 

as! r
“1%.?

 

Mum. H

5."!

:5:

 

.1. 3.12.! V .
s!...l::...
l:

 

 

x -
£325? .1.1 51.)... -...14 .
.. 1 xi.5.3.9.3.31_.=§._nn.;+. 3%. .
. L2 . .9. . . V . . V
.: . ,. ..., Emmi»?
. . . l, . ... .Mav...L

 

.1

10'“-

 

LIBRARY

Michigan

 

State
University

This is to certify that the
dissertation entitled

EXECUTIVE FUNCTION DEFICITS IN
PSYCHOPATHOLOGY

presented by

GILLIAN MARY STAVRO

has been accepted towards fulﬁllment
of the requirements for the

Doctoral degree in Psychology

 

 

 

 

éﬂi

Ma}? paraggargsignature
0319/5 7

Date

MSU is an aﬂ‘innative-action, equal-opportunity employer

0.-.-.--.-.------~--o---u-.-.----o--.-----c---o---o-n-c--o-.--o-o--o-u--o---o----u-o-:-v---.--.--->------.

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/CIRCIDateDue.indd~p.1

 

 

EXECUTIVE FUNCTION DEFICITS IN PSYCHOPATHOLOGY
By

Gillian Mary Stavro

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Psychology

2007

ABSTRACT
EXECUTIVE FUNCTION DEFICITS IN PSYCHOPATHOLOGY
By
Gillian Mary Stavro

Purpose: Striking parallelism in the literature links deﬁcits in executive functions
(EF) with a wide range of seemingly different psychiatric disorders. Questions have been
raised about how these diverse disorders can all be related to a similar cognitive and/or
neural dysﬁmction. Previous literature has focused primarily on effects associated with
individual, or few, disorders. As a result, the potentially signiﬁcant role of comorbidity,
as well as speciﬁcity of effects, is poorly understood. The present study was an attempt
to understand the speciﬁcity of EF deﬁcits to different types of psychopathology, taking
into account the effect of comorbidity. Four models examining the relationship at
different levels in the diagnostic hierarchy were tested as possible explanations for the
ﬁequent association of EF deﬁcits to psychiatric disorders. Methods: Adults participants
from two large, preexisting studies were combined (total n = 641). Diagnostic interviews
and neuropsychological testing were completed for all participants. Disorders
investigated were attention-deﬁcit/hyperactivity disorder (ADHD), childhood
externalizing disorders, antisocial personality disorder (ASPD), alcoholism, drug
dependence, depression, and anxiety disorders. EF tests were the Stroop Color-Word
Test (Stroop), Wisconsin Card Sorting Test (WCST), Trail Making Test (TMT), and Stop
Signal Test (Stop). Proﬁle analysis, linear regression, analysis of variance, and structural
equation modeling (SEM) were used to test the explanatory power of the following

models and their associated hypotheses: (1) componential — individual disorders were

associated with deﬁcits in different EF processes; (2) comorbidity-speciﬁc — only a few
speciﬁc disorders accounted for the association between EF and psychopathology, with
high rates of comorbidity making the relationship appear more widespread; (3)
comorbidity-nonspeciﬁcity — number of comorbid disorders, rather than type, was related
to EF impairment; and (4) dimension-speciﬁc — shared underlying dimensions of
psychopathology (i.e., internalizing versus externalizing) were differentially related to EF
deﬁcits. Results: The comorbidity-nonspeciﬁcity hypothesis was not supported. There
was support for comorbidity-speciﬁc and componential effects, as certain individual
disorders (ADHD predominantly, as well as alcoholism and ASPD) were associated with
poorer performance on measures from certain EF tests (TMT Residual Score, Stop
Response Inhibition and Response Variability). Individual-disorder analyses were ﬁnther
elucidated by structural models testing dimension-speciﬁc effects. Externalizing
disorders were associated with poorer performance on cognitive tests. Speciﬁcity was
found for types of disorders associated with cognitive ﬁmctioning, but not for cognitive
effects, as externalizing disorders were related to poorer performance on both EF and
processing speed tasks. Further, many EF effects disappeared after controlling for F SIQ
(although not those associated with ADHD), while speed-related effects were more
robust. Conclusions: Findings suggest that shared underlying effects associated with
externalizing disorders contribute to cognitive deﬁcits (or vice versa). Certain EF
processes and tasks are particularly sensitive to psychopathology, but neurocognitive
effects crossed cognitive domains in the present ﬁndings. Longitudinal and symptom-
based studies are needed to better understand the role of neurocognitive deﬁcits in the

etiology and maintenance of psychological disorders.

ACKNOWLEDGMENTS

This research was supported by National Institute of Mental Health Grant R01-
MH63146 to Joel Nigg, John Henderson, and Fernanda F erreira; R01 AA12217 to Robert
Zucker and Joel Nigg; and R37 AAO7065 to Robert Zucker and Hiram Fitzgerald.

I am extremely grateful to Joel Nigg for being involved in all aspects of this
manuscript. His time, effort, and commitment were greatly appreciated. Thank you as
well to Mark Ettenhofer, Bob Zucker, Norman Abeles, Christine Larson, and David Z.
Hambrick, whose thoughtful comments were invaluable. I am also grateful for the

assistance and support provided by Colleen Schmitt, Linna Leslie, and Leon Puttler.

iv

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... VII
LIST OF FIGURES .......................................................................................................... IX
INTRODUCTION .............................................................................................................. 1
Why Study Psychopathology and Executive Functions? ............................................ 7
Executive Functions ................................................................................................... 11
Measurement of Executive Functions ....................................................................... 17
Functional Signiﬁcance of Frontal-Subcortical Circuits ........................................... 20
EF in Key Psychopathologies .................................................................................... 24
ADHD and Extemalizing Disorders Typically Diagnosed in Childhood
(Manifestations in Adulthood) ........................................................................... 25
Antisocial personality disorder ........................................................................... 32
Alcoholism ......................................................................................................... 36
Drug Dependence ............................................................................................... 41
Depression .......................................................................................................... 45
Anxiety Disorders ............................................................................................... 49
Summary ............................................................................................................. 53
The Problem of Psychiatric Comorbidity .................................................................. 54
Study Objectives ........................................................................................................ 60
METHODS ....................................................................................................................... 63
Sample 1: ADHD Adults and Controls ..................................................................... 64
Participants ......................................................................................................... 64
Sample 2: Alcoholism Study ..................................................................................... 68
Participants ......................................................................................................... 68
Neuropsychological Test Battery .............................................................................. 74
Full Scale IQ (FSIQ) .......................................................................................... 74
Neuropsychological Tests .................................................................................. 75
RESULTS ......................................................................................................................... 86
Data Preparation ................................................................................................. 86
Sample Description ............................................................................................ 89
Checks on the Validity of Combining the Alcoholism and ADHD Samples ..... 92
Conﬁnnatory Factor Analyses to Establish Latent Variables for Psychiatric
Disorders and Cognitive Tests ............................................................................ 94
Tests of Hypotheses ................................................................................................. 101
Analysis 1: Testing the Componential Model for Disorder-Speciﬁc Patterns of
EF Test Performance ........................................................................................ 101
Analysis 2: Testing the Comorbidity-Speciﬁcity Model that Only One or a Few
Disorders Are Uniquely Associated with EF Deﬁcits ...................................... 106

Analysis 3: Testing the Comorbidity-Nonspeciﬁcity Model that Number of
Disorders Predicts Performance on EF Tests ................................................... 110
Analysis 4: Testing the Dimension-Speciﬁcity Model that Shared Underlying
Dimensions of Psychopathology are Differentially Related to Cognitive Task

Performance ...................................................................................................... 114
Overall Summary .............................................................................................. 1 18
DISCUSSION ................................................................................................................. 120
Examination of Hypotheses ..................................................................................... 120
Nonspeciﬁc EF Deﬁcits ................................................................................... 120
Disorder-Speciﬁc EF Deﬁcits .......................................................................... 122
Process-Speciﬁc Disorder-Related EF Deﬁcits ................................................ 123
Dimension-Speciﬁc EF Deﬁcits ....................................................................... 129
Conclusions on Main Models ........................................................................... 131
Additional Hypotheses ..................................................................................... 133

The present study provided information about other questions that further
elucidate the association between EF and psychiatric disorders. ..................... 133
Limitations ............................................................................................................... 138
Strengths .................................................................................................................. 142
Conclusions .............................................................................................................. 143
APPENDIX ..................................................................................................................... 179
Anatomy of the Frontal Lobes and Frontal Subcortical Circuits ............................. 179

Functional Signiﬁcance of Preﬁ'ontal Regions: Prefrontal Behavioral Syndromes 183

REFERENCES ............................................................................................................... 1 87

vi

LIST OF TABLES

Table 1. Summary of Major Findings in the Literature Regarding EF Component

Deﬁcits ........................................................................................................................... 145
Table 2. Demographic Information for the Combined, Alcoholism, and ADHD

Samples .......................................................................................................................... 148
Table 3. Numbers of Lifetime Disorders in the Combined, Alcoholism, and ADHD
Samples .......................................................................................................................... 149
Table 4. Numbers of Current Disorders in the Combined, Alcoholism, and ADHD
Samples .......................................................................................................................... 150
Table 5. Means and Standard Deviations of Cognitive Tasks for the Combined,
Alcoholism, and ADHD Samples .................................................................................. 151
Table 6. Mean Scores for Cognitive Tasks for Lifetime Disorders in the Combined
Sample ............................................................................................................................ 152
Table 7. Mean Scores for Cognitive Tasks for Current Disorders in the Combined
Sample ............................................................................................................................ 154
Table 8. Correlations Amongst Cognitive Tasks in the Combined Sample ................. 156
Table 9. Correlations Amongst Cognitive Tasks in the Alcoholism Sample ............... 157
Table 10. Correlations Amongst Cognitive Tasks in the ADHD Sample .................... 158
Table 11. Correlations Amongst Lifetime Disorders in the Combined Sample ........... 159
Table 12. Correlations Amongst Lifetime Disorders in the Alcoholism Sample ......... 160
Table 13. Correlations Amongst Lifetime Disorders in the ADHD Sample ................ 161
Table 14. Correlations Amongst Current Disorders in the Combined Sample ............. 162
Table 15. Correlations Amongst Current Disorders in the Alcoholism Sample ........... 163
Table 16. Correlations Amongst Current Disorders in the ADHD Sample .................. 164

Table 17. Summary of Regression Analysis for Lifetime Disorders Predicting
Performance on Executive Function Composite Score ................................................. 172

vii

LIST OF TABLES (continued)

Table 18. Summary of Regression Analysis for Current Disorders Predicting
Performance on Executive Function Composite Score ................................................. 172

Table 19. Summary of Regression Analysis for Lifetime Disorders Predicting
Performance on Processing Speed Composite Score ..................................................... 173

Table 20. Summary of Regression Analysis for Current Disorders Predicting
Performance on Processing Speed Composite Score ..................................................... 173

Table 21. Frequencies of Numbers of Individual Lifetime Disorders in the Combined,
Alcoholism, and ADHD Samples .................................................................................. 174

Table 22. Frequencies of Numbers of Individual Current Disorders in the Combined,
Alcoholism, and ADHD Samples .................................................................................. 174

viii

LIST OF FIGURES

Figure 1. Initial proposed model representing the relationships between DSM-IV

disorders based on Krueger’s (1999) ﬁndings ............................................................... 146
Figure 2. Proposed model to examine the relationships between underlying core

psychopathology domains and neuropsychological test performance ........................... 147
Figure 3. Final accepted latent model for internalizing disorders ................................ 165
Figure 4. Final accepted latent model for externalizing disorders ............................... 165
Figure 5. Final accepted latent model for executive function tests ............................... 166
Figure 6. Latent measurement model for cognitive test performance .......................... 167

Figure 7. Mean z-scores for performance on executive function tests for lifetime
disorders with nonoverlapping diagnostic groups ......................................................... 168

Figure 8. Mean z-scores for performance on executive function tests for current disorders
with nonoverlapping diagnostic groups ......................................................................... 169

Figure 9. Mean z-scores for performance on executive function tests for lifetime
disorders with overlapping diagnostic groups ............................................................... 170

Figure 10. Mean z-scores for performance on executive function tests for current
disorders with overlapping diagnostic groups ............................................................... 171

Figure 11. Measurement model for the relationships between lifetime internalizing and
externalizing disorders on cognitive task performance ................................................. 175

Figure 12. Model representing the predictive relationships between lifetime internalizing
and externalizing disorders on cognitive task performance ........................................... 176

Figure 13. Measurement model representing the relationships between current
internalizing and externalizing disorders on cognitive task performance ..................... 177

Figure 14. Model representing the predictive relationships between current internalizing
and externalizing disorders on cognitive task performance .......................................... 178

ix

Executive Function Deﬁcits in Psychopathology

Executive functions (EF) refer to cognitive abilities that are recruited in the
pursuit of goal-directed activity. Deﬁcits in this cognitive domain have been associated,
at least conceptually, with risk for the emergence of a wide range of psychological
disorders. These include, but are not limited to, attention-deﬁcit hyperactivity disorder
(ADHD), major depressive disorder (MDD), anxiety, obsessive-compulsive disorder
(OCD), schizophrenia, antisocial personality disorder (ASPD), alcoholism, and drug use.
To demonstrate interest in studying EF and psychopathology, each of these disorders
were entered as keywords into a PsycINFO search for articles dating from 1995 through
mid 2007, with the stem “executive function*” also appearing in the keyword ﬁeld. The
numbers of citations found for each disorder were: “schizophrenia” — 767; “depression” —
520; “ADHD” — 437; “anxiety” —— 193; “drug abuse” — 134; “alcoholism” — 91;
“antisocial” — 85; and “OCD” — 75. These numbers undoubtedly underestimate the actual
number of studies in the ﬁeld, given that different terms are often used for EF as well as
the individual disorders. However, they illustrate the strong interest in understanding the
involvement of EF across numerous psychiatric disorders. The speciﬁc results of these
studies vary considerably, and as I will emphasize, the effects of comorbidity have been
understudied and are poorly understood. Yet it is clear that the presence of some EF
impairment has been suggested for each of these disorders.

The parallelism in the literature, with EF deﬁcits being associated with such a
wide range of disorders, is striking and not easily explained. How can claims regarding
the involvement of EF deﬁcits be made across so many diverse disorders? Moreover,

going beyond simple associations, EF impairment has often been hypothesized to play a

causal role in the etiology of psychiatric disorders. However, if each of these hypotheses
were correct, then how the same underlying cognitive deﬁcit could lead to these very
different symptom presentations is perplexing. This question of disorder-deﬁcit
speciﬁcity, or the discriminant validity problem, has been posed by Pennington &
Ozonoff (1996): “How can symptomatically different complex behavior disorders all be
due to the same cognitive and/or neural dysﬁinction?” (p. 57). Such queries have led to
attempts to differentiate between EF as an etiologically speciﬁc deﬁcit versus a
nonspeciﬁc marker of dysfunction for individual psychiatric disorders.

In examining the discriminant validity problem, the goal of researchers was often
to ﬁnd the primary neurocognitive deﬁcit for each disorder, with primary referring to a
deﬁcit that is universal, speciﬁc, necessary, and sufﬁcient to cause the symptoms of the
disorder (Pennington & Ozonoff, 1996). Several problems with searching for this so-
called primary deﬁcit have precluded clarity on the issue and suggest the need to
reconceptualize.

The ﬁrst problem is that without the use of longitudinal study designs and
neuroimaging technology to test for localized changes in the structure and ﬁrnction of the
brain (which may be beyond the scope of present technology), the discriminant validity
problem cannot be solved at the level of brain mechanisms. Second, it is not realistic to
expect to ﬁnd a single cognitive deﬁcit that is sufﬁcient to cause all cases of what are
now recognized to be multifactorial psychiatric disorders (Garber & Hollon, 1991 ).
Finally, any search for a primary deﬁcit assumes that the current classiﬁcation system for
differentiating disorders, the Diagnostic and Statistical Manual of Mental Disorders,

Fourth Edition (DSM-IV; American Psychiatric Association, 1994) is accurate, which

would require placing greater faith in the present nosology than it may be due (Garber &
Hollon, 1991). Given the problems with searching for a primary deﬁcit, this line of
reasoning cannot be strictly applied in an attempt to understand the relationship between
EF deﬁcits and psychiatric disorders. It may be more realistic to demonstrate “partial
speciﬁcity” between psychopathology and EF deﬁcits. Such a relationship may be
suggested if EF deﬁcits were one of several factors consistently implicated in the
manifestation of certain disorders. As will be discussed, even this level of speciﬁcity has
been difﬁcult to support in the literature.

This study attempted to better understand the speciﬁcity of EF deﬁcits to wide-
ranging psychiatric disorders by examining the relationship at different levels in the
diagnostic hierarchy. The DSM-IV is the currently accepted diagnostic classiﬁcation
system for individual disorders, but there may be other ways to conceptualize psychiatric
disorders given their frequent comorbid presentations (Angold, Costello, & Erkanli,
1999; Kessler, Berglund, Demler, J in, & Walters, 20053; Kessler, Chiu, Demler, &
Walters, 2005b; Kessler, McGonagle, Zhao, Nelson et al., 1994) and similar
symptomatology and characteristics (Krueger, 1999). The focus of this study was to
examine the relationship between EF deﬁcits and exemplars of putative EF-related
psychopathology using four “models” or hypotheses to explain their associations. Each
hypothesis represented an alternative solution to the question of how EF deﬁcits could be
related to such a wide range of disorders. These four possible models incorporated
different levels of abstraction in a diagnostic classiﬁcation “hierarchy” — from individual
disorders, to multiple or comorbid diagnoses, to shared underlying psychopathological

processes. By using alternative conceptualizations of psychopathology, this study

attempted to understand both the cognitive levels and possible proﬁle differences
between these disorders without being constrained by only one deﬁnition of
psychopathology. The goal was to examine possible distinctions in types and extent of
EF deﬁcits that are associated with the presence versus absence of putatively different
kinds of psychopathology. The four main hypotheses are detailed as follows.

The ﬁrst possibility, at the level of the individual disorders, was that different
disorders are associated with different types of EF deﬁcits. In other words a
Componential Model, or proﬁle analysis of EF deﬁcits, may show that individual
psychiatric disorders, as deﬁned by the DSM-IV, are associated with deﬁcits in different
types of EF abilities so that several disorders may show some form of an EF deﬁcit, but
the affected component processes may vary between disorders.

Second, moving to the level of multiple disorders, two more possibilities are
apparent. One (second overall) is that speciﬁcity of EF deﬁcits is present for only one or
very few disorders. In this case, the appearance of EF deﬁcits with the other disorders
may be due to comorbidity between disorders that are and are not associated with EF
deﬁcits. As mentioned, comorbidity is oﬁen uncontrolled in neuropsychological studies
of mental disorders, and use of the appropriate controls (which, admittedly, are not easily
arrived at — see subsequent discussion) may provide a clearer picture of the relationship
between disorders and EF deﬁcits. Therefore, examination of this Specificity Model may
suggest that only a few disorders show EF deﬁcits while others do not, once comorbidity
is accounted for.

The other possibility at the level of multiple disorders (third overall) was that

there is no speciﬁcity of EF deﬁcits with the individual disorders, but that EF impairment

is related to the number of co-occuning disorders. That is, perhaps EF deﬁcits are seen
in relation to number, rather than form, of psychopathology (Kessler et al., 2005a). In
this case, the number of disorders alone would be a marker for impaired EF, regardless of
the particular disorder(s) that is (are) manifested. This Nonspeciﬁcity Model would
suggest that number, not type, of disorder could account for the ﬁndings of EF deﬁcits
across multiple disorders.

Finally, at the level of symptom dimensions, EF deﬁcits may be related to only
certain psychopathological processes and not to others. Recent research has suggested
that high levels of comorbidity amongst DSM-III-R (Kessler et al., 1994) and DSM-IV
disorders (Kessler et al., 2005a) may result from shared underlying core
psychopathological characteristics (Kessler, Crum, Warner, Nelson et al., 1997; Krueger,
1999, 2005). In a population-based prevalence study, Krueger (1999) found that two
broad, superordinate factors accounted for the pattern of correlations among liabilities to
have common mental disorders: an internalizing factor (that included two subfactors),
and an externalizing factor (note that these are Krueger’s (1999) labels for the factors).
In other words, the individual disorders appear to cluster together according to shared
underlying processes, with mood and anxiety disorders loading onto an “internalizing”
factor, and ASPD and substance use disorders loading onto an “externalizing” factor
(Krueger, 1999). This ﬁnding suggests that current nosology may be misleading to the
extent that it relies on completely separating behavioral syndromes into discrete disorders
(Krueger, 1999, 2005). Krueger (1999) recommended that research should focus on
these core processes rather than the individual disorders. It is possible that one of these

higher-order dimensions of psychopathology (i.e., internalizing versus externalizing

syndromes) may be related to EF deﬁcits, while the other may not. Therefore, according
to this Dimension-Speciﬁc Model, the ﬁndings of EF deﬁcits across multiple disorders
may be a result of shared underlying core processes that are related to EF deﬁcits.

Because EF are potentially relevant to so many disorders, to constrain the scope
of the study some decisions were made regarding which disorders to include. The
following disorders were selected for inclusion: childhood externalizing disorders (i.e.,
ADHD, conduct disorder (CD), and oppositional deﬁant disorder (ODD)), ASPD,
alcoholism, drug dependence, depression, and anxiety disorders. Theses disorders were
included for three reasons.

First, these are among the most common disorders diagnosed in the general
population and as such they are usually included in major prevalence studies (i.e., Kessler
et al., 2005a; Kessler et al., 2005b; Kessler et al., 1994; Krueger, 1999). Disorders that
were not included in this list, such as bipolar disorder, schizophrenia, and borderline
personality disorder have also been associated with EF impairments (Antonova, Sharma,
Morris, & Kumari, 2004; Henry & Crawford, 2005; Monarch, Saykin, & Flashrnan,
2004; Quraishi & Frangou, 2002), but their low base rates in community samples made it
unlikely that they will be adequately represented in the present study (Ekselius, Tillfors,
Furmark, & Fredrikson, 2001; Kessler et al., 2005a). Secondly, the disorders that have
been selected for inclusion ﬁequently co-occur with each other (Kessler et al., 2005b),
such that teasing apart speciﬁc EF correlates has been difﬁcult. Third, these disorders
share similar surface characteristics, such as impulsivity (i.e., ADHD, drug dependence,
ASPD), and withdrawal or lack of initiative (i.e., MDD and anxiety disorders), which

splits them conceptually into those disorders that are typically associated with

externalizing versus internalizing behaviors, respectively (Achenbach, 1966; Krueger,
1999). By the same token, in order to address the dimension-speciﬁc model (Hypothesis
4, above), it was necessary to include both “externalizing” and “internalizing” disorders
in the present study. Therefore, the selection of disorders for the present study was based
upon their associations with EF deﬁcits, prevalence in the population, high rates of
comorbidity, and similarities in symptom proﬁles.

The hierarchical approach to analyzing the relationship between EF and
psychopathology using the four proposed models (i.e., componential, comorbidity-
speciﬁc, comorbidity-nonspeciﬁc, and dimension-speciﬁc) underscores the importance of
looking at the broader concepts of behavior as well as their individual manifestations
(Gorenstein & Newman, 1980). These hypothesized models to explain the relationship
between EF deﬁcits and psychopathology are not necessarily mutually exclusive; one
may better explain the relationship than the others, or they may all contribute to our
understanding. By examining individual as well as comorbid disorders from multiple
perspectives, this study aimed to clarify the nature of EF deﬁcits in these conditions.
While a long-term goal for this research is to ﬁnd clues regarding the etiology of these
conditions, there are other more immediate beneﬁts associated with this line of inquiry.

Why Study Psychopathology and Executive Functions?

It is understandable that researchers in so many areas of psychopathology have
sought use the concept of EF. Conﬁrming a link between EF and individual psychiatric
diagnoses can provide objective support for classifying behavior as disordered. There is
much debate about the validity of various psychological disorders (see: Clark, 1999;

First, Spitzer, Gibbon, & Williams, 1997; Lilienfeld & Marino, 1995, 1999; Spitzer,

1999; Wakeﬁeld, 1999). For instance, the current diagnostic nosology, the DSM-IV
(American Psychiatric Association, 1994), has been criticized for being over-inclusive in
its criteria because it does not meet the dysfunction requirement in the “harmful
dysfunction” conceptualization of a mental disorder (Wakeﬁeld, 1997). Wakeﬁeld
(1992) indicated that while “harm” involves a societal value judgment, dysfunction
should ideally be a scientiﬁc term referring to “the failure of an internal mechanism to
perform a natural function for which it was designed” (p. 383). In this inﬂuential view,
both harm and dysfunction are required to label behavior as a disorder. Dysfunction, if it
can be accurately identiﬁed, has the potential to differentiate between the abnormal
functioning of internal mechanisms and nondisordered reactions to external stressors to
avoid pathologizing normal behavior (Wakeﬁeld, 1992). Although there have been
criticisms of Wakeﬁeld’s analysis (Clark, 1999; Lilienfeld & Marino, 1995, 1999), he
provides a useful perspective from which to begin to reﬁne the deﬁnition of disorder
(Cosmides & Tooby, 1999; Spitzer, 1999). Being able to substantiate a connection
between a disorder and EF deﬁcit could provide an objective index of dysﬁmction, thus
helping to validate behavior as disordered.

The potential gains that could result ﬁom studying the relationship between EF
deﬁcits and major disorders extend beyond this very general objective of being able to
validate a disorder with regard to dysfunctional internal mechanisms in the mind or the
brain. It may also provide insight into the validity of the speciﬁc criteria used to deﬁne
and differentiate various conditions. This is particularly so, however, when considering
more than one disorder at the same time. For instance, a problematic result of the current

deﬁnitions for various disorders is that there is a high degree of comorbidity between

disorders. Recent research using DSM-IV criteria suggested that while almost half of the
population had met criteria for a psychiatric disorder at some point in their lifetime, the
prevalence of having two or more lifetime conditions was 27.7%, and the prevalence of
having three or more disorders was 17.3% (Kessler et al., 2005a). Prevalence of
comorbidity was similar for the previous classiﬁcation system, the DSM-III-R, as well
(Kessler et al., 1994). Therefore, the majority of lifetime disorders are comorbid
conditions, which raises questions about the validity of our diagnostic criteria — or at least
about the assumption of discrete conditions.

The high rates of comorbidity may partially result from problems with the
nosology (i.e., overlapping symptomatology or the artiﬁcial separation of similar
disorders; Angold et al., 1999; Krueger, 1999) that an understanding of the speciﬁcity of
neuropsychological deﬁcits to individual disorders may help to clarify and reﬁne. In
other words, being able to differentiate disorders based upon patterns of
neuropsychological test performance could provide support for the validity of individual
disorders and might help to determine whether comorbid conditions represent some
“combination” of the individual disorders or a separate, third disorder (Angold et al.,
1999). On the other hand, if the same deﬁcits are seen for multiple individual disorders,
this may provide support for the idea that common underlying pathological processes
could lead to symptoms of several related conditions (Krueger, 1999). In short, an
understanding of the neuropsychological deﬁcits could guide the manner in which we
speciﬁcally deﬁne and conceptualize disorders, with implications for their assessment

and, potentially, treatment.

As mentioned, EF deﬁcits have been associated with a wide range of disorders.
Despite the problem of parallelism in the literature and the importance of clarifying it
before meaningful conclusions can be drawn about individual disorders, most research
has focused upon the EF deﬁcits associated with individual conditions. Fewer studies
have examined this issue with the intention being to understand the speciﬁcity versus
generalizability of EF deﬁcits to different psychological disorders. Those that have
involved comparisons across disorders have dealt with only a small number of disorders
(Airaksinen, Larsson, & Forsell, 2005; Boldrini, Del Pace, Placidi, Keilp, Ellis, Signori et
al., 2005; Fossati, Amar, Raoux, Ergis, & Allilaire, 1999; Moritz, Birkner, Kloss, J ahn,
Hand, Haasen et al., 2002; Selby & Azrin, 1998; Uekermann, Daum, Schlebusch, Wiebel,
& Trenckrnann, 2003; Weyandt, Rice, Linterman, Mitzlaff, & Emert, 1998), and/or
examined a number of domains of cognitive abilities without a speciﬁc focus upon EF
(Riordan, Flashman, Saykin, Frutiger, Carroll, & Huey, 1999). A small number of
review papers have integrated the ﬁndings on the neuropsychological correlates across
many individual disorders, and these have been helpful in highlighting the speciﬁcity of
EF deﬁcits in childhood (Pennington & Ozonoff, 1996; Sergeant, Geurts, & Oosterlaan,
2002). However, as several of those reviewers pointed out, comparing effects across
different studies which used different methodology makes interpretation difﬁcult, and
this has slowed progress in the study of EF deﬁcit speciﬁcity (Sergeant et al., 2002). An
empirical study considering key, frequently comorbid disorders is needed and is, for the
most part, unprecedented.

The present study was an attempt to address some of these issues to more clearly

understand the relationship between psychopathology and EF deﬁcits. The empirical

10

examination of multiple disorders within two large and well-deﬁned samples of adults
was guided by the following suggestions that were made by Sergeant et a1. (2002) to
improve research and enhance our knowledge on the speciﬁcity of EF deﬁcits: (a)
include multiple clinical group comparisons; (b) use EF tasks which show frontal lobe
involvement and some process speciﬁcity; (0) use the exact same EF tasks and dependent
variables across the different disorders/samples being studied; and (d) provide some form
of statistical control for comorbid disorders to increase conﬁdence that any observed
deﬁcits are related to the disorder in question and not to other co-occurring disorders.

Regarding this latter point, control of comorbidity is a conceptually complex
issue. Therefore, comorbidity was controlled in some analyses, but it was also a focus of
other analyses in order to understand how co-occurring disorders interacted to affect
neuropsychological performance. One issue here was that it was possible that the overlap
of symptoms and interactions between some disorders are necessary and core
components of their psychopathology. Controlling their covariance could remove key
aspects of the psychopathology that may be contributing to EF deﬁcits. Including
multiple levels of diagnostic analysis with the four models helped to address this issue.
Before shifting the focus to review previous literature on EF and psychopathology, it is
important to examine the EF construct and clarify how it was conceived in the present
study.

Executive Functions

The concept of EF is poorly deﬁned, often underspeciﬁed, and its deﬁnition tends

to vary between studies. As a result, the very term ‘executive functions’ is becoming

somewhat outdated, as there is (a) increased recognition of the need for more descriptive

ll

terminology to better specify and capture the components of EF, and (b) a need to avoid
the ‘meta-cognitive’ or hierarchical implications connoted by the use of the word
‘executive’ (for example, Denckla (1996) suggests “control processes”, p. 264). Given
that the term executive functions is well recognized by most neuropsychological
researchers as referring to a group of cognitive processes involved in goal-directed
activity, it was used herein despite its recognized meta-conceptual drawbacks.
Nonetheless, a number of conceptual clariﬁcations are necessary.

EF have often come to be synonymous with the brain’s frontal lobes, and have at
times been taken to be any process disrupted by damage to that region of the brain
(Denckla, 1996; Hayes, Gifford, & Ruckstuhl, 1996). Such an interpretation dates to
early descriptions of disorganized behavior, impulsivity, and lack of initiation following
frontal lobe damage, such as references to the well-known ﬁ'ontal lobe patient Phineas
Gage. This “frontal metaphor” (Pennington & Ozonoff, 1996), however, does not clarify
any shared mechanism amongst EF processes, nor provide a functional understanding of
the EF concept; instead, it explicitly avoids an operational deﬁnition of the concept
(Barkley, 1996). Further, while the frontal cortex is important to EF, its role is not
exclusive. Some patients with frontal lobe damage do not have any problems on EF tasks
(Shallice & Burgess, 1991), and damage to other key brain areas may also cause EF
impairment (Anderson, Darnasio, Jones, & Tranel, 1991). The vast circuitry connecting
the frontal lobes to other regions of the brain means that other forms of damage may
affect EF (Lichter & Cummings, 2001). Therefore, a sole focus on frontal localization
does not aid in understanding the EF concept or the related, more current concepts, such

as cognitive control.

12

Many different cognitive abilities have been subsumed under the heading
“executive functions,” and these tend to vary between theorists. Borkowski and Burke
(1996) observed that a major impediment to progress in the study of EF was the
ambiguity of the construct and resultant lack of shared meaning throughout different
disciplines within which EF are important (i.e., cognitive psychology, developmental
psychology, neuropsychology, and education). The central mechanism in EF has been
variously hypothesized as working memory (Pennington, Bennetto, McAleer, & Roberts,
1996; Smith & J onides, 1999), attention (Barkley, 1996), information processing
(Borkowski & Burke, 1996), and inhibition (Denckla, 1996). Various cognitive models
have combined some of these concepts in their deﬁnitions, such as Norman and
Shallice’s (1986) Control of Action model which included an executive Supervisory
Attentional System (SAS) to cope with novel information, and Baddeley’s (1986)
working memory model which included a central executive that was responsible for the
selection, initiation, and termination of processing routines (i.e., encoding, storing, and
retrieving). Each of these theoretical models is important in helping to understand the
concept of EF, but they are somewhat narrow in their purely cognitive interpretations of
the concept.

A broader conceptualization, involving multiple processes, was presented by
Pennington and Ozonoff (1996), and this deﬁnition was employed herein. The reason for
this selection was three-fold: their deﬁnition clearly outlines cognitive components of EF
that are accessible to direct measurement using clinical measures, they provide a theory
to conceptually link these component processes, and this deﬁnition of EF is relevant to

clinical problems/assessment involving EF and potential relationships between deﬁcits

13

and disorders. These authors (see also Welsh & Pennington, 1988), deﬁne EF as “the
ability to maintain an appropriate problem-solving set for attainment of a future goal.
This set can involve one or more of the following: (a) an intention to inhibit a response or
to defer it to a later more appropriate time, (b) a strategic plan of action sequences, and
(c) a mental representation of the task, including the relevant stimulus information
encoded into memory and the desired future goal-state” (pp. 201 -202). Thus, multiple
processes are included in this description.

Pennington and Ozonoff (1996) detail two additional concepts that are central to
their deﬁnition of EF. The ﬁrst is the idea that the selection of an action is speciﬁc to and
appropriate for the context within which the action is required. The ability to choose a
context—speciﬁc action is particularly important when other actions are available that
would be inappropriate to the particular context. Secondly, the selection of an action
depends upon the integration and satisfaction of constraints ﬁom a variety of domains
including, but not limited to, memory, perception, motivation, and affect. These concepts
help to explain the heterogeneity of EF processes, their particular relevance in novel
contexts, and the importance of other cognitive domains in performance on EF tasks. It
also explains the applicability of executive abilities to many aspects of human behavior,
as they integrate and inﬂuence cognitive as well as social, emotional, and motivational
drive states in the attainment of future goals.

While Pennington & Ozonoff’s (1996) deﬁnition guides a theoretical
understanding of the EF concept, it still leaves the speciﬁc EF processes somewhat vague
and abstract. Thus, more practically, tasks they consider to assess “executive functions”

are those that are thought to involve set-shifting and set maintenance, interference

l4

control, inhibition, integration across space and time, planning, and working memory
(Pennington & Ozonoff, 1996). These six cognitive operations thus comprise a single
conceptual model of EF that may be measured during a clinical neuropsychological
examination. That model guided the current work. An issue that it highlights, however,
is that EF is composed of multiple processes (Ward, Roberts, & Phillips, 2001), which
complicates the discussion of EF as a single unitary construct.

The question of unity versus diversity of processes within the EF construct and its
associated clinical measures has been examined recently in both normal adult (Miyake,
Friedman, Emerson, Witzki, & Howerter, 2000) and neurological populations (Duncan,
Johnson, Swales, & Freer, 1997). Both groups of researchers concluded that there is
support for both unity and diversity, or multiplicity, in understanding EF. Diversity is
exempliﬁed in EF tasks both through clinical observations as well as research ﬁndings.
For example, clinical observations that some people perform poorly on one EF task but
normally on others highlights the diversity of EF tasks. Duncan and colleagues (1997)
tested this ﬁnding empirically and found a similar effect in that performances on tests
assessing EF tended to correlate weakly with one another; however, note that is not a
consistent ﬁnding (Burgess, 1997; Hanes, Andrewes, Smith, & Pantelis, 1996; Miyake et
al., 2000). Providing additional support for the diversity of EF processes, subcomponent
analyses have revealed stronger relationships between tasks that assess a single
component of EF (i.e., set shifting) than between tasks that assess different components
(Miyake et al., 2000), and convergent validity has been demonstrated for these
component processes (Salthouse, Atkinson, & Berish, 2003). Therefore, there appear to

be multiple processes involved in the EF construct.

15

On the other hand, lending support to the unity of the EF construct, there also
appears to be an underlying shared mechanism or ability that contributes to performance
across different EF tasks (Duncan et al., 1997; Hanes et al., 1996; Miyake et al., 2000;
Salthouse et al., 2003). Some have suggested that this higher order factor is the general
intelligence factor or g (Salthouse et al., 2003), therefore criticizing the lack of divergent
validity of the EF construct. However, not everyone has agreed with this interpretation as
the opposite can be argued, that components of EF underlie IQ (Conway, Kane, & Engle,
2003; Kane & Engle, 2002). Regardless, while g contributes to performance across a
wide range of EF abilities as well as the measurement of other cognitive domains, other
shared mechanisms have been suggested which more uniquely differentiate EF from
other cognitive constructs. For instance, individuals with a head injury showed a
common deﬁcit of goal neglect, or a tendency to disregard the requirements of tasks
despite a conscious awareness of the rules, which contributed to impaired performance
across EF tasks (Duncan et al., 1997). Similarly, it has been hypothesized that the active
maintenance of goals and other task-relevant information in working memory is crucial to
performance across EF tasks (Miyake et al., 2000). Such ﬁndings provide support for the
unity of the EF construct.

Therefore, although there is not yet consensus regarding the number or nature of
separable components of EF (Salthouse et al., 2003), tasks assessing EF appear to
represent “unique aspects of executive functioning with some overlapping of variance”
(Delis, Kaplan, & Kramer, 2001 , p. 82). It appears as though multiple components are
involved in the EF construct, but these component processes may be partially uniﬁed by a

shared ability, such as goal maintenance, which is required across EF tasks. This unity

l6

and diversity of EF processes provides the opportunity to examine EF ﬁ'om a holistic
perspective, with a focus on the combined measurement of EF, as well as from a
component perspective, with more of a focus on individual processes.
Measurement of Executive Functions

Similar to the difﬁculties with understanding the concept of executive function, a
number of issues have been raised about the tasks used to measure EF. One main
problem in measuring EF is task impurity (Miyake et al., 2000). Since EF recruit other
cognitive processes in their activity, all tasks assessing EF involve other operations to
some degree. This is more of an issue with the molar tasks used in clinical practice than
cognitively-based molecular tasks. Molar tasks are complex and as such draw upon a
number of cognitive domains. These tasks were designed as “sign tests” to detect brain
damage, not to isolate component processes (Lezak, Howieson, & Loring, 2004).

So-called molar clinical measures have both strengths and weaknesses when it
comes to assessing psychopathology. With regard to weaknesses, four speciﬁc problems
with such molar tasks are that they may lack: (1) strong theoretical foundations; (2) the
ability to identify component processes that contribute to performance; (3) consistently
reliable and normally distributed performance; and (4) sensitivity to the same underlying
processes across the range of performance (Pennington & Ozonoff, 1996). Therefore,
despite obvious functional EF impairments in the “real world,” injured individuals may
perform normally on standard EF tests (Eslinger & Darnasio, 1985). It has been
suggested that the ability to accurately assess EF in laboratory tasks may be precluded by
the methods in which tasks are administered: clinicians provide structure and

organization to the task, and there is less emphasis on the participant discovering or

17

creating the solution (Burgess, 1997; Denckla, 1996; Rabbitt, 1997). Therefore, potential
problems with the sensitivity and speciﬁcity of molar EF tasks in head-injured
populations, and the manner in which they are administered, have called into question
how validly EF is measured by common clinical tasks.

Providing a balance to these issues, however, the molar tasks have a number of
strengths that enable them to provide unique contributions to clinical assessments.
Speciﬁcally, these include: (1) task complexity that enables for the simultaneous
measurement of multiple integrated processes; (2) widespread clinical applicability and
ease/portability of administration; (3) extensive clinical validation literature (detailed
later); (4) availability of national population norms in many instances; and (5) availability
of neuroimaging data so that their substrates are partially understood (detailed later).
Thus, there are distinct strengths for molar tasks.

To further elaborate on the strengths of molar tasks, it may be erroneous to
believe that what are considered problems with molar tasks could be ﬁxed without losing
some inherent aspects of the EF concept. In other words, the fact that EF tasks appear to
have low sensitivity and assess multiple interacting processes may not necessarily reﬂect
weaknesses in the tasks. Instead, they may result to some degree from exactly what we
are trying to measure, so that removing them may actually invalidate the measurement of
EF. For instance, the tasks’ low reliability, which has been criticized and likely
contributes to the observed low correlations between tasks (Salthouse et al., 2003), may
result in part from one of the main foundations of EF tasks: they are designed to assess
responses to novelty, and as such they are most valid the ﬁrst time the task is

administered (Burgess, 1997), or when the task is not remembered.

18

Further, since EF typically refers to the coordination of multiple cognitive
processes, and not one single operation, it may be difﬁcult, and perhaps defeat the
purpose of measuring EF, to distinguish the effects of constituent processes from EF
processes in performance (Salthouse et al., 2003). To split these components completely
in an effort to focus solely on molecular processes may result in a loss of the very
construct that we are attempting to measure. The clinical integrity of the EF processes
may be maintained with these molar tasks, more so than it would be if they were broken
into parts.

Finally, although measurement issues should always be kept in mind with regard
to interpretation of task performance, at this time molar EF tasks are widely used in
clinical practice and research (Lezak et al., 2004; Retzlaff, Butler, & Vanderploeg, 1992).
It is therefore important to understand how these particular tasks are affected by the
presence of psychiatric disorders. Their inclusion in this study provides clinical
applicability and facilitates comparison with previous research in the area.

Therefore, an important asset of molar tasks is that they are clinically applicable
and may be used to distinguish between normal and abnormal performance (Lezak et al.,
2004). They provide the opportunity to assess multiple interacting processes to
determine the overall integrity of EF processes in tandem. Molecular tasks, on the other
hand, can assess “purer” sub-components, which may remove the potentially
confounding effects of other cognitive processes and offset the difﬁculties in interpreting
performance on molar tasks. The complexities of EF measurement cannot be easily
solved; however, the advantages and disadvantages associated with the various

measurement approaches may be balanced by the use of multiple tests, which assess both

19

molar and molecular processes to some extent. Therefore, the approach adopted here was
to include representatives of both types of tasks, while attempting to capture the EF
processes suggested by Pennington & Ozonoff (1996). This may provide the needed
ﬂexibility to look at component as well as holistic EF processes.

Complications in the measurement of EF have led to increased use of latent
measurement techniques in recent years (Friedman, Miyake, Corley, Young, deFries, &
Hewitt, 2006; Friedman & Miyake, 2004; Nigg, Stavro, Ettenhofer, Hambrick, Miller, &
Henderson, 2005; Salthouse et al., 2003). Such methods provide a means to reduce the
effects of task impurity and heterogeneity amongst individual EF tests by pooling the
shared variance from several indices of EF. This maximizes both construct-relevant
variance and therefore increases reliability and interpretability when including multiple
different measures of EF. Such an approach was utilized, along with other methods, in
the present examination.

To appreciate how both component and holistic EF processes may be supported
by neural structures, some understanding of the anatomy of the ﬁ'ontal lobe region is
needed. This description can provide a conceptual anatomic basis for the componential
model of the relationship between EF and psychopathology, and aids in interpreting
regions of activation in neuroimaging data for tasks and disorders to be described later.

Functional Signiﬁcance of F rontal-Subcortical Circuits

The frontal lobes of the brain, particularly the prefrontal cortex, are most highly
developed in humans (Petrides & Pandya, 2002). They are the site where partially
overlapping systems integrate highly-processed external sensory and multimodal

information from posterior cortical areas (Petrides & Pandya, 2002) with internal

20

cognitive and emotional responses to modulate motivation and facilitate motor responses
(Lichter & Cummings, 2001). Prefrontal circuits processing cognitive and/or emotional
information may be involved in the manifestation of psychopathology.

It is possible that EF problems associated with frontal-subcortical regions
contribute to psychopathology in a top-down information-processing manner. Poor
modulation of attention and other cognitive processes may interact with temperament or
personality factors (i.e., negative affect or low positive affect) to cause pathological
changes in mood and behavior. Information-processing theories have received support in
the development of anxiety and depression (see Vasey, Dangleish, & Silverman, 2003).
On the other hand, positive affect has been shown to improve cognition, possibly through
increases in dopamine in frontal-subcortical pathways (Ashby, Isen, & Turken, 1999).
Thus, the lack of or generally low positive affect associated with various psychiatric
conditions may lead to problems with cognition. These theories provide means by which
cognition may lead to symptoms of psychopathology as well as for symptoms of
psychopathology to cause cognitive problems. While the speciﬁc direction of effects
remains unclear, these ideas highlight the importance of the prefrontal region of the brain.
A brief overview of the neuroanatomy and functional signiﬁcance of the prefrontal cortex
to EF processes and psychopathology is provided here. A more detailed description is
provided in the Appendix.

The circuitry linking the prefrontal cortex with other brain regions provides a
neuroanatomical basis for functional localization in terms of circuits, rather than single
structures. Distinctions in circuitry have been noted for the dorsolateral, medial (i.e.,

anterior cingulate), and orbitofrontal regions of the prefrontal cortex (Middleton & Strick,

21

2002). There appears to be partial speciﬁcity within the frontal lobes relating these
different circuits to different behaviors (Fuster, 1997) as well as EF component processes
(Rezai, Andreasen, Alliger, Cohen et al., 1993).

The dorsolateral prefrontal cortex appears to be involved in the implementation of
control and active manipulation of information (i.e., working memory; Smith & J onides,
1999). Injury in this region is associated with a “frontal abulic syndrome” (Mesularn,
2002), which involves a disruption of “intensive and selective” attention (Fuster, 1997, p.
172) and is characterized by a loss of initiative and creativity, reduced ability to
concentrate, and a tendency towards emotional apathy and ﬂat affect. Medial regions
such as the anterior cingulate cortex are activated in conditions requiring performance
monitoring, response selection (i.e., resolution of cognitive conﬂict or interference
control), and the modulation of attention and motivation (Devinsky, Morrell, & Vogt,
1995; Kems, Cohen, MacDonald, Cho, Stenger, & Carter, 2004; Smith & Jonides, 1999;
Stuss, Floden, Alexander, Levine, & Katz, 2001). Too much anterior cingulate activity
may be associated with obsessive-compulsive and tic-like symptoms, while too little
activity has been related to diminished self-awareness, apathy, and depression (Devinsky
et al., 1995; Fuster, 1997). The dorsolateral and anterior cingulate regions appear to
provide complementary cognitive activities, with the dorsolateral cortex executing
control, and the cingulate cortex selecting, monitoring, and assessing performance (Kems
et al., 2004).

Damage to orbitoﬁ'ontal regions seems to spare most cognitive functions (Fuster,
1997), but they appear to be involved in using reward to guide actions as lateral regions

are likely to be activated for suppression of a previously rewarded response (Elliott,

22

Dolan, & Frith, 2000). Thus, along with more ventrolateral regions (Aron, Fletcher,
Bullmore, Sahakian, & Robbins, 2003; Aron, Robbins, & Poldrack, 2004; Konishi,
Nakajima, Uchida, Hideyuki, Karneyama, & Miyashita, 1999), the orbitofrontal cortex
may play a role in response inhibition tasks (Casey, Castellanos, Giedd, & Marsh, 1997),
which are important to many executive models. Injury here results in a “frontal
disinhibition syndrome” (Mesulam, 2002), which is characterized by deﬁcits in the
“exclusionary” aspect of attention (F uster, 1997, p. 174). Individuals with this disorder
have difﬁculty suppressing interference from external or internal stimuli. The result is
behavioral excesses with too much drive and impulsivity, lack of judgment or ability to
learn ﬁ'om experience, and disregard for social conventions.

This interpretation of functional localization in the prefrontal cortex provides
some intriguing and potentially heuristic hypotheses regarding psychiatric behavior
disorders. For instance, based on its symptom links with damage to the dorsolateral and
medial regions, depression may involve deﬁcits on tasks assessing perseverative
responding, planning, working memory, verbal ﬂuency, temporal organization of
behavior, and interference control. ADHD, with its surface similarity to symptoms that
follow damage to dorsolateral and orbitofrontal regions, may be associated with similar
impairments to those hypothesized for depression, with the addition of response
inhibition and excluding interference control due to a focus on orbitofrontal, not medial,
dysfunctions. Individuals with ASPD may have problems with response inhibition tasks
because of the parallels with orbitofrontal lobe damage. Alcohol and drug abuse

disorders do not ﬁt neatly into this theory, but their symptomatology may correspond to

the behavioral effects following orbitofrontal and dorsolateral lobe damage, and therefore

23

they may be associated with many of the same deﬁcits as ADHD. Finally, the anterior
cingulate regions may be important for anxiety disorders.

How have these sorts of suppositions panned out? In reality, there are varying
levels of support for differential componential deﬁcits from studies of EF and the
individual disorders (to be detailed below). Further complicating clarity on this issue is
the frequent comorbidity between psychiatric disorders (Kessler et al., 2005a; Kessler et
al., 2005b), which is often not addressed adequately in neuropsychological studies. This
made pertinent an examination of EF deﬁcits in psychopathology that attempted to
explicitly assess the contributions of individual as well as comorbid disorders to more
fully understand the relationship between EF and psychopathology. Based on the above-
mentioned neuroanatomy literature, there could be some partial differentiation of EF
component processes mapping onto disorders. However, given that this was not a
straightforward association, other models were considered to explain EF and
psychopathology.

Following is a review of recent studies done on EF deﬁcits and brain function for
the individual disorders that were included in the present study. This overview will
illustrate the complexities and similarities in the literature across studies. The effects of
comorbidity will also be noted where relevant. Thereafter, the issue of comorbidity will
be addressed more directly.

EF in Key Psychopathologies

As mentioned, because EF are potentially relevant to so many disorders, some

decisions had to be made regarding which disorders to include in the present study.

Selection criteria included prevalence of the disorders in the general population,

24

frequency of comorbid presentations, and similarities in symptomatology. Further, all of
the disorders that were selected for inclusion were associated with EF deﬁcits in some
manner, and until now their frequent co-occurrence (Kessler et al., 2005a; Kessler et al.,
2005b; Kessler et al., 1994) had made it difﬁcult to isolate their individual
neuropsychological correlates. Thus, the disorders included are: childhood externalizing
disorders (i.e., ADHD, CD, and ODD), antisocial personality disorder (ASPD),
alcoholism, drug dependence, major depressive disorder (MDD), and anxiety disorders.
Selectivity was required in view of the size of some of these literatures. Therefore, recent
reviews, meta-analyses, and major studies conducted in the past ten years are
emphasized. Further, clinical neuropsychological measures are emphasized, with less
focus placed upon experimental studies. A table is presented in Table 1 of Appendix B to
summarize the main patterns of EF deﬁcits across disorders and regional involvement
that are suggested by the following reviews.

ADHD and Externalizing Disorders Typically Diagnosed in Childhood (Manifestations
in Adulthood)

Many articles have been published on neuropsychological functions and ADHD,
with a particular emphasis on EF task performance because of the frequent behavioral
comparisons between individuals with ADHD and those with ﬁ'ontal lobe damage
(Pennington & Ozonoff, 1996). The vast literature has covered the performance of
children, adolescents, and, more recently, adults on a very wide range of cognitive tests
with widely varying results. Recent reviews and meta-analyses have attempted to
synthesize the vast literature (Berlin, 2003; Sergeant et al., 2002; Willcutt, Doyle, Nigg,

Faraone, & Pennington, 2005). More recently, attention has turned towards adult ADHD,

25

and this summary will focus upon these studies as the present examination included
adults. Numerous articles also suggest there are associations between EF and disruptive
behavior disorders diagnosed in childhood, such as oppositional deﬁant disorder (ODD)
and conduct disorder (CD; Pennington & Ozonoff, 1996; Willcutt et al., 2005). These
disorders are frequently comorbid with childhood ADHD (Kuhne, Schachar, & Tannock,
1997), and shared genetic risk factors for these disorders and EF deﬁcits have been
suggested (Coolidge, Thede, & Young, 2000). However, the relationship between
childhood diagnoses of CD and ODD and adult neuropsychological functioning has not
been directly studied. Preliminary information is provided by studies of EF and
antisocial personality disorder, a possible adult manifestation of CD, which is reviewed
below. Therefore, this section will focus upon EF deﬁcits in adult ADHD.
Neuropsychological test performance. Adults with ADHD have shown deﬁcits
on a wide range of tasks assessing EF processes. Impairments have been seen on tasks
assessing response inhibition, cognitive ﬂexibility, set-shifting and maintenance, verbal
ﬂuency, working memory, planning, decision making, and strategy formation (Johnson,
Epstein, Waid, Latharn, Voronin, & Anton, 2001; Murphy, Barkley, & Bush, 2001; Nigg
et al., 2005; Schweitzer, Faber, Grafton, Tune, Hoffman, & Kilts, 2000; Seidman,
Biederman, Weber, Hatch, & Faraone, 1998; Walker, Shores, Trollor, Lee, & Sachdev,
2000). Other studies have found that individuals with ADHD have performed normally
on EF tasks which assessed interference control, working memory, and planning
(Holdnack, Moberg, Arnold, Gur, et al., 1995; Riccio, Wolfe, Davis, Romine, George, &

Lee, 2005; Riccio, Wolfe, Romine, Davis, & Sullivan, 2004; Weyandt et al., 1998), but

26

often the sample size was small and power was low. Thus, many EF deﬁcits have been
reported.

Similarly, in their review of the adult ADHD literature, Woods, Lovejoy, & Ball
(2002) found that compared to normal controls, adults with ADHD showed selective
deﬁcits in divided and sustained attention, verbal ﬂuency, auditory verbal list learning,
planning/organization, behavioral inhibition/impulsivity, cognitive ﬂexibility, and speed
of information processing. Most of these impairments involve EF abilities. Based upon
their ﬁndings for the neuropsychological proﬁle of adults with ADHD, Woods et a1.
(2002) concluded that the disorder represented a disruption of the dorsolateral prefrontal
cortex, and to a lesser extent the orbitofrontal cortex. Further, they (Woods et al., 2002)
suggested that ADHD adults could not be differentiated from individuals with other
disorders using neuropsychological test performance (Riccio et al., 2005; Riccio et al.,
2004; Walker et al., 2000; Weyandt et al., 1998). However, poor controls for subclinical
ADHD symptoms in these studies may have contributed to these ﬁndings. Therefore, EF
test results supported dorsolateral prefrontal, and possibly orbitofrontal, dysfunction.
Reduced sensitivitity to discriminate individuals with ADHD from other psychiatric
disorders may be partially related to poor controls for subclinical symptomatology.

Recent meta-analyses are particularly helpful in the present context because they
provide effect sizes for several widely used tasks that assess different EF processes. The
most relevant analyses to the present study, due to their focus on adults with ADHD,
were done by Hervey, Epstein, & Curry (2004) and Boonstra, Oosterlaan, Sergeant, &
Buitelaar (2005). An examination of different EF processes shows that certain tasks are

more likely to show group differences than others within each domain. Problems with

27

response inhibition have been considered central to ADHD in children (Pennington &
Ozonoff, 1996), and the largest effect size for tasks in this EF domain was found for the
Stop Signal Reaction Time (SSRT) variable (d = 0.85), a measure of response inhibition
ﬁ'om the Stop Signal Task (Hervey et al., 2004). This result is in accordance with the
ﬁndings from child cognitive studies, as group differences are most consistently seen for
SSRT in EF tasks (W illcutt et al., 2005) and group differences were recently found in the
largest study to date of adults with ADHD (N igg et al., 2005).

Such strong ﬁndings are not always seen for tasks that include some response
inhibition demands. Despite a small to large effect sizes for speed of information
processing on the color, word, and color-word trials of the Stroop task (d = 0.30, 0.23,
and 0.47, respectively for Hervey et al., 2004; d = 0.62, 0.60, and 0.89, respectively for
Boonstra et al., 2005) the effect size for Stroop Interference, which is primarily a measure
of interference control that involves some response inhibition, was almost negligible in
ADHD adults (d = 0.15 for Hervey et al., 2004; d = 0.13 for Boonsta et al., 2005). This
suggests that adults with ADHD may have slower processing speed times, consistent with
results from individual studies (Corbett & Stanczak, 1999; Johnson etal., 2001; Nigg et
al., 2005; Walker et al., 2000), but do not experience additional impairment from the
interference control demands of the task. However, this has, perhaps somewhat
misleadingly, been interpreted as support for deﬁcits simply in response inhibition
(Boonstra et al., 2005). Although a large study supported interference deﬁcits in adults
with ADHD (Murphy et al., 2001), most studies have generally suggested that Stroop
interference is not performed more poorly by adults with ADHD (Johnson et al., 2001;

Lovejoy, Ball, Keats, Stutts, Spain, J anda et al., 1999; Nigg et al., 2005; Riccio et al.,

28

2005; Walker et al., 2000). It is possible that the motor requirements in the Stop task
may have contributed to deﬁcits in the ADHD adults, as adults with ADHD have shown
deﬁcits on tasks assessing motor but not cognitive inhibition (N igg, Butler, Huang-
Pollock, & Henderson, 2002), and increased motor task demands contribute to deﬁcits in
performance across tasks with ADHD adults (Hervey et al., 2004). Therefore, ADHD
adults do not appear to have deﬁcits in cognitive interference control but do for more
motoric response inhibition tasks.

As for other aspects of EF, adults with ADHD showed moderate to large effect
sizes for tasks that assessed working memory (d = 0.44 - 0.83), verbal ﬂuency (d = 0.62 -
0.63), and planning (d = 1.09; Boonstra et al., 2005; Hervey et al., 2004). However, in
individual studies with smaller sample sizes these results were not always sustained
(Riccio et al., 2005; Riccio et al., 2004; Weyandt et al., 1998). With regards to set-
shifting abilities, adults with ADHD performed more poorly on some tasks in this domain
(i.e., Trail Making Test B; d = 0.65 - 0.68), but tended not to on others (i.e., Wisconsin
Card Sorting Task (WCST); d = 0.02 - 0.12; Boonstra et al., 2005; Hervey et al., 2004).
A similar pattern was seen in another meta-analytic review which included studies of
children, adolescents, and adults (Frazier, Demaree, & Youngstrom, 2004). Deﬁcits on
the WCST are not consistently observed in adults with ADHD (Johnson et al., 2001;
Riccio et al., 2005; Riccio et al., 2004), although they have been demonstrated in a large
study (N i gg et al., 2005). Trail Making Test B requires motor control and cognitive set-
shifting, while WCST requires working memory and conceptual-level responding in
addition to set-shifting. Therefore, similar to the results for response inhibition, it

appears as though the motor and time pressure demands for the Trail Making Test may

29

have contributed to increased impairments in the ADHD group. Hervey et a1. (2004)
noted that across domains, ADHD adults’ performance tended to deviate more and more
from controls as task demands such as motor, time, and complexity increased. Therefore,
although both WCST and Trail Making Test B are arguably set-shifting tasks, the
additional demands created by the time pressure and greater motor demand of the Trail
Making Test B appear to make it more sensitive to ADHD adults’ set-shifting
impairments.

The largest studies done to date suggest that there are deﬁcits across a wide range
of EF abilities such as ﬂuency, working memory, set-shifting, and response inhibition
(Murphy et al., 2001; Nigg et al., 2005). Therefore, some of the discrepant ﬁndings for
these processes may have resulted from small sample sizes in the individual studies.
Overall, tasks that appear to tap orbitoﬁontal and dorsolateral processes and which
involve more demands on speed, motor control, and complexity are better able to
differentiate between ADHD adults and healthy controls.

Eﬂects of comorbidity on neuropsychological test performance. High rates of
antisocial behavior, substance abuse, anxiety, and mood disorders are seen in adults with
ADHD (Biederman, F araone, Spencer, Wilens et al., 1993; Biederman, Wilens, Spencer,
Faraone et al., 1996). Individuals with comorbid disorders are often excluded from
studies of adults with ADHD. When this constraint is not imposed, high rates of the co-
occuning disorders are commonly reported, but not necessarily controlled, for adults with
ADHD (Corbett & Stanczak, 1999; Riccio et al., 2005; Riccio et al., 2004). When
neuropsychological test results were statistically adjusted for the presence of comorbid

conditions, ADHD deﬁcits remained robust in adult studies (N igg et al., 2005; Seidman

30

et al., 1998) and in a meta-analytic review of pediatric/adolescent studies (Willcutt et al.,
2005). While the presence of a comorbid Axis I disorder has resulted in elevated EF
deﬁcits (Downey, Stelson, Pomerleau, & Giordani, 1997), this is not a consistent ﬁnding
(Murphy et al., 2001), but this latter negative ﬁnding may have been due to low statistical
power after splitting a continuous variable to deﬁne comorbid depression in adults with
ADHD (Hervey et al., 2004). Overall, comorbidity effects have not been able to explain
ADHD deﬁcits, yet have not been fully mapped.

Neuroimaging studies. A recent review of neuroimaging studies in ADHD
suggested that abnormalities in structure and function have been found in the lateral
prefrontal cortex, dorsal anterior cingulate cortex, basal ganglia (expecially caudate),
corpus callosum, and cerebellum (Seidman, Valera, & Bush, 2004). An early study
suggested that ADHD adults had lower preﬁ'ontal activity in response to a cognitive
challenge compared to controls (Zametkin, Nordahl, Gross, King et al., 1990).
Differences in neural activation were seen with a working memory task, as adults with
ADHD had more diffuse activation and lacked task-related frontal activation compared to
controls (Schweitzer et al., 2000). Adolescents with ADHD showed reduced activation
of the right inferior prefrontal cortex during successful response inhibition on the Stop
Signal Task (Rubia, Smith, Brammer, Toone, & Taylor, 2005). Thus, abnormalities in
the fronto-striatal circuits appear to be relevant to the cognitive deﬁcits seen in ADHD
(Seidman et al., 2004).

Taken together, the results of neuropsychological testing and neuroimaging
implicate the involvement of fronto-striatal circuits in ADHD. Individual EF tasks show

differential sensitivity to impairments in cognitive performance, but the pattern of deﬁcits

31

suggest that the inferior/orbitofrontal and dorsolateral prefrontal cortex and associated
subcortical circuitry may be particularly important for the deﬁcits seen in ADHD.
Antisocial personality disorder

Antisocial personality disorder (ASPD), an adult version of CD, has received a
great deal of attention in the neuropsychological and neurological literature as researchers
have attempted to understand the cognitive, biological, and social mechanisms that
contribute to its development. However, inconsistencies in the deﬁnition and
operationalization of antisocial behavior have contributed to lack of cohesion in the
literature (Morgan & Lilienfeld, 2000). Three main approaches have dominated the
literature: (1) the clinical diagnosis of ASPD, deﬁned by the DSM-IV as a pervasive
pattern of disregarding and violating the rights of others with a childhood history of CD
(American Psychiatric Association, 1994); (2) the personality dimension of psychopathy,
which involves a constellation of personality traits such as lack of remorse or sincerity
and impoverished emotional reactions, and is assessed using multiple methods (Cleckley,
1941; Hare, 1996); and (3) the legal concepts of criminality and delinquency. These
deﬁnitions of antisocial behavior show considerable overlap and intercorrelations, but
they are distinct (Morgan & Lilienfeld, 2000). In the interest of consistency, the present
paper will focus on clinically-based DSM deﬁnitions of ASPD, but other deﬁnitions of
antisocial behavior will be covered when they help to elucidate the literature.

Neuropsychological test performance and eﬂects of comorbidity. A fairly recent
meta-analysis (Morgan & Lilienfeld, 2000) included studies using all of these
operationalizations of antisocial behavior and examined effect size differences between

individuals with antisocial behavior and controls on six widely used measures of EF:

32

Category Test, Porteus Mazes Test, Stroop Interference Test, Trailmaking Test Part B,
WCST perseverative errors, and verbal ﬂuency tests. EF processes that are assessed with
these tasks include set-shifting and maintenance, working memory, interference control,
cognitive ﬂexibility, and response inhibition. Across studies and different antisocial
behavior deﬁnitions, there was a medium weighted mean effect size (d = .62); the effect
sizes for criminality and delinquency were in the large range (d = 1.09 and .86,
respectively), CD and psychopathy were in the small to medium range (d = .40 and .29,
respectively), and ASPD (DSM-III or DSM-IIIR) had negligible but statistically
signiﬁcant effect sizes (d = .10) (Morgan & Lilienfeld, 2000). It should be noted that
very few studies have been conducted using DSM deﬁnitions of ASPD. Nonetheless,
these ﬁndings suggest a weak relationship between EF deﬁcits and ASPD.

Studies done since this meta-analysis have corroborated Morgan and Lilienfeld’s
(2000) ﬁndings that individuals with ASPD may not perform more poorly than controls
on typical tests of EF. Individuals with ASPD did not perform more poorly than normal
(Barkataki, Kumari, Das, Hill, Morris, O'Connell et al., 2005; Stevens, Kaplan, &
Hesselbrock, 2003) or psychiatric controls (Crowell, Kieffer, Kugeares, & Vanderploeg,
2003) on the WCST, a measure of concept formation and working memory. Similar null
effects were seen for tasks assessing set-shifting, planning, and interference control
(Barkataki et al., 2005; Stevens et al., 2003), although this is not a consistent ﬁnding
(Dolan & Park, 2002). Moreover, individuals with ASPD plus the personality dimension
of psychopathy have shown deﬁcits in performance on these processes (Dinn & Harris,

2000). It may be that individuals with ASPD are more likely to show deﬁcits in EF

33

performance when they have other comorbid psychopathology, as this was also found for
substance use disorder (Malloy, Noel, Longabaugh, & Beattie, 1990).

Unlike the null ﬁndings seen with most EF processes, males with ASPD alone
have shown differences from controls on response inhibition tasks (Dinn & Harris, 2000;
Dolan & Park, 2002; note that Dinn & Harris only approached signiﬁcance). These
results suggest that ASPD alone may not be associated with deﬁcits on tasks assessing
typical EF processes such as set-shifting, working memory, and interference control, but
do show deﬁcits on response inhibition. Response inhibition processes appear to rely
upon different neural networks in the prefrontal cortex (Aron et al., 2004), suggesting that
ASPD may primarily be associated with dysfunction in lateral-ventral (i.e., orbitoﬁontal)
regions of the prefrontal cortex. Similar discrepancies in ﬁndings with psychopaths
(Gorenstein, 1982; Hart, Forth, & Hare, 1990; Lapierre, Braun, & Hodgins, 1995; Pham,
Vanderstukken, Philippot, & Vanderlinden, 2003; Sutker & Allain, 1987) also provide
support for this regional involvement.

Neuroimaging and lesion studies. Brain lesion studies show parallels between
orbitofrontal lobe damage and antisocial behavior disorders. Cases of “acquired
sociopathy” have been reported following damage to the orbitofrontal region of the brain
(Blair & Cipolotti, 2000), and the resultant syndrome is characterized by personality
changes and cognitive deﬁcits on inhibition tasks (Fuster, 1997; Malloy, Bihrle, Duffy, &
Cimino, 1993), similar to those seen in individuals with ASPD (Dinn & Harris, 2000;
Dolan & Park, 2002). MRI has revealed reduced prefrontal gray matter volume in

uninstitutionalized men with ASPD, suggesting evidence of a structural brain deﬁcit in

34

that region of the brain (Raine, Lencz, Bihrle, LaCasse, & Colletti, 2000). Such ﬁndings
provide support for cerebral dysfunction being involved in the disorder.

However, when regions of the prefrontal cortex were assessed in adults with
ASPD and alcoholism, Laakso and colleagues (2002) found that only duration of alcohol
use and education predicted their volume deﬁcits. This suggests that comorbid pathology
may be important in brain dysﬁmctions and cognitive problems. Further supporting this
idea, while ASPD has been noted as an outcome for a subset of children with ADHD,
perhaps mediated by deﬁcits in EF (McKay & Halperin, 2001), antisocial behaviors in
adolescents with ADHD do not individually contribute to cognitive deﬁcits seen on any
typical EF processes (Déry, Toupin, Pauzé, Mercier, & Fortin, 1999). Therefore, ASPD
appears to be associated with certain speciﬁc cognitive deﬁcits that may partly involve
dysfunctions in the orbitofrontal prefrontal cortex. However, more general EF deﬁcits
may not been seen except in association with correlates of ASPD such as comorbid
disorders.

Taken together, these results suggested that individuals with ASPD would not
perform poorly on some of the more common tasks assessing major EF processes. More
general deﬁcits in EF may be seen when there are comorbid disorders such as ADHD and
alcoholism. Behavioral parallels exist between damage to the orbitofrontal cortex and the
symptomatology of ASPD, and response inhibition deﬁcits in individuals with ASPD
further support the possible involvement of this region of the brain. However, it may be
difﬁcult (or nearly impossible) to ﬁrlly separate ASPD from comorbid disorders such as

alcoholism and ADHD in order to determine their individual contributions to

35

neuropsychological impairment due to the strong overlap in symptomatology between
these disorders.

Alcoholism

Neuropsychological test performance. Alcohol dependence has been associated
with deﬁcits in a wide range of EF processes. These deﬁcits are so frequently observed
that a ‘frontal hypothesis’ has been proposed. As a possible explanation for the cognitive
impairments that follow chronic alcohol abuse, this hypothesis suggests that the frontal
lobes are particularly susceptible to alcohol-related damage (Ratti, Bo, Giardini, &
Soragna, 2002). Chronic alcoholics have performed more poorly than controls on tests
assessing working memory, cognitive set-shifting, abstract reasoning, ﬂuency tests,
response inhibition, and interference control (Adams, Gilman, Koeppe, Kluin et al., 1993;
Brokate, Hildebrandt, Eling, Fichtner, Runge, & Tim, 2003; Dao-Castellana, Samson,
Legault, Martinot, Aubin, Crouzel et al., 1998; Hoffman, Hall, & Bartsch, 1987; Poon,
Puttler, Zucker, Nigg, & Fitzgerald, 1999; Ratti et al., 2002; Uekerrnann et al., 2003).
Thus, deﬁcits have been found across a wide range of individual EF processes (Ratti et
al., 2002).

To increase power, many studies have examined combined or composite scores
from tests assessing general cognitive domains. When using this type of an analysis,
adults with alcohol dependence have shown deﬁcits on EF factor scores which included
processes such as working memory, set-shifting, conceptual reasoning, and interference
control (Goldstein, Leskovjan, Hoff, Hitzemann, Bashan, Khalsa et al., 2004b; Selby &

Azrin, 1998; Sullivan, Fama, Rosenbloom, & Pfefferbaum, 2002). It should be noted that

when results for individual tests were analyzed in these studies, few differences were

36

found in task performance, which suggested the need for a composite analysis. It may be
that the impairment was mild in these studies, and could only be detected through
analysis of the EF construct as a whole and not in component parts. Different levels of
severity for alcohol abuse/dependence may have resulted in weaker ﬁndings in some
studies versus others. However, although there appears to be some decreases in deﬁcits
with long-term abstinence (Selby & Azrin, 1998), EF impairment persists (Munro,
Saxton, & Butters, 2000). Therefore, similar to ADHD, EF deﬁcits are often noted in
adults with alcohol dependence, even with remittance of the disorder, although the
speciﬁc level and type of impairment tends to vary across studies.

Eﬂects of comorbidity and genetic liability on neuropsychological test
performance. Again, the issue of comorbidity is important for research on alcoholism.
In a recent study on the neuropsychological performance of 43 alcoholic women, only 1 l
were free of other lifetime Axis I comorbidities (Sullivan et al., 2002). Nine of the
comorbid women met criteria for one other disorder, while the remainder met criteria for
two or more lifetime Axis I disorders; depression and drug abuse were most frequently
diagnosed. Comorbid disorders were not taken into account in their analyses, however,
so it can be questioned whether their ﬁnding that alcoholic women performed more
poorly on an EF composite was due to alcohol dependence or other comorbid disorder.
Uekermann and colleagues (2003) examined the potential effect of depression upon
neuropsychological performance in alcoholics. While adult alcoholics performed more
poorly on EF measures than controls, there were no differences in comparison to
individuals with alcoholism and depression (Uekermann et al., 2003). In males, ASPD is

frequently comorbid with alcoholism (Kessler et al., 1997), and the association and

37

interaction between these two disorders has been a focus as a possible second type of
alcoholism (Zucker, Ellis, Bingham, & Fitzgerald, 1996; Zucker, Ellis, & Fitzgerald,
1992; Zucker, Ellis, Fitzgerald, & Bingham, 1996). In neuropsychological studies, it has
been shown that alcoholic individuals with ASPD perform more poorly on
neuropsychological tasks than pure alcoholics (Giancola & Moss, 1998; Malloy, Noel,
Rogers, Longabaugh et al., 1989), although this is not a consistent ﬁnding (Hoffman et
al., 1987; Sutker & Allain, 1987). Thus, results vary with regard to the effect that
comorbid disorders have on the cognitive performance of adults with alcoholism. ASPD
has been the most heavily studied disorder in this regard. Although results are not
consistent, they suggest that the presence of a comorbid disorder is likely to be important.

Further complicating the ﬁndings on alcoholism is a related line of research
suggesting that EF deﬁcits may actually precede alcohol use and that these deﬁcits are
risk factors for the development of problem drinking (N igg, Glass, Wong, Poon, Jester,
Fitzgerald et al., 2004). Alcoholism has been shown to be heritable (Knopik, Heath,
Madden, Bucholz, Slutske, Nelson et al., 2004), and a number of recent studies have used
the high-risk paradigm to examine neuropsychological functioning in the children of
alcoholic parents. Findings indicate that the sons of alcoholics show greater EF deﬁcits
compared to the offspring of non-alcoholic control parents (Pihl, Peterson, & Firm, 1990).
A family history of alcoholism has been associated with deﬁcits in EF abilities such as
response inhibition (N igg et al., 2004), while inconsistent results have been obtained for
the presence of conceptual set-shifting deﬁcits in children of alcoholics (Corral, Holguin,
& Cadaveira, 2003; Corral, Holguin, & Cadaveira, 1999). The presence of familial

comorbidity, particularly ASPD, may also contribute to certain cognitive deﬁcits. Family

38

history of both alcoholism and ASPD appears to be associated with a different set of
cognitive deﬁcits than those seen with alcoholism alone, as they involve general
intelligence and reward-response (N i gg et al., 2004). Stevens and colleagues (2003)
found that planning errors were increased in individuals with a family history of
alcoholism, but there was an interaction effect between family history of alcoholism and
ASPD, which resulted in difﬁculties inhibiting prepotent motor responses as well.

These precursor EF deﬁcits appear to increase risk for the development of alcohol
problems (Giancola, Shoal, & Mezzich, 2001; Nigg, Wong, Martel, Jester, Puttler, Glass
et al., 2006). Although it has been suggested that this relationship is fully mediated by
comorbid antisocial behavior (Giancola et al., 2001), such ﬁndings may only hold for
general substance disorders as opposed to alcoholism per se. For instance, even after
controlling for child externalizing behaviors, response inhibition deﬁcits in children
predicted later alcohol-related behaviors and problems (N igg et al., 2006). Therefore,
family history of alcoholism, or alcoholism and ASPD, appears to be associated with EF
deﬁcits which may place adolescents at increased risk for developing alcohol problems
themselves.

Neuroimaging studies. Neuroimaging research has supported ﬁ'ontal lobe
differences in alcoholism. While chronic alcoholics have lower global rates of
metabolism (Volkow, Hitzemann, Wang, Fowler et al., 1992), selective hypometabolism
has also been found in the frontal lobes (Volkow et al., 1992), and a post-mortem study
showed that alcoholics have fewer neurons in the frontal lobes compared with non-
alcoholic controls (Harper & Kril, 1985). A large percentage of recovered alcoholics,

studied after withdrawal symptoms had remitted, showed hypoperfusion of the frontal

39

lobes, with reduced rCBF being related to duration of drinking and presence of comorbid
antisocial personality disorder (Kuruoglu, Arikan, Vural, Karatas et al., 1996). Some
ﬁndings suggest that the frontal dysfunction is isolated to the medioﬁ'ontal region as
selective hypometabolism has been found in the medial part of the frontal lobes, which
includes the anterior cingulate gyrus (Adams et al., 1993; Dao-Castellana et al., 1998).
Dao-Castellana and colleagues (1998) found that healthy, neurologically intact adult
alcoholics primarily had hypometabolism in the medioﬁ'ontal region (including the
anterior cingulate gyrus), as well as dysfunction in the left dorsolateral preﬁ'ontal cortex,
but not the orbitoﬁ'ontal cortex (Dao-Castellana et al., 1998). Others have also found
differences in metabolism in the anterior cingulate cortex and dorsolateral preﬁ'ontal
cortex (Goldstein et al., 2004b). These ﬁndings had neurobehavioral implications as
hypometabolism in the mediofrontal lobes was correlated with impaired performance on
verbal ﬂuency and slower speed on the Stroop color-word condition, while the left
dorsolateral hypometabolism was related to increased errors on the Stroop test (Dao-
Castellana et al., 1998). Thus, neuropsychological deﬁcits are correlated with metabolic
abnormalities in the frontal lobes of chronic alcoholics. However, with time, frontal
brain abnormalities may subside following abstinence (Gansler, Harris, Oscar-Herman,
Streeter, Lewis, Ahmed et al., 2000).

Taken together, these results suggest that chronic alcohol dependence is
associated with deﬁcits in EF, but EF may be both a contributor and an outcome. First,
the consistency of these impairments, in conjunction with neuroimaging ﬁndings, has
prompted researchers to suggest a frontal lobe hypothesis, involving the dorsolateral,

medial, and orbitofrontal regions (the latter being based on response inhibition deﬁcits),

40

to explain some of the cognitive changes following heavy alcohol abuse. Second,
however, EF deﬁcits are associated with a family history of alcoholism, and therefore the
cognitive impairments may actually precede and increase risk for the onset of alcohol use
disorders. Finally, comorbid disorders such as ASPD may contribute to the cognitive
proﬁle by amplifying the deﬁcits associated with alcohol use as well as the familial risk
outcomes. Therefore, it appears as though there are multiple routes by which alcohol can
contribute to EF impairments. Each of these pathways - family history of
alcoholism/ASPD, chronic use, and antisocial comorbidity - may interact to produce the
observed deﬁcits on neuropsychological tests of EF.
Drug Dependence

Studies on drug use tend to include different deﬁnitions of substance use and
methods of diagnosis. A recent review included both abuse and dependence and noted
that there is a general paucity of studies on EF and drug use (Lundqvist, 2005). Based on
this, the present review will include ﬁndings for heavy abuse and dependence, although
the focus is upon dependence. Despite the inherent difﬁculties in the analysis, some
studies have been able to isolate the cognitive proﬁles for abuse of different drugs. Data
collected from a large epidemiological study suggested that marijuana and cocaine were
the two most commonly abused drugs (Petry, Stinson, & Grant, 2005), which has also
been seen in treatment populations (Miller, Klarnen, Hoffmann, & Flaherty, 1996).
Therefore, my review will focus on these two drugs.

Neuropsychological test performance. Generally mild and inconsistent
neurocognitive deﬁcits have been seen in groups with substance disorders. Chronic

cocaine users showed less severe deﬁcits on an EF composite than individuals with

41

alcoholism, although no differences were seen on the individual tests (Goldstein et al.,
2004b). Similar ﬁndings were seen in another study that attempted to isolate the speciﬁc
deﬁcits associated with cocaine abuse and dependence (Selby & Azrin, 1998). Cocaine
users did not perform differently ﬁom controls on tests assessing neuropsychological
functioning, while poly-substance and alcohol abuse/dependence were associated with a
number of cognitive impairments, including deﬁcits in set-shifting, interference control,
and verbal ﬂuency. A recent meta-analysis on the cognitive effects of cocaine abuse
(J ovanovski, Erb, & Zakzanis, 2005) found that the median effect size for a wide variety
of tests was 0.35. However, a number of EF tests had effect sizes greater than this
median effect; large effect sizes were seen for a general EF factor score, and a conceptual
reasoning test, while medium effect sizes were obtained for an interference control test
(Stroop interference), and set-shifting test (Trail Making Test, B), and small for another
set-shitting test that involved working memory (W CST). This suggested that chronic
cocaine users do show some mild dose-related (Bolla, Rothman, & Cadet, 1999)
neurocognitive deﬁcits, particularly in conceptual reasoning, psychomotor set-shifting,
and interference control. Differences in inclusion criteria (i.e., abuse versus dependence)
which affect amount and frequency of drug use, as well as small sample sizes, likely
contributed to a weak ability to detect differences between drug users and healthy
controls in individual studies.

Marijuana is also a commonly abused substance, but few studies have examined
its relationship with HF. Heavy marijuana users had more perseverative errors than light
users on the WCST, a test of working memory and set-shifting, following at least one day

of abstinence (Pope & Yurgelun-Todd, 1996). These deﬁcits appear to continue through

42

even longer periods of abstinence. Bolla and colleagues (2002) found that there were
performance decrements on tests of memory and EF processes, such as set-shifting and
interference control, in long-term users following a month of abstinence. Such deﬁcits
are not consistently seen (Eldreth, Matochik, Cadet, & Bolla, 2004; Gruber & Yurgelun-
Todd, 2005), but similar to cocaine there appears to be a dose-related effect of marijuana
use, with amount, not duration, being related to cognitive impairment (Bolla et al., 2002;
Bolla et al., 1999). Thus, dose-related EF deﬁcits are sometimes seen with marijuana
use; although deﬁcits are inconsistent, they appear to remain at least following short-term
periods of abstinence.

Ejfects of comorbidity on neuropsychological test performance. Few studies of
the neuropsychological effects of substance use have clear implications for the individual
deﬁcits associated with drug dependence, as it is difﬁcult to isolate participants who do
not have comorbid substance (i.e., alcoholism or other drugs) or other disorders. The
importance of these factors was evident from a recent latent variable analysis of cognitive
performance in drug-abusing patients (Pals-Stewart & Bates, 2003). It was shown that
alcohol use and number of current substance dependence diagnoses were important
contributors to current EF abilities, along with education and IQ. Further, these deﬁcits
have been proposed to precede and, along with family history of substance use, increase
risk for drug use (Giancola & Tarter, 1999). These ﬁndings suggest that comorbid
disorders contribute considerably to the cognitive deﬁcits seen in substance use disorders,
and make it difﬁcult to interpret any individual contributions.

Neuroimaging studies. Frontal lobe functioning and associated cognitive deﬁcits

have been related to the cycle of addiction. A recent review suggested that the

43

orbitofrontal and anterior cingulate circuits are dysfunctional in individuals with
addictions, thus contributing to lack of inhibitory control over reward-related behavior
(Lubman, Yiicel, & Pantelis, 2004). Others have also noted that these regions are most
consistently implicated in the stages of addiction such as intoxication, craving, and
binging, and they are deactivated during withdrawal (Goldstein & Volkow, 2002;
Volkow, Ding, Fowler, & Wang, 1996).

These ﬁndings have implications for functional imaging. Frequent cocaine users
had reduced activity in the anterior cingulate and right prefrontal cortex while completing
a response inhibition task which involved working memory demands (Hester & Garavan,
2004). Structural changes are also evident in prefrontal regions; decreased grey matter
was found for cocaine-dependent adults in the orbitoﬁontal and anterior cingulate regions
of the brain, along with the insula and some areas in the temporal lobe (Franklin, Acton,
Maldjian, Gray, Croft, Dackis et al., 2002). Frequent marijuana users also show
functional abnormalities in prefrontal activity. On an interference control task, the
Stroop, marijuana smokers demonstrated a different pattern of anterior cingulate activity
(as well as more diffuse dorsolateral activity) than control participants, although no
differences were noted in task performance between the groups (Gruber & Yurgelun-
Todd, 2005). These dysfunctions are still seen even following a 25-day period of
abstinence, as Eldreth and colleagues (2004) also found that chronic marijuana users
showed a pattern of hypoactivity in the left anterior cingulate and lateral prefrontal
cortex, and hyperactivity in the hippocampus even though they did not perform
differently from controls on a modiﬁed version of the Stroop interference task. These

results suggest that there are persistent metabolic changes which affect the prefrontal

44

regions of the brain (Eldreth et al., 2004). At rest, frequent marijuana users also have
hypoactivity in ventral regions of the brain following approximately 26 hours of
abstinence (Block, O'Leary, Hichwa, Augustinack, Ponto, Ghoneim et al., 2000), but no
structural changes are noted in the brain (Block, O'Leary, Ehrhardt, Augustinack,
Ghoneim, Arndt et al., 2000). Therefore, marijuana and cocaine use are associated with
functional and structural dysﬁmctions (the latter only seen for cocaine) in the prefrontal
regions of the brain. In particular, the anterior cingulate and orbitoﬁontal regions appear
to be involved in the cycle of addiction.

Taken together, mild EF deﬁcits have been observed with heavy marijuana and
cocaine use. These cognitive impairments appear to be dose-related and are related to
dysfunctions in prefrontal and associated subcortical regions of the brain, particularly the
orbitofrontal and anterior cingulate circuits. The pattern of deﬁcits suggests that
comorbid disorders, polysubstance use, familial risk, and preexisting EF deﬁcits may be
important to EF impairment. The similarities in the EF deﬁcits across the reviewed drugs
suggest that substance dependence disorders should be combined to form one group in
the present analyses (i.e., drug dependence). However, alcoholism is associated with
more consistent and speciﬁc effects, and thus was analyzed separately.

Depression

Neuropsychological test performance. Unipolar depression has been associated
with a range of neuropsychological deﬁcits in EF. Research has found that patients have
shown consistent deﬁcits on EF tasks during episodes of major depressive disorder.
Multiple studies have found impaired set-shifting and maintenance, working memory,

interference control, perseverative control, response inhibition, and concept formation

45

(Austin, Mitchell, Wilhehn, Parker, Hickie, Brodaty et al., 1999; Austin, Ross, Murray,
O'Carroll, et al., 1992; Channon, 1996; Fossati et al., 1999; George, Ketter, Parekh,
Rosinsky et al., 1997b; Kaiser, Unger, Kiefer, Markela, Mundt, & Weisbrod, 2003;
Merriam, Thase, Haas, Keshavan, & Sweeney, 1999; Moritz et al., 2002; Paradise,
Larnberty, Garvey, & Robinson, 1997; Trichard, Martinot, Alagille, Masure et al., 1995).
Severity of depressive symptomatology has also been correlated with performance on the
WCST, a task which involves set-shifting, concept formation, and working memory
(Merriam et al., 1999), and number of episodes and hospitalizations were related to
greater cognitive impairment (Purcell, Maruff, Kyrios, & Pantelis, 1998). Chronicity of
the disorder may be particularly important to cognitive deﬁcits (Richard et al., 2003).
Therefore, there appears to be a strong link between depressive symptoms and EF
impairment.

It is sensible that being depressed would weaken cognitive efﬁciency and problem
solving; however, such deﬁcits may last even after the disorder has remitted. Non-
syrnptomatic males with a chronic history of unipolar depression performed more poorly
than controls on tasks assessing set-shifting, working memory, and interference control
(Paradise et al., 1997). Again, severity appears to exacerbate this effect. Frequency of
affective episodes appears to be a factor in longer term deﬁcits, as patients with recurrent
affective episodes, as opposed to a single depressive episode, showed general cognitive
deﬁcits during the euthymic phase (Kessing, 1998). However, patients with recurrent
affective episodes also had higher levels of subclinical depression, which correlated with
some cognitive test results, suggesting that residual symptomatology may account for

some of the enduring neuropsychological deﬁcits.

46

Neuroimaging studies. Neuroimaging studies also support frontal lobe
dysfunctions in depressed patients. Unipolar depressed patients show reduced glucose
metabolism in the pre-frontal cortex (Baxter, Schwartz, Phelps, Mazziotta et al., 1989;
Martinot, Hardy, Feline, Huret et al., 1990), with a possible lateralization effect
(Kimbrell, Ketter, George, Little, Benson, Willis et al., 2002). In general, unipolar
depressed patients, compared to healthy controls, have decreased dorsolateral and
dorsomedial prefrontal regional activity, as well as decreased metabolism in areas of the
anterior cingulate gyrus (see reviews: Dougherty & Ranch, 1997; Drevets, 2000; Drevets,
Price, Simpson, Todd et al., 1997). Similar to the results from neuropsychological task
performance, it is difﬁcult to determine is whether this is a state or trait phenomenon as
some effects do not remit with cessation of symptoms. Increased ventrolateral
preﬁontal/paralimbic metabolism has also been observed (Drevets, Videen, Price,
Preskom et al., 1992); this effect tends to normalize following antidepressant treatment
(Drevets, 2000). Thus, patients with depression show widespread prefrontal
dysfunctions. Speciﬁcally, there is hypometabolism in the dorsolateral and dorsomedial
preﬁontal cortex and anterior cingulate regions, and increased activity in the ventrolateral
prefrontal area.

The ﬁndings from neuroimaging studies appear to have functional implications.
The dorsal anterior cingulate area appears to be involved in the attentional and cognitive
features of depression (Flint, Black, Campbell-Taylor, Gailey et al., 1993), and prefrontal
hypometabolism is correlated with severity of features of depression (Kimbrell et al.,
2002). Cortical activation has been evaluated during neuropsychological test

performance, and depressed patients demonstrate different patterns of activity than

47

controls. Depressed patients had normal cerebral activity, as measured by event-related
potentials, while completing a task which required simple responses to stimuli, but
frontotemportal activity was reduced when response inhibition demands were included
(Kaiser et al., 2003). During the Stroop task, which assesses interference control, patients
with depression had little activation in the anterior cingulate cortex, which is normally
activated by controls, but increased activation in the left dorsolateral prefrontal cortex
(George, Ketter, Parekh, Rosinsky et al., 1997a). While completing a planning task,
depressed patients had no signiﬁcant activation in the cingulate cortex and striatum and
low levels of activity in other prefrontal areas, all of which are typically highly involved
in performance by control participants (Elliott, Baker, Rogers, O'Leary et al., 1997).
Thus, a general prefrontal dysfunction marked by particular deﬁcits in the cingulate-
striatal circuit may be associated with some of the EF deﬁcits observed in depressed
patients (V idebach, Ravnkilde, Pedersen, Egander, Landbo, Rasmussen et al., 2001).
Taken together, depressed patients, even those in the euthymic phase, appear to
show consistent deﬁcits on a wide range of EF functions. Cognitive deﬁcits appear to be
related to severity and number of depressive episodes. Neuroimaging ﬁndings support a
generalized dysfunction in the prefrontal cortex, as abnormal patterns of activity have
been found in the dorsolateral, ventrolateral, and cingulate regions of the brain. These
results may be speciﬁc to EF deﬁcits or they may reﬂect a more generalized pattern of
cognitive deﬁcits. Given this, it will be important to determine whether any EF deﬁcits
can be better accounted for by generalized processing speed deﬁcits. No studies could be
located to speciﬁcally assess the effect of comorbid disorders upon performance on EF

tasks.

48

Anxiety Disorders

Neuropsychological test performance and neuroimaging studies. Compared to
the other disorders covered here, much less attention has been paid to the relationship
between EF impairment and anxiety disorders. Sub-clinical anxiety is widely recognized
as disruptive to testing as it can adversely affect test performance and cloud interpretation
of results (Lezak et al., 2004). However, research has not supported a detrimental effect
upon EF abilities such as working memory and set-shifting for experimentally-induced
anxiety (Martin & Franzen, 1989) or levels of state or trait anxiety (Gladsjo, Rapaport,
McKinney, Lucas, Rabin, Oliver et al., 1998; Waldstein, Ryan, Jennings, Muldoon et al.,
1997), although males’ performance on a task assessing interference control was
adversely affected by induced anxiety (Martin & Franzen, 1989). Therefore, the effects
of subclinical levels of anxiety upon neuropsychological test performance are variable
and may be somewhat gender speciﬁc.

Research on clinical anxiety disorders has predominantly focused upon OCD
(Airaksinen et al., 2005). The disorder has frequently been associated with speciﬁc EF
deﬁcits that are related to frontostriatal dysfunction. These suppositions are partially
based on ﬁndings of deﬁcits in response inhibition, planning, set-shifting and
maintenance, working memory, and interference control (Aycicegi, Dinn, Harris, &
Erkrnen, 2003; Basso, Bomstein, Carona, & Morton, 2001; Boldrini et al., 2005; Moritz
et al., 2002; Penadés, Catalan, Andres, Salamero, & Gasto, 2005; van den Heuvel,
Veltrnan, Groenewegen, Cath, van Balkom, van Hartskarnp et al., 2005). However, the
results for the individual studies were inconsistent and the deﬁcits often varied between

studies. Suggestions have been made that the primary dysfunction is in the orbitoﬁontal

49

cortex (Aycicegi et al., 2003; Basso et al., 2001), but neuroimaging has also shown
dysﬁmctions in the dorsolateral and anterior cingulate regions (van den Heuvel et al.,
2005; Van Veen & Carter, 2002). Therefore, general prefrontal dysfunction has been
supported by neuropsychological and neuroimaging research. The effects following
anterior cingulate damage, however, suggest this region may be of primary interest.

The relationship between EF deﬁcits and other anxiety disorders has received
much less attention through neuropsychological studies, and the results are somewhat less
compelling. Individuals with panic disorder/agoraphobia have demonstrated intact EF
abilities such as working memory, cognitive ﬂexibility, set-shifting, and planning
(Asmundson et al., 1995; Boldrini et al., 2005; Gladsjo et al., 1998; Purcell et al., 1998).
Compared to controls, adults with panic disorder have shown set-shifting deﬁcits, but this
effect disappeared after individuals with an alcohol abuse disorder were removed from
analyses, possibly as a result of reduced sample size (Airaksinen et al., 2005). One study
also noted a trend towards verbal ﬂuency deﬁcits (Gladsjo et al., 1998). Adults with
social phobia tended towards reduced verbal ﬂuency (Airaksinen et al., 2005) and
showed impaired performance on a set-shifting task (Cohen, Hollander, DeCaria, Stein et
al., 1996); however, other studies do not support such ﬁndings (Asmundson et al., 1995;
Sachs, Anderer, Margreiter, Semlitsch, Saletu, & Katschnig, 2004). Generalized anxiety
disorder (GAD) has been associated with deﬁcits in interference control (Dibartolo,
Brown, & Barlow, 1997). No EF impairment has been associated with speciﬁc phobia in
the literature, although only one study could be found that included this disorder

(Airaksinen et al., 2005). Thus, there are some weak ﬁndings for EF deﬁcits in

50

individuals with panic disorder/agoraphobia, social phobia, and GAD, while none have
been associated with speciﬁc phobia.

The association between EF deﬁcits and post-traumatic stress disorder (PTSD) is
somewhat stronger. Impaired performance has been observed on tasks assessing set-
shifting, set maintenance, and working memory (Beckharn, Crawford, & Feldman, 1998;
Koenen, Driver, Oscar-Berman, Wolfe, Folsom, Huang et al., 2001; Stein, Kennedy, &
Twarnley, 2002). However, such ﬁndings may be associated with severity of current
psychopathology and limited to lower functioning samples as college students with PTSD
did not show any deﬁcits across a wide range of EF tasks (Twamley, Hami, & Stein,
2004). Therefore, deﬁcits in anxiety disorders may be related to severity of the disorder,
as suggested by the ﬁndings with PTSD.

It has been suggested that anxiety as a trait or a disorder is related to the inability
to control cognitive interference, and thus that there are weaknesses in cognitive control
(Eysenck & Calvo, 1992). However, it was noted that such deﬁcits would be more
apparent on tasks that involved greater cognitive load, particularly those that require
efﬁciency as opposed to effectiveness (Eysenck & Calvo, 1992). This theory has
received some support (Calvo & Eysenck, 1996; Hopko, Ashcraft, Gute, Ruggiero, &
Lewis, 1998). Thus, previous studies may have missed this effect because the tasks they
used may not have put enough demands on speed, working memory, and interference
control.

The lack of strong connection between anxiety disorders and EF deﬁcits may be
due to emphasis on neural systems outside of the frontal region in maintenance of these

disorders. One of the most prominent theories of anxiety was originally proposed by

51

Gray (1982) and was recently updated (Gray & McNaughton, 1995). This theory
involves the septo-hippocampal system, which includes the hippocampus, septa] nuclei,
dentate gyrus, subiculum, and parahippocampal gyrus. The primary involvement of these
systems may explain the paucity of support for EF deﬁcits in anxiety disorders, and
ﬁndings of memory and learning deﬁcits (Airaksinen et al., 2005; Boldrini et al., 2005;
Sachs et al., 2004). However, a review of neuroimaging in anxiety disorders suggested
that the orbitofrontal and anterior cingulate cortices are implicated in nearly all
manifestations of anxiety (Malizia, 1999), with a focus on increased right frontal activity
(Keller, Nitschke, Bhargava, Deldin, Gergen, Miller et al., 2000). It is possible that along
with not including measures with large enough task demands to differentiate between
adults with anxiety disorders and controls, previous studies may have been hindered by
sample sizes that were too small to detect effects on measures associated with
ﬁ'ontostriatal functioning.

Taken together, only OCD has been strongly related to EF deﬁcits. Due to the
weak ﬁndings and lack of real divergent ﬁndings across the other anxiety disorder,
anxiety disorders were combined to examine the EF effects in related groups of disorders
for the present study. Despite the weak ﬁndings, anxiety disorders are the most
commonly diagnosed psychiatric disorders in the general population (Kessler et al.,
2005a), and therefore should be evaluated in some manner in the present study. The
large sample size in the present study increased the ability to detect any EF deﬁcits in
anxiety disorders. Further, their inclusion enhanced the ability to examine the effects of
internalizing symptoms on cognitive functioning. Thus, despite the weak support for EF

impairment, anxiety disorders were included in the present study.

52

Summary

As can be seen from this selective review of the vast literatures examining
neuropsychological functioning in individuals with psychiatric disorders, there appears to
be support for some deﬁcits in EF with each of these disorders. Yet, the ﬁndings also
tend to vary considerably, with less support for some disorders (i.e., anxiety disorders,
ASPD, and drug use) versus others (i.e., alcoholism, ADHD, and depression). Further,
within each of the individual disorders, support for EF deﬁcits is variable and
inconsistent, often affected by the speciﬁc EF tests used, sample sizes, and presence of
comorbid disorders. For instance, the apparent weak support for EF deﬁcits in ASPD
may simply be due to very speciﬁc regional involvement of the orbitofrontal cortex
which is not associated with typical EF processes. As well, while most anxiety disorders
are only weakly related to EF deﬁcits, few studies with small sample sizes have hindered
the ability to detect any signiﬁcant effects. Finally, strong relationships between
disorders, such as those seen between drug and alcohol use disorders, may create the
appearance of EF deﬁcits that are only related to one of these disorders. Despite these
drawbacks and variable ﬁndings, some differential patterns of deﬁcits can be seen
(summarized in Table 1 of Appendix B), providing a theoretical basis for the
componential model examining the individual disorders.

However, the results also demonstrated that there are parallels in the literature and
interactions between disorders. These substantiated the need to evaluate the relationship
between EF deﬁcits and psychopathology at multiple diagnostic levels. As mentioned,
inadequate controls for comorbidity have made it difﬁcult to tease apart the independent

effects for individual disorders. The parallels in the literature make it pertinent to

53

understand whether the deﬁcits were differentiable or shared amongst the disorders. The
focus now turns to a discussion of the complex problem of comorbidity.
The Problem of Psychiatric Comorbidity

Given the overlap in symptomatology and high rates of comorbidity between
many of these disorders, it is questionable whether they should be separated and analyzed
individually. To do so may remove core aspects of the disorder that contribute to
cognitive deﬁcits. On the other hand, failure to control these effects makes interpretation
difﬁcult as comorbidity is recognized as a major impediment to understanding the
speciﬁcity of deﬁcits to individual psychiatric disorders (Angold et al., 1999; Sergeant et
al., 2002).

The importance of understanding the effect of comorbidity is underscored by the
frequent occurrences of multiple disorders. The National Comorbidity Study has
assessed the patterns of lifetime comorbidity in a large, population-based sample for the
disorders of interest for this study. More than half of all individuals who have ever met
criteria for a psychiatric disorder have had two or more different lifetime conditions
(Kessler et al., 2005a; Kessler et al., 1994). It is certainly a major issue for the disorders
reviewed above. Commonly associated disorders include ADHD with ASPD, and mood,
anxiety, and substance use disorders (Biederrnan et al., 1993; Biedennan et al., 1996);
ASPD with anxiety disorders and substance abuse/dependence (Goodwin & Hamilton,
2003; Petry et al., 2005); alcohol abuse with anxiety and aﬁective disorders in females
and other substance and antisocial disorders in males (Kessler et al., 1997); substance
disorders with other substance, mood, and anxiety disorders, and ASPD (Agosti, Nunes,

& Levin, 2002; Falck, Wang, Siegal, & Carlson, 2004; Miller et al., 1996; Skinstad &

54

Swain, 2001); depression with anxiety and eating disorders, substance dependence
(alcohol and drugs), and ASPD (Biederman, Petty, Faraone, Hirshfeld-Becker, Henin,
Pollack et al., 2005; Niles, Mori, Lambert, & Wolf, 2005; Rush, Zimmerman,
Wisniewski, F ava, Hollon, Warden et al., 2005); and anxiety disorders with other anxiety,
eating, and mood disorders (Belzer & Schneier, 2004; Hunt, Slade, & Andrews, 2004;
Sateen, Stein, Cox, & Hassard, 2004). Thus, overlap and co-occurrence of disorders is
frequently seen, particularly amongst the disorders of interest in this proposed study

Yet, as noted, many psychiatric studies of EF fail to control for the presence of
comorbid syndromes, and without doing so it is impossible to determine whether the
outcome is speciﬁc to the disorder under question, an associated condition, or some
combination of both disorders. Several conceptual models have been proposed to explain
the co-occurrence of two disorders (whether concurrently or successively) in the same
individual (Wonderlich & Mitchell, 1997). Rejected explanations that cannot adequately
account for the rates of comorbidity in most instances include chance co-occurrences of
the individual disorders (Lilienfeld, Waldman, & Israel, 1994) and comorbidity as an
artifact of research methodology (Angold et al., 1999). Moreover, despite criticisms that
the diagnostic system inappropriately separates certain disorders (Krueger, 1999) and
includes too many overlapping symptoms (Angold et al., 1999), the taxonomic issues
likewise do not completely account for the high rates of comorbidity (Angold et al.,
1999)

Theoretical models for comorbid presentations that still remain viable at least for
the disorders being considered in the proposed study include the following: (1) the

presence of a disorder increasing risk for the development of another disorder; (2)

55

multiple behavioral manifestations of the same underlying risk factor; and (3)
quantitatively different expressions of the same underlying risk factor or disorder
(Krueger, 1999; Wonderlich & Mitchell, 1997). Therefore, comorbidity is
conceptualized herein as a real phenomenon that can occur for multiple reasons.

The importance of understanding comorbidity comes in part from ﬁndings that
suggest it can affect the presentation of individual disorders. A number of disorders
show different characteristics when they have been preceded by another disorder (i.e.,
dysthymia and MDD; Kovacs, 1996), and a worse overall outcome has been found for the
concurrent overlap of two disorders than would be found for either disorder alone (i.e.,
ADHD plus an internalizing or externalizing disorder; Angold et al., 1999). In general,
comorbidity has been associated with greater severity of psychopathology (Angold et al.,
1999; Kessler et al., 2005b). Angold and colleagues (1999) suggested that descriptions of
disorders and thus their treatment could be improved by examining the effects of
comorbidity. They indicated that certain disorders, such as ADHD and conduct disorder,
could be deﬁned by their comorbid conditions because the comorbid presentations were
strikingly different from those associated with either of the individual, “pure” forms of
the component disorders. Thus, presence of comorbidity has been shown to change the
symptom presentations and outcomes of disorders in a manner that may support the
diagnosis of a different disorder or a subtype of the disorders in question (Angold et al.,
1999). In light of this issue, an examination of EF deﬁcits in multiple psychiatric
disorders should speciﬁcally evaluate the effects of comorbidity.

Despite the potential confounding effect of comorbid disorders on EF, they have

often been neglected in studies of psychopathology and neuropsychological assessment.

56

An evaluation of psychopathology other than the main disorder in question is not often
referenced in studies of EF. When it is evaluated in some manner, it is usually to screen
out participants for substance use disorders which could affect sober testing. Attempts to
speciﬁcally address the potential confound of comorbid disorders often involves
excluding participants who meet criteria for any DSM-IV Axis I (or, less frequently, Axis
II) disorder aside from the main disorder under question. This effort to isolate the
neuropsychological ﬁndings to the disorder under study raises a problem, however. That
is, given the frequent presentations of comorbid disorders in the general population, the
ecological validity of the results may be compromised by excluding so many naturalistic
cases (Downey et al., 1997; Kessler et al., 2005a; Kessler et al., 1994; Krueger, 1999).
To do so may simply exclude the most severe cases (Angold et al., 1999; Kessler et al.,
2005b), and thus lead to an underestimation of EF impairment and poor generalizability.
Further, the sole exclusion of current, and not lifetime disorders, overlooks possible
persisting EF impairment after disorders have remitted (Kessing, 1998; Munro et al.,
2000; Paradiso et al., 1997), which could confound results. Therefore,
neuropsychological studies do not usually evaluate the effects of comorbidity in a manner
that could help to elucidate its effects upon performance.

As shown in the review of EF studies for the individual disorders, some studies
have speciﬁcally considered the effects of comorbidity in their evaluation of
neuropsychological test performance. Their results highlight the importance of including
some assessment of comorbidity as the possibility of an additive neuropsychological
dysfunction has received some support (i.e., Downey et al., 1997; Riordan et al., 1999).

These studies suggest that comorbid conditions may affect cognitive performance of

57

individuals with psychiatric disorders, perhaps by amplifying cognitive deﬁcits. On the
other hand, performance on neuropsychological tests has not always been affected by
comorbid conditions (i.e., Katz, Wood, Goldstein, Auchenbach, & Geckle, 1998;
Uekermann et al., 2003). Many of these studies with null effects have been hampered by
small sample sizes, particularly once diagnostic groups were separated into those with
and without a comorbid condition. Questionable methods for diagnosing comorbid
disorders have also been used. Although the few instances where comorbidity was
focused upon provided inconsistent results, their ﬁndings highlighted the fact that
comorbidity could be important to EF impairment. The role of comorbidity is too rarely
focused upon in neuropsychological studies, and more clarity is needed on this issue.
Multiple possible hypotheses may be proposed to account for the manner by
which comorbid disorders could affect neuropsychological test performance. Speciﬁc
effects could result, wherein only certain comorbid disorders contributed to EF
impairment. Another possibility is that comorbidity is related to increased deﬁcits,
regardless of the speciﬁc disorders under question. In other words, increased numbers of
comorbid disorders would be associated with increased neuropsychological deﬁcits.
Frost and colleagues (1989) found that although different disorders varied in their
cognitive impairment proﬁles, the presence of multiple conditions was the best predictor
of neuropsychological dysfunction. This suggests that just the presence of comorbidity
may increase neuropsychological dysﬁrnction, regardless of the results for the individual
disorders. Similar ﬁndings resulted ﬁ'om a study of adults with ADHD, where number of
comorbid disorders was related to the degree of attentional deﬁcits (Taylor & Miller,

1997). Comorbidity is strongly related to severity of psychopathology (Angold et al.,

58

1999; Kessler et al., 2005b), and this may be the important factor in neuropsychological
deﬁcits. Thus, it is unclear whether EF deﬁcits are associated with speciﬁc symptom
presentations or are merely a marker of nonspeciﬁc psychopathology associated with
increased numbers of disorders.

Taken together, the presence of comorbid disorders appears to have potentially
signiﬁcant effects upon neuropsychological test proﬁles, but results thus far have been
inconsistent and appropriate studies are few. The lack of focus upon the effects of
comorbidity, resulting lack of consistency in the manner with which comorbidity has
been dealt, different methods for diagnosing the primary and comorbid disorders, and
small sample sizes have complicated any conclusions about speciﬁcity in
neuropsychological studies. Remaining questions include whether comorbidity affects
cognitive performance, whether any effect is only apparent for certain disorders and not
for others, and whether EF deﬁcits are associated with speciﬁc syndrome constellations
or are simply related to psychopathology in a nonspeciﬁc manner as part of the “burden
of comorbidity,” as hinted at earlier in the basic hypothetical models outlined.

These issues were speciﬁcally addressed within the present study by focusing on a
selected subset of common psychiatric disorders as deﬁned by the DSM-IV diagnostic
criteria. In addition, broader, dimensional categories of psychopathology were used to
implement a shared-symptom-based analysis as described below. Four models of the
relationship between EF and psychopathology were tested to attempt to better understand
how claims can be made regarding such wide—ranging involvement of EF deﬁcits in
many different disorders. These models moved from speciﬁc to more generalized

deﬁnitions of psychOpathology.

59

Study Objectives

The present study sought to examine the proﬁle of EF deﬁcits and psychiatric
disorders in ﬁve forms of psychopathology that had been associated in some manner with
neuropsychological EF deﬁcits: ADHD, ASPD, alcohol dependence, drug dependence,
depression, and anxiety disorders. Previous research in this area failed to account well
for comorbidity; this was a speciﬁc focus of interest in this study, in order to understand
the individual as well as combined contributions to EF impairment. It was important to
determine whether EF simply represented a nonspeciﬁc marker of disturbance or a key
etiological clue, in order to understand why the relationship between EF and
psychopathology is apparently believed to be so wide-spread. Tests assessing processing
speed, considered a reﬂection of general cognitive functioning, were included as non-
executive neuropsychological measures to help differentiate whether any
neuropsychological deﬁcits were speciﬁc to EF or were simply a reﬂection of general
neurocognitive impairment.

To recap, four models were tested, which represent four possible explanations for
the puzzling extensiveness of EF and psychopathology. The Componential Model
hypothesized that each disorder was associated with a different pattern of EF deﬁcits. As
highlighted by the review of the neuropsychological literature, there was some suggestion
that this may be part of the story. The componential model was examined using a proﬁle
analysis to determine whether there are differentiable deﬁcits in individual EF processes
(i.e., set-shifting, working memory, response inhibition, and interference control) for each
of the psychiatric disorders. Thus, although most disorders appeared to be related to EF

deﬁcits, in reality the deﬁcits would be differentiable and not shared across disorders.

60

The second model, Comorbidity-Speciﬁcity, suggested that only some, but not all,
disorders were related to more globally-deﬁned EF deﬁcits. This solution also received
some support as noted in the literature review earlier. Such an explanation would suggest
that the high frequency of comorbidity between disorders, which has been poorly
controlled in most studies, had made it appear as though EF deﬁcits were widespread
across psychopathology. By looking at the unique contribution of the individual
disorders to EF deﬁcits, it was possible to determine whether certain, but not all,
disorders were associated with EF deﬁcits and thus accounting for the perceived
widespread effects.

The third model, Comorbidity—Nonspeciﬁcity, suggested that EF deﬁcits were
simply non-speciﬁc markers for dysﬁmction, so that number rather than form of
psychopathology contributes to EF deﬁcits. As noted earlier, this model has also
received support in some instances. This model suggested that regardless of the
particular conditions involved, having more disorders would lead to increased EF
impairment.

Finally, the fourth model, dimension-speciﬁcity, suggested that the reason that EF
deﬁcits were so widely associated with psychiatric disorders was due to shared
underlying dimensions of psychopathology. The idea was that the EF deﬁcits were
associated with a common dysfunction shared across many disorders, and thus each of
these disorders appeared to be individually associated with impairment. The
contributions of these superordinate factors, which may represent shared underlying

psychopathological dysfunctions amongst disorders, was assessed by examining the

61

differential effects of “internalizing” versus “externalizing” disorders (see Krueger, 1999)
upon the manifestation of EF deﬁcits.

It was possible that one or more of these models would be better able to capture
the relationship between EF deﬁcits and psychopathology, or that none would completely
capture the complexity of the association. Analyses attempted to take into account the
possible contributions of single and multiple disorders, and provide room for a broader
examination of the possible contributions of shared “symptomatology.” It was intended
that these multiple levels of analysis would help to clarify the reasons for the extensive

literature identifying relationships between EF and multiple forms of psychopathology.

62

Methods

Two large samples from two separate, but related, community-based studies were
combined for data analyses to assess the relationship between psychopathology and EF.
This was a secondary data analysis using already existing data. Fresh data collection was
not feasible given the difﬁculty in obtaining large samples with a wide range of
psychopathology to meet the study objectives. The ﬁrst sample that will be described,
the ADHD sample, consisted of adults who were primarily recruited based upon presence
of ADHD and controls without ADHD. The second sample consisted of men who were
recruited based upon presence of alcohol dependence (with a secondary focus upon
ASPD), control men without alcoholism or other such drug use disorder, and their
spouses. Many of the participants who met criteria for ADHD or alcoholism also met
criteria for additional comorbid disorders, so there was ample representation of ASPD,
drug dependence, depression, and anxiety. Further, the control groups for each sample
did not meet criteria for the primary disorder of interest (i.e., ADHD or alcoholism), but
other psychiatric disorders were allowed to vary (with some exceptions, as described
below). Disorders also were free to vary for the spouses in the alcoholism study.
Therefore, within the control groups and alcoholism study spouses, psychiatric disorders
were adequately represented, as were numbers of individuals with no psychiatric
diagnoses for comparison purposes. In total, 641 participants were included in analyses

(described in greater detail below).

63

Sample 1: ADHD Adults and Controls
Participants

Recruitment. Adult participants were recruited from the community via public
advertisements and then evaluated in a standard multistage screening and diagnostic
evaluation procedure. Prospective participants contacted the project office at which point
key rule-outs were checked (age 18-40, no sensory-motor handicap, no neurological
illness, no head injury with loss of consciousness, and native English-speaking). Eligible
participants were then scheduled for the diagnostic visit wherein they completed semi-
structured clinical interviews to further evaluate ADHD and comorbid conditions. The
study was described in full detail to all potential participants at their ﬁrst visit (i.e., prior
to the clinical interview). Written informed consent was then obtained from all
participants.

Assessment of psychopathology. Potential participants were assessed for current
and lifetime childhood and adult disorders, and they provided self-reported levels of
impairment for those disorders. Adult Axis I disorders were assessed during a face-to-
face interview with the Structured Clinical Interview for DSM-IV Axis I Disorders
(SCID-I; First et al., 1997). This interview assessed symptoms of the following
disorders: MDD, dysthymic disorder, bipolar disorder, substance abuse and dependence,
psychotic symptoms, GAD, PTSD, OCD, panic disorder, agoraphobia, simple phobia,
social phobia, and eating disorders. ASPD and other personality disorders were assessed
with the SCID-II. Disorders that are typically seen in childhood, such as ADHD, CD,
and ODD, were assessed with the Kiddie Schedule for Affective Disorders and

Schizophrenia (K-SADS; Puig-Antich & Ryan, 1986). Previously published procedures

64

for assessing adults for these typically childhood-related disorders were followed
(Biederman, Faraone, Keenan, Benjamin etal., 1992; Biederman, F araone, Keenan, Knee
et al., 1990), and the K-SADS interview was worded appropriately to assess both
childhood and adult symptoms. These interviews were administered by masters-level
clinicians following extensive training, and participants provided self-reported
information about their symptoms. Autistic disorder was screened by the clinician using
added symptom questions and was a rule-out.

The assessment of adult ADHD requires retrospective assessment of their
childhood ADHD status to establish childhood onset by age 12, as well as inclusion of
informant interviews to verify symptoms and impairment (Wender, Wolf, & Wasserstein,
2001). Due to the retrospective nature of the interview for the childhood disorders,
informants were contacted to verify symptoms and impairment. Two informants were
contacted: (1) a ‘retrospective’ informant (usually a parent) to report on childhood
symptoms ADHD, CD, and ODD using modules from the K-SADS; and (2) a ‘peer’
informant who could report on current levels of ADHD symptomatology using the K-
SADS module as well as other current antisocial symptomatology. All informant
interviews were conducted by telephone after appropriate consent procedures. To ensure
that prospective ADHD participants exceeded normative cutoffs for level of ADHD
symptoms, participants and informants completed additional ratings scales (see Nigg et
al., 2005).

Diagnostic procedures. A diagnostic team that included a licensed clinical social
worker, a licensed clinical psychologist, and a board certiﬁed psychiatrist evaluated the

item coding for ADHD. A best estimate diagnosis was reached for ADHD status based

65

upon self and informant ratings. Each team member independently reviewed all available
information from SCH), K-SADS, and rating scales to arrive at a clinical judgment about
all additional Axis I and II disorders. The DSM-IV criteria regarding comorbidity were
carefully followed in all cases, so that although comorbid disorders were diagnosed when
present, ADHD was not diagnosed if clinicians judged that symptoms were better
explained by a co-occurring mood or other major disorder (American Psychiatric
Association, 1994). Evidence of impairment was required to make diagnoses.
Individuals were classiﬁed for each adult disorder according to the following scale: (1)
Deﬁnite, meets ﬁrll criteria; (2) Probable, falls one symptom short of hill criteria; (3)
Possible, falls two symptoms short of full criteria; and (4) No diagnosis, falls three or
more symptoms short of full criteria. The rating scale for the K—SADS was different for
the childhood disorders: (1) Deﬁnite, meets full criteria; (2) Subthreshold, meets criteria
for more than half of the symptoms; and (3) No diagnosis, meets criteria for fewer than
half of the symptoms. Inter-rater agreement on presence or absence of ADHD (deﬁnite)
and other disorders were satisfactory (Nigg et al., 2005).

Exclusionary criteria. Potential participants were excluded from the ADHD and
the non-ADHD groups if they had a current major depressive or manic/hypomanic
episode, current severe substance dependence preventing sober testing, subthreshold
childhood ADHD, history of psychosis, history of autism, F SIQ < 75, history of head
injury with loss of consciousness (if determined to be of moderate severity), sensory-
motor handicap, neurological illness, native language not English, or currently prescribed
anti-psychotic, anti-depressant, or anti-convulsant medications. For the non-ADHD

group additional exclusions were antisocial or borderline personality disorder, past

66

bipolar disorder, or a previously diagnosed learning disorder. Bipolar disorder was an
exclusion for any individual in the present study due to its potential to confound
neuropsychological test results. Other psychiatric disorders were free to vary.

Criteria for testing. Participants that met inclusion criteria were asked to return
for neuropsychological testing at the lab ofﬁces on the Michigan State University
campus. Participants had to be sober during testing. They were checked for recent
alcohol, marijuana, and other medication consumption prior to testing. A number of
individuals were taking regular psychostimulant medications (Adderall, Ritalin,
Concerta, and Focalin in this sample). They were tested after a minimum of 24 hours (for
short acting preparations) to 48 hour washout (for long acting preparations); actual mean
washout time was 63.8 hours.

Final sample. 424 adults passed through the initial screen and completed the
screening rating scale and the diagnostic screen visit. The diagnostic procedures qualiﬁed
195 of them (46%) between the ages of 18 and 37 for the study, grouped into an ADHD
group (n = 105), and a non-ADHD control group (n = 90) and these 195 completed the
neuropsychological battery and were available for the present study. All participants
completed at least three of the four neuropsychological measures in this study. Presence
or absence (i.e., 0 or 1) of disorders were recoded for the following lifetime and current
diagnoses: MDD, dysthymia, bipolar 1, alcohol dependence, drug dependence (any type),
agoraphobia without panic disorder, panic disorder, social phobia, speciﬁc phobia, OCD,
PTSD, GAD, antisocial personality disorder, conduct disorder, ODD, and ADHD. Four
individuals met criteria for panic disorder; although DSM-IV panic disorder could not be

diagnosed retrospectively in the alcoholism study and thus was not included in the

67

present analyses, these individuals were retained in the study but not deﬁned as having
panic disorder. This was unlikely to affect analyses as there were no signiﬁcant
correlations between lifetime or current panic disorder diagnosis and neuropsychological
variables of interest. Two participants were removed from analyses because they met
criteria for bipolar I disorder (to avoid confounding neuropsychological test results since
this was not a focus of analyses), leaving n = 193 participants.
Sample 2: Alcoholism Study

Participants

The second sample consisted of parents (dads and moms) from families who
participated in the Michigan Longitudinal Study (MLS, also referred to herein as the
alcoholism study), an ongoing longitudinal study on the development and
stability/change of alcohol problems/abuse/dependence (Zucker et al., 2001). All
families were recruited based upon the alcoholism status of the father, and given the
interest in development of alcoholism, families were only included in the study if they
had at least one biological male child. All families were initially told that they were
being recruited to participate in a longitudinal study of child development and family
health. More detailed information was provided if requested. Neuropsychological
testing has been conducted for both children and adults; only data from the adult samples
were included in the present study, so the description of study methodology will focus
upon their recruitment.

Data have been collected at 3-year intervals since initial family contact (i.e.,
Wave 1). Parental consent and child assent was obtained from all participants by the

Field Coordinator at the ﬁrst contact visit of Wave 1. At the end of each wave of data

68

collection, request for permission to re-contact the family in three years was obtained,
and written consent was re-established at each new wave. Intellectual functioning data
were collected at Waves 2 and 4. At Wave 5, neuropsychological testing was
administered to all parents participating in the study. Wave 5 testing has been completed
for all families.

Three hundred and eleven families initially agreed to participate in the study. The
MLS has maintained contact with 100% of all still living participants from Wave 1 data
collection. Only a small percentage of families are no longer willing to participate in data
collection, and the sample currently contains more than 90% of the original participants.

Recruitment. Three subsets of families were originally recruited into the
Longitudinal Study. The three subsets were recruited to represent varying levels of risk
for development of alcoholism in the children. The subset of families included in the
present analyses are: (1) Court Alcoholic (n = 159), a court-recruited alcoholic family
group that included at least one parent with alcoholism, and often, with ASPD; (2)
Community Alcoholic (n = 91), a community-recruited alcoholic family group; and (3)
Control Group (n = 61), a matched group of community-recruited, non-substance abusing
parents and their offspring.

Families in the Court Alcoholic group were initially identiﬁed using court records
from four counties in Michigan. All men with drunk-driving convictions involving a
blood alcohol concentration of at least 0.15% for a ﬁrst conviction, or 0.12% if they had
been involved in multiple convictions, were potential study participants. Along with
conviction status, men who were included in the study had to meet Feighner and

colleagues (1972) criteria for probable or deﬁnite alcoholism (assessed using the Short

69

Michigan Alcoholism Screening Test (SMAST; Selzer, Vinokur, & van Rooijen, 1975),
have a male child between the ages of 3.0 and 6.0 currently living in the home, be living
with the boy’s biological mother at the time of ﬁrst contact, and identify as Caucasian.
Seventy-nine percent of the men identiﬁed through this method allowed the probation
ofﬁcers from the district courts to release their contact information. Of those contacted,
92% agreed to participate in the study. This recruitment strategy provided a large
number of alcoholic fathers with comorbid ASPD. Mother’s alcohol status was not a
factor in study inclusion or exclusion.

Control group families were of a similar composition (i.e., biological father and
mother, together at the time of ﬁrst contact, with their 3-5 year old male child living in
the home), but were recruited using a yoking procedure and door-to-door survey methods
in the same neighborhoods where the Court Alcoholic families resided. The canvas team
would begin a door-to-door search two blocks away from a court alcoholic family’s
residence for intact families with a male child whose age was within six months of the
alcoholic family’s son, but where neither the mother nor father met criteria for either
probable or deﬁnite alcoholism or other drug abuse/dependence using the Feighner et al.
(1972) criteria.

While canvassing the neighborhoods another subset of alcoholic families was
found, called the Community Alcoholic Group. These families were also intact with a
male child within six months of the court alcoholic family’s child, but the father met
criteria for probable or deﬁnite alcoholism during the initial screening procedure.
However, these fathers did not have any recent drunk driving or drug involved arrest

record. They were considered to fall into a moderate risk category for the male child to

70

develop problems with alcohol and provided a diagnostic comparison group. All 61
families agreed to participate in the study.

Note that familial conﬁgurations have changed over the tenure of the study, and
new family members (e. g., step-parents, younger siblings) have been added to assessment
protocols.

Diagnostic procedures. As mentioned, preliminary alcoholism diagnoses were
made using Feighner criteria for a probable or deﬁnite diagnosis of alcoholism. Although
DSM diagnoses were not a criterion for study inclusion, data that were collected through
SMAST, the Drinking and Drug History Questionnaire (DDHQ; Zucker, Fitzgerald, &
Noll, 1980), the Antisocial Behavior Inventory (ASB; Zucker, Noll, Harnm, Fitzgerald, &
Sullivan, 1994; Zucker et al., 1996) and the Diagnostic Interview Schedule, Version 111
(DIS; Robins, Helzer, Croughtan, & Ratcliff, 1980) were later used to establish DSM—IV
alcohol-related and ASPD diagnoses. Inter-rater reliability for these diagnoses was
excellent (Zucker et al., 1996).

During Wave 1, information on lifetime diagnoses was collected; at each follow-
up assessment, diagnostic information was obtained for the intervening three years
between data collection waves. Lifetime diagnoses were created by combining
information gathered at these ﬁve different time points for the disorders of interest in the
present study. “Current” diagnoses were those that the participants met criteria for near
the time of neuropsychological assessment (i.e., within three years prior to Wave 5).

At each wave of data collection, diagnostic information was collected using the
Diagnostic Interview Schedule (DIS) appropriate for the version of the DSM that was

current at that time. Therefore, diagnostic classiﬁcation systems used during the tenure

7l

of the study included the DSM-III, DSM-III-R, and DSM-IV. All diagnostic information
was recently converted to DSM-IV standards. Diagnostic information was collected for
the following disorders: MDD, dysthymia, mania/hypomania, schizophrenia, eating
disorders, speciﬁc phobia, social phobia, agoraphobia, panic, GAD, PTSD, OCD,
separation anxiety disorder, ODD, ADHD, and ASPD. As mentioned above, additional
collateral questionnaire and interview information was used along with DIS results to
reﬁne alcohol dependence and ASPD diagnoses since these were the focus of the
alcoholism study. Further, in the alcoholism group, participants were considered to meet
criteria for ASPD whether or not they met criteria for childhood CD. This exception was
made because the retrospective nature of the assessment made it difﬁcult to deﬁnitively
determine presence or absence of the childhood CD. (However, note that in the ADHD
group care had been taken to ensure the validity of retrospective CD diagnoses by
contacting a parent, so ASPD criteria did require a history of CD in that sample.)

Exclusionary criteria. Fetal alcohol effects in the target male child were
exclusionary for a family’s participation for all groups. Any participants who met criteria
for lifetime diagnoses of psychosis, bipolar disorder, neurological illness, and F SIQ < 75
at any time point were removed from the data sets due to the potentially confounding
effects upon neuropsychological data.

Criteria for testing. Examiners traveled to the families’ homes in order to
administer the neuropsychological test battery. Privacy and minimal distractions were
ensured throughout testing. Participants were not allowed to have drunk more than two
alcoholic beverages during the hour prior to testing. Information was collected on

medication and drug use prior to testing.

72

Final sample. Seven hundred and thirty-eight individuals participated in some
part of the alcoholism study. Since neuropsychological tests were administered at only
T5, and this was an important element of the present study, individuals who had
completed fewer than three out of the four tasks included in this study (i.e., Trail Making
Test, Stroop Interference, Stop Signal Test, and WCST) were eliminated from further
analyses. This resulted in the following participants being removed ﬁom the database:
participants who had not completed any of the neuropsychological tasks at T5 (n = 247),
and participants with only one (n = 4) or two (n = 22) tasks completed.

As mentioned, diagnostic information for the alcoholism sample was based upon
information evaluated and collapsed across ﬁve data collection waves. These diagnoses
were recoded into presence or absence (i.e., 0 or 1) of disorder for lifetime (i.e., at any of
the ﬁve time points), or current (i.e., since the previous data collection wave, or within
the three years prior to neuropsychological data collection at T5). If information on a
diagnosis was missing for any time point, the disorder was considered “absent” at that
time. Diagnoses recoded for analyses were: MDD (single or recurrent episodes),
dysthymia, bipolar I, schizophrenia, schizophreniform disorder, schizoaffective disorder,
agoraphobia without panic disorder, social phobia, speciﬁc phobia, OCD, PTSD, GAD,
alcohol dependence, drug dependence (any type), antisocial personality disorder (with or
without presence of conduct disorder), conduct disorder, ODD, and ADHD.

In order to match with exclusion criteria in the ADHD study, further rule cuts
were made for the following issues from the sample of n = 465 participants:
schizophrenia-spectrum disorder or history of psychosis (n = 10), bipolar I disorder (n =

9), epilepsy (n = 1), and anti-psychotic and anti-mania medications (n = 2). Thirty-nine

73

participants said they had experienced a head injury, and 29 participants endorsed a loss
of consciousness. However, these participants were retained in analyses since people
tend to misunderstand and over-endorse these experiences, the extent of these injuries
was not followed up on in the alcoholism study, and mild head injuries do not appear to
be associated with deﬁcits on tasks similar to those included in the present study
(Ettenhofer, 2006; note that there were also no signiﬁcant correlations with tasks in the
present data, r = -.06 - .07). Four hundred and forty-eight participants were included in
analyses.
Neuropsychological Test Battery

Except for the estimate of Full Scale IQ (F SIQ) and location of the testing, the
speciﬁc test forms and administration practices were very similar between the samples.
Therefore, the IQ estimation will be explained separately for each sample, but the EF
tasks will be explained together.
Full Scale IQ (FSIQ)

Sample I: ADHD adults. F SIQ was estimated with a ﬁve subtest short form of
the Wechsler Adult Intelligence Scales, Third Edition (WAIS-III; Sattler & Ryan, 1999;
Wechsler, 1997): Picture Completion, Vocabulary, Similarities, Arithmetic, and Matrix
Reasoning. Reliability and validity for this short form are adequate (Sattler & Ryan,
1999)

Sample 2: MLS/Alcohol parents. Four subtests of the WAIS, Revised Edition
(WAIS-R; Wechsler, 1981) were administered at Waves 2 and 4: Picture Completion,
Information, Arithmetic, and Block Design. This short form has been shown to be a valid

estimate of full scale IQ (Reynolds, Willson, & Clark, 1983). The results from Wave 4,

74

the most recent testing, were used in our analyses to provide the most accurate estimate
of parents’ intellectual functioning. There is an approximately six—year interval between
the ﬁrst and second test administrations for the parents in the alcohol study, suggesting
that retest effects were unlikely; IQ correlation between T2 and T4 assessments were .79,
alpha = .88.

Neuropsychological Tests

The tasks that were selected for inclusion in the present study are widely-used
clinical measures that assess EF and the integrity of the frontal lobes and associated
circuitry. Given the difﬁculties inherent in measuring EF and the importance of
demonstrating frontal involvement for tasks included in EF proﬁles (Sergeant et al.,
2002), the following is a brief review of lesion and imaging studies for the EF tasks
included in the present study, suggesting regions of activation that appear to be necessary
for task activity. This information is important to understand the basis for the
componential model, which examined the proﬁle of EF deﬁcits associated with the
individual disorders. Although the following review highlights the involvement of the
prefrontal cortex in each of these tasks, it also underscores the regional differentiation of
underlying neural circuits for different EF processes, which provides the theoretical basis
for the componential model.

Trail Making Test. The Trail Making Test (Trails) is a widely-used, timed paper-
and-pencil test consisting of two parts (Reitan, 1958). In Part A, the participant draws a
line connecting numbered circles in sequential order. In Part B, the participant draws a
line connecting numbered and lettered circles, switching between numbers and letters in

alphabetic-numerical order. Scores on each part of the Trails test are determined by the

75

time (in seconds) required to complete each trial. Whereas performance on Part A
depends largely upon psychomotor speed and visual search abilities, Part B also requires
working memory and cognitive ﬂexibility as the participant must maintain two mental
sets and alternate between them (Arbuthnott & Frank, 2000). For analyses using speed
variables, Trails A time was used as a measure of motor speed with higher scores
indicating slower speed. To assess set-shifting, a component of EF, a Trails B Residual
score was created by removing the effects of Trails A performance from Trails B (done
separately within the individual samples). Higher scores indicated poor performance.
Despite the widespread use of Trails as a frontal lobe task, few studies have been
done to validate such an application. No imaging studies are available for the Trail
Making Test due to its motor requirements, but a few studies have attempted to measure
the effect of lesion damage upon performance. One group of researchers found no
differences between patients with frontal and non-ﬁ'ontal lesion damage on Trails A, but
discrepancies in performance on Trails B (Ettlin, Kischka, Beckson, Gaggiotti,
Rauchﬂeisch, & Benson, 2000). This follows the expected pattern, as Trails B is thought
to be a measure of executive or frontal functioning due to its set-switching component,
while Trails A involves motor speed and visual scanning (Crowe, 1998). The ﬁndings of
a recent meta-analysis, however, were at odds with this hypothesis, as there was a much
larger effect size for Trails A in comparisons between frontal and non-frontal patients
(Demakis, 2004). Other researchers have also failed to differentiate ﬁ'ontal from more
posterior damage using the Trail Making Test (Anderson, Bigler, & Blatter, 1995; Reitan

& Wolfson, 1995).

76

Two explanations may account for this failure to ﬁnd frontal lobe speciﬁcity.
Firstly, the effect of damage to speciﬁc regions of the frontal lobes may have been lost by
combining studies that included any type of frontal lobe damage. Supporting this
possibility is the ﬁnding that patients with damage to the dorsolateral prefrontal cortex
showed the greatest impairment on Trails B, while damage to inferior frontal (ventral-
medial/orbitofrontal) and anterior cingulate regions had little effect on performance of
this task (Stuss, Bisschop, Alexander, Levine, Katz, & Izukawa, 2001a). The results of
this study also suggested that frontal lobe patients were slower than patients with
posterior lesions on Trails A, Trails B, and a proportional score intended to isolate the
executive aspects of Trails B performance. Further, only frontal lobe patients made more
than one error. These ﬁndings suggest that the Trail Making Test is sensitive to frontal
lobe damage, but perhaps more selectively for damage to the dorsolateral region.

Secondly, a derived score, similar to that used in the previously mentioned study
(Stuss et al., 2001a) may be more effective in differentiating frontal ﬁom non-frontal
damage. As mentioned, Trails A performance involves visual scanning abilities and
motor speed. By removing these effects from Trails B performance, a more pure measure
of executive abilities may be derived which more selectively assess the putative EF set-
shifting ﬁmctions of the frontal lobes (Arbuthnott & Frank, 2000).

Taken together, despite varied ﬁndings, the Trail Making Test appears to involve
the frontal lobes in performance, and perhaps selectively reﬂects the integrity of the
dorsolateral region. The isolation of the EF components of the task using a derived score

may better differentiate between frontal and non-frontal performance.

77

The Stroop Color- Word Test (Stroop). The Stroop Test (Golden, 1978) is a well-
known task which assesses the abilities to control interfering information and response
conﬂict. Three conditions are performed under a time pressure, each with separate
stimulus cards with ﬁve rows of twenty-ﬁve items. Participants are directed to read
down these rows as quickly as possible without making any errors. The number of items
completed after forty-ﬁve seconds is the total score for each condition. In the ﬁrst
condition, Word Reading, the participant reads color words (i.e., red, blue, and green)
that are printed in black ink. For the second condition, Color Naming, the stimulus card
has rows of X’s that are printed in different colored ink (i.e., XXXX) and participants
name the color of the ink for each item. In the ﬁnal condition, Color-Word, participants
must name the color of the ink for color words (i.e., red, blue, and green) that are printed
in an incongruent color. Naming the ink and not reading the word requires considerable
effort as reading is an overleamed activity, and thus participants must suppress the
prepotent response (Ward et al., 2001). Similar to the Trails derived score, an
interference residual score was created that removed Word Reading and Color Naming
performance from Color Word results (done separately in the individual samples). This
provided a purer measure of interference control as it removed the effects of processing
speed; higher scores indicated better performance. For analyses using speed measures,
Word Reading and Color Naming were included as measures of processing speed, with
higher scores indicating faster performance.

The primary EF component assessed by the Stroop task is interference control, as
measured by the incongruent condition. The primary region of activation during the

Stroop effect appears to be the anterior cingulate cortex (Smith & J onides, 1999).

78

Different aspects of the task appear to activate different frontal lobe regions, with the left
dorsolateral cortex activated during color naming for both congruent and incongruent
conditions (MacDonald, Cohen, Stenger, & Carter, 2000), while anterior cingulate
activation is speciﬁc to the incongruent naming condition (Bench, Frith, Grasby, Friston,
et al., 1993; MacDonald et al., 2000; Ravnkilde, Videbech, Rosenberg, Gjedde, & Gade,
2002). Activation has also been seen in the right superior mesial frontal lobe during the
Stroop effect (Larrue, Celsis, Bes, & Marc-Vergnes, 1994). Therefore, the medial region
and particularly the anterior cingulate cortex have been shown to be challenged during
the incongruent condition of the Stroop task.

Lesion studies have also suggested that frontal regions, particularly the anterior
cingulate cortex, are important to Stroop performance. A recent meta-analysis of lesion
studies found that there were moderate effects sizes for discriminating ﬁ'ontal ﬁom non-
frontal lesions on Stroop Word, Color, and Color-Word (Demakis, 2004). In one study,
damage to left mid-dorsal anterior cingulate cortex, but not the same region on in the
right hemisphere, resulted in consistently lower accuracy on incongruent trials, with
deﬁcits in maintaining task set and inhibiting the automatic response to read the words
(Swick & J ovanovic, 2002). Another study that included patients with single, focal brain
lesions in frontal and non-frontal regions found that only patients with frontal damage
had signiﬁcant impairment (Stuss et al., 2001b). They were slower overall on all
conditions, and similar to the results of imaging studies, damage to the left dorsolateral
frontal lobe resulted in increased errors and slowness in response speed for color naming,
while bilateral superior medial frontal damage (i.e., anterior cingulate) was associated

with increased errors and slowness in response time for the incongruent condition.

79

Increased errors have also been noted in right frontal lobe patients (V endrell, J unqué,
Pujol, Jurado et al., 1995).

Taken together, these imaging and lesion studies suggest that the Stroop task
selectively involves the frontal lobes. The task discriminates between frontal and non-
ﬁontal patients, and primarily activates regions of the superior frontal lobe such as the
anterior cingulate cortex. Therefore, the Stroop task appears to be a good measure of the
integrity of the anterior cingulate region of the brain.

Wisconsin Card Sorting Test (WCST). The WCST is a widely-used computer-
administered task assessing working memory, concept formation, and set-shifting
(Heaton, Chelune, Talley, Kay, & Curtiss, 1993). Participants were shown a computer
screen with four patterned “key” cards to which they needed to match a stimulus card
which appeared at the bottom of the screen. There were three different principles by
which the stimulus card could be matched to a key card: color, shape, or number. Using
feedback ﬁom the computer (i.e., “right” or “wrong”) participants had to deduce how to
sort the cards. Once they had ﬁgured out what principle to use in sorting the cards, the
category remained the same for ten cards, at which point the sorting principle switched.
In the alcohol group, up to 124 trials were administered. A shorter version of the task was
administered in the ADHD group, so up to 64 trials were administered. These tasks are
considered to be comparable (Axelrod, 2002; Love, Greve, Sherwin, & Mathias, 2003;
Sherer, Nick, Millis, & Novack, 2003). Participant performance was evaluated using the
standard scores for number of perseverative errors (i.e., continuing to sort using an

incorrect principle). Higher scores indicated better performance on this task.

80

Despite its widespread use (Retzlaff et al., 1992), the many and varied task
demands associated with the WCST have made it difficult to achieve consistent results
across lesion studies with regard to the involvement of the frontal lobes (Stuss, Levine,
Alexander, Hong, Palumbo, Hamer et al., 2000). This task involves working memory,
set-shifting and maintenance, abstract conceptualization, and responding to feedback.
While a number of studies have found that individuals with damage to frontal regions of
the brain perform more poorly on the task, speciﬁcally making increased perseverative
errors (Goldstein, Obrzut, John, Ledakis, & Armstrong, 2004a), other studies have
produced negative ﬁndings. Patients with posterior damage have also performed poorly
on the task, while frontal lobe patients have performed normally (Anderson et al., 1991).
In a study of patients with stable focal lesions there were no signiﬁcant differences
between frontal and non-frontal patients on WCST performance (Anderson et al., 1991).
These authors suggested that impaired performance on the WCST could not be
interpreted in isolation as a marker of frontal lobe damage.

However, similar to some of the studies with other EF tasks, it may be that
distinct regions of the ﬁ'ontal lobes are involved with task performance, and combining
frontal patients into a single group may result in a loss of information. While Anderson
et a1. (1991) did not ﬁnd any effect using sub-region analyses, perhaps due to small
sample sizes, others have found that damage to the dorsolateral (Milner, 1963) as well as
superior medial prefrontal regions (Stuss et al., 2000) were related to poor performance
on perseverative errors. Further, unilateral and bilateral lesions to the inferior medial
regions of the ﬁontal lobes (i.e., orbitofrontal) had little effect upon these measures of

task performance (Stuss et al., 2000). These authors suggested that the WCST was

81

sensitive to focal frontal lesions, but that their results supported a differentiation of
processes within the frontal lobes.

Support for these results come from a recent meta-analytic review which found
that participants with frontal damage showed poorer performance on all variables of the
WCST except for nonperseverative errors (Demakis, 2003). Strikingly, in moderator
analyses there was a large effect size for damage to the dorsolateral prefrontal cortex on a
composite WCST variable. These lesion results suggest that the dorsolateral prefrontal
cortex contributes to performance on commonly assessed variables of the WCST.

The lesion data is substantiated by the ﬁndings from imaging studies. They show
activation of the dorsolateral prefrontal cortex during task performance (Berman, Ostrem,
Randolph, Gold et al., 1995; Rezai et al., 1993). The right dorsolateral frontal-subcortical
(i.e., caudate nucleus) circuit was found to be critical for WCST performance as the
activation of this circuit was associated with reduced perseverative responses in patients
with a history of closed head injury (Lombardi, Andreason, Sirocco, Rio, Gross, Umhau
et al., 1999). Other researchers have found localized blood ﬂow to the left dorsolateral
frontal regions (Rezai et al., 1993), even following training and practice on the test
(Berman et al., 1995), suggesting that the working memory components of the task may
be largely responsible for the involvement of this region (Berman et al., 1995).
Therefore, the results of neuroimaging studies provide additional evidence for discrete
regional activation of the dorsolateral prefrontal cortex in task performance on the
WCST. Although different regions of the frontal lobes may be involved in performance
on different aspects of the task (Stuss et al., 2000), perseverative errors appears to be a

sensitive indicator of the functional integrity of the dorsolateral prefrontal cortex.

82

Logan Stop Task (Stop). The Stop task (Logan, 1994) is a dual-task computer
paradigm to assess response suppression or inhibition in a rapid decision context.
Procedures were the same as those used by Logan, Schachar, & Tannock (1997). The
computer screen displayed an X or an O on a black and white screen and individuals were
required to respond to these stimuli by pressing designated buttons labeled ‘X’ and ‘O’ as
quickly as possible using their dominant hand. They were to withhold responding when
they heard a tone. Four blocks of 64 trials were administered following two practice
blocks of 32 trials each. We used the tracking version of the stop task, which provides the
most valid estimates of stop signal reaction time (Band, Van Der Molen, & Logan, 2003).
The time of the stop signal tone was varied in a stochastic procedure to maintain accuracy
at 50%, so that stop signal reaction time (SSRT) was computed as the difference between
stop signal delay and go speed (Logan, 1994). In this study, we used stop signal reaction
time (response inhibition) as a measure of the time needed to inhibit a response (i.e., an
EF process), and variability of the go reaction time (RT variability) as a measure of
response variability on trials that they did not have to inhibit responding. Both of these
are related to EF, and higher score indicated poorer performance (i.e., slower inhibition
or more response variability).

Across studies, performance on the Stop task has been shown to be mediated by
the prefrontal cortex. Patients with frontal lobe damage have longer SSRTs than
orthopaedic and normal controls (Aron et al., 2003; Rieger, Gauggel, & Burmeister,
2003); in other words, frontal patients require a longer delay to successfully inhibit their
response. This is not a consistent ﬁnding (Dimitrov, Nakic, Elpem-Waxman, Granetz,

O'Grady, Phipps et al., 2003), but, as with other EF tasks, some variation may occur

83

based upon the speciﬁc location of the damage within the frontal lobes. For instance,
within frontal lobe patients those with right hemisphere or bilateral lesions have
signiﬁcantly longer SSRTs than patients with left lesions (Rieger et al., 2003).

Therefore, although the lesion results are not consistent in separating frontal from non-
frontal patients, they suggest that the right frontal circuits are involved in the inhibition of
ongoing responses.

Imaging data support the lesion results, and suggest discrete activation of a region
of the preﬁontal cortex in task performance. In healthy adults, a bilateral middle and
inferior frontal system appears to be predominantly involved in response inhibition
(Rubia, Overmeyer, Taylor, Brammer, Williams, Simmons et al., 2000). In a sample of
right ﬁontal lobe patients, Aron and colleagues (2003) found that damage to the inferior
cortex, speciﬁcally the pars triangularis, accounted for the variability in SSRT. These
authors indicated that low variability in damage to orbitofrontal and medial frontal
regions left open the possibility that other regions may be involved. The involvement of
the right inferior frontal cortex in successful response inhibition was also seen in a study
of normal controls (Rubia, Smith, Brammer, & Taylor, 2003), and a recent review
suggested that it was a commonly recruited region across studies of response inhibition
(Aron et al., 2004). In children, the activity in the prefrontal cortex as well as basal
ganglia has been shown to be important to task performance (Casey et al., 1997).
Therefore, lesion and imaging data have pointed to a speciﬁc role of the right frontal
cortex, speciﬁcally inferior and orbitofrontal regions and their striatal connections, in

response inhibition as measured by SSRT.

84

A measure that recently received attention as an index of regulatory control,
perhaps related to executive control but also possibily related to other processes such as
arousal, is variability of response time, such as that measured by the variability of the go
response (RT variability) in the Stop task. Intra-individual performance variability on an
executive function task has been associated with superior and dorsolateral preﬁrontal
brain lesions in adults (Stuss et al., 2003). Variability of response time has been shown
to be related to neurocognitive impairment in aging (Dixon, Garrett, Lentz, MacDonald,
Strauss, & Hultsch, 2007). Functional imaging using a Go/No-Go task revealed that go
response variability is related to response inhibition, and involves an overlapping neural
network that includes bilateral middle frontal regions, along with right inferior parietal
and thalarnic regions (Bellgrove, Hester, & Garavan, 2004). These results were
interpreted as suggesting that the increased frontal activation that is associated with
higher intra-individual response time variability reﬂects the need for executive control to
maintain task performance (Bell grove et al., 2004). Caudate and prefrontal regions were
also recruited in children with high response variability (Simmonds, Fotedar, Suskauer,
Pekar, Denckla, & Mostofsky, 2007). Again, this supports the idea that prefrontal areas
are needed for maintenance of task performance in conditions of behavioral inconsistency
(Simmonds et al., 2007). Therefore, response variability, or RT Variability, appears to be
a measure of executive control that is related to response inhibition and relies upon

preﬁ'ontal networks.

85

Results
Data Preparation

Data cleaning. As recommended in recent methodological texts, extreme outliers
(z > 4.0 and more than .5 SD ﬁ'om next score) were truncated to within .5 SD of the next
nearest score to prevent undue inﬂuence of single scores on linear models and reduce
type I and type 11 error (see Wilcox, Keselman, & Kowalchuk, 1998). Data preparation
was completed in the two groups separately.

Three scores were adjusted in the ADHD group using the above-described
method for removing outliers: one for Trails B (from 137.37 to 113.17), one for Stroop
Color-Word (102 truncated to 101), and one for Response Inhibition (from 625.33 to
480.67). The expectation maximization (EM) method of data imputation was used to
impute missing data in each of the samples individually, and 1.6% of the cognitive
variables were imputed in the ADHD group. Residual scores were created for the Stroop
and the Trails task as described above under task descriptions. Standardized z-scores
were also created for each task included in analyses using the mean and standard
deviation of the ADHD sample control group (i.e., individuals with no lifetime
disorders).

For the MLS/alcoholism group, a total of 10 outliers were adjusted: three for
Trails A (79.57, 88.00, and 100.00 truncated to 71.5), three for Trails B (196.56, 198.13,
and 200.00 truncated to 173.33), two for Response Inhibition (569.00 and 598.00 to
564.69), and two for WCST perseverative errors (55.00 and 76.00 to 51.55). Using EM,
5.8% of the cognitive variables were imputed in the alcoholism group. Residual scores

were created for the Stroop and Trails, and standardized z-scores were created for each

86

task based on the mean and standard deviation of the alcoholism sample control group
(i.e., individuals with no lifetime disorders).

Statistical Analyses. Different statistical approaches were used to examine the
explanatory power of the four models hypothesized to account for the extensive
relationship between psychiatric disorders and EF. With the Componential Model it was
hypothesized that deﬁcits in some, but not all, components of EF would be associated
with certain disorders. This was tested using proﬁle analysis, essentially a repeated
measures multivariate analysis of variance (MANOVA) except all observations are taken
at a single time point. This allows one to test for differences in the patterns of
performance across multiple tests (Atchinson, Bradshaw, & Massman, 2004; Tabachnick
& Fidell, 1996). Multiple regression analyses were used to examine the Comorbidity-
Speciﬁcity Model, or the possibility that only one or a few disorders were relatrxl to EF
deﬁcits. Both multiple regression analyses and univariate ANOVA tests were used to
examine whether number of disorders was important to neuropsychological test results
for the Comorbidity-Nonspecificity Model. Finally, structural equation modeling with
latent factors was used to examine the relative relationships between general
psychopathology and performance on neuropsychological tests assessing speed and EF
for the Dimension-Speciﬁcity Model.

Note that the primary focus for all analyses was upon lifetime diagnoses.
However, to examine whether there were differential effects, all analyses were repeated
to assess current disorders (i.e., disorders that were present when neuropsychological
testing was conducted for the ADHD sample, and within the past three years for the

alcoholism sample). Disorders were coded as being present or absent (i.e., 0 or 1). As

87

well, most analyses were conducted with an EF composite variable (consisting of Trails
Residual, Stroop Residual, Response Inhibition, RT Variability, and WCST perseverative
errors), a speed composite variable to test for differential cognitive effects (consisting of
Trails A, Stroop Word, and Stroop Color), and the individual EF tests.

The AMOS 5.0 (2003) statistical package, using the Maximum Likelihood
method, was employed for all latent variable and structural analyses. For analyses
relying on SEM analyses, multiple ﬁt analyses are reported and interpreted as outlined by
Kline (2004): (1) Pearson chi-square for which nonsigniﬁcant values signify good ﬁt, and
a xz/df ratio < 3 is acceptable; (2) Goodness of Fit Index (GFI; J oreskog & Sorbom,
1981) for which a value > .90 is considered a good ﬁt; (3) Comparative Fit Index (CPI;
Bentler, 1990) for which a value > .90 is considered a good ﬁt; and (4) Root Mean
Square Error of Approximation (RMSEA; Steiger, 1990) for which a value of .08 is
considered acceptable and .05 is considered good (lower is better). The SPSS 15.0
(2006) statistical package was used to perform all other statistical analyses.

Post-hoe power analyses were based on calculations provided by the power
program G*POWER 3.0.3 (Faul, Erdfelder, Lang, & Buchner, in press). A p value of .05
was used for all calculations. With a sample size of 641 , the present study had adequate
power to detect small effects (f = 0.10; f2 = 0.02) for all analyses. Power was greater than
.90 to detect a signiﬁcant overall omnibus F in the repeated measure MANOVA analysis
for both the within and between factor effects with seven comparison groups, and greater
than .77 for the test of the interaction. For regression analyses, power was greater than
.77 to detect a signiﬁcant overall R2 in the omnibus test with one to six predictors for a

small effect. The power to detect effects for the omnibus ANOVA F-test with four

88

groups was greater than .90. Power for medium effects exceeded .90 for all analyses.
For the structural model, Kline (2004) recommended that the ratio of number of cases to
free parameters should be 20: 1, but that a 10:1 ratio is more realistic and adequate for
power in SEM analyses. The largest structural model comparing the relative
relationships between psychiatric disorders and cognitive test performance contained 66
parameters, 44 of which were freely estimated, so the ratio of sample size to free
parameters exceeded Kline’s (2004) recommendations.
Sample Description

Data were merged between the two groups to create one large sample.
Demographic data are presented in Table 2. As is apparent, the samples differed on their
mean age and F SIQ score, as the ADHD sample was younger and had a higher mean IQ
score. The distribution of males and females was equal between the samples, and
although there was a trend towards the ADHD study participants having more years of
education than the alcoholism group, this difference did not reach signiﬁcance. As is
expected with an older sample (and due to the family-based recruitment strategy), a larger
proportion of the alcoholism sample reported that they were married or living with a
partner. Average personal income in the ADHD sample was $26,987. Given the young
nature of this sample, reported parental income is also informative as to SES (M=$72,000
(SD=$25,592)). Mean reported family income in the alcoholism sample at the time of
testing was $58,677 (SD=$26,355). Note that this represents a considerable improvement
over family income at the beginning of the study ($35,649 (SD=16,967)). Thus, the
ADHD sample was from a slightly higher SES, but both samples included considerable

variability in income levels.

89

The distribution of lifetime and current disorders in the two samples are presented
in Tables 3 and 4. Given the multiple differences in sample characteristics, particularly
age, SES, study recruitment objectives, and inclusion/exclusion criteria, differences in the
rates of disorders were not unexpected. In general, the alcoholism sample was a more
disordered population. For individual lifetime disorders, the alcoholism group had higher
rates of MDD, dysthymia, speciﬁc phobia, alcohol dependence, conduct disorder, and
ASPD, while the ADHD group had higher rates of ADHD and ODD. No differences
were seen for lifetime rates of GAD, agoraphobia, PTSD, OCD, or drug dependence. For
composite “any depression” and “any anxiety” variables, the alcoholism group had a
greater frequency of both types of disorders. Taken together, combining the samples
provided large distributions of a range of different disorders, particularly when using
composite variables.

The number of individuals who met criteria for any type of psychopathology near
the time of testing did not differ between the samples. Differences were nonetheless
noted in distribution of individual disorders between samples, as the alcoholism group
had higher rates of MDD (a considerable difference because current MDD was an
exclusionary criterion in the ADHD group), and a trend towards higher rates of
alcoholism. The ADHD group demonstrated higher rates of dysthymia, GAD, and
ADHD. This latter difference was striking because adult ADHD was not evaluated in the
alcoholism sample. Note that the variables for, and thus rates of, current and lifetime
ASPD are the same because it is considered a lifetime disorder. For composite variables,
the alcoholism group had higher rates of any depressive disorders, and the ADHD group

had higher rates of any anxiety disorders. For the combined group, there were relatively

90

high rates of most disorders when considering composite depressive and anxiety
disorders, although the rate of drug dependence was very low. It is likely that individuals
who were heavily involved with drugs at the time of testing able to commit to either
study. Therefore, there were many differences across the samples in distributions of
speciﬁc disorders for both lifetime and current diagnoses. However, this was expected
based on differences in individual study objectives and thus sample characteristics.

Cognitive test performance also differed across the two samples on many tasks
(see Table 5). Again, however, this was expected based upon differences in age, IQ, and
SES across the populations. Note that the means for Stroop Residual and Trails Residual
did not differ, despite differences in Stroop Color, Stroop Color-Word, and Trails A and
B, since the residuals were created separately in the two samples and sample-dependent
differences were effectively eliminated. Differences were also seen for RT Variability
and WCST perseverative errors. In all cases, the ADHD participants’ performance was
better than the alcoholism participants.

The mean cognitive test performances associated with the different lifetime and
current disorders in the combined sample are presented in Tables 6 and 7. Note that these
means are for overlapping diagnostic groups, and individuals in these groups may have
comorbid pathology. Correlations amongst the tasks included in this study are provided
in Tables 8 through 10 for the combined, alcoholism, and ADHD samples, respectively.
The sample is further characterized in correlation tables detailing relationships amongst

lifetime and current disorders in Tables 11 through 16.

91

Checks on the Validity of Combining the Alcoholism and ADHD Samples

The combination of two separate samples in the present study raised concerns
about conducting and interpreting analyses. While data were combined in order to obtain
a large sample with a range of disorders to conduct meaningful analyses, it clearly raises
concerns about the above-noted differences in the samples, and how those may contribute
to results. One means to deal with this issue is to control for “sample” in all analyses.
However, the two samples have very different distributions of disorders, and controlling
for the sample of origin may result in unintentionally controlling for presence of
disorders; in other words, controlling for the variable of interest. Therefore, alternative
means were implemented to control for sample differences in this study. These included:
(1) controlling for variables that differed notably between the samples in all analyses (i.e.,
FSIQ and age); (2) where appropriate, standardizing scores separately in the two samples
based on the mean of the individual samples’ control groups (i.e., individuals with no
disorders)‘; and (3) conﬁrmatory factor analyses (CFAs) were used with single- and
multiple-group analyses to examine the stability of the covariance structures between the
two samples.

CFAs were conducted on all composite/latent variables (e. g, EF, processing
speed, depression, anxiety disorders, childhood externalizing disorders, etc.; see results

below) to evaluate their structure in the combined and individual samples in order to

 

' Certain analyses required that test performance be standardized into a common metric (e. g., proﬁle
analysis, creation of EF composite score). When this was required (i.e., MANOVA for Analysis 1, linear
regressions for Analyses 2 and 3), participants’ performances on each of the tasks of interest were
standardized separately in the two samples to reduce sample-dependent differences. Therefore, participants
ﬁ’om the ADHD sample were standardized on the means and standard deviations of controls within the
ADHD sample, and participants in the alcoholism sample were standardized on controls within the
alcoholism sample. By standardizing these EF scores within the individual samples, sampledependent
differences were removed to ensure that test performance was not solely based upon differences in sample.
Similarly, residual variables for the Stroop and Trails EF measures were created individually in the separate
samples in order to reduce sample-based discrepancies.

92

justify their inclusion in structural analyses, as well as to provide a basis for creating
composite variables for other statistical analyses. Variables were ﬁrst analyzed with the
samples combined to determine a best-ﬁtting baseline model. Follow-up analyses were
then conducted using single- and multiple-group conﬁrmatory factor analyses to ensure
that there was measurement invariance, or equivalence in the covariance structure, for
latent variables between the alcoholism and ADHD samples. In other words, given that
two different samples were included in this study, these analyses evaluated the legitimacy
of interpreting combined analyses (Kline, 2004).

Therefore, in conﬁrmatory factor analyses, the following was conducted for each
latent variable: (1) Combined-group analyses were performed to determine a best-ﬁtting
baseline model using model ﬁt for each of the proposed latent/composite variables; (2)
Single-group analyses were performed in which the best-ﬁtting baseline model was
analyzed separately in the alcohol and ADHD groups and model ﬁt was evaluated; and
(3) Multiple-group analyses were performed in order to evaluate the measurement model
in both groups simultaneously with varying levels of cross-group equality constraints.
For the multiple-group analysis, if there was no signiﬁcant difference in ﬁt (as
determined by the difference in )8) of an unconstrained model to those with equality-
constrained loadings, then the indicators were judged to assess the factors comparably in
each group; conversely, signiﬁcant loss of ﬁt would suggest that group membership
moderated the relations speciﬁed in the model (Kline, 2004). Therefore, CFAs were
conducted prior to hypothesis testing in order to validate composite variables, evaluate
the structure of relationships within combined and separate samples, and determine the

legitimacy of interpreting analyses with the samples combined.

93

Conﬁrmatory Factor Analyses to Establish Latent Variables for Psychiatric Disorders
and Cognitive Tests

The structural model for psychopathology was initially based on a three-factor
hierarchical structure similar to that found by Krueger (1999) to explain patterns of
comorbidity. This model suggests that psychiatric disorders fall along two main
dimensions associated with internalizing (which included two sub-factors) and
externalizing symptoms. A variant of this model was replicated within the present study
to determine whether a similar pattern of comorbidity existed. The initially hypothesized
model with relations between disorders is presented in Figure 1. Some notable
differences between the model in Figure 1 and that presented by Krueger (1999) reﬂect
changes in the diagnostic criteria (he used DSM-III-R) and the inclusion of additional
disorders in the present study. For instance, Krueger (1999) did not include PTSD, OCD,
or the childhood externalizing disorders (i.e., ADHD, ODD, CD) in his study, and he did
include panic disorder but it was not be included in the present study as it could not be
diagnosed retrospectively in the alcoholism group. CFAs were used to determine
whether the model based on Krueger’s ﬁndings also represented the structure of
internalizing and externalizing DSM-IV disorders in the present study.

Internalizing disorders. As shown in Figure l, a hierarchical latent model based
on Krueger’s ﬁndings for internalizing disorders was initially tested. It included two
latent subfactors, one for “anxious-misery” (i.e., GAD, MDD, dysthymia), and another
for “fear” (i.e., social phobia, simple phobia, agoraphobia, OCD, and PTSD). In
Krueger’s model, the two latent factors (anxious-misery and fear) were subfactors of the

internalizing factor, so these factors were correlated in the present analyses. However,

94

this model was untenable, as it would not converge appropriately (i.e., the covariance
matrix was not positive deﬁnite) and estimates could not be interpreted. This occurred
even when the model only included the same disorders as Krueger (1999) included in his
internalizing model.2

To better understand the relationships in the data, an exploratory factor analysis
using principal components with varimax rotation was conducted on the remaining
internalizing disorders (i.e., GAD, social phobia, speciﬁc phobia, OCD, PTSD, MDD,
and dysthmia). Factor cutoffs were based on eigenvalues greater than or equal to 1.0.
Variables were considered to load signiﬁcantly on a factor when factor loadings exceeded
0.4 in magnitude. Two factors were found, comprising a total of 40.5% of the variance.
The ﬁrst and largest factor (eigenvalue = 1.72) included social phobia, OCD, and GAD,
while the second factor (eigenvalue = 1.12) included MDD, dysthymia, and PTSD.
Speciﬁc phobia did not load on either factor, but when a three factor solution was
allowed the third factor (eigenvalue = .98) included only speciﬁc phobia.

Therefore, the relationships amongst disorders in the present study differed
considerably from those found by Krueger (1999), as well as Watson (2005), both of
whom found that GAD loaded signiﬁcantly on a factor with MDD and dysthymia (as
well as PTSD when it was included in analyses (Cox, Clara, & Enns, 2002). The
inconsistency in ﬁndings may be due to differences in the nature of the samples included

in previous studies compared to this one (see Watson, 2005 for a full review of

 

2 In an attempt to improve the model, agoraphobia was removed from analyses due to extremely low factor
loading suggesting poor ﬁt ([3 = .08, p > .05), and small sample size (n = 8). There was a small negative
relationship between agoraphobia and Trails Residual (r = -.O8), but correlations with other cognitive
variables were not signiﬁcant (r = -.04 to .05), so dropping this variable was unlikely to affect later
analyses. However, the model remained untenable and did not converge appropriately (i.e., covariance
matrix was not positive deﬁnite).

95

limitations of diagnosis-based analyses). Speciﬁc issues that may have affected
relationships in the present study compared to Krueger’s (1999) report included the two-
sample design with differences in inclusion and exclusion criteria in the two samples,
resulting in differences in the distributions of disorders. Further, this may be a more
affected sample due to the clinically-based nature of the research questions. Importantly,
when Krueger (1999) reanalyzed his data within a treatment-seeking subsample, he found
that internalizing disorders were best described by a single latent factor. The present
study likely is most reﬂective of a treatment-seeking or clinical sample, in that there was
an emphasis upon recruiting speciﬁc disordered populations in both the alcoholic and the
ADHD samples, along with matched controls. The rates of disorders therefore obviously
do not match the population.

Other issues that may have affected relationships amongst internalizing disorders
included the changes in the diagnostic exclusions and criteria across the different editions
of the DSM that were used in the retrospective diagnosis of disorders in the alcoholism
sample. As well, the presence of multiple symptom dimensions in many of the anxiety
disorders tends to affect correlations within and between syndromes and makes it difficult
to create an adequate taxonomy for structural analyses (Watson, 2005). In particular,
PTSD did not ﬁt clearly into the present results in that it loaded with the depressive, not
anxiety, disorders in exploratory factor analyses, contrary to prior studies. Yet overall, the
appropriate conceptual placement of PTSD amongst other disorders remains somewhat
unclear (Watson, 2005). Theoretically, the appropriate placement of PTSD is
problematic, as it requires an external experience and is “therefore a less meaningful

indicator of a latent, endogenous ‘core psychopathological process’” (Krueger, 1999, p.

96

922). Therefore, Krueger (1999) did not include PTSD in his study, although a follow-up
re-analysis of his data suggested that PTSD loaded only weakly onto the general
“anxious-misery” factor (Cox et al., 2002).

Given these considerations, PTSD was dropped from the model and the remaining
internalizing disorders were analyzed as a single factor, which provided the best
representation of the current data. Because the exploratory factor analysis demonstrated
that MDD and dysthymia loaded together, their errors were correlated. The resultant
model is presented in Figure 3. The model ﬁt well (x2 (8) = 8.30, p > .05, GFI = .996,

CF I = .998, RMSEA = .008), and all factor loadings were signiﬁcant.

Running this model in the two samples individually demonstrated that model ﬁt
was acceptable for both samples (Alcohol: 112 (8) = 3.83, p > .05, GFI = .997, CFI = 1.00,
RMSEA = .00; ADHD: x2 (8) = 14.40, p > .05, GFI = .98, CFI = .88, RMSEA = .07),
although the model more accurately depicted the relationships in the alcoholism than in
the ADHD goup. When the multiple-goup analysis was conducted, there was a
signiﬁcant difference between the baseline, freely estimated model in the combined
sample and the constrained model (x2 (22) = 48.17, p < .01, GFI = .98, CF I = .82,
RMSEA = .04; x2 difference (14) = 39.87, p < .01 ). In the individual goups, the greatest
discrepancy in loadings appeared to be for the “speciﬁc phobia” indicator since it reached
signiﬁcance in the alcohol goup (r = .25), but not in the ADHD goup (r = .02).
Therefore, this loading was allowed to be freely estimated (x2 (21) = 39.97, p < .01, GFI =
.98, CFI = .87, RMSEA = .04; change in {(13) = 31.67, p < .01). The models were still
signiﬁcantly different between samples, so another constraint was released for the “social

phobia” factor loading but the ﬁt remained sigriﬁcantly different from baseline (752 (20) =

97

32.39p < .05, GFI = .99, CFI = .92, RMSEA = .03; change in x2 (12) = 24.09,p < .01).
Therefore, the factor loading for dysthymia was also released, and the difference from the
baseline unconstrained model was no longer sigriﬁcant (x2 (19) = 29.53, p > .05, GFI =
.99, CFI = .93, RMSEA = .03; change in 78(11) = 15.79, p > .05).

Thus, three out of six factor loadings had to be released to reach model ﬁt
equivalence between the samples, although ﬁt was quite good after releasing two
loadings and chi square difference is heavily susceptible sample size. This suggests that
there were differences in the structure of relations between the ADHD and alcoholism
samples. However, given the differences in recruitment and inclusion/exclusion criteria
in the two samples that contributed to differences in distributions of internalizing
disorders (see Table 3), along with the above-described problems associated with
analyzing the structure of anxiety disorders, this was a somewhat expected and was
viewed as an acceptable and interpretable result. Therefore, the best-ﬁtting baseline
model for internalizing disorders was retained and considered to be interpretable in SEM
analyses with the combined sample. Note, however, that separate composite variables
were used for “any depression” and “any anxiety” in all non-SEM analyses, with the
latter also including agoraphobia and PTSD to best reﬂect the full range of anxiety
disorder diagnoses in the sample (results were the same regardless of which anxiety
composite was included).

Externalizing disorders. A latent variable for externalizing disorders was tested
that included alcohol dependence, drug dependence, ASPD, and child externalizing

disorders (i.e., ADHD, CD, and ODD). The model is presented in Figure 4, and ﬁt

98

statistics suggested that it described the relationships in the model well (x2 (2) = 8.32, p <
.05; GFI = .99, CFI = .95, RMSEA = .07).

When the model was tested in the individual samples, model ﬁt was good and all
factor loadings were signiﬁcant (alcohol: 752 (2) = 5.60, p > .05, GFI = .99, CFI = .98,
RMSEA = .06; ADHD: x2 (2) = 1.27, p > .05, GFI = .997, CFI = 1.00, RMSEA = .00). A
multiple—goup analysis with all factor loadings constrained was not sigriﬁcantly
different from the baseline, combined sample model ()52 (7) = 14.44, p < .05, GFI = .99,
CFI = .96, RMSEA = .04; change in x2 (5) = 6.12, p > .05), suggesting that the model
could be interpreted in the combined sample.

Cognitive tasks. The latent variable including non-standardized scores for all of
the EF tasks (i.e., WCST, Stroop residual, Trails residual, Response Inhibition, and RT
Variability, with the two latter variables’ errors correlated due to their dependency) ﬁt
well and all factor loadings were signiﬁcant (712 (4) = 1.50, p > .05; GFI = .999, CFI =
1.00, RMSEA = 0.00). This model is presented in Figure 5. Non-standardized scores
were used because this analysis depends upon lack of equality in covariance structures.
The model ﬁt well in the individual samples (alcohol: )8 (4) = 4.20, p > .05; GFI = .996,
CF I = .999, RMSEA=0.01 although Response Inhibition was not signiﬁcant at .16;
ADHD: x2 (4) = 11.47, p < .05; GFI = .98, CFI = .85, RMSEA=0.01 and all factor
loadings were sigriﬁcant). F it for the multiple-goup analysis with all factor loadings
constrained was signiﬁcantly different from the freely-estimated baseline model (x2 (12)
= 19.50, p > .05; GFI = .99, CFI = .97, RMSEA=0.03; x2 difference (8) = 18.00, p < .05).
The difference in model ﬁt remained signiﬁcant even after removing one or two of the

constraints. Therefore, it appears that the difference between the goups was due to

99

differences in the structure of relationships due to consistent sample/methodological
differences that could not be easily overcome. As will be demonstrated below, however,
stability and equivalence were improved in a larger measurement model that also
included processing speed tasks.

Fit statistics for a latent variable including all of the speed tasks (i.e., Stroop
Color, Stroop Word, Trails A) could not be tested since the model was just-identiﬁed.
However, model ﬁt was adequate for a larger measurement model that correlated the EF
and processing speed latent variables (x2 (1 8) = 93.21, p < .05; GFI=.97, CFI=.92,
RMSEA=0.08). The model is presented in Figure 6.

Individual sample analysis of the EF and speed measurement model in Figure 6
demonstrated that ﬁt was adequate in both goups (alcohol: 752 (18) = 73.97, p < .05; GFI
= .96, CFI = .92, RMSEA = 0.08; ADHD: x2 (18) = 73.97, p < .05; GFI = .96, CFI = .92,
RMSEA = 0.08). However, it should be noted that the factor loading for Stroop Residual
on the EF variable was not signiﬁcant in either the ADHD or the alcohol goup. Despite
this, the task was retained in the model since this factor loading was signiﬁcant in the
combined sample, which was the focus of the present study. Therefore, taken together,
the model appeared to ﬁt adequately and similarly in both individual samples. This was
conﬁrmed by the multiple-goup analysis with all factor loadings constrained, as the ﬁt
was not signiﬁcantly different from the baseline, freely-estimated model (x2 (42) =
111.76, p < .05; GFI = .96, CFI = .92, RMSEA = 0.05; change in x2 (24) = 37.79, p >
.05). A more stringent test of sample equivalence was then conducted, and the
correlation between the EF and Speed factors was constrained in the multiple-goup; this

model was also not signiﬁcantly different from the baseline analysis ()6 (43) = 112.62, p

100

< .05; GFI = .96, CFI = .93, RMSEA = 0.05; change in )6 (25) = 19.41, p > .05). Taken
together, these ﬁndings suggest that the measurement model including both EF and speed
may be interpreted with the two samples combined.

Therefore, these preliminary stage analyses suggested that including both the EF
and speed latent variables together in a measurement model provides stability for each
individual cognitive factor, and thus results may be interpreted with the combined
samples. This sample equivalence could not be adequately achieved with the EF latent
factor alone, even after releasing a number of the constraints, suggesting there were
signiﬁcant structural differences moderated by goup membership even though the EF
factor ﬁt well in each of the individual samples and in the combined sample. Since
combined-sample analyses could not be clearly interpreted with only the EF factor, all
structural analyses included the full cognitive measurement model (i.e., both EF and
speed factors correlated), while analyses focusing on only one or the other cognitive
domain included manifest variables (i.e., standardized composite scores). With these
decisions about measurement models resolved, the analysis proceeded to test the four
models described in the hypotheses on page 60.

Tests of Hypotheses
Analysis 1 : Testing the Componential Model for Disorder-Speciﬁc Patterns of EF Test
Performance

Proﬁle analysis was used to test the hypothesis that there would be differences in
the pattern of deﬁcits between different disorders. In other words, this assessed the
possibility that different disorders were associated with deﬁcits in different EF processes.

This allowed for an examination of goup (i.e., disorder) differences on test performance

101

across Stroop Residual, Trails Residual, Response Inhibition, RT Variability, and WCST
perseverative errors. Disorders included were “any depression,” “any anxiety,”
alcoholism, drug dependence, ASPD, and childhood extemalizing/adult ADHD. Proﬁle
analysis using repeated measures MAN OVA is superior to using a series of univariate
analyses on individual tests because it allows a test of signiﬁcant patterns of performance
across multiple tests (Atchinson et al., 2004; Tabachnick & Fidell, 1996).

For this study, goup was entered into the analysis as a between-subjects factor,
and EF tests were entered as the within-subject factors. Proﬁle analysis tests three
separate aspects of the data: levels, ﬂatrness, and parallelism. Differences in the levels of
the proﬁles would suggest overall performance differences between disorder goups, and
were followed up with post-hoe Tukey tests. F latness was tested but was of less interest
for present purposes because it simply assessed whether the goups combined performed
differently on the EF tests (i.e., differences in the means of the EF tests), which were
interpreted by examining mean differences. Deviations from parallelism in the disorder
proﬁles were the main focus of this analysis; these would be suggested by a signiﬁcant
interaction of goup and test, and were interpreted with follow-up univariate ANOVAs
and Tukey tests. The interaction analysis provided the ability to determine whether
disorder goups performed differently across different components of EF processes.

Multivariate analysis of covariance (MANCOVA) was used to assess how the
effects of possible moderating variables such as age, gender, and IQ affected the pattern
of performance or interaction between disorder goup and test performance. In order to

make comparisons across tests, all scores were transformed into a common metric using

102

z-scores based on the mean and standard deviation of the control goups in the individual
samples.

Individuals were assigned to discrete, non—overlapping goups for analysis based
on their status for disorders/classes of disorders (i.e., controls, individuals with various
single disorders, and individuals with multiple disorders). For lifetime analyses, the
disorder variable deﬁned individuals as follows: controls (n = 170), any depression only
(n = 58), any anxiety only (n = 33), alcoholism only (n = 61), childhood externalizing
disorders only (n = 46), and comorbid disorders (n = 269). For current analyses
participants were deﬁned as current: controls (n = 360), any depression only (n = 61), any
anxiety only (n = 30), alcoholism only (n = 16), ASPD only (n = 36), adult ADHD only
(n = 49), and comorbid disorders (n = 88). Lifetime ASPD (n = 4) and current/lifetime
drug dependence (n = 0; n = 1) could not be included in analyses as few individuals met
criteria for only these diagnoses. Figures 7 and 8 present the standardized mean scores
for the EF tests for the non-overlapping lifetime and current diagnostic goups included
in this analysis. For informational purposes, the mean EF scores for the overlapping
diagnostic goups (i.e., individuals could meet criteria for multiple disorders, so a single
individual could be included in multiple disorder goups) are in Figures 9 and 10.

Lifetime analyses demonstrated that the main between-subject effect of goup was
signiﬁcant (F (1, 5, 631) = 3.22, p < .01). Post-hoe Tukey tests indicated that the anxiety
disorder goup performed signiﬁcantly better than both the child externalizing goup (p <
.05) and the comorbid goup (p < .05) across EF tests. The main within-subject effect of
test (F (4, 628) = .98, p > .05) and the test-by—disorder interaction were not signiﬁcant (F

(20, 2084) = 1.00, p > .05). These results suggest that for lifetime disorders there are

103

differences in overall EF test performance between certain disorders (but not signiﬁcantly
different from controls). However, the pattern of performance is not signiﬁcantly
different across tests based on disorder status. Results remained the same after covarying
age, gender, and IQ.

For current diagnoses, the main effect of disorder status was signiﬁcant (F (1, 6,
633) = 4.61, p < .01). Post-hoc Tukey tests indicated that there were signiﬁcant
differences between the control goup and the ADHD and ASPD goups (p < .05), and
the anxiety disorder goup and the alcoholism, ASPD, and ADHD goups (p < .05). The
main within-subjects effect of test was signiﬁcant (F (4, 630) = 3.70, p < .01), which
simply means that there were differences in the EF test means. Of geater interest,
however, was the signiﬁcant test-by-disorder interaction (F (24, 2199) = 1.63, p < .05).
Here, the quadratic and order-4 within-subj ects contrasts were signiﬁcant (quadratic: F
(6) = 2.30, p < .05; order-4: F(6) = 3.06, p < .01), suggesting there were quadratic and
quartic trends in the data. The curved proﬁles with poorer performance on Trails
Residual, Response Inhibition, and RT Variability for alcoholism, ADHD, and ASPD in
Figure 8 depict this relationship. Therefore, this analysis found differences in EF results
due to current disorder status and speciﬁc test, and the differences across tests varied
based upon goup membership.

Univariate ANOVAs with post-hoe Tukey analyses were conducted for each test
to better understand the test-by-disorder interaction. There were signiﬁcant goup (i.e.,
disorder) differences on Trails Residual (F (6, 633) = 2.78, p < .01), Response Inhibition
(F (6, 633) = 3.88, p < .01), and RT Variability (F (6, 633) = 2.68, p < .05). Follow-up

Tukey tests indicated that participants with ADHD performed signiﬁcantly worse than

104

those with anxiety disorders on Trails Residual (p < .05). On Response Inhibition,
alcoholics performed signiﬁcantly worse than controls, anxiety disordered, and comorbid
participants (p < .05). No individual comparisons reached signiﬁcance for RT
Variability, but there was a trend for the ADHD goup to perform more poorly than
controls (p = .07). Although differences did not reach signiﬁcance, the ASPD goup also
tended to perform more poorly than other goups on these three measures, while
depressed and, surprisingly, comorbid participants tended to perform similarly to
controls. When age, gender, and F SIQ were covaried in MANCOVA, the main effects of
disorder and test remained signiﬁcant, but the test-by-disorder interaction was no longer
signiﬁcant. Thus, these covariates appeared to account for the interaction, and when
controlled support for the componential model is reduced.

Summary of analysis for model 1. Partial support was provided for a
componential model in the proﬁle analysis, but only with current diagnoses, where
participants with ADHD and alcoholism demonstrated different patterns of performance
across EF tests from some other goups. Current ADHD was associated with poorer
Trails Residual and RT Variability (the latter was a trend), and current alcoholism was
associated with deﬁcits in Response Inhibition. RT Variability deﬁcits in the ASPD
goup approached, but did not reach, signiﬁcance. Support for the componential model
disappeared when age, gender, and FSIQ were covaried in current analyses. Differences
in global EF performance between disorders and between tests remained signiﬁcant, but
these did not interact. The componential model did not receive support in lifetime
diagnoses, as the pattern of performance across EF tests did not differ based on disorder

status, although there were differences in overall EF performance between individuals

105

with anxiety disorders and those with childhood externalizing or comorbid disorders. In
general, comorbid disorder presentation was not related to geater EF impairment.
Overall, the weak support for the componential model in the proﬁle analysis suggested
that this model failed to adequately explain the relationship between psychiatric disorders
and EF. Support for disorder-speciﬁc effects was suggested, however, and with these
results in mind, the following analysis further elucidated the effects of disorder status
upon global and speciﬁc EF test performance.
Analysis 2: Testing the Comorbidity-Specificity Model that Only One or a Few Disorders
Are Uniquely Associated with EF Deﬁcits

Multiple regession analyses were conducted to examine whether only certain
disorders were associated with EF deﬁcits. Then, high levels of comorbidity amongst
disorders may account for the appearance of more widespread deﬁcits. Predictors were
dummy-coded variables for presence/absence of each disorder (i.e., any depression, any
anxiety disorder, alcoholism, drug dependence, ASPD (lifetime), and childhood
externalizing disorders/adult ADHD). Predictors were entered in a single block, with the
outcome variable being the composite EF score. This allowed for an examination of the
unique contribution of individual disorders to EF test performance. To assess for
differential effects with other cognitive processes, this regession was repeated with the
processing speed composite variable and individual EF tests. The effects of age, gender,
and IQ were covaried.

The results of the linear multiple regession analysis for lifetime diagnoses and EF
composite are presented in Table 17. As can be seen, only childhood externalizing

disorders signiﬁcantly predicted tlne EF composite score, with presence of a childhood

106

externalizing disorder being related to poorer performance on EF tasks, and all lifetime
disorders together accounting for 3% of the variance in the EF composite score. To
better understand the contributions of individual disorders to this result, this analysis was
repeated with the childhood externalizing disorders separated into CD, ODD, and ADHD.
Here, only ADHD was signiﬁcantly predictive of EF composite score (B = -.20, p < .01),
suggesting that ADHD accounted for the predictive power of the childhood externalizing
disorders. Sinnilar results were found for current disorders, as only adult ADHD
signiﬁcantly predicted poorer performance on EF tests (see Table 18). Together, all of
the current disorders accounted for 4% of the variance in EF performance.

While age and IQ were also signiﬁcant predictors of EF composite score (B = .10,
p < .05 for age; B = -.35, p < .01 for IQ; B = -.07, p = .06 for gender with males trending
towards better performance), these results remained the same for lifetime disorders even
after age, gender, and IQ were controlled (B = -.14, p < .05 for lifetime childhood
externalizing; B = -.23, p < .01 for lifetime ADHD). The same results were seen for
current disorders, where both gender (B = -.09, p < .05, with males performing better) and
IQ (B = .36, p < .01) were signiﬁcant predictors, but adult ADHD remained signiﬁcant (B
= -.21, p < .01) even after controlling for age (B = .08, p = .08), gender, and IQ.
Therefore, for botln lifetime and current analyses, only ADHD was signiﬁcantly related to
performance on EF tests. Bivariate correlations also showed that only ADHD was related
to EF composite (results not shown).

Regession analyses were also performed with the individual tests, using the
standardized test performance. There were no signiﬁcant predictors for Stroop Residual.

Of the lifetime disorders, only childhood externalizing disorders signiﬁcantly predicted

107

poorer performance on Trails Residual (B = .16, p < .01), RT Variability (B = .16, p <
.01), and approached signiﬁcance for WCST (B = -.08, p = .06). Breaking the childhood
externalizing disorders down into the individual disorders (ADHD, ODD, and CD)
demonstrated that ADHD was accounting for tlnese relationships. Anxiety disorders
signiﬁcantly predicted Response Inhibition score (B = -.11, p < .01) and approached
signiﬁcance with RT Variability (B = -.08, p = .06), predicting better performance on
these measures; when broken into the individual disorders no one anxiety disorder was
signiﬁcantly related to either measure. When F SIQ, age, and gender were added to the
regession equation, childhood externalizing disorders remained signiﬁcantly related to
Trails Residual and RT Variability, although the relationship with WCST was no longer
signiﬁcant, and anxiety remained a signiﬁcant predictor of better Response Inhibition
score (results not shown). Surprisingly, alcoholism became a signiﬁcant predictor of
better Stroop Residual score after covarying these variables (B = .09, p < .05).

For current disorders, regession analyses suggested that ADHD signiﬁcantly
predicted Trails Residual (B = .13, p < .01), Response Inhibition (B = .11,p < .01), and
RT Variability (B = .19, p < .01). Alcoholism signiﬁcantly predicted Response Inhibition
(B = .08, p < .05) and RT Variability (B = .09, p < .05), and ASPD approached
signiﬁcance in predicting Trails Residual (B = .07, p = .07). No disorders signiﬁcantly
predicted Stroop or WCST. After controlling for age, F SIQ, and gender, ADHD
remained a signiﬁcant predictor of Trails Residual, Response Inlnibition, and RT
Variability. Alcoholism no longer signiﬁcantly predicted Response Inhibition but

predicted RT Variability, and ASPD demonstrated a trend towards better WCST

108

performance (B = .08, p = .06) while anxiety demonstrated a trend towards better
Response Inhibition (B = -.07, p = .08).

Finally, regession analyses were repeated with the speed composite score (see
Tables 19 and 20). For lifetime disorders, alcohol dependence and childhood
externalizing disorders (shown to be predominantly ADHD (B = -.10, p < .05) when
broken into component parts) were signiﬁcantly related to speed. When age (B = -.08, p
= .09), gender (B = -.17, p < .01 with males being faster), and IQ (B = .33, p < .01) were
controlled, only childhood externalizing disorders (i.e., ADHD) remained signiﬁcant.
For current disorders, only ASPD signiﬁcantly predicted slower speed, although this
result appeared to be due to sample-related issues as only adult ADHD signiﬁcantly
predicted slower speed after controlling for age (B = -.06, p > .05), gender (B = -.18, p <
.01 with males being faster), and IQ (B = .35, p < .01).

Summary of analysis for model 2. Only current and lifetime ADHD were unique
predictors of poorer performance on global EF measures when controlling for the effects
of other disorders, and these results remained after controlling for age, gender, and IQ.
While global effects were seen, only some of the individual EF tests were related to
ADHD, suggesting that these associations may be driving the overall, global score.
ADHD predicted poorer performance on Trails Residual and RT Variability (current and
lifetime) and Response Inhibition (current). A trend towards more WCST perseverative
errors disappeared after controlling for F SIQ. With regard to individual tests, current
alcoholism was also related to poorer RT Variability and Response Inhibition and a trend
was seen between ASPD and Trails Residual (none of these results remained with

covariates). Lifetime, but not current, anxiety predicted better performance on Response

109

Inhibition and RT Variability before controlling for covariates (the latter was a trend).
Thus, the robust relationship between ADHD and global and individual EF effects
provided strong support for the Comorbidity-Speciﬁcity Model.

The individual test effects also suggested that the Componential Model is
important to the relationship between EF and ADHD. Although weaker, the associations
between alcoholism and individual EF processes also provided support for both models.
However, ADHD was also associated with slower processing speed, suggesting that
cognitive effects are not speciﬁc to the EF domain.

Analysis 3 : Testing the Comorbidity-Nonspecificin Model that Number of Disorders
Predicts Performance on EF Tests

This analysis examined the possibility that number of disorders rather than type of
disorder, contributed to EF deﬁcits. A variable was created to deﬁne individuals with:
(1) no disorders; (2) one disorder; (3) two disorders; and (4) three or more disorders.
Disorders that were included in lifetime analyses were: MDD, dysthymia, GAD, OCD,
PTSD, agoraphobia without parnic disorder, social phobia, speciﬁc phobia, alcohol
dependence, drug dependence, ASPD, ODD, CD, and ADHD. In current-diagnosis
analyses, the following disorders were included in the determination of number of
disorders at the time of testing: MDD, dysthymia, GAD, OCD, PTSD, social phobia,
speciﬁc phobia, alcohol dependence, drug dependence, ASPD (lifetime), and adult
ADHD. Agoraphobia without panic disorder was not included in analyses on current
disorder status because no participants met criteria for agoraphobia near the time of

testing.

110

The effects of number of disorders on EF test performance were assessed using
multiple regession as well as mean difference (i.e., ANOVA with post-hoc Tukey tests)
analyses. Predictor variables included the composite EF variable, the composite speed
variable (to determine whether any effect of comorbidity cut across neurocognitive
domains), and the ﬁve individual EF tests (to determine whether certain tests were more
strongly related to comorbidity than others). Age (B = .18, p < .01), gender (B = -.07, p =
.05), and IQ (B = .34, p < .01) were covaried.

The distributions of numbers of lifetime and current disorders in the present study
are presented in Tables 21 and 22. Individuals who met criteria for three or more
disorders were collapsed into a single goup for analyses. The regession of the EF
composite variable upon number of lifetime disorders was not signiﬁcant (B = -0.052, p =
.193). When considering current disorders, the regression was signiﬁcant (B = -.104, p =
0.008). The relationship was small, but suggested that having more current comorbid
disorders was signiﬁcantly related to poorer performance on EF tasks. However, this
association was no longer signiﬁcant when controlling for age, gender, and F SIQ.

AN OVA also indicated that there was not a signiﬁcant difference in the mean EF
composite score between goups for lifetime disorders (F (3, 637) = .942, p = 0.42). The
omnibus test for current disorders suggested that there was a signiﬁcant difference in
mean EF composite score between the goups (F (3, 63 7) = 2.67, p = 0.047), but none of
the individual mean differences reached signiﬁcance in post-hoe Tukey’s tests.
Therefore, having a geater number of disorders near the time of testing, but not
throughout the lifetime, did appear to be linearly related to poorer performance on EF

tasks. However, there were no signiﬁcant goup differences in the EF composite score.

111

Further, independent samples t-tests demonstrated that simply having any disorder at all
was signiﬁcantly related to performance on EF tests for current diagnoses (t = 2.52, p =
.01) but not for lifetime disorders (t = 1.68, p = .09). These results suggested that,
perhaps, presence of a current disorder was more important than number of current
disorders in EF task performance. Further, presence versus absence of lifetime disorders
in general did not result in signiﬁcant differences in EF test performance.

When considering the individual EF tests, regession results indicated that there
was no predictive relationship between number of disorders across the lifetime and
performance on any of the individual tests (results not presented). AN OVA results
demonstrated no mean differences between goups (results not presented). For current
diagnoses, number of disorders signiﬁcantly predicted Trails Residual (B = .086, p =
.029), while Stroop (B = -.072, p = .070) and RT Variability (B = .076, p = .053)
approached signiﬁcance. WCST (B = -.001, p = .979) and Response Inhibition (B = .055,
p = .166) were not signiﬁcant predicted by number of disorders. No relationships
remained signiﬁcant when controlling for age, gender, and FSIQ. Signiﬁcant mean
differences were observed for Trails Residual (F (3, 63 7) = 3.09, p = .03) and Response
Inhibition (F (3, 637) = 3.96, p < .01). Post-hoc Tukey tests indicated that the difference
for Trails Residual was between participants with no disorders (M = -0.76, SD = 17.67)
and those with three or more disorders (M = 9.51, SD = 24.70), who performed
signiﬁcantly more poorly. For Response Inhibition, a signiﬁcant difference was seen
between individuals with no disorders (M = 241.37, SD = 68.34) and those with one
disorder (M = 262.97, SD = 82.93), with the latter perfomning signiﬁcantly worse.

Therefore, there was a linear relationship between number of disorders and performance

112

on Trails Residual, and mean differences between controls and individuals with three or
more disorders. A mean difference in Response Inhibition time was seen for controls and
individuals with one disorder.

Number of lifetime disorders did signiﬁcantly predict performance on the speed
composite variable (B = -.170, p < .01), as did number of current disorders (B = -.083, p =
.036), although the latter result was no longer signiﬁcant when controlling for age (B = -
.01,p > .05), gender (B = -.19,p < .01), and FSIQ (B = .34,p < .01). The omnibus
ANOVA F -test was signiﬁcant for lifetime disorders (F (3, 63 7) = 6.87, p < .01), with a
signiﬁcant difference between speed task means for individuals with no disorders (M =
0.00, SD = 2.48) and those with one (M = -0.67, SD = 2.34), two (M = -0.75, SD = 2.50),
and three or more disorders (M = -1.22, SD = 2.52). There were no mean differences for
current disorders (results not shown).

Summary of analysis for model 3. Taken together, number of lifetime disorders
was not associated with EF performance. A small but signiﬁcant linear relationship was
seen between frequencies of current disorders and performance upon EF (particularly
Trails Residual and Response Inhibition) and speed tasks; however, differences in age,
gender, and FSIQ appear to account for these relationships. Furtlner, presence of a
current disorder, rather than the number of disorders, appeared to be more important to
EF performance. A more robust and enduring relationship existed between lifetime
disorders and performance on speed tasks, with increasing number of disorders
contributing linearly to slower speed. In all, the Comorbidity-Nonspeciﬁcity hypothesis

that EF deﬁcits were nonspeciﬁcally related to number of disorders was not supported.

113

Analysis 4: Testing the Dimension-Specificity Model that Shared Underlying Dimensions
of Psychopathology are DiﬁferentiaUy Related to Cognitive Task Performance

Latent factors and structural equation modeling (S EM) were used to examine the
relative relationships between general dimensions of psychopathology and performance
on neuropsychological tests assessing speed and EF. Individual latent variables included
in this analysis were validated in the previously presented CFAs.3

The measurement model for the large model examining relationships between
lifetime disorders and performance on EF and processing speed tasks is presented in
Figure 11. Although not perfect, ﬁt for this model was adequate (x2 (127) = 287.37, p <
.01; GFI = .95, CFI = .88, RMSEA = 0.04), particularly considering that the focus of this
model was to analyze speciﬁcally-deﬁned a—priori relationships as opposed to capturing
all possible relationships between variables in the model. In other words, the reduced
model ﬁt indicates that there are relationships within the data that are not represented in
the model, but ﬁt was considered to be adequate to interpret the path coefﬁcients. All
factor loadings for the individual latent variables were signiﬁcant. Correlations between
the externalizing factor and both cognitive domains were signiﬁcant, as well as between
the two cognitive domains (i.e. EF and speed).

The structural model to examine the predictive relationships for lifetime
internalizing and externalizing disorders upon EF and speed latent variables is presented
in Figure 12. Fit was adequate, and ﬁt did not change from the measurement model (x2
(127) = 287.37, p < .01; GFI = .95, CFI = .88, RMSEA = 0.04). The externalizing latent

variable signiﬁcantly predicted both the executive and the speed latent factors. However,

 

3 Note that standardized test scores were no_t used in any of the latent, measurement, or structural models,
because SEM relies upon an analysis of covariance---that analysis would be undermined if these variances
were equalized.

114

the internalizing latent factor did not signiﬁcantly predict either of the cognitive
variables. Therefore, there was a differential relationship between disorder and cognitive
performance for the type of disorder (i.e., internalizing versus externalizing), but not for
the type of cognitive ability (i.e., speed versus executive). Together, the internalizing and
externalizing variables accounted for 8% of the variance in the EF factor and 11% of the
variance in the speed variable. Covariates were also entered into the structural model;
however, the sample size was not large enough to enter age, gender, and FSIQ
simultaneously so each covariate was entered individually. Gender (B = -.01, p > .05
with EF and B = -.09,p = .07 with speed) and age (B = .53,p < .01 with EF and B = .10,p
< .05 with speed) alone did not affect results. After controlling for IQ, only the
relationship between externalizing and speed remained signiﬁcant (B = .12, p > .05 for
EF; B = .26, p < .01 for speed), likely due to the very strong relationship between FSIQ
and EF (B = -.72,p < .01 with EF and B = -.30,p < .01 with speed).

When this same baseline measurement model was assessed with current disorders,
ﬁt was adequate (x2 (127) = 340.84, p < .001, GFI = .94, CFI = .83, RMSEA = .05), but
both dysthymia and ADHD demonstrated poor loadings on the internalizing and
externalizing factors, respectively (B = .06 and B = -.10). The poor dysthymia loading is
likely due to its small sample size, particularly in the alcoholism sample, as all of the
individuals with MDD are in the alcoholism sample. The poor ADHD loading is likely
also due to differences in distributions of disorders between the samples, as adult ADHD
was only evaluated in the ADHD sample, while the other externalizing disorders (i.e.,
alcoholism, ASPD, and drug dependence) were predominantly found in the alcoholism

sample. Therefore, there was almost no overlap between current diagnoses of

115

alcoholism, ASPD, and drug dependence and ADHD. Given the poor ﬁt for these factors
on their respective latent variables, dysthymia and ADHD were dropped from the current
model. There was a signiﬁcant improvement in ﬁt for the measurement model (x2 (97) =
213.07,p < .001, GFI = .96, CFI = .89, RMSEA = .04; change in 752 (30) = 127.77,p <
.01; see Figure 13) and all factor loadings were signiﬁcant. Correlations between
externalizing disorders and both cognitive domains were signiﬁcant, as was the
correlation between EF and speed.

When predictive relationships were examined in the structural model (see Figure
14), there was no change in ﬁt from the measurement model (x2 (97) = 213.07, p < .001 ,
GFI = .96, CFI = .89, RMSEA = .04). Similar to the results with lifetime diagnoses,
externalizing disorders signiﬁcantly predicted poorer performance on both EF and speed
measures. Internalizing disorders were not signiﬁcant predictors of either cognitive
domain. Therefore, again, there was speciﬁcity in the relationship between disorder and
cognitive performance for type of disorder (i.e., externalizing, not internalizing), but not
for type of cognitive task (i.e., externalizing disorders were related to both EF and speed).
Note that the predictive relationships between externalizing disorders and cognitive task
performance with current disorders were stronger than those for lifetime diagnoses.

Results were unchanged when gender was controlled in the model (B = -.03, p >
.05 with EF and B = .07, p > .05 with speed). When age was added as a covariate (B =
.47, p < .01 with EF and B = .11, p < .05 with speed), the relationship between
externalizing disorders and speed remained signiﬁcant (B = .32, p > .05), whereas the
relationship between externalizing and EF approached but did not reach signiﬁcance (B =

.33, p = .05). Similarly, when FSIQ was added as a covariate, the relationship between

116

externalizing and EF was no longer signiﬁcant (B = .14, p > .05), but the relationship
between externalizing and speed remained signiﬁcant (B = .25, p < .05). Again, tlnere
was a strong relationship between FSIQ and EF (B = -.69, p < .01), moreso than with
speed (B = -.27, p < .01). Irnternalizing disorders remained unrelated to either cognitive
domain in all of these covariate analyses.

Summary of analysis for model 4. The externalizing dimension of
psychopathology for both lifetime and current disorders predicted both poorer EF
performance and slower speed, whereas the internalizing dimension was unrelated to
eitlner cognitive test domain. Therefore, dimensional speciﬁcity was apparent with
regards to psychopathology, but not with regard to cognitive domain; that is, there was
not a differential cognitive deﬁcit. However, F SIQ was strongly related to EF, and when
controlled in the present analyses only the relationship between externalizing disorders
and speed remained signiﬁcant.

Separate Sample Analyses

It should be noted that linear regession analyses were conducted in the ADHD
and alcoholism samples separately to determine equivalence of ﬁndings. In the ADHD
sample, only childhood externalizing disorders/adult ADHD were related to poorer
performance on the EF composite (B = -.19, p < .05 for lifetime; B = -.l7, p < .05 for
current) even after including age, gender, and F SIQ. Lifetime drug dependence
approached signiﬁcance (B = -.14, p = .06). In the alcoholism sample, no disorder
goupings were related to EF composite score for lifetime analyses. Interestingly,
however, when childhood externalizing disorders were separated into ADHD, ODD, and

CD, ODD was signiﬁcantly predictive of better EF score (B = .19, p < .01). This unusual

117

result will be further investigated in later studies. Alcoholism approached signiﬁcance in
enurent analyses in the alcoholism sample (B = -.09, p = .08). The small sample size for
ADHD, and the fact that it was only assessed for childhood, likely prevented this disorder
from demonstrating BF effects in the alcoholism sample. Therefore, separate sample
analyses were similar to the ﬁndings from the combined sample, with ADHD
demonstrating robust effects.
Overall Summary

Taken together, these results suggested that there is speciﬁcity in the types of
disorders that are associated with EF deﬁcits. The Comorbidity-Nonspeciﬁcity Model
(Model 3) examining the effect of number of disorders from a non-speciﬁc perspective
was clearly ruled out by the present results. Before controlling for age, gender, and
FSIQ, a number of the other models received support and provided information about the
nature of the speciﬁc relationships. The Dimension-Speciﬁc Model (Model 4) perhaps
best summarizes the overall ﬁndings, with externalizing disorders, and not internalizing
disorders, being related to EF deﬁcits. Aspects of both the Comorbidity-Speciﬁcity
Model and Componential Model were supported as predominantly process-speciﬁc EF
deﬁcits were associated with only ADHD, alcoholism, and ASPD (all externalizing
disorders). Global EF effects appeared to be driven by individual-process weaknesses.

These results are tempered by the fact that most of the cognitive ﬁndings were
non-speciﬁc in nature and crossed neurocognitive domains. Similar impairments were
found for processing speed and psychopathology, and FSIQ appeared to account for
performance on EF tests. When IQ was controlled in analyses, only the global and

individual EF weaknesses associated with ADHD remained. Unlike the EF effects,

118

associations between psychiatric disorders and processing speed generally remained
signiﬁcant even after controlling for FSIQ. Thus, the processing speed deﬁcits were
more robust than EF effects in the present study. This suggests that slow processing
speed may be a general marker of disturbance, whereas EF may be related only to

particular forms of psychopathology when covariates are considered.

119

Discussion
The present study sought to examine the relationship between psychiatric
disorders and EF deﬁcits in six forms of psychopathology that have previously been
associated, in some manner, with neuropsychological EF deﬁcits: childhood-onset
disruptive behavior disorders (i.e., ADHD, CD, and ODD), adult ADHD, alcohol
dependence, drug dependence, ASPD, depression, and anxiety disorders. To better
understand the widespread association between psychopathology and EF, this study
focused on comorbidity amongst disorders and how they contribute to general and
speciﬁc cognitive impairments.
Four different models were tested as potential explanations for this relationship.
The models moved along the diagnostic hierarchy ﬁ‘om individual disorders, to multiple
or comorbid conditions, to dimensional patlnology. Two models appeared to best explain
the relationship between EF and psychopathology: the Dimensional Model and the
Componential Model. Before discussing the signiﬁcance of that conclusion, the results
of each of the main hypotheses will be brieﬂy examined in turn. Additional hypotlneses
that were examined within these main models, such as the speciﬁcity of EF deﬁcits, long-
term nature of disorder-related deﬁcits, sensitivity of EF tests to pathology, and
directional inﬂuences in the relationship between EF and psychiatric disorders, will also
be discussed.
Examination of Hypotheses

Nonspeciﬁc EF Deﬁcits

- Nonspeciﬁc comorbidity was not signiﬁcantly related to EF deﬁcits. In proﬁle

analyses, individuals with multiple lifetime disorders had signiﬁcantly poorer mean

120

global EF performance compared to individuals with anxiety disorders, but there were no
differences from controls. Participants with multiple disorders did not demonstrate any
deﬁcits in performance in current analyses compared to other disorder goups or controls.
In fact, as depicted in Figure 8, their performance on many tasks was better than that of
individuals with only current ADHD, alcoholism, or ASPD. Although there was a weak
linear relationship between number of current disorders and poorer EF performance, it
appeared as though the important effect was simply presence of a disorder near the time
of testing rather than comorbidity or the number of disorders.

Some differential effects suggested that current comorbidity predicted poorer
performance on certain EF processes, namely set-shifting and response inhibition. Again,
however, results were weak in that increasing numbers of disorders did not directly relate
to worse performance on these measures.

A number of potential explanations may account for the failure of the
nonspeciﬁcity hypothesis with regards to EF. Firstly, number of disorders may not be the
best marker of severity of psychopathology. Increasing numbers of disorders do not
necessarily reﬂect a concomitant increase in the level of impairment. Previous studies
have found that severity of depressive symptoms and frequency/Chronicity of episodes
were related to cognitive impairment (Merriam et al., 1999; Purcell et al., 1998; Richard
et al., 2003). Thus, other more informative markers of severity may be level of self-
reported impairment associated with individual disorders, Chronicity, success of
treatments, relapse rates, and other indicators. Second, the deﬁnition of “current”
disorders in the present study may have reduced the ability to detect effects. Current

disorders were not necessarily indicative of an “acute” condition at the time of testing,

121

but rather a “recent” condition because individuals ﬁ'om the alcoholism sample were
deﬁned as having a current disorder when they met criteria for a disorder witlnin the past
three years. Some “current” diagnoses may not have been present at the time of testing,
reducing their impact upon test performance. Further, current disorders may not have
been concurrently diagnosed, weakening any cumulative effects of having multiple
coexisting disorders. Finally, it may be that speciﬁc combinations of disorders are
important to EF effects (e.g., alcoholism and ASPD; Giancola & Moss, 1998; Malloy et
al., 1989). Overall, however, these results suggest that number of disorders alone does
not contribute to EF problems.
Disorder-Speciﬁc EF Deﬁcits

Given that nonspeciﬁc comorbidity was not associated with EF deﬁcits, a
comorbidity-speciﬁcity model was alternatively hypothesized to explain the relationship
between psychiatric disorders and EF deﬁcits at the level of multiple disorders. Here, it
was suggested that only certain disorders were related to EF deﬁcits, and comorbidity
amongst disorders made it appear more widespread. Analyses of global EF performance
showed that only certain disorders were associated with poorer performance. Mean
global EF differences for lifetime diagnoses suggested that individuals with ADHD
performed more poorly than individuals with anxiety disorders. No differences were
seen between controls and disorder goups for lifetime diagnoses, suggesting that this
ADHD effect did not reﬂect a clinical EF impairment. More clinically relevant results
were seen for current diagnoses, as current ADHD and ASPD goups demonstrated
global EF deﬁcits compared to controls. Current alcoholics (along with ADHD and

ASPD goups) performed more poorly than anxiety-disordered individuals. The small

122

sample size for current alcoholism may have prevented differences from controls from
reaching signiﬁcance. Thus, current and lifetime ADHD, ASPD, and current alcoholism
were associated with weaker global EF performance.

Many individuals in the present study met criteria for multiple lifetime/current
disorders, and the single-disorder analyses did not include this ﬁnll range of pathology.
When taking overlapping/comorbid conditions into account and controlling for their
presence in regession analyses, only current and lifetime ADHD were related to the EF
composite score. This was a robust effect, which remained even after controlling for
other factors such as age, gender, and FSIQ. Such ﬁndings are consistent with previous
ﬁndings for ADHD, as the EF deﬁcits associated with this disorder have generally
remained robust even after controlling for the presence of comorbid conditions (N igg et
al., 2005; Seidman et al., 1998; Willcutt etal., 2005). Therefore, ADHD, ASPD, and
alcoholism were associated with poorer global mean EF performance, particularly when
these conditions were diagnosed near the time of testing. In linear analyses, when
overlap and comorbidity amongst conditions were taken into account, only ADHD was
uniquely related to global EF deﬁcits.

Process-Speciﬁc Disorder-Related EF Deﬁcits

Speciﬁcity of the relationship between EF and psychopathology was further
examined by considering different aspects of EF processes. The hypothesis that
individual disorders may be associated with impaired performances in different aspects of
EF processes received some support. The same disorders that were shown to be related

to global effects had process-speciﬁc effects.

123

Proﬁle analysis highlighted the fact that participants with current ADHD or
current alcoholism tended to demonstrate relative decrements in their already poorer
global EF performance (described below) on set-shifting and response inhibition,
respectively. Overall, the somewhat weak componential effects in the proﬁle analysis
were surprising given that there was considerable support in the literature for process-
speciﬁc EF deﬁcits in the different disorders. F urtlner, the gaphs plotting the mean EF
task performance suggest that there were additional disorder-related differences in the
patterns of performance across tests for current diagnoses that did not reach signiﬁcance
in the proﬁle analysis (see Figure 8). The use of single-disorder goups (resulting in an
unbalanced design and reduced sample sizes) may have affected power to detect effects.

More powerful regession analyses examining the unique relationships between
disorders and individual EF test performances provided similar results, along with
additional effects that did not reach signiﬁcance in the proﬁle analysis. For instance,
lifetime ADHD was related to geater response variability and weak set-shifting. Current
disorder results were similar to those found with the proﬁle analysis, as current ADHD
was again related to deﬁcits in set-shifting (Trails Residual) and response variability, as
well as to response inhibition. Current alcoholism also predicted poorer response
inhibition (although not after controlling for F SIQ), as well as geater response
variability. Finally, a trend was seen for ASPD to be related to poorer set-shifting
performance (i.e., Trails Residual) in current-diagnosis analyses.

Therefore, the same disorders (i.e., current and lifetime ADHD, ASPD, and
current alcoholism) tended to be related to global as well as speciﬁc EF deﬁcits. The

process-speciﬁc weakrnesses overlapped amongst these disorders to some degee;

124

decrements on the residual score from Trail Making Test were seen for current and
lifetime ADHD and ASPD (the latter being a trend), and weak Response Inhibition and
RT Variability were seen for both ADHD and alcoholism. Thus, the processes most
sensitive to disorder-based differences were set-shifting, response inhibition, and
response variability. On the other hand, anxiety disorders, depression and drug
dependence did not demonstrate weaker global or component EF proceses. Other
childhood-onset disorders such as CD and ODD were also not uniquely related to EF
deﬁcits, likely because these disorders are not diagnosed in adultlnood and EF effects
were subsumed under the adult manifestations of these disorders (e.g., alcoholism,
ASPD, ADHD). Therefore, the same disorders demonstrated both global and speciﬁc EF
weaknesses, suggesting that the global effects were driven by component differences to
some degee. The deﬁcits associated with ADHD, alcoholism, and ASPD tended to
overlap. ADHD demonstrated the most widespread and robust effects, but the proﬁles of
performance on EF tests for these disorders were not entirely differentiable.
Consistencies and divergences exist between these results and past research with
respect to the individual disorder ﬁndings. Similar to the present study, past research on
ADHD has demonstrated deﬁcits on Trails B, Response Inhibition, and RT Variability,
while deﬁcits in WCST and Stroop Interference were rare (Boonstra et al., 2005;
Castellanos, Sonuga-Barke, Milham, & Tannock, 2006; Frazier et al., 2004; Hervey et al.,
2004; Johnson et al., 2001; Lijfﬁjt, Kenemans, Verbaten, & van Engeland, 2005; Lovejoy
et al., 1999; Murphy et al., 2001; Nigg et al., 2005; Riccio et al., 2005; Walker et al.,
2000; Willcut et al., 2005). Thus, EF results for participants with ADHD were highly

consistent with past literature.

125

More variability was noted between the present ﬁndings and past research for
alcoholism. A wide range of EF deﬁcits have been seen across previous studies on
alcoholism (Adams et al., 1993; Brokate etal., 2003; Dao-Castellana et al., 1998;
Hofﬁnan et al., 1987; Poon et al., 1999; Ratti et al., 2002; Uekermann et al., 2003), but
only deﬁcits in response inhibition (before controlling for IQ) and response variability
were found in the present study for alcoholism. The small sample size for current
alcoholism only in the present study may have reduced the power to detect additional
goup differences even though the alcoholism goup appeared to perform poorly on Trails
Residual as well (see Figure 8). While multiple deﬁcits have been seen across many
previous studies, the speciﬁc deﬁcits tended to differ between studies. Further, some
researchers found that individual tests lacked sensitivity in this population and global,
composite EF scores were required to demonstrate EF deﬁcits (Goldstein et al., 2004b;
Selby & Azrin, 1998; Sullivan et al., 2002). Thus, although speciﬁc deﬁcits were
conﬁned to only response variability and inhibition for alcoholism, these results are not
widely inconsistent with previous ﬁndings.

Contrary to expectations, ASPD was not related to deﬁcits in response inhibition.
Instead, individuals with ASPD demonstrated a trend towards poorer performance on a
set-shifting task. The expectation of response inhibition deﬁcits was based upon research
suggesting that similar regions of the brain contributed to antisocial symptoms and
response inhibition abilities (Blair & Cipolotti, 2000; Fuster, 1997; Rubia et al., 2000)
and some ﬁndings in children that aggession is related (Kerr, T remblay, Pagani, &
Vitaro, 1997). However, response inhibition deﬁcits have only been shown in a few

studies for DSM—IV diagnoses, and deﬁcits did not always reach full signiﬁcance (Dinn

126

& Harris, 2000; Dolan & Park, 2002). Furtlner, EF deﬁcits are more commonly seen in
individuals with ASPD and other comorbid conditions (Dinn & Harris, 2000; Malloy et
al., 1990), so controlling for other disorders in these analyses may have reduced cognitive
effects. Finally, although set-shifting deﬁcits were rare in the limited past research on
cognitive deﬁcits in ASPD, they were noted in at least one previous study (Dolan & Park,
2002). Therefore, the present ﬁndings are also not entirely inconsistent with previous
literature on ASPD.

Anxiety disorders, depression, and drug dependence were not related to EF
deﬁcits in this study. With respect to anxiety disorders, it was thought that because poor
interference control may theoretically contribute to the development of anxiety (Calvo &
Eysenck, 1996; Eysenck & Calvo, 1992; Hopko et al., 1998), deﬁcits would be seen on
EF tasks assessing the cognitive component of this process. However, this was not
supported. Interestingly, the presence of anxiety disorders, particularly lifetime
disorders, was associated with improved abilities to inhibit responses and maintain
consistency in response speeds (the latter being a trend). Further, individuals with
anxiety disorders performed signiﬁcantly better than individuals with ADHD on a set-
shifting task (i.e., Trails Residual). No individual anxiety disorders were accounting for
these ﬁndings. Taken together, it may be that individuals with a personality or response
styles that are vulnerable to developing anxiety disorders are more likely to be cautious,
attentive, and consistent on tasks that require sustained focus.

Note that the anxiety-based enhancements in set-shifting and response inhibition
and variability in the present study were generally relative to other disorders rather than

control participants, and tended to disappear after controlling for age, gender, and FSIQ.

127

Only OCD has been associated with consistent EF deﬁcits, and few individuals met
criteria for OCD in the anxiety disorder goup in this study. Thus, these results were
consistent with past ﬁndings suggesting lack of support for EF deﬁcits in anxiety. They
extend prior research by suggesting that anxiety-disordered individuals show relatively
better performance compared to indivduals with other pathologies on tasks assessing set-
shifting, response inhibition, and response variability.

Issues speciﬁc to the individual disorders as well as to the methodology of this
study may have reduced the likelihood of detecting EF effects with depression and drug
dependence. With regards to depression, no individuals with current MDD were included
in one of the samples (i.e., ADHD-based sample), and the deﬁnition of “current”
depression in the other sample (i.e., alcoholism) included individuals who met criteria
near the time of testing, not necessarily at the time of testing. Some studies have
suggested that cognitive deﬁcits continue even after depression has remitted (Kessing et
al., 1998). However, in general, there was geater support for EF deﬁcits in the acute
phase of the disorder, and these were related to severity of symptoms, recurrence of
depressive episodes, and Chronicity of the disorder (Kessing et al., 1998; Paradiso et al.,
1997). These key factors could not be assessed in the present examination and may have
limited ﬁndings for depression and EF.

Similar factors were important in past studies of drug dependence. For instance,
number of drug dependencies, presence of additional comorbid disorders, dose-related
drug effects, Chronicity of use, and recency of intake were shown to be signiﬁcant to past
associations between drug dependence and EF impairment (Bolla et al., 1999; Eldreth et

al., 2004; Fals-Stewart & Bates, 2003; Goldstein et al., 2004b; Gruber et al., 2005; Selby

128

& Azrin, 1998). The inability to account for these factors in the present study may have
impeded the ability to detect EF effects. Further, the present sample included almost no
individuals who met criteria for only drug dependence, disrupting the ability to isolate
cognitive deﬁcits associated with this disorder. The high rate of comorbidity with drug
dependence in the present study, and past ﬁndings that comorbidity was important to
cognitive effects, suggests that other disorders account for the link between drug
dependence and EF throughout the literature. Taken together, disorder-related and study-
speciﬁc issues may be contributing to the failure to detect an EF effect for anxiety
disorders, depression, and drug dependence.

To review, both disorder- and process-speciﬁc EF effects were found. The
deﬁcits in component processes across ADHD, ASPD, and alcoholism tended to overlap,
but were not identical, suggesting that certain EF tests were particularly sensitive to
psychopathology (i.e., Trail Making Test and Stop Signal). The presence of these
disorders near the time of testing increased the likelihood of deﬁcits, but a lifetime
diagnosis of ADHD was also associated with weaker performance. Anxiety disorders,
depression, and drug dependence were not associated with EF impairments.
Dimension-Specific EF Deﬁcits

The ﬁnal main hypothesis examined the relationship between psychopathology
and EF at the broadest level of the diagnostic hierarchy. Here, the question was whether
the wide-ranging association between psychiatric disorders and EF deﬁcits was due to
shared underlying dimensions of psychopathology. A common and accepted distinction
between “internalizing” and “externalizing” disorders was used to determine whether

traits/ symptoms that are shared in the development and manifestation of these disorders

129

are differentially related to EF deﬁcits. Previous analyses have highlighted that global
and speciﬁc EF deﬁcits were associated with the following disorders: ADHD,
alcoholism, and ASPD. Depression and anxiety disorders (along with drug dependence,
perhaps due to the highly comorbid nature of this disorder) did not demonstrate EF
deﬁcits. The dimensional model further emphasized the split between these types of
disorders, as only externalizing disorders were related to EF deﬁcits.

Given the strong relationship between ADHD and EF deﬁcits in prior analyses, it
is particularly striking that current externalizing disorders continued to signiﬁcantly
predict EF test performance even after removing the ADHD indicator from the latent
factor. As well, while individual lifetime analyses tended to demonstrate weaker links
between disorders and cognitive functioning, the latent lifetime externalizing factor
demonstrated a strong association between disorders and EF performance in the
dimensional analysis. This provides strong support for the possibility that a shared aspect
of these disorders, rather than just a single disorder or type of symptom, is contributing to
performance on EF tests.

ADHD was shown to be most strongly related to EF deﬁcits in previous analyses,
as the widespread relationship with component EF processes contributed to robust global
effects. Alcoholism and ASPD demonstrated more sporadic relationships were fewer EF
processes. It may be that a shared underlying feature of these disorders is stronger in
ADHD than in other externalizing disorders, thus accounting for the more consistent
relationship between ADHD and EF deﬁcits. However, EF effects disappeared after
controlling for IQ due to strong relationships between FSIQ, externalizing disorders, and

BF performance. It is difﬁcult to determine whether this is due to sample-based

130

differences, in that controlling for IQ actually removed disorder-related effects.
Alternatively, IQ differences may be mediating the association between externalizing
disorders and EF. This will be further discussed below.

Dimensional models also highlighted the lack of association between internalizing
disorders and cognitive performance. Anxiety and depression have been variably
associated with EF deﬁcits in past literature, although ﬁndings were much stronger for
depression. In this study, previously seen anxiety-based improvements in performance
on some EF processes were perhaps washed out by the lack of association between
depression and EF. Methodological issues may have also contributed to the limited
shared variance between internalizing disorders to some degee. It is possible that
limitations in measurement of these disorders (discussed under Limitations) weakened
the ability to detect EF effects in internalizing disorders. However, these results do
correspond with the ﬁndings from other analyses.

Taken togetlner, only externalizing disorders were related to deﬁcits in global EF
processes, and internalizing disorders were not associated with EF. However, these
effects disappeared after controlling for F SIQ, suggesting that FSIQ may account for the
relationship between externalizing disorders and EF deﬁcits.

Conclusions on Main Models

Integating the results across the four main models examined in the present study,
it appears as though shared underlying features of pathology and process-speciﬁc effects
are contributing to the widespread relationship between EF deﬁcits and psychiatric
disorders. Individuals with externalizing disorders tended to demonstrate poorer

performance on EF tests assessing set-shifting, response inhibition, and response

131

variability. One externalizing disorder, drug dependence, was not related to EF deﬁcits,
but detection of any associations may have been impeded by the difﬁculty isolating this
disorder in the present sample. Study-speciﬁc methodological issues may have
contributed to the failure of internalizing disorders to demonstrate an EF effect in the
present study (detailed below under Limitations). It is also possible, however, that the
failure to adequately assess for and control comorbid disorders in past studies produced
stronger associations between tlnese disorders and EF deﬁcits, particularly for depression.
Thus, tests of the main hypotheses provided support for dimensional and process-speciﬁc
effects. Since only speciﬁc disorders are associated with EF deﬁcits, past studies may
have failed to adequately test and control for comorbid conditions.

The nature of the disorder-speciﬁc relationships within the externalizing
dimension may reﬂect a developmental trajectory for the emergence of EF deﬁcits.
These extemaling disorders are all highly comorbid, but ADHD is the ﬁrst to deve10p in
childhood. In this study, ADHD was also the disorder with the strongest and broadest
relationship to EF deﬁcits, which remained even after controlling for IQ. Symptoms of
ADHD may contribute to and mediate the EF effects associated witln other disorders. For
instance, alcoholism and ASPD, which develop later in life, were associated with weaker
EF deﬁcits. Finally, drug dependence generally has the latest onset and usually occurs
within the context of other disorders, perhaps accounting for the failure of this disorder to
demonstrate unique EF effects. Thus, the developmental sequence of the externalizing

disorders may be contributing to the relationships with EF seen in this study.

132

Additional Hypotheses

The present study provided information about other questions that further
elucidate the association between EF and psychiatric disorders.

Speciﬁcity of EF deﬁcits. One of the goals in the present study was to determine
whether EF represented a nonspeciﬁc marker of disturbance or a potentially causal factor
in the development or maintenance of these disorders. Processing speed tasks were
included as non-executive indicators of general cognitive functioning. Processing speed
was consistently related to the same disorders that demonstrated EF effects. Therefore,
speciﬁcity was seen in the types of disorders related to cognitive deﬁcits (i.e., ADHD,
alcoholism, ASPD, externalizing), and in the EF processes/tests involved (i.e., Trails set-
shifting, Stop Signal response inhibition and variability), but domain-speciﬁc cognitive
deﬁcits were not supported. A more general neurocognitive weakness appears to be
associated with these disorders, rather than a speciﬁc or primary EF deﬁcit.

The relationship between speed and psychiatric disorders remained signiﬁcant
even after controlling for age, gender, and F SIQ. This indicates that this general
neurocognitive weakness was more robust than the EF impairment. It may be that the
extensive literature that links EF deﬁcits with psychiatric disorders stems in part from the
use of measures that did not isolate EF or control for component processes such as
processing speed. In this study, the prominent effects of speed were removed from two
very commonly used EF tasks, Stroop and Trail Making Test, and perhaps the fact that
this was almost never done in past studies contributed to wider support for EF deﬁcits
than is warranted. For instance, as mentioned previously, the large effect size for Stroop

Color-Word in individuals with ADHD has at times been presented as support for deﬁcits

133

in response inhibition or interference control (e.g., Boonstra et al., 2005); however, when
the effects of speed are controlled, ADHD subjects do not demonstrate deﬁcits in
interference control (Boonstra et al., 2005; Hervey et al., 2004). Studies of executive
functions also tend to rely upon the Trails B trial and fail to control for motor speed
assessed through Trails A; however, doing so would provide a more pure measure of
executive set-shifting abilities (Arbuthnott & Frank, 2000). The residual variables that
were included in this study better isolated EF processes.

Concerns have been raised about whether or not EF processes are differentiable
from other cognitive processes such as perceptual speed or the general intelligence factor
(g), which appears to underlie performance on IQ tests (Salthouse et al., 2003; 2005).
The BF effects that were found in this study withstood controls for processing speed
despite a high correlation between EF and processing speed. Both cognitive processes
were uniquely associated with psychopathology. The use of residual scores likely
reduced the correlation between EF and processing speed in this study, and helped to
differentiate these constructs to some degee. On the other hand, EF effects could not
withstand controls for FSIQ (although speed continued to be a signiﬁcant factor). This
suggests that there is divergent validity between EF and speed processes, at least as
related to psychopathology, but EF and g may be more difﬁcult to differentiate. Some
researchers have argued that components of EF underlie IQ (Conway et al., 2003;
Friedman et al., 2006; Kane & Engle, 2002). However, direction of effects cannot be
determined in the present study, and IQ remained the signiﬁcant predictor herein. It is

also possible that controlling for IQ inadvertently resulted in controlling for sample-based

134

differences in distributions of disorders. However, speed effects did withstand controls
for IQ although these effects were vulnerable to the same issues.

Thus, although externalizing disorders (ADHD, alcoholism, and ASPD) were
associated with EF weaknesses in this study, neuropsychological deﬁcits crossed
cognitive domains. Processing speed may be inﬂating the EF effects found in previous
studies when not adequately controlled. Finally, EF effects may be mediated by IQ.

Direction of eﬂects. The cross-sectional nature of the present study makes it
difficult to determine causality between disorders and cognition. Directional effects were
analyzed in this study from disorder (i.e., symptoms) to cognitive functioning. As
mentioned previously, affect may impact cognition (Ashby et al., 1999). However,
effects may also occur in the opposite direction. Top-down cognitive information-
processing may interact with personality factors and contribute to psychopathology
(V asey et al., 2003). Such cognitive models have been proposed for both ADHD and
alcoholism, wherein EF deﬁcits precede and contribute to the development of disordered
symptoms (Giancola et al., 2001; Nigg et al., 2006; Pennington & Ozonoff, 1996).
Another possibility is that neither behavioral symptoms nor cognitive functions are
“causal” in the etiology of these disorders, but that botln are manifestations of other
aberrant processes.

It should be noted that when the direction of causality was reversed in the
structural models, ﬁt did not change for either lifetime or current analyses (results not
shown). In the lifetime model, only speed signiﬁcantly predicted externalizing disorders
in this model. In the current model, there were no signiﬁcant predictive pathways.

Conclusions on directionality cannot be derived from these results given model

135

equivalence and the cross-sectional nature of the study. However, questions about
direction of effects are important and should continue to be examined. Prospective
longitudinal studies will ultimately be required to better understand the developmental
course of these disorders. Understanding etiological relationships will aid in prevention
and treatment of these conditions.

Current versus lifetime disorders. Across analyses, EF effects were stronger for
current versus lifetime disorders. Only lifetime ADHD contributed to EF weaknesses,
and this goup did not perform signiﬁcantly more poorly from controls, suggesting that
differences did not represent clinical deﬁcits. More current disorders were associated
with EF deﬁcits, and these performances did differ from controls. Some literature has
suggested that deﬁcits may be seen even following symptom remittance in depression
(Kessing, 1998; Paradiso et al., 1997) and abstinence in alcoholism (Munro et al., 2000).
However, factors such as subclinical symptoms and disorder severity may contribute to
these longer-term effects (Kessing, 1998), and effects generally remit over time (Selby &
Azrin, 1998). Therefore, the exacerbation of EF effects during the active disorder states
in the present study is consistent with past ﬁndings.

As mentioned, ADHD was the only lifetime disorder to demonstrate EF effects.
Many participants with ADHD were speciﬁcally recruited because they continued to
demonstrate symptoms of the disorder into adultlnood. The high rates of ADHD at the
time of testing likely enhanced the lifetime association to EF. On the other hand,
cognitive effects associated with other current disorders may have been attenuated by the

fact that, as previously mentioned, some “current” disorders may not have been present at

136

the time of testing. Despite these issues a strong “current” or recency effect was evident
across the study.

Sensitivity of tasks. As covered in the literature review in the introduction, wide
variability exists in the relationships between psychopathology and performance on
speciﬁc EF tasks and processes. In the present study, the gaplnical depictions in Figures
7 through 10 higlnlight the generally wider dispersion of means for Trails Residual,
Response Inhibition, and RT Variability. These variables were also associated with
ADHD, alcoholism, and ASPD. Thus, it appears as though the Trail Making Test and
Stop Signal Task were most sensitive to the negative effects of psychopathology in this
study. It may be that these tasks and the processes that they assess are most likely to
reveal the sometimes subtle neurocognitive changes that may be associated with
psychiatric disorders. The Stop Signal Task in particular assesses very speciﬁc processes
and is thus quite sensitive to differences in performance. As a result, both response
inhibition and variability were variables that consistently demonstrated sensitivity to
cognitive effects tlnroughout analyses. The Trails Residual score, with its emphasis upon
set-shifting, may have helped to make this task more “process pure.” Thus, purer tasks
may be more sensitive to the subtle effects of psychopathology.

The WCST, on the other hand, is a task that became popular after it was noted
that individuals with brain damage (particularly to the frontal lobe) performed poorly
upon it. It is a fairly simple task for “intact” individuals, and may be best suited to
distinguish normal ﬁ'om abnormal brain function. The cognitive effects associated with
the types of psychopathology included in this study are generally subtle and not

necessarily considered “abnormal.” As mentioned, the lack of sensitivity of the Stroop

137

Residual variable to psychopathology may be due to removal of speed effects. It was
also the weakest indicator on the latent EF factor, suggesting that it shares less variance
with other EF measures. Finally, individual EF tasks are notorious for poor reliability,
and this may have affected sensitivity to effects and signiﬁcance testing in the present
study. Thus, multiple factors may have contributed to the sensitivity of certain tasks and
processes to psychopathology. Set-shifting, as assessed by Trails Residual, and response
inhibition and RT variability ﬁom the Stop Signal Task revealed consistent relationships
with externalizing psychopathology and may be sensitive markers of psychopathology.
Limitations

A number of factors may limit the generalizability of the present results. First and
foremost is the fact that this study relied upon combirning two separate samples that
differed in a number of respects, from the methods of recruitment and inclusion/exclusion
criteria, to the speciﬁc diagnostic procedures (i.e., DIS versus SCID/KSADS,
retrospective diagnosis of DSM—IV disorders in the alcoholism sample), to the resultant
demogaphic characteristics of the samples. Further, different but similar measures were
used to assess FSIQ and WCST between the two samples. The two-sample issue has
been frequently referred to throughout this manuscript as potentially limiting the
interpretation of the results in both general and speciﬁc ways. All issues will not be
repeated here, but a few deserve further higlnlighting.

Different procedures were used to establish diagnoses in the two combined
samples, which may have affected the reliability and validity of diagnoses. Different
diagnostic interviews were used, as covered previously. Moreover, the extra attention

given to diagnosing externalizing disorders (i.e., ADHD, ASPD, and alcoholism) may

138

have improved and led to better ﬁt and appropriate representation for these disorders. In
other words, the increased attention to these disorders may have overridden the
potentially negative effects of using different diagnostic methods to some degee. On the
other hand, the latent factor for internalizing disorders had poorer ﬁt. It is possible tlnat
this factor was less stable and reliable a marker of true internalizing pathology than was
the externalizing factor for externalizing disorders. Perhaps the increased attention to
ensuring the reliability and validity of ADHD, alcoholism, and ASPD in the two samples
included in this study created better measures for these disorders, resulting in a more
consistent relationship with cognitive deﬁcits.

Also, the diagnosis of “current” disorders in the present study was based upon
disorders present at the time of testing in one sample (i.e., ADHD sample), and disorders
present within approximately the three years prior to testing in the other sample (i.e.,
alcoholism sample). This may have weakened the effects associated with current
disorders, but current disorders were still more strongly related to cognitive performance.
Current MDD was excluded ﬁ'om the ADHD-based study, so its deﬁnition was based
solely on the idea that the disorder was present within the past three years. Cognitive
effects for MDD are most apparent during the acute phase of the disorder, and tlnus the
deﬁnition of current in the present study likely weakened the ability to detect any
cognitive effects associated with this disorder.

Other sample-related differences, such as FSIQ and age, were controlled in order
to eliminate their inﬂuence upon the results. It is possible that differences in F SIQ and
the distributions of disorders varied together between the samples, and that in controlling

for FSIQ the effects of disorder were also inadvertently controlled. This could potentially

139

account for the removal or reduction of many effects after FSIQ was included in
analyses. On the other hand, as mentioned previously, the effects of speed did remain
even after controlling for FSIQ. As well, note that there were signiﬁcant differences in
IQ even between the “controls” in the two samples (alcoholism M = 107.13, SD=12.45;
ADHD M = 114.20, SD=9.50), suggesting that the IQ discrepancies were not simply due
to distributions of disorders but related to the different sample populations. Other
differences such as SES could not be directly controlled, but may have contributed to or
been related to the discrepancies in F SIQ and other variables. Taken together, the
difﬁculty disentangling sample-dependent effects ﬁom true variation in the variables of
interest complicates interpretation to some degee.

Another limit to generalizability is the fact that this was not a population-based
study. Recruitment for the two samples that were combined in this study was focused on
ADHD and alcoholism/ASPD. It is possible that the associations amongst disorders, and
between disorders and cognitive performance, were somewhat different than would occur
in the general population. For instance, there may be a geater preponderance of
depressive and anxious disorders that are secondary to other patlnology in the present
sample. This may have affected the relationships amongst internalizing disorders and
between internalizing disorders and cognitive effects. Despite this, a large number of
people still met criteria for only a depressive or anxious disorder. Assortative mating
may have affected the relationships amongst disorders and cognitive effects in the
alcoholism study as well. Finally, the sample was predominantly Caucasian, which
reduced generalizability. Thus, the fact that this was not a population-based study may

limit the generalizability of the results.

140

The use of categorical disorders rather than dimensional symptoms in this study
limited power to some degee. As well, individuals often present with quite
heterogeneous marnifestations of a particular diagnosis. It is possible that particular
symptom constellations within individual disorders may be related to cognitive effects,
and that these associations were removed by the use of categorical diagnoses in the
present study. Future studies would beneﬁt from examining how symptom counts and
constellations contribute to cognitive performance. Given that shared underlying features
of disorders appeared to be contributing to EF effects, this would help to elucidate the
speciﬁc pathological processes that are most strongly related to cognitive effects.

It should also be noted that certain aspects of EF were not included in the present
study. The present study used Pennington & Ozonoffs (1996) model of EF, although
other models exist and could have been used (Moscovitch & Winocur, 1996; West, 1996)
and some have questioned even the notion of the executive construct (Parkin, 1998). It
may also be that working memory, initiation/generativity (i.e., verbal ﬂuency), or
planning are key processes that are associated with the disorders included in the present
study. Otlner tasks such as the Halstead Category Test could have also provided useful
information. The low factor loadings on the EF latent factor also suggest that the tasks
included in this study did not have a lot of common variance. The use of tasks with
strong psychometric properties, such as those included in recent studies (i.e., Miyake et
al., 2000; Friedman etal., 2006), may improve reliability and speciﬁcity and, thus,
strengthen the measurement of the EF construct. Therefore, additional EF processes and
tasks should be included in future studies to more fully assess the relationship between

EF and psychopathology.

141

In sum, some limitations associated with the present study may affect the
generalizability of tlnese results. However, a number of signiﬁcant strengths help to
offset these effects and highlight the ways in which the present study adds to the extant
literature on EF and psychopathology.

Strengths

This study had numerous distinctive features. The high density of many different
types of disorders in tlnese at—risk samples provided a urnique opportunity to
simultaneously evaluate the cognitive effects associated with individual and classes of
disorders. This is perhaps the ﬁrst study to directly examine EF in this manner with so
many different disorders. Thus, it provides the best evidence for EF effects in
psychopathology, against which future studies may be evaluated.

The samples themselves were large, well-deﬁned, and community-based. The
integation of different sampling methods and a wide range of risk goups improved
generalizability (e. g, range of SES, high risk alcoholics and other disorders, controls that
are functioning well across a number of domains). Thus, although this was not a
population-based study, and disorder rates were higher than those in the general
population, this study included large and rather diverse samples from a socioeconomic
perspective.

Both EF and psychopathology were measured ﬁom multiple perspectives. EF
was considered as botln a global factor as well as multiple distinct component processes.
Psychopatlnology was examined at the level of individual disorders, comorbid conditions,
and shared underlying dimensions. This broad approach allowed for an extensive-

examination of the relationship between EF and psychopathology.

142

Along with including multiple theoretical orientations, both traditional and more
advanced statistical methods were used to examine associations. Of particular
signiﬁcance were latent modeling techniques, which were intended to reduce error and
increase reliability. By maximizing construct-relevant variance and excluding variance
unique to any single measure, latent modeling improved reliability of the single, global
EF variable used in SEM analyses. This technique allowed for examination of a “purer”
composite index of the EF, along with speed, internalizing, and externalizing constructs,
than would be possible working at the individual variable level. The improved reliability
associated with latent modeling helped to strengthen the results of the structural models.
Other statistical techniques to help isolate the effects of interest in EF included removing
component effects and creating residual scores, and controlling the effects of processing
speed and FSIQ. Together, tlnese methods helped to create purer measures to ensure to
the extent possible that we were focusing on the constructs of interest.

Therefore, the unique nature of this study and its many strengths provided
information that can advance the ﬁeld towards new methods of study. This is a useful
starting point for future examinations that should study large and diverse population-
based samples longitudinally to ﬁnrther evaluate the effects of psychopathology on EF.

Conclusions

Many conclusions about the association between psychopathology and cognitive
functioning may be drawn from this extensive examination. Firstly, nonspeciﬁc
comorbidity and internalizing psychopathology were ruled out as being associated with
EF. Extemalizing disorders were related to EF weakrnesses in what may be a

developmentally-mediated manner. ADHD, a childhood-onset externalizing disorder that

143

continued into adulthood in this study, had the strongest relationship with EF. Shared
features between early- and later-onset externalizing disorders may mediate and
contribute to EF effects. While adult-onset disorders such as alcoholism and ASPD
demonstrated weaker and more speciﬁc EF effects than ADHD, the overlap in affected
processes between these externalizing disorders points to shared disturbances. Thus,
common underlying pathways to EF deﬁcits in psychopathology may be
developmentally-mediated tlnrough early-onset externalizing disorders such as ADHD.

Irnportantly, cognitive effects in psychopatlnology were not isolated to the EF
domain. Processing speed was also related to externalizing disorders, with more robust
effects. Many effects disappeared after controlling for 1Q, suggesting that IQ may
mediate EF effects in some cases. Task impurity is a major concern in the measurement
of EF, and these ﬁndings further highlight the possibility that other cognitive processes
may underlie or account for apparent EF deﬁcits.

In conclusion, these results show that the relationship between EF and
psychopathology involves speciﬁc externalizing pathological processes, comorbidity
among disorders, speciﬁc components of EF, and multiple cognitive domains. All of
these factors have contributed to the seemingly wide-ranging relationship between
psychiatric disorders and EF throughout the literature. These results highlight important
issues that need to be accounted for in future studies to clarify the complex interplay

between behavioral and cognitive aspects of psychopathology.

144

Table 1

Summary of Major Findings in the Literature Regarding EF Component Deﬁcits

 

 

Working Planning Response Interference Set-shifting &
Memory Inhibition Control maintenance
Attention-Deﬁcit
/Hyperactivity X X X X
Disorder
Antisocial X
Personality Disorder
Alcohol
Dependence X X X X
Substance X X X X
Dependence
Obsessive-
Compulsive X X X X X
Disorder
Other Anxiety .7 .7 .,
Disorders °
Depression X X X X X

 

Notes to Table 1. An ‘X’ marks component processes with which a disorder has
ﬁ'equently been associated. Note that even though some disorders have similar deﬁcits,
there are often differences in the level of deﬁcits, if not in the components affected (i.e.,
alcohol and substance dependence).

145

 

I MDD

I Dysthymia I‘ Anxious-
, Miserv

 

 

 

 

 

I GAD

’ Internalizing
I Social Phobia
I Simple Phobia Iq‘ o

I Agoraphobia

 

 

 

 

 

 

 

 

I PTSD

 

 

 

I ocn

 

 

I Alcohol Dep

I Drug Dep
e
I ASPD

Childhood
Extemalizing/
Adult ADHD

 

 

 

 

 

 

 

 

 

Figure 1 . Initial proposed model representing the relationships between DSM-IV
disorders based on Krueger’s (1999) ﬁndings.

Note to Figure I . Panic disorder could not be included altlnough present in Krueger’s
model. Disorders that were added in the present analyses were childhood externalizing
disorders (ADHD, ODD, CD), PTSD, and OCD. Childhood externalizing disorders were
included in lifetime diagnosis analyses, and adult ADHD in current diagnosis analyses.
GAD=generalized anxiety disorder; PTSD=post-traumatic stress disorder;
OCD=obsessive-compulsive disorder; Dep=dependence; ASPD=antisocial personality
disorder; ADHD=attention-deﬁcit/hyperactivity disorder.

146

E manoou W855:
. >585-
039355 Emma?

a
EEBEEBW
monum— wrocmm ‘

Hamzm W885:

 

      

mxoocaé

. Womb Sac
mgosoam

mtg—o 3563 ﬁg. <33sz

vU\>moBgoEm
gamma. mom

wﬁmU

>332 Gnu mﬁeoou €03

E e was mass. 82
.9me

minim >

 

 

GEE
mxﬁoamzNEm

 

 

 

Emma N. 3388 Scan. 8 9655a :5 3533!? c328: cane—«Sm 88 amorous—9°89. acacia Ea aoﬁovﬁoraommom. 8m" volcanoes.

 

22m 8 $1 $56 N. O>Uumo=o3=~hm gonna. Emoaon EmUHuoaﬁmcamao 3.0mm Emoaon >616 knitnc: t n . .. in i. . “U Emoamn >vanm=amooE

unions—5v. E833 Unuuaavosaoaoo“ ZUUHBNHS nonnommmé 2832.“ $04210 ammoaon 20mg. wmnﬁmmoosmm: 08d measm Ham» 3320325 2.83“
ﬂame EDEN—womvoamn 53230: 8 58.0.58 3. mac mism— wowvoao ﬂan“ a. <m1mc=5u<m1mcz§ ammo antenna mom. wnov Ham—n.

147

Table 2

Demographic Information for the Combined, Alcoholism, and ADHD samples at the time
of neuropsychological testing

 

 

Variable Combined Alcoholism ADHD p

(n=641) (n=448) (n=193)
# and % 336 (52.4%) 233 (52.0%) 103 (53.4%) .75
male
# and % 615 (95.9%) 448 (100.0%) 167 (86.5%) <.01
white
# and % 405 (69.5%) 361 (83.8%) 44 (28.9%) <.01
married
Age in years 38.12 (10.39) 44.14 (5.00) 24.15 (4.53) <.01

(range=26.41 — (range=18.19 —
66.70) 37.59)

Education 14.07 (2.13) 13.97 (2.23) 14.30 (1.87) .07
FSIQ 106.29 (12.85) 103.85 (12.83) 111.98 (10.97) <.01

 

Notes to Table 2. Alcoholism=alcoholism sample; ADHD=attention-deﬁcit/hyperactivity
disorder sample; FSIQ=full scale intelligence quotient; SES=socioeconomic status. t-
tests and chi-square were used for analyses.

148

Table 3

Numbers of Lifetime Disorders in the Combined, Alcoholism, and ADHD Samples

 

T

 

Task Combined Alcoholism ADHD x p
(n=641) (n=448) (n=193)

No disorders 170 (26.5%) 106 (23.7%) 64 (33.2%) 6.25 .01

MDD 212 (33.1%) 169 (37.7%) 43 (22.3%) 14.53 <.001

Dysthynnia 63 (9.8%) 52 (11.6%) 11 (5.7%) 5.31 .02

Depressive 234 (36.5%) 183 (40.8%) 51 (26.4%) 12.10 .001

Disorders

GAD 39 (6.1%) 22 (4.9%) 17 (8.8%) 3.59 .06

Agoraphobia 8 (1.2%) 8 (1.8%) 0 (0%) 3.49 .06

without PD

Social Phobia 50 (7.8%) 34 (7.6%) 16 (8.3%) 0.09 .76

Speciﬁc 89 (13.9%) 83 (18.5%) 6 (3.1%) 26.82 <.001

Phobia

PT SD 26 (4.1%) 22 (4.9%) 4 (2.1%) 2.79 .13

OCD 11 (1.7%) 9 (2.0%) 2 (1.0%) 0.76 .52

Anxiety 174 (27.1%) 139 (31.0%) 35 (18.1%) 11.34 .001

Disorders

ADHD 112 (17.5%) 9 (2.0%) 103 (53.4%) 246.75 <.001

Conduct 58 (9.0%) 51 (11.4%) 7 (3.6%) 9.86 .002

Disorder

ODD 18 (2.8%) 2 (0.4%) 16 (8.3%) 30.41 <.001

Child 166 (25.9%) 59 (13.2%) 107 (55.4%) 125.60 <.001

Extemalizing

ASPD 72 (11.2%) 67 (15.0%) 5 (2.6%) 8.30 .004

Alcohol 208 (34.0%) 196 (43.8%) 22 (11.4%) 62.90 <.001

Dependence

Drug 44 (6.9%) 30 (6.7%) 14 (7.3%) 0.07 .80

Dependence

 

Notes to Table 3. MDD=major depressive disorder; GAD=generalized anxiety disorder; PD=panic
disorder; PTSD=post-traumatic stress disorder; OCD=obsessive compulsive disorder; ADHD=attention-
deﬁcit/hyperactivity disorder; ODD=oppositional deﬁant disorder; ASPD=antisocial personality disorder.

149

Table 4

Numbers of Current Disorders in the Combined, Alcoholism, and ADHD Samples

 

 

Task Combined Alcoholism ADHD x2 p
(n=641) (n=448) (n=193)

No disorder 360 (56.2%) 262 (58.5%) 98 (50.8%) 3.25 .07

MDD 100 (15.6%) 100 (22.3%) 0 (0.0%) 51.04 <.001

Dysthymia 14 (2.2%) 4 (0.9%) 10 (5.2%) 11.61 .002

Depressive 113 (17.6%) 103 (23.0%) 10 (5.2%) 29.46 <.001

Disorders

GAD 30 (4.7%) 13 (2.9%) 17 (8.8%) 10.55 .001

Agoraphobia 0 (0%) 0 (0.0%) 0 (0.0%) -- --

without PD

Social 24 (3.7%) 13 (2.9%) 11 (5.7%) 2.93 .09

Phobia

Speciﬁc 27 (4.2%) 21 (4.7%) 6 (3.1%) 0.83 .36

Phobia

PTSD 14 (2.2%) 13 (2.9%) 1 (0.5%) 3.59 .08

OCD 3 (0.5%) 1 (0.2%) 2 (1.0%) 1.91 .22

Anxiety 78 (12.2%) 48 (10.7%) 30 (15.5%) 2.94 .09

Disorders

ADHD 78 (12.2%) -- 78 (40.4%) -- --

ASPD 72 (11.2%) 67 (15.0%) 5 (2.6%) 20.68 .000

Alcohol 42 (6.6%) 35 (7.8%) 7 (3.6%) 3.86 .05

Dependence

Drug 8 (1.2%) 7 (1.6%) 1 (0.5%) 1.19 .45

Dependence

 

Notes to Table 4. MDD=major depressive disorder; GAD=generalized anxiety disorder;
PD=panic disorder; PTSD=post-traumatic stress disorder; OCD=obsessive compulsive
disorder; ADHD=attention-deﬁcit/hyperactivity disorder; ODD=oppositional deﬁant
disorder; ASPD=antisocial personality disorder.

150

Table 5

Means and Standard Deviations of Cognitive T asks for the Combined, Alcoholism, and
ADHD Samples

 

 

Task Combined Alcoholism ADHD t p
(n=641) (n=448) (n=193)

Resp Inhib 248.05 250.95 241.32 1.56 .12
(72.03) (76.05) (61.34)

RT 176.03 198.44 124.03 12.34 <.001

Variability (77.87) (79.71) (39.14)

Stroop 0.00 0.00 0.00 .00 l .00

Residual (8.39) (8.27) (8.68)

Trails 0.00 0.00 0.00 .00 1.00

Residual (18.19) (19.73) (14.04)

WCST PE 97.12 96.06 99.57 -3.15 .002
(13.04) (12.98) (12.89)

Trails A 27.37 28.39 24.98 4.38 <.001

(9.17) (9.55) (7.72)

Stroop Word 100.18 99.74 101.21 -1.02 .31
(16.76) (17.15) (15.83)

Stroop Color 74.89 73.46 78.21 -4.07 <.001
(13.72) (13.53) (13.62)

Stroop 44.44 40.52 53.53 -13.85 <.001

Color-Word (12.44) (10.53) (1 1.77)

Trails B 60.51 63.48 53.62 5.87 <.001
(23.62) (25.64) (16.17)

FSIQ 106.29 103.85 111.98 -8.17 <.001
(12.85) (12.83) (10.97)

 

Notes to Table 5. SSRT=stop signal reaction time; SDX=variability of Go response time; Stroop
Residual=unstandardized residual of Stroop Color-Word regessed on Stroop Color and Stroop Word;
Trails Residual=unstandardized residual of Trails B regessed on Trails A; WCST PE=standardized score
for Wisconsin Card Sorting Task perseverative errors; FSIQ=full scale intelligence quotient; Resp
Inhib=Response Inhibition as measured by Stop Signal Response Time; RT Variability=variability of go
response ﬁ'om Stop Task.

151

How—o a

San: hooxme.\o~. demonized Heumex h§m~§m Dacia m: Sm 065333 gain

 

 

Cicada mqoov Ham—m womb HS. <m~ (<03. mneou onou H8»? > mqoou 0. H85 w
wow Wow 536 area 8.2. g
20 Emoaoa 9% -_.o~ 393 393 3.3 393 qmbm ~98 3.3. m9w~
SEC GMuB 39ro Amohd 29qu C93; C935 3.3V 29%: A~ub~v
ECU 93 93 ~A9qm 3?: 093 393 3.3 ~33 3:3 3.:
$63 3903 Good 306$ C To: C9-v Quake 3. iv 2 Eu: Gar—3
GEEKBE 93 Fog ~m~.m~ 59.: 3.8 cm. 5 mohm ~93 s: .3 3.3
$.qu A-.~& $92: @9me SPEC C9~d :9me 3&8 C~u~v ﬁxed
Cap—.895 -93 9: ~m9~_ :98 c9: cob-no dub ~93 3.3 3.8
98:33 SM 5 Cab-c 393V QA.~8 C~bd Cabd Cwbwv 6.8V C #93 A~Powv
0% 98 :m ~no3m 39.3 3.3 on. 5 3.3 ~93 Amhw 8.2
Eb: ﬁrmov Gremv 895v 39m: 2908 Cm.~8 $.qu Chris @925
>moBuroEo 93 L93 ~m~.o_ 35.3 8.3 Ear—w 3.3. ~mbm s9~m H93
363 £908 G~.A~v Amrﬁv CP~3 3.me Coaov 95 6.93 C923
moon: 389m -9w~ Won ~w _ .3 _uu.~o 093 093 Kbm ~m._ _ t. _ u 3.3
can: SPAS 3903 09m& C~. 3V C9u$ 293V Amhmv 293V A~9m8
muoowmo 383m -93 9mg 39$. $93 3.8 8.2 3.3 ~93 3.3 8.3
3&3 A-b$ A363 Q9m3 Quid 289$ CPKV G..& 29ch A~9-V
Hog-m0 9mm 93 ~u~.m~ 399m cube 8. G q~.a~ ~93 ¢~.~w 3.3
C9~3 398v GS~3 393v GAE—V Chad C9u$ 3.3V 3 #03 $9un

 

152

Hug“ a 303.8.

 

Canvas. mndou H35 ”3.. .5, <8 €03. mneou mndon .435 > 8.8.. n- ...—.35 w
ﬂaw wow 5...... $52. 8.9. g
000 98 9.3 ~39~m 23.8 8. . m @980 «93¢ ~88 A~. .m u. .3
A. .83 A-.3~V A8. . u. 39 .3. 23.8. 29. 3 :98. A93. 398. 393.
>295. 98 ...q 3th .q..~o 3.8 8.3 3.8. ~98 3.8 8.8
0.82.08 A98. BN8. 393. 34.3. 23.9.. :92. 23.3. 3.8. :98. 393.
>030 -93 ~.3 ~8.3m 39m~ 09.3 8.3 3.... ~93 8.8 8....
A98. :93. 398. 098$ 293$ :93. :98. A938 298. 699..
02.9.8 -9~m u .8 ~A9mm .098 09. . 3.8 49.3 ~93 A9~w 8.3
0.82.9 3.38 33.8. a . .3. 39.3. 298V :98. A .93. :93. 6.33 A~9 .~.
OUU -. .~A 9~. ~A9wo . .98 .893 09?. 3.3 -.Am 8.8 m . .3
A98. A.3.-. 23.33 63 .3. C9 .3 $99.. 298. Sud A. ..A3 29~$
035 @203». -93 3.3 ~A9om 333 098 3... 3..~ ~83 3.8 8....
8.8V :98. 33.8. 3988 :98. :93. 23.3. 3.3. 33.93 @938
.9va -9~. Pom ~38” ~83... 3.3 3.9. 8.3 393 398 q9~.
3.33 393. $93. 89.3. 23.3. C9 .3. 33.8. 6&3 6.3. G... .3
28.5. Got 92 93 ~mm.3~ .3. .e 898 3.3 q . .3. ~98 5. . .~m 3.8.
G. . d A. 9.3. 393. 393. :98. A. .938 298. 8.3. A. ..3v 3388
Cam Dot ..3 ..ma ~82. 39 .c 3.8 8.3 8.3 ~98 A93 8.~.
A. ..a~. A. 98V 398. as... A. 95 C9 .8 A. 9~$ 3x3. 2938 @9qu

 

22mm 8 Shin 9 macaw wamngmsamaaaoa Ewan». o»- mndou OBS-€02. ..amnomman. o: mndov no.3 Ba madou €03. ...—dzm .wamugmgsama.Noa «8.98. om
H85 .w Rmammaa o: ...B..m >. .38 Eacuwomcoamo 532:0: Won. m8c San W... <m~H<~1mE=Q 0.. Do 8888. €0mdu€$oo=m5 0M5. 833m ...wmw
33203.20 9.33. macaw n-ﬁumndov mica-€02.“ EON—5&2 “.8852“. macaw... Q>Uuma=o8=~aq assoc. 9.82.2.“ EmUHuOandcamno 3.8m
Ema—do... OOU-I-oammmm.<o 83.2530 Ema—an... >UIUum2o=no=damo.QrvﬁaBoQSQ 982.0... 006338.325. mama: macaw... >meugmmoam.
3303.5 macaw... 038802.38.

153

.320 3

:33 M833®303m33§ 3.3333339. 033.333 983.933 3.3 :8 0939.333 MQSEm

 

 

Ummoaﬂ. macaw HEM—m W83 We. «<03, mqoov mndou 3.3% mndou H35
W8 Wow :53 <8. 203 83.. > 0-9. w
20 $8an coo -38 N333 333.3 @933 3:3. 395 N333 3.8 30.3
253 23.38 Emmi: Gmme A538 C3b8 C938 Sb 5 Quuwv A563
ZOO L .3 9% N353 53.03 093 59% 3_ .8 No.3 make 3. _ 3
3.03 Awohé wamV wa8 Cwbov :3.on 23b: Sch“: 3938 A233
9335.35 Emu -93 macaw .33.~A 5P3 Sm?“ 33.m3 ~3.~o 3.3 33.2
3.ro CNmS $0; 3 23.33 Cwbuv C 3.33 C fwd 3.va $.me G 333
088376 -93 o. 3 w 39.6 3.53 ombm 2: .Nm 3~3© mmam no.3 3.?
Q8303 3.38 6o; 3 3N.w3v A3333 2 n.3v 2 3 b8 C933 3.03 C _.3v 33.33
Q>U L .3 ~.3_ Nambm 39mm 093 2: .mo 33.33 ~83 33.3 3.3
3.03 A83 3 Gobov Amwué 2 938 33.38 393.3 Gme 33.25 0935
momma 35ch P3 3 .mm wwﬂwo 3N3 €730 She 3.3 um. I 3.3 ach
3.33 Awwbﬁ $3.va 33.33 23.38 235$ 993$ 3&3 23b3 GPOC
mwnommo 3533 Pam owe Nmobw _3N.om 3.3 Sobo 33.3 N93 2.3 393M
3.33 333$ A328 A3u.33v 29%: 2 0.3V 2 3.3V 8&8 C ruuv Awwbov
meU -w. 3 o who Nmm.w_ $33 8.3 _ co. 3 3a.: N333 3.3 93.3
6.33 2 mp8 ENNC A3908 2 3.8V 2 3.33 2 #38 3.38 GNMV C was:

 

154

HmEo u Anon—.3.

 

 

Dion—on maﬁov HEM—m W83 NH €03. onow $88 Hana mneou .33?
W8 W8 EEG <8. 20a 023 > 0)! w
OOU an: fuw 89$ 3 _ hm mobo 098 «9.3 ~93 mobo $33
God 3 903 Gobwv 69:: 3:5 2 9ch ©9ro $68 @9me Cwbuv
>353. L .oo ﬁg 59% Rube om. _ o och.» «Pam mmuu ﬁuo 8.3
Ummoaanm 3&3 ﬁroé 39ro 8930 Cam: 2 9va C903 3&3 Cnhuv 383v
>UEO -93 Pmm Nu H .3 $98 098 093 Q93 Nme m _ be 3...:
9&3 2.99: am. 5v $5va A568 293.3 Cwa 8.qu C #95 Cabcv
.9me -93 who NS .um moumu 3.3 3.3 ovum 3.3 wm.mm 3.8
9.ch 39$: 39qu 39%: A593 $9.3 Curmov GUS 3.33 98.9
>325— Uou L .3 -93 39$ 3 :3 Saw coho awn: u9~m ~93 man
:90 3 3 93v 2: .wov 39qu 2+th 2 9qu thov 3.9: C #03 AB .mé
Dam U8 L .3 9am 39am Erma 09% 8.3 3.3 ~93 3.3 m9:
Cobwv 29va $9on 8?me 2903 3533 29on Sud Copwv 69mg

 

223 8 ~6va N @306 Wamncsmgamamwoa Swag. o». 98% Gog—réoa BmRmmaa on $88 no.9. Ba mndou €29 Ham—m
Wamugmgamamwom Human». om HEW—m w namaammaa on Ham—m >w womc EEGHWmmuosmo 33338 8 59550; g mﬂou Emu”: Wamuosma
1:83 NA. <Bn<mnmE=Q o». mo 8.6030 #05 mﬁou Hmmww ﬁnmanﬁmmooamms 08d monwsm Hmmw Ramoénmaé 2.3%“ made“. 0.
ﬁnmcdou Oo_o7¢<oaw ZUCHBﬁQ. aawnammmé Emoaan Ogumgﬂm—ﬁaa ”552% Emoaan meUHdOm?Qm:Bmao 30mm E8305
OOUHocmommmé ooBcEmza Emoaon >63Cum—nasaozdomoWQEwuaBoniQ E839.“ OUUnoqumana momma Emoaon
>meHm=mmooE 3385:? ammo—don Unwumowgaanoo.

155

.920 m

chwﬁanca .3339. dominem Ham»... S :8 0.05333 9.9%? «aﬂoat

 

 

Hum—n ”a <8. made—V HEW—m (<an Hag—m > madou mndou mndov H3=m w v.20

”85E: x855. woqm. 23o $53 005.. 00—9..

madam éoa

Wow—u 53652. .31. ace... .8... -.om... bar. 33...... ..NN...... 93...... .NNI -. _ .7...
NH <8. -- Joe... .3... 25...... Mai. -. _ mi. 33...... 33...... but. 23......
madcw www.95— -- -- 35...... ._ 1.... -. _ NI. .8 .oo .31. -. . ,1... No.1.
Ham—m -- -- -- -.Noi. .oo 38...... -.Nm: 38...... .31. -.u _ ......
”85.5—
<<an Hum—m. -- -- -- -- 35...... be... .31. No.1. ANN: .31.
macaw
H85 > :30 -- -- .. -- -- 33...... ..31. 23...... .31. 9:1.
onou €02. -- -- -- -- -- -- .22.... .32. 33...... bat.
madon 00—0.. -- -- -- .. -- -- -- .33.. 33...... be:
mndou 02o..- -- -- -- -- -- -- -- -- -..—mi. .5 _ ......
(<03
H85 w -- .. -- : -- -- -- -- -- -..—mi.

 

156

208... 8 ﬁgk m. Womu Egmosuwgbosma r5350: 8 80858 3. mac Ema: Womvosmm .389 WA. <33.»ng om mo
momuosma #05 m8“. .355 $32. Womacm—chmgamamﬁa Swag. om mneou 0807303 ..omammoa on macaw 022. 8a macaw
$2.9 H35 Wommacw—ngmgbamaamNma Bang 0». H85 w nnmaammoa o: HEW—m >w gamﬁnﬁmmooammb 02d monmsm Hmmw Raga—dag
9.38“ mmBHmcz mom—o WEE—ﬁasco 98:9?

.92.... o

68.35265 858%.». ﬂowing Shin. S :8 E8855 Main? «3338

 

$.an

 

Hmmw ”a. <3 msdou H35 H35 > madam. £32. madov H85 w mEO
W888: ”855— woa. 38a €08 00.2 00.07
madam €02.
W8“. 85530: .33.. so; .3 -.oo .81. Ami... 33...... 33... but. -. — mi.
MG. <2. -- -. _ N... be Jog but. -. _ 9.... L _ .. -. _ u... but. ANN
93oz mamas»: : -- 33...... be -. _ u .oo ...oo 90...... ..No..... but.
H85 -- -- -- -. _ «...... .co swat. 38...... -.Nm: .31. Aunt.
www.58—
<<an -- -- -- : -.om .om ._ .... .5... -. ~ «...... .3
Han—m > 88 -- -- -- -- -- 33...... -..: ...... 33...... 68...... 38......
manoov €03 -- -- -- -- -- -- 60...... .31. 13...... .33..
macaw no.2 -- -- -- -- -- -- -- .31. 23...... be:
$32. 02o? -- -- -- -- -- -- -- -- J31. .33..
€08
H85 w -- -- -- -- -- -- -- -- -- 33......

 

157

228. 8 38% b. W88 Eacwmosnwguoama 839308 8 8038.8 3. m8“. Ems»— womuosma .289 .3. <38»?me om mo
88850 $08 mﬁou H85 onou Wammacm—chmgaﬁﬁuoa Homag— om mneou Oo_o~-€oa Rmammmoa on $33. 088. 88 £83.
€39 HEW—m WamEcaungbamamNoa 8895: om .585 w Rmaammaa 0: H85 >w €Oman€mm8=m8 03a madam Hmmw 60820895
9.88“ mmHOHwE 88180—538 scone?

.32”. 3

002.2885 mﬁozmh Gaming 3&8 S Sm .mblb MES»? «:1 a»

 

 

Hmmw WH <m~ 93% H8? €03. .58? > macaw waded wnooc Ham—m w mmE

W855— Womaca ma? 38a €03 00—3 00—07

madam €03

Wow—u 85320: .N _ ...... ..Nm...* .8 -. _ N .3... -.NN -. _ a... -.w _ ...... .3... -.N _ ......
.3. <2. -- -.3 .NA: -.Nwi. .8 ..NA: -.No: ..NA: .31. ..Nm.....
$32. womaca -- -- -.om .3... -.oc .oo .oo 98...... Joe .3...
H35 -- -- -- ..Nm..... .8 -.NN..... -.Nm: -.N . ...... .31. ..Nm.....
W83...»—
<<Om._. v08. -- -- -- -- J: .3... .3 .N__..... 33...... .uNi.
madam
.58? > :80 -- -- -- -- -- 33...... 33...... -.uN: .moi. .-. _ m...
msdon €03 -- -- -- .. -- -- .33.... .31. 33...... Nut.
98% 020.. -- -- -- -- -- -- -- .33.... 33...... .33..
@306 00—9.. -- -- -- -- -- -- -- -- Jun—...... .Nmi.
(<03
H85 w -- -- -- -- -- -- -- -- -- -01....

 

158

223 8 32m 3. W88 Signosuwguosmo 83288 8 808883 3. m8“. Emma ”86030 .389 NH «~333me c». mo
nomuoamo $08 meow .555 mnoow memacaugmﬁnmmamwaa 8&8»— om onou 00—2238 nomnommoa o: onou 028. 88 onou
$03“ .58? ﬂammacﬁucsmﬁbamaﬁoa 8&9.»— om 5.85 w nmmnammoa o: HEW—m >w gnmanémmoosmm: 08d momma—m Hmmw 3339.35
9.38“ mmEHmE moan 88:580." @8288.

.390 Z

OoumeQQE. 359%: ﬁgmmim 9.883me S 3w Goibmamm main? «unﬁt

 

 

9.839. me Doe O>U >moB was». mean 000 meU >3 >9 OD ODD GEE «53V 20 Dam
EU
:00 .Nm .eu .Nb bA .3 . 3 m .3 .5 .Nm .3 ..bm bA -.3 be .3 ba
0% -- .3 .: ..bA .5 bu bx .3 . _ A bu . 3 ._b bw bm .3 _ bm
Uneammm .. -- ..e be .3 . _ w .5 ._ e .Na bu -bbw be bN .3 ba bu
O>U -- .. -- ..m . 3 e bu .5 bm .AN .3 -.3 ._N . _ b -.3 ..be ba
>moa -- -- .. -- bu -b _, ..bN bu . 3 w -bm .3 ..bN ..bw .3 .3 bm
moor: -- -- -- I -- bu . _ e bu .Am bu bm -bm bm .3 -.3 bb
meoommn -- -- -- .. -- -- be ba bm -. _ _ -b3 -.3 -be bu bu -bbw
OOU -- -- I -- -- -- -- -bw .NN bbN bbb -bN bbA .3 .3 ._ _
3.2V -- -- -- -- -- -- -- .. .wA -bw ...3 -bA -bw bu be. be
323% -- -- .. -- -- -- -- -- -- -bN bw bN bb .3 bN be
>UIU -- -- -- -- .. -- -- -- -- -- -ba be .3 ..be .. _ w bm
OD -- -- -- -- : -- .. -- -- -- -- . _ g .m w .5 .3 .3 w
OUU -- -- -- -- .. -- -- .. -- -- -- -- .Ne ..bw -bh -b.
GEE -- -- -- -- -- -- -- u- -- -- .. -- -- . 3 m bb .5
.9me -- -- .. -- -- -- -- E -- -- -- -- -- -- .wm . _ a
28:2 -- -- -- .. -- -- -- -- -- -- -- -- -- -- -- ._ m

 

22mm 8 Hawk N N. mono—macaw Wbm mmmamoma m: .cAbM 09.85303 W: amamoma m” 3.3.

KUDHBmWS maenommmé Emoaon Demumwmngqamw Uaeaoeammuaaenommmé ammoaoaw Oénmasanﬂﬁaa 95x53. ammo—dam

>moBHmmoBeroEmw moomw—Hmooma 15ng meoneaomw—onmeoommo 30ng OOUHocmammzo 8335;." $839.“ meUHeOm?
35330 $83 ammoaon EgmoQﬂmH—xmﬂw 98:53“ EIUnm—ngaosdomo335880335. ammo—inn OUuoosmcoﬁ ammo—don
OUUHoEvOmEoE: momma E839.“ Dianna—broom axSBEWNEm Emoamamw >m3>menmbmmoo§ “Samoan—5. Emoaan

>_o\>_ooro_uw_ooro_ @393:an Camumam batman—80.

159

H35 5

02.83293.» 35933 533.38 Dacia; S Sm Encwomus Main? «333%

 

9839. 05 Dow Q>U >moB moor: mean OCD eHmU >3 >9 OD ODD GEE .5.—V >_o Dam

 

m5

250 .Nm .e3 .3 bu .3 .3 .3 .2 .me be -ba be -be. .3 -be be
Ova... -- .33 .3 -bm .3 bu .3 .2 .3 .Nb .2 ..e .3 bw .3 .3
U833 -- -- . _ e bu .3 Lu .3 .2 .Ne . 3 -b_ bm b3 .3 -bm be
Q>U -- .. .. .Nb .3 .3 .3 be .3 .3 -bN .ub bb bN -. _ N .3
>moB S -- -- .. bu -bN ..bN bu .Nb ..bN b_ -b_ ..be -b— -bw .3
moan. -- -- -- -- -- bm .3 bu .3u bu bm -bm ._ _ bN bu bu
meoommn -- -- -- -- -- -- .3 bu n: b_ -bu -bu -bw L: -b3 b _

OCD -- -- .. -- -- .. -- -bu ...: -be. -bb_ -b_ -.3 bu .38 be
3.2V -- : -- -- -- .. .. -- .3 bA -bN -bN bbu bu -b. . 3
>333. -- -- -- -- -- -- -- -- -- .3 bN .3 b3 -b_ -bu ._ _

>U=U -- -- -- -- -- -- : -- -- -- -bS .Nu .uq -bu bu ..3
OD -- -- u- -- -- -- -- -- -- : -- bm .ee. km .3 . _ a
ODD E E -- -- -- -- -- .. -- -- -- .. . _ e .3 -ba -bN
GEE -- -- -- -- -- -- -- -- -- -- -- -- -- .3 be . _ u

>meU -- -- -- -- -- -- -- -- -- -- -- -- -- -- .Ne . _ a
28.6— -- -- - .. .. -- -- -- -- -- -- -- -- -- -- .N _

160

 

223. 8 33m 3. Gonoﬁmoa W3 mmmamombﬁ a .mAbmw 09.85303 W._u mmmamoma m» hAbr

ZUUanue. maeaammmsw E830: menawmﬁrvqamw ereaenammuaaeammmé Emoaanmw mgumosaaﬁmnoa 8&an E830:
>moBummoBeroER moomanmoomm— 2529 weoneaommoumeoommo 252m“ OOUHOUmammmﬁw coats—mug E8305 meUueOm?
5:338 30% Emoaon Egmoaﬂgxmaa. Emoamnmw gmvumngmosdamomgéaamoaia. Emoaan OUHoosacoH Emoaan
OUUHoeeOmanm— gamma Emoaon Gianna—groom oﬁaBm—ﬁgm Emoaoaw >m3>mednm=mmooma Ramona—5. ammoaon
>_o\>_oo:o_nm_ooro_ aneonaaaoow UEmHBdm uneasagoo.

H520 ﬂu

00:05:05 33053.3 582.50 98303 5 S0 mbﬁb @5520 «:1 eue

 

 

Q0032 9.0 U03 O>U >m08 m005_ 300 000 meU >3 2! OD OCD GEE ..rm_V >_0 Dam
IU

250 .3 bu .3 E ._ _ -bN .3 b_ .3 .Nu J 3 .u _ .Nu Eb— .uu -b_
90 E .3 bu E be be -bu bu .3 .3 .3 .3 be -bA -bN .3
603000 E E .3 E . u u J3 bu bu .Nb .3 -bu . _ u .3 -bu . _ u bu
O>U E E E E .3 -bu . _ u -bu .uu .3 .3 .3 .3 -bu bu -bN
>m08 E E E E E E E E E E E E E E E E
meow». E E E E E bu .3 -b3 .u3 -bN -bu -be -bu ..bu J u _ .3
200030 E E E E E E -bN .3 .uu bu -b3 ._u .3 ..bu ..bu -bu
000 E E E E E E E -bN .NN .3 -bw -bu be -bu .b3 .3
EMU E E E E E E E E .u u -b. -bu J3 ..bu -bN bu -b3
>550? E E E E E E E E E . 3 -bm bu .3. -bu boo bw
>030 E E E E E E E E E E .3 . _ 3 .eu . u u .N _ .3
OD E E E E E E E E .. E E .3 . _ 3 . 3 bm bu
ODD E E E E E E E E E E E E .3 -bu . _ u J3
GEE E E E E E E E E E E E E E .u u .NN . _ u
>va E E E E E E E E E E E E E E .wu .N _
20032 E E E E E E E E E E E E E E E . _ u

 

230.0 8 3.230 C

. 09.85203 W. 3 mumauowa m» hAbuw 000853050 W3 053508: a bAbu.

KUUuBmu 00 05380070 E00305 meuuwmﬂgﬁumw U0ue0e8mmnu0u8mmm<0 E00253“ Q>Uum0=0~mmm0u 088$. E00303

>m08nmm08erou5w m005_u0005_ 2598 mu00\me0050um_00050 @3065“ 0930300030 02509—050 @5383 meUHeOm?

5:830 £88 98305 >=ﬁ>=xm0Qumbxm0Q 800808“ >UIUumz0=uoad0m3350080333. 900803 OUHooREQ E00305
OUUHoEVOmEosm— @353 $0025: OEEHoEEroou 9:035:55 E00258“ >mw\>meDumbamO05_ Ramon—Ema. E00305

>_0\>_0oro_nw_00:o_ u0u0=a0=00w UEmHuEm 83030300.

161

3520 33

09.85305. 338533 93.03 03.83.9303 3.: 338 605333.303 @3330 «auukt

 

0300032. 030 003 Q>0 >m08 moon: $80 000 130 >=x >03 >m3V >_0 02m

 

:0

300 -b3 .eu bu E .3 bu bu .3 .3 J3 .3 ._ _ bu
030 E .uu b3 E bu bw -bu -bm b3 .3 -bu bu bm
003800 E E b3 E .3 be bu .3 . _ u J 3 .3 ._ _ bu
Q>0 E E E E .3 .3 .Nb bN .uo .3 -bu -bu .33
>moB E E E E E E E E E E E E E
moo5. E E E E E .3 .uu -bu .uu bu bu -bu -bN
$000330 E E E E E E .3 .33 .uu bu b3 -bN .3
000 E E E E E E E -.3 .3 .3 3 -bw -bN J3
330 E E E E E E E E .3 -bu bu bu bu
3:3033 E E E E E E E E E . _ N bu -bN . 3 u
>020 E E E E E E E E E E -bu be be
>mw0 E E E E E E E E E E E ._ _ be
>_0o=o_ E E E E E E E E E E E E .3

 

220m 8 33330 3. 0985303 Wbm mum—3308: m" ﬁAbuw 000853030 WE 05333083 53 hAbH.

30035533 3030800730 33003303 030M3303355w 003300300030300080 33033083 Q>0nm0=083N03 0830.03 3309.305
>moBHmmonrou5w moo53nmo05_ @3035“ mu003mu00330Hmu00330 33035“ OO0Houm0mmu<0 0089338 302.303 3.30%?
#538530 03000 33000305 >=x3>=x30Qnmbx30Q 338308“ >0003303303730303333080933~ 3309.303.“ >m3>mw0nm=3moomm3
300895333 3000305 30320030730033 3030330300“ 0Emn32m 3030330300.

162

.320 3

00:65:03 magma 9%.me Dwain; N: Sm 58355 3.8%? «333%

 

 

Ummoaon 05 03 Q>U >mo3 moan. mvao OCD mam—u >=x “NW >mv >_n 02m
ECU .oo .3 .5. -- .3 .3 be .3 .8 -- .3 .3 .8
Ova. -- .3 -.ow -- -.oN -.cw .oc -.om -.cu -- ...3 .3 ..2
538mm -- -- .3 -- .3 .8 be .3 .3 -- .3 .3 .8.
Q>U -- -- -- -- .5 .Nm .3 .3 .mo -- Jog .8 .oo
>mo3 .. -- -- -- -- -- -- -- -- -- -- -- --
meow». -- -- -- -- -- .8 -.S ..8 .mo -- .3 new -.ow
muoommo -- -- .. -- -- -- .B .8 .3; -- .oo ...8 . 3
COO -- -- -- -- -- -- -- -.S .3 -- -.om ..o. -.S
360 -- -- -- -- -- -- -- -- .mo -- .3 .3 .8
>350. -- -- -- -- -- -- -- -- -- -- . _ o .3 . _ o
>036 -- -- -- -- .. -- -- -- -- : -- -- --
.9va -- -- -- -- -- -- -- -- -- -- -- .3 . _ o
2832 -- -- -- -- -- -- -- -- -- -- -- -- . _ q

 

22mm 8 New? C. 02.85325 WE mmmamoma m» hAbmw 02.85303 W. _ u mmmamnma m” EAS.

ZUUuBmHon amunammmaa E830: meuawmﬁrwammw Unwaacaommnaouaommmé ammoaaaw Q>Uuma=anﬁmwaa Edna? Emoaan
>mo8ummo§urogﬁ moomanmoomm— 2.53m“ mRQmuaomw—oumuoommo BEER OOUuocmammmaa ooBuEmmé Emoamn meUnuoﬁ-
"3:830 33m E839.“ EgmaQumsxmoQ ammo—dag >UICum—agnosdmmomgwuaamomSQ Emoaon OUHoosacon ammoaon
OUUuoqummmoE: momma Emoaan OEEHnE—Qrooa axﬂaammﬁnm ammoaoaw >m3>mwvumbmmoo§ @030:sz Ewen—on
>_o\>_ooro_nm_oo=o_ 9060:3058“ Damnaam 3060:3058.

163

ngn 3

Goxmeaoa 232%.: 05.2%: 02.23%; S Sm 2050 main? «nub ewe

 

 

9832. 0km 0% Q>0 >moB mega mean 000 v.30 >=x WW >2U >_o 02m
300 -- -- -- -- -- -- -- -- -- -- -- : --
0km -- _ b be -- bu be -bN -bN be be -bA bm .3
088mm -- -- be -- bu be -bN -bw be be -bA bu b _
Q>0 .. -- -- -- .NA -b3 .3 -bm ..NN . _ e -bm -b3 -bN
>m2.~ .. -- -- -- -- -- -- -- -- -- -- -- --
moor: -- -- -- -- -- be LN -bm .3 bu -ba -bm -bN
memommo -- -- -- -- -- -- -bN ..3 .AN .3 -bw -bA -b.
000 -- -- : -- -- -- -- -b. .NA .5 -bN -bN -b_
3&0 -- -- -- -- -- -- -- -- .3 -b3 -b. -b_ -b_
>333~ -- -- .. -- -- -- .. -- -- .3 ..be -bm ..bu
>030 -- -- -- -- -- -- -- -- -- -- .eb . _ e. be
>me0 -- -- -- -- -- -- -- -- -- -- -- .3 ..3
>832 -- -- -- -- -- -- -- -- -- -- -- -- -b _

 

22mm 8 Hawk 3. 003.2283 W. 3 imam—252 2 ﬁAbww 02.85325 Wbe mmmammoma 2 «Kb?

300358.02. 322.825 Emoaon 0<mua<m9v§mﬁ 0oe0oeqommuaae8mmm<a Emoaoamw Q>0umo=a3=Nab 232% 28303
>mo~mnmmo8erocmmw moomanmoamm— 3532 meoo\me02monmea2mo 1532 OO0Ho¢mmmmm<n 0235578 362.32.“ 130402-
352on 28% Emoaan >8Q>5£a€ﬂ§ia€ 9832.2 >0:0umnosmochamomgvﬁozﬁnsa. Emoaon 0032.92 ammo—awn
000338582 332: 28805 Gianna—groom Shamanism E8392 >m3>me0nm=amoo§ 332550. 28255
2228372852 manna—bosom“ 0an2dm aoegaasoa.

164

 

I Simple Phobia J\
i f .20
I Social Phobia |‘\
.43
’ .35
I OCD IA/
.44
MDD I/
7 .24
$| Dysthymia I/

Figure 3. Final accepted latent model for internalizing disorders.

 

 

 

 

 

.16

 

Notes to Figure 3. F it statistics: 752 (8) = 8.30, p > .05, GFI = .996, CFI = .998, RMSEA =
.008. All factor loadings were signiﬁcant (p<.01). GAD = generalized anxiety disorder;
OCD = obsessive-compulsive disorder; MDD = major depressive disorder.

 

I Alcohol Dependence I$

 

 

| Drug Dependence Iq. .30

I ASPD It ’

I Childhood Extemalizing IA’ '

 

Extemalizing

 

    

 

 

Figure 4. Final accepted latent model for externalizing disorders.

Notes to Figure 4. Fit statistics: x2 (2) = 8.32, p < .05; GF I = .99, CFI = .95, RMSEA =
.07. All factor loadings were signiﬁcant (p<.01). ASPD=antisocial personality disorder;
Childhood Extemalizing=ADHD, ODD, and CD.

165

 

 

I Stroop Residual |'\

’ -.28
I Trails Residual I“ .43 _¥
Resp Inhibition '4’ -21 ——
RT Variability I4/ '

-.43

l WCST PE |/

    
 

Executive
Functions

 

 

 

g"
\l

 

Figure 5. Final accepted latent model for executive function tests.

Notes to Figure 5. Fit statistics: x2 (4) = 1.50, p > .05; GFI = .999, CFI = 1.00, RMSEA
= 0.00. All factor loadings were signiﬁcant (p<.01). Stroop Residual=nonstandardized
residual of Strop Color-Word regressed on Stroop Color and Stroop Word; Trails
Residualmonstandardized residual of Trails B regressed on Trails A; Resp
Inhibition=Response Inhibition as measured by Stop Signal Response Time; RT
Variability=variability of go response from Stop Task; WCST PE = Wisconsin Card
Sorting Test perseverative errors standard score.

166

 

I Stroop Residual I‘\
: -.I7
I Trails Residual I“ .43 ‘

 

    
 
 
    

 

Executive
Functions

 

 

 

 

I WCSTPE IA/ .62

I Stroop Word I“ ‘80
I Stroop Color I‘—— _.85 ; Processing

Speed
L TrailsA '4’ -48 "’

 

 

 

 

 

Figure 6. Latent measurement model for cognitive test performance.

Notes to Figure 6. Fit statistics: 7808) = 93.21, p < .05; GFI=.97, CFI=.92,
RMSEA=0.08. All factor loadings were signiﬁcant (p<.01). Stroop
Residual=nonstandardized residual of Strop Color-Word regressed on Stroop Color and
Stroop Word; Trails Residual=nonstandardized residual of Trails B regressed on Trails
A; Resp Inhibition=Response Inhibition as measured by Stop Signal Response Time; RT

Variability=variability of go response from Stop Task; WCST PE=Wisconsin Card
Sorting Test perseverative errors standard score.

167

 

 

 

 

 

 

 

o . . + Dep
K + Anx
.o_1 N + Alc
/\ 2/ + Child
-o.2

 

 

 

 

 

 

 

 

 

‘x/ +Comorbid
-o.3 A

\ / \ /

V

-0.6

Stroop Trails Inhib Var WCST

Figure 7. Mean z-scores for performance on executive function tests for lifetime
disorders with nonoverlapping diagnostic groups.

Notes to Figure 7. Stroop=nonstandardized residual of Strop Color-Word regressed on
Stroop Color and Stroop Word; Trails=nonstandardized residual of Trails B regressed on
Trails A; Inhib=Response Inhibition as measured by Stop Signal Response Time;
Var=variability of go response from Stop Task; WCST=Wisconsin Card Sorting Test
perseverative errors standard score; Dep=depressive disorders; Anx=anxiety disorders;
Alc=alcohol dependence; Child=childhood externalizing disorders; Comorbid=multiple
disorders. Note that z-scores for Trails, SSRT, and SDX were reverse scored so that
lower scores reﬂected poorer performance.

168

0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
-0.1
-0.2
-0.3
-0.4
-0.5
-0.6
-0.7
-0.8
-0.9
-1
-1.1
-1.2

 

+ Control
+ Dep

+ Anx

—EI— Alc

+ ASPD

—l— ADHD

—>K— Comorbid

 

 

 

 

Stroop Trails Inhib Var WCST

Figure 8. Mean z-scores for performance on executive function tests for current
disorders with nonoverlapping diagnostic groups.

Notes to Figure 8. Stroop=nonstandardized residual of Strop Color-Word regressed on
Stroop Color and Stroop Word; Trails=nonstandardized residual of Trails B regressed on
Trails A; Inhib=Response Inhibition as measured by Stop Signal Response Time;
Var=variability of go response from Stop Task; WCST=Wisconsin Card Sorting Test
perseverative errors standard score; Dep=depressive disorders; Anx=anxiety disorders;
Alc=alcohol dependence; ASPD=antisocial personality disorder; ADHD=attention-
deﬁcit/hyperactivity disorder; Child=childhood externalizing disorders;
Comorbid=multiple disorders. Note that z-scores for Trails, SSRT, and SDX were
reverse scored so that lower scores reﬂected poorer performance.

169

0.5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.4
0.3
x + Dep
+ Anx
+ Alc
—l— Drug
+ ASPD
+ Child
-0.4 V
-o.5 ,.
Stroop Trails ‘ Inhib Var WCST

Figure 9. Mean z-scores for performance on executive function tests for lifetime
disorders with overlapping diagnostic groups.

Notes to Figure 9. Stroop=nonstandardized residual of Strop Color-Word regressed on
StrOOp Color and Stroop Word; Trails=nonstandardized residual of Trails B regressed on
Trails A; Inhib=Response Inhibition as measured by Stop Signal Response Time;
Var=variability of go response from Stop Task; WCST=Wisconsin Card Sorting Test
perseverative errors standard score; Dep=depressive disorders; Anx=anxiety disorders;
Alc=alcohol dependence; Drug=drug dependence; ASPD=antisocial personality disorder;
Child=childhood externalizing disorders. Note that z-scores for Trails, SSRT, and SDX
were reverse scored so that lower scores reﬂected poorer performance.

170

0.3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.2
0.1
0
+ Dep
-0.1 +Anx
-o.2 + A":
—l— Drug
-0.3 + ASPD
+ ADHD
-0.4
-0.5
-0.6
-0.7

 

Stroop Trails, Inhib Var wcsr

Figure 10. Mean z-scores for performance on executive function tests for current
disorders with overlapping diagnostic groups.

Notes to Figure 10. Stroop=nonstandardized residual of Strop Color-Word regressed on
Stroop Color and Stroop Word; Trails=nonstandardized residual of Trails B regressed on
Trails A; Inhib=Response Inhibition as measured by Stop Signal Response Time;
Var=variability of go response from Stop Task; WCST=Wisconsin Card Sorting Test
perseverative errors standard score; Dep=depressive disorders; Anx=anxiety disorders;
Alc=alcohol dependence; Drug=drug dependence; ASPD=antisocia1 personality disorder;
Child=childhood externalizing disorders. Note that z-scores for Trails, SSRT, and SDX
were reverse scored so that lower scores reﬂected poorer performance.

171

Table 17

Summary of Regression Analysis for Lifetime Disorders Predicting Performance on
Executive Function Composite Score

 

 

Disorders B SE B B p
Depressive 0.05 0.26 0.01 .8
Anxiety 0.38 0.28 0.05 .19
Alcohol 0.23 0.27 0.04 .40
Dependence

Drug ~0.26 0.50 -0.02 .61
Dependence

ASPD -0.10 0.42 -0.01 .81
Child -1.17 0.28 -0.17 <.001

 

Notes to Table I 7. R2 = .032. ASPD=antisocia1 personality disorder; Chi1d=chjldhood
externalizing diorders (attention-deﬁcit/hyperactivity disorder, conduct disorder,
oppositional deﬁant disorder). Results remain after controlling for age, gender, and

F SIQ.
Table 18

Summary of Regression Analysis for Current Disorders Predicting Performance on
Executive Function Composite Score

 

 

Disorders B SE B B p
Depressive 0.24 0.33 0.03 .46
Anxiety -0. 12 0.38 -0.01 .76
Alcohol -0.76 0.50 -0.06 .13
Dependence

Drug -0.18 1.11 -0.01 .87
Dependence

ASPD -0.37 0.39 -0.04 .34
Adult ADHD -1.18 0.38 -0.19 <.001

 

Notes to Table 18. R2 = .042. ASPD==antisocial personality disorder; ADHD=attention-
deﬁcit/hyperactivity disorder. Results remain after controlling for age, gender, and FSIQ.

172

Table 19

Summary of Regression Analysis for Lifetime Disorders Predicting Performance on
Processing Speed Composite Score

 

 

Disorders B SE B B p
Depression -0. 1 6 0.21 -0.03 .44
Anxiety 0.1 1 0.23 0.02 .63
Alcohol -0.57 0.22 -0.11 <.01
Dependence

Drug -0.67 0.39 -0.07 .09
Dependence

ASPD -0.49 0.33 -0.06 .14
Childhood -0.55 0.23 -O. 10 .02
Extemalizing

 

Notes to Table I 9. R2 = .045. ASPD=antisocial personality disorder; Childhood Extemalizing=attention—
deﬁcit/hyperactivity disorder, conduct disorder, oppositional deﬁant disorder. When controlling for age,
gender, and FSIQ, only child externalizing disorders were signiﬁcantly related to speed (B = -.15, p < .01),
and drug dependence approached signiﬁcance (B = -.07, p = .05).

Table 20

Summary of Regression Analysis for Current Disorders Predicting Performance on
Processing Speed Composite Score

 

 

Disorders B SE B [3 p
Depression -0.02 0.26 -0.003 .93
Anxiety 0.03 0.31 0.004 .93
Alcohol -0.64 0.40 -0.06 .1 1
Dependence

Drug -0.54 0.90 -0.02 .55
Dependence

ASPD -0.95 0.31 -0.12 <.01
Adult ADHD -0.08 0.30 -0.01 .79

 

Notes to Table 20. R2 = .042. ASPD=antisocial personality disorder; ADHD=attention-
deﬁcit/hyperactivity disorder. When controlling for age, gender, and F SIQ, ASPD was no longer
signiﬁcant and ADHD became signiﬁcant (B = -.1 l, p < .05).

173

Table 21

Frequencies of Numbers of Individual Lifetime Disorders in the Combined, Alcoholism,
and ADHD Samples

 

Number of Combined Alcoholism ADHD

 

Disorders (n=641) (n=448) (n=193)
O 170 (26.5%) 106 (23.7%) 64 (33.1%)
1 187 (29.2%) 133 (29.7%) 54 (28.0%)
2 127 (19.8%) 91 (20.3%) 36 (18.7%)
3 90 (14.0%) 70 (15.6%) 20 (10.4%)
4 43 (6.7%) 28 (6.3%) 15 (7.8%)
5 13 (2.0%) 9 (2.0%) 4 (2.1%)
6 7 (1.1%) 7 (1.6%) 0 (0%)

7 3 (0.5%) 3 (0.6%) 0 (0%)

8 O (0%) 0 (0%) O (0%)

9 1 (0.2%) l (0.2%) 0 (0%)

 

Notes to Table 21 . Disorders included were lifetime major depressive disorder, dysthymia, social
phobia, speciﬁc phobia, obsessive compulsive disorder, generalized anxiety disorder, post-
traumatic stress disorder, agoraphobia, alcohol dependence, drug dependence, ASPD, conduct
disorder, attention-deﬁcit/hyperactivity disorder, and oppositional deﬁant disorder.

Table 22

Frequencies of Numbers of Individual Current Disorders in the Combined, Alcoholism,
and ADHD Samples

 

Number of Combined Alcoholism ADHD

 

Disorders (n=641) (n=448) (n=193)
0 360 (56.2%) 262 (58.5%) 98 (50.8%)
1 188 (29.3%) 125 (27.9%) 63 (32.6%)
2 62 (9.7%) 39 (8.7%) 23 (11.9%)
3 24 (3.7%) 17 (3.8%) 7 (3.6%)
4 7(1.1%) 50.1%) 2(1.0%)

 

Notes to Table 22. Disorders included were current major depressive disorder, dysthymia, social
phobia, speciﬁc phobia, obsessive compulsive disorder, generalized anxiety disorder, post-
traumatic stress disorder, agoraphobia, alcohol dependence, drug dependence, antisocial
personality disorder, and attention-deﬁcit/hyperactivity disorder.

174

E
.3 .3

CELEBS 4’
H .0

magm— 352m

  

mmBEo 363m

003
I

2832 U3

Dam Cow

>mv0

GEE mg»

.3

.MN

.3
.uo

.Nu

  

   

mxﬁoamzNEm

     
  

mxoocmﬁw
mgnmosm

Ham—m ”83.5.

macaw Nuance.

 

wow—u EEG

3 <§sz=q

€03.

macaw €03

I
\l
\O

mbooa -.wm wqoov 00.2

.8 H35 >

+1 $56 :. ZomeoBoE Bong men 90 3520333 canine: 5.380 imam—Sum E5 gags—55m ammo—dowm o: nomads" Saw Ramayana—8 won 282:5

9:983. 2a .689

228 8 3.98 2. m: £35:an x~ 2N3 u wmqbq. h A b: 03 n .3. 02 n .3. Wamm.» H 0.0.“. map—88 833033338“<m1m2amwgaom 8382: 583

simian U n Ema—egg O>Uumgo3=Noa saga? Ewen—on >vauw=nmoo§ 8303—5 9832.

u ocmammmééoBvEmEo E830: GEE memes»: M 035308 mﬁoammuim 98302 30:5. 000
H35 w nomaommma 0: H85 >w mqoou Womacﬂgozmgzmﬁauoa Smaca o». mqoon 008795:— quommoa o: macaw (<03 Ea mqoou no.0: ﬁnmAn/Smoosmm:
Om:— moasm Ham" umnmo<o32<o 238” Wow“. ELEM—wnmvoamn 5.550: mm 532:8 3, mace Eman— womuoawo .239 WA. <m1wE:QH8mno=mo $1»ng $03

mﬁou 4.8—9

v

u

Unuuaovgamaoﬂ KUUHmeoq aauaammim 98:3: 000
9 OCH :3; Womacansosmsamamﬁa amaze. om

175

053530

.3
woos: 259.0

  

000

30050. One

03m 000

>200

03:0 manna»—

'C.‘
Q
l>ll
U

9.530 3.02»

    

0.3:». ”85.5.

 
   

   

.3
.3
.3 $00: Ewan—=0.
-. _ a
.8 58:00:55 mxoocmaa .uu wow: :53
mgomoa

 

a 553:5

. £03

.5

 

  

.mu

.uu macou $03

.3

930: 00.04
.mo

mxﬂoBEmNmH—m

  

mega -.Ma

.
'\l
o

.55
.Nm ABM—m >

ENEd E. 30%— 808835 90 9.8830 8530533 0038: 50:30 m:8_.:m=Nm:m ~50 oxSBm:i:m E8308 0: coming Smw 00:33:58 m2

980520 350303
208: S Emrxm C.

0:0 €80.
m: £0338” XNA 5.3 n $83. 0 A .o r 03 n .3. 02 H .mm. ESmm> u obs. mnzmwom 3088:. 30:48.“ Simian 0:29.. 8389: 58:"

<01».sz U M 0:81:28“ 0>Uuma:aB:Noa 0:0:wa Emoaon >vanmszmonma 00.30ng macaw: Dounnovoaogow ZUCnaag 0008mm?“ 0:325: 000
n 068329003258 05033 0E5 mﬁaaa M 0:20:00: mx830=i:m 98am; 90:6. OUU. on 00V“ 5.0% Wowacm_uamacm_ 0w .33; w Rmaomwaa 0:
H55 >w onou womacanamaca 0m w:.00_u 00_0?€0a 39880 0: @300 $08 0:0 @300 00.0: €0mﬂncﬁmogmm: 030 menim Ham» 003053910

94an W8: EEG"

”800:3 E5330: 8 5.35500 3. m8: 9mg— Wamnoaa .259 W... <5mc=5u3muoam <wlmc=§ was: 900 Hum—n.

176

0.30—m ”8.0.5.
.SUU

9300 ”8.0.5.

‘A
N

     

  

mxooﬁra
F5303

3880.05 .0“ Now“. 5.00

moam. 350.0

  

.3

. . W... <mlmc.—.Q
980.0 .000va

Q
I. 11>
U

ﬂaw...
000

.@m

o
Inge I

200.00. 00.0 @2000 £02.

  

Dam 00.0 .nm mﬁoam..N5m @000 La 9800 00.3
.3
3.00 a 3% >

.0550 ..w. Zommﬁnioa 300a. aauaomonmum 08 3.000333 00300: 95.2: .3830_.N.:m 000 0x003»..u.=m 0.82.08 00 coming Smr 001.020.0000 won
880005 050008 000 $0000.

208.0 8 Emuxm E. 2. 9000002 xw Gd n Nabﬂ u A .2... 0.”. u .00. 0.... u be. Em?» u .05. wagon 8330:" 500.3% $5022.. 0.3.00 89.82: .083
5013.8. U n 0.3500009 O>Uumn=oB=No0 .500? 0.82.0: EVE—u 5.5...5. . a . a k n n J 0.0302. >menm=0moaa 00303.5 0.82.0:
Uncu0ovaa00=oﬂ 2.00an..2 033%.?0 0.82.0: OOU u 0.083.500.5005?“ 0.80.00: 0.8.: www.0zm.n00=m§0e.0.uma Ema—5. 00. 03.; w 398.80 0:
H35 >. 9300 Wom.0=m.n=0=m§0m~0§o0 800:0. 00. was 00.03.32. nomnowmg 0: madou 00.3 ~00 mqoon €30. €0mﬂn€.m00=m.= 0&0 moi—0m ..dmn
00320305“ 0.40;. ”80 Eacuwomuoamo 50.0.00: 8 508500 0% m8“. 985. ”80008 H53. .5. <300...QI40%0=8 $503.5. 0.0:. m8“. H00».

177

ZOO H35 wnwaﬁm

C)
>
U

938 ”8&5:

\l
in
N

       

     

mega E59» .3 Macaw—Esm mxooché ma ”0% 53¢
F5953 .3
. . .2 a . . .
main 3865 WA; <mdmgrQ
.3 ..8
COO €03;
38:2 08 .8 -..3 mcdou $55
Cam 6% .Nm mxﬁmamﬁmum manna -.ma madou 00—3
um é ..

FWEd I. Zona— Suamoszzm :6 29:35 Haggai? 6038: 952: manna—FEM 8a mxHonszsm 98am; on Swiss" Saw Uaoawaoo m9. 88953
9:38am 25 $58.

228 8 +1 ﬁﬁm I. m: amazon xm Gd H 393. n A .oo_. 03 n b? OE n .3. Wme> u .3. mncmnnm 8382: 58.28» Siva—om“ Oman—am 89.82: 585
<m1mEmmw U u Emacncmsoow D>Unmn=2m:mmq 23:an Emoaon >UIU ..HSLS. “(1.an k w .. LO ammo—dam >mv0u§mm8m£ 6082.35 ammo—don
Douuaawmanasno“ ZUUHaﬂoﬁ mounommmé Egan: 000 n ocmommmééoBuEmZa $823 w H35 Womacausonmsamamuna n85.5— om ...ng w nomuommoa 0:
$3; >w macaw ﬂammacausosmﬁaamammoa ~855— o». mqon Oo_oT€o_d 898me o: mqoou 00—9. 25 macaw «<99 €OmHH€mmoo=m5 08d moi—Hm Ham»

8330335 233“ W8“. EEcuxomuoama 53533 8. 50858 3 man www.5— womcosmo ﬂan“ a. <3wc=§n8uuo=mo $5»ng m3:— wSu Hum—n.

78

l

APPENDIX
Anatomy of the Frontal Lobes and Frontal Subcortical Circuits

EF have historically been associated with activity in the prefrontal or frontal
cortex. The historical underpinnings for the connection between EF and the ﬁ'ontal lobes
come from early observations of seemingly disparate effects following frontal lobe
damage (Pennington & Ozonoﬁ‘, 1996). Therefore, the localization of these cognitive
functions preceded their conceptualization. Although the disrupted processes appeared to
be quite different from one another, they could all be understood as involving
dysregulation in goal-directed activity that could not be attributed to deﬁcits in more
basic cognitive processes; as a result it was thought that the ﬁontal lobes were involved
in “executive” or “supervisory” functions (Pennington & Ozonoff, 1996; quotes appeared
in original article).

The functional integrity of the frontal lobes (speciﬁcally prefrontal cortex) is
required to engage in “any series of purposive actions that deviates from rehearsed
automatic routing” (Fuster, 1997, p. 3), but an emphasis has also been on the vast
circuitry that connects the frontal lobes with cortical and subcortical regions (Alexander,
DeLong, & Strick, 1986; Lichter & Cummings, 2001; Stuss, Knight, & Ed, 2002). All
components of these networks are important for task performance, and lesions at any
point in the system may cause deﬁcits that look like ‘frontal’ damage, whether the lesion
site is cortical or subcortical (Lichter & Cummings, 2001; Mesulam, 2002). The
importance of other regions to EF abilities is a key reason why the ﬁontal metaphor is

inadequate for EF.

179

The frontal lobes are the area of the brain anterior to the central sulcus. Directly
in ﬁont of the central sulcus is the primary motor and premotor cortex, and anterior to
those areas are the prefrontal cortices, areas of the brain which are larger and most highly
developed in the human brain (Petrides & Pandya, 2002). This is the region of most
interest for cognitive control processes.

The prefrontal cortex appears to have several architectonically distinct regions
(Petrides & Pandya, 2002). In particular, the connections from the prefrontal cortex to
other cortical and subcortical regions are important, as they were observed to involve two
functionally and anatomically distinct systems (Pandya & Barnes, 1987). The ﬁrst
system mediates the sequential processing of sensory, spatially-related, and motivational
information through a dorsal stream, which involves the dorsolateral and medial areas of
the frontal lobes as well as interconnections with the posterior parietal lobe and cingulate
gyrus. The second system, which mediates emotional tone, is ventrally located and
involves the orbital surface of the frontal lobes as well as paralimbic regions. Thus, the
frontal lobes are the site where these partially overlapping systems involving external
sensory information and internal cognitive and emotional/limbic responses are integrated
and processed to modulate motivation and facilitate motor responses (Lichter &
Cummings, 2001). The functional signiﬁcance of these circuits will be elaborated upon
shortly.

A pattern of anatomical and functional duality has also emerged when examining
connections between the prefrontal cortex and subcortical structures, particularly the
basal ganglia. The basal ganglia is a subcortical system that is important in the regulation

and coordination of cortically-originated movement. It consists of the striatum (i.e.,

180

caudate nucleus, putamen, and nucleus accumbens), globus pallidus, substantia nigra, and
subthalarnic nucleus. Efferent projections from functionally-related areas of the
prefrontal lobe converge upon discrete areas of the striatum, which also appear to share
functional properties (Lichter & Cummings, 2001). Speciﬁcally, the dorsal system of the
frontal lobes connects to the dorsal caudate nucleus, while the ventral system maps onto
the ventromedial portion of the caudate and adjacent portions of the nucleus accumbens.
Thus, information processed by the cortex is received and processed by the striatum in a
manner that apparently maintains partial separation and specialization of functional
domains (Lichter & Cummings, 2001). Thus, the structure of the prefrontal cortex and
underlying circuitry provides the anatomical ﬁ'amework for differential processing.

The neuroanatomy of ﬁontostriatal circuits was recently extended by Middleton
and Strick (2002) to include seven hypothesized “categories” of circuits based upon
ﬁndings in primates. These circuits arise from extensive closed loop connections (i.e.,
both efferent and afferent connections) between the basal ganglia, ﬁ'ontal cortex, and
thalamus, thus supporting distinct anatomical and functional regions within the ﬁ'ontal
cortex. The term “categories” is used to denote the multiple parallel segregated circuits
that are contained within each category. Three of these categories of circuits are less
relevant to the present discussion because they involve the motor areas of the ﬁontal
lobes and posterior cortex. The main circuits of interest here are those that involve
projections between the basal ganglia and the following regions of the prefrontal cortex:
(1) dorsolateral; (2) lateral orbitoﬁontal; (3) media] orbitofrontal; and (4) anterior
cingulate (note that (3) and (4) are often combined in other conceptions). The circuits

traverse from these prefrontal regions to areas of the striatum ((1) dorsolateral caudate

181

head; (2) ventromedial caudate head; (3) ventromedial caudate and ventral striatum; and
(4) ventromedial caudate head, respectively). These striatal regions project to regions of
the globus pallidus and pars reticulata of the substantia nigra, which send efferents to
speciﬁc nuclei of the thalamus, which then project back to the prefrontal region, forming
a closed loop (Middleton & Strick, 2002). Middleton and Strick (2002) note that while
these projections mean that the basal ganglia can inﬂuence many cortical regions and
thus involve many functions, the “highly topographic and closed-loop nature” (p. 56) of
these circuits means that very speciﬁc ﬁinctional impairment can result ﬁom damage to
the structures in basal ganglia.

The cerebellum also appears to be involved in prefrontal circuitry.
Cerebrocerebellar circuits are heavily connected to the dorsolateral and dorsomedial
regions of the prefrontal cortex (Middleton & Strick, 2001; Schmahmann & Pandya,
1995). Medial regions appear to be less heavily connected with the cerebellum, while no
connections have been observed from the ventral prefrontal and orbitofrontal cortices
(Schmahmann & Pandya, 1995). In conjunction with its role in motor activity, the
cerebellum appears to be important to the same cognitive functions as the dorsolateral
preﬁ'ontal cortex. The cerebellum and this region of the prefrontal cortex appear to
interact with and modulate each other throughout development and during cognitive task
performance (Diamond, 2000). As such, the cerebellum is another important subcortical
structure that contributes to prefrontal cortex activity.

Along with the vast subcortical input into the prefrontal cortex, the frontal lobes
also receive reciprocal connections from posterior cortical areas (i.e., areas posterior to

the central sulcus). Afferent connections from posterior association cortices provide

182

highly processed sensory-speciﬁc or multimodal information to areas of the prefrontal
cortex, whose reciprocal efferent connections allow the prefrontal cortex to regulate
information processing in these posterior cortical areas (Petrides & Pandya, 2002). Thus
the preﬁ'ontal cortex exchanges information with other cortical regions.

Therefore, the ﬁontal lobes serve to integrate and modulate sensory and
perceptual information ﬁ'om distributed networks to aid in motor output. Reciprocal
circuits connect regions of the prefrontal cortex to posterior cortical association areas and
subcortical regions such as the limbic lobe, basal ganglia, and cerebellum. The prefrontal
cortex thus appears to be a site of convergence for information from overlapping yet
distinct circuits that are involved in emotion, motivation, cognitive processes, and motor
output. Recent advances in the study of frontostriatal circuits have supported the
existence of four (or, more commonly, three) distinct circuits connecting the prefrontal
cortex to subcortical regions. It is important to view these circuits as forming a “frontal
network system” (Mesulam, 2002) in order to recognize that problems at any stage in
processing may cause deﬁcits that look similar to those typically associated with the
frontal lobes. Further, due to the segregated and closed-loop nature of these circuits,
damage may result in the loss of very speciﬁc abilities.

Functional Signiﬁcance of Preﬁ'ontal Regions: Prefrontal Behavioral Syndromes

Three prototypical behavioral neurological syndromes have been observed
following injury to certain areas of the prefrontal cortex. Damage to the dorsolateral
region may result in a “ﬁontal abulic syndrome” (Mesulam, 2002), which is characterized
by a loss of initiative and creativity, reduced ability to concentrate, and a tendency

towards emotional apathy and ﬂat affect. Fuster (1997) noted that a disruption of

183

“intensive and selective” attention is at the forefront of this syndrome (p. 172), which
impairs the ability to direct or focus general arousal upon a particular sensory or internal
experience. Some patients suffer from depression, but it is difﬁcult to determine whether
it is primary or secondary to the cognitive disorder. Cognitively, these symptoms
manifest as the traditional “dysexecutive syndrome,” which involves problems with
perseveration, planning, working memory, verbal ﬂuency, and temporal organization of
behavior (Fuster, 1997).

Damage to the orbitofrontal region may result in a “frontal disinhibition
syndrome” (Mesulam, 2002), which is characterized by deﬁcits in the “exclusionary”
aspect of attention (Fuster, 1997, p. 174). Individuals with this disorder have diﬁiculty
suppressing interference from external or internal stimuli. The result is behavioral
excesses and too much drive, but these behaviors are characterized by impulsivity with
little judgment, insight, foresight, or ability to learn ﬁom experience. These patients may
disregard social conventions and show impaired moral judgment. Their affect is
generally euphoric, with irritability, contentiousness, and paranoia. Cognitively, they
have problems with focused attention (Fuster, 1997).

A third, less common syndrome, whose behavioral presentation is similar to
ﬁontal abulia has been called “akinetic mutism” (Pennington & Ozonoff, 1996) and may
result from damage to the medial/cingulate cortex (Fuster, 1997). This disorder is poorly
deﬁned, but it involves apathy and deﬁcits in the ability to initiate speech and other
spontaneous behavior. Too much anterior cingulate activity may be associated with
obsessive-compulsive and tic-like symptoms, while too little activity has been related to

diminished self-awareness and depression (Devinsky et al., 1995). Cognitively, the

184

anterior cingulate is important for response selection, motivation, and initiation
(Devinsky et al., 1995).

This summary greatly simpliﬁes the processes of the prefrontal cortex, as certain
regions are consistently activated during diverse EF processes (Duncan & Owen, 2000).
However, damage to different regions of the prefrontal cortex may result in distinct
behavioral/cognitive syndromes. Many of the ﬁmctional changes that are observed
following frontal lobe damage parallel the behavioral sequelae of psychiatric disorders,
providing a conceptual connection between ﬁ'ontal lobe circuits and psychopathology.
The dual realms of the preﬁontal cortex, which integrate cognitive as well as
motivational and emotional information (Lichter & Cummings, 2001), highlight the
means by which EF deﬁcits may be associated with both behavioral and emotional
symptomatology. Damage to or dysfunction in these regions of the brain may result not
only in cognitive changes, which would primarily involve EF, but also directly affect
emotional and behavioral functioning (Fuster, 1997; Lichter & Cummings, 2001).

Clinical parallels to psychological disorders may be drawn for each of these
neurological prefrontal syndromes. The dorsolateral-frontal abulic syndrome resembles
the symptomatology of ADHD, as well as depression to some degree. The orbitoﬁontal-
disinhibition syndrome resembles mania, antisocial personality disorder (ASPD), and the
hyperactive symptomatology sometimes associated with ADHD. Finally, the cingulate-
akinetic mutism syndrome resembles the symptoms of major depressive disorder (MDD),
while excessive cingulate activity is analogous to symptoms of anxiety disorders.
Therefore, the surface similarities between these frontal syndromes and symptom

presentations in key psychiatric disorders suggest some hypotheses for how

185

psychopathology may be related to components of EF dysfunction. Imaging research has
also provided varying degrees of support for dysfunctions in frontal regions of the brain
in ADHD, ASPD, alcoholism, substance dependence, depression, and anxiety (see
section on EF in Key Psychopathologies), and it is therefore unsurprising that EF deﬁcits

have also been associated with each of these disorders.

186

References

Achenbach, T. M. (1966). The classiﬁcation of children's psychiatric symptoms: A
factor-analytic study. Psychological Monographs: General & Applied, 80(7), 37.

Adams, K. M., Gilman, S., Koeppe, R. A., Kluin, K. J. et a1. (1993). Neuropsychological
deﬁcits are correlated with frontal hypometabolism in positron emission
tomography studies of older alcoholic patients. Alcoholism: Clinical &
Experimental Research, 1 7(2), 205-210.

Agosti, V., Nunes, E., & Levin, F. (2002). Rates of psychiatric comorbidity among US.
residents with lifetime cannabis dependence. American Journal of Drug &
Alcohol Abuse Vol, 28(4), 643-652.

Airaksinen, E., Larsson, M., & Forsell, Y. (2005). Neuropsychological ﬁrnctions in
anxiety disorders in population-based samples: Evidence of episodic memory
dysﬁlnction. Journal of Psychiatric Research, 39(2), 207-214.

Alexander, G. E., DeLong, M. R., & Strick, P. L. (1986). Parallel organization of
functionally segregated circuits linking basal ganglia and cortex. Annual Review
of Neuroscience 9(1986), 357-381.

American Psychiatric Association. (1994). Diagnostic and Statistical Manual of Mental
Disorders, Fourth Edition. Washington, DC: American Psychiatric Association.

Anderson, C. V., Bigler, E. D., & Blatter, D. D. (1995). Frontal lobe lesions, diffuse
damage, and neuropsychological functioning in traumatic brain-injured patients.
Journal of Clinical & Experimental Neuropsychology 1 7(6), 900-908.

Anderson, S. W., Damasio, H., Jones, R. D., & Tranel, D. (1991). Wisconsin Card
Sorting Test performance as a measure of frontal lobe damage. Journal of
Clinical & Experimental Neuropsychology 13(6), 909-922.

Angold, A., Costello, E. J ., & Erkanli, A. (1999). Comorbidity. Journal of Child
Psychology & Psychiatry, 40(1), 57-87.

Antonova, E., Sharma, T., Morris, R., & Kumari, V. (2004). The relationship between
brain structure and neurocognition in schizophrenia: A selective review.
Schizophrenia Research, 70(2-3), 117-145.

Arbuthnott, K., & Frank, J. (2000). Trail Making Test, Part B as a measure of executive

control: Validation using a set-switching paradigm. Journal of Clinical &
Experimental Neuropsychology 22(4), 518-528.

187

Aron, A. R., Fletcher, P. C., Bullmore, E. T., Sahakian, B. J ., & Robbins, T. W. (2003).
Stop-signal inhibition disrupted by damage to right inferior frontal gyrus in
humans. Nature Neuroscience, 6, 115-1 16.

Aron, A. R., Robbins, T. W., & Poldrack, R. A. (2004). Inhibition and the right inferior
frontal cortex. Trends in Cognitive Sciences, 8(4), 170-177.

Ashby F.G., Isen A.M., Turken A.U. (1999). A neuropsychological theory of positive
affect and its inﬂuence on cognition. Psychological Review, 106(3), 529-50.

Atchison T.B., Bradshaw M., & Massman PJ. (2004). Investigation of proﬁle
difference between Alzheimer's disease patients declining at different rates:

examination of baseline neuropsychological data. Archives of Clinical
Neuropsychology, 19(8): 1007-101 5.

Austin, M. R, Mitchell, P., Wilhelm, K., Parker, G., Hickie, I., Brodaty, H., Chan, J .,
Eyers, K., Milic, M., & Hadzi-Pavlovic, D. (1999). Cognitive function in
depression: A distinct pattern of frontal impairment in melancholia?
Psychological Medicine, 29(1), 73-85.

Austin, M. P., Ross, M., Murray, C., O'Carroll, R. E. et al. (1992). Cognitive function in
major depression. Journal of A ﬂective Disorders 25(1), 21-29.

Axelrod, EN. (2002). Are normative data from the 64-card version of the WCST
comparable to the full WCST? Clinical Neuropsychologist, 16(1), 7-11.

Aycicegi, A., Dinn, W. M., Harris, C. L., & Erkrnen, H. (2003). Neuropsychological
function in obsessive-compulsive disorder: Effects of comorbid conditions on task
performance. European Psychiatry Vol, 18(5), 241-248.

Baddeley, A. (1986). Working memory. New York, NY: Clarendon Press/Oxford
University Press.

Band, G. P. H., Van Der Molen, M. W., & Logan, G. D. (2003). Horse-race model
simulations of the stop-signal procedure. Acta Psychologica, 112(2), 105-142.

Barkataki, I., Kumari, V., Das, M., Hill, M., Morris, R., O'Connell, P., Taylor, R, &
Sharma, T. (2005). A neuropsychological investigation into violence and mental
illness. Schizophrenia Research, 74(1), 1-13.

Barkley, R. A. (1996). Linkages between attention and executive functions, In: Lyon, G

Reid; Krasnegor, Norman A ; Attention, memory, and executive function ,' 3 07-
325; Baltimore, MD, USublishing Co, 1996, xvii, 424.

188

Basso, M. R., Bomstein, R. A., Carona, F ., & Morton, R. (2001). Depression accounts for
executive function deﬁcits in obsessive-compulsive disorder. Neuropsychiatry,
Neuropsychology, & Behavioral Neurology Vol, 14(4), 241-245.

Baxter, L. R., Schwartz, J. M., Phelps, M. E., Mazziotta, J. C. et al. (1989). Reduction of
prefrontal cortex glucose metabolism common to three types of depression.
Archives of General Psychiatry, 46(3), 243-250.

Beckham, J. C., Crawford, A. L., & F eldman, M. E. (1998). Trail Making Test
performance in Vietnam combat veterans with and without posttraumatic stress
disorder. Journal of Traumatic Stress 1 1(4), 81 1-81 9.

Bellgrove, M.A., Hester, R., & Garavan, H. (2004). The functional neuroanatomical
correlates of response variability: evidence ﬁom a response inhibition task.
Neuropsychologia, 42(14), 1910-1916.

Belzer, K., & Schneier, F. R. (2004). Comorbidity of anxiety and depressive disorders:
Issues in conceptualization, assessment, and treatment. Journal of Psychiatric
Practice Vol, 10(5), 296-306.

Bench, C. J ., Frith, C. D., Grasby, P. M., Friston, K. J. et al. (1993). Investigations of the
ﬁinctional anatomy of attention using the Stroop test. Neuropsychologia, 31(9),
907-922.

Bentler, RM. (1990). Comparative ﬁt indexes in structural models. Psychological
Bulletin, 88, 588-606.

Berlin, L. (2003). Review of the role of inhibitory control and executive functioning in
hyperactivity/ADHD. Nordic Journal of Psychiatry, 5 7(5), 393.

Berman, K. F., Ostrem, J. L., Randolph, C., Gold, J. et al. (1995). Physiological
activation of a cortical network during performance of the Wisconsin Card
Sorting Test: A positron emission tomography study. Neuropsychologia, 33(8),
1 027- 1 046.

Biederman, J ., Faraone, S. V., Keenan, K., Benjamin, J. et al. (1992). Further evidence
for family-genetic risk factors in attention deﬁcit hyperactivity disorder: Patterns
of comorbidity in probands and relatives in psychiatrically and pediatrically
referred samples. Archives of General Psychiatry, 49(9), 728-738.

Biederman, J ., Faraone, S. V., Keenan, K., Knee, D. et al. (1990). Family-genetic and
psychosocial risk factors in DSM-III attention deﬁcit disorder. Journal of the
American Academy of Child & Adolescent Psychiatry, 29(4), 526-533.

Biederman, J ., Faraone, S. V., Spencer, T., Wilens, T. et al. (1993). Patterns of
psychiatric comorbidity, cognition, and psychosocial functioning in adults with

189

attention deﬁcit hyperactivity disorder. American Journal of Psychiatry, 15 0(12),
1792-1 798.

Biederman, J ., Petty, C., Faraone, S. V., Hirshfeld-Becker, D. R., Henin, A., Pollack, M.
H., & Rosenbaum, J. F. (2005). Patterns of Comorbidity in Panic Disorder and
Major Depression: Findings From a Nonreferred Sample. Depression & Anxiety
Vol, 21(2), 55-60.

Biederman, J ., Wilens, T. E., Spencer, T. J ., F araone, S. et al. (1996). Diagnosis and
treatment of adult attention-deﬁcit/hyperactivity disorder, In: Pollack, Mark H ;

Otto, Michael W; Challenges in clinical practice: Pharmacologic and
psychosocial strategies ,° 380-407; New York, NY, US, 1996, xv, 504.

Blair, R. J. R., & Cipolotti, L. (2000). Impaired social response reversal. A case of
'acquired sociopathy'. Brain, 123(6), 1122-1141.

Block, R. I., O'Leary, D. S., Ehrhardt, J. C., Augustinack, J. C., Ghoneim, M. M., Amdt,
S., & Hall, J. A. (2000). Effects of frequent marijuana use on brain tissue volume

and composition. Neuroreport: For Rapid Communication of Neuroscience
Research 11(3), 491-496.

Block, R. I., O'Leary, D. S., Hichwa, R. D., Augustinack, J. C., Ponto, L. L. B., Ghoneim,
M. M., Amdt, S., Ehrhardt, J. C., Hurtig, R. R., Watkins, G. L., Hall, J. A.,
Nathan, P. E., & Andreasen, N. C. (2000). Cerebellar hypoactivity in frequent
marijuana users. Neuroreport: For Rapid Communication of Neuroscience
Research, 11(4), 749-753.

Boldrini, M., Del Pace, L., Placidi, G. P. A., Keilp, J ., Ellis, S. P., Signori, S., Placidi, G.
F., & Cappa, S. F. (2005). Selective cognitive deﬁcits in obsessive-compulsive
disorder compared to panic disorder with agoraphobia. Acta Psychiatrica

Scandinavica, 1 1 1(2), 150-158.

Bolla, K. I., Brown, K., Eldreth, D., Tate, K., & Cadet, J. L. (2002). Dose-related
neurocognitive effects of marijuana use. Neurology, 59(9), 1337-1343.

Bolla, K. 1., Rothman, R., & Cadet, J. L. (1999). Dose-related neurobehavioral effects of
chronic cocaine use. Journal of Neuropsychiatry & Clinical Neurosciences 11(3),
361-369.

Boonstra, A.M., Oosterlaan, J ., Sergeant, J .A., & Buitelaar, J .K. (2005). Executive
functioning in adult ADHD: a meta-analytic review. Psychological Medicine,
35(8), 1097-1108.

Borkowski, J. G., & Burke, J. E. (1996). Theories, models, and measurements of
executive functioning: An information processing perspective. In 1996 (Ed.), In:

190

Lyon, G Reid; Krasnegor, Norman A ,' Attention, memory, and executive function;
235-261; Baltimore, MD, Paul H. Brookes Publishing Co, 1996, xvii, 424.

Brokate, B., Hildebrandt, H., Eling, P., Fichtrrer, H., Runge, K., & T irnm, C. (2003).
Frontal lobe dysfunctions in Korsakoffs syndrome and chronic alcoholism:
Continuity or discontinuity? Neuropsychology 1 7(3), 420-428.

Burgess, P. W. (1997). Theory and methodology in executive function research. Hove,
U.K.: Psychology Press.

Calvo, M. G., & Eysenck, M. W. (1996). Phonological Working Memory and Reading in
Test Anxiety. Memory Vol, 4(3), 289-305.

Castellanos, F .X., Sonuga-Barke, E., Milham, M.P., & Tannock, R. (2006).
Characterizing cognition in ADHD: beyond executive dysﬁmction. Trends in
Cognitive Sciences, 10(3), 117-123.

Casey, B. J ., Castellanos, F. X., Giedd, J. N., & Marsh, W. L. (1997). Implication of right
ﬁ'ontostriatal circuitry in response inhibition and attention-deﬁcit/hyperactivity
disorder. Journal of the American Academy of Child & Adolescent Psychiatry
36(3), 374-383.

Channon, S. (1996). Executive dysﬁrnction in depression: The Wisconsin Card Sorting
Test. Journal of Aﬂective Disorders, 3 9(2), 107-1 14.

Clark, L. A. (1999). Introduction to the special section on the concept of disorder.
Journal of A bnormal Psychology, 108(3), 371 -3 73.

Cleckley, H. (1941). The mask of sanity: an attempt to reinterpret the so-called
psychopathic personality. Oxford, England: Mosby.

Cohen, L. J ., Hollander, E., DeCaria, C. M., Stein, D. J. et al. (1996). Speciﬁcity of
neuropsychological impairment in obsessive-compulsive disorder: A comparison
with social phobic and normal control subjects. Journal of Neuropsychiatry &
Clinical Neurosciences 8(1), 82-85.

Conway, A. R. A., Kane, M. J ., & Engle, R. W. (2003). Working memory capacity and
its relation to general intelligence. Trends in Cognitive Sciences, 7(12), 547-552.

Coolidge, F., Theded, L., & Young, SE. (2000). Heritability and the comorbidity of
attention deﬁcit hyperactivity disorder with behavioral disorders and executive
function deﬁcits: a preliminary investigation. Developmental Neuropsychology,
1 7(3), 273-287.

191

Corbett, B., & Stanczak, D. E. (1999). Neuropsychological performance of adults
evidencing attention-deﬁcit hyperactivity disorder. Archives of Clinical
Neuropsychology, 14(4), 373-387.

Corral, M., Holguin, S. R., & Cadaveira, F. (2003). Neuropsychological characteristics of
young children from high-density alcoholism families: A three-year follow-up.
Journal of Studies on Alcohol, 64(2), 195-199.

Corral, M. M., Holguin, S. R., & Cadaveira, F. (1999). Neuropsychological
characteristics in children of alcoholics: Familial density. Journal of Studies on
Alcohol, 60(4), 509-513.

Cosmides, L., & Tooby, J. (1999). Toward an evolutionary taxonomy of treatable
conditions. Journal of Abnormal Psychology, 108(3), 453-464.

Cox, B.J., Clara, I.P., & Enns, M.W. (2002). Posttraumatic stress disorder and the
structure of common mental disorders. Depression and Anxiety, 15(4), 168-171.

Crowe, S. F. (1998). The differential contribution of mental tracking, cognitive
ﬂexibility, visual search, and motor speed to performance on Parts A and B of the
Trail Making Test. Journal of Clinical Psychology, 54(5), 585-591.

Crowell, T. A., Kieffer, K. M., Kugeares, S., & Vanderploeg, R. D. (2003). Executive
and nonexecutive neuropsychological functioning in antisocial personality
disorder. Cognitive & Behavioral Neurology, 16(2), 100-109.

Dao-Castellana, M. H., Samson, Y., Legault, F., Martinot, J. L., Aubin, H. J ., Crouzel, C.,
Feldman, L., Barrucand, D., Rancurel, G., Feline, A., & Syrota, A. (1998). Frontal
dysfunction in neurologically normal chronic alcoholic subjects: Metabolic and
neuropsychological ﬁndings. Psychological Medicine, 28(5), 1039-1048.

Delis, D. C., Kaplan, E., & Kramer, J. H. (2001). Delis-Kaplan Executive Function
System Manual. San Antonio, Texas: Harcourt Assessment, Inc.

Demakis, G. J. (2003). A meta-analytic review of the sensitivity of the Wisconsin Card
Sorting Test to frontal and lateralized frontal brain damage. Neuropsychology,
1 7(2), 255-264.

Demakis, G. J. (2004). Frontal Lobe Damage and Tests of Executive Processing: A Meta-
Analysis of the Category Test, Stroop Test, and Trail-Making Test. Journal of
Clinical & Experimental Neuropsychology, 26(3), 441-450.

Denckla, M. B. (1996). A theory and model of executive function: A neuropsychological
perspective, In: Lyon, G Reid; Krasnegor, Norman A ; Attention, memory, and
executive function ; 263-2 78; Baltimore, MD, Paul H. Brookes Publishing Co,
1996, xvii, 424.

192

Déry, M., Toupin, J ., Pauzé, R., Mercier, H., & Fortin, L. (1999). Neuropsychological
characteristics of adolescents with conduct disorder: Association with attention-
deﬁcit-hyperactivity and aggression. Journal of A bnormal Child Psychology,
27(3), 225-236.

Devinsky, O., Morrell, M. J ., & Vogt, B. A. (1995). Contributions of anterior cingulate
cortex to behaviour. Brain 118(1), 279-306.

Diamond, A. (2000). Close interrelation of motor development and cognitive
development and of the cerebellum and prefrontal cortex. Child Development
71(1), 44-56.

Dibartolo, P.M., Brown, T.A., & Barlow, DH. (1997). Effects of anxiety on attention
allocation and task performance: an information processing analysis. Behaviour
Research and Therapy, 35(12), 1101-1111.

Dimitrov, M., Nakic, M., Elpem-Waxman, J ., Granetz, J ., O'Grady, J ., Phipps, M., Milne,
E., Logan, G. D., Hasher, L., & Graﬁnan, J. (2003). Inhibitory attentional control
in patients with frontal lobe damage. Brain & Cognition 52(2), 258—270.

Dinn, W. M., & Harris, C. L. (2000). Neurocognitive function in antisocial personality
disorder. Psychiatry Research 97(2-3) (Dec 2000), 173-190.

Dixon, R.A., Garret, D.D., Lentz, T.L., MacDonald, S., Strauss, E., & Hultsch, DP.
(2007). Neurocognitive markers of cognitive impairment: exploring the roles of
speed and inconsistency. Neuropsychology, 21(3), 381—399.

Dolan, M., & Park, I. (2002). The neuropsychology of antisocial personality disorder.
Psychological Medicine, 32(3), 417-427.

Dougherty, D., & Rauch, S. L. (1997). Neuroimaging and neurobiological models of
depression. Harvard Review of Psychiatry, 5(3), 138-159.

Downey, K. K., Stelson, F. W., Pomerleau, O. F., & Giordani, B. (1997). Adult attention

deﬁcit hyperactivity disorder: Psychological test proﬁles in a clinical population.
Journal of Nervous & Mental Disease 185(1), 32-38.

Drevets, W. C. (2000). Neuroimaging studies of mood disorders. Biological Psychiatry,
48(8), 813-829.

Drevets, W. C., Price, J. L., Simpson, J. R., Jr., Todd, R. D. et al. (1997). Subgenual
preﬁ'ontal cortex abnormalities in mood disorders. Nature 386(6627), 824-827.

193

Drevets, W. C., Videen, T. 0., Price, J. L., Preskom, S. H. et al. (1992). A functional
anatomical study of unipolar depression. Journal of Neuroscience 12(9), 3628-
3641.

Duncan, J ., Johnson, R., Swales, M., & Freer, C. (1997). Frontal lobe deﬁcits after head
injury: Unity and diversity of ﬁmction. Cognitive Neuropsychology 14(5), 713-
741.

Duncan, J., & Owen, A. M. (2000). Common regions of the human frontal lobe recruited
by diverse cognitive demands. Trends in Neurosciences, 23(10), 475-483.

Ekselius, L., Tillfors, M., Furmark, T., & Fredrikson, M. (2001). Personality disorders in
the general population: DSM-IV and ICD-lO deﬁned prevalence related to
sociodemographic proﬁle. Personality & Individual Diﬂerences, 30(2), 31 1-320.

Eldreth, D. A., Matochik, J. A., Cadet, J. L., & L., B. K. (2004). Abnormal brain activity
in prefrontal brain regions in abstinent marijuana users. . NeuroImage, 23, 914-
920.

Elliott, R., Baker, S. C., Rogers, R. D., O'Leary, D. A. et al. (1997). Preﬁ'ontal
dysfunction in depressed patients performing a complex planning task: A study
using positron emission tomography. Psychological Medicine 27(4), 931-942.

Elliott, R., Dolan, R. J ., & Frith, C. D. (2000). Dissociable functions in the medial and
lateral orbitoﬁ'ontal cortex: Evidence from human neuroimaging studies. Cerebral
Cortex Special Issue: The mysterious orbitoﬁontal cortex, 10(3), 308-317.

Eslinger, P.J. & Damasio, AR. (1985). Severe disturbance of higher cognition after
bilateral frontal lobe abloation: patient EVR. Neurology, 35(12), 1731-1741.

Ettenhofer, ML. (2006). Mild traumatic brain injury, executive functions, and post-
concussive symptoms. Unpublished dissertation, Michigan State University.

Ettlin, T. M., Kischka, U., Beckson, M., Gaggiotti, M., Rauchﬂeisch, U., & Benson, D. F.
(2000). The Frontal Lobe Score: Part 1: Construction of a mental status of frontal
systems. Clinical Rehabilitation, 14(3), 260-271.

Eysenck, M. W., & Calvo, M. G. (1992). Anxiety and performance: The processing
efﬁciency theory. Cognition & Emotion Vol, 6(6), 409-434.

F alck, R. S., Wang, J ., Siegal, H. A., & Carlson, R. G. (2004). The prevalence of
psychiatric disorder among a community sample of crack cocaine users: An

exploratory study with practical implications. Journal of Nervous & Mental
Disease Vol, 192(7), 503-507.

194

Pals-Stewart, W., & Bates, M. E. (2003). The neuropsychological test performance of
drug-abusing patients: An examination of latent cognitive abilities and associated
risk factors. Experimental & Clinical Psychopharmacology, 11(1), 34-45.

F an], F ., Erdfelder, E., Lang, A.-G., & Buchner, A. (in press). G*Power 3: A ﬂexible
statistical power analysis program for the social, behavioral, and biomedical
sciences. Behavior Research Methods.

F eighner, J. P., Robins, E., Guze, S.B., Woodruff, R.A., Winokur, G., & Munoz, R.
(1972). Diagnostic criteria for use in psychiatric research. Archives of General
Psychiatry, 26, 57-63.

First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1997). Structured Clinical
Interview for DSM-1 VAxis I Disorders: Clinician Version: Administration
Booklet Washington, DC: American Psychiatric Press.

Flint, A. J ., Black, S. E., Campbell-Taylor, I., Gailey, G. F. et al. (1993). Abnormal
speech articulation, psychomotor retardation, and subcortical dysfunction in major
depression. Journal of Psychiatric Research Vol, 27(3), 309-319.

Fossati, P., Amar, G., Raoux, N., Ergis, A. M., & Allilaire, J. F. (1999). Executive
functioning and verbal memory in young patients with unipolar depression and
schizophrenia. Psychiatry Research, 89(3), 171-187.

Franklin, T. R., Acton, P. D., Maldjian, J. A., Gray, J. D., Croft, J. R., Dackis, C. A.,
O'Brien, C. P., & Childress, A. R. (2002). Decreased gray matter concentration in
the insular, orbitofrontal, cingulate, and temporal cortices of cocaine patients.
Biological Psychiatry 51(2), 134-142.

Frazier, T. W., Demaree, H. A., & Youngstrom, E. A. (2004). Meta-Analysis of
Intellectual and Neuropsychological Test Performance in Attention-
Deﬁcit/Hyperactivity Disorder. Neuropsychology 18(3), 543-555.

Friedman, N.P. & Miyake, A. (2004). The relations among inhibition and interference
control functions: a latent-variable analysis. Journal of Experimental Psychology,
133(1), 101-135.

Friedman, N.P., Miyake, A., Corley, R.P., Young, S.E., DeFries, J ., & Hewitt, J .K.
(2006). Not all executive functions are related to intelligence. Psychological
Science, 1 7(2), 172-179.

Frost, L. A., Mofﬁtt, T. E., & McGee, R. (1989). Neuropsychological correlates of

psychopathology in an unselected cohort of young adolescents. Journal of
Abnormal Psychology 98(3), 307-313.

195

Fuster, J. M. (1997). The prefrontal cortex: Anatomy, physiology,k and neuropsychology
of the ﬁontal lobe (Third Edition ed.). Philadelphia, NY: Lippincott-Raven
Publishers.

Gansler, D. A., Harris, G. J ., Oscar-Berman, M., Streeter, C., Lewis, R. F ., Ahmed, 1., &
Achong, D. (2000). Hypoperfusion of inferior frontal brain regions in abstinent
alcoholics: A pilot SPECT study. Journal of Studies on Alcohol 61 (1 ), 32-3 7.

Garber, J ., & Hollon, S. D. (1991). What can speciﬁcity designs say about causality in
psychopathology research? Psychological Bulletin 110(1), 129-136.

George, M. S., Ketter, T. A., Parekh, P. I., Rosinsky, N. et al. (1997a). Blunted left
cingulate activation in mood disorder subjects during a response interference task
(the Stroop). Journal of Neuropsychiatry & Clinical Neurosciences 9(1), 55-63.

George, M. S., Ketter, T. A., Parekh, P. I., Rosinsky, N. et al. (1997b). Blunted left
cingulate activation in mood disorder subjects during a response interference task

(the Stroop). Journal of Neuropsychiatry & Clinical Neurosciences Vol, 9(1), 55-
63.

Giancola, P. R., & Moss, H. B. (1998). Executive cognitive functioning in alcohol use
disorders, In: Galanter, Marc; Recent developments in alcoholism, Vol 14: The
consequences of alcoholism: Medical neuropsychiatric economic cross-cultural ;
22 7-251; New York, NY, US, 1998, xxii, 499 Series: Recent developments in
alcoholism.

Giancola, P. R., Shoal, G. D., & Mezzich, A. C. (2001). Constructive thinking, executive
functioning, antisocial behavior, and drug use involvement in adolescent females

with a substance use disorder. Experimental & Clinical Psychopharmacology,
9(2), 215-227.

Giancola, P. R., & Tarter, R. E. (1999). Executive cognitive functioning and risk for
substance abuse. Psychological Science Vol, 10(3), 203-205.

Gladsjo, J. A., Rapaport, M. H., McKinney, R., Lucas, J. A., Rabin, A., Oliver, T., Davis,
J ., Auerbach, M., & Judd, L. L. (1998). A neuropsychological study of panic
disorder: Negative ﬁndings. Journal of Aﬂective Disorders 49(2), 123-131.

Golden, C. J. (1978). Stroop Color and Word Test. Wood Dale, IL: Stoelting Co.

Goldstein, B., Obrzut, J. B, John, C., Ledakis, G., & Armstrong, C. L. (2004a). The
impact of frontal and non-frontal brain tumor lesions on Wisconsin Card Sorting

Test performance. Brain & Cognition 54(2), 110-116.

Goldstein, R. Z., Leskovjan, A. C., Hoff, A. L., Hitzemann, R., Bashan, F., Khalsa, S. S.,
Wang, G.-J., Fowler, J. S., & Volkow, N. D. (2004b). Severity of

196

neuropsychological impairment in cocaine and alcohol addiction: Association
with metabolism in the prefrontal cortex. Neuropsychologia 42(11), 1447-1458.

Goldstein, R. Z., & Volkow, N. D. (2002). Drug addiction and its underlying
neurobiological basis: Neuroimaging evidence for the involvement of the ﬁontal
cortex. American Journal of Psychiatry 159(10), 1642-1652.

Goodwin, R. D., & Hamilton, S. P. (2003). Lifetime comorbidity of antisocial personality
disorder and anxiety disorders among adults in the community. Psychiatry
Research Vol, 11 7(2), 159-166.

Gorenstein, E. E. (1982). Frontal lobe functions in psychopaths. Journal of Abnormal
Psychology 91(5), 368-3 79.

Gorenstein, E. E., & Newman, J. P. (1980). Disinhibitory psychopathology: A new
perspective and a model for research. Psychological Review 87(3), 301-315.

Gray, J. A. (1982). The neuropsychology of anxiety: An enquiry into the functions of the
septo-hippocampal system. New York, NY: Clarendon Press/Oxford University
Press.

Gray, J. A., & McNaughton, N. (1995). The neuropsychology of anxiety: Reprise, In:
Hope, Debra A ,° Nebraska Symposium on Motivation, 1995: Perspectives on
anxiety, panic, and fear ,' 61-134; Lincoln, NE, USebraska Press, 1996, xiii, 351
Series: Current theory and research in motivation, Vol 43.

Gruber, S. A., & Yurgelun-Todd, D. A. (2005). Neuroimaging of marijuana smokers
during inhibitory processing: A pilot investigation. Cognitive Brain Research
Special Issue: Multiple Perspectives on Decision Making Vol, 23(1), 107-118.

Hanes, K. R., Andrewes, D. G., Smith, D. J ., & Pantelis, C. (1996). A brief assessment of
executive control dysfunction: Discriminant validity and homogeneity of
planning, set shift, and ﬂuency measures. Archives of Clinical Neuropsychology,

11(3), 185-191.

Hare, R. D. (1996). Psychopathy: A clinical construct whose time has come. Criminal
Justice & Behavior 23(1), 25-54.

Harper, C. G., & Kril, J. (1985). Brain atrophy in chronic alcoholic patients: A
quantitative pathological study. Journal of Neurology, Neurosurgery &
Psychiatry 48(3), 21 1-217.

Hart, S. D., Forth, A. E., & Hare, R. D. (1990). Performance of criminal psychopaths on

selected neuropsychological tests. Journal of Abnormal Psychology 99(4), 374-
379.

197

Hayes, S. C., Gifford, E. V., & Ruckstuhl, L. E., Jr. (1996). Relational frame theory and
executive function: A behavioral approach, In: Lyon, G Reid; Krasnegor, Norman
A ; Attention, memory, and executive function ; 279-305; Baltimore, MD, Paul H.
Brookes Publishing Co, 1996, xvii, 424.

Heaton, R. K., Chelune, G. J ., Talley, J. L., Kay, G. G., & Curtiss, G. (1993). Wisconsin
Card Sorting Test Manual. USA: Publishing Assessment Resources.

Henry, J. D., & Crawford, J. R. (2005). A meta-analytic review of verbal ﬂuency deﬁcits
in schizophrenia relative to other neurocognitive deﬁcits. Cognitive
Neuropsychiatry 1 0(1), 1-3 3.

Hervey, A. S., Epstein, J. N., & Curry, J. F. (2004). Neuropsychology of Adults With
Attention-Deﬁcit/Hyperactivity Disorder: A Meta-Analytic Review.
Neuropsychology 18(3), 485-503.

Hester, R., & Garavan, H. (2004). Executive Dysfunction in Cocaine Addiction:
Evidence for Discordant Frontal, Cingulate, and Cerebellar Activity. Journal of
Neuroscience 24(49), 1 1017-1 1022.

Hoffman, J. J ., Hall, R. W., & Bartsch, T. W. (1987). On the relative importance of
"psychopathic" personality and alcoholism on neuropsychological measures of
ﬁ'ontal lobe dysfunction. Journal of Abnormal Psychology 96(2), 158-160.

Holdnack, J. A., Moberg, P. J ., Arnold, S. B, Our, R.C. et al. (1995). Speed of processing
and verbal learning deﬁcits in adults diagnosed with attention deﬁcit disorder.
Neuropsychiatry, Neuropsychology, & Behavioral Neurology 8(4), 282-292.

Hopko, D. R., Ashcraft, M. H., Gute, J ., Ruggiero, K. J ., & Lewis, C. (1998).
Mathematics anxiety and working memory: Support for the existence of a
deﬁcient inhibition mechanism. Journal of Anxiety Disorders Vol, 12(4), 343-355.

Hunt, C., Slade, T., & Andrews, G. (2004). Generalized Anxiety Disorder and Major
Depressive Disorder Comorbidity in the National Survey of Mental Health and
Well-Being. Depression & Anxiety Vol, 20(1), 23-31.

Johnson, D. E., Epstein, J. N., Waid, L. R., Latham, P. K., Voronin, K. E., & Anton, R. F.
(2001). Neuropsychological performance deﬁcits in adults with attention-
deﬁcit/hyperactivity disorder. Archives of Clinical Neuropsychology 16(6), 587-
604.

J oreskog, K.G.,, & Forbom, D. (1981). Analysis of linear structural relationships by

maximum likelihood and least squares methods (Research Report No 81-8).
Uppsala, Sweden: University of Uppsala.

198

Jovanovski, D., Erb, S., & Zakzanis, K. K. (2005). Neurocognitive Deﬁcits in Cocaine
Users: A Quantitative Review of the Evidence. Journal of Clinical &
Experimental Neuropsychology 27(2), 189-204.

Kuhne, M., Schachar, R., & Tannock, R. (1997). Impact of comorbid oppositional or
conduct problems on attention-deﬁcit hyperactivity disorder. Journal of the
American Academy of Child & Adolescent Psychiatry, 36(12), 1715-1725.

Kaiser, S., Unger, J ., Kiefer, M., Markela, J ., Mundt, C., & Weisbrod, M. (2003).
Executive control deﬁcit in depression: Event-related potentials in a Go/Nogo
task. Psychiatry Research: Neuroimaging 122(3), 169-184.

Kane, M. J ., & Engle, R. W. (2002). The role of prefrontal cortex in working-memory
capacity, executive attention, and general ﬂuid intelligence: An individual-
differences perspective. Psychonomic Bulletin & Review 9(4), 63 7-671.

Katz, L. J ., Wood, D. S., Goldstein, G., Auchenbach, R. C., & Geckle, M. (1998). The
utility of neuropsychological tests in evaluation of attention-deﬁcit/hyperactivity
disorder (ADHD) versus depression in adults. Assessment Vol, 5(1), 45-51.

Keller, J ., Nitschke, J. B., Bhargava, T., Deldin, P. J ., Gergen, J. A., Miller, G. A., &
Heller, W. (2000). Neuropsychological differentiation of depression and anxiety.
Journal of Abnormal Psychology 109(1), 3-10.

Kems, J. 6., Cohen, J. D., MacDonald, A. W., 111, Cho, R. Y., Stenger, V. A., & Carter,
C. S. (2004). Anterior Cingulate Conﬂict Monitoring and Adjustments in Control.
Science 3 03 (5660), 1023-1026.

Kerr, M., Tremblay, R.E., Pagani, L., & Vitaro, F. (1997). Boys’ behavioral inhibition
and the risk of later delinquency. Archives of General Psychiatry, 54(9), 809-816.

Kessing, L. V. (1998). Cognitive impairment in the euthymic phase of affective disorder.
Psychological Medicine 28(5), 1027-1038.

Kessler, R. C., Berglund, P., Demler, O., Jin, R., & Walters, E. E. (2005a). Lifetime
prevalence and age-of-onset distributions of DSM-IV disorders in the National
Comorbidity Survey Replication. Archives of General Psychiatry 62(6), 593-602.

Kessler, R. C., Chiu, W. T., Demler, O., & Walters, E. E. (2005b). Prevalence, severity,
and comorbidity of 12-Month DSM-IV Disorders in the National Comorbidity
Survey Replication. Archives of General Psychiatry 62(6), 617-627.

Kessler, R. C., Crum, R. M., Warner, L. A., Nelson, C. B. et al. (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.

199

Kessler, R. C., McGonagle, K. A., Zhao, S., Nelson, C. B., Hughes, M., Eshleman, S.,
Wittchen, H., & Kendler, K. (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.

Kimbrell, T. A., Ketter, T. A., George, M. S., Little, J. T., Benson, B. E., Willis, M. W.,
Herscovitch, P., & Post, R. M. (2002). Regional cerebral glucose utilization in

patients with a range of severities of unipolar depression. Biological Psychiatry
Vol, 51(3), 237-252.

Kline. (2004). Principles and Practice of Structural Equation Modeling, Second Edition.
New York, NY: The Guildford Press.

Knopik, V. S., Heath, A. C., Madden, P. A. F., Bucholz, K. K., Slutske, W. S., Nelson, E.
C., Statham, D., Whitﬁeld, J. B., & Martin, N. G. (2004). Genetic effects on
alcohol dependence risk: Re-evaluating the importance of psychiatric and other
heritable risk factors. Psychological Medicine Vol, 34(8), 1519-1530.

Koenen, K. C., Driver, K. L., Oscar-Berman, M., Wolfe, J ., Folsom, S., Huang, M. T., &
Schlesinger, L. (2001). Measures of preﬁ'ontal system dysfunction in

posttraumatic stress disorder. Brain & Cognition 45(1), 64-78.

Konishi, S., Nakajima, K., Uchida, I., Hideyuki, K., Kameyama, M., & Miyashita, Y.
(1999). Common inhibitory michanism in human inferior preﬁonal cortex
revealed by event-related functional MRI. Brain, 122, 981-991.

Kovacs, M. (1996). The course of childhood-onset depressive disorders. Psychiatric
Annals 26(6), 326-330.

Krueger, R. F. (1999). The structure of common mental disorders. Archives of General
Psychiatry, 56(10), 921-926.

Krueger, R. F. (2005). Continuity of Axes I and II: Toward a Uniﬁed Model of
Personality, Personality Disorders, and Clinical Disorders. Journal of Personality
Disorders 19(3), 233-261.

Kuruoglu, A. C., Arikan, Z., Vural, G., Karatas, M. et al. (1996). Single photon emission
computerised tomography in chronic alcoholism: Antisocial personality disorder

may be associated with decreased frontal perfusion. British Journal of Psychiatry
169(3), 348-3 54.

Laakso, M. P., Gunning-Dixon, F., Vaurio, O., Repo-Tiihonen, E., Soininen, H., &
Tiihonen, J. (2002). Prefrontal volumes in habitually violent subjects with
antisocial personality disorder and type 2 alcoholism. Psychiatry Research:
Neuroimaging 114(2), 95-102.

200

Lapierre, D., Braun, C. M. J ., & Hodgins, S. (1995). Ventral frontal deﬁcits in
psychopathy: Neuropsychological test ﬁndings. Neuropsychologia 33(2), 139-
151.

Larrue, V., Celsis, P., Bes, A., Marc-Vergnes, J .P. (1994). The functional anatomy of
attention in humans: cerebral blood ﬂow changes induced by reading, naming,
and the Stroop effect. Journal of Cerebral Blood Flow and Metabolism, 14, 958-
962.

Lezak, M., Howieson, D. B., & Loring, D. W. (2004). Neuropsychological Assessment,
Fourth Edition. Oxford: Oxford University Press.

Lichter, D. G. E., & Cummings, J. L. (Eds.). (2001). Frontal-subcortical circuits in
psychiatric and neurological disorders. New York, NY: Guilford Press.

Lijfﬁjt, M., Kenemans, J .L., Verbaten, M.N., van Engeland, H. (2005). A meta-analytic
review of stopping performance in attention-deﬁcit/hyperactivity disorder:
deﬁcient inhibitory motor control? Journal of Abnormal Psychology, 114(2),
216-222.

Lilienfeld, S. O., & Marino, L. (1995). Mental disorder as a Roschian concept: A critique
of Wakeﬁeld's "harmful dysfunction" analysis. Journal of Abnormal Psychology
104(3), 411-420.

Lilienfeld, S. O., & Marino, L. (1999). Essentialism revisited: Evolutionary theory and
the concept of mental disorder. Journal of Abnormal Psychology 108(3), 400-411.

Lilienfeld, S. O., Waldman, I. D., & Israel, A. C. (1994). A critical examination of the
use of the term and concept of comorbidity in psychopathology research. Clinical
Psychology: Science & Practice 1(1), 71-83.

Logan, G. D. (1994). On the ability to inhibit thought and action: A users' guide to the
stop signal paradigm, In: Dagenbach, Dale; Carr, Thomas H ; Inhibitory
processes in attention, memory, and language ,' 189-239; San Diego, CA, USnc,
1994, xiv, 461.

Logan, G. D., Schachar, R. J ., & Tannock, R. (1997). Irnpulsivity and inhibitory control.
Psychological Science 8(1), 60-64.

Lombardi, W. J ., Andreason, P. J ., Sirocco, K. Y., Rio, D. E., Gross, R. E., Umhau, J. C.,
& Hommer, D. W. (1999). Wisconsin Card Sorting Test performance following
head injury: Dorsolateral fronto-striatal circuit activity predicts perseveration.
Journal of Clinical &. Experimental Neuropsychology 21(1), 2-16.

201

Love, J .M., Greve, K.W., Sherwin, E., & Mathias, C. (2003). Comparability of the
standard WCST and WCST-64 in traumatic brain injury. Applied
Neuropsychology, 10(4), 246-251.

Lovejoy, D. W., Ball, J. D., Keats, M., Stutts, M., Spain, E. H., Janda, L., & Janusz, J.
(1999). Neuropsychological performance of adults with attention deﬁcit
hyperactivity disorder (ADHD): Diagnostic classiﬁcation estimates for measures
of frontal lobe/executive functioning. Journal of the International
Neuropsychological Society 5(3), 222-233.

Lubman, D. I., Yiicel, M., & Pantelis, C. (2004). Addiction, 3 condition of compulsive
behaviour? Neuroimaging and neuropsychological evidence of inhibitory
dysregulation. Addiction 99(12), 1491-1502.

Lundqvist, T. (2005). Cognitive consequences of cannabis use: Comparison with abuse of
stimulants and heroin with regard to attention, memory and executive functions.
Pharmacology, Biochemistry & Behavior Special Issue: Cannabinoids:
Interactions with other drugs 81(2), 319-330.

MacDonald, A. W., Cohen, J. D., Stenger, V. A., & Carter, C. S. (2000). Dissociating the
role of the dorsolateral prefrontal and anterior cingulate cortex in cognitive
control. Science, 288, 1835-1838.

Malizia, A. L. (1999). What do brain imaging studies tell us about anxiety disorders?
Journal of Psychopharmacology 13(4), 372-378.

Malloy, P., Bihrle, A., Duffy, J ., & Cimino, C. (1993). The orbitomedial frontal
syndrome. Archives of Clinical Neuropsychology 8(3), 185-201.

Malloy, P., Noel, N., Longabaugh, R., & Beattie, M. (1990). Determinants of
neuropsychological impairment in antisocial substance abusers. Addictive
Behaviors 15(5), 431-438.

Malloy, P., Noel, N., Rogers, S., Longabaugh, R. et a1. (1989). Risk factors for
neuropsychological impairment in alcoholics: Antisocial personality, age, years of
drinking and gender. Journal of Studies on Alcohol 50(5), 422-426.

Martin, N. J ., & Franzen, M. D. (1989). The effect of anxiety on neuropsychological
function. International Journal of Clinical Neuropsychology 11(1), 1-8.

Martinot, J .-L., Hardy, P., Feline, A., Huret, J .-D. et al. (1990). Left preﬁontal glucose
hypometabolism in the depressed state: A conﬁrmation. American Journal of
Psychiatry 147(10), 1313-1317.

McKay, K. E., & Halperin, J. M. (2001). ADHD, aggression, and antisocial behavior

across the lifespan: Interactions with neurochemical and cognitive function, In:
Wasserstein, Jeanette; Wolff Lorraine E ,' Adult attention deﬁcit disorder: Brain

202

mechanisms and life outcomes ; 84-96; New York, NY, US of Sciences, 2001, x,
409 Series: Annals of the New York Academy Sciences, vol 931 .

Merriam, E. P., Thase, M. E., Haas, G. L., Keshavan, M. S., & Sweeney, J. A. (1999).
Prefrontal cortical dysﬁmction in depression determined by Wisconsin Card
Sorting Test performance. American Journal of Psychiatry 156(5), 780-782.

Mesulam, M. M. (2002). The human frontal lobes: Transcending the default mode
through contingent encoding, In: Stuss, Donald T ,' Knight, Robert T ,° Principles
of ﬁontal lobe function ,° 8-3 0; London University Press, 2002, xxi, 616.

Middleton, F. A., & Strick, P. L. (2001). Cerebellar projections to the prefrontal cortex in
the primate. Journal of Neuroscience 21 (2), 700-712.

Middleton, F. A., & Strick, P. L. (2002). Basal-ganglia 'projections' to the prefrontal
cortex of the primate. Cerebral Cortex 12(9), 926-935.

Miller, N. S., Klamen, D., Hoffmann, N. G., & Flaherty, J. A. (1996). Prevalence of
depression and alcohol and other drug dependence in addictions treatment
populations. Journal of Psychoactive Drugs 28(2), 11 1-124.

Milner, B. (1963). Effects of different brain lesions on card sorting. Archives of
Neurology, 9, 90-100.

Miyake, A., Friedman, N. P., Emerson, M. J ., Witzki, A. H., & Howerter, A. (2000). The
unity and diversity of executive functions and their contributions to complex
"frontal lobe" tasks: A latent variable analysis. Cognitive Psychology 41(1), 49-
100.

Monarch, E. S., Saykin, A. J ., & Flashman, L. A. (2004). Neuropsychological impairment
in borderline personality disorder. Psychiatric Clinics of North America 27(1), 67-
82.

Morgan, A. B., & Lilienfeld, S. O. (2000). A meta-analytic review of the relation
between antisocial behavior and neuropsychological measures of executive
ﬁmction. Clinical Psychology Review 20(1), 1 13-156.

Moritz, S., Birkner, C., Kloss, M., Jahn, H., Hand, 1., Haasen, C., & Krausz, M. (2002).
Executive functioning in obsessive-compulsive disorder, unipolar depression, and
schizophrenia. Archives of Clinical Neuropsychology 1 7(5), 477-483.

Moscovitch, M., Winocur, G. (1995). Frontal lobes, memory, and aging. Annals of the
New York Academy of Sciences, 769, 119-150.

Munro, C. A., Saxton, J ., & Butters, M. A. (2000). The neuropsychological consequences

of abstinence among older alcoholics: A cross-sectional study. Alcoholism:
Clinical & Experimental Research Vol, 24(10), 1 5 10-1 516.

203

Murphy, K. R., Barkley, R. A., & Bush, T. (2001). Executive ﬁmctioning and olfactory
identiﬁcation in young adults with attention deﬁcit-hyperactivity disorder.
Neuropsychology 15(2), 211-220.

Nigg, J. T., Butler, K. M., Huang-Pollock, C. L., & Henderson, J. M. (2002). Inhibitory
processes in adults with persistent childhood onset ADHD. Journal of Consulting
& Clinical Psychology 70(1), 153-157.

Nigg, J. T., Glass, J. M., Wong, M. M., Poon, E., Jester, J. M., Fitzgerald, H. E., Puttler,
L. 1., Adams, K. M., & Zucker, R. A. (2004). Neuropsychological Executive

Functioning in Children at Elevated Risk for Alcoholism: Findings in Early
Adolescence. Journal of Abnormal Psychology 113(2), 302-3 14.

Nigg, J. T., Stavro, G. M., Ettenhofer, M. L., Hambrick, D. Z., Miller, T., & Henderson,
J. M. (2005). Executive functions and ADHD in adults: evidence for selective
effects on ADHD symptom domains. Journal of Abnormal Psychology, 114(4),

706-717.

Nigg, J. T., Wong, M., Martel, M., Jester, J ., Puttler, L. 1., Glass, J ., Adams, K.,
Fitzgerald, H. E., & Zucker, R. A. (2006). Poor response inhibition as a predictor
of problem drinking and illicit drug use in adolescents at risk for alcoholism and
other substance use disorders. Journal of the American Academy of Child and

Adolescent Psychiatry, 45, 468-475.

Niles, B. L., Mori, D. L., Lambert, J. F., & Wolf, E. J. (2005). Depression in Primary
Care: Comorbid Disorders and Related Problems. Journal of Clinical Psychology
in Medical Settings Vol, 12(1), 71-77.

Norman, D., & Shallice, T. (1986). Attention to action: Willed and automatic control of
behavior. New York, NY: Plenum.

Pandya, D. N., & Barnes, C. L. (1987). Architecture and connections of the frontal lobe,
In: Perecman, Ellen; The ﬁontal lobes revisited ; 41 -72; New York, NY, US,
198 7, xv, 309.

Paradiso, S., Larnberty, G. J ., Garvey, M. J ., & Robinson, R. G. (1997). Cognitive
impairment in the euthymic phase of chronic unipolar depression. Journal of
Nervous & Mental Disease 185(12), 748-754.

Parkin, A.J. (1998). The central executive does not exist. Journal of the International
Neuropsychological Society, 4, 518-522.

Penadés, R., Catalan, R., Andrés, S., Salamero, M., & Gasto, C. (2005). Executive

function and nonverbal memory in obsessive-compulsive disorder. Psychiatry
Research Vol, 133(1), 81-90.

204

Pennington, B. F., Bennetto, L., McAleer, O., & Roberts, R. J ., Jr. (1996). Executive
functions and working memory: Theoretical and measurement issues, In: Lyon, G

Reid; Krasnegor, Norman A ; Attention, memory, and executive function ; 32 7-
348; Baltimore, MD, PAul H. Brookes Publishing Co, 1996, xvii, 424.

Pennington, B. F., & Ozonoff, S. (1996). Executive functions and developmental
psychopathology. Journal of Child Psychology & Psychiatry Special Issue:
Annual research review, 3 7(1), 51-87.

Petrides, M., & Pandya, D. N. (2002). Association pathways of the prefrontal cortex and
functional observations, In: Stuss, Donald T ,° Knight, Robert T ; Principles of
ﬁ'ontal lobe function ; 31-50; London University Press, 2002, xxi, 616.

Petry, N. M., Stinson, F. S., & Grant, B. F. (2005). Comorbidity of DSM-IV Pathological
Gambling and Other Psychiatric Disorders: Results From the National

Epidemiologic Survey on Alcohol and Related Conditions. Journal of Clinical
Psychiatry Vol, 66(5), 564-574.

Pham, T. H., Vanderstukken, O., Philippot, P., & Vanderlinden, M. (2003). Selective
Attention and Executive Functions Deﬁcits Among Criminal Psychopaths.
Aggressive Behavior Vol, 29(5), 393-405.

Pihl, R. 0., Peterson, J ., & Finn, P. R. (1990). Inherited predisposition to alcoholism:
Characteristics of sons of male alcoholics. Journal of Abnormal Psychology
99(3), 291-301.

Poon, E., Puttler, L. 1., Zucker, R. A., Nigg, J. T., & Fitzgerald, H. E. (1999). Prefrontal
cortex neuropsychological functioning in a community sample of alcoholic men,

Annual Research Society on Alcoholism Meeting. Santa Barbara, CA.

Pope, H. 6., Jr., & Yurgelun-Todd, D. (1996). The residual cognitive effects of heavy
marijuana use in college students. JAMA: Journal of the American Medical
Association 275(7), 521-527.

Puig-Antich, J ., & Ryan, N. (1986). Schedule for Aﬂective Disorders and Schizophrenia
for School-Age Children. . Pittsburgh, PA: Western Psychiatric Institute and
Clinic, University of Pittsburgh School of Medicine.

Purcell, R., Maruff, P., Kyrios, M., & Pantelis, C. (1998). Neuropsychological deﬁcits in
obsessive-compulsive disorder: A comparison with unipolar depression, panic

disorder, and normal controls. Archives of General Psychiatry, 55(5), 415-423.

Quraishi, S., & Frangou, S. (2002). Neuropsychology of bipolar disorder: A review.
Journal of Aﬂective Disorders 72(3), 209-226.

205

Rabbitt, P. (1997). Introduction: Methodologies and Models in the Study of Executive
Function. East Sussex, UK: Psychology Press.

Raine, A., Lencz, T., Bihrle, S., LaCasse, L., & Colletti, P. (2000). Reduced prefrontal
gray matter volume and reduced autonomic activity in antisocial personality
disorder. Archives of General Psychiatry 57(2), 119-127.

Ratti, M. T., Bo, P., Giardini, A., & Soragna, D. (2002). Chronic alcoholism and the
frontal lobe: Which executive functions are impaired? Acta Neurologica
Scandinavica 105(4), 276-281.

Ravnkilde, B., Videbech, P., Rosenberg, R., Gjedde, A., & Gade, A. (2002). Putative
tests of ﬁ'ontal lobe function: A PET-study of brain activation during Stroop's
Test and verbal ﬂuency. Journal of Clinical & Experimental Neuropsychology
24(4), 534-547.

Reitan, R. M. (1958). Validity of the Trail Making Test as an indicator of organic brain
damage. Perceptual & Motor Skills, 8, 271-276.

Reitan, R. M., & Wolfson, D. (1995). Category Test and Trail Making Test as measures
of frontal lobe functions. Clinical Neuropsychologist 9(1), 50-56.

Retzlaff, P., Butler, M., & Vanderploeg, R. D. (1992). Neuropsychological battery choice
and theoretical orientation: A multivariate analysis. Journal of Clinical
Psychology, 48, 666-672. I

Reynolds, C. R., Willson, V. L., & Clark, P. L. (1983). A four-test short form of the
WAIS-R for clinical screening. Clinical Neuropsychology, 5, 114.

Rezai, K., Andreasen, N. C., Alliger, R., Cohen, G. et a1. (1993). The neuropsychology of
the preﬁ'ontal cortex. Archives of Neurology 50(6), 63 6-642.

Riccio, C. A., Wolfe, M., Davis, B., Romine, C., George, C., & Lee, D. (2005). Attention
Deﬁcit Hyperactivity Disorder: Manifestation in adulthood. Archives of Clinical
Neuropsychology 20(2), 249-269.

Riccio, C. A., Wolfe, M. E., Romine, C., Davis, B., & Sullivan, J. R. (2004). The Tower
of London and neuropsychological assessment of ADHD in adults. Archives of
Clinical Neuropsychology 19(5), 661-671.

Rieger, M., Gauggel, S., & Burmeister, K. (2003). Inhibition of ongoing responses
following frontal, nonfrontal, and basal ganglia lesions. Neuropsychology 1 7(2),
272-282.

Riordan, H. J ., Flashman, L. A., Saykin, A. J ., Frutiger, S. A., Carroll, K. E., & Huey, L.
(1999). Neuropsychological correlates of methylphenidate treatment in adult

206

ADHD with and without depression. Archives of Clinical Neuropsychology 14(2),
21 7-233.

Robins, L. M., Helzer, J. H., Crougho, J ., & Ratcliffe, K. S. (1980). National Institute of
Mental Health Diagnostic Interview Schedule: its history, characteristics, and
validity. Archives of General Psychiatry, 38, 381-3 89.

Rubia, K., Overmeyer, S., Taylor, R., Brammer, M., Williams, S. C. R., Simmons, A.,
Andrew, C., & Bullmore, E. T. (2000). Functional ﬁ'ontalisation with age:
Mapping neurodevelopmental trajectories with MRI. Neuroscience &
Biobehavioral Reviews 24(1), 13-19.

Rubia, K., Smith, A. B., Brammer, M. J ., & Taylor, E. (2003). Right inferior preﬁ'ontal
cortex mediates response inhibition while mesial prefrontal cortex is responsible
for error detection. NeuroImage, 20, 351-358.

Rubia, K., Smith, A. B., Brammer, M. J ., Toone, B., & Taylor, E. (2005). Abnormal
Brain Activation During Inhibition and Error Detection in Medication-Naive
Adolescents With ADHD. American Journal of Psychiatry 162(6), 1067-1075.

Rush, A. J ., Zimmerman, M., Wisniewski, S. R., Fava, M., Hollon, S. D., Warden, D.,
Biggs, M. M., Shores-Wilson, K., Shelton, R. C., Luther, J. F ., Thomas, B., &
Trivedi, M. H. (2005). Comorbid psychiatric disorders in depressed outpatients:
Demographic and clinical features. Journal of Aﬂective Disorders Vol, 87(1), 43-
55.

Sachs, G., Anderer, P., Margreiter, N., Semlitsch, H., Saletu, B., & Katschnig, H. (2004).
P300 event-related potentials and cognitive function in social phobia. Psychiatry
Research: Neuroimaging 131(3), 249-261.

Salthouse, TA. (2005). Relations between cognitive abilities and measures of executive
functioning. Neuropsychology, 19(4), 532-545.

Salthouse, T. A., Atkinson, T. M., & Berish, D. E. (2003). Executive Functioning as a
Potential Mediator of Age-Related Cognitive Decline in Normal Adults. Journal
of Experimental Psychology: General 132(4), 566-594.

Sareen, J ., Stein, M. B., Cox, B. J ., & Hassard, S. T. (2004). Understanding comorbidity
of anxiety disorders with antisocial behavior: Findings from two large community

surveys. Journal of Nervous & Mental Disease Vol, 192(3), 178-186.

Sattler, J. M., & Ryan, J. J. (1999). Assessment of Children: WAIS-III supplement San
Diego, CA: Jerome M. Sattler, Publisher, Inc.

Schweitzer, J. B., Faber, T. L., Grafton, S. T., Tune, L. E., Hoffman, J. M., & Kilts, C. D.
(2000). Alterations in the functional anatomy of working memory in adult

207

attention deﬁcit hyperactivity disorder. American Journal of Psychiatry 15 7(2),
278-280.

Seidman, L]. (2006). Neuropsychological functioning in people with ADHD across the
lifespan. Clinical Psychology Review, 26(4), 466-485.

Seidman, L. J ., Biederman, J ., Weber, W., Hatch, M., & Faraone, S. V. (1998).
Neuropsychological function in adults with attention-deﬁcit hyperactivity
disorder. Biological Psychiatry 44(4), 260-268.

Seidman, L. J ., Valera, E. M., & Bush, G. (2004). Brain function and structure in adults
with attention-deﬁcit/hyperactivity disorder. Psychiatric Clinics of North America
Special Issue: Adult Attention Deﬁcit Hyperactivity Disorder 2 7(2), 323-347.

Selby, M. J ., & Azrin, R. L. (1998). Neuropsychological functioning in drug abusers.
Drug & Alcohol Dependence, 50(1), 39-45.

- Selzer, M. L., Vinokur, A., & van Rooijen, L. (1975). A self administered short Michigan
Alcoholism Screening Test (SMAST). Journal of Studies on Alcohol, 36, 117-
126.

Sergeant, J. A., Geurts, H., & Oosterlaan, J. (2002). How speciﬁc is a deﬁcit of executive
functioning for attention-deﬁcit/hyperactivity disorder? Behavioural Brain
Research Special Issue: Neurobehavioural mechanisms in ADHD, 130(1-2) (Mar
2002), 3-28.

Shallice, T., & Burgess, P. (1991). Higher-order cognitive impairments and ﬁontal lobe
lesions in man, In: Levin, Harvey S ; Eisenberg, Howard M ; Frontal lobe
function and dysfunction ; 125-138; London University Press, 1991, xv, 42 7.

Sherer, M., Nick, T.G., Millis, S.R., & Novack, T.A. (2003). Use of the WCST and the
WCST-64 in the assessment of traumatic brain injury. Journal of Clinical and
Experimental Neuropsychology, 25(4), 512-520.

Simmonds, D.J., Fotedar, S.G., Suskauer, S.J., Pekar, J .J ., Denckla, M.B., & Mostofsky,
SH. (2007). Functional brain correlates of response time variability in children.
Neuropsychologia, 45(9), 2147-2157.

Skinstad, A. H., & Swain, A. (2001). Comorbidity in a clinical sample of substance
abusers. American Journal of Drug & Alcohol Abuse Vol, 27(1), 45-64.

Smith, E. E., & J onides, J. (1999). Storage and executive processes in the frontal lobes.
Science, 283(5408), 1657-1661.

Spitzer, R. L. (1999). Harmful dysfunction and the DSM deﬁnition of mental disorder.
Journal of Abnormal Psychology 108(3), 430-432.

208

Steiger, J .H. (1990). Structural model evaluation and modiﬁcation: An interval estimation
approach. Multivariate Behavioral Research, 25, 173-180.

Stein, M. B., Kennedy, C. M., & Twamley, E. W. (2002). Neuropsychological function in
female victims of intimate partner violence with and without posttraumatic stress
disorder. Biological Psychiatry 5 2(1 1), 1079-1088.

Stevens, M. C., Kaplan, R. F., & Hesselbrock, V. M. (2003). Executive-cognitive
functioning in the development of antisocial personality disorder. Addictive
Behaviors 28(2), 285-300.

Stuss, D. T., Benson, D. F ., Kaplan, E. F., Weir, W. S., & Della, M. C. (1981).
Leucotomized and nonleucotomized schizophrenics: Comparison on tests of
attention Biological Psychiatry, 16, 1085-1 100.

Stuss, D. T., Bisschop, S.M., Alexander, M.P., Levine, B., Katz, D., & Izukawa, D.
(2001a). The Trail Making Test: a study in focal lesion patients. Psychological
Assessment, 13(2), 230-239.

Stuss, D. T., Floden, D., Alexander, M. R, Levine, B., & Katz, D. (2001b). Stroop
performance in focal lesion patients: Dissociation of processes and frontal lobe
lesion location. Neuropsychologia 39(8), 771-786.

Stuss, D. T., Levine, B., Alexander, M. P., Hong, J ., Palumbo, C., Hamer, L., Murphy, K.
J ., & Isukawa, D. (2000). Wisconsin Card Sorting Test performance in patients
with focal ﬁ'ontal and posterior brain damage: Effects of lesion location and test
structure on separable cognitive processes. Neuropsychologia 38(4), 388-402.

Stuss, D. T. E., Knight, R. T., & Ed (Eds.). (2002). Principles of frontal lobe function.
London: Oxford University Press.

Stuss, D.T., Murphy, K.J., Binns, M.A., & Alexander, MP. (2003). Staying on the job:
the frontal lobes control individual performance variability. Brain: A Journal of
Neurology, 126(11), 2363-2380.

Sullivan, E. V., Fama, R., Rosenbloom, M. J ., & Pfefferbaum, A. (2002). A proﬁle of
neuropsychological deﬁcits in alcoholic women. Neumpsychology 16(1), 74-83.

Sutker, P. B., & Allain, A. N. (1987). Cognitive abstraction, shifting, and control:
Clinical sample comparisons of psychopaths and nonpsychopaths. Journal of
Abnormal Psychology 96(1), 73-75.

Swick, D., & J ovanovic, J. (2002). Anterior cingulate cortex and the Stroop task:

Neuropsychological evidence for topographic speciﬁcity. Neuropsychologia
40(8), 1240-1253.

209

Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd ed.). New
York: Harper Collins.

Taylor, C. J ., & Miller, D. C. (1997). Neuropsychological assessment of attention in
ADHD adults. Journal of Attention Disorders 2(2), 77-88.

Trichard, C., Martinot, J. L., Alagille, M., Masure, M. C. et al. (1995). Time course of
prefrontal lobe dysﬁmction in severely depressed in-patients: A longitudinal
neuropsychological study. Psychological Medicine 25(1), 79-85.

Twamley, E. W., Hami, S., & Stein, M. B. (2004). Neuropsychological function in
college students with and without posttraumatic stress disorder. Psychiatry
Research 126(3), 265-274.

Uekermann, J ., Daum, 1., Schlebusch, P., Wiebel, B., & Trenckrnann, U. (2003).
Depression and cognitive functioning in alcoholism. Addiction 98(11), 1521-
1529.

van den Heuvel, O. A., Veltrnan, D. J ., Groenewegen, H. J ., Cath, D. C., van Balkom, A.
J. L. M., van Hartskamp, J ., Barkhof, F., & van Dyck, R. (2005). Frontal-Striatal
Dysfunction During Planning in Obsessive-Compulsive Disorder. Archives of
General Psychiatry Vol, 62(3), 301-310.

Van Veen, V., & Carter, C. S. (2002). The anterior cingulate as a conﬂict monitor: MRI
and ERP studies. Physiology & Behavior Special Issue: The Pittsburgh special
issue Vol, 7 7(4-5) (Dec 2002), 477-482.

Vasey, M. W., Dalgleish, T., Silverman, W. K. (2003). Research on information-
processing factors in child and adolescent psychopathology: A critical
commentary. Journal of Clinical Child and Adolescent Psychology, 32, 81-93.

Vendrell, P., Junqué, C., Pujol, J ., Jurado, M. A. et a1. (1995). The role of prefrontal
regions in the Stroop task. Neuropsychologia 33(3), 341-352.

Videbach, P., Ravnkilde, B., Pedersen, A. R., Egander, A., Landbo, B., Rasmussen, N.
A., Andersen, F., Stodkilde-Jorgensen, H., Gjedde, A., & Rosenberg, R. (2001).
The Danish PET/depression project: PET ﬁndings in patients with major
depression. Psychological Medicine Vol, 31(7), 1147-1158.

Volkow, N. D., Ding, Y.-S., Fowler, J ., & Wang, G.-J. (1996). Cocaine addiction:
Hypothesis derived from imaging studies with PET. Journal of Addictive
Diseases 15(4), 55-72.

Volkow, N. D., Hitzemann, R., Wang, G.-J., Fowler, J. S. et al. (1992). Decreased brain

metabolism in neurologically intact healthy alcoholics. American Journal of
Psychiatry 149(8), 1016-1022.

210

Wakeﬁeld, J. C. (1992). The concept of mental disorder: On the boundary between
biological facts and social values. American Psychologist 47(3), 373-3 88.

Wakeﬁeld, J. C. (1997). Diagnosing DSM-IV--Part I: DSM-IV and the concept of
disorder. Behaviour Research & Therapy 35(7), 633-649.

Wakeﬁeld, J. C. (1999). Evolutionary versus prototype analyses of the concept of
disorder. Journal of Abnormal Psychology 108(3), 374-399.

Waldstein, S. R., Ryan, C. M., Jennings, J. R., Muldoon, M. F. et al. (1997). Self-reported
levels of anxiety do not predict neuropsychological performance in healthy men.
Archives of Clinical Neuropsychology 12(6), 567-574.

Walker, A. J ., Shores, E. A., Trollor, J. N., Ice, T., & Sachdev, P. S. (2000).
Neuropsychological functioning of adults with attention deﬁcit hyperactivity
disorder. Journal of Clinical & Experimental Neuropsychology 22(1), 115-124.

Ward, G., Roberts, M. J ., & Phillips, L. H. (2001). Task-switching costs, Stroop-costs,
and executive control: A correlational study. Quarterly Journal of Experimental
Psychology: Human Experimental Psychology, 54(2), 491 -51 1.

Watson, D. (2005). Rethinking the mood and anxiety disorders: a quantitative
hierarchical model for DSM-V. Journal of Abnormal Psychology, 114(4), 522-
536.

Wechsler, D. (1981). WAIS-R Manual. New York: Psychological Corporation.

Wechsler, D. (1997). WAIS-III Manual. San Antonio, Texas: Psychological Corporation.

Welsh, M. C., & Pennington, B. F. (1988). Assessing frontal lobe ﬁrnctioning in children:
Views ﬁ'om developmental psychology. Developmental Neuropsychology 4(3),
199-230.

Wender, P. H., Wolf, L. E., & Wasserstein, J. (2001). Adults with ADHD: An overview.
Annals of the New York Academy Sciences, 931, 1-16.

West, R.L. (1996). An application of prefrontal cortex ﬁrnction theory to cognitive
aging. Psychological Bulletin, 120, 272-292.

Weyandt, L. L., Rice, J. A., Linterman, 1., Mitzlaff, L., & Emert, E. (1998).
Neuropsychological performance of a sample of adults with ADHD,
Developmental Reading Disorder, and controls. Developmental Neuropsychology
14(4), 643-656.

211

Wilcox, R.R., Keselman, H.J., & Kowalchuk, R.K. (1998). Can tests for treatment group
equality be improved?: The bootstrap and trimmed means conjecture. British
Journal of Mathematical and Statistical Psychology, 51, 123-134.

Willcutt, E. G., Doyle, A. E., Nigg, J. T., Faraone, S. V., & Pennington, B. F. (2005).
Validity of the executive function theory of attention-deﬁcit/hyperactivity
disorder: A meta-analytic review. Biological Psychiatry.

Wonderlich, S. A., & Mitchell, J. E. (1997). Eating disorders and comorbidity: Empirical,
conceptual, and clinical implications. Psychopharmacology Bulletin Special
Issue: Research Priorities in Eating Disorders, 33(3), 381-390.

Woods, S. P., Lovejoy, D. W., & Ball, J. D. (2002). Neuropsychological characteristics
of adults with ADHD: A comprehensive review of initial studies. Clinical
Neuropsychologist, 16(1), 12-34.

Zametkin, A. J ., Nordahl, T. E., Gross, M., King, A. C. et al. (1990). Cerebral glucose
metabolism in adults with hyperactivity of childhood onset. New England Journal
of Medicine 323(20), 1361-1366.

Zucker, R. A., Ellis, D. A., Bingham, C. R., & Fitzgerald, H. E. (1996). The development
of alcoholic subtypes: Risk variation among alcoholic families during the early
childhood years. Alcohol Health & Research World 20(1), 46-55.

Zucker, R. A., Ellis, D. A., & Fitzgerald, H. E. (1992). Developmental evidence for at
least two alcoholisms: I. Biopsychosocial variation among pathways into
symptomatic difﬁculty, In: Babor, Thomas F ,' Hesselbrock, Victor M ,° Types of
alcoholics: Evidence ﬁ'om clinical, experimental, and genetic research ,° 134-146;
New York, NY.

Zucker, R. A., Ellis, D. A., Fitzgerald, H. E., & Bingham, C. R. (1996). Other evidence
for at least two alcoholism 11. Life course variation in antisociality and
heterogeneity of alcoholic outcome. Development & Psychopathology, 8(4), 831-
848.

Zucker, R. A., Fitzgerald, H. E., & Noll, R. B. (1990). Drinking and drug history (rev.
ed., version 4). Unpublished instrument, Michigan State University, Department
of Psychology, East Lansing, MI.

Zucker, R. A., Noll, R. B., Ham, H. P., Fitzgerald, H. E., & Sullivan, L. S. (1994).
Assessing antisociality with the Antisocial Behavior Checklist: reliability and
validity studies. Unpublished manuscript, University of Michigan, Department of
Psychiatry, Ann Arbor, MI, and Michigan State University, Department of
Psychology, East Lansing, MI.

212

   

uljltjjjlyjjj\njjjr