Internalizing-Externalizing Psychopathology and Personality

INTERNALIZING-EXTERNALIZING PSYCHOPATHOLOGY AND PERSONALITY

PATHOLOGY AS PREDICTORS OF TREATMENT REJECTION IN

SUBSTANCE USERS

Jonathan James Lewis

Thesis Prepared for the Degree of

MASTER OF SCIENCE

UNIVERSITY OF NORTH TEXAS

August 2013

APPROVED:

Craig Neumann, Major Professor Jennifer Callahan, Committee Member Thomas Parsons, Committee Member Vicki Campbell, Chair of the Department of Psychology Mark Wardell, Dean of the Toulouse Graduate School Lewis, Jonathan James. Internalizing-externalizing psychopathology and personality

pathology as predictors of treatment rejection in substance users. Master of Science

(Psychology), August 2013, 108 pp., 18 tables, 10 illustrations, references, 166 titles.

Substance use disorders (SUDs) are often comorbid with other psychopathology such as

mood disorders, anxiety disorders, and personality disorders. While some research suggests

individuals with comorbid psychopathology are more likely to seek substance use treatment than

those with independent disorders, other studies have also shown many individuals with dual

diagnoses still never seek treatment. Moreover, few studies have tried to elucidate the

underlying structure of SUD treatment rejection, and instead examined it in more simplistic

terms. In addition, studies have tended to examine the impact of individual disorders on

treatment rejection, but have not incorporated an empirically supported approach to

conceptualizing psychopathology in terms of comorbidity between broad latent dimensions

referred to as internalizing (e.g., depression, anxiety) and externalizing (e.g., antisocial personality disorder, polysubstance use) psychopathology. Modeling psychopathology in terms of internalizing and externalizing psychopathology is becoming a prominent approach to understanding mental disorders, yet little research to date has investigated the effects these broad dimensions have on SUD treatment rejection.

The current study utilized latent variable modeling techniques to (1) determine the latent

structure of SUD treatment rejection in a large U.S. sample, and investigate whether treatment

rejection is a multidimensional construct; and (2), to explore the ability of internalizing

psychopathology, externalizing psychopathology, and personality pathology to predict the SUD

treatment rejection factor(s). The current study relied on use of a general population sample of 43,093 individuals from the first wave of National Institute on Alcohol Abuse and Alcoholism’s

National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) study.

Support was found for the hypothesis that SUD treatment rejection would be a multidimensional construct. Exploratory structural equation modeling indicated a three-factor model best fit the data. Operational definitions and clinical implications of these three treatment rejection factors (“Objective barriers,” “Psychological barriers,” and “Self-focused barriers”) are discussed.

Among internalizing psychopathology, externalizing psychopathology, and personality pathology, structural equation modeling identified internalizing psychopathology as the most robust predictor of these three factors for alcohol treatment rejection (n = 1063), indicating endorsement of treatment barriers increased as levels of internalizing psychopathology increased.

This pattern also held true for externalizing psychopathology, while personality pathology only negatively predicted objective treatment barriers. For drug treatment rejection (n = 562), only internalizing psychopathology significantly predicted the treatment rejection factors, indicating treatment endorsement of drug treatment barriers increased as levels of internalizing psychopathology increased. Implications of these findings and directions for future research are discussed.

Jonathan James Lewis

TABLE OF CONTENTS

Page

LIST OF TABLES ...... vi

LIST OF ILLUSTRATIONS ...... viii

CHAPTER 1. INTRODUCTION ...... 1 Substance Use Disorders...... 3 Prevalence, Consequences, and Comorbidity of SUDs ...... 5 Internalizing Psychopathology and SUDs ...... 7 Comorbidity ...... 7 Treatment Rejection ...... 8 Externalizing Psychopathology and SUDs ...... 9 Comorbidity ...... 9 Treatment Rejection ...... 10 SUDs and Personality Disorders ...... 11 Comorbidity ...... 11 Treatment Rejection ...... 12 Addiction Theories...... 13 The Self-Medication Hypothesis ...... 13 Heritability Theories of SUDs ...... 14 SUD Treatment Rejection ...... 15 Motivation and Attitudes toward Treatment ...... 17 Conclusion ...... 19

CHAPTER 2. METHODS ...... 20 Participants ...... 20 Materials ...... 20 Data Analysis ...... 22 Descriptive Statistics ...... 22 Primary Data Analysis ...... 22 Research Questions and Hypotheses ...... 24 Research Question and Hypothesis 1: Which Aspects of Psychopathology are more Prevalent in Those Who Do Not Seek SUD Treatment? ...... 24

iii

Research Question and Hypothesis 2: Are Certain Disorders More Prevalent in Treatment Neglecters than Treatment Utilizers? ...... 24 Research Question and Hypothesis 3: How Strong is the Relationship between Treatment Rejection and Broad/Specific Composites among Individuals with a Specific SUD Diagnosis? ...... 25 Research Question and Hypothesis 4: Can Treatment Rejection be Understood in Terms of Underlying Dimensions? ...... 25 Research Question and Hypothesis 5: After Determining the Factor Structure of Treatment Rejection, What Variables Predict Specific Treatment Rejection Factors? ...... 26

CHAPTER 3. RESULTS ...... 27 Demographics ...... 27 Descriptive Statistics ...... 27 Internalizing Psychopathology...... 27 Externalizing Psychopathology ...... 28 PD Pathology ...... 28 Treatment Utilization and Rejection ...... 29 Prevalence of Psychopathology in Treatment Neglecters and Treatment Utilizers ...... 30 Prevalence of DSM-IV Disorders in Treatment Neglecters and Utilizers ...... 30 The Relationship between Treatment Neglect and Psychopathology ...... 32 The Factor Structure of Substance Use Disorder Treatment Rejection ...... 33 The Prediction of Substance Use Disorder Treatment Rejection ...... 35 Model Specification ...... 36 Alcohol Treatment Rejection SEM ...... 36 Drug Treatment Rejection SEM ...... 38

CHAPTER 4. DISCUSSION ...... 41 Do Treatment Neglecters Differ from Treatment Utilizers with Respect to Internalizing and Externalizing Psychopathology, Broad SUD Pathology, or PD Pathology? ...... 41 Internalizing Psychopathology...... 42 Externalizing Psychopathology ...... 44 Personality Pathology ...... 46 How does Comorbidity between Internalizing, Externalizing, and Personality Psychopathology Affect Treatment-Seeking Behavior? ...... 48 What is the Typology of Alcohol and Drug Treatment Rejection? ...... 50

What Variables Best Predict Treatment Rejection? ...... 53 Conclusions ...... 57 Limitations and Future Directions ...... 58

APPENDIX: NESARC ALCOHOL AND DRUG TREATMENT REJECTION ITEMS ...... 90

REFERENCES ...... 94

LIST OF TABLES

Page

1. Descriptive Statistics for Psychopathology and Treatment Rejection Composite Variables ...... 62

2. Demographics ...... 63

3. Descriptive Statistics – Prevalence of DSM-IV Disorders in Sample (N = 12,707) ...... 64

4. Type of SUD Treatment Obtained (n = 2212) ...... 65

5. Reasons for Treatment Rejection among Those with an AUD or DUD...... 66

6. Research Question 1 – Comparing Treatment Neglecters vs. Utilizers on Broad Psychopathology Composite Scores ...... 67

7. Research Question 2 – Comparing INT Psychopathology Symptoms in Treatment Neglecters and Utilizers ...... 68

8. Research Question 2 – Comparing EXT Psychopathology Symptoms in Treatment Neglecters and Utilizers ...... 69

9. Research Question 2 – Comparing DSM-IV Diagnoses in Treatment Neglecters and Utilizers ...... 70

10. Correlations between Symptom Composites among Treatment Neglecters (n = 10,061) and Treatment Utilizers (n = 2,094) ...... 71

11. Fisher’s r-to-Z Transformation Z-values and p-Values for Research Question 3 ...... 72

12. Exploratory Structural Equation Modeling for Alcohol Treatment Rejection Items (n = 1063) ...... 73

13. Exploratory Structural Equation Modeling for Drug Treatment Rejection Items (n = 562) ...... 74

14. Correlations among Demographic Variables and Treatment Rejection Factors in Alcohol (n = 1063) and Drug (n = 562) Treatment Rejecters ...... 75

15. Parameter and Standard Error Estimates for Psychopathology Composites in the Alcohol Treatment Rejection SEM...... 76

16. Parameter and Standard Error Estimates for Treatment Rejection Composites in the Alcohol Treatment Rejection SEM ...... 77

17. Parameter and Standard Error Estimates for Psychopathology Composites in the Drug Treatment Rejection SEM...... 78

18. Parameter and Standard Error Estimates for Treatment Rejection Composites in the Drug Treatment Rejection SEM...... 79

vii

LIST OF ILLUSTRATIONS

Page

1. Research Question 1 – Comparing neglectors and utilizers on broad psychopathology composite scores...... 80

2. Research Question 2 – Comparing internalizing psychopathology symptoms in treatment neglectors and utilizers...... 81

3. Research Question 3 – Comparing externalizing psychopathology symptoms in treatment neglectors and utilizers...... 82

4. Research Question 3 – Does comorbidity explain alcohol or drug treatment rejection? ...... 83

5. Research Question 4 – What is the typology of alcohol treatment rejection? ...... 84

6. Research Question 4 – What is the typology of drug treatment rejection? ...... 85

7. Confirmatory factor analysis for drug treatment rejection items...... 86

8. Confirmatory factor analysis for alcohol treatment rejection items...... 87

9. Alcohol treatment rejection structural equation model...... 88

10. Drug treatment rejection structural equation model...... 89

viii

CHAPTER 1

INTRODUCTION

Research has long documented the relationship between mental disorders and substance

use (e.g., Grant et al., 2004b; Hasin et al., 2011; Kessler, Crum, Warner, & Nelson, 1997;

Kessler, McGonagle, Zhao, & Nelson, 1994; Tate, Drapkin, & Brown, 2009), with 50% of

mentally disordered individuals meeting diagnostic criteria for a substance use disorder (SUD) at

some point during their lifetime (Thornton et al., 2012). SUDs are highly positively associated

with mood disorders, anxiety disorders, personality disorders, suicide attempts, and other severe

forms of psychopathology (e.g., Bolton, Belik, Enns, Cox, & Sareen, 2008; Grant et al., 2004a,

2004b, 2004c; Regier et al., 1990). Comorbidity of SUDs with other Axis I and II conditions

often result in multiple treatment occurrences, poor treatment prognosis, and costly health

services such as emergency room visits and psychiatric hospitalizations (Dickey & Azeni, 1996;

Lundgren, Sullivan, &Amodeo, 2006; McLellan, Lewis, O’Brien, & Kleber, 2000; Ziedonis &

Nickou, 2001). SUDs also have concerning economic, social, and other medical costs (e.g.,

Cartwright, 2008; Rice, Kelman, & Miller, 1991). For instance, a 2011 National Drug

Intelligence Center report asserted SUDs cost the United States more than $193 billion in 2007,

largely resulting from criminal justice system, healthcare, and general economic productivity

costs (National Drug Intelligence Center, 2011).

Despite these costs, treating individuals with SUDs is challenging because many are unlikely to seek treatment (Wu et al., 1999; Wu, Ringwalt, & Williams, 2003; Wang et al., 2005;

Cohen, Feinn, Arias, & Kranzler, 2007; Mojtabai, Olfson, & Mechanic, 2002; Ilgen et al., 2011).

For instance, Grant and colleagues (2004b) reported that among individuals meeting criteria for a

12-month drug use disorder (DUD) or 12-month alcohol use disorder (AUD) in the NESARC

sample, only 13.10% and 5.81%, respectively, sought treatment for their respective conditions.

Similar results have emerged from the National Comorbidity Survey (NCS), reporting that

among the 386 individuals who met criteria for an independent SUD, only 7 % sought

professional help and 3% sought help from mental health professionals (Mojtabai et al., 2002).

Additional epidemiological results have documented that 85% of Wave 1 NESARC individuals with a lifetime AUD diagnosis never sought formal treatment for alcohol problems (Cohen et al.,

2007). Gaining a better understanding of what factors prevent those with SUDs from seeking substance use treatment would not only benefit the individuals suffering from such mental disorders, but would greatly aid in addressing an important research topic, which may lead to suggestions to help lessen the additional social and economic costs to our larger society.

In contrast to studying treatment rejection, some studies have identified factors that predict treatment utilization and treatment prognosis, but findings have been inconsistent and are difficult to generalize when trying to control for the vast amount of comorbid psychopathology

(Hasin et al., 2011). Additionally, existing research examining treatment rejection in substance users has traditionally focused on specific disorder substances (e.g., alcohol) rather than examining broad psychopathology and substance use constructs that have been proposed in recent years. These methodological practices potentially exclude important factors (e.g., polysubstance abuse) that may influence treatment utilization. As it turns out, multiple studies using large national datasets have shown most forms of mental disorders can be mathematically represented in terms of broad externalizing (e.g., SUDs, antisocial personality disorder), and internalizing (e.g., depression, anxiety) psychopathology constructs (Krueger, 1999; Krueger &

Markon, 2006; Markon, 2010).

Overall, most studies focus on individuals who select to receive treatment yet do not

investigate those who neglect or reject treatment. Individuals who do not seek treatment can be

bifurcated into two groups: (1) those with significant impairment resulting from substance use

but believe their symptoms are not severe enough to warrant treatment (labeled “treatment neglecters”), and (2) individuals acknowledging their substance use symptoms are clinically relevant and deserve treatment, but knowingly reject treatment for a variety of reasons (labeled

“treatment rejecters”). The former group likely never comes to the attention of clinicians or researchers, but the latter, while still refraining from treatment, can potentially be identified and

perhaps studied. At the same time, because treatment rejecters typically do not take part in

treatment studies, understanding the reasons why they neglect or reject treatment has been

difficult, particularly because the samples sizes of such participants are not optimal for research.

However, large-scale epidemiological studies such as the NESARC, which provide amble

number of such individuals who have been assessed on a number of treatment-related variables,

offer great opportunity for better understanding why they reject SUD treatment.

The present study seeks to investigate how internalizing, externalizing, and personality psychopathology, will predict the decision to reject substance use treatment. For individuals who reject treatment, the current study investigated the latent structure of reported reasons for treatment rejection and analyze how these predictors serve as barriers to different dimensions of treatment rejection.

Substance Use Disorders

Overview of SUDs

Substance use disorders (SUDs) are defined as a broad array of disorders “related to the taking of a drug of abuse (including alcohol), to the side effects of a medication, and to toxin

exposure” Diagnostic and Statistical Manual of Mental Disorders – 4th edition; DSM-IV; APA,

2000, p. 191). SUDs are among the most frequently observed psychiatric conditions, with epidemiological prevalence rates estimated around 9 % (Grant et al., 2004a). As previously mentioned, SUDs are highly comorbid with a wide variety of internalizing and externalizing disorders (Grant et al., 2004b; Trull, Jahng, Tomko, Wood, & Sher, 2010). While the relationship between SUDs and externalizing psychopathology is not surprising given their shared relationship with impulsivity and disinhibition (Sher & Trull, 1994; Trull, Waudby, &

Sher, 2004), the relationship between SUDs and internalizing psychopathology is less understood.

At the same time, research has long demonstrated the comorbidity between other mental disorders and the substance use disorders (e.g., Dinwiddie, Cottler, Compton, & Abdallah, 1996).

Multiple studies have demonstrated high rates of comorbidity across alcohol and other substances (e.g., Compton, Thomas, Stinson, & Grant, 2007; Hasin, Stinson, Ogburn, & Grant,

2007; Morgenstern, Langenbucher, Labouvie, & Miller, 1997), with 20 % of those with an alcohol use disorder (AUD) also meeting diagnostic criteria for a drug use disorder (DUD)

(Nace, 2005). Such a relationship appears to maintain its significance even when controlling for sociodemographic factors and various psychiatric disorders (Hasin et al., 2007). Despite the significant comorbidity across AUDs and DUDs, the vast majority of literature, especially that which covers treatment utilization, tends to focus primarily on AUDs. Treatment (acceptance or rejection) research on DUDs lacks comprehensiveness, as many studies evaluate the sole impact of a single substance rather than assessing the full spectrum of DUDs. Broadening the scope of

SUD research to include a broad (externalizing) substance use factor is crucial given high rates

of polysubstance abuse in the presence of other psychopathology (Fenton et al., 2012; Grant et

al., 2004c; Trull et al., 2004).

Recognizing the far-reaching effects of SUDs is of particular importance for the current

study, with multiple studies indicating excessive cocaine, marijuana, and opiate use are

associated with long-term cognitive impairment (Jovanovski, Erb, & Zakzanis, 2005; Lundqvist,

2005; O’Malley, Adamse, Heaton, & Gawin, 1992). Some studies suggest these cognitive

impairments might be permanent and could therefore affect treatment, given cognitive

impairment resulting from long term substance abuse may limit ability to gain insight and

learning new skills from psychotherapy (Clark & Haughton, 1975; Moselhy, Georgiou, & Kahn,

2001; Rourke & Loberg, 1996; Tate et al., 2009). Further understanding of the comorbid nature of SUDs and the role of substance use psychopathology in treatment rejection is an important framework for the current study.

Prevalence, Consequences, and Comorbidity of SUDs

Alcohol is one of the most widely abused substances, with approximately 8.46% of the general population meeting diagnostic criteria for an alcohol use disorder (AUD) within the past

12 months (Grant et al., 2004a). Similarly, over 27% of the 43,093 individuals in the first

NESARC wave met diagnostic criteria for an AUD at some point in their lifetime (Oleski, Mota,

Cox, & Sareen, 2010), and lifetime prevalence rates for alcohol dependence have been estimated to be around 15% (Ray, Hutchison, & Hauser, 2008). Such high prevalence rates carry many negative social, legal, and health related consequences (Caetano & Cunradi, 2002; Harwood,

2000).

Epidemiological evidence indicates AUDs are associated with poor mental, emotional, and social functioning, and have also been shown to predict suicidal behavior, violence, and

aggression (Bolton et al., 2008; Hasin et al., 2007; Greenfield & Henneberg, 2001; Leonard &

Quigley, 1999; O’Farrell, Murphy, Neavins, & Van Hutton, 2000; Roizen, 1993). Likewise, epidemiological data have shown DUDs are also associated with significantly poorer mental, social, and emotional health (Compton et al., 2007). Other studies investigating tobacco, cocaine, marijuana, and hallucinogen use have also demonstrated such substances are strongly related to negative consequences such as suicide attempts and intimate partner violence

(Feingold, Kerr, & Capaldi, 2008; Oquendo et al., 2004; Reingle, Staras, Jennings, Branchini, &

Maldonado-Molina, 2012; Smith, Homish, Leonard, & Cornelious, 2012;). Multiple studies have also shown drug use to be associated with the contraction of human immunodeficiency virus (HIV), sexually transmitted diseases, and other infections resulting from hazardous drug injections, all of which also result in significant health care costs (Leukefeld & Haverkos, 1993;

Leukefeld, 1998).

Despite such negative effects, NESARC data have shown individuals with AUDs and/or

DUDs are unlikely to seek treatment. Wave 1 NESARC data demonstrated only 24.1% of those with lifetime alcohol dependence ever sought treatment, and a mere 12.1% of individuals with

12-month alcohol dependence sought treatment within the same 12 months (Hasin et al., 2007).

Hasin and colleagues (2007) report such rates are not significantly different than those observed ten years prior (reported in Grant et al., 2004a). Similarly, Compton et al. (2007) found low rates of treatment utilization among those who qualified for both lifetime and 12-month DUD diagnoses. As noted, extensive comorbidity has not only been established among SUDs, but has extended to various other forms of mental disorders. The following sections explore the concurrence of SUDs with various disorders, along with their impact on SUD treatment rejection.

Internalizing Psychopathology and SUDs

The internalizing-externalizing model of psychopathology was developed as a means of

explaining mental disorder comorbidity using data from NCS (Krueger, 1999). The internalizing

domain contains two lower-order factors, one characterized by fear (e.g., panic disorder, phobia)

and the other by anxiety/misery (e.g., major depression, dysthymic disorder). The externalizing domain is characterized by disorders related to impulsivity and disinhibition (e.g., antisocial

personality disorder, AUDs, other SUDs; see Krueger, 1999 for further detail). These two broad

domains of psychopathology are generally moderately correlated.

Because the conceptualization of mental disorders as internalizing and externalizing disorders is new relative to the commonplace DSM standard of diagnosing independent disorders, literature investigating the role of internalizing psychopathology within SUDS is somewhat disjointed and could benefit from a structured and concentrated approach that can comprehensively evaluate the relationship between internalizing psychopathology and SUDs.

As such, the following review aims to connect the existing research on mood and anxiety

disorders to highlight the nature of SUDs and internalizing psychopathology, along with the

main focus on how they may be linked to SUD treatment rejection.

Comorbidity

Various epidemiological studies have demonstrated significant comorbidity between

SUDs and mood disorders such as major depressive disorder (MDD), bipolar disorder, and

dysthymic disorder (Cohen et al., 2007; Grant et al., 2004b; Kessler et al., 1997; Regier et al.,

1990). Notably, some studies have indicated that when combined with certain forms of

internalizing psychopathology such as bipolar disorder, substance use may exacerbate manic

symptoms, increase the likelihood of psychiatric hospitalization, and put one at greater risk of

suicide (Brady & Sonne, 1995; Dalton, Cate-Carter, Mundo, Parikh, & Kennedy, 2003; Sbrana et al., 2007). Additionally, other research has shown cocaine, amphetamine, and alcohol dependence symptoms are typically accompanied by depression, with some treatment research suggesting MDD is the most comorbid disorder seen in alcohol and drug-dependent individuals

(Niciu et al., 2009; Nunes, Hennessy, & Selzer, 2010; Tate et al., 2009).

Similarly, Epidemiological findings from both the Epidemiological Catchment Area study (ECA) and NESARC studies repeatedly show significant comorbidity also exists between

SUDs and anxiety disorders such as panic disorder, social phobia, generalized anxiety disorder, and simple phobias (Grant et al., 2004b; Grant et al., 2005; Kessler, Chiu, Demler, & Walters,

2005; Magidson, Shang-Min, Lejuez, & Blanco, 2012; Morris, Stewart, & Ham, 2005; Regier et al., 1990; Smith & Book, 2010; Tómasson & Vaglum, 1997). Furthermore, individuals with anxiety disorders who abuse substances are more likely to have additional psychopathology, a higher likelihood of experiencing panic attacks, poorer general health and psychosocial functioning, and worse treatment prognosis than those with independent anxiety disorders

(Breslau & Klein, 1999; Grant et al., 2004b; Magidson et al., 2012; McCabe et al., 2004; Smith

& Book, 2010). Because mood disorders, anxiety disorders, and SUDs are all comorbid with one another (e.g., Grant et al., 2004b), looking at treatment rejection from the context of internalizing psychopathology (i.e., mood and anxiety disorders) rather than from independent disorders provides a viable theoretical and analytic framework for the current study.

Treatment Rejection

Despite such high comorbidity rates between internalizing disorders and SUDs, few individuals with these disorders seek treatment (Wu et al., 1999). This finding has been repeatedly supported by results from both the NCS and NESARC studies, which have indicated

those with internalizing psychopathology and comorbid SUD are less likely to recognize a need

for SUD treatment and are subsequently more apt to reject treatment (e.g., Cohen et al., 2007;

Hasin et al., 2007; Grant et al., 2004b; Mojtabai et al., 2002). Although research has identified

some specific disorders that may impact treatment utilization, the extent of comorbidity across

disorders obscures the nature of their relationship to treatment rejection. Utilizing latent

variables analyses that allows the internalizing factor to be precisely modeled with large

epidemiological datasets may offer promise to understanding relevant latent constructs that

influence treatment rejection.

Externalizing Psychopathology and SUDs

Comorbidity

Externalizing psychopathology encompasses disorders characterized by antisocial

behavioral tendencies, disinhibition, and substance abuse and dependence (Krueger, 1999;

Krueger et al., 2002). SUDs are strongly related to various components on the externalizing

psychopathology spectrum such as conduct disorder and antisocial personality disorder (ASPD;

Krueger et al., 2002; Regier et al., 1990; Vrieze, Perlman, Krueger, & Iacono, 2012). Various

NESARC studies have also shown significant overlap occurs between DUDs and AUDs (e.g.,

Hasin et al., 2007; Falk, Hsiao-ye, & Hiller-Sturmhöfel, 2006; see Jané-Llopis & Matytsina,

2006 for review). However, the majority of literature examining SUDs focuses primarily on

AUDs (Tate et al., 2009), especially in studies examining SUD treatment utilization (e.g., Ilgen et al., 2011; Wu et al., 1999;). Despite AUDs being more common than DUDs (e.g., Grant et al.,

2004a; 2004b), including DUDs to create a broad polysubstance latent factor could help account for the significant overlap that occurs between the two classes of disorders. Although it is now evident that significant comorbidity occurs between all the components subsuming the

externalizing spectrum of psychopathology, few studies have aggregated externalizing symptoms

to predict their impact on SUD treatment.

Treatment Rejection

The disinhibited and impulsive nature of individuals with externalizing psychopathology

makes treatment rejection likely and worsens treatment prognosis (Straussner & Nemenzik,

2007). Moreover, studying SUD treatment in individuals with comorbid ASPD and SUDs is

difficult because the combination of such externalizing psychopathology makes voluntary treatment utilization unlikely. However, a general population-based epidemiological study such

as the NESARC provides a large enough sample to study enough individuals who may have

comorbid ASPD and SUDs but have not sought treatment yet.

As previously mentioned, much of the research examining treatment utilization in

substance use populations has focused on using AUD symptoms as predictors of treatment

utilization (e.g., Cohen et al., 2007; Ilgen et al., 2011; Wu et al., 1999), but broadening the scope

of focus to include drug use disorders and other psychopathology seems fruitful. Compared to

prevalence rates, treatment utilization rates by individuals with AUDs and DUDs are comparably

low (Compton et al., 2007; Wu et al., 1999). The impact of comorbid substance symptoms on

treatment utilization remains largely unknown and understudied (Compton et al., 2007; Oser et

al., 2011). For instance, Wu et al. (1999) reported individuals who experienced three or more

alcohol dependence symptoms during their lifetime were more likely to seek treatment, while

those with only one or two alcohol dependence symptoms were significantly less likely to seek

treatment. Other research has shown that consequences resulting from substance misuse may

predict treatment utilization better than dependence symptoms (Kaskutas, Weisner, & Caetano,

1997).

Another problematic factor inherent to the study of treatment in SUD populations is a

sampling bias best explained by Berkson’s fallacy (Berkson, 1949). Comorbidity estimates

obtained from SUD treatment research may be inflated because such data are likely obtained

from treatment settings, which could result in further estimate inflation (Galanter, Castaneda, &

Ferman, 1988). The far-reaching epidemiological nature of the NESARC offers a compromise to this problem given its ability to reach both those who have (i.e., outpatient and inpatient treatment) and have not been in treatment.

SUDs and Personality Disorders

Comorbidity

Personality disorders (PDs) are characterized by maladaptive, inflexible, and enduring

patterns of problematic interpersonal behavior resulting in psychosocial impairment (APA,

2000). Those with PDs are more likely to experience interpersonal, employment, social,

emotional, and legal problems (Bland, Stebelsky, Orn, & Newman, 1988; Grant et al., 2004d;

McCranie, & Kahan, 1986; Skodol et al., 2005a;). Similarly, PD patients also tend to report lower Global Assessment of Functioning scores (Skodol et al., 2005a; Grant et al., 2004d), and consume large amounts of general health, mental health, legal, and social welfare services such as Medicaid and Supplemental Security Income (Bender et al., 2002; Coid, 2002; Dolan, Warren,

Menzies, & Norton, 1996; Vaughn et al., 2010). Taken together, these findings suggest that higher PD pathology would predict less treatment rejection.

The onerous nature of PDs is worsened by significant comorbidity between PDs and

SUDs (Compton et al., 2007; Grant et al., 2004c; Hasin et al., 2007; Regier et al., 1990).

Individuals with comorbid SUDs and PDs often demonstrate greater impairment, are less likely to seek treatment for their SUD, and have worse treatment prognosis than those without PDs

(Ekleberry, 2009; Kasen et al., 2007; Reich & Green, 1991; Ruegg & Frances, 1995; Pulay et al.,

2008). Thus, one might also pose the notion that those with PDs may become less likely to seek

treatment if they also possess a comorbid SUD. Additionally those with comorbid SUDs and

PDs typically require longer and more frequent treatment episodes and continued care after

treatment cessation (Moggi, Ouimette, Finney, & Moos, 1999). The question that remains is,

when the occurrence between SUDs and PDs is taken into account, via latent variable modeling,

does PD pathology lead to an increase or decrease in treatment rejection?

NESARC data have suggested that the majority of personality disorders are related to

SUDs, perhaps most notably ASPD (Hasin et al., 2011; Trull et al., 2010). While PD/SUD

literature has focused almost exclusively on ASPD (e.g., Grande, Wolf, Schubert, Patterson, &

Brocco, 1984; Regier et al., 1990; Strain, 1995; Morganstern et al., 2007), recent research has

suggested other PDs may also be strongly linked to SUDs, such as obsessive-compulsive PD, schizotypal PD, and avoidant PD (Grant et al., 2004c; Grant, Mooney, & Kushner, 2012; Hasin et al., 2011; Skodol et al., 2005b; Trull et al., 2010). Broadening the scope of SUD/PD research to include other PDs could be useful in the understanding of how personality pathology impacts

SUD treatment utilization.

Treatment Rejection

Empirical research investigating the influence of underlying personality patterns on the decision to seek or reject treatment is limited, but some research has found having a PD may make one more likely to enter treatment (e.g., Tyrer, Mitchard, Methuen, & Ranger, 2003), and thus conversely, less likely to reject treatment. However, accessing personality-disordered individuals with SUDs who have rejected treatment is inherently difficult because treatment utilization is usually conducted in treatment populations. While reasons for treatment rejection

are somewhat idiosyncratic (Ekleberry, 2009), utilizing a latent variable modeling approach may help identify how personality pathology may predict SUD treatment rejection, while also accounting for its comorbidity with internalizing and externalizing psychopathology. This approach can also be useful in its potential ability to parse out comorbidity frequently observed across PDs (e.g., Clark, 2005).

Although the current study primarily focuses on trying to determine why substance users might reject treatment, attaining a basic understanding of why such individuals abuse substances may be helpful to understanding why many do not seek treatment.

Addiction Theories

No specific factor (e.g., genetics) or broad theory fully explains why individuals abuse substances, but a great deal of collaborative research has aided in understanding the development of addiction. Moreover, motivation for why individuals continue to abuse substances despite functional impairment varies greatly. Because motivation for substance use as well as SUD treatment rejection may vary, understanding basic theories of addiction could help provide a framework for understanding why individuals may persistently abuse substances despite adverse consequences. Reviewing the overwhelmingly large number of addiction theories in depth is impractical and strays from the main purpose of this project. Therefore, a few select theories were summarized to highlight how addiction has been operationalized from different perspectives, and may help provide ideas on why high rates of treatment rejection are observed in SUD populations.

The Self-Medication Hypothesis

Khantzian’s (1985) Self-Medication Hypothesis (SMH) proposes those with comorbid mental disorders and SUDs use substances in order to eliminate painful emotions and other

psychological experiences. In addition to finding support for this hypothesis, recent evidence has found certain drugs may serve specific purposes in the presence of certain aversive states.

For instance, Thorton and colleagues (2012) reported depressed individuals in their study often used tobacco to relieve boredom, alcohol for social reasons, and marijuana to relax. Others further investigating the SMH have suggested SUDs are often comorbid with anxiety disorders because the individual self-medicates with various substances to reduce anxiety symptoms

(Robinson, Sareen, Cox, & Bolton, 2011). While reasons for substance use are often idiosyncratic, these studies nonetheless elucidate the varying functional nature of substances.

Although an individual’s latent psychopathology and perceived level of distress may govern the need for relief through substance use, many other individuals still choose different methods of coping with psychological suffering such as social support, hobbies and interests, and psychotherapy. Nevertheless, treatment rejection for some individuals may be linked with a desire to continue to use substances as the means of self-regulating unwanted psychological states.

Heritability Theories of SUDs

Given various studies that have demonstrated most individuals who experiment with illicit substances do not develop patterns of dependence (Kandel, 1992; WHO, 2002), better understanding how biological and environmental aspects intertwine to create patterns of substance abuse and dependence is an essential component of treatment research.

Research has shown heritable genetic and environmental characteristics strongly influence the likelihood of developing abuse and dependence (e.g., Prescott & Kendler, 1999;

Mares, van der Vorst, Engels, & Lichtwark-Aschoff, 2011; Tarter, Alterman, & Edwards, 1985).

Twin studies have also demonstrated the high genetic heritability of addiction (e.g., Gynther,

Carey, Gottesman, & Vogler, 1995; Uhl, Elmer, Labuda, & Pickens, 1995; Merikangas et al.,

1998; Kendler, Aggen, Tambs, & Reichborn-Kjennerud, 2006;), with some evidence even finding addictive factors such as sensitivity or tolerance to alcohol and tobacco to be genetically influenced (e.g., Merikangas et al., 1998; Hinckers et al., 2006). Moreover, the risk for substance misuse is likely to increase when other genetically related temperamental factors such as impulsivity emerge (Bezdjian, Baker, & Tuvblad, 2011). Thus, treatment rejection may be associated with various genetic (e.g., increased sensitivity to SUDs) or environmental (e.g., SUD use as social learning coping mechanism) factors. The next section of this review takes a more in depth look at how SUD treatment rejection is currently conceptualized.

SUD Treatment Rejection

While some research has suggested the likelihood of treatment utilization increases as comorbidity rates increase (Wu et al., 1999), numerous epidemiological studies have shown most individuals with dual diagnoses never seek treatment (Regier, 1990; Wu et al., 1999; Ilgen et al.,

2011; Edlund, Booth, & Han, 2012). For instance, Tucker’s (2001) proposal that more than 75% of substance users never use any type of formal or informal treatment has been supported by epidemiological evidence reporting twelve-month treatment rates for drug abuse and dependence to be around 6% and 30%, respectively (Compton et al., 2007).

It is also important to note the prevalence of “natural recovery” attitudes among substance users (i.e., no informal or formal treatment). Although such attitudes are common among addictive disorders (Slutske, 2010), mixed recovery rate findings and high rates of relapse in individuals who self-treat SUD problems on their own suggest that self-focused, informal treatment of SUDs may lack the long term effectiveness gained through formal treatment from a mental health professional (Bischof, Rumpf, Meyer, Hapke, & John, 2007; Project MATCH

Research Group, 1997). While some researchers have suggested most SUDs remit on their own

and that individuals who receive treatment typically do not recover in the long run (e.g., Price,

Risk, & Spitznagel, 2001; Sobell, Cunningham, & Sobell, 1996), such results can be confounded

by the fact many treatment samples include individuals with SUD problems so severe that they

are unlikely to improve with treatment (Boschloo et al., 2012). Besides the advantages of

treating SUDs directly (i.e., symptom elimination and overall functioning improvement),

economic incentives also exist. For instance, Ettner et al. (2006) reported substance use

treatment in California saved the state nearly seven dollars for every one dollar spent on SUD

treatment. Finding ways to communicate the broad utility of formal substance use treatment should be a key target for clinicians and policy makers.

Despite the obvious personal and economic advantages of treating individual with substance use problems, understanding why many do not seek treatment for periods as long as 10 years after experiencing impairment is yet to be understood (Tucker & Simpson, 2011).

Moreover, the exact typology of SUD treatment rejection is yet to be understood. Understanding the factor structure of individuals’ reported reasons for substance disorder treatment rejection could help researchers discover and eliminate treatment barriers. Such accomplishments could substantially increase the number of people who seek treatment and eliminate their reliance on psychoactive substances. Studies demonstrating factors such as stigmatization of treatment

(Cunningham, Sobell, Sobell, & Agrawal, 1993; George & Tucker, 1996) and a fear that substance abuse symptoms may worsen with treatment (Klingemann, 1991) offer some explanation, but more research is warranted given these factors do not fully explain why most individuals do not utilize treatment.

Mixed results have emerged regarding the predictive power of sociodemographic and

economic variables for treatment utilization. For instance, some have found age, gender, race, to

predict treatment utilization (Luekefeld et al., 1998) while others have not (e.g., Compton et al.,

2007), and some have argued individuals with greater social and economic resources are more

likely to utilize treatment (Chassler, Lundgren, & Lonsdale, 2006). However, some studies have

suggested that factors such as accessibility to treatment, costs associated with treatment, and

insurance coverage have little to do with treatment utilization (Tucker, 1995; George & Tucker,

1996; Oleski et al., 2010). For instance, a study comparing problem drinkers as they entered

treatment to untreated problem drinkers in the community found insurance coverage was not

significantly related to treatment entrance, even when controlling for sociodemographic and

motivational variables (Weisner, Matzger, Tam, & Schmidt, 2002). Moreover, other studies

have demonstrated that even if treatment is readily accessible and affordable, social context and

stigma surrounding SUD treatment overshadow and lower the likelihood of treatment utilization

(George & Tucker, 1996; Room, 1989). At the same time, these studies have not examined how the broad domains of internalizing and externalizing psychopathology, or personality pathology, may predict this type of treatment rejection factor (i.e., perceived objective barriers).

Motivation and Attitudes toward Treatment

Several factors have been shown to be positively associated with treatment utilization,

such as work problems, legal issues, and social pressure (Wu et al., 1999; Ross, Lin, &

Cunningham, 1999; Weisner et al., 2002; Tucker, Vuchinich, & Rippens, 2004). Conversely,

other factors such as negative attitudes towards treatment or beliefs that conditions will improve

without treatment are negatively associated with treatment utilization (Cohen et al., 2007).

Similarly, fearing unknown consequences related to treatment (Klingemann, 1991), viewing that

one should be capable of handling their own drinking problem, and believing that the severity of substance use problems do not necessitate treatment have all emerged as potential barriers to treatment utilization (Cohen et al., 2007). The role of these attitudes and potential subsequent stigmatization of SUD treatment utilization is of great importance, yet need to be further studied

(Tucker, 2001). Moreover, the current study attempted to delineate broad latent factors that may

represent these specific components of treatment rejection.

Not surprisingly, previous research has found individuals who perceive a need for

treatment utilization are much more likely to have received SUD treatment services than those

who do not perceive a need for treatment (Katz, Kessler, Frank, Leaf, & Lin, 1997). Moreover,

researching factors such as positive attitudes towards SUD treatment are important to explore

because perceptions of treatment necessity and attitudes may be factors that ultimately lead one

to seek treatment (Mojtabai et al., 2002). Despite evidence suggesting that negative attitudes and

perceived necessity of help become more salient as comorbidity with other disorders also

increases, the majority of individuals with SUDs who recognize and address this perception is

still low (Mojtabai et al., 2002). This gap between actual and perceived need for SUD treatment

may be partially explained by the presence of comorbid psychopathology (Wu et al., 1999;

Bischof et al., 2005), but further research should explore and more fully explicate the role of

other factors, such as attitudes towards treatment and perceived necessity of help, in the decision

to seek or reject treatment. This challenge can be accomplished by using more psychometrically

sophisticated methodology (e.g., latent variable analyses) to first define and subsequently predict

treatment rejection factors.

Conclusion

The concurrence of SUDs and psychopathology have important personal and societal implications (Grant et al., 2004b; Cartwright, 2008;), but such individuals are unlikely to seek treatment (Cohen et al., 2007, Wu et al., 2003). Previous studies have partially examined the role of psychopathology in the decision to enter or reject treatment, but research has been

conducted only on specific disorders and substances. The internalizing and externalizing model

of psychopathology is becoming increasingly useful in the conceptualization of mental disorders

and offers promise for understanding the interplay between these two factors and SUD treatment rejection. Moreover, few studies have investigated the role of personality pathology in SUD treatment rejection despite some evidence claiming comorbid PD/SUD worsens the likelihood of treatment and treatment prognosis (e.g., Ekleberry, 2009; Pulay et al., 2008)

The current study attempts examine the typology and factor structure of treatment rejection in individuals with SUDs. Once the structure and dimensionality of treatment rejection has been formulated, internalizing and externalizing psychopathology composites, as well as personality disorder composites, were used to predict various components of SUD treatment

rejection.

CHAPTER 2

METHODS

Participants

The current study utilizes Wave 1 data from the National Institute on Alcohol Abuse and

Alcoholism’s (NIAAA) National Epidemiologic Survey on Alcohol and Related Conditions

(NESARC). Wave 1 of the NESARC included 43,093 non-institutionalized Americans ages 18 and older, and includes a variety of individuals based on gender, age, race, and ethnicity.

Moreover, African Americans and Latino populations were oversampled to assuage any concerns regarding the representativeness of minority populations in the study.

Inclusion in the current study necessitated that an individual met criteria for a DSM-IV alcohol use disorder (AUD; abuse and/or dependence) or drug use disorder (DUD; abuse and/or dependence) was selected for inclusion (n = 12,707). Since this project aims to understand substance use treatment rejection at a broad level, different analyses involved the formation of different treatment neglect or rejection groups. Figures 1 through 6 demonstrate the steps taken to create subsamples for each research question. The majority of the 12,707 individuals who met criteria for a substance use disorder (SUD) did not seek treatment (n = 10,415, or 82%) and are deemed “treatment neglecters” throughout the rest of this document. The term “treatment rejecters” is reserved for a subgroup of these 10,415 individuals who recognized a need for treatment yet for some reason did not utilize either alcohol (n = 1,063) and/or drug (n = 562) treatment resources. Figures 5 and 6 visually demonstrate these selection procedures.

Materials

The diagnostic interview used in the NESARC study was the NIAAA’s Alcohol Use

Disorder and Associated Disabilities Interview Schedule–DSM-IV Version (AUDADIS-IV;

Grant, Dawson, & Hasin, 2000), an interview designed to be used by the lay interviewers

responsible for collecting the NESARC data. The AUDADIS-IV assesses DSM-IV symptoms for alcohol abuse and dependence, along with drug abuse and dependence (except nicotine dependence). The AUDADIS-IV includes a list of questions designed to assess DSM-IV diagnostic criteria for substance use disorders, such as alcohol abuse and dependence, as well as drug–specific abuse and dependence diagnoses for the following eight drug classes: cocaine

(including crack-cocaine), cannabis, sedatives, hallucinogens, stimulants, tranquilizers, opiates

(excluding heroin and methadone), and inhalants/solvents.

Additionally, the AUDADIS-IV also assesses the following DSM-IV mood and anxiety disorders: major depressive disorder, dysthymic disorder, social phobia, specific phobia, mania, hypomanic, panic disorder without agoraphobia, panic disorder without agoraphobia, and generalized anxiety disorder. Due to time constraints, only seven of the ten DSM-IV personality disorders were assessed during Wave 1 of the NESARC study (antisocial PD, avoidant PD, dependent PD, histrionic PD, obsessive-compulsive PD, paranoid PD, and schizoid PD).

Importantly for this study, the AUDADIS-IV has a section of questions asking participants whether they have either sought alcohol or drug treatment. If the participant acknowledged they perceived a need for help yet for some reason they did not seek treatment

(i.e., treatment rejection), 27 questions assessing various reasons for treatment rejection were asked. These 27 items serve the basis for determining the factor structure of SUD treatment rejection (See Table 1 of Appendix for specific drug and alcohol treatment rejection items).

Previous NESARC analyses have shown the AUDADIS-IV demonstrates excellent test- retest reliability for its drug and alcohol symptoms and disorders (Grant, Harford, Dawson,

Chou, & Pickering, 1995; Grant et al., 2003). Likewise, other studies have shown the

AUDADIS-IV demonstrates good construct (Grant, Harford, Dawon, & Chou, 1995; Grant et al.,

2004b; Hasin, Muthen, & Grant, 1997; Nelson, Rehm, Usten, Grant, & Chatterji, 1999) convergent, and discriminate validity (Cottler et al., 1997; Hasin & Paykin, 1999; Grant, 1992).

Several studies have found test-retest reliabilities for mood and anxiety disorders range from fair to good (Canino et al., 1999; Grant, Dawson, Stinson, Chou, Kay, & Pickering, 2003). Test- retest reliabilities for personality disorders assessed by the AUDADIS-IV range from fair to good (Grant et al., 2004c).

Data Analysis

Descriptive Statistics

Descriptive statistics were obtained for all participants on all variables. Means and standard deviations, as well as skewness and kurtosis, were computed for all necessary variables.

Gender differences on variables of interest were also analyzed.

Primary Data Analysis

Many of the research questions in this proposal were assessed using latent variable analyses, such as exploratory structural equation modeling (ESEM), confirmatory factor analysis

(CFA), and other forms of structural equation modeling (SEM). In addition, a wide array of symptoms for specific disorders were assessed in the NESARC study, and therefore composite scores based on DSM-IV criteria were calculated and used in all manifest variable analyses. mean inter-item correlations (MICs) were calculated to ensure the composite scores represent homogenous constructs.

Composite scores were created to test hypotheses about certain forms of psychopathology

(e.g., internalizing psychopathology), specific disorders (e.g., major depressive disorder), and substance use treatment rejection. If a specific symptom was assessed with multiple items, these

items were aggregated into single composite variables representing the original criterion in order

to reduce redundancy. Participants were required to have at least 85 % complete data on the

items subsuming a specific composite variable in order to receive a value for that specific

composite score variable. For instance, in order for a participant to obtain a score for the Panic

Disorder Composite Symptoms scale (13 items), they must have data (i.e., a non-missing value)

for at least 11 of 13 items (~85%). Symptom composite scale scores were calculated as means

rather than sums due to the varying number of items subsuming each composite scale.

Exploratory factor analyses (EFA) were conducted to investigate whether certain items

did not load onto an underlying factor. Depending on the results of these EFAs, certain items

were removed to if they did not load onto a factor, and also to increase item homogeneity.

Based on recommendations by Clark and Watson (1995), composite scale

unidimensionality was evaluated through a two-step process using tetrachoric correlations instead of Pearson’s product-moment correlations. First, Mean Inter-item Correlations (MICs) were calculated to assess scale homogeneity (an indicator of unidimensionality) rather than

Cronbach’s alpha because the latter is a measure of internal consistency, not unidimensionality.

Next, individual inter-item correlations were evaluated to ensure sufficient clustering around the mean for specific composite variables. Clark and Watson (1995) recommend both of these steps to ensure internal consistency and unidimensionality. MICs for internalizing and externalizing psychopathology, personality pathology, and treatment rejection composite scores are provided in Table 1.

Many of the variables used in the following analyses demonstrated a non-normal distribution. Despite conducting transformations in an attempt to achieve normality, many of the variables maintained their non-normal distribution. It is plausible that the clinical nature of the

variables used in the following analyses (e.g., internalizing psychopathology, personality

pathology) are not normally-distributed phenomena. Due to this non-normality, many of the

analyses were conducted with nonparametric analyses. For some research questions, both

parametric and nonparametric analyses were conducted in order to compare differences in results

among the two methods.

Research Questions and Hypotheses

The primary goals of the current project were to identify latent treatment rejection factors, using exploratory structural equation modeling (ESEM) and confirmatory factor analysis, and examine how they are predicted by internalizing, externalizing, and personality pathology among those with substance use disorders. More broadly, the current project sought to delineate the nature of the psychopathology in those who more generally choose not to enter treatment

(neglecters, rejecters). These goals were accomplished by examining the following questions:

Research Question and Hypothesis 1: Which Aspects of Psychopathology are more Prevalent in Those Who Do Not Seek SUD Treatment?

This research question focuses on individuals with SUD diagnoses to determine if treatment neglecters are higher on dimensionalized composite scores of broad internalizing and externalizing psychopathology, SUD pathology, and PD pathology than treatment utilizers. It was hypothesized that individuals who utilize treatment would exhibit higher levels of internalizing and externalizing psychopathology. Contrarily, it was expected that individuals who neglect treatment would possess higher levels of broad PD pathology compared to treatment utilizers.

Research Question and Hypothesis 2: Are Certain Disorders More Prevalent in Treatment Neglecters than Treatment Utilizers?

This research question investigated whether certain disorders are more prevalent in

individuals who utilize treatment versus individuals who neglect treatment. It was expected that

treatment utilizers would possess greater levels of alcohol dependence, drug dependence, and all

internalizing psychopathology disorder psychopathology. Moreover, it was also predicted that

alcohol dependence, drug dependence, and internalizing disorder diagnoses would increase the

likelihood of treatment utilization. It was hypothesized that individuals who neglect treatment

would exhibit higher levels of ASPD and other PD pathology, and that PD diagnoses would

decrease the likelihood of treatment rejection. Mann-Whitney U analyses and binary logistic

regressions were conducted to address this research question.

Research Question and Hypothesis 3: How Strong is the Relationship between Treatment Rejection and Broad/Specific Composites among Individuals with a Specific SUD Diagnosis?

This research question explored the extent to which a comorbid diagnosis can predict

treatment rejection or utilization in individuals with SUDs. Both manifest and latent variable

analyses were conducted to determine if the associations between internalizing, externalizing,

and PD pathology are stronger in treatment neglecters (or rejectors in the case of latent variable analyses) or utilizers. It was hypothesized that comorbidity would be greater in treatment utilizers than treatment rejecters.

Research Question and Hypothesis 4: Can Treatment Rejection be Understood in Terms of Underlying Dimensions?

This research question used exploratory structural equation modeling (ESEM) and confirmatory factor analysis (CFA) to determine the factor structure of reasons individuals give for SUD treatment rejection (see the appendix for specific drug and alcohol treatment rejection items). Ultimately, the results of the ESEM and CFA helped determine whether treatment rejection was a unidimensional or multidimensional construct. It was predicted that treatment rejection would be multifactorial in nature based on the abundance of factors (e.g., motivation,

attitudes towards treatment, etc.) likely to be considered when an individual is determining

whether to seek treatment.

Research Question and Hypothesis 5: After Determining the Factor Structure of Treatment Rejection, What Variables Predict Specific Treatment Rejection Factors?

Because the current study proposed multiple treatment factors would emerge, the

proposed study variables (internalizing psychopathology, externalizing psychopathology, and PD pathology) were employed to investigate which of these latent variables best predicts treatment rejection factors. It was expected that all variables would significantly contribute to the prediction of SUD treatment rejection.

CHAPTER 3

RESULTS

Demographics

Demographic results are presented in Table 1. The sample (N = 12, 707) was primarily male (61.9%) and the age of participants ranged from 18 to 92 (M = 42.6, SD = 15.0). The sample was primarily Caucasian (67.6%), with relatively equal proportions of Hispanic (14.4%) and African-American (14.2%) individuals. The majority of the sample reported being married

(47.6%), followed by those reporting that they were either single (24.8%) or divorced (15.7%).

Concerning education, 41.3% reporting obtaining their high school diploma/GED or less, 23.7% attended college but did not obtain their degree, 9.5% obtained their 2-year associate’s degree, and 25.5% obtained a bachelor’s degree or had post-bachelor’s degree experience. Prevalence rate differences between treatment neglecters and treatment utilizers on key demographic

variables are also shown in Table 2.

Descriptive Statistics

Internalizing Psychopathology

Prevalence rates of internalizing disorders in the current sample are presented in Table 3.

Of the total study sample (N = 12, 707), 40.2% met criteria for at least one internalizing disorder

and 18.2% met criteria for more than one internalizing disorder. Major depressive disorder

(MDD) was the most common form of internalizing psychopathology, with 28.1% of the sample

meeting criteria for MDD. Panic disorder without agoraphobia (PD w/out) was the least

common form of internalizing psychopathology, with 6.7% of participants meeting criteria for

PD w/out.

On average, a participant could be expected to endorse approximately 14 of the 59 symptoms (roughly 23 %) which were used to form the internalizing composite scale (Mdn =

.19, M = .23, SD = .22). Composite scores were highest for the dysthymia symptom composite

(Mdn = 1.00, M = .85, SD = .24), while composite scores were lowest for the simple phobia symptom composite (Mdn = 0.00, M = .10, SD = .15). The latter statistic is likely because approximately one-half of the sample did not endorse any simple phobia items. Descriptive statistics for the broad internalizing psychopathology composite along with its subscales (e.g.,

MDD symptomatology) are presented in Table 1.

Externalizing Psychopathology

Prevalence rates of externalizing disorders in the current sample are presented in Table 3.

Given the main criterion for sample inclusion is the diagnosis of a SUD, it is not surprising that the majority of the sample met criteria for an AUD (93.2%) or a DUD (32.0%). Similarly,

25.2% of the sample had met criteria for both disorders at some point in their lifetime.

Approximately 9 % of participants met criteria for antisocial personality disorder (ASPD).

On average, a participant could be expected to endorse 7 of the 34 (approximately 21 %) symptoms subsuming the externalizing composite scale (Mdn = .18, M = .21, SD = .17).

Composite scores were higher for the alcohol dependence composite (Mdn = .33, M = .38, SD =

.28) than the drug dependence (Mdn = .11, M = .25, SD = .29) and ASPD (Mdn = .13, M = .16,

SD = .13) composite scales.

PD Pathology

Prevalence rates of personality disorders (PD) in the current sample are presented in

Table 3. Approximately one-fourth of the sample met criteria for a PD, with approximately 10

% of the sample meeting criteria for more than one PD. The most commonly diagnosed PD was

obsessive-compulsive personality disorder (12.2%), while dependent personality disorder was

the least commonly diagnosed PD (0.8%). On average, a participant could be expected to

endorse approximately 10 (roughly 13%) of the symptoms subsuming the PD pathology

composite scale (Mdn = .13, M = .16, SD = .13).

Treatment Utilization and Rejection

As seen in Table 2, approximately 82 % of the sample had never sought treatment for

their SUD, while approximately 17 %of the sample had sought SUD treatment during their

lifetime. Of these treatment utilizers, nearly 84 %had sought treatment for an AUD, 34 %had sought treatment for a DUD, and approximately 18 % sought treatment for both disorders. Table

4 demonstrates that the most popular avenues of treatment were 12-step programs such as alcoholics or narcotics anonymous (70.8%), alcohol or drug rehabilitation programs (44.2%), or services received from a private physician, psychiatrist, psychologist, social worker or other mental health professional (40.5%). Likewise, few individuals utilized treatment resources at crisis centers (4.8%), employee assistance programs (7.4%), and halfway houses/therapeutic communities (8.6%).

Among those who rejected treatment even after acknowledging a need for treatment, follow up questions were administered to determine why treatment was rejected. Because many of the individuals in the sample did not recognize a need treatment (i.e., answering “No” to the item, “Was there ever a time when you thought you should see someone for your alcohol/drug use, but you didn't go?”), non-missing responses on treatment rejection items were limited.

Positive endorsement rates of treatment rejection reasons are presented in Table 5.

Individuals with data on these questions are deemed “treatment rejecters.” The two most common reasons for treatment rejection were thinking that one should be strong enough to

handle substance use issues alone (41% for AUD treatment rejecters and 43% for DUD treatment

rejecters), and thinking the problem would get better by itself (33% for AUD treatment rejecters

and 34% for DUD treatment rejecters).

Prevalence of Psychopathology in Treatment Neglecters and Treatment Utilizers

Figure 1 demonstrates the selection procedure used for the first research question. As

previously mentioned, many of the variables appear to have a non-normal distribution.

Acknowledging this limitation, Mann-Whitney U tests were conducted to determine if those who

neglect treatment, compared to those who utilize treatment, were higher on dimensionalized

composite scores of internalizing and externalizing psychopathology, broad SUD pathology, and

PD pathology. Results from these Mann-Whitney U tests are presented in Table 6.

Mann-Whitney U tests indicated a small statistically significant difference between

treatment neglecters and utilizers on internalizing psychopathology (r = .13) and PD pathology (r

= .09) composite scores, indicating treatment utilizers scored higher than treatment neglecters on these symptom composites. Mann-Whitney U tests indicated a moderate statistically significant difference between treatment neglecters and utilizers on externalizing psychopathology (r = .40) and Broad SUD pathology (r = .36) composite scores, again indicating treatment utilizers scored higher than treatment neglecters on these symptom composites.

Prevalence of DSM-IV Disorders in Treatment Neglecters and Utilizers

Figures 2 and 3 represent the subsample selection procedure used for the second research question. Mann-Whitney U tests were conducted to investigate whether certain disorder composites were higher in individuals who neglect treatment versus individuals who utilize treatment. Results for internalizing and externalizing disorder symptomatology are presented in

Tables 7 and 8, respectively.

With respect to internalizing disorder symptomatology, treatment neglecters

demonstrated somewhat higher levels of psychopathology on all disorder composites than

treatment utilizers. Mann-Whitney U tests yielded small, statistically significant effect sizes for

MDD (r = .13, p < .001), dysthymia (r = .10, p < .001), GAD (r = .06, p < .05), social phobia (r

= .08, p < .01), and simple phobia symptoms (r = .04, p < .001). Compared to the other

internalizing disorders, panic disorder without agoraphobia yielded the largest effect size, r =

.20, p < .001. Concerning externalizing disorder symptomatology, Mann-Whitney U tests

yielded moderate effect sizes for alcohol dependence (r = .33, p < .001), drug dependence (r =

.32, p < .001), and ASPD (r = .29, p < .001). As before, treatment utilizers displayed higher symptom composite scores, compared to the neglecters.

Logistic regressions were conducted to determine whether certain disorders were more

predictive in individuals who reject treatment versus those who utilize treatment. Treatment

rejection (“yes” = 1, “no” = 0) was used as the dependent variable and specific diagnoses (e.g.,

MDD, ASPD) were used as independent variables. These results are presented in Table 9.

The results indicate substance users with any given internalizing disorder are more likely

to seek substance use treatment than substance users without an internalizing disorder, OR =

1.68, [95% CI 1.54, 1.85]. Specifically, the likelihood of seeking treatment was especially

higher when the cases with SUD also met criteria for dysthymia, OR = 2.79 [CI 2.43, 3.19], or

panic disorder, OR = 2.00 [CI 1.70, 2.33]. Analyses also indicated that social phobia, OR = 1.29

[CI 1.10, 1.51], and simple phobia, OR = 1.17 [CI 1.02, 1.33], also increased the likelihood of

seeking treatment when comorbid with a SUD, but these disorders represented the smallest

increase in odds among all the internalizing disorders.

Among externalizing disorders, the likelihood of seeking treatment was substantially

increased when an alcohol use disorder, OR = 12.91 [CI 7.10, 23.45], or DUD was present, OR =

10.73 [CI 7.99, 14.39]. Surprisingly, treatment also became more likely when a SUD was

comorbid with ASPD, OR = 3.03 [CI 2.66, 3.50].

Additionally, the results highlight that the comorbidity of a SUD with a personality

disorder more than doubled the likelihood of seeking treatment, OR = 2.18, [CI 1.98, 2.40].

When a SUD was present, the utilization of treatment was especially likely with the addition of

dependent, OR = 4.00, CI [2.70, 5.93], antisocial (referenced above), or avoidant PD, OR = 2.68,

CI [2.21, 3.24].

The Relationship between Treatment Neglect and Psychopathology

Figure 4 demonstrates the subsample selection procedure used for the third research

question. To study the extent to which comorbidity (e.g., externalizing and internalizing

psychopathology) predicted treatment neglect or utilization, point-biserial correlations were

conducted on both treatment groups and subsequently compared using Fisher’s r-to-Z

transformations. These results are presented in Tables 10 and 11.

Results yielded moderate correlations between internalizing psychopathology and

externalizing psychopathology for both treatment neglecters (r = .33, p < .001) and treatment

utilizers (r = .44, p < .001). Furthermore, a Fisher’s r-to-Z transformation found this difference

to be statistically significant (Z = -5.39, p < .001), suggesting that the likelihood of seeking

treatment increases as the association (i.e., comorbidity) between internalizing and externalizing psychopathology increases.

Similarly, analyses demonstrated moderately sized correlations between internalizing

psychopathology and substance dependence pathology for treatment neglecters (r = .24, p <

.001) and treatment utilizers (r = .31, p < .001). Again, a Fisher’s r-to-Z transformation found

this difference to be statistically significant (Z = -3.06, p < .01), indicating treatment utilization

may become more likely when internalizing psychopathology and substance dependence

pathology become more strongly correlated.

Lastly, results yielded large correlations between internalizing psychopathology and

personality pathology for treatment neglecters (r = .49, p < .001) and treatment utilizers (r = .56,

p < .001). A Fisher’s r-to-Z transformation found the difference between these two correlations

to be statistically significant (Z = -3.95, p < .001), suggesting one may become more inclined to

seek treatment as internalizing psychopathology and personality pathology become more

strongly associated.

Results indicate increased levels of internalizing, externalizing, and personality

psychopathology are associated with increased rates of SUD treatment utilization. Moreover, the

SUD treatment utilization became even more likely when individuals possessed multiple forms

of psychopathology (i.e., comorbidity).

The Factor Structure of Substance Use Disorder Treatment Rejection

Figures 5 and 6 represent the subsample selection procedure used for the fourth research

question. Exploratory structural equation modeling (ESEM) and confirmatory factor analyses

(CFA) were conducted to investigate the broad structure of SUD treatment rejection items. weighted least square mean- and variance- adjusted (WLSMV) was used as the model estimator

for the structural equation modeling (SEM) because of its ability to accommodate both nonlinear

(i.e., binary, ordinal) and continuous data (Muthén, du Toit, & Spisic, 1997). Geomin rotations were used because of their utility when the true structure of a latent construct is unknown

(Asparouhov & Muthén, 2009).

Initial ESEMs with geomin rotations (a correlated factor approach to an oblique rotation)

suggested a three-factor model fit the data best (RMSEA = .02, CFI = .95, TLI = .94). Factor

loadings for the alcohol treatment rejection EFA are presented in Table 11. The three-factor

ESEM model fit for the drug treatment rejection items was poorer than the alcohol treatment

rejection items, but the model still achieved good fit (RMSEA = .02, CFI = .94, TLI = .93).

Factor loadings for the drug treatment rejection ESEM are presented in Table 12. As indicated

by Tables 11 and 12, items that worsened model fit, loaded poorly on all factors, or demonstrated

cross-loadings were removed from the final three-factor solution. As were discussed later, these

factors were labeled as “objective barriers,” “psychological barriers,” and “self-focused barriers”

factors.

Next, separate three-factor CFAs were conducted on both sets of treatment rejection

items (i.e., drug and alcohol treatment rejection) after the aforementioned ESEMs suggested that

a three-factor model may provide the best model fit.

CFA fit indices indicated acceptable model fit for the drug treatment rejection items (TLI

= .95, CFI = .96, RMSEA = .03) and alcohol treatment rejection items (TLI = .91, CFI = .92,

RMSEA = .04). These fit indices satisfy traditional model fit standards set forth by many (e.g.,

MacCallum, Browne, & Sugawara, 1996; Hooper, Coughlan, & Mullen, 2008). For the alcohol

treatment rejection CFA, all three factors were demonstrated strong positive interfactor

correlations with each other.

For the alcohol treatment rejection items, the objective barriers factor demonstrated

strong positive associations with the psychological barriers factor (r = .63), as well as the self- focused barriers factor (r = .50). The psychological barriers factor also demonstrated a strong positive association with the self-focused barriers factor (r = .67).

For the drug treatment rejection items, only the correlation between the psychological barriers and self-focused barriers factors was positive and strong in magnitude (r = .57). Results yielded a moderate, positive association between objective barriers and psychological barriers (r

= .39). The objective barriers and self-focused barriers factors demonstrated the smallest positive interfactor correlation in the drug treatment rejection CFA (r = .21). Factor loadings for the drug and alcohol treatment rejection items are presented in Figures 7 and 8, respectively.

Additionally, correlations between demographic variables and the treatment rejection factors are shown in Table 13.

The Prediction of Substance Use Disorder Treatment Rejection

Structural equation modeling (SEM) analyses were conducted for both alcohol (n = 1063) and drug (n = 562) treatment rejecters using MPlus version 6.1 (Muthén & Muthén, 1998-2010).

Results of these SEMs are presented in Figures 9 and 10. Unobserved variables (i.e., latent constructs) are represented with circle or ovals, and observed variables (i.e., indicator variables) are represented with rectangular boxes. As previously mentioned, the mean- and variance- adjusted weighted least squares (WLSMV) was the chosen estimator for these SEM analyses because of its robust estimation ability when dealing with a combination of dichotomous and continuous variables (Muthén & Muthén, 1997; Muthén et al., 1997).

Coefficients were standardized using both the variance of outcome and latent variables.

The standardized coefficients represent the level of change in the outcome variables per standard deviation unit of a specific predictor variable.

While researchers traditionally recommend a sample size of at least 200 for SEM analyses, current SEM researchers tend to evaluate the N:q ratio, where ‘N’ represents the total number of cases and ‘q’ represents the number of free parameters in the model (Jackson, 2003;

Kline, 2011). Most SEM researchers suggest using a N:q ratio of 10:1 (Schreiber et al., 2006), but others have indicated a 5:1 N:q ratio is still acceptable (Bentler & Chou, 1987). The alcohol treatment rejection SEM utilized 83 total parameters and thus satisfied both N:q ratio suggestions. The drug treatment rejection SEM utilized 77 total parameters, thereby satisfying only one of the aforementioned ratio suggestions and potentially suggesting this SEM analysis may be slightly under-powered.

Model Specification

Diagnoses of major depressive disorder, dysthymia, panic disorder without agoraphobia, social phobia, simple phobia, and generalized anxiety disorder, all of which were code “yes” or

“no,” served as indicator variables for the latent construct of internalizing psychopathology.

Avoidant PD, dependent PD, obsessive-compulsive PD, paranoid PD, and schizoid PD, all of which were coded “yes” or “no,” served as indicator variables for the latent construct PD pathology. Lastly, alcohol, nicotine, cocaine, and marijuana dependence diagnoses, along with antisocial PD, were all coded “yes” or “no” and served as indicator variables for the latent construct of externalizing psychopathology.

The three treatment rejection factors yielded from the CFAs were chosen as the latent variables representing the latent construct of “treatment rejection.” Items (coded “yes” or “no”) that created the best fit in the CFAs were chosen as indicator variables for their respective factors.

Alcohol Treatment Rejection SEM

Standardized and unstandardized parameter estimates for the latent psychopathology and treatment rejection factors, as well as standard errors for these parameters, are presented in

Tables 15 and 16. The full model is presented in Figure 9 model fit for the alcohol treatment

rejection SEM model was acceptable (CFI = .92, RMSEA = .02). Factor loadings were moderate

to large in magnitude for the internalizing psychopathology composite (M = .62, range = .41 –

.78), externalizing psychopathology composite (M = .61, range = .47 – .75), and PD pathology

composite (M = .77, range = .63 – .85). Average standard errors for the internalizing

psychopathology (Average SE = .04), externalizing psychopathology (Average SE = .07), and

personality pathology (Average SE = .04) indicator variables were low, suggesting good

reliability for the items subsuming their respective psychopathology composites.

As shown in Table 15, average standardized loadings for indicator variables on the

internalizing (Average β = .62) and externalizing (Average β = .61) psychopathology composites

were roughly equivalent, suggesting these two latent factors performed similarly well in

explaining the amount of variance of their respective indicator variables (~ 37%). However, the

average standardized loading on PD pathology was much greater (Average β = .77), indicating

the latent PD pathology factor explained more of the variance of its indicator variables (~ 59%),

compared to the other two psychopathology composites.

Factor loadings were moderate to large in magnitude for the Objective Barriers (M = .64, range = .61 – .68), Psychological Barriers (M = .64, range = .40 – .79), and Self-Focused Barriers

(M = .68, range = .41 – .83) composites. Furthermore, standard errors for the Objective Barriers

(SE = .07), Psychological Barriers (SE = .05), and Self-Focused barriers (SE = .05) composites

indicated good reliability for the parameters subsuming these composites.

The average standardized loadings for indicator variables on the Objective (Average β =

.64), Psychological (Average β = .64), and Self-focused (Average β = .68) Barriers treatment

rejection factors were roughly equivalent, indicating the treatment rejection factors three factors

performed similarly well in explaining the amount of variance of their respective indicator variables (~ 41 – 46%).

The results yielded a strong positive correlation between internalizing psychopathology composite and the PD pathology composite (r = .75), a moderate positive correlation between

PD pathology and externalizing Psychopathology(r = .49), and a moderate positive correlation between internalizing psychopathology and externalizing psychopathology (r = .32). Results also demonstrated moderate to strong positive correlations between Objective Barriers and

Psychological Barriers (r = .57), Psychological Barriers and Self-Focused Barriers (r = .63), and

Objective Barriers and Self-Focused Barriers (r = .42).

All Structural parameter estimates shown in Figure 9 were statistically significant. These structural parameters represent the expect level of change in the dependent variable associated with a one unit change in the predictor variable, while simultaneously controlling for correlational effects created by other predictors. It should be noted that the structural path coefficients from PD pathology  Objective treatment barriers and PD pathology  Self- focused barriers were not statistically significant and thus not were included in the final model.

Internalizing psychopathology significantly predicted Objective (β = .48, p < .001),

Psychological (β = .35, p = .001), and Self-Focused (β = .29, p < .01) Treatment Barriers.

Externalizing psychopathology also significantly predicted Objective (β = .43, p < .001),

Psychological (β = .20, p = .01), and Self-Focused (β = .18, p < .05) treatment Barriers. PD pathology significantly negatively predicted Objective Treatment Barriers (β = -.37, p = .01).

Drug Treatment Rejection SEM

Standardized and unstandardized parameter estimates for the latent psychopathology and treatment rejection factors, as well as Standard Errors for these parameters, are presented in

Tables 17 and 18. The full model is presented in Figure 10. Model fit for the alcohol treatment

rejection SEM model was acceptable (CFI = .90, RMSEA = .02). Factor loadings were moderate

to large in magnitude for the internalizing psychopathology composite (M = .64, range = .33 –

.79), externalizing psychopathology composite (M = .59, range = .45 – .75), and PD pathology

composite (M = .74, range = .62 – .86). Average Standard Errors for the internalizing

psychopathology (Average SE = .05), externalizing psychopathology (Average SE = .08), and personality pathology (Average SE = .05) indicator variables were low, suggesting good reliability for the items subsuming their respective psychopathology composites.

The average standardized loadings for indicator variables on the internalizing (Average β

= .64) and externalizing (Average β = .59) psychopathology composites were roughly equivalent, suggesting these two latent factors performed similarly well in explaining the amount of variance of their respective indicator variables. However, the average standardized loading on PD pathology was much greater (Average β = .74), again indicating the latent PD pathology factor better explained the variance of its indicator variables when compared to the other two psychopathology composites.

Factor loadings were moderate to large in magnitude for the Objective Barriers (M = .70, range = .59 – .79), Psychological Barriers (M = .65, range = .46 – .76), and Self-Focused Barriers

(M = .70, range = .50 – .84) composites. Furthermore, average Standard Errors for the Objective

Barriers (Average SE = .10), Psychological Barriers (Average SE = .06), and Self-Focused barriers (Average SE = .06) indicated good reliability for the parameters subsuming these composites.

The average standardized loadings for indicator variables on the Objective (Average β =

.70), Psychological (Average β = .65), and Self-focused (Average β = .70) Barriers treatment

rejection factors were roughly equivalent, indicating the treatment rejection factors three factors performed similarly well in explaining the amount of variance in their respective indicator variables.

Results yielded a strong positive correlation between the internalizing psychopathology composite and the PD Pathology composite (r = .78), a moderate positive correlation between the PD pathology and externalizing Psychopathology composites (r = .47), and a small positive correlation between internalizing psychopathology and externalizing psychopathology composites (r = .22). Results demonstrated moderate to strong positive correlations between

Objective Barriers and Psychological Barriers (r = .30), Psychological Barriers and Self-Focused

Barriers (r = .52), and a small positive correlation between Objective Barriers and Self-Focused

Barriers (r = .14).

Results only yielded three statistically significant structural parameter estimates for the drug treatment rejection CFA. Internalizing psychopathology significantly predicted Objective

(β = .35, p < .05), Psychological (β = .58, p = .001), and Self-Focused (β = .37, p < .05)

Treatment Barriers. Neither externalizing psychopathology nor PD pathology significantly predicted any of the three drug treatment rejection factors.

CHAPTER 4

DISCUSSION

As mentioned in the introduction, multiple studies have demonstrated that individuals with substance use problems never seek treatment (Grant et al., 2004b; Wu et al., 1999).

Mirroring results that used other epidemiological data (e.g., Mojtabai et al., 2002; Regier et al.,

1990), the current study also found the majority of individuals who qualified for a substance use disorder (SUD) had never sought treatment during their lifetime. As pointed out by Wu et al.

(1999), research must focus on identifying barriers to treatment given the vast number of people who never seek treatment for their substance use problems. As such, the goals of the current project were to understand if those who utilized treatment were different from those who neglected or rejected treatment, to identify the latent factor structure of SUD treatment rejection, and to identify factors that may serve useful in the prediction of SUD treatment rejection.

For the current study, the term “treatment neglecters” referred to individuals who possessed clinically relevant substance use symptoms but never perceived a need for treatment and thus never sought treatment. “Treatment rejecters” was used to refer to a subgroup of treatment neglecters who perceived a need for treatment for their substance use symptoms, but never sought treatment because for a variety of reasons.

Do Treatment Neglecters Differ from Treatment Utilizers with Respect to Internalizing and Externalizing Psychopathology, Broad SUD Pathology, or PD Pathology?

One of the primary goals of the current study was to evaluate whether having high levels of comorbid psychopathology (i.e., a SUD diagnosis along with another mental disorder) would drive individuals towards treatment (addressed by Hypotheses 1-3). While prior research has tangentially explored this question and found many with dual diagnoses are unlikely to seek treatment (e.g., Wu et al., 1999), the current project expands on past research by investigating

whether specific forms of psychopathology predict SUD treatment rejection. To accomplish this,

treatment neglecters and utilizers were compared on dimensionalized symptom composite

variables, as well as specific DSM diagnoses.

Internalizing Psychopathology

The first hypothesis proposed that treatment utilizers would exhibit higher levels of

certain forms of psychopathology (i.e., internalizing and broad SUD psychopathology) than

treatment neglecters, but results only partially supported this hypothesis, in that results indicated

treatment utilizers were actually higher on all forms of broad psychopathology when compared

to treatment neglecters.

Moderate support was found for the first component of the second hypothesis, which

proposed internalizing disorder composites would be higher in treatment utilizers than treatment

neglecters. While treatment utilizers demonstrated higher scores on all internalizing

psychopathology composites than treatment neglecters, effect sizes from all disorder specific

analyses were comparably small in magnitude with the exception of panic disorder without

agoraphobia. Despite reaching statistical significance, such small effect sizes should not be

interpreted as concrete evidence that treatment utilizers significantly differ from treatment

neglecters with respect to certain internalizing disorder symptomatology.

Support was found for the secondary component of the second hypothesis, which

proposed that having an internalizing disorder diagnosis would increase the odds of seeking

treatment. Logistic regressions revealed that presence of an internalizing disorder diagnosis

(along with a SUD diagnosis) increased the odds (OR=1.5) of treatment seeking. Notably, those with comorbid SUD/dysthymia or comorbid SUD/panic disorder without agoraphobia were especially likely to seek treatment when compared to individuals without the presence of such

disorders. Other research has noted high rates of substance use among individuals diagnosed with dysthymia, suggesting that the substance use exacerbates pre-existing dysthymia symptoms

(Eames, Westermeyer, & Crosby, 1998; Thomasius, Sack, & Petersen, 2010). That is, the addition of SUD symptomatology to preexisting mood symptoms could propel an individual to seek treatment for their comorbid psychopathology. However, few studies have examined the function of this comorbidity in the context of actual treatment utilization or neglect.

The high levels of panic disorder symptomatology observed in treatment utilizers, along with the increased likelihood of treatment utilization when comorbid with a SUD, may be explained within the context of withdrawal symptoms. As were discussed later, both drug and alcohol withdrawal symptoms were significantly higher in treatment utilizers than treatment neglecters. A comparison of substance use physical withdrawal symptoms to panic attack symptoms in the DSM-IV (APA, 2000) will show a high degree of overlap, especially since anxiety can result from substance withdrawal (APA, 2000). Thus, the increased odds of treatment utilization among those with comorbid SUD/panic disorder without agoraphobia could result from a combination of anxiety and substance use withdrawal symptoms. Some research supports this notion (Tómasson & Vaglum, 1996), but future research should address this hypothesis directly.

It is interesting that differences between treatment utilizers and treatment neglecters on internalizing psychopathology disorder composites were not moderate or large in magnitude, yet having such disorders increased the likelihood of seeking treatment. It is plausible that treatment utilizers with a comorbid SUD/internalizing disorder represent a group of individuals that have crossed a symptom- or impairment- specific threshold. Within the current sample, a certain threshold of comorbid symptomatology could exist and be an informal requirement for one to

“surpass” before treatment utilization becomes more likely. Future research could closely

examine internalizing symptoms to determine whether certain symptoms possess more “weight”

in the decision to seek treatment when they co-occur with a SUD, but this proposal is beyond the scope of the current project.

Externalizing Psychopathology

As previously mentioned, the first hypothesis predicted treatment utilizers would possess greater levels of broad externalizing psychopathology than treatment neglecters. Not only did

results yield support for this hypothesis, but the difference between the two groups on the Broad

Externalizing composite was far greater than the differences observed between treatment

neglecters and utilizers on both the internalizing psychopathology composite and the PD

pathology composite.

Partial support was found for the first component of the second hypothesis, in that results

confirmed the prediction that alcohol and drug dependence pathology would be greater in

treatment utilizers, but it was discovered that levels of ASPD symptomatology were greater in

treatment utilizers rather than treatment neglecters. Specifically, moderate effect sizes indicated

the two groups greatly differed in endorsement of ASPD symptomatology, alcohol and drug

dependence symptomatology, and substance use withdrawal symptoms. An examination of

medians between treatment utilizers versus neglecters revealed that treatment utilizers endorsed

twice as many alcohol dependence symptoms as treatment neglecters, and four times as many

drug dependence symptoms than treatment neglecters. Results confirmed the prediction from the

latter part of the second hypothesis, in that increased levels of alcohol or drug dependence

symptoms substantially increased the likelihood of seeking treatment.

Current findings demonstrating increased levels of substance use pathology in treatment

utilizers mirror results from past studies (e.g., Cohen et al., 2007; Salloum et al., 1998), further

reinforcing the notion that the likelihood of SUD treatment utilization increases as substance use

symptomatology increases. Given individuals with comorbid disorders are likely to experience

greater levels of distress and impairment compared to those with independent disorders (Lewis,

Bucholz, Spitznagel, & Shayka, 1996), it should not be surprising that comorbidity resulted in

greater treatment utilization results. Moreover, it would be surprising to find individuals in

substance use treatment setting who do not possess substance use problems, as those who seek

treatment should theoretically possess moderate levels of substance use (Cohen et al., 2007; Wu

et al., 1999).

It is interesting that treatment utilizers possessed twice as many ASPD symptoms as

treatment neglecters. These results are intriguing given the wide array of literature describing

the treatment-resistant and generally oppositional nature of this disorder (e.g., Pettinati, Pierece,

Belden, & Meyers, 1999; Powell & Peveler, 1996; Reich & Russell, 1993). However, logistic

regressions demonstrated individuals with presence of ASPD were three times more likely to

seek SUD treatment than without this disorder. This finding may speak to the high degree of

overlap in latent pathology between the two disorder such as impulsivity (Trull, Waudby, &

Sher, 2004), in addition to high levels of substance use frequently observed within ASPD (Hasin

et al., 2011; Regier et al., 1990). Another potential reason for the high rates of SUD treatment utilization among those with ASPD could be that those with this disorder may have sought treatment because it was court ordered. While some research has reported high rates of court-

ordered SUD treatment (e.g., Saunders, Zygowicz, & D'Angelo, 2006), further research must

investigate whether those with comorbid ASPD/SUD are independently seeking treatment or if they are being forced to do so.

Personality Pathology

Small, statistically significant effect sizes yielded from analyses comparing treatment utilizers to neglecters on a dimensionalized personality disorder (PD) pathology variable provided little support that the two groups greatly differed in mean PD pathology. However, logistic regressions revealed those who possessed a comorbid PD diagnosis were significantly more likely to seek treatment than those without a PD diagnosis. This increased likelihood was particularly amplified when a SUD was comorbid with Dependent PD, with results indicating that those with comorbid DPD/SUD were four times more likely to seek treatment than those without comorbid DPD.

Again, it may that appear strange that treatment utilizers did not greatly differ in PD pathology from treatment neglecters, yet those with actual PD diagnoses were more likely to seek treatment than those who did not. However, there are multiple explanations for this finding.

As with the case of internalizing psychopathology, it is quite possible that individuals who received PD diagnoses represented an extreme level of PD pathology, that when comorbid disorders with substance use pathology, coalesced to create a sufficient level of impairment that would drive an individual towards treatment utilization. Additionally, it is also plausible that a certain threshold of PD pathology must be passed in order to significantly predict whether an individual decides to seek treatment. This notion is supported by past research demonstrating higher levels of impairment among individuals with comorbid SUD/PDs than independent PDs

(Ruegg & Frances, 1995; Pulay et al., 2008), but should be subjected to further investigation.

Nonetheless, results still reaffirm past findings that those with comorbid PDs and SUDs tend to consume larger amounts of mental health resources compared to individuals with independent SUDs (Bender et al., 2002; Vaughn et al., 2010). Moreover, others have suggested that a combination of SUD and PD pathology requires excessive amounts of treatment resources to achieve outcomes similar to those who only possess a SUD (Moggi et al., 2011). Thus, the current study demonstrates that those with comorbid SUD/PD pathology are actually more likely to seek treatment than those without comorbid PDs, presumably because of the combinatory impairment resulting from both PD pathology and substance use pathology (Ekleberry, 2009;

Pulay et al., 2008). Similarly, one could propose that the high levels of comorbidity between internalizing symptoms (e.g., guilt, anxiety) that typically accompany substance use (e.g., Cohen et al., 2007) and PD pathology may create a unique form of impairment that motivates individuals to seek treatment.

It is also important to mention that some individuals have expressed concern that the methodology used to create PD diagnoses in the NESARC dataset have inflated true national prevalence rate of PDs (e.g., Trull et al., 2010). While NESARC procedures followed DSM-IV

(APA, 2000) guidelines regarding the number of criteria required to be given PD diagnosis (e.g., four of seven criteria are required for a diagnosis of Paranoid PD), NESARC procedures only required that one of these criterion was accompanied by self-reported impairment (i.e., asking the participant whether a given symptom caused impairment in work, school, or broad interpersonal functioning). Thus, an individual could possess “pathological” personality traits, not report much impairment, but still be given a PD diagnosis. As previously stated, this procedure could purportedly result in a gross inflation of PD prevalence rates. Although beyond the scope of the

current project, it would be useful for future research to create more reliable and valid PD

diagnosis algorithms in order to compare treatment rejecters and utilizers on such diagnoses.

How does Comorbidity between Internalizing, Externalizing, and Personality Psychopathology Affect Treatment-Seeking Behavior?

The third hypothesis proposed comorbidity between psychopathology composites would be greater in treatment utilizers than treatment neglecters, thereby supporting that notion that treatment utilization should become more likely as comorbidity increases. Results partially supported this hypothesis, in that not all combinations of comorbidity significantly differed between the two groups. First, it should be noted that strong correlations between various forms of psychopathology, such as internalizing and externalizing psychopathology, further attest to the high rates of comorbidity observed among substance users.

When examining comorbidity between internalizing psychopathology and other psychopathology composites, such as externalizing psychopathology, broad SUD pathology,

ASPD symptoms, and PD pathology, point-biserial correlations consistently yielded moderate to large effect sizes for both groups. Furthermore, Fisher’s r-to-Z transformations revealed these correlations (i.e., comorbidity) between internalizing and respective psychopathology composites were repeatedly stronger in treatment utilizers than treatment neglecters. In particular, correlations between internalizing psychopathology and externalizing psychopathology, along with those between internalizing psychopathology and PD pathology, were especially stronger in treatment utilizers.

Broadly speaking, these data provide preliminary support for the notion that internalizing psychopathology serves as a powerful catalyst of treatment utilization when combined with various other forms of psychopathology. The finding that the addition of internalizing psychopathology to other forms of psychopathology repeatedly resulted in increased rates of

treatment utilization may again indicate that the negative affect captured by internalizing disorders (e.g., anxiety, fear, guilt) may be an especially powerful motivator for treatment utilization. Support for this notion may be found in studies that have found high rates of internalizing disorders in treatment samples (Compton et al, 2007; Wu et al., 1999).

Results yielded moderately sized, positive associations for both treatment utilizers and neglecters when correlating broad SUD symptomatology and ASPD symptomatology. Fisher’s r-to-Z transformations indicated the associations between these two composites were stronger in treatment utilizers than neglecters, indicating that the relationship between these two forms of externalizing psychopathology may be stronger in treatment utilizers than treatment neglecters.

This finding further supports the notion that treatment becomes more likely as comorbidity increases (Regier et al., 1990).

Results indicated the strength of the relationship between externalizing psychopathology and PD Pathology did not significantly differ between the two groups. Likewise, results did not indicate that comorbidity between PD pathology and ASPD symptoms significantly differed between the two groups.

As previously mentioned, some of these results may indicate that the addition of internalizing psychopathology to other forms of psychopathology is likely to incite motivation to seek treatment. One might posit that the negative affect (e.g., anxiety, guilt, worry) that may accompany substance use becomes a driving force towards treatment because of internalizing disorders’ often ruminative and worrisome nature. Generally speaking, the current findings expand on past research (e.g., Compton et al., 2007; Regier et al., 1990; Wu et al., 1999) by affirming treatment utilization becomes more likely in most cases of comorbidity.

What is the Typology of Alcohol and Drug Treatment Rejection?

Results support the fourth hypothesis, which proposed that treatment rejection would be multifactorial. The number of items subsuming the three factors were almost identical for both alcohol and drug treatment rejection, with the exception of one item (“Didn’t think anyone could help”) that was excluded from the drug treatment CFA because it loaded weakly on its projected factor and worsened model fit.

The Objective Barriers factor was comprised of items perceived to be external to the individual’s immediate control, such as financial restrictions (e.g., lack of insurance, could not pay bill), a perceived inability to find treatment (e.g., lack of knowing where to seek help or not having transportation), or general pessimism regarding whether anyone could help. The items subsuming this factor demonstrated excellent loading homogeneity across both the alcohol and drug treatment rejection samples.

The Psychological Barriers factor was largely characterized by content suggesting a fear of processes that may be involved in treatment (e.g., treatment modality, hospitalization) and how others might perceive the individual after learning they sought treatment. In addition, the

items subsuming this factor can be explained by the level of social stigma that surrounds

substance use treatment (Saunders, et al., 2006). Additionally, other studies have found moral

pressure or fears of social judgment may create barriers to treatment (Klingemann, 1991).

Interestingly, items with the lowest loadings for the alcohol Psychological Barriers factor reflected beliefs that (1) the problem would get better by itself, and (2) that one should be strong enough to handle the drinking problem. That is, it appears the extent to which these items contribute to the meaning of this factor (for alcohol treatment rejction) is somewhat more limited compared to the other items subsuming this Factor. Loadings for the other five items that

comprised the Psychological Barriers factor in the alcohol CFA were quite homogenous with

respect to factor loadings, whereas the aforementioned items greatly differed in loading

magnitude. This is likely because endorsement rates for these two items were significantly

greater than the other items subsuming the treatment rejection factors (see Table 5), and thus these items became less discriminating because many potential reasons, not just treatment rejection, are associated with item endorsement.

While the loadings for these items were somewhat discrepant among the two samples

(drug and alcohol treatment rejecters), they were still above conservative cutoffs for factor analysis (i.e., > .40; Ford, MacCallum, & Tait, 1986). This observation does not serve as a suggestion that these two items represent their own factor or construct, especially since a minimum of three indicator variables is usually recommended to accurately measure a latent factor in a multi-factor model (Velicer & Fava, 1998). Nevertheless, the meaning of this factor is best understood in terms of those items that had strong item loadings and thus were good item discriminators on the underlying latent treatment rejection trait.

Lastly, the “Self-Focused Barriers” factor consisted of items insinuating resistance (e.g.,

“wanted to keep using drug,” “did not want to go to treatment”) or a lack of acknowledgement that a problem existed (e.g., “family thought treatment was necessary but I disagreed,” and

“didn’t think problem was serious enough to warrant treatment”). It appears the items subsuming this factor tap into the aspects of denial and disconnection of responsibility that is frequently observe in those with substance use (DiClemente, 2003). Many substance users may view others (e.g., family) as the individuals with “the problem,” rather than appreciating the possibility that their behavior may be problematic.

The patterns of interfactor correlations differed somewhat between the alcohol treatment

rejection sample and the drug treatment rejection sample. Significantly higher interfactor

correlations were observed in the alcohol treatment rejection CFA than in the drug treatment

rejection CFA. A follow-up Fisher’s r-to-Z transformation demonstrated these intercorrelations

significantly differed across the two groups. Given the exploratory nature of these analyses,

even within the context of the CFAs, it is difficult to identify a sole reason for this finding. One

potential reason could be the high degree of overlap between the two groups with respect to

substance use.

It was discovered that a larger percentage of the drug treatment rejection group also endorsed alcohol treatment rejection items (roughly 39 %), while smaller percentage of alcohol

treatment rejecters endorsed drug treatment rejection (approximately 21 %). This overlap could

reduce the “purity” of each respective sample by artificially inflating (within the alcohol group)

or suppressing (within the drug group) the true relationship between these three factors. This

finding is likely because the majority of alcohol users never become illicit drug users, while most

drug users began their substance use patterns with alcohol (Freshman, 2004). Obtaining a more

balanced and homogenous group of substance users may serve as a better basis of comparing the

factor structure of treatment rejection across different types of substance users.

Similarly, it may be that drug treatment rejection factors are naturally more orthogonal,

suggesting that increases or decreases in the number of certain types of rejection reasons (e.g.,

Objective Barriers) does not necessitate that similar patterns will occur in other factors (e.g.,

Psychological Barriers). The finding that intercorrelations between the Psychological Barriers

and Self-Focused Barriers were similar in magnitude across both samples is likely because both

factors are non-objective in nature.

What Variables Best Predict Treatment Rejection?

Results indicated internalizing psychopathology predicted increased levels of objective barriers for both alcohol and drug treatment rejection. Low mood, anhedonia, pessimism, and other symptoms that typically characterize internalizing psychopathology could be responsible for the finding that high levels of internalizing psychopathology are positively associated with perceived objective treatment barriers, or at least that these types of reasons are given for failing to seek treatment when a treatment need was also avowed. Given past research has found strong relationships between mood disorders such as depression and an external locus of control (e.g.,

Harrow, Hansford, & Astrachan-Fletcher, 2009), one might even propose that an individual who is depressed or anxious is likely to perceive the decision to enter treatment as out of their control, presumably resulting in higher endorsement rates of objective treatment barriers. This proposal may be further supported by other research demonstrating that a perceived need for treatment is more likely when anxiety and depression are comorbid with substance use problems (Green-

Hennessy, 2002). It would be useful for future research to directly focus on the roles negative affect and general pessimism may serve in the decision to reject treatment.

Results indicating externalizing psychopathology predicted higher levels of perceived objective treatment barriers could be explained by the impulsivity and lack of responsibility often seen in individuals with high levels of both substance use and ASPD symptomatology

(Sher & Trull, 1994). Similarly, it is possible that the monetary costs associated with excessive alcohol and drug use could result in a lower amount of income that could potentially be used towards substance treatment, but this notion must be addressed in future studies.

The finding that PD pathology predicted lower levels of perceived objective alcohol treatment barriers could be the product of a suppression effect produced by the high

interrelationship between internalizing psychopathology and PD pathology. It should be reiterated that structural path coefficients measure change in the dependent variable while controlling for other potential correlational influences. Thus, the high degree of overlap (i.e., correlation) between internalizing psychopathology and PD pathology, along with the robust ability of internalizing psychopathology to predict perceived objective barriers, could be responsible for the negative relationship between PD pathology and objective barriers to treatment. On the other hand, if the negative predictive path is a valid finding, then it suggests that more PD pathology is associated with less endorsement of objective barriers to seek treatment. Again, this notion must be further tested due to the exploratory nature of these analyses.

Results indicated internalizing psychopathology was predictive of greater perceived psychological barriers to both alcohol and drug treatment. This finding is understandable given that the psychological barriers factor is largely characterized by a fear of social stigma surrounding SUD treatment utilization, as well as general unease about processes involved in

SUD treatment. Notably, the ability of internalizing psychopathology to predict increased levels of psychological barriers was much greater in the drug treatment item model, potentially representing a higher degree of stigma surrounding drug treatment when compared to alcohol treatment. Because internalizing psychopathology is largely characterized by fear and anxiety, one could easily understand why fear and worry surrounding the social implications of SUD treatment could be predicted by internalizing psychopathology. Accordingly, past literature has also found that attitudes reflecting a fear of social stigma from SUD treatment can be a significant barrier to treatment utilization (Cunningham, Sobell, & Chow, 1993; Klingemann,

1991; Saunders et al., 2006).

Structural analyses also indicated that externalizing psychopathology predicted increased

levels of psychological barriers for alcohol treatment rejection, but to a far lesser extent when

compared to internalizing psychopathology. Nonetheless, the exact reason for why externalizing

psychopathology predicted higher levels of psychological barriers for alcohol treatment is

somewhat enigmatic, but it is plausible that the treatment-resistant nature of externalizing

disorders like ASPD and substance dependence (Straussner & Nemenzik, 2007; Tucker, 2001)

could create preferences for self-help rather than seeking treatment.

Unfortunately, it is unknown why path coefficients from externalizing psychopathology

onto drug treatment rejection factors were non-significant. This non-significant finding could be

explained by the non-normality of the data, reduced variance given the nature of the sample, or

possibly model misspecification with respect to the role of externalizing disorder in treatment

rejection outcomes. Nonetheless, future research must further address whether externalizing

psychopathology plays a specific role in the decision to reject drug treatment.

The alcohol treatment rejection model found internalizing and externalizing

psychopathology were predictive of greater levels of self-focused treatment barriers, while the drug treatment rejection model only found internalizing psychopathology predicted increase levels of self-focused barriers. Moreover, the ability of internalizing psychopathology to predict

increased levels of self-focused barriers was greater in the drug treatment rejection item model

than in the alcohol treatment rejection item model. As such, these results suggest that symptoms

of psychopathology play a significant role in rejecting treatment for a SUD.

As previously mentioned, self-focused treatment rejection items largely assess a general lack of desire to treat substance use problems (e.g., “Didn’t want to go to treatment”), along with general pessimism regarding potential benefits from seeking treatment (“Tried to get help before

but it didn’t work”). The relatively weaker ability of externalizing psychopathology to predict self-focused alcohol treatment barriers could reflect a latent threshold level of substance use problems that may be required before one enters treatment. That is, past research has shown individuals with alcohol use problems are not typically motivated to seek treatment until impairment from alcohol use has arose (Cunningham, Sobell, & Sobell, 1993) or they have been forced to seek treatment (Tucker & Simpson, 2001). Similarly, the finding that internalizing psychopathology predicted increases in self-focused barriers across both alcohol treatment rejection samples might be best explained by examining the latent nature of these barriers. The second component of this treatment rejection factor reflects general pessimism regarding the efficacy of treatment utilization. As previously stated, the negative affect comprising the internalizing psychopathology spectrum may largely explain such pessimistic attitudes towards treatment.

It is interesting that when comparing treatment neglecters to treatment utilizers, the latter group was found to have higher levels of psychopathology, indicating that the comorbidity of

SUD with internalizing and externalizing psychopathology could lead to treatment utilization.

However, exploratory structural equation modeling analyses indicated that more psychopathology led to greater endorsement of treatment barriers. That is, it seems that high levels of psychopathology can lead to either treatment rejection or utilization. Accordingly, one might propose treatment rejection is likely during the beginning stages use because psychological and physical symptoms associated with substance use and comorbid psychopathology have not been present long enough to motivate an individual towards treatment.

Thus, it would be interesting to determine if long periods of substance use impairment have a direct impact on whether the individual decides to seek or reject treatment. Future research

should investigate whether the likelihood of treatment rejection depends on how long the individual has been experiencing symptoms associated with substance use and comorbid psychopathology.

Conclusions

The current study had two overall goals, the first of which was to determine whether individuals who utilized treatment possessed greater levels of internalizing, externalizing, substance use disorder (SUD), and personality disorder (PD) psychopathology than individuals who neglected treatment. Results confirmed the notion that individuals who had sought alcohol and or drug treatment possessed greater levels of the aforementioned forms of psychopathology.

While treatment utilizers demonstrated higher levels of internalizing and externalizing psychopathology, broad SUD pathology, and PD pathology, these differences were most found for externalizing psychopathology. This finding is not surprising given substance use partially subsumes the externalizing psychopathology spectrum, and thus those who are seeking treatment for substance use problems would likely possess great levels of externalizing psychopathology.

Moreover, results also suggested that SUD treatment utilization becomes more likely as

SUD pathology becomes comorbid with other forms of psychopathology, especially internalizing psychopathology. These findings implicate internalizing psychopathology as a robust predictor of treatment utilization, presumably because of negative affect linked to internalizing psychopathology combined with impairment resulting from substance use.

The second goal of this study was to better understand the latent structure of SUD treatment rejection, and to investigate what variables best predicted these treatment rejction factors. Confirmatory factor analyses (CFAs) indicated a three-factor model fit best in accounting for the various reasons individuals report for rejecting treatment. These three factors

were labeled as follows: Objective Barriers, Psychological Barriers, and Self-focused Barriers.

Structural equation modeling (SEM) analyses for the alcohol and drug treatment rejection items indicated internalizing psychopathology may be especially powerful when trying to predict with an individual with a SUD will reject treatment.

Limitations and Future Directions

The current study is not without limitations. As with most studies examining treatment utilization rates in SUD populations (e.g., Cohen et al., 2007; Wu et al., 1999), an overwhelming majority of the current sample never sought treatment. As previously mentioned, the current study aimed to understand latent treatment rejection processes within a subgroup of treatment neglecters who deliberately did not seek treatment (treatment rejecters). Although a large percentage of the total NESARC study sample (63 %) was asked the screening question gauging drug and/or alcohol treatment rejection (“was there every a time where you thought you should go to treatment, but didn’t go for some reason?”), only those who provided an affirmative response were asked follow-up questions on treatment rejection. Thus, only a small subgroup of the treatment neglecters actually provided data on why they rejected treatment. Obtaining a better balance between the number of treatment rejecters and those who never thought they should have sought treatment despite clinically relevant substance use may allow for more valid cross-group comparisons. Because denial and/or a lack of insight into clinically relevant substance use problems precluded individuals from providing data on treatment rejection, future research should consider administering treatment rejection items to those with substance use problems, regardless of whether the individuals believed they should have sought treatment.

For the treatment rejection analyses (i.e., research questions four and five), it was discovered that some of these treatment rejecters had sought informal treatment in the past.

While this may insinuate that these individuals should be classified as treatment utilizers rather

than rejecters, it is important to reiterate that these research questions focused on why substance

users rejected treatment in the first place, regardless of whether treatment rejection occurred

before or after informal treatment. Nonetheless, future research must ensure treatment rejection

and utilization groups are as mutually exclusive as possible in order to maximize external

validity.

Similarly, some individuals in the drug treatment rejection group also reported rejecting

the notion of alcohol treatment during their lifetime, and thus their data was included in both

treatment rejection groups. This finding is expected because alcohol, rather than illicit drugs, is

much more likely to be a substance user’s initial substance (Freshman, 2004). However, this

overlap may have skewed results because the drug treatment rejection group was less

homogenous than originally intended. Designing a study that included two independent groups

(i.e., drug-only and alcohol-only), along with a “mixed” group, might increase the homogeneity

of groups. One might expect an increase in the external validity of the findings with sounder

group homogeneity, potentially allowing for cross-group factor structure comparison on treatment rejection models.

The epidemiological nature of the NESARC study has considerable strengths with respect to its over-inclusion of individuals often excluded from such studies (e.g., ethnic minorities), diverse geographic sampling, and the inclusion of many Axis I and II diagnostic inquiries. However, future research interested in the measuring the broad scope of psychopathology should consider including all major Axis I and II conditions. Such a procedure would be undoubtedly taxing and lengthy, but is crucial to truly understanding the latent structure of mental disorders given the vast comorbidity and symptom overlap between

disorders. While this limitation is especially relevant to the current study, one should keep in

mind that the main focus of the NESARC study was alcohol and drug related conditions, not

internalizing and externalizing psychopathology.

Another limitation of the current study is the extent of missing data within the NESARC

dataset. To be conservative, participants needed 85 % complete data for composite variables to

be included in the analyses. However, the mix of missing completely at random (MCAR) and

missing not at random (MNAR, as with the treatment rejection items), data limits its utility and

the generalizability of the results from the current study.

Additionally, a limitation inherent to the NESARC dataset and the current study is the

inability to identify simultaneous periods of substance use and psychopathology comorbidity with absolute certainty. That is, NESARC questions were designed in a manner that made it difficult to ascertain whether certain periods of psychopathology were undoubtedly experienced at the same time as SUD periods. Most studies examining concurrent disorders within the

NESARC dataset use diagnoses original to the dataset. Given others’ concerns about the formulation of diagnoses in the NESARC data (Trull et al., 2010), it would have been useful to obtain the exact algorithms used to create these diagnoses. However, these algorithms were not

made publicly available and the creation of such diagnostic variables would be too onerous for

the scope of the current project.

Lastly, an overwhelming majority of treatment rejecters reported their reasons for

rejecting treatment were only relevant during unspecified time only known as “before the last 12

months.” Although it would have been optimal to match up time of substance use,

psychopathology, and treatment rejection, the degree of which treatment rejection was related to

various forms of psychopathology is notable nonetheless. Moreover, given the normally

distributed nature of externalizing psychopathology, a lifetime diagnosis may serve as the best tool for detecting a diathesis for psychopathology rather than a current psychological condition.

However, future research must improve measurement of psychopathology and treatment rejection to establish a more reliable temporal framework.

Table 1

Descriptive Statistics for Psychopathology and Treatment Rejection Composite Variables

Variable (# of items) n MIC Mdn M SD Skew Kurt Internalizing Psychopathology (59) 12,433 .18 .19 .23 .22 0.70 -0.33 Major Depressive Disorder (9) 5,304 .39 .78 .71 .25 -0.74 -0.39 Dysthymia (6) 1,126 .54 1.00 .85 .24 -1.87 3.23 Generalized Anxiety Disorder (6) 1,421 .52 .83 .71 .29 -0.85 -0.18 Social Phobia(13) 2,272 .44 .38 .40 .25 0.40 -0.55 Simple Phobia(12) 12,322 .47 .00 .10 .15 1.97 4.61 Panic Disorder w/out Agoraphobia 2,403 .40 .54 .51 .26 0.07 -0.81 (13) Externalizing Psychopathology(34) 12,691 .39 .18 .21 .17 1.34 2.16 Substance Use Pathology (18) 12,576 .53 .28 .34 .25 0.72 -0.14 Alcohol Dependence(9) 12,612 .70 .33 .38 .28 0.45 -0.72 Drug Dependence(9) 6,598 .75 .11 .25 .29 1.24 0.47 Antisocial Personality Disorder(16) 12,409 .47 .06 .10 .14 2.02 4.93 Personality Pathology (64) 12,311 .31 .13 .16 .13 1.21 1.77 Alcohol Treatment Rejection (19) 1,063 .31 2.00 2.68 2.68 2.14 5.34 Objective barriers(5) 1,063 .40 0.00 .49 .86 2.07 4.62 Psychological barriers(7) 1,063 .39 1.00 1.27 1.38 1.51 2.43 Self-focused barriers (7) 1,063 .44 0.00 .91 1.34 1.77 2.97 Drug Treatment Rejection (18) 562 .27 2.00 2.97 2.77 1.68 3.37 Objective barriers(4) 562 .48 0.00 .45 .81 2.03 3.94 Psychological barriers(7) 562 .41 1.00 1.60 1.63 1.19 .91 Self-focused barriers (7) 562 .49 0.00 .93 1.40 1.79 3.06

Note. MIC = Mean Inter-item Correlation for composite scales. Mdn = Mean; M = mean; SD = Standard Deviation; Skew = Skewness; Kurt = Kurtosis. Because all symptom variables in the NESARC dataset were coded a ‘0’ (absence of symptom) or ‘1’ (presence of symptom), the means in this table represent the average number of symptoms endorsed for a specific scale. For instance, individuals who had data on the alcohol dependence items (9 items) would be expected to endorse 3 symptoms (3/9 = .33) on average.

Table 2

Demographics

Frequency (%)

Treatment neglecters Treatment Utilizers Entire sample Variable (n = 10,415) (n = 2212) (N = 12,707) (82.0%) (17.4%) Male 7864 (61.9) 6326 (60.7) 1485 (67.1) Gender Female 4843 (38.1) 4089 (39.3) 727 (32.9) Caucasian 8584 (67.6) 7112 (68.3) 1419 (64.2) Hispanic 1836 (14.4) 1477 (14.2) 345 (15.6) Ethnicity African-American 1789 (14.1) 1433 (13.8) 346 (15.6) American Indian 311 (2.4) 234 (2.2) 75 (3.4) Asian 187 (1.5) 159 (1.5) 27 (1.2) Single 3145 (24.8) 2546 (24.4) 587 (26.5) Married 6050 (47.6) 5225 (50.2) 791 (35.8) Living with partner 547 (4.3) 407 (3.9) 137 (6.2) Marital Status Separated 484 (3.8) 356 (3.4) 124 (5.6) Divorced 1994 (15.7) 1479 (14.2) 493 (22.3) Widowed 487 (3.8) 402 (3.9) 80 (3.6) < High school 1764 (13.9) 1326 (12.7) 422 (19.1) High school/GED 3476 (27.4) 2763 (26.5) 690 (31.2) Some college 3013 (23.7) 2417 (23.2) 574 (25.9) Education Associate’s degree 1213 (9.5) 1018 (9.8) 186 (8.4) Bachelor’s degree 1699 (13.4) 1523 (14.6) 174 (7.9) Post-Bachelor’s 1542 (12.1) 1368 (13.1) 166 (7.5)

Note. 80 (0.62%) participants with a SUD did not have data indicating whether they ever sought Substance Use Disorder treatment.

Table 3

Descriptive Statistics – Prevalence of DSM-IV Disorders in Sample (N = 12,707)

Prevalence of DSM-IV Disorder Diagnosis n % Any Internalizing disorder 5111 40.2 Major Depressive Disorder 3567 28.1 Dysthymia 1075 8.5 Generalized Anxiety 932 7.3 Disorder Social Phobia 997 7.8 Simple Phobia 1708 13.4 Panic Disorder w/out 846 6.7 Agoraphobia Any Substance Use Disorder 12,707 100.0 Alcohol Use Disorder 11, 843 93.2 Drug Use Disorder 4068 32.0 Comorbid AUD and DUD 3204 25.2 Any Personality Disorder 3201 25.2 (PD) Antisocial PD 1135 8.9 Avoidant PD 490 3.9 Dependent PD 101 0.8 Obsessive-Compulsive PD 1552 12.2 Paranoid PD 1043 8.2 Schizoid PD 690 5.4 Histrionic PD 494 3.9

Note. AUD = Alcohol Use Disorder; DUD = Drug Use Disorder. Percentages represent the proportion of individuals in the current study’s sample who met criteria for a respective disorder.

Table 4

Type of SUD Treatment Obtained (n = 2212)

Reported use of treatment Type of treatment form (%) Alcoholics/narcotics/cocaine anonymous or 12-step meeting 70.8 Alcohol/drug rehabilitation program 44.2 Private physician, psychiatrist, psychologist, social worker or other professional 40.5 Alcohol/drug detoxiﬁcation ward/clinic 33.8 Outpatient clinic, outreach and day/partial patient program 30.1 Emergency Room 26.1 Inpatient ward of psychiatric/general hospital/community mental health program 25.9 Family services or other social service agency 23.0 Clergyman, priest or rabbi 15.4 Other agency or professional 13.2 Halfway house/therapeutic community 8.6 Employee assistance program (EAP) 7.4 Crisis center 4.8 Methadone Maintenance Clinic 1.9

Note. These treatment categories are not mutually exclusive with respect to alcohol and drug treatment. The “Methadone Maintenance Clinic” category represents individuals with a DUD who sought treatment (n = 741), as alcohol use disorders are not the primary treatment focus of these clinics.

Table 5

Reasons for Treatment Rejection among Those with an AUD or DUD

Frequency (%) Reason given AUD (n = 1063) DUD (n = 562) Wanted to go, but health insurance didn’t cover 8.5 10.9 Didn’t think anyone could help 15.0 18.0 Didn’t know any place to go for help 7.4 9.8 Couldn’t afford to pay the bill 14.1 19.6 Didn’t have any way to get there 4.3 4.4 Didn’t have time 8.7 10.7 Thought the problem would get better by itself 32.5 33.8 Was too embarrassed to discuss it with anyone 19.0 25.8 Was afraid of what others would think 8.8 16.7 Thought I should be strong enough to handle problem alone 40.8 43.1 Was afraid they would put me into the hospital 7.8 13.5 Was afraid of the treatment they would give me 7.7 9.1 Hated answering personal questions 10.7 18.1 The hours were inconvenient 2.6 3.0 A member of my family objected 1.0 0.9 Family thought I should go but I didn’t agree 10.3 6.9 Can’t speak English very well 0.4 0.4 Was afraid I would lose my job 2.9 5.0 Couldn’t arrange for child care 0.7 2.0 Had to wait too long to get into a program 1.5 3.2 Wanted to keep drinking/using drug 16.1 17.4 Didn’t think drinking/drug problem was serious enough 19.5 16.9 Didn’t want to go 16.3 19.9 Stopped drinking/using drug on my own 19.0 19.0 Friends or family helped me stop drinking/using drug 5.0 5.2 Tried getting help before and it didn’t work 5.3 6.9 Other reason 7.8 5.7 Note. AUD = Alcohol Use Disorder. DUD = Drug Use Disorder. Because a response on these items necessitated an acknowledgment that a SUD problem existed (i.e., “Was there ever a time when you thought you should see someone for your alcohol/drug use, but you didn't go?”), a large proportion of the sample did not respond to these items.

Table 6

Research Question 1 – Comparing Treatment Neglecters vs. Utilizers on Broad Psychopathology Composite Scores

Effect Size Composite variable n Median U statistic Z (r = Z /√N)

INT Treatment neglecters 10,238 .18 8,885,027 -14.29 r = .13 psychopathology Treatment utilizers 2152 .29

EXT Treatment neglecters 10,409 .15 4,581,034 -44.51 r = .40 psychopathology Treatment utilizers 2209 .35

Broad SUD Treatment neglecters 10,314 .22 5,174,547 -40.01 r = .36 pathology Treatment utilizers 2192 .56

Treatment neglecters 10,161 .13 PD pathology 9,237,852 -10.05 r = .09 Treatment utilizers 2111 .16

Note. All effect sizes statistically significant at p < .001. INT = Internalizing psychopathology; EXT = Externalizing psychopathology; SUD = broad Substance Use Disorder pathology; PD = Personality Disorder.

Table 7

Research Question 2 – Comparing INT Psychopathology Symptoms in Treatment Neglecters and Utilizers

Effect Size Composite variable n Median U statistic Z (r = Z /√N) MDD Treatment neglecters 4107 .78 2,008,624 -9.12 r = .13*** Treatment utilizers 1183 .89 Dysthymia Treatment neglecters 739 1.00 126,427 -3.47 r = .10*** Treatment utilizers 386 1.00 GAD Treatment neglecters 1051 .83 177,707 -2.29 r = .06* Treatment utilizers 367 .83 Social Phobia Treatment neglecters 1816 .39 376,199 -2.69 r = .08** Treatment utilizers 451 .39 Specific Phobia Treatment neglecters 10167 .00 10,176,219 -4.07 r = .04*** Treatment utilizers 2112 .08 PD w/out Treatment neglecters 1812 .46 392,351 -9.69 r = .20*** Treatment utilizers 589 .62

Note. INT = Internalizing psychopathology; MDD = Major Depressive Disorder; GAD = Generalized Anxiety Disorder; PD w/out = Panic Disorder without Agoraphobia. *= p <.05; **= p < .01;***= p < .001.

Table 8

Research Question 2 – Comparing EXT Psychopathology Symptoms in Treatment Neglecters and Utilizers

Effect Size Composite variable n Median U statistic Z (r = Z /√N)

Alcohol Treatment neglecters 10,415 .33 5,742,911 -37.35 r = .33 Dependence Treatment utilizers 2212 .67

Drug Treatment neglecters 5069 .11 2,200,270 -26.29 r = .32 Dependence Treatment utilizers 1529 .44

Treatment neglecters 10235 .06 ASPD 6,352,172 -31.76 r = .29 Treatment utilizers 2134 .13

Note. EXT = Externalizing psychopathology; ASPD = Antisocial Personality Disorder. All effect sizes statistically significant at p < .001.

Table 9

Research Question 2 – Comparing DSM-IV Diagnoses in Treatment Neglecters and Utilizers

Diagnosis B SE B OR 95% CI Any Internalizing disorder 0.52 0.05 1.68 [1.54, 1.85] Major Depressive Disorder 0.63 0.05 1.88 [1.70, 2.06] Dysthymia 1.02 0.07 2.79 [2.43, 3.19] Generalized Anxiety Disorder 0.53 0.08 1.69 [1.45, 1.98] Social Phobia 0.25 0.08 1.29 [1.10, 1.51] Simple Phobia 0.15 0.07 1.17 [1.02, 1.33] Panic Disorder 0.69 0.08 2.00 [1.70, 2.33] Any Externalizing disorder* † † † † Alcohol Use Disorder 2.56 0.31 12.91 [7.10, 23.45] Drug Use Disorder 2.37 0.15 10.73 [7.99, 14.39] Antisocial Personality Disorder 1.11 0.07 3.03 [2.66, 3.50] Any Personality Disorder (PD) 0.78 0.05 2.18 [1.98, 2.40] Antisocial PD 1.10 0.07 3.03 [2.66, 3.50] Avoidant PD 0.99 0.09 2.68 [2.21, 3.24] Dependent PD 1.39 0.20 4.00 [2.70, 5.93] Obsessive-Compulsive PD 0.34 0.07 1.40 [1.23, 1.60] Paranoid PD 0.70 0.07 2.01 [1.74, 2.32] Schizoid PD 0.72 0.09 2.04 [1.72, 2.43] Histrionic PD 0.71 0.10 2.03 [1.66, 2.48]

Note. † = No values are listed for “Any Externalizing disorder” because having a SUD was the main criterion for sample inclusion. All variables were entered into their own logistic regression due to multicollinearity concerns. That is, multiple logistic regressions are represented in this table. n = 12, 627. All models significant at p < .001.

Table 10

Correlations between Symptom Composites among Treatment Neglecters (n = 10,061) and Treatment Utilizers (n = 2,094)

Internalizing Externalizing Substance dependence Antisocial PD

TN TU TN TU TN TU TN TU

Internalizing

Externalizing .331 .441

Substance dependence .242 .310 † †

ASPD .262 .322 † † .303 .356

PD pathology .487 .556 .392 .385 .259 .235 .374 .374

Note. Internalizing = Internalizing psychopathology; Externalizing = Externalizing psychopathology; Antisocial PD = Antisocial Personality Disorder; PD pathology = Personality Disorder pathology; TN = Treatment Neglecters, reflecting individuals who deliberately rejected alcohol and/or drug treatment; TU = Treatment Utilization. 552 cases were excluded from this analysis because of missing data. All correlations significant at p < .001. Bolded correlations indicate a statistically significant difference (p < .001) between the two groups based on Fisher’s r-to-Z transformations (see Table 11). † = Correlations that were excluded because of multicollinearity.

Table 11

Fisher’s r-to-Z Transformation Z-values and p-Values for Research Question 3

Substance Antisocial PD Internalizing Externalizing Dependence PD Pathology

Internalizing --

Externalizing -3.91*** --

Substance dependence -3.06** † --

ASPD -2.58* † -2.45* --

PD Pathology -4.01*** 0.29 1.06 -0.44 --

Note. Internalizing = Internalizing psychopathology; Externalizing = Externalizing psychopathology; Antisocial PD = Antisocial Personality Disorder; PD pathology = Personality Disorder pathology; 552 cases were excluded from this analysis because of missing data. *= p ≤ .05; **= p < .01; ***= p < .001. No asterisk indicates p > .05. † = Z-values that were excluded because of multicollinearity.

Table 12

Exploratory Structural Equation Modeling for Alcohol Treatment Rejection Items (n = 1063)

Reason given F1 F2 F3 Wanted to go, but health insurance didn’t cover .68 -.05 -.08 Didn’t think anyone could help .44 .20 .06 Didn’t know any place to go for help .62 .17 -.07 Couldn’t afford to pay the bill .80 -.01 -.19 Didn’t have any way to get there .76 .00 -.14 Didn’t have time* .35 .23 .10 Thought the problem would get better by itself -.06 .40 .18 Was too embarrassed to discuss it with anyone .13 .71 .00 Was afraid of what others would think .30 .71 -.04 Thought I should be strong enough to handle problem alone -.04 .40 .19 Was afraid they would put me into the hospital .45 .47 .01 Was afraid of the treatment they would give me .45 .47 .06 Hated answering personal questions .05 .24 .57 The hours were inconvenient* .36 .06 .42 A member of my family objected* .56 .14 .24 Family thought I should go but I didn’t agree* -.01 -.03 .80 Can’t speak English very well* .02 -.42 -.01 Was afraid I would lose my job* .46 .32 .18 Couldn’t arrange for child care* .32 .28 .43 Had to wait too long to get into a program* .69 -.25 .15 Wanted to keep drinking .13 -.01 .76 Didn’t think drinking problem was serious enough -.11 .03 .72 Didn’t want to go .01 .02 .78 Stopped drinking on my own -.15 -.05 .58 Friends or family helped me stop drinking .12 .01 .44 Tried getting help before and it didn’t work .35 -.01 .51 Other reason .32 -.34 .19

Note. Loadings > |40| bolded. CFI = .95, TLI = .94, RMSEA = .02. * = Items excluded from the final three factor model because of poor fit and/or cross- loadings.

Table 13

Exploratory Structural Equation Modeling for Drug Treatment Rejection Items (n = 562)

Reason given F1 F2 F3 Wanted to go, but health insurance didn’t cover .77 -.13 -.06 Didn’t think anyone could help* .25 -.02 .31 Didn’t know any place to go for help .67 .03 -.02 Couldn’t afford to pay the bill .70 .07 -.14 Didn’t have any way to get there .76 -.07 .01 Didn’t have time* .32 .17 .28 Thought the problem would get better by itself -.01 .38 .23 Was too embarrassed to discuss it with anyone .30 .77 -.07 Was afraid of what others would think .30 .56 .03 Thought I should be strong enough to handle problem alone .02 .45 .25 Was afraid they would put me into the hospital .45 .48 .01 Was afraid of the treatment they would give me .33 .59 .03 Hated answering personal questions .15 .54 .10 The hours were inconvenient* .53 .01 .31 A member of my family objected* .45 .02 .46 Family thought I should go but I didn’t agree* .00 .09 .80 Can’t speak English very well* .74 -.13 .54 Was afraid I would lose my job* .46 .06 .39 Couldn’t arrange for child care* .42 .05 .32 Had to wait too long to get into a program* .31 -.20 .45 Wanted to keep using drug .13 .08 .83 Didn’t think drug problem was serious enough .02 -.09 .77 Didn’t want to go .12 .03 .82 Stopped using drug on my own .01 -.09 .59 Friends or family helped me stop using drug .15 -.25 .55 Tried getting help before and it didn’t work .01 -.09 .80 Other reason .03 -.50 .43

Note. Loadings > |40| bolded. CFI = .94, TLI = .93, RMSEA = .02. * = Items excluded from the final three factor model because of poor fit and/or cross- loadings.

Table 14

Correlations among Demographic Variables and Treatment Rejection Factors in Alcohol (n = 1063) and Drug (n = 562) Treatment Rejecters

Objective Psychological Self-focused

Alc Drug Alc Drug Alc Drug

Sex .05 .00 .11** .05 .04 -.01

Race .03 .05 -.03 .00 -.01 -.04

Education -.09** -.13** .07* -.06 .03 .02

Income -.17*** -.12** .00 .02 -.05 .02

Note. Objective = Objective treatment barriers factor. Psychological = Psychological treatment barriers factor. Self-focused = Self-focused treatment barriers factor. Alc = Alcohol treatment rejection sample. Drug = Drug treatment rejection sample. * = p < .05; ** = p < .01; *** = p < .001. No asterisk = p > .05

Table 15

Parameter and Standard Error Estimates for Psychopathology Composites in the Alcohol Treatment Rejection SEM

Standardized Unstandardized Standard Error Model Parameters Estimate (β) Estimate (B) (SE) Major Depressive Disorder .76 1.00 .03 Dysthymia .78 1.03 .03 Loadings/effects on Panic Disorder w/out Agoraphobia .41 .54 .05 Internalizing Social Phobia .73 .96 .04 Psychopathology Specific Phobia .64 .84 .04 Generalized Anxiety Disorder .41 .97 .04 Average loading (or SE) on Internalizing Psychopathology .62 .89 .04 Antisocial Personality Disorder .66 1.00 .07 Alcohol Dependence .47 .71 .05 Loadings/effects on Nicotine Dependence .46 .69 .06 Externalizing Psychopathology Cocaine Dependence .75 1.14 .12 Marijuana Dependence .71 1.07 .08 Average loading (or SE) on Externalizing Psychopathology .61 .92 .08 Avoidant Personality Disorder .85 1.00 .04 Dependent Personality Disorder .76 .89 .06 Loadings/effects on Obsessive-Compulsive Personality Disorder .63 .75 .04 PD Pathology Paranoid Personality Disorder .84 .99 .03 Schizoid Personality Disorder .75 .88 .04 Average loading (or SE) on PD Pathology .77 .90 .04

Note. Internalizing = Internalizing psychopathology; Externalizing = Externalizing psychopathology; PD pathology = Personality Disorder pathology. SE = Standard Error. All coefficients significant at p < .001.

Table 16

Parameter and Standard Error Estimates for Treatment Rejection Composites in the Alcohol Treatment Rejection SEM

Standardized Unstandardize Standard Model Parameters Estimate (β) d Estimate (B) Error (SE) Wanted to go, but health insurance didn’t cover .61 1.00 .07 Didn’t think anyone could help .66 1.08 .06 Loadings/effects on Didn’t know where to get help .62 1.02 .08 Objective Barriers Couldn’t pay the bill .68 1.11 .05 No way to get there .64 1.05 .09 Average loading (or SE) on Objective Barriers .64 1.05 .07 Thought problems would get better by itself .36 1.00 .05 Too embarrassed to discuss it with anyone .63 1.76 .04 Afraid of what others would think .73 2.02 .05 Loadings/effects on Afraid of hospitalization .40 1.13 .04 Psychological Barriers I should be strong enough to handle it alone .77 2.15 .05 Afraid of treatment they would give me .79 2.18 .05 Hated answering personal questions .77 2.13 .05 Average loading (or SE) on Psychological Barriers .64 1.78 .05 Want to keep drinking or using drug/medicine .83 1.00 .03 Didn’t think problem was serious enough .60 .73 .05 Didn’t want to go .77 .92 .04 Loadings/effects on Self- Stopped drinking or using drug on my own .41 .49 .06 focused Barriers Friends/family helped me stop drinking/using drug .56 .68 .08 Tried getting help before and it didn’t work .79 .95 .06 Family thought I should go; I didn’t think so .78 .93 .04 Average loading (or SE) on Self-focused Barriers .68 .81 .05

Note. SE = Standard Error. All coefficients significant at p < .001.

Table 17

Parameter and Standard Error Estimates for Psychopathology Composites in the Drug Treatment Rejection SEM

Standardized Unstandardized Standard Model Parameters Estimate (β) Estimate (B) Error (SE) Major Depressive Disorder .74 1.00 .05 Dysthymia .68 .93 .05 Loadings/effects on Panic Disorder w/out Agoraphobia .33 .45 .07 Internalizing Social Phobia .79 1.07 .05 Psychopathology Specific Phobia .62 .84 .05 Generalized Anxiety Disorder .65 .88 .05 Average loading (or SE) on Internalizing Psychopathology .64 .86 .05 Antisocial Personality Disorder .45 1.00 .08 Alcohol Dependence .58 1.28 .07 Loadings/effects on Nicotine Dependence .53 1.17 .08 Externalizing Cocaine Dependence .62 1.38 .09 Psychopathology Marijuana Dependence .75 1.66 .08 Average loading (or SE) on Externalizing Psychopathology .59 1.30 .08 Avoidant Personality Disorder .86 1.00 .04 Dependent Personality Disorder .76 .89 .05 Loadings/effects on Obsessive-Compulsive Personality Disorder .62 .72 .06 PD Pathology Paranoid Personality Disorder .79 .92 .04 Schizoid Personality Disorder .69 .80 .05 Average loading (or SE) on PD Pathology .74 .87 .05

Table 18

Parameter and Standard Error Estimates for Treatment Rejection Composites in the Drug Treatment Rejection SEM

Standardized Unstandardized Standard Error Model Parameters Estimate (β) Estimate (B) (SE) Wanted to go, but health insurance didn’t cover .69 1.00 .09 Didn’t think anyone could help * * * Loadings/effects on Didn’t know where to get help .71 1.03 .10 Objective Barriers Couldn’t pay the bill .79 1.31 .07 No way to get there .59 .86 .14 Average loading (or SE) on Objective Barriers .70 1.05 .10 Thought problems would get better by itself .46 1.00 .06 Too embarrassed to discuss it with anyone .76 1.63 .05 Afraid of what others would think .67 1.43 .06 Loadings/effects on Afraid of hospitalization .59 1.27 .05 Psychological Barriers I should be strong enough to handle it alone .71 1.53 .06 Afraid of treatment they would give me .76 1.64 .07 Hated answering personal questions .60 1.30 .06 Average loading (or SE) on Psychological Barriers .65 1.40 .06 Want to keep drinking or using drug/medicine .83 1.00 .05 Didn’t think problem was serious enough .68 .82 .06 Didn’t want to go .80 .97 .05 Loadings/effects on Stopped drinking or using drug on my own .56 .67 .06 Self-focused Barriers Friends/family helped me stop drinking/using drug .50 .60 .11 Tried getting help before and it didn’t work .73 .87 .07 Family thought I should go; I didn’t think so .84 1.00 .06 Average loading (or SE) on Self-focused Barriers .70 .85 .07

Note. SE = Standard Error. * = the item loading was not significant and thus not included in the final model. All other coefficients significant at p < .001.

Figure 1. Research Question 1 – Comparing neglectors and utilizers on broad psychopathology composite scores. (Note. Percentages represent the proportion of substance abusers (12,707) with complete data on the composite variable used in the analysis. SUD Dx refers to a lifetime substance use disorder (i.e., abuse and/or dependence).

Figure 2. Research Question 2 – Comparing internalizing psychopathology symptoms in treatment neglectors and utilizers. (Note. Percentages represent the proportion of substance users with complete data [and thus included in the analysis] on the composit variable. MDD = major depressive disorder; DYS = dysthymia; GAD = generalized anxiety disorder; Social p = social phobia’ Specific p = specific phobia; Panic w/out = panic disorder without agoraphobia.)

Figure 3. Research Question 3 – Comparing externalizing psychopathology symptoms in treatment neglectors and utilizers. (Note. Percentages represent the proportion of substance users with complete data [and thus included in analyses] on the composite variable. ASPD = antisocial personality disorder.)

Figure 4. Research Question 3 – Does comorbidity explain alcohol or drug treatment rejection? (Note. The number of treatment neglectors is smaller than previously reported in this document because individuals with missing data for internalizing, externalizing, substance dependence, or personality disorder pathology composite variables were excluded [via listwise] from analyses.)

Figure 5. Research Question 4 – What is the typology of alcohol treatment rejection? (Note. *Individuals who met criteria for an alcohol disorder were asked, “Was there ever a time when you thought you should see someone for your drinking but you didn’t go?” Individuals who responded yes were asked about why they rejected treatment.)

Figure 6. Research Question 4 – What is the typology of drug treatment rejection? (Note. The large amount of data is presumably due to the low endorsement rate of drug use [i.e., treatment was never considered because illicit drugs were rarely, if ever, used]. *Individuals who met criteria for a drug use disorder were asked, “Was there ever a time when you thought you should see someone for your drug use, but you didn’t go?” Individuals who responded yes were asked a series of questions and thus provided data why treatment was rejected.)

Figure 7. Confirmatory factor analysis for drug treatment rejection items.

Figure 8. Confirmatory factor analysis for alcohol treatment rejection items.

Figure 9. Alcohol treatment rejection structural equation model.

Figure 10. Drug treatment rejection structural equation model.

APPENDIX

NESARC ALCOHOL AND DRUG TREATMENT REJECTION ITEMS

Alcohol Treatment Rejection Items 4a. Was there ever a time when you thought 1. Yes you should see a doctor, counselor, or other 2. No health professional or seek any other help for your drinking, but you didn’t go? 4b. Did this happen during the last 12 1. Yes months? 2. No

4c. Did this happen before 12 months ago, 1. Yes that is, before last (Month/one year ago)? 2. No

4d. What were your reasons for not getting 1. Wanted to go, but health insurance didn’t help? cover (Check all that apply.) 2. Didn’t think anyone could help 3. Didn’t know any place to go for help 4. Couldn’t afford to pay the bill 5. Didn’t have any way to get there 6. Didn’t have time

7. Thought the problem would get better by

itself

8. Was too embarrassed to discuss it with anyone 9. Was afraid of what my boss, friends, family, or others would think 10. Thought it was something I should be strong enough to handle alone 11. Was afraid they would put me into the hospital 12. Was afraid of the treatment they would give me 13. Hated answering personal questions 14. The hours were inconvenient 15. A member of my family objected 16. My family thought I should go but I didn’t think it was necessary 17. Can’t speak English very well 18. Was afraid I would lose my job 19. Couldn’t arrange for child care

20. Had to wait too long to get into a program

21. Wanted to keep drinking or got drunk

22. Didn’t think drinking problem was serious enough 23. Didn’t want to go 24. Stopped drinking on my own

25. Friends or family helped me stop drinking 26. Tried getting help before and it didn’t work 27. Other reason

Drug/Medicine Treatment Rejection Items 4a. Was there ever a time when you thought 1. Yes you should see a doctor, counselor, or other 2. No health professional or seek any other help for your drug use, but you didn’t go? 4b. Did this happen during the last 12 1. Yes months? 2. No 4c. Did this happen before 12 months ago, 1. Yes that is, before last (Month/one year ago)? 2. No 4d. What were your reasons for not getting 1 .Wanted to go, but health insurance didn’t help? cover (Check all that apply.) 2. Didn’t think anyone could help 3. Didn’t know any place to go for help 4. Couldn’t afford to pay the bill 5. Didn’t have any way to get there 6. Didn’t have time

7. Thought the problem would get better by

itself

8. Was too embarrassed to discuss it with anyone

9. Was afraid of what my boss, friends, family, or others would think 10. Thought it was something I should be strong enough to handle alone 11. Was afraid they would put me into the hospital 12. Was afraid of the treatment they would give me 13. Hated answering personal questions 14. The hours were inconvenient 15. A member of my family objected 16. My family thought I should go but I didn’t think it was necessary 17. Can’t speak English very well 18. Was afraid I would lose my job 19. Couldn’t arrange for child care 20. Had to wait too long to get into a program 92

21. Wanted to keep using a medicine or drug 22. Didn’t think medicine or drug problem was serious enough 23. Didn’t want to go 24. Stopped using a medicine or drug on my own 25. Friends or family helped me stop using a drug or medicine 26. Tried getting help before and it didn’t work 27. Other reason

REFERENCES

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed.; text revision). Washington, DC: Author.

Asparouhov, T., & Muthén, B. (2009). Exploratory structural equation modeling. Structural Equation Modeling, 16(3), 397-438. doi:10.1080/10705510903008204

Bank, B. (1985). Comparative research on the social determinants of adolescent drinking. Social Psychology Quarterly, 48(2), 164-177. doi:10.2307/3033611

Bender, D. S., Dolan, R. T., Skodol, A. E., Sanislow, C. A., Dyck, I. R., McGlashan, T. H., & ... Gunderson, J. G. (2001). Treatment utilization by patients with personality disorders. American Journal of Psychiatry, 158(2), 295-302. doi:10.1176/appi.ajp.158.2.295

Bentler, P. M., & Chou, C. P. (1987) Practical issues in structural modeling. Sociological Methods & Research, 16, 78-117.

Berkson, J. (1949). Limitations of the application of four-fold tables to hospital data. Biological Bulletin, 2, 47–53.

Bezdjian, S., Baker, L. A., & Tuvblad, C. (2011). Genetic and environmental influences on impulsivity: A meta-analysis of twin, family and adoption studies. Clinical Psychology Review, 31(7), 1209-1223. doi:10.1016/j.cpr.2011.07.005

Bischof, G. G., Rumpf, H. J., Meyer, C. C., Hapke, U. U., & John, U. U. (2005). Influence of psychiatric comorbidity in alcohol dependent subjects in a representative population survey on treatment utilization and natural recovery. Addiction, 100(3), 405-413. doi:10.1111/j.1360-0443.2005.01008.x

Bland, R. C., Stebelsky, G. G., Orn, H., & Newman, S. C. (1988). Psychiatric disorders and unemployment in Edmonton. Acta Psychiatrica Scandinavica,77(Suppl 338), 72-80. doi:10.1111/j.1600-0447.1988.tb08550.x

Bolton, J. M., Belik, S., Enns, M. W., Cox, B. J., & Sareen, J. (2008). Exploring the correlates of suicide attempts among individuals with major depressive disorder: Findings from the National Epidemiologic Survey on Alcohol and Related Conditions. Journal of Clinical Psychiatry, 69(7), 1139-1149. doi:10.4088/JCP.v69n0714

Boschloo, L., Vogelzangs, N., van den Brink, W., Smit, J. H., Beekman, A. F., & Penninx, B. H. (2012). Predictors of the 2‐year recurrence and persistence of alcohol dependence. Addiction, 107(9), 1639-1640. doi:10.1111/j.1360-0443.2012.03860.x

Brady, K. T., & Sonne, S. C. (1995). The relationship between substance abuse and bipolar disorder. Journal of Clinical Psychiatry, 56(Suppl 3), 19-24.

Breslau, N., & Klein, D. F. (1999). Smoking and panic attacks: An epidemiologic investigation. Archives of General Psychiatry, 56(12), 1141-1147. doi:10.1001/archpsyc.56.12.1141

Caetano, R., & Cunradi, C. (2002). Alcohol dependence: A public health perspective. Addiction, 97(6), 633-645. doi:10.1046/j.1360-0443.2002.00184.x

Canino, G., Bravo, M., Ramírez, R., Febo, V. E., Rubio-Stipec, M., Fernández, R., & Hasin, D. (1999). The Spanish Alcohol Use Disorder and Associated Disabilities Interview Schedule (AUDADIS): Reliability and concordance with clinical diagnoses in a Hispanic population. Journal of Studies on Alcohol, 60(6), 790-799.

Cartwright, W. S. (2008). Economic costs of drug abuse: Financial, cost of illness, and services. Journal of Substance Abuse Treatment, 34(2), 224-233. doi:10.1016/j.jsat.2007.04.003

Chassler, D., Lundgren, L., & Lonsdale, J. (2006). What factors are associated with high- frequency drug treatment use among a racially and ethnically diverse population of injection drug users?.American Journal on Addictions, 15(6), 440-449. doi:10.1080/10550490600998500

Clark, L. (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

Clarke, J. & Haughton, H. (1975). Study of intellectual impairment and recovery rates in heavy drinkers in Ireland. British Journal of Psychiatry, 126, 178-184.

Cohen, E., Feinn, R., Arias, A., & Kranzler, H. R. (2007). Alcohol treatment utilization: Findings from the National Epidemiologic Survey on Alcohol and Related Conditions. Drug and Alcohol Dependence, 86(2-3), 214-221. doi:10.1016/j.drugalcdep.2006.06.008

Coid, J. W. (2002). Personality disorders in prisoners and their motivation for dangerous and disruptive behaviour. Criminal Behaviour and Mental Health, 12(3), 209-226. doi:10.1002/cbm.497

Compton, W. M., Thomas, Y. F., Stinson, F. S., & Grant, B. F. (2007). Prevalence, correlates, disability, and comorbidity of DSM-IV drug abuse and dependence in the United States: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. Archives of General Psychiatry, 64(5), 566-576. doi:10.1001/archpsyc.64.5.566

Cottler, L. B., Grant, B. F., Blaine, J., Mavreas, V., Pull, C., Hasin, D., & ... Mager, D. (1997). Concordance of the DSM-IV alcohol and drug use disorder criteria and diagnoses as measured by AUDADIS-ADR, CIDI and SCAN. Drug and Alcohol Dependence, 47(3), 195-205. doi:10.1016/S0376-8716(97)00090-2

Cunningham, J. A., Sobell, L. C., Sobell, M. B., Agrawal, S., & Toneatto, T. (1993). Barriers to treatment: Why alcohol and drug abusers delay or never seek treatment. Addictive Behaviors, 18(3), 347-353. doi:10.1016/0306-4603(93)90036-9

Cunningham, J. A., Sobell, L. C., & Chow, V. M. (1993). What's in a label? The effects of substance types and labels on treatment considerations and stigma. Journal of Studies on Alcohol, 54(6), 693-699.

Dalton, E., Cate-Carter, T. D., Mundo, E., Parikh, S. V., & Kennedy, J. L. (2003). Suicide risk in bipolar patients: The role of co-morbid substance use disorders. Bipolar Disorders, 5(1), 58-61. doi:10.1034/j.1399-5618.2003.00017.x

Dickey, B., & Azeni, H. (1996). Persons with dual diagnoses of substance abuse and major mental illness: Their excess costs of psychiatric care. American Journal of Public Health, 86(7), 973-977. doi:10.2105/AJPH.86.7.973

DiClemente, C. C. (2003). Addiction and change: How addictions develop and addicted people recover. New York, NY US: Guilford Press.

Dinwiddie, S. H., Cottler, L., Compton, W., & Abdallah, A. (1996). Psychopathology and HIV risk behaviors among injection drug users in and out of treatment. Drug and Alcohol Dependence, 43(1-2), 1-11. doi:10.1016/S0376-8716(96)01290-2

Dolan, B. M., Warren, F. M., Menzies, D. D., & Norton, K. K. (1996). Cost-offset following specialist treatment of severe personality disorders. Psychiatric Bulletin, 20(7), 413-417. doi:10.1192/pb.20.7.413

Eames, S. L., Westermeyer, J., & Crosby, R. D. (1998). Substance use and abuse among patients with comorbid dysthymia and substance disorder. American Journal of Drug and Alcohol Abuse, 24(4), 541-550. doi:10.3109/00952999809019606

Edlund, M. J., Booth, B. M., & Han, X. (2012). Who seeks care where? Utilization of mental health and substance use disorder treatment in two national samples of individuals with alcohol use disorders. Journal of Studies on Alcohol and Drugs, 73(4), 635-646.

Ekleberry, S. C. (2009). Integrated treatment for co-occurring disorders: Personality disorders and addiction. New York, NY US: Routledge/Taylor & Francis Group.

Ettner, S. L., Huang, D., Evans, E., Ash, D., Hardy, M., Jourabchi, M., & Hser, Y. (2006). Benefit-Cost in the California Treatment Outcome project: Does substance abuse treatment 'pay for itself'?.Health Services Research, 41(1), 192-213.

Falk, D., Yi, H., & Hiller-Sturmhöfel, S. (2008). An epidemiologic analysis of co-occurring alcohol and drug use and disorders. Alcohol Research & Health, 31(2), 100-110.

Feingold, A., Kerr, D. R., & Capaldi, D. M. (2008). Associations of substance use problems with intimate partner violence for at-risk men in long-term relationships. Psychology,22(3), 429-438. doi:10.1037/0893-3200.22.3.429

Fenton, M. C., Keyes, K., Geier, T., Greenstein, E., Skodol, A., Krueger, B., & ... Hasin, D. S. (2012). Psychiatric comorbidity and the persistence of drug use disorders in the United States. Addiction, 107(3), 599-609. doi:10.1111/j.1360-0443.2011.03638.x

Freshman, A. (2004). Assessment and treatment of adolescent substance abusers. In S. Straussner (Ed.), Clinical work with substance-abusing clients (2nd ed.); (pp. 305-329). New York, NY US: Guilford Press.

Galanter, M., Castaneda, R., & Ferman, J. (1988). Substance abuse among general psychiatric patients: Place of presentation, diagnosis, and treatment. American Journal pf Drug And Alcohol Abuse, 14(2), 211-235. doi:10.3109/00952999809001548

George, A. A., & Tucker, J. A. (1996). Help-seeking for alcohol-related problems: Social contexts surrounding entry into alcoholism treatment or Alcoholics Anonymous. Journal of Studies on Alcohol, 57(4), 449-457.

Grande, T. P., Wolf, A. W., Schubert, D. S. P., Patterson, M. B., & Brocco, K. (1984). Associations within the diagnostic triad of alcoholism, drug abuse, and antisocial personality: A review of the literature. Psychological Reports, 55, 455-474.

Grant, B. F. (1992). DSM-III—R and proposed DSM-IV alcohol abuse and dependence, United States 1988: A nosological comparison. Alcoholism: Clinical And Experimental Research, 16(6), 1068-1077. doi:10.1111/j.1530-0277.1992.tb00701.x

Grant, B. F., Harford, T. C., Dawson, D. D., & Chou, P. S. (1995). The Alcohol Use Disorder and Associated Disabilities Interview Schedule (AUDADIS): Reliability of alcohol and drug modules in a general population sample. Drug and Alcohol Dependence, 39(1), 37- 44. doi:10.1016/0376-8716(95)01134-K

Grant, B. F.; Dawson, D. A.; and Hasin, D. S. The Alcohol Use Disorder and Associated Disabilities Interview Schedule–DSM-IV Version. Bethesda, MD: National Institute on Alcohol Abuse and Alcoholism, 2001.

Grant, B. F., Dawson, D. A., Stinson, F. S., Chou, P. S., Kay, W., & Pickering, R. (2003). The Alcohol Use Disorder and Associated Disabilities Interview Schedule-IV (AUDADIS- IV): Reliability of alcohol consumption, tobacco use, family history of depression and psychiatric diagnostic modules in a general population sample. Drug and Alcohol Dependence, 71(1), 7-16. doi:10.1016/S0376-8716(03)00070-X

Grant, B. F., Dawson, D. A., Stinson, F. S., Chou, S., Dufour, M. C., & Pickering, R. P. (2004a). The 12-month prevalence and trends in DSM-IV alcohol abuse and dependence: United States, 1991-1992 and 2001-2002. Drug and Alcohol Dependence, 74(3), 223-234. doi:10.1016/j.drugalcdep.2004.02.004

Grant B. F., Stinson F. S., Dawson, D. A., Chou, S. P., Dufour, M. C., Compton, W., Pickering R. P., & Kaplan, K. (2004b). Prevalence and co-occurrence of substance use disorders and Independent mood and anxiety disorders: Results from the national epidemiologic survey on alcohol and related conditions. Archives of General Psychiatry, 61(8), 807- 816. doi:10.1001/archpsyc.61.8.807

Grant, B. F., Stinson, F. S., Dawson, D. A., Chou, S., Ruan, W., & Pickering, R. P. (2004c). Co- occurrence of 12-month alcohol and drug use disorders and personality disorders in the

United States: Results From the National Epidemiologic Survey on Alcohol and Related Conditions. Archives of General Psychiatry, 61(4), 361-368. doi:10.1001/archpsyc.61.4.361

Grant, B. F., Hasin, D. S., Stinson, F. S., Dawson, D. A., Chou, S., Ruan, W., & Pickering, R. P. (2004d). Prevalence, correlates, and disability of personality disorders in the United States: Results From the National Epidemiologic Survey on Alcohol and Related Conditions. Journal of Clinical Psychiatry, 65(7), 948-958. doi:10.4088/JCP.v65n0711

Grant, B. F., Hasin, D. S., Blanco, C., Stinson, F. S., Chou, S., Goldstein, R. B., & ... Huang, B. (2005). The epidemiology of social anxiety disorder in the United States: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. Journal of Clinical Psychiatry,66(11), 1351-1361. doi:10.4088/JCP.v66n1102

Grant, J. E., Mooney, M. E., & Kushner, M. G. (2012). Prevalence, correlates, and comorbidity of DSM-IV obsessive-compulsive personality disorder: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. Journal of Psychiatric Research, 46(4), 469-475. doi:10.1016/j.jpsychires.2012.01.009

Green-Hennessy, S. (2002). Factors associated with receipt of behavioral health services among persons with substance dependence. Psychiatric Services, 53(12), 1592-1598. doi:10.1176/appi.ps.53.12.1592

Greenfeld, L. A., & Henneberg, M. A. (2001). Victim and offender self-reports of alcohol involvement in crime. Alcohol Research & Health, 25(1), 20-31.

Gynther, L. M., Carey, G., Gottesman, I. I., & Vogler, G. P. (1995). A twin study of non-alcohol substance abuse. Psychiatry Research, 56(3), 213-220. doi:10.1016/0165- 1781(94)02609-M

Hajema, K., Knibbe, R. A., & Drop, M. J. (1999). Social resources and alcohol-related losses as predictors of help seeking among male problem drinkers. Journal of Studies on Alcohol, 60(1), 120-129.

Harrow, M., Hansford, B. G., & Astrachan-Fletcher, E. B. (2009). Locus of control: Relation to schizophrenia, to recovery, and to depression and psychosis—A 15-year longitudinal study. Psychiatry Research, 168(3), 186-192. doi:10.1016/j.psychres.2008.06.002

Harwood, H., (2000). Report prepared by the Lewin Group for the National Institute on Alcohol Abuse and Alcoholism; National Institutes of Health, NIH Publication No. 98-4327. Rockville, MD, http://www.niaaa.nih.gov

Hasin, D. S., & Grant, B. F. (1995). AA and other helpseeking for alcohol problems: Former drinkers in the U.S. general population. Journal of Substance Abuse, 7(3), 281-292. doi:10.1016/0899-3289(95)90022-5

Hasin, D. S., Muthen, B., Grant, B. F. The dimensionality of DSM-IV alcohol abuse and

dependence: factor analysis in a clinical sample. In: Vrosti R, (ed). Alcoholism: New

Research Perspectives. Munich, Germany: Hogrefe and Hubner; 1997:27-39.

Hasin, D., & Paykin, A. (1999). Alcohol dependence and abuse diagnoses: Concurrent validity in a nationally representative sample. Alcoholism: Clinical and Experimental Research, 23(1), 144-150. doi:10.1097/00000374-199901000-00021

Hasin, D. S., Stinson, F. S., Ogburn, E., & Grant, B. F. (2007). Prevalence, correlates, disability, and comorbidity of DSM-IV alcohol abuse and dependence in the United States: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. Archives of General Psychiatry, 64(7), 830-842. doi:10.1001/archpsyc.64.7.830

Hasin, D., Fenton, M. C., Skodol, A., Krueger, R., Keyes, K., Geier, T., & ... Grant, B. (2011). Personality disorders and the 3-year course of alcohol, drug, and nicotine use disorders. Archives of General Psychiatry, 68(11), 1158-1167. doi:10.1001/archgenpsychiatry.2011.136

Hinckers, A. S., Laucht, M., Schmidt, M. H., Mann, K. F., Schumann, G., Schuckit, M. A., & Heinz, A. (2006). Low level of response to alcohol as associated with serotonin transporter genotype and high alcohol intake in adolescents. Biological Psychiatry, 60(3), 282-287. doi:10.1016/j.biopsych.2005.12.009

Ilgen, M. A., Price, A. M., Burnett-Zeigler, I., Perron, B., Islam, K., Bohnert, A. B., & Zivin, K. (2011). 'Longitudinal predictors of addictions treatment utilization in treatment-naïve adults with alcohol use disorders': Corrigendum. Drug and Alcohol Dependence, 119(3), doi:10.1016/j.drugalcdep.2011.08.002

Jackson, D. L. (2003). Revisiting sample size and number of parameter estimates: Some support for the N:q hypothesis. Structural Equation Modeling, 10(1), 128-141. doi:10.1207/S15328007SEM1001_6

Jané-Llopis, E., & Matytsina, I. (2006). Mental health and alcohol, drugs and tobacco: A review of the comorbidity between mental disorders and the use of alcohol, tobacco and illicit drugs. Drug and Alcohol Review, 25(6), 515-536. doi:10.1080/09595230600944461

Jovanovski, D., Erb, S., & Zakzanis, K. K. (2005). Neurocognitive Deficits in Cocaine Users: A Quantitative Review of the Evidence. Journal of Clinical And Experimental Neuropsychology, 27(2), 189-204. doi:10.1080/13803390490515694

Kandel, D. B. (1992). Epidemiological trends and implications for understanding the nature of addiction. In C. P. O'Brien, J. H. Jaffe (Eds.) , Addictive states (pp. 23-40). New York, NY US: Raven Press.

Kasen, S. S., Cohen, P. P., Skodol, A. E., First, M. B., Johnson, J. G., Brook, J. S., & Oldham, J. M. (2007). Comorbid personality disorder and treatment use in a community sample of youths: A 20-year follow-up. Acta Psychiatrica Scandinavica,115(1), 56-65. doi:10.1111/j.1600-0447.2006.00842.x

Kaskutas, L., Weisner, C., & Caetano, R. (1997). Predictors of help seeking among a longitudinal sample of the general public, 1984-1992. Journal of Studies on Alcohol, 58(2), 155-161.

Katz, S. J., Kessler, R. C., Frank, R. G., Leaf, P., Lin, E., & Edlund, M. (1997). The use of outpatient mental health services in the United States and Ontario: The impact of mental morbidity and perceived need for care. American Journal of Public Health, 87(7), 1136- 1143. doi:10.2105/AJPH.87.7.1136

Kendler, K. S., Aggen, S. H., Tambs, K., & Reichborn-Kjennerud, T. (2006). Illicit psychoactive substance use, abuse and dependence in a population-based sample of Norwegian twins. Psychological Medicine, 36(7), 955-962. doi:10.1017/S0033291706007720

Kessler, R. C., McGonagle, K. A., Zhao, S., & Nelson, C. B. (1994). Lifetime and 12-month prevalence of DSM-III—R psychiatric disorders in the United States: Results from the National Comorbidity Study. Archives of General Psychiatry, 51(1), 8-19. doi:10.1001/archpsyc.1994.03950010008002

Kessler, R. C., Nelson, C. B., McGonagle, K. A., Edlund, M. J., Frank, R. G., & Leaf, P. J. (1996). The epidemiology of co-occurring addictive and mental disorders: Implications for prevention and service utilization. American Journal of Orthopsychiatry, 66(1), 17- 31. doi:10.1037/h0080151

Kessler, R. C., Crum, R. M., Warner, L. A., & Nelson, C. B. (1997). Lifetime co-occurrence of DSM-III-R alcohol abuse and dependence with other psychiatric disorders in the National Comorbidity Survey. Archives of General Psychiatry, 54(4), 313-321. doi:10.1001/archpsyc.1997.01830160031005

Kessler, R. C., Chiu, W., Demler, O., & 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(6), 617-627. doi:10.1001/archpsyc.62.6.617

Khantzian, E. J. (1985). The self-medication hypothesis of addictive disorders: Focus on heroin and cocaine dependence. American Journal of Psychiatry, 142, 1259-1264.

Kline, R. B. (2011). Principles and practice of structural equation modeling (3rd ed.). New York, NY US: Guilford Press.

100

Klingemann, H. K. (1991). The motivation for change from problem alcohol and heroin use. British Journal of Addiction, 86(6), 727-744. doi:10.1111/j.1360- 0443.1991.tb03099.x

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., Hicks, B. M., Patrick, C. J., Carlson, S. R., Iacono, W. G., & McGue, M. (2002). Etiologic connections among substance dependence, antisocial behavior and personality: Modeling the externalizing spectrum. Journal of Abnormal Psychology, 111(3), 411-424. doi:10.1037/0021-843X.111.3.411

Krueger, R. F., & 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

Leonard, K. E., & Quigley, B. M. (1999). Drinking and marital aggression in newlyweds: An event-based analysis of drinking and the occurrence of husband marital aggression. Journal of Studies on Alcohol, 60(4), 537-545.

Lewis, C. E., Bucholz, K. K., Spitznagel, E., & Shayka, J. J. (1996). Effects of gender and comorbidity on problem drinking in a community sample. Alcoholism: Clinical And Experimental Research, 20(3), 466-476. doi:10.1111/j.1530-0277.1996.tb01077.x

Leukefeld, C. G., & Haverkos, H. W. (1993). Sexually transmitted diseases. In T. P. Gullotta, G. R. Adams, R. Montemayor (Eds.) , Adolescent sexuality (pp. 161-180). Thousand Oaks, CA US: Sage Publications, Inc.

Leukefeld, C. G., Logan, T. K., Martin, S. S., Purvis, R. T., & Farabee, D. (1998). A health services use framework for drug-abusing offenders. American Behavioral Scientist, 41(8), 1123-1135. doi:10.1177/0002764298041008008

Lundgren, L. M., Sullivan, L., & Amodeo, M. (2006). How do treatment repeaters use the drug treatment system? An analysis of injection drug users in Massachusetts. Journal of Substance Abuse Treatment, 30(2), 121-128. doi:10.1016/j.jsat.2005.10.007

Magidson, J. F., Liu, S., Lejuez, C. W., & Blanco, C. (2012). Comparison of the course of substance use disorders among individuals with and without generalized anxiety disorder in a nationally representative sample. Journal of Psychiatric Research, 46(5), 659-666. doi:10.1016/j.jpsychires.2012.02.011

Mares, S. W., van der Vorst, H., Engels, R. E., & Lichtwarck-Aschoff, A. (2011). Parental alcohol use, alcohol-related problems, and alcohol-specific attitudes, alcohol-specific communication, and adolescent excessive alcohol use and alcohol-related problems: An indirect path model. Addictive Behaviors, 36(3), 209-216. doi:10.1016/j.addbeh.2010.10.013

101

Markon, K. E. (2010). Modeling psychopathology structure: A symptom-level analysis of Axis I and II disorders. Psychological Medicine, 40(2), 273-288. doi:10.1017/S0033291709990183

McCabe, R. E., Chudzik, S. M., Antony, M. M., Young, L., Swinson, R. P., & Zolvensky, M. J. (2004). Smoking behaviors across anxiety disorders. Journal of Anxiety Disorders, 18(1), 7-18. doi:10.1016/j.janxdis.2003.07.003

McCranie, E. W., & Kahan, J. (1986). Personality and multiple divorce: A prospective study. Journal of Nervous And Mental Disease, 174(3), 161-164. doi:10.1097/00005053- 198603000-00006

McLellan, A. T., Lewis, D. C., O’Brien, C. P., & Kleber, H. D. (2000). Drug dependence, a chronic medical illness: Implications for treatment, insurance, and outcome evaluation. Journal of American Medical Association, 284, 1689-1695. doi:10.1001/jama.284.13.1689

Merikangas, K. R., Stolar, M., Stevens, D. E., Goulet, J., Preisig, M. A., Fenton, B., Zhang, H., O’Malley, S. S., & Rounsaville, B. J. (1998). Familial transmission of substance use disorders. Archives of General Psychiatry, 55(11), 973-979. doi:10.1001/archpsyc.55.11.973

Moggi, F., Ouimette, P., Moos, R. H., & Finney, J. W. (1999). Dual diagnosis patients in substance abuse treatment: Relationship of general coping and substance-specific coping to 1-year outcomes. Addiction, 94(12), 1805-1816. doi:10.1046/j.1360- 0443.1999.941218056.x

Mojtabai, R., Olfson, M., & Mechanic, D. (2002). Perceived need and help-seeking in adults with mood, anxiety, or substance use disorder. Archives of General Psychiatry, 59(1), 77- 84. doi:10.1001/archpsyc.59.1.77

Morgenstern, J., Langenbucher, J., Labouvie, E., & Miller, K. J. (1997). The comorbidity of alcoholism and personality disorders in a clinical population: Prevalence and relation to alcohol typology variables. Journal of Abnormal Psychology, 106(1), 74-84. doi:10.1037/0021-843X.106.1.74

Morris, E. P., Stewart, S. H., & Ham, L. S. (2005). The relationship between social anxiety disorder and alcohol use disorders: A critical review. Clinical Psychology Review, 25(6), 734-760. doi:10.1016/j.cpr.2005.05.004

Moselhy, H. F., Georgiou, G., & Kahn, A. (2001). Frontal lobe changes in alcoholism: A review of the literature. Alcohol And Alcoholism, 36(5), 357-368. doi:10.1093/alcalc/36.5.357

Nace, E. P. (2005). Alcohol. In R. J. Frances, S. I. Miller, A. H. Mack (Eds.) , Clinical textbook of addictive disorders, 3rd ed (pp. 75-104). New York, NY US: Guilford Publications.

National Drug Intelligence Center. The Economic Impact of Illicit Drug Use on American Society. Washington, DC: United States Department of Justice, 2011.

102

Nelson C. B., Rehm J., Ustun T. B., Grant B., Chatterji S. (1999). Factor structures for DSM-IV substance disorder criteria endorsed by alcohol, cannabis, cocaine and opiate users: Results from the WHO reliability and validity study. Addiction,94, 843–855. DOI: 10.1046/j.1360-0443.1999.9468438.x

Niciu, M. J., Chan, G., Gelernter, J., Arias, A. J., Douglas, K., Weiss, R., & ... Kranzler, H. R. (2009). Subtypes of major depression in substance dependence. Addiction, 104(10), 1700-1709. doi:10.1111/j.1360-0443.2009.02672.x

Nunes, E. V., Hennessy, G., & Selzer, J. (2010). Depression in patients with substance use disorders. In E. V. Nunes, J. Selzer, P. Levounis, C. A. Davies (Eds.) , Substance dependence and co-occurring psychiatric disorders: Best practices for diagnosis and treatment (pp. 1-36). Kingston, NJ US: Civic Research Institute.

O'Farrell, T. J., Murphy, C. M., Neavins, T. M., & Van Hutton, V. (2000). Verbal aggression among male alcoholic patients and their wives in the year before and two years after alcoholism treatment. Journal of Family Violence, 15(4), 295-310. doi:10.1023/A:1007503411845

Oleski, J., Mota, N., Cox, B. J., & Sareen, J. (2010). Perceived need for care, help seeking, and perceived barriers to care for alcohol use disorders in a national sample. Psychiatric Services, 61(12), 1223-1231. doi:10.1176/appi.ps.61.12.1223

O'Malley, S., Adamse, M., Heaton, R. K., & Gawin, F. H. (1992). Neuropsychological impairment in chronic cocaine abusers. American Journal of Drug and Alcohol Abuse, 18(2), 131-144. doi:10.3109/00952999208992826

Oquendo, M. A., Galfalvy, H., Russo, S., Ellis, S. P., Grunebaum, M. F., Burke, A., & Mann, J. (2004). Prospective study of clinical predictors of suicidal acts after a major depressive episode in patients with major depressive disorder or bipolar disorder. American Journal of Psychiatry, 161(8), 1433-1441. doi:10.1176/appi.ajp.161.8.1433

Oser, C. B., Leukefeld, C. G., Staton Tindall, M., Garrity, T. F., Carlson, R. G., Falck, R., & ... Booth, B. M. (2011). Rural drug users: Factors associated with substance abuse treatment utilization. International Journal of Offender Therapy And Comparative Criminology, 55(4), 567-586. doi:10.1177/0306624X10366012

Pettinati, H. M., Pierce, J. r., Belden, P. P., & Meyers, K. (1999). The relationship of Axis II personality disorders to other known predictors of addiction treatment outcome. American Journal on Addictions, 8(2), 136-147. doi:10.1080/105504999305947

Powell, G., & Peveler, R. (1996). Nature and prevalence of personality disorders amongst patients receiving treatment for alcohol dependence. Journal of Mental Health, 5(3), 305- 314. doi:10.1080/09638239650036965

103

Prescott, C. A., & Kendler, K. S. (1999). Genetic and environmental contributions to alcohol abuse and dependence in a population-based sample of male twins. American Journal of Psychiatry, 156(1), 34-40.

Price, R., Risk, N. K., & Spitznagel, E. L. (2001). Remission from drug abuse over a 25-year period: Patterns of remission and treatment use. American Journal of Public Health, 91(7), 1107-1113. doi:10.2105/AJPH.91.7.1107

Project MATCH Research Group. (1997). Matching alcoholism treatments to client heterogeneity: Project MATCH Posttreatment drinking outcomes. Journal of Studies on Alcohol, 58(1), 7-29.

Pulay, A. J., Dawson, D. A., Ruan, W., Pickering, R. P., Huang, B., Chou, S., & Grant, B. F. (2008). The relationship of impairment to personality disorder severity among individuals with specific axis I disorders: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. Journal of Personality Disorders, 22(4), 405-417. doi:10.1521/pedi.2008.22.4.405

Ray, L. A., Hutchison, K. E., & Hauser, M. (2008). Alcohol use disorders: History, theory, and diagnosis. In W. Craighead, D. J. Miklowitz, L. W. Craighead (Eds.) , Psychopathology: History, diagnosis, and empirical foundations (pp. 495-523). Hoboken, NJ US: John Wiley & Sons Inc.

Regier, D. A., Farmer, M. E., Rae, D. S., Locke, B. Z., Keith, B. J., Judd, L. L., et al. (1990). Comorbidity of mental disorders with alcohol and other drug abuse. Results from the Epidemiologic Catchment Area (ECA) study. Journal of American Medical Association, 264, 2511-2518. doi:10.1001/jama.1990.03450190043026.

Reich, J. H., & Green, A. I. (1991). Effect of personality disorders on outcome of treatment. Journal of Nervous and Mental Disease, 179(2), 74-82. doi:10.1097/00005053-199102000-00003

Reich, J. H., & Vasile, R. G. (1993). Effect of personality disorders on the treatment outcome of Axis I conditions: An update. Journal of Nervous and Mental Disease,181(8), 475-484. doi:10.1097/00005053-199308000-00002

Reingle, J. M., Staras, S. S., Jennings, W. G., Branchini, J., & Maldonado-Molina, M. M. (2012). The relationship between marijuana use and intimate partner violence in a nationally representative, longitudinal sample. Journal of Interpersonal Violence, 27(8), 1562-1578. doi:10.1177/0886260511425787

Rice, D. P., Kelman, S., & Miller, L. S. (1991). The economic cost of alcohol abuse. Alcohol Health & Research World, 15(4), 307-316.

Robinson, J., Sareen, J., Cox, B. J., & Bolton, J. M. (2011). Role of self-medication in the development of comorbid anxiety and substance use disorders: A longitudinal investigation. Archives of General Psychiatry, 68(8), 800-807. doi:10.1001/archgenpsychiatry.2011.75

104

Roizen, J. (1993). Issues in the epidemiology of alcohol and violence. In: Martin S (ed.). Alcohol and interpersonal violence: Fostering multidisciplinary perspectives. (NIAAA Research Monograph No. 24, pp. 3-36). Bethesda, MD: US Department of Health and Human Services.

Room, R. (1989). The U.S. general population's experiences of responding to alcohol problems. British Journal of Addiction, 84(11), 1291-1304. doi:10.1111/j.1360- 0443.1989.tb00731.x

Ross, H. E., Lin, E., & Cunningham, J. (1999). Mental health service use: A comparison of treated and untreated individuals with substance use disorders in Ontario. The Canadian Journal of Psychiatry / La Revue Canadienne De Psychiatrie, 44(6), 570-577.

Rourke, S. B., & Løberg, T. (1996). Neurobehavioral correlates of alcoholism. In I. Grant, K. M. Adams (Eds.) , Neuropsychological assessment of neuropsychiatric disorders (2nd ed.) (pp. 423-485). New York, NY US: Oxford University Press.

Ruegg, R., & Frances, A. (1995). New research in personality disorders. Journal of Personality Disorders, 9(1), 1-48. doi:10.1521/pedi.1995.9.1.1

Salloum, I. M., Moss, H. B., Daley, D. C., Cornelius, J. R., Kirisci, L., & Al-Maalouf, M. (1998). Drug use problem awareness and treatment readiness in dual-diagnosis patients. American Journal on Addictions, 7(1), 35-42. doi:10.3109/10550499809034710

Saunders, S. M., Zygowicz, K. M., & D'Angelo, B. R. (2006). Person-related and treatment- related barriers to alcohol treatment. Journal of Substance Abuse Treatment, 30(3), 261- 270. doi:10.1016/j.jsat.2006.01.003

Sbrana, A., Bizzarri, J. V., Rucci, P., Gonnelli, C., Massei, J. G., Ravani, L., & ... Cassano, G. B. (2007). Family history of psychiatric disorders and alcohol and substance misuse in patients with bipolar I disorder, substance use disorder, or both. American Journal on Addictions, 16(3), 227-231. doi:10.1080/10550490701375699

Sher, K. J., & Trull, T. J. (1994). Personality and disinhibitory psychopathology: Alcoholism and antisocial personality disorder. Journal of Abnormal Psychology, 103(1), 92-102. doi:10.1037/0021-843X.103.1.92

Skodol, A. E., Oldham, J. M., Bender, D. S., Dyck, I. R., Stout, R. L., Morey, L. C., & ... Gunderson, J. G. (2005a). Dimensional representations of DSM-IV personality disorders: Relationships to functional impairment. American Journal of Psychiatry, 162(10), 1919- 1925. doi:10.1176/appi.ajp.162.10.1919

Skodol, A. E., Gunderson, J. G., Shea, M. T., McGlashan, T. H., Morey, L. C., Sanislow, C. A., Bender, D. S., Grilo, C. M., Zanarini, M. C., Yen, S., Pagano, M. E., & Stout, R. L. (2005b). The Collaborative Longitudinal Personality Disorders Study (CLPS): Overview and implications. Journal of Personality Disorders, 19(5), 487-504. doi:10.1521/pedi.2005.19.5.487

105

Slutske, W. S. (2010). Why is natural recovery so common for addictive disorders? Addiction, 105(9), 1520-1521. doi:10.1111/j.1360-0443.2010.03035.x

Smith, J. P., & Book, S. W. (2010). Comorbidity of generalized anxiety disorder and alcohol use disorders among individuals seeking outpatient substance abuse treatment. Addictive Behaviors, 35(1), 42-45. doi:10.1016/j.addbeh.2009.07.002

Smith, P. H., Homish, G. G., Leonard, K. E., & Cornelius, J. R. (2012). Intimate partner violence and specific substance use disorders: Findings from the National Epidemiologic Survey on Alcohol and Related Conditions. Psychology of Addictive Behaviors, 26(2), 236-245. doi:10.1037/a0024855

Sobell, L. C., Cunningham, J. A., Sobell, M. B., & Agrawal, S. (1996). Fostering self-change among problem drinkers: A proactive community intervention. Addictive Behaviors, 21(6), 817-833. doi:10.1016/0306-4603(96)00039-1

Straussner, S., & Nemenzik, J. M. (2007). Co-occurring substance use and personality disorders: Current thinking on etiology, diagnosis, and treatment. Journal of Social Work Practice In The Addictions, 7(1-2), 5-23. doi:10.1300/J160v07n01_02

Strain, E. C. (1995). Antisocial personality disorder, misbehavior, and drug abuse. Journal of Nervous and Mental Disease, 183(3), 162-165. doi:10.1097/00005053-199503000-00006

Tarter, R. E., Alterman, A. I., & Edwards, K. L. (1985). Vulnerability to alcoholism in men: A behavior-genetic perspective. Journal of Studies on Alcohol, 46(4), 329-356.

Tate, S. R., Drapkin, M., & Brown, S. A. (2009). Substance abuse: Diagnosis, comorbidity, and psychopathology. In P. H. Blaney, T. Millon (Eds.) , Oxford textbook of psychopathology (2nd ed.) (pp. 280-297). New York, NY US: Oxford University Press.

Thomasius, R., Sack, P., & Petersen, K. (2010). DSM-IV Axis-I comorbidity among illicit drug users seeking treatment for substance use disorders: Results from the multi-centre study of psychiatric comorbidity in drug addicts (MUPCDA). Mental Health and Substance Use: Dual Diagnosis, 3(3), 182-192. doi:10.1080/17523281.2010.504643

Thornton, L. K., Baker, A. L., Lewin, T. J., Kay-Lambkin, F. J., Kavanagh, D., Richmond, R., & ... Johnson, M. P. (2012). Reasons for substance use among people with mental disorders. Addictive Behaviors, 37(4), 427-434. doi:10.1016/j.addbeh.2011.11.039

Tómasson, K., & Vaglum, P. (1996). Psychopathology and alcohol consumption among treatment-seeking alcoholics: A prospective study. Addiction, 91(7), 1019-1030. doi:10.1111/j.1360-0443.1996.tb03598.x

106

Tómasson, K., & Vaglum, P. (1997). The 2-year course following detoxification treatment of substance abuse: The possible influence of psychiatric comorbidity. European Archives of Psychiatry and Clinical Neuroscience, 247(6), 320-327. doi:10.1007/BF02922262

Trull, T. J., Waudby, C. J., & Sher, K. J. (2004). Alcohol, tobacco, and drug use disorders and personality disorder symptoms. Experimental and Clinical Psychopharmacology, 12(1), 65-75. doi:10.1037/1064-1297.12.1.65

Trull, T. J., Jahng, S., Tomko, R. L., Wood, P. K., & Sher, K. J. (2010). Revised NESARC personality disorder diagnoses: Gender, prevalence, and comorbidity with substance dependence disorders. Journal of Personality Disorders, 24(4), 412-426. doi:10.1521/pedi.2010.24.4.412

Tucker, J. A. (1995). Predictors of help-seeking and the temporal relationship of help to recovery among treated and untreated recovered problem drinkers. Addiction, 90(6), 805-809. doi:10.1111/j.1360-0443.1995.tb02228.x

Tucker, J. A. (2001). Resolving problems associated with alcohol and drug misuse: Understanding relations between addictive behavior change and the use of services. Substance Use & Misuse, 36(11), 1501-1518. doi:10.1081/JA-100106961

Tucker, J. A., & Simpson, C. A. (2011). The recovery spectrum: From self-change to seeking treatment. Alcohol Research & Health, 33(4), 371-379.

Tucker, J. A., Vuchinich, R. E., & Rippens, P. D. (2004). A factor analytic study of influences on patterns of help-seeking among treated and untreated alcohol dependent persons. Journal of Substance Abuse Treatment, 26(3), 237-242. doi:10.1016/S0740-5472(03)00209-5

Tyrer, P., Mitchard, S., Methuen, C., & Ranger, M. (2003). Treatment rejecting and treatment seeking personality disorders: Type R and type S. Journal of Personality Disorders, 17(3), 263-267. doi:10.1521/pedi.17.3.263.22152

Uhl, G. R., Elmer, G. I., Labuda, M. C., Pickens, R. W. (1995). Genetic influences in drug abuse. In Gloom, F., Kupfer, D. (Eds.) Psychopharmacology: The Fourth Generation of Progress. New York , Raven Press; 1995:1793-2783.

Vaughn, M. G., Fu, Q., Beaver, D., DeLisi, M., Perron, B., & Howard, M. (2010). Are personality disorders associated with social welfare burden in the United States?. Journal of Personality Disorders, 24(6), 709-720. doi:10.1521/pedi.2010.24.6.709

Vrieze, S. I., Perlman, G., Krueger, R. F., & Iacono, W. G. (2012). Is the continuity of externalizing psychopathology the same in adolescents and middle–aged adults? A test of the externalizing spectrum’s developmental coherence. Journal of Abnormal Child Psychology, 40(3), 459-470. doi:10.1007/s10802-011-9571-x

Wang, P. S., Lane, M., Olfson, M., Pincus, H. A., Wells, K. B., & Kessler, R. C. (2005). Twelve- month use of mental health services in the United States: Results from the National

107

Comorbidity Survey replication. Archives of General Psychiatry, 62(6), 629-640. doi:10.1001/archpsyc.62.6.629

Weisner, C., Matzger, H., Tam, T., & Schmidt, L. (2002). Who goes to alcohol and drug treatment? Understanding utilization within the context of insurance. Journal of Studies on Alcohol, 63(6), 673-682.

World Health Organization (2002). The world health report 2002. Geneva, World Health Organization.

Wu, L., Kouzis, A. C., & Leaf, P. J. (1999). Influence of comorbid alcohol and psychiatric disorders on utilization of mental health services in the National Comorbidity Survey. American Journal of Psychiatry, 156(8), 1230-1236.

Wu, L., Ringwalt, C. L., & Williams, C. E. (2003). Use of substance abuse treatment services by persons with mental health and substance use problems. Psychiatric Services, 54(3), 363- 369. doi:10.1176/appi.ps.54.3.363

Wu, L., & Ringwalt, C. L. (2004). Alcohol dependence and use of treatment services among women in the community. American Journal of Psychiatry, 161(10), 1790-1797. doi:10.1176/appi.ajp.161.10.1790

Ziedonis, D., & Nickou, C. (2001). Substance abuse in patients with schizophrenia. In M. Y. Hwang, P. C. Bermanzohn (Eds.) , Schizophrenia and comorbid conditions: Diagnosis and treatment Arlington, VA US: American Psychiatric Publishing, Inc.

108