42.3% of the sample and showed an intercept at 3.7 of 5 on the DTS and
moderately increased (slope= -0.04, quadratic= 0.09) across the timepoints (Class 3 in Figure B3,
reference group). Table B4 shows participant characteristics by distress tolerance class.

Latent Growth Models for Each Substance
We observed significant variability in initial past 30-day substance use levels and in rates of
change over time separately in latent growth models for each substance. When estimating the
average substance use trajectories (separately for alcohol, cigarette, e-cigarette, and cannabis) the

45
slope of alcohol and cannabis use significantly increased over time. The slope of e-cigarette use
significantly decreased over time. The slope of cigarette use was not significant. See
Supplemental Table B2 for latent growth model estimates for each substance.

Associations Between Substance Use Intercept and Slope with Emergent DTS Classes
Table B5 shows the fully adjusted final model associations between intercept and slope of each
substance use frequency predicting odds of belonging to DT Moderate Stable or Poor
Declining versus High Increasing. For each additional day where participants reported using
alcohol estimated at baseline, there was a 10% increased odds of belonging to the Moderate
Stable (OR [95% CI]= 1.10 [1.03, 1.18]) and a 20% increased odds of belonging to the Poor
Declining (OR [95% CI]= 1.20 [1.06, 1.36]) versus the High Increasing class. Additionally,
increases in alcohol use at each addition timepoint (rate of acceleration) was associated with
10.05 higher odds of belonging to the Poor Declining class versus the High Increasing class
(OR [95% CI]=10.05 [2.42, 41.75]). For each additional day where participants reported using
cannabis estimated at baseline there was an 8% increased odds of belonging to the Poor
Declining versus the High Increasing (OR [95% CIs]= 1.08 [1.02, 1.14]). For each additional
day where participants reported using e-cigarettes estimated at baseline there was a marginally
significant 8% increased odds of belonging to the Poor Declining versus the High Increasing
(OR [95% CIs]= 1.08 [1.01, 1.16]). Finally, increases in alcohol use at each additional timepoint
was associated with 4.77 higher odds of belonging to the Poor Declining class versus the
Moderate Stable class ( OR [95% CI]=4.77[1.14, 20.05]). All other substance use intercept and
slope estimates were not significantly associated with distress tolerance class belonging. Gender

46
and RCADS were significant covariates associated with DTS class in all models, while
race/ethnicity and age were not significantly associated with DTS classes.

47
B.4 Discussion
The current study classified a sample of 14 to 18-year olds based on their development of self-
reported distress tolerance across four years of high school and examined associations between
substance use trajectories across those same four years with distress tolerance class belonging.
Findings suggest (1) there are three qualitatively distinct trajectories of perceived distress
tolerance based on this diverse sample of high school adolescents and (2) greater baseline levels
of alcohol, e-cigarettes, and cannabis use at the start of high school and greater change at each
timepoint of alcohol use was associated with belonging to the Poor Declining (vs. High
Increasing and vs. Moderate Stab le ) distress tolerance trajectories. The current study
advances understanding in the etiology of perceived distress tolerance and trajectory association
with substance use during mid-adolescent, or high school, years.
We identified three emergent classes of perceived distress tolerance measured by The
Distress Tolerance Scale (Simons & Gaher, 2005) at the beginning of each year in high school.
To the best of our knowledge, this is the first study to examine such patterns of the DTS among
this age group, and findings advance previous work that has examined the developmental
trajectory of distress tolerance. For example, Cummings et al. examined the developmental
trajectory of distress tolerance (measured by the Behavioral Indicator of Resiliency to Distress
[BIRD]) among a sample of 277 adolescent boys and girls (mean age=11 years old) and found
that distress tolerance was temporally stable (from a univariate growth trajectory model;
Cummings et al., 2013). From our growth mixture modeling approach, we found the majority of
adolescents in our sample were relatively high or stable yet there was a risky group. The
statistical approach used in the current study allowed for such specification, which holds clinical

48
relevance for assessing and identifying what risky distress tolerance development may look like
across high school.
Overall, those who reported greater alcohol, e-cigarette use, and cannabis at baseline
were more likely to be in the Poor Declining ve rsus the High Increasing distress tolerance
class. Additionally, increases in only alcohol use at each timepoint associated with greater odds
of belonging to poorer distress tolerance classes versus the other two classes. These findings
align with previous literature identifying a cross-sectional association between distress tolerance
and alcohol consumption and problems (Simons, Sistad, Simons, & Hansen, 2018) and advances
the longitudinal relation suggesting early alcohol use may alter distress tolerance development.
Previous research indicates early substance use, defined as before the age of 15, has detrimental
effects on the adolescent brain development and remodeling (Ewing, Sakhardande, &
Blakemore, 2014) as well as other consequences such as substance disorders, poorer academic
performance, and higher rates of mental health symptoms and disorders (Otten et al., 2019). All
of which suggest substance use during adolescence may affect their ability to navigate
socioemotional demands in an adaptive, and healthy manner. However, additional research
shows adolescents with increased stress-reactivity are vulnerable to maladaptive behaviors such
as being more likely to initiate substance use and progress to addiction or relapse (Chaplin et al.,
2010; Liu et al., 2012). While additional research is warranted for replication and temporal
sequence, these findings suggest greater use of commonly used substances at the start of high
school could potentially derail optimal development of distress tolerance.
Limitations of the current study primarily surround inference of temporal nature. Given
that distress tolerance and substance use were measured at overlapping timepoints, we are only
able to infer directionality for substance use intercept (baseline) and association with distress

49
tolerance class. Additional research is needed to further examine the temporal nature of distress
tolerance trajectory and substance use by measuring them at different (non-overlapping)
timepoints. Thus, this study reports the association between the slope or rate of change in
substance use across high school with distress tolerance class belonging and the direct cause or
temporal nature of cannot be inferred.
These findings contribute greatly to current gaps and call to action for research to
examine distress tolerance with more sophisticated analyses and in association with real-world
outcomes (Glassman et al., 2016). This is the first study to characterize qualitatively distinct
developmental trajectories of perceived distress tolerance growth during high school and
estimate the association with latent growth trajectories of substance use frequency in a large,
diverse sample. Overall, findings suggest adolescents who report greater early substance use are
more likely to belong to the Poor Declining or risky distress tolerance group than the Hig h
Increasing or resilient distress tolerance group. Additional research is needed to replicate
findings before further interpretation of directionality.

50
B.5 Figures and Tables

Figure B1. Analytic Sample Based on Study Accrual and Distress Tolerance Scale Respondents

Note. Analytic sample consists of 2,163 high school students surveyed from 2013-2017 with complete
data for the Distress Tolerance Scale (DTS).

51

Figure B2. Conceptual Model of Latent Substance Use Intercept and Linear Slope on Distress Tolerance Scale Classes Across High
School

Note. Substance use measured by past 30-day use of alcohol, cigarettes, e-cigarettes, and cannabis and modeled separately for each substance.

52
Figure B3. Distress Tolerance Scale Estimated Means Across Time for Each of the Three Classes Across High School (N=2,163)

Note. Latent growth mixture modeling (LGMM) was used to identify heterogenous classes to characterize growth trajectories of DTS. A 3-class
quadratic model was determined to fit the data best, which is displayed here.

53
Table B1. Descriptive Statistics, Internal Consistency and Inter-Correlation of Distress
Tolerance Scale at Each Timepoint

Distress Tolerance Scale Overall
M(SD)
1. 2. 3. 4.
1. Baseline (n=3,205) 3.43 (0.85) (.91) - - -
2. 12-month Follow-up (n=2,946) 3.47 (0.88) .47** (.93) - -
3. 24-month Follow-up (n=2,671) 3.52 (0.89) .40** .52** (.93) -
4. 36-month Follow-up (n=2,490) 3.55 (0.89) .37** .46** .55** (.94)

Note. Cronbach alphas along diagonal and Pearson correlations in bottom half of table to show stability
across time. Response for the Distress Tolerance Scale range from 1.00-5.00.
P-values * <.05 ** <.01 *** <.001

54

Table B2. Linear versus Quadratic Growth Model Comparison for Distress Tolerance Scale Across High School (N=2,163)

Log Likelihood Degrees of Freedom -2 Log Likelihood P-Value
LINEAR -9952.24 9 19904.47 <.0001
QUADRATIC -9941.87 13 19883.73

Note. P-value < .05 suggests a quadratic term significantly improves model fit and will be used in DTS LGMM to estimate classes.

55
Table B3. Quadratic Growth Mixture Model Fit Statistics for the Distress Tolerance Scale Across High School (N=2,163)

Log
Likelihood
AIC BIC ADJ. BIC ENTROPY VLRT P-value LRT LRT
P-Value
BLRT BLRT
P-Value
1 CLASS -9941.865 19909.730 19983.561 19942.258 - - - - - - -
2 CLASS -9916.691 19867.382 19963.929 19909.918 0.479 -9941.865 <.001 48.761 <.001 -9941.865 < 0.001
3 CLASS -9801.485 19644.969 19764.233 19697.514 0.793 -9888.799 < 0.001 169.123 < 0.001 -9888.799 < 0.001
4 CLASS -9863.342 19776.683 19918.664 19839.236 0.770 -9888.799 .240 49.985 .240 -9888.799 < 0.001
5 CLASS -9769.349 19596.699 19761.397 19669.260 0.825 -9784.276 .247 28.912 .257 -9784.276 < 0.001

Note. VLRT= Vuong-Lo-Mendell-Rubin Likelihood Ratio Test; LRT= Lo-Mendell-Rubin Adjusted Lrt Test; BLRT= Bootstrapped Parametric
Likelihood Ratio Test. GMM with 4 and 5 classes added groups with less than 2% of sample and thus 3 class solution fit these data best.

56
Table B4. Participant Characteristics by Distress Tolerance Classes

Note. Sample sizes vary slightly in this table because the nominal class variable (vs. fixed logits) was
used to analyze participant characteristics in SPSS.
*Signifies groups are significantly different based on ANOVA and T-tests.

DT Class 1
47.6% (n=1034),
Moderate Stable

DT Class 2
10.1% (n=205),
Poor Declining

DT Class 3
42.3% (n=924),
High Increasing

Total
(N=2,163)
M(SD) or N(%)
Age 14.56 (0.39) 14.54 (0.27) 14.57 (0.39) 14.56 (0.39)
Gender*
Female 625 (60.4%) 142 (69.3%) 443 (47.9%) 1210 (55.9%)
Male 409 (39.6%) 63 (30.7%) 481 (52.1%) 953 (44.1%)
Race/Ethnicity
Hispanic 468 (45.3%) 79 (38.5%) 421 (45.5%) 968 (44.8%)
Asian 210 (20.3%) 50 (24.4%) 190 (20.6%) 450 (20.8%)
Black 44 (4.3%) 7 (3.4%) 35 (3.8%) 86 (4.0%)
Multiracial 62 (6.0%) 17 (8.3%) 54 (5.8%) 133 (6.1%)
Other 81 (7.8%) 16 (7.8%) 56 (6.1%) 153 (7.1%)
White 169 (16.3%) 36 (17.6%) 168 (18.2%) 373 (17.2%)
RCADS*
Non-Clinical 848 (82.0%) 156 (76.1%) 835 (90.4%) 1839 (85.0%)
Clinical 186 (18.0%) 49 (23.9%) 89 (9.6%) 324 (15.0%)

57
Table B5. Past 30-day Use Frequency Baseline Intercept and Linear Slope Across High School
on Distress Tolerance Class (N =2,108)

Note. Substance use outcomes re-coded from 0-8 count to 0-30 (0, 2, 4, 8, 12, 18, 22, 27, 30). Model
specified for negative binomial count data. Alcohol, cigarette, and cannabis use collected at timepoints 1-
8; e-cigarette use collected at timepoints 3-8. Models run separately to use all available data and examine
differences across substances. Models adjusted for baseline sociodemographics (age, gender,
race/ethnicity) and mental health status (RCADS 0/1).
P < .05* P <.01** P<.001***

DT Moderate Stable
(vs. DT High Increasing)
DT Poor Declining
(vs. DT High Increasing)
DT Poor Declining
(vs. DT Moderate Stable)
OR (95% CI)
Alcohol Use
Intercept 1.10 (1.03, 1.18)* 1.20 (1.06, 1.36)* 1.09 (0.96, 1.24)
Slope 2.11 (0.96, 4.61) 10.05 (2.42, 41.75)** 4.77 (1.14 , 20.05)*
Cigarette Use
Intercept 1.03 (0.99, 1.06) 1.05 (1.00, 1.11) 1.02 (0.97, 1.08)
Slope 1.16 (0.77, 1.76) 1.22 (0.63, 1.76) 1.03 (0.53, 2.03)
E-Cigarette Use
Intercept 1.02 (0.98, 1.06) 1.08 (1.01, 1.16)* 1.06 (0.98, 1.13)
Slope 1.02 (0.78, 1.35) 0.72 (0.45, 1.15) 0.70 (44, 1.14)
Cannabis Use
Intercept 1.03 (0.99, 1.07) 1.08 (1.02, 1.14)* 1.05 (0.99, 1.11)
Slope 1.48 (0.93, 2.34) 2.06 (0.98, 4.31) 1.39 (0.65, 2.97)

58
Supplemental Figure B1. Distress Tolerance Scale Sample Means for Two Class Trajectories Across High
School

Note. Latent growth mixture modeling (LGMM) was used to identify heterogenous classes to characterize growth
trajectories of DTS.

59
Supplemental Figure B2. Distress Tolerance Scale Sample Means for Four Class Trajectories Across High
School

Note. Latent growth mixture modeling (LGMM) was used to identify heterogenous classes to characterize growth
trajectories of DTS.

60
Supplemental Table B1. Model Fit Statistics for Alcohol Use Growth Model Comparisons
Across High School

Log
Likelihood
AIC Adj. BIC
Linear Poisson -16631.220 33272.440 33284.951
Linear Negative Binomial -14251.800 28529.599 28562.127
Quadratic Negative Binomial -14226.625 28487.250 28529.786
Linear Zero-Inflated Negative Binomial -14056.920 28155.839 28208.384
Linear Negative Binomial Hurdle -15164.932 30371.864 30424.408

Note. This supplemental table only displays a snapshot of the model fit indices for alcohol growth model
comparisons. Although the quadratic function showed slightly better model fit statistics, the quadratic
term was not significant, and a linear model was retained for parsimony. Model estimates were compared
for zero-inflated negative binomial and negative binomial model and in the output there was an error
message suggesting there may be untrustworthy estimates likely due to model nonidentification so linear
negative binomial was retained as best fitting model.

61
Supplemental Table B2. Past 30-day Substance Use Linear Growth Models Across High
School

Estimate S.E. P-Value
Alcohol
Means
Intercept -3.64 0.17 <.001
Slope 0.33 0.03 <.001
Variances
Intercept 12.189 0.82 <.001
Slope 0.14 0.01 <.001
Cigarette
Means
Intercept -8.14 0.86 <.001
Slope -0.07 0.17 .70
Variances
Intercept 25.57 4.38 <.001
Slope 0.48 0.12 <.001
E-Cigarette
Means
Intercept -5.19 0.39 <.001
Slope -0.56 0.14 <.001
Variances
Intercept 15.88 1.70 <.001
Slope 0.72 0.12 <.001
Cannabis
Means
Intercept -7.32 0.37 <.001
Slope 0.61 0.06 <.001
Variances
Intercept 29.90 2.33 <.001
Slope 0.36 0.04 <.001

Note. Use days in past month from 0-30 (0, 2, 4, 8, 12, 18, 22, 27, 30) and specified as negative binomial
count. Alcohol, marijuana, and cigarette data were collected at waves 1-8; e-cigarette data were collected at
waves 3-8.

62
Chapter 4. Facets of Mindfulness Disposition and Frequency of
Substance Use After High School

Abstract

Background: Mindfulness has been associated with reduced risk for substance use among youth
and young adults. While two primary components of mindfulness – attention monitoring and
acceptance – have been hypothesized to reduce the risk of substance use, the associations of each
component, alone and in combination, have not been examined.

Methods: Data for the current study were collected via online surveys conducted in 2018-2019
from participants enrolled in an ongoing prospective cohort in Southern California (N=2,273;
58.6% female; 55.2% Hispanic). Mindfulness was measured by the Five-Facet Mindfulness
Questionnaire-15, which includes subscales for attention monitoring (observing, acting with
awareness) and acceptance (nonreactivity, nonjudging, and describing). Participants reported
number of days they used alcohol, cigarettes, e-cigarettes, and cannabis in the past month (0-30
days). Zero-inflated negative binomial regression models were used to evaluate the main effects
of each subscale and the interactive effects of attention monitoring and acceptance subscales
with frequency of use of each substance, with adjustment for sociodemographic characteristics
and mental health variables. Regression coefficients were exponentiated and are presented as
Incident Rate Ratios (IRR).

Results: Among those with mean levels of non-reactivity, each 5-point increase in acting with
awareness was associated with a 21% decrease in frequency of alcohol use in past 30 days
(IRR[95% CI]=0.79 [0.69, 0.92]). This association differed for those with lower or higher levels

63
of non-reactivity (see Figure C1). Among those in 25th percentile for non-reactivity (low levels
of non-reactivity), acting with awareness was not significantly associated with frequency of
alcohol use in the past 30-days (IRR[95% CI]=0.93 [0.79, 1.11]). Among those in the 75th
percentile of non-reactivity (high levels of non-reactivity), each 5-point increase in acting with
awareness was associated with a 28% decrease in the frequency of alcohol use in the past 30
days (IRR[95% CI]=0.72 [0.62, 0.84]). All other individual scale and interactions were not
significant after accounting for multiple comparisons.

Conclusions: Findings suggest not all mindfulness facets are protective against frequency of
substance use. While longitudinal and psychometric research is needed, training the synergistic
skills of acting with awareness and adopting a non-reactive stance may be beneficial in
preventing high rates of alcohol use in emerging adults.

64
C.1 Introduction

Mindfulness disposition refers specifically to the degree to which one is purposefully attentive to
and accepting of the present moment (Brown & Ryan, 2003; Crane et al., 2017; Kabat-Zinn,
1982). Greater mindfulness disposition has been implicated as a protective factor against
substance use in both adolescents and adults (Bowen & Enkema, 2014; Robinson et al., 2014). A
recent meta-analytic quantification of the relation between overall trait mindfulness and
substance use from an aggregated sample found a weighted r = -.13, which was statistically
significant (Karyadi, Vanderveen, & Cyders, 2014). However, psychometric research has
established that mindfulness is a multidimensional construct with five distinct facets: observing,
acting with awareness, non-reactivity, non-judging, and describing which are commonly
assessed using the Five-Facet Mindfulness Questionnaire (FFMQ; Baer, Hopkins, et al., 2006).
In the same meta-analysis, associations of mindfulness with substance use behaviors were
strongest for acting with awareness, non-judging, and non-reactivity (vs. observing and
describing; Karyadi et al., 2014). The current literature suggests that certain mechanisms of
mindfulness may be driving the protective effects against greater substance use, but each facets
specific role or mechanisms through which it contributes toward health outcomes are not yet
clear (Cortazar & Calvete, 2019).
Monitor and Acceptance Theory (MAT) is an emerging theory proposing that not all five
mindfulness facets may be equally protective against substance use and other adverse risk
behaviors and suggests there are two active mechanisms that explain mindfulness-related
benefits (Lindsay & Creswell, 2017). The first is attention monitoring, which refers to skills that
enhance awareness of present-moment experience (the what captured by the observing
subscale in the FFMQ) and the second is acceptance, which refers to skills that modify the way

65
one relates to the present-moment experience, regulating reactivity to affective experience (the
how captured by the non -judging and non-reactivity subscales in the FFMQ; Lindsay &
Creswell, 2017). Preliminary evidence suggests that only when both facets are high do they
explain how mindfulness improves negative affect, stress, and health outcomes. For example, in
a sample of college students, controlling for personality traits and using the (full 39-item) FFMQ,
there was a synergistic interaction between observing and non-reactivity suggesting that the
tendency to be focused on the present-moment experience may be protective against alcohol and
tobacco use when the tendency to simultaneously be non-reactive toward the observed stimuli is
also high (vs. low; Eisenlohr-Moul et al., 2012). However, non-judging did not act as a
moderator in the same way that non-reactivity did with observing in this study.
Research examining mindfulness disposition and substance use in late adolescence and
emerging adulthood has mostly evaluated alcohol, traditional tobacco products, and some with
cannabis. As such, it is unclear whether interactive mechanistic effects extend to novel products
(e.g., e-cigarettes). Given the concerns regarding adverse effects of e-cigarette use in adolescence
(e.g., risk for transition to combustible products, risk of dependence with high-nicotine content
products, risk of adverse respiratory health effects; Barrington-Trimis & Leventhal, 2018; Gotts,
Jordt, Mcconnell, & Tarran, 2019; Layden et al., 2019), research is needed to provide insight into
theoretical, clinical, and policy targets for preventing use of these highly addictive products
(such as scalable, low-cost interventions like facets of mindfulness skills) and reduce the later
public health burden of addiction.
The current study tested the association of each mindfulness facet on frequency of past
30-day use of alcohol, cigarettes, e-cigarettes, and cannabis in a large, diverse sample of

66
emerging adults, and examined the interactive effect of attention monitoring facets with
acceptance facets on frequency of past 30-day substance use.

67
C.2 Methods

Participants
Data were collected via online surveys from a prospective cohort of young adults in Southern
California. The cohort was originally recruited when participants were in 9th grade in Fall 2013
(N=3396). The current analysis includes data from the most recent survey, conducted online
between October 2018 and November 2019 (N=2422; 71%). The analytic sample includes
participants with data on the FFMQ (N=2,273, 94% of Wave 9 respondents).

Measures
Five-Facet Mindfulness Questionnaire-Short Form (FFMQ-SF; Baer, Smith, Hopkins,
Krietemeyer, & Toney, 2006; Baer et al., 2008) is a 15-item measure of five mindfulness facets:
observing (3-items, Cronbachs alpha=.71 ), describing (3-items, Cronbachs alpha=.41 ), acting
with awareness (3-items, Cronbachs alpha=.81 ), non-judging (3-items, Cronbachs alpha=.86 ),
and non-reactivity (3-items, Cronbachs alpha=.77 ). Items are rated on a 5-point Likert scale (1 =
Never/Rarely true to 5 = Very Often/Always true), where a higher score reflects a higher level of
mindfulness disposition. Example items include, When I take a shower or a bath, I stay alert to
the sensations of water on my body (observing= attending to or noticing internal and external
experiences such as sounds, emotions, thoughts, body sensations, and scent), Im good at
finding words to describe my feelings (describing= the ability to express in words ones
experience), and When I have distressing thoughts or images, I step back and am aware of the
thought or image without getting taken over by it (non -reactivity= ability to detach from
thoughts and emotions, allowing them to come and go without getting involved or carried away
by them), I dont pay attention to what Im doing b ecause Im daydreaming, worrying, or

68
otherwise distracted reverse scored ( acting with awareness=attending to ones present moment
activity, rather than being on autopilot while attention is focused elsewhere). I believe some of
my thoughts are abnormal or bad and I shouldnt think that way reverse scored (non-judging=
accepting and not evaluating thoughts and emotions as good or bad). Sum and subscale scores
were standardized by their interquartile ranges (7 and 5, respectively). See Appendix 2 for full
scale and subscale breakdown.

Substance Use: Participants reported on frequency of past 30-day use (0 days, 1 to 2 days, 3 to 5
days, 6 to 9 days, 10 to 19 days, 20 to 29 days, all 30 days) of alcohol, cigarettes, electronic
cigarettes, and cannabis.

Mental Health and Sociodemographic Covariates
Symptoms of anxiety were assessed using the 7-item Generalized Anxiety Subscale (GAD;
Spitzer, Kroenke, Williams, & Lowe, 2006). Items are rated on a 4-point Likert scale (never,
sometimes, often, and always) where a higher score indicates higher levels of anxiety symptoms.
Symptoms of depression were assessed using Center for Epidemiologic Studies Depression Scale
(CESD; Radloff, 1977). CESD includes 10-items about past week symptoms. Items are rated on
a 4-point Likert scale (Rarely or None of the Time, Some or Little of the Time, Moderately or
Much of the time, Most or Almost All the Time).
Measures of age, gender (Female, Male), ethnicity (Hispanic or non-Hispanic), race
(American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or
Pacific Islander, White, Multi-ethnic or Multi-racial, Other), and whether they were currently

69
enrolled in a degree program (yes, no, dont know) were collected using investigator-defined
forced-choice items.

Statistical Analyses
Descriptive Analyses
Mean or sum scores, distribution, and inter-correlations of FFMQ, substance use
frequency, and mental health variables were first examined.

Primary analyses: FFMQ Sum and Subscale and Substance Use Frequency
Due to overdispersion and zero-inflation of substance use variables, a series of zero inflated
negative binomial (ZINB) models were used to examine cross-sectional associations of each
FFMQ subscale and past 30-day substance use frequency, where the median value of each
substance use category was used to model the outcome variable. In each set of models for each
substance (alcohol, cigarette, e-cigarette, cannabis), a single FFMQ predictor was tested,
adjusting for age, gender, ethnicity/race, current enrollment in degree program, CESD, and
GAD.

Secondary analyses: FFMQ Attention Monitoring Acceptance Subscale Interaction and
Substance Use Frequency
Then interactions between two FFMQ subscales with one another were added in subsequent
steps of the modeling. Subscales were categorized into attention monitoring (i.e., FFMQ
observing and FFMQ acting with awareness) and acceptance (i.e., FFMQ non -reactivity,
FFMQ non-judging, FFMQ describing). All combinations of attention monitoring

70
acceptance were examined (i.e., observing non-reactivity, observing non-judging,
observing describing, acting with awareness non-reactivity, acting with awareness non-
judging, acting with awareness describing). Separate sets of models were tested for each
substance outcome (i.e., alcohol, cigarette, e-cigarette, cannabis).
Missing baseline sociodemographic covariates were imputed (SAS Institute Inc., 2016) to
avoid listwise deletion of missing data. All ZINB regression coefficients were exponentiated and
are presented as Incident Rate Ratios (IRR). Models were corrected for false discovery rate using
the Benjamini Hochberg method (Jafari & Ansari-Pour, 2019). All analyses were conducted in
SAS 9.4.

71
C.3 Results
Descriptive Data
Participants in the current sample were 55% Hispanic, 59% female, and 65% currently enrolled
in a degree program, with a mean age of 19.2 years (SD=0.7) at survey completion. See Table
C1. FFMQ subscales observing and non-reactivity were inversely associated with FFMQ
subscales non-judging and acting with awareness and positively associated with anxiety and
depression symptoms. All other FFMQ subscales were negatively associated with anxiety and
depression symptoms. See Table C2.

FFMQ Subscale and Substance Use Frequency
No statistically significant associations of FFMQ (either sum or subscale) with any substance use
(alcohol, cigarette, e-cigarette, or cannabis) were observed when adjusting for all covariates and
accounting for multiple comparisons (all ps>0.05) (see Table C3).

FFMQ Subscale Interaction and Substance Use Frequency
After correction for multiple comparisons, the interaction between acting with awareness non-
reactivity was statistically significant in models evaluating frequency of past 30-day alcohol use
(Table C4). Among those with mean levels of non-reactivity, each 5-point increase in acting with
awareness was associated with a 21% decrease in the frequency of alcohol use in the past 30
days (IRR[95% CI]=0.79 [0.69, 0.92]). This association differed for those with lower or higher
levels of non-reactivity (see Figure C1). Among those in 25th percentile for non-reactivity (low
levels of non-reactivity), acting with awareness was not significantly associated with frequency
of alcohol use in the past 30 days (IRR[95% CI]=0.93 [0.79, 1.11]). Among those in the 75th

72
percentile of non-reactivity (high levels of non-reactivity), each 5-point increase in acting with
awareness was associated with a 28% decrease in the frequency of alcohol use in the past 30
days (IRR[95% CI]=0.72 [0.62, 0.84]). All other interactions were not significant after
accounting for multiple comparisons.

73
C.4 Discussion
The current study is the first to directly examine all interactions between attention monitoring
by acceptance mindfulness disposition facets (from the Monitor and Acceptance Theory,
MAT; Lindsay & Creswell, 2017) in association with substance use frequency. We tested
associations of mindfulness mechanisms with frequency of alcohol, cigarette, e-cigarette, and
cannabis use (separately) in a large, racially/ethnically, diverse sample of emerging adults. Our
findings suggest the association between mindfulness facets and substance use is multifaceted
and only one combination of attention monitoring and acceptance facets was protective against
frequency of alcohol use (i.e., acting with awareness nonreactivity).
Findings from the current study showed some evidence that greater sum mindfulness and
acting with awareness subscale was associated with lower rates of alcohol use only (before
multiple tests correction). This finding aligns with previous work showing overall mindfulness
was more protective against alcohol use than tobacco and cannabis use (Karyadi et al., 2014).
Furthermore, acting with awareness has emerged as a prominent facet predicting fewer
psychological problems and an important dimension in adolescent adaptive coping (Cortazar &
Calvete, 2019). However, contrary to hypotheses and related findings (Cortazar & Calvete,
2019), greater describing subscale was associated with greater e-cigarette use (also before
multiple tests correction). At large, our study among diverse emerging adults did not replicate
findings of the previously documented (Robinson et al., 2014) protective association between
greater mindfulness and lower substance use.
The interaction between acting with awareness nonreactivity indicated the greatest
evidence for protective patterns against past 30-day substance use for alcohol (and e-cigarette
use although no longer significant [p=0.07] after multiple tests correction), such that greater

74
levels of acting with awareness predicted lowest levels of use among those with mean and high
levels of non-reactivity but not low levels of non-reactivity. This finding is most similar to
previous literature that reported the tendency to be focused on the present-moment experience
(using the observing versus acting with awareness subscale), was protective against alcohol and
tobacco use when the tendency to simultaneously be non-reactive toward the observed stimuli
was also high (vs. low; Eisenlohr-Moul et al., 2012). The current study is the first study to our
knowledge that directly tests acting with awareness nonreactivity on substance use frequency.
Previous work alludes to these two mindfulness mechanisms aligning with experiential avoidant
substance use or avoidant-based coping in the self-medication model (Bowen & Enkema, 2014),
which is characterized by substance use in attempt to alter or alleviate negative thoughts and
feelings (Lindsay & Creswell, 2017). While longitudinal and psychometric research is needed,
training the synergistic skills of acting with awareness and adopting a non-reactive stance may be
beneficial in preventing high rates of alcohol use in emerging adults.
After multiple tests correction no other mindfulness subscale interactions significantly
associated with past 30-day substance use frequency. While one plausible explanation may be
that different drugs serve different purposes (i.e., relief vs. reward), recent research is emerging
that suggests protective effects of mindfulness mechanisms are quite complex and likely to differ
by health behaviors. For example, one study found that higher levels of non-judging predicted
lower internalizing and externalizing symptoms in youth with higher levels of acting with
awareness (but not among those with low non-judging; Cortazar & Calvete, 2019). This
complexity may be due in part to the participants developmental stage of mindfulness skills or
status of health behavior (e.g., substance use stage) which is not captured in these measures and

75
there is much left to be learned about how protective effects of mindfulness may differ by health
behaviors.
Limitations of the current study include the observational, cross-sectional nature of the
data which do not allow inference of causality or directionality regarding the mechanisms
driving mindfulness disposition effects. Longitudinal research is needed to better understand
temporal nature in mindfulness skill development especially in the context of state versus trait
(Kiken, Garland, Bluth, Palsson, & Gaylord, 2015) as well as replicate these substance specific
differences before further interpretation. The current study contributes to the complex findings
on the relation of mindfulness disposition and substance use in milieu of new products and a
large diverse sample of emerging adults. At large, findings from this study do not fully support
MAT – using the theory driven subscales (observing, non-reactivity, non-judging) nor our
expansion to all five FFMQ subscales (to include acting with awareness and describing). Given
the evolution and theory development of active mindfulness mechanisms, new scales should be
developed to more accurately capture facets of attention monitoring and acceptance.

76
C.5 Figures and Tables
Figure C1. Acting with Awareness Non-reactivity on Past 30-Day Alcohol Use

Note. Five-Facet Mindfulness Questionnaire-Short Form (FFMQ-SF) subscale (Acting with Awareness and Non-Reactivity) raw scores were
standardized by their interquartile range of 5 and centered on the mean. Pink group of 1.5 represents those with greatest level of non-reactivity
mindfulness skills while -0.9 represents those with lowest levels of non-reactivity. Alcohol use represents number of reported days of use in past
30-days. Estimates from model controlling for age, gender, race, ethnicity, enrollment in higher education program, and depression and anxiety
symptoms (measured by the Center for Epidemiologic Studies Depression and the Generalized Anxiety Disorder scales).

77

Table C1. Baseline Participant Descriptive Characteristics (N=2,273)

Study Variable Overall Sample
N(%) or M(SD)
Age 19.21 (0.70)
Gender
Male 941 (41.4%)
Female 1332 (58.6%)
Hispanic Ethnicity
No 1019 (44.8%)
Yes 1254 (55.2%)
Race
American Indian or Alaska Native 80 (3.5%)
Asian 441 (19.4%)
Black or African American 115 (5.1%)
Native Hawaiian or Pacific Islander 77 (3.4%)
White 655 (28.8%)
Multi-ethnic or Multi-racial 452 (19.9%)
Other 453 (19.9%)
Currently Enrolled in Higher Ed
No 648 (28.5%)
Yes 1488 (65.5%)
Unsure 137 (6.0%)

Note. Analytic sample consists of participants who responded to the Five-Facet Mindfulness
Questionnaire-Short Form (FFMQ-SF), which was 94% of Wave 9 respondents.

78
Table C2. Correlation matrix of Five-Facet Mindfulness Questionnaire Subscales, Substance Use, and Mental Health Variables
(N=2,273)

Variable Sum or
M(SD)
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
1. FFMQ-SF Sum 47.67 (6.52) (.53) - - - - - - - - - -
2. FFMQ-SF Observing 8.54 (3.06) .47* (.71) - - - - - - - - -
3. FFMQ-SF Non-judging 11.29 (3.16) .44* -.35* (.86) - - - - - - - -
4. FFMQ-SF Non-reactivity 7.48 (3.00) .33* .58* -.50* (.77) - - - - - - -
5. FFMQ-SF Act with awareness 11.05 (2.83) .36* -.44* .67* -.51* (.81) - - - - - -
6. FFMQ-SF Describing 9.31 (2.38) .70* .27* .17* .18* .10* (.41) - - - - -
7. Alcohol 2.36 (4.56) -.05* .04 -.08* .03 -.09* -.01 - - - - -
8. Cigarettes 0.47 (2.75) -.02 .04 -.07* .03 -.06* .02 .20* - - - -
9. E-cigarettes 2.03 (6.51) .03 .04* -.02 -.00 -.01 .07* .24* .24* - - -
10. Cannabis 3.23 (7.60) -.02 .08* -.05* .01 -.06* -.01 .33* .18* .24* - -
11. CESD Mean 1.36 (0.62) -.44* .14* -.63* .24* -.50* -.27* .08* .03 -.04* .06* -
12. GAD Mean 1.84 (0.78) -.31* .24* -.58* .27* -.47* -.19* .08* .04 -.02 .07* .69*

Note. Five-Facet Mindfulness Questionnaire-Short Form (FFMQ-SF), unstandardized sum scores range from 22.0-74.0 while subscale scores
range from 3.0-15.0, where higher scores represent greater report of mindfulness skills. Cronbach alphas in parentheses along diagonal. Substance
use variables are past 30-day use frequency coded as count values ranging from 1=no use to 7=used every day transformed to 0-30 days for
interpretation. Past 6-month use was coded as 0 to account for all participant responses as they were only prompted to past 30-day use survey if
they reported use in past 6-months. Cannabis use is referring to smoking cannabis.
CESD= Center for Epidemiologic Studies Depression Scale (10 items, mean range 0.4-3.4) and GAD=Generalized Anxiety Disorder (7 items,
mean range 1.0-4.0).
P-value * <.05

79
Table C3. Effect of Mindfulness Disposition Skills on Frequency of Past 30-day Substance Use Adjusted for Sociodemographics and
Mental Health (N=2,273)

Model Predictor Alcohol Cigarette E-Cigarette Cannabis
IRR (95% CI) P-value IRR (95% CI) P-value IRR (95% CI) P-value IRR (95% CI) P-value
FFMQ-SF Sum 0.92 (0.86, 0.997) .04 0.84 (0.65, 1.08) .17 1.08 (0.92, 1.27) .36 1.01 (0.91, 1.11) .91
FFMQ Observing 0.99 (0.89, 1.10) .85 0.82 (0.54, 1.23) .34 1.19 (0.96, 1.47) .11 1.12 (0.96, 1.30) .14
FFMQ Non-judging 0.88 (0.77, 1.01) .08 0.79 (0.47, 1.34) .40 0.86 (0.65, 1.13) .27 0.99 (0.84, 1.19) .99
FFMQ Non-
reactivity
1.01 (0.90, 1.14) .83 0.89 (0.59, 1.32) .55 1.01 (0.79, 1.29) .94 0.93 (0.80, 1.09) .37
FFMQ Acting with
Awareness
0.83 (0.73, 0.95) .005 0.67 (0.39, 1.15) .14 0.88 (0.68, 1.14) .34 1.01 (0.85, 1.19) .94
FFMQ Describing 0.97 (0.85, 1.12) .71 1.24 (0.73, 2.10) .43 1.39 (1.04, 1.84) .02 0.94 (0.78, 1.14) .54

Note. All zero-inflated negative binomial models were run separately and adjusted for age, gender, ethnicity, race, highest education level they
completed, and mental health variables for anxiety and depression (GAD and CESD). Five-Facet Mindfulness Questionnaire-Short Form (FFMQ-
SF or FFMQ) sum and subscale scores were standardized by their interquartile ranges (7 and 5, respectively) and then centered on the mean.

Significant terms are bolded prior to Benjamini-Hochberg correction. After correction, none remained significant.

80
Table C4. Interactive Effects of Mindfulness Disposition Skills on Frequency of Past 30-day Substance Use Adjusted for
Sociodemographics and Mental Health (N=2,273)

Alcohol Cigarette E-Cigarette Cannabis
IRR (95% CI) P-value IRR (95% CI) P-value IRR (95% CI) P-value IRR (95% CI) P-value
Observing 0.98 (0.86, 1.12) .79 0.83 (0.50, 1.36) .45 1.32 (1.01, 1.72) .04 1.28 (1.07, 1.52) .008
Non-reactivity 1.02 (0.88, 1.17) .82 0.88 (0.51, 1.53) .66 0.82 (0.60, 1.11) .20 0.77 (0.64, 0.93) .008
Observing Non-reactivity 1.07 (0.91, 1.25) .43 1.21 (0.69, 2.13) .51 1.19 (0.87, 1.62) .28 1.31 (1.05, 1.62) .01
Observing 0.97 (0.86, 1.08) .55 0.78 (0.52, 1.16) .22 1.17 (0.94, 1.46) .15 1.14 (0.98, 1.34) .09
Non-judging 0.87 (0.76, 1.00) .058 0.79 (0.48, 1.29) .35 0.91 (0.69, 1.21) .53 1.06 (0.88, 1.28) .54
Observing x Non-judging 1.05 (0.91, 1.22) .50 0.83 (0.52, 1.33) .45 0.90 (0.68, 1.19) .47 0.89 (0.73, 1.10) .29
Observing 0.99 (0.89, 1.12) .92 0.76 (0.51, 1.15) .20 1.12 (0.90, 1.40) .31 1.14 (0.98, 1.33) .09
Describing 0.95 (0.82, 1.10) .52 1.50 (0.88, 2.57) .14 1.34 (0.99, 1.81) .056 0.85 (0.70, 1.05) .14
Observing x Describing 1.19 (0.97, 1.46) .09 0.70 (0.34, 1.43) .32 0.93 (0.62, 1.39) .72 1.19 (0.91, 1.56) .21
Acting w/ Awareness 0.79 (0.69, 0.92) .002 0.65 (0.39, 1.09) .09 0.89 (0.66, 1.20) .44 0.98 (0.81, 1.19) .83
Non-reactivity 0.93 (0.82, 1.06) .28 0.74 (0.49, 1.13) .16 0.94 (0.72, 1.24) .68 0.92 (0.77, 1.09) .32
Acting w/ Awareness Non-reactivity 0.72 (0.60, 0.87) .0004 0.76 (0.43, 1.34) .35 0.69 (0.48, 0.98) .04 0.84 (0.68, 1.05) .13
Acting w/ Awareness 0.84 (0.72, 0.98) .03 0.69 (0.35, 1.37) .29 0.97 (0.71, 1.33) .86 1.02 (0.84, 1.25) .84
Nonjudging 0.99 (0.85, 1.16) .93 1.06 (0.54, 2.07) .86 0.90 (0.65, 1.23) .49 1.03 (0.84, 1.28) .77
Acting w/ Awareness Nonjudging 1.22 (1.02, 1.46) .03 1.27 (0.69, 2.33) .44 1.30 (0.94, 1.79) .12 1.19 (0.96, 1.48) .12
Acting w/Awareness 0.83 (0.72, 0.94) .004 0.68 (0.40, 1.15) .15 0.93 (0.71, 1.21) .57 1.00 (0.84, 1.19) .99
Describing 0.96 (0.84, 1.10) .58 1.24 (0.69, 2.24) .47 1.38 (1.04, 1.83) .03 0.91 (0.75, 1.11) .35
Acting w/Awareness x Describing 0.85 (0.69, 1.05) .13 1.05 (0.41, 2.69) .91 0.80 (0.49, 1.30) .37 0.84 (0.64, 1.12) .24

Note. All zero-inflated negative binomial models were run separately and adjusted for age, gender, ethnicity, race, highest education level they
completed, and mental health variables for anxiety and depression (GAD and CESD). Five-Facet Mindfulness Questionnaire-Short Form (FFMQ-
SF or FFMQ) sum and subscale scores were standardized by their interquartile ranges (7 and 5, respectively) and then centered on the mean score.

Significant interaction terms are bolded prior to Benjamini-Hochberg correction. After correction, only remaining significant interaction was
Acting w/ Awareness Non-reactivity on Alcohol Use (Benjamini-Hochberg P-value = 0.0216).

81
Supplemental Figure C1. Acting with Awareness Non-reactivity on Past 30-Day E-cigarette Use

Note. Five-Facet Mindfulness Questionnaire-Short Form (FFMQ-SF) subscale (Acting with Awareness and Non-Reactivity) raw scores were
standardized by their interquartile range of 5 and centered on the mean. Pink group of 1.5 represents those with greatest level of non-reactivity
mindfulness skills while -0.9 represents those with lowest levels of non-reactivity. E-cigarette use represents number of reported days of use in
past 30-days. Estimates from model controlling for age, gender, race, ethnicity, enrollment in higher education program, and depression and
anxiety symptoms (measured by the Center for Epidemiologic Studies Depression Scale Generalized Anxiety Disorder).

82
Supplemental Figure C2. Observing Non-reactivity on Cannabis Use on Past 30-Day Cannabis Use

Note. Five-Facet Mindfulness Questionnaire-Short Form (FFMQ-SF) subscale (Observing and Non-Reactivity) raw scores were standardized by
their interquartile range of 5 and centered on the mean. Pink group of 1.5 represents those with greatest level of non-reactivity mindfulness skills
while -0.9 represents those with lowest levels of non-reactivity. Cannabis use represents number of reported days of use in past 30-days. Cannabis
use is referring to smoking cannabis. Estimates from model controlling for age, gender, race, ethnicity, enrollment in higher education program,
and depression and anxiety symptoms (measured by the Center for Epidemiologic Studies Depression Scale Generalized Anxiety Disorder).

83

Supplemental Figure C3. Acting with Awareness Non-judging on Past 30-Day Alcohol Use

Note. Five-Facet Mindfulness Questionnaire-Short Form (FFMQ-SF) subscale (Acting with Awareness and Non-Judging) raw scores were
standardized by their interquartile range of 5 and centered on the mean. Pink group of 0.74 represents those with greatest level of non-judging
mindfulness skills while -1.7 represents those with lowest levels of non-judging. Alcohol use represents number of reported days of use in past 30-
days. Estimates from model controlling for age, gender, race, ethnicity, enrollment in higher education program, and depression and anxiety
symptoms (measured by the Center for Epidemiologic Studies Depression Scale Generalized Anxiety Disorder).

84
Supplemental Table C1. Acting with Awareness and Non-Reactivity on Past 30-Day Use Frequency of Alcohol and E-Cigarettes
(N=2,273)

Parameter Alcohol E-Cigarette
Estimate SE P-Value Estimate SE P-Value
Intercept .03 .94 .98 1.01 2.23 .65
FFMQ Acting w/ Awareness -0.23 0.07 .002 -.12 0.15 .44
FFMQ Non-reactivity -0.07 0.07 .28 -0.06 0.14 .68
FFMQ Acting w/ Awareness Non-reactivity -0.32 0.09 .0004 -0.38 0.14 .04
Age 0.07 0.05 .13 0.06 0.12 .60
Female -0.01 0.07 .84 0.12 0.13 .37
Hispanic 0.08 0.09 .37 0.48 0.17 <.01
American Indian or Alaska Native 0.09 0.19 .65 0.06 0.40 .88
Asian -0.01 0.14 .94 0.16 024 .50
Black or African American -0.14 0.20 .49 -0.25 0.49 .61
Native Hawaiian or Pacific Islander 0.21 0.19 .26 0.04 0.32 .90
White 0.01 0.11 .90 0.27 0.20 .19
Multi-ethnic or Multi-racial 0.03 0.10 .80 0.06 0.21 .77
Enrolled in Program -0.20 0.07 .01 -0.07 0.15 .64
Dont Know -0.29 0.16 .08 -0.17 0.34 .62
CESD 0.10 0.07 .19 -0.21 0.16 .18
GAD -0.003 0.06 .95 -0.05 0.12 .68
Dispersion 0.82 0.08 - 1.43 0.20 -

Note. All zero-inflated negative binomial models were run separately and adjusted for age, gender, ethnicity, race, highest education level they
completed, and mental health variables for anxiety and depression (GAD and CESD). Five-Facet Mindfulness Questionnaire-Short Form (FFMQ-
SF or FFMQ) sum and subscale scores were standardized by their interquartile ranges (7 and 5, respectively) and centered on means. Female (vs.
Male); Hispanic (vs. Non-Hispanic); American Indian, Asian, Black, White, Multi-racial (vs. other), Enrolled in Program, Dont Know ( vs. Not
Enrolled).
P-values * <.05 ** <.01 *** <.001

85

Supplemental Table C2. Observing and Non-Reactivity on Past 30-Day Use Frequency of Cannabis

Parameter Cannabis
Estimate SE P-Value
Intercept 1.70 1.30 .19
FFMQ Observing 0.24 0.09 .008
FFMQ Non-Reactivity -0.26 0.08 .008
FFMQ Observing Non-Reactivity 0.27 0.11 .01
Age 0.03 0.07 .67
Female 0.19 0.09 .050
Hispanic 0.22 0.13 .09
American Indian or Alaska Native 0.13 0.26 .63
Asian -0.45 0.19 .02
Black or African American -0.45 0.23 .052
Native Hawaiian or Pacific Islander -0.33 0.25 .18
White 0.06 0.14 .66
Multi-ethnic or Multi-racial 0.03 0.13 .80
Enrolled in Program -0.31 0.20 .003
Dont Know -0.33 0.10 .10
CESD 0.05 0.11 .65
GAD 0.06 0.08 .46
Dispersion 1.12 0.11 -

Note. All zero-inflated negative binomial models were run separately and adjusted for age, gender, ethnicity, race, highest education level they
completed, and mental health variables for anxiety and depression (GAD and CESD). Five-Facet Mindfulness Questionnaire-Short Form (FFMQ-
SF or FFMQ) sum and subscale scores were standardized by their interquartile ranges (7 and 5, respectively) and centered on means. Female (vs.
Male); Hispanic (vs. Non-Hispanic); American Indian, Asian, Black, White, Multi-racial (vs. other), Enrolled in Program, Dont Know ( vs. Not
Enrolled).

86
Supplemental Table C3. Acting with Awareness and Non-Judging on Past 30-Day Use Frequency of Alcohol

Parameter Alcohol
Estimate SE P-Value
Intercept 0.11 0.94 .91
FFMQ Acting w/ Awareness -0.17 0.08 .03
FFMQ Non-Judging -0.01 0.08 .93
FFMQ Acting w/ Awareness Non-Judging 0.20 0.09 .03
Age 0.07 0.05 .14
Female -0.02 0.07 .82
Hispanic 0.07 0.09 .44
American Indian or Alaska Native 0.13 0.19 .51
Asian 0.01 0.14 .92
Black or African American -0.14 0.20 .49
Native Hawaiian or Pacific Islander 0.25 0.19 .20
White 0.02 0.11 .86
Multi-ethnic or Multi-racial 0.03 0.10 .74
Enrolled in Program -0.22 0.08 <.01
Dont Know -0.27 0.16 .11
CESD 0.09 0.08 .23
GAD -0.02 0.06 .77
Dispersion 0.84 0.08 -

Note. All zero-inflated negative binomial models were run separately and adjusted for age, gender, ethnicity, race, highest education level they
completed, and mental health variables for anxiety and depression (GAD and CESD). Five-Facet Mindfulness Questionnaire-Short Form (FFMQ-
SF or FFMQ) sum and subscale scores were standardized by their interquartile ranges (7 and 5, respectively) and centered on means. Female (vs.
Male); Hispanic (vs. Non-Hispanic); American Indian, Asian, Black, White, Multi-racial (vs. other), Enrolled in Program, Dont Know ( vs. Not
Enrolled)

87
Chapter 5. Discussion
Overview
The goal of the current dissertation was to deepen the scientific understanding of adolescent
substance use etiology, particularly in regard to malleable individual traits/skills related to stress
reactivity (i.e., distress tolerance and mindfulness disposition). Findings are especially relevant
during a time period of new products (e.g., e-cigarettes) and recent policy changes (e.g., cannabis
legalization, tobacco purchasing age increased). In a large, diverse, sample of adolescents and
emerging adults, I examined (1) bidirectional trajectories of distress tolerance and substance use
during high school and (2) interactive mechanisms of mindfulness disposition associated with
substance use frequency after high school. The specific research objectives were to: (1) examine
the association of distress tolerance with subsequent substance use by baseline use and mental
health disorder statuses for each respective substance throughout high school; (2) characterize
developmental growth trajectories of DTS across high school and examine associations between
baseline and slope of substance use across high school with DTS membership of trajectory; and
(3) examine the association of mindfulness disposition by mechanisms (subscales of attention
monitoring and acceptance described in Monitor and Acceptance Theory) with frequency of
substance use after high school. Findings hold implications for specific components of distress
tolerance and mindfulness disposition in order to target and continue to update and align
substance use prevention efforts during adolescence suitably.
Key Findings
Transdiagnostic Stress Reactivity Factors in Adolescent/Emerging Adulthood Substance Use

88
Distress tolerance: Scientific literature currently lacks developmental work evaluating (1)
distress tolerance as a prospective predictor of substance use escalation and the competing
explanation that (2) substance use lowers or changes the course of distress tolerance (Veilleux,
2019). Studies 1 and 2 aim to fill these gaps and bridge research on self-regulation skill
development with addiction research.
Is distress tolerance associated with substance use frequency in adolescents?
Study 1 found that greater DTS score was protective against past month substance use frequency
across high school among those who had never used substances before. DTS was not associated
with substance use frequency among those who had used substances before. In other words,
higher perceived distress tolerance is associated with lower substance use frequency (among
never-users) but lower distress tolerance was not associated with higher substance use frequency.
The current study cannot ascertain whether those who have used substances (i.e., ever-users) are
more susceptible to initial use due to lower distress tolerance. Future research is needed to better
understand the potential differences in the temporal nature by use status groups. Additionally,
mental health status did not moderate the association between DTS and substance use frequency.
This finding suggests that whether or not adolescents are experiencing clinical or subclinical
anxiety and mood symptomology, greater distress tolerance may be a predictor of lower
substance use frequency above and beyond mental health. Specifically, DTS subscales of
absorption and appraisal seemed to drive this effect. I will discuss these mechanisms more in the
following future research section in combination with mindfulness mechanisms.
Does substance use lower or change the course of distress tolerance development during
adolescence? Study 2 takes into consideration that distress tolerance is likely a dynamic and
contextually changing construct over time, especially during high school. Results showed three

89
main trajectories of distress tolerance development from 9th to 12th grade (i.e., high Increasing,
moderate stable, poor declining). Participants who reported greater alcohol, e-cigarette, and
cannabis use at baseline were more likely to belong in the poor declining distress tolerance group
compared to the high Increasing group. Additionally, those who reported a greater acceleration
of alcohol use across timepoints were more likely to belong to the poor declining distress
tolerance group compared to both the moderate stable and high Increasing group. Findings
suggest that early substance use may negatively alter the developmental course of optimal
distress tolerance.
Collectively, studies 1 and 2 contribute to the literature in several critical ways. Prior to
substance use, distress tolerance seems to be a protective factor against substance use frequency.
After initial substance use and with greater rate of use (specifically alcohol), odds of optimal
distress tolerance development seem to decline. These findings offer further theoretical
conceptualization of distress tolerance as a construct and suggestions for refining future
measurements. Findings in this dissertation contribute evidence to the theoretical debate in the
literature that distress tolerance refers to the general ability to withstand aversive experiences
encompassed by multiple facets rather than one of many related constructs assessing coping
abilities (Evanovich et al., 2019). Future research is needed to enhance the way in which we are
measuring and interpreting those underlying facets. Findings in this dissertation also contribute
evidence regarding which DTS facets may best cluster together and associate with lower rates of
substance use in adolescents (i.e., absorption and appraisal). This finding partially aligns with a
recent study that examined the factor structure of the DTS and found that absorption, appraisal,
and tolerance clustered together while regulation subscale was differentially associated due to the
items being worded as action-oriented (Tonarely & Ehrenreich-May, 2019). In order to move

90
toward more pragmatic screening tools, the DTS subscales and identified cut off scores could be
used to distinguish at risk youth (as reported by Tonarely & Ehrenreich-May upon replication)
and assist in identifying candidates that may benefit from distress tolerance focused interventions
(e.g., mindfulness- and acceptance based therapies, dialectical behavioral therapy), ideally prior
to early substance use.

Mindfulness disposition: Previous work suggests mindfulness disposition, another stress
reactivity trait/skill, has multifaceted associations with substance use. For example, previous
research has reported that awareness or attention to mind-body was associated with greater
substance use (Leigh, 2009) while others facets, or the interaction thereof - observing non-
reactivity - was associated with lower substance use (Eisenlohr-Moul et al., 2012). Monitor and
Acceptance Theory (MAT; Lindsay & Creswell, 2017) followed in attempt to explain that there
may be two components attention monitoring and acceptance that are necessary for the
protective benefits from mindfulness.
Which mechanisms of mindfulness are protective against substance use frequency in
emerging adults? In study 3, I expand previous work by testing this theory in the largest and
most diverse sample to date. Findings at large did not support MAT other than the interaction
between acting with awareness non-reactivity on past 30-day alcohol use frequency. Trends
indicated among those with mean or higher levels of non-reactivity that higher levels of acting
with awareness associated with lower rates of past 30-day use for all substances tested (alcohol,
cigarettes, e-cigarettes, cannabis). Training skills of acting with awareness and non-reactivity
may assist in alcohol (and potentially other substance) use prevention during emerging
adulthood.

91
We identified complex FFMQ subscale associations in our sample of emerging adults.
FFMQ observing and non-reactivity were negatively associated with FFMQ non-judging and
acting with awareness, while FFMQ describing and sum score were positively associated with all
other FFMQ subscales. Additionally, FFMQ observing and non-reactivity were positively
associated with greater anxiety and depression symptoms, while all other subscales were
negatively associated with worse mental health symptoms. Previous literature has seen this
unexpected association for FFMQ observing (Baer, Smith, et al., 2006; Barnes & Lynn, 2010;
Karyadi et al., 2014), which called into question the scales integrity. A recent study using a
sample of minority adults in a residential substance abuse treatment facility also showed negative
associations between non-reactivity with acting with awareness and non-judging (Temme &
Wang, 2018). Therefore, a possible explanation of these negative associations could be
differences in participants interpretations and thus responses to FFMQ items. Additionally,
future research should consider the potentially critical differences between acting with awareness
and observing subscales. It is possible that acting with awareness simultaneously captures
intention or ability to delay gratification while observing does not assess any purposeful
awareness. It is likely that this is why current research is showing the protective effects of acting
with awareness but not the observing subscale. Given the longstanding psychometric debate in
the field regarding the validity of these measures across samples, context, and function (Bowen
& Enkema, 2014) and the newly developing theories of mechanisms driving mindfulness-related
protective effects (Lindsay & Creswell, 2017), a new measure with all of this in mind should be
developed.

92
Limitations
All three studies are limited in ability to make causal and/or directional inferences based on
survey-based data, which only provides evidence of association. Studies 1 and 2 are strengthened
by their repeated measures while this is needed for study 3. To our knowledge, the Happiness &
Health cohort is the most diverse sample to test associations between both distress tolerance and
mindfulness disposition with substance use in adolescents and emerging adults. This large-scale
study could potentially impact our findings to contradict the research literature due to issues of
generalizability. For example, the majority of the literature examining the association between
mindfulness disposition and substance use has been among majority white females (Karyadi et
al., 2014). Therefore, previously documented protective effects among that group may be washed
out in our sample which has nearly an even split of females and males and majority Hispanic
participants. While this was outside the scope of this dissertation and proposed study aims, future
research is needed to test these potential sociodemographic moderating effects in order to better
explain whether contradicting findings are due to differences in for whom and under what
conditions mindfulness disposition may be protective against substance use and other poor health
behaviors.
The number of theories, constructs, and measures used to capture constructs associated
with adolescent and emerging adult substance use continues to grow. Such complexity has
limited researchers ability to develop a clear and comprehensive picture regarding the causes and
prevention of early stage substance use (Petraitis, 1995). Specifically, some theories hold
incompatible underlying assumptions and it is challenging to integrate all constructs into one all-
encompassing theoretical explanation, particularly with partial (versus consistent) or
heterogenous (versus homogenous) findings across studies (Jonson, McArthur, Cullen, &

93
Wilcox, 2012; Petraitis, 1995). Furthermore, Jonson et al. recommended that future research on
adolescent substance use should select a limited number of the most salient variables from
multiple rival theories and research findings (Jonson et al., 2012). The current dissertation
examined individual factors of stress reactivity (Schulte & Hser, 2014) in association with
substance use frequency. In attempt to synthesize the protective subscales across distress
tolerance and mindfulness disposition measures in this dissertation the following section
condenses constructs that may be of particular importance in the prevention of substance use to
assist in the development of pragmatic screening tool.

Future Research: Integrating Subscales Driving Associations across the DTS and FFMQ-SF
Lower order subscales (from higher order DTS and FFMQ) that showed the strongest protective
associations against substance use frequencies in this dissertation were the DTS absorption, DTS
appraisal, FFMQ acting with awareness, and FFMQ non-reactivity (see original items in
Appendices 1 and 2). Specifically, in Study 1 supplemental analyses suggested among baseline
never-users greater DTS absorption and DTS appraisal drove the association of lower substance
use frequency in past 30-days across all available follow-up timepoints. In Study 3, findings
suggested greater FFMQ acting with awareness was associated with lower frequency of past 30-
day alcohol use among those with greater FFMQ non-reactivity. When examining these 15-items
and four subscale constructs for themes, it appears there are three key components that arise: (1)
being aware (with intentionality) in the present moment, (2) ability (vs. inability) to accept and
tolerate stressors, and (3) choosing an adaptive (vs. maladaptive) response.
First, being aware (with intentionality) in the present moment captured by the FFMQ
acting with awareness subscale. Example item (1=never or rarely true to 5=very often or always

94
true; reverse coded), I do jobs or tasks automatically without being aware of what Im doing.
Awareness with intentionality is specified for this theme following the discussion on the passive
language in the FFMQ observing which seems to be associated with greater health risk outcomes
versus active language in the FFMQ acting with awareness which generally speaking seems to be
associated with greater protective health outcomes. Second, the ability (vs. inability) to accept
and tolerate stressors is captured by the DTS appraisal. Example item (1=strongly agree to
strongly disagree = 5), My feelings of distress or being upset are not acceptable. Third,
choosing an adaptive (vs. maladaptive response) is captured by FFMQ non-reactivity and DTS
absorption. FFMQ non-reactivity example item (1=never or rarely true to 5=very often or always
true), When I have distressing thoughts or images I just notice them and let them go. DTS
absorption example item (1=strongly agree to strongly disagree = 5), When I feel distressed or
upset, I cannot help but concentrate on how bad the distress actually feels. These are both
tapping into the internal coping self-efficacy (vs. ruminative responses). Future research may
consider the integration of these items into a new measure addressing these dynamic emotion
regulation traits/skills for adaptive stress responses.
Distress tolerance and mindfulness disposition are both cognitive mechanisms that are
involved in stress reactivity and the multidimensional construct emotion regulation. Emotion
regulation includes awareness and acceptance of emotions as well as the ability to manage
behavior in response to emotions in order to achieve longer term goals (Gratz & Roemer, 2004).
The relation between and integration of these constructs is being explored in intervention and lab
studies (e.g., Carpenter et al., 2019). Some researchers have referred to the combination of
emotion regulation and mindfulness as mindful emotion regulation (Chambers, Gullone, &
Allen, 2009), yet much research is still needed to properly define and test this construct

95
(Guendelman, Medeiros, & Rampes, 2017). The current dissertation may help inform such
efforts in the development of a mindful emotion regulation – or more specifically a mindful
distress tolerance measure. To exemplify, I condensed the subscales DTS absorption, DTS
appraisal, FFMQ acting with awareness, and FFMQ non-reactivity into a single table (Table D1).

Table D1. Protective Subscales across the DTS and FFMQ-SF

Intentional Awareness
FFMQ-SF Acting with awareness
(R) I dont pay attention to what Im doing because Im daydreaming, worrying, or otherwise distracted.
I do jobs or tasks intentionally (vs. automatically) with awareness of what Im doing.
(R) I find myself doing things without paying attention.
Acceptance and Tolerance
DTS Appraisal
I can tolerate being distressed or upset as well as most people.
My feelings of distress or being upset are acceptable.
(R) Other people seem to be able to tolerate feeling distressed or upset better than I can.
Being distressed or upset is not a major ordeal for me.
(R) I am ashamed of myself when I feel distressed or upset.
(R) My feelings of distress or being upset scare me.
Adaptive Response
FFMQ-SF Non-reactivity
When I have distressing thoughts or images, I step back and am aware of the thought or image without
getting taken over by it.
When I have distressing thoughts or images I am able just to notice them without reacting.
When I have distressing thoughts or images I just notice them and let them go.
DTS Absorption
(R) When I feel distressed or upset, all I can think about is how bad I feel.
(R) My feelings of distress are so intense that they completely take over.
When I feel distressed or upset, I do not concentrate on how bad the distress feels.

Note. Items re-worded from the Distress Tolerance Scale (DTS; Simons & Gaher, 2005) and Five-Facet
Mindfulness Questionnaire-Short Form (FFMQ-SF; Baer et al., 2006) so that all items are positive and

96
scored on a scale of 1=never true to 5=always true so that higher scores indicate better skills. Items with a
(R) in front should be reverse scored such that a 5 becomes a 1 and a 1 becomes a 5.
Implications and Conclusions
Findings from the current dissertation suggest greater levels of specific mechanisms of distress
tolerance and mindfulness disposition may drive associations with lower substance use.
Screening, targeting, and monitoring these stress reactivity traits/skills may influence youths
substance use decisions and early prevention (Krank & Robinson, 2017). Testing the second
order factor structure using confirmatory and exploratory factor analyses may be a useful next
step in the mindful distress tolerance assessment tool. Following those findings, future
research may consider targeting the assessment and enhancement of these four skills and
overarching themes of (1) being aware in the present moment, (2) accepting and tolerating
stressors, and (3) choosing an adaptive response. All of these malleable traits are protective
character-building skills (Heckman & Kautz, 2013) that have the potential to be beneficial
beyond substance use prevention and may assist the adolescent in internal coping self-efficacy.
Furthermore, these skills/traits could be screened at the start of each academic year to identify at
risk youth whereby matching them to an appropriate intervention (i.e., mindfulness- and
acceptance-based programs) could be beneficial for adolescents to develop adaptive coping skills
in order to prevent youth from engaging in risky substance use and other poor health behaviors.
A better understanding of malleable stress reactivity traits/skills in relation to
maladaptive behaviors or behavioral determinants (alcohol use, smoking, diet, and physical
activity) of leading medical conditions (heart disease, cancer, stroke, diabetes) holds potential to
increase the effectiveness of behavioral interventions by precisely targeting mechanisms that
could enhance health outcomes thereby reducing the global burden of disease (Stein et al., 2019).
Given the transdiagnostic processes of the protective skills identified in this dissertation, an

97
integrative assessment tool has the potential to more precisely test associations between
protective stress reactivity traits/skills and transdiagnostic health outcomes to inform larger
public health prevention program initiatives.

98
Bibliography

Baer, R. A., Hopkins, J., Krietemeyer, J., Smith, G. T., & Toney, L. (2006). Using Self-Report
Assessment Methods to Explore Facets of Mindfulness. Assessment, 13(1), 27-45.
Baer, R. A., Smith, G. T., Hopkins, J., Krietemeyer, J., & Toney, L. (2006). Using self-report
assessment methods to explore facets of mindfulness. Assessment, 13(1), 27-45.
doi:10.1177/1073191105283504
Baer, R. A., Smith, G. T., Lykins, E., Button, D., Krietemeyer, J., Sauer, S., . . . Williams, J. M.
(2008). Construct validity of the five facet mindfulness questionnaire in meditating and
nonmeditating samples. Assessment, 15(3), 329-342. doi:10.1177/1073191107313003
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. doi:10.1037/0033-295X.111.1.33
Ballinger, G. A. (2004). Using Generalized Estimating Equations for Longitudinal Data
Analysis. Organizational Research Methods, 7(2), 127 150. .
doi:https://doi.org/10.1177/1094428104263672
Banerjee, N. (2014). Neurotransmitters in alcoholism: A review of neurobiological and genetic
studies. Indian J Hum Genet, 20(1), 20-31. doi:10.4103/0971-6866.132750
Barnes, S. M., & Lynn, S. J. (2010). Mindfulness skills and depressive symptoms: a longitudinal
study. IMAGINATION, COGNITION AND PERSONALITY, 30(1), 77-91.
Barrington-, J. B., & L, A. M. (2018). A P M E -
Cigarettes Urgent Concerns. New England Journal of Medicine, 379(12).

99
Berlin, K. S., Parra, G. R., & Williams, N. A. (2014). An introduction to latent variable mixture
modeling (part 2): longitudinal latent class growth analysis and growth mixture models. J
Pediatr Psychol, 39(2), 188-203. doi:10.1093/jpepsy/jst085
Black, D. S. (2012). Mindfulness and substance use intervention. Subst Use Misuse, 47(3), 199-
201. doi:10.3109/10826084.2011.635461
Black, D. S. (2014). Mindfulness-based interventions: an antidote to suffering in the context of
substance use, misuse, and addiction. Substance Use & Misuse, 49(5), 487-491.
doi:10.3109/10826084.2014.860749
Bowen, S., & Enkema, M. (2014). Relationship between dispositional mindfulness and substance
use: Findings from a clinical sample. Addictive behaviors, 39(3), 532 537.
doi:https://doi.org/10.1016/j.addbeh.2013.10.026
Brown, K. W., & Ryan, R. M. (2003). The benefits of being present: mindfulness and its role in
psychological well-being. J Pers Soc Psychol, 84(4), 822-848. Retrieved from
https://www.ncbi.nlm.nih.gov/pubmed/12703651
Carpenter, J. K., Sanford, J., & Hofmann, S. G. (2019). The Effect of a Brief Mindfulness
Training on Distress Tolerance and Stress Reactivity. Behav Ther, 50(3), 630-645.
doi:10.1016/j.beth.2018.10.003
Chambers, R., Gullone, E., & Allen, N. B. (2009). Mindful emotion regulation: An integrative
review. Clin Psychol Rev, 29(6), 560-572. doi:10.1016/j.cpr.2009.06.005
Chaplin, T. M., Hong, K., Fox, H. C., Siedlarz, K. M., Bergquist, K., & Sinha, R. (2010).
Behavioral arousal in response to stress and drug cue in alcohol and cocaine addicted
individuals versus healthy controls. Hum Psychopharmacol, 25(5), 368-376.
doi:10.1002/hup.1127

100
Chorpita, B. F., Yim, L., Moffitt, C., Umemoto, L. A., & Francis, S. E. (2000). Assessment of
symptoms of DSM-IV anxiety and depression in children: a revised child anxiety and
depression scale. Behav Res Ther, 38(8), 835-855. Retrieved from
https://www.ncbi.nlm.nih.gov/pubmed/10937431
Compas, B. E., Jaser, S. S., Dunbar, J. P., Watson, K. H., Bettis, A. H., Gruhn, M. A., &
Williams, E. K. (2014). Coping and Emotion Regulation from Childhood to Early
Adulthood: Points of Convergence and Divergence. Aust J Psychol, 66(2), 71-81.
doi:10.1111/ajpy.12043
Cortazar, N., & Calvete, E. (2019). Dispositional Mindfulness and Its Moderating Role in the
Predictive Association Between Stressors and Psychological Symptoms in Adolescents.
Mindfulness, 10(10), 2046 2059. doi:https://doi.org/10.1007/s12671-019-01175-x
Crane, R. S., Brewer, J., Feldman, C., Kabat-Zinn, J., Santorelli, S., Williams, J. M., & Kuyken,
W. (2017). What defines mindfulness-based programs? The warp and the weft. Psychol
Med, 47(6), 990-999. doi:10.1017/S0033291716003317
Creswell, J. D. (2017). Mindfulness Interventions. Annu Rev Psychol, 68, 491-516.
doi:10.1146/annurev-psych-042716-051139
Cummings, J. R., Bornovalova, M. A., Ojanen, T., Hunt, E., MacPherson, L., & Lejuez, C.
(2013). Time doesn't change everything: the longitudinal course of distress tolerance and
its relationship with externalizing and internalizing symptoms during early adolescence. J
Abnorm Child Psychol, 41(5), 735-748. doi:10.1007/s10802-012-9704-x
Dahl, R. E., Allen, N. B., Wilbrecht, L., & Suleiman, A. B. (2018). Importance of investing in
adolescence from a developmental science perspective. Nature, 554(7693), 441-450.
doi:10.1038/nature25770

101
Daughters, S. B., Reynolds, E. K., MacPherson, L., Kahler, C. W., Danielson, C. K., Zvolensky,
M., & Lejuez, C. W. (2009). Distress tolerance and early adolescent externalizing and
internalizing symptoms: the moderating role of gender and ethnicity. Behav Res Ther,
47(3), 198-205. doi:10.1016/j.brat.2008.12.001
Davis, J. P., Kendall, P. C., & Suveg, C. M. (2019). Emotional Awareness Predicts Specific
Cognitive-Behavioral Therapy Outcomes for Anxious Youth. Child Psychiatry Hum Dev.
doi:10.1007/s10578-018-00863-4
Eisenlohr-M, ., , E., C, R., L, D., & B, R. (2012).
H D M: I A S F -
Facet Mindfulness Questionnaire to Understand Its Relation To Substance Use.
Assessment, 19(3), 276 286. doi:https://doi.org/10.1177/1073191112446658
Evanovich, E. K., Marshall, A. J., David, S. J., & Mumma, G. H. (2019). A network
conceptualization of the multiple facets of distress tolerance. Anxiety Stress Coping,
32(6), 654-669. doi:10.1080/10615806.2019.1641799
Ewing, S. W., Sakhardande, A., & Blakemore, S. J. (2014). The effect of alcohol consumption
on the adolescent brain: A systematic review of MRI and fMRI studies of alcohol-using
youth. Neuroimage Clin, 5, 420-437. doi:10.1016/j.nicl.2014.06.011
Felton, J. W., Banducci, A. N., Shadur, J. M., Stadnik, R., MacPherson, L., & Lejuez, C. W.
(2017). The developmental trajectory of perceived stress mediates the relations between
distress tolerance and internalizing symptoms among youth. Dev Psychopathol, 29(4),
1391-1401. doi:10.1017/S0954579417000335

102
Felton, J. W., Collado, A., Havewala, M., Shadur, J. M., MacPherson, L., & Lejuez, C. W.
(2018). Distress Tolerance Interacts With Negative Life Events to Predict Depressive
Symptoms Across Adolescence. J Clin Child Adolesc Psychol, 1-10.
doi:10.1080/15374416.2017.1405354
Garland, Boettiger, & Howard. (2011). Targeting cognitive-affective risk mechanisms in stress-
precipitated alcohol dependence: an integrated, biopsychosocial model of automaticity,
allostasis, and addiction. Med Hypotheses, 76(5), 745-754.
doi:10.1016/j.mehy.2011.02.014
Garland, & Howard. (2018). Mindfulness-based treatment of addiction: current state of the field
and envisioning the next wave of research. Addiction Science & Clinical Practice, 13(1),
14. doi:10.1186/s13722-018-0115-3
Garland, E., Gaylord, S., & Park, J. (2009). The role of mindfulness in positive reappraisal.
Explore (NY), 5(1), 37-44. doi:10.1016/j.explore.2008.10.001
Glassman, L. H., Martin, L. M., Bradley, L. E., Ibrahim, A., Goldstein, S. P., Forman, E. M., &
Herbert, J. D. (2016). A Brief Report on the Assessment of Distress Tolerance: Are We
Measuring the Same Construct? J Rat-Emo Cognitive-Behav Ther, 34, 87-99.
doi:10.1007/s10942-015-0224-9
Gotts, J., Jordt, S., Mcconnell, R., & Tarran, R. (2019). What are the respiratory effects of e-
cigarettes? BMJ : British Medical Journal (Online), 366. doi:doi:10.1136/bmj.l5275
Gratz, K., & Roemer, L. (2004). Multidimensional Assessment of Emotion Regulation and
Dysregulation: Development, Factor Structure, and Initial Validation of the Difficulties in
Emotion Regulation Scale. Journal of Psychopathology and Behavioral Assessment,
26(1), 41 54. doi:https://doi.org/10.1023/B:JOBA.0000007455.08539.94

103
Greenberg, L., Martindale, S., Fils-Aimé, L., & Dolan, S. (2016). Distress Tolerance and
Impulsivity Are Associated With Drug and Alcohol Use Consequences in an Online
Community Sample. Journal of Cognitive Psychotherapy, 30(1), 50 59.
doi:https://doi.org/10.1891/0889-8391.30.1.50
Guendelman, S., Medeiros, S., & Rampes, H. (2017). Mindfulness and Emotion Regulation:
Insights from Neurobiological, Psychological, and Clinical Studies. Front Psychol, 8,
220. doi:10.3389/fpsyg.2017.00220
Heckman, J., & Kautz, T. (2013). Fostering and Measuring Skills: Interventions That Improve
Character and Cognition. IDEAS Working Paper Series from RePEc.
Jafari, M., & Ansari-Pour, N. (2019). Why, When and How to Adjust Your P Values? Cell J,
20(4), 604-607. doi:10.22074/cellj.2019.5992
Jonson, C., McArthur, R., Cullen, F., & Wilcox, P. (2012). Unravelling the Sources of
Adolescent Substance Use: A Test of Rival Theories. . International Journal on School
Disaffection, 9(2), 53 90. doi:https://doi.org/10.18546/IJSD.09.2.05
Jordan, C. J., & Andersen, S. L. (2017). Sensitive periods of substance abuse: Early risk for the
transition to dependence. Dev Cogn Neurosci, 25, 29-44. doi:10.1016/j.dcn.2016.10.004
Juarascio, A. S., Felton, J. W., Borges, A. M., Manasse, S. M., Murray, H. B., & Lejuez, C. W.
(2016). An investigation of negative affect, reactivity, and distress tolerance as predictors
of disordered eating attitudes across adolescence. J Adolesc, 49, 91-98.
doi:10.1016/j.adolescence.2016.02.005
Jung, T., & Wickrama, K. (2008). An Introduction to Latent Class Growth Analysis and Growth
Mixture Modeling. . Social and Personality Psychology Compass, 2(1), 302 317. doi:
https://doi.org/10.1111/j.1751-9004.2007.00054.x

104
Kabat-Zinn, J. (1982). An outpatient program in behavioral medicine for chronic pain patients
based on the practice of mindfulness meditation: theoretical considerations and
preliminary results. Gen Hosp Psychiatry, 4(1), 33-47. Retrieved from
https://www.ncbi.nlm.nih.gov/pubmed/7042457
Karyadi, K., Vanderveen, J., & Cyders, M. (2014). A meta-analysis of the relationship between
trait mindfulness and substance use behaviors. Drug and alcohol dependence, 143(1), 1-
10. doi:https://doi.org/10.1016/j.drugalcdep.2014.07.014
Kechter, A., Black, D. S., Riggs, N. R., Warren, C., Ritt-Olson, A., Chou, C. P., & Pentz, M. A.
(2019). Factors in the perceived stress scale differentially associate with mindfulness
disposition and executive function in early adolescents. Journal of Child and Family
Studies. doi:10.1007/s10826-018-01313-4
Kechter, A., & Leventhal, A. M. (2018). Longitudinal Association of Sleep Problems and
Distress Tolerance During Adolescence. Behav Med, 1-9.
doi:10.1080/08964289.2018.1514362
Kiken, L. G., Garland, E. L., Bluth, K., Palsson, O. S., & Gaylord, S. A. (2015). From a state to a
trait: Trajectories of state mindfulness in meditation during intervention predict changes
in trait mindfulness. Pers Individ Dif, 81, 41-46. doi:10.1016/j.paid.2014.12.044
Krank, M., & Robinson, J. (2017). Automatic Cognitive Processes and Youth Substance Use:
Risks and Prevention. Current Addiction Reports, 4(4), 386 396.
doi:https://doi.org/10.1007/s40429-017-0168-5

105
Layden, J., Ghinai, I., Pray, I., Kimball, A., Layer, M., Tenforde, M., & Meiman, J. (2019).
Pulmonary Illness Related to E-Cigarette Use in Illinois and Wisconsin - Preliminary
Report. The New England Journal of Medicine.
doi:https://doi.org/10.1056/NEJMoa1911614
Leigh, J., & Neighbors, C. (2009). ENHANCEMENT MOTIVES MEDIATE THE POSITIVE
ASSOCIATION BETWEEN MIND/BODY AWARENESS AND COLLEGE
STUDENT DRINKING. Journal of Social and Clinical Psychology, 28(5), 650-669.
doi:doi:http://dx.doi.org.libproxy2.usc.edu/10.1521/jscp.2009.28.5.650
Leventhal, A., Strong, D., Sussman, S., Kirkpatrick, M., Unger, J., Barrington-Trimis, J.;, &
Audrain-Mcgovern, J. (2016). Psychiatric comorbidity in adolescent electronic and
conventional cigarette use. . Journal of Psychiatric Research, 73, 71 78.
Leventhal, A., & Zvolensky, M. (2015). Anxiety, depression, and cigarette smoking: a
transdiagnostic vulnerability framework to understanding emotion-smoking comorbidity.
Psychol Bull, 141(1), 176-212. doi:10.1037/bul0000003
Leyro, T. M., Zvolensky, M. J., & Bernstein, A. (2010). Distress tolerance and
psychopathological symptoms and disorders: a review of the empirical literature among
adults. Psychol Bull, 136(4), 576-600. doi:10.1037/a0019712
Lindsay, E. K., & Creswell, J. D. (2017). Mechanisms of mindfulness training: Monitor and
Acceptance Theory (MAT). Clin Psychol Rev, 51, 48-59. doi:10.1016/j.cpr.2016.10.011
Liu, J., Chaplin, T. M., Wang, F., Sinha, R., Mayes, L. C., & Blumberg, H. P. (2012). Stress
reactivity and corticolimbic response to emotional faces in adolescents. J Am Acad Child
Adolesc Psychiatry, 51(3), 304-312. doi:10.1016/j.jaac.2011.12.014

106
Manning, K., Rogers, A., Bakhshaie, J., Hogan, J., Buckner, J., Ditre, J., & Zvolensky, M.
(2018). The association between perceived distress tolerance and cannabis use problems,
cannabis withdrawal symptoms, and self-efficacy for quitting cannabis: The explanatory
role of pain-related affective distress. Addictive behaviors, 85, 1-7.
doi:https://doi.org/10.1016/j.addbeh.2018.05.009
Miettunen, J., Murray, G. K., Jones, P. B., Maki, P., Ebeling, H., Taanila, A., . . . Moilanen, I.
(2014). Longitudinal associations between childhood and adulthood externalizing and
internalizing psychopathology and adolescent substance use. Psychol Med, 44(8), 1727-
1738. doi:10.1017/S0033291713002328
Miettunen, J., Murray, G. K., Jones, P. B., Maki, P., Ebeling, H., Taanila, A., & Moilanen, I.
(2014). Longitudinal associations between childhood and adulthood externalizing and
internalizing psychopathology and adolescent substance use. . Psychological Medicine,
44(8), 1727-1738.
doi:doi:http://dx.doi.org.libproxy1.usc.edu/10.1017/S0033291713002328
Muthén, L. K., & Muthén, B. O. (Eds.). (1998-2017). Mplus User’s Guide. Eighth Edition:
Muthén & Muthén.
Petraitis, J., Flay, B., Miller, T., & Petraitis, J. . (1995). Reviewing theories of adolescent
substance use: organizing pieces in the puzzle. . Psychological Bulletin, 117(1), 67 86.
doi:https://doi.org/10.1037/0033-2909.117.1.67
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.
doi:https://doi.org/10.1177/014662167700100306

107
Robinson, J. M., Ladd, B. O., & Anderson, K. G. (2014). When you see it, let it be: Urgency,
mindfulness and adolescent substance use. Addictive behaviors, 39(6), 1038-1041.
Rubin, D. B. (1987). Multiple Imputation for Nonresponse in Surveys. New York: John Wiley &
Sons Inc.
SAS Institute Inc. (2016). SAS/CONNEC 9.4 G, F E. C, NC: SAS
Institute Inc.
Sandler, I., Wolchik, S., Cruden, G., Mahrer, N., Ahn, S., Brincks, A., & Brown, C. . (2014).
Overview of Meta-Analyses of the Prevention of Mental Health, Substance Use, and
Conduct Problems. . Annual Review of Clinical Psychology, 10(1), 243 273. doi:
https://doi.org/10.1146/annurev-clinpsy-050212-185524
Sawyer, S. M., Azzopardi, P. S., Wickremarathne, D., & Patton, G. C. (2018). The age of
adolescence. Lancet Child Adolesc Health, 2(3), 223-228. doi:10.1016/S2352-
4642(18)30022-1
Schulte, M., & Hser, Y. (2014). Substance Use and Associated Health Conditions throughout the
Lifespan. Public health reviews, 35(2).
Simons, & Gaher. (2005). The Distress Tolerance Scale: Development and Validation of a Self-
Report Measure. Motivation and Emotion, 29(2), 83-102.
Simons, Sistad, R., Simons, J., & Hansen, J. (2018). The role of distress tolerance in the
relationship between cognitive schemas and alcohol problems among college students.
Addictive behaviors, 78.
Skara, S., & Sussman, S. (2003). A review of 25 long-term adolescent tobacco and other drug
use prevention program evaluations. Prev Med, 37(5), 451-474. Retrieved from
https://www.ncbi.nlm.nih.gov/pubmed/14572430

108
Skinner, E. A., & Zimmer-Gembeck, M. J. (2007). The development of coping. Annu Rev
Psychol, 58, 119-144. doi:10.1146/annurev.psych.58.110405.085705
Spitzer, R. L., Kroenke, K., Williams, J. B. W., & Lowe, B. (2006). A brief measure for
assessing generalized anxiety disorder. . Arch Inern Med. (166), 1092-1097.
Stein, A., Freeman, S., & Smits, J. (2019). Cognitive Mechanisms Underlying the Link Between
Anxiety, Health Behaviors, and Illness Outcomes: Commentary on a Special Issue.
Cognitive Therapy and Research,, 43(1), 131 138. doi:https://doi.org/10.1007/s10608-
019-09999-6
Temme, L., & Wang, D. (2018). Relationship Between the Five Facets of Mindfulness on Mood
and Substance Use Relapse. Families in Society: The Journal of Contemporary Social
Services, 99(3), 209 218. . doi:https://doi.org/10.1177/1044389418784961
Tomkins, D. M., & Sellers, E. M. (2001). Addiction and the brain: the role of neurotransmitters
in the cause and treatment of drug dependence. CMAJ, 164(6), 817-821. Retrieved from
https://www.ncbi.nlm.nih.gov/pubmed/11276551
Tomlinson, E., Yousaf, O., Vittersø, A., & Jones, L. (2017). Dispositional Mindfulness and
Psychological Health: a Systematic Review. Mindfulness, 9(1), 23-43.
doi:https://doi.org/10.1007/s12671-017-0762-6
Tonarely, N. A., & Ehrenreich-May, J. (2019). Confirming the Factor Structure and Validity of
the Distress Tolerance Scale (DTS) in Youth. Child Psychiatry Hum Dev.
doi:10.1007/s10578-019-00935-z
Trafton, J. A., & Gifford, E. V. . (2011). Biological bases of distress tolerance. In A. B. M. J.
Zvolensky, & A. A. Vujanovic (Ed.), Distress tolerance: Theory, research, and clinical
applications (pp. 80-102). New York, NY: US: Guilford Press.

109
Veilleux, J. C. (2019). The relationship between distress tolerance and cigarette smoking: A
systematic review and synthesi. Clinical Psychology Review.
doi:https://doi.org/10.1016/j.cpr.2019.01.003
Warren, C., Kechter, A., Christodoulou, G., Cappelli, C., & Pentz, M. (2020). Psychosocial
behaviors, obesity risk, and substance use risk in early adolescents. Journal of Behavioral
Medicine.
Whyte, A. J., Torregrossa, M. M., Barker, J. M., & Gourley, S. L. (2018). Editorial: Long-Term
Consequences of Adolescent Drug Use: Evidence From Pre-clinical and Clinical Models.
. Frontiers in behavioral neuroscience, 12(83). doi:doi:10.3389/fnbeh.2018.00083
Winters, K. C., & Arria, A. (2011). Adolescent Brain Development and Drugs. Prev Res, 18(2),
21-24. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/22822298
Winward, J. L., Bekman, N. M., Hanson, K. L., Lejuez, C. W., & Brown, S. A. (2014). Changes
in emotional reactivity and distress tolerance among heavy drinking adolescents during
sustained abstinence. Alcohol Clin Exp Res, 38(6), 1761-1769. doi:10.1111/acer.12415
Zubrick, S. R., Lawrence, D., Mitrou, F., Christensen, D., & Taylor, C. L. . (2012). Early mental
health morbidity and later smoking at age 17 years. . Psychological Medicine, 42(5),
1103-1115. doi:doi:http://dx.doi.org.libproxy1.usc.edu/10.1017/S0033291711002182

110
Appendix

Appendix 1. Distress Tolerance Scale

Tolerance
1. Feeling distressed or upset is unbearable to me.

3. I cant handle feeling distressed or upset.

5. Theres nothing worse than feeling distressed or upset.

Absorption
2. When I feel distressed or upset, all I can think about is how bad I feel.

4. My feelings of distress are so intense that they completely take over.

15. When I feel distressed or upset, I cannot help but concentrate on how bad the distress actually feels.

Appraisal
6R. I can tolerate being distressed or upset as well as most people.

7. My feelings of distress or being upset are not acceptable.

9. Other people seem to be able to tolerate feeling distressed or upset better than I can.

10. Being distressed or upset is always a major ordeal for me.

11. I am ashamed of myself when I feel distressed or upset.

12. My feelings of distress or being upset scare me.

Regulation
8. Ill do anything to avoid feeling distressed or upset.

13. Ill do anything to stop feeling distressed or upset.

14. When I feel distressed or upset, I must do something about it immediately.

111
Appendix 2. Five-Facet Mindfulness Questionnaire (Short Form)

Observing items: 1, 6, 11.

1. When I take a shower or a bath, I stay alert to the sensations of water on my body.

6. I notice how foods and drinks affect my thoughts, bodily sensations, and emotions.

11. I pay attention to sensations, such as the wind in my hair or sun on my face.

Describe items: 2, 7R, 12.

2. Im good at finding words to describe my feelings.

7. I have trouble thinking of the right words to express how I feel about things.

12. Even when Im feeling terribly upset I can find a way to put it into words.

Acting with awareness items: 3R, 8R, 13R.

3. I dont pay attention to what Im doing because Im daydreaming, worrying, or otherwise distracted.

8. I do jobs or tasks automatically without being aware of what Im doing.

13. I find myself doing things without paying attention.

Non-judging items: 4R, 9R, 14R.

4. I believe some of my thoughts are abnormal or bad and I shouldnt think that way.

9. I think some of my emotions are bad or inappropriate and I shouldnt feel them.

14. I tell myself I shouldnt be feelin g the way Im feeling.

Non-reactivity items: 5, 10, 15.

5. When I have distressing thoughts or images, I step back and am aware of the thought or image
without getting taken over by it.

10. When I have distressing thoughts or images I am able just to notice them without reacting.

15. When I have distressing thoughts or images I just notice them and let them go.

112
Appendix 3. Average Past 30-Day Substance Use by Distress Tolerance Class (Corresponds to
Appendix 4-7)

Note. Past 30-day use scale transformed from 0-8 to 0-30 count (0, 2, 4, 8, 12, 18, 22, 27, 30). Alcohol,
cannabis, and cigarette data were collected at timepoints 1-8; e-cigarette data were collected at timepoints
3-8. P-value from ANOVA tests.

DT Class 1
47.6% (n=1034),
Moderate Stable

DT Class 2
10.1% (n=205),
Poor Declining

DT Class 3
42.3% (n=924),
High Increasing

Total Difference
(p)
M(SD)
Alcohol
W1 0.34 (1.67) 0.72 (3.50) 0.27 (1.67) 0.39 (0.71) .03
W2 0.75 (2.88) 0.92 (3.37) 0.58 (2.20) 0.69 (2.66) .15
W3 0.79 (2.66) 0.87 (3.28) 0.69 (2.46) 0.76 (2.63) .58
W4 0.78 (2.50) 0.79 (2.27) 0.76 (2.83) 0.77 (2.67) .99
W5 0.78 (2.58) 0.91 (3.34) 0.80 (3.11) 0.80 (2.76) .84
W6 0.74 (2.12) 0.92 (2.92) 0.71 (2.15) 0.75 (2.22) .46
W7 0.91 (2.15) 1.31 (3.72) 0.77 (2.17) 0.89 (2.36) .01
W8 1.02 (2.11) 1.56 (3.66) 0.93 (2.42) 1.04 (2.42) <.01
Cigarette
W1 0.08 (1.01) 0.50 (3.71) 0.04 (0.38) 0.11 (1.36) <.001
W2 0.07 (0.61) 0.20 (2.12) 0.13 (1.45) 0.11 (1.23) .36
W3 0.12 (1.19) 0.36 (2.74) 0.19 (1.77) 0.17 (1.65) .14
W4 0.19 (1.73) 0.21 (1.64) 0.27 (2.39) 0.23 (2.04) .69
W5 0.12 (1.16) 0.40 (3.10) 0.31 (2.81) 0.23 (2.22) .09
W6 0.19 (1.77) 0.33 (2.64) 0.25 (2.32) 0.23 (2.11) .62
W7 0.15 (3.14) 0.42 (3.14) 0.25 (2.25) 0.22 (1.99) .16
W8 0.26 (1.87) 0.39 (2.59) 0.40 (2.73) 0.33 (2.34) .41
E-Cigarette
W3 0.34 (1.80) 0.99 (3.91) 0.43 (2.43) 0.44 (2.35) <.01
W4 0.41 (2.34) 0.52 (2.53) 0.41 (2.47) 0.42 (2.41) .84
W5 0.32 (2.16) 0.53 (2.80) 0.32 (2.46) 0.34 (2.56) .49
W6 0.47 (2.93) 0.55 (3.23) 0.33 (2.34) 0.42 (2.73) .41
W7 0.29 (2.28) 0.46 (3.12) 0.28 (2.52) 0.30 (2.48) .63
W8 0.53 (3.12) 0.72 (3.90) 0.58 (3.33) 0.57 (3.29) .75
Cannabis
W1 0.26 (1.87) 0.80 (4.04) 0.30 (2.13) 0.33 (2.28) <.01
W2 0.44 (2.82) 0.68 (3.04) 0.50 (2.78) 0.49 (2.82) .55
W3 0.64 (3.26) 0.82 (3.39) 0.65 (3.28) 0.66 (3.28) .78
W4 0.78 (3.50) 0.87 (3.70) 0.81 (3.84) 0.80 (3.57) .95
W5 0.70 (3.42) 0.96 (3.81) 0.76 (3.64) 0.75 (3.55) .65
W6 1.08 (4.14) 1.02 (3.86) 1.08 (4.12) 1.07 (4.10) .98
W7 1.04 (4.10) 1.31 (4.63) 1.27 (4.86) 1.16 (4.49) .49
W8 1.41 (4.68) 1.89 (5.54) 1.75 (5.55) 1.60 (5.15) .25

113
Appendix 4. Alcohol Use by Distress Tolerance Class Across High School (Corresponds to Supplemental Table B1)

Note. Past 30-day use scale transformed from 0-8 to 0-30 count (0, 2, 4, 8, 12, 18, 22, 27, 30). Alcohol use data were collected at timepoints 1-8.
Distress tolerance classes were identified from latent growth mixture modeling (LGMM; refer to Figure B3).

114

Appendix 5. Cigarette Use by Distress Tolerance Class Across High School (Corresponds to Supplemental Table B1)

Note. Past 30-day use scale transformed from 0-8 to 0-30 count (0, 2, 4, 8, 12, 18, 22, 27, 30). Cigarette use data were collected at timepoints 1-8.
Distress tolerance classes were identified from latent growth mixture modeling (LGMM; refer to Figure B3).

115
Appendix 6. E-Cigarette Use by Distress Tolerance Class Across High School (Corresponds to Supplemental Table B1)

Note. Past 30-day use scale transformed from 0-8 to 0-30 count (0, 2, 4, 8, 12, 18, 22, 27, 30). E-cigarette use data were collected at timepoints 3-
8. Distress tolerance classes were identified from latent growth mixture modeling (LGMM; refer to Figure B3).

116
Appendix 7. Cannabis Use by Distress Tolerance Class Across High School (Corresponds to Supplemental Table B1)

Note. Past 30-day use scale transformed from 0-8 to 0-30 count (0, 2, 4, 8, 12, 18, 22, 27, 30). Cannabis (smoking) use data were collected at
timepoints 1-8. Distress tolerance classes were identified from latent growth mixture modeling (LGMM; refer to Figure B3

Abstract (if available)

Abstract The goal of the current dissertation was to deepen the scientific understanding of adolescent substance use etiology, particularly in regard to malleable individual traits/skills related to stress reactivity (i.e., distress tolerance and mindfulness disposition). In a large, diverse, cohort of adolescents and emerging adults from the Happiness & Health Study, the specific research objectives were to: (1) examine the association of distress tolerance with subsequent substance use throughout high school

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Sociocultural stress, coping and substance use among Hispanic/Latino adolescents

PDF

The dynamic relationship of emerging adulthood and substance use

PDF

The role of social support in the relationship between adverse childhood experiences and addictive behaviors across adolescence and young adulthood

PDF

Effects of flavorings in electronic cigarettes on the use and appeal of e-cigarettes among youth and adults

PDF

Social self-control and adolescent substance use

PDF

The Internet activities, gratifications, and health consequences related to compulsive Internet use

PDF

The effects of mindfulness on adolescent cigarette smoking: Measurement, mechanisms, and theory

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

Opioid withdrawal symptoms, opioid use, and injection risk behaviors among people who inject drugs (PWID)

PDF

Adolescent cigarette use and psychological distress: relation to adult obesity risk

PDF

Understanding the methodological limitations In the ecological momentary assessment of physical activity

PDF

The impact of childhood trauma on substance use and mental health during the SARS-CoV-2 pandemic among young adults

PDF

Understanding the dynamic relationships between physical activity and affective states using real-time data capture techniques

PDF

The acute and longitudinal associations between sedentary behaviors, affective states, and emotional disorder symptoms among youth

PDF

Contextualizing social network influences on substance use among high risk adolescents

PDF

The influence of contextual factors on the processes of adoption and implementation of evidence-based substance use prevention and tobacco cessation programs in schools

PDF

Motivational interviewing with adolescent substance users: a closer look

PDF

Diffusion of a peer‐led suicide preventive intervention in secondary schools: strategies to increase effectiveness of peer‐led interventions

PDF

Pathways of drug use among people who inject drugs

Asset Metadata

Creator Kechter, Afton Victoria (author)

Core Title Distress tolerance and mindfulness disposition: associations with substance use during adolescence and emerging adulthood

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Preventive Medicine (Health Behavior Research)

Publication Date 05/07/2020

Defense Date 02/11/2020

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag adolescence,distress tolerance,emerging adulthood,mindfulness disposition,OAI-PMH Harvest,stress reactivity,substance use

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Leventhal, Adam M. (committee chair), Barrington-Trimis, Jessica L. (committee member), Black, David S. (committee member), Davis, Jordan P. (committee member), Huh, Jimi (committee member)

Creator Email aftonkechter@gmail.com,kechter@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-303867

Unique identifier UC11664345

Identifier etd-KechterAft-8458.pdf (filename),usctheses-c89-303867 (legacy record id)

Legacy Identifier etd-KechterAft-8458.pdf

Dmrecord 303867

Document Type Dissertation

Rights Kechter, Afton Victoria

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 a...

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