University of Southern California Dissertations and Theses

Psychopathic traits and the fronto-amygdala circuit in a community sample: the role of trait anxiety

(USC Thesis Other)

Psychopathic traits and the fronto-amygdala circuit in a community sample: the role of trait anxiety

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

PSYCHOPATHIC TRAITS AND THE FRONTO-AMYGDALA CIRCUIT IN A
COMMUNITY SAMPLE: THE ROLE OF TRAIT ANXIETY

By

Mona Sobhani

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

December 2013

Copyright 2013 Mona Sobhani

ii

EPIGRAPH

“In the desert, an old monk once advised a traveler that the voices of God and the
Devil are barely distinguishable.” ~ Loren Eiseley

iii

DEDICATION

For my mother, father, and brother

iv

ACKNOWLEDGEMENTS

There are so many elements that go into any given experience, and the writing of
a dissertation is no exception. I hardly know where to begin acknowledging all
the people, relationships, and circumstances that have influenced me as I
completed my doctoral work in neuroscience at the University of Southern
California. I cannot end this adventure without thanking the many individuals
who have walked with me and helped along the way.

First and foremost, I cannot thank enough my primary advisor, Dr. Lisa Aziz-
Zadeh, for patiently bringing me into the world of neuroimaging. I am forever
indebted to her for taking a chance and allowing me to design my research
project on my primary research interest (psychopathy). It has been a pleasure
working alongside you, and learning how to make science approachable for
others. I must thank my co-advisor, Dr. Laura Baker, for her endless
encouragement and advice, as well as the generous donation of her Twin Study
participant pool, around which my entire dissertation work revolves. Laura has
been a true inspiration to me, not only as a researcher and as a person, but also
as a successful woman in science. I must thank the members of my advisement
committee: Dr. Michael Dawson; for teaching me the principles of fear
conditioning and holding my science to a high standard; to Dr. Antoine Bechara,
for being the first person interested in my dissertation project and for exposing

v

me to the complex world of publishing science; and Dr. Hanna Damasio, for her
kindness, wisdom, boundless encouragement and for inspiring me to be a better
scientist and leader. A special thank you to Elyn Saks for providing me the
opportunity to join The Saks Institute at USC during my last year of graduate
school, and for agreeing to be on my dissertation committee. The experience
within the institute opened my eyes to a whole other world, and reminded me of
the importance of integrating neuroscience with other fields, and I am grateful for
it. I extend the warmest thank you to Anne Schell for her enormous knowledge
base, and willingness to help.

I will always have a special place in my heart for Glenn Fox, with whom I
struggled through our first fMRI project (from scratch!). I also must thank Jonas
Kaplan for his collaboration with Glenn and I on our first fMRI paper (which would
have been impossible without him), as well as Savio Wong for happily answering
all my questions about skin conductance recordings. A warm thank you to Assal
Habibi, Sarah Gimbel, and Gil Carvalho for brightening lab and making the Brain
and Creativity Institute a wonderful place to work.

I need to thank my labmates Sook-Lei Liew, Kathleen Garrison, Tong Sheng, and
Julie Werner for inspiring me to master my craft. Without their advice,
encouragement, commiseration, hugs, and laughs this thesis never would have

vi

been possible. A special thanks to Lei for, oh, so many things, but especially the
company in lab, words of wisdom, guidance, friendship, and inspiration to always
work harder (as well as the cute handwritten notes).

Among the graduate students in the Brain and Creativity Institute, we always say
that we are lucky to work with such amiable people, who fulfill dual roles as
friends and colleagues. I truly believe this, and thus, must acknowledge that
graduate school was made an enjoyable experience through all the wonderful
graduate students I had the pleasure of knowing. Each of you has contributed to
making my years in graduate school some of the best yet. This list includes (in
alphabetical order): Madeline Andrews, Helder Araujo, Farhan Baluch, Victor
Barres, Jimmy Bonaiuto, Henryk Bukowski, Dave Clewett, Kevins Crimi, Sims
Dulai, Ryan Essex, Glenn Fox, Braddy Gasser, Dave Herman, Panthea Heydari,
Anna Kamitakahara, Vilay Khandelwal, Natalie Kintz, Leigh Komperda, Lei Liew,
Nichole Lighthall, Kingson Man, Jenny McGrady, Meghen Miles, Raina Pang,
Shane Roach, Bella Rozenkrants, Jillian Shaw, Tong Sheng, Erin Zomber, Julie
Werner, and Fei Yang. A special thank you to Brad Gasser for showing me what
true passion for science and politics looks like, and for sharing a love of the
Beatles and basketball (even though you are a Lebron fan). Another warm thank
you to Simren Dulai for being my Westside buddy and for getting me to admit I
love trail running. I must separately thank Helder for always being the best

vii

person to lean on, Fei for being the nicest person ever and a great carpool
buddy, and Kingson for our study breaks. Kevins also gets a special thank you
for providing me with so many laughs (and drawings for my quotes!).

Without the many research assistants I have had the pleasure to work with, I
would have been at USC years longer! Thank you, Benny Heikali, Jenna
Elkington, Alex Gong, Brad Martins, Rachel Sisson-Rosenstein, Lauren Cutler,
Amy Tressan, Janelle Wang, and Zade Shakir. Wishing you the best of luck in
all your endeavors!

There are no words to express the gratitude I feel towards my family and friends,
who supported and encouraged me over the years. I have to thank my mother
and father for instilling in me the desire to continue learning and for teaching me
the importance of higher education. Without their love and support, I never would
have kept on trudging. While this work is very important to me, you should know
that much of it is for you, Mom and Dad. A special loving thank you goes to my
big brother, Rama, for being the smartest person I know, and inspiring me to be
smart, too. The advice he has given me over the years is likely much more
valuable to me than he ever knew. Also, we make really cool cartoons together
(Mermoo lives!). It is impossible that my childhood Saturday morning lab

viii

experiments (white coats and all!) with my cousin, Yasha, did not instill a love of
science within me (thanks for starting that, Pooshes).

I have to thank my closest friends. Thank you to (in alphabetical order): Amna,
for Tuesday reminders and for keeping me motivated to be excellent and well-
rounded; Chantalle, for everything under the sun, but especially our talks, our
hikes, and our wonderful friendship; Chrystal, for “living life” with me, especially in
the southern states; Erica, for being my best friend since 7
th
grade and still telling
it to me like it is; Judy, for our adventures, for celebrating the holidays
(extravagantly) with me, and especially for tolerating my moods during graduate
school; Matt, for the comfort that comes from our friendship; Niousha, for always
understanding exactly what I mean and for the 22 years of laughs; Stephanie, for
being an awesome person with whom to share experiences; and Tina, for your
encouragement and for always knowing the perfect thing to say in any given
situation. I cherish all your friendships more than words can say and have to
give a big collective thank you for providing those sparkling moments of meaning
to my life.

My graduate work was supported by the Neuroscience Graduate Program
Fellowship, the Dornsife Imaging Center, the Brain and Creativity Institute, and
The Saks Institute.

ix

TABLE OF CONTENTS

EPIGRAPH ii
DEDICATION iii
ACKNOWLEDGEMENTS iv
LIST OF TABLES xi
LIST OF FIGURES xiv
ABSTRACT xvii
CHAPTER 1: Introduction 1
1.1 Psychopathy 2
1.1.1 Behavioral Evidence 5
1.1.2 Physiological Evidence 6
1.2 Psychopathy and the Brain 7
1.2.1 Comparison with Brain Lesion Patients 7
1.2.2 Ventromedial Prefrontal Cortex 9
1.2.2.1 Anterior Cingulate Cortex 12
1.2.3 Ventromedial Prefrontal Cortex and
Psychopathy 14
1.2.3.1 Ventromedial Prefrontal Cortex
Function 14
1.2.4 Amygdala 15
1.2.5 Amygdala-Ventromedial Prefrontal Circuit 18
1.2.6 Amygdala and Psychopathy 19
1.2.6.1 Amygdala Function 19
1.2.7 Amygdala-Ventromedial Prefrontal and
Psychopathy 21
1.2.8 Children and Adolescents with Conduct
Disorder/Oppositional Defiant Disorder 22
1.3 Subtypes of Psychopathy Based on Trait Anxiety 25
1.3.1 Behavioral and Physiological Evidence 27
1.3.2 Subtypes and the Brain 32

CHAPTER 2: Trait anxiety interacts with psychopathic traits to predict
amygdala BOLD activity during fear conditioning
2.1 Abstract 35
2.2 Introduction 36
2.3 Materials and Methods 37
2.4 Results 57
2.5 Discussion 76

x

CHAPTER 3: Psychopathic traits, trait anxiety, and amygdala-ventromedial
prefrontal cortex functional connectivity
3.1 Abstract 89
3.2 Introduction 90
3.3 Materials and Methods 93
3.4 Results 102
3.5 Discussion 115

CHAPTER 4: Psychopathic traits modulate microstructural integrity of right
uncinate fasciculus
4.1 Abstract 123
4.2 Introduction 124
4.3 Materials and Methods 126
4.4 Results 130
4.5 Discussion 134

CHAPTER 5: DISCUSSION 141
REFERENCES 154
APPENDICES
Appendix I: Recruitment Information 180
Appendix II: Zygosity and Co-twin Information 182
Appendix III: Welsh Anxiety Scale as a measure of 184
trait anxiety

xi

LIST OF TABLES

Table 1-1. Items from Hervey Cleckley’s original psychopathy checklist 3
Table 1-2. Items from Robert Hare’s Psychopathy Checklist-Revised 4
Table 1-3. DSM-IV Criteria for Oppositional Defiant Disorder (ODD) and
Conduct Disorder (CD) 23

Table 2-1. Correlation Matrix for youth psychopathy measures used in the
composite score 39

Table 2-2. Age, composite psychopathy score, and race information for
potential participant pool and actual participant pool 43

Table 2-3. Descriptive statistics and correlations for trait anxiety and
psychopathy measures 58

Table 2-4. Descriptive statistics for trait anxiety and psychopathy subscales 58

Table 2-5. Means and standard deviations for skin conductance responses
(SCR) across both runs, by condition and SCR phase 62

Table 2-6. Unstandardized and standardized coefficients and t-values for
regression model of the interaction between trait anxiety and
psychopathic traits to predict BOLD activity in bilateral amygdala 70

Table 2-7. Unstandardized and Standardized Coefficients and T-values for
the regression model of the interaction between trait anxiety and
psychopathic traits in predicting skin conductance responses
during fear conditioning 74

Table 3-1. Localization of clusters within VMPFC significantly functionally
connected to Left Amygdala during Resting State. 103

Table 3-2. Localization of clusters within VMPFC where right
amygdala-VMPFC connectivity is negatively correlated with
total PPI scores. 104

Table 3-3. Localization of clusters within Left Amygdala where significantly
functionally connected to Right Amygdala during Resting State 106

xii

Table 3-4. Localization of clusters within Right STG where significantly
functionally connected to Right Amygdala during Resting State 106

Table 3-5. Unstandardized and standardized coefficients and t-values for
the interaction between trait anxiety and psychopathic traits that
predicts amygdala-VMPFC functional connectivity during resting
state 108

Table 3-6. Localization of clusters of amygdala-VMPFC connectivity during
fear conditioning that are negatively correlated with Blame
Externalization. 110

Table 3-7. Localization of clusters of right amygdala connectivity with
other fear-conditioning related regions during fear conditioning
that are negatively correlated with total PPI score 113

Table 1. Number of participation invitations by level of psychopathic traits
and wave of recruitment 180

Table 2. Participation by wave of recruitment and level of psychopathic traits. 181

Table 1. Zygosity of potential participant pool and actual participant pool. 182

Table 1. Descriptive statistics and correlations for trait anxiety and
psychopathy measures 185

Table 2. Descriptive statistics for trait anxiety and psychopathy subscales 186

Table 3. Model statistics for reanalyzed significant multiple regression
models of brain data with trait anxiety items of WAS 188

Table 4. Unstandardized and standardized coefficients and t-values of
reanalyzed significant multiple regression models of brain data
with trait anxiety items of WAS 189

Table 5. Model statistics for reanalyzed significant multiple regression
models of skin conductance data with trait anxiety items of WAS 189

xiii

Table 6. Unstandardized and standardized coefficients and t-values of
reanalyzed significant multiple regression models of skin
conductance with trait anxiety items of WAS 190

xiv

LIST OF FIGURES

Figure 1-1. Location of the ventromedial prefrontal cortex (VMPFC) 10
Figure 1-2. Location of anterior cingulate cortex (ACC) 13
Figure 1-3. Location of left and right amygdala 16
Figure 1-4. Amygdala and ventromedial prefrontal cortex (VMPFC) are
structurally and functionally reciprocally interconnected 18

Figure 2-1. Distribution of the composite psychopathy score in entire male
twin sample 40

Figure 2-2. Distribution of the composite psychopathy score within the MRI
participant pool 41

Figure 2-3. Schematic illustration of the experimental design 47

Figure 2-4. Masks of ventromedial prefrontal cortex (VMPFC) 54

Figure 2-5. Masks of amygdala 55

Figure 2-6. Skin conductance responses by phase and condition 61

Figure 2-7. Skin conductance orienting response (OR) difference scores
correlate with right amygdala activity during first half of fear
conditioning. 63

Figure 2-8. Skin conductance orienting response (OR) difference scores
correlate with right amygdala activity during second half of fear 64

Figure 2-9. Skin conductance orienting response (OR) difference scores
correlate with left amygdala activity during second half of fear 64

Figure 2-10. Parameter estimates by region of interest (ROI) for the first half
of fear conditioning. 66

Figure 2-11. Parameter estimates by region of interest (ROI) for the second
half of fear conditioning 67

xv

Figure 2-12. Left Amygdala BOLD activity negatively correlates with
Machiavellian Egocentricity 68

Figure 2-13. Trait anxiety interacts with Machiavellian Egocentricity to predict
BOLD activity in right amygdala 70

Figure 2-14. Trait anxiety interacts with Machiavellian Egocentricity to predict
BOLD activity in left amygdala. 71

Figure 2-15. Trait anxiety interacts with Blame Externalization to predict
orienting skin conductance responses during fear conditioning 75

Figure 2-16. Trait anxiety interacts with Carefree Nonplanfulness to predict
orienting skin conductance responses during fear conditioning. 76

Figure 3-1. Example of a handdrawn amygdala region of interest on an
individual subject’s brain. 96

Figure 3-2. Regions within ventromedial prefrontal cortex (VMPFC) that
show decreased resting state connectivity to right amygdala
with increased psychopathy scores 104

Figure 3-3. Total Psychopathic Personality Inventory (PPI) score negatively
correlates with right amygdala-ventromedial prefrontal cortex
(VMPFC) connectivity at rest. 105

Figure 3-4. Trait anxiety interacts with total Psychopathic Personality
Inventory (PPI) score to predict resting state functional
connectivity between amygdala and ventromedial prefrontal
cortex (VMPFC). 109

Figure 3-5. Regions within ventromedial prefrontal cortex (VMPFC) that
show decreased resting state connectivity with right amygdala
with increased Blame Externalization scores 111

Figure 3-6. Right amygdala functional connectivity during fear conditioning
to key fear conditioning regions modulated by total
Psychopathic Personality Inventory (PPI) scores. 114

Figure 4-1. Example of a reconstructed uncinate fasciculus (UF) tract using 128
diffusion tensor imaging.

xvi

Figure 4-2. Example of a reconstructed inferior longitudinal fasciculus (ILF). 129

Figure 4-3. Psychopathy scores from ages 14-16 negatively correlate with
right uncinate fasciculus fractional anisotropy (FA) values. 132

Figure 4-4. Adult psychopathy scores negatively correlate with right uncinate
fasciculus fractional anisotropy (FA) values when psychopathy
score switchers are removed. 133

Figure 5-1. Summary of main findings. 145

Figure 5-2. Other neural regions possibly implicated in psychopathy. 150

Figure 1. Quartile placement of co-twins by participant quartile. 183

xvii

ABSTRACT

Psychopathy is a personality disorder comprised of a constellation of traits, such
as lack of remorse, lack of empathy, impulsivity, and irresponsibility. Although
much research has been conducted on this personality disorder, the neural
correlates remain poorly understood. Progress in this realm has been hindered
by two factors. First, given the large body of work supporting delineation of
psychopathy into subtypes (based on trait anxiety), the exploration of dissociable
neural correlates between subtypes has been insufficiently studied. Second,
many studies utilize a single neuroimaging modality to explore a given neural
property, and thus, cannot answer comprehensive questions about neural
network dynamics. The current work addresses both limitations in the literature
by examining, in a community sample, the relationships between psychopathic
traits, trait anxiety, and neural properties (function, functional connectivity,
microstructural integrity) using multiple neuroimaging methods. The two
overarching aims of this dissertation are to determine whether the relationship
between psychopathic traits and neural properties is dependent upon trait
anxiety, and whether these relationships are evident across several different
neuroimaging modalities.

1

CHAPTER 1: INTRODUCTION
Psychopathy is a personality disorder typically represented by two dimensions of
traits: interpersonal-affective and antisocial-impulsivity. The exact functional and
structural neural underpinnings of the behavioral profile remain unknown,
although the fronto-amygdala circuit (ventromedial prefrontal cortex (VMPFC)
and amygdala) has been frequently implicated. Progress in understanding the
disorder has been hindered by the fact that psychopathy has been viewed as a
single construct, though much research shows it to be a multifaceted personality
disorder (J. L. Skeem, Mulvey, & Grisso, 2003). Understanding of the personality
disorder has also been hindered by neuroimaging research that focuses on a
single imaging modality (e.g. function or structure) in one population, making
legitimate comparisons across studies difficult, since most studies also differ on
many other factors. Part of the heterogeneity of the disorder may be explained
by accounting for trait anxiety when considering variants of the disorder (J.
Skeem, Johansson, Andershed, Kerr, & Louden, 2007). For instance, in tasks
such as fear conditioning, when classified by trait anxiety, Individuals high on
psychopathy but low on trait anxiety display behavioral and autonomic deficits
that individuals high on psychopathy and high on trait anxiety do not (Lykken,
1957). Additionally, behavioral studies reveal that while individuals with
psychopathy behave similarly to patients with lesions to the VMPFC and
amygdala (R. J. Blair, 2008; A.R. Damasio, 1994), the behavior of individuals
with high psychopathy and low trait anxiety more closely resembles that of

2

patients with lesions to the VMPFC than those with high psychopathy and high
trait anxiety (Koenigs, Kruepke, & Newman, 2010). Thus, it is reasonable to
hypothesize that the relationship between psychopathic traits and the fronto-
amygdala circuit, in terms of neural structure and function (as well as behavioral
and autonomic responses) may be dependent on levels of trait anxiety. In this
chapter, I will introduce psychopathy, the fronto-amygdala circuit (referred to as
amygdala-VMPFC circuit from hereforth), and background studies (including
neuroimaging) on the neurobiology of psychopathy and the link between trait
anxiety and subtypes of psychopathy.

1.1 PSYCHOPATHY
In hospitalized patients of a psychiatric institution, Hervey Cleckley identified a
previously ignored subgroup of patients as having some mental illness, which
was not readily detectable from superficial interactions with the individuals.
Informed by encounters with these suave, charming, and yet deviant individuals,
whom he called psychopaths, Cleckley established a list of defining characteristic
traits to aid the identification of the disorder (Cleckley, 1982). A sample of these
characteristics include: superficial charm, lack of remorse or shame, poor
judgment, failure to learn from experience, egocentricity, poverty of emotions,
lack of anxiety, unreliability, failure to plan ahead, pathological lying and failure to
maintain interpersonal relationships (see Table 1-1 for full checklist) (Cleckley,
1982).

3

Table 1-1. Items from Hervey Cleckley’s original psychopathy checklist
# Items
1 Superficial charm and good “intelligence”
2 Absence of delusions and other signs of irrational thinking
3 Absence of “nervousness” or psychoneurotic manifestations
4 Unreliability
5 Untruthfulness and insincerity
6 Lack of remorse and shame
7 Inadequately motivated antisocial behavior
8 Poor judgment and failure to learn by experience
9 Pathologic egocentricity and incapacity for love
10 General poverty in major affective reactions
11 Specific loss of insight
12 Unresponsiveness in general interpersonal relations
13 Fantastic and uninviting behavior with drink and sometimes without
14 Suicide rarely carried out
15 Sex life impersonal, trivial, and poorly integrated
16 Failure to follow any life plan

Based on Cleckley’s initial trait list, Robert Hare created a checklist designed to
capture the main personality features of psychopathy, called the Psychopathy
Checklist, which was eventually refined and revised (PCL-R; (R.D. Hare, 1991).
The PCL-R was designed using offender populations, and is thus, most fitting for
assessing psychopathy in forensic settings. The personality construct is usually
viewed as having two main components: an interpersonal and affective
component and an antisocial and deviant component (see Table 1-2 for full
checklist) (Robert D. Hare, 1999).

4

Table 1-2. Items from Robert Hare’s Psychopathy Checklist-Revised
Factor 1: Personality (Affective-Interpersonal)
Glibness/superficial charm
Grandiose sense of self-worth
Pathological lying
Cunning/manipulative
Lack of remorse or guilt
Shallow affect
Callousness; lack of empathy
Failure to accept responsibility for his or her own actions
Factor 2: Case history (Socially deviant lifestyle)
Need for stimulation/proneness to boredom
Parasitic lifestyle
Poor behavioral control
Lack of realistic long-term goals
Impulsivity
Irresponsibility
Juvenile delinquency
Early behavior problems
Revocation of conditional release
Traits that do not correlate with either factor
Promiscuous sexual behavior
Many short-term (marital) relationships
Criminal versatility

True to Cleckley’s understanding of the disorder, individuals with psychopathy
are experts at wreaking havoc in their own lives, as well as in the lives of people
they encounter. They are masterful manipulators, unrestrained by any morsel of
a conscience that would allow them to feel remorse or guilt for the consequences
of their actions. As such, individuals with psychopathy comprise approximately
25-35% of US prison populations (Hare, 1990). The estimated annual cost of
crime in the United States is approximately $1 trillion (Anderson, 1999), and
offenders with psychopathy contribute greatly to this cost as they have the

5

highest rates of recidivism of all offenders (Hare, 1990). Psychopaths are also
notoriously untreatable, leading a persistent problem for society. Since Cleckley’s
groundbreaking book (Hare, 1990), much research on the etiology, nature and
biomechanisms of the disorder has been conducted, yet the precise neural
underpinnings remain elusive. The elucidation of a neurobiological profile may
enable targeted, effective treatment of this troubled population. Effective
treatment may alleviate the strain these individuals cause the justice and health
care systems, as well as society at large.

1.1.1 BEHAVIORAL EVIDENCE
There is a wealth of behavioral evidence supporting the notion of abnormal
affective processes and antisocial behavior in individuals with psychopathy. In
one study, motor responses were recorded after emotional distractors were
presented immediately before and after rapidly presented target stimuli.
Typically developed controls showed increased response latency if stimuli were
coupled with an emotional distractor, whereas individuals with criminal
psychopathy did not (Mitchell, Richell, Leonard, & Blair, 2006). Adults with
psychopathy and children with psychopathic traits have also shown impairments
in identifying and naming emotional facial expressions (R. J. Blair, 2004; R. J.
Blair & Cipolotti, 2000; Dadds, El Masry, Wimalaweera, & Guastella, 2008; Dolan
& Fullam, 2006; Kosson, Suchy, Mayer, & Libby, 2002). Both psychopathic
populations have also shown impairments in recognizing emotional vocal affect

6

(R. J. Blair, Colledge, Murray, & Mitchell, 2001; Stevens, Charman, & Blair,
2001)). When it comes to moral reasoning, adult psychopaths and children with
psychopathic tendencies have difficulties making distinctions between moral
transgressions and more conventional transgressions and they are more likely to
deem moral transgressions acceptable (R.J.R. Blair, 1997; R.J.R Blair, Jones,
Clark, & Smith, 1995). Taken together, these findings imply that individuals with
high levels of psychopathic traits differ from healthy controls in basic affective
processing in multiple sensory domains, as well as in social behavior.

1.1.2 PHYSIOLOGICAL EVIDENCE
There is also an extensive literature on the psychophysiology of individuals with
psychopathy, indexing deficits in affective arousal. Much of this work has
included the recording of skin conductance responses (SCR) during experimental
tasks. SCRs are the recorded changes of moisture levels on the skin, which
confer electrical conductance (Dawson, Schell, Filion, 2000). Since the sweat
glands that produce the measured moisture are controlled by the sympathetic
nervous system, it is believed that changes in skin conductivity are a proxy to
emotional responses (Carlson, 2013). One of the most well established findings
in psychopathy research is that individuals with psychopathy do not display, or
show only greatly reduced, SCRs in a number of situations that would typically
induce emotional autonomic responding in normal individuals, including
responses to noxious stimuli (i.e. insertion of hypodermic needle, mutilated faces

7

(R. D. Hare, 1972; Mathis, 1970), and in anticipation of receiving painful stimuli
(R. D. Hare, 1965; R. D. Hare & Quinn, 1971). Individuals with psychopathy
have also failed to exhibit anticipatory SCRs in classic fear conditioning tasks,
during anticipation of loud noises or shock (Flor, Birbaumer, Hermann, Ziegler, &
Patrick, 2002; R. D. Hare, 1978; R. D. Hare & Thorvaldson, 1970; Lykken, 1957).
One of the defining criteria of psychopathy is having reduced empathic
responding to victims; highly antisocial individuals typically display this behavior
towards the distress of their victims (Chaplin, Rice, & Harris, 1995; Perry & Perry,
1974). In two studies where individuals with psychopathy viewed confederates
that they believed were receiving painful electrical shocks, reduced skin
conductance responses were recorded, when compared to controls (Aniskiewicz,
1979; House & Milligan, 1976). In another study, adults with psychopathy and
children with psychopathic tendencies viewed images of others in distress, and
displayed reduced autonomic responding when compared to controls (R.J.R.
Blair, 1997; R.J.R. Blair, 1999). In sum, individuals with psychopathy, when
compared to controls, display reduced physiological responses, alongside the
previously mentioned behavioral differences.

1.2 PSYCHOPATHY AND THE BRAIN
1.2.1 COMPARISON WITH BRAIN LESION PATIENTS
A fair amount of evidence purports that individuals with psychopathy display
abnormal behaviors, both in laboratory settings and in real-life situations.

8

Evidence also supports that they possess abnormal autonomic responding to
emotional or stressful situations. But, is there something wrong in the brain?
Similarities in behavior between individuals with psychopathy and ventromedial
prefrontal cortex lesions (VMPFC; see Figure 1 below) have been recognized for
decades and continue to be discussed in various studies (A. R. Damasio, Tranel,
& Damasio, 1990; Koenigs et al., 2010). Indeed, Antonio Damasio labeled the
collection of newly observable behaviors in VMPFC patients as “acquired
sociopathy” (A. R. Damasio et al., 1990). A sample of these overlapping
characteristics include: lack of empathy, irresponsibility, poor decision-making,
inappropriate social behavior, and failure to plan ahead (Koenigs et al., 2010).
Like VMPFC patients, individuals with psychopathy can know and say “the right
thing,” but they will do “the wrong thing” (Cleckley, 1982). Evidence suggests
impairment of moral judgment in both groups, such that both are more likely to
use utilitarian approaches to moral dilemmas than control subjects (R.J.R. Blair,
1997; Koenigs et al., 2007), and are more likely to approve of harmful actions in
situations they deem as appropriate or reasonable (Arsenio & Fleiss, 1996;
Young et al., 2010).

In addition to the parallels between VMPFC patients and individuals with
psychopathy, it has been suggested that amygdala dysfunction is also involved in
the disorder (See Figure 3 below). Behaviorally, individuals with psychopathy
display deficits in fear conditioning (Birbaumer et al., 2005), fearful facial

9

expression recognition and processing (R. J. Blair et al., 2001), passive
avoidance learning (J. P. Newman & Schmitt, 1998), augmentation of startle
reflex by visual threat primes (Levenston, Patrick, Bradley, & Lang, 2000), and
less interference by emotional distracters (Mitchell et al., 2006) – all of which
have been described as consequences of dysfunction of the amygdala. Besides
behavioral similarities, there are some parallels in physiological data, as well. As
a result of these behavioral similarities between individuals with psychopathy and
VMPFC and amygdala patients, it has been suggested that individuals with
psychopathy have abnormal neural function or even altered structure in these
regions (J. Blair, Mitchell, & Blair, 2005; A. R. Damasio et al., 1990).

1.2.2 VENTROMEDIAL PREFRONTAL CORTEX
Before we try to understand the role of the VMPFC in human behavior, it is
necessary to first define the ventromedial prefrontal cortex neuroanatomically,
and to describe its structural connections to other brain regions. When creating
regions of interest (ROIs) for neuroimaging studies, especially for the purposes of
psychopathy studies, the VMPFC is often defined as the inferior portion of the
prefrontal cortex, extending from the inter-hemispheric fissure to the medial
orbital sulcus, defined superiorly by a straight line from the genu of the corpus
collosum, to the most anterior portion of the frontal pole, including the pre- and
sub-callosal sectors of the anterior cingulate cortex (ACC) (Motzkin, Newman,
Kiehl, & Koenigs, 2011) (See Figure 1-1). These regions include Brodmann

10

areas 14, and parts of 10, 11, 12, 13, 25, and 32. For the purposes of this
discussion about VMPFC function (which does not need to be defined in terms of
an ROI) all prefrontal areas of the VMPFC ‘ROI’ will be referred to as VMPFC,
excluding the ACC, which will be discussed in a separate section below.

Figure 1-1. Location of the ventromedial prefrontal cortex (VMPFC). The
region of interest is shown as it is often defined in neuroimaging studies.

The VMPFC has direct structural connections with multiple other brain regions. It
receives projections from higher association areas of all sensory modalities (e.g.
visual, auditory, gustatory, olfactory, somatosensory) (Carmichael & Price,
1995c; Cavada, Compañy, Tejedor, Cruz-Rizzolo, & Reinoso-Suárez, 2000;
Romanski et al., 1999), is weakly connected to the motor system (Carmichael &
Ventromedial Prefrontal Cortex (VMPFC)

11

Price 1995), is extensively connected with the limbic system (e.g. amygdala,
hippocampus, cingulate) (Carmichael & Price, 1995), has influence over
autonomic functioning through connections to the hypothalamus, periaqueductal
gray and other brainstem regions (Ongür, An, & Price, 1998), and is connected to
the nucleus accumbens (Haber, Kunishio, Mizobuchi, & Lynd-Balta, 1995). The
region is also densely connected to other parts of the prefrontal cortex (Barbas &
Pandya, 1989; Carmichael & Price, 1995c; Cavada et al., 2000). In summary, the
VMPFC is organized to receive multimodal information about an organism’s
environment, as well as information about the organism’s internal motivational
and emotional states.

Human lesion studies and non-human primate studies have elucidated many
functions of the VMPFC. In humans, lesions to the VMPFC cause cognitive and
emotional deficits, while leaving overall intelligence and executive functioning
intact (Zald & Kim, 2001). One of the defining features of VMPFC lesions, which
led to much of the animal work on this topic, was the disability in response
reversal. In a response reversal task, an animal will learn that one of two objects
is associated with a reward, only to have that association changed in the middle
of the task such that the previously unrewarded object becomes the rewarded
object. Healthy animals will learn the new contingencies over a few trials,
whereas animals with lesions to VMPFC (which is referred to as
orbitofrontal/ventromedial prefrontal cortex in the animal literature) cannot adjust

12

to the new contingencies, and repeatedly choose the previously rewarded item
(Mishkin, 1964). Non-human primate work also helped in establishing the role of
VMPFC in assigning and updating values related to stimuli (both positive and
negative; Baxter, Parker, Lindner, Izquierdo, & Murray, 2000; Wallis & Miller,
2003; Walton, Behrens, Buckley, Rudebeck, & Rushworth, 2010; Morrison &
Salzman, 2009; Salzman, Paton, Belova, & Morrison, 2007). Evidence that
VMPFC and amygdala work closely together in establishing, sharing, and
updating reinforcement values suggests that, in conjunction with other brain
regions, VMPFC can modulate goal-directed (and emotion-related) behavior
toward maintaining homeostatic balance (Schoenbaum & Roesch, 2005;
Schoenbaum, Setlow, & Ramus, 2003; Schoenbaum, Setlow, Saddoris, &
Gallagher, 2003; Wallis & Miller, 2003).

1.2.2.1 ANTERIOR CINGULATE CORTEX
The anterior cingulate cortex is a large neural region, with many functional and
structural subdivisions (Huster, Westerhausen, Kreuder, Schweiger, & Wittling,
2007; Ongür, Ferry, & Price, 2003; Paus et al., 1996). For the purposes of the
current work, and psychopathy work more generally, the focus will be on the pre-
and sub-callosal portion of the anterior cingulate, which has a superior boundary
of a line drawn from the genu of the corpus callosum to the foremost part of the
frontal pole (see Figure 1-2). The ACC has inputs from, and outputs to, many
neural regions, including VMPFC, lateral prefrontal cortex, anterior insular cortex,

13

amygdala, inferior parietal lobe, thalamus, among other regions (Fujii, 1983;
Pandya, Van Hoesen, & Mesulam, 1981).

Figure 1-2. Location of anterior cingulate cortex (ACC). ACC (purple) is
shown within the region of interest for ventromedial prefrontal cortex (red).

Lesions of the anterior cingulate cortex (ACC) result in reduced motivation,
blunted affect, social disinhibition, disrupted decision-making and increased
irritability (Hadland, Rushworth, Gaffan, & Passingham, 2003; Kennerley, Walton,
Behrens, Buckley, & Rushworth, 2006; Tow & Whitty, 1953). It has been
implicated in error detection, performance monitoring, cognitive control, reward,
punishment, and empathy (for reviews see (Devinsky, Morrell, & Vogt, 1995;
Etkin, Egner, & Kalisch, 2011; Rushworth & Behrens, 2008; Shackman et al.,
2011).
Anterior Cingulate Cortex (ACC)
Anterior Cingulate Cortex (ACC)

14

1.2.3. VENTROMEDIAL PREFRONTAL CORTEX AND PSYCHOPATHY
1.2.3.1. VENTROMEDIAL PREFRONTAL CORTEX FUNCTION
In addition to neuropsychological evidence, neuroimaging data suggests VMPFC
(and other prefrontal areas) dysfunction in individuals with psychopathy. Studies
that have examined VMPFC function during task performances show a wide
range of diversity. In criminal populations, decreased VMPFC function has been
observed during emotional memory tasks (Kiehl et al., 2001), viewing emotional
pictures (Müller et al., 2003), social aggression tasks (Veit et al., 2010), target
detection tasks (Juárez, Kiehl, & Calhoun, 2012), moral judgment (Pujol et al.,
2012), processing emotional facial expressions (Contreras-Rodríguez et al.,
2013), and fear conditioning (Birbaumer et al., 2005). Increased activation has
been observed in the VMPFC when individuals with psychopathy watch a
conspecific being punished (Veit et al., 2010).

In noncriminal populations, abnormal VMPFC functioning has been observed in
subjects with psychopathic traits during situations of social cooperation (Rilling et
al., 2007), moral judgment (Marsh & Cardinale, 2012), cognitive control (Sadeh
et al., 2013), incentive motivation (Bjork, Chen, & Hommer, 2012), and tasks
involving prosodic speech (Sheng, Gheytanchi, & Aziz-Zadeh, 2010). Of
particular note, Sadeh and colleagues (2010) found a positive correlation
between VMPFC activity during cognitive control tasks and both factors of

15

psychopathy (impulsive-antisocial and interpersonal-affective) (Sadeh et al.,
2013). Additionally, Sommer 2010 found increased activity in VMPFC in relation
to controls during affective theory of mind tasks (Müller et al., 2003; Sommer et
al., 2010). Total PPI scores have been positively correlated with ACC during
instrumental reward anticipation (Bjork et al., 2012).

In summary, abnormal VMPFC functioning has been observed in criminal and
non-criminal individuals as a function of psychopathic traits. It does appear that
studies using criminal populations find more consistent reductions in VMPFC
functioning, however, it is difficult to attribute these differences specifically to
population characteristics when task paradigms and analysis methods vary so
widely across studies.

1.2.4. AMYGDALA
The amygdala is a small almond-shaped structure situated in the medial
temporal lobe (see Figure 1-3). The structure is comprised of multiple nuclei,
each with distinct functions.

16

Figure 1-3. Location of left and right amygdala. Amygdalae (red) are situated
within the medial temporal lobes.

Again, it may be useful to first lay out the structural connections of the amygdala,
before describing its known functions. Like the VMPFC, sensory information
reaches the amygdala from higher association sensory cortices (e.g. visual,
auditory, somatosensory, olfactory), as well as the perirhinal cortex and
parahippocampal gyrus (D.G. Amaral, Price, Pitkanen, & Carmichael, 1992;
McDonald, 1998; Stefanacci & Amaral, 2002). Information from each sensory
modality reaches the amygdala at the level of the lateral nucleus. A multi-modal
representation results from actions in the lateral, basal, accessory basal, and
other nuclei of the amygdala (Pitkänen & Amaral, 1998; Stefanacci & Amaral,
2002). Projections from the amygdala reach both cortical (e.g. sensory cortices,
prefrontal cortex) and subcortical targets (e.g. striatum, hippocampus, perirhinal
cortex, entorhinal)(Davis & Shi, 2000). Projections to subcortical regions, which
Amygdala

17

originate from the central nucleus, are responsible for initiating autonomic
responses (Carmichael & Price, 1995a; Davis & Shi, 2000). Connections to
cortical regions, as well as striatum, stem from the basal, accessory basal, and
lateral nuclei (D. G. Amaral & Dent, 1981; D.G. Amaral et al., 1992; Carmichael &
Price, 1995b). Just as the VMPFC, the amygdala is in a prime location to receive
information from all sensory modalities and to participate in orchestrating
behavioral outputs.

The amygdala is also involved in stimulus-reinforcement associations. It is
probably best known for its required role in Pavlovian conditioning paradigms
(e.g. fear conditioning) (Antoniadis, Winslow, Davis, & Amaral, 2009; Davis &
Shi, 2000; LeDoux, 1998, 2007), but is also involved in appetitive conditioning
(when new rewards are learned and acquire their value) (Everitt, Cardinal,
Parkinson, & Robbins, 2003; Gottfried, O'Doherty, & Dolan, 2002), encoding
valence of stimuli (positive or negative) (Belova, Paton, & Salzman, 2008; Paton,
Belova, Morrison, & Salzman, 2006), encoding emotional intensity or arousal of
stimuli (Belova, Paton, Morrison, & Salzman, 2007), response reversal,
recognizing salient information (e.g. detecting threat) (Davis & Whalen, 2001),
assigning and updating the value of an object (Machado & Bachevalier, 2007;
Murray & Izquierdo, 2007; Málková, Gaffan, & Murray, 1997), and coordinating
physiological and hormonal responses to stimuli that evoke emotions, such as
fear and anxiety (Izquierdo, Suda, & Murray, 2005; Kalin, Shelton, & Davidson,

18

2004; Machado, Kazama, & Bachevalier, 2009). In sum, previous research
indicates that the amygdala is responsible for using information from the
environment to assign rewarding or threatening values.

1.2.5 AMYGDALA-VENTROMEDIAL PREFRONTAL CIRCUIT
The amygdala and VMPFC are densely interconnected (D.G. Amaral et al.,
1992) (see Figure 1-4).

Figure 1-4. Amygdala and ventromedial prefrontal cortex (VMPFC) are
structurally and functionally reciprocally interconnected.

Amygdala-VMPFC connectivity

19

Many of the connections from the VMPFC are received in the basolateral nucleus
of the amygdala (Carmichael & Price, 1995a; Stefanacci & Amaral, 2002).
VMPFC also has connections to intercalated cells in the amygdala, which are
situated between the basolateral and central nuclei of the amygdala. Through
this connection, VMPFC can inhibit output to autonomic and hormonal control
centers from the central nucleus (Quirk, Likhtik, Pelletier, & Paré, 2003). One of
the major white matter tracts connecting VMPFC to amygdala is the uncinate
fasciculus.

In terms of function, both the amygdala and VMPFC are necessary to update the
value of a reinforcer (Baxter et al., 2000; Hampton et al, 2007), and are involved
in processing and regulating emotion (Davidson, 2002; Kalin & Shelton, 2003;
Ochsner, Bunge, Gross, & Gabrieli, 2002; Urry et al., 2006). In functional
imaging studies, the two regions are functionally connected during rest (Roy et
al., 2009), and shown to be anti-correlated during emotion appraisal and
regulation tasks (Delgado, Nearing, Ledoux, & Phelps, 2008; Pezawas et al.,
2005).

1.2.6 AMYGDALA AND PSYCHOPATHY
1.2.6.1 AMYGDALA FUNCTION
Neuroimaging data has provided support for behavior observations in individuals
with psychopathy that suggest amygdala dysfunction.

20

Functionally, reduced activity in the amygdala has been reported in incarcerated
individuals with psychopathy during tasks involving the processing of emotional
stimuli (Kiehl et al., 2001) and fear conditioning (Birbaumer et al., 2005).

There is also evidence from community populations that psychopathic traits and
amygdala function are negatively correlated during tasks such as viewing
unpleasant pictures (Harenski, Kim, & Hamann, 2009), moral judgment (Marsh &
Cardinale, 2012), social cooperation (Rilling et al., 2007), emotional moral
decision-making (Glenn, Raine, & Schug, 2009), and an affect recognition task
(Gordon, Baird, & End, 2004). However, some studies report increased activity in
the amygdala during negative visual content viewing (Müller et al., 2003), fear
conditioning (Schneider et al., 2000), emotional moral decision making (Glenn et
al., 2009), punishment of a conspecific (Veit et al., 2010), fearful facial
expression viewing (Marsh et al., 2008), and cognitive control tasks (Sadeh et al.,
2013).

In summary, there are reports of abnormal amygdala functioning in criminal and
noncriminal populations as a function of psychopathic traits. Similar to studies of
VMPFC function in psychopathy, studies of amygdala in criminal populations
tend to report reductions in activity, whereas in noncriminal populations both
increased and decreased activity has been reported. Again, it is difficult to say

21

whether the differences in findings are due to true population differences, or
differences in experimental design.

While many of the impairments seen in patients with amygdala lesions are similar
to those seen in psychopathy, this is not true for all of them. For example, the
formation of a stimulus-reward association is preserved in psychopaths, as well
as the intactness of certain aspects of social cognition (R. J. Blair, 2008). This
may imply that there are impairments in specific sub-regions of the amygdala or
even impairments in specific neurotransmitter signaling. These could be the
reasons for the mixed findings and it remains unclear how exactly the amygdala
relates to psychopathy, but there does appear to be some evidence suggesting a
possible role.

1.2.7 AMYGDALA-VENTROMEDIAL PREFRONTAL AND PSYCHOPATHY
In addition to task-based fMRI studies, there are resting state fMRI studies, which
evaluate the correlations between different neural regions while the brain is at
‘rest’ (e.g. relaxing and not performing a task). Amygdala-VMPFC functional
connectivity has been assessed during resting state fMRI, but only in criminal
populations with psychopathy. Motzkin and colleagues (2011) found a reduction
in resting state amygdala-VMPFC connectivity in offenders with psychopathy, as
compared to offenders without psychopathy. Assessment of the white matter
tract connecting amygdala and VMPFC (i.e. uncinate fasciculus) in individuals

22

with psychopathy has also been conducted. Using diffusion tensor imaging (an
analysis method used to measure white matter microstructural integrity), two
studies have found reduced microstructural integrity of the uncinate fasciculus in
criminal individuals with psychopathy (Craig et al., 2009; Motzkin et al., 2011).

1.2.8 CHILDREN AND ADOLESCENTS WITH CONDUCT
DISORDER/OPPOSITIONAL DEFIANT DISORDER
The topic of psychopathic traits in children is controversial due to ethical
concerns of premature personality and behavioral labels. However, discussing
neuroimaging findings from this population are worthwhile as they provide some
insight into possible trajectories of neural development in psychopathy. That
said, with regard to assessing antisocial behavior in children, two disorders are
typically discussed: conduct disorder (CD) and oppositional defiant disorder
(ODD) (the criteria of each are listed in Table 1-3). A comparison of the criteria
for these two disorders and psychopathy criteria show strikingly differenences,
with very little “personality trait” overlap, although CD (rightly more so than ODD)
is often mentioned as a precursor to psychopathy.

23

Table 1-3. DSM-IV Criteria for Oppositional Defiant Disorder (ODD) and
Conduct Disorder (CD)
DSM-IV Criteria for Oppositional
Defiant Disorder
DSM-IV Criteria for Conduct Disorder
A pattern of negativistic, hostile and
deviant behavior lasting at least 6
months, during which four (or more) of
the following are present:
1. Often bullies, threatens, or
intimidates others
1. Often loses temper 2. Often initiates physical fights
2. Often argues with adults 3. Has used a weapon
3. Often actively defies or refuses to
comply with adults’ request or rules
4. Has been physically cruel to people
4. Often deliberately annoys people 5. Has been physically cruel to animals
5. Often blames others for his or her
mistakes or misbehavior
6. Has stolen while confronting a victim
6. Is often touchy or easily annoyed by
others
7. Has forced someone into sexual
activity
7. Is often angry and resentful 8. Has deliberately engaged in fire
setting
8. Is often spiteful or vindictive 9. Has deliberately destroyed others’
property
10. Has broken into someone else’s
house, building or car
11. Often lies to con others
12. Has stolen items of nontrivial value
without confronting the victim
13. Often out late without permission,
starting before age 13
14. Has run away from home overnight
at least twice
15. Often truant from school, starting
before age 13

In any case, all studies of the neural correlates of antisocial behavior in children
have used CD and/or ODD populations. Recently, some investigators have
suggested that in order to make a true comparison to psychopathy (and not, e.g.,

24

antisocial personality disorder), an assessment of callous-unemotional traits (or
psychopathic traits, more generally) must be included.

In terms of microstructural integrity of the uncinate fasciculus in adolescent
populations, there appear to be no studies that account for CD/ODD with
psychopathic traits. Two studies do show increased microstructural integrity of
UF in CD/ODD populations when compared to healthy controls (Passamonti et
al., 2012; Sarkar et al., 2013).

Studies of functional data that examine ODD/CD populations with psychopathic
traits compared to healthy controls reveal reduced amygdala activation when
subjects watch others in pain (Marsh et al., 2013), fearful facial expressions
(Jones, Laurens, Herba, Barker, & Viding, 2009; White et al., 2012), and make
moral judgments (Marsh, Finger, Fowler, Jurkowitz et al., 2011). Reduced
VMPFC activity has been observed during a passive avoidance task (Finger et
al., 2011) and response reversal (Finger et al., 2008). Additionally, rostral ACC
activity while watching others in pain was negatively correlated with psychopathic
traits (Marsh et al., 2013).

There appears to be more evidence for reduced amygdala and VMPFC activity in
children with CD/ODD plus psychopathic traits, than in adults with psychopathy.
In adults, the results are more variable. This raises the question of whether

25

development plays a role in the neurobiological underpinnings of psychopathic
traits, or if the differences in results are attributable to the personality disorder
criteria or differing experimental methods used.

1.3 SUBTYPES OF PSYCHOPATHY BASED ON TRAIT ANXIETY
Usually, psychopathy is viewed as a unitary construct, but theoretical and
empirical research suggest that there are variants of the disorder, specifically
primary and secondary subtypes. Cleckley was the first to suggest that a lack of
anxiety was core to the conceptualization of psychopathy, and could distinguish
individuals truly afflicted with the disorder from those with a similar, but different
condition. He listed “absence of ‘nervousness’ as one of the criteria on his
checklist. Benjamin Karpman’s classic theory (1947) also proposed, based on
clinical observations, the idea of subtypes of the disorder. He suggested that
primary psychopathy is the result of a genetic abnormality that results in the
affective deficits observed in these individuals, whereas secondary psychopathy
is the product of environmental influences that result in the affective disturbances
seen in secondary psychopathy (KARPMAN, 1947). Since Cleckley’s and
Karpman’s initial proposals, there has been much theoretical and empirical work
to expand on the notion of two groups that are phenotypically similar, yet
etiologically distinct (Karpman, 1947; Lykken, 1957; J. P. Newman, MacCoon,
Vaughn, & Sadeh, 2005). It has been suggested that an adequate way to
describe their condition is that they are in “lack of a conscience”, possibly due to

26

the fact they lack the biological foundation to support one (Hare, 1991).
Individuals with primary psychopathy are extraverted, confident, dominant, low in
negative affect, and low to average in anxiety (Blackburn, 1998). Individuals with
secondary psychopathy, on the other hand, are believed to have an affective
disturbance that is acquired from aversive conditions in their environment, e.g.
parental abuse (Karpman 1947). It is believed that individuals with secondary
psychopathy do possess the hardware, so to speak, to support a conscience, but
that their social learning environment has shaped them in such a way that it is
kept from functioning properly (Vidal, Skeem, & Camp, 2010). This second
subtype is emotionally disturbed, socially anxious, withdrawn, moody, more
submissive, high in negative affect and low in self-esteem (Blackburn, 1998;
Blackburn & Lee-Evans, 1985). In the setting of research, after being classified
as psychopathic by the classic criteria, the two groups are usually distinguished
by trait anxiety, with primary psychopaths being characterized by low anxiety,
and secondary psychopaths being characterized by normal, or high anxiety (J.
Skeem et al., 2007).

Despite these observed behavioral differences, the most commonly used
psychopathy scale, the PCL-R, does not directly assess trait anxiety. Although
much research has been conducted on this disorder, the exact functional and
structural neural abnormalities that contribute to the behavioral profile remain
unknown. It seems possible that part of the heterogeneity found in psychopathy

27

may be explained by accounting for trait anxiety when considering variants of the
disorder (J. Skeem et al., 2007).

1.3.1 BEHAVIORAL AND PHYSIOLOGICAL EVIDENCE
Research from various domains provides a foundation for the theories listed
above. Multiple studies have identified inmates with psychopathy whose
behavioral profiles reflect personality traits of primary and secondary
psychopaths. Typically, model-based clustering statistical methods are used in
such grouping analyses, which assume a dataset (e.g. a population) is
comprised of multiple subgroups, and aims to identify the subgroups based on
defined characteristics (e.g. anxiety, psychopathic traits, etc) (Stahl & Sallis,
2012). For example, Blackburn (1975) discovered 4 offender profile types in a
sample of British male offenders, of which two were identified as primary
psychopaths (extroverted, but not neurotic) and secondary psychopaths
(neurotic, but not extroverted) (Blackburn, 1975). Many studies have replicated
these results, identifying two groups of offenders with psychopathy within
incarcerated populations that display either low anxiety and dominant features
(e.g. bold, competitive, arrogant), or high anxiety and negative emotionality
(Henderson, 1983; J. Skeem et al., 2007; Swogger, Walsh, & Kosson, 2008;
Vassileva, Kosson, Abramowitz, & Conrod, 2005). This pattern of subtypes has
also been found in incarcerated youth (E.R. Kimonis, Skeem, Cauffman, &
Dmitrieva, 2011; Vaughn, Edens, Howard, & Smith, 2009) and in non-

28

incarcerated samples (Christian, Frick, Hill, & Tyler, 1997; Falkenbach,
Poythress, & Creevey, 2008). In addition to differences in personality traits, the
two subtypes also differ in their criminal behavior. Researchers have found that
low-anxious criminals with psychopathy committed more crimes (violent and
nonviolent), had higher levels of criminal versatility (Brinkley, Newman, Widiger,
& Lynam, 2004), and had more frequent and severe antisocial behavior (Fagan &
Lira, 1980).

In laboratory settings, empirical evidence suggests that low-anxious individuals
with psychopathy display abnormalities in emotional processing and behavior
that high-anxious individuals with psychopathy do not. For instance, as a group,
individuals with psychopathy are notorious for failing to correct behaviors that
have led to punishment in prior instances, a behavior that is known as passive
avoidance learning (Cleckley, 1982; Lykken, 1957; J. P. Newman & Kosson,
1986; Schmauk, 1970). When a group of offenders with psychopathy is parsed
by trait anxiety, it is low-anxious offenders with psychopathy that display deficits
in passive avoidance learning at significantly greater rates (Arnett, Smith, &
Newman, 1997; J.P. Newman, Patterson, Howland, & Nichols, 1990).

Differences between the two groups also emerge during economic decision-
making tasks. On two tasks of social economic decision-making, offenders with
psychopathy who were also low-anxious, displayed significantly less pro-social

29

behavior (e.g. amounts of money offers; rejected money offers) towards their
partners, than offenders with psychopathy who were high-anxious, as well as
control individuals (Koenigs et al., 2010). In studies of moral decision-making,
participants are often asked to judge transgressions that are either ‘moral’ (i.e.
acts that violate the welfare of others), or ‘conventional’ (i.e. do not directly affect
the welfare of others) in character. Further, transgressions may be either
‘personal’, which involves committing direct physical harm to another (e.g.
pushing a person off a bridge to stop a runaway train car from hitting five people)
or ‘impersonal’, which involves more indirect harm (e.g. pulling a switch to divert
a runaway boxcar from hitting five people). Personal actions tend to be rated as
significantly more emotionally unpleasant than impersonal ones (Koenigs et al.,
2007). In one study, as expected, since individuals with psychopathy tend to
rank high on amoral behavior, both low- and high-anxious offenders with
psychopathy were more likely to approve of the impersonal harm in moral
dilemmas, than the controls without psychopathy. However, low-anxious
offenders with psychopathy were also more likely to approve of personal harm in
moral dilemmas if it was a means to a particular self-serving end (Koenigs,
Kruepke, Zeier, & Newman, 2011).

In addition to these behavioral differences, low- and high-anxious offenders with
psychopathy have been observed to differ in physiological responses. For
example, the low-anxious subtype has been shown to display reduced autonomic

30

responses to emotional and neutral stimuli (Hiatt, Lorenz, & Newman, 2002;
Lorenz & Newman, 2002; J. P. Newman, Schmitt, & Voss, 1997), as well as
reduced fear-potentiated startle responses (Sutton, Vitale, & Newman, 2002).
Additionally, psychopathy is frequently associated with impaired fear conditioning
and it has been demonstrated that the low-anxious subtype displays significantly
reduced autonomic responding in anticipation of an aversive stimulus (Lykken,
1957).

In studies using juvenile offenders differences have been found between low-
and high-anxious individuals with high levels of psychopathic traits. One study
found that juvenile offenders with high psychopathic traits and low trait anxiety
were significantly less engaged by emotionally distressing pictures (E. R.
Kimonis, Frick, Cauffman, Goldweber, & Skeem, 2012), and another found that
they performed more poorly on passive avoidance tasks (Vitale et al., 2005).

Research on subtypes of psychopathy in non-incarcerated subjects are fewer,
but nonetheless informative, given that the studies of this dissertation use non-
incarcerated populations. Within an undergraduate sample, primary
psychopathic traits were significantly positively correlated with positive affect and
negatively correlated with negative affect, while secondary psychopathic traits
were significantly positively correlated with negative affect (Del Gaizo &
Falkenbach, 2008), in agreement with published results. Similar findings were

31

reported in another study within undergraduate males; high-anxious individuals
with high levels of psychopathic traits had significantly lower emotional
intelligence than low-anxious individuals with psychopathic traits (who had intact
emotional intelligence), particularly in domains related to managing emotions,
facilitating thoughts, and violence (Vidal et al., 2010). Another recent study found
that undergraduates who were low on anxiety and high on psychopathy were
significantly more likely to punish a defector in an altruistic punishment game,
where players share points with each other, or punish players both at costly and
non-costly conditions (Masui, Iriguchi, Nomura, & Ura, 2011).

Taken together, these findings suggest that many deficits typically ascribed to
groups with psychopathy are, in fact, specific to a low-anxious subtype.
Individuals with low trait anxiety and high psychopathic traits possesses greater
deficits in emotional learning that are expressed as abnormal moral decision-
making and pro-social behavior, as well as being marked by altered autonomic
responding. Moreover, these differences in primary and secondary traits and
behaviors can be seen in juvenile offenders, as well as non-incarcerated
populations. One area of research that stands to be explored is whether the
autonomic differences between low- and high- anxiety individuals with high
psychopathic traits extends to both incarcerated and non-incarcerated
populations. It is possible, as Robert Hare once suggested, that autonomic

32

dysfunction may be unique to “pure” primary psychopathy (R. D. Hare &
Thorvaldson, 1970).

1.3.2. SUBTYPES AND THE BRAIN
Until recently, although behavioral similarities had been noted between brain
damaged patients and individuals with psychopathy, no direct comparisons
between the groups had been conducted. Further, despite the fact that the
theory about subtypes of psychopathy posited foundational neural abnormalities
for primary, but not secondary psychopaths, no empirical research has
addressed the issue. These still open questions have just begun to be
addressed. In a recent study, it was shown that individuals with primary
psychopathy (low-anxious) were quantitatively similar to VMPFC lesion patients
in their response patterns on two decision-making tasks (Koenigs et al., 2010),
when compared to individuals with secondary psychopathy (high-anxious) and
controls. Individuals with psychopathy and low trait anxiety, and VMPFC lesion
patients shared a deficit in basic processes of self-insight and self-reflection that
were displayed as exaggerated anger and irritability, as well as diminished
empathy and guilt. This was the first study to directly compare VMPFC lesion
patients and criminals with psychopathy. Another recent study, was the first
neuroimaging study to directly compare the neural properties of low-anxious and
high-anxious offenders with psychopathy (Motzkin et al., 2011). This study found
no differences in microstructural integrity of the uncinate fasciculus between low-

33

and high- anxious individuals with psychopathy, although the sample size was
modest. They did not find significant differences in amygdala-VMPFC resting
state functional connectivity between the subtypes, but did find an interaction
between psychopathy and anxiety, such that higher anxiety was associated with
greater functional connectivity in criminals without psychopathy, but lower in
criminals with psychopathy.
In summary, while neural data to support the distinction between low-anxious
and high-anxious individuals with psychopathy is sparse, some headway has
been made. Most interesting, the neural regions often implicated in psychopathy
(e.g. amygdala, VMPFC) also underlie trait anxiety (Kalin et al., 2004; Kalin,
Shelton, & Davidson, 2007; Koenigs et al., 2008), as well as autonomic
responding (Critchley, Elliott, Mathias, & Dolan, 2000; LaBar, LeDoux, Spencer,
& Phelps, 1995). It is possible that general dysfunction in the amygdala-VMPFC
circuit may explain the differences observed in all these domains in low-anxious
individuals with psychopathy. It is also likely that dysfunction in this neural circuit
will occur along a continuum. While it has previously been proposed that
psychopathic traits are spread over a continuum (Benning et al, 2003), it is
plausible to propose that this continuum is comprised of interactions between
psychopathic traits and trait anxiety. In this latter scenario, individuals with high
psychopathic traits and low anxiety will display the greatest neural abnormalities,
which manifest themselves in significantly different behaviors. More research is
needed to investigate whether the divergent continuum is only evident in extreme

34

cases of psychopathy (e.g. high scoring incarcerated samples), or whether the
relationship between psychopathy, trait anxiety, and neural properties extends
into community samples.

35

CHAPTER 2: Trait anxiety interacts with psychopathic
traits to predict amygdala BOLD activity during fear
conditioning

2.1 ABSTRACT
Impaired fear conditioning has long been associated with psychopathy, a
personality disorder of interpersonal deficits and antisocial behavior. Classic
studies revealed that criminals with psychopathy display reduced autonomic
responses to fear conditioning (Hare & Quinn, 1971; Hare & Thorvaldson, 1970),
although when subtypes are distinguished using trait anxiety, low-anxious
individuals with psychopathy exhibit the greatest fear conditioning impairments
(Lykken, 1957). Neuroimaging of fear conditioning in incarcerated criminals with
psychopathy suggests that neural regions involved in fear conditioning (e.g.
amygdala, ventromedial prefrontal cortex (VMPFC)) exhibit significantly reduced
BOLD responses when compared to healthy controls (Birbaumer et al., 2005).
However, whether neural activity during fear conditioning is modulated by
psychopathic traits in a non-incarcerated population has yet to be examined.
Furthermore, the relationship between trait anxiety, psychopathic traits, and
neural activity during fear conditioning has not yet been addressed. In the
current study, in a community sample we explore the relationship between
psychopathic traits and fear conditioning-related BOLD activity in amygdala and
VMPFC, as well as whether the relationship is conditional upon levels of trait
anxiety. The findings indicate that trait anxiety moderates the negative
association between psychopathic traits (Machiavellian Egocentricity, ME) and

36

BOLD activity in the right amygdala, such that increases in ME coupled with
decreases in trait anxiety result in decreased right amygdala activity, similar to
previous findings in autonomic responding to fear conditioning. The results
support previously observed negative associations between psychopathic traits
and amygdala activity during fear conditioning, but reveal that the relationship
can be moderated by trait anxiety. Similar relationships were observed between
trait anxiety, psychopathic traits and autonomic responses during fear
conditioning. The results support the notion of altered neural and physiological
functioning in individuals with higher levels of psychopathic traits, but highlight
the important of trait anxiety.

2.2 INTRODUCTION
Neuroimaging studies of fear conditioning in criminals with psychopathy, which
do not account for subtypes, have revealed significantly reduced activation of
multiple regions involved in fear conditioning, including the amygdala-VMPFC
circuit (Veit et al., 2002). However, the modulation of BOLD activity in fear
conditioning related regions in a non-incarcerated sample has not yet been
investigated. It is important to study psychopathic personality traits in community
samples as these traits are also present in the general population (e.g.
successful psychopaths; (Levenson, Kiehl, & Fitzpatrick, 1995; Widom, 1977;
Yang et al., 2005). Additionally, problems arise in making valid comparisons
between criminal and non-criminal populations, as criminal populations are

37

qualitatively different from non-criminal populations in important ways (e.g. risk
aversion;(Raine, 1993). Furthermore, the question of whether trait anxiety can
moderate the relationship between psychopathic traits and BOLD activity during
fear conditioning in the amygdala-VMPFC circuit has not been studied.

In the current study, fMRI is used to assess BOLD activity in the amygdala and
VMPFC during fear conditioning, using a young adult, community sample with a
wide distribution of psychopathic traits. No direction is predicted for the
relationship between psychopathic traits and BOLD activity in the amygdala and
VMPFC during fear conditioning, as the relationship is expected to be moderated
by trait anxiety, and thus, a result in one direction or the other would be solely a
product of the given sample. Further, a multiple regression analysis is used to
assess whether trait anxiety can moderate the relationship between psychopathic
traits (those unrelated to trait anxiety) and BOLD activity during fear conditioning
in such a way that increased psychopathic traits, coupled with decreased trait
anxiety, would lead to a reduction in BOLD activity in the amygdala and VMPFC
during fear conditioning.

2.3 MATERIALS AND METHODS
Participants
Twenty-four healthy participants (24 male, 18 to 21 years of age, mean ± SD =
18.9 ± 0.7), were recruited from a larger study at the University of Southern

38

California, the Social and Moral Development of Twins (Baker, Barton, & Raine,
2002). The goal of the twin study is to investigate the environmental and genetic
contributions to the development of aggressive and antisocial behavior. The Twin
Study began assessment of twins from the general Los Angeles area in 2001,
and has invited each twin pair to return every 2-3 years. The study is currently in
Wave 5, with subjects between 19-20 years old. The current investigation is a
supplement to this larger study. This population was chosen because of
available longitudinal psychopathy scores, allowing for participant selection
based on previously acquired psychopathy scores. This was crucial due to the
fact that recruiting participants from the general population scoring high on
psychopathy measures can be extremely difficult and time consuming.
Participants were invited in as singletons (i.e. one random twin of the pair) in
order to avoid the statistical dependency issue that exists between twins.
Invitations to participate in the study were mailed to potential participants, and a
follow-up call was made two weeks following the mailing (see Appendix I for
recruitment mailing information). Participants were told that the study was
investigating the relationship between personality traits and brain function and
structure. In order to ensure a wide range of scores for psychopathic traits in the
analysis, we recruited half of the participants (12 participants) from the bottom
quartile (0-25
th
percentile) and half (12 participants) from the top quartile (75
th
-
100
th
percentile) of a composite psychopathy score. The composite score was
computed from the combination of the standardized scores of the following three

39

psychopathy measures from the Twin Study’s Wave 3 data (all based on self-
report by the participants at age 14-16): Childhood Psychopathy Scale (youth
report; (Lynam, 1997), Antisocial Process Screening Device (APSD, youth report;
(Frick & Hare, 2001), and PCL-R youth version (PCL:YV; (Forth, Hart, & Hare,
1990). High significant correlations between the three psychopathy measures
provided the rationale for making the composite score (See Table 2-1 for
correlations).
Table 2-1. Correlation Matrix for youth psychopathy measures used in the
composite score
CPS-
youth
APSD -
Youth PCL-YV
CPS – Youth
-
APSD – Youth
.64* -
PCL – YV
.45* .49* -
Note. CPS-youth – Child Psychopathy Scale, Youth Report; APSD-youth = Anti-
Social Process Screening Device, Youth Report; PCL-YV = Hare Psychopathy
Checklist Revised, Youth Version. * indicates significance at p < 0.01.

Although the psychopathy measures were from an earlier time frame, they were
considered a reliable representative of psychopathic traits in adulthood, as these
traits are known to emerge early in life and remain stable through adulthood
(Hare & Neumann, 2008). The distribution of the composite score for males in
the entire initial twin sample can be seen in Figure 2-1, with the final range for the
current study of the 0-25
th
percentile (cut-off: z = -0.62; range: z = -0.62 to -1.51,
mean ± SD = -1.02 ± .28) and the 75
th
to 100
th
percentile (cut-off: z = 0.47; range:

40

z = 0.53 to 1.63, mean ± SD = .91 ± .35) marked by red circles. The final
distribution of composite psychopathy scores for participants of the MRI study
can be seen in Figure 2-2.

Figure 2-1. Distribution of the composite psychopathy score in entire male
twin sample. n = 511. This sample includes the initial potential male participant
pool (including twin siblings) for illustration of score distributions. The red circles
indicate the range of composite psychopathy scores in the recruited participants
for the current MRI study.

41

Figure 2-2. Distribution of the composite psychopathy score within the MRI
participant pool. n = 24.

Mean composite scores for the high quartile were significantly higher than means
for the low quartile, t(23) = -14.93, p < 0.001. All participants had at least two of
the psychopathy measures completed (14 participants had all 3 measures
completed). The missing scores came solely from the PCL-YV. A chi-square
test determined that individuals who had only completed two psychopathy
measures were more likely to end up in the lower quartile, Χ
2
(2, N = 24) = 9.88,
p = 0.002. Therefore, in order to ensure group differences in total composite
psychopathy scores were not being driven by the PCL-YV scores, it was tested
whether differences in overall psychopathy scores remained significant after
excluding PCL-YV from the analysis. A t-test between the two quartiles’ new
0

1

2

3

4

5

6

1.5
-‐1.3
-‐1.1
-‐0.9
-‐0.7
-‐0.5
0
0.5
0.7
0.9
1.1
1.3
1.5
1.7

Frequency

Composite
Psychopathy
Score

42

composite scores, comprised of only APSD and CPS, revealed that scores of the
low quartile remained significantly lower than those in the high quartile, t(23) = -
6.56, p < 0.001. The two quartiles are race-matched, and there are no significant
differences of age between quartiles, t(23) = -0.74, p > 0.05). All ethnicities were
included in recruitment, except for African-Americans, as it has been reliably
shown that this ethnic group exhibits lower basal skin conductance levels (higher
skin resistance; (Johnson & Corah, 1963; Korol & Kane, 1978), and could thus
make determination of fear conditioning (from skin conductance responses)
difficult in these participants. Additionally, only males were invited for
participation, as a majority of studies have found psychopathy to be more
prevalent in males than in females (Forth et al., 1990; Hamburger, Lilienfeld, &
Hogben, 1996). Age and composite psychopathy scores of both recruited
quartiles are comparable to those of the two quartiles in the potential participant
pool, representing the community sample (see Table 2-2).

43

Table 2-2. Age, composite psychopathy score, and race information for
potential participant pool and actual participant pool

Potential Participants Actual Participants
Quartile 1 Quartile 4 Quartile 1 Quartile 4
n 90 91 12 12

age mean 18.6 18.9 18.9 19.1
Std. dev. 0.63 0.71 0.67 0.9
range 16-21 16-21 18-20 18-21
Comp
score

mean -1.03 1.11 -1.02 0.91
Std. dev. 0.29 0.55 0.28 0.35
range -0.65 to -1.99 0.48 to 3.22 -0.62 to -1.51 0.53 to
1.63
Race
Caucasian 42.7% 28% 25% 25%
Hispanic 20% 41.6% 66.6% 66.6%
African-
American
7.3% 11.2% - -
Asian 4.5% 2.4% 8.3% 8.3%
Mixed 25.5% 16.8% - -
Note: Comp = Composite.

All participants were right-handed, had normal or corrected-to-normal vision, and
had no neurological or psychiatric history. Participants were compensated for
their time. Written informed consent was obtained from all participants before
inclusion in the study. The study was approved by the Institutional Review Board
of the University of Southern California.

44

Although this study uses only one twin of a twin pair, information about the co-
twins and zygosity can be found in Appendix II.

Behavioral procedure
The composite psychopathy score was used as assessment of psychopathic
traits. In order to determine the stability of psychopathic traits over time, the
Psychopathic Personality Inventory, Revised (Lilienfeld & Andrews, 1996), a 154-
item, adult self-report psychopathy measure, was administered to the participants
when they agreed to participation. The PPI was designed to be used in the
general population, and to focus on the theorized core personality traits of
psychopathy, rather than antisocial behavior. Responses to each item are on a
4-point Likert scale (True, Mostly True, Mostly False, False). The PPI is
comprised of eight subscales: Machiavellian Egocentricity (ME), Rebellious
Nonconformity (RN), Blame Externalization (BE), Carefree Nonplanfulness (CN),
Social Influence (SOI), Fearlessness (F), Stress Immunity (STI), and
Coldheartedness (C). Benning and colleagues (2003) used Exploratory Factor
Analysis on the original PPI subscales and found that seven of the eight
subscales loaded onto two factors, PPI-FD (Fearless Dominance, also called
PPI-I; containing: STI, SOI, F), and PPI-SCI (Self-Centered Impulsivity, also
called PPI-II; containing: ME, RN, BE, CN). The last subscale, Coldheartedness
(PPI-C), did not load onto either factor. Although many studies use the two
higher order factors, there has been some concern that the PPI-I does not

45

properly capture the core features of psychopathy (e.g. meanness or antisocial
behavior; Lynam & Miller, 2012). Additionally, although there has been
substantial evidence that PPI-II maps onto Hare’s Psychopathy Checklist
Revised (PCL-R, the gold standard for assessing psychopathy), Factor 2 (the
antisocial factor) (Benning, Patrick, Hicks, Blonigen, & Krueger, 2003; Patrick,
Edens, Poythress, Lilienfeld, & Benning, 2006), there is much less evidence that
PPI-I maps onto the PCL-R’s Factor 1 (interpersonal/affective factor)(Hughes,
Stout, & Dolan, 2013; Malterer, Lilienfeld, Neumann, & Newman, 2010;
Zolondek, Lilienfeld, Patrick, & Fowler, 2006). Due to the above reasons, as well
as due to the fact that PPI-I tends to be negatively correlated with trait anxiety
(Edens & McDermott, 2010) and cannot be used as a co-dependent variable in a
multiple regression analysis, analyses of the current study will focus on the total
PPI score and the individual components of the PPI-II scale for multiple
regression analyses.
Psychopathy studies that include trait anxiety measures to define subtypes
typically use the MMPI-derived Welsh Anxiety Scale (WAS; Welsh, 1956)
(Kosson, Smith, & Newman, 1990; J. P. Newman & Kosson, 1986; J.P. Newman,
Patterson, Howland, & Nichols, 1990) to assess trait anxiety (see Appendix II for
comparison with other trait anxiety measures). The WAS is a 39-item, self-report
scale that is derived from the MMPI (Hathaway, Engel, & MacKinley, 2000) (see
Appendix III for more information on WAS). Statistical analysis of all
questionnaire data was performed using Statistical Product and Service

46

Solutions (SPSS; version 19, SPSS, inc., Chicago).
Stimuli
Two geometrical shapes (circle/square, green/blue) were presented as
conditioned stimuli (CS). The CS+ (28 trials) co-terminated with a 200-ms
aversive event, electrocutaneous stimulation with 100% pairing. The CS- (28
trials) was never paired with the aversive stimulus. The electrocutaneous
stimulation was presented through two electrodes placed on the participant's left
shin. Stimulation was achieved with use of a Grass Instruments Stimulator
delivering a DC pulse through the electrodes on the shin. Before the brain scan
session, the participants themselves set the voltage level of the electrocutaneous
stimulation (mean ± SD = 33.3V ± 0.79). Participants were told that stimulation of
the set intensity will be delivered during the experiment, and that the intensity
should be at a level that the participant finds “annoying, but not painful”. The
shin electrodes were attached to wires connected through a wall to a computer in
the scanner control room during the brain scan. The CS stimuli were presented in
a pseudorandom order, such that no more than 2 trials of the same stimulus
were presented in a row. The CS duration was 6 seconds, followed by a jittered
8-12 sec (average 10-sec) rest condition (black screen with a crosshair fixation).
There were a total of 56 trials: 28 trials of CS+ and 28 trials of CS-, divided over
two functional runs. Participants were given four habituation trials (two CS+ and
two CS-) in a pseudorandom order before the brain scan in order to habituate
orienting responses to the CSs. The fear conditioning task was programmed in

47

Matlab using PsychToolbox (Brainard, 1997). Acquisition of conditioning
occurred within the scanner. Participants were instructed that the task was
passive, and that we were searching for differences in brain activity responses
that are explained by differences in personality. They were told before the brain
scan which of the geometric figures would be paired with the electrocutaneous
stimulation. Participant awareness of the reinforcement contingencies, as well as
valence and arousal of the CSs (Likert scales of 1-4), was assessed post-scan
outside the magnet. Figure 3 summarizes the conditioning parameters used in
the present study.

Figure 2-3. Schematic illustration of the experimental design. Two different trials
are illustrated. One geometric figure (CS+) was paired with the unconditioned
stimulus (US) and the other geometric figure (CS-) was never followed by the
unconditioned stimulus. The US co-terminated with the CS+ with a 200-msec
duration. Each CS was presented for 6-sec and the rest periods (white fixation
cross on black background) were jittered with length durations between 8-12-sec
(10-sec average).

CS+$
(6$sec)$
Rest$$
(8.10$sec)$
CS.$
(6$sec)$
Rest$$
(8.10$sec)$

48

SCR data acquisition and analysis
Skin conductance was recorded during the classical fear conditioning task within
the scanner to ensure that fear conditioning was successful. Skin conductance
responses (SCR) were recorded by MRI-safe pre-gelled (isotonic electrode gel)
silver-silver chloride electrodes (Biopac model EL508) attached by adhesive
disks to the palmar surface of the second and third digits of the non-dominant
hand (BIOPAC Systems, Santa Barbara, CA). Participants were told that the
electrodes attached to their fingers would only be used to record their SCR
during the experiment, and not for stimulation of any kind. Data was amplified
and acquired with BioPac instrumentation and hand scored with the
accompanying AcqKnowledge software (BIOPAC Systems, Santa Barbara, CA),
by a rater who was blind to psychopathy score. The SCR signal was down-
sampled to 15 samples/second, and a low pass filter fixed at 1 HZ was used in
order to remove high frequency noise from the data. Research on SCRs to fear
conditioning has revealed three distinct phases of responding: orienting response
(OR), anticipatory response (AR), and unconditioned response (UCR) (Boucsein,
1992). In this experiment, the timeframes for these phases are as follows: OR
(1-3s), AR (3-6.5s), and UCR (6.5-9.5s). The OR is often interpreted as an
orienting response to CS presentation, whereas the AR is generally considered
an emotional response, elicited by UCS anticipation, that reflects learning the
CS–UCS association (Boucsein, 1992; Prokasy, Williams, Kumpfer, Lee, &
Jenson, 1973; Wolter & Lachnit, 1993). The level of SCR response was

49

assessed for each phase and for each trial as the base to peak amplitude
difference in skin conductance of the largest deflection in the 0.5-6.5 sec latency
window after onset of the CS (LaBar et al., 1998). A minimum response criterion
of 0.02 µS was used with lower responses scored as 0. Raw scores were
square-root transformed prior to statistical analysis to normalize the distributions
(LaBar et al., 1998). Acquired SCRs during the CS phases were averaged per
subject and per type of trial (CS+, CS-). The first trial of each stimulus type was
discarded due to the fact that no learning had yet occurred at this stage. A two-
way repeated-measures analysis of variance (ANOVA) was used to analyze the
SCR data with the SCR phases (OR, AR, UCR) and condition type (CS+, CS-) as
within-group factors. Follow-up two-tailed t-tests were used to directly compare
the SCR phase responses. Additionally, a difference score was derived as a
measure of differential conditioning by subtracting CS- responses from CS+
responses (LaBar et al., 1995), and was used in analyses with brain data.
According to this measure, difference scores greater than zero reflect greater
relative conditioning to the CS+, difference scores equal to zero reflect no
differential conditioning, and difference scores less than zero reflect greater
relative conditioning to the CS-. SCR data was available for 21 out of 24
participants (3 subjects had missing data due to technical difficulties). Subject
awareness of the reinforcement contingencies was assessed immediately
following the scanner session in the control room outside the magnet, by asking
whether the stimulation followed just one stimulus, both, or neither; they were

50

also asked to rate their certainty of this response. Participants were also asked
to rate the arousal and pleasantness of each stimulus type using a Likert scale
from 1 = “not at all arousing/pleasant” to 4 = “very arousing/pleasant”.

Previous research has shown a positive correlation between amygdala BOLD
activity during acquisition of fear conditioning and measures of autonomic
conditionability (LaBar et al., 1998; Phelps et al., 2001). In order to assess this
relationship, a correlation analysis was conducted between amygdala BOLD
activity during fear conditioning and conditioned SCRs (the difference scores, as
described above). Amygdala BOLD activity was represented by extracted
parameter estimates from the fear conditioning functional map of the CS+>CS-
contrast in the left and right amygdala separately. Due to the predicted
attenuation of BOLD signal in the amygdala to conditioning over time, the SCR
difference scores were also correlated with parameter estimates from the left and
right amygdala during the first and second half of the first run.
As previous research has shown that individuals with high psychopathic traits
and low anxiety exhibit deficient fear conditioning as measured by SCRs
(Lykken, 1957), the relationship between psychopathic traits (as measured by the
composite psychopathy score, PPI total and subscales) and SCR during fear
conditioning was assessed using correlation analyses. Further, to examine
whether trait anxiety would moderate the relationship between SCR during fear

51

conditioning and psychopathic traits, multiple regression analyses were
performed, containing an interaction term between psychopathic traits and trait
anxiety. In these parallel analyses, the differential SCR conditioning score was
entered as the dependent variable, while the predictors were as follows:
psychopathic traits (composite psychopathy, PPI total and subscales), trait
anxiety (WAS) and the interaction term of psychopathic trait scores and trait
anxiety scores. All predictors were centered (the group mean was subtracted
from each score) to reduce multicolinearity issues (Aiken & West, 1991).
Creation of the interaction term involved the multiplication of the two centered
predictors. It was predicted that the interaction term would be significant within
the multiple regression model, indicating that the relationship between
psychopathic traits and differential SCR during fear conditioning is conditional,
and depends on levels of trait anxiety. Specifically, it was predicted that an
increase in psychopathic traits coupled with a decrease in trait anxiety would
result in a decrease in differential SCR scores during fear conditioning.
Scanning Procedure
Participants were scanned at the Dana and David Dornsife Cognitive
Neuroscience Imaging Center at the Univeristy of Southern California in a 3T
Siemens MAGNETOM Trio MRI scanner. All images were collected using a 32-
channel head coil. A high resolution T1-weighted anatomical scan was
performed on all participants structural scan (MPRAGE; TR = 2530ms; TE = 3.37
ms; flip angle = 7°; 256 mm FoV; 256x256 voxel matrix; 208 coronal slices; 1mm

52

isotropic voxels). Two functional scans were acquired during the fear
conditioning paradigm. 229 volumes of echo-planar volumes were acquired
continuously with 41 slices per volume, and with the following parameters: TR =
2000 ms, TE = 35 ms, flip angle = 90°, 64×64 matrix with a spatial resolution of
3.5×3.5×3.5 mm, slice thickness: 3mm and interslice time = 54 ms, with no slice
gap.

Functional Image Analysis
Functional data analysis was performed using FEAT (FMRI Expert Analysis
Tool), part of FSL (FMRIB's Software Library, www.fmrib.ox.ac.uk/fsl). Standard
pre-processing steps were carried out on individual participants, and include;
interleaved slice timing correction, motion correction using MCFLIRT (Jenkinson
& Smith, 2001); non-brain removal using BET (S. M. Smith, 2002); spatial
smoothing using a Gaussian kernel of full-width-half-maximum 6mm; and high
pass temporal filtering (Gaussian-weighted least-squares straight line fitting, with
sigma = 50.0 s). The estimated motion parameters for each participant were
included as covariates of no interest to reduce spurious activations due to head
motion and scanner drift. BOLD response was modeled using a separate
explanatory variable (EV) for each stimulus condition. For fear conditioning,
these conditions include: a) conditioned stimulus paired with aversive stimuli
(CS+; modeled as 5.8 s duration); b) conditioned stimuli not paired with
conditioned stimulus (CS-; modeled as 5.8 s duration); c) unconditioned stimulus

53

for CS+ (UCS; modeled as 3 s duration); d) non-unconditioned stimulus for CS-
(non-UCS; modeled as 3 s duration). For each stimulus type, the stimulus timing
files were convolved with a double gamma hemodynamic response function to
produce an expected BOLD response. Data was then fitted to the model using
FSL’s implementation of the general linear model. The resulting z statistic
images were thresholded at p < 0.001 (uncorrected), and warped into a standard
space based on the MNI-152 atlas (2 x 2 x 2 mm
3
Montreal Neurological Institute
(MNI) brain) using FLIRT (FMRIB’s Linear Image Registration Tool) (Jenkinson &
Smith, 2001; Jenkinson et al., 2002). A second-level analysis to average across
the two runs was carried out using a fixed effects model, by forcing the random
effects variance to zero in FLAME (FMRIB’s Local Analysis of Mixed Effects)
(Beckman, Jenkinson, & Smith, 2003; Woolrich, Behrens, Beckmann, Jenkinson,
& Smith, 2004; Woolrich, 2008). Mixed effects higher-level analysis for the
participants was carried out using FLAME 1 (FMRIB's Local Analysis of Mixed
Effects) (Beckmann, Jenkinson, & Smith, 2003). FLAME 1 uses both fixed
effects and random effects from cross session/subject variance (Beckmann et al.,
2003; Woolrich et al. 2004; Woolrich, 2008). Z statistic images at this level were
thresholded using clusters determined by z > 2.3 and a corrected cluster
significance threshold of p < 0.05 (Jezzard, Matthews, & Smith, 2001).

54

Regions of Interest Analysis
Given the a priori hypotheses, anatomical Regions of Interest (ROI) for the
ventromedial prefrontal cortex (VMPFC) and amygdala were created in MNI
space. The VMPFC ROIs were drawn with boundaries as follows: superior
boundary defined by a line drawn from the genu of the corpus callosum to the
frontal most portion of the prefrontal cortex; the lateral boundaries were defined
by the orbital sulci (See Figure 2-4).

Figure 2.4 Masks of ventromedial prefrontal cortex (VMPFC). Handdrawn
masks were used to constrain functional analyses and are displayed in red on
the MNI averaged brain.
The FSL Harvard-Oxford probabilistic atlas was used to create amygdala ROIs,
thresholded at 60% (See Figure 2-5).

55

Figure 2.5 Masks of amygdala. Masks used to constrain functional analyses
were created from the FSL Harvard-Oxford probabilistic atlas (thresholded at
60%) and are displayed in green on the MNI averaged brain.
Featquery (Smith et al., 2004) in FSL was used to extract the parameter
estimates from the four ROIs during the four conditions: CS+, CS-, UCS, non-
UCS. To determine whether fear conditioning was successful, the parameter
estimates from the CS+ and CS- conditions were contrasted with a paired t-test
for each ROI.

It has been previously demonstrated that amygdala responses rapidly habituate
during conditioning (Büchel, Morris, Dolan, & Friston, 1998; Quirk, Armony, &
LeDoux, 1997). For this reason, CS+ and CS- parameter estimates for the first
half of the first run and the second half of the first run were extracted from each
of the four ROIs. Data from one subject had to be removed due to extreme
BOLD values that were deemed outliers using the outlier labeling rule (Hoaglin &

56

Iglewicz, 1987; Tukey, 1977). A two-way repeated measures ANOVA was
conducted, using the two stimulus conditions (CS+, CS-) and “half of
conditioning” (first half, second half) as factors. Follow-up paired t-tests were
used to directly examine fear conditioning within each half of conditioning. To
assess the relationship between psychopathic traits and fronto-amygdala activity
during fear conditioning, correlation analyses were performed between the
parameter estimates of the CS+>CS- contrast from each of the 4 ROIs
separately and the psychopathy assessments (composite psychopathy score and
PPI). To assess whether trait anxiety would moderate the relationship between
psychopathic traits and neural function in the fronto-amygdala circuit during fear
conditioning, multiple regression models were implemented. In parallel models,
the parameter estimates from the CS+>CS- contrast from each ROI was used as
a dependent variable, while psychopathic traits (composite psychopathy score,
PPI), trait anxiety (WAS), and the interaction term between psychopathy and trait
anxiety were used as predictors in the model. Psychopathic trait scores and trait
anxiety scores were centered by subtracting the mean from each value.
Centering predictors in multiple regression models reduces problems raised by
multicollinearity among predictor variables when computing interaction terms
(Jaccard, Turrisi, & Wan, 1990). The interaction term contained the product of
the centered psychopathic trait scores and trait anxiety scores. The interaction
term investigates whether the correlation between psychopathic trait scores and
neural activity in the fronto-limbic circuit is conditional upon trait anxiety levels.

57

Specifically, it was predicted that an increase in psychopathic traits coupled with
a decrease in trait anxiety would result in a decrease in fronto-amygdala neural
activity during fear conditioning. Correlations and multiple regression analyses
were performed with SPSS (Release Version 18.0, © SPSS, Inc., 2009, Chicago,
IL, www.spss.com).

2.4 RESULTS
Sample Characteristics
Means, standard deviations, and bivariate correlations between trait anxiety and
total psychopathy measures for the sample are shown in Table 2-3, and between
trait anxiety and the psychopathy subscales are shown in Table 2-4. Total PPI
scores (mean ± SD = 286.72 ± .34.3) were comparable to total PPI scores for
males aged 18-24 years from community/college samples (Lilienfeld; mean =
301.06). Means for the WAS scores (mean ± SD = 11.38 ± 8.72) were
comparable to those previously reported for non-clinical samples of adult males
(Mean = 10.5; Colligan & Offord, 1988). Age of participants was not related to
the composite psychopathy score, the PPI (total or subscales), or WAS.

58

Table 2-3. Descriptive Statistics and correlations for trait anxiety and
psychopathy measures
n Mean SD Correlations
PPI Comp-P WAS
PPI 24 287.5 34.8 - .35 -0.08
Comp-P 24 -0.05 1 - -0.01
WAS 24 11.38 8.72 -
Note: PPI = Psychopathic Personality Inventory; Comp-P = Composite
Psychopathy Score; WAS = Welsh Anxiety Scale. * indicates significance at p <
.01.
Table 2-4. Descriptive Statistics for trait anxiety and psychopathy
subscales

Note: WAS = Welsh Anxiety Inventory; PPI-ME = Machiavallian Egocentricity;
PPI-RN = Rebellious Nonconformity; PPI-BE = Blame Externalization; PPI-CN =
Carefree Nonplanfulness. * indicates significance at p < .05; ** indicates
significance at p < .01. n= 24 for all subscales.

Mean SD
WAS
PPI-
ME
PPI-
RN
PPI-
BE
PPI-
CN

WAS 66.67 8.78 - .17 -.10 .04 .27
PPI-ME 41.29 8.43 - .37 .31 .25
PPI-RN 30.83 5.87 - .57** .08
PPI-BE 26.75 5.89 - .28
PPI-CN 37.71 8.95 -

59

The PPI and the composite psychopathy score were moderately correlated, with
only a trend toward significance (p = 0.09). However, closer inspection of the
scores revealed that two individuals had switched their pattern of psychopathic
traits from the previously collected scores: one person switched from a low to
high score, and one person switched from a high to low score. When these
individuals were removed from the correlation analysis, PPI and the composite
psychopathy score were strongly significantly correlated, r(23) = 0.68, p < 0.001).
Even so, the two individuals’ scores for both scales were included in all
subsequent analyses because their psychopathy scores were not deemed
outliers in their respective measures.
Importantly, none of the total psychopathy measures were correlated with the
trait anxiety measure. Psychopathic traits and trait anxiety scores were to be
used in multiple regression models, and non-correlated dependent variables in
multiple regression analyses are crucial to avoid issues of multicollinearity (Aiken
& West, 1991). Also, as expected, PPI-II subscales were not correlated with
WAS. Additionally, in conjunction with prior findings, PPI-FD, and its subscales,
were significantly negatively correlated with WAS (with the exception of the
correlation between Fearlessness and WAS which did not reach significance).
Behavioral Results
Although all participants were informed of the CS-UCS contingency prior to the
start of the experiment, 78.3% of participants reported being aware of the CS-US

60

contingencies as assessed by the post-scan questionnaire, 8.7% thought the
pairing was not systematic, and another 8.7% could not recognize the pairing.
Additionally, subjects rated the CS+ (mean rating ±SD = 2.17 ± .78) as
significantly less pleasant than the CS- (mean rating ±SD = 3.30 ± .82) on a 4-
point Likert scale from 1 = “very unpleasant” to 4 = “very pleasant’), t(23) = 5.15,
p < 0.001. Arousal ratings for the stimuli revealed the CS+ to be significantly
more arousing (mean rating ±SD = 2.35 ± 1.07) than the CS- (mean rating ±SD =
1.74 ± 86) on a 4-point Likert scale from 1 = “very unarousing” and 4 = “very
arousing”, t(23) = 3.42, p < 0.01.

The composite psychopathy score, the total PPI score, or PPI-II subscales were
not correlated with arousal or valence ratings.

SCR results
SCRs from the OR and AR phases were not significantly correlated with age,
although UCRs were significantly negatively correlated with age (r(19) = -0.57, p
< 0.01).

The ANOVA revealed significantly higher SCRs to the CS+ condition than the
CS- condition (F(1, 20) = 64.23, p < .000). Additionally, there was a significant
main effect for SCR phase (F(1,20) = 32.78, p <.000). Follow-up paired t-tests
revealed no significant differences between OR and AR phases (t(20) = -0.04, p

61

> 0.05). There were significant differences between the UCR phase and the OR
phase (t(20) = -4.98, p < .000), as well as the AR phase (t(20) = -4.74, p < .000).
There was also a significant interaction between condition and phase (F(1,20) =
32.04, p < .000), indicating greater activity during the CS+ condition than the CS-
condition in all three SCR phases (See Figure 2-6; Table 2-5 for means).

Figure 2-6. Skin conductance responses by phase and condition
(continued). CS+ denotes the conditioned stimulus paired with the
unconditioned stimulus and CS- denotes the unpaired conditioned stimulus. OR
= orienting response; AR = anticipatory response; UCR = unconditioned
response. * indicates significance at p < 0.05.

0"
0.2"
0.4"
0.6"
0.8"
1"
1.2"
1.4"
1.6"
Square Root transformed SCR Values
CS+$
CS%$
*$
*$
Orienting
Response
(OR)
Anticipatory
Response
(AR)
Unconditioned
Response
(UCR)
*$

62

Table 2-5. Means and standard deviations for skin conductance responses
(SCR) across both runs, by condition and SCR phase.

CS + CS-
OR AR UCR OR AR UCR
Mean .64 .61 1.34 .43 .47 .50
Standard deviation .32 .30 .52 .25 .34 .30
Note: Values reported as square root-transformed micromho units. OR =
orienting phase; AR = anticipatory phase; UCR = unconditioned response phase,
CS+ = conditioned stimulus paired with aversive stimulus; CS- = conditioned
stimulus unpaired with aversive stimulus.

As previous work has shown a positive relationship between amygdala activity
and SCR responses during fear conditioning (Phelps, 2001; LaBar, 1998),
difference scores for all phases of SCR were correlated with left and right
amygdala BOLD activity during fear conditioning in the corresponding SCR time
window. During fear conditioning, SCRs of the orienting phase (OR) were
positively correlated with right amygdala BOLD activity, r(20) = 0.41, p = 0.07,
with a trend toward significance.

When the first run was divided into halves to account for time-dependent
amygdala attenuation effects, the SCRs from the orienting response were
positively correlated with BOLD activity in the right amygdala during the first half
r(19) = 0.41, p = 0.075 (see Figure 2-7), and second half of conditioning, r(19) =
0.64, p = 0.002 (see Figure 2-8). SCRs of the orienting phase (OR) were also

63

positively correlated with left amygdala BOLD responses, r(19) = 0.57, p = 0.008,
during the second half of conditioning (see Figure 2-9).

Figure 2-7. Skin conductance orienting response (OR) difference scores
correlate with right amygdala activity during first half of fear conditioning.
Right amygdala activity measured by parameter estimates, r(19) = 0.41, p =
0.075.

-‐60

-‐40

-‐20

0

20

40

60

80

-‐0.8
-‐0.6
-‐0.4
-‐0.2
0
0.2
0.4
0.6
0.8

Right Amygdala BOLD
response (parameter
estimates)
Square Root Transformed OR Difference Score

64

Figure 2-8. Skin conductance orienting response (OR) difference scores
correlate with right amygdala activity during second half of fear
conditioning. Left amygdala activity measured by parameter estimates, r(19) =
0.64, p = 0.002.

Figure 2-9. Skin conductance orienting response (OR) difference scores
correlate with left amygdala activity during second half of fear
conditioning. Left amygdala activity measured by parameter estimates, r(19) =
0.57, p = 0.008.

-‐80

-‐60

-‐40

-‐20

0

20

40

60

80

100

-‐0.8
-‐0.6
-‐0.4
-‐0.2
0
0.2
0.4
0.6
0.8

Right AMygdala BOLD
response (parameter
estimates)
Square Root Transformed OR Difference Score
-‐80

-‐60

-‐40

-‐20

0

20

40

60

80

-‐0.8
-‐0.6
-‐0.4
-‐0.2
0
0.2
0.4
0.6
0.8

Left Amygdala BOLD
response (parameter
estimates)
Square Root Transformed OR Difference Scores

65

fMRI Results
Because the amygdala has been shown to have time-dependent responses
during fear conditioning (LaBar, 1998), the first and second halves of conditioning
were examined separately for all four ROIs. The repeated measures two-way
ANOVA results from the two stimulus conditions (CS+, CS-) within the first and
second halves of conditioning (referred to as ‘condition half’ from here on)
revealed no significant main effect for stimulus condition, F(1, 22) = 1.8, p > 0.05
for the left VMPFC, but a significant effect for conditioning half, F(1, 22) = 5.24, p
< 0.05; the second half of conditioning had higher levels of BOLD activity. The
stimulus condition x conditioning half interaction was also significant, F(1, 22) =
7.18, p < 0.05). Follow-up paired t-tests revealed significantly more BOLD
activity in the left VMPFC for the CS- condition during the first half of
conditioning, t(22) = -2.66, p < 0.01 (see Figures 2-10 and 2-11). There were no
significant main effects for stimulus condition, condition half, or for the stimulus
condition x conditioning half interaction in left amygdala, right amygdala, or right
VMPFC.

66

Figure 2-10. Parameter estimates by region of interest (ROI) for the first half
of fear conditioning. * indicates significance at p < 0.05. VMPFC =
ventromedial prefrontal cortex; CS+ = conditioned stimulus paired with shock;
CS- = conditioned stimulus unpaired with shock.

-‐15

-‐10

-‐5

0

5

10

ROI Parameter Estimates
CS+ 1st half
CS- 1st half
Left
Amygdala
Right
Amygdala
Left
VMPFC
Right
VMPFC
*

67

Figure 2-11. Parameter estimates by region of interest (ROI) for the second
half of fear conditioning. VMPFC = ventromedial prefrontal cortex; CS+ =
conditioned stimulus paired with shock; CS- = conditioned stimulus unpaired with
shock.

Relationships between Psychopathic Traits and fMRI during Fear
conditioning
There were no significant correlations between fronto-amygdala activity and the
composite psychopathy score or the total PPI score during fear conditioning in
any of the four ROIs. In the examination of the subscales of the PPI-II, a
significant negative correlation emerged between the Machiavellian Egocentricity
subscale and BOLD activity in the left amygdala (r(23) = -0.42, p < 0.05) during
fear conditioning (contrast of CS+>CS-) (see Figure 2-12). Machiavellian
Egocentricity was also negatively correlated to BOLD activity during fear
-‐10

-‐5

0

5

10

15

ROI Parameter Estimates
CS+ 2nd half
CS- 2nd half
Left
Amygdala
Right
Amygdala
Left
VMPFC
Right
VMPFC

68

conditioning in the right amygdala, although this did not reach significance, r(23)
= -0.17, p > 0.05.

Figure 2-12. Left Amygdala BOLD activity negatively correlates with
Machiavellian Egocentricity. r(23) = -0.42, p < 0.05.

Relationships between fMRI during Fear conditioning, Psychopathic Traits,
and Trait Anxiety
Multiple regression analyses were used to examine the independent and
interaction effects of psychopathic traits and trait anxiety on BOLD activity during
fear conditioning in each of the four ROIs. Parallel analyses were conducted
using one of the psychopathy measures (composite psychopathy score, total
PPI, or PPI-II subscales) as one independent variable and the trait anxiety
measures (WAS) as the other independent variable, while using parameter
estimates from one of the four ROIs as the dependent variable in a multiple
regression model. We predicted that trait anxiety would moderate the
relationship between BOLD activity and psychopathic traits during fear
-‐60

-‐50

-‐40

-‐30

-‐20

-‐10

0

10

20

30

40

50

0
10
20
30
40
50
60

Left Amygdala BOLD
activity (parameter
estimates)
PPI
Machiavellian
Egocentricity

69

conditioning such that individuals with the highest levels of psychopathic traits
and lowest levels of anxiety would display the least BOLD signal. The model
using the Machiavellian Egocentricity subscale of the PPI and the WAS to predict
BOLD activity in the right amygdala was significant, R
2
= 0.36 (adjusted R
2
=

0.26), F(3,20) = 3.75, p < 0.05 (see Figure 2-13). The raw and standardized
regression coefficients of the predictors are reported in Table 2-6. While the two
main predictors of Machiavellian Egocentricity and Welsh Trait Anxiety did not
significantly predict BOLD activity in the right amygdala during fear conditioning,
the Machiavellian Egocentricity x WAS interaction term did. The beta value was
positive, indicating that as psychopathic traits increased and trait anxiety
decreased, there was an observed decrease in BOLD activity in the right
amygdala during fear conditioning, supporting the predictions. The WAS also
moderated the relationship between Machiavellian Egocentricity and left
amygdala BOLD activity during fear conditioning, although the model only
trended toward significance, R
2
= 0.30 (adjusted R
2
=

0.20), F(3,20) = 2.87, p =
0.06 (see Table 2-6 for unstandardized and standardized coefficients and t-
values; see Figure 2-14). No other regression models were significant.

70

Table 2-6. Unstandardized and standardized coefficients and t-values for
regression model of the interaction between trait anxiety and psychopathic
traits to predict BOLD activity in bilateral amygdala
Brain
Region
Unstandardized
Coefficients
Standardized
Coefficients

B SE β t
Right
Amygdala
(Constant) -4.17 2.93 -1.44
PPI_ME -0.33 0.36 -0.17 -0.93
WAS 0.29 0.35 0.15 0.82
PPI_ME*
WAS
0.13 0.05 0.52 2.82*
Left
Amygdala
(Constant) -1.87 3.29 -0.57
PPI_ME -0.90 0.40 -0.43 -2.25*
WAS 0.37 0.39 0.18 0.93
PPI_ME*
WAS
0.07 0.05 0.26 1.37
Note. PPI_ME = Machiavellian Egocentricity; WAS = Welsh Anxiety Scale; *
indicates significance at p < .05.

Figure 2-13. Trait anxiety interacts with Machiavellian Egocentricity to
predict BOLD activity in right amygdala. PPI_ME = Machiavellian

71

(Figure 2-13 continued) Egocentricity; trait anxiety indexed by Welsh Anxiety
Scale. The sample mean of trait anxiety is represented by the solid blue line.
Dotted lines represent trait anxiety scores above the sample mean (red) or below
the sample mean (black). F(3,20) = 3.75, p < 0.05.

Figure 2-14. Trait anxiety interacts with Machiavellian Egocentricity to
predict BOLD activity in left amygdala. PPI_ME = Machiavellian Egocentricity;
trait anxiety indexed by Welsh Anxiety Scale. The sample mean of trait anxiety is
represented by the solid blue line. Dotted lines represent trait anxiety scores
above the sample mean (red) or below the sample mean (black). F(3,20) = 2.87,
p = 0.06.

Relationships between SCR during Fear conditioning, Psychopathic Traits,
and Trait Anxiety
Because previous results have shown a negative relationship between
psychopathic traits and SCRs during fear conditioning, SCR difference scores
(CS+ minus CS-) from each of the three phases (OR, AR, UCR) were also

72

correlated with the composite psychopathy score, the total PPI, and the PPI-II
subscales. SCRs from the OR phase were significantly negatively correlated
with the composite psychopathy score (r(20) = -0.46, p < 0.05), and three PPI
subscales: Blame Externalization, r(20) = -0.48, p < 0.05; Carefree
Nonplanfulness, r(20) = -0.43, p < 0.05). SCRs from the AR phase were
significantly positively correlated with the total PPI score, r(20) = 0.49, p < 0.05.
Thus, it appears that as total psychopathy scores and certain subscales of
psychopathy increase, there is a decrease in amplitudes of SCR during the OR
phase of fear conditioning, replicating previous findings. However, it appears
that during the AR and UCR phases of SCR to fear conditioning there is a
positive relationship between total scores and certain psychopathy subscales.

Multiple regression analyses were used to examine the independent and
interaction effects of psychopathic traits and trait anxiety on SCRs during three
different phases of fear conditioning. Parallel analyses were conducted using
one of the psychopathy measures (composite psychopathy score, total PPI, or
PPI-II subscales) as one independent variable and the trait anxiety measure
(WAS) as the other independent variable, while using difference scores from one
of the phases of SCRs as the dependent variable. It was predicted that trait
anxiety would moderate the relationship between the OR and AR phase
amplitudes, but not UCR, of SCR during fear conditioning and psychopathic traits
such that an increase in psychopathic traits coupled with a decrease in trait

73

anxiety would result in an observed decrease in SCR amplitude in these phases.
The results of the regression analyses revealed three significant models, and the
raw and standardized regression coefficients of the predictors of each model are
reported in Table 2-7. The model using the Blame Externalization subscale of
the PPI and the WAS to predict OR SCR amplitudes during fear conditioning was
significant, R
2
= 0.53 (adjusted R
2
=

0.45), F(3,17) = 6.46, p = 0.004 (see Figure
2-15). Independently, Blame Externalization is negatively associated, and WAS
positively associated, with OR SCR amplitudes. However, the interaction term of
Blame Externalization x WAS is also significant, indicating that the value of one
predictor depends on the other. The beta value of the interaction term is positive,
meaning that as Blame Externalization increases and trait anxiety decreases,
there is a decrease in amplitude in OR SCRs during fear conditioning, which
supports the prediction. The second significant regression model included the
Carefree Nonplanfulness subscale of the PPI and WAS as independent variables
and OR SCR amplitudes as the dependent variable, R
2
= 0.51 (adjusted R
2
=

0.42), F(3,17) = 5.87, p = 0.006 (see Figure 2-16). This model was similar to the
previous one in that the significant interaction term, which also has a positive
beta value, indicates that an increase in carefree nonplanfulness coupled with a
decrease in trait anxiety result in decreased OR SCR amplitudes during fear
conditioning. Overall, it appears that the predicted effect of trait anxiety on the
relationship between psychopathic traits and SCRs to fear conditioning in the
orienting SCR phase are supported.

74

Table 2-7. Unstandardized and Standardized Coefficients and T-values for
the regression model of the interaction between trait anxiety and
psychopathic traits in predicting skin conductance responses during fear
conditioning
SCR
Phase
Unstandardized
Coefficients
Standardized
Coefficients

B SE β t
OR (Constant) 0.19 0.04 4.55*
PPI_BE -0.02 0.01 -0.34 -1.92
WAS 0.01 0.01 0.47 2.54*
PPI_BE*
WAS
0.00 0.00 0.59 3.05*
OR (Constant) 0.19 0.05 3.95*
PPI_CN -0.01 0.01 -0.48 -2.66*
WAS 0.01 0.01 0.27 1.47
PPI_CN*
WAS
0.00 0.00 0.46 2.63*
Note. PPI_BE = Blame Externalization; PPI_CN = Carefree Nonplanfulness;
WAS = Welsh Anxiety Scale; * indicates significance at p < .05.

75

Figure 2-15. Trait anxiety interacts with Blame Externalization to predict
orienting skin conductance responses during fear conditioning. PPI_BE =
Blame Externalization; trait anxiety indexed by Welsh Anxiety Scale. The sample
mean of trait anxiety is represented by the solid blue line. Dotted lines represent
trait anxiety scores above the sample mean (red) or below the sample mean
(black). F(3,17) = 6.46, p = 0.004.

76

Figure 2-16. Trait anxiety interacts with Carefree Nonplanfulness to predict
orienting skin conductance responses during fear conditioning. PPI_CN =
Carefree Nonplanfulness; trait anxiety indexed by Welsh Anxiety Scale. The
sample mean of trait anxiety is represented by the solid blue line. Dotted lines
represent trait anxiety scores above the sample mean (red) or below the sample
mean (black). F(3,17) = 5.87, p = 0.006.

2.5 DISCUSSION
In this study, trait anxiety moderated the relationship between psychopathic traits
and two separate physiological measures during fear conditioning: skin
conductance responses (SCRs) and BOLD activity in the amygdala. Each of
these two findings will be discussed separately below.

77

Behavior
Valence and arousal ratings confirm that participants recognized the CS-UCS
pairing, and rated the conditioned stimulus paired with electrocutaneous
stimulation as more arousing and less pleasant than the unpaired conditioned
stimulus. Additionally, skin conductance responses over the course of the
experiment, although not available for all participants, indicated successful fear
conditioning, with higher amplitudes of skin conductance responding to the CS+
than to the CS-. Thus, judging from the behavior measures, conditioning was
successful.

Fronto-amygdala activity during Fear Conditioning
When the first run was divided into halves to account for amygdala time-
dependent responses, the left VMPFC was the only region that displayed
significantly different BOLD activity between the CS+ and CS- conditions. Left
VMPFC was significantly more active for the CS- condition than the CS+
condition in the first half of the first run. The VMPFC is involved in forming, as
well as inhibiting, responses to fear (Burgos-Robles, Vidal-Gonzalez, Santini, &
Quirk, 2007; Laurent & Westbrook, 2008; Marek, Strobel, Bredy, & Sah, 2013),
as well as generally being involved in assigning and updating values related to
stimuli (Baxter, Parker, Lindner, Izquierdo, & Murray, 2000; Wallis & Miller, 2003;
Walton, Behrens, Buckley, Rudebeck, & Rushworth, 2010). These roles may
explain the increased BOLD activation during the early trials of the CS- condition,

78

when contingencies and responses were being learned. Prior studies have also
found higher levels of ventromedial prefrontal cortex activity (including anterior
cingulate cortex) when learning new contingencies during fear conditioning
(Morris & Dolan, 2004; Schiller, Levy, Niv, LeDoux, & Phelps, 2008), during early
versus late phases of acquisition (LaBar et al., 1998), and during CS- during
acquisition (Phelps et al., 2004).

Psychopathic Traits negatively correlated with BOLD activity in fronto-
amygdala circuit during Fear Conditioning
Negative relationships between Machiavellian Egocentricity and BOLD activity in
bilateral amygdala emerged during fear conditioning, although only the left
amygdala reached significance. Although an a priori prediction about the
direction of this relationship was not made due to the expected moderating effect
of trait anxiety, these findings are concordant with previous fMRI studies
examining fear conditioning in criminals with psychopathy, which found
significantly less activity in the amygdala in criminals with psychopathy, when
compared to healthy controls (Birbaumer et al., 2005; Veit et al., 2002). The
current study differs from previous studies in one important aspect making it a
valuable contribution to the literature. Both prior studies did not use criminal non-
psychopathic control groups, making it difficult to ascertain whether the
differences in neural activity during fear conditioning were due to incarceration
and its effects, or to psychopathy. The current study examines a wide range of

79

psychopathic traits within one community sample, and is thus in a better position
to permit more precise conclusions about the association between psychopathic
traits and amygdala BOLD activity during fear conditioning.

With regard to the laterality of results, negative associations between
psychopathic traits were discovered in the amygdala in both hemispheres,
although only the left amygdala reached significance. Additionally, the
moderating effect of trait anxiety was also revealed bilaterally, although it only
reached significance in the right hemisphere. One previous fMRI fear
conditioning study on psychopathy revealed significantly less activity in the left
amygdala (compared to healthy controls; Birbaumer et al., 2005), while the other
revealed significantly less activity in the right amygdala (compared to healthy
controls; Veit et al, 2002). These two studies are from the same research group
and have nearly identical protocols and populations. Therefore, the differing
results highlight the effect of individual variability on studies with small sample
sizes (10 subjects in one study and 4 subjects in the other). Additionally, the
medial temporal lobe is highly prone to susceptibility artifacts (Chen, Dickey,
Yoo, Guttmann, & Panych, 2003), which may contribute to an apparent activation
variability. Thus, it has been suggested recently that lateralization effects in the
amygdala should be interpreted with caution (Mathiak, Zvyagintsev, Ackermann,
& Mathiak, 2012). With these facts in mind, it is difficult to draw conclusions
about laterality from the current results. However, it is encouraging that the

80

results of the current study show trends that go in the same direction for both
amygdalae. Theses amygdala results will need to be replicated in future studies
with larger sample sizes to confirm the findings.

The VMPFC did not display any significant relationship with psychopathic traits
during fear conditioning, nor was there a moderating effect of trait anxiety. A
comparison with two previous studies using fMRI of fear conditioning in
psychopathy is difficult, as those studies did not have proper control groups, thus
limiting the interpretation of the results. Despite this problem, it has been shown
that individuals high on psychopathy display reduced functional activity in the
VMPFC as measured by fMRI during certain tasks, e.g. during tasks using
emotional memory (Kiehl et al., 2001) and social cooperation (Rilling et al.,
2007), although increased activity has also been observed during other tasks,
such as tasks involving affective theory of mind (Sommer et al., 2010) and
cognitive control (Sadeh et al., 2013). In the case of fear conditioning, classic
studies reveal that the VMPFC is a necessary region for fear conditioning
(Burgos-Robles et al., 2007; Corcoran & Quirk, 2007; Laurent & Westbrook,
2008; Sotres-Bayon & Quirk, 2010). One of the primary roles of the VMPFC
includes the evaluation of the reinforcement value of a stimulus and the learning
of stimulus–reinforcement associations (O'Doherty, Hampton, & Kim, 2007; Rolls,
Critchley, Browning, Hernadi, & Lenard, 1999; Rolls & Grabenhorst, 2008). It is
possible that individuals with high psychopathic traits from a community sample

81

have difficulties in linking emotional responses to aversive stimuli via the
amygdala, but that they are able to properly make explicit associations between
stimulus and outcome through VMPFC. The evidence for this statement comes
from the accurate reports of valence and arousal ratings across all participants.
Thus, although reductions in VMPFC activity have been shown in other
neuroimaging studies of psychopathy, a difference in activation based on
psychopathic traits was not revealed in the current study by fear conditioning.
Although the mean of total PPI and PPI-II subscales are comparable to those
reported for offender samples (Hughes et al., 2013; Lilienfeld & Andrews, 1996),
it is possible that they are not high enough to reveal functional differences in
VMPFC.
Trait Anxiety moderates the relationship between psychopathic traits and
fear conditioning related neural responses in fronto-amygdala circuit
Our results indicate for the first time that trait anxiety, as measured by the MMPI-
derived Welsh Anxiety Scale, interacts with Machiavellian Egocentricity to predict
BOLD activity in the right amygdala during fear conditioning, in the predicted
direction. In other words, as scores on Machiavellian Egocentricity increased
and trait anxiety decreased, there was a decrease in BOLD activity in the right
amygdala during fear conditioning. A similar effect was seen in the left
amygdala, although here the relation only tended toward significance. These
results are concordant with previous theories that proposed that individuals with

82

high psychopathic traits, but low trait anxiety possess greater abnormalities in the
neural function of the amygdala-VMPFC circuit.

The present study expands previous findings in a number of ways. First, it is the
first to investigate whether the relationship between psychopathic traits and
amygdala-VMPFC activity during fear conditioning is dependent upon levels of
trait anxiety. By examining this critical question, and particularly by using a
multiple regression model to do so, we are able to address the possibility for
psychopathic subtypes that are delineated by levels of trait anxiety on a
continuum. The results demonstrate that trait anxiety should be taken into
account when examining the relationship between psychopathic traits and neural
responses, and may help explain why neuroimaging of psychopathy has
produced inconsistent findings. Machiavellian Egocentricity is a subscale of the
PPI that loads onto PPI-II, thought to represent the antisocial aspect of
psychopathy. In the population used here, as social deviance increases (as
measured by ME) in lower anxious individuals, significantly reduced amygdala
BOLD activity is observed. Although the same effect is not viewed in the VMPFC
(similar to the correlation results above), the results support the conclusions from
previous studies (Koenigs et al, 2010), namely that individuals with high
psychopathic traits and low anxiety possess the greatest neural abnormalities. It
is possible that in community samples with lower levels of psychopathic traits,
such as in the current sample, these neural abnormalities are present in the

83

amygdala, rather than the entire amygdala-VMPFC circuit, as may be possible in
individuals with extremely high psychopathic traits. The results are concordant
with those of another study that found amygdala BOLD responses to emotional
stimuli moderated by trait anxiety in an adolescent population with severe
conduct disorder (thought to be a precursor to psychopathy). In that sample,
similar to the one in the current study, individuals with the highest psychopathic
traits and lowest trait anxiety showed the lowest neural responses (Sterzer,
Stadler, Krebs, Kleinschmidt, & Poustka, 2005). Taken together, the results
suggest that trait anxiety, in conjunction with psychopathic traits, should be
considered in the determination of neurobiological profiles.

Additionally, previous studies of fear conditioning in psychopathy, and
neuroimaging studies of psychopathy in general, have heavily relied upon
criminal populations due to the high prevalence of psychopathy rates in prisons
and jails. While these populations are informative, interpretation of results from
such studies is limited, as criminal and non-criminal populations are substantially
different (Raine, 1993). This is the first study to examine the relationship
between psychopathic traits and the neural correlates of fear conditioning in a
non-criminal sample. What makes the results particularly interesting is that, in
this twin sample, the range of psychopathic traits, although wide, does not
include the individuals with the highest scores. The subject with the highest
score in the MRI sample is only 1.5 standard deviations away from the group

84

mean, indicating that participants are high scorers for a community sample (the
group mean was comparable to that of offender samples (Lilienfeld, 2006), even
so it may still not represent the entire range of traits. Even within this range of
psychopathic traits, differences in neural reactivity to fear conditioning emerged.
Neural activation differences based on levels of psychopathic traits from
community samples have been observed before during a variety of tasks (Bjork,
Chen, & Hommer, 2012; Glenn, Raine, & Schug, 2009; Sadeh et al., 2013;
Sommer et al., 2010), and have actually shown more mixed results than studies
using criminal populations; this may be likely due to the wide and different levels
of psychopathic traits, as well as unaccounted trait anxiety, in the different
studies. It is possible that taking trait anxiety into account and including
individuals with a range of scores on psychopathy measures can explain some of
the differences in brain activation found in previous studies.

SCR responses during fear conditioning and psychopathic traits
Skin conductance responding during the orienting phase was positively
correlated with amygdala BOLD activity, similar to previous findings (Cheng,
Richards, & Helmstetter, 2007; LaBar et al., 1998; Phelps et al., 2001), and
supporting the role of the amygdala in regulating SCRs during conditioning.

In past studies, autonomic responses to fear conditioning were modulated by
psychopathic traits (Hare & Quinn, 1971; Hare & Thorvaldson, 1970; Lykken,

85

1957), with individuals with psychopathy having smaller SCRs when they had
higher scores of psychopathic traits. In the current study similar relationships
between autonomic responding and the composite psychopathy score was found
as well as with two subscales of the PPI (Blame Externalization, Carefree
Nonplanfulness). It is interesting that measures of psychopathic traits from ages
14 and 16 were negatively correlated with autonomic responding during fear
conditioning at ages 18 to 21, and that measures of psychopathic traits taken at
this later age also displayed a negative association. These results provide
further support for the stability of psychopathic traits, and their behavioral
correlates, over time.

Although a negative correlation was found between SCRs of the orienting phase
during fear conditioning and all three of the psychopathic trait subscales, it was
also found that trait anxiety moderates these relationships. The moderation for
BE and CN were similar to the fMRI results, meaning that an increase in these
particular psychopathic traits coupled with a decrease in trait anxiety, resulted in
decreased amplitudes in skin conductance responding during orienting
responses. While orienting responses are often interpreted as an attentional
orientation (Boucsein, 1992; Prokasy et al., 1973; Wolter & Lachnit, 1993), it has
been shown that they can be conditioned (Gale & Stern, 1967). The current
results support previous findings that show significant differences in OR
conditioning between criminals with psychopathy and criminals without

86

psychopathy (Hare & Quinn, 1971). Both BE and CN load onto PPI-II, which is
thought to correspond to Factor 2 (Antisocial) of the PCL-R (Lilienfeld &
Andrews, 1996). On the PCL-R, the correlate of the Blame Externalization
subscale is “failure to accept responsibility”, which loads onto the
interpersonal/affective facet of psychopathy (Factor 1). However, Benning et al
(2003) showed that the PPI Blame Externalization subscale was better
represented as a proneness to social deviance (Benning et al., 2003). Thus,
individuals with higher levels of social deviance, but lower levels of trait anxiety,
display reduced autonomic responding during fear conditioning. These results
are concordant with the fMRI results in the present study, as well as with the
notion that the amygdala, as part of its regulatory role, initiates SCR orienting
responses (Bagshaw, Kimble, & Pribram, 1965; Critchley et al., 2000; LANG,
TUOVINEN, & VALLEALA, 1964). Thus, including trait anxiety in the account of
psychopathy not only aids in predicting neurobiological profiles, but physiological
ones as well. Furthermore, it appears from these results, that although low-
anxious individuals with higher levels of psychopathic traits are believed to have
more severe emotional deficits that are usually represented by the Fearless
Dominance factor of the PPI (Factor 1 of the PCL-R), traits that load onto the
antisocial factor (though in conjunction with trait anxiety levels) are also important
to assess emotion-related physiological differences.

87

In summary, in a young adult community sample, neural activity during fear
conditioning was modulated by psychopathic traits in the amygdala, but not the
entire amygdala-VMPFC circuit. Furthermore, trait anxiety interacts with
psychopathic traits to predict neural activity, such that individuals with the highest
psychopathic traits and lowest trait anxiety display the greatest reduction in
BOLD signal. Additionally, similar results were found for skin conductance
responses during fear conditioning, indicating lower autonomic responding in
individuals with high psychopathic traits and low trait anxiety. Taken together,
these results reveal that individuals with higher levels of psychopathic traits and
lower trait anxiety have significantly reduced function of the amygdala and
reduced amygdala-related physiological responses (i.e. skin conductance
responses), but no changes in the VMPFC. Given the fact that VMPFC is
commonly implicated in fMRI studies of psychopathy, it is possible that VMPFC
dysfunction was not observed in the current population due to the fact that the
highest score on psychopathic traits might simply not have been high enough.
We need further studies investigating the interaction between trait anxiety and
psychopathic traits on neural activity in criminal populations to elucidate the role
of VMPFC in psychopathy.

Results support the theory that subtypes of psychopathy can be distinguished by
levels of trait anxiety, and that the subtypes may possess unique neurobiological
and physiological profiles. While studies of psychopathic subtypes typically have

88

used criminal individuals with psychopathy, this study is the first to show that trait
anxiety can be used to parse neural and physiological differences in a community
sample with a wide range of psychopathic traits, and that psychopathic traits
need not be at the extreme end of the spectrum to show differences. The current
results have implications for the recognition of subtypes of psychopathy, and
particularly for choices of treatment with customized treatment plans for each
profile.

89

CHAPTER 3: Psychopathic traits, trait anxiety and
amygdala-ventromedial prefrontal cortex functional
connectivity

3.1 ABSTRACT
Individuals with psychopathy possess an array of personality traits that include
callousness, lack of remorse, and impulsiveness. Abnormal functioning of
emotional regulation and behavioral control regions, such as amygdala and
ventromedial prefrontal cortex (VMPFC), are thought to contribute to the
disorder. Recent evidence suggests that amygdala-VMPFC functional
connectivity is reduced during resting state fMRI in criminals with psychopathy
(Motzkin, Newman, Kiehl, & Koenigs, 2011), and task-based fMRI in adolescents
(Marsh et al., 2008). However, it remains unknown whether the negative
relationship between psychopathic traits and amygdala-VMPFC connectivity
exists in a non-incarcerated sample during rest, and in an adult population during
task-based fMRI. Furthermore, despite evidence of subtypes of psychopathy
(defined by trait anxiety), the relationship between psychopathic traits, trait
anxiety, and amygdala-VMPFC functional connectivity have not been examined
in a community population. In this study, it is investigated whether psychopathic
traits modulate amygdala-VMPFC functional connectivity in a community sample
during: a) resting state fMRI, and b) task-based fMRI (i.e. fear conditioning), and
whether these relationships are influenced by levels of trait anxiety. Results
reveal that trait anxiety moderates the association between psychopathic traits

90

and amygdala-VMPFC functional connectivity at rest, such that individuals with
the highest psychopathic traits and lowest trait anxiety had the highest levels of
functional connectivity. Amygdala functional connectivity to other fear
conditioning-related neural regions was also negatively modulated by
psychopathic traits. The results support previous findings of psychopathic traits
modulating amygdala functional connectivity during task and rest, and provide
new evidence of trait anxiety as a moderator of connectivity during rest. Results
further support the notion of altered neural networks in individuals with higher
psychopathic traits, and provides support for subtypes of psychopathy.

3.1 INTRODUCTION
Contributing to evidence of possible neural differences between subtypes of
psychopathy in criminal populations, results from Chapter 2 demonstrated that
trait anxiety interacted with psychopathic traits to predict BOLD activity in the
amygdala during a fear conditioning task within a community population. Thus,
emerging evidence supports the notion of neural differences, specifically in the
amygdala and VMPFC, between low- and high-anxious individuals with
psychopathic traits, in both criminal and community samples. However, the
relationship between psychopathic traits, trait anxiety, and the functional
connectivity of amygdala and VMPFC has been little studied.

91

As mentioned previously, it has been suggested that connectivity between the
two neural regions may be altered in individuals with psychopathy (Blair, 2008),
which would explain many of the emotional and behavioral deficits observed in
this population. One study reported a reduction in amygdala-VMPFC functional
connectivity during a facial affect processing task in adolescents with
psychopathic traits (Marsh et al., 2008). In another study, functional connectivity
between the two regions was examined in criminals with psychopathy. Motzkin
and colleagues demonstrated a reduction in amygdala-VMPFC connectivity in
criminals with psychopathy as compared to criminals without psychopathy,
seeming to corroborate previous findings (Motzkin et al., 2011).
Taking into consideration the behavioral deficits of the low-anxious subtype of
psychopathy, it might be expected that this subtype would possess less
amygdala-VMPFC functional connectivity, since increased amygdala-VMPFC
connectivity has been associated with increased emotional regulation and
aggression (Davidson, Jackson, & Kalin, 2000; Ochsner & Gross, 2005; Quirk &
Beer, 2006). However, Motzkin and colleagues (2011) reported an interaction
between psychopathic traits and trait anxiety in their criminal population, such
that lower anxiety was associated with greater amygdala-VMPFC functional
connectivity in individuals with psychopathy. Nonetheless, results of this study
cannot be considered definitive due to the modest size of the sample population
(n = 12 and n = 8). While important, the results from these do not allow direct

92

comparison because one examined resting state amygdala-VMPFC functional
connectivity, while the other examined functional connectivity of these two neural
regions during a task performance. Furthermore, the sample populations of the
two studies differ on both incarceration and age, as the first study used a criminal
adult population, while the second study used a non-criminal, adolescent
population. Finally, the study with adolescents did not address subtypes based
on trait anxiety, and it is unknown whether the trait anxiety findings of a criminal
population can be extended to a community population.
The current study aims to address the above problems. Specifically, both resting
state and task-based (i.e. fear conditioning) amygdala-VMPFC functional
connectivity is examined in a young adult, community sample with a wide range
of psychopathic traits. This study examines whether psychopathic traits
modulate amygdala-VMPFC functional connectivity during resting state and task-
based fMRI in order to be able to fairly compare across imaging modalities.
Furthermore, it is investigated whether trait anxiety moderates the relationship
between psychopathic traits and resting state, as well as task-based, amygdala-
VMPFC functional connectivity. It is predicted that an increase in psychopathic
traits, coupled with a decrease in trait anxiety, will result in decreased functional
connectivity in both modalities.

93

3.3 MATERIALS AND METHODS
The population, behavioral procedures, and fear conditioning paradigm for the
analyses proposed here are the same as described in Chapter 2 Materials and
Methods.
Scanning Procedure
Resting State fMRI
Functional images were acquired using an echo-planar T2*-weighted imaging
sequence. A total of 210 echo planar images (EPI) were obtained during a 7-
minute scan (axial slices = 38; TR/TE = 2500/30ms; FOV =
216mm; flip angle = 80 degrees; in-plane resolution = 3x3mm2; slice thickness =
3mm). Participants were instructed to remain awake, and keep their eyes open
for the duration of the scan. They were also instructed to look at a white
crosshair on a black background that was presented through a projector onto a
rear-projection screen attached to the head coil and located above the
participant’s head.

Image Preprocessing
Functional images were preprocessed using FSL Version 6 (FMRIB’s Software
Library, www.fmrib.ox.ac.uk/fsl). The EPI images were preprocessed using the
following steps: motion correction, removal of non-brain tissue, spatial smoothing
using a Gaussian kernel of 6mm full-width at half-maximum (FWHM), grand-

94

mean intensity normalization of the entire 4D dataset by a single multiplicative
factor, and a high-pass temporal filter of 100 seconds (i.e. 0.01 Hz).

Nuisance Variables and De-noising of fMRI Data
In order to account for physiological noise (e.g. cardiac and respiratory cycles)
and head motion artifact, 8 nuisance variables, including white matter (WM),
cerebrospinal fluid (CSF), and the 6 motion parameters, were included as
covariates of no interest in the regression model. In addition, a single-session
independent component analysis (ICA) was performed to further de-noise the
functional data of extreme head motion and physiological artifact (Beckmann and
Smith, 2004), by searching the data for noise patterns. Noise components were
classified according to a list of criteria developed in the lab of Dr. Mara Mather at
the University of Southern California. These criteria included: examining whether
signal was localized in white matter or ventricles; reviewing the power spectra to
help filter out high or scrambled frequency components; and scrutinizing the
principle eigenvector timeseries to exclude components with spikes that
exceeded 6 scale units (usually indicative of sudden head motion). Nuisance
covariates of WM and CSF were also created (Chang et al., 2009) to filter out
physiological noise that is similar to grey matter (Fox et al., 2005). WM and CSF
ROIs were first defined as 6mm-diameter spheres in MNI space. The center of
the WM ROI was located in a deep white matter tract (right superior longitudinal
fasciculus; [26, -12, 35]), and the center of the CSF ROI was in a ventricle (right

95

superior lateral ventricle; [19, -33, 18]). Different brain tissue-types were obtained
by segmenting each individual’s high- resolution structural image with the FAST
tool in FSL. Both the two spherical and two tissue ROIs were reverse-registered
to functional space resolution (2mm
3
) by applying the corresponding
transformation matrices created during fMRI data preprocessing. The re-
sampled tissue masks were threshold at 90% partial volume fraction to minimize
overlap with grey matter. Finally, the thresholded WM and CSF masks were
intersected with their respective spheres to acquire subject-specific CSF-only
and WM-only nuisance masks. The resultant tissue masks were applied to each
individual’s preprocessed fMRI data, and a mean timecourse was extracted by
averaging the timeseries across all voxels within each mask.

Amygdala Seed-Based fMRI Analysis
In order to investigate the connectivity of amygdala to VMPFC at rest, a seed-
based correlation analysis was executed. The left and right amygdala were
hand-drawn on each subject’s anatomical scan using a detailed protocol for
proper delineation of amygdala in MR images, based on the steps outlined in
(Convit et al., 1999) (see Figure 3-1).

96

Figure 3-1. Example of a handdrawn amygdala region of interest on an
individual subject’s brain.

First-level analyses were performed using FSL FEAT. The mean timeseries for
left and right amygdala was used as a temporal predictor in separate whole-brain
linear regressions. As a result, amygdala resting-state functional connectivity
spatial maps are produced for each subject such that significant clusters
represent a high degree of coherence with BOLD signal in the respective
amygdala. In addition, 8 nuisance regressors (CSF, WM, and 6 motion
parameters) were modeled as confound variables. In order to examine the
relationship between amygdala-VMPFC connectivity and psychopathic traits,
resulting Z-score statistical parametric maps were processed in a higher-level
one-sample t-test, with demeaned psychopathic trait scores (psychopathy
composite, PPI, PPI-II subscales) modeled as a covariate of interest. The higher-
level group analysis was thresholded at a cluster size of Z > 2.3 and a
significance of P < 0.05 to control for multiple comparisons. Analyses were also
constrained to VMPFC ROIs. Two contrasts were constructed to examine

97

whether psychopathic traits were positively or negatively related to functional
connectivity between the left/right amygdala and the bilateral VMPFC.

To assess whether psychopathic traits modulate the resting state connectivity of
right amygdala to other neural regions, or whether modulation is specific to
amygdala-VMPFC connectivity, we used two additional ROIs, the left amygdala
and right superior temporal gyrus (STG). It has been demonstrated previously
that both regions are functionally connected to right amygdala during resting
state fMRI (Roy et al, 2009). The STG ROI was created by expanding a 6-mm
sphere around the peak coordinates of right STG functional connectivity with
right amygdala (34, 12, -30) from Roy et al., 2009. For each of the ROIs, it was
first assessed whether right amygdala was indeed functionally connected to
these regions during resting state. Then, it was assessed whether psychopathic
traits moderated amygdala connectivity to these neural regions.

Next, Featquery in FSL was used to extract parameter estimates representing
connectivity values from VMPFC. These values were used as the dependent
variable in a multiple regression analysis, where psychopathic traits (composite
psychopathy score, PPI, or PPI-II subscales), trait anxiety (WAS), and an
interaction term of psychopathic trait scores and trait anxiety scores were used
as predictors. All predictors were centered to reduce multicollinearity issues
(Aiken & West, 1991). Creation of the interaction term involved the multiplication

98

of the two centered predictors. The interaction term investigates whether the
correlation between psychopathic trait scores and neural activity in the fronto-
limbic circuit is dependent upon trait anxiety levels. It was predicted that an
increase in psychopathic traits coupled with a decrease in trait anxiety would
result in a decrease in amygdala-VMPFC connectivity at rest. Lastly, parameter
estimates (as indices of mean connectivity) were extracted from each control ROI
(left amygdala, right STG) and included in similar multiple regression analyses. It
was predicted there would be no significant interactions between psychopathic
traits and trait anxiety for these connections. Correlations and multiple
regression analyses were performed with SPSS (Release Version 18.0, © SPSS,
Inc., 2009, Chicago, IL, www.spss.com).

Task-based Functional Connectivity Analysis
Parameters for the two fear conditioning functional scans are described in
Chapter 2 Materials and Methods.

Task-Based Connectivity Analysis
Psychophysiological Interaction (PPI) is a task-based functional connectivity
analysis. Using the timeseries from an a priori defined seed region of interest,
the analysis searches for voxels that have increased coherence with the
timeseries of the seed region during a given behavioral task.
Functional data analysis was performed using FEAT (FMRI Expert Analysis

99

Tool), part of FSL (FMRIB's Software Library, www.fmrib.ox.ac.uk/fsl). Standard
pre-processing steps were carried out on individual participants, and include;
interleaved slice timing correction, motion correction using MCFLIRT (Jenkinson
& Smith, 2001); non-brain removal using BET (S. M. Smith, 2002); spatial
smoothing using a Gaussian kernel of full-width-half-maximum 6mm; and high
pass temporal filtering (Gaussian-weighted least-squares straight line fitting, with
sigma = 50.0 s). The estimated motion parameters for each participant were
included as covariates of no interest to reduce spurious activations due to head
motion and scanner drift.
In order to investigate whether functional connectivity of amygdala-VMPFC
during fear conditioning is modulated by psychopathic traits, a generalized
psycho-physiological interaction was conducted (gPPI) (McLaren, Ries, Xu, &
Johnson, 2012; Mednick & Christiansen, 1977; O'Reilly, Woolrich, Behrens,
Smith, & Johansen-Berg, 2012). To accomplish this, anatomical regions of
interest (ROI) for bilateral amygdala were hand-drawn onto each subject’s high
resolution anatomical scan (Convit et al., 1999). Each ROI was then transformed
into the subject’s native functional space using a transformation matrix from a
previous FEAT analysis. ROIs were then used to extract amygdala timeseries
from the filtered functional data of each individual subject, which were then used
in the regression model as the physiological regressor. Additional regressors
from the original GLM model were also included in the model as psychological
regressors and were defined as follows: a) conditioned stimulus paired with

100

aversive stimuli (CS+; modeled as 5.8 s duration); b) unpaired conditioned
stimulus (CS-; modeled as 5.8 s duration); c) unconditioned stimulus for CS+
(UCS; modeled as 3 s duration); d) non-unconditioned stimulus for CS- (non-
UCS; modeled as 3 s duration). For each stimulus type, the stimulus timing files
were convolved with a double gamma hemodynamic response function to
produce an expected BOLD response. Each psychological regressor from the
original GLM model was mean centered and multiplied with the demeaned
amygdala timeseries to create four interaction regressors representing the gPPI.
At the individual level, to assess functional connectivity differences specific to
fear conditioning, contrasts between the interaction terms containing the CS+
and CS- conditions were created. Data was then fitted to the model using FSL’s
implementation of the general linear model. The resulting z statistic images were
thresholded at p < 0.001 (uncorrected), and warped into a standard space based
on the MNI-152 atlas (2 x 2 x 2 mm
3
Montreal Neurological Institute (MNI) brain)
using FLIRT (FMRIB’s Linear Image Registration Tool) (Jenkinson & Smith,
2001; Jenkinson et al., 2002). A second-level analysis to average across the two
runs was carried out using a fixed effects model, by forcing the random effects
variance to zero in FLAME (FMRIB’s Local Analysis of Mixed Effects) (Beckman,
Jenkinson, & Smith, 2003; Woolrich, Behrens, Beckmann, Jenkinson, & Smith,
2004; Woolrich, 2008). Mixed effects higher-level analysis for the participants
was carried out using FLAME 1 (FMRIB's Local Analysis of Mixed Effects)
(Beckmann et al., 2003). FLAME 1 uses both fixed effects and random effects

101

from cross session/subject variance (Beckmann et al., 2003; Woolrich et al.
2004; Woolrich, 2008). Z statistic images at this level were thresholded using
clusters determined by z > 2.3 and a corrected cluster significance threshold of p
< 0.05 (Jezzard, Matthews, & Smith, 2001). In order to determine whether
amygdala-VMPFC connectivity is modulated by psychopathic traits, composite
psychopathy scores or PPI scores (total and PPI-II subscales) were used as
covariates in the higher-level analysis. Additionally, to examine differences in
amygdala-VMPFC connectivity specifically, the ROI for bilateral VMPFC
(described above) was used to mask the connectivity analysis.

Previous research has indicated that other regions besides amygdala and
VMPFC are involved in fear conditioning, such as hippocampus, caudate, insula,
somatosensory cortex (see Sehlmeyer 2010 for review). Therefore, in order to
assess whether connectivity of the amygdala to other neural regions involved in
fear conditioning were influenced by psychopathic traits, a whole brain
connectivity analysis was also performed. To test whether trait anxiety would
moderate the relationship between psychopathic traits and amygdala connectivity
to the fear circuit during fear conditioning, Featquery from FSL was used to
extract values from ROI masks (of regions that did show modulation, which
includes right hippocampus and left caudate) of the higher-level analyses. ROIs
for right hippocampus and left caudate were defined anatomically from the FSL
Harvard-Oxford probabilistic subcortical atlas, with both regions thresholded at

102

70%. Connectivity values were used as the dependent variable in multiple
regression analyses in SPSS (Release Version 18.0, © SPSS, Inc., 2009,
Chicago, IL, www.spss.com), where psychopathic traits (total PPI score), trait
anxiety (WAS), and an interaction term of psychopathic trait scores and trait
anxiety scores were used as predictors. It was predicted that an increase in
psychopathic traits coupled with a decrease in trait anxiety would result in a
decrease in amygdala connectivity with the fear circuit during fear conditioning.

3.4 RESULTS
Resting State fMRI Functional Connectivity
Age was unrelated to amygdala-VMPFC resting state functional connectivity.
In order to assess functional connectivity between the amygdala and VMPFC at
rest, seed regions were defined in the left and right amygdala. Time series were
extracted from these two seed regions, and each time series was used in a
separate voxelwise analysis to assess whether VMPFC activity was significantly
correlated with the respective seed region’s activity. The left amygdala seed was
found to be positively correlated with regions in VMPFC (Table 3-1), whereas no
significant findings emerged for the right amygdala.
Table 3-1. Localization of clusters within VMPFC significantly functionally
connected to left amygdala during rest. Group-level random effects analyses,
thresholded at z > 2.3, p < 0.05, cluster corrected for multiple comparisons.
Coordinates [x y
z]
Brain Region Z-stat Cluster Size
[2mm
3
voxels]
Cluster Index
[-2 40 -18] L VMPFC 4.14 1296 2
[-8 40 -16] L VMPFC 4.12 - 2

103

Table 3-1. Localization of clusters within VMPFC significantly functionally
connected to Left Amygdala during Resting State (continued). Group-level
random effects analyses, thresholded at z > 2.3, p < 0.05, cluster corrected for
multiple comparisons.
Coordinates [x y
z]
Brain Region Z-stat Cluster Size
[2mm
3
voxels]
Cluster
Index

[-4 48 -6] L
Paracingulate
Gyrus
3.92 - 2
[-16 8 -22] L
Orbitofrontal
Cortex
3.89 - 2
[-8 30 -16] L Subcallosal
Cortex
3.88 - 2
[-2 44 -4] L
Paracingulate
Gyrus
3.87 - 2
[2 40 -18] R VMPFC 4.15 676 1
[18 8 -18] R
Orbitofrontal
Cortex
3.8 - 1
[2 42 -2] R Cingulate
Gyrus
3.67 - 1
[6 40 -4] R Cingulate
Gyrus
3.57 - 1
[2 50 -10] R VMPFC 3.42 - 1
[8 34 0] R Cingulate
Gyrus
3.33 - 1

Next, we used scores from the composite psychopathy scale, the PPI, and the
PPI-II subscales as separate covariates to assess whether psychopathic traits
modulate amygdala-VMPFC functional connectivity at rest. Supporting our
hypothesis, and replicating the results of Motzkin et al. 2011, we found that the
right amygdala had decreased functional connectivity with VMPFC as the total

104

PPI scores increased, r(23)= -0.53, p < 0.01 (see Table 3-2 and Figure 3-2 for
localization, and Figure 3-3)).

Table 3-2. Localization of clusters within VMPFC where right amygdala-
VMPFC connectivity is negatively correlated with total PPI scores. Group-
level random effects analyses, thresholded at z > 2.3, p < 0.05, cluster corrected
for multiple comparisons.
Coordinates [x y z] Z-stat Cluster Size
[2mm
3
voxels]
Cluster
Index
[0 56 -4] 3.35 2211 1
[12 50 -2] 3.29 - 1
[-4 46 -20] 3.26 - 1
[12 56 -6] 3.17 - 1
[-8 46 -20] 3.13 - 1
[2 36 -14] 3.11 - 1

Figure 3-2. Regions within ventromedial prefrontal cortex (VMPFC) that
show decreased resting state connectivity to right amygdala with increased
psychopathy scores. Psychopathy is indexed by total Psychopathic Personality
Inventory scores. Analysis masked by VMPFC ROI and thresholded at z > 2.3
and cluster corrected for multiple comparisons with spatial extent p < 0.05.

x"="$4" y"="48" z"="$20"
0" 1.6"

105

Figure 3-3. Total Psychopathic Personality Inventory (PPI) score negatively
correlates with right amygdala-ventromedial prefrontal cortex (VMPFC)
functional connectivity at rest. r(23)= -0.53, p < 0.01.

Due to the fact that it appeared the distribution of the connectivity (z) values may
have had an outlier at the higher end, we conducted a post-hoc analysis using
the outlier labeling rule (Tukey, 1977; Hoaglin, Iglewicz, 1987) and found this
value was indeed a statistically significant outlier. However, since this value
corresponds to the highest PPI score, it is difficult to determine whether it is a
true effect, or a measurement error, and will thus be kept in the remaining
analyses. The subject whose data was the outlier in this case is different from
the previously mentioned outlier subject in Chapter 2. In any case, when the
outlier was removed, the negative correlation persisted, albeit non-significantly,
r(22) = -0.12, p > 0.05.

-‐1.4

-‐1.2

-‐1

-‐0.8

-‐0.6

-‐0.4

-‐0.2

0

0.2

0.4

190
240
290
340
390

Right
Amygdala
-‐

VMPFC

functional
connectivity
(z)

Total
PPI
Score

106

In order to assess whether psychopathic traits modulated other connections of
the amygdala, or whether the effect was specific to right amygdala-VMPFC
connectivity, two other seeds that are known to be functionally connected to the
right amygdala at rest were investigated: the left amygdala and the superior
temporal gyrus (STG; Roy et al., 2009). The right amygdala was used as the
seed in an analysis that used the left amygdala as a mask. The voxelwise
analysis revealed that the left and right amygdala were indeed functionally
connected (Table 3-3), however the relationship was not modulated by any of the
psychopathic trait scores. A separate analysis used the right amygdala as a
seed, and used the right STG ROI as a mask. Confirming previous findings, the
right amygdala and right STG were functionally connected at rest, however (and
in agreement with Motzkin et al., 2011) (See Table 3-4), this connectivity was not
modulated by psychopathic traits.
Table 3-3. Localization of clusters within Left Amygdala where significantly
functionally connected to Right Amygdala during Resting State

Coordinates [x y z] Z-stat Cluster Size
[2mm
3
voxels]
Cluster Index
[-20 -4 -20] 5.1 291 1
[-24 0 -28] 3.78 - 1

Table 3-4. Localization of clusters within Right STG where significantly
functionally connected to Right Amygdala during Resting State

Coordinates [x y z] Z-stat Cluster Size
[2mm
3
voxels]
Cluster Index
[30 10 -28] 5.38 122 1
[34 16 -28] 4.84 - 1

107

Next, in order to assess whether trait anxiety could moderate the relationship
between amygdala-VMPFC connectivity at rest and psychopathic traits, multiple
regression analyses were conducted. Parallel analyses were conducted using
one of the psychopathy measures (composite psychopathy score, total PPI, or
PPI-II subscales) as one independent variable and the trait anxiety measure
(WAS) as the other independent variable, while using parameter estimates from
the VMPFC (indices of amygdala-VMPFC connectivity) as the dependent
variable. The raw and standardized regression coefficients of the predictors are
reported in Table 3-5. The model that used the total PPI score and the WAS to
predict right amygdala connectivity with VMPFC at rest was significant R
2
= 0.54
(adjusted R
2
=

0.48), F(3,20) = 7.94, p < 0.001 (Figure 3-4). It appears that
neither the PPI score, nor WAS, are independently associated with right
amygdala-VMPFC functional connectivity. The interaction term is significant,
however, indicating that the value of one predictor depends on the value of the
other. The beta value of the PPI x WAS interaction term is negative, indicating
that as psychopathic traits increase and trait anxiety decreases, there is an
increase in amygdala-VMPFC functional connectivity. No other regression
models were significant.

108

Table 3-5 Unstandardized and standardized coefficients and t-values for
the interaction between trait anxiety and psychopathic traits that predicts
amygdala-VMPFC functional connectivity during resting state
Unstandardized
Coefficients
Standardized
Coefficients

B SE β t
Right
Amygdala -
VMPFC
(Constant) 0.00 0.04 0.00
PPI -0.003 0.001 -0.30 -1.75
WAS 0.01 0.01 -0.22 -1.42
PPI * WAS 0.00 0.00 -0.52 -3.02*
Note. PPI = Psychopathic Personality Inventory; WAS = Welsh Anxiety Inventory;
* indicates significance at p < .01.

109

Figure 3-4. Trait anxiety interacts with total Psychopathic Personality
Inventory (PPI) score to predict resting state functional connectivity
between amygdala and ventromedial prefrontal cortex (VMPFC). The mean
of trait anxiety scores is represented by the solid black line. Dotted lines
represent trait anxiety scores 1 standard deviation (SD) above (red) or below
(blue) the mean. Solid colored lines represent trait anxiety scores 2 SD above
(red) or below (blue) the mean. F(3,20) = 7.94, p < 0.001

Functional Connectivity during Fear conditioning
The VMPFC and amygdala are known to be involved in fear conditioning from
animal studies (LeDoux, 2000), as well as human studies (LaBar), however the
functional connectivity of the two regions during fear conditioning in humans has
not been reported. Additionally, as the dysfunction of these two neural regions

110

are thought to be implicated in psychopathy, especially with regard to fear
conditioning (Hare, 1970, Birbaumer et al., 2005), it is necessary to examine how
psychopathic traits are related to the functional connectivity of the amygdala and
VMPFC during fear conditioning. The left and right amygdala were used as seed
regions in separate voxelwise functional connectivity analyses, which were
constrained to bilateral VMPFC with a bilateral VMPFC ROI mask. The
amygdala-VMPFC connectivity was compared between CS+ and CS- conditions.
There was no significant amygdala-VMPFC connectivity using left or right
amygdala for the CS+>CS- contrast.

Next, psychopathy scores (composite psychopathy score, PPI, PPI-II subscales)
were included in separate spatial map analyses as covariates to examine how
psychopathic traits are related to amygdala-VMPFC functional connectivity
during fear conditioning (CS+>CS-). Neither of the two total psychopathy scores
were found to modulate amygdala-VMPFC functional connectivity during fear
conditioning. However, when examining subscales from the PPI, a negative
relationship between the PPI subscale BE and right amygdala connectivity with
VMPFC (see Table 3-6 and Figure 3-5).
Table 3-6. Localization of clusters of amygdala-VMPFC connectivity during
fear conditioning that are negatively correlated with Blame Externalization.

Coordinates [x
y z]
Brain Region Z-stat Cluster Size
[2mm
3

voxels]
Cluster Index
[-6 52 -6] L VMPFC 3.62 289 2
[-12 60 -6] L Frontal Pole 3.14 - 2

111

Table 3-6. Localization of clusters of amygdala-VMPFC connectivity during
fear conditioning that are negatively correlated with Blame Externalization
(continued).
Coordinates [x
y z]
Brain Region Z-stat Cluster Size
[2mm
3

voxels]
Cluster Index

[-18 52 -6] L White
Matter/Frontal
Pole
2.44 - 2
[6 44 -4] R Paracingulate
Gyrus
3.05 98 1
[16 54 -6] R White
Matter/Frontal
Pole
2.73 - 1

Figure 3-5. Regions within ventromedial prefrontal cortex (VMPFC) that
show decreased resting state connectivity with right amygdala with
increased Blame Externalization scores. Blame Externalization is indexed by
the PPI-II subscale with the same name. Analysis masked by VMPFC ROI and
thresholded at z > 2.3 and cluster corrected for multiple comparisons with spatial
extent p < 0.05.

Next, in order to assess whether trait anxiety could be moderating the
relationship between amygdala-VMPFC functional connectivity during fear
0 1.6
x = -3 y = 48 Z = -7

112

conditioning and psychopathic traits, multiple regression analyses were
conducted. Parallel multiple regression analyses were conducted using one of
the psychopathy measures (composite psychopathy score, total PPI, or PPI-II
subscales) as one independent variable and the trait anxiety measure (WAS) as
the other independent variable, while using parameter estimates from the
VMPFC (indices of amygdala-VMPFC connectivity) as the dependent variable.
None of the regression models were significant, failing to show that trait anxiety
interacts with psychopathic traits to predict amygdala-VMPFC functional
connectivity during fear conditioning.

Finally, the left and right amygdala were used in a whole brain voxelwise
connectivity analysis to assess whether connectivity with other regions of the fear
conditioning circuit were observed, and subsequently modulated by psychopathic
traits and trait anxiety. Results from the whole brain analysis revealed that as
psychopathic traits (as assessed by the total PPI score) increased, the right
amygdala was less functionally connected to right amygdala/hippocampus and
left caudate/putamen, as well as inferior temporal gyrus, during fear conditioning
(See Table 3-7 and Figure 3-6). There was no modulation of left amygdala
connectivity with other fear conditioning neural regions by psychopathic traits.

113

Table 3-7. Localization of clusters of right amygdala connectivity with other
fear-conditioning related regions during fear conditioning that are
negatively correlated with total PPI score

Coordinates [x
y z]
Brain Region Z-stat Cluster
Size [2mm
3

voxels]
Cluster
Index
[24 -2 -22] R Amygdala 3.14 254 3
[30 4 -30] R White
Matter/Parahippocampal
Gyrus
3.08 -
[38 -6 -28] R White
Matter/Temporal
Fusiform Cortex
3.02 -
[36 8 -30] R White
Matter/Temporal Pole
2.94 -
[42 6 -24] R Temporal Pole 2.86 -
[12 4 -22] R Parahippocampal
Gyrus, anterior
2.73 -
[-46 -2 -32] L Inferior Temporal
Gyrus
3.35 203 2
[-54 10 -28] L Temporal Pole 2.49 -
[-44 -16 -26] L Inferior Temporal
Gyrus
2.46 -
[-56 -4 -32] L Middle Temporal
Gyrus
2.46 -
[-14 8 16] L Caudate 3.33 198 1
[-26 8 10] L Putamen 2.99 -
[-26 6 14] L White Matter/Putamen 2.96 -
[-6 6 10] L Caudate 2.59 -

114

Figure 3-6. Right amygdala functional connectivity during fear conditioning
to key fear conditioning regions modulated by total Psychopathic
Personality Inventory (PPI) scores. Whole brain analysis thresholded at z > 2.3
and cluster corrected for multiple comparisons with spatial extent p < 0.05.

To examine whether trait anxiety would interact with psychopathic traits to predict
amygdala functional connectivity to other neural regions involved in fear
conditioning, parameter estimates from right hippocampus and left caudate were
used in multiple regression models (similar to above). None of the regression
models were significant.

1.6 0
x = 25 y = 1
x = -15 Z = 15

115

3.5 DISCUSSION
In this study, psychopathic traits were found to modulate amygdala-VMPFC
resting state functional connectivity and that trait anxiety interacted with
psychopathic traits to predict the degree of amygdala-VMPFC connectivity. It
was also demonstrated that functional connectivity during fear conditioning
between amygdala and other neural regions of the fear circuit was also
modulated by psychopathic traits, but that the degree of connectivity during fear
conditioning was not modulated by trait anxiety. Each finding will be discussed in
turn.

Resting state functional connectivity
In agreement with previous findings, our results revealed that the left amygdala
was functionally connected to bilateral VMPFC during resting state (Roy et al.,
2009). However, the right amygdala did not show overall functional connectivity
to VMPFC during rest. This is likely due to the fact that right amygdala-VMPFC
functional connectivity during rest was negatively associated with psychopathic
traits (indexed by total PPI score). It should be noted that this association was
greatly reduced, and became statistically non-significant, when an extreme
outlier was removed from the analysis, although the negative trend persisted.
This result will need to be replicated in a larger sample size to ascertain validity.
However, the findings provide the first evidence of the modulatory effect of

116

psychopathic traits on amygdala-VMPFC resting state functional connectivity in a
community sample.

The modulation of resting state amygdala-VMPFC functional connectivity
supports previous results in criminals with psychopathy (Motzkin et al., 2011).
This study is the first to show reduced resting state amygdala-VMFPC coherence
in a community sample with a wide range of psychopathic traits. The results
extend previous findings from criminal samples in important ways, such as
demonstrating that neural differences due to psychopathic traits are not reserved
exclusive to criminal populations.

To assess the specificity of resting state connectivity differences to the right
amygdala-VMPFC circuit, we tested whether connectivity of the amygdala to
other neural regions was modulated by psychopathic traits. In accord with prior
research, left and right amygdala were significantly functionally connected at rest
(Roy et al., 2009), but were not modulated by psychopathic traits. Additionally,
functional connectivity between right amygdala and right superior temporal gyrus
was also significant during rest, but not modulated by psychopathic traits. Taken
together, these results confirm the specificity of the modulatory effect of
psychopathic traits on amygdala-VMPFC resting state functional connectivity.
While we cannot rule out disruption in other neural networks in psychopathy, the

117

findings once again highlight the importance of the amygdala-VMPFC circuit in
the personality disorder.

Role of Trait Anxiety in Resting State Connectivity
In the current study, the relationship between psychopathic traits and amygdala-
VMPFC resting state functional connectivity is dependent upon levels of trait
anxiety. In this sample, as psychopathic traits increased and trait anxiety
decreased, there was an increase in amygdala-VMPFC resting state functional
connectivity. This is the first study to take trait anxiety into account when
assessing psychopathic traits and amygdala-VMPFC functional connectivity in a
community sample, although the findings do support previous results from a
criminal sample. Motzkin and colleagues (2011) also found that trait anxiety and
psychopathic traits interacted in such a way that offenders with the highest levels
of psychopathic traits and lowest trait anxiety, had the strongest amygdala-
VMPFC functional connectivity.

Strong amygdala-VMPFC functional connectivity is thought to result in better
emotional appraisal and regulation (Bishop, Duncan, Brett, & Lawrence, 2004;
Ochsner et al., 2004). Thus, it is possible that increased resting state functional
connectivity between these two neural regions and the combination of high
psychopathic traits and low anxiety actually provide a calmer disposition due to
improved resting arousal control. Having a strong baseline arousal control fits
the notion of psychopaths being “cool, calm and collected”. Additionally,

118

increased amygdala-VMPFC functional connectivity at rest has been associated
with decreased trait anxiety, purportedly due to prefrontal control (M. J. Kim &
Whalen, 2009). Results garner support for the notion that individuals with higher
psychopathic traits and lower trait anxiety possess greater neural abnormalities
that may explain the observed differences in behavior (Koenigs et al., 2010).
The results also further support theories of abnormal neural functioning in
emotional regulation and behavioral control centers in psychopathy (Kiehl et al.,
2004; Motzkin et al., 2011)

Functional Connectivity during Fear conditioning
The amygdala and VMPFC are known to work together during fear conditioning
in assigning valence to stimuli, as well as regulating stimulus-response
associations and orchestrating fear responses (Antoniadis, Winslow, Davis, &
Amaral, 2009; Davis & Shi, 2000; Quirk & Beer, 2006). Both neural regions have
been found to be active during MRI studies of fear conditioning in humans
(Knight, Smith, Cheng, Stein, & Helmstetter, 2004; LaBar et al., 1998; Phelps et
al., 2004). In criminals with psychopathy, it has been demonstrated that both
amygdala and VMPFC display reduced activity during fear conditioning
(Birbaumer et al., 2005; Veit et al., 2002). However, amygdala-VMPFC
functional connectivity during fear conditioning has not been assessed in healthy
individuals, or in non-criminal individuals with psychopathy.

119

Using total psychopathy scores, as well as subscales from PPI-II, modulation of
amygdala-VMPFC functional connectivity during fear conditioning was assessed.
The PPI subscale Blame Externalization was found to be negatively correlated
with right amygdala-VMPFC functional connectivity during fear conditioning,
indicating that there was less connectivity between these two key fear
conditioning regions in individuals with higher psychopathic traits. Findings
support a previous study that demonstrated reduced amygdala-VMPFC
functional connectivity in a group of adolescents with psychopathic traits (Marsh
et al., 2008), as well as, more generally, multiple neuroimaging studies reporting
functional abnormalities in both amygdala and VMPFC in individuals with
psychopathy (Glenn et al., 2009; Gordon, Baird, & End, 2004; Juárez, Kiehl, &
Calhoun, 2012; Rilling et al., 2007; Veit et al., 2010). An advantage of using
psychopathy scale subscales is that one can begin to parse the neural correlates
of the many personality traits that comprise psychopathy. For example, Blame
Externalization involves self-other and self-referential processing (i.e. not being
able to personally take responsibility, and consistently externalizing it to others).
These psychological processes have been associated with medial prefrontal and
VMPFC cortices (Kelley et al., 2002; Northoff & Bermpohl, 2004), the same
regions displaying reduced functional connectivity to amygdala during a task in
the current study. Thus, findings provide further support for an association
between a particular psychopathic trait and neural region.

120

Furthermore, the right amygdala also displayed reduced functional connections
to other neural regions, such as the hippocampus/parahippocampal gyrus,
caudate, and middle temporal gyrus, some of which have been previously
implicated in fear conditioning (Carlsson et al., 2006; Jensen et al., 2003). These
results corroborate previous findings of reduced function in key fear conditioning
neural regions in offenders with psychopathy (Birbaumer et al., 2005; Veit et al.,
2002). The current findings also expand previous research by providing the first
evidence of altered fear conditioning related neural connectivity in association
with psychopathic traits in a community sample.

During fear conditioning, it appears that functional connections of the amygdala
are modulated by psychopathic traits, but that these relationships are not
moderated by trait anxiety in this community sample. Other work from the
current sample demonstrates that trait anxiety interacts with psychopathic traits
to predict amygdala-VMPFC resting state functional connectivity, as well as
amygdala functioning during fear conditioning. Taken together, the combination
of results highlights the importance of network interactions. It is possible that the
moderating effect of trait anxiety on neural function and psychopathic traits is
specific to the amygdala in the current community sample, and that this
moderation does not necessarily affect all functional connections of the amygdala
to other regions in a one to one manner. For example, in the current population,
individuals high on psychopathic traits and low on trait anxiety demonstrate

121

decreased amygdala functioning during fear conditioning, but increased
amygdala-VMPFC functional connectivity during resting state. If amygdala-
VMPFC functional connectivity would be driven by amygdala alone, one would
expect a decrease in functional connectivity. The fact that this is not the case,
underlines the importance of reciprocal amygdala-VMPFC network dynamics.
Thus, with regard to functional connectivity between the two neural regions, it is
possible that having one normally functioning region may compensate for an
underperforming counterpart. Furthermore, the amygdala is comprised of
approximately 13 nuclei (Amaral et al., 1992), and likewise the VMPFC has many
functionally distinct subregions (Wallis, 2007). The two neural regions are
densely interconnected, and future studies, especially with higher resolution
fMRI, will be needed to parse the relationship between amygdala-VMPFC
functional connectivity, psychopathic traits, and trait anxiety.

Alternatively, it may be that fear conditioning is not the most optimal task to
assess the relationship between psychopathic traits, trait anxiety, and task-based
amygdala-VMPFC functional connectivity. Results from the functional
connectivity contrast CS+>CS- support this fact and imply that the theorized
sustained amygdala-VMPFC connectivity in CS+ does not occur, or is not
detectable with the current method of analysis. Although amygdala and VMPFC
have been shown to work interactively during fear conditioning in animal studies

122

(Sotres-Bayon & Quirk, 2010), future studies may utilize tasks designed to better
target amygdala-VMPFC interactions, such as emotion regulation.

The lack of extremely high psychopathy scores in the community sample may
also contribute to the current findings. While PPI scores of the current sample
are comparable to offender populations (Lilienfeld et al, 2002), it is plausible that
individuals scoring in the extreme ranges of psychopathy scores are needed to
observe functional VMPFC differences.

In the past, it was difficult to make legitimate comparisons across studies,
populations, and acquisition modalities to draw conclusions about a given
construct, such as psychopathy. In this study, we use two different acquisition
modalities in the same population to assess differences in neural connectivity as
a function of a personality trait. Our results indicate that even within one
population, differences in neural connectivity due to personality traits vary across
tasks, highlighting the importance of network dynamics and interactions.

123

CHAPTER 4: Psychopathic traits modulate
microstructural integrity of right uncinate fasciculus

4.1 ABSTRACT
Individuals with psychopathy possess emotional and behavioral abnormalities.
Two neural regions, involved in behavioral control and emotion regulation, are
often implicated: amygdala and VMPFC. Recently, in studies using adult criminal
populations (mean ages around 30), reductions in microstructural integrity of the
white matter connections (i.e. uncinate fasciculus (UF)) between these two
neural regions have been discovered in criminals with psychopathy (Craig et al.,
2009), supporting the notion of neural dysfunction in the amygdala-VMPFC
circuit. In contrast, studies using adolescent populations with conduct disorder
(believed to be a precursor to psychopathy), have reported increased
microstructural integrity of UF in comparison to healthy controls (Passamonti et
al., 2012; Sarkar et al., 2013), raising the question of whether differences
between studies are due to neurobiological development. Furthermore, despite
evidence of possible subtypes of psychopathy that can be delineated based on
trait anxiety (Koenigs et al., 2010; Lykken, 1957), psychopathy is viewed as a
homogeneous construct and trait anxiety is often unaccounted. Here, a young
adult (mean age = 19), community sample is used to assess whether
psychopathic traits modulate microstructural integrity of UF, and whether this
relationship is dependent upon levels of trait anxiety. Results reveal a negative
association between psychopathic traits and microstructural integrity of UF,

124

similar to what has been described in adult populations. However, no moderation
of the relationship by trait anxiety was discovered. Findings provide further
support for the notion of altered amygdala-VMPFC connectivity in association
with higher psychopathic traits.

4.2 INTRODUCTION
Studies examining the association between psychopathic traits and the white
matter tract that connects the VMPFC to the amygdala (uncinate fasciculus; UF),
show a dichotomy between studies using adult populations, and those using
adolescent populations. Results from populations using adults with psychopathy
(typically with age means around 30 years) reveal decreased microstructural
integrity of the UF (Craig et al., 2009; Motzkin et al., 2011), whereas results from
populations of adolescents with conduct disorder (thought to be a precursor of
psychopathy) report increased values for microstructural integrity (Passamonti et
al., 2012; Sarkar et al, 2013). The difference in findings between adult and
adolescent populations prompts the query of whether the neurobiological profile
of psychopathic traits changes over development. For example, increased
microstructural integrity of adolescent populations with high psychopathic traits
may represent a lack of typical developmental neural pruning, whereas the
abnormal reduction of microstructural integrity in adulthood could indicate a
delayed, over-pruning of white matter pathways. Alternatively, the difference in

125

criteria between the two personality disorders may be the underlying reason for
the dichotomy of study results.

Another inadequately addressed issue is that of possible differences between
neurobiological profiles of subtypes of psychopathy. As stated before, the
literature suggests that subtypes of psychopathy exist and may be parsed using
low and high levels of trait anxiety. It is theorized that low-anxious individuals
with psychopathy may have greater neural abnormalities, based on reports of no
behavioral differences between this subtype and certain brain lesion patients
(VMPFC) on certain laboratory tasks, such as economic decision-making
Koenigs, Kruepke, & Newman, 2010. Motzkin and colleagues (2011) published
the only study to examine microstructural integrity of UF in low-anxious and high-
anxious male inmates with psychopathy, and found no significant differences
between the two groups. However, interpretation of the results is limited due to
the small sample size of 7 individuals per group.

Here, previous findings are expanded to examine the relationship between
psychopathic traits and microstructural integrity of the uncinate fasciculus, as
measured with diffusion tensor imaging (DTI), in a community sample of young
adults with a wide range of psychopathic traits. The age range of our population
(18-21 yrs) situates the group between adolescence and older adult criminal
populations, allowing an examination of the time window during continued neural

126

development. Additionally, a multiple regression analysis is used to examine
whether trait anxiety can moderate the relationship between psychopathic traits
and microstructural integrity of UF in such a way that the highest scorers on
psychopathy, but least anxious individuals possess the lowest microstructural
integrity of the UF.

4.3 MATERIALS AND METHODS
The population, behavioral procedures, and fear conditioning paradigm for the
analyses proposed here are the same as described in Chapter 2 Materials and
Methods.
Scanning Procedure
DTI Acquisition
A 64-direction diffusion tensor imaging sequence was implemented using parallel
imaging acceleration (GRAPPA), factor = 2. A total of 64 diffusion-weighted
images were acquired (axial slices = 60; TR/TE = 10000/88ms; FOV = 256mm;
b-value = 1000 s/mm2; in-plane resolution = 2x2mm2; slice thickness = 2mm).

DTI analysis
Preprocessing of diffusion data was performed using FSL (FMRIB's Software
Library, www.fmrib.ox.ac.uk/fsl). Preprocessing included: skull-stripping to
remove non-brain tissue, eddy current correction, head movement correction,
and realignment of diffusion-weighted data using the mean of the B0 volumes as

127

a reference. Diffusion tensor calculation and tracking was performed with
Diffusion Toolkit 0.6.1 and TrackVis 0.5.1 (http//trackvis.org). Tracts were
reconstructed by using a fiber assignment by continuous tracking (FACT)
algorithm and an angular threshold of 35º. A DWI (diffusion-weighted image)
mask was used to remove CSF, and an FA threshold of 0.02 was used for tract
reconstruction. Tracts were isolated in all subjects by a research assistant who
was blind to psychopathy score. The manual “obligatory passage” two-ROI
approach (Catani & Thiebaut de Schotten, 2008) was used to isolate the fibers of
the UF. With this method, one ROI was drawn in the anterior temporal lobe, with
the other being defined around the white matter of the anterior floor of the
external/extreme capsule. Each resulting set of tracts was edited by exclusion
ROIs to ensure tracts reliably represented the characteristic “c” shape of UF
fibers (see Figure 4-1) (Wakana et al., 2007).

128

Figure 4-1. Example of a reconstructed uncinate fasciculus (UF) tract using
diffusion tensor imaging.

In order to isolate the inferior longitudinal fasciculus (ILF) as a control tract, a
third ROI was drawn in the occipital lobes according to a previously detailed
procedure (Catani, 2008). Briefly, an ROI was drawn in the occipital lobes on 13-
15 axial slices, and ILF was defined as tracts that only pass through both the
occipital and temporal ROIs (see Figure 4-2).

Uncinate Fasciculus (UF)
Prefrontal
Cortex
Temporal
Cortex

129

Figure 4-2. Example of a reconstructed inferior longitudinal fasciculus
(ILF).

ROIs were drawn on the participant’s high resolution T1 anatomical scan for
more precise accuracy. The T1-weighted scan was manually aligned with the
participant’s b0 image using FSL’s Nudge. After virtual alignment, FSL’s FLIRT
(Jenkinson et al., 2002) was used to register the T1 image to diffusion space
using an affine registration with 12 degrees of freedom. Using the tracts derived
from the ROI masking, FA values for the UF from both hemispheres were
calculated by averaging the FA values of all voxels of the tract.

Temporal
Cortex
Occipital
Cortex
Inferior Longitudinal Fasciculus
(ILF)

130

The relationship between psychopathic traits (composite psychopathy score, PPI
or PPI-II subscales) and UF values was first assessed using correlation
analyses, which are typically used in psychopathy studies. To test whether trait
anxiety moderates the relationship between amygdala-VMPFC microstructural
integrity and psychopathic traits, the FA values were used as the dependent
variable in multiple regression analyses in SPSS (Release Version 18.0, ©
SPSS, Inc., 2009, Chicago, IL, www.spss.com). In the multiple regression model,
psychopathic traits (composite psychopathy score, PPI, or PPI-II subscales), trait
anxiety (WAS), and an interaction term of psychopathic trait scores and trait
anxiety scores were used as predictors. All predictors were centered to reduce
multicollinearity issues (Aiken & West, 1991). Creation of the interaction term
involved the multiplication of the two centered predictors. The interaction term
investigates whether the correlation between psychopathic trait scores and FA
values of the UF is dependent upon trait anxiety levels. It was predicted that an
increase in psychopathic traits coupled with a decrease in trait anxiety would
result in a decrease in microstructural integrity of the UF, as measured through
FA values.

4.4 RESULTS
Sample Characteristics
Sample characteristics are reported in Chapter 2 Results.

131

DTI Results
Previous results have shown reduced fractional anisotropy (FA) values in the
uncinate fasciculus (UF, a white matter tract that connects the amygdala to the
VMPFC) in criminal psychopaths (Craig et al., 2009), but increased FA in
adolescents with conduct disorder. Left and right UF ROIs were used to extract
mean FA values to assess whether they are modulated by psychopathic traits in
a community sample with a wide range of psychopathic traits. Separate
correlation analyses were used for the two psychopathy measures (composite
psychopathy scale, PPI), as well as the PPI-II subscales, with the left and right
FA values of the UF. The composite psychopathy score was found to be
negatively correlated with FA values of the right UF, r(23) = -0.52, p < 0.01 (see
Figure 4-3). No other correlations with PPI-II subscales were significant.

132

Figure 4-3. Psychopathy scores from ages 14-16 negatively correlate with
right uncinate fasciculus fractional anisotropy (FA) values. PsychoComp =
composite psychopathy score. r(23) = -0.52, p < 0.01.

Given that the composite psychopathy score and total PPI score are only
significantly correlated when the two individuals who switched levels of
psychopathy are removed, a second correlation was run between right UF values
and total PPI score after removing these two subjects. In this case, total PPI
score and right UF also showed a significant negative correlation, r(21) = -0.43, p
< 0.05 (Figure 4-4), while the original correlation between composite
psychopathy score and right UF remained significant, r(21) = -0.40, p < 0.05.

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

-‐2
-‐1.5
-‐1
-‐0.5
0
0.5
1
1.5
2

Right Uncinate Fasciculus FA
Values
PsychoComp

133

Figure 4-4. Adult psychopathy scores negatively correlate with right
uncinate fasciculus fractional anisotropy (FA) values when psychopathy
score switchers are removed. PPI = Psychopathic Personality Inventory. r(21)
= -0.43, p < 0.05.

Psychopathic traits were not correlated with FA values from the left or right ILF
(with or without the two individuals who switched psychopathy levels), indicating
that differences in microstructural integrity associated with psychopathic traits do
not extend to all white matter tracts of the brain.

In order to assess whether trait anxiety would moderate the relationship between
FA values of the UF and psychopathic traits, multiple regression models were
conducted. FA values of the left and right UF were used as dependent variables
in separate multiple regression analyses, with psychopathic traits (composite
psychopathy score, PPI, PPI-II subscales) and trait anxiety (WAS) as predictors.
None of the regression models was significant.
200

220

240

260

280

300

320

340

0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5

Right
Uncinate
Fasciculus
FA

values

PPI

134

4.5 DISCUSSION
In accordance with the adult psychopathy literature, a negative correlation was
seen between psychopathic traits (indexed by the composite psychopathy score)
and microstructural integrity (indexed by FA value) of the right uncinate
fasciculus (UF), the white matter tract connecting the amygdala and ventromedial
prefrontal cortex. Furthermore, the possible anatomical specificity of the finding
was supported by the fact that psychopathic traits did not modulate the control
tract (inferior longitudinal fasciculus).

As mentioned previously, the current study is novel in that the narrow age range
of participants (18-21) falls between previous adolescent and adult samples.
This provides the opportunity to examine the unique window of neural
development, which has been identified as a period of continued prefrontal
growth and pruning (Bourgeois, Goldman-Rakic, & Rakic, 1994; Gogtay et al.,
2004; Huttenlocher, 1979; Zecevic & Rakic, 2001). Results indicate that with
increases in psychopathic traits, microstructural integrity profiles of young adults
more closely resemble that of adults with psychopathy, rather than that of
adolescents with conduct disorder. If differences between previous studies are
due to developmental trajectories, the present study provides evidence that by
young adulthood, the relationship between microstructural integrity of UF and
psychopathic traits is not any more positively correlated as in younger subjects,

135

but rather negatively correlated as in adulthood. The best way to assess the
neural development in relation to psychopathic traits would be, of course, to use
a longitudinal experimental design.

It is possible that the reason the results of the current study are more similar to
adult criminal population studies is because there is a difference in the criteria
used to define each of the personality disorders. Adult criminal population
studies have used adult psychopathy measures (e.g. Psychopathy Checklist –
Revised (PCL-R), Hare, 1990) to assess levels of psychopathy, while adolescent
population studies have used definitions of conduct disorder to define groups.
Conduct disorder is a psychiatric disorder (as defined by DSM-IV (Association,
1994)) that includes such traits as aggressiveness and antisocial behavior. The
definition of conduct disorder does not include the interpersonal and affective
facets that underlie psychopathy. Typically, studies that wish to compare
individuals with conduct disorder to individuals with psychopathy will specifically
do an additional assessment of callous-unemotional traits, which represent the
affective-interpersonal component of psychopathy. The current study used an
adult psychopathy measure, which includes assessment of callous-unemotional
traits. The amygdala-VMPFC circuit is involved in emotional learning, emotion
regulation and behavioral control (LeDoux, 2007; Ochsner, Bunge, Gross, &
Gabrieli, 2002; Ochsner et al., 2004). Thus, dysfunction of white matter
pathways in the circuit may lead to emotion related deficits that are not

136

encompassed by measures of conduct disorder, but are encompassed by
measures of psychopathy. Passamonti and colleagues (2012) do assess
callous-unemotional traits, but the conduct disorder and control groups do not
seem to differ on levels of these traits. This suggests that observed neural
differences may be due solely to the aggressive nature of conduct disorder,
which is typically captured by the antisocial-impulsivity factor of psychopathy.
This difference may explain the discrepancy in brain imaging results of UF
between conduct disorder studies, and those using individuals with psychopathy.
It is to be expected that the findings of the current study match those of other
studies assessing psychopathy because these studies capture the callous-
unemotional traits, in addition to the antisocial behaviors.

The current study differs from previous studies in that a community sample is
utilized, rather than offender samples. Our results extend previous findings by
demonstrating a negative relationship between psychopathic traits and
microstructural integrity of UF in a non-offending, non-incarcerated population.
They suggest, once more, that to understand the neurobiological underpinnings
of psychopathy, it is important to investigate both incarcerated and non-
incarcerated populations, as they are qualitatively different and can provide
unique insights to the development of this personality disorder (Raine, 1993).

137

The current study also adds to previous knowledge by demonstrating that
psychopathic traits are relatively stable across the development from
adolescence, into young adulthood. Furthermore, it shows that psychopathy
scores in younger subjects can be used to predict microstructural integrity of UF
in young adulthood. Some amount of variation in scores should be expected
over time, and was observed in the current sample in the two individuals who
switched levels of psychopathic traits between adolescent and adult psychopathy
assessments. However, when the two individuals are removed from the
analysis, the negative relationship between total adult psychopathy score
(indexed by total PPI score) and microstructural integrity of UF is maintained,
strongly suggesting stability throughout development.

Role of Trait Anxiety
The relationship between psychopathic traits and microstructural integrity of UF
was not moderated by trait anxiety in the current study, counter to our
predictions. Motzkin and colleagues (2011) also investigated differences in
microstructural integrity of UF between subtypes of psychopathy in offenders
based on trait anxiety, but did not find significant differences between low-
anxious and high-anxious offenders with psychopathy. Thus, although trait
anxiety has been demonstrated to interact with psychopathic traits to predict
amygdala-VMPFC function (see Chapter 2) and functional connectivity (see
Chapter 3; Motzkin et al., 2011), it is possible that trait anxiety does not moderate

138

the relationship between amygdala-VMPFC structural connectivity and
psychopathic traits. However, it should also be noted that the offender study
mentioned utilized a small sample size of only 7 individuals per group, and that
the current study, on the other hand, does not include individuals with extremely
high scores of psychopathy, thus limiting the interpretation of results for both
studies. It is possible that larger sample sizes and the inclusion of subjects with
higher psychopathy scores may reveal the moderating effect of trait anxiety on
psychopathic traits and neural structure.

The normal evelopmental trajectory of the uncinate fasciculus begins with
increasing fractional anisotropy (FA) values in early childhood, continues through
adolescence and young adulthood (Lebel, Walker, Leemans, Phillips, &
Beaulieu, 2008; Mabbott, Rovet, Noseworthy, Smith, & Rockel, 2009; Mukherjee
& McKinstry, 2006), and declines after young adulthood (Lebel, Caverhill-
Godkewitsch, & Beaulieu, 2010). Factors such as premature birth (Constable et
al., 2008) and early socioemotional deprivation have been shown to be
associated with reduced FA values of UF (Eluvathingal et al., 2006). Given the
role of amygdala-VMPFC connections in emotion regulation and behavioral
control, it is plausible that disruption of this pathway early in life results in
increased antisocial behavior and affective deficits. In the future, studies that
examine differences in upbringing and early childhood experience could help

139

explain microstructural integrity differences in the UF, and parse how these relate
to psychopathic traits.

A few limitations of the study should be noted. FA values are determined by
multiple neuronal properties and should be considered an indirect measure of
“microstructural integrity”. Multiple factors contribute to the FA value, such as
degree of myelination and cellular arrangement around axons (Assaf &
Pasternak, 2008), thus it is difficult to say for certain what is the specific cause of
a difference in FA value. However, while the underlying cellular mechanisms
may remain unknown, meaningful differences may still be extracted from
diffusion data, such as differences between groups. Additionally, while our
recruitment efforts attempted to target more extreme scores, the final sample did
not reach the highest scores of the entire community sample, and thus fails to
represent the possible entire community range. In the future, studies should
include more extreme scores, especially in the higher range. Lastly, it is
important to bear in mind that the UF undergoes substantial continued
development in young adulthood (the age range of the current sample) (Lebel &
Beaulieu, 2011). Therefore, more research is required, especially with larger
sample sizes, to elucidate the true relationship between developmental trajectory
of white matter tracts, psychopathic traits, and trait anxiety.

140

In conclusion, the current study provides further evidence supporting a negative
association between psychopathic traits and microstructural integrity of the
uncinate fasciculus. The findings add an age group of particular developmental
importance to the literature, and highlight the need for longitudinal studies of the
relationship between psychopathic traits and neural structure.

141

CHAPTER 5. Discussion

Some researchers have argued that if subtypes of psychopathy (based on levels
of trait anxiety) are taken into account, a clearer picture of the neural
underpinnings of the disorder will emerge. The current work starts to address
this issue in that it is the first to examine in a community sample the relationship
between psychopathic traits, trait anxiety, and a neural circuit thought to be
involved in psychopathy (i.e. amygdala-VMPFC circuit) using multiple
neuroimaging techniques. Previous neuroimaging studies have focused on
simple correlation analyses between psychopathic traits and neural properties,
with a majority (although not all) reporting decreased function, structure or both
functional and structural connectivity. The body of work in this dissertation
supports these previous findings, but also indicates that the relationship between
psychopathic traits and some (but not all) neural properties are dependent upon
trait anxiety. This data highlights the importance of using multiple analyses
techniques to analyze neural circuits, as they, and their relations to personality,
are extremely complex.

The studies described in this dissertation used a community population with a
wide range of psychopathic trait scores to examine the relationship between
neural properties of the amygdala-VMPFC circuit, psychopathic traits, and trait
anxiety. The overarching goals were two-fold: (1) to address whether trait

142

anxiety helps explain the way psychopathic traits and neural properties are
related to each other; and (2) whether the relationship between psychopathic
traits, trait anxiety and amygdala-VMPFC could be evidenced across different
types of neural properties (e.g. structure, function). The results in relation to the
first goal would allow better distinction between subtypes of psychopathy. The
second goal was achieved through the use of multiple neuroimaging methods to
assess various characteristics of the amygdala-VMPFC circuit (i.e. function,
functional connectivity, microstructural integrity). The results support the notion
that trait anxiety is necessary to fully understand how psychopathic traits and the
brain (as well as physiological responses) are related to each other.
Furthermore, the consistency of findings across the studies underline the
importance and usefulness of the use of multiple neuroimaging modalities to
elucidate the relationship between psychopathy and the amygdala-VMPFC
circuit.

In Chapter 2, we describe that BOLD signal from bilateral amygdala during a fear
conditioning task was negatively correlated with a key psychopathic trait,
Machiavellian Egocentricity, in our community population. This finding supports
previous studies of criminals with psychopathy which demonstrated reduced
activity in the amygdala-VMPFC circuit during fear conditioning (Birbaumer et al.,
2005, Veit et al., 2002). However, counter to these previous findings, our study,
described in Chapter 2, showed no correlation between psychopathic traits and

143

VMPFC. We also found that during fear conditioning, trait anxiety interacted with
Machiavellian Egocentricity to predict BOLD responses in the right amygdala
(and a trend toward significance in the left amygdala). Finally, Chapter 1
revealed a negative correlation between skin conductance responses during fear
conditioning and two other psychopathic traits, Blame Externalization and
Carefree Nonplanfulness, with both these relationships being moderated by trait
anxiety.

In Chapter 3, we show that amygdala-VMPFC resting state functional
connectivity is negatively correlated with psychopathic traits (total score),
although this relationship is moderated by trait anxiety. In other words,
individuals with the highest psychopathy scores and lowest trait anxiety showed
the highest functional connectivity between amygdala and VMPFC. These
results support previous findings from a criminal population (Motzkin et al., 2011).
Functional connectivity of amygdala and VMPFC during fear conditioning was
also modulated by one of the PPI subscales, Blame Externalization. Lastly,
during fear conditioning, connectivity of the right amygdala with other key fear
conditioning neural regions was negatively correlated with psychopathic traits
(total score). This is a novel finding, that in fact supports previous work indicating
reduced fear circuit activation during fear conditioning in criminals with
psychopathy (Birbaumer et al., 2005, Veit et al., 2005).

144

Finally, in Chapter 4, we describe that psychopathic traits (composite score)
obtained at ages 14-16 were negatively correlated with microstructural integrity of
the right uncinate fasciculus, the white matter tract connecting amygdala and
VMPFC. These results support previous findings from adult psychopathy
populations indicating reduced uncinate fasciculus integrity in criminals with
psychopathy (Craig et al., 2009, Motzkin et al., 2011).

Results are summarized in a schematic fashion in Figure 5-1.

145

Figure 5-1. Summary of main findings. Locations of amygdala and ventromedial
prefrontal cortex (VMPFC) in relation to each other are mapped on the model brain (top),
with arrows illustrating the bidirectional functional and structural interconnections. The
box contains a schematic of results. Connections between the two regions are
represented by bidirectional arrows with text indicating the neuroimaging modality used
to assess connectivity (except the connection between amygdala and fear conditioning
circuit which corresponds to connectivity during fear conditioning). Orienting skin
conductance response (SCR) results are also represented (the black arrow indicates
amygdala’s involvement in SCR generation, but is not a result of the current work).
Items highlighted in blue outlines were negatively modulated by psychopathic traits.
!P!($)!
!P!+!Anx!(+)!
!P!+!Anx!($)!
VMPFC Amygdala
Resting State Cxn
SCR$
Fear Conditioning Cxn
Structural Cxn
Fear$Condi-oning$Circuit$
(hippocampus,$parahippocampal$
gyrus,$caudate)$

146

Items highlighted in a Figure 5-1. Summary of main findings (continued). red outline
have their relationships with psychopathic traits moderated by trait anxiety. Amygdala
function and SCRs during fear conditioning were both predicted by an interaction
between psychopathic traits and trait anxiety, such that individuals with the highest
psychopathy scores and lowest anxiety had the least amygdala activity and SCRs
(indicated by ‘-‘ in legend). Amygdala-VMPFC functional connectivity during resting
state was also predicted by an interaction between psychopathic traits and trait anxiety,
but such that individuals with the highest psychopathy scores and lowest anxiety had the
greatest connectivity (indicated by ‘+’ in legend). Amygdala-VMPFC functional
connectivity during fear conditioning and microstructural integrity (“structural cxn” in the
figure) were negatively modulated by psychopathic traits. Finally, amygdala functional
connectivity to fear conditioning circuit regions was also negatively modulated by
psychopathic traits. P = psychopathic traits; P + anx = Psychopathic traits and trait
anxiety; Cxn = connectivity.

Results from the current work illustrate the usefulness of community samples in
examining the neural correlates of psychopathy. However, it should be noted
that in our sample, the predicted association between VMPFC and psychopathic
traits was not observed. One possible explanation for this is that only individuals
with extremely high psychopathy scores (not available in the current population)
possess abnormalities in VMPFC. A direct comparison with other studies using
task-based fMRI to investigate psychopathic traits in community samples is
difficult due to differences in tasks, psychopathy assessment and reporting
practices for means and ranges of scores (Bjork et al., 2012; Harenski, Kim, &
Hamann, 2009; Marsh & Cardinale, 2012; Sheng, Gheytanchi, & Aziz-Zadeh,
2010). Comparisons with fMRI studies of fear conditioning in criminal samples
are also difficult due to substantial differences in population characteristics. We
thus conclude that further studies are needed to examine the relationship
between VMPFC and psychopathic traits in community samples.

147

The benefit of using multiple different neuroimaging techniques in one single
population is that it provides a more complete picture of a given neural network’s
structure and dynamics and its interactions with measures of psychopathy. For
example, in the current work, the function, functional connectivity, and
microstructural integrity of the amygdala-VMPFC circuit were analyzed. Despite
the reference in the psychopathy literature to the modulatory effect of
psychopathic traits on the amygdala-VMPFC circuit, described in general as
negative, results of the current analyses reveal a more nuanced story. The
present work demonstrates that individuals with the highest psychopathic traits
and lowest trait anxiety display decreased function in amygdala during fear
conditioning, increased amygdala-VMPFC functional connectivity during resting
state and no effect on functional connectivity during fear conditioning. As
mentioned in the discussion of Chapter 3, the combined results highlight the
importance of remembering that networks are comprised of multiple connections,
and that “dysfunction” of a network may not apply to all of its connections. Both
amygdala and VMPFC are complex structures with multiple connections between
each other, as well as with other brain regions. For example, it is possible that
an abnormality in one region of the amygdala, in this population, is related to a
particular psychopathic trait increase, and that connections to and from the
region might also be altered, but that connections with other regions of the
amygdala may be normal and fully functional. Through the extensive animal

148

work on fear conditioning, we know that the basolateral nucleus of the amygdala
is necessary for fear conditioning (Falls, Miserendino, & Davis, 1992; M. Kim,
Campeau, Falls, & Davis, 1993), and thus we can speculate that deficits in
acquisition of fear conditioning may stem from basolateral abnormalities. Future
work, possibly with higher resolution fMRI, will need to parse these fine
connections and their relationships with psychopathic traits and trait anxiety.

Many of the results from recent studies implicate abnormal function, functional
connectivity and microstructural integrity in the right hemisphere. In the
psychopathy literature, laterality effects of amygdala function and structure are,
however, inconclusive. In a metanalysis of neuroimaging data addressing the
laterality of amygdala activation, the left amygdala was found to be more active
during emotional processing (Baas, Aleman, & Kahn, 2004), the opposite of what
was found in this dissertation.

Psychopathy is a multifaceted disorder, being comprised of about 20 identifiable
personality traits. Despite the complex nature of the disorder, studies typically
use total psychopathy scores to explore neural correlates of the disorder. While
there have been groundbreaking studies identifying a couple of key neural
regions in the expression of the disorder, they have not provided a definitive
neurobiological profile for psychopathy. It would be beneficial to the
understanding of psychopathy to begin investigating the neural correlates of

149

individual personality traits that comprise the disorder. In the present work, we
found informative relationships between subscales of the Psychopathic
Personality Inventory and the neural circuits related to them. For example,
Blame Externalization was negatively correlated with amygdala-VMFPC
functional connectivity during fear conditioning. This trait is likely related to self-
other and self-referential processing, which has been associated with activity in
medial prefrontal and ventromedial prefrontal cortex (Kelley et al., 2002; Northoff
& Bermpohl, 2004). In the future, more studies should expand this work and
investigate the neural correlates of individual traits of psychopathy.

The present work targets the amygdala-VMPFC circuit as it is the most
commonly implicated neural circuit in discussions on psychopathy. However, the
brain has a large number of neural networks, and it is unlikely that deficits of the
amygdala-VMPFC circuit would be restricted to this circuit alone (as is evident by
the disruption of the fear conditioning circuit in the present work). The two main
models of psychopathy both include the amygdala-VMPFC circuit. James Blair’s
model of psychopathy, in fact, only accounts for functional and structural
abnormalities in these two neural regions (Blair, 2008). Kent Kiehl’s model of
psychopathy extends the network to a “paralimbic” circuit, including cingulate,
insula, parahippocampal gyrus, and superior temporal cortex (see Figure
2)(Kiehl, 2006). Amygdala and VMPFC are the two most studied regions with
respect to psychopathy, and, even so, their roles remain elusive. In the future,

150

studies should systematically look at network connections between these two
neural regions and with other regions, across populations and tasks to better
elucidate the possible or probable network profile of psychopathy.

Figure 5-2. Other neural regions possibly implicated in psychopathy. Other
regions of the “paralimbic” circuit, which are involved in sensory and emotional
processing, are indicated above, and include insula (blue), parahippocampal
gyrus (pink), and superior temporal gyrus (orange). Ins = insula; pHIP =
parahippocampal gyrus; STG = superior temporal gyrus.

Although psychopathy has garnered much attention in the scientific world
recently, many questions remain unanswered. It is a complex personality
disorder comprised of a constellation of traits and behaviors. A systems level
approach is necessary, as is the case with many psychiatric disorders, to
Ins$
Ins$ Ins$
STG$
pHIP$$
y$=$'7$ x$=$41$
STG$
pHIP$$

151

deconstruct and identify its contributing factors. There are many levels at which
the construct can be approached, such as molecular, genetic, neurological, or
environmental. For example animal studies have demonstrated that early
environmental stress (e.g. maternal deprivation) can cause changes in various
neurotransmitter systems (e.g. serotonin and dopamine) that persist into
adulthood, and are sometimes associated with aggressive behavior (Kraemer,
Ebert, Schmidt, & McKinney, 1989). In human research it has been suggested
that some psychopathic personality traits have about 50% heritability (Bezdjian,
Raine, Baker, & Lynam, 2011; Waldman & Rhee, 2007). However, environment
is also an important factor as shown in one study in which children who have
birth complications, as well as early maternal rejection (at one year of age), are
more likely to be convicted of violent crimes by the time they are 19 years old
(Raine, Brennan, & Mednick, 1994). Recent work from the USC Twin Study
project also demonstrates the importance of environment-biology interactions as
psychopathic traits were found to moderate the relationship between parental
affect and aggression (Yeh, Chen, Raine, Baker, & Jacobson, 2011). One
optimal way to systematically examine many of these factors is to use
longitudinal study designs that include neural assessments. In conclusion, more
research is needed at every level to achieve a complete picture of psychopathy.

Implications of the present work, and of psychopathy research in general, include
improvement in the development of tailored treatment plans for individuals with

152

this condition. Psychopathy is currently viewed as an untreatable disorder and
part of the reason for this dismal view is the fact that the neurobiology of the
disorder is not completely understood. As a better understanding of precise
deficits emerge, behavioral programs can be created explicitly to address the
issues. Determining whether subtypes of psychopathy exist, and what
distinguishing features can be used to separate the groups, will also deeply
benefit development of treatment programs, as each subtype is likely to need a
different kind of approach. A neuroscientific understanding of psychopathy also
has implications for the legal system, as evidence of brain based deficits brings
into question the responsibility of actions which the perpetrator may not
understand from a general moral perspective. The argument of “moral insanity”
would not likely stand in a court of law currently, however, who is to say it would
not in the future if more supporting behavioral and neural evidence is uncovered
to justify such a position.

This is a legal question, which is well beyond the scope of this dissertation. A
more practical question is what to do with such people? Previous research has
shown that prison is not a permanent solution as individuals with psychopathy
are particularly adept at manipulating parole boards into early release (Hare,
1990). A possible solution would be an institution where the aforementioned
personalized behavioral treatment programs could be implemented. However,
there is still a great deal of work to be done before reaching solutions, but it is

153

obvious that the implications for the understanding of this personality disorder are
far reaching.

Finally, psychopathy is interesting to study not only because it is difficult to
comprehend the motives and processes behind the oftentimes destructive
crimes, but also because this condition serves as a model for what happens
when emotional dysfunctions occur. Their confounding behavior stems from,
presumably, some neurobiological deficit that when completely understood will
help the further understanding of emotional processing mechanisms. However,
the problem must be addressed from a multifaceted systems level approach and
neuroimaging is just one tool, albeit a necessary one, to aid in the search.

154

REFERENCES
Aiken,
L.
S.,
&
West,
S.
G.
(1991).
Multiple
Regression:
Testing
and
interpreting

interactions.:
Sage
Publications,
Inc.

Amaral,
D.
G.,
&
Dent,
J.
A.
(1981).
Development
of
the
mossy
fibers
of
the
dentate

gyrus:
I.
A
light
and
electron
microscopic
study
of
the
mossy
fibers
and
their

expansions.
J
Comp
Neurol,
195(1),
51-‐86.
doi:
10.1002/cne.901950106

Amaral,
D.
G.,
Price,
J.
L.,
Pitkanen,
A.,
&
Carmichael,
S.
T.
(1992).
Anatomical

Organization
of
the
Primary
Amygdaloid
Complex.
In
J.
P.
Aggleton
(Ed.),
The

amygdala:
neurobiological
aspects
of
emotion,
memory,
and
mental

dysfunction
(pp.
1-‐66).
New
York:
Wiley-‐Liss.

Aniskiewicz,
A.
S.
(1979).
Autonomic
components
of
vicarious
conditioning
and

psychopathy.
J
Clin
Psychol,
35(1),
60-‐67.

Antoniadis,
E.
A.,
Winslow,
J.
T.,
Davis,
M.,
&
Amaral,
D.
G.
(2009).
The
nonhuman

primate
amygdala
is
necessary
for
the
acquisition
but
not
the
retention
of

fear-‐potentiated
startle.
Biol
Psychiatry,
65(3),
241-‐248.
doi:

10.1016/j.biopsych.2008.07.007

Arnett,
P.
A.,
Smith,
S.
S.,
&
Newman,
J.
P.
(1997).
Approach
and
avoidance

motivation
in
psychopathic
criminal
offenders
during
passive
avoidance.
J

Pers
Soc
Psychol,
72(6),
1413-‐1428.

Arsenio,
W.
F.,
&
Fleiss,
K.
(1996).
Typical
and
behaviourally
disruptive
children's

understanding
of
the
emotion
consequences
of
socio-‐moral
events
(Vol.
14,

pp.
173-‐186):
British
Journal
of
Developmental
Psychology.

Assaf,
Y.,
&
Pasternak,
O.
(2008).
Diffusion
tensor
imaging
(DTI)-‐based
white
matter

mapping
in
brain
research:
a
review.
J
Mol
Neurosci,
34(1),
51-‐61.
doi:

10.1007/s12031-‐007-‐0029-‐0

Association,
A.
A.
P.
(1994).
Diagnostic
and
statistical
manual
of
mental
disorders

(DSM-‐IV).
American
Psychiatric
Association.

Baas,
D.,
Aleman,
A.,
&
Kahn,
R.
S.
(2004).
Lateralization
of
amygdala
activation:
a

systematic
review
of
functional
neuroimaging
studies.
Brain
Res
Brain
Res

Rev,
45(2),
96-‐103.
doi:
10.1016/j.brainresrev.2004.02.004

155

Bagshaw,
M.
H.,
Kimble,
D.
P.,
&
Pribram,
K.
H.
(1965).
The
GSR
of
monkeys
during

orienting
and
habituation
and
after
ablation
of
the
amygdala,
hippocampus,

and
inferotemporal
cortex
(Vol.
3,
pp.
111-‐119):
Neuropsychologia.

Baker,
L.
A.,
Barton,
M.,
&
Raine,
A.
(2002).
The
Southern
California
Twin
Register
at

the
University
of
Southern
California.
Twin
Res,
5(5),
456-‐459.
doi:

10.1375/136905202320906273

Barbas,
H.,
&
Pandya,
D.
N.
(1989).
Architecture
and
intrinsic
connections
of
the

prefrontal
cortex
in
the
rhesus
monkey.
J
Comp
Neurol,
286(3),
353-‐375.
doi:

10.1002/cne.902860306

Baxter,
M.
G.,
Parker,
A.,
Lindner,
C.
C.,
Izquierdo,
A.
D.,
&
Murray,
E.
A.
(2000).
Control

of
response
selection
by
reinforcer
value
requires
interaction
of
amygdala

and
orbital
prefrontal
cortex.
J
Neurosci,
20(11),
4311-‐4319.

Beckmann,
C.
F.,
Jenkinson,
M.,
&
Smith,
S.
M.
(2003).
General
multilevel
linear

modeling
for
group
analysis
in
FMRI.
Neuroimage,
20(2),
1052-‐1063.
doi:

10.1016/S1053-‐8119(03)00435-‐XS105381190300435X
[pii]

Belova,
M.
A.,
Paton,
J.
J.,
Morrison,
S.
E.,
&
Salzman,
C.
D.
(2007).
Expectation

modulates
neural
responses
to
pleasant
and
aversive
stimuli
in
primate

amygdala.
Neuron,
55(6),
970-‐984.
doi:
10.1016/j.neuron.2007.08.004

Belova,
M.
A.,
Paton,
J.
J.,
&
Salzman,
C.
D.
(2008).
Moment-‐to-‐moment
tracking
of

state
value
in
the
amygdala.
J
Neurosci,
28(40),
10023-‐10030.
doi:

10.1523/JNEUROSCI.1400-‐08.2008

Benning,
S.
D.,
Patrick,
C.
J.,
Hicks,
B.
M.,
Blonigen,
D.
M.,
&
Krueger,
R.
F.
(2003).

Factor
structure
of
the
psychopathic
personality
inventory:
validity
and

implications
for
clinical
assessment.
Psychol
Assess,
15(3),
340-‐350.
doi:

10.1037/1040-‐3590.15.3.340

Bezdjian,
S.,
Raine,
A.,
Baker,
L.
A.,
&
Lynam,
D.
R.
(2011).
Psychopathic
personality
in

children:
genetic
and
environmental
contributions.
Psychol
Med,
41(3),
589-‐
600.
doi:
10.1017/S0033291710000966

Birbaumer,
N.,
Veit,
R.,
Lotze,
M.,
Erb,
M.,
Hermann,
C.,
Grodd,
W.,
&
Flor,
H.
(2005).

Deficient
fear
conditioning
in
psychopathy:
a
functional
magnetic
resonance

imaging
study.
Arch
Gen
Psychiatry,
62(7),
799-‐805.
doi:
62/7/799

[pii]10.1001/archpsyc.62.7.799

Bishop,
S.,
Duncan,
J.,
Brett,
M.,
&
Lawrence,
A.
D.
(2004).
Prefrontal
cortical
function

156

and
anxiety:
controlling
attention
to
threat-‐related
stimuli.
Nat
Neurosci,

7(2),
184188.
doi:
nn1173
[pii]10.1038/nn1173

Bjork,
J.
M.,
Chen,
G.,
&
Hommer,
D.
W.
(2012).
Psychopathic
tendencies
and

mesolimbic
recruitment
by
cues
for
instrumental
and
passively
obtained

rewards.
Biol
Psychol,
89(2),
408-‐415.
doi:
10.1016/j.biopsycho.2011.12.003

Blackburn,
R.
(1998).
Psychopathy
and
personality
disorder:
Implications
of

interpersonal
theory.
In
D.
J.
Cooke,
A.
E.
Forth
&
R.
D.
Hare
(Eds.),

Psychopathy:
Theory,
research,
and
implications
for
society
(pp.
269-‐301).

Dordrecht,
The
Netherlands:
Kluwer.

Blackburn,
R.,
&
Lee-‐Evans,
J.
M.
(1985).
Reactions
of
primary
and
secondary

psychopaths
to
anger-‐evoking
situations.
Br
J
Clin
Psychol,
24
(
Pt
2),
93-‐100.

Blair,
J.,
Mitchell,
D.
R.,
&
Blair,
K.
(2005).
The
psychopath
:
emotion
and
the
brain.

Malden,
MA:
Blackwell
Pub.

Blair,
R.
J.
(2004).
The
roles
of
orbital
frontal
cortex
in
the
modulation
of
antisocial

behavior.
Brain
Cogn,
55(1),
198-‐208.
doi:
10.1016/S0278-‐2626(03)00276-‐8

Blair,
R.
J.
(2008).
The
amygdala
and
ventromedial
prefrontal
cortex:
functional

contributions
and
dysfunction
in
psychopathy.
Philos
Trans
R
Soc
Lond
B
Biol

Sci,
363(1503),
2557-‐2565.
doi:
FJ4U862M1642N201

[pii]10.1098/rstb.2008.0027

Blair,
R.
J.
(2009).
Too
much
of
a
good
thing:
increased
grey
matter
in
boys
with

conduct
problems
and
callous-‐unemotional
traits.
Brain,
132(Pt
4),
831-‐832.

doi:
awp051
[pii]10.1093/brain/awp051

Blair,
R.
J.,
&
Cipolotti,
L.
(2000).
Impaired
social
response
reversal.
A
case
of

'acquired
sociopathy'.
Brain,
123
(
Pt
6),
1122-‐1141.

Blair,
R.
J.,
Colledge,
E.,
Murray,
L.,
&
Mitchell,
D.
G.
(2001).
A
selective
impairment
in

the
processing
of
sad
and
fearful
expressions
in
children
with
psychopathic

tendencies.
J
Abnorm
Child
Psychol,
29(6),
491-‐498.

Blair,
R.
J.
R.
(1997).
Moral
reasoning
in
the
child
with
psychopathic
tendencies
(Vol.

22,
pp.
731-‐739):
Personality
and
Individual
Differences.

Blair,
R.
J.
R.
(1999).
Responsiveness
to
distress
cues
in
the
child
with
psychopathic

tendencies.
(Vol.
27,
pp.
135-‐145):
Personality
and
Individual
Differences.

157

Blair,
R.
J.
R.,
Jones,
L.,
Clark,
F.,
&
Smith,
M.
(1995).
Is
the
psychopath
"morally

insane"?
(Vol.
19,
pp.
741-‐752):
Personality
and
Individual
Differences.

Boucsein,
W.
(1992).
Electrodermal
activity.
New
York:
Plenum
Press.

Bourgeois,
J.
P.,
Goldman-‐Rakic,
P.
S.,
&
Rakic,
P.
(1994).
Synaptogenesis
in
the

prefrontal
cortex
of
rhesus
monkeys.
Cereb
Cortex,
4(1),
78-‐96.

Brainard,
D.
(1997).
The
Psychophysics
Toolbox
(Vol.
10,
pp.
433-‐436).
Spatial

Vision.

Brinkley,
C.
A.,
Newman,
J.
P.,
Widiger,
T.
A.,
&
Lynam,
D.
R.
(2004).
Two
approaches

to
parsing
the
hterogeneity
of
psychopathy.
(Vol.
11,
pp.
69-‐94):
Clinical

Psychology:
Science
and
Practice.

Burgos-‐Robles,
A.,
Vidal-‐Gonzalez,
I.,
Santini,
E.,
&
Quirk,
G.
J.
(2007).
Consolidation

of
fear
extinction
requires
NMDA
receptor-‐dependent
bursting
in
the

ventromedial
prefrontal
cortex.
Neuron,
53(6),
871-‐880.
doi:

10.1016/j.neuron.2007.02.021

Büchel,
C.,
Morris,
J.,
Dolan,
R.
J.,
&
Friston,
K.
J.
(1998).
Brain
systems
mediating

aversive
conditioning:
an
event-‐related
fMRI
study.
Neuron,
20(5),
947-‐957.

doi:
S0896-‐6273(00)80476-‐6
[pii]

Carlsson,
K.,
Andersson,
J.,
Petrovic,
P.,
Petersson,
K.
M.,
Ohman,
A.,
&
Ingvar,
M.

(2006).
Predictability
modulates
the
affective
and
sensory-‐discriminative

neural
processing
of
pain.
Neuroimage,
32(4),
1804-‐1814.
doi:

10.1016/j.neuroimage.2006.05.027

Carmichael,
S.
T.,
&
Price,
J.
L.
(1995a).
Limbic
connections
of
the
orbital
and
medial

prefrontal
cortex
in
macaque
monkeys.
J
Comp
Neurol,
363(4),
615-‐641.
doi:

10.1002/cne.903630408

Carmichael,
S.
T.,
&
Price,
J.
L.
(1995b).
Limbic
connections
of
the
orbital
and
medial

prefrontal
cortex
in
macaque
monkeys.
J
Comp
Neurol,
363(4),
615-‐641.
doi:

10.1002/cne.903630408

Carmichael,
S.
T.,
&
Price,
J.
L.
(1995c).
Sensory
and
premotor
connections
of
the

orbital
and
medial
prefrontal
cortex
of
macaque
monkeys.
J
Comp
Neurol,

363(4),
642-‐664.
doi:
10.1002/cne.903630409

Catani,
M.,
&
Thiebaut
de
Schotten,
M.
(2008).
A
diffusion
tensor
imaging

158

tractography
atlas
for
virtual
in
vivo
dissections.
Cortex,
44(8),
1105-‐1132.

doi:
S0010-‐9452(08)00123-‐8
[pii]10.1016/j.cortex.2008.05.004

Cavada,
C.,
Compañy,
T.,
Tejedor,
J.,
Cruz-‐Rizzolo,
R.
J.,
&
Reinoso-‐Suárez,
F.
(2000).

The
anatomical
connections
of
the
macaque
monkey
orbitofrontal
cortex.
A

review.
Cereb
Cortex,
10(3),
220-‐242.

Chaplin,
T.
C.,
Rice,
M.
E.,
&
Harris,
G.
T.
(1995).
Salient
victim
suffering
and
the

sexual
responses
of
child
molesters.
J
Consult
Clin
Psychol,
63(2),
249-‐255.

Chen,
N.
K.,
Dickey,
C.
C.,
Yoo,
S.
S.,
Guttmann,
C.
R.,
&
Panych,
L.
P.
(2003).
Selection

of
voxel
size
and
slice
orientation
for
fMRI
in
the
presence
of
susceptibility

field
gradients:
application
to
imaging
of
the
amygdala.
Neuroimage,
19(3),

817-‐825.

Cheng,
D.
T.,
Richards,
J.,
&
Helmstetter,
F.
J.
(2007).
Activity
in
the
human
amygdala

corresponds
to
early,
rather
than
late
period
autonomic
responses
to
a
signal

for
shock.
Learn
Mem,
14(7),
485-‐490.
doi:
10.1101/lm.632007

Christian,
R.
E.,
Frick,
P.
J.,
Hill,
N.
L.,
&
Tyler,
L.
(1997).
Psychopathy
and
conduct

problems
in
children:
II.

Implications
for
subtyping
children
with
conduct

problems:
Journal
of
the
American
Academy
of
Child
&
Adolescent

Psychiatry.

Cleckley,
H.
M.
(1982).
The
mask
of
sanity
(Rev.
ed.).
New
York

St.
Louis:
New
American
Library
;

Mosby.

Constable,
R.
T.,
Ment,
L.
R.,
Vohr,
B.
R.,
Kesler,
S.
R.,
Fulbright,
R.
K.,
Lacadie,
C.,
.
.
.

Reiss,
A.
R.
(2008).
Prematurely
born
children
demonstrate
white
matter

microstructural
differences
at
12
years
of
age,
relative
to
term
control

subjects:
an
investigation
of
group
and
gender
effects.
Pediatrics,
121(2),
306-‐
316.
doi:
10.1542/peds.2007-‐0414

Contreras-‐Rodríguez,
O.,
Pujol,
J.,
Batalla,
I.,
Harrison,
B.
J.,
Bosque,
J.,
Ibern-‐Regàs,
I.,

.
.
Cardoner,
N.
(2013).
Disrupted
neural
processing
of
emotional
faces
in

psychopathy.
Soc
Cogn
Affect
Neurosci.
doi:
10.1093/scan/nst014

Convit,
A.,
McHugh,
P.,
Wolf,
O.
T.,
de
Leon,
M.
J.,
Bobinski,
M.,
De
Santi,
S.,
.
.
.
Tsui,
W.

(1999).
MRI
volume
of
the
amygdala:
a
reliable
method
allowing
separation

from
the
hippocampal
formation.
Psychiatry
Res,
90(2),
113-‐123.

Corcoran,
K.
A.,
&
Quirk,
G.
J.
(2007).
Activity
in
prelimbic
cortex
is
necessary
for
the

159

expression
of
learned,
but
not
innate,
fears.
J
Neurosci,
27(4),
840-‐844.
doi:

10.1523/JNEUROSCI.5327-‐06.2007

Craig,
M.
C.,
Catani,
M.,
Deeley,
Q.,
Latham,
R.,
Daly,
E.,
Kanaan,
R.,
.
.
.
Fahy,
T.
(2009).

Altered
connections
on
the
road
to
psychopathy
(Vol.
14,
pp.
946-‐953

907):
Molecular
Psychiatry.

Critchley,
H.
D.,
Elliott,
R.,
Mathias,
C.
J.,
&
Dolan,
R.
J.
(2000).
Neural
activity
relating

to
generation
and
representation
of
galvanic
skin
conductance
responses:
a

functional
magnetic
resonance
imaging
study.
J
Neurosci,
20(8),
3033-‐3040.

Dadds,
M.
R.,
El
Masry,
Y.,
Wimalaweera,
S.,
&
Guastella,
A.
J.
(2008).
Reduced
eye

gaze
explains
"fear
blindness"
in
childhood
psychopathic
traits.
J
Am
Acad

Child
Adolesc
Psychiatry,
47(4),
455-‐463.
doi:

10.1097/CHI.0b013e31816407f1

Damasio,
A.
R.
(1994).
Descartes'
error
:
emotion,
reason,
and
the
human
brain.
New

York:
G.P.
Putnam.

Damasio,
A.
R.,
Tranel,
D.,
&
Damasio,
H.
(1990).
Individuals
with
sociopathic

behavior
caused
by
frontal
damage
fail
to
respond
autonomically
to
social

stimuli.
Behav
Brain
Res,
41(2),
81-‐94.
doi:
0166-‐4328(90)90144-‐4
[pii]

Damasio,
A.
R.,
Tranel,
D.,
&
Damasio,
H.
(1990).
Individuals
with
sociopathic

behavior
caused
by
frontal
damage
fail
to
respond
autonomically
to
social

stimuli.
Behav
Brain
Res,
41(2),
81-‐94.
doi:
0166-‐4328(90)90144-‐4
[pii]

Davidson,
R.
J.,
Jackson,
D.
C.,
&
Kalin,
N.
H.
(2000).
Emotion,
plasticity,
context,
and

regulation:
perspectives
from
affective
neuroscience.
Psychol
Bull,
126(6),

890-‐909.

Davidson,
R.
J.
(2002).
Anxiety
and
affective
style:
role
of
prefrontal
cortex
and

amygdala.
Biol
Psychiatry,
51(1),
68-‐80.

Davis,
M.,
&
Shi,
C.
(2000).
The
amygdala.
Curr
Biol,
10(4),
R131.

Davis,
M.,
&
Whalen,
P.
J.
(2001).
The
amygdala:
vigilance
and
emotion.
Mol

Psychiatry,
6(1),
13-‐34.

De
Brito,
S.
A.,
McCrory,
E.
J.,
Mechelli,
A.,
Wilke,
M.,
Jones,
A.
P.,
Hodgins,
S.,
&
Viding,

E.
(2011).
Small,
but
not
perfectly
formed:
decreased
white
matter

concentration
in
boys
with
psychopathic
tendencies.
Mol
Psychiatry,
16(5),

476-‐477.
doi:
10.1038/mp.2010.74

160

De
Brito,
S.
A.,
Mechelli,
A.,
Wilke,
M.,
Laurens,
K.
R.,
Jones,
A.
P.,
Barker,
G.
J.,
.
.
.

Viding,
E.
(2009).
Size
matters:
increased
grey
matter
in
boys
with
conduct

problems
and
callous-‐unemotional
traits.
Brain,
132(Pt
4),
843-‐852.
doi:

awp011
[pii]10.1093/brain/awp011

Del
Gaizo,
A.
L.,
&
Falkenbach,
D.
M.
(2008).
Primary
and
secondary
psychopathic

traits
and
their
relationship
to
perception
and
experience
of
emotion
(Vol.

45,
pp.
206-‐212):
Personality
and
Individual
Differences.

Delgado,
M.
R.,
Nearing,
K.
I.,
Ledoux,
J.
E.,
&
Phelps,
E.
A.
(2008).
Neural
circuitry

underlying
the
regulation
of
conditioned
fear
and
its
relation
to
extinction.

Neuron,
59(5),
829-‐838.
doi:
10.1016/j.neuron.2008.06.029

Devinsky,
O.,
Morrell,
M.
J.,
&
Vogt,
B.
A.
(1995).
Contributions
of
anterior
cingulate

cortex
to
behaviour.
Brain,
118
(
Pt
1),
279-‐306.

Dolan,
M.,
&
Fullam,
R.
(2006).
Face
affect
recognition
deficits
in
personality

disordered
offenders:
association
with
psychopathy.
Psychol
Med,
36(11),

1563-‐1569.
doi:
10.1017/S0033291706008634

Edens,
J.
F.,
&
McDermott,
B.
E.
(2010).
Examining
the
construct
validity
of
the

Psychopathic
Personality
Inventory-‐Revised:
preferential
correlates
of

fearless
dominance
and
self-‐centered
impulsivity.
Psychol
Assess,
22(1),
32-‐
42.
doi:
10.1037/a0018220

Eluvathingal,
T.
J.,
Chugani,
H.
T.,
Behen,
M.
E.,
Juhász,
C.,
Muzik,
O.,
Maqbool,
M.,
.
.
.

Makki,
M.
(2006).
Abnormal
brain
connectivity
in
children
after
early
severe

socioemotional
deprivation:
a
diffusion
tensor
imaging
study.
Pediatrics,

117(6),
2093-‐2100.
doi:
10.1542/peds.2005-‐1727

Ermer,
E.,
&
Kiehl,
K.
A.
(2010).
Psychopaths
are
impaired
in
social
exchange
and

precautionary
reasoning.
Psychol
Sci,
21(10),
1399-‐1405.
doi:

10.1177/0956797610384148

Etkin,
A.,
Egner,
T.,
&
Kalisch,
R.
(2011).
Emotional
processing
in
anterior
cingulate

and
medial
prefrontal
cortex.
Trends
Cogn
Sci,
15(2),
85-‐93.
doi:

10.1016/j.tics.2010.11.004

Everitt,
B.
J.,
Cardinal,
R.
N.,
Parkinson,
J.
A.,
&
Robbins,
T.
W.
(2003).
Appetitive

behavior:
impact
of
amygdala-‐dependent
mechanisms
of
emotional
learning.

Ann
N
Y
Acad
Sci,
985,
233-‐250.

Fagan,
T.
J.,
&
Lira,
F.
(1980).
The
primary
and
secondary
sociopathic
personality:

161

Differences
in
frequency
and
severity
of
antisocial
behaviors
(Vol.
89,
pp.

493-‐496):
Journal
of
abnormal
psychology.

Fairchild,
G.,
Passamonti,
L.,
Hurford,
G.,
Hagan,
C.
C.,
von
dem
Hagen,
E.
A.,
van

Falkenbach,
D.,
Poythress,
N.,
&
Creevey,
C.
(2008).
The
exploration
of
subclinical

psychopathic
subtypes
and
the
relationship
with
types
of
aggression.
(Vol.

44,
pp.
821-‐832):
Personality
and
Individual
Differences.

Falls,
W.
A.,
Miserendino,
M.
J.,
&
Davis,
M.
(1992).
Extinction
of
fear-‐potentiated

startle:
blockade
by
infusion
of
an
NMDA
antagonist
into
the
amygdala.
J

Neurosci,
12(3),
854-‐863.

Finger,
E.
C.,
Marsh,
A.
A.,
Blair,
K.
S.,
Reid,
M.
E.,
Sims,
C.,
Ng,
P.,
.
.
.
Blair,
R.
J.
(2011).

Disrupted
reinforcement
signaling
in
the
orbitofrontal
cortex
and
caudate
in

youths
with
conduct
disorder
or
oppositional
defiant
disorder
and
a
high

level
of
psychopathic
traits.
Am
J
Psychiatry,
168(2),
152-‐162.
doi:

10.1176/appi.ajp.2010.10010129

Finger,
E.
C.,
Marsh,
A.
A.,
Mitchell,
D.
G.,
Reid,
M.
E.,
Sims,
C.,
Budhani,
S.,
.
.
.
Blair,
J.
R.

(2008).
Abnormal
ventromedial
prefrontal
cortex
function
in
children
with

psychopathic
traits
during
reversal
learning.
Arch
Gen
Psychiatry,
65(5),
586-‐
594.
doi:
10.1001/archpsyc.65.5.586

Flor,
H.,
Birbaumer,
N.,
Hermann,
C.,
Ziegler,
S.,
&
Patrick,
C.
J.
(2002).
Aversive

Pavlovian
conditioning
in
psychopaths:
peripheral
and
central
correlates.

Psychophysiology,
39(4),
505-‐518.
doi:
S0048577202394046
[pii]

10.1017.S0048577202394046

Forth,
A.
E.,
Hart,
S.
D.,
&
Hare,
R.
D.
(1990).
Assessment
of
psychopathy
in
male

young
offenders.
(Vol.
2,
pp.
342-‐344):
Psychological
Assessment.

Frick,
P.
J.,
&
Hare,
R.
D.
(2001).
THe
Antisocial
Process
Screening
Device
(APSD).

Toronto,
Canada:
Multi-‐Health
Systems.

Fujii,
M.
(1983).
Fiber
connections
between
the
thalamic
posterior
lateral
nucleus

and
the
cingulate
gyrus
in
the
cat.
Neurosci
Lett,
39(2),
137-‐142.

Glenn,
A.
L.,
Raine,
A.,
&
Schug,
R.
A.
(2009).
The
neural
correlates
of
moral
decision

making
in
psychopathy.
Mol
Psychiatry,
14(1),
5-‐6.
doi:
mp2008104
[pii]

10.1038/mp.2008.104

Gogtay,
N.,
Giedd,
J.
N.,
Lusk,
L.,
Hayashi,
K.
M.,
Greenstein,
D.,
Vaituzis,
A.
C.,
.
.
.
Toga,

162

A.
W.
(2004).
Dynamic
mapping
of
human
cortical
development
during

childhood
through
early
adulthood.
Proceedings
of
the
National
Academy
of

Sciences
of
the
United
States
of
America,
101(21),
8174-‐8179.

Gordon,
H.
L.,
Baird,
A.
A.,
&
End,
A.
(2004).
Functional
differences
among
those
high

and
low
on
a
trait
measure
of
psychopathy.
Biol
Psychiatry,
56(7),
516-‐521.

doi:
S0006-‐3223(04)00708-‐5
[pii]10.1016/j.biopsych.2004.06.030

Gottfried,
J.
A.,
O'Doherty,
J.,
&
Dolan,
R.
J.
(2002).
Appetitive
and
aversive
olfactory

learning
in
humans
studied
using
event-‐related
functional
magnetic

resonance
imaging.
J
Neurosci,
22(24),
10829-‐10837.

Haber,
S.
N.,
Kunishio,
K.,
Mizobuchi,
M.,
&
Lynd-‐Balta,
E.
(1995).
The
orbital
and

medial
prefrontal
circuit
through
the
primate
basal
ganglia.
J
Neurosci,
15(7

Pt
1),
4851-‐4867.

Hadland,
K.
A.,
Rushworth,
M.
F.,
Gaffan,
D.,
&
Passingham,
R.
E.
(2003).
The
effect
of

cingulate
lesions
on
social
behaviour
and
emotion.
Neuropsychologia,
41(8),

919-‐931.

Hamburger,
M.
E.,
Lilienfeld,
S.
O.,
&
Hogben,
M.
(1996).
Psychopathy,
gender,
and

gender
roles:Implications
for
antisocial
and
histrionic
personality
disorders.

(Vol.
10,
pp.
41-‐55):
Journal
of
Personality
Disorders.

Hare,
R.
D.
(1965).
Temporal
gradient
of
fear
arousal
in
psychopaths.
J
Abnorm

Psychol,
70(6),
442-‐445.

Hare,
R.
D.
(1972).
Psychopathy
and
physiological
responses
to
adrenalin.
J
Abnorm

Psychol,
79(2),
138-‐147.

Hare,
R.
D.
(1978).
Psychopathy
and
electrodermal
responses
to
nonsignal

stimulation.
Biol
Psychol,
6(4),
237-‐246.

Hare,
R.
D.,
&
Neumann,
C.
S.
(2008).
Psychopathy
as
a
clinical
and
empirical

construct.
Annu
Rev
Clin
Psychol,
4,
217-‐246.
doi:

10.1146/annurev.clinpsy.3.022806.091452

Hare,
R.
D.,
&
Quinn,
M.
J.
(1971).
Psychopathy
and
autonomic
conditioning.
J

Abnorm
Psychol,
77(3),
223-‐235.

Hare,
R.
D.,
&
Thorvaldson,
S.
A.
(1970).
Psychopathy
and
response
to
electrical

stimulation.
J
Abnorm
Psychol,
76(3),
370-‐374.

163

Hare,
R.
D.
(1991).
Psychopathy
Checklist:
Revised.
908
Niagara
Falls
Blvd,
North

Tonawanda,
New
York,
USA,
14120-‐2060:
Multi-‐Health
Systems.

Hare,
R.
D.
(1999).
Without
conscience
:
the
disturbing
world
of
the
psychopaths

among
us.
New
York:
Guilford
Press.

Harenski,
C.
L.,
Kim,
S.
H.,
&
Hamann,
S.
(2009).
Neuroticism
and
psychopathy

predict
brain
activation
during
moral
and
nonmoral
emotion
regulation.
Cogn

Affect
Behav
Neurosci,
9(1),
1-‐15.
doi:
9/1/1
[pii]

10.3758/CABN.9.1.1

Hathaway,
S.
R.,
Engel,
R.
R.,
&
MacKinley,
J.
C.
(2000).
Minnesota
Multiphasic

Personality
Inventory
2:
MMPI
2:
Huber.

Henderson,
M.
(1983).
An
empirical
classification
of
non-‐violent
offenders
using
the

MMPI
(Vol.
4,
pp.
671-‐677):
Personality
and
Individual
Differences.

Hiatt,
K.
D.,
Lorenz,
A.
R.,
&
Newman,
J.
P.
(2002).
Assessment
of
emotion
and

language
processing
in
psychopathic
offenders:
Results
from
a
dichotic

listening
task.
(Vol.
32(7),
pp.
1255-‐1268):
Personality
and
Individual

Differences.

Hoaglin,
D.
C.,
&
Iglewicz,
B.
(1987).
Fine-‐tuning
some
resistant
rules
for
outlier

labeling
(Vol.
82,
pp.
1147-‐1149):
Journal
of
the
American
Statistical

Assocation.

House,
T.
H.,
&
Milligan,
W.
L.
(1976).
Autonomic
responses
to
modeled
distress
in

prison
psychopaths.
J
Pers
Soc
Psychol,
34(4),
556-‐560.

Huebner,
T.,
Vloet,
T.
D.,
Marx,
I.,
Konrad,
K.,
Fink,
G.
R.,
Herpertz,
S.
C.,
&
Herpertz

Dahlmann,
B.
(2008).
Morphometric
brain
abnormalities
in
boys
with

conduct
disorder.
J
Am
Acad
Child
Adolesc
Psychiatry,
47(5),
540-‐547.
doi:

10.1097/CHI.0b013e3181676545

Hughes,
M.
A.,
Stout,
J.
C.,
&
Dolan,
M.
C.
(2013).
Concurrent
validity
of
the

Psychopathic
Personality
Inventory-‐Revised
and
the
Psychopathy
Checklist

Screening
Version
in
an
Australian
Offender
Sample
(Vol.
40,
pp.
802-‐813):

Criminal
Justice
and
Behavior.

Huster,
R.
J.,
Westerhausen,
R.,
Kreuder,
F.,
Schweiger,
E.,
&
Wittling,
W.
(2007).

Morphologic
asymmetry
of
the
human
anterior
cingulate
cortex.
Neuroimage,

34(3),
888-‐895.
doi:
10.1016/j.neuroimage.2006.10.023

164

Huttenlocher,
P.
R.
(1979).
Synaptic
density
in
human
frontal
cortex
-‐
developmental

changes
and
effects
of
aging.
Brain
Res,
163(2),
195-‐205.

Hyatt,
C.
J.,
Haney-‐Caron,
E.,
&
Stevens,
M.
C.
(2012).
Cortical
thickness
and
folding

deficits
in
conduct-‐disordered
adolescents.
Biol
Psychiatry,
72(3),
207-‐214.

doi:
10.1016/j.biopsych.2011.11.017

Izquierdo,
A.,
Suda,
R.
K.,
&
Murray,
E.
A.
(2005).
Comparison
of
the
effects
of

bilateral
orbital
prefrontal
cortex
lesions
and
amygdala
lesions
on
emotional

responses
in
rhesus
monkeys.
J
Neurosci,
25(37),
8534-‐8542.
doi:

10.1523/JNEUROSCI.1232-‐05.2005

Jenkinson,
M.,
&
Smith,
S.
(2001).
A
global
optimisation
method
for
robust
affine

registration
of
brain
images.
Med
Image
Anal,
5(2),
143-‐156.
doi:

S1361841501000366
[pii]

Jensen,
J.,
McIntosh,
A.
R.,
Crawley,
A.
P.,
Mikulis,
D.
J.,
Remington,
G.,
&
Kapur,
S.

(2003).
Direct
activation
of
the
ventral
striatum
in
anticipation
of
aversive

stimuli.
Neuron,
40(6),
1251-‐1257.

Johnson,
L.
C.,
&
Corah,
N.
L.
(1963).
Racial
Differences
in
Skin
Resistance.
Science,

139(3556),
766-‐767.
doi:
139/3556/766
[pii]

10.1126/science.139.3556.766

Jones,
A.
P.,
Laurens,
K.
R.,
Herba,
C.
M.,
Barker,
G.
J.,
&
Viding,
E.
(2009).
Amygdala

hypoactivity
to
fearful
faces
in
boys
with
conduct
problems
and
callous-‐
unemotional
traits.
Am
J
Psychiatry,
166(1),
95-‐102.
doi:

appi.ajp.2008.07071050
[pii]10.1176/appi.ajp.2008.07071050

Juárez,
M.,
Kiehl,
K.
A.,
&
Calhoun,
V.
D.
(2012).
Intrinsic
limbic
and
paralimbic

networks
are
associated
with
criminal
psychopathy.
Hum
Brain
Mapp.
doi:

10.1002/hbm.22037

Kalin,
N.
H.,
&
Shelton,
S.
E.
(2003).
Nonhuman
primate
models
to
study
anxiety,

emotion
regulation,
and
psychopathology.
Ann
N
Y
Acad
Sci,
1008,
189-‐200.

Kalin,
N.
H.,
Shelton,
S.
E.,
&
Davidson,
R.
J.
(2004).
The
role
of
the
central
nucleus
of

the
amygdala
in
mediating
fear
and
anxiety
in
the
primate.
J
Neurosci,
24(24),

5506-‐5515.
doi:
24/24/5506
[pii]10.1523/JNEUROSCI.0292-‐04.2004

Kalin,
N.
H.,
Shelton,
S.
E.,
&
Davidson,
R.
J.
(2007).
Role
of
the
primate
orbitofrontal

cortex
in
mediating
anxious
temperament.
Biol
Psychiatry,
62(10),
1134-‐
1139.
doi:
10.1016/j.biopsych.2007.04.004

165

Karpman,
B.
(1947).
Passive
parasitic
psychopathy:
toward
the
personality

structure
and
psychogenesis
of
idiopathic
psychopathy
(anethopathy).

Psychoanal
Rev,
34(2),
198;
passim.

Kelley,
W.
M.,
Macrae,
C.
N.,
Wyland,
C.
L.,
Caglar,
S.,
Inati,
S.,
&
Heatherton,
T.
F.

(2002).
Finding
the
self?
An
event-‐related
fMRI
study.
J
Cogn
Neurosci,
14(5),

785-‐794.
doi:
10.1162/08989290260138672

Kennerley,
S.
W.,
Walton,
M.
E.,
Behrens,
T.
E.,
Buckley,
M.
J.,
&
Rushworth,
M.
F.

(2006).
Optimal
decision
making
and
the
anterior
cingulate
cortex.
Nat

Neurosci,
9(7),
940-‐947.
doi:
10.1038/nn1724

Kiehl,
K.
A.
(2006).
A
cognitive
neuroscience
perspective
on
psychopathy:
evidence

for
paralimbic
system
dysfunction.
Psychiatry
Res,
142(2-‐3),
107-‐128.
doi:

S0165-‐1781(05)00290-‐8
[pii]10.1016/j.psychres.2005.09.013

Kiehl,
K.
A.,
Smith,
A.
M.,
Hare,
R.
D.,
Mendrek,
A.,
Forster,
B.
B.,
Brink,
J.,
&
Liddle,
P.

F.(2001).
Limbic
abnormalities
in
affective
processing
by
criminal

psychopaths
as
revealed
by
functional
magnetic
resonance
imaging.
Biol

Psychiatry,
50(9),
677-‐684.
doi:
S0006322301012227
[pii]

Kiehl,
K.
A.,
Smith,
A.
M.,
Mendrek,
A.,
Forster,
B.
B.,
Hare,
R.
D.,
&
Liddle,
P.
F.
(2004).

Temporal
lobe
abnormalities
in
semantic
processing
by
criminal
psychopaths

as
revealed
by
functional
magnetic
resonance
imaging.
Psychiatry
Res,
130(1),

27-‐42.
doi:
S0925492703001069
[pii]10.1016/S0925-‐4927(03)00106-‐9

Kim,
M.,
Campeau,
S.,
Falls,
W.
A.,
&
Davis,
M.
(1993).
Infusion
of
the
non-‐NMDA

receptor
antagonist
CNQX
into
the
amygdala
blocks
the
expression
of
fear-‐
potentiated
startle.
Behav
Neural
Biol,
59(1),
5-‐8.

Kim,
M.
J.,
&
Whalen,
P.
J.
(2009).
The
structural
integrity
of
an
amygdala-‐prefrontal

pathway
predicts
trait
anxiety.
J
Neurosci,
29(37),
11614-‐11618.
doi:

29/37/11614
[pii]10.1523/JNEUROSCI.2335-‐09.2009

Kimonis,
E.
R.,
Frick,
P.
J.,
Cauffman,
E.,
Goldweber,
A.,
&
Skeem,
J.
(2012).
Primary

and
secondary
variants
of
juvenile
psychopathy
differ
in
emotional

processing.
Dev
Psychopathol,
24(3),
1091-‐1103.
doi:

10.1017/S0954579412000557

Kimonis,
E.
R.,
Skeem,
J.
L.,
Cauffman,
E.,
&
Dmitrieva,
J.
(2011).
Are
secondary

166

variants
of
juvenile
psychopathy
more
reactively
violent
and
less

psychosocially
mature
than
primary
variants?
(Vol.
35,
pp.
381-‐391):
Law

and
Human
Behavior.

Knight,
D.
C.,
Smith,
C.
N.,
Cheng,
D.
T.,
Stein,
E.
A.,
&
Helmstetter,
F.
J.
(2004).

Amygdala
and
hippocampal
activity
during
acquisition
and
extinction
of

human
fear
conditioning.
Cogn
Affect
Behav
Neurosci,
4(3),
317-‐325.

Koenigs,
M.,
Kruepke,
M.,
Zeier,
J.,
&
Newman,
J.
P.
(2011).
Utilitarian
moral

judgment
in
psychopathy.
Soc
Cogn
Affect
Neurosci.
doi:
nsr048
[pii]

10.1093/scan/nsr048

Koenigs,
M.,
Young,
L.,
Adolphs,
R.,
Tranel,
D.,
Cushman,
F.,
Hauser,
M.,
&
Damasio,
A.

(2007).
Damage
to
the
prefrontal
cortex
increases
utilitarian
moral

judgements.
Nature,
446(7138),
908-‐911.
doi:
nature05631

[pii]10.1038/nature05631

Koenigs,
M.,
Baskin-‐Sommers,
A.,
Zeier,
J.,
&
Newman,
J.
P.
(2011).
Investigating
the

neural
correlates
of
psychopathy:
a
critical
review.
Mol
Psychiatry,
16(8),
792-‐
799.
doi:
mp2010124
[pii]10.1038/mp.2010.124

Koenigs,
M.,
Huey,
E.
D.,
Raymont,
V.,
Cheon,
B.,
Solomon,
J.,
Wassermann,
E.
M.,
&

Grafman,
J.
(2008).
Focal
brain
damage
protects
against
post-‐traumatic
stress

disorder
in
combat
veterans.
Nat
Neurosci,
11(2),
232-‐237.
doi:
nn2032

[pii]10.1038/nn2032

Koenigs,
M.,
Kruepke,
M.,
&
Newman,
J.
P.
(2010).
Economic
decision-‐making
in

psychopathy:
a
comparison
with
ventromedial
prefrontal
lesion
patients.

Neuropsychologia,
48(7),
2198-‐2204.
doi:
S0028-‐3932(10)00145-‐4

[pii]10.1016/j.neuropsychologia.2010.04.012

Koenigs,
M.,
Kruepke,
M.,
Zeier,
J.,
&
Newman,
J.
P.
(2012).
Utilitarian
moral

judgment

in
psychopathy.
Soc
Cogn
Affect
Neurosci,
7(6),
708-‐714.
doi:

nsr048
[pii]10.1093/scan/nsr048

Korol,
B.,
&
Kane,
R.
E.
(1978).
An
examination
of
the
relationship
between
race,
skin

color
and
a
series
of
autonomic
nervous
system
measures.
Pavlov
J
Biol
Sci,

13(2),
121-‐132.

Kosson,
D.
S.,
Smith,
S.
S.,
&
Newman,
J.
P.
(1990).
Evaluating
the
construct
validity
of

psychopathy
in
black
and
white
male
inmates:
three
preliminary
studies.
J

Abnorm
Psychol,
99(3),
250-‐259.

167

Kosson,
D.
S.,
Suchy,
Y.,
Mayer,
A.
R.,
&
Libby,
J.
(2002).
Facial
affect
recognition
in

criminal
psychopaths.
Emotion,
2(4),
398-‐411.

Kraemer,
G.
W.,
Ebert,
M.
H.,
Schmidt,
D.
E.,
&
McKinney,
W.
T.
(1989).
A
longitudinal

study
of
the
effect
of
different
social
rearing
conditions
on
cerebrospinal
fluid

norepinephrine
and
biogenic
amine
metabolites
in
rhesus
monkeys.

Neuropsychopharmacology,
2(3),
175-‐189.
doi:
0893-‐133X(89)90021-‐3
[pii]

LaBar,
K.
S.,
Gatenby,
J.
C.,
Gore,
J.
C.,
LeDoux,
J.
E.,
&
Phelps,
E.
A.
(1998).
Human

amygdala
activation
during
conditioned
fear
acquisition
and
extinction:
a

mixed-‐trial
fMRI
study.
Neuron,
20(5),
937-‐945.
doi:
S0896-‐6273(00)80475-‐4

[pii]

LaBar,
K.
S.,
LeDoux,
J.
E.,
Spencer,
D.
D.,
&
Phelps,
E.
A.
(1995).
Impaired
fear

conditioning
following
unilateral
temporal
lobectomy
in
humans.
J
Neurosci,

15(10),
6846-‐6855.

Lang,
H.,
Tuovinen,
T.,
&
Valleala,
P.
(1964).
Amygdaloid
after
discharge
and
galvanic

skin
response.
Electroencephalogr
Clin
Neurophysiol,
16,
366-‐374.

Laurent,
V.,
&
Westbrook,
R.
F.
(2008).
Distinct
contributions
of
the
basolateral

amygdala
and
the
medial
prefrontal
cortex
to
learning
and
relearning

extinction
of
context
conditioned
fear.
Learn
Mem,
15(9),
657-‐666.
doi:

10.1101/lm.1080108

Lebel,
C.,
&
Beaulieu,
C.
(2011).
Longitudinal
development
of
human
brain
wiring

continues
from
childhood
into
adulthood.
J
Neurosci,
31(30),
10937-‐10947.

doi:
10.1523/JNEUROSCI.5302-‐10.2011

Lebel,
C.,
Caverhill-‐Godkewitsch,
S.,
&
Beaulieu,
C.
(2010).
Age-‐related
regional

variations
of
the
corpus
callosum
identified
by
diffusion
tensor
tractography.

Neuroimage,
52(1),
20-‐31.
doi:
10.1016/j.neuroimage.2010.03.072

Lebel,
C.,
Walker,
L.,
Leemans,
A.,
Phillips,
L.,
&
Beaulieu,
C.
(2008).
Microstructural

maturation
of
the
human
brain
from
childhood
to
adulthood.
Neuroimage,

40(3),
1044-‐1055.
doi:
10.1016/j.neuroimage.2007.12.053

LeDoux,
J.
(2007).
The
amygdala.
Curr
Biol,
17(20),
R868-‐874.
doi:

10.1016/j.cub.2007.08.005

Levenson,
M.
R.,
Kiehl,
K.
A.,
&
Fitzpatrick,
C.
M.
(1995).
Assessing
psychopathic

attributes
in
a
noninstitutionalized
population.
J
Pers
Soc
Psychol,
68(1),
151-‐
158.

168

Levenston,
G.
K.,
Patrick,
C.
J.,
Bradley,
M.
M.,
&
Lang,
P.
J.
(2000).
The
psychopath
as

observer:
emotion
and
attention
in
picture
processing.
J
Abnorm
Psychol,

109(3),
373-‐385.

Lilienfeld,
S.
O.,
&
Andrews,
B.
P.
(1996).
Development
and
preliminary
validation
of

a
self-‐report
measure
of
psychopathic
personality
traits
in
noncriminal

populations.
J
Pers
Assess,
66(3),
488-‐524.

Lorenz,
A.
R.,
&
Newman,
J.
P.
(2002).
Deficient
response
modulation
and
emotion

processing
in
low-‐anxious
Caucasian
psychopathic
offenders:
results
from
a

lexical
decision
task.
Emotion,
2(2),
91-‐104.

Lykken,
D.
T.
(1957).
A
study
of
anxiety
in
the
sociopathic
personality.
J
Abnorm

Psychol,
55(1),
6-‐10.

Lynam,
D.
R.
(1997).
Pursuing
the
psychopath:
capturing
the
fledgling
psychopath
in

a
nomological
net.
J
Abnorm
Psychol,
106(3),
425-‐438.

Mabbott,
D.
J.,
Rovet,
J.,
Noseworthy,
M.
D.,
Smith,
M.
L.,
&
Rockel,
C.
(2009).
The

relations
between
white
matter
and
declarative
memory
in
older
children

and
adolescents.
Brain
Res,
1294,
80-‐90.
doi:
10.1016/j.brainres.2009.07.046

Machado,
C.
J.,
&
Bachevalier,
J.
(2007).
Measuring
reward
assessment
in
a
semi

naturalistic
context:
the
effects
of
selective
amygdala,
orbital
frontal
or

hippocampal
lesions.
Neuroscience,
148(3),
599-‐611.
doi:

10.1016/j.neuroscience.2007.06.035

Machado,
C.
J.,
Kazama,
A.
M.,
&
Bachevalier,
J.
(2009).
Impact
of
amygdala,
orbital

frontal,
or
hippocampal
lesions
on
threat
avoidance
and
emotional
reactivity

in
nonhuman
primates.
Emotion,
9(2),
147-‐163.
doi:
10.1037/a0014539

Málková,
L.,
Gaffan,
D.,
&
Murray,
E.
A.
(1997).
Excitotoxic
lesions
of
the
amygdala

fail
to
produce
impairment
in
visual
learning
for
auditory
secondary

reinforcement
but
interfere
with
reinforcer
devaluation
effects
in
rhesus

monkeys.
J
Neurosci,
17(15),
6011-‐6020.

Malterer,
M.
B.,
Lilienfeld,
S.
O.,
Neumann,
C.
S.,
&
Newman,
J.
P.
(2010).
Concurrent

validity
of
the
psychopathic
personality
inventory
with
offender
and

community
samples.
Assessment,
17(1),
3-‐15.
doi:

10.1177/1073191109349743

Marek,
R.,
Strobel,
C.,
Bredy,
T.
W.,
&
Sah,
P.
(2013).
The
amygdala
and
medial

169

prefrontal
cortex:
partners
in
the
fear
circuit.
J
Physiol,
591(Pt
10),
2381-‐
2391.
doi:
10.1113/jphysiol.2012.248575

Marsh,
A.
A.,
&
Cardinale,
E.
M.
(2012).
When
psychopathy
impairs
moral
judgments:

neural
responses
during
judgments
about
causing
fear.
Soc
Cogn
Affect

Neurosci.
doi:
10.1093/scan/nss097

Marsh,
A.
A.,
Finger,
E.
C.,
Fowler,
K.
A.,
Adalio,
C.
J.,
Jurkowitz,
I.
T.,
Schechter,
J.
C.,
.
.
.

Blair,
R.
J.
(2013).
Empathic
responsiveness
in
amygdala
and
anterior

cingulate
cortex
in
youths
with
psychopathic
traits.
J
Child
Psychol

Psychiatry.
doi:
10.1111/jcpp.12063

Marsh,
A.
A.,
Finger,
E.
C.,
Mitchell,
D.
G.,
Reid,
M.
E.,
Sims,
C.,
Kosson,
D.
S.,
.
.
.
Blair,
R.

J.
(2008).
Reduced
amygdala
response
to
fearful
expressions
in
children
and

adolescents
with
callous-‐unemotional
traits
and
disruptive
behavior

disorders.
Am
J
Psychiatry,
165(6),
712-‐720.
doi:
appi.ajp.2007.07071145

[pii]10.1176/appi.ajp.2007.07071145

Masui,
K.,
Iriguchi,
S.,
Nomura,
M.,
&
Ura,
M.
(2011).
Amount
of
altruistic
punishment

accounts
for
subsequent
emotional
gratification
in
participants
with
priary

psychopathy
(Vol.
51,
pp.
823-‐828):
Personality
and
Individual
Differences.

Mathiak,
K.
A.,
Zvyagintsev,
M.,
Ackermann,
H.,
&
Mathiak,
K.
(2012).
Lateralization

of
amygdala
activation
in
fMRI
may
depend
on
phase-‐encoding
polarity.

MAGMA,
25(3),
177-‐182.
doi:
10.1007/s10334-‐011-‐0285-‐4

Mathis,
H.
(1970).
Emotional
responsivity
in
the
antisocial
personality.
Unpublished

Doctoral
Dissertation.
Washington,
DC:
George
Washington
University.

McDonald,
A.
J.
(1998).
Cortical
pathways
to
the
mammalian
amygdala.
Prog

Neurobiol,
55(3),
257-‐332.

McLaren,
D.
G.,
Ries,
M.
L.,
Xu,
G.,
&
Johnson,
S.
C.
(2012).
A
generalized
form
of

context-‐dependent
psychophysiological
interactions
(gPPI):
a
comparison
to

standard
approaches.
Neuroimage,
61(4),
1277-‐1286.
doi:

10.1016/j.neuroimage.2012.03.068

Mednick,
S.
A.,
&
Christiansen,
K.
O.
(1977).
Biosocial
Bases
of
Criminal
Behavior.

New
York:
Gardner
Press.

Milad,
M.
R.,
Rauch,
S.
L.,
Pitman,
R.
K.,
&
Quirk,
G.
J.
(2006).
Fear
extinction
in
rats:

implications
for
human
brain
imaging
and
anxiety
disorders.
Biol
Psychol,

73(1),
61-‐71.
doi:
10.1016/j.biopsycho.2006.01.008

170

Mishkin,
S.
(1964).
The
interdependence
of
clinical
neurology
and
neurophysiology.

An
historical
review
of
the
vestibuilo-‐ocular
reflex.
McGill
Med
J,
33,
80-‐97.

Mitchell,
D.
G.,
Richell,
R.
A.,
Leonard,
A.,
&
Blair,
R.
J.
(2006).
Emotion
at
the
expense

of
cognition:
psychopathic
individuals
outperform
controls
on
an
operant

response
task.
J
Abnorm
Psychol,
115(3),
559-‐566.
doi:
2006-‐09167-‐017
[pii]

10.1037/0021-‐843X.115.3.559

Morris,
J.
S.,
&
Dolan,
R.
J.
(2004).
Dissociable
amygdala
and
orbitofrontal
responses

during
reversal
fear
conditioning.
Neuroimage,
22(1),
372-‐380.
doi:

10.1016/j.neuroimage.2004.01.012

Morrison,
S.
E.,
&
Salzman,
C.
D.
(2009).
The
convergence
of
information
about

rewarding
and
aversive
stimuli
in
single
neurons.
J
Neurosci,
29(37),
11471-‐
11483.
doi:
10.1523/JNEUROSCI.1815-‐09.2009

Motzkin,
J.
C.,
Newman,
J.
P.,
Kiehl,
K.
A.,
&
Koenigs,
M.
(2011).
Reduced
prefrontal

connectivity
in
psychopathy.
J
Neurosci,
31(48),
17348-‐17357.
doi:

31/48/17348
[pii]10.1523/JNEUROSCI.4215-‐11.2011

Mukherjee,
P.,
&
McKinstry,
R.
C.
(2006).
Diffusion
tensor
imaging
and
tractography

of
human
brain
development.
Neuroimaging
Clin
N
Am,
16(1),
19-‐43,
vii.
doi:

10.1016/j.nic.2005.11.004

Müller,
J.
L.,
Sommer,
M.,
Wagner,
V.,
Lange,
K.,
Taschler,
H.,
Röder,
C.
H.,
.
.
.
Hajak,
G.

(2003).
Abnormalities
in
emotion
processing
within
cortical
and
subcortical

regions
in
criminal
psychopaths:
evidence
from
a
functional
magnetic

resonance
imaging
study
using
pictures
with
emotional
content.
Biol

Psychiatry,
54(2),
152-‐162.
doi:
S0006322302017493
[pii]

Murray,
E.
A.,
&
Izquierdo,
A.
(2007).
Orbitofrontal
cortex
and
amygdala

contributions
to
affect
and
action
in
primates.
Ann
N
Y
Acad
Sci,
1121,
273-‐
296.
doi:
10.1196/annals.1401.021

Newman,
J.
P.,
&
Kosson,
D.
S.
(1986).
Passive
avoidance
learning
in
psychopathic

and
nonpsychopathic
offenders.
J
Abnorm
Psychol,
95(3),
252-‐256.

Newman,
J.
P.,
Patterson,
C.
M.,
Howland,
E.
W.,
&
Nichols,
S.
L.
(1990).
Passive

avoidance
in
psychopaths:
the
effects
of
reward
(Vol.
11,
pp.
1101-‐1114):

Personal
Individual
Differences.

171

Newman,
J.
P.,
MacCoon,
D.
G.,
Vaughn,
L.
J.,
&
Sadeh,
N.
(2005).
Validating
a

distinction
between
primary
and
secondary
psychopathy
with
measures
of

Gray's
BIS
and
BAS
constructs.
J
Abnorm
Psychol,
114(2),
319-‐323.
doi:
2005-‐
04292-‐014
[pii]10.1037/0021-‐843X.114.2.319

Newman,
J.
P.,
&
Schmitt,
W.
A.
(1998).
Passive
avoidance
in
psychopathic
offenders:

a
replication
and
extension.
J
Abnorm
Psychol,
107(3),
527-‐532.

Newman,
J.
P.,
Schmitt,
W.
A.,
&
Voss,
W.
D.
(1997).
The
impact
of
motivationally

neutral
cues
on
psychopathic
individuals:
assessing
the
generality
of
the

response
modulation
hypothesis.
J
Abnorm
Psychol,
106(4),
563-‐575.

Northoff,
G.,
&
Bermpohl,
F.
(2004).
Cortical
midline
structures
and
the
self.
Trends

Cogn
Sci,
8(3),
102-‐107.
doi:
10.1016/j.tics.2004.01.004

O'Doherty,
J.
P.,
Hampton,
A.,
&
Kim,
H.
(2007).
Model-‐based
fMRI
and
its
application

to
reward
learning
and
decision
making.
Ann
N
Y
Acad
Sci,
1104,
35-‐53.
doi:

10.1196/annals.1390.022

O'Reilly,
J.
X.,
Woolrich,
M.
W.,
Behrens,
T.
E.,
Smith,
S.
M.,
&
Johansen-‐Berg,
H.

(2012).
Tools
of
the
trade:
psychophysiological
interactions
and
functional

connectivity.
Soc
Cogn
Affect
Neurosci,
7(5),
604-‐609.
doi:

10.1093/scan/nss055

Ochsner,
K.
N.,
Bunge,
S.
A.,
Gross,
J.
J.,
&
Gabrieli,
J.
D.
(2002).
Rethinking
feelings:
an

FMRI
study
of
the
cognitive
regulation
of
emotion.
J
Cogn
Neurosci,
14(8),

1215-‐1229.
doi:
10.1162/089892902760807212

Ochsner,
K.
N.,
&
Gross,
J.
J.
(2005).
The
cognitive
control
of
emotion.
Trends
Cogn

Sci,
9(5),
242-‐249.
doi:
10.1016/j.tics.2005.03.010

Ochsner,
K.
N.,
Ray,
R.
D.,
Cooper,
J.
C.,
Robertson,
E.
R.,
Chopra,
S.,
Gabrieli,
J.
D.,
&

Gross,
J.
J.
(2004).
For
better
or
for
worse:
neural
systems
supporting
the

cognitive
down-‐
and
up-‐regulation
of
negative
emotion.
Neuroimage,
23(2),

483-‐499.
doi:
10.1016/j.neuroimage.2004.06.030

Ongür,
D.,
An,
X.,
&
Price,
J.
L.
(1998).
Prefrontal
cortical
projections
to
the

hypothalamus
in
macaque
monkeys.
J
Comp
Neurol,
401(4),
480-‐505.

Ongür,
D.,
Ferry,
A.
T.,
&
Price,
J.
L.
(2003).
Architectonic
subdivision
of
the
human

orbital
and
medial
prefrontal
cortex.
J
Comp
Neurol,
460(3),
425-‐449.
doi:

10.1002/cne.10609

172

Pandya,
D.
N.,
Van
Hoesen,
G.
W.,
&
Mesulam,
M.
M.
(1981).
Efferent
connections
of

the
cingulate
gyrus
in
the
rhesus
monkey.
Exp
Brain
Res,
42(3-‐4),
319-‐330.

Passamonti,
L.,
Fairchild,
G.,
Fornito,
A.,
Goodyer,
I.
M.,
Nimmo-‐Smith,
I.,
Hagan,
C.
C.,

&
Calder,
A.
J.
(2012).
Abnormal
anatomical
connectivity
between
the

amygdala
and
orbitofrontal
cortex
in
conduct
disorder.
PLoS
One,
7(11),

e48789.
doi:
10.1371/journal.pone.0048789

Paton,
J.
J.,
Belova,
M.
A.,
Morrison,
S.
E.,
&
Salzman,
C.
D.
(2006).
The
primate

amygdala
represents
the
positive
and
negative
value
of
visual
stimuli
during

learning.
Nature,
439(7078),
865-‐870.
doi:
10.1038/nature04490

Patrick,
C.
J.,
Edens,
J.
F.,
Poythress,
N.
G.,
Lilienfeld,
S.
O.,
&
Benning,
S.
D.
(2006).

Construct
validity
of
the
psychopathic
personality
inventory
two-‐factor

model
with
offenders.
Psychol
Assess,
18(2),
204-‐208.
doi:
10.1037/1040-‐
3590.18.2.204

Paus,
T.,
Tomaiuolo,
F.,
Otaky,
N.,
MacDonald,
D.,
Petrides,
M.,
Atlas,
J.,
.
.
.
Evans,
A.
C.

(1996).
Human
cingulate
and
paracingulate
sulci:
pattern,
variability,

asymmetry,
and
probabilistic
map.
Cereb
Cortex,
6(2),
207-‐214.

Perry,
D.
G.,
&
Perry,
L.
C.
(1974).
Denial
of
suffering
in
the
victim
as
a
stimulus
to

violence
in
aggressive
boys.
Child
Dev,
45(1),
55-‐62.

Pezawas,
L.,
Meyer-‐Lindenberg,
A.,
Drabant,
E.
M.,
Verchinski,
B.
A.,
Munoz,
K.
E.,

Kolachana,
B.
S.,
.
.
.
Weinberger,
D.
R.
(2005).
5-‐HTTLPR
polymorphism

impacts
human
cingulate-‐amygdala
interactions:
a
genetic
susceptibility

mechanism
for
depression.
Nat
Neurosci,
8(6),
828-‐834.
doi:

10.1038/nn1463

Phelps,
E.
A.,
Delgado,
M.
R.,
Nearing,
K.
I.,
&
LeDoux,
J.
E.
(2004).
Extinction
learning

in
humans:
role
of
the
amygdala
and
vmPFC.
Neuron,
43(6),
897-‐905.
doi:

10.1016/j.neuron.2004.08.042

Phelps,
E.
A.,
O'Connor,
K.
J.,
Gatenby,
J.
C.,
Gore,
J.
C.,
Grillon,
C.,
&
Davis,
M.
(2001).

Activation
of
the
left
amygdala
to
a
cognitive
representation
of
fear.
Nat

Neurosci,
4(4),
437-‐441.
doi:
86110
[pii]10.1038/86110

Pitkänen,
A.,
&
Amaral,
D.
G.
(1998).
Organization
of
the
intrinsic
connections
of
the

monkey
amygdaloid
complex:
projections
originating
in
the
lateral
nucleus.
J

Comp
Neurol,
398(3),
431-‐458.

Prokasy,
W.
F.,
Williams,
W.
C.,
Kumpfer,
K.
L.,
Lee,
W.
Y.,
&
Jenson,
W.
R.
(1973).

173

Differential
SCR
conditioning
with
two
control
baselines:
random
signal
and

signal
absent.
Psychophysiology,
10(2),
145-‐153.

Pujol,
J.,
Batalla,
I.,
Contreras-‐Rodríguez,
O.,
Harrison,
B.
J.,
Pera,
V.,
Hernández-‐Ribas,

R.,
.
.
.
Cardoner,
N.
(2012).
Breakdown
in
the
brain
network
subserving

moral
judgment
in
criminal
psychopathy.
Soc
Cogn
Affect
Neurosci,
7(8),
917-‐
923.
doi:
10.1093/scan/nsr075

Quirk,
G.
J.,
Armony,
J.
L.,
&
LeDoux,
J.
E.
(1997).
Fear
conditioning
enhances

different
temporal
components
of
tone-‐evoked
spike
trains
in
auditory
cortex

and
lateral
amygdala.
Neuron,
19(3),
613-‐624.
doi:
S0896-‐6273(00)80375-‐X

[pii]

Quirk,
G.
J.,
Likhtik,
E.,
Pelletier,
J.
G.,
&
Paré,
D.
(2003).
Stimulation
of
medial

prefrontal
cortex
decreases
the
responsiveness
of
central
amygdala
output

neurons.
J
Neurosci,
23(25),
8800-‐8807.

Quirk,
G.
J.,
&
Beer,
J.
S.
(2006).
Prefrontal
involvement
in
the
regulation
of
emotion:

convergence
of
rat
and
human
studies.
Curr
Opin
Neurobiol,
16(6),
723-‐727.

doi:
10.1016/j.conb.2006.07.004

Raine,
A.
(1993).
The
psychopathology
of
crime.
New
York:
Academic
Press.

Raine,
A.,
Brennan,
P.,
&
Mednick,
S.
A.
(1994).
Birth
complications
combined
with

early
maternal
rejection
at
age
1
year
predispose
to
violent
crime
at
age
18

years.
Arch
Gen
Psychiatry,
51(12),
984-‐988.

Rilling,
J.
K.,
Glenn,
A.
L.,
Jairam,
M.
R.,
Pagnoni,
G.,
Goldsmith,
D.
R.,
Elfenbein,
H.
A.,
&

Lilienfeld,
S.
O.
(2007).
Neural
correlates
of
social
cooperation
and
non-‐
cooperation
as
a
function
of
psychopathy.
Biol
Psychiatry,
61(11),
1260-‐1271.

doi:
10.1016/j.biopsych.2006.07.021

Rolls,
E.
T.,
Critchley,
H.
D.,
Browning,
A.
S.,
Hernadi,
I.,
&
Lenard,
L.
(1999).

Responses
to
the
sensory
properties
of
fat
of
neurons
in
the
primate

orbitofrontal
cortex.
J
Neurosci,
19(4),
1532-‐1540.

Rolls,
E.
T.,
&
Grabenhorst,
F.
(2008).
The
orbitofrontal
cortex
and
beyond:
from

affect
to
decision-‐making.
Prog
Neurobiol,
86(3),
216-‐244.
doi:

10.1016/j.pneurobio.2008.09.001

Romanski,
L.
M.,
Tian,
B.,
Fritz,
J.,
Mishkin,
M.,
Goldman-‐Rakic,
P.
S.,
&
Rauschecker,
J.

174

P.
(1999).
Dual
streams
of
auditory
afferents
target
multiple
domains
in
the

primate
prefrontal
cortex.
Nat
Neurosci,
2(12),
1131-‐1136.
doi:

10.1038/16056

Roy,
A.
K.,
Shehzad,
Z.,
Margulies,
D.
S.,
Kelly,
A.
M.,
Uddin,
L.
Q.,
Gotimer,
K.,
.
.
.

Milham,
M.
P.
(2009).
Functional
connectivity
of
the
human
amygdala
using

resting
state
fMRI.
Neuroimage,
45(2),
614-‐626.
doi:

10.1016/j.neuroimage.2008.11.030

Rushworth,
M.
F.,
&
Behrens,
T.
E.
(2008).
Choice,
uncertainty
and
value
in

prefrontal
and
cingulate
cortex.
Nat
Neurosci,
11(4),
389-‐397.
doi:

10.1038/nn2066

Sadeh,
N.,
Spielberg,
J.
M.,
Heller,
W.,
Herrington,
J.
D.,
Engels,
A.
S.,
Warren,
S.
L.,
.
.
.

Miller,
G.
A.
(2013).
Emotion
disrupts
neural
activity
during
selective

attention
in
psychopathy.
Soc
Cogn
Affect
Neurosci,
8(3),
235-‐246.
doi:

10.1093/scan/nsr092

Salzman,
C.
D.,
Paton,
J.
J.,
Belova,
M.
A.,
&
Morrison,
S.
E.
(2007).
Flexible
neural

representations
of
value
in
the
primate
brain
(Vol.
1121,
pp.
336-‐354).
Ann
N

Y
Acad
Sci.

Sarkar,
S.,
Craig,
M.
C.,
Catani,
M.,
Dell'acqua,
F.,
Fahy,
T.,
Deeley,
Q.,
&
Murphy,
D.
G.

(2013).
Frontotemporal
white-‐matter
microstructural
abnormalities
in

adolescents
with
conduct
disorder:
a
diffusion
tensor
imaging
study.
Psychol

Med,
43(2),
401-‐411.
doi:
10.1017/S003329171200116X

Schiller,
D.,
Levy,
I.,
Niv,
Y.,
LeDoux,
J.
E.,
&
Phelps,
E.
A.
(2008).
From
fear
to
safety

and
back:
reversal
of
fear
in
the
human
brain.
J
Neurosci,
28(45),
11517-‐
11525.
doi:
10.1523/JNEUROSCI.2265-‐08.2008

Schmauk,
E.
J.
(1970).
Punishment,
arousal,
and
avoidance
learning
in
sociopaths

(Vol.
76,
pp.
325-‐335):
Journal
of
Abnormal
Psychology.

Schneider,
F.,
Habel,
U.,
Kessler,
C.,
Posse,
S.,
Grodd,
W.,
&
Müller-‐Gärtner,
H.
W.

(2000).
Functional
imaging
of
conditioned
aversive
emotional
responses
in

antisocial
personality
disorder.
Neuropsychobiology,
42(4),
192-‐201.
doi:

nps42192
[pii]

Schoenbaum,
G.,
&
Roesch,
M.
(2005).
Orbitofrontal
cortex,
associative
learning,
and

expectancies.
Neuron,
47(5),
633-‐636.
doi:
10.1016/j.neuron.2005.07.018

Schoenbaum,
G.,
Setlow,
B.,
&
Ramus,
S.
J.
(2003).
A
systems
approach
to

175

orbitofrontal
cortex
function:
recordings
in
rat
orbitofrontal
cortex
reveal

interactions
with
different
learning
systems.
Behav
Brain
Res,
146(1-‐2),
19-‐
29.

Schoenbaum,
G.,
Setlow,
B.,
Saddoris,
M.
P.,
&
Gallagher,
M.
(2003).
Encoding

predicted
outcome
and
acquired
value
in
orbitofrontal
cortex
during
cue

sampling
depends
upon
input
from
basolateral
amygdala.
Neuron,
39(5),

855-‐867.

Shackman,
A.
J.,
Salomons,
T.
V.,
Slagter,
H.
A.,
Fox,
A.
S.,
Winter,
J.
J.,
&
Davidson,
R.
J.

(2011).
The
integration
of
negative
affect,
pain
and
cognitive
control
in
the

cingulate
cortex.
Nat
Rev
Neurosci,
12(3),
154-‐167.
doi:
10.1038/nrn2994

Sheng,
T.,
Gheytanchi,
A.,
&
Aziz-‐Zadeh,
L.
(2010).
Default
network
deactivations
are

correlated
with
psychopathic
personality
traits.
PLoS
One,
5(9),
e12611.
doi:

10.1371/journal.pone.0012611

Skeem,
J.,
Johansson,
P.,
Andershed,
H.,
Kerr,
M.,
&
Louden,
J.
E.
(2007).
Two

subtypes
of
psychopathic
violent
offenders
that
parallel
primary
and

secondary
variants.
J
Abnorm
Psychol,
116(2),
395-‐409.
doi:
2007-‐06673-‐015

[pii]10.1037/0021-‐843X.116.2.395

Skeem,
J.
L.,
Mulvey,
E.
P.,
&
Grisso,
T.
(2003).
Applicability
of
traditional
and
revised

models
of
psychopathy
to
the
Psychopathy
Checklist:
screening
version.

Psychol
Assess,
15(1),
41-‐55.

Sloper,
D.S.
(2013).
Interaction
software.

www.danielsoper.com/Interaction/default.aspx

Smith,
S.
M.
(2002).
Fast
robust
automated
brain
extraction.
Hum
Brain
Mapp,
17(3),

143-‐155.
doi:
10.1002/hbm.10062

Smith,
S.
S.,
Arnett,
P.
A.,
&
Newman,
J.
P.
(1992).
Neuropsychological
differentiation

of
psychopathic
and
nonpsychopathic
criminal
offenders.
(Vol.
13,
pp.
1233-‐
1245):
Personal
Individual
Differences.

Sommer,
M.,
Sodian,
B.,
Döhnel,
K.,
Schwerdtner,
J.,
Meinhardt,
J.,
&
Hajak,
G.
(2010).

In
psychopathic
patients
emotion
attribution
modulates
activity
in
outcome-‐
related
brain
areas.
Psychiatry
Res,
182(2),
88-‐95.
doi:

10.1016/j.pscychresns.2010.01.007

Sotres-‐Bayon,
F.,
&
Quirk,
G.
J.
(2010).
Prefrontal
control
of
fear:
more
than
just

176

extinction.
Curr
Opin
Neurobiol,
20(2),
231-‐235.
doi:

10.1016/j.conb.2010.02.005

Stahl,
D.,
&
Sallis,
H.
(2012).
Model-‐based
cluster
analysis
(Vol.
4,
pp.
341-‐358):

Wiley
Interdisciplinary
Reviews:
Computational
Statistics.

Stefanacci,
L.,
&
Amaral,
D.
G.
(2002).
Some
observations
on
cortical
inputs
to
the

macaque
monkey
amygdala:
an
anterograde
tracing
study.
J
Comp
Neurol,

451(4),
301-‐323.
doi:
10.1002/cne.10339

Sterzer,
P.,
Stadler,
C.,
Krebs,
A.,
Kleinschmidt,
A.,
&
Poustka,
F.
(2005).
Abnormal

neural
responses
to
emotional
visual
stimuli
in
adolescents
with
conduct

disorder.
Biol
Psychiatry,
57(1),
7-‐15.
doi:
10.1016/j.biopsych.2004.10.008

Sterzer,
P.,
Stadler,
C.,
Poustka,
F.,
&
Kleinschmidt,
A.
(2007).
A
structural
neural

deficit
in
adolescents
with
conduct
disorder
and
its
association
with
lack
of

empathy.
Neuroimage,
37(1),
335-‐342.
doi:

10.1016/j.neuroimage.2007.04.043

Stevens,
D.,
Charman,
T.,
&
Blair,
R.
J.
(2001).
Recognition
of
emotion
in
facial

expressions
and
vocal
tones
in
children
with
psychopathic
tendencies.
J

Genet
Psychol,
162(2),
201-‐211.

Sutton,
S.
K.,
Vitale,
J.
E.,
&
Newman,
J.
P.
(2002).
Emotion
among
women
with

psychopathy
during
picture
perception.
J
Abnorm
Psychol,
111(4),
610-‐619.

Swogger,
M.
T.,
Walsh,
Z.,
&
Kosson,
D.
S.
(2008).
Psychopathy
subtypes
among

African
American
county
jail
inmates
(Vol.
35,
pp.
1484-‐1499):
Criminal

justice
and
behavior.

Tow,
P.
M.,
&
Whitty,
C.
W.
(1953).
Personality
changes
after
operations
on
the

cingulate
gyrus
in
man.
J
Neurol
Neurosurg
Psychiatry,
16(3),
186-‐193.

Tukey,
J.
W.
(1977).
Some
thoughts
on
clinical
trials,
especially
problems
of

multiplicity.
Science,
198(4318),
679-‐684.

Urry,
H.
L.,
van
Reekum,
C.
M.,
Johnstone,
T.,
Kalin,
N.
H.,
Thurow,
M.
E.,
Schaefer,
H.

S.,
.
.
.
Davidson,
R.
J.
(2006).
Amygdala
and
ventromedial
prefrontal
cortex

are
inversely
coupled
during
regulation
of
negative
affect
and
predict
the

diurnal
pattern
of
cortisol
secretion
among
older
adults.
J
Neurosci,
26(16),

4415-‐4425.
doi:
10.1523/JNEUROSCI.3215-‐05.2006

177

Vassileva,
J.,
Kosson,
D.
S.,
Abramowitz,
C.,
&
Conrod,
P.
(2005).
Psychopathy
versus

psychopathies
in
classifying
criminal
offenders
(Vol.
10,
pp.
27-‐43):
Legal
and

Criminological
Psychology.

Vaughn,
M.
G.,
Edens,
J.
F.,
Howard,
M.
O.,
&
Smith,
S.
T.
(2009).
An
investigation
of

primary
and
secondary
psychopathy
in
a
statewide
sample
of
incarcerated

youth
(Vol.
7,
pp.
172-‐188).
Youth
Violence
and
Juvenile
Justice.

Veit,
R.,
Flor,
H.,
Erb,
M.,
Hermann,
C.,
Lotze,
M.,
Grodd,
W.,
&
Birbaumer,
N.
(2002).

Brain
circuits
involved
in
emotional
learning
in
antisocial
behavior
and
social

phobia
in
humans.
Neurosci
Lett,
328(3),
233-‐236.
doi:
S0304394002005190

[pii]

Veit,
R.,
Lotze,
M.,
Sewing,
S.,
Missenhardt,
H.,
Gaber,
T.,
&
Birbaumer,
N.
(2010).

Aberrant
social
and
cerebral
responding
in
a
competitive
reaction
time

paradigm
in
criminal
psychopaths.
Neuroimage,
49(4),
3365-‐3372.
doi:

10.1016/j.neuroimage.2009.11.040

Vidal,
S.,
Skeem,
J.,
&
Camp,
J.
(2010).
Emotional
intelligence:
painting
different
paths

for
low-‐anxious
and
high-‐anxious
psychopathic
variants.
Law
Hum
Behav,

34(2),
150-‐163.
doi:
10.1007/s10979-‐009-‐9175-‐y

Vitale,
J.
E.,
Newman,
J.
P.,
Bates,
J.
E.,
Goodnight,
J.,
Dodge,
K.
A.,
&
Pettit,
G.
S.
(2005).

Deficient
behavioral
inhibition
and
anomalous
selective
attention
in
a

community
sample
of
adolescents
with
psychopathic
traits
and
low-‐anxiety

traits
(Vol.
33,
pp.
461-‐470).
Journal

of
Abnormal
Child
Psychology.

Waldman,
I.
D.,
&
Rhee,
S.
H.
(2007).
Genetic
and
environmental
influences
on

psychopathy
and
antisocial
behavior.
In
C.
J.
Patrick
(Ed.),
Handbook
of

psychopathy
(pp.
205-‐228.).
New
York:
Guilford.

Wallis,
J.
D.
(2007).
Orbitofrontal
cortex
and
its
contribution
to
decision-‐making.

Annu
Rev
Neurosci,
30,
31-‐56.
doi:

10.1146/annurev.neuro.30.051606.094334

Wallis,
J.
D.,
&
Miller,
E.
K.
(2003).
Neuronal
activity
in
primate
dorsolateral
and

orbital
prefrontal
cortex
during
performance
of
a
reward
preference
task.
Eur

J
Neurosci,
18(7),
2069-‐2081.

Wallis,
J.
D.,
&
Miller,
E.
K.
(2003).
From
rule
to
response:
neuronal
processes
in
the

178

premotor
and
prefrontal
cortex.
J
Neurophysiol,
90(3),
1790-‐1806.
doi:

10.1152/jn.00086.2003

Walton,
M.
E.,
Behrens,
T.
E.,
Buckley,
M.
J.,
Rudebeck,
P.
H.,
&
Rushworth,
M.
F.

(2010).
Separable
learning
systems
in
the
macaque
brain
and
the
role
of

orbitofrontal
cortex
in
contingent
learning.
Neuron,
65(6),
927-‐939.
doi:

10.1016/j.neuron.2010.02.027

White,
S.
F.,
Marsh,
A.
A.,
Fowler,
K.
A.,
Schechter,
J.
C.,
Adalio,
C.,
Pope,
K.,
.
.
.
Blair,
R.

J.
(2012).
Reduced
amygdala
response
in
youths
with
disruptive
behavior

disorders
and
psychopathic
traits:
decreased
emotional
response
versus

increased
top-‐down
attention
to
nonemotional
features.
Am
J
Psychiatry,

169(7),
750-‐758.
doi:
10.1176/appi.ajp.2012.11081270

Widom,
C.
S.
(1977).
A
methodology
for
studying
noninstitutionalized
psychopaths.
J

Consult
Clin
Psychol,
45(4),
674-‐683.

Wolter,
J.,
&
Lachnit,
H.
(1993).
Are
anticipatory
first
and
second
interval
skin

conductance
responses
indicators
of
predicted
aversiveness?
Integr
Physiol

Behav
Sci,
28(2),
163-‐166.

Yang,
Y.,
Raine,
A.,
Lencz,
T.,
Bihrle,
S.,
LaCasse,
L.,
&
Colletti,
P.
(2005).
Volume

reduction
in
prefrontal
gray
matter
in
unsuccessful
criminal
psychopaths.

Biol
Psychiatry,
57(10),
1103-‐1108.
doi:
S0006-‐3223(05)00098-‐3

[pii]10.1016/j.biopsych.2005.01.021

Yang,
Y.,
Raine,
A.,
Narr,
K.
L.,
Colletti,
P.,
&
Toga,
A.
W.
(2009).
Localization
of

deformations
within
the
amygdala
in
individuals
with
psychopathy.
Arch
Gen

Psychiatry,
66(9),
986-‐994.
doi:
66/9/986

[pii]10.1001/archgenpsychiatry.2009.110

Yeh,
M.
T.,
Chen,
P.,
Raine,
A.,
Baker,
L.
A.,
&
Jacobson,
K.
C.
(2011).
Child

psychopathic
traits
moderate
relationships
between
parental
affect
and
child

aggression.
J
Am
Acad
Child
Adolesc
Psychiatry,
50(10),
1054-‐1064.
doi:

10.1016/j.jaac.2011.06.013

Young,
L.,
Bechara,
A.,
Tranel,
D.,
Damasio,
H.,
Hauser,
M.,
&
Damasio,
A.
(2010).

Damage
to
ventromedial
prefrontal
cortex
impairs
judgment
of
harmful

intent.
Neuron,
65(6),
845-‐851.
doi:
S0896-‐6273(10)00172-‐8
[pii]

10.1016/j.neuron.2010.03.003

Zald, D. H., & Kim, S. W. (2001). The orbitofrontal cortex.

179

Zecevic,
N.,
&
Rakic,
P.
(2001).
Development
of
layer
I
neurons
in
the
primate

cerebral
cortex.
J
Neurosci,
21(15),
5607-‐5619.

Zeier,
J.
D.,
Maxwell,
J.
S.,
&
Newman,
J.
P.
(2009).
Attention
moderates
the

processing
of
inhibitory
information
in
primary
psychopathy.
J
Abnorm

Psychol,
118(3),
554-‐563.
doi:
2009-‐12104-‐011
[pii]10.1037/a0016480

Zolondek,
S.,
Lilienfeld,
S.
O.,
Patrick,
C.
J.,
&
Fowler,
K.
A.
(2006).
The
Interpersonal

Measure
of
Psychopathy
Construct
and
Incremental
Validity
in
Male

Prisoners
(Vol.
13,
pp.
470-‐482):
Assessment.

180

APPENDIX I: Recruitment Information

Participants were recruited over six mailings. All potential participants were
drawn from Wave 3 of the USC Twin Study. The first two mailings were sent to
18 year olds only, to avoid overlap with Wave 5 of the USC Twin Study (age 19-
21), which was concurrently being run. The third through sixth mailings
continued to invite 18 year olds (who had turned 18 since the previous mailings),
as well as including 19-21 year olds who had already taken part in Wave 5 of the
Twin Study. Recruitment and participation outcomes broken down by levels of
psychopathic from the various mailings can be seen in Tables 1 and 2.

Table 1. Number of participation invitations by level of psychopathic traits
and wave of recruitment.
TOTAL INVITED
Mailing LOW HIGH TOTAL
1
st
20 (44%) 25 (56%) 45
2
nd
9 (56%) 7 (44%) 16
3
rd
10 (50%) 10 (50%) 20
4
th
5 (100%) 0 (0%) 5
5
th
5 (50%) 5 (50%) 10
6
th
11 (61%) 7 (39%) 18
Note. Low = low psychopathic traits (i.e. recruited from Quartile 1, 0-25
th

percentile). High = high psychopathic traits (i.e. recruited from Quartile 4, 75-100
th

percentile).

A total of 114 individuals were invited to participate in the studies of this
dissertation, with 24 final participants (21% of those invited) being scanned.
Reasons for not participating included: not being interested, interested but
stopped replying, no response, not being MRI eligible, and being away at college.

181

182

APPENDIX
II:
Zygosity
and
Co-‐Twin
Information

While the current work is not a twin study, the participant pool was comprised of
individuals who were twins. Due to statistical dependency issues, only one twin
of a pair was invited to participate, and the participant was chosen at random if
both were eligible. However, provided here is information on zygosity for
participants as compared to the original participant pool (see Table 1), as well as
psychopathy quartile information for co-twins (see Figure 1). In the potential
participant pool, quartile 1 had roughly half monozygotic (MZ) and dizygotic (DZ)
twins and quartile 4 had slightly more DZ than MZ twins. The actual participant
pool differed from this breakdown. In actual participants, three quarters of
quartile 1 was comprised of MZ twins, and quartile 4 had slightly more MZ than
DZ twins. Future studies with larger sample sizes are needed to investigate the
how zygosity is related to the expression of psychopathic traits in the brain.
Table 1. Zygosity of potential participant pool and actual participant pool.
Potential Participants Actual Participants
Quartile 1 Quartile 4 Quartile 1 Quartile 4

Monozygotic 45% 40.63% 75% 58.3%
Dizygotic
(same sex)
23.35% 34.38% 8.3% 25%
Dizygotic
(Different sex)
28.74% 23.44% 16.6% 16.6%

183

The co-twins of participants from quartile 1 were mostly also in quartiles 1 or 2.
The co-twins of participants from quartile 4 were more varied in their quartiles,
with most being in quartile 3 and the rest being spread out over quartiles 1, 2,
and 4.

Figure 1. Quartile placement of co-twins by participant quartile

0

1

2

3

4

5

6

7

8

1
2
3
4

Number
of
individuals

Co-‐twin
quartile

Participant
Quartile
1

Participant
Quartile
4

184

APPENDIX III: Welsh Anxiety Scale as a measure of trait
anxiety

Psychopathy studies that include trait anxiety measures to define subtypes
typically use the MMPI-derived Welsh Anxiety Scale (WAS; Welsh, 1956)
(Kosson, Smith & Newman, 1990; Lykken, 1957; Newman & Kosson, 1986;
Newman, Patterson, Howland, & Nichols, 1990) to assess trait anxiety.
However, other trait anxiety measures are more typically used in the anxiety
literature, such as the trait scale of the State-Trait Anxiety Inventory (STAI-T;
Spielberger, 1983). In the current study, both WAS and STAI-T were
administered to participants. The WAS is a 39-item, self-report scale that is
derived from the MMPI (Minnesota Multiphasic Personality Inventory; Greene,
1999). The trait scale of the State-Trait Anxiety Inventory (STAI; Spielberger,
1983) is a self-report questionnaire designed to assess an individual’s proneness
to general anxiety and negative affect. Some argue that the items of the WAS
reflect negative affect more so than trait anxiety, and that the behavioral
differences observed between high- and low- anxious psychopathy are unique to
classification using WAS, and not robust to other measures of trait anxiety (Hale,
2002). Therefore, we administer both WAS and STAI-T to examine whether
STAI-T also reveals trait anxiety-psychopathy relationships. Statistical analysis
of all questionnaire data was performed using Statistical Product and Service
Solutions (SPSS; version 19, SPSS, inc., Chicago).

185

Results
Means for the WAS scores (mean ± SD = 11.38 ± 8.72) were comparable to
those previously reported for non-clinical samples of adult males (Mean = 10.5;
Colligan & Offord, 1988). The mean STAI-T scores were also comparable to
those found in college aged male samples (Spielberger; mean = 38.3). Internal
consistency was high for both WAS (α = 0.93) and STAI-T (α = 0.91).
Correlations reveal that WAS and STAI-T were significantly positively correlated,
in line with previous results (Hale, 2002). Also in line with previous findings
(Patrick, 1994; Schmitt & Newman, 1999), neither trait anxiety score was
significantly correlated with either total psychopathy score (see Table 1).

Table 1. Descriptive statistics and correlations for trait anxiety and
psychopathy measures
Mean SD Correlations
PPI Comp-P STAI-T WAS
PPI 287.5 34.8 - .35 -.24 -0.08
Comp-P -0.05 1 - -.12 -0.01
STAI-T 35.4 8.5 - 0.55*
WAS 11.38 8.72 -
Note: PPI = Psychopathic Personality Inventory; Comp-P = Composite
Psychopathy Score; STAI-T = Spielberger State-Trait Anxiety Inventory, Trait
version; WAS = Welsh Anxiety Scale. * indicates significance at p < 0.01. n = 24
for all measures.

186

With regard to subscales of the PPI (see Table 2), both trait anxiety scales were
negatively correlated with the total PPI-I score and PPI-I subscales, except the
correlation between WAS and Fearlessness, which did not reach significance.
These findings also corroborate previous findings between trait anxiety scales
and psychopathy factor 1 scores, which reflect affective facets of psychopathy,
and are thought to be related to fearlessness/low anxiety (Frick, 1998). In
summary, although the degree of correlations between each trait anxiety scale
and psychopathy scales slightly differs, the relatively same relationships emerge,
suggesting shared variance.

The series of studies in this dissertation found a number of significant effects of
trait anxiety (as measured by WAS) interacting with psychopathic traits in
predicting brain activity, as well as skin conductance responding. To assess
Table 2. Descriptive Statistics for trait anxiety and psychopathy subscales

Note: Comp-P = Composite Psychopathy Score; STAI-T = Spielberger State-Trait Anxiety Inventory, Trait version; WAI = Welsh
Anxiety Inventory; PPI-ME = Machiavallian Egocentricity; PPI-RN = Rebellious Nonconformity; PPI-BE = Blame Externalization;
PPI-CN = Carefree Nonplanfulness; PPI-SOI = Social Influence; PPI-F = Fearlessness; PPI-STI = Stress Immunity; PPI-C =
Coldheartedness. * indicates significance at p < .05; ** indicates significance at p < .01. n = 24 for all scales.
Mean SD
Comp-P STAI-T WAI
PPI-
ME
PPI-
RN
PPI-
BE
PPI-
CN
PPI-
SOI
PPI-
F
PPI-
STI
PPI-
C

Comp-P -0.05 1 - -.12 .00 .57** .23 .33 .27 .25 .45* .27 .60**
STAI-T 35.38 8.54 - -.55** -.09 -.10 .017 .60** -.60** -.43* -.63** .07
WAI 66.67 8.78 - .17 -.10 .04 .27 -.54** -.18 -.59** -.19
PPI-ME 41.29 8.43 - .37 .31 .25 .15 .29 -.04 .11
PPI-RN 30.83 5.87 - .57** .08 .51* .40 .29 .01
PPI-BE 26.75 5.89 - .28 .17 -.07 -.01 .42*
PPI-CN 37.71 8.95 - -.48* -.21 -.57** .53**
PPI-SOI 37.46 7.47 - .52** .82** .15
PPI-F 37.25 9.86 - .57** .02
PPI-STI 37.58 6.85 - .12
PPI-C 34.75 7.49 -

187

whether WAS and STAI-T are tapping enough of the same construct to replicate
the multiple regression results, each significant multiple regression model was
run again with STAI-T scores. None of the models, or the interaction terms were
significant.

These results raise the question of what exact construct the WAS is measuring.
A closer inspection of the WAS items reveal items varying in classification from
depression to attention to anxiety (as does the STAI-T). While many of these
questions are often included on trait anxiety scales because the constructs they
represent are consequences of anxiety or are closely related, they make
interpretation of results difficult. Thus, for the current studies, in order to have
confidence in interpreting WAS as a trait anxiety scale, the WAS items thought to
represent pure trait anxiety were selected and combined to form a new WAS
score (WAS-anx; included items: 6, 7, 10, 11,12, 14, 15, 16, 17, 20, 22, 24, 25,
27, 28, 29, 30, 31, 32, 33, 34, 37, 38). Items thought to represent the following
constructs were removed: depression (e.g. “Most of the time I feel blue”),
attention (e.g. “I find it hard to keep my mind on a task or job”), fantasy (e.g. “I
have a daydream life about which I do not tell other people”), and remorse (e.g. “I
do many things which I regret afterwards”).

188

Using the new score, all significant models from the studies were reanalyzed.
Every previously significant multiple regression model and interaction term of the
brain analyses remained significant using WAS-anx. (see Table 3 for model
summary; see Table 4 for unstandardized and standardized coefficients, and t-
values). However, the two findings for skin conductance responding were no
longer significant (see Table 5 for model summaries; see Table 6 for
unstandardized and standardized coefficients, and t-values).

Table 3. Model statistics for reanalyzed significant multiple regression
models of brain data with trait anxiety items of WAS

R
2

Adjusted
R
2
F p
Right
Amygdala
PPI_ME,
WAS-anx 0.40 0.31 4.49 0.02
Right
Amygdala -
VMPFC

PPI, WAS-
anx 0.53 0.46 7.57 0.001

Note. PPI = Psychopathic Personality Inventory; PPI_ME = Machiavellian
Egocentricity; WAS-anx = Welsh Anxiety Inventory, trait anxiety items only.

189

Table 4. Unstandardized and standardized coefficients and t-values of
reanalyzed significant multiple regression models of brain data with trait
anxiety items of WAS
Unstandardized
Coefficients
Standardized
Coefficients

B SE β t
Right
Amygdala
(Constant) -4.16 2.84 -1.46
PPI_ME --0.32 0.35 -0.17 -0.93
WAS-anx 0.71 0.57 0.23 1.26
PPI_ME *
WAS-anx
0.20 0.07 0.50 2.75*
Right
Amygdala -
VMPFC
(Constant) 0.00 0.04 0.10
PPI -0.003 0.001 -0.33 -1.95
WAS-anx -0.01 0.01 -0.20 -1.32
PPI * WAS-
anx
0.00 0.00 -0.52 -3.07*

Note. PPI = Psychopathic Personality Inventory; PPI_ME = Machiavellian
Egocentricity; WAS-anx = Welsh Anxiety Inventory, trait anxiety items only.

Table 5. Model statistics for reanalyzed significant multiple regression
models of skin conductance data with trait anxiety items of WAS

R
2

Adjusted
R
2
F p
OR
PPI_BE,
WAS-anx 0.26 0.13 1.98 0.16
OR

PPI_CN,
WAS-anx 0.25 0.09 1.63 0.22

Note. OR = orienting response difference score; PPI_BE = Blame
Externalization; PPI_CN = Carefree Nonplanfulness; WAS-anx = Welsh Anxiety
Inventory, trait anxiety items only.

190

Table 6. Unstandardized and standardized coefficients and t-values of
reanalyzed significant multiple regression models of skin conductance with
trait anxiety items of WAS
Unstandardized
Coefficients
Standardized
Coefficients

B SE β t
OR (Constant) 0.19 0.06 3.59
PPI_BE --0.02 0.01 -0.44 -2.07
WAS-anx 0.01 0.01 0.14 0.59
PPI_BE *
WAS-anx
0.00 0.00 0.21 0.83
OR (Constant) 0.21 0.06 3.32
PPI_CN -0.01 0.01 -0.46 -2.02
WAS-anx 0.01 0.01 0.14 0.61
PPI_CN *
WAS-anx
0.00 0.00 0.12 0.59

Note. PPI = Psychopathic Personality Inventory; PPI_ME = Machiavellian
Egocentricity; WAS-anx = Welsh Anxiety Inventory, trait anxiety items only.

In summary, although the STAI-T was unable to significantly moderate the
relationships between psychopathic traits and brain activity or skin conductance
responding, it can be reported with confidence that the trait anxiety items of WAS
are responsible for the observed moderating effects in all brain results of this
dissertation. Results from skin conductance analyses suggest that the other
constructs represented by the WAS (e.g. depression), which are consequences
of trait anxiety, are important in moderating the relationship between
psychopathic traits and skin conductance responding. Further studies will be
needed to parse the unique and shared variances between STAI-T and WAS.

Abstract (if available)

Abstract Psychopathy is a personality disorder comprised of a constellation of traits, such as lack of remorse, lack of empathy, impulsivity, and irresponsibility. Although much research has been conducted on this personality disorder, the neural correlates remain poorly understood. Progress in this realm has been hindered by two factors. First, given the large body of work supporting delineation of psychopathy into subtypes (based on trait anxiety), the exploration of dissociable neural correlates between subtypes has been insufficiently studied. Second, many studies utilize a single neuroimaging modality to explore a given neural property, and thus, cannot answer comprehensive questions about neural network dynamics. The current work addresses both limitations in the literature by examining, in a community sample, the relationships between psychopathic traits, trait anxiety, and neural properties (function, functional connectivity, microstructural integrity) using multiple neuroimaging methods. The two overarching aims of this dissertation are to determine whether the relationship between psychopathic traits and neural properties is dependent upon trait anxiety, and whether these relationships are evident across several different neuroimaging modalities.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The brain's virtuous cycle: an investigation of gratitude and good human conduct

PDF

Mapping the neural architecture of concepts

PDF

Engagement of the action observation network through functional magnetic resonance imaging with implications for stroke rehabilitation

PDF

Psychometric and biometric evaluation of psychopathic personality traits in 9 and 10 year old twins

PDF

Behavioral and neural influences of interoception and alexithymia on emotional empathy in autism spectrum disorder

PDF

Structural and functional neural correlates of developmental dyspraxia in the mirror neuron system

PDF

The neural correlates of creativity and perceptual pleasure: from simple shapes to humor

PDF

Genotype and sex effects on anxiety behavior and functional activation in corticolimbic circuits in serotonin transporter knockout mice (5-HITT KO)

PDF

The neural correlates of face recognition

PDF

Functional models of fMRI BOLD signal in the visual cortex

PDF

The behavioral and neural correlates of music-evoked nostalgia across the lifespan

PDF

Biometric models of psychopathic traits in adolescence: a comparison of item-level and sum-score approaches

PDF

Exploring the genetic basis of complex traits

PDF

Psychophysiological reactivity to stress, psychopathic traits, and developmental trajectories of antisocial behavior: a prospective, longitudinal study

PDF

Heart, brain, and breath: studies on the neuromodulation of interoceptive systems

PDF

Pattern detection in medical imaging: pathology specific imaging contrast, features, and statistical models

PDF

Using neuroinformatics to identify genomic and proteomic markers of suboptimal aging and Alzheimer's disease

PDF

The role of blood pressure variability in cognitive decline, cerebrovascular disease and Alzheimer’s disease

PDF

Different genetic and environmental structures for the overlap of three antisocial behavior factors with alcohol initiation

PDF

Examining the neuroanatomical correlates of reading in developmental dyslexia

Asset Metadata

Creator Sobhani, Mona (author)

Core Title Psychopathic traits and the fronto-amygdala circuit in a community sample: the role of trait anxiety

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Neuroscience

Publication Date 08/28/2013

Defense Date 07/10/2013

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

Tag amygdala,community,DTI,fMRI,OAI-PMH Harvest,prefrontal,psychopathy

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Aziz-Zadeh, Lisa (committee chair), Baker, Laura A. (committee chair), Damasio, Hanna (committee member), Saks, Elyn (committee member)

Creator Email mssobhani@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-324192

Unique identifier UC11293201

Identifier etd-SobhaniMon-2025.pdf (filename),usctheses-c3-324192 (legacy record id)

Legacy Identifier etd-SobhaniMon-2025.pdf

Dmrecord 324192

Document Type Dissertation

Format application/pdf (imt)

Rights Sobhani, Mona

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA