Citation
Emotional Reactivity to Picture Stimuli in Children with and without Attention-Deficit/Hyperactivity Disorder

Material Information

Title:
Emotional Reactivity to Picture Stimuli in Children with and without Attention-Deficit/Hyperactivity Disorder
Copyright Date:
2008

Subjects

Subjects / Keywords:
Attention deficit hyperactivity disorder ( jstor )
Child psychology ( jstor )
Conduct disorder ( jstor )
Galvanic skin response ( jstor )
Mental stimulation ( jstor )
Nurture ( jstor )
Reactivity ( jstor )
Self reports ( jstor )
Startle reflex ( jstor )
Symptomatology ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright the author. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
8/31/2007
Resource Identifier:
53997096 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

EMOTIONAL REACTIVITY TO PICTURE STIMULI IN CHIL DREN WITH AND WITHOUT ATTENTION-DEFICIT/HYPERACTIVITY DISORDER By GREGG JONATHAN SELKE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Gregg Jonathan Selke

PAGE 3

I would like to dedicate my dissertation to two women, one who now watches over me from above, and the other who looks after me with unconditional l ove and inspiration. My always loving grandmother, Sophie Koffe r Rippel, who taught me so much about what is truly important in life. The love of my life, Sarah Brinton Clark, for whom I am ever thankful for her encour agement, patience, and love.

PAGE 4

ACKNOWLEDGMENTS I would like to thank my two mentors, Eileen Fennell, Ph.D., and Dawn Bowers, Ph.D., for their support, kindness, and wisdom. This project could not have come to fruition without their direction and research expertise. Additionally, I would like to thank my many research assistants, including Diane Harrell, Laura Williams, Sharmeen Qudrot, Melanie Brewster, Amber Beckwith, and Lisa Hurley. I would like to thank my family, my loving parents, Marlene and Gerald, and my sisters, Melissa, Jodi, and Kara, who are always there unconditionally. I thank them for keeping things in perspective for me. I would also like to thank my nieces and nephews, Marissa, Jeffrey, Jaimie, Carly, and Amanda, and Jessica who always have the ability to keep a smile on my face and a glow in my heart. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................ix LIST OF FIGURES.............................................................................................................x ABSTRACT.....................................................................................................................xiii CHAPTER 1 INTRODUCTION........................................................................................................1 2 REVIEW OF LITERATURE.......................................................................................4 Startle Eyeblink Reflex.................................................................................................4 Relationship to Other Psychophysiologic Measures....................................................4 Neuroanatomic Circuitry of Startle..............................................................................5 Aversion Enhancement of the Startle Eyeblink Response...........................................6 Dimensions of Emotion................................................................................................8 Startle Reactivity in Adults.........................................................................................10 Startle Reactivity in Children.....................................................................................12 Attention-Deficit/Hyperactivity Disorder...................................................................14 Prevalence and Definition...................................................................................14 Outcomes.............................................................................................................15 Neurobiologic Correlates of ADHD....................................................................16 Emotional Dysfunction in ADHD.......................................................................17 Conduct Disorder........................................................................................................22 Prevalence and Definition...................................................................................22 Development and Outcomes of Conduct Disorder..............................................23 Neurobiologic Correlates of CD..........................................................................24 Physiologic Correlates of Conduct Disorder.......................................................24 ADHD with CD comorbidity......................................................................................25 Prevalence and Risk Factors................................................................................25 Psychophysiologic Correlates of ADHD/CD......................................................26 Psychopathy................................................................................................................27 Oppositional Defiant Disorder (ODD) Comorbidity with ADHD.............................28 v

PAGE 6

3 SUMMARY, AIMS, AND HYPOTHESES..............................................................30 Summary and Aims....................................................................................................30 Hypotheses/Predictions...............................................................................................31 ADHD Compared to Healthy Controls...............................................................31 Psychopathy, CD Symptoms, ODD Symptoms, and Impulsivity Within ADHD..............................................................................................................34 4 RESEARCH DESIGN AND METHODS..................................................................36 Diagnostic, Screening, and Descriptive Measures.....................................................36 Participants.................................................................................................................38 Emotion Psychophysiology Task Apparatus and Procedures....................................45 Physiologic Recordings..............................................................................................49 Data Reduction...........................................................................................................49 Study Design and Analyses........................................................................................50 Series One: ADHD Versus Controls...................................................................50 Series Two: Variables within ADHD..................................................................53 Series Three: ADHD, ADHD + ODD, ADHD + CD.........................................55 5 SERIES ONE RESULTS: ADHD VERSUS CONTROLS.......................................56 Series One Hypotheses...............................................................................................56 Startle Eyeblink Magnitude........................................................................................57 Picture Category..................................................................................................57 Unpleasant Picture Subcategory..........................................................................58 Pleasant Picture Subcategory..............................................................................61 Skin Conductance Response.......................................................................................62 Picture Category..................................................................................................62 Unpleasant Picture Subcategory..........................................................................64 Pleasant Picture Subcategory..............................................................................65 Self-Report Ratings....................................................................................................65 Picture Category..................................................................................................65 Unpleasant Picture subcategory..........................................................................69 Pleasant Picture subcategory...............................................................................71 Raw Startle Eyeblink Response..................................................................................72 Startle Eyeblink Onset Latency..................................................................................74 Series One Results Summary Table...........................................................................74 6 SERIES TWO RESULTS: VARIABLES WITHIN ADHD......................................76 Series Two Hypotheses...............................................................................................76 Conduct Disorder Symptoms......................................................................................76 Startle Eyeblink Magnitude.................................................................................76 Skin Conductance Response................................................................................79 vi

PAGE 7

Self-Report Ratings.............................................................................................80 Raw Startle Eyeblink Response...........................................................................82 Oppositional Defiant Disorder Symptoms..................................................................82 Startle Eyeblink Magnitude.................................................................................82 Skin Conductance Response................................................................................84 Self-Report Ratings.............................................................................................87 Raw Startle Eyeblink Response...........................................................................89 Psychopathy Symptoms..............................................................................................89 Startle Eyeblink Magnitude.................................................................................89 Skin Conductance Response................................................................................91 Self-Report Ratings.............................................................................................92 Raw Startle Eyeblink Response...........................................................................95 Impulsivity..................................................................................................................95 Startle Eyeblink Magnitude.................................................................................95 Skin Conductance Response................................................................................97 Self-Report Ratings.............................................................................................99 Raw Startle Eyeblink Response.........................................................................102 Series Two Results Summary Table.........................................................................103 7 SERIES THREE RESULTS: ADHD, ADHD/ODD, ADHD/CD............................105 Startle Eyeblink Magnitude......................................................................................105 Picture Category................................................................................................105 Unpleasant Picture Subcategory........................................................................108 Pleasant Picture Subcategory............................................................................111 Skin Conductance Response.....................................................................................112 Picture Category................................................................................................112 Unpleasant Picture Subcategory........................................................................115 Pleasant Picture Subcategory............................................................................115 Self-Report Ratings..................................................................................................115 Picture Category................................................................................................115 Unpleasant Picture subcategory........................................................................118 Pleasant Picture subcategory.............................................................................118 Raw Startle Eyeblink Response................................................................................121 Startle Eyeblink Onset Latency................................................................................122 Series Three Results Summary Table.......................................................................122 8 DISCUSSION...........................................................................................................124 ADHD Versus Controls............................................................................................125 Picture Category................................................................................................125 Unpleasant Picture Subcategory........................................................................127 Pleasant Picture Subcategory............................................................................133 Other Theoretical Frameworks..........................................................................136 Variables Within ADHD..........................................................................................137 Startle Reactivity...............................................................................................138 Unpleasant picture category.......................................................................138 vii

PAGE 8

Direct threat pictures..................................................................................138 Victim pictures...........................................................................................139 Injury pictures............................................................................................139 Pleasant picture category............................................................................140 Neutral picture category.............................................................................140 Startle and comorbidity overview..............................................................142 Skin Conductance..............................................................................................143 Self-Report Ratings...........................................................................................144 Impulsivity.........................................................................................................145 Summary............................................................................................................147 Methodological Limitations..............................................................................149 Future Research Directions...............................................................................150 APPENDIX A DSM-IV DIAGNOSTIC CRITERIA FOR ADHD, CD, & ODD............................153 B SELF-ASSESSMENT MANIKIN...........................................................................156 C SELF-ASSESSMENT MANIKIN INSTRUCTIONS.............................................157 D EXAMPLE PICTURE TRIAL.................................................................................159 LIST OF REFERENCES.................................................................................................160 BIOGRAPHICAL SKETCH...........................................................................................172 viii

PAGE 9

LIST OF TABLES Table page 4-1. P-ChIPS symptoms in control and ADHD children...................................................44 4-2. Conners’ Parent Rating Scale T-scores......................................................................44 4-3. Connors’ Continuous Performance Test T-scores......................................................45 5-1. Series one hypotheses: ADHD versus healthy controls.............................................56 5-2. Dependent measure means by picture category.........................................................59 5-3. Dependent measure means across unpleasant picture subcategory............................61 5-4. Dependent measure means across pleasant picture subcategory................................63 5-5. Series one results summary: ADHD versus healthy controls.....................................75 6-1. Series two hypotheses: Dimensional analysis within ADHD....................................76 6-2. Series two results summary: Dimensional analysis within ADHD..........................103 7-1. Series Three Results Summary Table: ADHD, ADHD/ODD, ADHD/CD.............122 ix

PAGE 10

LIST OF FIGURES Figure page 2-1. Neuroanatomic circuitry of the startle reflex.Adapted from Lang et al., 1997............6 5-1. Mean startle eyeblink response for control and ADHD groups across picture category....................................................................................................................58 5-2. Mean startle eyeblink response for control and ADHD groups across unpleasant picture subcategory..................................................................................................59 5-3. Mean startle eyeblink response for ADHD and control groups across pleasant picture subcategory..................................................................................................62 5-4. Mean skin conductance response for control and ADHD groups across picture categories..................................................................................................................63 5-5. Mean skin conductance response for control and AHD groups across unpleasant picture subcategory..................................................................................................64 5-6. Mean skin conductance response for control and ADHD groups across pleasant picture subcategory..................................................................................................65 5-7. Distribution of self-report valence and arousal affective judgments for 36 picture stimuli by group.......................................................................................................66 5-8. Mean self-report valence rating for ADHD and control groups across picture category....................................................................................................................67 5-9. Mean self-report arousal rating for ADHD and control groups across picture category....................................................................................................................68 5-10. Mean self-report valence rating for ADHD and control groups across unpleasant picture subcategory..................................................................................................70 5-11. Mean self-report arousal rating for ADHD and control groups across unpleasant picture subcategory..................................................................................................70 5-12. Mean self-report valence rating for ADHD and control groups across pleasant picture subcategory..................................................................................................71 x

PAGE 11

5-13. Mean self-report arousal rating for ADHD and control groups across pleasant picture subcategory..................................................................................................72 5-14. Mean absolute startle response for control and ADHD groups across picture category....................................................................................................................73 6-1. Neutral picture startle response by number of Conduct Disorder symptoms in ADHD children........................................................................................................77 6-2. Injury picture startle response by number of Conduct Disorder symptoms in ADHD children........................................................................................................78 6-3. Self-report valence averaged across all picture stimuli by number of CD symptoms in ADHD.................................................................................................80 6-4. Unpleasant picture startle response by number of Oppositional Defiant Disorder symptoms in ADHD.................................................................................................83 6-5. Neutral picture startle response by number of Oppositional Defiant Disorder symptoms in ADHD.................................................................................................83 6-6. Skin conductance response collapsed across picture categories by number of ODD symptoms in ADHD.......................................................................................85 6-7. Unpleasant picture skin conductance response by number of ODD symptoms in ADHD......................................................................................................................86 6-8. Pleasant picture skin conductance response by number of ODD symptoms in ADHD......................................................................................................................86 6-9. Self-report unpleasant picture arousal by number of ODD symptoms in ADHD children.....................................................................................................................88 6-10. Neutral picture startle response by total psychopathy score in ADHD....................90 6-11. Injury picture startle response by total psychopathy score in ADHD......................91 6-12. Self-report valence averaged across all picture stimuli by total psychopathy score in ADHD.........................................................................................................93 6-13. Self-report attraction picture arousal by psychopathy score in ADHD children......95 6-14. Skin conductance response to attraction pictures by Beta score in ADHD children.....................................................................................................................98 7-1. Mean startle eyeblink response for ADHD, ADHD+ODD, and ADHD+CD groups across picture category...............................................................................106 xi

PAGE 12

7-2. Mean startle eyeblink response for ADHD, ADHD+ODD, and ADHD+CD groups across unpleasant picture subcategory.......................................................109 7-4. Mean self-report valence rating for ADHD, ADHD/ODD, and ADHD/CD groups across picture category...........................................................................................116 7-5. Mean self-report arousal rating for ADHD, ADHD/ODD, and ADHD/CD groups across picture category...........................................................................................117 7-6. Mean self-report arousal rating for ADHD, ADHD/ODD, and ADHD/CD groups across pleasant picture subcategory.......................................................................119 7-7. Mean absolute startle response for ADHD, ADHD+ODD, and ADHD+CD across picture category......................................................................................................121 xii

PAGE 13

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EMOTIONAL REACTIVITY TO PICTURE STIMULI IN CHILDREN WITH AND WITHOUT ATTENTION-DEFICIT/HYPERACTIVITY DISORDER By Gregg Jonathan Selke August 2005 Chair: Eileen Fennell Cochair: Dawn Bowers Major Department: Clinical and Health Psychology Accumulating evidence suggests emotional dysregulation in Attention-Deficit/Hyperactivity Disorder (ADHD). The co-occurrence of early psychopathic and/or conduct-related symptoms may place ADHD children at greater risk for poor outcomes. Affective modulation of the startle reflex is a valuable tool for measuring emotional reactivity. Normal adults show smaller startle reflex responses to pleasant stimuli and larger startle responses to unpleasant stimuli. Enhanced startle to unpleasant stimuli is also influenced by content type (i.e., direct threat versus sad). The study aimed to primarily compare the emotional reactivity of ADHD to healthy control children. A secondary aim was to investigate whether the severity of conduct-related, psychopathy, and impulsivity symptoms impact emotional reactivity within ADHD. Thirty-four ADHD and 23 healthy control children were presented with a 100 dB acoustic startle probe while viewing pleasant, unpleasant, and neutral picture stimuli. Unpleasant pictures consisted of direct threat, victimization, and injury pictures. Dependent measures xiii

PAGE 14

included startle eyeblink magnitude, skin conductance, and subjective valence and arousal ratings. Results revealed that ADHD and control children show different patterns of startle reactivity across different types of unpleasant picture stimuli. Both ADHD and healthy control children show highly reliable aversion enhanced startle modulation to direct threat and injury pictures. However, only controls show potentiated startle to victim pictures, and in contrast, ADHD children display inhibited startle reactivity to victim pictures. This occurs despite ADHD children showing appropriate physiological arousal while viewing victim pictures, and being successful at cognitively appraising their emotional experience associated with viewing victim pictures. Moreover, comorbid Oppositional Defiant Disorder symptomatology appears to impact physiological arousal (i.e., lower skin conductance) whereas Conduct Disorder (CD) and psychopathy symptomatology appear to impact affective startle modulation (i.e., reduced startle response to injury and victim pictures). These findings provide support for the existence of emotional dysregulation in children with ADHD, and suggest that affective modulation of the startle may represent an early biologic marker for increased risk of developing psychopathy in adulthood. xiv

PAGE 15

CHAPTER 1 INTRODUCTION Attention-Deficit Hyperactivity Disorder (ADHD) is one of the most commonly diagnosed psychiatric disorders of childhood, with an estimated prevalence rate of 4-12% in school-age children (Brown et al., 2001; Kaplan & Sadock, 1998). Children diagnosed with ADHD are at increased risk for the development of a wide range of chronic behavioral problems during adolescence and adulthood. The co-occurrence of early psychopathic and/or conduct-related symptoms may place ADHD children at even greater risk for poor outcomes. Psychophysiological techniques that integrate physiology with behavior may be useful for investigating emotional and physiological abnormalities in ADHD. Over the last decade, studies with normal adults have consistently shown that the startle eyeblink response is larger during negative emotional states (e.g., fear, anxiety) compared to pleasant emotional states (Bradley, 2000; Lang et al., 1990; Vrana et al., 1988). This phenomenon, known as affective modulation of the startle reflex (AMSR), has become a valuable tool for measuring emotional and physiological reactivity in normal, psychiatric, and neurologically impaired individuals. Affective modulation of the startle reflex has significantly increased our understanding of the basic neural mechanisms of emotional processing and behavior. Affective modulation of the startle reflex has also demonstrated its usefulness by extending our knowledge of underlying processes that mediate psychopathology. 1

PAGE 16

2 Notably, only one study has examined affective modulation of startle in children with disruptive behavior disorders, and relatively few such studies have been published investigating AMSR in normal children. Specifically, “adult-like” patterns of AMSR have been found in infants (Balaban, 1995) and adolescents (13-18 years old: Grillon et al., 1999), but only inconsistently in male children between the ages of 7 to 10 years (McManis et al., 2001; Van Goozen et al., 2004). In addition, Selke et al. (2005) found “adult-like” patterns of AMSR in females ages 7 to 14 and males ages 11 to 14 years. However, an “adult-like” pattern was only found in males ages 7-10 years for certain types of stimuli (i.e., direct threat but not victim or injury pictures). More recently, Van Goozen et al. (2004) found an “adult-like” pattern of startle reactivity (i.e., unpleasant > pleasant) in both 33 healthy control children and 21 disruptive behavior disorder children (15 with Oppositional Defiant Disorder, 6 with Conduct Disorder, 16 of 21 with comorbid ADHD). The study also found that the disruptive behavior disorder children showed lower absolute startle values, and found an association between increased delinquency and decreased startle reactivity to unpleasant pictures. Of particular relevance to the current project, adult criminal psychopaths fail to show a normal pattern of reactivity, and instead show a reduced eyeblink startle response to unpleasant stimuli (Herpertz, Werth, et al., 2001; Levenston et al., 2000; Patrick et al., 1993). This occurs even though the psychopath verbally appraises the stimuli as aversive. This finding suggests a defect in translating the results of cognitive appraisal into somato-motor changes that are associated with an unpleasant emotional stimulus. This leads to the question of whether ADHD children with and without comorbid conduct problems, who are at increased risk for developing adult psychopathy and/or antisocial

PAGE 17

3 personality disorder (Burke et al., 2000; Loeber et al., 2000; Lynam, 1996), might also fail to show decreased startle reactivity to unpleasant stimuli. If so, those children with aberrant reactivity may be at particular risk for more chronic antisocial behaviors in the future. This was the first study to specifically compare the emotional reactivity of normal children to those diagnosed with ADHD from ages 7 to 14 years-old using the startle reflex as the primary measure and skin conductance (SC) and subjective ratings as secondary measures. The secondary measures were collected to aid in the interpretation of startle reactivity findings. The study also examined whether early psychopathic and conduct-related symptoms impact emotional reactivity within ADHD. The project also explored whether measures of impulsivity were associated with emotional reactivity within ADHD.

PAGE 18

CHAPTER 2 REVIEW OF LITERATURE Startle Eyeblink Reflex Of central importance to this study is the use of a relatively new “emotion” measure involving the startle reflex. A reflex is an involuntary response to a specific sensory stimulus, generally thought to be adaptive and protective. Startle responses are triggered by an abrupt stimulus (e.g., burst of white noise) and can involve flexor movement in the arms and shoulders, as well as an eyeblink. Typically, the eyeblink has been used to measure the startle reflex response since it is the most reliable, easily recorded, and quickest component of the startle response (Bradley, 2000; Lang et al., 1990). The startle eyeblink response is also valuable in experimental settings because it is an involuntary response, and therefore does not rely on an individual’s cooperation, training, motivation, or attention (Bradley, 2000). Overall, the startle eyeblink response functions as a direct link to the nervous system, allowing for the objective study of physiological, sensory, and attentional processes. Relationship to Other Psychophysiologic Measures Relative to other psychophysiologic measures (skin conductance, heart rate), the startle eyeblink response offers some unique advantages as a measure of emotional reactivity. In adults, startle reactivity is relatively stable and reliably observed (Bradley et al., 1999). In contrast, approximately 15-20% of normal adults are electrodermal “nonresponders,” making this measure less than optimal if one is attempting to use electrodermal reactivity as a classification variable. Further, gross motor activity can 4

PAGE 19

5 interfere with heart rate (HR) and skin conductance (SC) recordings, yet it has minimal effects on startle eyeblink measures (Bradley, 2000). Children may have difficulty remaining still during prolonged testing procedures. Therefore, it is less than optimal to rely on HR or SC as the primary measure of physiological reactivity in children. Additionally, certain psychophysiologic measures seem better suited for indexing arousal (i.e., high to low), whereas others are more sensitive to valence (i.e., degree of pleasantness). Skin conductance correlates with self-reported arousal in normal adults, regardless of the valence of the stimulus (Bradley et al., 1990; Greenwald et al., 1989; Lang et al., 1993). Heart rate is more sensitive to attentional demands and varies depending on whether the emotional reaction is elicited via imagery (deceleratory with unpleasant imagery; Lang et al., 1990, Lang et al., 1993) or via pictures (acceleratory with unpleasant pictures; Lang, 1990). Thus, HR does not correlate well with rated arousal and SC does not correlate well with rated valence (Bradley, 2000). In contrast, patterns of startle reactivity correlate highly with the valence (i.e., pleasantness) of the stimulus independent of foreground modality, especially at higher levels of arousal. Skin conductance will be used as a secondary measure in the proposed study because it may aid in interpreting the startle modulation data and affective picture ratings. Neuroanatomic Circuitry of Startle The neuroanatomic circuitry underlying the startle reflex has been mapped out using ablation and stimulation techniques in non-human animals (Davis, 1998; Koch & Schnitzler, 1997; see Figure 2-1 below adapted from Lang et al., 1997). The startle reflex circuitry per se is entirely subcortical, involving afferent and efferent pathways in the brainstem and spinal cord. Even anencephalic infants show normal startle responses.

PAGE 20

6 Although the startle circuitry is subcortical, the amygdala appears to directly modulate startle circuitry via projections from the central nucleus of the amygdala to the brainstem. Ablations of the amygdala in rodents do not affect basic startle responses, but do abolish fear potentiation of the startle response. Even studies of humans with temporal Figure 2-1. Neuroanatomic circuitry of the startle reflex.Adapted from Lang et al., 1997. Potentiated Startle Potentiated Startle Stimulus Input Sensory Cortex Sensory Thalamus Lateral Region Hypothalamus Lateral Central N ucleus Nucleus Amygdala Autonomic N S HR, BP Dorsal Central Gray (Fight/Flight) Ventral Central Gray (Freezing) N ucleus Reticularis Pontis Caudalis Stimulus Sensory Cortex Sensory Thalamus Lateral Region Hypothalamus Lateral Central N ucleus Nucleus Amygdala Autonomic N S HR, BP Dorsal Central Gray (Fight/Flight) Ventral Central Gray (Freezing) N ucleus Reticularis Pontis Caudalis Input lobe ablations involving the amygdala have found that startle eyeblink responses are no longer primed when subjects view unpleasant pictures or think about the meaning of sentences conveying fear (Bowers et al., in press). Taken together, such findings have been interpreted as evidence that increases in the startle response during unpleasant/aversive emotional states may reflect the amygdala’s dual role in (1) danger/fear detection and in (2) modulating the subcortical startle circuitry (Lang, 1995; Lang et al., 1997). Aversion Enhancement of the Startle Eyeblink Response Extensive research over the past 10 years suggests that the startle eyeblink may serve as a motivational or valence marker in humans. Numerous studies have shown that the startle reflex is primed under conditions of fear, anxiety and aversive states (Bradley, 2000; Hamm et al., 1997; Lang et al., 1990; Vrana et al., 1988). Even normal individuals

PAGE 21

7 show heightened startle reactions during transient episodes of fear or aversion. For example, imagine watching a scary movie like Jaws . If the person behind you unexpectedly taps your shoulder during a tense scene, very likely a startle response will be elicited. The startle reaction elicited during Jaws will be much larger than a startle reaction that is elicited while you watch home movies of your sister’s wedding. In more detail, affective modulation of the startle reflex (AMSR) involves two different types of stimuli that occur independently in time. The first stimulus influences the affective response (e.g., scary scene) and is called the “foreground stimulus.” The second stimulus triggers the startle eyeblink response (e.g., tap on shoulder) and is called the “startle probe.” The foreground stimulus is presented to an individual over a relatively extended period of time (e.g., 6 seconds). Although it can be emotional in nature and affect heart rate, sweating, or facial expression, it is never a strong enough stimulus to elicit a startle eyeblink response. Rather, it is evaluated or judged over a period of time. Next, the startle probe is brief and intense, and is followed 30-40 milliseconds later by an involuntary eyeblink response. A startle probe is presented to an individual while he/she is seeing, hearing, smelling, or thinking about the foreground stimulus. In the laboratory, startle probes are typically used at moderate intensities (e.g., 95-100 dB burst of white noise) to reliably produce an eyeblink response without overt movements in the torso, arms, or shoulders. The unifying theme across affective startle modulation studies is that normal adults show smaller startle eyeblinks when engaged in an emotionally pleasant task and larger startle eyeblinks when engaged in an emotionally unpleasant task. This enhancement of startle has been viewed as a “motivational priming” effect (Lang, 1995), such that an

PAGE 22

8 essentially protective withdrawal reflex is potentiated or primed during unpleasant emotional/motivational states. In the context of pleasant emotional/motivational states, the defensive reflex is inhibited. Facilitation or inhibition of the defensive reflex is most likely to occur when the foreground stimulus is highly arousing (i.e., when the motivational system is most engaged; Cuthbert et al., 1996). Consequently, a startle probe produces an eyeblink response that is largest for highly arousing unpleasant pictures, smallest for highly arousing pleasant pictures, and intermediate for neutral pictures. Before reviewing the literature on AMSR in adult and child populations in greater detail, it is important to review how emotions are defined and operationalized for the purposes of the current dissertation study. Dimensions of Emotion Contemporary accounts view emotional behavior as multidimensional and consisting of (1) verbal reports and cognitions about one’s emotional experience; (2) overt behaviors ranging from specific social displays to behavioral dispositions like avoidance, withdrawal, attack, and (3) physiologic arousal and activation. It is often assumed that these three components should closely correspond, particularly if they are measuring the same underlying emotional state. Frequently, however, they are divergent as with the case of right-hemisphere lobectomy patients (Bowers, in press) or criminal psychopaths (Herpertz, Werth, et al., 2001; Levenston et al., 2000; Patrick et al., 1993). The presence of such discordances reinforces the view that these verbal, behavioral, and physiologic components are discrete response systems that reflect the output of different neural processors within the brain. Each of these behavioral components of emotion (verbal report, physiology, behavior) can vary along the dimensions of valence (i.e., degree of pleasantness) and

PAGE 23

9 arousal (i.e., highly arousing to not arousing). Valence and arousal have been shown to be fundamental in organizing emotional language. Osgood (1969) used factor analysis to show that 2 factors, hedonic valence and arousal, consistently accounted for the most variability in word judgments. Motivated by these and other findings, Lang (1995) has proposed a dimensional theory of affect. In this theory, two orthogonal dimensions define the emotional responses associated with an event or stimuli. The first dimension, affective valence, corresponds to the appetitive (i.e., approach) versus aversive (i.e., avoidance) disposition associated with the emotional response. The second dimension, arousal, corresponds to the degree of activation associated with the emotional response. The valence and arousal dimensions of affect have been operationalized by the Self-Assessment Manikin or SAM ratings (Bradley & Lang, 1994). The SAM is a pictorial assessment technique that uses cartoon figures to measure subjective emotional reactivity to stimuli. Using the SAM, affective stimuli are rated along the valence and arousal dimensions, each typically on a 1-9 ordinal scale. Graphically, the valence dimension varies from a smiling cartoon figure to a frowning figure. The arousal dimension varies from a highly excited or nervous looking cartoon figure to a bored and sleepy looking figure. The current study used a modified SAM, in that the valence and arousal dimensions were rated along a 1-7 ordinal scale (See Appendix A). When pictures from the International Affective Picture System (IAPS: over 800 color photographs derived from a broad range of semantic categories) are plotted using the SAM, picture judgments span the entire valence range and vary extensively across arousal (Lang et al., 1999). The fundamental shape of this distribution pattern has been found to be identical for pictures, words, brief films, instrumental music, and sounds

PAGE 24

10 (Bradley, 2000). These findings suggest that this organization is relevant to emotion as a whole, and support the use of the Self-Assessment Manikin as a valid measure of one’s verbal report about one’s emotional experience. Startle Reactivity in Adults In adults, affective modulation of the startle reflex has been reliably demonstrated in the laboratory using a variety of startle probe modalities, including acoustic (i.e., white noise burst; Vrana et al., 1988), visual (i.e., flash of light; Bradley et al., 1990), and tactile (Hawk & Cook, 1997). A variety of foreground stimuli have also been shown to induce emotional states that modulate startle reactivity, such as viewing affective pictures (Vrana et al., 1988), viewing affective films (Jansen & Frijda, 1994), reading emotional sentences (Spence & Lang, 1990), listening to emotional sounds (Bradley & Lang, 2000), perceiving hedonic odors (Miltner, 1994), using imagery (Vrana & Lang, 1990), exposing participants to darkness (Grillon et al., 1997), and anticipating shock (Grillon et al., 1994). Affective modulation of the startle reflex has also been used to study emotions in various adult disorders, including phobias (Hamm et al., 1997), panic disorder (Grillon et al., 1994), schizophrenia (Schlenker et al., 1995), psychopathy (Herpertz et al., 2001; Levenston et al., 2000; Patrick et al., 1993) and neurologic impairments (Morris et al., 1991). Many of these studies have shown that AMSR has the power to discriminate among different groups of adults (e.g., phobics from controls, panic disorder patients from controls, psychopaths from non-psychopaths). As stated in the introduction, of particular relevance to the current project, adult criminal psychopaths fail to show a normal pattern of reactivity, and instead show a reduced eyeblink startle response to unpleasant stimuli (Herpertz et al., 2001; Levenston

PAGE 25

11 et al., 2000; Patrick et al., 1993). This occurs even though the psychopath verbally appraises the stimuli as aversive. Moreover, both Patrick et al. (1993) and Levenston et al. (2000) found that across unpleasant, neutral, and pleasant picture categories, psychopaths showed a quadratric response pattern with neutral pictures eliciting greater startle reactivity than pleasant and unpleasant pictures. These findings suggest a defect in translating the results of cognitive appraisal into somato-motor changes that are associated with an unpleasant emotional stimulus. Moreover, with respect to study design, Levenston’s (2000) study is directly applicable to the current dissertation study. Specifically, Levenston not only compared unpleasant to pleasant to neutral categories pictures, but also examined startle reactivity across different types or subcategories of unpleasant (direct threat, mutilation/injury, assault/victimization) and pleasant (erotic, thrill) pictures. With respect to overall picture category, nonpsychopaths showed a linear pattern of startle reactivity (i.e., unpleasant > pleasant), whereas psychopaths showed both a linear (upleasant > pleasant) and quadratric pattern (neutral pictures > pleasant and unpleasant). With respect to unpleasant picture subcategory, Levenston et al. (2000) found that nonpsychopaths demonstrated increased or potentiated startle magnitude across all three unpleasant subcategories relative to pleasant pictures, with mutilation/injury and assault/victim pictures showing similar levels of potentiation, and direct threat showing even greater potentiation than mutilation and assault pictures. On the other hand, psychopaths demonstrated an aberrant pattern of response. Specifically, although reduced relative to nonpsychopaths, psychopaths demonstrated potentiated startle to only

PAGE 26

12 direct threat pictures, but inhibited startle to pictures depicting mutilation/injury or assault/victimization relative to pleasant pictures. With respect to pleasant picture subcategories (erotic, thrill), Levenston et al. (2000) found that nonpsychopaths demonstrated decreased or inhibited startle for erotic pictures but increased startle for thrill pictures, whereas psychopaths demonstrated decreased startle for both erotic and thrill pictures. Moreover, with respect to affective self-report ratings, psychopaths appraised the aversive pictures as less unpleasant and the pleasant pictures as more pleasant compared to nonpsychopaths. There were no differences between psychopaths and nonpsychopaths on rated arousal. Results of Levenston’s (2000) study demonstrated the importance of not only investigating startle reactivity across overall unpleasant, neutral, and pleasant categories, but also across different types of unpleasant and pleasant pictures (e.g., direct versus vicarious threat, thrill versus attaction/erotica). Results of all three psychopathy studies (Herpertz et al., 2001; Levenston et al., 2000; Patrick et al., 1993) lead to the question of whether ADHD children with and without comorbid conduct problems, who are at increased risk for developing adult psychopathy and/or antisocial personality disorder (Burke et al., 2000; Loeber et al., 2000; Lynam, 1996) might also fail to show decreased startle reactivity to unpleasant stimuli. If so, those children with aberrant reactivity may be at particular risk for more chronic antisocial behaviors in the future. Startle Reactivity in Children Over the years, many studies have used the startle eyeblink response in children to examine various physiological phenomenon unrelated to emotional reactivity. The vast majority have used the startle reflex to study attentional and sensory processes in normal

PAGE 27

13 children (Ornitz et al., 1991; Orntiz et al., 1996), those with ADHD (Goldstein & Blumenthal, 1995; Herpertz et al., 2001; Ornitz et al., 1997; Ornitz et al., 1992), and those at risk for psychopathology (children at risk for anxiety disorders and/or alcoholism: Grillon et al., 1997; posttraumatic stress disorder: Ornitz & Pynoos, 1989). While affective modulation of the startle reflex has proved to be valuable for learning about the processing and organization of emotions in the nervous system in normal and clinical populations, most of this research has primarily focused on adult populations. Still, several studies have been published examining emotional modulation of the startle reflex in children. Balaban (1995) demonstrated larger startle eyeblink responses when five-month-old infants viewed angry facial expressions compared to happy expressions. Grillon et al. (1999) demonstrated that the threat of an air blast to the larynx and during darkness elicited larger eyeblink responses in adolescents ages 13 to 17. McManis et al. (2001) found an adult-like pattern in females age 7-10 (trend level), but not males, using picture stimuli and 95dB acoustic startle probes. However, McManis did not select gender specific picture stimuli. McManis also standardized the picture stimuli for a broad age range, from seven years old to college-age adults. This may have resulted in some of the picture stimuli not being developmentally appropriate or relevant for young children. Cook (1995) failed to find the adult-like pattern using imagery as foreground stimuli in school-age children. Unfortunately, imagery studies such as Cook’s may not ensure successful imagery or the desired emotional state in children. More recently, Selke et al. (2005) completed a normative study and found “adult-like” patterns of AMSR in healthy “normal” females ages 7 to 14 and males ages 11 to 14

PAGE 28

14 years. However, an “adult-like” pattern was only found in males ages 7-10 years for certain types of stimuli (i.e., direct threat but not victimization or injury pictures). Based on these findings, it appears to be critical to account for age-based effects as well as the specific unpleasant picture subcategories when examining AMSR in children. Of particular relevance to the current dissertation study, Van Goozen et al. (2004) recently published a study examining affective modulation of the startle reflex using picture stimuli in 33 normal control children (19 girls, 14 boys) and 21 disruptive behavior disorder (DBD) children (19 boys, 2 girls) ages 7 to 12 years old. The DBD group consisted of 15 children with Oppositional Defiant Disorder (ODD) and 6 children with Conduct Disorder (CD), of which 16 also met criteria for comorbid ADHD. Ten of the DBD children were on stimulant medication. The study found an adult-like linear pattern of startle reactivity (unpleasant > pleasant) using raw startle responses for both controls and DBD children. However, DBD children showed lower absolute startle values (i.e., not converted to T-scores) across all picture categories, which was interpreted by the authors as evidence of fearlessness. The study also found that greater delinquency (Achenbach Child Behavior Checklist) was associated with decreased startle reactivity to unpleasant pictures. Moreover, there was no effect of stimulant medication on startle reactivity. The study did not report skin conductance response or self-report valence and arousal ratings. Attention-Deficit/Hyperactivity Disorder Prevalence and Definition Attention-Deficit/Hyperactivity Disorder (ADHD) is one of the most commonly diagnosed psychiatric disorders of childhood, with prevalence rates ranging from 4% to 12% in epidemiologic samples of 6 to 12 year olds (Brown et al., 2001). According to

PAGE 29

15 the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders Fourth Edition (1994), there are three subtypes of ADHD: ADHD, Predominantly Hyperactive-Impulsive; ADHD, Predominantly Inattentive Type; ADHD, Combined Type. Symptoms fall under 3 domains: (1) inattention (e.g., difficulty sustaining attention, easily distracted, forgetful), (2) hyperactivity (e.g., fidgets, “on the go,” talks excessively), and (3) impulsivity (e.g., difficulty waiting turn, interrupts others). In addition, symptoms of hyperactivity-impulsivity and/or inattention must “have persisted for at least 6 months to a degree that is maladaptive and inconsistent with developmental level,” with some impairment present before age 7, and “clinically significant impairment in social, academic, or occupational functioning” in at least 2 settings. Please see Appendix A for the complete DSM-IV diagnostic criteria for ADHD. Outcomes Children with ADHD are at increased risk for low academic achievement, school suspensions and expulsions, depression, driving accidents, early substance use and abuse, peer and family relationship problems, poor social and occupational outcomes in adulthood, and significant antisocial and criminal activity in adolescents and early adulthood (Barkley, 1997, 1998; Satterfield et al., 1982; Satterfield et al., 1987; Weiss et al., 1971). It is estimated that up to 70% of children with ADHD will continue to meet full diagnostic criteria in adolescence, and up to 66% will continue to meet full diagnostic criteria in adulthood (Barkley, 1998). In addition, it is estimated that between 18-28% of children with ADHD meet criteria for Antisocial Personality Disorder (APD) in adulthood (Barkley, 1998; Kaplan & Sadock, 1998). Impulsivity may play an important role in functional outcomes of ADHD. Specifically, impulse control deficits may contribute to the development of delinquent

PAGE 30

16 behavior by interfering with a child’s ability to (1) actively inhibit their behaviors, and (2) consider the consequences of their behaviors. Impulsivity at kindergarten (Tremblay et al., 1994) and at 4th grade (White et al., 1994) has been shown to be predictive of delinquent behavior at ages 10-13. In adolescent aged 12-18 years, impulsivity was also closely correlated with antisocial behavior (Luengo et al. 1994). Neurobiologic Correlates of ADHD Although the neural basis of ADHD is unclear, research in neuropsychology, neuroimaging, and molecular genetics seems to imply a multifactorial etiology. There is robust support for a hereditary basis to ADHD. The concordance rate ranges from 51-80% for monozygotic twins and 29-33% for dizygotic twins (Kaplan & Sadock, 1998). If one parent has ADHD, there is 57% chance an offspring will develop ADHD (Biederman et al., 1995). Moreover, neuroimaging studies have suggested smaller brain volumes in the globus pallidus (Castellanos et al., 1996; Filipek et al., 1997; Hynd et al, 1993), frontal/prefrontal cortices (Castellanos et al., 1996; Filipek et al., 1997; Mostofsky et al., 2002), cerebellar vermis (Berquin et al., 1998; Castellanos, 2001; Mostofsky et al., 1998) and the corpus callosum (Baumgardner et al., 1996; Castellanos, 2001; Giedd et al., 1994; Hynd et al., 1991). In addition, neuropsychological studies have pointed to deficits in attention (Barkley, 1998; Barkley et al., 1992; Brodeur & Pond, 2001; Carter et al., 1995; Chabildas et al. , 2001; Corkum & Siegal, 1993, Heaton et al., 2001; Seidman et al., 1997), inhibition (Barkley et al., 1992; Carter, 1995; Oosterlann et al., 1998; Sergeant et al., 2002), planning and organization (Sergeant et al., 2002), time reproduction (Kerns et al., 2001), and temporal discounting (Barkley et al., 2001). Although clearly controversial, some evidence supports dysfunction of frontal-striatal dopaminergic and noradrenergic systems (Castellanos, 1997, 2001; Swanson et al., 1998).

PAGE 31

17 Emotional Dysfunction in ADHD Accumulating evidence suggests a significant emotional component in ADHD. Specifically, there appears to be dysfunction in affect recognition, motivation, affect avoidance, and response to rewards in ADHD (Ornitz et al., 1997). Still, very little research has directly investigated expressive and receptive processing of emotions (e.g., affect, empathy) in ADHD. For instance, after an extensive literature search, only one study was found to directly examine empathy in ADHD. Specifically, Braaten and Rosen (2000) presented vignettes to 24 ADHD and 19 control boys ages 6 to 12. The vignettes were intended to elicit positive or negative emotions, as something positive or negative happened to a child in each story. Results indicated that boys with ADHD expressed less empathy when something negative happened the child in the story. Relating these findings to the current dissertation study, if children with ADHD have reduced ability to verbally express and physiologically experience empathy it would be predicted that they would rate victim pictures as less arousing and more pleasant than direct threat pictures, would rate victim pictures as less arousing and more pleasant than healthy controls, and would subsequently show decreased startle response to victim pictures relative to direct threat pictures and relative to healthy controls. In another study, Shapiro et al. (1993) found only minimal differences between children with ADHD and those without ADHD ages 6 to 11 years-old on battery designed to measure evaluation of emotional cues. Specifically, perception and recognition of affective stimuli (facial expression, language, and speech prosody) were generally equivalent between groups across most tasks. However, ADHD children performed worse when greater attentional and working memory demands were required for affective processing, such as with a task requiring matching of orally presented prosody to visually

PAGE 32

18 presented affective facial expressions. Moreover, Rapport et al. (2002) found that adults with ADHD adults tend to experience their own emotions with stronger intensity and are less accurate at identifying other’s emotional states than adults without ADHD. Based on these findings, it would be predicted that in the current dissertation study, ADHD children would show less reliable differences between picture categories (unpleasant, neutral, pleasant) on self-report valence (i.e., pleasantness), but show higher arousal ratings and skin conductance responses to unpleasant and pleasant pictures, relative to controls. Several scientists have proposed that individuals with ADHD display dysfunction in their reward system. For instance, Haenlein and Caul (1987) propose an “Elevated Reward Threshold Hypothesis,” whereby the magnitude of the experienced reward in response to positive reinforcement is reduced in ADHD compared to non-ADHD individuals. For instance, continuous reinforcement schedules have been shown to enhance performance on reaction time and problem-solving tasks over partial reinforcement schedules to a larger degree in ADHD children compared to control children (Douglas & Parry, 1983; Parry & Douglas, 1983). In a more recent study, children with ADHD performed worse (i.e., more errors, longer reaction times) than controls on a stop-signal task under low incentive conditions, but displayed equivalent performance to controls under high incentive conditions (Slusarek et al., 2001). In addition, Sonuga-Barke’s (1992) Delay Aversion Hypothesis proposes that ADHD symptoms (i.e., hyperactivity, inattention, and impulsivity) are expressions of delay aversion. Specifically, on tasks where children can make a choice between small immediate rewards and a large delayed reward, ADHD children choose the small immediate reward more often than non-ADHD children (Solanto et al., 2001). When it is

PAGE 33

19 not possible to reduce a delay period, children with ADHD maximize their attention to nontemporal stimuli (i.e., spend time off-task, elicit stimulation from their environment, or increase their motor activity level) in an attempt to reduce the perceived length of time or subjective experience of the delay. Considering theories of reward dysfunction in the context of the current dissertation study, pleasant picture stimuli could be conceptualized as positively rewarding. Therefore, it would be predicted that pleasant picture stimuli would activate the appetive or approach motivational system to a lesser degree (i.e., increased threshold) in ADHD children compared to non-ADHD children, resulting in lower valence ratings (i.e., less pleasant) and decreased startle inhibition in response to pleasant picture stimuli in ADHD children. Next, other scientists have proposed that ADHD behaviors are manifestations of a physiologically under-aroused nervous system. For instance, Zentall & Zentall (1976, 1983) have put forward the optimal stimulation theory, whereby hyperactive-impulsive behaviors serve to maintain optimal arousal levels in an under-aroused system. Under this model, children with ADHD are also more likely to seek out stimulation in the service of maintaining optimal arousal levels. Limited support of this model come from psychophysiological and behavioral studies. For instance, Zentall & Zentall (1976) found that decreases in activity level were associated with a highly stimulating environmental condition compared to stimulus-poor condition. Zentall et al. (1985) found that performance on a copying task improved when stimuli was presented in a more highly stimulating format for ADHD adolescents but not for controls. In addition, several studies have suggested that children with ADHD have lower autonomic arousal.

PAGE 34

20 Specifically, children with ADHD show reduced phasic electrodermal reactivity to both signal and nonsignal stimuli (Zahn et al., 1975; Shibagaki et al., 1993). However, differences in resting levels of skin conductance (SC) and heart rate (HR) have not been consistently found (Satterfield & Dawson, 1971; Zahn et al., 1975; Zahn & Kruesi, 1993). Unfortunately, relatively few studies examining autonomic arousal in ADHD have accounted for comorbid conduct symptoms. Please see below for a review of the psychophysiologic correlates of ADHD with comorbid CD. In the context of the current dissertation study, under-arousal theories of ADHD would predict that children with ADHD would show lower absolute startle values, skin conductance magnitude, and self-report arousal ratings in response to affective picture stimuli. In addition, with respect to pleasant picture stimuli and stimulation seeking, ADHD would be predicted to show higher arousal and valence (more pleasant) ratings, stronger skin conductance response, and greater inhibition of startle to thrilling pictures, compared attraction and nurture pictures. More recently, researchers have implicated dysfunction in the behavioral inhibition system as the fundamental problem in ADHD (Barkley, 1997, 1998; Quay, 1997; Schachar et al., 1995). Quay’s (1997) model proposes that behavior is explained by the activity of two opposing brain systems, a behavioral inhibition system (BIS) and a behavioral activation system (BAS). Signals of reward or removal of punishment activate the BAS and result in approach behavior, whereas signals of punishment or loss of positive reinforcement activate the BIS and result in inhibition or extinction of behavior. Under Quay’s model, children with ADHD have an underactive BIS, resulting in diminished sensitivity to signals of punishment or negative stimuli, as well as

PAGE 35

21 hyperactive-impulsive symptomatology. In addition, children with conduct problems are proposed to have an overactive BAS that dominates their BIS. Under Barkley’s (1997, 1998) model, impaired behavioral inhibition (i.e., inhibiting a prepotent response, stopping an ongoing response, and inhibiting interference) leads to impairment of four “executive” functions: 1) working memory, 2) internalization of speech, 3) self-regulation of affect/motivation/arousal, and 4) reconstitution. These impaired executive functions lead to disturbances in self-control and self-directed behavior, and ultimately result in the behaviors seen in ADHD. With respect to emotional processing, deficits in inhibitory control lead children with ADHD to have poor emotional self-control (such as in provoking situations), poor social perspective taking and decreased empathy, and poor self-regulation of drive, motivation, and arousal in service of goal-directed activity (i.e., the “self-regulation of affect-motivation-arousal” executive function). As a consequence, the behaviors of ADHD children are controlled more by the immediate context and its consequences, rather than the behavior of others (Houghton, 1999). Whereas the behaviors of non-ADHD are more controlled by internally represented information (i.e., hindsight, forethought, time, plans, rules, self-motivating stimuli). In the context of the current dissertation study, theories of disinhibition would predict that atypical patterns of emotional reactivity (i.e., startle reactivity, skin conductance) would be directly associated with decreased behavioral inhibition (i.e., worse performance on continuous performance test commission errors or Beta, greater hyperactive-impulsive symptomatology).

PAGE 36

22 Next, deficits in attention are clearly associated with ADHD, as demonstrated across many neuropsychology studies (Barkley, 1998; Barkley et al., 1992; Brodeur & Pond, 2001; Carter et al., 1995; Chabildas et al., 2001; Corkum & Siegal, 1993, Heaton et al., 2001; Seidman et al., 1997). With respect to emotional processing, inattention may potentially lead to misinterpretation or less efficient processing of emotional stimuli, particularly when stimuli is complex, ambiguous, or requiring high attentional demands. In the context of the current dissertation study, an attentional framework of ADHD would predict that atypical patterns of emotional reactivity (i.e., startle reactivity, skin conductance) would be directly associated with greater evidence of inattention (i.e., worse performance on continuous performance test ommission errors, greater inattentive symptomatology). Prior to reviewing the literature on ADHD with comorbid CD, the definition, development, outcomes, and physiologic correlates of CD-alone will be addressed. Conduct Disorder Prevalence and Definition Conduct Disorder affects 6-16% of males and 2-9% of females under the age of 18 (DSM-IV, 1994) or between 1.3 and 4 million children and adolescents in the United States (Kaplan & Sadock, 1995). Conduct Disorder is characterized by “a repetitive and persistent pattern of behavior in which the basic rights of others or major age-appropriate societal norms or rules are violated” (DSM-IV, 1994). Symptoms fall under 4 domains: (1) serious violations of rules (e.g., defiance, truancy), (2) deceitfulness or theft (e.g., stealing, conning), (3) destruction of property (e.g., fire setting, vandalism), and (4) aggression to people and animals (e.g., fighting, animal cruelty). See Appendix A for the complete DSM-IV diagnostic criteria for CD. Children with CD also appear to have

PAGE 37

23 serious emotional deficits. For instance, they may lack empathy and appropriate feelings of guilt, have little concern for feelings and well-being of others, misperceive the intentions of others in ambiguous situations as more hostile and threatening, and fail to inhibit antisocial behavior regardless of potential punishment (Kaplan & Sadock, 1995). Development and Outcomes of Conduct Disorder There appear to be multiple interacting etiologies in the development of CD. Typically, mild delinquent behaviors emerge first, with more severe behaviors gradually surfacing later (Kaplan & Sadock, 1995). Conduct Disorder is very heterogeneous with respect to etiology, symptomatology, course, and treatment response (Frick, 1988). Child risk factors include: inappropriate early aggression, hyperactivity, difficult temperament, neuropsychological deficits, and male gender. Family risk factors include: inconsistent and authoritarian parenting, poverty, and a history of parental alcohol dependence, mental illness, ADHD, or CD (Barkley, 1998; Kaplan & Sadock, 1995). Conduct Disorder is associated with an early onset of drinking, smoking, sexual behavior, reckless behavior, and use of illegal substances, as well as an increased risk for future criminal behavior, incarceration, alcohol abuse, marital discord, occupational impairment, and social impairment (Robins, 1966; DSM-IV, 1994). Further, conduct behavior problems may be the greatest childhood behavioral risk factor for later antisocial behavior (Loeber & Dishion, 1983; Robins, 1966, 1978; White et al., 1990). It is estimated that up to 40% of children with CD will meet criteria for Antisocial Personality Disorder (APD) in adulthood (Barkley, 1998; Kaplan & Sadock, 1998; Lynam, 1996; Robins 1966, 1978).

PAGE 38

24 Neurobiologic Correlates of CD Unfortunately, relatively little is known about the neuroanatomy of conduct disorder in children. However, there appears to be limited evidence for reduced right temporal lobe volume reductions (Kruesi et al., 2004) and frontal lobe abnormalities based on neuropsychological studies of adolescents with conduct disorder (Lueger & Gill, 1990). There is also limited evidence for frontal and temporal lobe abnormalities in adults with criminal/antisocial behavior based on neuropsychological, CT, PET, and MRI studies (Raine, 1993). Physiologic Correlates of Conduct Disorder Numerous studies have indicated that non-clinical samples of children with antisocial symptoms have lower resting heart rate levels, decreased heart rate reactivity, and decreased electrodermal reactivity (Delamater & Lahey, 1983: decreased SC and HR reactivity in CD children; Maliphant et al., 1990: lower resting HR and during Ravens Matrices Test; Raine & Venables, 1984: lower resting HR levels in antisocial adolescents; Rogeness et al., 1990: lower resting HR in boys and girls with CD; Schmidt et al.1985: decreased SC reactivity to a 90dB sound in CD). Still, others have failed to find differences between children with and without CD on measures of basal HR and SC, and SC reactivity to 84 dB tones (Garralda et al., 1991). Overall, most findings still suggest that children with CD may be under-aroused or biologically under-responsive. Unfortunately, many of these studies did not account for comorbid ADHD in their samples. Furthermore, Raine et al. (1990) examined a non-clinical sample of 15 year old adolescents and found that lower resting HR and SC, and reduced HR and SC orienting to neutral tones predicted later criminal activity at age 24 years. Raine et al. (1995)

PAGE 39

25 prospectively examined the criminal records of these individuals again at age 29 years. He found that those antisocial adolescents who had subsequently desisted from adult crime had higher levels of autonomic arousal and orienting. Together, these studies suggest that autonomic arousal levels in childhood and adolescence may be predictive of adult antisocial behaviors. Still, these issues need to be examined prior to adolescence and in clinical samples, so that biological markers can be directly applied to treatment prevention strategies. ADHD with CD comorbidity Prevalence and Risk Factors Approximately 25% of children with ADHD are also diagnosed with Conduct Disorder (Brown et al., 2001). The presence of conduct problems appears to influence the manifestation and course of ADHD. Specifically, the presence of CD in children with ADHD is associated with more aggressive, severe, and persistent behaviors (Frick, 1988). Children with ADHD/CD typically have more serious and earlier onset antisocial behaviors, traffic offenses, failing a grade, suspension and expulsion from school, compared to ADHD alone and CD alone (Babinski, 1999; Barkley, 1998). Additionally, increased risk for substance use/abuse in adolescents with ADHD/CD has been attributed to the presence of CD symptoms, and not ADHD symptoms (Barkley, 1998). Farrington et al. (1989) found that while CD and ADHD independently predicted delinquency (35% for CD alone, 24% for ADHD alone, 12.6% controls), ADHD/CD most highly predicted delinquency (46%). Satterfield et al. (1994) compared aggressive versus non-aggressive children with ADHD. They found that 43% of the aggressive group versus 26% of the non-aggressive group were convicted of one or more felonies in early adulthood. Overall, children with CD/ADHD have higher rates of antisocial behavior in adolescence and

PAGE 40

26 Antisocial Personality Disorder (APD) in adulthood, compared to CD alone or ADHD alone (Burke et al., 2000; Loeber et al., 2000; Lynam, 1996). Psychophysiologic Correlates of ADHD/CD Again, most studies of autonomic arousal in ADHD have not accounted for comorbid conduct symptoms. This appears to be an extremely important issue. Specifically, Delamater and Lahey (1983) found that children who had both hyperactivity and conduct problems showed lower skin conductance reactivity than children with only hyperactivity. Pelham et al. (1991) found that high-aggressive ADHD boys responded to provocation with lower physiological arousal (heart rate) than low-aggressive ADHD boys. More recently, Herpertz et al. (2001) examined psychophysiological responses in 8 to 13 year-old ADHD boys with and without conduct disorder (CD), and healthy controls. They found no significant difference between the ADHD and ADHD/CD groups for resting heart rate and the number of spontaneously occurring skin conductance responses in the absence of a stimulus over a 3-minute rest period. There were also no group differences for amplitude of electromyographic (EMG) eye-blink response to acoustic startle stimuli (i.e., 100dB-50ms burst of white noise). However, ADHD/CD children showed reduced autonomic orienting (i.e., lower skin conductance amplitudes and more rapid habituation to 65dB tones), compared ADHD and control children. The ADHD/CD group also showed reduced skin conductance amplitude and more rapid habituation to acoustic startle stimuli, compared to ADHD children. Overall, research suggests that co-occurring conduct symptoms significantly impact psychophysiological responding within ADHD. In addition, children with ADHD and conduct-related symptoms appear to have physiological similarities to antisocial adults. Specifically, Raine (1993) extensively reviewed the psychophysiological correlates of

PAGE 41

27 criminal and antisocial behavior in adults. Raine concluded that (1) some evidence has shown reduced resting SC activity in antisocial individuals, (2) early studies showing decreased SC reactivity to aversive stimuli (90-120 dB sounds) have not been replicated in recent studies of either psychopathic or non-psychopathic antisocial individuals, and (3) non-clinical samples of antisocial individuals consistently show lower resting HR levels. Psychopathy Lynam (1996) has argued that the group of children who manifest both CD and ADHD symptoms contain the future psychopathic adult. Lynam (1997) found that boys who had both CD and ADHD symptoms had elevated scores on a childhood psychopathy scale compared to normal boys and boys with only ADHD symptoms, supporting the argument that these children may be at the greatest risk for developing psychopathy. Psychopathy is conceptualized as a personality type and is related, but unique from anti-social personality disorder (APD), which is behaviorally based. Psychopaths are characterized as impulsive, manipulative, lacking in empathy and remorse, cold, selfish, shallow, irresponsible, and involved in criminal acts. The psychopathic adult criminal offender is more violent, commits a greater variety of and more crimes, and has a higher recidivism than the average non-psychopathic criminal. Approximately 90% of psychopaths meet criteria for APD, but only 25% of individuals with APD meet criteria for psychopathy (Lynam, 1998). Overall, there is a great likelihood that there are subgroups of children with ADHD and comorbid conduct problems who will develop APD, psychopathy, or both APD and psychopathy. Since individuals with psychophathic personality traits appear be responsible for a disproportionate amount of societal harm, it is worthwhile to attempt to identify children at highest risk for psychopathy. Pertinent to

PAGE 42

28 this dissertation study, adult psychopaths show abnormal affective startle modulation (Herpertz et al., 2001; Levenston et al., 1999; Patrick et al., 1993). Children who show a similar pattern of reactivity maybe tomorrow’s adult psychopath. Affective modulation of the startle reflex has the potential to be reliable and valuable tool for identifying high-risk children. The first step in this line of research must be to identify whether aberrant patterns of reactivity exist in specific subgroups of children. Oppositional Defiant Disorder (ODD) Comorbidity with ADHD Approximately 35% of patients with ADHD also have oppositional defiant disorder, (Brown et al., 2001). Prevalence rates of ODD during childhood across several epidemiological studies ranges from 2.1% to 15.4% depending on the population sample (Loeber et al., 2000). The essential features of ODD are a recurrent “pattern of negativistic, hostile, and defiant behavior” lasting for at least 6 months and leading to significant impairment in social, academic, or occupational functioning (DSM-IV, 1994). Please see Appendix A for the complete DSM-IV diagnostic criteria for ODD. The role of ODD symptoms may play an important role in the development, course, and severity of conduct problems within ADHD children. Specifically, persistent ODD symptoms often precede and predict the early onset of CD (Loeber et al., 2000). Some scientists have also questioned whether ODD and CD are truly distinct disorders or whether new classifications are needed to better characterize these children. For example, some have proposed separating ODD into non-aggressive versus aggressive subtypes, while other’s have supported one syndrome for ODD with aggressive CD behaviors versus a second syndrome for nonaggressive CD behaviors (Loeber et al., 2000). Related to this issue, Bierderman et al. (1996) found evidence for two subtypes of ODD within ADHD in a 4-year follow-up study of 140 ADHD children. One subtype

PAGE 43

29 consisted of ADHD/ODD children with comorbid CD (32% of ADHD/ODD group), with ODD onset preceding CD onset by several years. Of the children with ADHD and CD, only one did not meet criteria for comorbid ODD at baseline and follow-up. The second subtype consisted of ADHD/ODD children (68% of ADHD/ODD) without CD at baseline. They were characterized as subsyndromal to CD, and were not at increased risk for meeting CD criteria at the 4-year follow-up. This ADHD/ODD group also showed less severe ODD than the ADHD/ODD/CD group, which was partially explained by less frequent endorsement of the DSM-IIIR ODD symptoms “bullying others,” “swearing or using bad language,” and “often spiteful and vindictive.” Given that ODD symptoms often precede CD, play a significant role in the development of CD, and represent a subset of conduct-related problems, it seems worthwhile to investigate whether severity of ODD symptomatology impacts emotional reactivity in ADHD children.

PAGE 44

CHAPTER 3 SUMMARY, AIMS, AND HYPOTHESES Summary and Aims The overall purpose of the study was to compare patterns of emotional modulation of the startle reflex of children with ADHD to children without ADHD. A secondary aim was to investigate whether the severity of conduct-related (i.e., CD and ODD) and psychopathy symptoms impact emotional reactivity within ADHD. In addition, since impulsivity appears to play a critical role in the development of maladaptive behavior in ADHD, this study also examined the relationship between impulsivity and emotional reactivity (subjective and physiological) in ADHD. Few studies have examined affective modulation of the startle reflex in normal children or children with ADHD. Affective modulation of the startle eyeblink reflex has the potential to be a valuable tool for identifying which ADHD children may be at increased risk for developing antisocial personality disorder and/or psychopathy in adulthood. Therefore, the first step in this line of research must be to identify whether aberrant patterns of reactivity exist in specific subgroups of children. Such an understanding of brain behavior linkages in childhood neuropsychiatric disorders would potentially contribute to a better conceptualization ADHD and comorbid conduct problems. 30

PAGE 45

31 Hypotheses/Predictions Based on prior research investigating affective modulation of the startle reflex in children and adults, research examining other psychophysiologic correlates of ADHD and conduct problems, and models of ADHD, the following are predicted: ADHD Compared to Healthy Controls 1. Affective modulation of the startle reflex (primary emotion measure): The ADHD and healthy control groups will differ in their pattern of eyeblink magnitudes (T-scores) across the three picture valence categories (i.e., pleasant, neutral, unpleasant). There will also be differences across the unpleasant picture subcategories (i.e., direct threat, injury, victimization) and pleasant picture subcategories (i.e., nurture, attraction, thrill). The healthy control group will exhibit patterns of reactivity similar to what is typically seen in normal adults. Specifically, they will show an increase (i.e., potentiation) in startle eyeblink magnitude for unpleasant pictures, and a decrease for pleasant pictures, relative to neutral pictures (i.e., linear pattern). Within the unpleasant picture subcategory, they will show the greatest potentiated startle to direct threat pictures, relative to injury and victimization pictures. Within the pleasant picture subcategory, they will show greater inhibition (i.e., more of a decrease) of startle for nurture and attraction pictures, compared to thrill pictures. The ADHD group will exhibit decreased startle potentiation for unpleasant pictures compared to healthy controls. Although reduced compared to controls, they will still show greater overall reactivity to unpleasant pictures compared to pleasant pictures. Within the unpleasant picture subcategory, they will show potentiated startle response to only direct threat pictures, but inhibited startle

PAGE 46

32 response to injury and victimization pictures. Within the pleasant picture subcategory, they will show similar levels inhibition of startle across all three subcategories (i.e., nurture, attraction, thrill). Based on results of the normative study (Selke, 2001), there will also be an interaction of age (i.e., older, younger children) with startle reactivity. Specifically, younger children in both healthy control and ADHD groups will show less “normal adult-like” patterns of startle reactivity compared to older children in the healthy control and ADHD groups, explained by reduced potentiated startle response to unpleasant pictures. 2. Skin conductance (secondary emotion measure): The healthy control group will show greater skin conductance response for both unpleasant and pleasant pictures, relative to neutral pictures (i.e., quadratic pattern). There will be no differences in skin conductance across the different unpleasant picture subcategories or across the pleasant picture subcategories. The ADHD group will also show greater skin conductance response for both unpleasant and pleasant pictures, relative to neutral pictures (i.e., quadratic pattern), but will show overall reduced reactivity to unpleasant, neutral, and pleasant pictures compared to healthy controls. There will be no differences in skin conductance across the different unpleasant picture subcategories or across the pleasant picture subcategories. 3. Subjective ratings (secondary emotion measures): There will be no group effects for valence and arousal ratings. For the valence subjective ratings, both groups will rate unpleasant pictures (a

PAGE 47

33 priori selected) as more negative on valence, and pleasant pictures as more positive on valence, relative to neutral pictures (i.e., linear pattern). There will be no differences in valence ratings across the different unpleasant picture subcategories or across the pleasant picture subcategories. For the arousal subjective ratings, both groups will rate the unpleasant and pleasant pictures as more arousing than the neutral pictures (i.e., quadratic pattern). There will be no differences in arousal ratings across the different unpleasant picture subcategories or across the pleasant picture subcategories. 4. Raw startle response (exploratory measure): Most researchers do not use absolute or raw startle magnitude to examine patterns of reactivity across affective pictures because of between-subject variability in overall startle magnitude (i.e., due to electrode placement and individual differences). Use of absolute values can result in some participants having greater “weight” or contribution to group effects than other participants, reducing the statistical meaningfulness of findings. Nonetheless, examining absolute startle values may be useful in investigating physiological and emotion processing in the context of ADHD. Specifically, there is some limited but inconsistent evidence to predict differences in absolute startle response between ADHD and controls. Several studies (Goldstein & Blumenthal, 1995; Herpertz et al., 2001; Ornitz et al., 1997; Ornitz et al, 1992) have not found absolute startle magnitude differences between ADHD and controls, although these studies examined startle in the context of attention, habituation, or general response, and not emotion modulation. In addition, one very recent study (Van Goozen et al., 2004) did find absolute startle magnitude differences between controls and a disruptive behavior

PAGE 48

34 disorder children (included children with ADHD) in a study of affective modulation of the startle reflex using picture stimuli. Consistent with models of under-arousal in ADHD, it was predicted that compared to controls, ADHD children will show decreased absolute startle magnitude across all picture categories (unpleasant, neutral, pleasant). Unpleasant and pleasant subcategories will not be directly examined. 5. Startle eyeblink onset latency (exploratory measure): Startle onset latency has only rarely been reported in the ADHD literature. In studies of attentional processing, Ornitz et al. (1992) did not find differences between ADHD and healthy controls, whereas Goldstein & Blumenthal (1995) found that 7 ADHD children had quicker starlet onset latency than 4 children with ODD or Posttraumatic Stress Disorder. Based on the limited evidence for group differences, it was predicted that there would be no group effects or interactions with picture category or subcategory. Psychopathy, CD Symptoms, ODD Symptoms, and Impulsivity Within ADHD 1. Psychopathy: Higher levels of psychopathy (Total CPS score) will be associated with reduced reactivity to affective stimuli (eyeblink magnitude and SC) within the ADHD group. Specifically, ADHD children with more psychopathy symptoms will show reduced startle potentiation to aversive stimuli, and reduced skin responses to both unpleasant and pleasant pictures. Psychopathy symptoms will not be associated with valence and arousal self-report ratings. 2. CD symptoms: A higher number of CD symptoms (Total from P-ChIPS) will be associated with reduced reactivity to affective stimuli (eyeblink magnitude and SC) within the ADHD group. Specifically, ADHD children with more CD symptoms will show reduced startle potentiation to aversive stimuli, and reduced skin responses to both unpleasant and pleasant pictures. Conduct Disorder symptoms will not be

PAGE 49

35 associated with valence and arousal self-report ratings. 3. ODD symptoms: A higher number of ODD symptoms (Total from P-ChIPS) will be associated with reduced reactivity to affective stimuli (eyeblink magnitude and SC) within the ADHD group. Specifically, ADHD children with more ODD symptoms will show reduced startle potentiation to aversive stimuli, and reduced skin responses to both unpleasant and pleasant pictures. ODD symptoms will not be associated with valence and arousal self-report ratings. 4. Impulsivity: Greater impulsivity as determined by the CPT (Beta) and the Restless-Impulsive Subscale score (CPRS-R:L) will be associated with reduced reactivity to affective stimuli (eyeblink magnitude and SC). Specifically, ADHD children with lower Beta scores (i.e., greater risk-taking tendency) and Restless-Impulsive Indices will show reduced startle potentiation to aversive stimuli, and reduced skin responses to both unpleasant and pleasant pictures.

PAGE 50

CHAPTER 4 RESEARCH DESIGN AND METHODS Diagnostic, Screening, and Descriptive Measures All testing took place in the Cognitive Neuroscience laboratory at the University of Florida Brain Institute. Following informed consent and assent, the caregiver completed the Demographic Questionnaire, Children’s Interview for Psychiatric Syndromes-Parent Version (P-ChIPS), Childhood Psychopathy Scale (CPS), and the Conners’ Parent Rating Scales-Revised, Long Form (CPRS-R), while the child completed the Wechsler Abbreviated Scale of Intelligence (WASI), Conners’ Continuous Performance – II (CPT-II), and the emotion psychophysiology task. Measures are described below. Children were administered the following: Wechsler Abbreviated Scale of Intelligence (WASI): The WASI (Wechsler, 1999) is a brief measure of intellectual functioning for ages 6 through 89 years. The WASI consists of 4 subtests, two making up the Verbal Scale (Vocabulary and Similarities) and a two the Performance Scale (Block Design and Matrix Reasoning). The WASI was standardized on a large and highly representative sample, and has shown high test-retest reliability, internal consistency, and validity (i.e., high correlation with WISC-III). Children were administered only Vocabulary and Matrix Reasoning due to time considerations. For inclusion, subjects obtained an estimated IQ Scores > 80. 36

PAGE 51

37 Conners’ Continuous Performance – II (CPT-II): The CPT (Conners, 1995, 2001) was chosen because it appears to directly measure impulsiveness without contamination from other cognitive factors, and is routinely given when diagnosing ADHD (Barkley, 1998). Children look at a computer screen and press a key as quickly as they can whenever they see any letter of the alphabet except for the letter “x.” The CPT has normative data based on 520 general population individuals and 670 clinic-referred individuals age 4-70 years, and is able to discriminate between individuals with and without diagnosed ADHD. The CPT will be used to generate an index associated with impulsivity (Beta). Beta incorporates both the proportion of commission errors and hit rate, and is recognized as a measure of risk-taking or impulsive response tendency. The Parent/Caregiver completed the following: Demographic Questionnaire: This questionnaire will be used to collect background information about the child (educational, medical, and psychiatric history, etc.) and family members (education level, occupation, age, race, etc.) Children’s Interview for Psychiatric Syndromes-Parent Version (P-ChIPS): The P-ChIPS (Rooney et al., 1999) is a structured diagnostic interview developed by psychiatrists and psychologists for administration to parents. There are separate interview modules for 20 common DSM-IV Axis I psychiatric disorders seen in children and adolescents. Each module is composed of a series of questions that assess whether strict DSM-IV criteria are met for its respective disorder. Individual modules can be administered separately. The P-ChIPS has been shown to have good reliability and validity. For the proposed study, 4 modules (ADHD, CD, and Oppositional Defiant

PAGE 52

38 Disorder, Generalized Anxiety Disorder) were administered. The ADHD module was administered three times, one reflecting current symptomatology on medication, the second reflecting current symptomatology off medication, and the third reflecting symptoms at age of onset prior to medication treatment. The total number of current CD symptoms and total number of current ODD symptoms endorsed were used in the statistical analysis. Childhood Psychopathy Scale (CPS): The CPS (Lynam, 1997) is a 41-item scale designed to assess childhood psychopathy based on Hare’s (1991) Adult Psychopathy Checklist. The CPS has shown high overall internal consistency and good construct validity (i.e., total score correlated with antisocial behavior, impulsivity, & delinquency). While subscales exist (Glibness, Untruthfulness, Boredom Susceptibility, Manipulation, Lack of Guilt, Poverty of Affect, Callousness, Parasitic Lifestyle, Behavioral Dyscontrol, Lack of Planning, Impulsiveness, Unreliability, Failure to Accept Responsibility, and Grandiosity), Total CPS score was used in the analysis. Conners’ Parent Rating Scales-Revised, Long Form (CPRS-R): The CPRS (Conners, 1997) is an 80-item scale designed to assess ADHD and related behavior problems in children ages 3-17 years. The CPRS has shown excellent reliability and validity. The Restless-Impulsive Subscale score was used as an additional measure of behavioral impulsivity in the analysis. Participants The research protocol was approved by the Human Studies Committee at the University of Florida. All participants provided written informed consent (caregivers) and assent (children) prior to participation in the study according to the University of

PAGE 53

39 Florida IRB guidelines. Two overall groups of children participated in this study: (1) a target sample of children diagnosed with ADHD, and (2) healthy controls. Children with ADHD were recruited through the University of Florida (UF) ADHD Program, the UF Psychology Clinic, the UF Department of Child Psychiatry, and community advertisement. Healthy control children were recruited through community advertisement. Children in the ADHD group were all previously diagnosed with and medically treated for ADHD prior to study entry. The study aimed to recruit ADHD group with a wide range of CD and/or ODD symptomatology (i.e., from no CD/ODD to severe CD/ODD symptomatology). Children were paid $15 for their participation in the study and $5 for parking. Exclusionary criteria included significant hearing or visual impairment, mental retardation, <80 Estimated IQ Score on the Wechsler Abbreviated Scale of Intelligence (WASI), psychiatric diagnoses (i.e., major depression, anxiety disorders) other than ADHD, CD, or ODD, and any serious medical or neurological disorder (i.e., epilepsy, sickle cell anemia, diabetes, traumatic brain injury). Children with suspected or previously diagnosed learning disabilities were not excluded from the study. For inclusion into the study, children recruited into the ADHD group had to meet current DSM-IV diagnostic criteria for ADHD based on a structured diagnostic interview (Children’s Interview for Psychiatric Syndromes-Parent Version: P-ChIPS). Children were also evaluated for the presence of comorbid CD and/or ODD. Children with ADHD were included whether or not they had comorbid CD and/or ODD. Children recruited into the healthy control group could not meet current or prior DSM-IV diagnostic criteria for ADHD, CD, or ODD based on the P-ChIPS. In addition, children with ADHD were

PAGE 54

40 required to currently be taking medication for their ADHD symptomatology (e.g., stimulant medication, Strattera). Twenty-nine healthy control children and 40 children previously diagnosed and medicated for ADHD were recruited and provided written informed consent and assent. However, six healthy control and six ADHD children were not included in the statistical analysis. Specifically, four children recruited into the healthy control group were not included because they met current DSM-IV criteria for Attention-Deficit/Hyperactivity Disorder (all Inattentive Type) based on the structured diagnostic interview (P-ChIPS). These four children had never been formally diagnosed with or medically treated for symptoms of ADHD, although caregivers reported a history of inattention and/or hyperactivity in home and school settings. These four children are not included in the statistical analyses. In addition, one control and one ADHD child were not included in the data analysis because their startle eyeblink magnitudes exceeded the maximal EMG level that the equipment was capable of detecting at the set calibration. Another control child chose not to participate in the psychophysiological portion of the study after completing the WASI. One ADHD child became too uncooperative early into the psychophysiological procedures to complete the study. An additional two ADHD children did not complete the psychophysiology portion of the study because of equipment problems (i.e., program could not be accessed), and were unable to return another day to complete study procedures. Two other ADHD children were taking medication during the initial recruitment process, but on the day of testing it was discovered that they had recently discontinued taking their medication (> 1 week). While these two children completed all study procedures, they were not included in the

PAGE 55

41 statistical analyses. Overall, 23 healthy control children and 34 ADHD children were included in the statistical analyses. Children were between the ages of 7-years, 0-months and 15-years, 1-month. The ADHD and healthy control groups were well matched for age, gender, and ethnicity. Specifically, unpaired t-tests (2-tailed) did not reveal a reliable difference between the control (mean age = 11.10 + 2.11 s.d.) and ADHD (mean age = 10.80 + 2.53 s.d.) groups for age t(55)=0.46, p=0.65. Chi-square tests did not reveal reliable frequency differences between the control and ADHD groups for gender, 2(3, N=57)=0.04, p=.85, or ethnicity (i.e., Caucasian, non-Caucasian), 2(1, N=57)=0.54, p=.46. The ADHD group consisted of 67.6% males (n=23) and 32.4% females (n=11), and the healthy control group consisted of 65.2% males (n=15) and 34.8% females (n=8). In addition, the ADHD group consisted of 79.4% Caucasian (n=27), 17.6% African American (n=6), 2.9% Hispanic (n=1) children, whereas the healthy control group consisted of 87.0% Caucasian (n=20), 8.7% Hispanic (n=2), and 4.3% Asian American (n=1) children. There were no reliable differences between the ADHD and control groups for male caregiver years of education (t(42)=1.16, p=0.25; ADHD mean = 14.26 + 2.54 s.d., control mean = 15.05 + 1.86 s.d.). However, there were reliable differences for female caregiver years of education (t(42)=1.16, p=0.25; ADHD mean = 14.17 + 1.89 s.d., control mean = 15.35 + 1.61 s.d.), WASI Vocabulary subtest (t(55)= 2.06, p=0.04; ADHD mean T-score = 52.27 + 12.06 s.d., control mean T-score = 58.61 + 10.29 s.d), WASI Matrix Reasoning subtest (t(55)= 2.45, p=0.02; ADHD mean T-score = 52.88 + 10.24 s.d., control mean T-score = 58.61 + 5.45 s.d.), and WASI estimated IQ (t(55)= 2.47, p=0.02; ADHD mean IQ = 105.12 + 16.42 s.d., control mean IQ = 115.22 + 12.98

PAGE 56

42 s.d.), explained by the ADHD group having lower female caregiver education and lower IQ scores. Moreover, compared to the control group, ADHD children showed reliably higher elevation of DSM-IV ADHD, CD, and ODD symptoms (all p<.0001, P-ChIPS, see Table XX), psychopathy symptoms (p<.0001, Child Psychopathy Scale), and all Connors’ Parent Rating Scale subscale scores (all p<0.003, see Table 4-2). In contrast, there were no reliable differences between controls and ADHD children on all the Connors’ Continuous Performance Test variables, with the exception of mean Hit Reaction Time (see Table 4-3). Specifically, the ADHD group showed slower hit reaction time than the control group, t(53)= -2.56, p=0.01, although mean reaction time was not within the clinically significant range for either group. Based on these results, for the most part, the medicated ADHD sample performed similarly to healthy control children with respect to visual sustained attention. Within the ADHD group, 13 children met DSM-IV diagnostic criteria for only ADHD, 10 children met DSM-IV diagnostic criteria for comorbid Oppositional Defiant Disorder (ODD) but not Conduct Disorder (CD), and 11 children met DSM-IV diagnostic criteria for comorbid CD (8 of these 11 children also met criteria for ODD). There were no reliable differences between the ADHD, ADHD + ODD only, and ADHD + CD children for male or female caregiver years of education (all p>.24). However, while the ADHD + ODD children were younger than the ADHD children (t(21)= 2.18, p=0.04; ADHD mean = 11.58 + 2.23 s.d. years, ADHD + ODD mean = 9.59 + 2.11 s.d. years, ADHD + CD mean = 10.98 + 2.98 s.d. years), there were no reliable differences between ADHD + CD and ADHD (p=.58) or ADHD + CD and ADHD +ODD (p=.24) children. In addition, the ADHD + CD children had reliably lower WASI estimated IQ than the

PAGE 57

43 ADHD children (t(22)= 2.16, p=0.04; ADHD mean IQ = 108.85 + 12.30 s.d., ADHD + CD mean IQ = 96.64 + 15.45), although there were no reliable differences between the ADHD + ODD and ADHD (p=.91) or ADHD + ODD and ADHD +CD (p=.11) children. The ADHD children included 8 males and 5 females, the ADHD +ODD children included 5 males and 5 females, and the ADHD + CD children included 10 males and 1 female. Within the ADHD group, 17 children were classified as Predominantly Inattentive Type ADHD and 17 were classified as Combined Type ADHD, based on parent report of current “off-medication” ADHD symptoms using DSM-IV diagnostic criteria (P-ChIPS). Notable, 8 of the 11 children diagnosed with ADHD + CD were classified as Combined Type ADHD. Reported mean age of ADHD symptom onset was 4.85 + 1.48 s.d. years old. All ADHD children took their regularly scheduled medication the day of testing, which included at least one stimulant medication (i.e., Concerta, Adderall, Ritalin, Dexedrine, Focaline, and/or Stattera). Twenty ADHD children were prescribed only one medication, ten were prescribed two medications, three were prescribed three medications, and one was prescribed four medications. Stimulant and non-stimulant medications included Strattera, Ritalin, Concerta, Adderall XR, Focaline, Dexedrine, Clonidine, Tenex, Prozac, Depakote, Remeron, and Wellbutrin.

PAGE 58

44 Table 4-1. P-ChIPS symptoms in control and ADHD children Control (n=23) ADHD (n=34) Variable Mean S.D. Mean S.D. DSM-IV ADHD Symptoms At age of onset Inattentive ----8.22 1.24 Hyperactive-Impulsive ----5.97 2.91 Current on medication Inattentive ----5.89 2.48 Hyperactive-Impulsive ----3.85 3.46 Current off medication Inattentive 0.57 1.04 8.43 0.94 Hyperactive-Impulsive 0.44 0.95 5.00 3.49 DSM-IV CD Symptoms 0.09 0.29 1.74 1.64 DSM-IV ODD Symptoms 0.57 0.95 3.71 2.69 Table 4-2. Conners’ Parent Rating Scale T-scores Control (n=23) ADHD (n=34) Subscale Mean S.D. Mean S.D. Oppositional 48.74 8.29 65.24 13.76 Cognitive Problems/Inattention 46.52 4.77 70.12 9.77 Hyperactivity 46.26 4.20 68.06 15.10 Anxious-Shy 46.26 9.40 58.06 15.10 Perfectionism 46.22 4.09 56.94 12.51 Social Problems 45.91 1.78 64.71 16.99 Psychosomatic 45.78 13.10 61.53 11.46 Connors’ADHD Index 45.61 4.53 70.85 9.35 Connors’ Restless-Impulsive 45.17 4.90 70.97 11.5 Connors’ Emotional Lability 46.17 10.32 60.44 14.77 Connors’ Global Index: Total 45.26 7.32 68.77 12.58 DSM-IV: Inattentive 46.87 5.90 70.32 9.77 DSM-IV: Hyperactive-Impulsive 47.70 6.06 68.53 14.37 DSM-IV: Total 47.26 5.48 71.12 11.09

PAGE 59

45 Table 4-3. Connors’ Continuous Performance Test T-scores Control (n=23) ADHD (n=33) Variable Mean S.D. Mean S.D. Omission Errors 48.79 8.96 49.49 10.14 Commission Errors 55.44 8.93 49.82 12.22 Hit Reaction Time 39.50 9.24 48.37 14.40 Hit Reaction Time Standard Error 47.74 8.74 51.28 10.29 Variability 48.99 8.49 51.16 10.51 Detectability (d’) 55.08 10.21 48.31 14.49 Response Style (Beta) 50.71 13.42 50.75 10.68 Perseverations 62.24 37.30 55.37 19.95 Hit Reaction Time Block Change 51.49 10.75 49.29 10.08 Hit Standard Error Block Change 52.81 8.77 50.56 11.75 Hit Reaction Time Inter-Stimulus Interval Change 49.74 8.53 53.40 12.00 Hit Standard Error Inter-Stimulus Interval Change 48.08 9.81 51.05 10.47 Confidence Index (Percentile) 45.0 24.8 48.7 18.2 Emotion Psychophysiology Task Apparatus and Procedures Physiological sensors were attached to the child while he/she was sitting in an electrically shielded, sound attenuated room. Surface Ag-AgCl electrodes filled with the appropriate electrode gel (e.g., isotonic electrolyte for SC) were positioned under the left and right eyes to record EMG activity from orbicularis oculi eye muscle activity, and on the thenar and hypothenar eminence of left and right palm to measure skin conductance. After electrode placement, headphones were comfortably positioned over participants’ ears. Participants then sat quietly for 5 minutes while the psychophysiology equipment was calibrated. Still sitting quietly, participants were presented with a set of 6 acoustic startle probes delivered every 10-15 seconds. This initial procedure was used primarily to correct any problems with electrode placement in order to obtain optimal recordings,

PAGE 60

46 and to acclimate subjects to the startle probe noise and their reflexive eyeblink response. The startle eyeblink response during this initial period and the remainder of the experiment was elicited by a 50ms burst of white noise (i.e., startle probe) reaching approximately 100dB with instantaneous rise time. The white noise bursts were produced by a Colbourn S81-02 module and then gated through a Colbourn S82-24 amplifier and delivered binaurally through Telephonics (TD-591c) stereo headphones. Next, participants were provided with instructions about using the Self-Assessment-Manikin (see description in Chapter 2, Pictorial in Appendix B, and instructions in Appendix C). Subjects then completed practice valence and arousal ratings on a pleasant, unpleasant, and neutral picture. The experimental session began after subjects were instructed to ignore any loud noise presented into the headphones and to attend to each picture during its entire presentation. Two preliminary trials consisting of a neutral picture and an acoustic startle probe were presented to again acclimate the participants to the experimental procedures. This was followed by presentation of 51 (36 primary startle probe trials + 15 trials without the startle probe noise) pictures arranged in 3 blocks of 17 pictures. Each block consisted of 12 startle probe picture trials plus 5 interspaced non-startle-probe picture trials. The ratio of startle to no-startle probe trials (70.6 % startle) was consistent with the majority of past studies examining affective modulation of the startle reflex using picture stimuli (66-85% of trials have startle probe) (Cuthbert et al., 1996; Hamm et al., 1997; Levenston et al., 2000; Patrick et al., 1993; Vrana et al., 1988). Each block contained 4 exemplars of each valence (4 pleasant, 4 unpleasant, 4 neutral) for the probe trials. The slide block order was repositioned to create

PAGE 61

47 two different presentation orders counterbalanced across participants. There were also an equal number of pleasant, unpleasant, and neutral pictures across the no-probe trials. Overall, the 36 primary startle probe trials consisted of 12 pleasant, 12 unpleasant, 12 neutral pictures. For males, 32 of the 36 primary picture stimuli were chosen from the International Affective Picture System (Lang et al., 1999). For females, 30 of the 36 primary picture stimuli were chosen from the International Affective Picture System. The remaining primary picture stimuli were selected from other public sources to increase the number of pictures relevant to the interests of children. The pleasant and unpleasant picture stimuli were all previously used and evaluated in two preliminary studies with children (Selke et al., 2005). Pictures were selected based on specific content, and with the following considerations: a) that they would be developmentally and ethically appropriate for children between the ages of 7 to 14 years, b) that there would be a variety of pictures relevant to the interests of both males and females, and c) that the pleasant and unpleasant pictures would be roughly equivalent on rated arousal based on the preliminary studies. Moreover, the 12 unpleasant pictures were a priori selected to be divided into 3 subcategories, including 4 “direct threat,” 4 “victim,” and 4 “injury” pictures. Direct threat pictures depicted a weapon such as a gun pointed toward the participant/viewer. Victim pictures depicted one individual threatening or assaulting a second individual or the “victim.” Injury pictures depicted an individual who has been injured (i.e., burn victim) or an injured part of a body (i.e., an infected tongue). The 12 pleasant pictures were also a priori selected to be divided into 3 subcategories, including 4 “nurture,” 4 “thrill,” and 4 “attraction” pictures. Nurture pictures depicted cute animals (kittens,

PAGE 62

48 puppies, dolphins) or a human baby. Thrill pictures depicted action or adventure, such as individuals on a waterslide, whitewater rafting, trick skating, or windsurfing. Attraction pictures depicted attractive individuals of the opposite sex than the participant, such as man lifting weights for female participants or a woman swimming in a pool for male participants (no nudity was shown to participants). Two standardized sets of pictures were used, one for each gender. Thirty out of the thirty-six primary pictures were identical for each gender. The six different pictures were four “attraction” pictures and two “thrill” pictures. Each picture was shown for 6 seconds on a 21-inch monitor located 3 feet in front of the subject. During slide viewing, startle eyeblinks were elicited by a 50 ms burst of white noise (100db: startle probe) with instantaneous rise time. Startle probe stimuli were randomly delivered at one of 3 intervals after slide onset (4200 ms, 5000 ms, 5800 ms) to reduce anticipatory responses, and occurred equally often across all slide categories. No acoustic startle probes were delivered during the “no-probe” picture trials. Following each picture, participants viewed a blank screen for 6 seconds. Participants then verbally rated the picture along the dimensions of valence and arousal using two independent 1-7 ordinal rating scales (i.e., the Self-Assessment Manikin). These scales were vertically arrayed on a slide and depicted by versions of a cartoon figure (see Appendix B). For valence, the cartoon figure ranged from very unpleasant to neutral to very pleasant, and for arousal the figure ranged from sleepy or bored to highly excited. Subjects verbally reported their ratings over an intercom. A research assistant situated in an adjacent control room manually recorded ratings. After completing the valence and arousal ratings for each picture, there was a 6 second intertrial interval (i.e., blank screen) prior to

PAGE 63

49 presentation of the next picture (Please See Appendix D for a visual display of a picture trial). Physiologic Recordings Eyeblinks were measured by recording EMG activity from the orbicularis oculi muscle beneath each eye using Ag-AgCl electrodes. The raw EMG signal was amplified (30,000 gain) and frequencies below 90 Hz and above 1000 Hz were filtered using Colbourn bioamplifiers. The raw signal was then rectified and integrated using a Colbourn Contour Following Integrator with a time constant of 200 milliseconds. Digital sampling at 1000 Hz began 50 ms before presentation of the acoustic startle probe and continued for 250 ms after startle probe offset. Skin conductance, measured from the thenar and hypothenar eminence of each palm, was be sampled at 20 Hz using two Coulbourn Isolated Skin Conductance couplers in DC mode. This information was sent to a Scientific Solutions A/D board interconnected with a Gateway PII-500 PC. Data Reduction The startle magnitude and skin conductance data was reduced offline using a custom software program for data condensing. For startle eyeblinks, each trial was scored for amplitude (i.e., peak – baseline in microvolts) during the 30-130 ms interval following the acoustic stimulus onset. Raw scores were converted into standard z-scores [(trial amplitude – mean amplitude)/ standard deviation] and then into T-scores (mean of 50, standard deviation of 10) for each subject’s left and right eye separately. T-score conversions were used to minimize the between-subject variability in overall startle magnitude. Trials with T-scores greater than 90 or less than 10 were eliminated to reduce the impact of outliers. Left and right eyeblink response T-score values were then averaged into a composite score for each of the 36 trials. These composite scores were

PAGE 64

50 used to produce average startle eyeblink responses for each picture category (i.e., pleasant, unpleasant, neutral). If a signal was not detected for one eye on an individual trial, the eyeblink response from the other eye was used in place of an average response. For skin conductance, each trial was scored for the largest change in SC between 0.9 and 4.0 seconds after picture onset. Raw values for the left and right palms were averaged into a composite score for each picture trial. The composite scores had a value of ” added to eliminate negative values, and then be log transformed to reduce skewness (Levenston, et al., 2000; Hamm, et al., 1997). Study Design and Analyses The study design incorporated two main statistical analyses, with one series comparing ADHD children to Healthy Control Children, and the second series addressing the impact of psychopathy, CD symptoms, ODD symptoms, and impulsivity within the ADHD group. In addition, a third series of analyses was performed to aid in the interpretation of the second series (i.e., dimensional approach to comorbidity). Specifically, the third series compared children diagnosed with only ADHD to those diagnosed with ADHD plus comorbid ODD, to those diagnosed with ADHD plus comorbid CD. There were also a number of exploratory analyses completed, including examination of startle latency and raw startle magnitude, several logistic regressions to examine the predictive value of startle reactivity, as well as examination of a number of other independent variables such as inattentive symptomatology and ADHD subtype. Series One: ADHD Versus Controls The first analysis incorporated both withinand between-subject variables. The within-subjects factor was Picture Category with 3 levels (pleasant, unpleasant, and neutral pictures). The between-subjects factors were Group with 2 levels (Controls;

PAGE 65

51 ADHD) and Age with 2 levels (Younger: 7-10 years; Older: 11-14 years). Age was initially used as a between-subjects variable because the Normative Study (Selke, 2005) suggested that age impacted startle eyeblink reactivity during childhood. However, as discussed later in the results section, age effects (i.e., main effect, interactions) were not found with startle eyeblink magnitude, and therefore age was dropped from subsequent analyses. The primary dependent repeated measure was startle eyeblink magnitude, and the secondary dependent repeated measures were skin conductance response and self-report picture ratings (valence and arousal). Exploratory dependent repeated measures were raw startle eyeblink magnitude and startle eyeblink latency. Dependent measures were examined across overall Picture Category (unpleasant, neutral, pleasant), as well as across Unpleasant Picture Subcategory (direct threat, injury, victim) and Pleasant Picture Subcategory (nurture, attraction, thrill). Again, subcategory unpleasant and pleasant pictures were all a priori selected for content. Analyses were performed using SPSS Version 10.0 statistical software. The hypotheses predicted different patterns of change in (1) startle eyeblink response, (2) skin conductance response, and (3) affective self-report ratings across Picture Category for the ADHD and healthy control groups. As these hypotheses concerned the “form” of response to Picture Category, planned tests of trend over Picture Category (i.e., unpleasant, neutral, pleasant) were tested for both linear and quadratic components (Keppel, 1982). These single-df tests evaluated “main” effects of trend and “Group by trend” interactions. For instance, a significant Group (ADHD, control) by linear (i.e., pleasant versus unpleasant) trend interaction over Picture Category was predicted for the startle eyeblink repeated measure, and a significant Group by quadratic

PAGE 66

52 (i.e., unpleasant-pleasant versus neutral) trend interaction over Picture Category was predicted for the skin conductance repeated measure. Main effects of group were also examined when appropriate. Further, significant (p< 0.05) Group by trend interactions were followed up by decomposing the model into separate ADHD and Control tests of trend over Picture Category. For instance, this allowed or “gave permission” for the predicted linear (i.e., unpleasant versus pleasant) trend over Picture Category for startle eyeblink magnitude to be examined within the ADHD and control groups separately. If a significant Group by trend interaction was found for skin conductance, this allowed the predicted quadratic (i.e., unpleasant-pleasant versus neutral) trend over Picture Category to be examined within the ADHD and control groups separately. Significant Group by trend interactions were also followed up with univariate ANOVAs comparing the groups within each picture category separately. In addition, significant “main” effects (p< 0.05) of trend were followed up by pairwise comparisons. An index of effect size (2) was reported for significant main effects and interactions. Cohen’s “d” was reported for significant pairwise comparisons as an indication of change in standard deviation units. A similar statistical approach was used for Valence and Arousal Subjective Ratings. Moreover, the dependent measures (i.e., startle, skin conductance, self-report ratings) were examined across Unpleasant and Pleasant Picture Subcategories using repeated measures ANOVAs. Trend analyses were not used, as differences within subcategories were of primary interest rather than specific patterns of reactivity. Significant (p< 0.05) Group by Picture Subcategory interactions were followed up by decomposing the model into separate repeated measures ANOVAs for the ADHD and

PAGE 67

53 Control groups, and by decomposing the model into separated univariate ANOVAs for each picture subcategory (i.e., threat, injury, victim). In addition, significant main effects (p< 0.05) of Picture subcategory were followed up by pairwise comparisons. An index of effect size (2) was reported for significant main effects and interactions. Cohen’s “d” was reported for significant pairwise comparisons as an indication of change in standard deviation units. Please note that figures examining unpleasant picture subcategories for all dependent variables (ratings, startle, skin conductance) include the overall pleasant picture category as a point of reference within each figure, although pleasant picture category is not incorporated into the respective statistical analysis. Series Two: Variables within ADHD The second series of analyses addressed the impact of psychopathy, CD symptoms, ODD symptoms, and impulsivity (CPT-beta and Restlessness-Impulsivity Index) within the ADHD group. This set incorporated two within-subject independent variables. Specifically, Picture Category with 3 levels (i.e., pleasant, unpleasant, and neutral pictures) was a within-subjects factor. Total psychopathy score, number of P-ChIPS CD symptoms, number of P-ChIPS ODD symptoms, CPT-Beta score, and Conners’ Restlessness/Impulsive Index were each a within-subjects continuous variable (i.e., covariate) within its respective analyses. Again, the dependent repeated measures included startle eyeblink magnitude, skin conductance reactivity, and self-report picture ratings (valence and arousal). Absolute startle magnitude was an exploratory dependent variable to investigate whether comorbid CD, ODD, and psychopathy symptoms as well as measures of impulsivity were associated with decreased absolute values of startle reactivity. Most relevant to this question, the presence of a significant main effect for each of the independent variables (i.e., CD, ODD symptoms) across Picture Category

PAGE 68

54 (unpleasant, neutral, pleasant) for absolute startle was examined. Main effects of Picture Category and interactions with the independent variables are most appropriately examined using the standardized startle magnitude scores (i.e., t-scores), and are therefore are not reported for absolute startle magnitude. Please note that startle eyeblink latency was analyzed but not reported in the series two results section, as there were no associations with any of the independent variables. In addition, Connors’ DSM-IV Inattentive Index, Connors’ DSM-IV Hyperactive-Impulsive Index, Connors’ DSM-IV Total Index, CPT commission errors, and ADHD subtype were exploratory independent variables. Similar to series one, for Picture Category, as the primary hypotheses concern the “form” of response to Picture Category, planned tests of trend over Picture category (i.e., unpleasant, neutral, pleasant) was tested for both linear and quadratic components (Keppel, 1982). These single-df tests evaluated “main” effects of trend over Picture Category, as well as “Psychopathy by trend” or “CD by trend” or “ODD by trend” or “Beta by trend” or “Restless-Impulsive by trend” over Picture Category interactions. Most relevant in these analyses was whether interactions were present. Significant “main” effects of trend were typically redundant with findings in the series one analyses, and therefore not systematically decomposed into pairwise comparisons of the different picture categories (i.e., unpleasant versus pleasant, unpleasant versus neutral, pleasant versus neutral) within the ADHD group. Moreover, just as in the first series of analyses, the dependent measures (startle, skin conductance, self-report ratings) were examined across Unpleasant and Pleasant Picture Subcategories using repeated measures ANOVAs and not trend analyses.

PAGE 69

55 Series Three: ADHD, ADHD + ODD, ADHD + CD To aid in the interpretation of the dimensional statistical approach (i.e., number of ODD and CD symptoms), follow-up analyses were completed examining emotional reactivity across discrete groups within ADHD, including 13 children with only ADHD, 10 ADHD children with only comorbid ODD, and 11 ADHD children with comorbid CD (8 of these 11 also met criteria for ODD). The approach to these analyses was identical to the series one analysis.

PAGE 70

CHAPTER 5 SERIES ONE RESULTS: ADHD VERSUS CONTROLS Series One Hypotheses The hypotheses for the first series of results are presented in Table 5-1. Table 5-1. Series one hypotheses: ADHD versus healthy controls Picture Category Unpleasant Picture Subcategory Pleasant Picture Subcategory Startle Both Groups Unpleasant>Pleasant Unpleasant Pictures ADHDVictim, Injury All Potentiated ADHD Threat >Victim, Injury Threat Potentiated Victim, Injury Inhibited Controls Attraction, Nurture Neutral Controls Overall >ADHD Both Groups Threat =Victim=Injury Controls Overall >ADHD Both Groups Attraction=Nurture =Thrill Controls Overall >ADHD Self-Report Valence No Group Differences Both Groups Pleas>Neutral>Unpleas No Group Differences Both Groups Threat =Victim=Injury No Group Differences Both Groups Attraction=Nurture =Thrill Self-Report Arousal No Group Differences Both Groups Pleas, Unpleas >Neutral No Group Differences Both Groups Threat =Victim=Injury No Group Differences Both Groups Attraction=Nurture =Thrill Raw Startle Overall ADHD
PAGE 71

57 Startle Eyeblink Magnitude Picture Category Age was first examined as a covariate since a previous normative study (Selke, 2005) found an interaction of age with startle reactivity, with younger boys showing decreased startle potentiation to unpleasant pictures than older boys. For startle eyeblink magnitude, trend tests did not reveal a significant interaction of Age with linear (p=.97) or quadratic (p=.67) trend over Picture Category, and did not reveal a significant interaction of Age by Group by linear (p=.90) or quadratic (p=.22) trend. There was also no significant main effect of Age (p=.67) or significant Age by Group interaction (p=.97). There were similarly non-significant findings for Unpleasant Picture Subcategory (all p>.16). As previously noted, since age did not interact with the primary dependent measure, startle reactivity, age was not included as a variable in subsequent analyses. For startle eyeblink magnitude (Figure 5-1, Table 5-1), the trend tests revealed the predicted linear (i.e., pleasant versus unpleasant) trend over Picture Category (F[1,55]=28.34, p<.0001, power=1.00, 2=.34), but not a quadratic effect (p=.51). Pairwise comparisons revealed reliable differences between each of the picture categories (unpleasant>neutral: p=.009, d=.64; unpleasant>pleasant: p<.0001, d=1.20; pleasant
PAGE 72

58 46 47 48 49 50 51 52 53 54 55T-score + S.D. Unpleasant Neutral PleasantPicture Category ADHD Control Figure 5-1. Mean startle eyeblink response for control and ADHD groups across picture category Overall, both ADHD and control children showed the predicted linear “adult-like” pattern of affective modulation of the startle reflex, that is larger startle eyeblink reactivity to unpleasant pictures relative to unpleasant pictures. Unpleasant Picture Subcategory For startle eyeblink magnitude (Figure 5-2, Table 5-2), repeated measures ANOVA revealed a reliable main effect of Unpleasant Picture Subcategory (F[2,110]=10.38, p<.0001, power=.98, 2=.16). This main effect was explained by decreased startle response to victim pictures compared to direct threat (p<.0001, d =1.00) and injury (p<.0001, d=.81) pictures, but no difference between direct threat and injury pictures (p=.59), collapsed across groups. There was no main effect of group (p=.47). However, there was a reliable Group by Unpleasant Picture Subcategory interaction (F[2,110]=4.46, p=.014, power=.75, 2=.08). Each group and each unpleasant subcategory was then analyzed separately to determine the source of the interaction.

PAGE 73

59 44 46 48 50 52 54 56 58 60T-score + S.D. Threat Injury Victim PleasantUnpleasant Picture Subcategory ADHD Control Figure 5-2. Mean startle eyeblink response for control and ADHD groups across unpleasant picture subcategory. Table 5-2. Dependent measure means by picture category _Unpleasant_ __Neutral__ __Pleasant__ Measures Mean S.D. Mean S.D. Mean S.D. Self-Report Valence Control 2.37 .90 3.88 .60 5.20 .67 ADHD 2.73 1.01 4.21 1.01 5.27 .97 Self-Report Arousal Control 5.11 1.42 1.66 .84 4.18 1.24 ADHD 3.93 1.31 2.26 1.63 4.19 1.38 Startle Magnitude Control 51.62 2.30 49.06 2.50 48.78 1.85 ADHD 51.10 2.57 50.20 2.42 48.31 2.45 Raw Startle Value Control 7.62 4.86 6.59 4.47 6.26 3.79 ADHD 7.22 5.36 6.80 4.61 6.18 4.32 Startle Latency Control 66.95 7.11 66.55 8.65 65.12 7.78 ADHD 63.47 12.30 60.66 11.35 63.50 13.08 Skin Conductance Control .097 .075 .056 .047 .067 .043 ADHD .077 .097 .030 .044 .048 .060 Note. Units of measurement are as follows: startle magnitude T-score, raw startle value V, startle latency ms, and skin conductance log(1 + Siemens).

PAGE 74

60 For the control group, repeated measures ANOVA did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.35), and there were no reliable differences between the direct threat, injury, and victim pictures (all paired comparisons p>.17). In contrast, for the ADHD group, repeated measures ANOVA revealed a highly reliable main effect of Unpleasant Picture Subcategory (F[2,66]=15.51, p<.0001, power=1.00, 2=.32). Pairwise comparisons revealed highly reliable differences between victim and threat pictures (p<.0001, d=1.42; threat>victim) and between victim and injury pictures (p<.0001, d=1.29; injury>victim), but not between threat and injury pictures (p=.98). Moreover, univariate ANOVAs revealed no main effect of group for direct threat (p=.68) and injury (p=.23) pictures, but a significant main effect of group for victim pictures (F[1,55]=11.96, p=.001, power=.93, 2=.18, d=.92), explained by decreased startle response to victim pictures in the ADHD group compared to the control group. Moreover, a logistic regression was performed to inspect how well startle response to victim pictures predicted group membership. The logistic regression revealed that startle response to victim pictures was a significant predictor of group membership (Wald=8.48, p=.0036; Likeli hood Ratio Test=10.94, p=.0009), and resulted in 71.9% of the children being correctly classified, a sensitivity of 85.3%, a specificity of 52.2%, a negative predictive value of 70.6%, and a positive predictive value of 72.5%. Overall, there were no significant differences in startle reactivity across the different unpleasant subcategories for healthy control children, whereas there were significant differences across the subcategories for ADHD children. Controls showed increased or potentiated startle for all three unpleasant subcategories, relative to the

PAGE 75

61 pleasant picture category. In contrast, ADHD children showed potentiated startle for direct threat and injury pictures, but inhibited startle for victim pictures, relative to the pleasant picture category. Moreover, startle response to victim pictures showed good ability to predict group membership. Table 5-3. Dependent measure means across unpleasant picture subcategory Direct Threat Injury Victim Measures Mean S.D. Mean S.D. Mean S.D. Self-Report Valence Control 2.60 1.08 2.27 1.09 2.23 .97 ADHD 2.96 1.32 2.65 1.27 2.59 1.09 Self-Report Arousal Control 5.16 1.50 4.98 1.50 5.20 1.68 ADHD 3.85 1.47 3.83 1.51 4.12 1.41 Startle Magnitude Control 52.71 4.24 51.38 4.83 50.69 4.50 ADHD 53.25 5.23 53.20 6.00 46.89 3.76 Raw Startle Value Control 8.14 5.58 7.50 4.70 7.18 5.23 ADHD 8.13 6.76 7.65 5.70 5.90 4.39 Startle Latency Control 68.48 7.42 68.93 9.36 63.86 8.63 ADHD 63.18 11.91 66.40 19.96 60.93 14.78 Skin Conductance Control .111 .092 .096 .086 .085 .069 ADHD .066 .090 .086 .146 .069 .083 Note. Units of measurement are as follows: startle magnitude T-score, raw startle value V, startle latency ms, and skin conductance log(1 + Siemens). Pleasant Picture Subcategory For startle eyeblink magnitude (Figure 5-3, Table 5-3), repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,110]=7.29, p=.001, power=.93, 2=.12). Paired comparisons revealed reliable differences between attraction and thrill pictures (p<.0001, d=.79) and between attraction and nurture pictures (p=.05, d=.41), but not between nurture and thrill pictures (p=.08), collapsed across groups.

PAGE 76

62 Specifically, children showed greater inhibition of startle in response to attraction pictures compared to thrill and nurture pictures. There was also no reliable Group by Pleasant Picture Subcategory interaction (p=.77) or main effect of Group (p=.61). Overall, healthy control and ADHD children showed a similar degree of startle reactivity to each of the different pleasant picture subcategories. Control 56 ADHD Figure 5-3. Mean startle eyeblink response for ADHD and control groups across pleasant picture subcategory Skin Conductance Response Picture Category For skin conductance (Figure 5-4, Table 5-1), the trend tests revealed the predicted quadratic (i.e., pleasant-unpleasant versus neutral: F[1,55]=18.95, p<0.0001, power=0.99, 2=0.26), as well as a linear (i.e., pleasant versus unpleasant: F[1,55]=14.57, p<0.0001, power=0.96, 2=0.21) trend over Picture Category. Pairwise comparisons (combining ADHD and controls) revealed reliable differences between the unpleasant and neutral picture categories (p<0.0001, d=0.66; unpleasant>neutral), between the unpleasant and pleasant picture categories (p<.0001, d=0.42; unpleasant>pleasant), and between the 44 46 48 50 52 54 + S.D. N urture Attraction T score Thrill Pleasant Picture Subcategory

PAGE 77

63 neutral and pleasant categories (p=.014, d=.31; pleasant>neutral). The trend tests did not reveal a significant interaction of Group with linear or quadratic trend over Picture Category (p=.94, p=.64, respectively), and there was no main effect of Group (p=.17). Overall, there were no group differences, as both groups of children showed greater skin conductance response to unpleasant pictures than to pleasant pictures, and greater skin conductance response to pleasant pictures than to neutral pictures. 0 .02 .04 .06 .08 .1 .12 .14 .16 .18log(1 + uSiemens) + S.D. Unpleasant Neutral PleasantPicture Category ADHD Control Figure 5-4. Mean skin conductance response for control and ADHD groups across picture categories. Table 5-4. Dependent measure means across pleasant picture subcategory Nurture Attraction Thrill Measures Mean S.D. Mean S.D. Mean S.D. Self-Report Valence Control 5.83 .80 4.58 1.09 5.18 .65 ADHD 5.65 1.08 4.71 1.39 5.44 1.29 Self-Report Arousal Control 4.32 1.45 3.49 1.76 4.74 1.33 ADHD 4.08 1.88 3.71 2.00 4.79 1.45 Startle Magnitude Control 48.47 4.19 46.97 2.90 50.49 4.41 ADHD 48.55 4.54 46.98 3.46 49.55 4.14 Raw Startle Value Control 6.07 3.74 5.71 4.08 6.95 4.29

PAGE 78

64 Table 5-4. Continued. Nurture Attraction Thrill Measures Mean S.D. Mean S.D. Mean S.D. ADHD 6.41 4.92 5.47 3.81 6.70 4.87 Startle Latency Control 66.14 10.52 64.40 7.62 64.98 7.63 ADHD 62.97 15.80 61.57 14.57 65.84 17.40 Skin Conductance Control .066 .069 .079 .075 .055 .056 ADHD .054 .071 .054 .090 .031 .047 Note. Units of measurement are as follows: startle magnitude T-score, raw startle value V, startle latency ms, and skin conductance log(1 + Siemens). Unpleasant Picture Subcategory For skin conductance (Figure 5-5, Table 5-2), repeated measures ANOVA did not reveal a significant main effect of Unpleasant Picture Subcategory (p=.56), main effect of group (p=.41), or a Group by Unpleasant Picture Subcategory interaction (p=.24). Overall, there were no reliable group differences within or across different unpleasant subcategories for skin conductance response. 0 .05 .1 .15 .2 .25log(1 + uSiemens) + S.D. Threat Injury Victim PleasantUnpleasant Picture Subcategory ADHD Control Figure 5-5. Mean skin conductance response for control and AHD groups across unpleasant picture subcategory.

PAGE 79

65 Pleasant Picture Subcategory For skin conductance (Figure 5-6, Table 5-3), repeated measures ANOVA did not reveal a reliable main effect of Group (p=.21), a Group by Pleasant Picture Subcategory interaction (p=.85), or a main effect of Pleasant Picture Subcategory (p=.08). Overall, there were no group differences group differences within or across different pleasant subcategories for skin conductance response. 0 .02 .04 .06 .08 .1 .12 .14 .16log(1 + uSiemens) + S.D. Nurture Attraction ThrillPleasant Picture Subcategory ADHD Control Figure 5-6. Mean skin conductance response for control and ADHD groups across pleasant picture subcategory. Self-Report Ratings Picture Category The distribution of mean self-report ratings for ADHD and control children is displayed in Figure 5-7. Children tended to rate higher arousal pictures as more extreme on the valence dimension, as expected. Group differences in self-ratings are discussed below.

PAGE 80

66 1 2 3 4 5 6 7Valence Rating 1 2 3 4 5 6 7Arousal Rating ADHD Controls Figure 5-7. Distribution of self-report valence and arousal affective judgments for 36 picture stimuli by group. For valence ratings (Figure 5-8, Table 5-1), the trend tests revealed the predicted linear (i.e., pleasant versus unpleasant) trend over Picture Category (F[1,54]=185.49, p<0.0001, power=1.00, 2=.78), but not a quadratic effect (p=.23). Pairwise comparisons revealed highly reliable differences between each of the picture categories (all p<0.0001; unpleasantneutral: d=1.34). There was no significant interaction of group with linear (p=.46) or quadratic (p=.63) trend over Picture Category, or reliable main effect of Group (p=.09).

PAGE 81

67 0 1 2 3 4 5 6 7Valence Rating + S.D. Unpleasant Neutral PleasantPicture Category ADHD Control Figure 5-8. Mean self-report valence rating for ADHD and control groups across picture category Overall, both ADHD and control children judged the pleasant pictures as more pleasant than the neutral pictures, and the neutral pictures as more pleasant than the unpleasant pictures. For arousal ratings (Figure 5-9, Table 5-1), the trend tests revealed the predicted quadratic (i.e., unpleasant-pleasant versus neutral) trend over Picture Category (F[1,54]=142.88, p<0.0001, power=1.00, 2=.73), but not a linear effect (p=.12). Pairwise comparisons revealed highly reliable differences between unpleasant and neutral picture categories (p<0.0001, d=1.66; unpleasant>neutral), and between pleasant and neutral picture categories (p<0.0001; d=1.60; pleasant>neutral), but not between pleasant and unpleasant picture categories (p=0.35). For arousal ratings, trend tests also revealed an interaction of Group with linear trend over Picture Category (F[1,31]=8.95, p=0.005, power=0.83, 2=0.22) and an interaction of Group with quadratic trend over Picture Category (F[1,31]=8.95, p=0.005, power=0.83, 2=0.22), but no main effect of Group (p=.48). Each group (ADHD,

PAGE 82

68 control) and category (unpleasant, neutral, pleasant) was then analyzed separately to determine the source of the interactions. 0 1 2 3 4 5 6 7Arousal Rating + S.D. Unpleasant Neutral PleasantPicture Category ADHD Control Figure 5-9. Mean self-report arousal rating for ADHD and control groups across picture category. For the control group, the trend tests revealed a reliable quadratic (i.e., unpleasant-pleasant versus neutral) trend over Picture Category (F[1,21]=118.54, p<0.0001, power=1.00, 2=.85), and a linear (i.e., pleasant versus unpleasant) trend over Picture Category (F[1,54]=12.52, p=.002, power=.92, 2=.37). Pairwise comparisons for the controls revealed reliable differences between the unpleasant and neutral pictures (p<0.0001, d=3.07; unpleasant>neutral), between the pleasant and neutral pictures (p<0.0001, d=.70; pleasant>neutral), and between the unpleasant and pleasant pictures (p=0.002, d=2.44; unpleasant>pleasant). For the ADHD group, the trend tests revealed a reliable quadratic trend over Picture Category (F[1,33]=44.95, p<0.0001, power=1.00, 2=.58), but not a linear trend over Picture Category (p=.39). Pairwise comparisons for the ADHD group revealed reliable differences between the unpleasant and neutral

PAGE 83

69 pictures (p<0.0001, d=1.14; unpleasant>neutral) and between the pleasant and neutral pictures (p<0.0001, d=1.28; pleasant>neutral), but not between the unpleasant and pleasant pictures (p=0.39). Moreover, univariate ANOVAs revealed that the control group rated unpleasant pictures higher on arousal than the ADHD group (F[1,54]=10.18, p=.002, power=.88, 2=.16, d=.87), whereas there were no reliable differences between the groups for pleasant (p=.98) and neutral (p=.12) pictures. Overall, both ADHD and control groups rated the neutral pictures as less arousing than the unpleasant and pleasant pictures. However, control group children rated unpleasant pictures as more highly arousing than pleasant pictures, whereas ADHD children rated unpleasant and pleasant pictures similarly on arousal. Unpleasant Picture subcategory For valence ratings (Figure 5-10, Table 5-2), repeated measures ANOVA revealed a reliable main effect of Unpleasant Picture Subcategory (F[2,108]=3.43, p=.04, power=.63, 2=.06), explained by the victim pictures being judged as more unpleasant than the direct threat pictures (pairwise comparison, p=.005, d=.33), collapsed across groups. In addition, there was no main effect of Group (p=.18) or Group by Unpleasant Picture Subcategory interaction (p=.99). Overall, the both ADHD and control children similarly judged the valence of the different unpleasant picture subcategories. For arousal ratings (Figure 5-11, Table 5-2), repeated measures ANOVA revealed a reliable main effect of Group (F[1,54]=10.18, p=.002, power=.88, 2=.16), explained by the control group judging the unpleasant pictures (across the subcategories) as more arousing than the ADHD group. In addition, there was no main effect of Unpleasant Picture Subcategory (p=.18) or Group by Unpleasant Picture Subcategory interaction (p=.99).

PAGE 84

70 0 1 2 3 4 5 6 7Valence Rating + S.D. Threat Injury Victim PleasantUnpleasant Picture Subcategory ADHD Control Figure 5-10. Mean self-report valence rating for ADHD and control groups across unpleasant picture subcategory. 0 1 2 3 4 5 6 7 8Arousal Rating + S.D. Threat Injury Victim PleasantUnpleasant Picture Subcategory ADHD Control Figure 5-11. Mean self-report arousal rating for ADHD and control groups across unpleasant picture subcategory Overall, while the control group tended to judge all the unpleasant picture subcategories as more arousing than the ADHD group, ADHD children judged all

PAGE 85

71 unpleasant subcategories similarly, and the controls also judged all unpleasant subcategories similarly. Pleasant Picture subcategory For valence ratings (Figure 5-12, Table 5-3), repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,108]=21.36, p<.0001, power=1.00, 2=.28), explained by the nurture pictures being judged as more pleasant than the attraction (p<.0001, d=.95) and thrill (p=.007, d=.37) pictures, and the thrill pictures being judged at more pleasant than attraction pictures (p<.0001, d=.58), collapsed across groups. In addition, there was no main effect of Group (p=.76) or Group by Pleasant Picture Subcategory interaction (p=.42). Overall, both the ADHD and control children similarly judged the valence of the different pleasant picture subcategories. 0 1 2 3 4 5 6 7Valence Rating + S.D. Nurture Attraction ThrillPleasant Picture Subcategory ADHD Control Figure 5-12. Mean self-report valence rating for ADHD and control groups across pleasant picture subcategory.

PAGE 86

72 For arousal ratings (Figure 5-13, Table 5-3), repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,108]=10.94, p<.0001, power=.99, 2=.17), explained by the thrill pictures being judged as more highly arousing than the attraction (p<.0001, d=.70) and nurture (p=.003, d=.39) pictures, and the nurture pictures being judged at more highly arousing than the attraction pictures (p=.05, d=.31), collapsed across groups. In addition, there was no main effect of Group (p=.98) or Group by Pleasant Picture Subcategory interaction (p=.65). Overall, both the ADHD and control children similarly judged the arousal of the different pleasant picture subcategories. 0 1 2 3 4 5 6 7Arousal Rating + S.D. Nurture Attraction ThrillPleasant Picture Subcategory ADHD Control Figure 5-13. Mean self-report arousal rating for ADHD and control groups across pleasant picture subcategory. Raw Startle Eyeblink Response Raw startle eyeblink was examined as an exploratory analysis primarily to investigate whether ADHD children show decreased absolute values of startle reactivity.

PAGE 87

73 While an overall main effect of group across all pictures is most relevant for addressing this, main effect for Picture Category and its interaction with group are reported. Specifically, for Picture Category (Figure 5-14) consistent with the standardized startle magnitude results reported earlier, trend tests revealed a linear (i.e., pleasant versus unpleasant) trend over Picture Category (F[1,55]=28.19, p<.0001, power=1.00, 2=.34), but not a quadratic effect (p=.49). Pairwise comparisons revealed reliable differences between each of the picture categories (unpleasant>neutral: p=.008; unpleasant>pleasant: p<.0001; pleasant
PAGE 88

74 Most relevant to the motivation behind this analysis, there were no group differences for absolute startle magnitude values across the picture categories. Startle Eyeblink Onset Latency Startle eyeblink onset latency was examined as an exploratory analysis. For picture category (Table 5-1), repeated measures ANOVA did not reveal a reliable main effect of Group (p=.19), main effect of Picture Category (p=.26), or a Group by Picture Category interaction (p=.10). For pleasant picture subcategory (Table 5-3), repeated measures ANOVA also did not reveal a reliable main effect of Group (p=.59), main effect of Picture Subcategory (p=.39), or a Group by Picture Subcategory interaction (p=.45). For unpleasant picture subcategory (Table 5-2), repeated measures ANOVA did not reveal a reliable main effect of Group (p=.23) or a Group by Unpleasant Picture Subcategory interaction (p=.76). However, there was a reliable main effect of Unpleasant Picture Subcategory (F[2,106]=3.49, p=.034, power=0.64, 2=0.06), explained by faster startle eyeblinks in response to victim pictures compared to injury (p=.02, d=.36) and direct threat (p=.01, d=.27) pictures. Overall, there were no group differences across the different picture categories or subcategories. Series One Results Summary Table A summary of the first series results is presented in Table 5-5.

PAGE 89

75 Table 5-5. Series one results summary: ADHD versus healthy controls Picture Category Unpleasant Picture Subcategory Pleasant Picture Subcategory Startle Both Groups Unpleasant>Pleasant Controls No difference ADHD Threat, Injury>Victim Victim: Controls>ADHD No Group Differences Both groups AttractionPleas> Neutral No Group Differences No Subcategory Differences No Group Differences Self-Report Valence No Group Differences Both Groups Pleas>Neutral> Unpleas No Group Differences Both Groups VictimThrill>Attraction (Nurture most pleasant) Self-Report Arousal Controls Unpleas>Pleas> Neutral ADHD Unpleas, Pleas>Neutral Unpleasant: Control>ADHD Across All Subcategories Controls>ADHD Both Groups No Subcategory Differences No Group Differences Both Groups Thrill>Nurture>Attraction (Thrill most arousing) Raw Startle No Group Differences Not Analyzed Not Analyzed Startle Latency No Group Differences No Category Differences No Group Differences Both Groups Victim
PAGE 90

CHAPTER 6 SERIES TWO RESULTS: VARIABLES WITHIN ADHD Series Two Hypotheses The hypotheses for the second series of results are presented in Table 6-1. Table 6-1. Series two hypotheses: Dimensional analysis within ADHD Picture Category Unpleasant Picture Subcategory Pleasant Picture Subcategory Startle CD, CPS, ODD, & Impulsivity Associated with Startle Unpleasant (Reduced Potentiation to Unpleasant Pictures) CD, CPS, ODD, & Impulsivity Associated with Startle Across All Subcategories No associations with CD, CPS, ODD, or Impulsivity Skin CD, CPS, ODD, & Impulsivity Associated with Skin Response to Pleasant and Unpleasant Pictures CD, CPS, ODD, & Impulsivity Associated with Skin Across All Subcategories CD, CPS, ODD, & Impulsivity Associated with Skin Across All Subcategories Self-Report Valence No associations with CD, CPS, ODD, or Impulsivity No associations with CD, CPS, ODD, or Impulsivity No associations with CD, CPS, ODD, or Impulsivity Self-Report Arousal No associations with CD, CPS, ODD, or Impulsivity No associations with CD, CPS, ODD, or Impulsivity No associations with CD, CPS, ODD, or Impulsivity Raw Startle No associations with CD, CPS, ODD, or Impulsivity None None Conduct Disorder Symptoms Startle Eyeblink Magnitude For startle eyeblink magnitude, trend tests revealed a linear (i.e., unpleasant > pleasant) trend over Picture Category (F[1,32]=6.37, p=.017, power=.69, 2=.17), but not 76

PAGE 91

77 a quadratic effect (p=.15), consistent with the startle reactivity results from the analysis presented earlier in series one. Pairwise comparisons within the ADHD group were reported earlier. More relevant to the purpose of this analysis, there was a reliable interaction of Conduct Disorder Symptoms with quadratic trend over Picture Category (F[1,32]=7.87, p=.008, power=.78, 2=.20), but no interaction with linear trend (p=.95), and no main effect of Conduct Disorder Symptoms (p=.16). Each picture category (unpleasant, neutral, pleasant) was then analyzed separately to determine the source of the quadratic interaction. Univariate ANOVAs revealed a significant main effect of Conduct Disorder Symptoms for neutral pictures (F[1,32]=5.90, p=.021, power=.65, 2=.16), but not for unpleasant (p=.16) or pleasant (p=.11) pictures. Examining Figure 6-1, higher number of conduct disorder symptoms within the ADHD group was associated with increased startle response to neutral pictures. 44 46 48 50 52 54 56Startle Magnitude -1 0 1 2 3 4 5 6# CD Symptoms Figure 6-1. Neutral picture startle response by number of Conduct Disorder symptoms in ADHD children. Next, repeated measures ANOVA revealed a reliable main effect of Unpleasant Picture Subcategory (F[2,64]=9.17, p<.0001, power=.97, 2=.22). Pairwise comparisons

PAGE 92

78 were reported for ADHD children in the series one analysis section (i.e., victim < injury & direct threat pictures). Again, more relevant, there was also a reliable Conduct Disorder Symptoms by Unpleasant Picture Subcategory interaction (F[2,64]=3.89, p=.026, power=.68, 2=.11). Each unpleasant picture subcategory (direct threat, injury, victim) was then analyzed separately to determine the source of the interaction. Univariate ANOVAs revealed a significant main effect of Conduct Disorder Symptoms for injury pictures (F[1,32]=6.48, p=.016, power=.69, 2=.17), but not for direct threat (p=.24) or victim (p=.15) pictures. Examining Figure 6-2, higher number of conduct disorder symptoms within the ADHD group was associated with decreased startle response to injury pictures. 40 45 50 55 60 65 70 75Startle Magnitude -1 0 1 2 3 4 5 6# CD Symptoms Figure 6-2. Injury picture startle response by number of Conduct Disorder symptoms in ADHD children. Next, repeated measures ANOVA did not reveal a reliable main effect of Pleasant Picture Subcategory (p=.08), or a Conduct Disorder Symptoms by Pleasant Picture Subcategory interaction (p=.28).

PAGE 93

79 Overall, higher number of comorbid Conduct Disorder symptoms in ADHD was associated with increased startle reactivity to neutral pictures and decreased startle reactivity to injury pictures. Skin Conductance Response For skin conductance response, trend tests revealed linear (F[1,32]=5.71, p=.023, power=.64, 2=.15) and quadratic (F[1,32]=15.52, p<.001, power=.97, 2=.33) trends over Picture Category, consistent with the skin conductance results from the series one analysis (i.e., unpleasant > pleasant > neutral pictures). More relevant, there was no reliable interaction of Conduct Disorder Symptoms with linear (p=.50) or quadratic (p=.18) trend over Picture Category, and no main effect of Conduct Disorder Symptoms (p=.37). Next, repeated measures ANOVA revealed a reliable main effect of Unpleasant Picture Subcategory (F[2,64]=4.18, p=.02, power=.72, 2=.12), consistent with the series one results. More relevant, there was a reliable Conduct Disorder Symptoms by Unpleasant Picture Subcategory interaction (F[2,64]=3.37, p=.04, power=.62, 2=.10). Each unpleasant picture subcategory (direct threat, injury, victim) was then analyzed separately to determine the source of the interaction. However, univariate ANOVAs did not reveal significant main effects of Conduct Disorder Symptoms for direct threat (p=.80), injury (p=.13), or victim (p=.30) pictures. Repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,64]=4.23, p=.02, power=.72, 2=.12), consistent with the skin conductance results from the series one analysis. More relevant, there was not a reliable Conduct Disorder Symptoms by Pleasant Picture Subcategory interaction (p=.14).

PAGE 94

80 Overall, number of comorbid Conduct Disorder symptoms within ADHD did not appear to be reliably associated with skin conductance reactivity. Self-Report Ratings For self-report valence ratings, trend tests revealed a linear (F[1,32]=41.24, p<.0001, power=1.00, 2=.56) but not a quadratic (p=.79) trend over Picture Category, consistent with the valence results from the series one analysis (i.e., pleasant > neutral > unpleasant pictures). More relevant, there was no reliable interaction of Conduct Disorder Symptoms with linear (p=.85) or quadratic (p=.45) trend over Picture Category. However, there was a main effect of Conduct Disorder Symptoms (F[1,32]=10.08, p=.003, power=.87, 2=.24) explained by higher number of comorbid CD symptoms being associated with higher valence ratings (i.e., more pleasant) collapsed across picture stimuli (Figure 6-3). 0 1 2 3 4 5 6 7Valence Rating + S.D. -1 0 1 2 3 4 5 6# CD Symptoms Figure 6-3. Self-report valence averaged across all picture stimuli by number of CD symptoms in ADHD. For self-report valence ratings, repeated measures ANOVA did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.52). More relevant, there was

PAGE 95

81 not a reliable Conduct Disorder Symptoms by Unpleasant Picture Subcategory interaction (p=.96). For self-report valence ratings, repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,64]=10.23, p<.0001, power=.98, 2=.24), consistent with the series one analysis (i.e., nurture more pleasant than attraction and thrill pictures). More relevant, there was not a reliable Conduct Disorder Symptoms by Unpleasant Picture Subcategory interaction (p=.083). Overall, children with greater comorbid CD symptomatology tended to rate all pictures as more pleasant, irrespective of picture category or subcategory. For self-report arousal ratings, trend tests revealed a quadratic (F[1,32]=37.17, p<.0001, power=1.00, 2=.56) but not a linear (p=.57) trend over Picture Category, consistent with the arousal results from the series one analysis (i.e., pleasant & unpleasant > neutral pictures). More relevant, there was not a reliable interaction of Conduct Disorder Symptoms with quadratic (p=.072) or linear (p=.12) trend over Picture Category, or a main effect of Conduct Disorder Symptoms (p>.10). For self-report arousal ratings, repeated measures ANOVA did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.77). More relevant, there was not a reliable Conduct Disorder Symptoms by Unpleasant Picture Subcategory interaction (p=.87). Next, for self-report arousal ratings, repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,64]=7.24, p=.001, power=.92, 2=.18), consistent with the series one analysis. More relevant, there was not a reliable Conduct Disorder Symptoms by Unpleasant Picture Subcategory interaction (p=.063).

PAGE 96

82 Overall, number of comorbid Conduct Disorder symptoms within ADHD did not appear to be reliably associated with self-report arousal ratings. Raw Startle Eyeblink Response As discussed earlier, raw startle eyeblink was examined as an exploratory analysis primarily to investigate whether comorbid CD symptoms were associated with decreased absolute values of startle reactivity. Overall, there was no main effect of CD symptoms (p=.61) across Picture Category. Raw startle eyeblink values for different unpleasant and pleasant subcategories were not analyzed. Overall, there was no association between overall absolute startle eyeblink values and the number of comorbid CD symptoms in ADHD. Oppositional Defiant Disorder Symptoms Startle Eyeblink Magnitude For startle eyeblink magnitude, trend tests revealed a linear trend over Picture Category (F[1,32]=9.94, p=.004, power=.86, 2=.24), and a quadratic trend over Picture Category (F[1,32]=6.28, p=.018, power=.68, 2=.16). Again, pairwise comparisons within the ADHD group were reported in the series one results section. More relevant, there was a reliable interaction of Oppositional Defiant Disorder Symptoms with quadratic trend over Picture Category (F[1,32]=14.05, p=.001, power=.95, 2=.31), but not a significant linear interaction (p=.27), and not a significant main effect of Oppositional Defiant Disorder Symptoms (p=.30). Each picture category (unpleasant, neutral, pleasant) was then analyzed separately to determine the source of the quadratic interaction. Univariate ANOVAs revealed a significant main effect of Oppositional Defiant Disorder Symptoms for neutral pictures (F[1,32]=11.88, p=.002, power=.92, 2=.27) and unpleasant pictures (F[1,32]=8.29, p=.007, power=.80, 2=.21), but not for

PAGE 97

83 pleasant pictures (p=.44). Examining Figures 6-4 and 6-5, higher number of Oppositional Defiant Disorder symptoms within the ADHD group is associated with increased startle response to neutral pictures but decreased startle response to unpleasant pictures. 46 48 50 52 54 56 58Startle Magnitude -1 0 1 2 3 4 5 6 7 8 9# ODD Symptoms Figure 6-4. Unpleasant picture startle response by number of Oppositional Defiant Disorder symptoms in ADHD. 44 46 48 50 52 54 56Startle Magnitude -1 0 1 2 3 4 5 6 7 8 9# ODD Symptoms Figure 6-5. Neutral picture startle response by number of Oppositional Defiant Disorder symptoms in ADHD.

PAGE 98

84 Next, repeated measures ANOVA revealed a reliable main effect of Unpleasant Picture Subcategory (F[2,64]=6.94, p=.002, power=.91, 2=.18). However, there was not a reliable ODD Symptoms by Unpleasant Picture Subcategory interaction (p=.45), and therefore each unpleasant picture subcategory was not analyzed separately. In addition, repeated measures ANOVA did not reveal a reliable main effect of Pleasant Picture Subcategory (p=.34), or an ODD Symptoms by Pleasant Picture Subcategory interaction (p=.95). Overall, higher number of comorbid ODD symptoms in ADHD was associated with increased startle reactivity to neutral pictures and decreased reactivity to unpleasant pictures. Skin Conductance Response For skin conductance response, trend tests revealed a linear trend over Picture Category (F[1,32]=8.89, p=.005, power=.82, 2=.22), and a quadratic trend over Picture Category (F[1,32]=24.34, p<.0001, power=1.00, 2=.43), consistent with previous analyses. More relevant, there was a reliable interaction of Oppositional Defiant Disorder Symptoms with quadratic trend over Picture Category (F[1,32]=7.53, p=.01, power=.76, 2=.19), but not a significant linear interaction (p=.12). Notably, there was also a significant main effect of ODD Symptoms (F[1,32]=6.64, p=.015, power=.71, 2=.17). Therefore, the relationship between ODD Symptoms and skin conductance was examined by collapsing across Picture Category. Examining Figure 6-6, higher number of ODD symptoms within the ADHD group was associated with overall decreased skin conductance response.

PAGE 99

85 -.05 .05 .15 .25 .35 .45log (1 + uSiemens) + S.D. -1 0 1 2 3 4 5 6 7 8 9# ODD Symptoms Figure 6-6. Skin conductance response collapsed across picture categories by number of ODD symptoms in ADHD. Moreover, each picture category (unpleasant, neutral, pleasant) was analyzed separately to determine the source of the quadratic interaction. Univariate ANOVAs revealed a significant main effect of Oppositional Defiant Disorder Symptoms for unpleasant pictures (F[1,32]=6.80, p=.013, power=.72, 2=.18) and pleasant pictures (F[1,32]=6.05, p=.02, power=.66, 2=.16), but only a trend level effect for neutral pictures (p=.09). Examining Figures 6-7 and 6-8, higher number of oppositional defiant disorder symptoms within the ADHD group was associated with decreased skin conductance response to both unpleasant and pleasant pictures. Next, repeated measures ANOVA did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.07) or an ODD Symptoms by Unpleasant Picture Subcategory interaction (p=.18). In addition, repeated measures ANOVA revealed a main effect of Pleasant Picture Subcategory (F[2,64]=3.55, p=.035, power=.64, 2=.10), consistent with the series one analysis. However, there was not a reliable ODD Symptoms by Pleasant Picture Subcategory interaction (p=.24).

PAGE 100

86 -.05 .05 .15 .25 .35 .45log (1 + uSiemens) + S.D. -1 0 1 2 3 4 5 6 7 8 9# ODD Symptoms Figure 6-7. Unpleasant picture skin conductance response by number of ODD symptoms in ADHD. -.05 .05 .15 .25 .35 .45log (1 + uSiemens) + S.D. -1 0 1 2 3 4 5 6 7 8 9# ODD Symptoms Figure 6-8. Pleasant picture skin conductance response by number of ODD symptoms in ADHD. Overall, higher number of comorbid ODD symptoms in ADHD was associated with decreased skin conductance reactivity to picture stimuli in general, but with the greater reduction in response to unpleasant and pleasant pictures, compared neutral pictures. There was no evidence of comorbid ODD symptoms influencing direct threat,

PAGE 101

87 injury, and victim pictures differently, or influencing nurture, attraction, and thrill pictures differently. Self-Report Ratings For self-report valence ratings, trend tests revealed a linear (F[1,32]=30.03, p<.0001, power=1.00, 2=.48) but not a quadratic (p=.48) trend over Picture Category, consistent with the valence results from the series one analysis (i.e., pleasant > neutral > unpleasant pictures). More relevant, there was not a reliable interaction of Oppositional Defiant Disorder symptoms with quadratic (p=.09) or linear (p>.10) trend over Picture Categor, or a main effect of ODD symptoms (p=.37). For self-report valence ratings, repeated measures ANOVA did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.52). More relevant, there was not a reliable ODD Symptoms by Unpleasant Picture Subcategory interaction (p=.93), similar to the CD symptoms analysis. For self-report valence ratings, repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,64]=6.06, p=.004, power=.87, 2=.16), consistent with the series one analysis (i.e., nurture more pleasant than attraction and thrill pictures). More relevant, there was not a reliable ODD Symptoms by Unpleasant Picture Subcategory interaction (p=.22). Overall, comorbid ODD symptomatology did not impact valence ratings across picture categories or subcategories. For self-report arousal ratings, trend tests revealed a quadratic (F[1,32]=43.17, p<.0001, power=1.00, 2=.57) but not a linear (p=.33) trend over Picture Category, consistent with the arousal results from the series one analysis (i.e., pleasant & unpleasant > neutral pictures). More relevant, there was a reliable interaction of ODD Symptoms

PAGE 102

88 with quadratic trend over Picture Category (F[1,32]=7.88, p=.008, power=.78, 2=.20), but not a relieable linear interaction (p=.07) or main effect of ODD Symptoms (p=1.00). 0 1 2 3 4 5 6 7Arousal Rating + S.D. -1 0 1 2 3 4 5 6 7 8 9# ODD Symptoms Figure 6-9. Self-report unpleasant picture arousal by number of ODD symptoms in ADHD children. Each picture category was analyzed separately to determine the source of the interaction. Univariate ANOVAs revealed a significant main effect of ODD Symptoms for unpleasant pictures (Figure 6-9; F[1,32]=5.52, p=.025, power=.63, 2=.15), explained by unpleasant pictures being rated as lower on arousal with increasing number of ODD symptoms. There was no such relationship with neutral (p>.10) or pleasant (p=.88) pictures. For self-report arousal ratings, repeated measures ANOVA did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.91). More relevant, there was not a reliable ODD Symptoms by Unpleasant Picture Subcategory interaction (p=.26), similar to the CD symptoms analysis. Next, for self-report arousal ratings, repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,64]=5.54, p=.006, power=.84,

PAGE 103

89 2=.15), consistent with the series one analysis (i.e., thrill more arousing than attraction and nurture pictures). More relevant, there was not a reliable ODD Symptoms by Unpleasant Picture Subcategory interaction (p=.21). Overall, children with greater comorbid ODD symptomatology tended to rate unpleasant pictures as less arousing than children with fewer comorbid ODD symptoms. Raw Startle Eyeblink Response Overall, there was no main effect of CD symptoms (p=.94) across Picture Category. Raw startle eyeblink values for different unpleasant and pleasant subcategories were not analyzed. Overall, there was no association between overall absolute startle eyeblink values and the number of comorbid ODD symptoms in ADHD. Psychopathy Symptoms Startle Eyeblink Magnitude For startle eyeblink magnitude, trend tests revealed a significant quadratic trend over Picture Category (F[1,32]=4.58, p=.04, power=.55, 2=.13), but not a linear effect (p=.16). More relevant to the aim of this analysis, there was a reliable interaction of Psychopathy Score with quadratic trend over Picture Category (F[1,32]=6.58, p=.015, power=.70, 2=.17), but not a significant linear effect (p=.85). Each picture category (unpleasant, neutral, pleasant) was then analyzed separately to determine the source of the quadratic interaction. Univariate ANOVAs revealed a significant main effect of Psychopathy Score for neutral pictures (F[1,32]=4.860, p=.035, power=.57, 2=.13), but not for unpleasant (p=.12) or pleasant (p=.20) pictures. Examining Figure 6-10, within the ADHD group, similar to CD and ODD symptoms, higher total psychopathy score was associated with increased startle response to neutral pictures. Notable, total psychopathy score was strongly correlated to both

PAGE 104

90 number of CD symptoms (r=.76, p<.0001) and number of ODD symptoms (r=.72, p<.0001), and number of CD symptoms was strongly correlated to number of ODD symptoms (r=.63, p<.0001). 44 46 48 50 52 54 56Startle Magnitude 5 10 15 20 25 30 35 40 45 50Total Psychopathy Score Figure 6-10. Neutral picture startle response by total psychopathy score in ADHD. Next, repeated measures ANOVA revealed a reliable main effect of Unpleasant Picture Subcategory (F[2,64]=6.42, p=.003, power=.89, 2=.17). More relevant, there was also a reliable Total Psychopathy Score by Unpleasant Picture Subcategory interaction (F[2,64]=4.31, p=.018, power=.73, 2=.12). Each unpleasant picture subcategory (direct threat, injury, victim) was then analyzed separately to determine the source of the interaction. Similar to CD symptoms, univariate ANOVAs revealed a significant main effect of Total Psychopathy Score for injury pictures (F[1,32]=8.69, p=.006, power=.82, 2=.21), but not for direct threat (p=.32) or victim (p=.49) pictures. Examining Figure 6-11, similar to CD symptoms, increased psychopathy symptomatology within ADHD was associated with decreased startle response to injury pictures.

PAGE 105

91 40 45 50 55 60 65 70 75Startle Magnitude 5 10 15 20 25 30 35 40 45 50Total Psychopathy Score Figure 6-11. Injury picture startle response by total psychopathy score in ADHD. Next, repeated measures ANOVA did not reveal a reliable main effect of Pleasant Picture Subcategory (p=.08), or a Total Psychopathy Score by Pleasant Picture Subcategory interaction (p=.28). Overall, similar to CD and ODD symptoms, greater psychopathy symptomatology in ADHD was associated with increased startle reactivity to neutral pictures, and similar to CD symptoms, greater psychopathy was associated with decreased startle reactivity to injury pictures. Skin Conductance Response For skin conductance response, trend tests revealed a reliable quadratic trend over Picture Category(F[1,32]=12.56, p=.001, power=.93, 2=.28), but not a reliable linear interaction (p=.09). More relevant, there was a reliable interaction of Psychopathy Score with quadratic trend over Picture Category (F[1,32]=4.79, p=.036, power=.57, 2=.13), but no significant linear interaction (p=.12) or main effect of Psychopathy Score (p=.12). Each picture category (unpleasant, neutral, pleasant) was then analyzed separately to determine the source of the quadratic interaction. However, univariate ANOVAs did not

PAGE 106

92 reveal reliable main effects of Psychopathy Score for unpleasant (p=.11), neutral (p=.43), or pleasant (p=.10) pictures. Next, repeated measures ANOVA revealed a reliable main effect of Unpleasant Picture Subcategory (F[2,64]=3.20, p=.048, power=.59, 2=.09). As reported earlier under the CD symptom section, pairwise comparisons did not reveal reliable differences in skin conductance between the different unpleasant subcategories (all p>.26), consistent with the series one results. More relevant, there was not a reliable Psychopathy Score by Unpleasant Picture Subcategory interaction (p=.10). Repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,64]=4.22, p=.02, power=.72, 2=.12), consistent with the skin conductance results from the CD symptoms analysis as well as the series one analysis. More relevant, there was not a reliable Psychopathy Score by Pleasant Picture Subcategory interaction (p=.08). Overall, comorbid psychopathy symptomatology within ADHD did not appear to be reliably associated with skin conductance reactivity, similar to CD symptomatology. Self-Report Ratings For self-report valence ratings, trend tests revealed a linear (F[1,32]=7.93, p=.008, power=.78, 2=.20) but not a quadratic (p=.67) trend over Picture Category, consistent with the valence results from the series one analysis (i.e., pleasant > neutral > unpleasant pictures). More relevant, there was no reliable interaction of Psychopathy Score with linear (p=.82) or quadratic (p=.39) trend over Picture Category, similar to CD symptoms. However, similar to CD symptoms, there was a main effect of Psychopathy Score (Figure 6-12; F[1,32]=4.54, p=.041, power=.54, 2=.12) explained by higher psychopathy score

PAGE 107

93 being associated with higher valence ratings (i.e., more pleasant) collapsed across picture stimuli. 0 1 2 3 4 5 6 7Valence Rating + S.D. 0 5 10 15 20 25 30 35 40 45 50Total CPS Figure 6-12. Self-report valence averaged across all picture stimuli by total psychopathy score in ADHD. For self-report valence ratings, repeated measures ANOVA did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.65). More relevant, there was not a reliable Psychopathy Score by Unpleasant Picture Subcategory interaction (p=.32), similar to the CD and ODD symptoms analyses. For self-report valence ratings, repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,64]=5.85, p=.005, power=.86, 2=.16), consistent with the series one analysis (i.e., nurture more pleasant than attraction and thrill pictures). More relevant, similar to CD symptoms, there was no reliable Psychopathy Score by Unpleasant Picture Subcategory interaction (p=.094). Overall, children with greater comorbid psychopathy symptomatology tended to rate all pictures as more pleasant, irrespective of picture category or subcategory.

PAGE 108

94 For self-report arousal ratings, trend tests revealed a quadratic (F[1,32]=11.30, p=.002, power=.90, 2=.26) but not a linear (p=.37) trend over Picture Category, consistent with the arousal results from the series one analysis (i.e., pleasant & unpleasant more arousing than neutral pictures). More relevant, there was not a reliable interaction of Psychopathy Score with linear (p=22) or quadratic (p=.24) trend over Picture Category or a reliable main effect of Psychopathy Score (p=.12). For self-report arousal ratings, repeated measures ANOVA did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.74). More relevant, there was not a reliable Psychopathy Score by Unpleasant Picture Subcategory interaction (p=.41). Next, for self-report arousal ratings, repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,64]=7.24, p=.001, power=.92, 2=.18), consistent with the series one analysis (i.e., thrill more arousing than attraction and nurture pictures). More relevant, there was a reliable Psychopathy Score by Pleasant Picture Subcategory interaction (F[2,64]=3.07, p=.05, power=.57, 2=.09), explained by attraction pictures being rated as higher on arousal with increasing psychopathy score (Figure 6-13; F[1,32]=6.67, p=.015, power=.71, 2=.17). There was no such relationship with nurture (p=.53) or thrill (p=.86) pictures.

PAGE 109

95 0 1 2 3 4 5 6 7Arousal Rating + S.D. 0 5 10 15 20 25 30 35 40 45 50Total CPS Figure 6-13. Self-report attraction picture arousal by psychopathy score in ADHD children. Overall, ADHD children with higher psychopathy scores tended to rate attraction pictures as more arousing than ADHD children with lower psychopathy scores. There were no other reliable relationships between psychopathy and ratings of arousal for other picture categories or subcategories. Raw Startle Eyeblink Response Overall, there was no main effect of Psychopathy Score (p=.94) across Picture Category. Raw startle eyeblink values for different unpleasant and pleasant subcategories were not analyzed. Overall, there was no association between overall absolute startle eyeblink values and total psychopathy score in ADHD. Impulsivity Startle Eyeblink Magnitude Examining Connors’ Continuous Performance Test (CPT) Beta Score, for startle reactivity, trend tests did not reveal a linear (p=.30) or quadratic (p=.98) trend over Picture Category. More relevant, there was no reliable interaction of Beta with linear

PAGE 110

96 (p=.80) or quadratic (p=.95) trend over Picture Category, and no main effect of Beta (p=.43). In addition, repeated measures ANOVAs did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.36), Beta by Unpleasant Picture Subcategory interaction (p=.84), main effect of Pleasant Picture Subcategory (p=.67), or Beta by Pleasant Picture Subcategory interaction (p=.60). In an exploratory analysis, there was similarly no interaction of startle reactivity with Connors’ CPT commission errors across Picture Category or across Picture Subcategories (all p>.45). In addition, on parent ratings of impulsivity (i.e., Connors’ Restless-Impulsive Index), trend tests again did not reveal a linear (p=.49) or quadratic (p=.94) trend over Picture Category, or a reliable interaction of Restless-Impulsive Index score with linear (p=.92) or quadratic (p=.96) trend over Picture Category. In addition, repeated measures ANOVAs did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.25), Restless-Impulsive Index score by Unpleasant Picture Subcategory interaction (p=.54), main effect of Pleasant Picture Subcategory (p=.99), or Restless-Impulsive Index score by Pleasant Picture Subcategory interaction (p=.94). Moreover, in exploratory analyses of parent report of ADHD symptomatology, there was similarly no reliable interactions of startle reactivity with Connors’ DSM-IV Inattentive scale, DSM-IV Hyperactive-Impulsive scale, DSM-IV Total scale, or ADHD subtype (i.e., Inattentive versus Combined) across Picture Category or Subcategories (all interactions p>.17). Overall, within the ADHD group, there were no reliable relationships between startle reactivity and parent report impulsivity or direct measures of impulsivity. Moreover, exploratory analyses did not reveal relationships between startle reactivity and

PAGE 111

97 severity of DSM-IV inattention, DSM-IV hyperactivity-impulsivity, DSM-IV total, or ADHD subtype. Skin Conductance Response Examining Connors’ Continuous Performance Test (CPT) Beta Score, for skin conductance, trend tests did not reveal a linear (p=.44) or quadratic (p=.81) trend over Picture Category. More relevant, there was no reliable interaction of Beta with linear (p=.81) or quadratic (p=.26) trend over Picture Category, and no significant main effect of Beta (p=.20). Repeated measures ANOVA revealed a reliable main effect of Unpleasant Picture Subcategory (F[2,64]=3.46, p=.038, power=.63, 2=.10), and a Beta by Unpleasant Picture Subcategory interaction (F[2,64]=4.45, p=.016, power=.74, 2=.13). Each unpleasant subcategory was examined separately to determine the source of the interaction. However, no reliable association between Beta and skin conductance response to direct threat (p=.60), injury (p=.12), or victim (p=.30) pictures emerged. Repeated measures ANOVA also revealed a reliable main effect of Pleasant Picture Subcategory (F[2,64]=3.35, p=.041, power=.61, 2=.10), and a Beta by Pleasant Picture Subcategory interaction (F[2,64]=4.77, p=.012, power=.78, 2=.13). Each pleasant subcategory was examined separately to determine the source of the interaction. Notable, a reliable association between Beta and skin conductance response to attraction pictures emerged (Figure 6-14; F[1,31]=5.65, p=.024, power=.63, 2=.15), explained by lower Beta scores (greater risk-taking response style) being associated with lower skin conductance response to attraction pictures. There was no such relationship with nurture (p=.11) or thrill (p=.91) pictures. In an exploratory analysis, there was no interaction of

PAGE 112

98 skin conductance response with Connors’ CPT commission errors across Picture Category or across Picture Subcategories (all p>.51). -.05 0 .05 .1 .15 .2 .25 .3 .35 .4log (1 + uSiemens)+ S.D. 30 40 50 60 70 80 90Response Style B T Figure 6-14. Skin conductance response to attraction pictures by Beta score in ADHD children. In addition, on parent ratings of impulsivity (i.e., Connors’ Restless-Impulsive Index), for skin conductance, trend tests did not reveal a reliable quadratic (p=.06) or linear (p=.94) trend over Picture Category. More relevant, there was no reliable interaction of Restless-Impulsive Index score with linear (p=.96) or quadratic (p=.20) trend over Picture Category. In addition, repeated measures ANOVAs did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.09), a Restless-Impulsive Index score by Unpleasant Picture Subcategory interaction (p=.13), a main effect of Pleasant Picture Subcategory (p=.29), or a Restless-Impulsive Index score by Pleasant Picture Subcategory interaction (p=.45). Moreover, in exploratory analyses of parent report of ADHD symptomatology, there was similarly no interaction of skin conductance response with Connors’ DSM-IV Inattentive symptoms, DSM-IV Hyperactive-Impulsive symptoms, or DSM-IV Total

PAGE 113

99 symptoms across Picture Category or Pleasant Picture Subcategory (all interactions p>.43). In addition, for Unpleasant Picture Subcategory, there was a significant interaction with DSM-IV Inattentive symptoms (F[2,64]=3.15, p=.049, power=.59, 2=.09) and DSM-IV Total symptoms (F[2,64]=3.05, p=.054, power=.57, 2=.09), but not with DSM-IV Hyperactive-Impulsive symptoms (p=.24). Each unpleasant picture subcategory (direct threat, injury, victim) was then analyzed separately for DSM-IV Inattentive and Total symptoms determine the source of the interactions. However, univariate ANOVAs did not reveal any reliable main effects of DSM-IV Inattentive or Total symptoms for direct threat, injury, or victim pictures (all p>.21). In an additional exploratory analysis, there was no reliable interaction of ADHD subtype (i.e., Inattentive versus Combined) with linear (p=.64) or quadratic (p=.17) trend over Picture Category, or ADHD subtype by Unpleasant ( p=.43) or Pleasant (p=.45) Picture Subcategory. Overall, within the ADHD group, there were no reliable relationships between skin conductance response and parent report impulsivity or direct measures of impulsivity, with the exception of lower Beta scores being associated with smaller skin conductance response to attraction pictures. Moreover, exploratory analyses did not reveal reliable relationships between skin conductance and severity of DSM-IV inattention, DSM-IV hyperactivity-impulsivity, DSM-IV total, or ADHD subtype. Self-Report Ratings Examining Connors’ Continuous Performance Test (CPT) Beta Score, for self-report valence ratings, trend tests revealed a linear (F[1,32]=4.56, p=.04, power=.54, 2=.13) but not a quadratic (p=.94) trend over Picture Category, consistent with the valence results from the series one analysis (i.e., pleasant > neutral > unpleasant pictures). More relevant, there was no reliable interaction of Beta with linear (p=.79) or

PAGE 114

100 quadratic (p=.82) trend over Picture Category, and no main effect of Beta (p=.72). In addition, repeated measures ANOVAs did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.92), Beta by Unpleasant Picture Subcategory interaction (p=.89), main effect of Pleasant Picture Subcategory (p=.29), or Beta by Pleasant Picture Subcategory interaction (p=.78). In an exploratory analysis, there was similarly no interaction of startle reactivity with Connors’ CPT commission errors across Picture Category or across Picture Subcategories (all p>.27). In addition, on parent ratings of impulsivity (i.e., Connors’ Restless-Impulsive Index), for self-report valence ratings, trend tests did not reveal a linear (p=.28) or quadratic (p=.75) trend over Picture Category, or a reliable interaction of Restless-Impulsive Index score with linear (p=.27) or quadratic (p=.75) trend over Picture Category. Furthermore, repeated measures ANOVAs did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.80), Restless-Impulsive Index score by Unpleasant Picture Subcategory interaction (p=.57), main effect of Pleasant Picture Subcategory (p=.82), or Restless-Impulsive Index score by Pleasant Picture Subcategory interaction (p=.86). Moreover, in exploratory analyses of parent report of ADHD symptomatology, there was similarly no interactions of self-report valence ratings with Connors’ DSM-IV Inattentive scale, DSM-IV Hyperactive-Impulsive scale, or DSM-IV Total scale across Picture Category or Subcategories (all interactions p>.23). In an additional exploratory analysis, there was no reliable interaction of ADHD subtype (i.e., Inattentive versus Combined) with linear (p=.81) or quadratic (p=.97) trend over Picture Category, or ADHD subtype by Unpleasant ( p=.80) or Pleasant (p=.69) Picture Subcategory.

PAGE 115

101 Overall, within the ADHD group, there were no reliable relationships between self-report valence ratings and parent report impulsivity or direct measures of impulsivity. Moreover, exploratory analyses did not reveal relationships between valence ratings and severity of DSM-IV inattention, DSM-IV hyperactivity-impulsivity, DSM-IV total, or ADHD subtype. Examining Connors’ Continuous Performance Test (CPT) Beta Score, for self-report arousal ratings, trend tests did not reveal a linear (p=.68) or a quadratic (p=.49) trend over Picture Category. More relevant, there was no reliable interaction of Beta with linear (p=.78) or quadratic (p=.55) trend over Picture Category, and no main effect of Beta (p=.42). In addition, repeated measures ANOVAs did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.53), Beta by Unpleasant Picture Subcategory interaction (p=.75), main effect of Pleasant Picture Subcategory (p=.28), or Beta by Pleasant Picture Subcategory interaction (p=.66). In an exploratory analysis, there was similarly no interaction of arousal ratings with Connors’ CPT commission errors across Picture Category or across Picture Subcategories (all p> .19). In addition, on parent ratings of impulsivity (i.e., Connors’ Restless-Impulsive Index), for self-report arousal ratings, trend tests did not reveal a linear (p=.57) or quadratic (p=.22) trend over Picture Category, or a reliable interaction of Restless-Impulsive Index score with linear (p=.37) or quadratic (p=.29) trend over Picture Category. In addition, repeated measures ANOVAs did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.68), Restless-Impulsive Index score by Unpleasant Picture Subcategory interaction (p=.50), main effect of Pleasant Picture

PAGE 116

102 Subcategory (p=.29), or Restless-Impulsive Index score by Pleasant Picture Subcategory interaction (p=.53). Moreover, in exploratory analyses of parent report of ADHD symptomatology, there was similarly no interactions of self-report arousal ratings with Connors’ DSM-IV Inattentive scale, DSM-IV Hyperactive-Impulsive scale, or DSM-IV Total scale across Picture Category or Subcategories (all interactions p> .19). In an additional exploratory analysis, there was no reliable interaction of ADHD subtype (i.e., Inattentive versus Combined) with linear (p=.09) or quadratic (p=.17) trend over Picture Category, or ADHD subtype by Unpleasant ( p=.59) or Pleasant (p=.55) Picture Subcategory. Overall, within the ADHD group, there were no reliable relationships between self-report arousal ratings and parent report impulsivity or direct measures of impulsivity. Furthermore, exploratory analyses did not reveal any reliable relationships between arousal ratings and severity of DSM-IV inattention, DSM-IV hyperactivity-impulsivity, DSM-IV total, or ADHD subtype. Raw Startle Eyeblink Response Overall, there was no main effect of CPT Beta score (p=.79) or Connors’ Restless-Impulsive Index (p=.10) across Picture Category. Raw startle eyeblink values for different unpleasant and pleasant subcategories were not analyzed. Overall, there was no association between overall absolute startle eyeblink values and parent report or direct measures of impulsivity in ADHD. Moreover, there were also no main effects of Connors’ CPT commission errors, Connors’ DSM-IV Inattentive scale, Connors’ DSM-IV Hyperactive-Impulsive scale, Connors’ DSM-IV Total, or ADHD subtype (all p>.36).

PAGE 117

103 Series Two Results Summary Table A summary of the series two results is presented in Table 6-2. Table 6-2. Series two results summary: Dimensional analysis within ADHD Picture Category Unpleasant Picture Subcategory Pleasant Picture Subcategory Startle CD: Startle Neutral CPS: Startle Neutral ODD: Startle Neutral ODD: Startle Unpleasant No Impulsivity association CD: Startle Injury CPS: Startle Injury No associations with ODD No Impulsivity association No associations with CD No associations with Psychopathy (CPS) No associations with ODD No Impulsivity association Skin No associations with CD No associations with CPS ODD: Skin Unpleasant ODD: Skin Pleasant ODD: Skin Neutral: trend No Impulsivity association No associations with CD No associations with CPS No associations with ODD No Impulsivity association No associations with CD No associations with CPS No associations with ODD Beta: Skin Attraction Self-Report Valence CD: Overall Valence (more pleasant) CPS: Overall Valence No associations with ODD No Impulsivity association No associations with CD No associations with CPS No associations with ODD No Impulsivity association No associations with CD No associations with CPS No associations with ODD No Impulsivity association Self-Report Arousal No associations with CD No associations with CPS ODD: Arous. Unpleasant No Impulsivity association No associations with CD No associations with CPS No associations with ODD No Impulsivity association No associations with CD CPS: Arousal Attraction No associations with ODD No Impulsivity association

PAGE 118

104 Table 6-2. Continued. Picture Category Unpleasant Picture Subcategory Pleasant Picture Subcategory Raw Startle No associations with CD No associations with CPS No associations with ODD No Impulsivity association Not Analyzed Not Analyzed

PAGE 119

CHAPTER 7 SERIES THREE RESULTS: ADHD, ADHD/ODD, ADHD/CD The following series of analysis were completed primarily to aid in the interpretation of the dimensional statistical approach (i.e., series two results). While connected to the series two hypotheses, these analyses are exploratory and therefore do not have specific hypotheses of their own. These follow-up analyses were completed examining emotional reactivity across discrete groups within ADHD, including 13 children with only ADHD, 10 ADHD children with only comorbid ODD, and 11 ADHD children with comorbid CD (8 of these 11 also met criteria for ODD). The approach to these analyses was identical to the series one analysis. However, unlike previous analyses, trend-level findings were explored and reported, and therefore should be viewed with some degree of caution. Startle Eyeblink Magnitude Picture Category For startle eyeblink magnitude (Figure 7-1), the trend tests revealed a linear (i.e., pleasant versus unpleasant) trend over Picture Category (F[1,31]=13.44, p=.001, power=.9, 2=.30), but not a quadratic effect (p=.31), consistent with the series one analysis. Pairwise comparisons within the ADHD group were reported earlier in the series one analysis (i.e., unpleasant > pleasant pictures, & neutral > pleasant pictures). More relevant, there was a reliable interaction of Group (i.e., ADHD, ADHD/ODD, ADHD/CD) with quadratic trend over Picture Category (F[2,31]=3.49, p=.04, power=.61, 2=.18), but no linear interaction (p=.96) or main effect of group (p=.23). Each group 105

PAGE 120

106 and picture category was examined separately to examine the source of the quadratic interaction. 45 47 49 51 53 55T-score + S.D. Unpleasant Neutral PleasantPicture Category ADHD + CD ADHD + ODD ADHD Figure 7-1. Mean startle eyeblink response for ADHD, ADHD+ODD, and ADHD+CD groups across picture category. Examining each group separately, for the ADHD group, trend tests revealed a reliable linear trend over Picture Category (F[1,12]=7.83, p=.016, power=.73, 2=.40), but not a quadratic effect (p=.27). Pairwise comparisons for the ADHD group revealed reliable differences between unpleasant and pleasant pictures (p=.016, d=1.03; unpleasant>pleasant), and between unpleasant and neutral pictures (p=.05, d=1.27; unpleasant>neutral), but not between pleasant and neutral pictures (p=.82). For the ADHD/CD group, trend tests revealed both a reliable linear trend over Picture Category (F[1,10]=4.92, p=.05, power=.52, 2=.33) and a quadratic trend over Picture Category (F[1,10]=6.85, p=.03, power=.66, 2=.41). Pairwise comparisons for the CD group revealed reliable differences between unpleasant and pleasant pictures (p=.05, d=1.22 ; unpleasant>pleasant), and between neutral and pleasant pictures (p=.001; d=1.79; neutral>pleasant), but not between neutral and unpleasant pictures (p=.41). For the

PAGE 121

107 ADHD/ODD group, trend tests did not reveal a reliable linear (p=.15) or quadratic (p=.48) trend over Picture Category. However, this was likely due to lack of statistical power as the effect size approached a 1 standard deviation difference between unpleasant and pleasant pictures for the ADHD/ODD group (p=.15, d=.96; unpleasant>pleasant). Examining each picture category separately, there were no reliable differences between the groups for unpleasant (p=.24) and pleasant (.46) picture categories. However, univariate ANOVA revealed a trend-level main effect of Group for the neutral pictures (F[2,31]=2.55, p=.09, power=.46). While only trend-level, this finding was further examined to better understand the previously reported association between increased startle response to neutral pictures and increased CD, ODD, and psychopathy symptoms. Unpaired 2-sided t-tests showed a reliable difference between the ADHD and ADHD + CD groups (p=.047, d=.87), but not between ADHD and ADHD/ODD (p=.28) or ADHD/CD and ADHD/ODD (p=.26). Specifically, ADHD children with comorbid Conduct Disorder showed greater startle eyeblink response to neutral pictures than ADHD children without comorbid Conduct or Oppositional Defiant Disorders. Overall, ADHD, ADHD/ODD (based on effect size), and ADHD/CD children showed and a linear “adult-like” pattern of affective modulation of the startle reflex, that is larger startle eyeblink reactivity to unpleasant pictures relative to unpleasant pictures. Notable, similar to the series two analyses, ADHD children with comorbid CD showed greater startle eyeblink response to neutral pictures (i.e., potentiated startle relative to pleasant pictures) than ADHD children without comorbid disorders. To explore this finding, a logistic regression was performed to inspect how well startle response to neutral pictures predicted ADHD versus ADHD/CD group membership (i.e., excluded

PAGE 122

108 ADHD/ODD group). The logistic regression revealed that startle response to neutral pictures was a moderate predictor of group membership (Wald=3.45, p=.06; Likelihood Ratio Test=4.29, p=.04), and resulted in 62.50 % of the children being correctly classified, 69.2% or 9/13 ADHD children correctly classified, 54.6% or 6/11 ADHD/CD children correctly classified, a predictive value of 64.3% for ADHD, and a predictive value of 60.0% for ADHD/CD. Notable, when the ADHD/ODD children were included in the model, 0.0% of ADHD/ODD were correctly classified (6 were incorrectly classified as ADHD, 4 were incorrectly classified as ADHD/CD). Unpleasant Picture Subcategory For startle eyeblink magnitude (Figure 7-2), repeated measures ANOVA revealed a reliable main effect of Unpleasant Picture Subcategory (F[2,62]=16.29, p<.0001, power=1.00, 2=.34), consistent with the series one analysis. Pairwise comparisons within the ADHD group were reported earlier in the series one analysis i.e., threat & injury > victim pictures). More relevant, there was a reliable interaction of Group by Unpleasant Picture Subcategory interaction (F[4,62]=2.47, p=.05, power=.67, 2=.14). Each group and each unpleasant subcategory was then analyzed separately to determine the source of the interaction.

PAGE 123

109 42 46 50 54 58 62 66T-score + S.D. Threat Injury Victim PleasantUnpleasant Picture Subcategory ADHD + CD ADHD + ODD ADHD Figure 7-2. Mean startle eyeblink response for ADHD, ADHD+ODD, and ADHD+CD groups across unpleasant picture subcategory. Examining each group separately, for the ADHD group, repeated measures ANOVA revealed a main effect of Unpleasant Picture Subategory (F[2,24]=6.11, p=.007, power=.84, 2=.34), explained by reliable differences between direct threat and victim pictures (p=.03, d=1.02; threat>victim) and between injury and victim pictures (p=.007, d=1.56; injury>victim), but not between direct threat and injury pictures (p=.16). For the ADHD/ODD group, repeated measures ANOVA revealed a trend-level effect of Unpleasant Picture Subategory (F[2,18]=2.57, p=.10, power=.45, 2=.22), explained by trend-level differences between direct threat and victim pictures (p=.07, d=1.05; threat>victim) and between injury and victim pictures (p=.06, d=1.00; injury>victim), but not between direct threat and injury pictures (p=.80). For the ADHD/CD group, repeated measures ANOVA revealed a highly reliable main effect of Unpleasant Picture Subategory (F[2,20]=20.25, p<.0001, power=1.00, 2=.67), explained by reliable differences between direct threat and victim pictures (p<.0001, d=2.36; threat>victim), between injury and victim pictures (p=.001, d=1.73; injury>victim), and unlike the

PAGE 124

110 ADHD and ADHD/ODD groups, also between direct threat and injury pictures (p=.001, d=1.24; direct threat>injury). Examining each picture category separately, univariate ANOVA did not reveal a reliable main effect of Group for direct threat pictures (p=.44). However, univariate ANOVAs revealed a trend-level main effect of Group for victim pictures (F[2,31]=2.81, p=.076, power=.50), and a significant main effect of Group for injury pictures (F[2,31]=3.50, p=.04, power=.60). For victim pictures, unpaired 2-sided t-tests revealed a trend-level difference between the ADHD and ADHD/CD groups (p=.08, d=.75), and a more reliable difference between ADHD/ODD and ADHD/CD groups (p=.05, d=.94), but no difference between ADHD and ADHD/ODD groups (p=.69). For injury pictures, unpaired t-tests revealed a reliable difference between the ADHD and ADHD/CD groups (p=.03, d=1.14), but not between the ADHD and ADHD/ODD (p=.14) or the ADHD/ODD and ADHD/CD groups (p=.33). Specifically, ADHD children with comorbid Conduct Disorder tended to show decreased startle eyeblink response (i.e., greater inhibition) to victim pictures, compared to ADHD and ADHD/ODD children. In addition, ADHD/CD children tended to show decreased startle eyeblink response to injury pictures, compared to ADHD children with no comorbid disorders. Overall, ADHD/CD children showed a potentiated or increased startle response to direct threat pictures, and intermediate response to injury pictures, and an inhibited or decreased startle response to victim pictures relative to pleasant pictures. On the other hand, ADHD and ADHD/ODD children showed similarly potentiated or increased startle response to direct threat and injury pictures, and inhibited startle response to victim pictures relative to pleasant pictures. In addition, ADHD/CD children tended to show a

PAGE 125

111 greater degree of inhibited startle to victim pictures than ADHD or ADHD/ODD children. Moreover, similar to the series two analysis, startle response to injury pictures appears to most strongly differentiate the ADHD and ADHD/CD groups. Specifically, ADHD children showed reliably greater potentiated or increased startle to injury pictures compared to ADHD children with comorbid CD. To explore this finding, a logistic regression was performed to inspect how well startle response to threat pictures predicted ADHD versus ADHD/CD group membership (i.e., excluded ADHD/ODD group). The logistic regression revealed that startle response to victim pictures was a significant predictor of group membership (Wald=3.82, p=.05; Likelihood Ratio Test=6.44, p=.01), and resulted in 79.2% of the children being correctly classified, 76.92% or 10/13 ADHD children correctly classified, 81.8% or 9/11 ADHD/CD children correctly classified, a predictive value of 83.0% for ADHD, and a predictive value of 75.0% for ADHD/CD. Notable, when the ADHD/ODD children were included in the model, 0.0% of ADHD/ODD children were correctly classified (4 were incorrectly classified as ADHD, 6 were incorrectly classified as ADHD/CD). Pleasant Picture Subcategory Repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,62]=3.26, p=.05, power=.60, 2=.10), consistent with the series one analysis. Paired comparisons revealed reliable differences between attraction and thrill pictures (p=.02, d=.67; thrill>attraction), but not between attraction and nurture pictures (p=.13) or nurture and thrill pictures (p=.32), collapsed across groups. More relevant, there was no reliable Group by Pleasant Picture Subcategory interaction (p=.60).

PAGE 126

112 Overall, ADHD, ADHD/ODD, and ADHD/CD children showed a similar degree of startle reactivity to each of the different pleasant picture subcategories. Skin Conductance Response Picture Category For skin conductance (Figure 7-3), the trend tests revealed a quadratic (F[1,31]=19.51, p<0.0001, power=0.99, 2=0.39), as well as a linear (F[1,531]=7.01, p=.01, power=0.73, 2=0.18) trend over Picture Category, similar to the series one analysis. Pairwise comparisons (collapsed across groups) revealed reliable differences between the unpleasant and neutral picture categories (p<0.0001, d=0.67; unpleasant>neutral), between the unpleasant and pleasant picture categories (p=.009, d=0.37; unpleasant>neutal), and between the neutral and pleasant categories (p=.014, d=.34; pleasant>neutral). More relevant, there was a reliable interaction of Group (i.e., ADHD, ADHD/ODD, ADHD/CD) with quadratic trend over Picture Category (F[2,31]=5.63, p=.008, power=.82, 2=.27) but no linear interaction (p=.17). There was also trend-level main effect of Group (F[2,31]=3.10, p=.06, power=.56, 2=.17), explained by the ADHD/ODD group having lower overall skin response (collapsed across picture categories) than the ADHD group (p=.03, d=1.16), but no reliable differences between the ADHD and ADHD/CD groups (p=.21) or the ADHD/ODD and ADHD/CD groups (p=.20). Each group and picture category was examined separately to examine the source of the quadratic interaction.

PAGE 127

113 0 .05 .1 .15 .2 .25 .3log(1 + uSiemens) + S.D. Unpleasant Neutral PleasantPicture Category ADHD + CD ADHD + ODD ADHD Figure 7-3. Mean skin conductance response for ADHD, ADHD+ODD, and ADHD+CD groups across picture categories. Examining each group separately, for the ADHD group, trend tests revealed a reliable quadratic trend over Picture Category (F[1,12]=14.33, p=.003, power=.93, 2=.54), and a trend-level linear trend over Picture Category (F[1,12]=4.303, p=.06, power=.48, 2=.26). Pairwise comparisons for the ADHD group revealed reliable differences between unpleasant and neutral pictures (p=.004, d=.97; unpleasant>neutral), and between pleasant and neutral pictures (p=.03, d=.60; pleasant>neutral), but only a trend-level difference between unpleasant and pleasant pictures (p=.06, d=.49; unpleasant>pleasant). For the ADHD/CD group, trend tests revealed a reliable quadratic (F[1,10]=10.16, p=.01, power=.82, 2=.50) and linear (F[1,10]=10.90, p=.008, power=.84, 2=.52) trend over Picture Category. Pairwise comparisons for the ADHD/CD group revealed reliable differences between unpleasant and neutral pictures (p=.002, d=.58; unpleasant>neutral), and between unpleasant and pleasant pictures (p=.008, d=.56; unpleasant>pleasant), but not between neutral and pleasant pictures (p=.55).

PAGE 128

114 For the ADHD/ODD group, unlike the ADHD and ADHD/CD groups, trend tests did not reveal reliable quadratic (p=.18) linear (p=.97) trends over Picture Category, as there were no reliable skin conductance differences between any of the picture categories (all p>.15). Examining each picture category separately, univariate ANOVAs revealed a main effect of Group for unpleasant pictures (F[2,31]=3.75, p=.03, power=.64) and a trend-level main of Group for pleasant pictures (F[2,31]=2.82, p=.07, power=.50), but no main effect of Group for neutral pictures. For skin conductance response to unpleasant pictures, unpaired t-tests showed a reliable difference between the ADHD and ADHD/ODD groups (p=.02, d=.1.35; ADHD>ADHD/ODD), and a trend-level difference between the ADHD/CD and ADHD/ODD groups (p=.07, d=.95; ADHD/CD > ADHD/ODD), but no difference between the ADHD and ADHD/CD (p=.20) groups. For pleasant pictures, unpaired 2-sided t-tests only showed a trend-level difference between the ADHD and ADHD/ODD groups (p=.06, d=.89; ADHD>ADHD/ODD), but no difference between the ADHD and ADHD/CD groups (p=.12) or between the ADHD/ODD and ADHD/CD groups (p=.50). Overall, similar to the series two findings (i.e., greater ODD symptoms associated with decreased skin response), ADHD/ODD children tended to have weaker skin conductance responses to picture stimuli in general (largest difference with unpleasant pictures followed by pleasant pictures), compared to ADHD children without comorbidity. There were no reliable differences for skin conductance between ADHD and ADHD/CD children. In addition, children with ADHD without comorbidity showed increased skin conductance response to unpleasant and pleasant pictures relative to

PAGE 129

115 neutral pictures, whereas ADHD/CD children showed increased skin conductance response to unpleasant pictures relative to both neutral and pleasant pictures, and children with ADHD/ODD showed no reliable differences across unpleasant, neutral, and pleasant pictures. Unpleasant Picture Subcategory For skin conductance, repeated measures ANOVA did not reveal a significant main effect of Unpleasant Picture Subcategory (p=.47) or a Group by Unpleasant Picture Subcategory interaction (p=.25). Overall, there were no reliable group differences within or across different unpleasant subcategories for skin conductance response. Pleasant Picture Subcategory Repeated measures ANOVA did not reveal a reliable main effect of Pleasant Picture Subcategory (p=.13), or a Group by Pleasant Picture Subcategory interaction (p=.22). Overall, there were no reliable group differences within or across different pleasant subcategories for skin conductance response. Self-Report Ratings Picture Category For valence ratings (Figure 7-4), the trend tests revealed a linear trend over Picture Category (F[1,54]=185.49, p<0.0001, power=1.00, 2=.78), but not a quadratic effect (p=.23), similar to the series one analysis. Pairwise comparisons revealed highly reliable differences between each of the picture categories (all p<0.0001). More relevant, there was no significant interaction of group with linear (p=.27) or quadratic (p=.11) trend over Picture Category. However, there was a reliable main effect of Group (F[1,31]=3.66, p=.04, power=.63, 2=.19), explained by ADHD/CD children tending to rate pictures collapsed across all categories as more pleasant than ADHD (p=.03, d=.93) and

PAGE 130

116 ADHD/ODD (p=.03, d=1.0) children. Overall, while ADHD/CD children tended to rate picture stimuli as more pleasant, all the ADHD groups tended to judge the pleasant pictures as more pleasant than the neutral pictures, and the neutral pictures as more pleasant than the unpleasant pictures. 0 1 2 3 4 5 6 7Valence Rating + S.D. Unpleasant Neutral PleasantPicture Category ADHD + CD ADHD + ODD ADHD Figure 7-4. Mean self-report valence rating for ADHD, ADHD/ODD, and ADHD/CD groups across picture category. For arousal ratings (Figure 7-5), the trend tests revealed a quadratic trend over Picture Category (F[1,31]=51.17, p<0.0001, power=1.00, 2=.62), but not a linear effect (p=.35). Pairwise comparisons for the ADHD children were reported in the series one analysis (i.e., unpleasant and pleasant > neutral pictures). More relevant, trend tests revealed a reliable interaction of Group with quadratic trend over Picture Category (F[2,31]=5.32, p=0.01, power=0.80, 2=0.26), but no linear interaction or main effect of group (p=.73). Each group and category was then analyzed separately to determine the source of the quadratic interaction.

PAGE 131

117 0 1 2 3 4 5 6 7Arousal Rating + S.D. Unpleasant Neutral PleasantPicture Category ADHD + CD ADHD + ODD ADHD Figure 7-5. Mean self-report arousal rating for ADHD, ADHD/ODD, and ADHD/CD groups across picture category. For the ADHD without ODD/CD group, trend tests again revealed a reliable quadratic trend over Picture Category (F[1,12]=154.33, p<.0001, power=1.00, 2=.93), and no linear effect (p=.60), explained by reliable differences between the unpleasant and neutral pictures (p<.0001, d=3.24; unpleasant>neutral) and pleasant and neutral pictures (p<.0001, d=.2.85; pleasant>neutral), but not between the unpleasant and pleasant pictures (p=.60). For the ADHD/ODD group, the trend tests also revealed a reliable quadratic trend over Picture Category (F[1,9]=8.10, p=0.02, power=.72, 2=.47), and no linear effect (p=.71), explained by a reliable difference between pleasant and neutral pictures (p=.008, d=.79; pleasant>neutral), but not between unpleasant and neutral pictures (p=.11) or unpleasant and pleasant pictures (p=.71). For the ADHD/CD group, the trend tests revealed only a trend-level quadratic trend over Picture Category (F[1,10]=4.40, p=.06, power=.47, 2=.31) and no linear effect (p=.71), explained by a difference between pleasant and neutral pictures (p=.01, d=.91; pleasant>neutral), but not

PAGE 132

118 between unpleasant and neutral pictures (p=.31) or unpleasant and pleasant pictures (p=.20). Notable, univariate ANOVAs did not reveal any reliable main effects of Group within unpleasant (p=.20), neutral (p=.13), or pleasant (p=.52) picture categories. Overall, all groups generally showed similar patterns of self-report ratings of arousal across picture categories. However, the ADHD/ODD and ADHD/CD did not show reliable differences between unpleasant and neutral pictures, whereas the ADHD without comorbid ODD/CD did show reliable differences between unpleasant and neutral pictures. Unpleasant Picture subcategory For valence ratings, repeated measures ANOVA did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.20), or a Group by Unpleasant Picture Subcategory interaction (p=.99). Similarly, for arousal ratings, repeated measures ANOVA did not reveal a reliable main effect of Unpleasant Picture Subcategory (p=.21), or a Group by Unpleasant Picture Subcategory interaction (p=.41). Overall, there were no differences in self-report valence or arousal ratings between the ADHD, ADHD/ODD, and ADHD/CD groups across or within the different unpleasant picture subcategories. Pleasant Picture subcategory For valence ratings, repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,62]=9.17, p<.0001, power=.97, 2=.23), consistent with the series one analysis, and explained by the nurture pictures being judged as more pleasant than the attraction (p<.0001, d=.86) and thrill (p=.007, d=.32) pictures, and the thrill pictures being judged at more pleasant than attraction pictures (p=.0002, d=.51), collapsed across groups. In addition, there was no reliable Group by Pleasant Picture Subcategory interaction (p=.11). Overall, there were no differences in self-report valence

PAGE 133

119 ratings between the ADHD, ADHD/ODD, and ADHD/CD groups across or within the different pleasant picture subcategories. For arousal ratings (Figure 7-6), repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,62]=5.54, p=.006, power=.84, 2=.15), consistent with the series one analysis, and explained by the thrill pictures being judged as more highly arousing than the attraction (p<.0001, d=.67) and nurture (p=.003, d=.36) pictures, and the nurture pictures being judged at more highly arousing than the attraction pictures (p=.05, d=.28), collapsed across groups. In addition, there was a reliable Group by Pleasant Picture Subcategory interaction (F[4,62]=2.62, p=.04, power=.70, 2=.15). Each group and pleasant picture subcategory was then analyzed separately to determine the source of the interaction. 0 1 2 3 4 5 6 7Arousal Rating + S.D. Nurture Attraction ThrillPleasant Picture Subcategory ADHD + CD ADHD + ODD ADHD Figure 7-6. Mean self-report arousal rating for ADHD, ADHD/ODD, and ADHD/CD groups across pleasant picture subcategory. For the ADHD group without comorbid ODD/CD, repeated measures ANOVA revealed only a trend-level main effect of Pleasant Picture Subcategory (F[2,24]=3.17,

PAGE 134

120 p=.06, power=.55, 2=.21), explained thrill pictures being judged as more highly arousing than the attraction (p=.03, d=.91) and nurture (p=.02, d=.66) pictures, but no difference between nurture and attraction pictures (p=.56). For the ADHD/ODD group, repeated measures ANOVA revealed a reliable main effect of Pleasant Picture Subcategory (F[2,18]=5.77, p=.01, power=.80, 2=.39), explained both thrill (p=.02, d=.95) and nurture (p=.009,d=.88) pictures being judged as more highly arousing than attraction pictures, but no difference between thrill and nurture pictures (p=.83). For the ADHD/CD group, repeated measures ANOVA did not reveal a reliable main effect of Pleasant Picture Subcategory (p=.12), as there were no reliable differences between thrill, nurture, and attraction pictures (all p>.08). Examining each pleasant picture subcategory separately, univariate ANOVAs did not reveal reliable main effects of Group for nurture (p=.92) or thrill (p=.51) pictures. However, there was a reliable main effect of Group for attraction pictures (F[2,31]=3.26, p=.05, power=.57), explained by ADHD/CD children judging attraction pictures as more highly arousing than ADHD/ODD children (p=.01). There were no reliable differences between ADHD and ADHD/ODD (p=.32) or ADHD and ADHD/CD (p=.13) children. It is also important to note that the ADHD/CD children were an average of 1 year, 3 months older than the ADHD/ODD children (not statistically significant, p=.24). However, ADHD/CD children did not appear to rate attraction pictures as more arousing because they were older, as there was no association between age and attraction picture arousal (p=.59).

PAGE 135

121 Overall, ADHD, ADHD/ODD, and ADHD/CD children judged the arousal of nurture and thrill pictures similarly, whereas ADHD/CD children judged attraction pictures as more highly arousing than ADHD/ODD children. Raw Startle Eyeblink Response Raw startle eyeblink was examined as an exploratory analysis to investigate whether comorbid ODD or CD within ADHD was associated decreased absolute values of startle reactivity. Examining whether there was an overall main effect of group across all pictures was most relevant f or addressing this question. Overall, there was no main effect of Group (p=.99) across Picture Category, indicating that the comorbid ODD or CD did not impact the absolute startle magnitude in response to picture stimuli, consistent with the series two analysis. 0 2 4 6 8 10 12 14 16uV + S.D. Unpleasant Neutral PleasantPicture Category ADHD + CD ADHD + ODD ADHD Figure 7-7. Mean absolute startle response for ADHD, ADHD+ODD, and ADHD+CD across picture category.

PAGE 136

122 Startle Eyeblink Onset Latency Startle eyeblink onset latency was also examined as an exploratory analysis. For picture category, repeated measures ANOVA did not reveal a reliable main effect of Group (p=.91), main effect of Picture Category (p=.11), or a Group by Picture Category interaction (p=.22). For unpleasant and pleasant picture subcategories, repeated measures ANOVA also did not reveal reliable main effects of Picture Subcategory (p=.26, p=.29, respectively), or Group by Picture Subcategory interactions (p=.15, p=.40, respectively). Overall, there were no group differences across the different picture categories or subcategories. Series Three Results Summary Table A summary of series three results is presented in Table 7-1. Table 7-1. Series Three Results Summary Table: ADHD, ADHD/ODD, ADHD/CD Picture Category Unpleasant Picture Subcategory Pleasant Picture Subcategory Startle ADHD Unpleas.>Neutral, Pleas. ADHD/CD Unpleas., Neutral>Pleas. ADHD/ODD: lacked power Unpleasant>Pleasant Neutral Pictures: trend ADHD/CD>ADHD ADHD Threat, Injury>Victim ADHD/CD Threat>Injury>Victim ADHD/ODD: trend-level Threat, Injury>Victim Victim Pictures: ADHD & ADHD/ODD>ADHD/CD Injury Pictures: ADHD>ADHD/CD No Group Differences All groups Attraction
PAGE 137

123 Table 7-1. Continued. Picture Category Unpleasant Picture Subcategory Pleasant Picture Subcategory Skin Collapsed Across Pictures ADHD>ADHD/ODD ADHD Unpleas.> Pleas.>Neutral ADHD/CD Unpleas.>Pleas., Neutral ADHD/ODD No Category Differences Unpleasant Pictures: ADHD & ADHD/CD>ADHD/ODD No Group Differences No Subcategory Differences No Group Differences No Subcategory Differences Self-Report Valence Collapsed Across Pictures ADHD/CD>ADHD & ADHD/ODD All Groups: No interaction Pleas.>Neutral>Unpleas. No Group Differences No Subcategory Differences No Group Differences All Groups Nurture>Thrill>Attraction (Nurture most pleasant) Self-Report Arousal ADHD Unpleas., Pleas.>Neutral ADHD/CD Pleas.>Neutral ADHD/ODD Pleas.>Neutral No Group Differences No Subcategory Differences ADHD Thrill>Attraction, Nurture ADHD/CD No reliable differences ADHD/ODD Thrill, Nurture>Attraction Attraction Pictures ADHD/CD>ADHD/ODD

PAGE 138

CHAPTER 8 DISCUSSION Over the last decade, studies with normal adults have convincingly shown that the startle eyeblink response is larger during negative emotional states and smaller during positive emotional states. This phenomenon has proven useful for studying emotional processes in normal adults and those with psychopathology. Unfortunately, relatively few studies have examined affective modulation of the startle reflex (AMSR) in normal healthy children or children with presumed underlying emotional dysfunction (i.e., anxiety disorders, ADHD, Conduct Disorder). The primary aim of the current study was to compare patterns of AMSR of children with ADHD to children without ADHD. Specifically, do children with ADHD show aberrant patterns of reactivity compared to healthy control children? A secondary aim was to investigate whether the severity of conduct-related (i.e., CD and ODD), psychopathy, and impulsive symptoms impact emotional reactivity within ADHD. Specifically, if aberrant patterns of emotional reactivity exist in ADHD, are they explained by or influenced by the severity of comorbid conduct problems or impulsivity? Moreover, while the current study was not designed to specifically test different models of ADHD, findings will also be discussed in the context of inhibition, reward dysfunction, arousal, and attentional models of ADHD. 124

PAGE 139

125 ADHD Versus Controls Picture Category With respect to overall picture categories (i.e., unpleasant, neutral, pleasant), it was hypothesized that (a) both ADHD and control groups would show a linear pattern of startle eyeblink reactivity (i.e., unpleasant > pleasant), but the ADHD group would exhibit decreased startle potentiation to unpleasant pictures compared to controls; and (b) both ADHD and control groups would show a quadratic pattern of skin conductance reactivity (i.e., unpleasant, pleasant > neutral), but the ADHD group would exhibit overall reduced reactivity across all picture categories compared to controls; and (c) both ADHD and control groups would similarly show a linear pattern of self-report valence ratings (pleasant > neutral > unpleasant) and a quadratic pattern of self-report arousal ratings (unpleasant, pleasant > neutral). It was also predicted that there would be an interaction of age with startle reactivity, explained by reduced potentiated startle response to unpleasant pictures in younger children. These predictions are only partially supported by the results. Specifically, both ADHD and control children show the predicted linear “adult-like” pattern of affective modulation of the startle reflex, that is larger startle eyeblink reactivity to unpleasant pictures relative to pleasant pictures. However, contrary to the hypothesis, there is no reliable difference between the groups for startle reactivity to unpleasant pictures (not accounting for subcategories). In addition, both ADHD and control groups show the predicted quadratic pattern of skin conductance reactivity, although both groups also show a linear response pattern, explained by greater skin conductance response to unpleasant pictures than to pleasant pictures, and greater skin conductance response to pleasant pictures than to neutral pictures. Contrary to what was predicted, ADHD

PAGE 140

126 children do not exhibit overall reduced skin conductance reactivity across all picture categories compared to controls. Next, as predicted, both ADHD and control groups similarly show a linear pattern of self-report valence ratings, as both groups judge pleasant pictures as more pleasant than neutral pictures, and neutral pictures as more pleasant than unpleasant pictures. In addition, as predicted, both ADHD and control groups rate the neutral pictures as less arousing than the unpleasant and pleasant pictures. However, controls rate unpleasant pictures as more highly arousing than pleasant pictures, whereas ADHD children rate unpleasant and pleasant pictures similarly on arousal. Moreover, the current dissertation study does not replicate findings of a previous normative study (Selke, 2005), where younger normal healthy boys showed startle potentiation to direct threat but not victim or injury pictures, and older healthy boys showed potentiated startle to all unpleasant picture subcategories. Specifically, in the current study, age does not interact with startle reactivity in children with or without ADHD. In other words, older and younger children showed similar patterns of startle reactivity within each group. These findings may simply reflect differences in the picture stimuli used. Specifically, 45 picture stimuli (15 per category or 45/51 pictures had startle probes) were used in the original normative study. Notable, a number of the unpleasant pictures did not reliably elicit startle potentiation across older or younger children in the normative study, and therefore were not used in the more recent dissertation study. In addition, fewer picture stimuli (i.e., 12 per category or 36/47 pictures had startle probes) were used in the current study. This may have decreased overall habituation to the unpleasant and pleasant picture stimuli and to the actual startle

PAGE 141

127 probes (i.e., 100dB noises), resulting in magnifying the relative differences between picture categories, increasing reliability, and increasing the power to detect changes in the younger children. At this point, there appears to be minimal differences between healthy control and ADHD children with respect to physiologically and verbally responding to affective pictures. Most significant, both ADHD and healthy control children show highly reliable aversion enhanced startle modulation. Both groups also appear to be similarly successful at cognitively appraising their emotional experience associated with viewing affective picture stimuli. The only difference across variables was that controls tend to judge unpleasant pictures as more arousing than ADHD children, although this does not appear to lead to corresponding differences between the groups with regard to magnitude of startle or skin conductance response to unpleasant pictures. Unpleasant Picture Subcategory Results will now be examined with regards to the different types of unpleasant pictures, namely direct threat, injury, and victim pictures. With respect to unpleasant picture subcategories, it was hypothesized that (a) both ADHD and control groups would show potentiated startle to direct threat pictures, with controls also showing potentiated startle to injury and victim pictures but to a lesser degree than with direct threat pictures, and with ADHD children showing inhibited startle to injury and victims pictures (relative to pleasant pictures); and (b) the ADHD group would show overall reduced skin conductance across all three subcategories compared to controls, but there would be no differences in skin conductance across subcategories within each group; and (c) there would be no differences between ADHD and control groups or across unpleasant picture subcategories within each group for valence and arousal self-report ratings.

PAGE 142

128 Again, these predictions are only partially supported by the results. Specifically, contrary to what was predicted, there is no significant difference in startle reactivity across the different unpleasant subcategories for healthy control children, as direct threat, injury, and victim pictures are all similarly increased or potentiated relative to the pleasant picture category. In contrast, ADHD children show potentiated startle for direct threat and injury pictures, but inhibited startle for victim pictures, relative to the pleasant picture category. Next, as predicted, there are no reliable skin conductance differences across direct threat, injury, and victim pictures within each group. However, contrary to what was predicted, ADHD children do not show reduced skin conductance reactivity across the unpleasant picture subcategories compared to controls. In addition, while there are no group differences with respect to valence, contrary to what was predicted, both groups appraise victim pictures as being more unpleasant than direct threat pictures. Moreover, while within each group there are no differences in rated arousal across the three subcategories, healthy controls tend to judge all unpleasant picture subcategories as more arousing than ADHD children. Overall, both ADHD and healthy control children show highly reliable aversion enhanced startle modulation to direct threat and injury pictures. However, only controls show aversion enhanced startle to victim pictures, and in contrast, ADHD children actually display inhibited startle reactivity to victim pictures. Furthermore, inhibition of startle response to victim pictures occurs in the ADHD group despite victim pictures being judged as more unpleasant than direct threat, and victim pictures judged no differently than direct threat or injury pictures on arousal. This also occurs despite victim

PAGE 143

129 pictures showing similar levels of skin conductance response to direct threat and injury pictures, suggesting a similar degree of physiological arousal. Thus, ADHD children appear to become appropriately physiologically aroused by victim picture, and are also successful at cognitively appraising their emotional experience associated with viewing victim pictures. This “mismatch” or lack of correspondence between different components of emotional processing suggests that children with ADHD have a defect in translating the results of cognitive appraisal into somato-motor changes associated with unpleasant emotional stimuli that specifically depicts victimization. These results may be interpreted or considered from a number of different positions or theories. One of the most plausible explanations for these results relates to attention. Specifically, inhibition of startle to victim pictures may reflect a deficit in attentional processing in children with ADHD. Lang and colleagues (Lang et al., 1997; Bradley et al., 2001) have postulated the “defense cascade model” where defensive responding (i.e., to unpleasant stimuli) goes through a series of stages, dependent on the interaction between attention and emotion. In more detail, early in the defensive “cascade,” oriented attention dominates, inhibiting defensive activation in the service of information detection, processing, and retrieving from memory. This can be observed in the case of pre-pulse inhibition, where the startle reflex is inhibited after startle probes (i.e., 100 dB noise) are presented early into picture viewing (i.e., less than one second). In addition, picture stimuli that engage attentional systems to a greater degree (i.e., those with affective content) result in greater inhibition of the startle reflex compared to pictures with fewer attentional demands (Bradley, Cuthbert, & Lang, 1993). Next, as time elapses in the “defensive cascade,” the inhibitory effects of oriented attention decreases

PAGE 144

130 as defensive activation emerges. At this point, if an unpleasant stimulus is sufficiently intense or threatening, the threshold for defensive mobilization is reached, resulting in an increase in skin conductance and potentiated startle. If the emotional intensity and perceived threat is strong enough, overt behavioral action (i.e., fight or flight response) will also ultimately result (Bradley et al., 2001). Thus, there are two competing processes that impact the magnitude of the startle reflex during picture viewing. One is attentional processing that results in inhibition, and the second is the degree of defensive activation that results in potentiation. Notable, Lang and colleagues (Bradley et al., 2001; Levenston et al., 2000) posit that even at the later stages of the “defensive cascade” (e.g., after 3-4 seconds), high levels of attentional processing can potentially dominate over or inhibit startle response to even strongly aversive and arousing stimuli. For instance, Bradley et al. (2001) found that mutilation pictures elicited relatively smaller startle response than direct threat pictures, despite mutilation pictures eliciting similar levels of skin conductance response, self-report arousal, and self-report valence. The authors hypothesized that this occurred because the mutilation pictures required greater attentional processing than direct threat pictures, resulting in relative inhibition. Bradley’s (2001) findings parallel what was found in the current dissertation study. Specifically, ADHD children appear to become appropriately physiologically aroused (i.e., skin conductance) by victim pictures, and are also successful at cognitively appraising their emotional experience associated with viewing victim pictures, as these responses are similar to direct threat and injury pictures. Applying the principles of the “defensive cascade model,” it follows that startle reactivity to victim pictures in ADHD

PAGE 145

131 may have been inhibited because they required greater attentional demands than direct threat and injury pictures. In other words, the findings may be indicative of an increased threshold in ADHD for the shift from an orienting or attentional disposition to one of defensive activation (Levenston et al., 2000) for scenes depicting victimization. In a sense, the results of this study support the premise that ADHD children “get stuck” in the attentional engagement of the victimization scenes, where they are more dominated by attentional factors rather than fear or aversion induced emotion factors. This may occur because children with ADHD appear to have impairments in focused or selective attention, valuable for exploration of one’s environment (Brodeur & Pond, 2001; Chhabildas et al., 2001; Cooley & Morris, 1990; Mirsky et al., 1991; Seidman et al., 1997; Sergeant et al., 2002). In addition, while attentional dysfunction in ADHD appears to impact emotional processing of victim pictures, it does not appear to significantly dominate over defensive activation during direct threat or injury pictures. This finding may be due to victim pictures requiring greater attentional resources to analyze and synthesize picture content, as they appear to be qualitatively more complex than direct threat or injury pictures. In particular, victim pictures involve a social interaction between two individuals, and require the appraisal and integration of differing emotional states and physical actions exhibited by the two individuals (e.g., fear or helplessness by victim versus anger or callousness of perpetrator). Thus, children with ADHD may be more vulnerable than children without ADHD to the increased attentional demands of victimization pictures, resulting in inhibition of the startle reflex. Along these lines, attentional theories of ADHD would predict that more severe attentional deficits in ADHD would be associated with the greatest degree of startle inhibition to

PAGE 146

132 victim pictures. Contrary to this expectation, in exploratory analyses within the ADHD group, there were no reliable associations between measures of attention (i.e., CPT omission errors, Connors’ Connors’ DSM-IV Inattentive scale, or ADHD Inattentive versus Combined subtype) and startle reactivity to any of the picture categories or subcategories. There were also no associations between measures of attention and skin conductance response of self-report valence and arousal ratings. Future research should more directly test this attention-based hypothesis by comparing different types of unpleasant pictures in ADHD using pre-pulse inhibition and event-related potential (ERP) techniques. Future studies should also examine affective modulation of children with ADHD off and on medication. It is likely that ADHD children off their medication would have decreased attentional abilities. Under these conditions, it is possible that subsequent increased attentional demands across all unpleasant picture stimuli would further dominate over emotions, resulting in reduced or inhibited startle response not only to victim pictures, but also to direct threat and injury pictures. Next, these findings will be examined in the context of empathic responding. Specifically, if ADHD is associated with a reduced capacity to express or experience empathy, it would be predicted that ADHD children would (a) judge victim pictures as less arousing and more pleasant, compared to direct threat pictures and compared to healthy controls, (b) become less physiologically aroused relative to direct threat pictures and healthy controls, and (c) subsequently show decreased startle response to victim pictures relative to direct threat pictures and healthy controls. It is also possible that injury pictures would elicit emotional responses intermediate to victim and direct threat pictures, as injury

PAGE 147

133 pictures potentially elicit feelings of disgust (egocentric) as well as empathy for the injured individual. The results only partially support the hypothesis that ADHD is associated with dysfunction in empathic responding. Specifically, ADHD children verbally report feeling just as unpleasant, bad, scared, or upset (i.e., valence) as controls report during victim pictures. In fact, similar to controls, ADHD children verbally report that victim pictures actually make them feel more unpleasant than direct threat pictures, suggesting intact ability to express discomfort associated with viewing scenes depicting victimization. ADHD children also verbally report feeling just as aroused or activated in response to victim pictures as direct threat or injury pictures. In addition, ADHD children experience similar levels of physiological arousal (i.e., skin conductance) to victim pictures, as compared to direct threat and injury pictures, and as compared to healthy controls. Still, a similar level of skin conductance response may not be due to empathy, as feelings of excitement could also result in an increased skin conductance response. Most supportive of dysfunction in empathic responding is the finding that victimization pictures are associated with inhibited startle reactivity in ADHD. Still, it cannot be ruled out that heightened attentional demands of victimization scenes lead to this phenomenon. In other words, attentional deficits in ADHD may directly lead to dysfunction in empathic responding. Overall, there appears to be only tentative evidence for support of a deficit in empathy in ADHD. Pleasant Picture Subcategory Results will now be examined with regards to the different types of pleasant pictures, namely nurture, thrill, and attraction pictures. With respect to pleasant picture subcategories, it was hypothesized that (a) both ADHD and control groups would show inhibited startle across all three pleasant picture types, but the controls would show less

PAGE 148

134 inhibition of startle to thrill pictures compared to nurture and attraction pictures due to the element of risk in thrill pictures, whereas the element of risk would heighten the appetive experience of thrill pictures in ADHD (based on stimulation seeking theories and results of Levenston et al., 2000) resulting in similar levels of startle inhibition across all three pleasant picture subcategories for ADHD; and (b) the ADHD group would show overall reduced skin conductance across all three subcategories compared to controls, but there would be no differences in skin conductance across subcategories within each group; and (c) there would be no differences between ADHD and control groups or across pleasant picture subcategories within each group for valence and arousal self-report ratings. Again, these predictions are only partially supported by the results. Specifically, as predicted, both controls and ADHD children display inhibited startle across all three pleasant picture types relative to the unpleasant picture category (collapsed across subcategories). However, contrary to what was predicted, both groups similarly show greater startle inhibition to attraction pictures compared to thrill and nurture pictures. In addition, contrary to what was predicted, ADHD children show similar levels of skin conductance reactivity to healthy controls across the pleasant picture subcategories. Moreover, as predicted, there are no differences between ADHD and control groups across the pleasant picture subcategories for valence and arousal self-report ratings. However, both groups rated attraction pictures as less pleasant and less arousing than nurture and thrill pictures. Notable, normal adult populations have been found to display a similar pattern of startle reactivity. Specifically, erotica has been found to elicit greater inhibition of startle than thrill or nurture pictures in normal adults (Bradley et al., 2001). Overall, ADHD and healthy control children show similar levels of startle inhibition to

PAGE 149

135 different types of pleasant picture stimuli. Moreover, it appears that similar to normal adults, attraction pictures most strongly activate the appetitive motivational system in both ADHD and control children. While there are no group differences across all of the variables for pleasant pictures, one might expect differences in emotional reactivity to attraction pictures between younger and older participants (irrespective of group) due to maturational and social factors. In an exploratory analysis, children ages 7 to10 years-old were more likely to rate attraction pictures as less pleasant and less arousing than thrill and nurture pictures, compared to children ages 11 to 14 years-old (Younger children: all p<.003, attraction less pleasant and arousing than nurture & thrill; Older children: p<.004 for attraction less pleasant than nurture only, but no differences for arousal ratings). These findings are interesting considering attraction pictures elicited the greatest startle inhibition compared to both thrill and nurture pictures in both younger and older participants. These findings suggest that a highly reliable appetive response to opposite sex “attraction” pictures exists relatively early in development (ages 7-10), despite younger children appraising the pictures as less pleasant and less arousing than other types of pleasant pictures. Next, increased startle inhibition response to attraction pictures can be evaluated in the context of theories of reward dysfunction. As discussed in the literature review, theories of reward dysfunction in ADHD would predict that pleasant picture stimuli would activate the appetive or approach motivational system to a lesser degree (i.e., increased threshold) in ADHD children compared to non-ADHD children, resulting in lower valence ratings (i.e., less pleasant) and decreased startle inhibition in response to

PAGE 150

136 pleasant picture stimuli in ADHD children. The results of the study do not support these predictions, as ADHD and healthy control children show similar valence ratings and degree of inhibited across all types of pleasant pictures. Overall, there is no evidence for reward dysfunction in ADHD, at least in the context the emotion paradigm used in the current study. Other Theoretical Frameworks In the context of the current dissertation study, under-arousal theories of ADHD would predict that children with ADHD would show lower absolute startle values, skin conductance magnitude, and self-report arousal ratings in response to affective picture stimuli. Contrary to under-arousal theories, ADHD children in the current study are not physiologically under-aroused compared to healthy control children. Specifically, ADHD children do not display lower absolute startle responses or lower skin conductance responses while viewing unpleasant, neutral, or pleasant pictures. In addition, compared to healthy controls, ADHD children do not verbally report feeling less aroused in response to viewing neutral or pleasant pictures. The only support of under-arousal comes from the finding that ADHD children verbally report feeling less aroused than controls in response to viewing unpleasant pictures. As a whole, the findings do not strongly support under-arousal theories of ADHD. Moreover, with respect to stimulation seeking theories, ADHD children would be predicted to show higher arousal and valence (more pleasant) ratings, stronger skin conductance response, and greater inhibition of startle to thrilling pictures, compared attraction and nurture pictures, and compared to healthy controls. Results indicate that children with and without ADHD show similar responses to thrill pictures across all of the variables. In addition, both groups show similar levels of startle inhibition and skin

PAGE 151

137 conductance response to thrill and nurture pictures. Overall, the current study does not support theories linking elevated stimulation seeking to ADHD. Highly relevant, the findings of the current study are partially consistent with Levenston’s (2000) study examining AMSR in adult psychopaths. Specifically, psychopaths demonstrated potentiated startle to only direct threat pictures, but inhibited startle to pictures depicting mutilation/injury and assault/victimization, relative to pleasant pictures. As children with ADHD appear to be at increased risk for developing Antisocial Personality Disorder and possibly psychopathy (Barkley, 1998; Kaplan & Sadock, 1998; Lynam, 1996, 1997), it is likely that a least a subset of the psychopaths in Levenston’s study met criteria for ADHD as children. Considering the similarities between children with ADHD and adult psychopaths leads to the question of whether aberrant startle response patterns seen in psychopathic adults are related to ADHD symptomatology rather than to psychopathy per say. This clearly cannot be answered based on the current study. However, the results of the current study can at least bring into question whether studies of adult psychopathy should have accounted for symptoms of ADHD, which they appear not to do (Patrick et al., 1993; Levenston et al., 2000; Herpertz et al., 2001). As the children with ADHD in the current dissertation study displayed a varying range of conduct and psychopathy symptomatology, this question will be more directly considered in the following sections. Variables Within ADHD The secondary aim of the study was to investigate whether the severity of conduct-related (i.e., CD and ODD), psychopathy, and impulsive symptoms impact emotional reactivity within ADHD. In addition, is inhibition of startle to victim pictures in ADHD

PAGE 152

138 explained by or influenced by the severity of comorbid conduct problems, psychopathy symptoms, or impulsivity? It was predicted that greater conduct, oppositional defiant, and psychopathy symptoms would be associated with reduced startle potentiation to unpleasant picture stimuli. These predictions are only partially supported by the results, as discussed below. Startle Reactivity Unpleasant picture category With respect to startle reactivity, higher number of ODD symptoms, but not CD or psychopathy symptoms, are associated with decreased startle reactivity to unpleasant pictures as a whole (i.e., collapsed across all subcategories). This finding suggests that comorbid ODD symptomatology may increase the risk for emotional dysregulation in ADHD. Still, as discussed below, comorbid ODD symptoms do not appear to impact startle magnitude response within each of the separate unpleasant picture subcategories (i.e., direct threat, injury, victim pictures). Direct threat pictures Conduct, oppositional defiant, and psychopathy symptoms are not associated with the magnitude of startle response to direct threat pictures in ADHD. Specifically, irrespective of severity of comorbid symptoms ODD, CD, and psychopathy symptoms or comorbid ODD or CD diagnosis, children with ADHD reliably show potentiated startle to direct threat pictures. Therefore, ADHD children with and without comorbid conduct-related problems show highly reliable aversion enhanced startle modulation to direct threat.

PAGE 153

139 Victim pictures Next, conduct, oppositional defiant, and psychopathy symptoms are not reliably associated with the magnitude of startle response to victim pictures. In addition, ADHD children diagnosed with and without comorbid ODD or CD (most also with ODD), all showed inhibited startle to victim pictures. These findings support the premise that startle inhibition to victim pictures is primarily related to having ADHD, rather than to CD, ODD, or psychopathy symptomatology. Nonetheless, ADHD children with comorbid CD tend to show even further decreased startle eyeblink response (i.e., greater inhibition) to victim pictures, compared to ADHD and ADHD/ODD children. This suggests that the already dysfunctional emotional regulation to victim pictures in ADHD is amplified by the co-occurrence of Conduct Disorder, but not by Oppositional Defiant Disorder. Injury pictures Results indicate that higher number of comorbid CD and psychopathy symptoms, but not ODD symptoms, are associated with decreased startle reactivity to injury pictures in ADHD. This finding is not confounded by age, as age is not reliably correlated with number of CD symptoms (r=.14, p=.44) or psychophathy symptoms (r=.15, p=.41). In addition, ADHD children with comorbid CD diagnosis tend to show decreased startle eyeblink response to injury pictures, compared to ADHD children without comorbidity. While CD and psychopathy symptoms appear to reduce the degree of startle modulation to injury pictures, startle reactivity to injury pictures in ADHD/CD remains potentiated compared to pleasant pictures and victim pictures. Interestingly, results of the current study partially parallel results from Levenston’s (2000) study, in that adult psychopaths demonstrated potentiated startle to only direct

PAGE 154

140 threat pictures, but inhibited startle to pictures depicting mutilation/injury or assault/victimization relative to pleasant pictures. In addition, startle response to injury pictures is a relatively strong predictor of ADHD versus ADHD/CD group membership (i.e., excluding ADHD/ODD group: correct classification = 79.2%, 76.92% or 10/13 ADHD children correctly classified, 81.8% or 9/11 ADHD/CD children correctly classified). Therefore, startle response to injury pictures may not only be valuable for differentiating between ADHD children from those with comorbid Conduct Disorder, but also may represent an early biologic marker for increased risk of developing psychopathy. Pleasant picture category Conduct, oppositional defiant, and psychopathy symptoms are not associated with the magnitude of startle response to pleasant picture category or different pleasant picture subcategories in ADHD. Specifically, irrespective of severity of comorbid symptoms or diagnosis of comorbid ODD or CD, children with ADHD reliably show inhibited startle to pleasant pictures, relative to direct threat and injury pictures (but not relative to victim pictures). Therefore, ADHD children with and without comorbid conduct-related problems show highly reliable startle inhibition to pleasant stimuli (i.e., nurture, thrill, and attraction pictures), suggesting an intact appetitive motivational system. Neutral picture category Unexpectedly, higher number of CD symptoms, ODD symptoms, and psychopathy symptoms are all reliably associated with increased startle reactivity to neutral pictures. In addition, while ADHD children with comorbid CD diagnosis show an “adult-like” linear startle response pattern (unpleasant > pleasant pictures), they also show a quadratic response pattern, explained by greatest startle response to neutral pictures. Children with

PAGE 155

141 ADHD alone and ADHD/ODD do not show this quadratic response pattern. Furthermore, this unusual response pattern in ADHD/CD occurs despite ADHD, ADHD/ODD, and ADHD/CD groups showing similar levels of low arousal, neutral valence ratings, and low physiological arousal (i.e., skin conductance) in response to neutral pictures. Still, startle response to neutral pictures is only moderate predictor of ADHD versus ADHD/CD group membership (i.e., excluding ADHD/ODD group: correct classification = 62.50 %, 62.50 % correct for ADHD, 54.6% correct for ADHD/CD). Taken as a whole, startle response to neutral pictures may not only be valuable for differentiating between ADHD children from those with comorbid conduct problems, but like injury picture, may also represent an early biologic marker for increased risk of developing psychopathy. Interestingly, these results parallel what Patrick et al. (1993) and Levenston et al. (2000) found with adult psychopaths. Specifically, adult psychopaths also showed a quadratric response pattern, with neutral pictures eliciting greater startle reactivity than pleasant and unpleasant pictures. Such parallel response patterns further reinforce the hypothesis that ADHD children who display more severe forms of conduct-related problems (i.e., conduct disorder, high psychopathy), may be at increased risk for developing psychopathy in adulthood (Lynam, 1996, 1997, 1998). In addition, these findings suggest that children with ADHD/CD may be at increased risk for inappropriate defensive responding (i.e., fight or flight response) during low arousal, neutral situations. This premise is consistent with findings that CD children are at increased risk for misperceiving the intentions of others in ambiguous situations as more hostile and threatening than non-CD children (Kaplan & Sadock, 1995).

PAGE 156

142 Startle and comorbidity overview Examining comorbidity within a dimensional framework, severity of comorbid CD, ODD, and psychopathy symptomatology do not appear to impact the degree of potentiated startle for direct threat pictures, or the degree of inhibited startle for victim pictures. In contrast, higher number of comorbid CD and psychopathy symptoms, but not ODD symptoms, appear to reduce startle reactivity to injury pictures. Furthermore, examining discrete diagnostic groups, on the whole there are no reliable differences between the ADHD and ADHD/ODD groups. Specifically, both groups show similarly potentiated startle to direct threat and injury pictures (i.e., no reliable difference between direct threat and injury), and similar levels of inhibited startle to victim pictures. In comparison, ADHD/CD children show the greatest degree of potentiated startle to direct threat pictures, lower but still potentiated startle to injury pictures, and an inhibited startle to victim pictures, relative to pleasant pictures. In addition, ADHD/CD children tend to show a greater degree of inhibited startle to victim pictures than ADHD or ADHD/ODD children. These findings support the conclusion that comorbid conduct disorder symptoms, but not comorbid oppositional defiant symptoms, impact the pattern of affective modulation of the startle reflex in ADHD. Moreover, the current sample of ADHD/CD children consisted of children with mild to moderate conduct disorder, but none met criteria for the severe form of CD. This may have reduced the power and sensitivity to detect further conduct disorder related impact (i.e., may have found differences on direct threat pictures or even further reductions in startle response to injury and victim pictures). Therefore, it cannot be ruled out that children with the most severe forms of comorbid Conduct Disorder would show an even more aberrant pattern of affective modulation of the startle reflex. Future

PAGE 157

143 research should attempt to recruit ADHD children with more severe forms of comorbid ADHD. Along these lines, future studies should also examine children with Conduct Disorder with and without comorbid ADHD, to better evaluate the relative impact of ADHD versus CD on emotional reactivity. Skin Conductance It was predicted that greater conduct, oppositional defiant, and psychopathy symptoms would be associated with reduced skin conductance response to both unpleasant and pleasant picture stimuli. These predictions are only partially supported. Specifically, symptoms of conduct disorder and psychopathy are not associated with skin conductance response across unpleasant, pleasant, and neutral picture categories. In contrast, higher number of ODD symptoms is associated with decreased skin conductance response to unpleasant, pleasant, and neutral picture categories. Moreover, ADHD children diagnosed with comorbid ODD do not show differences in skin conductance across unpleasant, pleasant, and neutral pictures, as skin response is relatively low across all picture categories. In addition, skin conductance averaged across all pictures is a strong predictor of comorbid ODD diagnosis within ADHD. Specifically, comparing the 16 ADHD children without an ODD diagnosis (i.e., 13 ADHD-only and 3 ADHD/CD participants) to the 18 ADHD children with an ODD diagnosis (i.e., 10 ADHD/ODD and 8 ADHD/ODD/CD participants) resulted in an overall correct classification of 70.6%, 15/18 or 83.3% children correctly classified with comorbid ODD, and 56.3% or 9/16 children correctly classified as not having ODD. Therefore, skin conductance response appears to be valuable for differentiating between ADHD with and without comorbid oppositional defiant disorder.

PAGE 158

144 In conjunction with the startle reactivity findings, these findings suggest that ODD symptoms impact a different component of emotional processing than CD or psychopathy symptoms. Specifically, ODD symptomatology appears to impact physiological arousal (i.e., lower skin conductance) whereas CD and psychopathy symptomatology appear to impact affective startle modulation (i.e., reduced startle response to injury and victim pictures). Self-Report Ratings It was predicted that conduct, oppositional defiant, and psychopathy symptoms would not impact valence and arousal ratings in ADHD. These predictions are only partially supported. With respect to valence ratings, comorbid ODD symptomatology does not impact valence ratings across picture categories or subcategories. However, greater comorbid CD and psychopathy symptomatology is associated higher valence ratings collapsed across picture categories (i.e., all pictures as a whole judged as more pleasant), and higher valence ratings with attraction pictures. However, separate ADHD, ADHD/ODD, and ADHD/CD groups do not rate attraction pictures differently on valence. Most relevant, each of the separate ADHD groups tend to judge the pleasant pictures as more pleasant than the neutral pictures, and the neutral pictures as more pleasant than the unpleasant pictures, suggesting intact appraisal of valence across all ADHD participant groups. With respect to arousal ratings, there are no reliable associations between rated arousal and conduct disorder or psychopathy symptoms. However, greater comorbid ODD symptomatology is associated with lower arousal ratings for unpleasant pictures. As earlier reviewed, skin conductance typically correlates with self-reported arousal in normal adults, regardless of the valence of the stimulus (Bradley et al., 1990; Greenwald,

PAGE 159

145 Cook, & Lang, 1989; Lang, Greenwald, Bradley, & Hamm, 1993). In the context of the current study, while comorbid ODD symptoms is associated with reduced skin conductance response (i.e., physiological arousal) across unpleasant, pleasant, and neutral pictures, a corresponding relationship between ODD symptoms and self-report arousal exists only for unpleasant pictures. Specifically, greater ODD symptomatology is associated with decreased self-report arousal for unpleasant, but not pleasant or neutral pictures. Therefore, cognitive appraisal of arousal does not appear to fully account for the association between ODD symptoms and reduced skin conductance response in ADHD. Interestingly, greater psychopathy symptomatology is associated with increased arousal to attraction pictures. This finding is likely not related to age, as age is not reliably correlated with psychophathy symptoms in the current study. In addition, there are no corresponding associations between degree of psychopathy symptomatology and startle reactivity or skin conductance for attraction pictures. In addition, verbally expressing how one feels when seeing moderately provocative attraction pictures in the presence of a stranger (i.e., the examiner) may potentially lead to some degree of self-consciousness and subsequent under-reporting or arousal and pleasantness in children. In this context, this finding leaves open the possibility that comorbid psychopathy symptomatology is associated with decreased social inhibition. Impulsivity It was predicted that greater impulsivity as determined by lower CPT Beta and higher Restless-Impulsive Subscale score (CPRS-R:L) would be associated with reduced reactivity to affective stimuli (eyeblink magnitude and SC). Specifically, ADHD children with lower Beta and higher Restless-Impulsive Index scores would show

PAGE 160

146 reduced startle potentiation to aversive stimuli, and reduced skin responses to both unpleasant and pleasant pictures. For the most part, these predictions are not supported by the results. Specifically, the only reliable finding is that ADHD children who are more physiologically aroused by attraction pictures show greater risk-taking (Beta score) on a continuous performance test. It is possible that this finding is related to symptoms of conduct disorder or psychopathy. However, this hypothesis is not supported as CD and psychopathy symptoms are not reliably associated with greater risk-taking (Beta score) on the continuous performance test (p=.62, p=.45, respectively). Notably, there are no other reliable relationships between measures of emotional reactivity (i.e., startle reactivity, skin conductance, and self-report ratings) and parent report of impulsivity or direct measures of impulsivity. Despite prior research finding that impulsivity is closely correlated with antisocial behavior in adolescence (Luengo et al. 1994), and that level of impulsivity in kindergarten and fourth grade has been shown to be predictive of later delinquent behavior (Tremblay et al., 1994; White et al., 1994), based on the current results, impulsivity on the whole does not appear to reliably impact emotional reactivity. Moreover, theories of disinhibition would predict that atypical patterns of emotional reactivity (i.e., startle reactivity, skin conductance) would be directly associated with decreased behavioral inhibition. For instance, Barkley’s (1997, 1998) model posits that impaired behavioral inhibition (i.e., inhibiting a prepotent response, stopping an ongoing response, and inhibiting interference) leads to impairment of four “executive functions,” with one being self-regulation of affect/motivation/arousal. However, a variety of behavioral inhibition measures, including CPT commission errors, CPT Beta, Connors’ Restless-Impulsive Index, Connors’ DSM-IV Hyperactive

PAGE 161

147 Impulsive scale, are not associated with impaired self-regulation of affect (i.e., startle reactivity, self-report valence) or arousal (i.e., skin conductance, self-report arousal). In addition, startle latency (i.e., reaction time of the eyeblink reflex) could be cautiously viewed as a measure of physiological inhibition, with faster reaction times being indicative of greater disinhibition. However, the results of the current study do not reveal any differences between ADHD and healthy control children on startle latency. Overall, theories of inhibition, and in particular Barkley’s model of ADHD, are not supported by the results of the current study. Still, there remains the possibility that CPT Beta, CPT Commission Errors, Connors’ Restless-Impulsive Index, and Connors’ DSM-IV Hyperactive-Impulsive scale are poor measures of underlying behavioral and cognitive inhibition or impulsivity. Future studies should use other possible “purer” measures of inhibitory control, such as Stop-Signal Reaction Time on the Stop Task (Oosterlann et al., 1998) to better examine the relationship between inhibition and emotional reactivity. In addition, the ADHD children were all taking medication for their ADHD symptoms during the test procedures, which likely reduced the ability of the CPT to sensitively detect inherent impulsive response tendencies. Again, future studies should consider examining the relationship between impulsivity and emotional reactivity in ADHD children on and off stimulant medication. Summary In summary, both ADHD and healthy control children show highly reliable aversion enhanced startle modulation. Both groups also appear to be similarly successful at cognitively appraising their emotional experience associated with viewing affective picture stimuli. However, ADHD and control children show different patterns of startle

PAGE 162

148 reactivity across different types of unpleasant picture stimuli. Specifically, both ADHD and healthy control children show highly reliable aversion enhanced startle modulation to direct threat and injury pictures. However, only controls show potentiated startle to victim pictures, and in contrast, ADHD children display inhibited startle reactivity to victim pictures. This occurs despite ADHD children showing appropriate physiological arousal while viewing victim pictures, and being successful at cognitively appraising their emotional experience associated with viewing victim pictures. This “mismatch” or lack of correspondence between different components of emotional processing suggests that children with ADHD have a defect in translating the results of cognitive appraisal into somato-motor changes associated with unpleasant emotional stimuli that specifically depict scenes of victimization. These findings provide support for the existence of emotional dysregulation in children with ADHD. Moreover, ADHD children with and without comorbid conduct-related and psychopathy symptoms show highly reliable aversion enhanced startle modulation to direct threat pictures, and inhibited startle modulation to victim pictures, compared to pleasant pictures. Still, the severity of these comorbid conduct-related (i.e., CD and ODD) and psychopathy symptoms impact emotional reactivity within ADHD. In addition, they appear to impact different components of emotional processing in ADHD. Specifically, ODD symptomatology impacts physiological arousal (i.e., lower skin conductance) whereas CD and psychopathy symptomatology impact affective startle modulation (i.e., reduced startle response to injury and victim pictures). Furthermore, ADHD children with high levels of CD and psychopathy symptomatology display similar patterns of startle reactivity to adults with psychopathy (Herpertz et al., 2001; Levenston

PAGE 163

149 et al., 2000; Patrick et al., 1993). These findings suggest that affective modulation of the startle reflex may not only be valuable for differentiating between ADHD children with and without comorbid Conduct Disorder, but also may represent an early biologic marker for increased risk of developing psychopathy in adulthood. Methodological Limitations There are several methodological limitations that potentially impacted the findings of the study. First, as discussed earlier, the ADHD children were all taking medication for their ADHD symptoms during the test procedures. This may have facilitated efficiency of attentional processing while viewing picture stimuli, thereby decreasing the inhibitory effect of attention or decreasing the threshold for the shift from an orienting/attentional disposition to one of defensive activation in ADHD. As a result, if children with ADHD were off medication, additional (i.e., with direct threat or injury pictures) or larger (i.e., with victim pictures) differences between ADHD and controls may have emerged for startle reactivity. In addition, medication also likely reduced the ability of the CPT to sensitively detect inherent impulsive response tendencies in ADHD. Furthermore, parent report of impulsivity (i.e., Connors’ Restless-Impulsive Index and DSM-IV Hyperactive-Impulsive scale) was based on observed behaviors in general, and not specific to off-medication behavior. Therefore, both CPT and Connors’ parent report may not sensitively measure underlying behavioral and cognitive inhibition or impulsivity. Therefore, the lack of a relationship between impulsivity and emotional reactivity should be viewed cautiously. Next, the current sample of ADHD/CD children consisted of children with mild to moderate conduct disorder, but none met criteria for the severe form of CD. This may have reduced the power and sensitivity to detect further conduct disorder related impact

PAGE 164

150 (i.e., may have found differences on direct threat pictures or even further reductions in startle response to injury and victim pictures). Therefore, it cannot be ruled out that children with the most severe forms of comorbid Conduct Disorder would show an even more aberrant pattern of affective modulation of the startle reflex. Finally, the different unpleasant picture categories are likely not matched on complexity with respect to foreground properties and social interaction. Therefore, it is difficult to determine the source of the aberrant startle response pattern in ADHD (i.e., inhibited response to victim pictures). Specifically, the study is unable to answer whether this finding is related to 1) purely attention-based processes, 2) purely emotion-based processes, 3) emotion dysfunction leading to increased attentional demands for victim pictures, which in turn leads to startle inhibition, or 4) attentional dysfunction leading to emotional dysfunction, which in turn leads to startle inhibition of victim pictures. Future Research Directions There are a variety of potentially productive directions to take this line of research. First, for reasons described earlier, future studies should examine affective modulation of children with ADHD on and off medication. Next, future research should attempt to recruit ADHD children with more severe forms of comorbid ADHD. Along these lines, future studies should also examine children with Conduct Disorder with and without comorbid ADHD, to better evaluate the relative impact of ADHD versus CD on emotional reactivity. In addition, as the children with ADHD/CD in the current sample were almost all male (i.e., 10/11), future research should attempt to include more females with CD. To accomplish this, the age range would likely need to be increased up to age 18.

PAGE 165

151 Future research should also directly examine the attentional properties of the picture stimuli in ADHD on and off medication using pre-pulse inhibition and event-related potential (ERP) techniques. Another approach is to develop a picture set that systematically varies the degree of attentional or social complexity within a picture subcategory. For instance, the direct threat pictures in the current study incorporated a weapon being pointed out toward the viewer (participant) typically being held by one person in the picture. The attentional and cognitive-social demands of direct threat pictures could be increased by having one individual (person #1) continue to direct a weapon directed out at the viewer/participant, but also have a second individual in the picture pointing a weapon toward person #1 in the picture. This type of picture would continue to be directly threatening toward the viewer, but also incorporate a social interaction with potentially greater attentional demands. Along these lines, it would be interesting to develop a set of unpleasant pictures that were purposefully ambiguous in terms of perceived threat or danger. These ambiguous pictures could lead to potentially fruitful findings with respect to the thresholds for defensive activation in healthy children and children with ADHD, CD, ODD, and psychopathy. Next, it would be useful to examine emotional reactivity in children at even younger ages (prior to age 7). In particular, a study should target children who have not yet met criteria for ADHD, ODD, or CD, but are at high risk for developing these disorders based on genetic vulnerability. In addition, these high-risk children could be followed longitudinally to truly test the predictive value of affective modulation of the startle reflex. Another interesting research direction would be to examine emotional reactivity before and after behavioral and/or medication treatment in children with

PAGE 166

152 ADHD and/or conduct problems. Specifically, this type of study could address whether patterns of emotional reactivity within children are static or dynamic, and whether emotional reactivity could be useful as an indicator of treatment response. Overall, affective modulation of the startle reflex has been shown to be a valuable tool for increasing our understanding of emotional processes in normal and behaviorally disordered children. Still, future research and development of affective modulation of the startle reflex as an emotion measure is necessary to determine its clinical utility for identifying high-risk children, and potentially allowing for earlier and better treatment strategies for ADHD children with and without comorbid CD, ODD, and psychopathy symptomatology.

PAGE 167

APPENDIX A DSM-IV DIAGNOSTIC CRITERIA FOR ADHD, CD, & ODD DSM-IV (1994) Diagnostic Criteria for ADHD A. Either 1 or 2 1) Six (or more) of the following symptoms of inattention have persisted for at least 6 months to a degree that is maladaptive and inconsistent with developmental level: Inattention a) Often fails to give close attention to details or makes careless mistakes in schoolwork, work, or other activities b) Often has difficulty sustaining attention in tasks or play activities c) Often does not seem to listen when spoken to directly d) Often does not follow through on instructions and fails to finish schoolwork, chores, or duties in the workplace (not due to oppositional behavior or failure to understand instructions) e) Often has difficulty organizing tasks and activities f) Often avoids, dislikes, or is reluctant to engage in tasks that require sustained mental effort (such as schoolwork or homework) g) Often loses things necessary for tasks or activities (eg, toys, school assignments, pencils, books, or tools) h) Is often easily distracted by extraneous stimuli i) Is often forgetful in daily activities 2) Six (or more) of the following symptoms of hyperactivity-impulsivity have persisted for at least 6 months to a degree that is maladaptive and inconsistent with developmental level: Hyperactivity a) Often fidgets with hands or feet or squirms in seat b) Often leaves seat in classroom or in other situations in which remaining seated is expected c) Often runs about or climbs excessively in situations in which it is inappropriate (in adolescents or adults, may be limited to subjective feelings of restlessness) d) Often has difficulty playing or engaging in leisure activities quietly e) Is often "on the go" or often acts as if "driven by a motor" f) Often talks excessively Impulsivity g) Often blurts out answers before questions have been completed h) Often has difficulty awaiting turn i) Often interrupts or intrudes on others (eg, butts into conversations or games) B. Some hyperactive-impulsive or inattentive symptoms that caused impairment were present before 7 years of age. C. Some impairment from the symptoms is present in 2 or more settings (eg, at school [or work] or at home). D. There must be clear evidence of clinically significant impairment in social, academic, or occupational functioning. E. The symptoms do not occur exclusively during the course of a pervasive developmental disorder, 153

PAGE 168

154 schizophrenia, or other psychotic disorder and are not better accounted for by another mental disorder (eg, mood disorder, anxiety disorder, dissociative disorder, or personality disorder). DSM-IV (1994) Diagnostic Criteria for Conduct Disorder A. A repetitive and persistent pattern of behavior in which the basic rights of others or major age-appropriate societal norms or rules are violated, as manifested by the presence of three (or more) of the following criteria in the past 12 months, with at least one criterion present in the past 6 months: Aggression to people and animals a) often bullies, threatens, or intimidates others b) often initiates physical fights c) has used a weapon that can cause serious physical harm to others (e.g., a bat, brick, broken bottle, knife, gun) d) has been physically cruel to people e) has been physically cruel to animals f) has stolen while confronting a victim (e.g., mugging, purse snatching, extortion, armed robbery) g) has forced someone into sexual activity Destruction of property h) has deliberately engaged in fire setting with the intention of causing serious damage i) has deliberately destroyed others' property (other than by fire setting) Deceitfulness or theft j) has broken into someone else's house, building, or car k) often lies to obtain goods or favors or to avoid obligations (i.e., "cons" others) l) has stolen items of nontrivial value without confronting a victim (e.g., shoplifting, but without breaking and entering; forgery) Serious violations of rules m) often stays out at night despite parental prohibitions, beginning before age 13 years n) has run away from home overnight at least twice while living in parental or parental surrogate home (or once without returning for a lengthy period) o) is often truant from school, beginning before age 13 years B. The disturbance in behavior causes clinically significant impairment in social, academic, or occupational functioning. C. The disturbance in behavior causes clinically significant impairment in social, academic, or occupational functioning. D. If the individual is age 18 years or older, criteria are not met for Antisocial Personality Disorder. Specify type based on age at onset: Childhood-Onset Type: onset of at least one criterion characteristic of Conduct Disorder prior to age 10 Adolescent-Onset Type: absence of any criteria characteristic of Conduct Disorder prior to age 10 Specify severity: Mild: few if any conduct problems in excess of those required to make the diagnosis and conduct proble m cause only minor harm to others Moderate: number of conduct problems and effect on others intermediate between "mild" and "severe" Severe: many conduct problems in excess of those required to make the diagnosis or conduct problems cause considerable harm to others

PAGE 169

155 DSM-IV (1994) Diagnostic Criteria for Oppositional Defiant Disorder A pattern of negativistic, hostile, and defiant behavior lasting at least 6 months, during which four (or more) of the following are present: a) often loses temper b) often argues with adults c) often actively defies or refuses to comply with adults' requests or rules d) often deliberately annoys people e) often blames others for his or her mistakes or misbehavior f) is often touchy or easily annoyed by others g) is often angry and resentful h) is often spiteful or vindictive Note: Consider a criterion met only if the behavior occurs more frequently than is typically observed in individuals of comparable age and developmental level. The disturbance in behavior causes clinically significant impairment in social, academic, or occupational functioning. The behaviors do not occur exclusively during the course of a Psychotic or Mood Disorder. Criteria are not met for Conduct Disorder, and, if the individual is age 18 years or older, criteria are not met for Antisocial Personality Disorder.

PAGE 170

APPENDIX B SELF-ASSESSMENT MANIKIN 156

PAGE 171

APPENDIX C SELF-ASSESSMENT MANIKIN INSTRUCTIONS Introduction “SAM is to help you tell me how you were feeling while you were looking at or seeing a picture. Let’s look at a copy of SAM.” Valence “Looking at the page, you see that the left side of the page has seven pictures of SAM going the top of the page to the bottom of the page (pointing). Notice that at the very top, SAM has a big smile, and at the very bottom, SAM has a big frown. The pictures on the left go from a very happy or pleasant SAM to a very unhappy or unpleasant SAM. You can see that there is a number 7 at the top matching up with the very happy or pleasant SAM. This is the number you would say if the slide you had just seen made you feel very good, glad, pleasant, happy, or cheerful. Now if you look next the number 1 over here, you can see a picture of SAM frowning. This is what you would say if the slide made you feel very bad, unhappy, upset, unpleasant, yucky, scared, angry, or sad. If you feel neutral or in the middle, that is you didn’t feel either pleasant or unpleasant, then you would say the number 4, which matches with the SAM that is straight-faced. Also, if you felt in between very happy and neutral, you could say 5 or 6. For example, if the picture made you feel just a little bit happy you could say 5, or if it made you feel medium or pretty happy, you would say 6, and again, if you felt very pleasant or happy, you would a 7. And if the slide made you feel just a little bit unpleasant, yucky, or scared you would put a 3, or medium or pretty unpleasant you would put a 2, and so on. Does that make sense, do you have any questions so far?” Arousal “Now, let’s look at the second feeling. This feeling is a little trickier to understand, so please tell me if you get confused or this does not make sense. This feeling is on the right side. You can see that the pictures of SAM look different from the smiling to frowning SAM. These pictures are in order from a very relaxed or calm SAM to a very excited or nervous SAM. You can see next to number 7, SAM is jumping up and down and his stomach is very excited. This is like when you get excited and can’t sit still or like when you have butterflies in your stomach when you are very nervous. So, if the picture made you feel very excited, nervous, scared, jittery, active, wide awake, or enthusiastic, you would say the number 7. Now look at the picture next to the number 1, you can see that SAM is very still and his eyes are closed. This is what you would circle if the picture made you feel very calm, bored, relaxed, or sleepy. This is a little bit different from the smiling/frowning SAM because there is no neutral SAM on this side. But like the other SAM, you can mark in between very excited and very calm. So if the 157

PAGE 172

158 picture made you medium excited, you would say 4, and if it made you just a little bit excited you could put a 2, or a lot excited you may put 5 or 6, or very excited a 7,etc. Does that make sense? Also remember to always give me the number for the left side of the slide first for happy or unhappy, and then for the right side of the slide for how excited or nervous you were.” Practice “Ok, let me test you to make sure you all understand. What number would you say on the left side of the page if the picture made you feel very unpleasant or unhappy? And what number would you put if that same picture made you feel extremely nervous or excited? Now, we will practice using SAM on three slides. Remember, you will see each picture for about 6 seconds. Make sure you look at each picture the whole time it is showing, no talking when the picture is showing, and think about how you are feeling while you are seeing the picture. Then the picture will go away and you will see a blank screen for about 6 seconds. Then SAM will come on the screen. Then you are to first say a number for how happy or unhappy the picture made you feel, and then the a number for how excited or nervous the picture made you feel. Also, there are no right or wrong answers, just choose numbers for how you felt when you were looking at the picture.” Show first practice picture and say “What two numbers or ratings would you say for this picture.” Proceed to next practice slide if child appeared to understand rating system, if not, review directions again. Proceed to remaining two practice slides.

PAGE 173

APPENDIX D EXAMPLE PICTURE TRIAL 6 seconds (100 Db) Startle Probe Onset: 4.2s, 5.0s, or 5.8s Counterbalanced across Time, Category, and Subcategory 6 secVariable: 4-12 sec6 secSelf-Report Ratings VeryBad, Unhappy, Upset Very Good, Glad, HappyNeutral76543217654321 Very Bored, Calm, Relaxed, or SleepyVery, VeryExcited,Aroused, Nervous, or Scared MediumExcited,Aroused, Nervous, or Scared VeryBad, Unhappy, Upset Very Good, Glad, HappyNeutral76543217654321 7654321 Very Bored, Calm, Relaxed, or SleepyVery, VeryExcited,Aroused, Nervous, or Scared MediumExcited,Aroused, Nervous, or Scared 6 seconds (100 Db) Startle Probe Onset: 4.2s, 5.0s, or 5.8s Counterbalanced across Time, Category, and Subcategory 6 secVariable: 4-12 sec6 secSelf-Report Ratings VeryBad, Unhappy, Upset Very Good, Glad, HappyNeutral76543217654321 Very Bored, Calm, Relaxed, or SleepyVery, VeryExcited,Aroused, Nervous, or Scared MediumExcited,Aroused, Nervous, or Scared VeryBad, Unhappy, Upset Very Good, Glad, HappyNeutral76543217654321 7654321 Very Bored, Calm, Relaxed, or SleepyVery, VeryExcited,Aroused, Nervous, or Scared MediumExcited,Aroused, Nervous, or Scared VeryBad, Unhappy, Upset Very Good, Glad, HappyNeutral76543217654321 Very Bored, Calm, Relaxed, or SleepyVery, VeryExcited,Aroused, Nervous, or Scared MediumExcited,Aroused, Nervous, or Scared VeryBad, Unhappy, Upset Very Good, Glad, HappyNeutral76543217654321 7654321 Very Bored, Calm, Relaxed, or SleepyVery, VeryExcited,Aroused, Nervous, or Scared MediumExcited,Aroused, Nervous, or Scared VeryBad, Unhappy, Upset Very Good, Glad, HappyNeutral76543217654321 Very Bored, Calm, Relaxed, or SleepyVery, VeryExcited,Aroused, Nervous, or Scared MediumExcited,Aroused, Nervous, or Scared VeryBad, Unhappy, Upset Very Good, Glad, HappyNeutral76543217654321 7654321 Very Bored, Calm, Relaxed, or SleepyVery, VeryExcited,Aroused, Nervous, or Scared MediumExcited,Aroused, Nervous, or Scared 159

PAGE 174

LIST OF REFERENCES American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Balaban, M.T. (1995). Affective influences on startle in five-month-old infants: Reactions to facial expressions of emotion. Child Development, 66, 28-36. Babinski, L., Harsough, C., & Lambert, N. (1999). Childhood conduct problems, hyperactivity-impulsivity, and inattention as predictors of adult criminal activity. Journal of Child Psychology and Psychiatry, 40, 347-355. Barkley, R.A. (1997). Attention-deficit/hyperactivity disorder, self-regulation, and time: toward a more comprehensive theory. Journal of Developmental and Behavioral Pediatrics, 18 (4), 271-9. Barkley, R. (1998). Attention-deficit hyperactivity disorder: A handbook for diagnosis and treatment (2nd ed.). New York: Guilford Press. Barkley, R.A., Edwards, G., Laneri, M., Fletcher, K., & Metevia, L. (2001). Executive functioning, temporal discounting, and sense of time in adolescents with attention deficit hyperactivity disorder (ADHD) and oppositional defiant disorder (ODD). Journal of Abnormal Child Psychology, 29 (6), 541-56. Barkley, R.A., Grodzinsky, G., & DuPaul, G.J. (1992). Frontal lobe functions in attention deficit disorder with and without hyperactivity: a review and research report. Journal of Abnormal Child Psychology, 20 (2), 163-88. Berquin, P.C., Giedd, J.N., Jacobsen, L.K., Hamburger, S.D., Krain, A.L., Rapoport, J.L., & Castellanos, F.X. (1998). Cerebellum in attention-deficit hyperactivity disorder: a morphometric MRI study. Neurology, 50 (4), 1087-93. Baumgardner, T.L., Singer, H.S., Denckla, M.B., Rubin, M.A., Abrams, M.T., Colli, M.J., Reiss, A.L. (1996). Corpus callosum morphology in children with Tourette syndrome and attention deficit hyperactivity disorder. Neurology, 47 (2), 477-82 Biederman, J., Faraone, S., Mick, E., Spencer, T., Wilens, T., Kiely, K., Guite, J., Ablon, J.S., Reed, E., & Warburton. (1995). High risk for attention deficit hyperactivity disorder among children of parents with childhood onset of the disorder: A pilot study. American Journal of Psychiatry, 152, 431-435. 160

PAGE 175

161 Biederman, J., Faraone, S.V., Milberger, S., Jetton, J.G., Chen, L., Mick, E., Greene, R.W., & Russell, R.L. (1996). Is childhood oppositional defiant disorder a precursor to adolescent conduct disorder? Findings from a four-year follow-up study of children with ADHD. Journal of the American Academy of Child and Adolescent Psychiatry, 35, 1193-1204. Bowers D, Gilmore R, Bauer R, & Rogish M. (in press). Affect modulation o f startle before and after temporal resections involving the amygdala. Biological Psychiatry. Braaten, E.B., & Rosen, L.A. (2000). Self-regulation of affect in attention deficit-hyperactivity disorder (ADHD) and non-ADHD boys: Differences in empathic responding. Journal of Consulting and Clinical Psychology, 68 (2), 313-321. Bradley, M.M. (2000). Emotion and motivation. In L.G. Tassinary, J.T. Cacioppo, & G. Berntson (eds.) Handbook of psychophysiology. Stanford, CA: Cambridge University Press. Bradley, M.M., Codispoti, M., Cuthbert, B.N., & Lang, P.J. (2001). Emotion and motivation I: Defensive and appetive reactions in picture processing. Emotion, 1 (3), 276-298. Bradley, M.M., Cuthbert, B.N., & Lang, P.J. (1990). Startle reflex modification: Emotion or attention? Psychophysiology, 27, 513-522. Bradley, M.M., Cuthbert, B.N., & Lang, P.J. (1993). Pictures as prepulse: Attention and emotion in startle modification. Psychophysiology, 30, 541-545. Bradley, M.M., Cuthbert, B.N., & Lang, P.J. (1999). Affect and the startle reflex. In Dawson, M.E., Schell, A.M., Bohmelt, A.H. (eds.) Startle modification: Implications for neuroscience, cognitive science, and clinical science. Stanford, CA: Cambridge University Press. Bradley, M.M., & Lang, P.J. (1994). Measuring emotion: The self-assessment manikin and the semantic differential. Journal of Behavioral Therapy and Experimental Psychiatry, 25, 49-59. Bradley, M.M., & Lang, P.J. (2000). Affective reactions to acoustic stimuli. Psychophysiology, 37, 204-215. Brodeur, D.A., Pond, M. (2001). The development of selective attention in children with attention deficit hyperactivity disorder. Journal of Abnormal Child Psychology, 29 (3), 229-39. Brown, R.T., Freeman, W.S., Perrin, J.M., Stein, M.T., Amler, R.W., Feldman, H.M., Pierce, K., & Wolraich, M.L. (2001) Prevalence and assessment of attention-deficit/hyperactivity disorder in primary care settings. Pediatrics, 107 (3), 43.

PAGE 176

162 Burke, J.D., Loeber, R, Lahey, B.B., & Birmaher, B. (2000). Oppositional defiant and conduct disorder: A review of the past 10 years, part II. Journal of the American Academy of Child and Adolescent Psychiatry, 41:11, 1275-1293. Carter, C.S., Krener, P., Chaderjian, M., Northcutt, C., & Wolfe V. (1995). Abnormal processing of irrelevant information in attention deficit hyperactivity disorder. Psychiatry Research, 56 (1), 59-70. Castellanos, F. (1997). Toward a pathophysiology of attention-deficit/hyperactivity disorder. Clinical Pediatrics, 36, 381-393. Castellanos, F.X. (2001). Neural substrates of attention-deficit hyperactivity disorder. Advances in Neurology, 85, 197-206. Castellanos, F.X., Giedd, J.N., Marsh, W.L., Hamburger, S.D., Vaituzis, A.C., Dickstein, D.P., Sarfatti, S.E., Vauss, Y.C., Snell, J.W., Lange, N., Kaysen, D., Krain, A.L., Ritchie, G.F., Rajapakse, J.C., & Rapport, L.J. (1996). Quantitative brain magnetic resonance imaging in attention-deficit hyperactivitydisorder. Archives of General Psychiatry, 53 (7), 607-16. Chhabildas, N., Pennington, B.F., Willcutt, E.G. (2001). A comparison of the neuropsychological profiles of the DSM-IV subtypes of ADHD. Journal of Abnormal Child Psychology, 29 (6), 529-40. Conners, C. (1997). Conners' Rating Scales-Revised: Technical Manual . North Tonawanda, NY: Multi-Health Systems. Conners, C. (2001). Conners' Continuous Performance Test . North Tonawanda, NY: Multi-Health Systems. Cook, E.I., Hawk, L., Hawk, T., & Hummer, K. (1995). Affective modulation of startle in children [Abstract]. Psychophysiology, 32 (Suppl.1), S25. Cooley, E., & Morris, D. (1990). Attention in Children: A neuropsychologically based model for assessment. Developmental Neuropsychology, 6 (3), 239-274. Corkum, P.V., & Siegel, L.S. (1993). Is the continuous performance task a valuable research tool for use with children with Attention-Deficit-Hyperactivity Disorder? Journal of Child Psychology and Psychiatry, 34 (7), 1217-39. Cuthbert, B.N., Bradley, M.M., & Lang, P.J. (1996). Probing picture perception: Activation and emotion. Psychophysiology, 33, 103-111. Davis M. (1998). Are different parts of the extended amygdala involved in fear versus anxiety? Biological Psychiatry, 44, 1239-47.

PAGE 177

163 Delamater, A., & Lahey, B. (1983). Physiol ogical correlates of conduct problems in hyperactive and learning disabled children . Journal of Abnormal Child Psychology, 11, 311-320. Douglas, V.I., & Parry, P.A. (1983). E ffects of reward on delayed reaction time task performance of hyper active children. Journal of Abnormal Child Psychology, 11 (2), 313-326. Farrington, D., Loeber, R., & Van Ka mmen, W. (1989). Longterm criminal outcomes of hyperactivity-impulsivity-att ention deficit and conduct problems in childhood. In L Robins & M Rutter (Eds.), Stra ight and devious pathways to adulthood . NewYork: Cambridge University Press. Filipek, P.A., Semrud-Clikeman, M., Stei ngard, R.J., Renshaw, P.F., Kennedy, D.N., & Biederman, J. (1997). Volumetric MRI analysis comparing subjects having attention-deficit hyperactivity disord er with normal controls. Neurology, 48 (3), 589-601. Frick, P. (1988). Conduc t disorders. In T. Ollendick & M. Hersen (Eds.), Handbook of child psychopathology (3rd ed.) . New York: Plenum. Garralda, M.E., Connell, J., & Taylor, D. C. (1991). Psychophysiological anomalies in children with emotional and conduc t disorders. Psychol ogical Medicine, 21 (4), 94757. Giedd, J.N., Castellanos, F.X., Casey, B. J., Kozuch, P., King, A.C., Hamburger, S.D., & Rapoport, J.L. (1994). Quantita tive morphology of the corpus callosum in attention deficit hyperactiv ity disorder. American J ournal of Psychiatry, 151 (5), 665-9. Goldstein, D.J., & Blumenthal, T.D. (1995). Startle eyeblink elicit ation in attention deficit disordered children using low-intens ity acoustic stimuli. Perceptual and Motor Skills, 80, 227-231. Greenwald, M.K, Cook III, E.W., & Lang, P.J. (1989). Affective judgment and psychophysiological response: Dimensional cova riation in the eval uation of pictorial stimuli. Journal of Psychophysiology, 3, 51-64. Grillon, C., Ameli, R., Goddard, A., Woods , S.W., & Davis, M. (1994). Baseline and Fear-Potentiated Startle in Panic Disorder Patients. Society of Biological Psychiatry, 35, 431-439. Grillon, C., Dierker, L., & Merikangas, K. R. (1997). Startle modulation in children at risk for anxiety disorders and/or alcoho lism. Journal of the American Academy of Child and Adolescent Psychiatry, 36, 925-932.

PAGE 178

164 Grillon C., Merikangas K.R., Dierker L., Snidman N., Arriaga R.I., Kagan J., Donzella B., Dikel T., & Nelson C. (1999). Startle potentiation by threat of aversive stimuli and darkness in adolescents: A multi-site study. International Journal of Psychophysiology, 32, 63-73. Grillon, C., Pellowski, M., Merikangas, K.R., & Davis, M. (1997). Darkness facilitates the acoustic startle reflex in humans. Society of Biological Psychiatry, 42, 453-460. Haenlein, M., & Caul, W.F. (1987). Attention deficit disorder with hyperactivity: A specific hypothesis of reward dysfunction. Journal of the American Academy of Child and Adolescent Psychiatry, 26 (3), 356-62. Hamm, A.O., Cuthebert, B.N., Globisch, J., & Vaitl, D. (1997). Fear and the startle reflex: Blink modulation and autonomic response patterns in animal and mutilation fearful subjects. Psychophysiology, 34, 97-107. Hawk, L.W., & Cook E.W., III (1997). Affective modulation of tactile startle. Psychophysiogy, 34 , 23-31. Hare, R.D. (1991). Manual for the Hare Psychopathy Checklist –Revised. North Tonawanda, NY: Multi-Health Systems. Heaton, S.C., Reader, S.K., Preston, A.S., Fennell, E.B., Puyana, O.E., Gill, N., & Johnson, J.H. (2001). The test of everyday attention for children (TEA-Ch): Patterns of performance in children with ADHD and clinical controls. Child Neuropsychology, 7 (4), 251-264. Herpertz, S.C., Wenning, B., Mueller, B., Qunaibi, M., Sass, H., & Herpertz-Dahlmann, B. (2001). Psychophysiological responses in ADHD boys with and without conduct disorder: Implications for adult antisocial behavior. Journal of the American Academy of Child and Adolescent Psychiatry, 40:10, 1222-1228. Herpertz, S.C., Werth, U., Lukas, G., Qunaibi, M., Schuerkens, A., Kunert, H., Freese, R., Flesch, M., Mueller-Isberner, R., Osterheider, M., & Sass, H. (2001). Emotion in criminal offenders with psychopathy and borderline personality disorder. Archives of General Psychiatry, 58 , 737-745. Houghton, S., Douglas, G., West, J., Whiting, K., Wall, M., Langsford, S., Powell, L., & Carroll, A. (1999). Differential patterns of executive function in children with attention-deficit hyperactivity disorder according to gender and subtype. Journal of Child Neurology, 14 (12), 801-5. Hynd, G.W., Hern, K.L., Novey, E.S., Eliopulos, D., Marshall, R., Gonzalez, J.J., & Voeller, K.K. (1993). Attention deficit-hyperactivity disorder and asymmetry of the caudate nucleus. Journal of Child Neurology, 8 (4), 339-47.

PAGE 179

165 Hynd, G.W., Semrud-Clikeman, M., Lorys, A.R., Novey, E.S., Eliopulos, D., & Lyytinen, H. (1991). Corpus callosum morphology in attention deficit-hyperactivity disorder: Morphometric analysis of MRI. Journal of Learning Disabilities, 24 (3), 141-6. Jansen, D.M., & Frijda, N.H. (1994). Modulation of the acoustic startle response by film-induced fear and sexual arousal. Psychophysiology, 31, 565-571. Kaplan H, & Sadock J. (1995). Comprehensive textbook of psychiatry (7th ed.). Baltimore: Williams & Wilkins. Kaplan H, & Sadock J. (1998). Synopsis of psychiatry: Behavioral sciences, clinical psychiatry (8th ed). Baltimore: Williams & Wilkins. Keppel, G. (1982). Design and analysis: A researcher’s handbook. (2nd ed.). New Jersey: Prentice-Hall. Kerns, K.A., McInerney, R.J., & Wilde, N.J. (2001). Time reproduction, working memory, and behavioral inhibition in children with ADHD. Child Neuropsychology, 7 (1), 21-31. Koch, M., & Schnitzler, H. (1997). The acoustic startle response in rats-circuits mediating evocation, inhibition and potentiation. Behavioral Brain Research, 89, 35-49. Kruesi, M., Casanova, M.F., Mannheim, G., & Johnson-Bilder, A. (2004). Reduced temporal lobe volume in early onset conduct disorder. Psychiatry Research: Neuroimaging, 132, 1-11. Lang, P. (1995). The emotion probe: Studies of motivation and attention. American Psychologist, 50, 372-384. Lang, P., Bradley, M., & Cutherbert, B. (1990). Emotion, attention, and the startle reflex. Psychological Review, 97, 377-395. Lang, P.J., Bradley, M.M., & Cutherbert, B.N. (1997). Motivated attention: Affect, activation, and action. In Lang, P.J., Simons, R.F., Balaban, M.T. (eds.) Attention and orienting: Sensory and motivational processes. Mahwah, NJ: Lawrence Erlbaum Associates. Lang, P., Bradley, M., & Cuthbert, B. (1999). International affective picture system (IAPS): Instruction manual and affective ratings. Technical Report A-4, The Center for Research in Psychophysiology, University of Florida. Lang, P., Greenwald, M., Bradley, M., & Hamm, A. (1993). Looking at pictures: Affective, facial, visceral, and behavioral reactions. Psychophysiology, 30, 261-373.

PAGE 180

166 Levenston, G.K., Patrick, C.J., Bradley, M.M., & Lang, P.J. (2000). The Psychopath as Observer: Emotion and Attention in Picture Processing. Journal of Abnormal Psychology, 109, 373-385. Loeber, R., & Dishion, T. (1983). Early predictors of male delinquency: A review. Psychology Bulletin, 94, 68-99. Loeber, R, Burke, J.D., Lahey, B.B., Winters, A., & Zera, M. (2000). Oppositional defiant and conduct disorder: A review of the past 10 years, part 1. Journal of the American Academy of Child and Adolescent Psychiatry, 39:12 , 1468-1484. Lueger, R., & Gill, K. (1990). Frontal-lobe cognitive dysfunction in conduct disorder adolescents. Journal of Clinical Psychology, 46, 696-706. Luengo, M., Carrillo-de-la-Pena, M., Otero, J., & Romero, E. (1994). A short-term longitudinal study of impulsivity and antisocial behavior. Journal of Personality and Social Psychology, 3, 542-548. Lynam, D. (1996). Early identification of chronic offenders: Who is the fledgling psychopath. Psychology Bulletin, 120, 209-234. Lynam, D. (1997). Pursuing the psychopath: Capturing the fledgling psychopath in a nomological net. Journal of Abnormal Psychology, 106, 425-438. Lynam, D. (1998). Early identification of the fledgling psychopath: Locating the psychopathic child in the current nomenclature. Journal of Abnormal Psychology, 107, 566-575. Maliphant, R., Hume, F., & Furnham, A. (1990). Autonomic nervous system (ANS) activity, personality characteristics and disruptive behaviour in girls. Journal of Child Psychology and Psychiatry, 31, 619-628. McManis, M.H., Bradley, M.M., Berg, K., Cuthbert, B.N., & Lang, P.J. (2001). Emotional reactions in children: Verbal, physiological, and behavioural responses to affective pictures. Psychophysiology, 38, 222-231. Miltner W. (1994). Emotional qualities of odors and their influence on the startle reflex in humans. Psychophysiology, 31, 107-110. Mirsky, A.F., Anthony, B.J., Duncan, C.C., Ahearn, M.B., & Kellam S.G. (1991). Analysis of the elements of attention: A neuropsychological approach. Neuropsychology Review, 2 (2), 109-45. Morris, M., Bradley, M., Bowers, D., Lang, P., & Heilman, K. (1991). Valence-specific hypoarousal following right temporal lobectomy. Paper presented at the 19th Annual Meeting of the International Neuropsychological Society, San Antonio, Texas.

PAGE 181

167 Mostofsky, S.H., Cooper, K.L., Kates, W.R., Denckla, M.B., & Kaufmann, W.E. (2002). Smaller prefrontal and premotor volumes in boys with attention-deficit/hyperactivity disorder. Biological Psychiatry, 52 (8), 785-94. Mostofsky, S.H., Reiss, A.L., Lockhart, P., Denckla, M.B. (1998). Evaluation of cerebellar size in attention-deficit hyperactivity disorder. Journal of Child Neurology, 13 (9), 434-9. Oosterlaan, J., Logan, G.D., & Sergeant, J.A. (1998). Response inhibition in AD/HD, CD, comorbid AD/HD + CD, anxious, and control children: A meta-analysis of studies with the stop task. Journal of Child Psychology and Psychiatry, 39 (3), 411-25. Ornitz, E.M., Gabikian, P., Russell, A.T., Guthrie, D., Hirano, & C., Gehricke, J. (1997). Affective valence and arousal in ADHD and normal boys during a startle habituation experiment. Journal of the American Academy of Child and Adolescent Psychiatry, 36, 1698-1705. Ornitz, E.M., Guthrie, D., Sadeghpour, M., & Sugiyama, T. (1991). Maturation of pre-stimulation-induced startle modulation in girls. Psychophysiology, 28, 11-20. Ornitz, E.M., Hanna, G.L., & de Traversay, J. (1992). Prestimulation-induced startle modulation in attention-deficit hyperactivity disorder and nocturnal enuresis. Psychophysiology, 29, 437-451. Ornitz, E.M., & Pynoos, R.S., (1989). Startle modulation in children with posttraumatic stress disorder. American Journal of Psychiatry, 146, 866-870. Ornitz, E.M., Russell, A.T., Yuan, H., & Liu M. (1996). Autonomic, electroencephalographic, and myogenic activity accompanying startle and its habituation during mid-childhood. Psychophysiology, 33, 507-513. Osgood, C. (1969). The nature and measurement of meaning. In J.B. Snider &C.E. Osgood (Eds.) Semantic differential technique. Chicago: Aldine. Parry, P.A., & Douglas, V.I. (1983). Effects of reinforcement on concept identification in hyperactive children. Journal of Abnormal Child Psychology, 11 (2), 327-340. Patrick, J.C., Bradley, M.M., & Lang, P.J. (1993). Emotion in the criminal psychopath: Startle reflex modulation. Journal of Abnormal Psychology, 102, 82-92.

PAGE 182

168 Pelham, W., Milich, R., Cummings, M., Murphy, D., Schaughency, E.A., & Greiner, A.R. (1991). Effects of background anger, provocation, and methylphenidate on emotional arousal and aggressiveresponding in attention-deficit hyperactivity disordered boys with and without concurrent aggressiveness. Journal of Abnormal Child Psychology, 19, 407-426. Quay H. (1997). Inhibition and attention deficit hyperactivity disorder. Journal of Abnormal Child Psychology, 25, 7-14. Raine, A. (1993). The psychopathology of crime: Criminal behavior as a clinical disorder . San Diego: Academic Press. Raine, A., Venables, P., & Williams, M. (1990). Relationships between central and autonomic measures of arousal at age 15 years and criminality at age 24 years. Archives of General Psychiatry, 47, 1003-7. Raine, A., Venables, P., & Williams, M. (1995). High autonomic arousal and electrical orienting at age 15 years as protective factors against criminal behavior at age 29 years. American Journal of Psychiatry, 152, 1595-1600. Raine, A., & Venables, H. (1984). Tonic heart rate level, social class, and antisocial behavior in adolescents. Biological Psychology, 18, 123-132. Rapport, L.J., Friedman, S.L., Tzelepis, A., & Van Voorhis, A. (2002). Experienced emotion and affect recognition in adult attention-deficit hyperactivity disorder. Neuropsychology, 16 (1), 102-110. Robins, L. (1966). Deviant children grown up: A sociological and psychiatric study of sociopathic personality . Baltimore: Williams &Wilkins. Robins, L. (1978). Study childhood predictors of adult antisocial behavior: replications from longitudinal studies. Psychological Medicine, 8, 611-622. Rogeness, G., Cepeda, C., Macedo, C., Fischer, C., & Harris, W. (1990). Differences in heart rate and blood pressure in children with conduct disorder, major depression, and separation anxiety. Psychiatry Research, 33, 199-206. Rooney, M., Fristad, M., Weller, E., & Weller, R. (1999). Children's Interview for Psychiatric Syndromes: Administration Manual . Washington, DC: American Psychiatric Press, Inc. Satterfield, J., & Dawson, M. (1971). Electrodermal correlates of hyperactivity in children. Psychophysiology, 8, 191-197.

PAGE 183

169 Satterfield J, Hoppe C, & Schell A. (1982). A prospective study of delinquency in (110) adolescent boys with attention deficit disorder and 88 normal adolescent boys. American Journal of Psychiatry,139, 795-798. Satterfield J, Satterfield B, & Schell A. (1987). Therapeutic interventions to prevent delinquency in hyperactive boys. Journal of the American Academy of Child and Adolescent Psychiatry, 26, 56-64. Satterfield, J., Swanson, J., Schell, A., & Lee, F. (1994). Prediction of antisocial behavior in attention-deficit hyperactivity disorder boys from aggression/defiance scores. Journal of the American Academy of Child and Adolescent Psychiatry, 33 (2), 185-90. Schachar, R., Tannock, R., Marriott, M., & Logan, G. (1995). Deficient inhibitory control in attention deficit hyperactivity disorder. Journal of Abnormal Child Psychology, 23, 411-437. Schlenker, R., Cohen, R., & Hopmann, G. (1995) Affective modulation of the startle reflex in schizophrenic patients. European Archives of Psychiatry and Clinical Neuroscience, 245, 309-318. Schmidt, K., Solant, M., & Bridger, W. (1985). Electrodermal activity of undersocialized aggressive children: A pilot study. Journal of Child Psycholgy and Psychiatry, 26, 653-660. Seidman, L.J., Biederman, J., Faraone, S.V., Weber, W., & Ouellette, C. (1997). Toward defining a neuropsychology of attention deficit-hyperactivity disorder: performance of children and adolescents from a large clinically referred sample. Journal of Consulting and Clinical Psychology, 65 (1), 150-60. Selke, G., Fennell, E.B., Dikel, T., Bosch, W., Harrell, D., & Bowers D. (2005) Possible gender and age differences in affective modulation of the startle reflex in healthy control children . Unpublished manuscript. Sergeant, J.A., Geurts, H., & Oosterlaan, J. (2002). How specific is a deficit of executive functioning for attention-deficit/hyperactivity disorder? Behavioral Brain Research, 130 (1-2), 3-28. Shapiro, E.G., Hughes, S.J., August, G.J., & Bloomquist, M.L. (1993). Processing of emotional information in children with attention-deficit hyperactivity disorder. Developmental Neuropsychology, 9 (3-4), 207-224. Shibagaki M, Yamanaka T, & Furuya T. (1993). Attention state in electrodermal activity during auditory stimulation of children with attention-deficit hyperactivity disorder. Perceptual Motor Skills, 77, 331-8.

PAGE 184

170 Slusarek, M., Velling, S., Bunk, D., & Eggers, C. (2001). Motivational effects on inhibitory control in children with ADHD. Journal of the American Academy of Child and Adolescent Psychiatry, 40 (3), 355-363. Solanto, M.V., Abikoff, H., Sonuga-Barke, E., Schachar, R., Logan, G.D., Wigal, T., Hechtman, L., Hinshaw, S., & Turkel, E. (2001). The ecological validity of delay aversion and response inhibition as measures of impulsivity in AD/HD: A supplement to the NIMH multimodal treatment study of AD/HD. Journal of Abnormal Child Psychology, 29 (3), 215-28. Sonuga-Barke, E.J., Taylor, E., Sembi, S., & Smith, J. (1992). Hyperactivity and delay aversion-I. The effect of delay on choice. Journal of Child Psychology and Psychiatry, 33 (2), 387-98. Spence, K.W., & Lang, P.J. (1990). Reading affective text: The startle probe response [Abstract]. Psychophysiology, 27, S65. Swanson J, & Castellanos F. (1998). Cognitive neuroscience of attention deficit hyperactivity disorder and hyperkinetic disorder. Current Opinions in Neurobiology, 8, 263-271. Tremblay, R., Pihl, R., Vitaro, F., & Dobkin, P. (1994). Predicting early onset of male antisocial behavior from preschool behavior. Archives of General Psychiatry, 51, 732-739. Van Goozen, S.H.M, Snoek, H., Matthys, W., van Rossum, I., & van Engeland, H. (2004). Evidence of fearlessness in behaviourally disordered children: A study on startle reflex modulation. Journal of Child Psychology and Psychiatry, 45, 884-892. Vrana, S.R., & Lang, P.J. (1990). Fear imagery and the startle-probe response: A new measure of emotin? Journal of Abnormal Psychology, 97, 487-491. Vrana, S.R., Spence, E.L, & Lang, P.J. (1988). The startle probe response: A new measure of emotion? Journal of Abnormal Psychology, 97, 487-491. Wechsler. (1999). Wechsler Abbreviated Scale of Intelligence: Manual . San Antonio: The Psychological Corporation. Weiss, G., Minde, K., Werry, J., Douglas, V., & Nemeth, E. (1971). Studies on the hyperactive child: VII. Five-year follow-up. Archives of General Psychiatry, 24, 409-414. White, J., Moffitt, T., Earls. E., Robins, L., & Silva, P. (1990). How early can we tell? Predictors of childhood conduct disorder and adolescent delinquency. Criminology, 28, 507-533.

PAGE 185

171 White, J., Moffitt, T., Caspi, A., & Bartusch, D. (1994). Measuring impulsivity and examining its relationship to delinquency. Journal of Abnormal Psychology, 103, 192-205. Zahn, T., Abate, F., Little, B., & Wender, P. (1975). Minimal brain dysfunction, stimulant drugs, and autonomic nervous system activity. Archives of General Psychiatry, 32, 381-7. Zahn, T., & Kruesi, M. (1993). Autonomic activity in boys with disruptive behavior disorders. Psychophysiology, 30, 605-14. Zentall, S.S., & Zentall, T.R. (1976). Activity and task performance of hyperactive children as a function of environmental stimulation. Journal of Consulting and Clinical Psychology, 44, 693-697. Zentall, S.S., & Zentall, T.R. (1983). Optimal stimulation: a model of disordered activity and performance in normal and deviant children. Psychology Bulletin, 94 (3), 446-71. Zentall, S.S., Falkenberg, S.D., & Smith, L.B. (1985). Effects of Color Stimulation and Information on the Copying Performance of Attention-Problem Adolescents. Journal of Abnormal Child Psychology, 13, 501-511.

PAGE 186

BIOGRAPHICAL SKETCH I graduated from Washington University in St. Louis, Missouri, in 1993 with a Bachelor of Arts in psychology. Upon graduation, I worked in Dr. Newcomer’s cognitive neuroendocrinology laboratory for six years as a research coordinator and Statistical Data Analyst in the Department of Psychiatry at Washington University School of Medicine. My work at Washington University School of Medicine led to authorship on 6 peer reviewed journal articles. In 1999, I was accepted into the University of Florida’s doctoral program in clinical psychology in the Department of Clinical and Health Psychology. As an entering graduate student, I was awarded the J. Hillis Miller Presidential Fellowship. I received a Master of Science in clinical psychology from the University of Florida in May 2001. While at the University of Florida I was awarded a three year Ruth L. Kirschstein National Research Service Award, as well as the Department of Clinical and Health Psychology’s Scientist-Practitioner Award, Molly Harrower Award, and Robert and Phyllis Levitt Neuropsychology Research Award. Moreover, while at the University of Florida, I have had the opportunity to be mentored by Eileen Fennell, Ph.D., and Dawn Bowers, Ph.D. I have recently completed my predoctoral internship in pediatric psychology and neuropsychology at Kennedy Krieger Institute, Johns Hopkins University. After receiving my Ph.D. I will complete a NIH funded two-year postdoctoral fellowship at the University of Florida, Department of Clinical and Health Psychology. 172