University of Southern California Dissertations and Theses

Using a structural model of psychopathology to distinguish relations between shared and specific features of psychopathology, smoking, and underlying mechanisms

(USC Thesis Other)

Using a structural model of psychopathology to distinguish relations between shared and specific features of psychopathology, smoking, and underlying mechanisms

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

USING A STRUCTURAL MODEL OF PSYCHOPATHOLOGY TO DISTINGUISH
RELATIONS BETWEEN SHARED AND SPECIFIC FEATURES OF
PSYCHOPATHOLOGY, SMOKING, AND UNDERLYING MECHANISMS

By
Katherine J. Ameringer

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

December 2013

ii

Acknowledgements
I am very appreciative to the many individuals who have supported me and my work over
these last several years. First, I would like to thank my advisor, Dr. Adam Leventhal, for your
academic guidance, mentorship, and encouragement. You have provided me with the foundation
and opportunity to achieve a lot in the field. Thank you also to the other members of my
committee, Dr. Steve Sussman, Dr. Jennifer Unger, Dr. Genevieve Dunton, and Dr. Chih-Ping
Chou, for your time and advice during this process, especially Dr. Chou for always lending a
helping hand with statistics despite your heavy involvement in many other research activities. A
sincere thank you to Marny Barovich for your extremely quick help, at all times, on anything and
everything administrative related. You play a key role for all graduate students in this program.
Thank you to all the members of the HEAL laboratory, particularly Michael Trujillo, for working
so hard to carry out multiple studies, and most importantly, for making the days at lab fun.
On a more personal note, thank you to my parents and my sister for their love, support,
and encouragement. I am blessed to have such a wonderful family and incredible role-models in
my life. Thank you to Watson and Luna, your unconditional love and constant companionship
helped me to get through the long days writing at home. Last, a special thank you to Phil. You
are an amazing partner, and I am so lucky to have you in my life. Your optimism and belief in
me have helped me to succeed through these years. I am so excited to start the next part of our
lives together.

iii

Table of Contents
Acknowledgements ii
List of Tables iv
List of Figures v
Abstract vi
Chapter 1: Introduction
Specific Aims
Background and Significance
Overview of the Dissertation
1
1
4
16
Chapter 2: Contrasting Structural Models of Psychopathology to Elucidate Relations
between Shared and Specific Features of Psychopathology and Smoking
Abstract
Introduction
Methods
Results
Discussion

19
19
20
25
35
38
Chapter 3: Shared versus Specific Features of Psychopathology and Smoking Level:
Structural Relations and Mediation by Positive and Negative
Reinforcement Mechanisms
Abstract
Introduction
Methods
Results
Discussion

55
55
56
62
66
69
Chapter 4: Psychopathology, Motivation to Resume Smoking, and the Mediating
Effects of Nicotine Withdrawal Symptoms
Abstract
Introduction
Methods
Results
Discussion

81
81
82
88
96
99
Chapter 5: Conclusion 110
Bibliography 117
iv

List of Tables
Table 1: Select Demographic and Smoking Characteristics 47
Table 2: Descriptive Statistics of Manifest Psychopathological Symptom
Indices Included in the Structural Models

48
Table 3: Correlations between Manifest Psychopathological Symptom Indices
Included in the Structural Models

49
Table 4: Fit Indices for Comparative Structural Models of the Set of Manifest
Psychopathological Symptom Indices

50
Table 5: Associations between Manifest Psychopathological Symptom
Indices, Latent Factors, and Smoking Characteristics

51
Table 6: Descriptive Statistics and Correlations of Manifest Psychopathology
Indices, Reinforcement Smoking Scales, and Smoking Level

76
Table 7: Direct, Indirect, and Total Effect Results from the Negative and
Positive Reinforcement Models

77
Table 8: Descriptive Statistics and Correlations of Manifest Psychopathology
Indices, Withdrawal Symptoms, and Time to First Cigarette

105
Table 9: Direct, Indirect, and Total Effects from the Withdrawal Mediation
Model

106

v

List of Figures
Figure 1: Hypothesized Latent Factor Models of the Set of Psychopathological
Symptom Indices
52
Figure 2: Best-Fitting Latent Factor Model in Both Samples 53
Figure 3: Second-Order Factor Model in Both Samples 54
Figure 4: Final Model of Latent Factors and Residuals Predicting Smoking
Level

78
Figure 5: Psychopathology, Negative and Positive Reinforcement Smoking,
and Smoking Level Mediation Models

79
Figure 6: First-Order Three-Factor and Single Second-Order Factor Models of
Psychopathology

107
Figure 7: Final Model of Latent Factors and Residuals Predicting Time to First
Cigarette

108
Figure 8: Psychopathology, Withdrawal, and Time to First Cigarette Mediation
Model

109

vi

Abstract
Extensive psychopathological comorbidity makes it difficult to interpret the nature and
underlying mechanisms of psychopathology-smoking relations. Considerable research in the
psychiatric literature illustrates that latent shared dimensions of psychopathology may best
explain psychopathological comorbidity. Incorporating this conceptualization of
psychopathology into psychopathology-smoking research may represent one promising way to
refine the psychopathology-smoking link. This dissertation consists of three independent studies
testing this novel approach. The first study establishes a meaningful latent factor model for a
wide variety of manifest emotional and behavioral symptom- and syndrome-specific
psychopathological indicators commonly associated with smoking (happiness, anhedonia,
depression, anxiety, anxious arousal, ADHD symptoms, physical aggression, and alcohol use
problems). The second study uses this model to test the extent to which shared features of
psychopathology (latent dimensions) versus specific features of psychopathology (residuals of
the manifest indicators) associate with smoking level, as well as the underlying role of positive
and negative reinforcement motivations to smoke in these relations. Relatedly, the third study
uses the structural model to test whether shared or specific features of psychopathology associate
with ability to delay smoking in exchange for monetary reinforcement and whether withdrawal
symptom severity mediates these relations. Taken together, the findings from this dissertation (1)
emphasize the importance of accounting for psychopathological comorbidity when investigating
specific psychopathology-smoking relations and (2) suggest that focusing on core
psychopathological liabilities that associate with smoking but that are shared among many
different forms of manifest psychopathology warrants further consideration.

1

Chapter 1
Introduction
Specific Aims
Although considerable research has documented a link between psychopathology and
smoking, much of this research has not accounted for and examined the influence of
psychopathological comorbidity. This creates issues for interpreting many of the previously
established relations between psychopathology and smoking. Specifically, it remains unclear
whether associations evidenced with smoking were due to features shared across many forms of
psychopathology or to features specific to particular forms of psychopathology. Clarifying these
relations is important for increasing understanding of the link between psychopathology and
smoking and for developing more effective smoking prevention and cessation interventions for
individuals with psychopathology. Therefore, the overall aim of this dissertation was to examine
how a liability-spectrum conceptualization of psychopathology (Krueger & Markon, 2006) can
address the challenges psychopathological comorbidity creates for interpreting relations between
psychopathology and smoking. Three related studies are conducted to address this overall aim.
The first study establishes a latent factor model of a set of psychopathological symptom indices
commonly associated with smoking. The second two studies incorporate this model into
analyses between psychopathology and smoking. The specific aims for each of these studies are:
Study I
1. Use confirmatory factor analysis (CFA) to establish a parsimonious and meaningful latent
factor model of psychopathology relevant for psychopathology-smoking relations:
a. Test three alternate models: a 1-factor (general maladjustment) model, a 2-factor
(internalizing-externalizing) model, and a 3-factor (low positive affect-negative
2

affect-disinhibition) model, for a range of emotional and behavioral
psychopathological indicators commonly associated with smoking (happiness,
anhedonia, depression, anxiety, anxious arousal, physical aggression, ADHD
symptoms, alcohol use problems);
b. Test these models in two diverse samples: college students and adult daily smokers.
2. Investigate associations between each of the manifest psychopathology indicators and the
latent factor scores derived from the structural models with several smoking variables
including:
a. Smoking experimentation and smoking establishment in the college student sample;
b. Age of smoking onset, smoking heaviness, nicotine dependence, and retrospective
withdrawal symptom severity in the adult daily smoker sample.
Study II
3. With structural equation modeling (SEM) procedures, use the latent factor model of
psychopathology identified in Study I to test the extent to which shared versus specific
features of psychopathology associate with cigarettes per day (CPD):
a. Include regression paths from each of the latent factors to CPD to test shared relations.
b. Subsequently, include regression paths from each of the residuals of the
psychopathological indicators (the portion of variance within the indicator that does
not covary with the latent factor) to CPD, one at a time, to examine specific relations.
4. Investigate whether motivations to smoke for positive and negative reinforcement purposes
mediate any shared or specific links found between psychopathology and CPD.

3

Study III
5. With SEM procedures, use the latent factor model of psychopathology established in Study I
to test whether shared or specific features of psychopathology associate with ability to delay
smoking in exchange for monetary reinforcement:
a. To examine shared relations, investigate associations between the latent factors and
ability to delay smoking.
b. To examine specific relations, investigate associations between the residuals of each of
the manifest psychopathological indicators and ability to delay smoking.
6. Test whether overall withdrawal symptom severity mediates any associations found between
shared or specific features of psychopathology and ability to delay smoking.
7. Investigate whether abstinence status (within-subject: after 16 hours abstinence and
nonabstinence) moderates any psychopathology, withdrawal, and delay relations.

4

Background and Significance
Impact of Cigarette Smoking in the United States
Since the first Surgeon General’s report on cigarette smoking and health in 1964,
society’s view of cigarette smoking has turned considerably negative and the prevalence of
cigarette smoking has substantially declined (U.S. Department of Health and Human Services
[USDHSS], 1989). These significant changes were likely due to a combination of several factors,
including increased knowledge about the health consequences of smoking, anti-smoking public
health campaigns, anti-smoking legislation, and the development of new behavioral and
pharmacological smoking cessation interventions. Despite these numerous and intense efforts,
smoking prevalence (Centers for Disease Control and Prevention [CDC], 2011; Warner & Burns,
2003) and abstinence rates (Irvin & Brandon, 2000) have evidenced a slow decline since the
mid-1990s, and smoking continues to produce staggering negative consequences.
Approximately 19.3% of adults (an estimated 45.3 million people) in the U.S. smoke today
(CDC, 2011), cigarette smoking remains the leading cause of preventable death in the U.S.
(CDC, 2002) responsible for approximately 443,000 deaths every year (CDC, 2004, 2008), costs
the U.S. billions per year in health-related costs (i.e., direct medical costs and lost productivity),
and results in 5.1 million years of potential life lost annually (CDC, 2008). As such, continued
efforts to reduce smoking prevalence remain of significant public health importance.
The slowing decline of smoking prevalence and abstinence rates, despite the increasing
social pressure against smoking and the well-documented health consequences of smoking, has
led to the speculation that today’s smokers may have certain individual differences that put them
at risk for heavy and persistent smoking, perhaps because they obtain specific benefits from
smoking, are more likely to maintain smoking, and/or because they have greater difficulty
5

quitting (Emery, Gilpin, Ake, Farkas, & Pierce, 2000; Irvin & Brandon, 2000). One group of
individual difference characteristics that have received considerable attention in the literature as
a risk factor for smoking is psychopathology.

Psychopathology and Cigarette Smoking
The relationship between any one of a number of common psychopathological disorders
(e.g., depressive, anxiety, disruptive behavior, substance use disorders) and different aspects of
the tobacco dependence process (e.g., initiation, progression to heavy smoking, nicotine
dependence, patterns of withdrawal, and difficulties quitting) has been demonstrated across
numerous studies (e.g., Breslau, Novak, & Kessler, 2004; Degenhardt & Hall, 2001; Grant,
Hasin, Chou, Stinson, & Dawson, 2004; Lasser et al., 2000; Rohde, Kahler, Lewinsohn, &
Brown, 2004; Rohde, Lewinsohn, Brown, Gau, & Kahler, 2003). These associations are not only
present at full-blown clinical levels of these disorders but are also evident for many continuous
measures of psychopathological symptom dimensions associated with these disorders(e.g.,
Cinciripini et al., 1995; Fuemmeler, Kollins, & McClernon, 2007; Kollins, McClernon, &
Fuemmeler, 2005; Lipkus, Barefoot, Williams, & Siegler, 1994; McLeish, Zvolensky, Yartz, &
Leyro, 2008; Zvolensky, Stewart, Vujanovic, Gavric, & Steeves, 2009). In fact, psychopathology
and smoking relations have been found at very low levels of psychopathological symptoms (e.g.,
>1 ADHD symptom, Elkins, McGue, & Iacono, 2007; >2 depressive symptoms, Niaura et al.,
2001) and in individuals that do not have the respective current disorder (Heffner, Johnson,
Blom, & Anthenelli, 2010; Leventhal, Ramsey, Brown, LaChance, & Kahler, 2008). As such, it
appears that the entire spectrum of psychopathological disturbance, not necessarily only clinical
levels, may be important for smoking relations. This is critical as it extends the relevance of the
6

psychopathology-smoking relation to a wide range of the general public and increases the public
health importance of offsetting the link between psychopathology and smoking.
The psychopathology-smoking relationship is likely driven by a complex combination of
mechanisms. These include: 1) common factors (e.g., familial, genetic, environmental) that
predispose individuals to both psychopathology and smoking, 2) self-medication processes
(underlying psychopathology reinforces future smoking), 3) selective quitting (i.e.,
psychopathology is more prevalent among the remaining group of smokers because
psychopathology associates with higher relapse and fewer cessation attempts,), 4) smoking
exacerbates psychopathology, and 5) common neurobiological alterations (nicotine modulates
several neurotransmitters involved in both psychopathology and smoking) (Gilbert & Gilbert,
1995). Regarding hypotheses 2 and 4, research suggests a reciprocal and dynamic relationship
between psychopathology and smoking, with evidence that both may be risk factors for each
other (Breslau, Peterson, Schultz, Chilcoat, & Andreski, 1998; Dierker, Avenevoli, Merikangas,
Flaherty, & Stolar, 2001; Kandel, Huang, & Davies, 2001; Morisano, Bacher, Audrain-
McGovern, & George, 2009). Despite the complexity of the relationship between smoking and
psychopathology, there is sufficient evidence to indicate that psychopathology plays an
important role in the onset of smoking for many individuals (Breslau, Kilbey, & Andreski, 1993;
Orlando, Ellickson, & Jinnett, 2001; Patton et al., 1998; Rohde et al., 2004), emphasizing the
importance of understanding the link from psychopathology to smoking vulnerability.
In lieu of the consistently documented relations between psychopathology and smoking,
it has been widely acknowledged that smoking cessation efforts must account for comorbid
psychopathology. Specialized smoking cessation interventions meant to target a range of
different forms of psychopathology have been investigated, such as individualized
7

pharmacotherapy, cognitive behavioral therapy, and mood management strategies (El-Guebaly,
Cathcart, Currie, Brown, & Gloster, 2002; Ranney, Melvin, Lux, McClain, & Lohr, 2006). While
these types of more intensive, individualized cessation interventions hold promise for individuals
with comorbid psychopathology, these techniques have demonstrated limited effectiveness
(Brown et al., 2007; Winhusen et al., 2010) and face considerable challenges. One main barrier
is the cost and time associated with developing, testing, and disseminating a multitude of
different smoking cessation interventions targeted to different forms of psychopathology. A
second key barrier is the lack of understanding of how psychopathological comorbidity
influences relations with smoking. Because a large portion of the psychopathology-smoking
literature has not accounted for the extensive issue of psychopathological comorbidity, it is
unclear whether relations found across many different types of psychopathology and smoking
were due to features specific to the type of psychopathology being studied or to features shared
among many different forms of psychopathology. Without this knowledge, it is difficult to
determine which psychopathological features should be targeted in individualized cessation
interventions. Furthermore, because research indicates that “pure” cases of psychopathology
(i.e., no other forms of comorbid psychopathology present) are rare and are likely not
representative of the majority of individuals with psychopathology (Krueger, Caspi, Moffitt, &
Silva, 1998) the generalizability of smoking cessation interventions targeted to a specific “pure”
form of psychopathology would be greatly limited.
Based on these challenges, it may be helpful to take a more macroscopic and
transdiagnostic approach to the psychopathology-smoking relation by examining and targeting
psychopathological features that associate with smoking but that are shared among many
different forms of psychopathology. To examine the potential utility of this more broad-based
8

approach to psychopathology and smoking, it is first necessary to identify the most parsimonious
and meaningful underlying dimensions that represent the features shared across many forms of
psychopathology. Second, it is important to examine the extent to which these shared features
can account for the relationship with smoking. Third, mechanisms underlying the link between
psychopathology and smoking must be identified to better understand processes that maintain
this link and to identify modifiable factors that can be targeted in treatment to disrupt this link.

Identifying the Shared Features of Psychopathology
Comorbidity, referred to in this dissertation as two psychological disorders occurring in
the same person more often than expected by chance (Krueger & Markon, 2006) is markedly
common among individuals presenting with psychiatric disorders (Kessler, Chiu, Demler,
Merikangas, & Walters, 2005) and is a major issue in adult psychopathology(Clark, Watson, &
Reynolds, 1995). In the psychiatric literature, recognition of the widespread and problematic
issue of comorbidity, along with modern statistical techniques, has led to extensive research on
how to better account for and model patterns of comorbidity. Many of these studies conclude that
comorbidity among several common forms of psychopathology may best be understood in terms
of broad higher-order dimensions, such that specific forms of psychopathology reflect different
manifestations of underlying core psychopathological vulnerabilities (Krueger & Markon, 2006).
For example, a significant body of research has found evidence for a two factor internalizing-
externalizing model, wherein many common forms of psychopathology are best modeled by two
subordinate but correlated dimensions: internalizing (psychopathological maladjustment is
primarily expressed inward; unipolar mood and anxiety disorders) and externalizing
9

(psychopathological maladjustment is primarily expressed outward; disruptive behavior and
substance use disorders) (Krueger, 1999; Krueger et al., 1998; Slade & Watson, 2006).
Based on the increasing amount of evidence demonstrating the presence of broad higher-
order dimensions, modeling different forms of psychopathology based on their common
substrates may represent one promising way to address the challenges of psychopathological
comorbidity in relation to smoking. However, despite the widely documented issue of
comorbidity within and across many psychopathological disorders and symptoms and the
considerable attention and support for structural model-based approaches to psychopathology,
this conceptualization of psychopathology has largely not been incorporated into the smoking
literature. This hinders interpretation of the psychopathology-smoking relationship because
without accounting for the shared dimensions of psychopathology, it remains unknown whether
associations between psychopathology and smoking are due to shared features of
psychopathology and/or to specific features of particular forms of psychopathology.

Examining Relations between Shared versus Specific Features of Psychopathology and
Cigarette Smoking
Structural equation modeling provides a useful and appropriate analytic technique to
flush out the extent to which relations with smoking are due to shared versus specific features of
psychopathology. Specifically, this analytic technique can: 1) identify and model the underlying
latent dimensions that account for the covariance among a set of different psychopathological
syndromes, 2) use these latent dimensions to examine relations between shared features of
psychopathology and smoking, and 3) use the residuals of each different psychopathological
indicator (i.e., the portion of each psychopathological indicator that does not covary with the
10

latent dimensions and is unique relative to the other psychopathological indicators) to investigate
relations between specific features of psychopathology and smoking that are present above and
beyond the underlying latent factors.
Due to the prominence of psychopathology among today’s smokers, despite the social
pressures against smoking and the known health consequences of smoking, it is important to
examine relations between psychopathology and measures of smoking that reflect persistent
forms of smoking behavior. One important measure is heavy cigarette smoking as research has
shown that heavier smokers are less likely to alter their smoking behavior and have greater
difficulties quitting (Hyland et al., 2004; Nordstrom et al., 2000; Thompson, Thompson,
Thompson, Fredickson, & Bishop, 2003). Furthermore, heavier cigarette smoking is associated
with clinically relevant smoking variables, such as higher levels of nicotine dependence (Kandel
& Chen, 2000) and more severe craving and withdrawal (Fidler, Shahab, & West, 2011), as well
as risk for numerous types of diseases (Bagaitkar, Demuth, & Scott, 2008; Bazzano, He,
Muntner, Vupputuri, & Whelton, 2003; Cheng et al., 2000; Law, Morris, Watt, & Wald, 1997).
A second important measure is the ability to delay smoking when it is advantageous to do so.
McKee and colleagues (2006) developed a laboratory task which pits smoking a cigarette against
monetary incentives to objectively assess this construct. In this task, the ability to delay smoking
reflects a proximal, objective measure of initial lapse behavior (McKee, 2009), which is
particularly important as lapse is a strong predictor of relapse (Brandon, Tiffany, Obremski, &
Baker, 1990; Kenford et al., 1994). The ability to delay smoking when it is advantageous to do
so may also closely measure ongoing patterns of smoking following short periods of abstinent
and thus may be another relevant indicator for understanding persistent smoking behavior.

11

Elucidating the Etiologic Processes underlying Psychopathology and Persistent Smoking
In addition to identifying the features of psychopathology that drive associations with
smoking, it is also critical to elucidate mechanisms underlying the link between psychopathology
and smoking in order to develop more refined and effective smoking cessation interventions.
Clarifying these mechanisms may help provide insight into why individuals with higher levels of
psychopathology are more likely to maintain smoking, as well aid in identification of modifiable
factors that can be targeted in treatment to disrupt the links between psychopathology and
smoking. One group of models that have long been hypothesized to drive substance use
behavior in general, and may be particularly relevant for the link between psychopathology and
persistent smoking, are reinforcement models of addiction (Eissenberg, 2004; Glautier, 2004).
Reinforcement smoking is a broad construct that refers to a general tendency to smoke to
modulate affective and cognitive disturbances (Carmody, 1992; Pomerleau & Pomerleau, 1984).
Although a variety of different reinforcement models for smoking have been identified, such as
smoking for reduction of anxious and depressed mood, enhancement of pleasure and attention,
relaxation, and stimulation (Gilbert, Sharpe, Ramanaiah, Detwiler, & Anderson, 2000; Piper et
al., 2004; Tate, Pomerleau, & Pomerleau, 1994) research indicates that most of types of
reinforcement smoking can be more parsimoniously classified into two distinct groups
(Pomerleau, Fagerstrom, Marks, Tate, & Pomerleau, 2003): positive reinforcement (when a
certain behavior produces a rewarding outcome that would not have happened otherwise;
Glautier, 2004) and negative reinforcement (when a certain behavior terminates, ameliorates, or
avoids an aversive state; Eissenberg, 2004). Because nicotine has demonstrated the ability to
provide both stimulating and sedating effects (Domino, 2001; Gilbert, 1979; Murray, 1991) both
positive and negative reinforcement models are likely pertinent for cigarette smoking.
12

Furthermore, because nicotine can target a range of psychopathological symptoms (e.g., increase
arousal, alleviate emotional distress, enhance cognitive functioning, improve attention and
inhibition; Heishman, Kleykamp, & Singleton, 2010; Picciotto, Brunzell, & Caldarone, 2002;
Potter & Newhouse, 2004) and because individuals with psychopathology may be more like to
experience an exacerbation of their underlying psychopathology symptoms upon abstinence
(Pomerleau, Marks, & Pomerleau, 2000), reinforcement mechanisms may be amplified for
individuals with psychopathology. The alleviation of psychopathological symptoms upon
smoking may then reinforce future and ongoing smoking behavior as individuals learn to rely on
smoking to manage their psychopathology, and the re-emergence of their symptoms may come
to independently cue cravings for nicotine (Eissenberg, 2004).
Although considerable work has speculated that positive and negative reinforcement
mechanisms may underlie the relationship between psychopathology and smoking, limited
empirical work has examined the meditational role of these mechanisms. Furthermore, no
studies have investigated how accounting for the shared dimensions across a broad range of
emotional and behavioral psychopathology may influence these relations. The influence of these
shared dimensions is important to elucidate because this may illustrate modifiable factors that if
addressed may disrupt the link between several different forms of psychopathology and smoking.
In turn, this information may be used to develop cessation interventions that are applicable for
multiple different forms of psychopathology.

Implications of the Proposed Study and Contributions to the Literature
In conclusion, the primary aim of this dissertation is to examine how a structural model
conceptualization of psychopathology can be used to improve understanding of the
13

psychopathology-smoking link. To accomplish this goal, confirmatory factor analysis is first
used to establish a parsimonious and meaningful structural model from a wide range of
emotional and behavioral symptom- and syndrome-specific psychopathological indicators
commonly associated with cigarette smoking: happiness (Lepper, 1998), anhedonia (Leventhal et
al., 2008), depression (Kenney & Holahan, 2008), anxiety (Nabi et al., 2010), anxious arousal
(Johnson, Stewart, Zvolensky, & Steeves, 2009), ADHD symptoms (Kollins et al., 2005),
physical aggression (Audrain-McGovern, Rodriguez, Tercyak, Neuner, & Moss, 2006), and
alcohol use disorder symptomatology (Lasser et al., 2000). Second, this structural model is used
to examine shared (latent factor) and specific (indicator residual) relations between
psychopathology and smoking. Third, the influence of reinforcement mechanisms underlying
links between psychopathology and smoking is tested.
Although any one of a number of different types and measures of psychopathology could
have been included in the structural model, these particular indicators were chosen because the
main purpose was to examine how a variety of different forms of emotional and behavioral
psychopathology associated with smoking using more narrow symptom components or
syndrome specific indicators that have been consistently implicated in smoking. These indicators
were also selected so as not to bias confirmatory factor analytic results toward any particular
type of latent factor model. Symptom-specific indicators, versus syndrome-based indicators,
were included because they: 1) do not assume homogeneity across syndromes (e.g., presence vs.
absence of major depression), 2) address heterogeneity of symptoms within syndromes (e.g.,
positive and negative affect in depression), and 3) allow for the ability to examine whether
particular symptom components of a psychiatric disorder are more robustly associated with
smoking in comparison to other symptom components.
14

Notably, missing from this selected group of indicators are symptoms of schizophrenia
and personality disorders, both of which have demonstrated associations with smoking (de Leon
& Diaz, 2005; Zvolensky, Jenkins, Johnson, & Goodwin, 2011). This is primarily because these
measures were not included in the current studies. Therefore, the findings in this dissertation
regarding the structure of the latent factor model and subsequent relations with smoking
characteristics are limited to the specific emotional and behavioral psychopathological indicators
included in the model. This is important to consider because including indicators of
schizophrenia and personality disorders may change the structure of the underlying latent model.
For example, Markon (2010) found that in a confirmatory factor analysis of Axis I and Axis II
psychopathology, many symptoms of psychosis (e.g., paranoia, eccentric behavior) loaded onto a
‘thought disorder’ factor that was separate from internalizing and externalizing factors.
Consequently, future research incorporating a larger variety of psychopathology types is needed
to better understand the number and quality of latent factors underlying psychopathology and
importantly, to examine the influence of comorbidity across a wider range of psychopathology
on specific psychopathology-smoking relations.
To the best of my knowledge, this is the first in-depth examination of how accounting for
the shared dimensions underlying a broad variety of different forms of emotional and behavioral
psychopathology influences relations between psychopathology, smoking, and reinforcement
mechanisms. The results from this dissertation have the potential to have several important
implications for understanding relations between psychopathology and smoking. First, utilizing
a structural model of psychopathology can account for and investigate the influence of
psychopathological comorbidity on psychopathology-smoking relations, which poses a major
barrier to interpreting currently established relations between psychopathology and smoking.
15

Second, clarifying whether associations with smoking are attributable to shared and/or
specific features of psychopathology may aid in understanding etiological processes underlying
psychopathology and smoking. If results primarily support relations between shared features of
psychopathology and smoking, this may suggest that underlying, core psychopathological
liabilities to developing different manifest disorders (e.g., maladaptive temperament systems)
may directly associate with smoking, regardless of any impact of any particular form of manifest
psychopathology. On the other hand, if results support specific relations, this may provide
insight into particular aspects of specific psychopathological syndromes, independent from the
underlying core dimensions, that are uniquely important for smoking.
Third, the pattern of shared versus specific findings would also provide useful knowledge
on the most efficient and informative screening methods to assess psychopathological liabilities
in regular smokers (i.e., whether it is more useful to screen for underlying core dimensions of
psychopathology, specific features of different forms of psychopathology, or some combination
thereof). Additionally, this information may provide insight into which aspects of
psychopathology may be most effective and efficient to target in smoking cessation interventions
for individuals with psychopathology.
Fourth, elucidating mechanisms underlying the link between psychopathology and
persistent smoking may help identify modifiable factors that can be targeted in cessation
programs in order to offset the link between psychopathology and persistent smoking, which is
critical for the development of effective tailored cessation programs for smokers with
psychopathology. Furthermore, investigating the influence of shared dimensions of
psychopathology on these relations may identify factors that if addressed in smoking cessation
interventions may be applicable for a broad range of different forms of psychopathology. Hence,
16

using a structural model conceptualization of psychopathology to examine relations with
smoking may have several important scientific and clinical implications.

Overview of the Dissertation
The introduction of this dissertation emphasized the continued need to better understand
the psychopathology-smoking relation to reduce the public health burden of smoking among
those with psychopathology. Utilizing a structural model conceptualization of psychopathology
to isolate and incorporate underlying, shared dimensions of psychopathology into
psychopathology-smoking relations was presented as a novel approach to address the challenges
psychopathological comorbidity creates for clarifying which aspects of psychopathology
primarily associate with smoking. A brief review of the literature on psychopathological
comorbidity and relations among psychopathology, positive and negative reinforcement smoking
mechanisms, and smoking was presented.
Although prior research on psychopathology and smoking has controlled for comorbid
forms of psychopathology, to the best of my knowledge no studies to date have utilized latent
factor models of psychopathology to disentangle the extent to which smoking relations are due
to shared (i.e., latent factors that explain covariance among different forms of psychopathology)
or specific (i.e., the residual variance in each form of psychopathology that is left after partialling
out variance accounted for by latent factors) features of psychopathology. To address this gap in
the literature, this dissertation includes three independent studies that use this approach to
examine whether accounting for shared dimensions of psychopathology helps to (1) address the
challenges psychopathological comorbidity creates for interpreting relations between smoking
17

and psychopathology and (2) elucidate which shared and specific features of psychopathology
are primarily responsible for smoking relations.
Study I establishes the most parsimonious model of a set of several different emotional
and behavioral psychopathological symptom dimensions in two diverse samples: college
students and adult daily smokers. This study also provides exploratory information on relations
between the latent factor scores generated from this model and several smoking characteristics
(experimentation, established smoking, age onset, smoking heaviness, retrospective withdrawal
symptom severity). In Study II, the latent factor model of psychopathology identified in Study I
is used to test the extent to which shared versus specific features of psychopathology associate
with daily smoking level. Subsequently, motivations to smoke for positive and negative
reinforcement were added to this model to test whether these motivations mediated any links
found between shared and specific features of psychopathology and smoking heaviness. The
latent factor model established in Study I is also used in Study III to test whether shared and/or
specific features of psychopathology associate with ability to delay smoking in exchange for
monetary reinforcement. Self-report withdrawal symptoms were then added to the model to
examine whether they mediated any shared or specific links between psychopathology and
ability to delay smoking. Last, this study examines whether within-subject abstinence status
(after 16 hours smoking abstinence and non-abstinence) moderates any links between
psychopathology, withdrawal symptoms, and ability to delay smoking.
These studies have the opportunity to provide novel information on the importance and
utility of accounting for the shared dimensions of psychopathology in relation to smoking.
Although further research will be needed that incorporates larger sample sizes and a more
diverse range of type and severity of psychopathological indicators to validate the current results,
18

if these studies illustrate the presence and importance of shared dimensions of psychopathology
in relation to smoking, this may indicate that focusing on these core processes and developing
more broad-based screening assessments and cessation programs may be a meaningful and
efficient way to address the psychopathology-smoking relationship.

19

Chapter 2
Contrasting Structural Models of Psychopathology to Elucidate Relations between Shared
and Specific Features of Psychopathology and Smoking
Abstract
Psychopathological comorbidity creates challenges for interpreting relations between
psychopathology and cigarette smoking. Although latent dimensions of psychopathology have
been shown to account for psychopathological comorbidity, this conceptualization of
psychopathology has rarely been integrated into the smoking literature. In this study,
confirmatory factor analysis is used to compare three models: a 1-factor model, a 2-factor
(internalizing-externalizing) model, and a 3-factor (low positive affect-negative affect-
disinhibition) model, for nine manifest psychopathological symptom indices associated with
smoking. Subsequently, multiple regression is used to examine associations between each of the
nine manifest indices, factor scores derived from the latent factors in the structural model, and
several smoking characteristics (experimentation, established smoking, age onset, smoking
heaviness, nicotine dependence, and retrospective withdrawal symptom severity). Analyses
were conducted in two diverse samples: Study Sample 1: College Students (N = 288; mean age =
20; 75% female) and Study Sample 2: Adult Daily Smokers (N = 338; mean age = 44; 32%
female). In both samples, the 3-factor (low positive affect-negative affect-disinhibition) model
provided the best fit, and these latent factors illustrated similar patterns of associations with
smoking characteristics as their respective manifest indicators. In the college students, the
disinhibition factor and its’ respective indicators significantly associated with established
smoking. In the adult daily smokers, the low positive and negative affect factors and their
respective indicators associated with heavier smoking and retrospective withdrawal symptom
20

severity. These findings demonstrate the importance of considering shared features of
psychopathology in psychopathology-smoking relations.

Introduction
Psychiatric comorbidity is well-recognized and prevalent among nearly all types of
psychopathological disorders and symptom (Clark et al., 1995). The widespread issue of
comorbidity creates fundamental challenges not only in the psychiatric realm for classifying and
treating psychological disorders (Krueger & Markon, 2006) but also for interpreting and
addressing relations between psychopathology and important health behaviors, such as cigarette
smoking. In line with the traditional DSM approach to conceptualizing different types of
psychopathology as distinct diagnostic groups, the majority of smoking-psychopathology studies
have primarily studied and conceptualized different forms of psychopathology in isolation from
one another. This approach may be problematic because without accounting for the widespread
comorbidity, it is unclear whether the many relations documented across numerous types of
psychopathology disorders and symptoms (e.g., anxiety, depression, disruptive behavior,
substance use disorders) and aspects of the tobacco dependence process (e.g., initiation,
progression to heavy smoking, nicotine dependence, withdrawal, and quitting difficulties)
(Breslau, 1995; Breslau et al., 2004; Breslau, Peterson, Schultz, Andreski, & Chilcoat, 1996;
Elkins et al., 2007; Hu, Davies, & Kandel, 2006; Lasser et al., 2000; Niaura et al., 2001;
Pomerleau et al., 2000) were uniquely due to features specific to the form of psychopathology
being studied or to features shared among many different forms of psychopathology.
Over the last decade, significant advances in the psychiatric literature have led to the
proposal that comorbidity among many common forms of psychiatric disorders can best be
21

understood in terms of broad higher-order dimensions shared among many different forms of
psychopathology and specific disorders reflect different manifestations of underlying liabilities
(Brown & Barlow, 2009; Krueger & Markon, 2006). However, the type and number of
underlying dimensions is debatable. One hypothesis is that most psychiatric disorders and
symptoms are part of one broad problem and fall onto a single higher-order dimension of
psychological maladjustment that manifests itself as many differing forms of emotional-
behavioral disturbance (Krueger, Caspi, Moffitt, Silva, & McGee, 1996; Weiss, Susser, &
Catron, 1998). A second hypothesis is that many common psychiatric disorders are best modeled
by two underlying, correlated dimensions: internalizing (psychopathological maladjustment is
primarily expressed inward) and externalizing (psychopathological maladjustment is primarily
expressed outward) (Krueger, 1999; Krueger et al., 1998; Slade & Watson, 2006). The
internalizing dimension consists of unipolar mood and anxiety psychopathology and the
externalizing dimension consists of disruptive behavior psychopathology (e.g., ADHD,
oppositional defiant disorder, conduct disorder) and substance use problems (Krueger & Markon,
2006). A third hypothesis, proposed by Clark (2005), is that all forms of personality and
psychopathology develop from three underlying dimensions: two affective systems - positive
affectivity, which mainly associates with depression (Clark & Watson, 1991; Durbin, Klein,
Buckley, & Moerk, 2005), and negative affectivity, which underlies a broad range of
psychopathology (Clark & Watson, 1991; Krueger et al., 1996) - and one regulatory system -
disinhibition, which controls restraint and underlies the externalizing psychopathology (Krueger
& Piasecki, 2002; Lynam, Leukefeld, & Clayton, 2003).
Although a structural model-based approach to psychopathology has received
considerable attention and support in the psychiatric literature, this conceptualization of
22

psychopathology has largely not been incorporated into the smoking literature. Incorporating
this approach by isolating and integrating underlying dimensions of psychopathology into
psychopathology-smoking analyses may be particularly valuable for several reasons. First, this
technique provides a way to account for and examine the influence of comorbidity on
psychopathology-smoking relations. Second, this approach may be helpful for understanding
etiologic processes underlying the psychopathology-smoking link as these latent dimensions may
reflect underlying, core liabilities to developing manifest disorders (e.g., maladaptive
temperament systems) that directly influence smoking behavior irrespective of any impact of any
particular form of manifest symptoms. Furthermore, this approach may offer a more
parsimonious way to broadly conceptualize the psychopathology-smoking relationship and serve
as a framework to organize and integrate the diverse and large body of research on
psychopathology and smoking. Last, by elucidating psychopathological features that associate
with smoking but that are shared among many different forms of psychopathology, this approach
may provide insight into potentially useful transdiagnostic treatments.
In the current study, confirmatory factor analysis was used to test three competing
structural models to establish a parsimonious and meaningful latent factor model of a set of
psychopathological symptom indices. A major goal was to explain how a wide variety of
different forms of emotional and behavioral psychopathology were associated with smoking
utilizing more narrow symptom components or syndrome specific indicators that have been
consistently implicated in smoking. Hence, we investigated the following symptom- and
syndrome-specific indicators that have demonstrated associations with smoking: happiness,
anhedonia, depression, anxiety, anxious arousal, ADHD symptoms, physical aggression, and
23

alcohol use disorder symptomatology (Audrain-McGovern et al., 2006; Kollins et al., 2005;
Lasser et al., 2000; Lepper, 1998; Leventhal et al., 2008; Nabi et al., 2010).
Although different from most previous confirmatory factor analytic studies of
psychopathology that used syndrome-based indicators in their models, utilizing symptom-
specific indicators has several benefits. Unlike syndrome-based studies which assume
homogeneity across syndromes (e.g., presence vs. absence of major depression, composite
severity across all symptoms within a particular diagnostic category), symptom component level
analyses address heterogeneity of symptoms within syndromes (e.g., positive and negative affect
in depression; Watson, Clark, et al., 1995; Watson, Weber, et al., 1995) by not combining
different symptom forms into a common index. This is important for latent structural modeling
because it permits interpretation of how putatively distinct symptom components differentially
load onto latent factors, ultimately providing more thorough and precise information about the
composition of latent factors. It may be that particular symptom components of a psychiatric
disorder are robustly associated with smoking whereas other symptom components are only
weakly related to smoking. In those cases, exploring the overall effect of a composite indicator
that amalgamates many forms of psychopathology may obscure the underlying source of
psychiatric-smoking comorbidity. Furthermore, examining psychopathological symptom
components from a continuous (versus categorical) perspective and in community (versus only
psychiatric patient) samples is particularly important given previously demonstrated associations
between smoking and continuous measures of psychopathology at very low levels (e.g., >1
ADHD symptom, Elkins et al., 2007; >2 depressive symptoms, Niaura et al., 2001), and in
individuals that do not have the respective current psychiatric disorder (Heffner et al., 2010;
Leventhal et al., 2008).
24

Subsequently, the current study uses the identified structural model of psychopathology
to examine relations between psychopathology and smoking. In addition to investigating
associations between each of the nine different manifest psychopathological indicators and
smoking characteristics, factor scores derived from the latent factors in the structural model are
used to analyze the extent of the relationships between shared features of psychopathology and
smoking characteristics. Two diverse samples were included in the study: Study Sample 1:
College Students and Study Sample 2: Adult Daily Smokers, to examine the generalizability of
the structural model across diverse samples and to investigate relations across different stages of
the tobacco dependence process. Because college represents a time period when many
individuals are trying cigarettes and/or increasing their cigarette use (Costa, Jessor, & Turbin,
2007; Rigotti, Lee, & Wechsler, 2000), this sample provides the opportunity to examine
differences in risk of smoking experimentation and of established smoking; accordingly,
associations between psychopathology and 1) ever smoke a cigarette and 2) ever smoke 100
cigarettes, are examined in Study Sample 1. On the other hand, the adult daily smokers who have
already progressed to consistent patterns of smoking behavior allow for the ability to investigate
variation in aspects of dependence severity; consequently associations between psychopathology
and multiple indicators of dependence processes (age of smoking onset, smoking heaviness,
nicotine dependence severity, and past withdrawal symptoms) are examined in Study Sample 2.
Based on the amount of support demonstrated in prior studies across many samples
(Krueger & Markon, 2006; Krueger, 1999; Krueger et al., 1998; Slade & Watson, 2006), we
hypothesized that the psychopathological constructs would load onto two higher-order factors:
internalizing (e.g., happiness, anhedonia, depression, anxiety, anxious arousal) and externalizing
(e.g., physical aggression, ADHD symptoms, alcohol use problems). Second, due to the large
25

body of research documenting associations across many different types of psychopathology and
stages of the tobacco dependence process previously reviewed, we hypothesized that the
manifest psychopathological indicators would associate with each of the smoking characteristics.
Additionally, because we speculate that the shared features that likely account for
psychopathology comorbidity may underlie many psychopathology-smoking relationships, we
hypothesized that the factor scores derived from the structural model would also associate with
each of the smoking characteristics.

Methods
Participants and Procedures
Study Sample 1: College Students. This present study is part of a larger correlational
survey study of health behaviors and their relations with psychopathology, personality, and
genes. Participants were 288 undergraduate and graduate students enrolled at the University of
Southern California. Fliers, class announcements, electronic postings, and human subject
participation pools were used to recruit students. Inclusion criteria were: 1) ≥ 18 years of age, 2)
fluent in English (many questionnaires were not yet available in other languages and staff
resources were not able to translate and validate the measures in other languages), 3) able to
provide informed consent, and 4) currently enrolled at the University of Southern California.
The study was approved by the University’s Institutional Review Board.
Study announcements instructed participants to either correspond with research staff to
obtain additional information about the study or to sign up directly for a data collection session
through a web-based student research participation interface. Participation involved attending a
single weekday in-person data collection session that lasted approximately 90-120 minutes. All
26

study sessions were run in small groups on the USC campus and conducted by trained research
staff in which participants completed questionnaires and provided either a saliva or buccal cell
sample for DNA analysis.
Study Sample 2: Adult Daily Smokers. The present study is a secondary analysis based
on a study examining the influence of individual differences in psychopathology on sensitivity to
tobacco deprivation. Participants were current adult smokers recruited through community
advertisements (e.g., newspaper, online advertisements) and referrals. For inclusion, participants
must have met the following criteria: 1) ≥ 18 years of age, 2) regular cigarette smoker for ≥ 2
years, 3) currently smoke ≥ 10 cigarettes per day, 4) reported normal or corrected-to-normal
vision with no color blindness and 5) fluent in English. Participants were excluded if they
evidenced the following: 1) active DSM-IV non-nicotine substance dependence, 2) current DSM-
IV mood disorder or psychotic symptoms (to minimize cognition-impairing effects of acute and
severe psychiatric dysfunction), 3) breath carbon monoxide (CO) levels <10ppm at intake, 4) use
of non-cigarette forms of tobacco or nicotine products, 5) use of psychiatric medications, or 6)
currently pregnant. Participants were compensated ~$200 for completing the entire study. The
study was approved by the University’s Institutional Review Board.
Following an initial telephone screen, participants were invited to attend an in-person
baseline session involving informed consent, breath CO analysis, psychiatric screening interview
by a trained research assistant to further assess eligibility, and measures of psychopathology and
smoking, which served as the primary data utilized in this study. Eligible participants
subsequently completed overnight tobacco deprived and non-deprived experimental sessions. Of
the 515 smokers recruited, 165 were ineligible, 7 declined to participate, and 5 had unclear
27

responses on some of the main smoking outcomes, leaving a final sample of 338 who completed
the baseline session and were included in analyses.

Measures
Measures specific to Study Sample 1: College Students.
Tobacco and Alcohol Use History Survey (TAUHS; Pierucci-Lagha et al., 2005). The
TAUHS assesses lifetime tobacco and alcohol use characteristics, including age of onset,
patterns of heaviest use, current use, quit attempts, and family history of substance dependence.
This measure is based on the validated alcohol and tobacco modules of the Semi-structured
Assessment for Drug and Alcoholism and was altered to fit a paper-and-pencil survey rather than
the original interview format. The TAUHS was used in the current study to assess the following
smoking characteristics: 1) ever smoke a cigarette (yes/no), and 2) smoked 100+ cigarettes in
lifetime (yes/no), a marker of established smoking.
Measures specific to Study Sample 2: Adult Daily Smokers.
Structured Clinical Interview for DSM-IV-Axis I Disorders, Research Version, Non-
Patient Edition (SCID-I/NP; First, Spitzer, Gibbon, & Williams, 2002). The SCID-I/NP is a
well-established interview to evaluate psychiatric diagnoses and was administered to assess
lifetime and current prevalence of key Axis I disorders (i.e., lifetime psychotic disorder, current
mood disorder, current [hypo]manic disorder, past year non-nicotine substance dependence) for
eligibility purposes.
Smoking History Questionnaire (SHQ; Brown, Lejuez, Kahler, & Strong, 2002). The
SHQ measures general information about smoking history and smoking patterns. Variables
included in this study are: age onset of regular smoking, average number of cigarettes smoked
28

per day, and total mean severity of seven retrospective withdrawal symptoms. For withdrawal
symptoms, participants were asked how severely (1=Not at all to 5=Very severe) they
experienced seven withdrawal symptoms (i.e., cravings, irritability, nervousness, difficulty
concentrating, physical symptoms, difficulty sleeping, loss of interest or pleasure) in their most
recent attempt to quit smoking, which were averaged to create a reliable construct (α = .88).
Although retrospective assessments of withdrawal have limitations to consider when interpreting
results (e.g., recall bias; Shiffman et al., 1997), this measure illustrated good convergent validity
with nicotine dependence (r = .38, p < .0001) and significantly associated with prospectively
assessed withdrawal symptoms (composite score on the Minnesota Nicotine Withdrawal Scale;
Hughes & Hatsukami, 1986) following 16 hours of deprivation (r = .50, p < .0001) in the subset
of these smokers who completed the entire study (N = 286).
Fagerström Test of Nicotine Dependence (FTND; Heatherton, Kozlowski, Frecker, &
Fagerstrom, 1991). The FTND is a well-validated, 6-item self-report measure of nicotine
dependence that produces a composite score.
Measures administered in both studies.
Personal Information Questionnaire (PIQ). The PIQ is a self-report questionnaire that
collects sociodemographic information (e.g., age, gender, race, and ethnicity).
Subjective Happiness Scale (SHS; Lyubomirsky & Lepper, 1999). The SHS is a
previously validated, 4-item measure of global subjective happiness. Participants report what
they think is the most accurate description of themselves on a 7-point scale (e.g., “In general, I
consider myself”: 1=Not a very happy person to 7=A very happy person). A mean score across
the four items was computed.
29

Center for Epidemiologic Studies Depression Scale: Anhedonia Subscale (CESD:ANH;
.Radloff, 1977). The CESD:ANH scale is a subscale on the 20-item CESD, a widely used, well-
validated assessment of current levels of depressive symptomatology in clinical and non-clinical
samples. Participants report how often they have felt a certain way “during the past week” on a
4-point Likert-type scale (1=Rarely or none of the time [0-1 days] to 4=Most or all of the time
[5-7 days]). The CESD:ANH subscale contains four reverse-scored items pertaining to feelings
of happiness and hopefulness (e.g., “I enjoy life”) which were averaged to compute a composite
score, such that a higher score indicates more severe depressive symptomatology. Confirmatory
factor analyses have demonstrated support for the CESD:ANH subscale (Knight, Williams,
McGee, & Olaman, 1997; Nguyen, Kitner-Triolo, Evans, & Zonderman, 2004; Shafer, 2006) and
previous studies have demonstrated associations between this subscale and smoking variables
(Leventhal et al., 2008; Pomerleau, Zucker, & Stewart, 2003). Mean composite scores for the
raw (non-reversed) items were also calculated in order to compare levels on this subscale with
other similar study samples. To examine the prevalence of clinically meaningful low positive
affect depressive symptoms in the present sample, the standard scoring procedure for the CESD,
which classifies participants with a total score of 16+ (out of 20) as having depressive symptoms,
was tailored to the CESD:ANH scale used in the current study, such that a mean score of 0.8
(16/20) or higher on the CESD:ANH reflected presence of depressive symptoms.
Mood and Anxiety Symptom Questionnaire-Short Form (MASQ-SF; Watson, Clark, et
al., 1995; Watson, Weber, et al., 1995). The MASQ is a validated, 62-item self-report measure of
affective symptoms. Participants rate “how much” they have experienced each symptom “during
the past week, including today,” on a 5-point Likert-type scale (1=Not at all to 5=Extremely).
Based on the Tripartite Model of Anxiety and Depression as a conceptual guide (Watson, Clark,
30

et al., 1995; Watson, Weber, et al., 1995), the MASQ consists of four subscales: 1) Anxious
Arousal (MASQ:AA), a measure of somatic tension and physical hyperarousal (e.g., “felt
dizzy”), 2) Anhedonic Depression (MASQ:AD), a measure of loss of interest in life (e.g.,
“nothing felt enjoyable”) with reverse-keyed items measuring positive affect (“felt really
happy”), 3) General Distress-Depression (MASQ:GDD), a measure of non-specific depressed
mood commonly experienced in both anxiety and depression (e.g., “felt sad”), and 4) General
Distress-Anxiety (MASQ:GDA), a measure of non-specific anxious mood commonly
experienced in both anxiety and depression (e.g., “was unable to relax”). In line with traditional
scoring procedures, sum scores on each of the four subscales were used in the present study for
primary analyses. Mean scores on each of the subscales were also calculated to examine the
proportion of participants who surpassed clinically relevant cutpoints established in prior work
(≥ 1.8 on the MASQ:AA, ≥ 1.9 on the MAS:GDA and MASQ:GDD, and ≥ 2.3 on the
MASQ:AD; Schulte-van Maaren et al., 2012).
Aggression Questionnaire Revised: Physical Aggression (AQR:PA; Bryant & Smith,
2001)The AQR:PA scale is a subscale of the 12-item AQR, a validated and refined version of the
29-item Aggression Questionnaire (Buss & Perry, 1992), which measures multidimensional
aspects of dispositional aggression. Participants are asked to rate “how characteristic the
following statements are of you” on a 6-point Likert-type scale (1=Extremely uncharacteristic to
6=Extremely characteristic). The AQR:PA subscale contains three items that tap disposition to
physical aggression (e.g, “Given enough provocation, I may hit another person”) and has been
supported in prior confirmatory factor analysis (Bryant & Smith, 2001). Scores were based on
the mean across the three items.
31

Adult ADHD Self-Report Scale (ASRS; Kessler, Adler, et al., 2005). The ASRS is a
validated self-report rating scale for evaluating Attention Deficit Hyperactivity Disorder
(ADHD) symptoms in adults (Kessler, Adler, et al., 2005). This measure includes 18 items (9
inattentive items, e.g., “How often are you distracted by activity or noise around you” and 9
hyperactive-impulsive items, e.g., “How often do you feel restless or fidgety”), that correspond
directly to DSM-IV Criterion A symptoms of ADHD. Participants rate each item based on how
often they have felt and conducted themselves over the past 6 months on a 5-point Likert-type
scale (1=Never to 5=Very often). Following suggestions from prior work (Kessler, Adler, et al.,
2005), a total mean score was calculated across all 18 items for primary analyses. To examine
presence of clinical levels of ADHD symptoms in the sample (≥ 9 symptoms; Kessler, Adler, et
al., 2005), a sum score across the 18 items (possible range: 0-18) on the ASRS was computed
based on the standard scoring rubric for the ASRS, which dichotomizes each symptom as present
or absent based on severity level of the symptom endorsed.
Alcohol Use Disorders Identification Test (AUDIT; Babor, Higgins-Biddle, Saunders, &
Monteiro, 2001) . The AUDIT is a ten-item measure of alcohol use patterns, with items
pertaining to hazardous alcohol use (e.g., frequency of heavy drinking), dependence symptoms
(e.g., morning drinking), and harmful alcohol use (e.g., blackouts). The AUDIT has received
widespread validation across gender, age, and culture (Allen, Litten, Fertig, & Babor, 1997;
Saunders, Aasland, Amundsen, & Grant, 1993). Response options range from 0 to 4, with higher
scores reflecting more hazardous or harmful alcohol use. Participants received a sum score
(possible range: 0-40) across the ten items, which was also used to assess prevalence of alcohol
use problems in the current study. Prior work suggests that a sum score of 8+ reflects potential
problematic drinking (Babor et al., 2001).
32

Data Analysis
Preliminary analyses. Percentages, means, standard deviations, and Cronbach alpha
coefficients for scales were calculated for key variables. Bivariate correlations were calculated
to examine the intercorrelations among all psychopathological symptoms dimensions included in
the structural models. To examine severity level of psychopathology and representativeness of
the current samples, mean and sum scores for each of the scales were compared to other studies
involving comparable samples (i.e., college students and community/smoking samples) and
proportions of individuals who scored above established cut-off points on relevant scales (see
Measures section) were calculated. Because certain questionnaires were added in at different
time points during the two studies, there is a range of missing data across the measures in the two
samples. In Study Sample 1: College Students, N’s for each questionnaire are the following (out
of a possible 288): SHS (265), CESD (288), MASQ (287), AQR (283), ASRS (287), AUDIT
(285); and for Study Sample 2: Adult Smokers are (out of a possible 338): SHS (324), CESD
(338), MASQ (324), AQR (278), ASRS (252), AUDIT (202). All preliminary analyses were
conducted in SAS v9.2 (SAS Institue Inc., 2009) and based on complete, non-missing data.
Primary analyses.
Confirmatory factor analysis. Confirmatory factor analysis (CFA), based on the analysis
of covariance, was used to test several competing models of the set of psychopathology symptom
indices: (1) a 1-factor model in which all psychopathological symptom indices loaded onto one
factor reflecting general psychopathological maladjustment, (2) a 2-factor model in which SHS,
CESD:ANH, MASQ:AD, MASQ:GDD, MASQ:GDA, and MASQ:AA loaded onto an
internalizing factor and AQR:PA, ASRS, and AUDIT loaded onto an externalizing factor, and,
(3) a 3-factor model in which SHS, CESD:ANH, and MASQ:AD loaded onto a low positive
33

affect factor, MASQ:GDD, MASQ:GDA, MASQ:AA loaded onto a negative affect factor, and
AQR:PA, ASRS, and AUDIT loaded onto a disinhibition factor (see Figure 1 for a representation
of these models). If the best-fitting model consisted of multiple first-order latent factors, we also
examined whether these first-order factors loaded onto a second-order factor of general
psychological maladjustment representing broad concepts, such as overall severity of
psychopathological functioning or features of psychopathology widely shared across many
differing forms of emotional and behavioral disturbances.
All CFA were conducted in Mplus Version 6 (Muthen & Muthen, 1998-2011) with
maximum likelihood estimation (ML) to addresses missing data. ML has been recommended as
one efficient approach for managing missing data in SEM (Allison, 2003) and has demonstrated
an ability to outperform traditional methods of handling missing data (Enders & Bandalos, 2001;
Peters & Enders, 2002). Instead of deleting cases or imputing missing observations, ML
estimation partitions observations into subsets, each with the same pattern of missing
observations, and extracts important statistical information from each subset so all observations
are kept for analysis (Allison, 2003; Muthen & Muthen, 1998-2011). Therefore, full samples in
both data sets (Study Sample 1: College Students, N=288; Study Sample 2: Adult Daily
Smokers, N=338) were used in all primary analyses. Because certain measures were skewed
across the samples, ML estimation with robust standard errors (MLR) was used for all analyses
to account for potentially problematic multivariate non-normality.
Model fit evaluation was based on the following criteria recommended by Hu & Bentler
(1999): 1) a non-significant Chi-square goodness-of-fit statistic (p-value >.05), 2) a comparative
fit index (CFI) > .95, and 3) a root mean-square error of approximation (RMSEA) < .06. The
Chi-square difference test was used to compare competing models. Because MLR estimation
34

was used, the chi-square goodness-of-fit statistic is based on the Satorra-Benter scaled chi-square
(S-B χ2; Satorra & Bentler, 1994) and the Chi-square difference is based on the Satorra &
Bentler scaled difference chi-square statistic (Satorra & Bentler, 2001). Meaningful empirical
modifications to improve model fit were considered using the modindices output option in
Mplus. Once the best-fitting model was established, factor scores for each of the latent factors in
the final structural model were computed in Mplus, which uses the modal posterior estimate
regression method (Muthen & Muthen, 1998-2011).
Relationships between psychopathology and smoking characteristics. For each smoking
characteristic, separate models that included each manifest psychopathological symptom indice
and each factor score as the sole predictor were tested. In Study Sample 1: College Students, the
two smoking characteristics assessed (ever smoke and smoke 100+ cigarettes) were measured
dichotomously; therefore, multiple logistic regression was used for all models. In Study Sample
2: Adult Daily Smokers, the four smoking characteristics examined (age of onset, cigarettes per
day, FTND, retrospective withdrawal symptom severity) were measured continuously; thus,
multiple linear regression was utilized for all models. Due to significant associations between
each of age, gender, race/ethnicity and at least some of the manifest psychopathological
symptom indicators and some of the smoking characteristics in both samples, these variables
were included as covariates in all models. Because of low cell counts and to reduce model
complexity, race/ethnicity was reduced into fewer categories: Black or African American, White,
and Other in the adult smokers; Asian, White, and Other in the college students. Results are
reported as standardized odds ratios (ORs) in Study Sample 1: College Students and standardized
beta weights (βs) in Study Sample 2: Adult Daily Smokers. All regression analyses were
conducted in Mplus Version 6 (Muthen & Muthen, 1998-2011).
35

For all analyses, significance was set at p < .05 without adjustment for multiple tests to
provide a broad picture of the pattern of relations between manifest indicators of
psychopathology, shared features of psychopathology, and different smoking characteristics. All
tests were two-tailed except for the Satorra-Bentler scaled chi-square difference tests, which
were one-tailed as they assessed improvement in model fit.

Results
Preliminary results.
Shown in Table 1, Study Sample 1: College Students was mainly female, late
adolescence/young adult, White or Asian, and non-Hispanic. A little less than half of the college
students in this sample had ever smoked a cigarette and only a small portion had smoked 100+
cigarettes in their lifetime. Study Sample 2: Adult Daily Smokers was primarily male, middle-
aged, Black or White, and non-Hispanic. These participants began smoking regularly around
late adolescence/young adulthood, smoked approximately 1.5 packs per day, had moderate levels
of nicotine dependence, and experienced mild to moderate levels of withdrawal symptoms on
their most recent quit attempt. Means, standard deviations, and Cronbach’s alpha coefficients of
the psychopathological symptoms dimensions included in the structural models for both samples
are presented in Table 2. Bivariate correlations among the psychopathological symptom
indicators exhibited a wide degree of intercorrelation magnitudes across variables in both
samples (see Table 3).
Generally, scores on the psychopathology indices were higher for the college students;
however, across both samples, mean and sum scores on most of the psychopathological symptom
scales (SHS, CESD:ANH, MASQ scales, AQR:PA) largely reflected low severity of
36

psychopathology and were comparable to other similar study samples of college students and
community residents/adult smokers (Borders, Earleywine, & Jajodia, 2010; Fresco, Heimberg,
Mennin, & Turk, 2002; Gerevich, Bacskai, & Czobor, 2007; Leventhal et al., 2008;
Lyubomirsky & Lepper, 1999; Mazzeo & Espelage, 2002; Watson, Clark, et al., 1995). The
percent of participants who were above established clinically relevant cut-off points on
questionnaires were as follows (Study Sample 1: College Students / Study Sample 2: Adult
Smokers): CESD:ANH = 42.4% / 34.3%, MASQ:AD = 58.2% / 54.9%, MASQ:GDD = 33.8% /
19.4%, MASQ:GDA = 30.3% / 13.3%, MASQ:AA = 9.0% 8.0%, ASRS = 18.8% / 8.7%, and
AUDIT = 28.7% / 9.9% (see Measures section for a description of the cut-off points). For the
ASRS, these proportions are higher than estimated college student (Weyandt & DuPaul, 2006)
and adult (Kessler et al., 2006) ADHD prevalence rates. For the AUDIT, these proportions were
slightly lower than rates in college student samples (Saitz et al., 2007) and similar to rates in non
alcohol-dependent adult community samples (Saunders et al., 1993).
Primary results.
Confirmatory factor analyses. Illustrated in Table 4, the pattern of fit indices for each of
the three hypothesized models was relatively similar across both samples. For each subsequent
model, the fit indices improved and the Satorra-Bentler scaled difference chi-square statistic was
significant. The three factor model provided the best fit for the data in both samples; however,
fit indices still largely did not meet recommended criteria for good model fit. In both samples,
the output modification indices suggested that loading non-specific depressed mood
(MASQ:GDD) additionally onto the low positive affect factor would significantly improve
model fit. Because this modification was suggested in both samples, we decided this
modification was warranted and required further consideration. This modification significantly
37

improved model fit and almost all fit indices demonstrated excellent fit across both samples. All
psychopathological symptom indicators significantly loaded onto their respective factors (see
Figure 2; standardized loadings presented). In these models, the three latent factors
demonstrated significant moderate correlations indicating that they may be subfactors on a
second-order factor. Indeed, each first-order factor significantly loaded onto a second-order
factor in both samples when we tested this model (see Figure 3; standardized loadings
presented). Accordingly, factor scores were computed for this second-order factor and included
in regression analyses with smoking characteristics.
Relationships between psychopathology and smoking characteristics. Shown in Table 5,
in Study Sample 1: College Students, the AUDIT significantly associated with ever smoke a
cigarette and the ASRS, AUDIT, and disinhibition factor each significantly associated with ever
smoke 100 cigarettes. Neither the low positive and negative affect factors nor any of their
respective indicators associated with these smoking characteristics. Similarly, the second-order
general maladjustment factor did not associate with either of the smoking variables. In Study
Sample 2: Adult Daily Smokers, cigarettes per day positively associated with CESD:ANH, all
MASQ subscales, the low positive affect and negative affect factors, and inversely associated
with the SHS and AUDIT. Retrospective withdrawal symptom severity positively associated
with all MASQ subscales, ADHD symptoms, all three of the first-order latent factors, and the
second-order factor of general maladjustment. Age onset of regular smoking only associated
(positively) with MASQ:GDA and the negative affect factor. Nicotine dependence severity did
not associate with any of the manifest indicators or latent factors.

38

Discussion
Identifying a structural model of psychopathology. In accord with prior research
(Brown & Barlow, 2009; Krueger & Markon, 2006), underlying latent dimensions reflecting
core psychopathological processes accounted for shared variance across the range of manifest
psychopathological symptoms in this study. However, in contrast to prior research (Krueger,
1999; Krueger et al., 1998; Slade & Watson, 2006) and our hypothesis, a three factor model
provided the best fit for the data in both samples. This three factor model has not often been
directly tested in the psychiatric literature; rather one factor models, two factor (internalizing-
externalizing) models, and models that support the DSM organizational scheme, have primarily
been examined (Krueger, 1999; Krueger et al., 1998; Vollebergh et al., 2001), which may
account for this discrepancy. The current identified model does not directly cohere with DSM-
based syndrome conceptualizations as two distinct symptom manifestations of the same DSM-
based syndrome (i.e., unipolar depression) loaded onto separate factors of [low] positive and
negative affect. The make-up of this three factor model underscores the problem of within-
disorder heterogeneity and supports approaches using symptom-specific indicators that do not
combine heterogeneous manifestations of psychopathology.
This model is largely consistent with Clark’s (2005) proposed three factor model of
personality and psychopathology, which suggests that three latent factors representing innate
biobehavioral temperament systems – positive affectivity, negative affectivity, and disinhibition -
account for comorbidity across all combinations of personality and psychopathology. In Clark’s
model, positive affectivity refers to a person’s tendency to experience a wide range of positive
emotions, which reflects the strength of a behavioral approach system aimed at obtaining reward
and mainly inversely associates with depression (Clark & Watson, 1991; Durbin et al., 2005).
39

The [low] positive affect factor yielded in this study is consistent with the behavioral approach
system conceptualization, as it had loadings from several measures of happiness and anhedonia.
One notable finding was that the MASQ:GDD cross-loaded onto this factor. The relatively
lower loadings of the MASQ:GDD compared to the other indicators on this factor suggests that
the underlying behavioral approach system is primarily dominated by indicators of low positive
affect but may also encompass certain aspects of distress. Indeed, studies have found that the
behavioral approach system may be linked to certain low-arousal distress indicators which
appear on the MASQ:GDD scale (e.g., sadness, hopelessness, feelings of failure, and
discouragement; Carver, 2004; Higgins, Shah, & Friedman, 1997).
Negative affect, the second factor in Clark’s model, reflects a person’s tendency to
experience a diverse range of aversive emotional states, which reflects the strength of the
behavioral avoidance system aimed at avoiding threat and underlies a broad range of
psychopathology (Clark & Watson, 1991; Krueger et al., 1996). Negative affect has been
proposed as a central, shared component of anxiety and depression (Clark & Watson, 1991).
Consistent with this, the indicators that loaded onto this factor were measures of depressive and
anxious symptoms (MASQ:GDD, MASQ:GDA, MASQ:AA). The high loading of the
MASQ:GDA on this factor indicates that in the current study, this factor largely reflects a
general anxiety-centered negative affect.
The third factor in the model is disinhibition, which refers to a person’s lack of restraint
in response to incoming stimuli (Clark, 2005). Disinhibition underlies a range of externalizing
disorders (e.g., disruptive behavior and substance use disorders; Kendler, Prescott, Myers, &
Neale, 2003; Krueger, 1999; Lynam et al., 2003; Vollebergh et al., 2001). Unsurprisingly,
measures of hyperactivity-impulsivity, alcohol abuse, and physical aggression significantly
40

loaded onto this factor, each of which are components of externalizing disorders. The relatively
high factor loading of the ADHD symptom indicator on this factor suggests that in this study,
this factor may be more prominently characterized by the generalized impulse control across
many situations rather than impaired inhibition of drinking behavior or aggressive impulses.
Results also demonstrated that these three latent first-order factors significantly loaded on
a second-order factor representing their shared variance. In this second-order model, there are
two levels of shared features: 1) the first-order factors: low positive affect, negative affect, and
disinhibition, which consist of psychopathological features that differentiate these three factors
from one another but that are shared among the manifest indicators that load on to each factor,
and 2) the second-order factor: general maladjustment, which consists of psychopathological
features shared across many differing forms of emotional and behavioral psychopathology. This
second-order factor also likely represents general constructs, such as overall severity of
psychopathological functioning, regardless of type or quality of psychopathology.
Relations between psychopathology and smoking characteristics. In partial support of
our hypothesis, each of the manifest psychopathological symptom indices, except for physical
aggression, and each of the underlying latent factors associated with at least some of the smoking
characteristics; however, not all psychopathology-smoking relations were significant. In Study
Sample 1: College Students, only the latent factor of disinhibition and its respective indicators
demonstrated significant results. The AUDIT was the only psychopathology measure to
associate with cigarette smoking experimentation (ever smoke a cigarette). A large body of
research has found a strong link between smoking and drinking in college students (Nichter,
Nichter, Carkoglu, Lloyd-Richardson, & Tern, 2010; Weitzman & Chen, 2005); these results
support this link and further suggest that something important about alcohol use disorder
41

symptomatology, separate from the variance shared with other disinhibitory psychopathology,
such as associating with substance using peers (Lynskey, Fergusson, & Horwood, 1998) and
being in substance use situations (Nichter et al., 2010), may drive this relationship. With respect
to established smoking, alcohol use disorder symptomatology, ADHD symptoms, and the latent
disinhibition factor associated with ever smoke 100 cigarettes. This is consistent with prior
research demonstrating associations across a number of different externalizing disorders and
markers of established smoking (Burt, Dinh, Peterson, & Sarason, 2000; Kollins et al., 2005;
Masse & Tremblay, 1997; McMahon, 1999; Rohde et al., 2003) and suggests that many of these
relationships may stem from features of impulse control shared across many of these disorders.
While many college students experiment with cigarette smoking, particularly in combination
with alcohol use, these results illustrate that a much smaller portion demonstrate established
smoking and indicate that impulse control features shared across a broad range of disinhibitory
psychopathology may increase risk for this more established pattern of smoking. Furthermore,
as the second-order factor did not associate with either of these smoking characteristics it appears
that individual differences specifically in disinhibitory psychopathology, rather than overall
severity of psychopathological functioning or broad psychopathological features, were
responsible for these relations.
In the adult daily smoker sample, each of the low positive affect and negative affect
indicators significantly associated with cigarettes per day. Potentially, this finding reflects
individuals who smoke at heavier levels to self-medicate their affective disturbances (Carmody,
1992; Pomerleau & Pomerleau, 1984, Picciotto et al., 2002). It is also possible that continued,
heavy tobacco smoking may have led to exacerbations in low positive affect and negative affect
(Breslau et al., 1998). Generally, these results may indicate that underlying shared features of
42

low positive affect and negative affect largely account for many of the associations documented
between various different manifestations of affective and anxious psychopathology and heavier
smoking (Greenberg et al., 2012; Johnson et al., 2009; Kenney & Holahan, 2008; Kollins et al.,
2005; Lasser et al., 2000). Also in this sample, almost all of the low positive affect and negative
affect indicators, the ADHD symptom indicator, each of the first-order latent factors (low
positive affect, negative affect, disinhibition), and the second-order general maladjustment factor
associated with more severe retrospective composite withdrawal symptom severity. Taken
together, this broad pattern of findings may suggest that overall severity of psychopathology or
widespread psychopathological features, rather than individual differences in type or quality of
psychopathology, primarily explains relations found across many different types of
psychopathology and more severe nicotine withdrawal (Breslau, Kilbey, & Andreski, 1992;
Pomerleau et al., 2000; Weinberger, Desai, & McKee, 2010; Ameringer & Leventhal, 2012;
McClernon et al., 2011; Pomerleau, Downey, et al., 2003).
The present results also revealed several unexpected findings. First, alcohol use disorder
symptomatology inversely associated with cigarettes per day. Potentially, this finding might
capture heavy drinking, social smokers, who smoke considerably while drinking but have lighter
patterns of daily smoking (King & Epstein, 2005) compared to heavy smokers who likely smoke
more evenly day-by-day (Krukowski, Solomon, & Naud, 2005). Second, the MASQ:GDA and
the underlying factor of negative affect significantly associated with a later age of smoking
onset. Similarly, a previous study found that children with anxiety disorders, controlling for
comorbidity with other psychiatric disorders (e.g., ADHD, depression), had a later age onset of
smoking (Costello, Erkanli, Federman, & Angold, 1999). Because individuals with anxiety-
spectrum pathology are more likely to have higher levels of trait harm avoidance (Brown,
43

Svrakic, Przybeck, & Cloninger, 1992; Starcevic, Uhlenhuth, Fallon, & Pathak, 1996) and may
be more worried about the negative consequences of smoking (Baker, Brandon, & Chassin,
2004), this may reduce risk for early age of smoking onset. However, due to the retrospective
nature of this study measure, longitudinal research is needed to more thoroughly examine this
relation. Third, none of the manifest psychopathological indicators and latent factors associated
with nicotine dependence severity. This finding may reflect that fact that only individuals who
smoked ten or more cigarettes per day were included in the study. Prior research has found that
risk for nicotine dependence increases most significantly from less than one cigarette per day to
ten cigarettes per day and is minimal at levels higher than ten cigarettes per day (Kandel & Chen,
2000). This suggests that lower levels of smoking may be needed to demonstrate relations
between psychopathology and variation in nicotine dependence.
Overall, these results illustrated that the underlying dimensions of low positive affect,
negative affect, and disinhibition that link together many different types of manifest
psychopathology associated with the same smoking characteristics as their respective individual
manifest indicators, which emphasizes the importance of accounting for shared features of
psychopathology in smoking relations. The associations between these underlying dimensions
and smoking outcomes provides insight into etiological processes underlying the
psychopathology-smoking link, as they suggest that factors associated with innate biobehavioral
temperament systems (e.g., genetic risk factors or disrupted emotional processes; Watson,
Gamez, & Simms, 2005), which increase vulnerability to different forms of manifest
psychopathology (Clark, 2005), may directly influence smoking regardless of any influence of
any specific form of manifest psychopathology. Additionally, these findings indicate that
transdiagnostic treatment approaches that focus on core, underlying psychopathological
44

processes may be an effective and efficient way to offset the link between many different types
of psychopathology and smoking outcomes. Although these analyses are unable to demonstrate
if specific components of the different manifest indicators associate with smoking outcomes,
above and beyond the influence of the underlying dimensions, these results illustrate that taking
a more macroscopic and transdiagnostic approach to the psychopathology-smoking relation by
examining and targeting psychopathological features that may associate with smoking but that
may be shared among many different forms of psychopathology deserves further consideration.
Limitations and conclusions. Several limitations should be considered when interpreting
the current results. First, all psychopathology and smoking measures in this study were self-
report and therefore are subject to several biases, such as recall, desirability, and self-awareness.
Second, the cross-sectional design of the study does not allow for causal or temporal conclusions
of relations between the measures of psychopathology and smoking characteristics.
Additionally, the cross-sectional nature of the study leaves unclear whether the structural model
found in the current study will generalize across time as symptoms of psychopathology change.
However, this concern is somewhat offset because the same structural model was found in two
diverse samples and prior research has found evidence for the stability of core psychopathology
processes over three years (Krueger et al., 1998). Third, due to the relatively small sample sizes
for confirmatory factor analyses, the current structural model only included nine dimensional
measures of psychopathology that have often illustrated associations with smoking
characteristics. Any one of a number of different types, measures, and conceptualizations of
psychopathology could have been included (e.g., measures of schizophrenia and personality
disorders), which may have changed the structure of the model and factors. However, based on a
number of studies in the psychiatric literature (Krueger & Markon, 2006; Krueger, 1999;
45

Krueger et al., 1998; Slade & Watson, 2006) it appears that a factor of externalizing
disinhibition, which underlies the spectrum of disruptive behavior disorders, and one or two
factors of internalizing affectivity, which underlie(s) the spectrum of mood and anxiety
disorders, emerge across a variety of different types, measures, and conceptualizations of
psychopathology.
Fourth, because individuals who were currently on psychiatric medications or suffering
from current mood disorder episodes or substance dependence were excluded from the adult
smokers, it is unclear how these results will generalize to more severe levels of psychopathology
in this population. Fifth, significant sociodemographic differences do not allow for direct
comparisons of the psychopathology and smoking results between the two samples. As there
appeared to be an interesting disparate pattern of associations between the two samples (i.e.,
disinhibitory psychopathology more strongly associated with earlier smoking characteristics in
the college students whereas affective psychopathology more robustly associated with heavier,
more dependent smoking variables in the adult smokers), future longitudinal research using the
same sample is of importance to examine whether different types of psychopathology
differentially influence different stages of the tobacco dependence process. Last, because
structural models of psychopathology have largely not been incorporated into the smoking
literature, significance was kept at p < .05 to provide a broad picture of the patterns of
psychopathology-smoking relations using this type of analysis; therefore, findings are subject to
Type I error due to the multiple outcomes assessed.
Despite these limitations, this study is one of the first to examine how a structural model
of psychopathology may address the issues psychopathological comorbidity creates for
understanding psychopathology-smoking relations. Furthermore, this study is unique by
46

incorporating a range of symptom- and syndrome-specific psychopathology and by including
two diverse samples. These results further support the notion that underlying, core dimensions
of psychopathology likely account for the extensive comorbidity across many different types of
psychopathology, demonstrate that these dimensions may be generalizable across
sociodemographically diverse samples, and suggest that many of the relations found across
different manifestations of psychopathology and smoking may stem from latent, shared features
of psychopathology. Due to the early nature of this work, future research is needed to further
validate the underlying factors with different types of psychopathology indices and to use more
sophisticated analytic techniques to investigate how structural models of psychopathology can
help to elucidate the psychopathological correlates of smoking.

47

Table 1
Select Demographic and Smoking Characteristics
Variable Study Sample 1:
College Students
Study Sample 2:
Adult Daily Smokers
Demographic Characteristics M (SD) / % M (SD) / %
Gender (female) 74.7% 32.3%
Race
Black
White
Asian
Multi-Racial
Other

4.5%
37.7%
36.3%
11.4%
10.1%

51.9%
34.7%
0.9%
4.2%
8.3%
Ethnicity (Hispanic) 14.4% 14.8%
Age 19.8 (1.7) 43.8 (10.8)
Smoking Characteristics
Ever Smoke a Cigarette (yes) 39.5% ----
Ever Smoke 100 Cigarettes (yes) 10.5% ----
Age Onset ---- 19.2 (5.5)
Cigs/Day ---- 16.8 (6.8)
FTND ---- 5.3 (1.9)
Withdrawal Symptoms ---- 2.5 (0.9)
Note. N=288 in Study Sample 1: College Students. N=338 in Study Sample 2: Adult Daily Smokers. M M (SD) = Mean
(Standard Deviation). Ever Smoke = ever smoked a cigarette (yes/no). Smoke 100 = Smoked 100 cigarettes in lifetime (yes/no).
Age Onset = age first started smoking regularly. Cigs/Day = average number of cigarettes smoked per day. FTND = Fagerstrӧ m
Test of Nicotine Dependence. Withdrawal Symptoms = severity of withdrawal symptoms in most recent quit attempt.

48

Table 2
Descriptive Statistics of Manifest Psychopathological Symptom Indices Included in the
Structural Models

Study Sample 1: College Students Study Sample 2: Adult Daily Smokers
M (SD) Alpha M (SD) Alpha
SHS 5.05 (1.20) .85 5.28 (1.07) .76
CESD:ANH 0.88 (0.70) .83 0.72 (0.65) .67
MASQ:AD 55.56 (15.60) .93 52.58 (14.11) .89
MASQ:GDD 21.82 (9.52) .94 17.60 (7.02) .92
MASQ:GDA 18.59 (6.40) .83 15.17 (5.52) .87
MASQ:AA 22.89 (7.37) .88 21.66 (6.79) .89
AQR:PA 1.80 (0.94) .68 1.82 (1.06) .77
ASRS 2.60 (0.49) .85 2.18 (0.64) .92
AUDIT 5.56 (5.16) .84 3.50 (4.86) .88
Note. N=288 in Study Sample 1: College Students. N=338 in Study Sample 2: Adult Daily Smokers. M (SD) = Mean (Standard
Deviation). Alpha = Cronbach’s alpha. SHS=Subjective Happiness Scale; CESD:ANH=Center for Epidemiologic Studies
Depression-Anhedonia Scale; MASQ=Mood and Anxiety Symptom Scale: AD=Anhedonic Depression Scale, GDD= General
Distress Depression Scale, GDA=General Distress Anxious Scale, AA=Anxious Arousal Scale; AQR:PA=Aggression
Questionnaire Revised-Physical Aggression Scale; ASRS=Adult ADHD Self-Report Scale; AUDIT=Alcohol Use Disorders
Identification Test.
49

Table 3
Correlations between Manifest Psychopathological Symptom Indices Included in the Structural
Models

1. 2. 3. 4. 5. 6. 7. 8. 9.
1. SHS --- -.58† -.59† -.45† -.21† -.16** -.16** -.10 .08
2. CESD:ANH -.44† --- .76† .61† .37† .29† .13* .26** -.01
3. MASQ:AD -.59† .56† --- .74† .47† .40† .18** .28† -.01
4. MASQ:GDD -.34† .30† .43† --- .71† .59† .14* .38† .04
5. MASQ:GDA -.27† .22† .31† .81† --- .75† .17** .40† .03
6. MASQ:AA -.24† .19† .28† .67† .83† --- .19** .43† .06
7. AQR:PA -.16** .15* .12* .26† .31† .34† --- .24† .20†
8. ASRS -.22† .24† .33† .45† .53† .49† .36† --- .24†
9. AUDIT -.06 .14* .01 .07 .04 .08 .24† .18** ---
Note. Correlations on the upper right diagonal refer to Study Sample 1: College Students (N=288). Correlations on the lower left
diagonal refer to Study Sample 2: Adult Daily Smokers (N = 338). SHS =Subjective Happiness Scale; CESD:ANH=Center for
Epidemiologic Studies Depression-Anhedonia Scale; MASQ=Mood and Anxiety Symptom Scale: AD=Anhedonic Depression
Scale, GDD=General Distress Depression Scale, GDA=General Distress Anxious Scale, AA=Anxious Arousal Scale;
AQR:PA=Aggression Questionnaire Revised-Physical Aggression Scale; ASRS=Adult ADHD Self-Report Scale;
AUDIT=Alcohol Use Disorders Identification Test.
*p < .05, **p < .01, †p< .001.
50

Table 4
Fit Indices for Comparative Structural Models of the Set of Manifest Psychopathological
Symptom Indices

Fit Index
Factors S-B χ
2
P df CFI RMSEA S-B χ
2
Diff
Study Sample 1: College Students (N = 288)
1 331.60 .0000 27 .672 .198 ---
2 307.81 .0000 26 .696 .194 20.54†
3 167.12 .0000 24 .846 .144 146.58†
3-Refined 39.45 .0177 23 .982 .050 48.39†
Study Sample 2: Adult Daily Smokers (N = 338)
1 244.21 .0000 27 .741 .154 ---
2 233.68 .0000 26 .753 .154 10.45**
3 53.13 .0006 24 .965 .060 176.21†
3-Refined 28.50 .1975 23 .993 .027 20.05†
Note. 1 Factor Model = ‘Maladjustment’ Factor; 2 Factor Model = ‘Internalizing’ and ‘Externalizing’ Factors; 3 Factor Model =
‘Low Positive Affect,’ ‘Negative Affect,’ and ‘Disinhibition’ Factors; 3-Refined = 3 Factor Model with empirical modifications
based on modification indices output (MASQ:GDD loads onto both ‘Positive Affect’ and ‘Negative Affect’ factors). S-B

χ2 =
Satorra-Bentler model Chi-square goodness-of-fit statistic; P = Satorra-Bentler model Chi-square p-value; df = model degrees of
freedom; CFI = comparative fit index; RMSEA = root mean square error of approximation; S-B χ2 = Satorra-Bentler scaled chi-
square difference test.
* p < .05, **p < .01, † p< .001.

51

Table 5
Associations between Manifest Psychopathological Symptom Indices, Latent Factors, and
Smoking Characteristics

Study Sample 1:
College Students
Study Sample 2:
Adult Daily Smokers
Psychopathology Indices Ever
Smoke

Smoke
100

Age
Onset
Cigs/
Day
FTND Withdrawal
Manifest Indicators OR

OR

β

β

β

β

SHS 0.88 1.14 .03 -.12* -.03 -.06
CESD:ANH 0.97 0.88 -.03 .11* .10 .09
MASQ:AD 1.06 1.00 .07 .13* .01 .12*
MASQ:GDD 0.91 0.94 .10 .12* -.01 .21†
MASQ:GDA 0.91 1.02 .13* .12* -.00 .30†
MASQ:AA 1.00 1.20 .08 .13* .06 .28†
AQR:PA 1.08 1.38 -.01 .05 .03 .04
ASRS 1.15 1.65** .03 -.07 -.10 .15*
AUDIT 2.23† 2.72† -.07 -.18* -.08 -.03
Latent Factors
Low Positive Affect -1.03 -0.96 .04 .14** .03 .12*
Negative Affect 0.93 1.08 .12* .12* -.00 .29†
Disinhibition 1.17 1.70** .06 .05 -.04 .23†
General Maladjustment 0.97 1.13 .08 .09 -.02 .26†
Note. N=288 in Study Sample 1: College Students; N=338 in Study Sample 2: Adult Daily Smokers. SHS=Subjective Happiness Scale;
CESD:ANH=Center for Epidemiologic Studies Depression-Anhedonia Scale; MASQ=Mood and Anxiety Symptom Scale: AD=Anhedonic
Depression Scale, GDA=General Distress Anxious Scale, AA=Anxious Arousal Scale; AQR:PA=Aggression Questionnaire Revised-Physical
Aggression Scale; ASRS=Adult ADHD Self-report Scale; AUDIT=Alcohol Use Disorders Identification Test. Ever Smoke = ever smoked a
cigarette (yes/no); Smoke 100 = Smoked 100 cigarettes in lifetime (yes/no); Age Onset = age first started smoking; Cigs/Day = average number
of cigarettes per day; FTND=Fagerstrӧ m Test of Nicotine Dependence; Withdrawal = withdrawal symptom severity in most recent quit attempt.
β = standardized betas. OR = standardized odds ratios. All models are adjusted for age, gender, and race. Significant findings at p <.05 are in
bold.
*p < .05, **p < .01, †p < .001.
52

Figure 1. Hypothesized Latent Factor Models of the Set of Psychopathological Symptom Indices

Figure 1. Model A: Hypothesized 1-factor model of psychopathology symptom indices. Model B:
Hypothesized 2-factor model of psychopathology symptom indices. Model C: Hypothesized 3-factor
model of psychopathology symptom indices. SHS=Subjective Happiness Scale; CESD:ANH=Center for
Epidemiologic Studies Depression-Anhedonia Scale; MASQ=Mood and Anxiety Symptom Scale:
AD=Anhedonic Depression Scale, GDD=General Distress Depression Scale, GDA=General Distress
Anxious Scale, AA=Anxious Arousal Scale; AQR:PA=Aggression Questionnaire Revised-Physical
Aggression Scale; ASRS=Adult ADHD Self-Report Scale; AUDIT = Alcohol Use Disorders
Identification Test. Error terms associated with each psychopathological symptom index are not shown.

53

Figure 2. Best-Fitting Latent Factor Model in Both Samples

Figure 2. Model A: Best-fitting model of psychopathology symptom indices, factor loadings, and factor
correlations in Study Sample 1: College Students (N=288). Model B: Best-fitting model of
psychopathology symptom indices, factor loadings, and factor correlations in Study Sample 2: Adult
Daily Smokers (N=338). SHS=Subjective Happiness Scale; CESD:ANH=Center for Epidemiologic
Studies Depression-Anhedonia Scale; MASQ=Mood and Anxiety Symptom Scale: AD=Anhedonic
Depression Scale, GDD=General Distress Depression Scale, GDA=General Distress Anxious Scale,
AA=Anxious Arousal Scale; AQR:PA=Aggression Questionnaire Revised-Physical Aggression Scale;
ASRS=Adult ADHD Self-Report Scale; AUDIT=Alcohol Use Disorders Identification Test. All factor
loadings and factor correlations are standardized. Error terms associated with each psychopathological
symptom index are not shown.
*p < .05, **p < .01, †p < .001.

54

Figure 3. Second-Order Factor Model in Both Samples

Figure 3. Model A: Second-order factor model of psychopathology symptom indices, factor loadings, and
factor correlations in Study Sample 1: College Students (N=288). Model B: Second-order factor model of
psychopathology symptom indices, factor loadings, and factor correlations in Study Sample 2: Adult
Daily Smokers (N=338). SHS=Subjective Happiness Scale; CESD:ANH=Center for Epidemiologic
Studies Depression-Anhedonia Scale; MASQ=Mood and Anxiety Symptom Scale: AD=Anhedonic
Depression Scale, GDD=General Distress Depression Scale, GDA=General Distress Anxious Scale,
AA=Anxious Arousal Scale; AQR:PA=Aggression Questionnaire Revised-Physical Aggression Scale;
ASRS=Adult ADHD Self-Report Scale; AUDIT=Alcohol Use Disorders Identification Test. All factor
loadings and factor correlations are standardized. Error terms associated with each psychopathological
symptom index are not shown.
*p < .05, **p < .01, †p < .001.
55

Chapter 3
Shared versus Specific Features of Psychopathology and Smoking Level: Structural
Relations and Mediation by Positive and Negative Reinforcement Mechanisms
Abstract
The complex interrelations among several forms of psychopathology obscure the nature
and motivational mechanisms of psychopathology-smoking relations. The purposes of this study
is to examine: 1) the extent to which features of psychopathology that are shared (i.e., found
across multiple forms of psychopathology) versus specific (i.e., unique to a particular form of
psychopathology) associated with smoking level; and 2) whether motivation to smoke for
negative and positive reinforcement mediated psychopathology-smoking relations. Adult daily
smokers (N = 338) completed self-report measures of psychopathology constructs (e.g.,
depression, anxiety, alcohol abuse, ADHD, aggression), reinforcement smoking, and cigarettes
per day (CPD). Structural equation modeling was used to: 1) construct a three-factor model of
psychopathology consisting of latent low positive affect, negative affect, and disinhibition, 2)
test relations of each latent factor and manifest indicator residual to CPD, and 3) examine the
mediational role of positive and negative reinforcement smoking. The shared features of lower
positive affect, higher negative affect, and lower disinhibition associated with higher CPD. Of
the specific features, only alcohol use problems associated with lower CPD over and above the
latent shared features. Negative reinforcement smoking mediated the link between negative
affect and CPD. Positive reinforcement smoking did not mediate any relations. Shared
underlying dimensions rather than specific features of psychopathology may be most important
to the etiology of heavier smoking patterns.

56

Introduction
The slowing rate of decline in smoking prevalence over the last decade (CDC, 2002,
2011, 2012) has led to the speculation that a hardcore subgroup of smokers exist who are more
likely to continue smoking or who have particular difficulties quitting (Irvin & Brandon, 2000).
Although no formal definition of the hardcore smoker exists, heavy smoking has often been
associated with this concept (Costa et al., 2010; Emery et al., 2000) due to considerable research
demonstrating that heavier smokers are less likely to alter their smoking behavior, lack
confidence in quitting, and have reduced cessation success (Hyland et al., 2004; Nordstrom et al.,
2000; Thompson et al., 2003). Heavier cigarette smoking is also associated with clinically
relevant smoking variables, including more severe craving and withdrawal (Fidler et al., 2011)
and nicotine dependence (Kandel & Chen, 2000), as well as risk for several types of diseases,
including lung cancer (Law et al., 1997), cardiovascular disease (Bazzano et al., 2003),
susceptibility to bacterial infection (Bagaitkar et al., 2008), and ocular diseases (Cheng et al.,
2000). As such, considerable research has sought to identify the correlates of cigarettes per day
(CPD) to aid in the development of effective prevention and treatment efforts to offset the public
health burden of smoking.
One robust correlate of CPD is psychopathology, as a range of different psychiatric
disorders and symptom indices associate with higher CPD (Greenberg et al., 2012; Johnson et
al., 2009; Kenney & Holahan, 2008; Kollins et al., 2005; Lasser et al., 2000). Much of this work
has examined psychopathological covariation with smoking frequency in samples mixed with
both daily and non-daily smokers or heterogeneous samples of daily smokers CPD that include
individuals who smoke as little as one CPD. However, non-daily and very light smokers, whose
smoking is primarily driven by external motivations (e.g., social purposes, sensory motives,
57

specific situational influences), are qualitatively different from heavier smokers who are more
often driven by internal or pharmacological motivations to smoke (e.g., avoid withdrawal,
automaticity, reduce craving, control negative moods) (Shiffman, Dunbar, Scholl, & Tindle,
2012; Shiffman, Kassel, Paty, Gnys, & Zettler-Segal, 1994). Including these qualitatively
heterogenous samples makes examining relations with CPD difficult to interpret because
psychopathology may play different roles in CPD at different levels of smoking heaviness. For
example, Payne et al. (2013) found that although depressive symptoms associated with smoking
status, depressive symptoms inversely associated with CPD among those who smoked ≥20 CPD,
which they attributed to a successful antidepressant effect of nicotine with increased
consumption. Therefore, examining relations between psychopathology and CPD specifically
across heavier levels of daily smoking may clarify important individual differences in
psychopathology, as well as motivational mechanisms linking psychopathology and CPD, at
heavier smoking levels.
Additionally, although research has provided evidence for a likely relation across many
different types of psychopathology and CPD, much of this research has not accounted for the
extensive rate of comorbidity among many forms of psychopathology (Clark, Watson, Reynolds,
1995), which creates barriers for understanding which aspects of psychopathology specifically
associate with heavier smoking. Previously, we found that three latent dimensions (low positive
affect, negative affect, disinhibition) accounted for the covariance among a set of
psychopathological symptom indicators frequently associated with smoking: low happiness,
anhedonia, depression, anxiety, anxious arousal, ADHD symptoms, physical aggression, and
alcohol use (Audrain-McGovern et al., 2006; Kollins et al., 2005; Lasser et al., 2000; Lepper,
1998; Leventhal et al., 2008; Nabi et al., 2010) (see Figure 2B, pg. 53) in two independent
58

samples, including one sample that overlaps with the sample in the current report. As outlined in
Clark’s (2005) proposed three factor model of personality and psychopathology, these latent
factors may represent three different biobehavioral temperament systems: 1) positive affectivity,
an individual’s tendency to experience a wide range of positive emotions, which reflects the
strength of a behavioral approach system aimed at obtaining reward and mainly inversely
associates with depression (Clark & Watson, 1991; Durbin et al., 2005), 2) negative affectivity,
an individual’s tendency to experience a wide range of aversive emotional states, which reflects
the strength of the behavioral avoidance system aimed at avoiding threat and underlies a broad
range of psychopathology (Clark & Watson, 1991; Krueger et al., 1996), and 3) disinhibition, an
individual’s lack of restraint in response to incoming stimuli, which underlies the externalizing
psychopathology (Krueger & Piasecki, 2002; Lynam et al., 2003).
Incorporating this model and directly testing associations of the three latent factors
(“shared features” of psychopathology) and each of the manifest indicators’ residuals (i.e., the
portion of variance within the indicator that does not covary with the latent factor and is unique
from the other indicators on that factor: “specific features” of psychopathology) with smoking
level can help elucidate which particular components of psychopathology drive associations with
CPD. This may have important implications for scientific conceptualizations of the
psychopathology-smoking relationship. If results illustrate associations primarily with shared
features of psychopathology, this may indicate that underlying, core psychopathological
liabilities to developing different types of manifest psychopathology (e.g., maladaptive
temperament systems), directly associate with smoking level irrespective of any influence of any
particular manifest syndrome. On the other hand, associations with specific features of
psychopathology would provide insight into particular aspects of specific manifest
59

psychopathological syndromes (e.g., somatic symptoms of anxiety) that may be uniquely
important for heavy smoking.
This type of analysis is also beneficial because simultaneously testing associations of
different types of psychopathology to smoking level can reveal potentially important suppressor
effects, which occur when two correlated predictors have opposing relations with the dependent
variable (Paulhus, Robins, Trzesniewski, & Tracy, 2004). For instance, Hicks & Patrick (2006)
found evidence of a suppressor effect between two different factors of psychopathy (Factor 1:
interpersonal and affective traits associated with psychopathy, Factor 2: impulsive and antisocial
behaviors) predicting emotional distress, such that relations between both psychopathy factors
and emotional distress increased in opposite directions (negatively for F1 and positively for F2)
when both factors were simultaneously included as predictors. Thus, modeling multiple
psychopathology-smoking relations in the same model may unearth important links between
pure forms of psychopathology and smoking level that are not apparent until adjusting for the
shared variance across different types of psychopathology.
It is also important to elucidate mechanisms underlying links between psychopathology
and CPD to help identify factors that can be targeted in treatment. One proposed hypothesis for
the link between psychopathology and heavier, more dependent smoking is that individuals with
psychopathology are motivated to smoke to control their underlying psychopathological
symptoms (Carmody, 1992; Pomerleau & Pomerleau, 1984), which is supported by the ability of
nicotine to target a range of psychopathological symptoms (e.g., increase arousal, alleviate
emotional distress, enhance cognitive functioning, improve attention and inhibition; Heishman et
al., 2010; Picciotto et al., 2002; Potter & Newhouse, 2004). In turn, the successful alleviation of
their underlying symptoms may then reinforce future and heavier smoking as individuals learn to
60

rely on smoking to manage their psychopathology (Eissenberg, 2004; Glautier, 2004). Smoking
reinforcement motivations can be broadly classified as either negative (terminating/avoiding
negative outcomes; e.g., “Cigarettes help me deal with anxiety or worry”) or positive (producing
positive outcomes; e.g., “I smoke to get a sense of euphoria”) reinforcement smoking
motivations (Pomerleau, Fagerstrom, et al., 2003). A handful of studies have found support for a
mediational link between psychopathology, smoking for negative and/or positive reinforcement,
and heavier, more dependent smoking (Cohen, McCarthy, Brown, & Myers, 2002; Copeland,
Brandon, & Quinn, 1995; Schleicher, Harris, Catley, & Nazir, 2009); however, these studies
have only examined overall depressive and/or anxious symptom severity and dispositional
negative affect. Therefore, it currently remains unclear whether negative and/or positive
reinforcement smoking motivations mediate the link between other types of psychopathology
(e.g., low positive affect and/or disinhibitory psychopathology) and heavier smoking, and how
shared versus specific forms of psychopathology may influence these pathways.
One relevant question in negative and positive reinforcement models is whether there are
“trait-consistent” patterns of associations between specific psychopathological dimensions and
negative vs. positive reinforcement smoking motivations, which would be expected if individuals
with psychopathology smoke to control their underlying symptoms. In support of this, Lerman
and colleagues (1996) found a two-step mediational link in which depressive symptoms
associated with negative affect reduction smoking, which in turn led to stimulation smoking, and
stimulation smoking related to nicotine dependence. In contrast, Audrain and colleagues (1998)
found that smoking for negative affect reduction, but not for stimulation, mediated the link
between trait anxiety and nicotine dependence. Notably, negative affect reduction and
stimulation smoking were both important for depressed smokers whereas only negative affect
61

reduction smoking was relevant for anxious smokers, possibly because low positive affect plays
a much stronger role in depression compared to anxiety (Brown, Chorpita, & Barlow, 1998).
Because the ability to distinguish negative vs. positive reinforcement smokers based on their pre-
existing psychopathology may have important implications for tailored interventions, ongoing
research that adjusts for the covariation among many different types of psychopathology is
needed to clarify whether it is possible to distinguish types of reinforcement smokers based on
their pre-existing psychopathology.
The present study aims to refine the links between psychopathology and CPD across the
moderate to heavy range of smoking (≥ 10 CPD) in adult daily smokers and to elucidate the
underlying role of negative and positive reinforcement mechanisms in these relations. First, a
three factor model is used to examine the extent to which smoking relations are attributable to
shared versus specific features of psychopathology. Based on prior research illustrating stronger
relations between shared features of psychopathology and alcohol dependence (Kushner et al.,
2012), and the notion that core temperament systems may be more important for influencing
heavier smoking than the specific manifest symptoms, we hypothesize that the shared features of
psychopathology (low positive affect, negative affect, and disinhibition) will primarily associate
with heavier levels of smoking. Second, motivations to smoke for negative and positive
reinforcement were added to the model to examine whether these variables mediated any links
between psychopathology and CPD. Based on studies previously reviewed illustrating that
positive and negative reinforcement smoking mediated the link between depressive and anxious
symptomatology and heavier, more dependent smoking, we hypothesize that this relationship
extends to other types of psychopathology and motivation to smoke for negative and positive
reinforcement will mediate links found between psychopathology and CPD. Given that it has
62

been proposed that individuals smoke to address their specific psychopathological deficits
(Gilbert & Gilbert, 1995), one question was whether there were ‘trait-consistent’ relations
between specific types of psychopathology and negative vs. positive reinforcement smoking
motivations. There is some evidence that negative reinforcement smoking may be a primary
affective mechanism linking nicotine dependence to anxiety—a syndrome putatively underlied
by negative affect (Audrain et al., 1998), whereas both negative and positive reinforcement are
mechanisms linking smoking to depression—a syndrome putatively underlied by both low
positive affect and high negative affect (Lerman et al., 1996). We sought to explore whether
differential trait-consistent relations were apparent in this study.

Methods
Participants and Procedures
The present study is a secondary analysis derived from a study examining the influence
of individual differences in psychopathology on sensitivity to tobacco deprivation. Briefly,
participants were current adult smokers invited to participate in a study on personality and
smoking. Participation involved: 1) an initial telephone screen of general inclusion criteria, 2) a
baseline session involving informed consent, breath CO analysis, psychiatric screening interview
by a trained research assistant to further assess eligibility, and measures of psychopathology and
smoking, which were the main data used in the present study, and 3) if eligible, two in-person
overnight tobacco abstinent and nonabstinent experimental sessions. The study was approved by
the University’s Institutional Review Board. Please see pages 26-27 for a more detailed
description of the recruitment methods, eligibility criteria, and study procedures. Of the 515
smokers recruited, 165 were ineligible, 7 declined to participate, and 5 had unclear responses to
63

CPD, leaving a final sample of 338 who completed the baseline session at which all of the
measures used in the current report were administered and which served as the final sample for
analyses.

Measures
Demographic and eligibility information.
Personal Information Questionnaire (PIQ). The PIQ is a self-report questionnaire that
collects sociodemographic information (e.g., age, gender, ethnicity).
Structured Clinical Interview for DSM-IV-Axis I Disorders, Research Version, Non-
Patient Edition (SCID-I/NP; First et al., 2002). The SCID-I/NP is a well-established interview to
evaluate psychiatric diagnoses and was administered to assess lifetime and current prevalence of
key Axis I disorders (i.e., lifetime psychotic disorder, current mood disorder, current
[hypo]manic disorder, past year non-nicotine substance dependence) for eligibility purposes.
Independent variables.
The latent factors and residuals generated from the set of psychopathological symptom
indices represent the independent variables. Please see pages 27-31 for a description of these
variables.
Mediator variables.
Minnesota Nicotine Reinforcement Questionnaire (MNRQ; Pomerleau, Fagerstrom, et
al., 2003). The MNRQ is a 13-item self-report measure of smoking for positive reinforcement
(MNRQ:PR; 5 items) and negative reinforcement (MNRQ:NR; 8 items). Participants rate the
extent to which they agree with or experience several statements about smoking (e.g., “I smoke
because it is pleasurable” – positive reinforcement, or “I crave a cigarette to provide relief from
64

withdrawal” – negative reinforcement) on a 4-point Likert-type scale (0=Never to 3=Always).
The MNRQ has demonstrated acceptable psychometric properties (Pomerleau, Fagerstrom, et al.,
2003). Participants received a mean score across the items for each separate scale.
Dependent variable.
Smoking History Questionnaire (SHQ; Brown et al., 2002). The SHQ measures basic
information about smoking history and smoking patterns, including previous quit attempts, age
of smoking first cigarette, age of regular smoking, etc. On the SHQ participants were asked to
report the “exact” number of cigarettes they smoker per day, which serves as the continuous
dependent variable.
Data Analysis
Preliminary analyses. Descriptive statistics (e.g., percentages, means, and standard
deviations) of demographic characteristics, relevant smoking variables, and psychopathological
symptom dimensions were calculated. Cronbach alpha coefficients were assessed for each of the
psychopathological symptom indices and smoking reinforcement scales. Bivariate correlations
between smoking variables and psychopathological symptom indices included in the structural
model were calculated. To examine severity level of psychopathology and representativeness of
the current sample, mean and sum scores for each of the scales were compared to other studies
involving community and/or smoking samples, and the proportions of individuals who scored
above established cut-off points on relevant scales (see Measures section) were calculated.
Because certain psychopathology measures were added during the study, there is a range of
missing data across these measures. Out of a total possible 338, complete Ns for each of the
following questionnaires are: SHS = 324, CESD = 338, MASQ = 324, AQR = 238, ASRS = 220,
AUDIT = 178. Dependent t-tests revealed no significant differences in CPD across all
65

questionnaires between those who missed the questionnaire vs. those who did not miss the
questionnaire. All preliminary analyses were performed in SAS v9.2 (SAS Institute Inc., 2009)
and based on complete, non-missing data.
Primary analyses. All structural equation model (SEM) analyses were conducted in
MPLUS Version 6 (Muthen & Muthen, 1998-2011) and based on the analysis of covariance.
Maximum likelihood (ML) estimation, one of the most highly recommended approaches to deal
with missing data in SEM (Allison, 2003; Enders & Bandalos, 2001; Peters & Enders, 2002),
was used. ML estimation maintains all observations for analysis by partitioning patterns of
missing observations into subsets and extracting important information from each subset
(Allison, 2003; Muthen & Muthen, 1998-2011). Therefore, the full baseline sample (N=338) was
used for all primary SEM analyses. Because certain measures were skewed, ML estimation with
robust standard errors (MLR) was used for all analyses to account for potentially problematic
multivariate non-normality. Acceptable model fit was based on the following criteria
recommended by Hu & Bentler (1999): 1) a non-significant Chi-square goodness-of-fit statistic
(p-value >.05), 2) a comparative fit index (CFI) > .95, and 3) a root mean-square error of
approximation (RMSEA) < .06. Because MLR estimation was used, the chi-square goodness-of-
fit statistic is based on the Satorra-Benter scaled chi-square (S-B χ2; Satorra & Bentler, 1994)
Shared and specific relationships between psychopathology and CPD. To test the extent
to which shared features of psychopathology associated with CPD, regression paths from each of
the three latent factors (low positive affect, negative affect, disinhibition) to CPD were included
in the model. These paths were first added separately to examine their independent contribution
to CPD and then all paths were added simultaneously to examine the unique contribution to CPD
(i.e., the association between each factor and CPD after adjusting for the shared variance among
66

the factors). Subsequently, paths from the residuals of each psychopathological symptom
indicator to CPD were included, one at a time, to examine whether the residual of a particular
indicator demonstrated a significant relation to CPD above and beyond the influence of the
underlying factors. Residuals were created following the steps provided by Muthen & Muthen
(“Regression on Residual,” n.d.). Only those pathways (from the latent factors and residuals) that
were significant were included in the final model for mediational analyses. Age, gender, and
race were not significantly associated with CPD in the structural model and yielded no
substantive changes to the primary results; thus, they were not included as covariates in the final
models.
Mediation analyses. After associations between the latent factors and residuals with
CPD were established, motivations to smoke for negative and positive reinforcement were added
to the model to examine whether these motivations mediated any of the shared or specific links
found between the latent factors, residuals, and CPD. Although including both reinforcement
measures in one model is ideal, we opted for separate models to increase the parameter to sample
size ratio that is important for maintaining trustworthy results in SEM. The MODEL INDIRECT
command in Mplus was used to obtain the indirect (based on the product of coefficients method)
and total effects. All primary results are reported as standardized beta weights (βs), all tests are
two-tailed, and significance was set at p < .05.

Results
Preliminary results.
Participants were an average of 43.7 (SD = 10.8) years of age, primarily male (68%),
were: 52% Black, 35% Caucasian, and 14% other, with 15% identifying as Hispanic, and
67

reported smoking an average of 16.8 (SD = 6.8) cigarettes per day. Although participants had to
report smoking ≥ 10 CPD on their phone screen and had to pass a baseline CO level ≥ 10ppm to
be included in the study, some individuals reported smoking fewer than 10 CPD at the time of
the in-person visit (N = 22), resulting in a reported range of 6-60 CPD. Means, standard
deviations, Cronbach’s alpha coefficients, and bivariate correlations of all variables included in
the models are shown in Table 6. Mean and sum scores across most of the psychopathological
symptom scales (SHS, CESD:ANH, MASQ scales, AQR:PA) reflected low severity of
psychopathology, high between-participant variability, and were generally comparable to those
found in studies of community and/or smoking samples (Gerevich et al., 2007; Leventhal et al.,
2008; Lyubomirsky & Lepper, 1999; Watson, Clark, et al., 1995). The percentage of those who
scored above established clinically relevant cut-off points on questionnaires were similar or
slightly higher to prior community samples (Kessler et al., 2006; Saunders et al., 1993) and were
as follows: CESD:ANH = 34.3%, MASQ:AD = 54.9%, MASQ:GDD = 19.4%, MASQ:GDA =
13.36%, MASQ:AA = 8.0%, ASRS = 8.7%, and AUDIT = 9.9% (see measures section for a
description of cut-points).
Primary results.
Shared and specific relationships between psychopathology and CPD. When entered
independently in the model, low positive affect was the only latent factor to significantly
(positively) associate with CPD (β = .16, p < .01). However, when pathways from low positive
affect, negative affect, and disinhibition were simultaneously added to the model, negative affect
became positively associated with CPD (β = .24, p < .05) and disinhibition became inversely
associated with CPD (β = -.31, p < .01), indicating that shared variance among the factors may
have suppressed effects between pure, partialled factors of negative affect and disinhibition with
68

CPD. Low positive affect maintained significance with CPD (β = .21, p < .05) in this model.
When each indicator residual was added to the model, only the residual from harmful alcohol use
significantly (inversely) associated with CPD above and beyond the influence of the underlying
factors. Inclusion of this residual reduced the positive association between negative affect and
CPD to a non-significant trend level (β = .21, p < .10); associations between low positive affect,
disinhibition, and CPD reduced in strength but remained significant. We chose to keep the path
between negative affect and CPD in the final model since the reduction in strength was small and
because adjusting for this path reveals the significant relationship between disinhibition and CPD
in the final model. The final model of significant paths between the latent factors and residuals to
CPD is illustrated in Figure 4. This model had excellent fit: S-B χ
2
= 33.86, p = .21; CFI = .993;
RMSEA = .025.
Mediation analyses. As shown in Table 7 and Figure 5A, neither low positive affect,
disinhibition, nor the alcohol use (AUDIT) residual significantly associated with motivation to
smoke for negative reinforcement, and no significant indirect pathways were found for these
variables. On the other hand, negative affect positively associated with motivation to smoke for
negative reinforcement, which in turn associated with higher CPD, and the indirect effect of this
path was significant. As can be seen in Figure 5B, negative affect, disinhibition, and the AUDIT
residual did not associate with motivation to smoke for positive reinforcement. Low positive
affect inversely associated with positive reinforcement motivation to smoke, which in turn
associated with higher CPD. Furthermore, accounting for the influence of this mediator
strengthened the relationship between low positive affect and CPD, indicating that motivation to
smoke for positive reinforcement was potentially suppressing a stronger link between low
positive affect and CPD. Both mediational models demonstrated excellent fit - negative
69

reinforcement model: S-B χ
2
= 37.36, p = .28; CFI = .995; RMSEA = .020, positive
reinforcement model: S-B χ
2
= 37.18, p = .28; CFI = .996; RMSEA = .019.

Discussion
Shared and specific relationships between psychopathology and CPD. Consistent
with our hypothesis, the specific features of psychopathology demonstrated little association with
CPD above and beyond the influence of the latent shared features. On the other hand, the latent
factors of low positive affect, negative affect, and disinhibition each associated with CPD. As
expected, lower positive affect and higher negative affect associated with heavier smoking.
Because none of their respective residuals significantly associated with CPD, it appears that
these underlying dimensions of positive and negative affect more directly associate with CPD
regardless of any impact of any particular manifest symptom of an internalizing syndrome (e.g.,
psychomotor retardation in depression, autonomic arousal in anxiety). These results provide an
important context for interpreting prior studies that found that several different forms of manifest
affective and anxious psychopathology associate with smoking heaviness (Greenberg et al.,
2012; Johnson et al., 2009; Kenney & Holahan, 2008; Kollins et al., 2005; Lasser et al., 2000).
Perhaps an influence of shared psychopathological features (low positive affect and negative
affect) may explain the well-documented relation between internalizing symptoms and smoking.
Furthermore, these findings provide insight into the etiological process linking psychopathology
and heavy smoking. It is possible that maladaptive positive and negative temperament systems
(e.g., an increased tendency to experience a wide range of aversive emotional states), which
increase susceptibility to different types of internalizing manifest psychopathology, may directly
influence heavier smoking irrespective of any influence of any particular manifest internalizing
70

disorder. As such, focusing on the underlying, core dimensions at the center of internalizing
psychopathology, instead of the manifest internalizing symptoms and syndromes, may provide a
more fundamental and parsimonious way to interpret and to address the relationship between
internalizing psychopathology and heavier smoking.
After accounting for the influence of low positive affect and negative affect, disinhibition
was inversely associated with CPD. Prior studies have generally demonstrated positive
associations between facets of disinhibitory psychopathology and CPD (Grano, Virtanen,
Vahtera, Elovainio, & Kivimaki, 2004; Kollins et al., 2005; Shekelle, Gale, Ostfeld, & Paul,
1983). Two key differences between the current study and these studies may account for the
discrepant findings. First, the present study only included smokers who reported at least
moderate levels of daily smoking, whereas the other studies included individuals who smoked as
little as one CPD. Thus, it may be that disinhibition positively associates with CPD at lighter
levels of smoking but inversely associates with CPD at heavier levels of smoking. Future
empirical work could address this by examining whether quadratic effects are present between
disinhibition and CPD across the entire smoking level range. Second, these prior studies did not
adjust for comorbid psychopathology. In the present study, relations between disinhibition,
negative affect, and CPD only became significant when paths from all latent factors to CPD were
included in the model. This suggests that covariance among the latent factors may have
suppressed these associations and highlights that there may be important relations between pure
forms of psychopathology and smoking variables that are not apparent until accounting for
comorbid psychopathology. Taken together, these findings may suggest that individuals with
pure features of disinhibition (e.g., compulsivity, poor control, low persistence) smoke primarily
for specific instrumental purposes (e.g., to increase attention and inhibition) when necessary and
71

are successfully self-medicating these symptoms, leading to the inverse association between
disinhibition and CPD across this upper level of smoking. This is in contrast to individuals with
negative affective psychopathology who may need to smoke more heavily and evenly throughout
the day to maintain consistent nicotine levels to avoid withdrawal symptoms, which mediated the
link between negative affect and CPD.
In addition to disinhibition, the residual of harmful alcohol use (AUDIT) inversely
related to CPD, indicating that something particular about this indicator, apart from the shared
variance in generalized impulse control with ADHD symptoms and physical aggression, was
important for this relation. Potentially, this finding reflects individuals whose smoking is
specifically cued by drinking and thus have more irregular, lighter patterns of daily smoking
(King & Epstein, 2005). However, because the level of alcohol use was low due to study
exclusion of alcohol dependent individuals, it is unclear how these results generalize to more
severe levels of harmful alcohol use.
Negative and positive reinforcement models. Consistent with prior research (Audrain
et al., 1998; Cohen et al., 2002; Copeland et al., 1995; Lerman et al., 1996), motivation to smoke
for negative reinforcement mediated the link between negative affect and CPD. As the negative
reinforcement scale in this study primarily assesses DSM nicotine withdrawal symptoms, this
mediation model may accord with a withdrawal-based model of negative reinforcement. That is,
this finding may suggest that individuals with elevated negative affect may consume high levels
of CPD to avoid or attenuate withdrawal symptoms. Research illustrating that individuals with
affective and anxious psychopathology may be more likely to experience withdrawal symptoms
(Breslau et al., 1992) provides insight into a potential reason for the significance of this
relationship. By examining this mediational pathway simultaneously across all significant links
72

between psychopathology and CPD, these results were able to show that this path was specific to
negative affect. Thus, this trait-consistent pattern indicates that this relationship may reflect not
only motivation to smoke to avoid or relieve withdrawal symptoms but also that individuals with
negative affect may smoke to self-medicate their specific psychopathological deficits (Gilbert &
Gilbert, 1995; Pomerleau et al., 2000).
In contrast to our hypothesis, motivation to smoke for positive reinforcement did not
mediate any links between psychopathology and CPD. In fact, it appears that this mediator
suppressed a stronger relation between low positive affect and CPD. Perhaps individuals with
low positive affect perceive experiencing low pleasure from smoking due to their inherent
diminished reward capacity and tendency to experience most positive reinforcers as less
pleasurable (Hasler, Drevets, Manji, & Charney, 2004; Pizzagalli, Jahn, & O'Shea, 2005);
however, some prior work examining the real-time effects of smoking has shown that anhedonic
individuals actually experience greater positive affect enhancement from nicotine (Cook, Spring,
& McChargue, 2007). Additional research examining whether low positive affect modulates
real-time effects of smoking on mood would clarify this. The suppressor effect showed that after
partialling out variation due to individuals with low positive affect who reported greater positive
reinforcement smoking, the relationship between low positive affect and CPD became even
stronger. Hence, it is clear that there are likely other important mediators underlying the link
between low positive affect and higher CPD. For example, it may be that individuals with low
positive affect are more strongly driven by primary dependence motives (e.g., automaticity,
craving, loss of control, tolerance) that are associated with heavier smoking patterns (Piper et al.,
2008). This should be addressed in future work.
73

Limitations and conclusions. The current study is not without its limitations. First, all
measures were self-report and therefore have several biases, such as recall and self-awareness, to
consider when interpreting results. Second, the outcome variable was based on a single item:
CPD. Although CPD is a variable of significant public health importance due to associations
with continued smoking behavior and quitting difficulties (Hyland et al., 2004; Nordstrom et al.,
2000; Thompson et al., 2003) and with numerous types of disease (Bagaitkar et al., 2008;
Bazzano et al., 2003; Cheng et al., 2000; Law et al., 1997), multi-item assessments of CPD over
several days would have been preferable to ensure greater accuracy in individual’s report of their
CPD. Also, other important and clinically relevant aspects of heavy smoking behavior (level of
nicotine dependence, years of smoking) should be considered in future research. Additionally,
these results are limited to the moderate to heavy end of the smoking level range. Although this
provides unique insight into important psychopathological variation across this upper range,
these results may not generalize to light smokers. Third, the cross-sectional design of this study
does not allow for temporal or causal conclusions. Fourth, although the current sample size is
acceptable for structural equation modeling, larger sample sizes are ideal and should be utilized
to validate the current findings. Additionally, although we included these particular dimensional
measures of psychopathology indices because they have illustrated associations with smoking,
any one of a number of different types, measures, and conceptualizations of psychopathology
could have been included that may have changed the structure of the model (e.g., indicators of
schizophrenia and personality disorders). Last, as this sample included adult smokers without
current psychiatric and substance dependence, it is unclear how these results will generalize to
adolescents and to smokers with more severe levels of psychopathology.
74

Limitations notwithstanding, to the best of our knowledge, this is the first study to use a
structural model of psychopathology to elucidate which aspects of psychopathology underlie
associations with heavier smoking. This type of analysis is critical because understanding which
aspects of psychopathology associate with CPD will aid in the development of more efficient
screening assessments of individual psychopathological variation in regular cigarette smokers.
These results suggest that evaluating underlying dimensions of negative and low positive affect,
instead of the many different manifest internalizing disorders, may be the most parsimonious and
informative way to assess psychopathological liabilities of heavier cigarette smoking. For
example, measures that assess temperament systems of positive affectivity, negative affectivity,
and disinhibition, such as Clark’s (1993) Schedule for Nonadaptive and Adaptive Personality
(SNAP), may offer an efficient and meaningful way to capture psychopathological variation that
is important for heavier smoking.
Relatedly, these results suggest that smoking cessation interventions that seek to alleviate
negative affect and enhance positive affect may be effective transdiagnostic strategies for
reducing cigarette consumption that are applicable for multiple forms of internalizing disorders.
For example, interventions that focus on increasing the ability to tolerate nicotine withdrawal and
negative affect and teach skills to reduce avoidance or escape of aversive internal states (Brown
et al., 2008) may be useful for many types of internalizing disorders comprised of negative
affect. Future work elucidating the mechanisms underlying the association between low positive
affect and CPD will help provide insight into transdiagnostic strategies that may offset this
particular link. Finally, these results highlight the possibility that externalizing disorders may
not uniformly predict greater CPD under all circumstances, and that their impact on CPD may be
more complex among moderate-to-heavy daily smokers, especially after taking into account
75

other forms of psychopathology. Overall, this line of research highlights the complexity of
explaining how a broad spectrum of different forms of psychopathology may drive heavier
smoking behavior and underscore the utility to examining multiple psychopathologies
simultaneously to attain the goal of reducing the public health burden of heavy smoking among
the psychologically vulnerable.

76

Table 6
Descriptive Statistics and Correlations of Manifest Psychopathology Indices, Reinforcement Smoking Scales, and Smoking Level

M (SD) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
1.SHS 5.28 (1.07) (.76)
2. CESD:ANH 0.72 (0.65) -.44† (.67)
3. MASQ:AD 52.58 (14.11) -.59† .56† (.89)
4. MASQ:GDD 17.60 (7.02) -.34† .30† .43† (.92)
5. MASQ:GDA 15.17 (5.52) -.27† .22† .31† .81† (.87)
6. MASQ:AA 21.66 (6.79) -.24† .19† .28† .67† .83† (.89)
7. AQR:PA 1.82 (1.06) -.16** .15* .12* .26† .31† .34† (.77)
8. ASRS 2.18 (0.64) -.22† .24† .33† .45† .53† .49† .36† (.92)
9. AUDIT 3.50 (4.86) -.06 .14* .01 .07 .04 .08 .24† .18* (.88)
10. MNRQ:NR 1.60 (0.64) -.07 .01 .08 .20† .27† .27† .08 .15* -.08 (.75)
11. MNRQ:PR 1.22 (.60) .13* -.06 -.13* .02 .04 .04 .04 -.06 .02 .37† (.81)
12. Cigs/Day 16.76 (6.85) -.13* .11* .13* .12* .11* .12* .03 -.07 -.18* .21† .28† ---
Note. N = 338. M (SD) = Mean (Standard Deviation). Cronbach’s alpha on the diagonal. SHS=Subjective Happiness Scale; CESD:ANH=Center for Epidemiologic Studies Depression-
Anhedonia Scale; MASQ=Mood and Anxiety Symptom Questionnaire: AD=Anhedonic Depression Scale, GDD=General Distress Depression Scale, GDD=General Distress Anxious Scale,
AA=Anxious Arousal Scale; AQR:PA=Aggression Questionnaire Revised-Physical Aggression Scale; ASRS=Adult ADHD Self-Report Scale; AUDI =Alcohol Use Disorders Identification
Test. MNRQ = Minnesota Nicotine Reinforcement Questionnaire: NR = Negative Reinforcement, PR = Positive Reinforcement; Cigs/Day = number of cigarettes smoked per day.
*p < .05, **p < .01, †p < .001.

77

Table 7
Direct, Indirect, and Total Effect Results from the Negative and Positive Reinforcement Models

Model 1:
Negative Reinforcement
Model 2:
Positive Reinforcement
Psychopathology Indices MNRQ:NR Cigs/Day MNRQ:PR Cigs/Day
Low Positive Affect

Direct Effect --- .21** --- .24**
Indirect Effect --- -.01 --- -.04*
Total Effect -.04 .20* -.17* .20*
Negative Affect
Direct Effect --- .13 --- .17
Indirect Effect --- .07** --- .03
Total Effect .25** .20 .13 .20
Disinhibition
Direct Effect --- -.27** --- -.25*
Indirect Effect --- .02 --- -.01
Total Effect .07 -.26* -.05 -.26*
AUDIT Residual
Direct Effect --- -.13** --- -.16**
Indirect Effect --- -.03 --- .01
Total Effect -.10 -.16** .05 -.15**
MNRQ:PR --- --- --- .24†
MNRQ:NR --- .27† --- ---
R
2
.09** .16† .03 .14†
Note. N = 338. All results are standardized betas. MNRQ = Minnesota Nicotine Reinforcement Questionnaire: NR = Negative
Reinforcement, PR = Positive Reinforcement; AUDIT=Alcohol Use Disorders Identification Test, Residual = portion of the
AUDIT that is independent from the latent factors. Cigs/Day = average number of cigarettes smoked per day. R
2
= amount of
variance explained in the index variable by all of the variables in the models. Significant findings are in bold.
*p < .05, **p < .01, †p < .001.

78

Figure 4. Final Model of Latent Factors and Residuals Predicting Smoking Level

Figure 4. Final model of significant paths from latent factors and residuals to cigarettes per day.
N=338. SHS=Subjective Happiness Scale; CESD:ANH=Center for Epidemiologic Studies
Depression-Anhedonia Scale; MASQ=Mood and Anxiety Symptom Questionnaire:
AD=Anhedonic Depression Scale, GDD=General Distress Depression Scale, GDA=General
Distress Anxious Scale, AA=Anxious Arousal Scale; AQR:PA=Aggression Questionnaire
Revised-Physical Aggression Scale; ASRS=Adult ADHD Self-Report Scale; AUDIT=Alcohol
Use Disorders Identification Test. All path betas are standardized. Significant relations between
psychopathology and cigarettes per day at p<.05 are in bold. Factor correlations and error terms
associated with each psychopathological symptom index are not shown.
⁰p < .10, *p < .05, **p < .01, †p < .001.

79

Figure 5. Psychopathology, Negative and Positive Reinforcement Smoking, and Smoking Level
Mediation Models

Figure 5. Model A: Negative reinforcement mediation model. Model B: Positive reinforcement
mediation model. N=338. SHS=Subjective Happiness Scale; CESD:ANH=Center for
Epidemiologic Studies Depression-Anhedonia Scale; MASQ=Mood and Anxiety Symptom
Questionnaire: AD=Anhedonic Depression Scale, GDD=General Distress Depression Scale,
GDA=General Distress Anxious Scale, AA=Anxious Arousal Scale; AQR:PA=Aggression
Questionnaire Revised-Physical Aggression Scale; ASRS=Adult ADHD Self-Report Scale;
AUDIT=Alcohol Use Disorders Identification Test. MNRQ = Minnesota Nicotine
Reinforcement Questionnaire: NR = Negative Reinforcement, PR = Positive Reinforcement. All
80

results are standardized. Significant relations between psychopathology, reinforcement scales,
and cigarettes per day at p<.05 are in bold. Factor correlations and error terms associated with
each psychopathological symptom index are not shown.
*p < .05, **p < .01, †p < .001.

81

Chapter 4
Psychopathology, Motivation to Resume Smoking, and the Mediating Effects of Nicotine
Withdrawal Symptoms
Abstract
The interrelation between psychopathology, tobacco withdrawal symptoms, and
persistent smoking can be obscured by the comorbidity across multiple forms of
psychopathology. To address this barrier, this study uses a structural model of nine
psychopathological symptom indices commonly associated with smoking and tests their relations
to withdrawal and smoking motivation. Adult daily smokers (N = 286) completed a baseline
session at which psychopathology was assessed and two counterbalanced lab sessions (16 hours
smoking abstinence and ad lib smoking), during which withdrawal symptoms and ability to delay
smoking in exchange for monetary reinforcement for up to 50 minutes, which may reflect a
motivational process underlying smoking persistence, were measured. A single second-order
factor of general psychopathological maladjustment identified in the model associated with more
severe nicotine withdrawal symptoms, which in turn inversely associated with smoking delay.
The first-order factors (low positive affect, negative affect, disinhibition) and the manifest
indicators provided little predictive power over and above the general psychopathological
maladjustment factor. Experimentally-manipulated abstinence did not moderate these relations.
These results suggest that a shared risk pathway may drive associations evidenced across many
different manifest psychopathological syndromes to tobacco withdrawal and smoking
motivation, highlighting the potential for effective and efficient transdiagnostic smoking
prevention and cessation strategies.

82

Introduction
Although the health, economic, and social costs of cigarette smoking are well-recognized,
smoking rates remain disproportionately high among individuals with full-blown or sub-clinical
psychopathological syndromes, particularly the depressive, anxiety, disruptive behavior, and
substance use disorders (Anda et al., 1990; Bunde & Suls, 2006; Degenhardt & Hall, 2001;
Kollins et al., 2005; Lasser et al., 2000; McCrae, Costa, & Bosse, 1978). One likely reason for
this relationship is that individuals with elevated psychopathology are more likely to maintain
daily smoking and have greater difficulties quitting (Anda et al., 1990; Cinciripini et al., 1995;
Lasser et al., 2000). Notably, associations with difficulties quitting have been found even at very
low levels of psychopathology (e.g., >2 depressive symptoms; Niaura et al., 2001) and among
individuals not diagnosed with the respective psychiatric disorder (Leventhal et al., 2008).
Cessation of chronic tobacco use often produces several physical and affective signs and
symptoms of nicotine withdrawal, such as anger, depression, decreased heart rate, and
restlessness (American Psychiatric Association, 1994; Hughes, Gust, Skoog, Keenan, &
Fenwick, 1991; Hughes & Hatsukami, 1986). These symptoms are thought to be due, at least in
part, to molecular or neurochemical adaptations (e.g., alterations in receptor number or function)
in many different systems (e.g., cholinergic, dopaminergic, serotonergic) resulting from chronic
nicotine administration (Eissenberg, 2004; Hughes, 2007; Watkins, Koob, & Markou, 2000).
One hypothesis for the link between psychopathology and persistent smoking (i.e., quitting
difficulties or maintenance of daily smoking) is that psychopathology gives rise to more severe
nicotine withdrawal (McClernon et al., 2011; Pomerleau et al., 2000; Weinberger et al., 2010;
Xian et al., 2005). In turn, more severe withdrawal may maintain ongoing smoking behavior or
undermine quit attempts (Chandra, Shiffman, Scharf, Dang, & Shadel, 2007; Piasecki, 2006) due
83

to the motivation to avoid or to suppress withdrawal symptoms (Baker, Piper, McCarthy,
Majeskie, & Fiore, 2004). Although a large body of research provides support for associations
between psychopathology, withdrawal symptoms, and persistent smoking, several important
aspects of these interrelationships require further investigation.
First, the majority of this research has examined these associations across independent
manifest syndromes, without accounting for the extensive rate of psychopathological
comorbidity (Clark et al., 1995). Considerable research indicates that underlying, shared features
of psychopathology likely account for psychopathological comorbidity (Brown & Barlow, 2009;
Krueger & Markon, 2006; Krueger, 1999). Failing to consider these shared features in analyses
may obscure the extent to which relations with smoking are due to components of
psychopathology that are shared across many different syndromes or to components of
psychopathology that are specific to a certain psychopathological syndrome. To address this
issue, structural models of psychopathology can be utilized.
Previously, we found that a set of nine manifest psychopathological symptom indicators,
reflective of many psychopathological constructs commonly associated with smoking (i.e., low
happiness, anhedonia, depression, anxiety, anxious arousal, ADHD symptoms, physical
aggression, and alcohol use disorder symptomatology), loaded onto three latent factors (see
Figure 2B, pg. 53) (Ameringer et al., in preparation). Based on Clark’s (2005) proposed three
factor model of psychopathology and personality, these latent factors may represent three
different biobehavioral temperament systems: 1) positive affectivity, an individual’s tendency to
experience a wide range of positive emotions, which reflects the strength of a behavioral
approach system aimed at obtaining reward and mainly inversely associates with depression
(Clark & Watson, 1991; Durbin et al., 2005), 2) negative affectivity, an individual’s tendency to
84

experience a wide range of aversive emotional states, which reflects the strength of the
behavioral avoidance system aimed at avoiding threat and underlies a broad range of
psychopathology (Clark & Watson, 1991; Krueger et al., 1996), and 3) disinhibition, an
individual’s lack of restraint in response to incoming stimuli, which underlies the externalizing
psychopathology (Krueger & Piasecki, 2002; Lynam et al., 2003). The three-factor model we
previously identified also illustrated that these three latent factors can be alternately expressed as
subfactors on a second-order factor (see Figure 6B). In this second-order model, there are two
levels of shared features: 1) the first-order factors: low positive affect, negative affect, and
disinhibition, which consist of psychopathological features that differentiate these three factors
from one another but that are shared among the manifest indicators that load on to each factor,
and 2) the second-order factor: general maladjustment, which consists of psychopathological
features shared across many differing forms of emotional and behavioral psychopathology.
The results obtained by incorporating this structural model into analyses can help clarify
which components of psychopathology drive associations with smoking. For example,
associations between the residuals of the manifest indicators (i.e., the portion of variance within
the indicator that is unique from the other indicators and does not covary with the latent factors)
and smoking would shed light on specific, exclusive aspects of certain psychopathological
disorders (e.g., somatic symptoms of anxiety separate from general tendency toward negative
affectivity) that may be particularly important for smoking. Associations between the first-order
factors and smoking may indicate that features of psychopathology shared among small groups
of manifest syndromes that tend to cluster together may directly associate with smoking,
irrespective of the particular specific components of each manifest syndrome (e.g., general
tendency toward negative affectivity separate from specific manifest symptoms, such as somatic
85

anxiety). From a conceptual perspective, evidence that the latent factors are prepotent correlates
of smoking over and above manifest residuals could suggest that underlying biobehavioral
temperament systems increase vulnerability to both smoking and manifest psychopathology and
drive the comorbidity between smoking and psychopathology more generally. It is also possible
that a second-order general maladjustment factor is the most powerful psychopathological
correlate of smoking over and above first-order factors and manifest residuals, which may
suggest that shared features found across many differing dimensions of psychopathology (e.g.,
depressive, anxious, disruptive behavior) directly associate with smoking. That is, it may be that
the simply the severity, but not necessarily the quality, of psychological disturbance is most
important with regards to explaining variance in smoking behavior.
It is also important to consider the phenomenological overlap between psychopathology
and tobacco withdrawal symptoms (e.g., irritability and anxiety are symptoms of
psychopathological disorders and the tobacco withdrawal symptom) in the relationship
evidenced between psychopathology and withdrawal. This overlap creates difficulties for
understanding whether smokers with psychopathology are: (1) experiencing more severe true
withdrawal (i.e., withdrawal symptoms arising from neuroadaptations caused by chronic nicotine
exposure and the discontinuation of that nicotine exposure), (2) experiencing a masking and
unmasking or exacerbation of their underlying psychopathology from nicotine use and
abstinence, respectively, or (3) consistently reporting higher levels of withdrawal symptoms,
regardless of their nicotine use, due to the symptomatic overlap between their underlying
psychopathology and withdrawal symptoms; or, some combination of these possibilities.
Prior research on psychopathology-withdrawal relations has often utilized retrospective
reports of withdrawal symptoms or has not compared withdrawal symptom reports between
86

abstinent and nonabstinent states, which make it difficult to disentangle these alternate
possibilities. Specifically, retrospective reports of withdrawal symptoms may be confounded by
recall bias, particularly if individuals are currently emotionally distressed (Shiffman et al., 1997).
Because memories are subject to influence by one’s present mood state (Teasdale & Fogarty,
1979), past reports of withdrawal symptoms may be exaggerated during times of emotional
distress, thereby creating a spurious relationship. Additionally, without directly comparing
withdrawal reports between abstinent and nonabstinent states, it is hard to distinguish whether
withdrawal symptom reports represent true abstinence-induced provocation in withdrawal
symptoms or baseline underlying psychopathological disturbance. The current study addresses
these issues by prospectively examining withdrawal symptoms in a within-subject experiment
wherein withdrawal symptoms were measured in both abstinent and non-abstinent conditions.
This creates the ability to test whether abstinence status moderates the association between
psychopathology and withdrawal, which can help identify whether individuals with
psychopathology report more severe withdrawal-like symptoms based on tobacco use.
Last, it remains unclear in this research whether more severe withdrawal symptoms
reported in individuals with psychopathology in turn contribute to persistent smoking. Although
considerable research has examined links between psychopathology and withdrawal and between
withdrawal and persistent smoking behavior; relatively little research has examined this full
mediational pathway and those that have found inconsistent results (Cinciripini et al., 2003;
Doran, Spring, McChargue, Pergadia, & Richmond, 2004; Thorndike et al., 2008). Notably, each
of these studies included one or more intervention components, which may have disrupted the
natural experience of withdrawal. Furthermore, as each of these studies used smoking relapse as
the outcome, it remains unclear how withdrawal may maintain day-to-day smoking among
87

individuals with psychopathology not attempting to quit. To extend this research, the current
study examines whether withdrawal symptoms mediate the link between psychopathology and
ability to delay smoking in exchange for monetary reinforcement (McKee et al., 2006). In this
human laboratory paradigm, the ability to delay smoking represents a proximal, objective
assessment of initial lapse behavior (McKee, 2009), which is particularly important as lapse is a
strong predictor of relapse (Brandon et al., 1990; Kenford et al., 1994). The ability to delay
smoking may also closely measure ongoing patterns of smoking following short periods of
abstinence and thus may provide insight into the maintenance of day-to-day smoking among
those not attempting to quit.
To address these three important aspects of the interrelations among psychopathology,
withdrawal symptoms, and persistent smoking, we: 1) use a structural model of psychopathology
to elucidate the extent to which shared versus specific features of psychopathology associate
with ability to delay smoking, 2) test the mediational pathway between psychopathology,
withdrawal symptoms, and ability to delay smoking, and 3) examine whether experimentally-
induced abstinence moderates associations between psychopathology, withdrawal, and ability to
delay smoking. We carry out these aims using nine psychopathologic indicators shown to be
related with smoking characteristics in previous work (i.e., low happiness, anhedonia,
depression, anxiety, anxious arousal, ADHD symptoms, physical aggression, and alcohol use
disorder symptomatology; Kollins et al., 2005; Lasser et al., 2000; Lepper, 1998; Leventhal et
al., 2008; Nabi et al., 2010) and are common features of the anxiety, depressive, disruptive
behavior, and substance use disorders, which are strongly comorbid with smoking (Lasser et al.,
2000). Based on prior research demonstrating stronger relations between the shared (vs.
specific) features of psychopathology and alcohol dependence (Kushner et al. 2011), we
88

hypothesize that shared (vs. specific) features of psychopathology will primarily associate with a
reduced ability to delay smoking when it is advantageous to do so. Because low positive affect,
negative affect, and disinhibition have important differences in etiology, clinical correlates, and
genetic basis (Clark, 2005), we hypothesize that these first-order shared features of
psychopathology will incrementally associate with delay (e.g., above and beyond one another).
Second, due to the speculation that an individual’s underlying psychopathology symptoms may
become exacerbated upon abstinence (Pomerleau et al., 2000), we hypothesize that the relation
between psychopathology and withdrawal will be stronger when individuals are abstinent. Last,
based on research previously reviewed that illustrates links among psychopathology, withdrawal
symptoms, and abstinence difficulties, we hypothesize that withdrawal symptoms will mediate
the link between psychopathology and reduced ability to delay smoking.

Methods
Participants and Procedures
This study is a secondary analysis of a study designed to examine the influence of
individual differences in psychopathology on sensitivity to the effects of tobacco deprivation.
Participants were current adult smokers invited to participate in a study on personality and
smoking. Please see pages 26-27 for a more detailed description of the recruitment methods and
eligibility criteria. Participation involved: 1) an initial telephone screen of general inclusion
criteria, 2) a baseline session involving informed consent, breath CO analysis, psychiatric
screening interview by a trained research assistant to further assess eligibility, and measures of
psychopathology and smoking, and 3) if eligible, two in-person overnight tobacco abstinent and
nonabstinent experimental sessions (order counterbalanced) that started at 12pm.
89

For their experimental sessions, participants were instructed to not smoke after 8pm the
day before for their abstinent session and to smoke normally for their nonabstinent session.
Procedures were equivalent across the two sessions except that participants smoked a cigarette of
their preferred brand at the start of their nonabstinent session, prior to CO assessment, to
standardize time from last cigarette whereas their abstinent session began with CO assessment.
Based on prior research and recommendations that a CO ≥ 10ppm indicates recent smoking
(Benowitz et al., 2002), participants’ with a CO ≥ 10ppm at their abstinent session could return
for a second attempt to complete their abstinent session. Those with a CO ≥ 10ppm on their
second attempt were discontinued (n = 6). Subsequently, participants completed measures of
nicotine withdrawal followed by other assessments. Approximately an hour after the start of
their session, participants completed a smoking lapse analogue task (described below) that
consists of a 0-50 minute delay period, a one hour self-administration period, and a rest period
that ended 2 hours and 50 minutes after the start of the delay period. Compensation was ~$200
for completing the entire study. The study was approved by the University of Southern
California’s Institutional Review Board.
Of the 515 participants enrolled in the study after passing the telephone screen, 165
individuals were found ineligible at the baseline session (primary reasons included low CO (n =
103) and a current psychiatric disorder (n = 39). Of the 350 eligible participants, 58 dropped
from the study after entry and 6 twice failed to meet abstinence criteria at the abstinence session.
This left a final sample of 286 that was included for analyses.

90

Measures
Baseline session measures.
Personal Information Questionnaire (PIQ). The PIQ is a self-report questionnaire that
collects sociodemographic information (e.g., age, gender, ethnicity).
Structured Clinical Interview for DSM-IV-Axis I Disorders, Research Version, Non-
Patient Edition (SCID-I/NP; First et al., 2002). The SCID-I/NP is a well-established interview to
evaluate psychiatric diagnoses and was administered to assess lifetime and current prevalence of
key Axis I disorders (i.e., lifetime psychotic disorder, current mood disorder, current
[hypo]manic disorder, past year non-nicotine substance dependence) for eligibility purposes.
Fagerström Test of Nicotine Dependence (FTND; Heatherton et al., 1991). The FTND is
a well-validated six-item measure of nicotine dependence severity.
Smoking History Questionnaire (SHQ; Brown et al., 2002). The SHQ measures general
information about smoking history and smoking patterns, including cigarettes/day, previous quit
attempts, age of smoking first cigarette, age of regular smoking, etc. The FTND and SHQ are
used in the current study to describe the sample and to investigate as potential confounders in the
relationships between psychopathology, withdrawal symptoms, and ability to delay smoking.
Psychopathology Indices.
The latent factors and residuals generated from the set of psychopathological symptom
indices represent the independent variables. Please see pages 27-31 for a description of these
variables.
Experimental session measures.
The Minnesota Nicotine Withdrawal Scale (MNWS; Hughes & Hatsukami, 1986). An
11-item variant of the MNWS, a widely used assessment of tobacco withdrawal that has been
91

previously validated several studies (Hughes, 1992; Hughes et al., 1991; Hughes & Hatsukami,
1986) was used to asses withdrawal symptoms. Participants were asked to rate a range of
withdrawal symptoms they experienced “so far today” on a scale of 0 (Never) to 5 (Severe). Due
to prior research suggesting craving is distinct from other withdrawal symptoms (Hughes &
Hatsukami, 1998), the symptom ‘craving for nicotine’ was eliminated from analyses. The
remaining withdrawal symptoms included were: irritable/angry, anxious/tense, difficulty
concentrating, restlessness, excessive hunger, physiological symptoms, increased eating,
drowsiness, and headaches. Participants received a composite mean withdrawal symptom score
across the ten items for both abstinent and nonabstinent sessions.
Time to First Cigarette: Relapse Analogue Task (McKee et al., 2006). The relapse
analogue task measures individual differences in the ability to delay smoking initiation when
delaying is monetarily reinforced. In the first part of this task (the delay period), participants
were presented with a box of eight cigarettes and were instructed that they could begin smoking
at any point during the next 50 minutes, but for each five minutes they delayed smoking they
could earn $0.20 up to a maximum of $2.00. The delay period ended when participants indicated
they wanted to smoke. For participants who chose not to smoke, the delay period ended at the
end of the 50 minute time period. Previous studies utilizing this type of task have found that
factors such as abstinence and individual differences (e.g., delay discounting, stress, nicotine
dependence severity) were able to predict differences in time of delay to first cigarette (Dallery
& Raiff, 2007; Leeman, O'Malley, White, & McKee, 2010; McKee, 2009; McKee et al., 2006;
Sweitzer, Denlinger, & Donny, 2013). After the delay period, participants began a self-
administration period in which they were told they had a $1.60 credit and could smoke as little or
as much as they wanted over the following 60 minutes, however, each cigarette would cost them
92

$0.20. Participants were told that after this hour, they would not be allowed to smoke until the
study session ended at 4:00pm. Because the primary interest was to examine psychopathological
variation in smoking persistence and initial lapse behavior, time delayed (continuous range of 0-
50 minutes) in the first part of the relapse analogue task serves as the primary outcome. Since
the relapse analogue task was completed at both sessions, participants have measures of abstinent
time delayed and nonabstinent time delayed.
Data Analysis
Preliminary analyses. Descriptive statistics were calculated for the following:
demographics, smoking characteristics (age onset, cigarettes per day, FTND), psychopathology
symptom indices, withdrawal symptoms (abstinent and nonabstinent), and length of delay to first
cigarette (abstinent and nonabstinent). To examine severity level of psychopathology and
representativeness of the current sample, mean and sum scores for each of the scales were
compared to other studies involving community and/or smoking samples and the proportion of
individuals who scored above established cut-off points on relevant scales was calculated (see
Measures section). Correlations and Cronbach alpha coefficients for all variables included the
structural models were computed. Independent t-tests were used to test for significant
differences on the psychopathological symptom indices between these who completed the study
and those who did not complete the study after enrollment, and paired t-tests were used to test for
differences on the experimental session measures between abstinent and nonabstinent conditions.
All preliminary analyses were performed in SAS v9.2 (Inc., 2009) and based on complete
non-missing data. Because certain psychopathology measures were added throughout the study,
there is a range of missing data across these measures. Complete N’s for each of the
93

questionnaires are (out of 286): SHS (275), CESD:ANH (286), MASQ (275), AQR:PA (238),
ASRS (220), AUDIT (178).
Primary analyses. All confirmatory factor analysis (CFA) and structural equation
models (SEM) were conducted in Mplus Version 6 (Muthen & Muthen, 1998-2011), based on
the analysis of covariance, and utilized maximum likelihood (ML) estimation to handle missing
data. ML estimation, recommended to deal with missing data in SEM (Allison, 2003; Enders &
Bandalos, 2001; Peters & Enders, 2002), maintains all observations for analysis by partitioning
patterns of missing observations into subsets and extracting important information from each
subset (Allison, 2003; Muthen & Muthen, 1998-2011). Therefore, the full sample of completers
(N = 286) was used for primary analyses. Because certain measures were skewed, ML
estimation with robust standard errors (MLR) was used for all analyses to account for potentially
problematic multivariate non-normality. Model fit evaluation was based on the following
criteria recommended by Hu & Bentler (1999): 1) a non-significant Chi-square goodness-of-fit
statistic (p-value >.05), 2) a comparative fit index (CFI) > .95, and 3) a root mean-square error of
approximation (RMSEA) < .06. Because MLR estimation was used, the chi-square goodness-of-
fit statistic was based on the Satorra-Benter scaled chi-square (S-B χ2; Satorra & Bentler, 1994)
and chi-square difference tests were based on the Satorra & Bentler scaled difference chi-square
statistic (Satorra & Bentler, 2001). All variables were previously standardized (M = 0, SD = 1)
prior to inclusion in the structural models.
Confirmatory factor analysis. Confirmatory factor analysis was first used to examine the
fit of the previously identified 3-factor first-order model and second-order model, as well as
factor loadings and factor correlations, in the sample of completers included in the present study.
94

Shared relations between psychopathology and length of delay. Next, regression paths
from each of the three first-order factors (low positive affect, negative affect, and disinhibition)
to delay were included in the model. These paths were first added separately to examine their
independent relationship with delay and then all paths were added simultaneously to examine
their unique relation with delay (i.e., the association between each factor and length of delay
after adjusting for the shared variance among the factors). If the factors independently
associated with delay but not uniquely (i.e., these associations dropped below significance when
all paths were added simultaneously to the model suggesting that a shared source of variance was
contribution to their relations with delay), the relationship between the second-order factor and
delay was tested and further analyses utilized the second-order factor.
Specific relations between psychopathology and length of delay. Once shared
associations between psychopathology and delay were established, paths from the residuals of
each manifest psychopathological indicator to length of delay were successively included in the
model to examine whether the residual of a particular indicator demonstrated a significant
relation to delay above and beyond the influence of the underlying factors. Residuals were
created following the steps provided by Muthen & Muthen (“Regression on Residual,” n.d.).
Only those paths from the latent factors or residuals to delay that were significant were included
in the final model for mediation analyses.
Withdrawal mediation model. After identifying all shared and specific relations between
psychopathology and length of delay, mean withdrawal symptom severity scores were included
in the models to examine their potential mediational role. Separate variables were included to
reflect withdrawal at each condition (abstinent and nonabstinent). The MODEL INDIRECT
95

command in Mplus was used to obtain the indirect (based on the product of coefficients method)
and total effects.
Abstinence condition interaction analyses. For all models previously described,
variables measured in both experimental sessions (i.e., length of delay and withdrawal
symptoms) were included in the same model. To test whether abstinence moderated any of the
identified relations among psychopathology, length of delay, and withdrawal symptoms, the
regression paths representing the same relationship across abstinent and nonabstinent conditions
were constrained to be equal. The same model was then rerun after releasing the constraints and
model fit improvement was assessed with the Satorra-Bentler scaled chi-square difference test.
If model fit significantly improved, suggesting a significant difference in the regression paths
and moderation by abstinence status (Muthen & Muthen, 1998-2011), these paths were left freed
in the model and interpreted as such. If model fit did not improve, implying that the two
regression paths were not significantly different suggesting no evidence of moderation by
abstinence status, these paths were constrained in the model and consequently interpreted.
Covariates. In the structural models, demographics (age, gender, race) and smoking
characteristics (age of smoking onset, FTND, cigarettes per day) were independently regressed
on withdrawal symptoms and delay and allowed to covary with the latent factors. None of the
demographics or smoking characteristics were associated with withdrawal symptoms or delay in
these models and yielded no substantive changes to the primary results. Therefore, these
variables were not included as covariates in the final models to reduced model complexity.
All primary results are reported as standardized beta weights (βs) and significance was
set at p < .05. All tests were two-tailed except for the Satorra-Bentler scaled chi-square tests,
which were one-tailed as they assessed improvement in model fit.
96

Results
Preliminary results.
On average, participants were 44.0 (SD = 10.6) years of age, 68% male, and were: 52%
African American, 34% Caucasian, and 14% other, with 15% identifying as Hispanic.
Participants smoked approximately 16.7 (SD = 7.0) cigarettes per day, had an FTND level of 5.3
(SD = 1.9), and started smoking regularly around the age of 19.5 (SD = 5.6). There were no
significant differences on any of the psychopathological symptom indices between study
completers and non-completers after enrollment. Carbon monoxide levels were significantly
different between nonabstinent (M = 27.98, SD = 12.84) and abstinent (M = 5.55, SD = 2.14)
conditions (T = -30.14, p < .0001) and abstinence also had a main effect on withdrawal
symptoms (T = 10.99, p < .0001) and length of delay (T = -11.32, p < .0001) (see Table 8 for
mean withdrawal symptom and delay scores in abstinent and nonabstinent conditions).
Table 8 presents descriptive statistics, Cronbach’s alpha coefficients, and
intercorrelations of all variables included in the structural models. Mean and sum scores across
most of the psychopathological symptom scales (SHS, CESD:ANH, MASQ scales, AQR:PA)
reflected low severity of psychopathology, high between-participant variability, and were
generally comparable to those found in studies of community and/or smoking samples (Gerevich
et al., 2007; Leventhal et al., 2008; Lyubomirsky & Lepper, 1999; Watson, Clark, et al., 1995).
The percentage of those who scored above established clinically relevant cut-off points on
questionnaires were similar or slightly higher to prior community samples (Kessler et al., 2006;
Saunders et al., 1993) and were as follows: CESD:ANH = 33.6%, MASQ:AD = 55.3%,
MASQ:GDD = 18.6%, MASQ:GDA = 12.7%, MASQ:AA = 9.1%, ASRS = 6.4%, and AUDIT =
10.1% (see measures section for a description of cut-off points).
97

Primary results.
Confirmatory factor analysis. The first-order and second-order models of
psychopathology in the current study sample of completers are shown in Figure 6. Because the
second-order model is an alternate way of expressing the significant correlations among the first-
order factors, model fit for these two models is identical. Their fit was excellent: S-B χ
2
= 30.66,
p = .13; CFI = .989; RMSEA = .034. Although the factor loading for the AUDIT was a non-
significant trend (p < .10), we kept this indicator in the model because alcohol use disorder is a
theoretically important aspect of disinhibition (Clark, 2005), the model maintained excellent fit,
and the modification indices output did not suggest that the AUDIT should load onto any other
factor. All other indicators significantly loaded onto their respective factors. In the second-order
model, each of the three first-order factors significantly loaded onto the second-order factor.
Shared relations between psychopathology and length of delay. When entered
independently in the models, low positive affect and disinhibition each significantly associated
with reduced delay (p’s < .05), and negative affect associated with reduced delay at a non-
significant trend level (p < .10). However, when paths from low positive affect, negative affect,
and disinhibition were simultaneously added to the model, all associations reduced well below
significance (p’s > .10), indicating that each factor did not incrementally associate with delay
above and beyond one another. Thus, we tested the relation between the second-order factor and
delay and found a significant inverse relationship (p < .05), such that more severe general
psychological maladjustment on the second-order factor associated with reduced ability to delay
smoking. Abstinence status did not moderate the relationship between the second-order factor
and delay (i.e., model fit did not significantly improve when paths from this factor to delay
98

abstinent and to delay nonabstinent were freed); thus, these pathways were constrained and
consequently interpreted in the model.
Specific relations between psychopathology and length of delay. When the residuals
from each of the indicators were individually added to the model, the MASQ:GDD was the only
indicator to significantly predict delay above and beyond the second-order factor. Model fit
improved significantly when the paths from MASQ:GDD to delay abstinent and to delay
nonabstinent were freed, indicating that these paths were significantly different (i.e., moderation
by abstinent status). Upon freeing these pathways, higher MASQ:GDD illustrated a significant
positive association with longer delay when abstinent but did not associate with length of delay
when nonabstinent, thus this later path was dropped from the final model. Inclusion of this path
did not reduce the strength of the inverse association between the second-order factor and delay.
The final model of significant paths between latent factors and residuals to delay retained for the
mediation model is presented in Figure 7, which had excellent fit (S-B χ
2
= 43.06, p = .30; CFI =
.995; RMSEA = .019).
Withdrawal mediation model. In the mediation model, the second-order factor of general
maladjustment significantly associated with more severe withdrawal symptoms in both abstinent
and nonabstinent conditions. In turn, more severe withdrawal symptoms inversely associated
with length of delay in both abstinent and nonabstinent conditions (see Figure 8). Withdrawal
symptom severity significantly mediated the association between general maladjustment and
length of delay, as the indirect effect of general maladjustment on delay was significant (Table 9)
and the association between general maladjustment and delay dropped below significance in the
model that included withdrawal paths. Abstinent status did not moderate any of these relations
(i.e., psychopathology to withdrawal and withdrawal to delay). The MASQ:GDD did not
99

significantly associate with withdrawal symptoms and maintained significance with delay,
indicating that withdrawal symptoms did not mediate this relationship. All indirect, direct, and
total effects in the mediation model can be found in Table 9. This final model demonstrated
excellent fit (S-B χ
2
= 60.82, p = .37; CFI = .997; RMSEA = .013).

Discussion
Shared and specific relations between psychopathology and length of delay. The
current study extends the literature on the interrelations among psychopathology, withdrawal
symptoms, and persistent smoking behavior in several important ways. In contrast to
hypotheses, the use of structural equation modeling revealed that the second-order factor
(general maladjustment) was most important and meaningful for explaining the relationship
between the psychopathological indicators in the model and length of delay. In other words,
because each first-order factor (low positive affect, negative affect, disinhibition) did not
incrementally associate with delay above and beyond one another, this suggests that the effects
of the first-order factors on delay can best be explained by a second-order factor that represents
the shared variance of the three first-order factors. This finding indicates that it may be the
overall severity of psychopathology, rather than the type or quality of psychopathology, which is
prepotent in accounting for variation in willingness to delay smoking. Furthermore, the residuals
of each of the psychopathological manifest indicators, representing the specific features of
psychopathology, primarily demonstrated little association with length of delay above and
beyond the influence of the second-order factor. As such, it seems that shared features found
across nearly all types of emotional and behavioral psychopathology may directly influence
motivation to smoke, irrespective of the influence of any specific form of manifest syndrome or a
100

cluster of syndromes that may reflect a shared qualitatively unique dimension of
psychopathology (e.g., negative affectivity).
The majority of prior psychopathology-smoking research has focused on independent
manifest disorders, thus, knowledge of these widespread, shared features and their associations
with smoking warrants further consideration. The current findings illustrate the importance of
accounting for and considering shared features of psychopathology when analyzing and
interpreting previously established relations evidenced across a wide range of manifest
psychopathological syndromes and cessation difficulties in separate studies that have focused on
one or two psychopathological syndromes and their relation with smoking characteristics (Anda
et al., 1990; Cinciripini et al., 1995; Doran et al., 2004; Leventhal et al., 2008). Furthermore,
these results emphasize the need to better understand, identify, and measure features of
psychopathology that are shared across many forms of manifest emotional-behavior disturbance.
We also found that the second-order factor equally associated with ability to delay
smoking across abstinent and nonabstinent conditions. A reduced ability to delay smoking may
represent not only a proximal measure of initial lapse behavior following a quit attempt (McKee,
2009; McKee et al., 2006) but also motivation to smoke (1) following brief periods of abstinence
and (2) consistently throughout the day. Hence, the current findings may indicate that general
psychopathological maladjustment motivates smoking across many situations and levels of
abstinence, including undermining quit attempts and maintaining daily smoking behavior among
those not attempting to quit. Future research examining whether psychopathology influences
smoking motivation across these different contexts is needed to clarify the current findings.
One unexpected finding was that the MASQ:GDD residual associated with a longer
ability to delay smoking in the abstinent condition. Although the exact reasons for this
101

relationship are unclear, this finding highlights how associations with pure measures of
psychopathology may be masked until parceling out the variance shared with other
psychopathological indices. For example, perhaps there was something about this specific task
that was particularly motivating to delay smoking, such as the monetary reward or a type of
internal motivation, which associated with pure depressive-centered negative affect symptoms
but was not apparent until parceling out variance shared with other symptom indices. Future
work is necessary to examine whether this association translates to lapse and relapse following
naturalistic quit attempts.
Psychopathology and withdrawal. The results also illustrated that the general
maladjustment factor significantly associated with more severe self-report withdrawal symptoms
equally across abstinent and nonabstinent conditions. By prospectively examining withdrawal
symptoms in real-time and investigating reports of withdrawal in both abstinent and nonabstinent
conditions, these results provide important insight into the complex relationship between
psychopathology and withdrawal. Specifically, these findings could indicate that underlying
psychopathology directly impacts reports of withdrawal symptoms due to the symptomatic
overlap between the two entities. Perhaps, individuals with psychopathology consistently
experience these psychological and physiological withdrawal-like symptoms at elevated levels,
regardless of their abstinence, due to their underlying psychopathology (Gray, Baker, Carpenter,
Lewis, & Upadhyaya, 2010). An alternate explanation for these findings may be that these
individuals experience more severe withdrawal much more quickly, even within one hour since
smoking their last cigarette. This explanation is plausible as prior work has demonstrated that
certain symptoms of nicotine withdrawal can onset in as little as 30 minutes after the last
cigarette (Hendricks, Ditre, Drobes, & Brandon, 2006). Nonetheless, these findings illustrate the
102

importance of more thoroughly understanding the withdrawal symptom experience among
individuals with psychopathology.
Withdrawal mediation model. Clarifying the nature of the withdrawal experience
among individuals with psychopathology is also particularly important since these results
illustrated that withdrawal symptom severity mediated the link between psychopathology and
ability to delay smoking, in both abstinent and nonabstinent conditions. These findings suggest
that psychopathology experienced over prior longer periods (during the past week; such as on the
MASQ) or more generally as a trait (e.g., happiness, measured by the SHS) predicts self-report
of elevated acute psychological and physiological symptoms of withdrawal, regardless of
whether these symptoms are a result of abstinence, their underlying psychopathology, or some
alternative mechanism. In turn, the experience of these withdrawal-like symptoms influences
their ability to delay smoking, irrespective of abstinence status. Thus, it may be that individuals
with psychopathology are motivated to maintain or resume smoking to manage acute
withdrawal-like symptoms (Carmody, 1992; Pomerleau & Pomerleau, 1984) that could be
attributed to their underlying psychopathology and not abstinence-provoked withdrawal per se.
Limitations and conclusions. Several limitations should be considered when interpreting
the present results. First, the particular nine psychopathological symptom indices were included
in the structural model because they represent important components of psychopathological
syndromes (depressive, anxious, disruptive behavior, substance use) that have consistently
demonstrated relations with smoking (Anda et al., 1990; Bunde & Suls, 2006; Degenhardt &
Hall, 2001; Kollins et al., 2005; Lasser et al., 2000; McCrae et al., 1978); however, many
different types, measures, and conceptualizations of psychopathology could have been included
(e.g., schizophrenia and personality disorder indicators), that may have changed the model
103

structure and subsequent results. Individuals who were currently on psychiatric medications or
who met criteria for a current mood disorder or substance dependence were excluded to reduce
the potential effects of these factors on nicotine withdrawal. Thus, although the current results
provide insight into important subclinical variation of psychopathology in relation to smoking,
this leaves unclear how these results will generalize to more severe levels of psychopathology in
this population. Third, the self-report nature of the psychopathology and withdrawal symptom
measures lends itself to possible biases, such as limited self-awareness and/or minimization or
exaggeration of these symptoms. Furthermore, withdrawal symptoms were only measured at one
time point, which leaves unclear the relation between psychopathology and withdrawal across
different time points and lengths of abstinence. Participants in this study were not attempting to
quit and the delay outcome only assesses one potential aspect of persistent smoking behavior, at
one specific time. Future research is needed to examine whether these results translate to
individuals attempting to quit and to naturalistic quit attempts. As the current study utilized a
model with many parameters relative to sample size, research that incorporates larger, more ideal
samples sizes for this type of analysis is needed to validate the current results. Last, we opted to
maintain the significance level at p < .05 to provide a broad picture of the interrelations of
psychopathology, withdrawal, and ability to delay smoking utilizing a structural model of
psychopathology; as such, these findings are subject to Type I error due to the multiple tests run.
Despite these limitations, to our knowledge this is the first study to use a structural model
to examine the interrelations among psychopathology, withdrawal, and persistent smoking
behavior. These results point to a potential shared risk pathway across many different types of
psychopathology, withdrawal, and ability to delay smoking when it is advantageous to do so.
This emphasizes the need for future research to identify and to create meaningful assessments of
104

shared features of psychopathology found across many types of manifest emotional and
behavioral syndromes. This study also uniquely benefitted from the use of prospective
withdrawal measures across abstinent and nonabstinent conditions. Results illustrated that
general maladjustment significantly associated with withdrawal symptoms across abstinent and
nonabstinent conditions, which in turn led to a reduced ability to delay smoking. Thus,
individuals with underlying psychopathology may require enhanced treatment, beyond typical
nicotine replacement therapy, during a cessation attempt as their withdrawal symptom
experience may be confounded by their underlying psychopathological disturbance. For
example, interventions that combine traditional nicotine replacement therapy with strategies that
help individuals manage their reactions to and increase their tolerance of withdrawal-like
experiences (e.g., Brown et al., 2008) may prove beneficial for smokers with underlying
psychopathology.
105

Table 8
Descriptive Statistics and Correlations of Manifest Psychopathology Indices, Withdrawal Symptoms, and Time to First Cigarette
M (SD) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
1. SHS 5.27 (1.13) (.79)
2. CESD:ANH 0.72 (0.66) -.42† (.70)
3. MASQ:AD 52.94 (14.03) -.62† .55† (.90)
4. MASQ:GDD 17.68 (7.11) -.35† .28† .46† (.92)
5. MASQ:GDA 15.17 (5.51) -.28† .20† .35† .80† (.87)
6. MASQ:AA 21.68 (6.95) -.27† .19** .30† .68† .83† (.90)
7. AQR:PA 1.80 (1.04) -.20** .15* .13 .26† .30† .37† (.79)
8. ASRS 2.18 (0.64) -.29† .27† .38† .46† .54† .51† .36† (.92)
9. AUDIT 3.47 (4.98) -.05 .13 -.04 .03 -.01 .04 .19* .16* (.89)
10. MNWS:A 1.64 (1.15) -.11 .10 .18** .20† .29† .27† .19** .32† -.02 (.89)
11. MNWS:NA 0.94 (0.98) -.16** .21† .28† .37† .45† .40† .25† .45† .13 .49† (.90)
12. DELAY:A 23.93 (22.92) .05 -.05 -.07 .05 -.04 -.01 -.10 -.08 .10 -.23† -.17** ---
13. DELAY:NA 39.50 (17.63) .09 -.07 -.10 -.13* -.13* -.07 -.04 -.14* -.08 -.10 -.20** .36† ---
Note. N=286. M (SD) = Mean (Standard Deviation). Cronbach’s alphas are on the diagonal. SHS=Subjective Happiness Scale; CESD:ANH=Center for Epidemiologic Studies Depression-Anhedonia
Scale; MASQ=Mood and Anxiety Symptom Questionnaire: AD=Anhedonic Depression Scale, GDD=General Distress Depression Scale, GDA=General Distress Anxious Scale, AA = Anxious Arousal
Scale; AQR:PA=Aggression Questionnaire Revised-Physical Aggression Scale; ASRS=Adult ADHD Self-Report Scale; AUDIT=Alcohol Use Disorders Identification Test. MNWS = Minnesota
Nicotine Withdrawal Scale; Delay = length of time delayed to first cigarette in the smoking laboratory task (0-50 minutes); :A = Abstinent, :NA = Nonabstinent.
*p < .05, **p < .01, †p < .001.
106

Table 9
Direct, Indirect, and Total Effects from the Withdrawal Mediation Model

Psychopathology Indices
MNWS:
Abstinent
Delay:
Abstinent
MNWS:
Nonabstinent
Delay:
Nonabstinent
General Maladjustment

Direct Effects .52† -.05 .54† -.05
Indirect Effects --- -.09** --- -.09**
Total Effects --- -.14* --- -.14*
MASQ:GDD Residual
Direct Effects -.08 .17** --- ---
Indirect Effects --- .01 --- ---
Total Effects --- .19** --- ---
MNWS: Abstinent --- -.16** --- ---
MNWS: Nonabstinent --- --- --- -.17**
R
2
.27† .08** .30† .04*
Note. N = 286. MNWS = Minnesota Nicotine Withdrawal Scale; Delay = length of time delayed to first cigarette in the smoking
laboratory task (0-50 minutes); A = Abstinent; NA = Nonabstinent; General Maladjustment = Single second-order latent factor in
the structural model; MASQ:GDD = Mood and Anxiety Symptom Questionnaire: General Distress Depression Scale; Residual =
portion of the MASQ:GDD that is independent from the latent factors. R
2
= amount of variance explained in the index variable
by all of the variables in the final mediation model. All results are standardized. Significant findings are in bold.
*p < .05, **p < .01, †p < .001.

107

Figure 6. First-Order Three-Factor and Single Second-Order Factor Models of Psychopathology

Figure 6. Model A: First-order 3-factor model of psychopathology. Model B: Single second-order factor
model of psychopathology. N=286. SHS=Subjective Happiness Scale; CESD:ANH=Center for
Epidemiologic Studies Depression-Anhedonia Scale; MASQ=Mood and Anxiety Symptom Scale:
AD=Anhedonic Depression Scale, GDD=General Distress Depression Scale, GDA=General Distress
Anxious Scale, AA=Anxious Arousal Scale; AQR:PA=Aggression Questionnaire Revised-Physical
Aggression Scale; ASRS=Adult ADHD Self-Report Scale; AUDIT=Alcohol Use Disorders Identification
Test. Error terms and disturbances are not shown.
°p < .10, *p < .05, **p < .01, †p < .001.

108

Figure 7. Final Model of Latent Factors and Residuals Predicting Time to First Cigarette

Figure 7. Final model of significant paths from latent factors and residuals to delay. Only
residuals significantly associated with delay are shown in the model. N=286. SHS=Subjective
Happiness Scale; CESD:ANH=Center for Epidemiologic Studies Depression-Anhedonia Scale;
MASQ=Mood and Anxiety Symptom Scale: AD=Anhedonic Depression Scale, GDA=General
Distress Anxious Scale, AA=Anxious Arousal Scale; AQR:PA=Aggression Questionnaire
Revised-Physical Aggression Scale; ASRS=Adult ADHD Self-Report Scale; AUDIT=Alcohol
Use Disorders Identification Test. Delay=time to first cigarette. All results are standardized.
Error terms and disturbances are not shown.
*p < .05, **p < .01, †p < .001.

109

Figure 8. Psychopathology, Withdrawal, and Time to First Cigarette Mediation Model

Figure 8. Model of withdrawal symptom severity mediating links between psychopathology and
delay. N=286. SHS=Subjective Happiness Scale; CESD:ANH=Center for Epidemiologic Studies
Depression-Anhedonia Scale; MASQ=Mood and Anxiety Symptom Scale: AD=Anhedonic
Depression Scale, GDA=General Distress Anxious Scale, AA=Anxious Arousal Scale;
AQR:PA=Aggression Questionnaire Revised-Physical Aggression Scale; ASRS=Adult ADHD
Self-Report Scale; AUDIT=Alcohol Use Disorders Identification Test. All results are
standardized. Error terms and disturbances are not shown.
*p < .05, **p < .01, †p < .001.

110

Chapter 5
Conclusion
Growing research in the field of psychology evidences that extensive psychopathological
comorbidity may best be explained by a smaller number of latent liability constructs that underlie
multiple different forms of manifest psychopathology (Krueger & Markon, 2006). The main
purpose of this dissertation was to examine how this liability-spectrum conceptualization of
psychopathology can be used to address the challenges psychopathological comorbidity creates
for interpreting relations between psychopathology, smoking, and underlying mechanisms.
Overall, the studies in this dissertation supported this conceptualization of psychopathology and
demonstrated that latent liability constructs may account for many of the relations found across
several different types of psychopathology and heavy, persistent smoking behavior. These
findings have several important implications concerning the association between
psychopathology and smoking and provide guidance for future areas of research on this topic.
First, these results emphasize the importance of accounting for psychopathological
comorbidity when investigating specific psychopathology-smoking relations. Much of the
psychopathology-smoking literature to date has examined and conceptualized different forms of
psychopathology in isolation. This leaves unclear whether the many relations evidenced across
multiple different forms of manifest psychopathology and smoking were due to features specific
to particular forms of psychopathology or to features shared across multiple forms of
psychopathology. The first study in this dissertation demonstrated that three latent dimensions
(low positive affect, negative affect, disinhibition) accounted for the comorbidity across a wide
range of emotional and behavioral psychopathological indices commonly associated with
smoking. The make-up of this three factor model, which is in contrast to the two factor

111

(internalizing-externalizing) model primarily supported in the psychology literature, underscores
the problem of within-disorder heterogeneity and supports approaches using symptom-specific
indicators that do not combine heterogeneous manifestations of psychopathology. In the second
two dissertation studies, using this model in psychopathology-smoking relations elucidated the
extent to which relations with heavy and persistent smoking were due to shared versus specific
features of psychopathology. Taken together, these findings illustrate that structural models of
psychopathology may be a promising way to account for and examine the influence of
psychopathological comorbidity on specific psychopathology-smoking relations.
Second, this dissertation provides insight into potential etiological processes underlying
the link between psychopathology and heavy, persistent smoking. In the second two studies, the
residuals of the manifest indicators (i.e., “specific” features of psychopathology) demonstrated
little association with smoking variables above and beyond the influence of the underlying latent
factors (i.e., “shared” features of psychopathology). Specifically, the second dissertation study
illustrated that features shared within clusters of low positive affect (e.g., happiness, anhedonia)
and within clusters of high negative affect (e.g., depression, anxiety) psychopathology
syndromes primarily associated with heavier smoking. More broadly, the third dissertation study
demonstrated the features shared across nearly all types of emotional and behavioral
psychopathology most directly influenced motivation to smoke. Taken together, these findings
suggest that underlying core psychopathological processes (e.g., maladaptive temperament
systems) may increase vulnerability to both heavy, persistent smoking behaviors and manifest
psychopathology, whether to specific clusters of psychopathological syndromes or to
psychopathology in general, and largely drive the relationship between psychopathology and
smoking more generally (Gilbert & Gilbert, 1995). Furthermore, these findings indicate that

112

genetic and environmental (e.g. adverse life events, parenting style) determinants of poor
underlying psychopathological processes may be important, indirect risk factors for heavy,
persistent smoking. Future research examining these relations is needed to test these notions,
particularly because they may highlight important avenues for prevention efforts.
Third, the current findings provide knowledge on screening methods that may be efficient
and informative for assessing psychopathological variation in regular cigarette smokers. Overall,
results suggest that evaluating underlying core liabilities, such as low positive affect, negative
affect, and disinhibition temperament systems, instead of the many different manifest disorders,
may be the most parsimonious and informative way to assess psychopathological liabilities of
heavy and persistent cigarette smoking. This type of assessment may also be particularly
beneficial for tailored smoking cessation treatment planning because an individual would receive
dimensional scores on the multiple different underlying liability constructs. This has the
opportunity to provide more comprehensive information on an individual’s psychopathological
vulnerability to smoking compared to information only about whether an individual passes a
diagnostic cutpoint for a particular disorder. As much of the psychopathology-smoking literature
has focused on independent manifest disorders, future research is critical to better understand,
identify, and create valid assessments of underlying, core psychopathological processes that are
important for smoking.
One potential approach is to examine whether the latent constructs identified in the model
can be conceptualized and measured as manifest variables on currently available questionnaires.
For example, Clark’s (1993) Schedule for Nonadaptive and Adaptive Personality (SNAP) which
assesses the three temperament systems may be one prospect. A likely more promising approach
is to create new questionnaires or scoring algorithms that closely assess these latent constructs.

113

In other words, conducting studies that include large sample sizes and multiple
psychopathological assessments would help to identify which items across the psychopathology
measures most strongly load onto the latent constructs. This information can then be used to
create shorter questionnaires or scoring algorithms for weighting individual items that closely
capture these latent constructs. The growing support and research of underlying liability
constructs in the field of psychopathology will likely offer promising guidance for this research.
Fourth, the findings from this dissertation have potential important clinical implications.
These results support the notion that individuals with elevated psychopathology likely require
more intense, individualized cessation efforts to offset the link between psychopathology and
smoking. However, they indicate that transdiagnostic treatment efforts that focus on core,
underlying psychopathological vulnerabilities may be applicable for a broad range of different
forms of manifest psychopathology. Additionally, these findings illustrate that a much broader
range of the general public – those with underlying psychopathological liabilities and elevated
psychopathology, not necessarily only those with diagnostic levels of psychopathology – may
require and benefit from these targeted treatments. The mediational models in studies two and
three illustrated that targeting negative reinforcement processes, particularly those that focus on
withdrawal-like symptoms and experiences, may prove beneficial. Therefore, interventions that
focus on increasing the ability to tolerate nicotine withdrawal and negative affect and teach skills
to reduce avoidance or escape of aversive internal states (Brown et al., 2008) may be useful for
many different types and levels of psychopathology associated with persistent smoking.
Last, these results have implications for other addictive behaviors. Though limited,
growing research illustrates support for the importance of latent, shared features of
psychopathology in relation to other types of substance addictions. Kushner and colleagues

114

(2012) found that a latent internalizing factor associated with alcohol dependence whereas none
of the residuals of individual manifest internalizing disorders (e.g. panic disorder, major
depression) associated with alcohol dependence. Similarly, Jahng et al. (2011) illustrated that a
latent general personality disorder factor, underlying multiple DSM-IV personality disorders
(e.g., paranoid, antisocial behavior, avoidant), significantly associated with alcohol dependence,
nicotine dependence, and other drug dependence. Taken together, these studies and the current
findings demonstrate that the issues of psychopathological comorbidity and the presence of latent
shared factors of psychopathology are important to consider for all psychopathology-substance
addiction research. Although the current body of research is limited to substance-related
addictions, the pattern of findings indicate that it is likely similarly important to account for
underlying, shared psychopathological processes in relation to the behavioral addictions (e.g.,
gambling).
This dissertation has several limitations to consider when interpreting the current
findings. While there are many study specific limitations discussed in each section, four main
limitations are important for the overall dissertation. First, almost all measures of
psychopathology and smoking were self-report and are therefore subject to possible biases (e.g.,
desirability, recall, self-awareness). Second, the majority of analyses in this dissertation were
cross-sectional, which does not allow for causal or temporal conclusions of relations between the
measures of psychopathology and smoking characteristics. As the current studies indicate that
underlying psychopathological liabilities may directly impact smoking, prospective research
examining temperament, manifest psychopathology, and smoking would be particularly useful
for further clarifying the nature of the psychopathology-smoking relationship. Third, individuals
who were currently on psychiatric medications or who were suffering from a current mood

115

disorder episode or substance dependence were excluded from the adult daily smoker sample in
order to reduce the potential effects of these factors on nicotine withdrawal. Therefore, although
the current results illustrate important subclinical variation of psychopathology in relation to
smoking, it is unclear how these results will generalize to more severe levels of psychopathology
in this population.
Fourth, due to the relatively small sample sizes for structural equation modeling, this
dissertation only includes nine particular indices of psychopathology. Although these particular
indices were chosen (1) because they are components of psychopathology syndromes often
associated with smoking and (2) to allow all tested confirmatory factor analytic models an equal
chance, any number of different types and measures of psychopathology (e.g., symptoms of
schizophrenia and personality disorders) could have been included, which may have changed the
structure of the model and the subsequent results. Although it appears that a factor of
externalizing disinhibition, which underlies the spectrum of disruptive behavior and substance
use disorders, and one or two factors of internalizing affectivity, which underlie(s) the spectrum
of mood and anxiety disorders, emerge across many different types, measures, and
conceptualizations of psychopathology (Krueger & Markon, 2006; Krueger, 1999; Slade &
Watson, 2006), the current findings are limited to the specific emotional and behavioral
psychopathological indicators included in the models. Future research incorporating more types
and measures of psychopathology is needed to better understand the number and quality of latent
factors underlying psychopathology and importantly, to examine the influence of comorbidity
across a broad range of psychopathology on the relations between different types of
psychopathology and smoking.

116

Despite these limitations, these studies are among the first to use a structural model
conceptualization of psychopathology to examine how multiple psychopathologies
simultaneously associate with smoking. Incorporating new approaches such as this to better
understand the psychopathology-smoking relation is critical to reduce the public health burden of
smoking among the psychologically vulnerable. Due to the early nature of this work, future
research is needed to validate and to extend the current findings. However, these studies indicate
that taking a more macroscopic and transdiagnostic approach to the psychopathology-smoking
relation by examining and targeting psychopathological features that associate with smoking but
that are shared among many different forms of psychopathology deserves further consideration.

117

Bibliography

Allen, J. P., Litten, R. Z., Fertig, J. B., & Babor, T. (1997). A review of research on the Alcohol
Use Disorders Identification Test (AUDIT). Alcoholism-Clinical and Experimental
Research, 21(4), 613-619. doi: 10.1097/00000374-199706000-00008
Allison, P. D. (2003). Missing data techniques for structural equation modeling. Journal of
Abnormal Psychology, 112(4), 545-557. doi:10.1037/0021-843x.112.4.545
Ameringer, et. al. (in preparation). Shared versus specific features of psychopathology and
smoking level: Structual relations and mediation by positive and negative reinforcement
mechanisms.
Ameringer, K. J., & Leventhal, A. M. (2012). Symptom dimensions of attention deficit
hyperactivity disorder and nicotine withdrawal symptoms. Journal of Addictive Diseases,
31(4), 363-375. doi: 10.1080/10550887.2012.735568
Anda, R. F., Williamson, D. F., Escobedo, L. G., Mast, E. E., Giovino, G. A., & Remington, P.
L. (1990). Depression and the dynamics of smoking. A national perspective. Journal of
the American Medical Association, 264(12), 1541-1545.
American Psychiatric Association. (2000) Diagnostic and statistical manual of mental disorders
(4
th
ed., text rev.). Washington, DC: Author.
Audrain-McGovern, J., Rodriguez, D., Tercyak, K. P., Neuner, G., & Moss, H. B. (2006). The
impact of self-control indices on peer smoking and adolescent smoking progression.
Journal of Pediatric Psychology, 31(2), 139-151. doi: 10.1093/jpepsy/jsi079
Audrain, J., Lerman, C., Gomez-Caminero, A., Boyd, N. R., & Orleans, C. T. (1998). The role of
trait anxiety in nicotine dependence. Journal of Applied Biobehavioral Research, 3, 29-
42.
Babor, T. F., Higgins-Biddle, J. C., Saunders, J. B., & Monteiro, M. G. (2001). AUDIT: The
Alcohol Use Disorders Identification Test: Guidelines for Use in Primary Health Care.
2
nd
ed. World Health Organization, Geneva.
Bagaitkar, J., Demuth, D. R., & Scott, D. A. (2008). Tobacco use increases susceptibility to
bacterial infection. Tobacco Induced Diseases, 4, 12. doi: 10.1186/1617-9625-4-12
Baker, T. B., Brandon, T. H., & Chassin, L. (2004). Motivational influences on cigarette
smoking. Annual Review of Psychology, 55, 463-491. doi:
10.1146/annurev.psych.55.090902.142054
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.
Psychological Review, 111(1), 33-51. doi: 10.1037/0033-295X.111.1.33

118

Bazzano, L. A., He, J., Muntner, P., Vupputuri, S., & Whelton, P. K. (2003). Relationship
between cigarette smoking and novel risk factors for cardiovascular disease in the United
States. Annals of Internal Medicine, 138(11), 891-897.
Benowitz, N. L., Jacob, I. P., Ahijevych, K., Jarvis, M. J., Hall, S., LeHouezec, J.,…Velicer W.
(2002). Biochemical verificiation of tobacco use and cessation. Nicotine & Tobacco
Research, 4(2), 149-159. doi: 10.1080/14622200210123581
Borders, A., Earleywine, M., & Jajodia, A. (2010). Could Mindfulness Decrease Anger,
Hostility, and Aggression by Decreasing Rumination? Aggressive Behavior, 36(1), 28-44.
doi: 10.1002/Ab.20327
Brandon, T. H., Tiffany, S. T., Obremski, K. M., & Baker, T. B. (1990). Postcessation cigarette
use: the process of relapse. Addictive Behaviors, 15(2), 105-114.
Breslau, N. (1995). Psychiatric comorbidity of smoking and nicotine dependence. Behavior
Genetics, 25(2), 95-101.
Breslau, N., Kilbey, M. M., & Andreski, P. (1992). Nicotine Withdrawal Symptoms and
Psychiatric-Disorders - Findings from an Epidemiologic-Study of Young-Adults.
American Journal of Psychiatry, 149(4), 464-469.
Breslau, N., Kilbey, M. M., & Andreski, P. (1993). Nicotine dependence and major depression.
New evidence from a prospective investigation. Archives of General Psychiatry, 50(1),
31-35.
Breslau, N., Novak, S. P., & Kessler, R. C. (2004). Psychiatric disorders and stages of smoking.
Biological Psychiatry, 55(1), 69-76.
Breslau, N., Peterson, E., Schultz, L., Andreski, P., & Chilcoat, H. (1996). Are smokers with
alcohol disorders less likely to quit? American Journal of Public Health, 86(7), 985-990.
Breslau, N., Peterson, E. L., Schultz, L. R., Chilcoat, H. D., & Andreski, P. (1998). Major
depression and stages of smoking. A longitudinal investigation. Archives of General
Psychiatry, 55(2), 161-166.
Brown, R. A., Lejuez, C. W., Kahler, C. W., & Strong, D. R. (2002). Distress tolerance and
duration of past smoking cessation attempts. Journal of Abnormal Psychology, 111(1),
180-185.
Brown, R. A., Niaura, R., Lloyd-Richardson, E. E., Strong, D. R., Kahler, C. W., Abrantes, A.
M., . . . Miller, I. W. (2007). Bupropion and cognitive-behavioral treatment for
depression in smoking cessation. Nicotine & Tobacco Research, 9(7), 721-730. doi:
10.1080/14622200701416955
Brown, R. A., Palm, K. M., Strong, D. R., Lejuez, C. W., Kahler, C. W., Zvolensky, M. J., . . .
Gifford, E. V. (2008). Distress tolerance treatment for early-lapse smokers: rationale,

119

program description, and preliminary findings. Behavior Modification, 32(3), 302-332.
doi: 10.1177/0145445507309024
Brown, S. L., Svrakic, D. M., Przybeck, T. R., & Cloninger, C. R. (1992). The relationship of
personality to mood and anxiety states: a dimensional approach. Journal of Psychiatric
Research, 26(3), 197-211.
Brown, T. A., & Barlow, D. H. (2009). A proposal for a dimensional classification system based
on the shared features of the DSM-IV anxiety and mood disorders: Implications for
assessment and treatment. Psychological Assessment, 21(3), 256-271. doi:
10.1037/A0016608
Brown, T. A., Chorpita, B. F., & Barlow, D. H. (1998). Structural relationships among
dimensions of the DSM-IV anxiety and mood disorders and dimensions of negative
affect, positive affect, and autonomic arousal. Journal of Abnormal Psychology, 107(2),
179-192. doi: 10.1037/0021-843x.107.2.179
Bryant, F. B., & Smith, B. D. (2001). Refining the architecture of aggression: A measurement
model for the Buss-Perry aggression questionnaire. Journal of Research in Personality,
35(2), 138-167. doi: 10.1006/jrpe.2000.2302
Bunde, J., & Suls, J. (2006). A quantitative analysis of the relationship between the Cook-
Medley Hostility Scale and traditional coronary artery disease risk factors. Health
Psychology, 25(4), 493-500. doi:10.1037/0278-6133.25.4.493
Burt, R. D., Dinh, K. T., Peterson, A. V., & Sarason, I. G. (2000). Predicting adolescent
smoking: A prospective study of personality variables. Preventive Medicine, 30(2), 115-
125. doi:10.1006/pmed.1999.0605
Buss, A. H., & Perry, M. (1992). The Aggression Questionnaire. Journal of Personality and
Social Psychology, 63(3), 452-459. doi:10.1037//0022-3514.63.3.452
Carmody, T. P. (1992). Affect regulation, nicotine addiction, and smoking cessation. Journal of
Psychoactive Drugs, 24(2), 111-122.
Carver, C. S. (2004). Negative affects deriving from the behavioral approach system. Emotion,
4(1), 3-22. doi: 10.1037/1528-3542.4.1.3
Centers for Disease Control and Prevention. (2002). Annual smoking-attributable mortality,
years of potential life lost, and economic costs - United States, 1995-1999. Morbidity and
Mortality Weekly Report, 51(14), 300-303.
Centers for Disase Control and Prevention. (2008). Smoking-attributable mortality, years of
potential life lost, and productivity losses - United States, 2000-2004. Morbidity and
Mortality Weekly Report, 57(45), 1226-1228.

120

Centers for Disease Control and Prevention. (2011). Current cigarette smoking among adults
aged >= 18 years - United States, 2005 - 2010. Morbidity and Mortality Weekly Report,
60(33), 1207-1212.
Centers for Disease Control and Prevention. (2012). Cigarette smoking among adults - United
States, 2011. Morbidity and Mortality Weekly Report, 61(44), 889-894.
Chandra, S., Shiffman, S., Scharf, D. M., Dang, Q., & Shadel, W. G. (2007). Daily smoking
patterns, their determinants, and implications for quitting. Experimental and Clinical
Psychopharmacology, 15(1), 67-80. doi: 10.1037/1064-1297.15.1.67
Cheng, A. C., Pang, C. P., Leung, A. T., Chua, J. K., Fan, D. S., & Lam, D. S. (2000). The
association between cigarette smoking and ocular diseases. Hong Kong Medical Journal,
6(2), 195-202.
Cinciripini, P. M., Lapitsky, L., Seay, S., Wallfisch, A., Meyer, W. J., & Vanvunakis, H. (1995).
A placebo-controlled evaluation of the effects of buspirone on smoking cessation -
differences between high-anxiety and low-anxiety smokers. Journal of Clinical
Psychopharmacology, 15(3), 182-191. doi: 10.1097/00004714-199506000-00006
Cinciripini, P. M., Wetter, D. W., Fouladi, R. T., Blalock, J. A., Carter, B. L., Cinciripini, L. G.,
& Baile, W. F. (2003). The effects of depressed mood on smoking cessation: mediation
by postcessation self-efficacy. Journal of Consulting and Clinical Psychology, 71(2),
292-301.
Clark, L. A. (1993). Schedule for nonadaptive and adaptive personality: Manual for
administration, scoring, and interpretation. Minneapolis, MN: University of Minnesota
Press.
Clark, L. A. (2005). Temperament as a unifying basis for personality and psychopathology.
Journal of Abnormal Psychology, 114(4), 505-521. doi: 10.1037/0021-843x.114.4.505
Clark, L. A., & Watson, D. (1991). Tripartite model of anxiety and depression: psychometric
evidence and taxonomic implications. Journal of Abnormal Psychology, 100(3), 316-336.
Clark, L. A., Watson, D., & Reynolds, S. (1995). Diagnosis and classification of
psychopathology - Challenges to the current system and future directions. Annual Review
of Psychology, 46, 121-153. doi:10.1146/annurev.psych.46.1.121
Cohen, L. M., McCarthy, D. M., Brown, S. A., & Myers, M. G. (2002). Negative affect
combines with smoking outcome expectancies to predict smoking behavior over time.
Psychology of Addictive Behaviors, 16(2), 91-97. doi:10.1037//0893-164x.16.2.91
Cook, J. W., Spring, B., & McChargue, D. (2007). Influence of nicotine on positive affect in
anhedonic smokers. Psychopharmacology (Berlinin), 192(1), 87-95. doi:
10.1007/s00213-006-0688-5

121

Copeland, A. L., Brandon, T. H., & Quinn, E. P. (1995). The smoking consequences
questionnaire adult: Measurement of smoking outcome expectancies of experienced
smokers. Psychological Assessment, 7(4), 484-494. doi: 10.1037/1040-3590.7.4.484
Costa, F. M., Jessor, R., & Turbin, M. S. (2007). College student involvement in cigarette
smoking: the role of psychosocial and behavioral protection and risk. Nicotine &
Tobacco Research, 9(2), 213-224. doi: 10.1080/14622200601078558
Costa, M. L., Cohen, J. E., Chaiton, M. O., Ip, D., McDonald, P., & Ferrence, R. (2010).
"Hardcore" definitions and their application to a population-based sample of smokers.
Nicotine & Tobacco Research, 12(8), 860-864. doi:10.1093/Ntr/Ntq103
Costello, E. J., Erkanli, A., Federman, E., & Angold, A. (1999). Development of psychiatric
comorbidity with substance abuse in adolescents: effects of timing and sex. Journal of
Clinical Child Psychology, 28(3), 298-311. doi: 10.1207/S15374424jccp280302
Dallery, J., & Raiff, B. R. (2007). Delay discounting predicts cigarette smoking in a laboratory
model of abstinence reinforcement. Psychopharmacology (Berlinin), 190(4), 485-496.
doi: 10.1007/s00213-006-0627-5
Degenhardt, L., & Hall, W. (2001). The relationship between tobacco use, substance-use
disorders and mental health: results from the National Survey of Mental Health and Well-
being. Nicotine & Tobacco Research, 3(3), 225-234. doi: 10.1080/14622200110050457
De Leon, J., & Diaz, F. J. (2005). A meta-analysis of worldwide studies demonstrates an
association between schizophrenia and tobacco smoking behaviors. Schizophrenia
Research, 76, 135-157.
Dierker, L. C., Avenevoli, S., Merikangas, K. R., Flaherty, B. P., & Stolar, M. (2001).
Association between psychiatric disorders and the progression of tobacco use behaviors.
Journal of the American Academy of Child and Adolescent Psychiatry, 40(10), 1159-
1167. doi: 10.1097/00004583-200110000-00009
Domino, E. F. (2001). Nicotine and tobacco dependence: normalization or stimulation? Alcohol,
24(2), 83-86.
Doran, N., Spring, B., McChargue, D., Pergadia, M., & Richmond, M. (2004). Impulsivity and
smoking relapse. Nicotine & Tobacco Research, 6(4), 641-647. doi:
10.1080/14622200410001727939
Durbin, C. E., Klein, D. N., Buckley, M. E., & Moerk, K. C. (2005). Temperamental
emotionality in preschoolers and parental mood disorders. Journal of Abnormal
Psychology, 114(1), 28-37. doi: 10.1037/0021-843x.114.1.28
Eissenberg, T. (2004). Measuring the emergence of tobacco dependence: the contribution of
negative reinforcement models. Addiction, 99 Suppl 1, 5-29. doi: 10.1111/j.1360-
0443.2004.00735.x

122

El-Guebaly, N., Cathcart, J., Currie, S., Brown, D., & Gloster, S. (2002). Smoking cessation
approaches for persons with mental illness or addictive disorders. Psychiatric Services,
53(9), 1166-1170.
Elkins, I. J., McGue, M., & Iacono, W. G. (2007). Prospective effects of attention-
deficit/hyperactivity disorder, conduct disorder, and sex on adolescent substance use and
abuse. Archives of General Psychiatry, 64(10), 1145-1152. doi:
10.1001/archpsyc.64.10.1145
Emery, S., Gilpin, E. A., Ake, C., Farkas, A. J., & Pierce, J. P. (2000). Characterizing and
identifying "Hard-Core" smokers: Implications for further reducing smoking prevalence.
American Journal of Public Health, 90(3), 387-394. doi:10.2105/Ajph.90.3.387
Enders, C. K., & Bandalos, D. L. (2001). The Relative Performance of Full Information
Maximum Likelihood Estimation for Missing Data in Structural Equation Models.
Structural Equation Modeling-a Multidisciplinary Journal, 8(3), 430-457. doi:
10.1207/S15328007sem0803_5
Fidler, J. A., Shahab, L., & West, R. (2011). Strength of urges to smoke as a measure of severity
of cigarette dependence: comparison with the Fagerstrom Test for Nicotine Dependence
and its components. Addiction, 106(3), 631-638. doi: 10.1111/j.1360-0443.2010.03226.x
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B.W. Structured clinical interview for
DSM-IV-TR Axis I disorders, research version, non-patient edition (SCID-I/NP).
Biometrics Research, New York State Psychiatric Institute, New York (2002).
Fresco, D. M., Heimberg, R. G., Mennin, D. S., & Turk, C. L. (2002). Confirmatory factor
analysis of the Penn State Worry Questionnaire. Behavior Research and Therapy, 40(3),
313-323.
Fuemmeler, B. F., Kollins, S. H., & McClernon, F. J. (2007). Attention deficit hyperactivity
disorder symptoms predict nicotine dependence and progression to regular smoking from
adolescence to young adulthood. Journal of Pediatric Psychology, 32(10), 1203-1213.
doi: 10.1093/jpepsy/jsm051
Gerevich, J., Bacskai, E., & Czobor, P. (2007). The generalizability of the Buss-Perry
Aggression Questionnaire. International Journal of Methods in Psychiatric Research,
16(3), 124-136. doi: 10.1002/Mpr.221
Gilbert, D. G. (1979). Paradoxical tranquilizing and emotion-reducing effects of nicotine.
Psychological Bulletin, 86(4), 643-661. doi: 10.1037/0033-2909.86.4.643
Gilbert, D. G., & Gilbert, B. O. (1995). Personality, psychopathology, and nicotine response as
mediators of the genetics of smoking. Behavior Genetics, 25(2), 133-147. doi:
10.1007/Bf02196923
Gilbert, D. G., Sharpe, J. P., Ramanaiah, N. V., Detwiler, F. R. J., & Anderson, A. E. (2000).
Development of a Situation x Trait Adaptive Response (STAR) model-based smoking

123

motivation questionnaire. Personality and Individual Differences, 29(1), 65-84. doi:
10.1016/S0191-8869(99)00178-6
Glautier, S. (2004). Measures and models of nicotine dependence: positive reinforcement.
Addiction, 99 Suppl 1, 30-50. doi: 10.1111/j.1360-0443.2004.00736.x
Grano, N., Virtanen, M., Vahtera, J., Elovainio, M., & Kivimaki, M. (2004). Impulsivity as a
predictor of smoking and alcohol consumption. Personality and Individual Differences,
37(8), 1693-1700. doi: 10.1016/j.paid.2004.03.004
Grant, B. F., Hasin, D. S., Chou, S. P., Stinson, F. S., & Dawson, D. A. (2004). Nicotine
dependence and psychiatric disorders in the United States: results from the national
epidemiologic survey on alcohol and related conditions. Archives of General Psychiatry,
61(11), 1107-1115. doi: 10.1001/archpsyc.61.11.1107
Gray, K. M., Baker, N. L., Carpenter, K. C., Lewis, A. L., & Upadhyaya, H. P. (2010).
Attention-Deficit/Hyperactivity Disorder Confounds Nicotine Withdrawal Self-Report in
Adolescent Smokers. The American Journal on Addictions, 19, 325-331.
Greenberg, J. B., Ameringer, K. J., Trujillo, M. A., Sun, P., Sussman, S., Brightman, M., . . .
Leventhal, A. M. (2012). Associations Between Posttraumatic Stress Disorder Symptom
Clusters and Cigarette Smoking. Psychology of Addictive Behaviors, 26(1), 89-98.
doi:10.1037/A0024328
Hasler, G., Drevets, W. C., Manji, H. K., & Charney, D. S. (2004). Discovering endophenotypes
for major depression. Neuropsychopharmacology, 29(10), 1765-1781. doi:
10.1038/sj.npp.1300506
Heatherton, T. F., Kozlowski, L. T., Frecker, R. C., & Fagerstrom, K. O. (1991). The Fagerstrom
Test for Nicotine Dependence: a revision of the Fagerstrom Tolerance Questionnaire.
British Journal of Addiction, 86(9), 1119-1127.
Heffner, J. L., Johnson, C. S., Blom, T. J., & Anthenelli, R. M. (2010). Relationship between
cigarette smoking and childhood symptoms of inattention and hyperactivity/impulsivity
in alcohol-dependent adults without attention-deficit hyperactivity disorder. Nicotine &
Tobacco Research, 12(3), 243-250. doi: 10.1093/Ntr/Ntp200
Heishman, S. J., Kleykamp, B. A., & Singleton, E. G. (2010). Meta-analysis of the acute effects
of nicotine and smoking on human performance. Psychopharmacology (Berlin), 210(4),
453-469. doi: 10.1007/s00213-010-1848-1
Hendricks, P. S., Ditre, J. W., Drobes, D. J., & Brandon, T. H. (2006). The early time course of
smoking withdrawal effects. Psychopharmacology (Berlin), 187(3), 385-396. doi:
10.1007/s00213-006-0429-9
Hicks, B. M., & Patrick, C. J. (2006). Psychopathy and negative emotionality: analyses of
suppressor effects reveal distinct relations with emotional distress, fearfulness, and anger-

124

hostility. Journal of Abnormal Psychology, 115(2), 276-287. doi: 10.1037/0021-
843X.115.2.276
Higgins, E. T., Shah, J., & Friedman, R. (1997). Emotional responses to goal attainment: strength
of regulatory focus as moderator. Journal of Personality and Social Psychology, 72(3),
515-525.
Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modeling-a
Multidisciplinary Journal, 6(1), 1-55. doi: 10.1080/10705519909540118
Hu, M. C., Davies, M., & Kandel, D. B. (2006). Epidemiology and correlates of daily smoking
and nicotine dependence among young adults in the United States. American Journal of
Public Health, 96(2), 299-308. doi: 10.2105/AJPH.2004.057232
Hughes, J., & Hatsukami, D. K. (1998). Errors in using tobacco withdrawal scale. Tobacco
Control, 7(1), 92-93.
Hughes, J. R. (1992). Tobacco withdrawal in self-quitters. Journal of Consulting and Clinical
Psychology, 60(5), 689-697.
Hughes, J. R. (2007). Effects of abstinence from tobacco: etiology, animal models,
epidemiology, and significance: a subjective review. Nicotine & Tobacco Research, 9(3),
329-339. doi: 10.1080/14622200701188927
Hughes, J. R., Gust, S. W., Skoog, K., Keenan, R. M., & Fenwick, J. W. (1991). Symptoms of
tobacco withdrawal. A replication and extension. Archives of General Psychiatry, 48(1),
52-59.
Hughes, J. R., & Hatsukami, D. (1986). Signs and symptoms of tobacco withdrawal. Archives of
General Psychiatry, 43(3), 289-294.
Hyland, A., Li, Q., Bauer, J. E., Giovino, G. A., Steger, C., & Cummings, K. M. (2004).
Predictors of cessation in a cohort of current and former smokers followed over 13 years.
Nicotine & Tobacco Research, 6 Suppl 3, S363-369.
SAS Institue Inc. (2009). The SAS System for Windows (Version 9.2). Cary, NC: SAS Institute
Inc.
Irvin, J. E., & Brandon, T. H. (2000). The increasing recalcitrance of smokers in clinical trials.
Nicotine & Tobacco Research, 2(1), 79-84.
Jahng, S., Trull, T. J., Wood, P. K., Tragesser, S. L., Tomko, R., Grant, J. D., & Bucholz, K.
(2011). Distinguishing general and specific personality disorder features and
implications for substance dependence comorbidity. Journal of Abnormal Psychology,
120(3), 656-669.

125

Johnson, K. A., Stewart, S. H., Zvolensky, M. J., & Steeves, D. (2009). Evaluating the mediating
role of coping-based smoking motives among treatment-seeking adult smokers. Nicotine
& Tobacco Research, 11(11), 1296-1303. doi: 10.1093/Ntr/Ntp134
Kandel, D. B., & Chen, K. (2000). Extent of smoking and nicotine dependence in the United
States: 1991-1993. Nicotine & Tobacco Research, 2(3), 263-274.
Kandel, D. B., Huang, F. Y., & Davies, M. (2001). Comorbidity between patterns of substance
use dependence and psychiatric syndromes. Drug and Alcohol Dependence, 64(2), 233-
241.
Kendler, K. S., Prescott, C. A., Myers, J., & Neale, M. C. (2003). The structure of genetic and
environmental risk factors for common psychiatric and substance use disorders in men
and women. Archives of General Psychiatry, 60(9), 929-937. doi:
10.1001/archpsyc.60.9.929
Kenford, S. L., Fiore, M. C., Jorenby, D. E., Smith, S. S., Wetter, D., & Baker, T. B. (1994).
Predicting smoking cessation. Who will quit with and without the nicotine patch. Journal
of the American Medical Association, 271(8), 589-594.
Kenney, B. A., & Holahan, C. J. (2008). Depressive symptoms and cigarette smoking in a
college sample. Journal of American College Health, 56(4), 409-414.
Kessler, R. C., Adler, L., Ames, M., Demler, O., Faraone, S., Hiripi, E., . . . Walters, E. E.
(2005). The World Health Organization Adult ADHD Self-Report Scale (ASRS): a short
screening scale for use in the general population. Psychological Medicine, 35(2), 245-
256.
Kessler, R. C., Adler, L., Barkley, R., Biederman, J., Conners, C. K., Demler, O., . . . Zaslavsky,
A. M. (2006). The prevalence and correlates of adult ADHD in the United States: results
from the National Comorbidity Survey Replication. American Journal of Psychiatry,
163(4), 716-723. doi: 10.1176/appi.ajp.163.4.716
Kessler, R. C., Chiu, W. T., Demler, O., Merikangas, K. R., & Walters, E. E. (2005). Prevalence,
severity, and comorbidity of 12-month DSM-IV disorders in the national comorbidity
survey replication. Archives of General Psychiatry, 62(7), 709-709.
King, A. C., & Epstein, A. M. (2005). Alcohol dose-dependent increases in smoking urge in light
smokers. Alcoholism-Clinical and Experimental Research, 29(4), 547-552. doi:
10.1097/01.Alc.0000158839.65251.Fe
Knight, R. G., Williams, S., McGee, R., & Olaman, S. (1997). Psychometric properties of the
Centre for Epidemiologic Studies Depression Scale (CES-D) in a sample of women in
middle life. Behavior Research and Therapy, 35(4), 373-380. doi: 10.1016/S0005-
7967(96)00107-6
Kollins, S. H., McClernon, F. J., & Fuemmeler, B. F. (2005). Association between smoking and
attention-deficit/hyperactivity disorder symptoms in a population-based sample of young

126

adults. Archives of General Psychiatry, 62(10), 1142-1147. doi:
10.1001/archpsyc.62.10.1142
Krueger, R. E., & Markon, K. E. (2006). Reinterpreting comorbidity: A model-based approach to
understanding and classifying psychopathology. Annual Review of Clinical Psychology,
2, 111-133. doi: 10.1146/annurev.clinpsy.2.022305.095213
Krueger, R. F. (1999). The structure of common mental disorders. Archives of General
Psychiatry, 56(10), 921-926. doi: 10.1001/archpsyc.56.10.921
Krueger, R. F., Caspi, A., Moffitt, T. E., & Silva, P. A. (1998). The structure and stability of
common mental disorders (DSM-III-R): A longitudinal-epidemiological study. Journal of
Abnormal Psychology, 107(2), 216-227. doi: 10.1037/0021-843x.107.2.216
Krueger, R. F., Caspi, A., Moffitt, T. E., Silva, P. A., & McGee, R. (1996). Personality traits are
differentially linked to mental disorders: A multitrait-multidiagnosis study of an
adolescent birth cohort. Journal of Abnormal Psychology, 105(3), 299-312. doi:
10.1037/0021-843x.105.3.299
Krueger, R. F., & Piasecki, T. M. (2002). Toward a dimensional and psychometrically-informed
approach to conceptualizing psychopathology. Behavior Research and Therapy, 40(5),
485-499.
Krukowski, R. A., Solomon, L. J., & Naud, S. (2005). Triggers of heavier and lighter cigarette
smoking in college students. Journal of Behavioral Medicine, 28(4), 335-345. doi:
10.1007/s10865-005-9003-x
Kushner, M. G., Wall, M. M., Krueger, R. F., Sher, K. J., Maurer, E., Thuras, P., & Lee, S.
(2012). Alcohol dependence is related to overall internalizing psychopathology load
rather than to particular internalizing disorders: evidence from a national sample.
Alcoholism-Clinical and Experimental Research, 36(2), 325-331. doi: 10.1111/j.1530-
0277.2011.01604.x
Lasser, K., Boyd, J. W., Woolhandler, S., Himmelstein, D. U., McCormick, D., & Bor, D. H.
(2000). Smoking and mental illness: A population-based prevalence study. Journal of the
American Medical Association, 284(20), 2606-2610.
Law, M. R., Morris, J. K., Watt, H. C., & Wald, N. J. (1997). The dose-response relationship
between cigarette consumption, biochemical markers and risk of lung cancer. British
Journal of Cancer, 75(11), 1690-1693.
Leeman, R. F., O'Malley, S. S., White, M. A., & McKee, S. A. (2010). Nicotine and food
deprivation decrease the ability to resist smoking. Psychopharmacology (Berlin), 212(1),
25-32. doi: 10.1007/s00213-010-1902-z
Lepper, H. S. (1998). Use of other reports to validate subjective wellbeing measures. Social
Indicators Research, 44, 367-369.

127

Lerman, C., Audrain, J., Orleans, C. T., Boyd, R., Gold, K., Main, D., & Caporaso, N. (1996).
Investigation of mechanisms linking depressed mood to nicotine dependence. Addictive
Behaviors, 21(1), 9-19. doi: 10.1016/0306-4603(95)00032-1
Leventhal, A. M., Ramsey, S. E., Brown, R. A., LaChance, H. R., & Kahler, C. W. (2008).
Dimensions of depressive symptoms and smoking cessation. Nicotine & Tobacco
Research, 10(3), 507-517. doi: 10.1080/14622200801901971
Lipkus, I. M., Barefoot, J. C., Williams, R. B., & Siegler, I. C. (1994). Personality measures as
predictors of smoking initiation and cessation in the UNC Alumni Heart-Study. Health
Psychology, 13(2), 149-155. doi: 10.1037/0278-6133.13.2.149
Lynam, D. R., Leukefeld, C., & Clayton, R. R. (2003). The contribution of personality to the
overlap between antisocial behavior and substance use/misuse. Aggressive Behavior,
29(4), 316-331. doi: 10.1002/Ab.10073
Lynskey, M. T., Fergusson, D. M., & Horwood, L. J. (1998). The origins of the correlations
between tobacco, alcohol, and cannabis use during adolescence. Journal of Child
Psychology and Psychiatry and Allied Disciplines, 39(7), 995-1005. doi:
10.1017/S0021963098002960
Lyubomirsky, S., & Lepper, H. S. (1999). A measure of subjective happiness: Preliminary
reliability and construct validiation. Social Indicators Research, 46, 137-155.
Markon, K. E. (2010). Modeling psychopathology structure: a symptom-level analysis of Axis I
and Axis II disorders. Psychological Medicine, 40, 273-288.
Masse, L. C., & Tremblay, R. E. (1997). Behavior of boys in kindergarten and the onset of
substance use during adolescence. Archives of General Psychiatry, 54(1), 62-68.
Mazzeo, S. E., & Espelage, D. L. (2002). Association between childhood physical and emotional
abuse and disordered eating behaviors in female undergraduates: An investigation of the
mediating role of alexithymia and depression. Journal of Counseling Psychology, 49(1),
86-100. doi: 10.1037//0022-0167.49.1.86
McClernon, F. J., Van Voorhees, E. E., English, J., Hallyburton, M., Holdaway, A., & Kollins, S.
H. (2011). Smoking withdrawal symptoms are more severe among smokers with ADHD
and independent of ADHD symptom change: Results from a 12-day contingency-
managed abstinence trial. Nicotine & Tobacco Research, 13(9), 784-792. doi:
10.1093/Ntr/Ntr073
McCrae, R. R., Costa, P. T., Jr., & Bosse, R. (1978). Anxiety, extraversion and smoking. British
Journal of Social and Clinical Psychology, 17(3), 269-273.
McKee, S. A. (2009). Developing human laboratory models of smoking lapse behavior for
medication screening. Addiction Biology, 14(1), 99-107. doi: 10.1111/j.1369-
1600.2008.00135.x

128

McKee, S. A., Krishnan-Sarin, S., Shi, J., Mase, T., & O'Malley, S. S. (2006). Modeling the
effect of alcohol on smoking lapse behavior. Psychopharmacology (Berlin), 189(2), 201-
210. doi: 10.1007/s00213-006-0551-8
McLeish, A. C., Zvolensky, M. J., Yartz, A. R., & Leyro, T. M. (2008). Anxiety sensitivity as a
moderator of the association between smoking status and anxiety symptoms and bodily
vigilance: Replication and extension in a young adult sample. Addictive Behaviors, 33(2),
315-327. doi: 10.1016/j.addbeh.2007.09.018
McMahon, R. J. (1999). Child and adolescent psychopathology as risk factors for smoking
initiation: an overview. Nicotine & Tobacco Research, 1(Suppl.), 45-50.
Morisano, D., Bacher, I., Audrain-McGovern, J., & George, T. P. (2009). Mechanisms
Underlying the Comorbidity of Tobacco Use in Mental Health and Addictive Disorders.
Canadian Journal of Psychiatry-Revue Canadienne De Psychiatrie, 54(6), 356-367.
Murray, J. B. (1991). Nicotine as a Psychoactive Drug. Journal of Psychology, 125(1), 5-25.
Muthen, L. K., & Muthen, B. O. (1998-2011). Mplus User's Guide. Sixth Edition. Los Angeles,
CA: Muthen & Muthen.
Muthen, L. K., & Muthen, B. O. (n.d.). "Regressing on a Residual" [PDF document] Retrieved
2013, from http://www.statmodel.com/faq.shtml
Nabi, H., Hall, M., Koskenvuo, M., Singh-Manoux, A., Oksanen, T., Suominen, S., . . . Vahtera,
J. (2010). Psychological and somatic symptoms of anxiety and risk of coronary heart
disease: the health and social support prospective cohort study. Biological Psychiatry,
67(4), 378-385. doi: 10.1016/j.biopsych.2009.07.040
Nguyen, H. T., Kitner-Triolo, M., Evans, M. K., & Zonderman, A. B. (2004). Factorial
invariance of the CES-D in low socioeconomic status African Americans compared with
a nationally representative sample. Psychiatry Research, 126(2), 177-187. doi:
10.1016/j.psychres.2004.02.004
Niaura, R., Britt, D. M., Shadel, W. G., Goldstein, M., Abrams, D., & Brown, R. (2001).
Symptoms of depression and survival experience among three samples of smokers trying
to quit. Psychology of Addictive Behaviors, 15(1), 13-17. doi: 10.1037//0893-
164x.15.1.13
Nichter, M., Nichter, M., Carkoglu, A., Lloyd-Richardson, E., & Tern. (2010). Smoking and
drinking among college students: "It's a package deal". Drug and Alcohol Dependence,
106(1), 16-20. doi: 10.1016/j.drugalcdep.2009.07.025
Nordstrom, B. L., Kinnunen, T., Utman, C. H., Krall, E. A., Vokonas, P. S., & Garvey, A. J.
(2000). Predictors of continued smoking over 25 years of follow-up in the normative
aging study. American Journal of Public Health, 90(3), 404-406. doi:
10.2105/Ajph.90.3.404

129

Orlando, M., Ellickson, P. L., & Jinnett, K. (2001). The temporal relationship between emotional
distress and cigarette smoking during adolescence and young adulthood. Journal of
Consulting and Clinical Psychology, 69(6), 959-970.
Patton, G. C., Carlin, J. B., Coffey, C., Wolfe, R., Hibbert, M., & Bowes, G. (1998). Depression,
anxiety, and smoking initiation: a prospective study over 3 years. American Journal of
Public Health, 88(10), 1518-1522.
Paulhus, D. L., Robins, R. W., Trzesniewski, K. H., & Tracy, J. L. (2004). Two replicable
suppressor situations in personality research. Multivariate Behavioral Research, 39(2),
303-328. doi: 10.1207/s15327906mbr3902_7
Payne, T. J., Ma, J. Z., Crews, K. M., & Li, M. D. (2013). Depressive symptoms among heavy
cigarette smokers: The influence of daily rate, gender, and race. Nicotine & Tobacco
Research. doi: 10.1093/ntr/ntt047
Peters, C. L. O., & Enders, C. (2002). A primer for estimation of structural equation models in
the presence of missing data: Maximum likelihood algorithms. Journal of Targeting,
Measurement, and Analysis for Marketing, 81-95.
Piasecki, T. M. (2006). Relapse to smoking. Clinical Psychology Review, 26(2), 196-215. doi:
10.1016/j.cpr.2005.11.007
Picciotto, M. R., Brunzell, D. H., & Caldarone, B. J. (2002). Effect of nicotine and nicotinic
receptors on anxiety and depression. Neuroreport, 13(9), 1097-1106.
Pierucci-Lagha, A., Gelernter, J., Feinn, R., Cubells, J. F., Pearson, D., Pollastri, A., . . .
Kranzler, H. R. (2005). Diagnostic reliability of the Semi-structured Assessment for Drug
Dependence and Alcoholism (SSADDA). Drug and Alcohol Dependence, 80(3), 303-
312. doi: 10.1016/j.drugalcdep.2005.04.005
Piper, M. E., Bolt, D. M., Kim, S. Y., Japuntich, S. J., Smith, S. S., Niederdeppe, J., . . . Baker,
T. B. (2008). Refining the tobacco dependence phenotype using the Wisconsin Inventory
of Smoking Dependence Motives. Journal of Abnormal Psychology, 117(4), 747-761.
doi: 10.1037/a0013298
Piper, M. E., Piasecki, T. M., Federman, E. B., Bolt, D. M., Smith, S. S., Fiore, M. C., & Baker,
T. B. (2004). A multiple motives approach to tobacco dependence: The Wisconsin
Inventory of Smoking Dependence Motives (WISDM-68). Journal of Consulting and
Clinical Psychology, 72(2), 139-154. doi: 10.1037/0022-006x.72.2.139
Pizzagalli, D. A., Jahn, A. L., & O'Shea, J. P. (2005). Toward an objective characterization of an
anhedonic phenotype: a signal-detection approach. Biological Psychiatry, 57(4), 319-327.
doi: 10.1016/j.biopsych.2004.11.026
Pomerleau, C. S., Downey, K. K., Snedecor, S. M., Mehringer, A. M., Marks, J. L., &
Pomerleau, O. F. (2003). Smoking patterns and abstinence effects in smokers with no

130

ADHD, childhood ADHD, and adult ADHD symptomatology. Addictive Behaviors,
28(6), 1149-1157.
Pomerleau, C. S., Marks, J. L., & Pomerleau, O. F. (2000). Who gets what symptom? Effects of
psychiatric cofactors and nicotine dependence on patterns of smoking withdrawal
symptomatology. Nicotine & Tobacco Research, 2(3), 275-280.
Pomerleau, C. S., Zucker, A. N., & Stewart, A. J. (2003). Patterns of depressive symptomatology
in women smokers, ex-smokers, and never-smokers. Addictive Behaviors, 28(3), 575-
582.
Pomerleau, O. F., Fagerstrom, K. O., Marks, J. L., Tate, J. C., & Pomerleau, C. S. (2003).
Development and validation of a self-rating scale for positive- and negative-
reinforcement smoking: The Michigan Nicotine Reinforcement Questionnaire. Nicotine
& Tobacco Research, 5(5), 711-718.
Pomerleau, O. F., & Pomerleau, C. S. (1984). Neuroregulators and the reinforcement of
smoking: towards a biobehavioral explanation. Neuroscience and Biobehavioral
Reviews, 8(4), 503-513.
Potter, A. S., & Newhouse, P. A. (2004). Effects of acute nicotine administration on behavioral
inhibition in adolescents with attention-deficit/hyperactivity disorder.
Psychopharmacology (Berlin), 176(2), 182-194. doi: 10.1007/s00213-004-1874-y
Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for research in the
general population. Applied Psychological Measurement 1, 385–401
Ranney, L., Melvin, C., Lux, L., McClain, E., & Lohr, K. N. (2006). Systematic review: smoking
cessation intervention strategies for adults and adults in special populations. Annals of
Internal Medicine, 145(11), 845-856.
Rigotti, N. A., Lee, J. E., & Wechsler, H. (2000). US college students' use of tobacco products:
results of a national survey. Journal of the American Medical Association, 284(6), 699-
705.
Rohde, P., Kahler, C. W., Lewinsohn, P. M., & Brown, R. A. (2004). Psychiatric disorders,
familial factors, and cigarette smoking: II. Associations with progression to daily
smoking. Nicotine & Tobacco Research, 6(1), 119-132. doi:
10.1080/14622200310001656948
Rohde, P., Lewinsohn, P. M., Brown, R. A., Gau, J. M., & Kahler, C. W. (2003). Psychiatric
disorders, familial factors and cigarette smoking: I. Associations with smoking initiation.
Nicotine & Tobacco Research, 5(1), 85-98. doi: 10.1080/1462220031000070507
Saitz, R., Palfai, T. P., Freedner, N., Winter, M. R., Macdonald, A., Lu, J., . . . Dejong, W.
(2007). Screening and brief intervention online for college students: the ihealth study.
Alcohol and Alcoholism, 42(1), 28-36. doi: 10.1093/alcalc/agl092

131

Satorra, A., & Bentler, P. M. (1994). Corrections to test statistics and standard errors in
covariance structure analysis. In Alexander von Eye and Clifford C. Cloggs (Eds.) ,
Latent Variable Analysis: Appliciations to Developmental Research (pp. 399-419).
Thousand Oaks, CA: Sage Publications Inc.
Satorra, A., & Bentler, P. M. (2001). A scaled difference chi-square test statistic for moment
structure analysis. Psychometrika, 66(4), 507-514. doi: 10.1007/Bf02296192
Saunders, J. B., Aasland, O. G., Amundsen, A., & Grant, M. (1993). Alcohol-Consumption and
Related Problems among Primary Health-Care Patients - Who Collaborative Project on
Early Detection of Persons with Harmful Alcohol-Consumption .1. Addiction, 88(3), 349-
362. doi: 10.1111/j.1360-0443.1993.tb00822.x
Schleicher, H. E., Harris, K. J., Catley, D., & Nazir, N. (2009). The role of depression and
negative affect regulation expectancies in tobacco smoking among college students.
Journal of American College Health, 57(5), 507-512. doi: 10.3200/JACH.57.5.507-512
Schulte-van Maaren, Y. W. M., Carlier, I. V. E., Zitman, F. G., van Hemert, A. M., de Waal, M.
W. M., van Noorden, M. S., & Giltay, E. J. (2012). Reference values for generic
instruments used in routine outcome monitoring: the leiden routine outcome monitoring
study. BMC Psychiatry, 12. doi:10.1186/1471-244x-12-203
Shafer, A. B. (2006). Meta-analysis of the factor structures of four depression questionnaires:
Beck, CES-D, Hamilton, and Zung. Journal of Clinical Psychology, 62(1), 123-146. doi:
10.1002/Jclp.20213
Shekelle, R. B., Gale, M., Ostfeld, A. M., & Paul, O. (1983). Hostility, Risk of Coronary Heart-
Disease, and Mortality. Psychosomatic Medicine, 45(2), 109-114.
Shiffman, S., Dunbar, M. S., Scholl, S. M., & Tindle, H. A. (2012). Smoking motives of daily
and non-daily smokers: a profile analysis. Drug and Alcohol Dependence, 126(3), 362-
368. doi: 10.1016/j.drugalcdep.2012.05.037
Shiffman, S., Hufford, M., Hickcox, M., Paty, J. A., Gnys, M., & Kassel, J. D. (1997).
Remember that? A comparison of real-time versus retrospective recall of smoking lapses.
Journal of Consulting and Clinical Psychology, 65(2), 292-300. doi: 10.1037/0022-
006x.65.2.292.A
Shiffman, S., Kassel, J. D., Paty, J., Gnys, M., & Zettler-Segal, M. (1994). Smoking typology
profiles of chippers and regular smokers. Journal of Substance Abuse, 6(1), 21-35.
Slade, T., & Watson, D. (2006). The structure of common DSM-IV and ICD-10 mental disorders
in the Australian general population. Psychological Medicine, 36(11), 1593-1600. doi:
10.1017/S0033291706008452
Starcevic, V., Uhlenhuth, E. H., Fallon, S., & Pathak, D. (1996). Personality dimensions in panic
disorder and generalized anxiety disorder. Journal of Affective Disorders, 37(2-3), 75-79.
doi: 10.1016/0165-0327(95)00058-5

132

Sweitzer, M. M., Denlinger, R. L., & Donny, E. C. (2013). Dependence and withdrawal-induced
craving predict abstinence in an incentive-based model of smoking relapse. Nicotine &
Tobacco Research, 15(1), 36-43. doi: 10.1093/ntr/nts080
Tate, J. C., Pomerleau, C. S., & Pomerleau, O. F. (1994). Pharmacological and non-
pharmacological smoking motives: a replication and extension. Addiction, 89(3), 321-
330.
Teasdale, J. D., & Fogarty, S. J. (1979). Differential effects of induced mood on retrieval of
pleasant and unpleasant events from episodic memory. Journal of Abnormal Psychology,
88(3), 248-257. doi: 10.1037//0021-843x.88.3.248
Thompson, B., Thompson, L. A., Thompson, J., Fredickson, C., & Bishop, S. (2003). Heavy
smokers: A qualitative analysis of attitudes and beliefs concerning cessation and
continued smoking. Nicotine & Tobacco Research, 5(6), 923-933. doi:
10.1080/14622200310001615277
Thorndike, A. N., Regan, S., McKool, K., Pasternak, R. C., Swartz, S., Torres-Finnerty, N., &
Rigotti, N. A. (2008). Depressive symptoms and smoking cessation after hospitalization
for cardiovascular disease. Archives of Internal Medicine, 168(2), 186-191. doi:
10.1001/archinternmed.2007.60
U.S. Department of Health and Human Services. Reducing the Health Consequences of
Smoking: 25 Years of Progress. A Report of the Surgeon General. Atlanta: U.S.
Department of Health and Human Services, Centers for Disease Control and Prevention,
National Center for Chronic Disease Prevention and Health Promotion, Office on
Smoking and Health, 1989. DHHS Publication No. (CD0 89-8411).
U.S. Department of Health and Human Services. The Health Consequences of Smoking: A
Report of the Surgeon General. Atlanta: U.S. Department of Health and Human Services,
Centers for Disease Control and Prevention, National Center for Chronic Disease
Prevention and Health Promotion, Office on Smoking and Health, 2004.
Vollebergh, W. A., Iedema, J., Bijl, R. V., de Graaf, R., Smit, F., & Ormel, J. (2001). The
structure and stability of common mental disorders: the NEMESIS study. Archives of
General Psychiatry, 58(6), 597-603.
Warner, K. E., & Burns, D. M. (2003). Hardening and the hard-core smoker: concepts, evidence,
and implications. Nicotine & Tobacco Research, 5(1), 37-48.
Watkins, S. S., Koob, G. F., & Markou, A. (2000). Neural mechanisms underlying nicotine
addiction: acute positive reinforcement and withdrawal. Nicotine & Tobacco Research,
2(1), 19-37.
Watson, D., Clark, L. A., Weber, K., Assenheimer, J. S., Strauss, M. E., & McCormick, R. A.
(1995). Testing a tripartite model: II. Exploring the symptom structure of anxiety and
depression in student, adult, and patient samples. Journal of Abnormal Psychology,
104(1), 15-25.

133

Watson, D., Gamez, W., & Simms, L. J. (2005). Basic dimensions of temperament and their
relation to anxiety and depression: A symptom-based perspective. Journal of Research in
Personality, 39(1), 46-66. doi: 10.1016/j.jrp.2004.09.006
Watson, D., Weber, K., Assenheimer, J. S., Clark, L. A., Strauss, M. E., & McCormick, R. A.
(1995). Testing a tripartite model: I. Evaluating the convergent and discriminant validity
of anxiety and depression symptom scales. Journal of Abnormal Psychology, 104(1), 3-
14.
Weinberger, A. H., Desai, R. A., & McKee, S. A. (2010). Nicotine withdrawal in U.S. smokers
with current mood, anxiety, alcohol use, and substance use disorders. Drug and Alcohol
Dependence, 108(1-2), 7-12. doi: 10.1016/j.drugalcdep.2009.11.004
Weiss, B., Susser, K., & Catron, T. (1998). Common and specific features of childhood
psychopathology. Journal of Abnormal Psychology, 107(1), 118-127. doi:
10.1037//0021-843x.107.1.118
Weitzman, E. R., & Chen, Y. Y. (2005). The co-occurrence of smoking and drinking among
young adults in college: National survey results from the United States. Drug and
Alcohol Dependence, 80(3), 377-386. doi: 10.1016/j.drugalcdep.2005.05.008
Weyandt, L. L., & DuPaul, G. (2006). ADHD in college students. Journal of Attention
Disorders, 10(1), 9-19. doi: 10.1177/1087054705286061
Winhusen, T. M., Somoza, E. C., Brigham, G. S., Liu, D. S., Green, C. A., Covey, L. S., . . .
Dorer, E. M. (2010). Impact of attention-deficit/hyperactivity disorder (ADHD) treatment
on smoking cessation intervention in ADHD smokers: a randomized, double-blind,
placebo-controlled trial. Journal of Clinical Psychiatry, 71(12), 1680-1688. doi:
10.4088/JCP.09m05089gry
Xian, H., Scherrer, J. F., Madden, P. A., Lyons, M. J., Tsuang, M., True, W. R., & Eisen, S. A.
(2005). Latent class typology of nicotine withdrawal: genetic contributions and
association with failed smoking cessation and psychiatric disorders. Psychological
Medicine, 35(3), 409-419.
Zvolensky, M. J., Jenkins, E. F., Johnson, K. A., & Goodwin, R. D. (2011). Personality disorders
and cigarette smoking among adults in the United States. Journal of Psychiatry Research,
45(6), 835-841.
Zvolensky, M. J., Stewart, S. H., Vujanovic, A. A., Gavric, D., & Steeves, D. (2009). Anxiety
sensitivity and anxiety and depressive symptoms in the prediction of early smoking lapse
and relapse during smoking cessation treatment.. Nicotine & Tobacco Research, 11(3),
323-331. doi: 10.1093/ntr/ntn037

Abstract (if available)

Abstract Extensive psychopathological comorbidity makes it difficult to interpret the nature and underlying mechanisms of psychopathology-smoking relations. Considerable research in the psychiatric literature illustrates that latent shared dimensions of psychopathology may best explain psychopathological comorbidity. Incorporating this conceptualization of psychopathology into psychopathology-smoking research may represent one promising way to refine the psychopathology-smoking link. This dissertation consists of three independent studies testing this novel approach. The first study establishes a meaningful latent factor model for a wide variety of manifest emotional and behavioral symptom- and syndrome-specific psychopathological indicators commonly associated with smoking (happiness, anhedonia, depression, anxiety, anxious arousal, ADHD symptoms, physical aggression, and alcohol use problems). The second study uses this model to test the extent to which shared features of psychopathology (latent dimensions) versus specific features of psychopathology (residuals of the manifest indicators) associate with smoking level, as well as the underlying role of positive and negative reinforcement motivations to smoke in these relations. Relatedly, the third study uses the structural model to test whether shared or specific features of psychopathology associate with ability to delay smoking in exchange for monetary reinforcement and whether withdrawal symptom severity mediates these relations. Taken together, the findings from this dissertation (1) emphasize the importance of accounting for psychopathological comorbidity when investigating specific psychopathology-smoking relations and (2) suggest that focusing on core psychopathological liabilities that associate with smoking but that are shared among many different forms of manifest psychopathology warrants further consideration.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

A network analysis of online and offline social influence processes in relation to adolescent smoking and alcohol use

PDF

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

PDF

Mixed methods investigation of user engagement with a smoking cessation app

PDF

Smoke-free housing policies and secondhand smoke exposure in low income multiunit housing in Los Angeles County

PDF

Cultural risk and protective factors for tobacco use behaviors and depressive symptoms among American Indian adolescents in California

PDF

A sociocultural and developmental approach to intimate partner violence among a sample of Hispanic emerging adults

PDF

Effects of parent stress on weight-related parenting practices and child obesity risk

PDF

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

PDF

Problematic alcohol use in Hispanic emerging adults: the role of perceived discrimination, cultural identity, and salient cultural values

PDF

The dynamic relationship of emerging adulthood and substance use

PDF

The role of depression symptoms on social information processing and tobacco use among adolescents

PDF

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

PDF

Affect, digital media use, physical activity, and ADHD in youth

PDF

Role transitions, past life events, and their associations with multiple categories of substance use among emerging adults

PDF

Influences of specific environmental domains on childhood obesity and related behaviors

PDF

Addressing federal pain research priorities: drug policy, pain mechanisms, and integrative treatment

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

Examining tobacco regulation opinions and policy acceptance among key opinion leaders and tobacco retailers in low socioeconomic status African American, Hispanic, and non-Hispanic White communities

PDF

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

Asset Metadata

Creator Ameringer, Katherine J. (author)

Core Title Using a structural model of psychopathology to distinguish relations between shared and specific features of psychopathology, smoking, and underlying mechanisms

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Preventive Medicine (Health Behavior)

Publication Date 10/07/2013

Defense Date 07/26/2013

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

Tag comorbidity,disinhibition,negative affect,OAI-PMH Harvest,positive affect,psychopathology,smoking

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Leventhal, Adam M. (committee chair), Chou, Chih-Ping (committee member), Sussman, Steven (committee member), Unger, Jennifer B. (committee member)

Creator Email kameringer1@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-335495

Unique identifier UC11297454

Identifier etd-AmeringerK-2085.pdf (filename),usctheses-c3-335495 (legacy record id)

Legacy Identifier etd-AmeringerK-2085.pdf

Dmrecord 335495

Document Type Dissertation

Format application/pdf (imt)

Rights Ameringer, Katherine J.

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