STATE

ms mjlzll lllllllllllllllllllllllllll

2 301565 4399

                       

 

 

 

 

 

 

 

 

This is to certify that the

thesis entitled

Serotonergic Function And Impulsive Aggression In
Alcoholics And Their Male And Female Offspring

presented by

Geoffrey Raymond 'IWitchell

has been accepted towards fulfillment
of the requirements for

 

 

Masters degree in Psychology

    

Major pr e or

Date_Ma;L_L_129]___

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

LIBRARY

Michigan State
University

 

 

 

PLACE Ii RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

ll DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU In An Affirmative ActioNEquei Opportunity Instituion

Wane-9.!

 

SEROTONERGIC FUNCTION AND IMPULSIVE AGGRESSION IN ALCOHOLICS
AND THEIR MALE AND FEMALE OFFSPRING
By

Geofi’ey Raymond Twitchell

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

MASTER OF ARTS

Department of Psychology

1997

ABSTRACT
SEROTONERGIC FUNCTION AND IMPULSIVE AGGRESSION IN ALCOHOLICS
AND THEIR MALE AND FEMALE OFFSPRING
By

Geofl‘rey Raymond Twitchell

A large literature describes an inverse relationship between a variety of measures of
serotonergic function and both alcoholism and impulsive aggression in adults recruited
fiom clinical settings. However, studies among children with these behavioral
characteristics are rare, have only minimally sampled female subjects, and the findings
have been inconsistent. This study achieved two goals. First, it examined the relationship
between serotonergic fimction, as measured by whole blood 5-HT and impulsive
aggression in community recruited antisocial alcoholic, non-antisocial alcoholic, and non-
alcoholic control men. Second, this study examined the relationship between serotonergic
function and impulsive aggression in 33 male and 12 female offspring (Mean age = 10.47
i 1.54 years) from these families. Results indicated that current child serotonergic
fimction was significantly related to current child impulsive aggression as measured by
Total Behavior Problem CBCL scores (1 = -.36, p = .007, N = 45). High impulsive
aggressive children exhibited greater serotonergic dysfimction (M = 171.13 ng/ml, SD =
62.13, N = 8) compared to low impulsive aggressive children (M = 229.81 ng/ml, SD =
58.67, N = 37; E (1,43) = 6.45, p = .02). Expected difi‘erences among adult groups
identified by impulsive aggressiveness were not found. Limitations of the study were

presented. Clinical and research implications were discussed.

To Clark, my family, and friends, for their support during the completion of this project.

iii

ACKNOWLEDGMENTS

This work could not have been completed without the assistance and support of
the following people:

Dr. Hiram E. Fitzgerald, whose dedication to this project and commitment to my
learning has been most valuable. His example of scientific excellence within the context of
a balanced life has been an important contribution to my graduate training.

Dr. Robert A. Zucker, who has increased the quality of my work by leaps and
bounds through his rigorous pursuit of excellence. His comprehensive understanding Of
the field of alcoholism research has guided my work in important and valuable ways.

Dr. Robert Caldwell, who enriched this project by challenging me to explore the
variety of ways my basic ideas might fit into the context Of the larger picture. His
scientific mind provided me with the opportunity to consolidate what I had leaned and to
grapple with alternative possibilities.

Dr. Edwin Cook, Department of Psychiatry, University of Chicago, for his
collaboration on this project, and his role in performing the sample assays at the Brain
Research Institute.

Drs. Gregory Hanna and Karley Little, Department of Psychiatry, University of
Michigan, for their continued scientific collaboration and for their combined wealth of

knowledge and experience in this field.

iv

Dr. William Atchison and Ravindra Hajela for providing me the use of their lab,
equipment, and not infiequently, their time.

Drs. Tom Smith and Raymond Bingharn who graciously provided invaluable
methodological and statistical support. Their knowledge and patience were greatly
appreciated.

Susan Refior, whose clinical skill and investment in the Family Study helped build
the positive relationships with these families that made the present study painless for
everyone involved.

My grandmother, Dorothy Lawrence, for her expert editing of this document.

My parents, who have both successfirlly completed their own advanced degrees
and who tirelessly gave me moral support even when they may have wondered if I’d ever
finish this project. I would also like to thank other family members, fiiends, and
colleagues, who provided needed emotional support throughout this lengthy process.

Most of all, I am especially grateful to Clark for his enduring patience, love, and
support through four long Michigan winters. I would also like to thank him for his ability
to help me set limits and keep things in perspective. Being able to share this

accomplishment with him has made the experience all the more valuable.

This work was supported in part by grants to RA Zucker and HE. Fitzgerald
from the National Institute on Alcohol Abuse and Alcoholism (ROI AAO7065), a
Biomedical Research Grant fi'orn Michigan State University, and by way of pilot firnd

support from the University of Michigan Alcohol Research Center.

TABLE OF CONTENTS

LIST OF TABLES .................................................... x
LIST OF FIGURES ................................................. xiii
INTRODUCTION .................................................... 1
Background Of the Proposed Model ................................. 3
Serotonergic Functioning ......................................... 5
Neurotransmission ......................................... 5

Clinical Advances .......................................... 8

Indices of Serotonergic Activity .............................. 11

REVIEW OF THE LITERATURE ....................................... 14
Alcohol and Serotonergic Function ................................. 14
Animal Studies ........................................... 14

Human Studies ........................................... 15

Aggression and Serotonergic Function ............................... 17
Animal Studies ........................................... 17

Human Studies ........................................... 18

Impulsivity, Aggressiveness, and Serotonergic Function .................. 19
Impulsive, Aggressive Alcoholics ................................... 21
Studies With Children And Adolescents .............................. 24
STATEMENT OF THE PROBLEM ...................................... 27
HYPOTHESES ...................................................... 28

METHOD .......................................................... 29

SUBJECTS ................................................... 29
PROCEDURE ................................................. 32
Participant Consent ....................................... 32
Materials and Preparation ................................... 32

Health History Questionnaire .......................... 32

Venipuncture ...................................... 33

Sample Storage and Transportation ...................... 33

Whole Blood 5-HT Assay ............................. 33
Methodological Considerations ............................... 34
MEASURES .................................................. 35
Demographic Variables ..................................... 35
Parental Socioeconomic Status ......................... 35

Parental Education .................................. 35

Family Income ..................................... 35

Current Alcohol and Cigarette Use ............................ 36
Paternal Impulsive Aggression Indicator ........................ 36
Alcoholic Subtype ........................................ 37
Childhood Impulsive Aggression Indicator ...................... 38
RESULTS .......................................................... 4O
Exclusionary Criteria ............................................ 40
Demographic Characteristics ...................................... 40

vii

Hypothesis 1 .................................................. 46

Adult Impulsive Aggression ................................. 46
Alcoholics versus Controls .................................. 48
Covariates .............................................. 48
Hypothesis 2 .................................................. 50
Childhood Impulsive Aggression ............................. 50

Time 1 Ratings of Son’s Impulsive Aggression ............. 50

Current Ratings of Son’s Impulsive Aggression ............. 52

Current Ratings of Daughter’s Impulsive Aggression ........ 54

Is There a Gender Efi‘ect in Children for Whole Blood 5-HT? . . 54
Current Ratings of Son’s and Daughter’s Impulsive Aggression
Combined ................................... 56

Correlations Between Child 5-HT and Impulsive Aggression . . . 59

Hypothesis 3 .................................................. 63
Intrafarnilial Correlations in Serotonergic Function ................ 63
DISCUSSION ....................................................... 66
Childhood Impulsive Aggression ................................... 67
Intrafarnilial Correlations of Serotonergic Function ..................... 68

Adult Impulsive Aggression ....................................... 69
Methodological Limitations ....................................... 7O
Implications ............................................. 72

Future Directions ......................................... 73

LIST OF TABLES

p—a

. Sampling Design ................................................... 31

N

. Group Membership Characteristics of Family Members With Analyzable Whole Blood
5-HT Samples ................................................. 41

3. Time 1 Parental Sociodemographic Characteristics of Families Classified by Father’s
Time 1 Indices of Antisociality and Alcoholism ........................ 42

4. Time 1 Sociodemographics of Sons Classified by Father’s Time 1 Rating of Impulsive
Aggressiveness ................................................. 44

5. Current Parental Sociodemographic Characteristics of Families Classified by Father’s
Time 1 Indices Of Impulsive Aggressiveness and Alcoholism .............. 45

6. Summary F Table For One Way ANOVA “frth The Independent Variable of Father’s
Alcoholism Status at Time 1 (AAL, NAAL, NC) And The Dependent Variable of
Father’s Current Whole Blood 5-HT(ng/ml) ........................... 47

\l

. Summary F Table for 2 (Mother’s Classification of Son at Time 1=High or Low
Impulsive Aggressive) X 3 (Father’s Alcoholism Status at Time 1=AAL, NAAL,
NC) ANOVA With Son’s Current S-HT (ng/ml) as The Dependent Variable . . 51

00

. Summary F Table For One Way AN OVA With The Independent Variable of Mother’s
Classification of Son’s Current Behavior (High or Low Impulsive Aggressive),
Dependent Variable of Son’s Current Whole Blood 5-HT (ng/ml) .......... 53

9. Summary F Table For One Way AN OVA Vlfrth The Independent Variable of Mother’s
Classification of Daughter’s Current Behavior (High or Low Impulsive
Aggressive), and Dependent Variable of Daughter’s Current Whole Blood
5-HT (ng/ml) .................................................. 55

10. Summary F Table For One Way ANOVA With The Independent Variable of Mother’s
Classification of Child’s Current Behavior (High or Low Impulsive Aggressive)
and The Dependent Variable of Child’s Current Whole Blood 5-HT (ng/ml) . . 57

11. Sociodemographics of High and Low Impulsive Aggressive Children (N=45;
Boys=33, Girls=12) as Classified by Maternal Ratings of Current Child
Behavior ..................................................... 58

12. Correlations Between Child’s Current 5-HT (ng/ml) and Time 1 Total Behavior

Problem Ratings ................................................ 60
13. Correlations Between Child’s 5-HT (ng/ml) and Current Ratings of Child Total

Behavior Problems (N = 45; Sons = 33, Daughters = 12) ................. 62
14. Intrafamilial Correlations in Serotonergic Function ......................... 64
Al. Parent Health History Questionnaire ................................... 75

B 1. Summary F Table For One Way AN OVA “With The Independent Variable of Father’s
Alcoholism Classification at Time 1 (Alcoholic or Non-Alcoholic) “With The
Dependent Variable of Father’s Current Whole Blood 5-HT (ng/ml) ........ 83

B2. Correlations Between Father’s Current 5-HT and Potential Covariates, (N = 38) . . 84

B3. Summary F Table For Simple Factorial ANOVA With The Independent Variable of
Father’s Alcoholism Status at Time 1 (AAL, NAAL, NC), The Dependent
Variable of Father’s Current Whole Blood S-HT (ng/ml), and Covariate of
Father’s Average Number of Days Smoking ........................... 85

C 1. Summary F Table For 2 (Mother’s Classification of Son at Time 1=High or Low
Impulsive Aggressive) X 3 (Father’s Alcoholism Status at Time 1=AAL, NAAL,
NC) ANOVA “With The Dependent Variable of Son’s Current S-HT (ng/ml), and
The Covariates of Age and Season .................................. 86

C2. Summary F Table For Simple Factorial ANOVA With The Independent Variable of
Mother’s Classification of Son’s Current Behavior (High or Low Impulsive
Aggressive), Dependent Variable of Son’s Current Whole Blood 5-HT (ng/ml),
Covariates of Age and Season ..................................... 87

C3. Summary F Table For Simple Factorial ANOVA “With The Independent Variable of
Mother’s Classification of Daughter’s Current Behavior (High or Low Impulsive
Aggressive), Dependent Variable of Daughter’s Current Whole Blood S-HT
(ng/ml), and Covariates of Age and Season ........................... 88

C4. Summary F Table For One Way ANOVA With The Independent Variable of Gender
and the Dependent Variable Of Child’s Current Whole Blood 5-HT (ng/ml) . . . 89

C5. Summary F Table For Simple Factorial ANOVA With The Independent Variable of
Gender, Dependent Variable of Child’s Current Whole Blood 5-HT (ng/ml), and
Covariates of Age and Season ..................................... 90

C6. Summary F Table For One Way ANOVA “With The Independent Variable of Mother’s
Classification of Child’s Current Behavior (High or Low Impulsive Aggressive),
Dependent Variable of Child’s Current Whole Blood 5-HT (ng/ml), and Covariates
of Age and Season .............................................. 91

C7. Correlations Between Child’s Current 5-HT (ng/ml) and Time 1 Total Behavior
Problem Ratings, Controlling For Childs’ Age and Season of Blood Draw
(N = 40; Sons = 34, Daughters = 6) ................................. 92

C8. Correlations Between Child’s Current 5-HT (ng/ml) and Current Behavior Problem

Ratings, Controlling For Childs’ Age and Season of Blood Draw
(N = 45; Sons = 33, Daughters = 12) ............................... 93

xii

LIST OF FIGURES

Figure l The Synapse .................................................. 6

Figure 2 The Mechanisms of Action for Psychiatric Medications ................. 10

xiii

INTRODUCTION

The precise ways that parental alcoholism afl‘ect the development of children are
still largely unknown, but they are most likely multifactorial with biological, as well as
psychosocial components (Sher, 1991; Sher, Walitzer, Wood, & Brent, 1991; Zucker &
Gomberg, 1986). However, most psychosocial studies neglect the biological bases of
transmission and association whereas, interestingly, many of the genetic studies focus
primarily on biological factors at the exclusion of a psychosocial contribution. The current
study attempts to address this void by exploring the important biological role serotonergic
dysfunction may play in the development of children of alcoholics who may be at increased
risk for the development of impulsivity, aggressiveness, and alcoholism.

One model suggests that irnpulsivity in childhood can be precursive to the
development of aggressiveness, and aggressiveness can be precursive to earlier and more
severe alcoholism in later adolescence and adulthood (Zucker & Fitzgerald, 1991a;
1991b). There is an increasing literature suggesting that serotonergic dysfirnction is
associated with alcoholism, aggressiveness, and irnpulsivity in adults manifesting in lower
whole blood serotonin (5-HT), lower cerebrospinal fluid monoamine metabolites of
serotonin (S-HIAA), and increased serotonergic uptake (Moss, 1987). The current study
was designed to examine the biologic contribution that serotonergic dysfirnction may
make to the development and emergence of impulsivity, aggressiveness, and ultimately
alcoholism in adults and their Ofi‘spring.

The current research builds on the notion that there is more than one kind of

alcoholism, which in turn means that diagnostic specificity is necessary to fully understand

2

the symptomatic and etiologic variations of the disease over the life course (Zucker, Ellis,
Fitzgerald, Bingharn, & Sanford, 1996). This study relied on one such classification
scheme which suggests that there are at the least two alcoholisms which can be
difi‘erentiated by the level of antisociality present (Zucker, 1987; Zucker et al., 1996;
Zucker, Fitzgerald, & Moses, 1995). It is hypothesized that alcoholics with higher levels
of antisociality spanning childhood and continuing into adulthood (AALs) have a life
course that is etiologically difl’erent and more severe than non-antisocial alcoholics

(N AALs).

When comparing AALs with NAALs, Zucker et al. (1996) found AALs to have
increased overall lifetime alcohol problem scores, including age first drunk, lifetime variety
of alcohol problems, and life percent involving alcohol problems. Additionally, NAALs
reported achieving higher socioeconomic status in adulthood, although childhood status
did not differ between the two groups. AALs had higher rates of separation and divorce,
increased levels of depression, and greater drug involvement in comparison to NAALs.
On the drinking variables of current frequency and quantity-variability of drinking, there
was no difl‘erence between AALs and NAALs.

Few studies have gathered data on serotonergic function in children, and none to
date have examined offspring of AALs and NAALs to help determine what role
serotonergic firnction may play in later development. Furthermore, studies of adults have
been inconclusive as to whether serotonergic dysfunction in alcoholics is heritable or
caused by previous alcohol consumption (Moss, 1987). Hopefully, data obtained from

children before the onset of alcohol use can help clear up that question.

Background Of the Proposed Model

The model used in this study came fi'om developmental data exploring early risk
pathways for future development of alcoholism (Zucker & Fitzgerald, 1991a; 1991b).
When exploring the etiology of alcoholism, the co-occurrence of other dificulties and
disorders may give clues to causation. To date, the strongest association with alcoholism
is the cO-action of antisocial personality disorder. The co-action of antisocial personality
disorder has been cited at 14% in samples of alcoholics (Regier et al., 1990). While
prevalence data are confounded by the issues of primary versus secondary occurrence and
base rate expression, using odds-ratios helps correct for the latter attenuation. When
using this technique, antisocial personality disorder shows the strongest possibility of
being causally connected to alcoholism at an Odds ratio of 21:1 (Zucker & Fitzgerald,
1991). This means antisocial personality disorder is 21 times more likely to be found in
alcoholics than in the general population. Considering the strength of this connection, it
seems reasonable to hypothesize that such a connection may begin as early as childhood.
Additionally, a review by Zucker and Gomberg (1986) cited several studies which also
indicated that antisocial behavior is a risk pathway to the manifestation of alcoholism.

Longitudinal studies using a high risk design compare individuals at known risk for
developing alcoholism (e.g. offspring of alcoholics) with lower-risk individuals to track
difl‘erential alcoholic outcome and premorbid indicators that proved predictive of such
outcome. Several of these studies have found that antisocial behavior is often associated
with the development of alcohol problems, with increased antisocial and aggressive

activity among later alcoholics, and with increased achievement related difficulties such as

4

poor school performance, poor productivity, high truancy, and high drop out rates
(Zucker & Gomberg, 1986). These data support the hypothesis that antisocial behavior is
connected to the development of alcoholism and also suggest that parental antisocial
behavior may play a role in child development (Zucker & Fitzgerald, 1991).

While these data come fi'om longitudinal studies spanning adolescence through
adulthood, it is still necessary to determine the pathway from childhood risk to adolescent
manifestation. Using data fi'om the Michigan State University/University of Michigan
Family Study (Zucker, Noll, & Fitzgerald, 1986), childhood risk factors from as early as
age 3-5 were explored (Zucker & Fitzgerald, 1991). Using the Child Behavior Checklist
(CBCL: Achenbach & Edelbrock, 1983), those children considered at risk showed
consistently elevated symptomatology in all behavioral areas except depression. The
differences between high risk and control children was 3:1 for overall problem behavior,
global externalizing symptoms including impulsivity, hyperactivity, and aggressiveness, as
well as a content specific measure of aggressiveness (Zucker & Fitzgerald, 1991).

This evidence Of increased levels of difiiculties in early childhood parallels the
evidence of later childhood and adolescent dificulty and may indicate an extension of a
continuous lifetime process. These data may also support the hypothesis of increased
impulsivity, as defined by CBCL scores, leading to greater aggressiveness in adolescence,
and finally, to the development of greater life difliculty with alcohol. The current study
used biological data to test the hypothesis that serotonergic dysfunction is involved in the

development of impulsivity, aggressiveness, and ultimately, alcoholism.

Serotonergic Functioning

The central serotonergic system reaches to the cortex and the limbic structures.
The limbic system is an area of the brain implicated in the coordination of feelings and
perception through its many connections to the frontal lobe. Consequently, the
serotonergic system is involved in the regulation of several physiologic and affective
firnctions. These include sleep, appetite, cognition, mood, sexuaL and motoric activity
(Vogt, 1982). Therefore, it has been suggested that any disturbance or dysfunction in this
system may contribute to the development of various psychiatric behavior disorders
involving abnormalities in these firnctions, such as afl‘ect (depression), anxiety (Obsessive
compulsive disorder, alcoholism), eating, and personality disorders (Coccaro & Murphy,
1990).

W The basic firnction of the central nervous system is the
transfer of information from one area of the brain to another. This system of intercellular
communication is composed of billions of neurons which interact with one another
through characteristic processes. Since these interconnections can work in synchrony, this
system is capable of producing graded responses which result in a vast array of cognitive,
afl‘ective, and behavioral responses (Gitlin, 1990).

The synaptic unit is the firnctional building block of the central nervous system and
consists of two neurons (presynaptic and postsynaptic) and a physical gap between them
called the synapse, see Figure l (Gitlin, 1990). For these neurons to communicate with
each other, a combination of bioelectrical and chemical processes must take place.

Neurotransmitter is released when an electrical impulse called the action potential anives

SYNAPTIC CLEFT
NERVE TERMINAL

 

 

 

 

@

 

 

ELECTRICAL IMPULSE

VESICLES POST SYNAPTIC RECEPTORS
WITH STORED
NEUROTRANSMI'ITER

  

Figure 1 The Synapse

7
at the axon terminal. This depolarizes the presynaptic terminal membrane, opening
voltage gated calcium channels. The resulting elevation in the internal calcium ion
concentration signals the release of neurotransmitter fiom the synaptic vesicles into the
synaptic cleft in a process called exocytosis (Bear, Connors, & Paradiso, 1996). The
neurotransmitter spreads across the gap and binds to specialized receptor sites on the
membrane of the postsynaptic neuron. The postsynaptic receptor then interprets the
information within the neurotransmitter and begins a physiologic response that can be
either inhibitory or excitatory (Martin, Owen, & Morihisa, 1987).

The neurotransmitter serotonin is found in large quantities within blood platelets
and the central nervous system and is synthesized within the presynaptic neuron from the
common amino acid L-tryptophan (Levitan & Kaczrnarek, 1997). Synthesis of serotonin
occurs in two steps. First, tryptophan is converted into 5-HTP (S-Hydroxytryptophan) by
the enzyme tryptophan hydroxylase. Next, S-HTP is converted to 5-HT (serotonin) by the
enzyme 5-HTP decarboxylase (Bear, Connors, & Paradiso, 1996).

Serotonin is then packaged into secretory granules by special proteins embedded in
the vesicle membrane called transporters and stored in the terminal bouton of the
presynaptic neuronal axon Once the released serotonin has interacted with the
postsynaptic receptors, it must be cleared fiom the synaptic cleft to allow future
transmission. Serotonin can be removed from the synaptic cleft by a specific transporter
through reuptake into the presynaptic terminal where it can either be recycled and
repackaged into synaptic vesicles or degraded by monoamine oxidase, resulting in the

metabolite 5-Hydroxyindoleacetic acid (5-HIAA).

8

Central 5-HT neurons are localized in tracts within the brainstem, but afl‘ect most
brain areas (Levitan & Kaczrnarek, 1997). The neurotransmitter serotonin appears to be
involved in several modulatory physiologic processes (Tollefson, 1989). There is a
widespread consensus that serotonin has an inhibitory effect on behavior. Specifically,
decrements in serotonergic function (decreased neurotransmission shown by lowered 5-
HT turnover) have been linked to behavioral disinhibition (irnpulsivity), low frustration
tolerance, (Moss, 1987; Jones, 1968; Goodwin, Schulsinger, Hennansen, et al., 1975;
Robins, 1966) attention deficits, (Rydelius, 1983; Tarter, Hegedus, Goldstein et al., 1984)
increased motor activity, (Jones, 1968; McCord, McCord, & Gudeman, 1968)
emotionality, (Rosenberg, 1969; Tarter, 1982) and aggressiveness (Guze, Goodwin, &
Crane, 1969).

W: One of the most important clinical advances to date is the use
of specific serotonergic uptake inhibitors to treat depressive symptomatology. Several
compounds such as fluoxetine (Prozac), citalopram, and zimeldine have been shown to
have antidepressant effects through their ability to inhibit the serotonin uptake carrier
(Fuller, 1992). While the antidepressant effects are clear, it is unknown how this specific
alteration in the dynamic, homeostatic, and complex serotonergic system produces these
positive efl‘ects.

Data on enhancing serotonergic function fi'om rat studies show that serotonergic
uptake inhibitors increase serotonergic firnction by decreasing serotonin synthesis, but at
the same time, increasing extracellular serotonin. It is behaved that the increased

extracellular serotonin leads to increased neurotransmission which causes positive

9

firnctional changes and decreased serotonin synthesis and neuron firing (Fuller, 1992).
Uptake inhibitors increase serotonin concentration in the synaptic cleft which results in
increased activation of all post-synaptic serotonin receptors.

Most psychiatric medications work by affecting some aspect of the
neurotransmitter/receptor system. The mechanisms of action can be explained by seven
difi‘erent efi‘ects of this system as diagramed in Figure 2 (Gitlin, 1990). 1.) First, a
medication can directly bind to the receptor site. A receptor agonist is one that mimics the
neurotransmitter by stimulating the receptor. Also, a medication can bind to the receptor
causing no response and efl‘ectively block the neurotransmitter fi'om binding to the
receptor (antagonists). 2.) Medications can also cause the release of more
neurotransmitter, firnctionally increasing the efi‘ect of the system. 3.) Medications can also
block the reuptake of neurotransmitters back into the presynaptic neuron which allows the
neurotransmitter to have more time in the synapse, increasing the possibility of stimulating
the receptor in the postsynaptic neuron and thereby increasing neurotransnrission. This is
best illustrated by the action of selective serotonergic reuptake inhibitors (S SRIs) such as
the widely used antidepressant, Prozac. 4.) Medications can also cause an increase or
decrease in the number of receptor sites (up or down regulation). 5.) Medications may
also alter the sensitivity of receptors, changing the magnitude of response that results from
receptor stimulation. 6.) Medications can alter the metabolism of a neurotransmitter
which changes the amount of neurotransmitter that is available for release. The
monoamine oxidase inhibitors work as antidepressants primarily by decreasing the

metabolism of a variety of neurotransmitters including serotonin, which are metabolized by

10

 

Q) RECEPTOR SITE

(a). MAC
SYNTHESIS --o. 4a. . ®@
.__°; ® -’ EFFECT

/_

 

 

  

 

 

@ REUPTAKE

Figure 2 The Mechanisms of Action for Psychiatric Medications

11

the enzyme MAO. 7.) The amount of available neurotransmitter can be altered by
increasing the amount of the precursor ingredients that are used in synthesizing
neurotransmitters. For example, a dietary increase in L-tryptophan, the amino acid
synthesized into serotonin, can result in an increase in serotonin (Gitlin, 1990).

The brain reacts to internal influences and external influences such as medications
by trying to maintain homeostasis. For example, a medication such as Prozac that blocks
the reuptake of serotonin, thereby firnctionally increasing the amount of available
neurotransmitter, produces a decrease in the amount of serotonin released in an attempt to
provide homeostasis (Gitlin, 1990). In essence, medications compete with a naturally
adapting system whose goal is to prevent changes. Other than agonists which mimic the
neurotransmitter by stimulating the receptor, medications appear to provide their clinical
efl‘ects by exerting a continual influence over neurotransmitter systems for extended
periods of time, thereby altering the regulatory mechanisms.

W11: Serotonergic activity can be studied by
measuring the concentration of the neurotransmitter, its metabolites, or metabolic
enzymes, in plasma, cerebrospinal fluid, and urine. All studies using measures of 5-HT
concentration are evaluating the presynaptic processes (e.g., metabolic turnover of the
neurotransmitter; Martin, Owen, & Morihisa, 1987). All are peripheral measures of brain
serotonin, and as such provide only an indirect measure of brain levels of the
neurotransmitter. Cerebrospinal fluid 5-HIAA concentrations have been used, but it is
unclear if CSF 5-HIAA adequately reflects brain serotonin firnction (Murphy, 1990). CSF

samples are particularly problematic because they reflect an average of brain and spinal

12

cord CNS concentrations and, therefore, cannot provide information regarding
localization of an observed abnormality (Martin, Owen, & Morihisa, 1987). Additionally,
many subjects refuse to consent to the invasive and sometimes painful spinal tap required.

Since most of the body's serotonin is synthesized within the gastrointestinal tract, it
has been suggested that measuring serotonin or its metabolites in urine or blood plasma
are inadequate (Fuller, 1992). Blood platelets, however, contain common constituents
with serotonin neurons such as serotonin storage granules and serotonin transporters
(Pletscher, 1987) which make it a more valid measure of central serotonergic firnction.
Measures from blood platelets such as serotonin content, serotonin transport, and
radioligand-binding to transporter have been used successfirlly to study autism, aggression
in males, and alcoholism, respectively (Fuller, 1992). Because of these benefits blood
, platelets have been called "models of brain serotonin neurons" (DaPrada, Cesrrra, Launay,
& Richards, 1988). Additionally, since blood platelets have been found to contain
serotonin receptors, blood platelet studies allow an examination of the less well
understood connection between post-synaptic S-HT firnctioning and suicidal and/or
impulsive aggressive behavior (Coccaro, Kavoussi, Sheline, Berman, & Csemansky,
1997). Blood platelets are also easily obtainable and accessible which increases the
attractiveness of their use.

Blood platelets do differ from brain serotonin in significant ways. Since blood
platelets do not synthesize serotonin but derive it completely by uptake, continuous
inhibition of the uptake carrier results in marked depletion of serotonin content in

platelets. However, brain serotonin content is not decreased by uptake inhibition, and

l3

extracellular serotonin concentration in the brain is actually increased (Fuller, 1992).
Given the similarities between blood platelets and serotonin neurons and the ease of

Obtaining such samples, blood platelets were deemed the most promising measure of

serotonergic function for this study.

REVIEW OF THE LITERATURE
Alcohol and Serotonergic Function

The association between alcoholism and serotonergic firnction has been shown in
both animal and human studies. Research suggests that alcohol consumption and
preference in humans may be related to decreased serotonergic neurotransnrission as
measured by the presynaptic measures of increased serotonin uptake and higher mean
mardmal-velocity-of-serotonin-transport (Vmax, density of platelet uptake sites; Boismare
et al., 1987; Bokii, Kiseleva, Lapin, Prakhe, Rybakova, 1984; Daoust, et al., 1991;
McBride, Murphy, Lumeng, & Li, 1989; Naranjo, et al., 1984; Naranjo, et al., 1987;
Naranjo, Sellers, & Larwin, 1986; Neiman, Beving, Malmgren, 1987; Rausch, Monteiro,
& Schuckit, 1991; Tollefson, 1989). Given this data, it should be possible to identify such
a dysfunction and educate those individuals in this group to their increased risk for the
development of alcoholism should they expose themselves to alcohol.

W: Data from animal studies have shown an association between
alcohol preference in rats and low brain serotonin turnover (Li et al., 1989). Increased
serotonin uptake has also been found in alcohol preferring rats (Daoust et al., 1985).
Additionally, serotonin reuptake inhibitors have been found to reduce ethanol
consumption in rats (Daoust et al., 1984; Murphy et al., 1985; Waller, Murphy, McBride,
Lumeng, & Li, 1985).

5-HT level has also been linked to alcohol preference and consumption in animal
studies, but the results are somewhat contradictory. One group found lower 5-HT levels

in brains of inbred alcohol preferring rats versus non alcohol preferring rats (Murphy,

14

15

McBride, Lumeng, & Li, 1982) Another study showed an ability to decrease ethanol
consumption in rats by administration of a 5-HT inhibitor (Rockman, Amit, Carr, Brown,
& Ogren, 1979). However, Kiianamaa (1976) found no difference in voluntary alcohol
consumption in rats after lowering brain serotonin by three difl‘erent methods.

Therefore, evidence for the link between serotonin and alcohol in the animal
literature is strongest in regard to alcohol preference, serotonergic uptake, and serotonin
turnover. The link between serotonin and alcohol is less strong in regard to S-HT
concentration's connection to voluntary consumption of alcohol. One reviewer has
suggested that rat strain difl‘erences, methodologic considerations, and differential food
intake could be responsible for some of the discrepancies in the animal literature (Moss,
1987)

W: Serotonin reuptake inhibitors have been found to reduce ethanol
consumption in male drinkers (Naranjo et al., 1984; Naranjo et al., 1987). In aggregate,
the data support the hypothesis that decreased serotonergic neurotransnrission might
facilitate alcohol intake, whereas increases in S-HT function may inhibit alcohol intake
(McBride, Murphy, Lumeng, & Li, 1989; Naranjo et al., 1986; Tollefson, 1989).

In a group of men with and without family histories of alcoholism, the family
history positive subgroup had a higher mean density of platelet serotonin uptake sites
(V max; Rausch, Monteiro, & Schuckit, 1991). Another study showed increased Vrnax of
serotonin uptake in alcoholics versus controls (Daoust et al., 1991). The same group also

noted an increase in platelet 3H-serotonin uptake in former alcoholics with eleven years

16

abstinence versus non-alcoholics (Boismare et al., 1987). D'aoust ct al., (1991) suggest
that serotonin uptake is altered during chronic alcohol consumption.

Two other studies have also shown abnormalities in serotonin uptake in the
platelets of detoxifying alcoholics (Bokii, Kiseleva, Lapin, Prakhe, & Rybakova, 1984;
Neiman, Beving, & Malmgren, 1987). It should be noted that one study found contrasting
results with a sample of withdrawaled alcoholic patients showing a decrease in platelet
serotonin uptake (Kent at al., 1985).

One hypothesis that has evolved fi'om these data suggests that alcoholics may
have a pre-existing low brain serotonin level that can be transiently raised by alcohol
consumption, but in the long run, actually leads to further depletion. This has lead to the
suggestion that the alcoholic may drink repeatedly to pharmacologically modify a
serotonin deficiency in the brain (Ballenger, Goodwin, Major, & Brown, 1979). D'aoust
et al., (1991) cite their previous study of former alcoholics with 11 years abstinence where
serotonin uptake was always increased (Boismare et al., 1987) as supporting evidence of
this hypothesis.

Platelet S-HT concentration is another measure that can be used as a measure of
central serotonergic activity. One study found reduced platelet S-HT content in alcoholics
(Banki, 1987) and another found lower mean platelet 5-HT levels in a sample of male and
female alcoholic patients during withdrawal and afier 2 weeks of abstinence versus non-
alcoholic controls (Bailly et al., 1990).

Cerebrospinal fluid (CSF) monoamine metabolites of serotonin such as 5-

hydroxyindoleacetic acid (5-I-lIAA) provide another measure of central serotonergic

17

function. Research on 5-I-IIAA in alcoholics is rare, but one study has shown abstinent
alcoholic men to have lower CSF-S-HIAA levels than controls (Ballenger et al., 1979).
Ballenger studied alcoholic men entering an inpatient program for treatment and measured
CSF S-HIAA in the immediate post-intoxication period and after 4 weeks of abstinence.
While CSF 5-HIAA measures in alcoholics could not be differentiated fi'om controls in the
post-intoxication period, there was a difference shown after four weeks of abstinence,
with the alcoholic group having lower concentrations of CSF 5-HIAA.
Aggression and Serotonergic Function

mm: Data linking serotonin with aggressive behavior in animals have
existed since a 1959 study that showed 5-HT reduced isolation-induced aggression in mice
(Yen, Stangler, & Millman, 1959). Currently, there exists a large literature of animal
studies supporting the hypothesis that central 5-HT is involved in the regulation of
aggressive behavior (Eichelman, 1979; Soubrie, 1986; Valzelli, 1981). Data cited by
Soubrie show an inverse correlation with central 5-HT activity and shock-induced fighting
(Karitak, Hegstrand, & Eichelrnan, 1981; Sewell, Gallus, Gault, & Cleary, 1982),
muricidal or mouse killing behavior (Katz, 1980; Waldbillig, 1979), and filicidal or pup
killing behavior (Copenhaver, Schalock, & Carver, 1978). It appears that each of these
behaviors can be increased or decreased by neurochemical manipulations of central 5-HT
activity.

Studies of isolation-induced fighting in male mice and mouse-killing behavior in
rats found an association between violence and a low 5-HIAA to serotonin ratio (V alzelli,

1969, 1971). A low ratio indicates a low serotonin turnover rate, and reducing CNS

18

serotonin by diet or pharmacological means leads to increased shock induced fighting or
mouse-killing behaviors (Kantak et al., 1981; Katz, 1980). Further evidence of this was
shown when mouse-killing behavior that is produced by p—chlorophenylalanine which
decreases the synthesis of serotonin could also be reduced by the serotonin reuptake
inhibitor fluoxetine (Berzenyi, Galateo, & Valzelli, 1983).

Interestingly, some evidence suggests that central S-HT activity is associated more
with reactive aggressive behavior versus the flank expression of aggression (Coccaro,
1989). When rats were previously exposed to mice the muricidal response produced by
decreased 5-HT was prevented (Marks, O’Brien, & Paxinos, 1977). These findings
suggest that when sufficient arousal is not present, central 5-HT may not be associated
with aggression. In a study of non-human primates, no aggression was found unless
conditions of increased arousal were present (Chamberlain, Pihl, & Young, 1987) as
occurs upon the insertion of a nasogastric tube.

Similarly, Soubrie (1986) argues that serotonin acts as an endogenous inhibitor for
various behaviors. Reduced serotonin may not lead to aggressive behaviors but may be a
catalyst for the expression of aggressive impulses. This model suggests that serotonin
depletion results in a primary increase in irnpulsiveness, and the increased aggressiveness is
just one of the more observable manifestations of behavioral change.

W: Many human studies have consistently shown reduced levels of
CSF S-HIAA in patients with a history of aggressive behavior as manifested by outward
violence (Brown, Goodwin, Ballenger, Goyer, & Major, 1979; Brown et al., 1982;

Linnoila et al., 1983; Roy, Adinoff, & Linnoila, 1988). Additionally, aggressiveness

19
directed inward which may be conceptualized as suicide attempts are also associated with
reduced CSF S-HIAA (Asberg, Traskrnan, & Thoren 1976; Banki & Arato, 1983;
Traskrnan, Asberg, Bertilsson, & Sjostrand, 1981; van Praag, 1982, 1983). For a
comprehensive review of this literature, please see Asberg et al., 1987.
Impulsivity, Aggressiveness, and Serotonergic Function

Many of the human studies on serotonergic dysfunction focus on impulsive
aggression (Coccaro, 1989; Asberg et al., 1987; Roy & Linnoila, 1988). In a study of 36
male murderers and attempted murderers under forensic evaluation, lower levels of CSF
S-HIAA were found among those whose crimes were impulsive versus premeditated
(Linnoila, \Wirkkunen, Scheinen, Nuutila, Rimon, & Goodwin, 1983). Interestingly,
further analyses found that those ofl‘enders who had committed more than one violent
crime had significantly lower CSF 5-HIAA levels than the ofl'enders who had committed
only one violent crime. Additionally, the impulsive offenders who had attempted suicide
had significantly lower CSF 5-HIAA levels than the non-impulsive ofl‘enders who had not
attempted suicide. These findings support the hypothesis that low CSF 5-HIAA may be
associated with impulsive violent behavior either towards others or oneself.

A later study was conducted of incarcerated male arsonists (mean age =29.8 3; 9.6
years) to determine whether low CSF S-HIAA levels were associated with aggressiveness
or impulsivity (Virkkunen, Nuutila, Goodwin, & Linnoila, 1987). Arsonists whose crimes
were motivated by uncontrollable impulsive urges to set fires were selected because they
show little interpersonal aggressiveness. Using a subset of the male violent ofl‘enders fiom

the previous study as well as a control group of 10 community recruited healthy subjects

20

(3 female, 7 male) who were flee of any history of mental illness, CSF S-HIAA levels were
found to be lower in the arsonist group than either the violent offenders or controls. The
authors suggest that this gives further evidence that low CSF S-HIAA is associated with
violent behavior through an efl‘ect on poor impulse control.

In one review, Coccaro (1989) cites various studies of clinical samples comprised
largely of male subjects in their 20's giving evidence that CSF SHIAA levels are negatively
correlated with both clinician and self-reported aggression (Brown et al., 1979; Brown et
al., 1982; Lidberg, Tuck, Asberg, et al., 1985; Linnoila et al., 1983), irritability and
hostility (Brown & Goodwin, 1984; Roy et al., 1988; Rydin et al., 1982; van Praag, 1986)
as well as criminal behavior (Lide et al., 1985; Linnoila et al., 1983; van Praag, 1983;
\Wirkkunen et al., 1987). A negative correlation was also found in retrospective self
reports of childhood and adolescent behavioral difficulties (Brown et al., 1985) which may
also indicate that reduced 5-HT activity is associated with early behavioral dificulties.

Coccaro (1989) also cites evidence fiom other peripheral measures of 5-HT
functioning (whole blood 5-HT, plasma tryptophan/neutral amino acid ratio, and platelet
3H-IMI binding) which also indicate an inverse correlation with aggressive behaviors.
Three out of 4 of these studies showed a negative correlation with clinician rated measures
Of current aggressive behavior or historic accounts of serious aggressive acts that led to
jail time. One study found an inverse relationship between whole blood 5-HT and
hyperactivity and aggression in a sample of 30 mentally retarded patients (age 4-3 9) with
hyperactivity and aggressive behavior (Greenberg & Coleman, 1976). A second study

found a low ratio of tryptophan to amino acids which signifies lowered synthesis of

21

sertonin in male alcoholics who had histories of aggression when compared to male
alcoholics without such histories (Branchey, Branchey, Shaw, & Lieber, 1984). A third
study of pre-pubertal children found that a group diagnosed with DSM-III conduct
disorder (14 male/3 female) had lower levels of H-IMI binding (Bmax) compared to an
age and sex matched group of control children (Stofl‘, Pollack, \Wrtiello, Behar, & Bridger,
1987). However, one study showed a positive correlation between whole blood 5-HT and
ratings of conduct disorder in male adolescents (Pliszka et al., 1988).

These data suggest that indices of 5-HT function correlate more strongly with
irritable, impulsive, aggression versus premeditated aggression or violence. This is
bolstered by the Linnoila studies (1983, 1985) of impulsive violent ofl‘enders versus non-
impulsive violent ofl‘enders, with the non-impulsive violent offenders exhibiting normal
CSF 5-HIAA concentrations. Coccaro also points out that several studies have noted that
low CSF 5-HIAA patients exhibit increased irritability and hostility (Brown & Goodwin,
1984; Roy et al., 1988; Rydin et al., 1982; van Praag, 1986). Coccaro argues that reduced
S-HT function may lead to vulnerability to increased response to noxious stimuli. Brown
and Linnoila agree with Coccaro that there is increasing evidence for the hypothesis that
impulsivity, disinhibition, or dyscontrol is the behavioral variable linked to low levels of
CSF 5-HIAA rather than antisocial or violent acts themselves.

Impulsive, Aggressive Alcoholies

Studies of impulsivity, aggressiveness, and alcoholism have found abnormal

serotonergic function in early onset male alcoholics with paternal history of alcoholism and

impulsive aggressiveness when drunk (Buydens-Branchey, Branchey, Noumair, Lieber,

22

1989; Virkkunen & Linnoila, 1990). This subgroup of alcoholics has been classified by
Cloninger as the Type II alcoholic (Cloninger, Bohman, & Sigvardsson, 1981; Cloninger,
Sigvardsson, von Knorring, & Bohman, 1988). However, several research groups note
that these characteristics are associated with individuals who are suffering fi'om antisocial
personality disorder (Irwin, Schuckit, & Smith, 1990), borderline personality disorder, and
intermittent explosive disorder (Linnoila et al., 1983; Linnoila et al., 1989; Virkkunen,
DeJong, Goodwin, & Linnoila, 1989; Virkkunen, Nuutila, Goodwin, & Linnoila, 1987).
All of these disorders have been associated with reduced brain serotonergic function.
Additionally, one group has noted that there is strong evidence that antisocial personality
disorder and alcoholism are distinct entities (Schuckit, Klein, Twitchell, & Smith, 1994),
and others stress that alcoholism accompanied with antisocial personality disorder should
be differentiated fi'om alcoholism outside the context of antisocial personality disorder
(von Knorn'ng, von Knorring, Smigan, Lindberg, & Edholm, 1987).

Evidence given by Virkkunen and Linnoila (1990) linking abnormal serotonergic
metabolism with irnpulsivity in the group of early onset alcoholics comes fi'om Monoamine
Oxidase (MAO) and CSF 5-HIAA studies. Low platelet MAO has been linked with
alcoholism since the 1980's (Oreland, 1983; Oreland, Gottfiies, Kiianmaa, Wiberg, &
“melad 1983). When using Cloninger's proposed classification schema of Type I (history
of maternal alcoholism, later onset of problems, less life dificulty, low aggressiveness) and
Type II alcoholics (paternal history of alcoholism, early onset of life problems, greater
severity of life interference, and increased aggressiveness), it was found that Type II‘s

were primarily characterized by low platelet MAO activity (Oreland, von Knoriing, von

23

Knorring, & Bohman, 1985; von Knorring, Bohman, von Knorring, & Oreland, 1985)
which has also been found in criminals with antisocial personality disorder (Lidberg,
Modin, Oreland, Tuck, & Gillner, 1985). Interestingly, only 1.5% of the population
belong to this group characterized by low platelet MAO (Cloninger, von Knorring, &
Oreland, 1985). Moreover, there is an over-representation of Type II's within this
subgroup that demonstrates low platelet MAO, which comprises between 15-25% of
male alcoholics (Linnoila & Virkkunen, 1989). There is also evidence that personality
characteristics such as irnpulsiveness and sensation seeking which are typical of Type 11's
and antisocial personality disorder are also linked to low platelet MAO activity (Von
Knorring, Oreland, & Winblad, 1984).

Virkkunen and Linnoila (1990) also note from previously cited studies that
habitually violent, impulsive ofl‘enders who behave violently when dnrnk and have a low
concentration of CSF S-HIAA also fit the criteria for Cloninger's Type II alcoholics
(Linnoila et al., 1983; Virkkunen et al., 1987). These authors also point out that in a
study of violent ofl‘enders and their families, low CSF SHIAA was associated with paternal
alcoholism (Linnoila et al., 1989).

Another study looked at early onset of alcohol related problems in male alcoholics
(Buydens-Branchey, Branchey, Noumair, & Lieber, 1989). Those who were age 20 or
below when problem drinking began comprised early onset alcoholics, while males whose
problems emerged after age 20 comprised the late onset alcoholics. These authors noted
that the group with early alcohol problems had higher rates of incarceration for violent

ofl’enses, more suicide attempts, and more depression. To see if they could differentiate

24

between the two groups by serotonergic function, tryptophan availability was measured.
The ratio of tryptophan over other amino acids competing for entrance into the brain
determines the amount of tryptophan available for synthesis into serotonin. It was found
that patients with early onset alcoholism had lower ratios of tryptophan compared with
other amino acids. Therefore, these authors hypothesize that altered serotonergic
neurotransmission plays a role in alcohol-seeking behavior by early onset drinkers.
Studies With Children And Adolescents

While studies with adults show that serotonergic function is negatively associated
with aggressiveness and impulsivity, results with children and adolescents are less
consistent. In one review of CSF serotonin metabolites, Brown and Linnoila, (1990) point
out that high levels of CSF 5-HIAA have been found in infancy and adolescence (Riddle et
al., 1986; Kruesi, Swede, Hamburger, Potter, & Rapoport, 1988). These authors conclude
that it is unclear how early age-associated changes in CSF SI-IIAA relate to
aggi'essivefnnpulsive and self-destructive behaviors.

Brown et al., (1986) showed that low CSF 5-HIAA in young adult males was
associated with childhood conduct problems, and Kruesi et al., (1990, 1992) found lower
CSF SHIAA in children with disruptive behavior disorders (27 male/2 female) versus age,
sex, and race matched controls both at assessment (mean age = 11.3 i 3.6 years) and 2
year follow-up (mean age = 13.8 i 3 .9 years). Another study used platelet irniprarnine
binding as yet another measure of peripheral serotonergic firnction in a study of 23, 10-16
year old males recruited from an inpatient psychiatric unit, and who all firlfilled criteria for

conduct disorder, and 16 who also fiilfilled criteria for attention deficit hyperactivity

25

disorder (Birmaher et al., 1990). Results showed that the number of imiprarnine binding
sites correlated inversely with total problems scores on the Child Behavior Checklist
(CBCL: Achenbach, & Edelbrock, 1983), as well as externalizing, hostility, and
aggressiveness scores on the CBCL. Data fi'om these three studies coincide with the mass
of adult literature suggesting a negative correlation between serotonergic function and
aggressiveness. However, it should be noted that one additional study found conflicting
results with an increase in whole blood 5-HT among incarcerated adolescent males with
conduct disorder with or without anxiety or depression (mean age = 15.5 _t 1.43 years)
compared to a male adolescent sample recruited from a community mental health clinic
with anxiety or depressive symptoms alone (mean age = 12.4 i 2.0 years; Pliszka et al.,
1988). Importantly, this study did not have a control sample of healthy controls to
determine how whole blood 5-HT in conduct disordered males compares to a healthy,
non-clinical sample. However, the same study also noted that 5-HT correlated mm
with clinician ratings of conduct symptoms, as well as a trend for higher 5-HT
concentration in violent versus non-violent ofl‘enders in this conduct disorder sample.

Still other studies have shown consistent results with the adult data. 3H-
Imipramine binding, the distribution of which closely parallels the distribution of serotonin
neurons, was reduced in children with mixed conduct disorder and ADHD (Stofl‘, Pollack,
Vitiello, Behar, & Bridger, 1987), but was not found in a sample of children with only
ADI-ID diagnosis without conduct disorder (W eizrnan, Bernhout, Weitz, Tyano, &
Rehavi, 1988). Central 5-HT function was studied in 25, 7-11 year old boys with

aggressive and non-aggressive ADI-ID by Prolactin response to fenfleuramine challenge

26

(Halperin et al., 1994). The findings showed the aggressive group had greater response to
the challenge which is consistent with altered central S-HT function. It must be noted that
controls were not used and results cannot unequivocally identify which group had the
deviant response to fenfleurarnine challenge. It must also be noted that the peripheral
measures such as platelet 5-HT, plasma Ml-IPG (3 -methoxy-4hydroxy-phenylglycol, a
norepinephrine metabolite), or plasma HV A (homovanillic acid, a dopamine metabolite)

did not show any between group differences.

STATEMENT OF THE PROBLEM

The literature reviewed indicates that serotonergic dysfirnction may be a factor in
the development of irnpulsivity, aggressiveness, and alcoholism in samples recruited fi'0m
clinical settings. Therefore, the current study used classification schemas based on
variables thought to be related to irnpulsivity and aggressiveness (Antisocial Behavior
Scores for fathers and CBCL Total Behavior Problem scores for children) to explore the
relationship between impulsive aggression and serotonergic firnction. While the adult data
clearly show that alcoholism and impulsive aggression are marked by decreased
serotonergic neurotransmission as measured by low CSF SI-IIAA, low whole blood
platelet 5-HT, and increased 5-HT uptake, the data from children and adolescents are less
clear. While one study found a positive correlation between serotonergic
neurotransmission and aggressive and impulsive children and adolescents, the majority of
the studies agree with the adult data which show a negative correlation.

First, the present study hoped to add to the sparse literature of children’s studies
by exarning serotonergic function and impulsive aggression in both male and female
children Second, this study was designed to further contribute to the existing literature by
examining this relationship in a sample of individuals who difl‘er from previous samples in
that they are community recruited and were not comprised of clinical treatment or
incarcerated groups. Given the wealth of consistent data from adult samples and the
similar findings in the large majority of child and adolescent studies, the current study
predicted a negative relationship between both adult and childhood impulsive aggression

and serotonergic firnction.

27

HYPOTHESES

1.) Greater serotonergic dysfunction will be demonstrated in AALs when
compared with NAALs. Specifically, this dysfirnction will manifest in AALs having lower
serotonergic neurotransmission as indicated by lower levels of whole blood 5-HT.
Relatedly, it is expected that there will be a negative correlation between father’s
serotonergic function and impulsive aggression, as measured by the Antisocial Behavior
Checklist (Zucker & Noll, 1980)

2.) Greater serotonergic dysfunction will be demonstrated in children who are
rated as high impulsive aggressive when compared to children rated as low impulsive
aggressive. Relatedly, it is expected that with all children combined, there will be a
negative correlation between serotonergic firnction (as measured by whole blood 5-HT)
and scores on the Child Behavior Checklist (Achenbach & Edelbrock, 1983).

3 .) Children's serotonergic function will be positively correlated with parent’s

serotonergic function.

28

METHOD
SUBJECTS

Subjects for the present study were drawn fi'om an already existing longitudinal
data set of the Michigan State University/University of Michigan Family Study which is
tracking the etiology of alcoholism in a sample of 215 alcoholic families and 96 non-
alcoholic control families. Alcoholic fathers for the original study were recruited fi'om
local district court records of all male drunk driving offenders who had a blood alcohol
concentration ofO. 15% or higher at the time of arrest, or 0.12% or higher if that arrest
was the second or more documented offense. Court personnel approached these men and
obtained consent to release their phone numbers to research stafl‘ for potential involvement
in a study of child development and family health.

Participation criteria for the larger study included having a biological son between
age 3.0 and 6.0 living in an intact family with the biological mother. Evidence of fetal
alcohol syndrome resulted in exclusion from study participation (Fitzgerald et al., 1993).
Subjects were informed that participation was entirely voluntary, confidential, and
unrelated to court outcome. Seventy-nine percent of the men agreed to release their
names to study personnel and 91% of these men and their wives agrwd to participate.
The original sample was comprised of non-Hispanic Caucasians due to the limited
ethnic/racial make-up of the catchment area. Additionally, it should be noted that the
larger sample is comprised of convicted drunk drivers which may indicate that an
antisocial component is also present in the sample. All of these men were later screened

and met diagnosis of probable or definite alcoholism using Feighner diagnostic criteria

29

30
(Feighner et al., 1972). This diagnosis incorporated information from the Short Michigan
Alcohol Screening Test (Selzer, 197 5) and the Diagnostic Interview Schedule (Robins et
al., 1981). Diagnostic status of the mothers in these families was flee to vary, although a
subset (41%) made a lifetime DSMIII-R alcohol abuse/dependence diagnosis.

To insure demographic comparability, non-alcoholic control families were
recruited through door to door canvassing which started one block away fiom each
alcoholic family and stayed within the same census area. Other criteria for control families
included the presence of a male biological child who was within 6 months of the same age
as the experimental target child. Feighner criteria were used for both fathers and mothers
to rule out probable or definite alcohol or other drug abuse/dependence.

The sampling design for the present study is shown in Table 1. Given that some
families were not able to produce usable blood samples for a complete family (father,
mother, and son), the final sample included 13 AALs, 13 NAALs, and 12 NCs. Children's
ages ranged fiom 7-15 (sons and daughters combined, N = 45, M = 10.48, SD = 1.65).
Within each father's group we selected 6 fathers who had sons with high Total Behavior
Problem scores on the CBCL and 6 fathers who had sons with low Total Behavior
Problem scores on the CBCL. AAL families were those that had fathers with Antisocial
Behavior Scores of 24 or greater on the Antisocial Behavior Checklist (Zucker & Noll,
1980) while NAAL families were characterized by fathers with ASB scores below 24.
High impulsive aggressive children were classified by CBCL T-scores of 60 or greater,

and low impulsive aggressive children were characterized by scores of 59 or less. Subjects

31

for the current study included the following: biological father, biological mother, male

target child, and a female child if available and age appropriate.

Table 1

Father’s Classification

 

 

Son’s Hi TBP 6 6 6
Classification Lo TBP 6 6 6

 

 

 

 

 

 

 

32

PROCEDURE
Participant Consent

Prior to sample collection, the study coordinator contacted each participating
family and gave an introduction to the proposed study. Then a follow-up phone call was
made by the author, outlining detailed procedures of this separate in-home session and the
voluntary nature of their participation. Participants were then asked to give written
consent to this procedure with the understanding that they could withdraw at any time
without risk of penalty.
Materials and Preparation

Each subject had 24ml of whole blood collected by venipuncture in polystyrene
ethylenediarninetetraacetic acid (EDTA) Vacutainer brand collection tubes by the author
who is a trained phlebotomist. Twenty-one gauge Vacutainer brand needles were used to
insure platelet integrity. All Vacutainer tubes were labeled beforehand with an indelible
marker with the subject's 6-digit identification number, date, and time of blood draw. All
identifying information, collection time, and number of tubes were logged in a lab book
immediately following venipuncture.

MW: While the MSU/U of M Family Study already
has a self reported life history of physical health for each study participant, a general
medical screen in the form of a health history questionnaire was administered to each
participant prior to venipuncture to ascertain any current medications, illnesses, or
difliculties that might have confounded sample results, see Appendix A. Relatedly, this

health screen also inquired about recent alcohol abuse of the respondent and their spouse

33

as well as recent cigarette use. All family members were informed that they would be
paid $10 for their participation. Each family member was informed that if they were
injured as a result of participation in this study, Michigan State University would provide
emergency medical care if necessary. Sample collection proceeded following a full review
and approval by Internal Review Boards at Michigan State University and at the
collaborating institution, the University of Michigan.

W Each subject was seated in a comfortable supine position for the
single blood draw. The antecubital area of the most appropriate arm for draw was
prepared for venipuncture with alcohol pad and tourniquet. Venipuncture proceeded with
a single needle stick. A total of 24 mls of whole blood was collected into (6), 4ml EDTA
Vacutainer tubes. As soon as a steady flow of blood was apparent, the tourniquet was
undone to prevent any subject discomfort. Following venipuncture completion and needle
withdrawal, the puncture site was observed, and the subject was instructed to place sterile
gauze with pressure to the site to prevent any firrther bleeding. The site was then covered
with a band-aid to prevent any irritation of the area. All tubes were then gently inverted
approximately 5-6 times to thoroughly mix blood with EDTA anti-coagulant.

W Blood for whole blood serotonin assay
was immediately flow in dry ice pellets, then stored in a -70° C fieezer upon arrival at
Michigan State University. Later, these samples were batch shipped in dry ice to the
University of Chicago for laboratory analysis.

W Whole blood 5-HT was analyzed by high pressure

liquid chromatography with fluoremetric detection (Anderson, Young, Cohen, & Schlicht,

34

1981). 5-Hydroxytryptophan was used as an internal standard. Intra-assay and interassay
coeflicients of variation were 2.3% and 4.0%, respectively. Assays were performed blind
to all demographic and diagnostic information in the Developmental Neuroscience
Laboratory of the Brain Research Institute at the University of Chicago.
Methodological Considerations

Some evidence suggests that whole blood 5-HT decreases moderately between
ages 10 and 12, but alter this time period, there is no correlation of whole blood S-HT and
age (Ritvo et al., 1971). Since the children for this study were within this age range,
statistical analyses of this dataset took this variable into account. Other evidence suggests
that whole blood 5-HT is not afl‘ected by ordinary changes in diet (Ritvo et al., 1970),
therefore fasting was not required in the present study. Prior work has shown that there is
no significant diurnal variation in blood 5-HT content in adults (Kramer, Goekoop, & Van
Kempen, (1990). However, seasoml effects on platelet S-I-IT content have been found in
children and adolescents with OCD and matched normal controls (Brewerton, Flament,
Rapoport, & Murphy, 1993). Therefore, analyses controlled for season of blood draw
through the use of a dichotomous covariate. This covariate was created by dividing the
calendar year by equinox. Thus, blood samples drawn in “Winter (September 21-March
20) were coded as low photo period and samples drawn in Summer (March 21-September

20) were coded as high photo period.

35

MEASURES
Demographic Variables

The families in this study were characterized by the demographic measures of
socioeconomic status, current educational degree, years of education, family income, and
age. All measures were reported by respondents on a demographic instrument
administered as part of the protocol of the parent study except the variable of age which
was calculated using date of birth and date of blood draw.

Wm: Information regarding parental socioeconomic
status (SES) was measured by the Revised Duncan Socioeconomic Index (TSE12,
Stevens & Featherrnan, 1981). The TSE12 is a measure of occupational attainment which
has been suggested as a more contemporary indicator of SES than measures based solely
on income.

2mm: Two measures of education were used to describe the
sample. The first measure, years of education, denotes the number of years of academic
or vocational education achieved. Scores ranged fi'om 10-20 years of education. The
second measure denotes the highest degree achieved. Scores ranged from 0 to 4 (0=high
school diploma or less, lwocational or technical degree, 2=Bachelors degree, 3=Masters
degree, 4=Doctorate, Medical degree, or Veterinarian degree).

W: Family income was measured as each family’s gross annual
income in dollars as reported by father. Subjects responded to a lO-point scale in which a
score of 1 represented income of $4,000 or less; 2=$4,001-$7,000; 3=$7,001-$10,000;

4=$10,001-313,000; 5=$13,001-$l6,000; 6=$l6,001-$20,000; 7=$20,00l-$30,000;

36

8=$30,001-$50,000; 9=$50,001-$75,000; 10=$75,001-$100,000; 11=Over $100,000. The
midpoint dollar level of each interval was used in computing mean income level for the
sample. For example, an interval score of 3 was coded as $8500.
Current Alcohol and Cigarette Use

Quantity and frequency measures of alcohol and cigarette use for all subjects were
gathered at the time of blood draw through subject responses to related items on the
Health History Questionnaire (see Appendix A). Parents were asked to provide a self
report of their own alcohol use as well as a collateral report of their spouse’s alcohol use
in an attempt to reduce possible reporting bias. These data were coded such that the
highest report of alcohol use by either the subject or the subject’s spouse was used for
analyses. Children were asked to report on their own alcohol and cigarette use in an
abbreviated version of this instrument which was administered privately.
Paternal Impulsive Aggression Indicator

The Antisocial Behavior Checklist (Zucker & Noll, 1980) was used to assess
father's level of impulsive aggression. The ASB is a 46-item self report inventory that
measures the fi'equency of participation in various aggressive and antisocial activities both
in childhood and adulthood. The current version is a revision of an earlier instrument used
in the Rutgers Community Study (Zucker & Barron, 1973) and was modified to make
items salient for both childhood and adulthood activities.

Reliability and validity studies of college students and inmates have found the ASB
to have adequate test-retest reliability (.91 over 4 weeks) and internal consistency

(coefiicient alpha =93) (Zucker & Noll, 1980). The ASB has also been shown to

37
difl‘erentiate between groups with heavy antisocial histories (inmates), individuals with
minor district court ofl‘enses, and university students (Zucker, Noll, Ham, Fitzgerald, &
Sullivan, 1992; Fitzgerald, Davies, Zucker, & Klinger, 1993 ), and between those with
antisocial personality disorder and those without it (Zucker, Ellis, Fitzgerald, Bingharn, &
Sanford, 1996).
Alcoholic Subtype

Family classification as AAL or NAAL was assigned using the father's Time 1
score on the Antisocial Behavior Checklist. The overall ASB index was computed by first
summing the father‘s childhood and adulthood domain scores. This procedure was used to
insure tlmt high scoring fathers had evidenced a lifetime pattern of aggressive and
antisocial behavior beginning in childhood and continuing into adulthood as evidenced by
alcoholism and sociopathy. Thus, high-scoring fathers (AALs) are those who have
evidenced a lifetime pattern of sustained antisociality rather than a sporadic or more
intermittent history of antisocial difficulties.

Scores of 24 or greater on the ASB served as the cutofl‘ for classification as AALs
while those less than 24 were classified as NAALs. This cutofl‘ was chosen by computing
sensitivity and specificity of scores using DSM-IIIR codings of antisociality as a standard.
The sensitivity of the ASB score of 24 was calculated to be .85 while specificity was
calculated to be .83. By comparison, cutoff scores of 21 and 27 yielded sensitivity and
specificity scores of .94 and .75 and .79 and .87 respectively. Therefore, the ASB score of
24 proved to be the best combination of sensitivity and specificity and was also a

comparable classification compared to the DSM-IIIR category for antisocial personality.

38

Childhood Impulsive Aggression Indicator

The 4-16 year old version of the Child Behavior Checklist (Achenbach &
Edelbrock, 1983) was used to classify sons as high or low on a behavioral phenotype
indicator of impulsive aggression. The CBCL provides an objective measure of the child's
social and emotional functioning. It is comprised of 113 items regarding prevalence and
fi'equency of child problem behavior. When summed, these items yield a total behavior
problem score (TBP) as well as two broadband scores (internalizing and externalizing
behaviors) and eight narrow band scores. For child classification in the present study,
fathers’ (TBP) ratings of their sons when they were 3-6 years of age were used, so that
this indicator is one of the early appearance of impulsive aggression. Although mother’s
ratings on their children’s behavior have been more heavily utilized in the literature
(Barber, Olsen, & Shagle 1993; Fitzgerald, Zucker, Maguin, & Raider, 1994; Prior,
Smart, Sanson, Pedlow, & Oberklaid, 1992), data fi'om the larger parent study indicate
that mothers and fathers rate their son’s behavior similarly (Bingham, Fitzgerald, Zucker,
under review). Since the primary focus was on fathers and sons, it was decided that
fathers' ratings of their children were most appropriate for this classification.

The Total Behavior Problem Score was selected for this categorization because it
is the indicator with the greatest range, and because it correlates very highly with both the
CBCL Aggression subscale (r=.88, p S .01) and the CBCL Externalizing Broad Band
scale (r=.91, p S .01) in analyses of the larger longitudinal sample; correlations of similar

magnitude are also reported in the norming of the instrument (Achenbach & Edelbrock,

39

1983). In addition, in the larger longitudinal study the TBP score correlated moderately
with ratings of irnpulsivity and hyperactivity on the Conners Impulsivity/Hyperactivity
Scale (r=.31, p, 5.01, Conners, 1990).

Using sample distribution characteristics and CBCL clinical scoring criteria as a
base, we established a TBP T-score of 60 or greater (borderline clinical cutofl) as the
cutofl‘ for the high impulsive aggression categorization; boys with TBP T-scores of 59 or
less were classified as being low on this behavioral dimension. Reliability coeflicients for

the CBCL range from .84 to .93 (Achenbach and Edelbrock, 1983).

RESULTS
Exclusionary Criteria

Before beginning analyses, subjects taking any potentially serotonergic altering
medications were identified from self report data in the Health History Questionnaire. The
5-HT value for each of these 12 individuals was then excluded from analyses. The
composition of these exclusions by medication status were broken down into the following
groups: Boys (3-Ritalin, l-Tofranil, l-Ritalin and Buspar), Fathers (l-Calan), Mothers (1-
Xanax and Dyazide, l-Xanax and Cardizem, l-Prozac, l-Valium, l-Efl‘exor and Parnelor,
I-Prozac and Efl‘exor).

Given that this study examined only firlly biologically related family members, one
girl's 5-HT value was excluded from analyses based on her father‘s report of questionable
paternity. Overall, a total of 13 individual S-HT values (10% of total samples drawn) were
excluded from analyses. This resulted in a total of 117 analyzable whole blood 5-HT
values (F athers-3 8, Mothers-32, Boys-34, Girls-13). Table 2 shows their group
membership characteristics.

Demographic Characteristics

Table 3 presents the Tirna 1 parental sociodemographic and educational
characteristics of this study’s families which were classified by father’s Time 1 indices of
impulsive aggression and alcoholism. Parental SES as measured by the Revised Duncan
Socioeconomic Index (TSE12, Stevens & Featherman, 1981) revealed that mean
occupational attainment scores for AAL fathers were 28. 55 which reflect technical repair

occupations (e.g., structural metal craftsmen). Mean occupational attainment scores for

40

41

 

 

 

 

 

 

 

Table 2
J 0 _ 0 U ’ H h“) ‘0 1-5. [531 ill ' -.I.| U ‘-m kit A .I 61.3. 3. ” All 0 ' : h‘r‘
FATHER’S CLASSIFICATION
AAL NAAL NC Total
Father 7 Father 5 Father 6 F A 18
HI Mother 5 Mother 5 Mother 6 MO 16
TBP Son 6 Son 6 Son 5 S 17
Daughter 3 Daughter 2 Daughter 4 D 9
SON’S
CLASS Father 6 Father 8 Father 6 FA 20
L 0 Mother 5 Mother 7 Mother 4 MO 16
TBP Son 5 Son 7 Son 5 S 17
Daughter 1 Daughter 1 Daughter 2 D 4
FA 13 F A 13 FA 12
MO 10 MO 12 MO 10
S 1 l S 13 S 10
D 4 D 3 D 6

42

 

 

 

 

 

 

 

Table3
II‘ ': in; .orum'uourur 1-: FM'I‘l! .‘ o u ' mac n i-i‘ ii
111' {5' 'I' “III
AAL NAAL NC
N=l3 N=13 N=12
M SD M SD M SD F
Father
Age 32.14 3.66 3127‘ 3.88 3488" 4.17 3.64*
Education (Years) 1277‘ 2.17 13.23d 1.79 15.25“! 2.70 4.30“
Degree'I .15 .55 .46 .78 1.00 1.60 1.50
Socioeconomic
Status” 28.55 14.13 30.09 18.46 42.28 22.41 2.03
Family Income 24§23 11,755 36,230 23,398 43,833 19,793 3.17
AAL NAAL NC
N=10 N=12 N=10
M SD M SD M SD F
Mother
Age 28.90 3.39 31.72 4.76 31.71 3.52 1.71
Education (Years) 13.00 1.33 13.25 2.83 13.90 2.13 .43
Dagree‘I .30 .67 .58 .90 .60 .97 .21
Socioeconomic
Status” 26.49 7.92 28.91 16.96 24.59 5.08 .38
'See text for coding formula

I’Duncan TSE12 (Stevens & Featherman, 1981)
°"'Means labeled with the same superscript difl‘er fiom one another at the p<.05 level of

significance.
'p<.05

43

NAAL fathers were 30.09 which reflect low level clerical occupations and craftsmen (e.g.,
postal clerks and oflice machine repairman). Mean occupational attainment scores for NC
fathers were 42.28 [e.g., health technician occupations, (radiological techriician)].

Results fi'om individual one-way AN OVAs found that NAAL fathers were
significantly younger than NC fathers. Additionally, results found that AAL fathers and
NAAL fathers had achieved significantly less years of education than NC fathers. These
difl‘erences in paternal education parallel those found in the larger sample (Ellis, Bingharn,
Zucker, & Fitzgerald, 1996). Univariate analyses failed to find any significant difl‘erences
between demographic variables for mothers in these families.

Table 4 presents the Time 1 sociodemographic characteristics of the children in the
sample who were originally classified by father’s Time 1 ratings Of impulsive aggression.
Univariate analyses of these variables did not show any significant difl‘erences between
these two groups.

Table 5 presents the current sociodemographic characteristics of the parents classified
by father’s Time 1 indices of antisociality and alcoholism. Univariate analyses found AAL
and NAAL fathers to have significantly less years of education than NC fathers. No

significant differences were found for mothers.

1:

1H

 

 

 

 

Table 4
rr‘ germanium” r in z. is; r r-i ii {'11 0 HO

Aggressixeness

High Impulsive Low Impulsive

Aggressive Aggressive

N=17 N=17

M SD M SD F
Son’s Age 4.04 .95 4.08 .86 .01
Father’s Years
of Education 14.29 3.06 13.47 1.77 .92
Mother’s Years
ofEducation 13.18 2.53 13.82 1.94 .70
Father’s Degree’ .88 1.41 .29 .69 2.40
Mother’s Degree' .47 .87 .65 .86 .35
Father’s Socioeconomic
Status” 32.20 19.83 34.88 19.77 .16
Mother’s Socioeconomic
Status” 24.01 4.93 32.04 15.82 3.99
Family Income 32,618 17,755 39,029 22,673 .84
'See text for coding formula

l’Duncan TSE12 (Stevens & Featherman, 1981)

Table 5

.I HI

3 it!!!)

‘31 '1' Fill 9 ‘1

45

1.: rm arr. .

I' 111' El 1.! . ill 11'

0
rep 0

3.1"!

 

 

 

 

 

 

 

AAL NAAL NC
N=13 N=13 N=12
M SD M SD M SD F
Father
Age 37.90 4.12 37.96 4.09 41.18 4.04 2.59
Education 1223" 2.45 13.62" 1.66 15.58“" 2.43 404*
Degree‘ .46 1.13 .54 .88 1.08 1.56 .96
Socioeconomic
Statusb 29.61 18.76 33.75 22.86 48.58 19.33 2.94
Family Income 45,885 27,502 44,962 23,164 53,125 19,746 .44
AAL NAAL NC
N=10 N=12 N=10
M SD M SD M SD F
Mother
Age 33.51 3.37 38.58 4.59 38.04 3.84 1.74
Education 13.20 1.87 13.67 2.67 13.90 2.13 .24
Degree’ .30 .67 .67 .89 .60 .97 .56
Socioeconomic
Status” 27.03 8.42 32.18 19.05 33.11 14.54 .48
‘See text for coding formula

l’Duncan TSE12 (Stevens & Featherman, 1981)
”Means labeled with the same superscript difl‘er from one another at the p<.05 level of

significance.

‘p<.05

46

Hypothesis 1
Adult Impulsive Aggression

To test hypothesis 1, which predicted lower 5-HT in AALs when compared to
NAALs, two analytic strategies were utilized. Given the literature’s focus on categorical
comparisons of clinical samples and the resulting database from which comparisons
between various clinical groups can then be made, it was decided to first use AN OVAs to
test mean difl‘erences in whole blood S-HT. Results fiom this approach could then
contribute data to the existing literature regarding mean level whole blood 5-HT data fi'om
not only two well documented subtypes of alcoholics (AALs and NAALs), but also the
much less studied population of normal community controls (NCs). Ultimately, this
approach provides a metric from which comparisons can be made between a wide variety
of clinical and normal control samples. Secondly, correlational analyses using the range of
ASB scores were deemed necessary in examining the relationship between whole blood 5-
HT and impulsive aggression due to increased statistical power. Additionally, it was
assumed that the level of pathology differs between the individuals within each
dichotomous grouping and that correlational analyses allow meaningful clinical
information to be obtained.

First, a one way ANOVA with the independent variable of father’s alcohol
classification at Time 1 (AAL, NAAL, or NC) and father’s S-HT as the dependent variable
was performed. Results failed to show a significant main efl’ect and did not support a

difl‘erence in 5-HT level between AALs and NAALs, see Table 6.

47

 

Table6
-11 ms :.0‘ 0 011' i: as. s A .1 li‘ “atrium :r.:.r ‘ r z..I‘-i‘
s .010". :1. : 11‘ as. ass a an Ii‘hmranm :r.:.-‘0
W
Source of Variation Sum of DF Mean F Sig.
Squares Square of F
Between Groups 441.12 2 220.56 .13 .87
Within Groups 57526.25 35 1643.61
Total 57967.37 37
new: .=.it.:0.i‘r=...01 :H\’ 0 .1" r: at. s r .1 Ii“ Hug-ricer
r 0 0 ’ . 101011 1 ii .. \.. k =10 Ii‘h‘u‘emmi

 

 

AAL NAAL NC Total
N=13 N=13 N=12 N=38
Father’s 5-HT M=157.69 M=157.46 M=150.25 M=155.26

(ng/ml) SD=34.1 1 SD=48.06 SD=37.96 SD=39.58

48

WW: To test whether alcoholism at Time 1 was
significantly related to decreased current S-HT, a one way AN OVA with the independent
variable of father’s alcoholism classification at Time 1 (alcoholic or non-alcoholic) and
father’s 5-HT as the dependent variable was performed. Thus, AAL and NAAL fathers
were combined into a Time 1 alcoholic group while non-alcoholic controls were used as
the comparison. Results failed to show a significant main effect and did not support the
hypothesis of differences in current 5-HT between Time 1 alcoholic fathers and Time 1
non-alcoholic fathers, see Appendix B, Table B1.

m: Given the lack of significant differences in father’s S-HT by father’s
classification as AAL, NAAL, or NC in the initial analysis, the potential effect of
covariates was examined. First, a correlation between father’s 5-HT and variables other
than AAL, NAAL, or NC classification hypothesized to be related to father’s serotonin
was conducted. These variables included, measures of father’s current alcohol use as
reported on the health history questionnaire administered at blood draw (1. quantity X
frequency=average drinking days per week in last 4 weeks multiplied by usual number of
drinks consumed per drinking occasion; 2. in last 4 weeks, average number of days alcohol
consumed per week; 3. average number of drinks consumed per occasion in the last 4
weeks), measures of dad’s current cigarette use as reported on the health history
questionnaire (1. quantity X frequency= average days smoking in last 4 weeks multiplied
by average number of cigarettes smoked per occasion; 2. average number of days smoked
cigarettes in past 4 weeks, 3. average number of cigarettes per occasion in last 4 weeks ),

father’s age, and seasonal variation (amount of daylight calculated by splitting the calendar

49

year by the equinox). Average number of days smoking was the only variable significantly
correlated with father's 5-HT, see Appendix B, Table BZ. A simple factorial AN OVA,
with father’s classification as the independent variable, father’s 5-HT as the dependent
variable, and father’s average number of days smoking as a covariate failed to show a
significant main effect for father’s classification, see Appendix B, Table B3.

Next, the relationship between father’s impulsive aggression and serotonergic
firnction was axarnined by two one-tailed Pearson product moments correlations. First, a
correlation was conducted between father’s 5-HT and father’s Time 1 total score on the
Antisocial Behavior Checklist. Results did not show a significant correlation (N=38, r = -
.07, p = .34). Second, a one-tailed PPMC was conducted between father’s 5-HT and
current total score on the ASB. Results failed to show a significant correlation between
fathers’ 5-HT and current impulsive aggression as measured by the ASB (N=3 8, r = .19, p

= .13). These results failed to support Hypothesis 1.

50
Hypothesis 2
Childhood Impulsive Aggression

Hypothesis 2, which predicted greater serotonergic dysfunction in high impulsive
aggressive children when compared to low impulsive aggressive children was also tested
by categorical comparisons and correlations for reasons identical to those outlined in
Hypothesis 1.

While father’s ratings were used to categorize their sons into high and low impulsive
aggressive groups during the design stage of this study, later findings fi'om the longitudinal
study indicated that maternal ratings of children were more accurate indicators of
impulsive aggression (Loukas at al., 1997). Therefore, in all subsequent analyses,
maternal ratings of child behavior were used to categorize the children.

WW: First, a 2 (mother’s classification
of son at Time l=high or low impulsive aggression) X 3 (father’s alcoholism status at
Time 1=AAL, NAAL, NC) AN OVA with son’s whole blood 5-HT as the dependent
variable was conducted. While neither of the main effects were significant, results
indicated a significant interaction. See Table 7 for AN OVA table, means, and standard
deviations.

Next, the potential effects of covariates were examined. Results from analyses with
sons corroborated findings in the literature supporting a significant negative correlation
between serotonergic function and age for children in the 10-12 age range [N = 34,

L = -.36, p = .02; (PPMC)] as well as a positive relationship between 5-HT and amount

51

 

 

 

 

 

 

 

 

Table 7
4111.13 :_1 ‘ t “10131 .1510. t 01:, =I ‘1 t
s'°1'-:~' . l:1°-r a .11111 .=1 - 11‘ -s. \66 L
Sum of Mean
Source of Variation Squares DF Square
Main Effects 16828.163 3 5609.388
MBOYGRPl 4234.220 1 4234.220
RISK 16290.463 2 8145.232
2-Way Interactions 46073.764 2 23036.882
MBOYGRPI RISK 46073.764 2 23036.882
Explained 46811.241 5 9362.248
Residual 106292.524 28 3796.162
Total 153103.765 33 4639.508
U'JI‘. .11L:1 I; iti11 k, 1 hat is, .Lfrro
-e°11 .1.111.'s,°°1‘-:~. . ‘21‘ $.11011. :i
as; k as. s r 1 11 1611-“ 1'11 'kmz-ru'vri
AAL NAAL NC
Low Impulsive =237.33 =237.86 M=l98.75 M=224.65
Aggression SD=60.48 SD=63.37 SD=62.92 SD=62.26
N=6 N=7 =8 N=21
High Impulsive M=194.00 M=203.17 M=351.00 M=249.39
Aggression SD=15.23 SD=83.65 SD=16.97 SD=38.62
N=5 N=6 N=2 N=13
M=216.67 M=220.52 M=274.88
SD=37.86 SD=73.51 SD=39.95
N=11 N=13 N=10

,1. 111

as.

F

.478
.115
.146

.068
.068

.466

6 A
Sig
of F
.242
.300

.136

.006
.006

.057

52
of available sunlight (N = 34, r = .29, p = .049, Spearman Correlation).

Therefore, the same 2 X 3 AN OVA with son’s 5-HT as the dependent variable was
performed adding both age and season as covariates. Results with these covariates added
did not change the relationship (see Appendix C, Table CI for AN OVA table and adjusted
means). Neither the covariates nor the main effects were significant. As previously found,
results indicated a significant interaction. Contrary to expectations, for non-alcoholic
control fathers, sons who were rated as low impulsive aggressive had significantly lower
mean 5-HT than those sons rated as high impulsive aggressive. This result may be
spurious and may be related to the small number of subjects in the high impulsive group
(N = 2).

WW: Analyses were then performed
with mother’s ratings of son’s current behavior. First, a one-way simple factorial AN OVA
with mother’s classification of son’s current impulsive aggression as the independent
variable (high impulsive aggression=clinical range as measured by CBCL Total Behavior
Problem T-scores of 63 or greater, low impulsive aggression=borderline clinical and
normal range as measured by Total Behavior Problem T-scores of 62 or less) and son’s 5-
HT as the dependent variable was conducted. The main efl‘ect was not significant, see
Table 8. Second, the same ANOVA using son’s 5-HT as the dependent variable was run,
adding both son’s age and season of blood draw as the covariates. While the covariate of
son’s age was significant, the second covariate, season, and the main effect were not. See

Appendix C, Table C2 for ANOVA table and adjusted means.

53

 

Table8
, 4111.3‘ A" 0 .11' A: it. 5 A .1 '1' 1161111611 isn‘t (111;;
_. .1.:1111 11 .1 611 3‘12“ 11,111 .1. 111 , ' ‘ 5'91“... ‘ hertz-111m
W
Source of Variation Sum of DF Mean F Sig.
Squares Square of F
Between Groups 15435.66 1 15435.66 3.52 .07
Within Groups 135965.85 31 43865.00
Total 151401.52 32
112:1. .51'.3'.x'lil.01 :11, 1 .11'11 as. s A .1I1‘ 11311131114
11:.1‘1 1111‘ . 13-111 11 .-1‘1 3‘13“ 11,911 .1. 111..

 

 

HI IMPULSIVE LOW IMPULSIVE
AGGRESSIVE AGGRESSIVE Total
N=6 N=27 N=33
Son’s S-HT M=178.00 M=234.07 M=206.04

(rig/ml) SD=71.70 SD=65.12 SD=68.41

54

WW: Given these suggestive
findings, similar analyses were conducted on the sample of daughters. First, a one-way
AN OVA with mother’s rating of current behavior as the independent variable and
daughter’s 5-HT as the dependent variable was conducted. The main effect was
significant, with high impulsive aggressive daughters having lower whole blood S-HT than
low impulsive aggressive daughters, see Table 9. Second, an AN OVA with daughter’s 5-
HT as the depcndent variable was performed adding daughter’s age and season of blood
draw as the covariates. While neither the covariate of daughter’s age or season was
significant, the main effect remained highly significant, see Appendix C, Table C3. Given
that the high impulsive aggressive group consisted of only two subjects, a non-parametric
test was run on the two groups. Results from a Mann Whitney U-Wilcoxon test
confirmed the previous analyses (W = 3.0, two-tailed P = .03).

W: Before
performing analyses on the larger combined sample of sons and daughters (N=45), we
explored the possibility of a significant main efl‘ect by gender for the dependent variable of
child’s 5-HT. While the literature does not support a significant difl‘erence in S-HT related
to gender, we ran a series of analyses on our own sample to verify this observation. As
expected, an analysis of variance using gender as the independent variable and child’s 5-
HT as the dependent variable failed to show a significant main efl‘ect, see Appendix C,
Table C4. Second, an AN OVA using gender as the independent variable and child’s S-HT
as the dependent variable was run adding both age and season as covariates. Neither the

covariates, nor the main efl‘ect were significant, see Appendix C, Table C5 for

55

° ’
1111.21 :.1' 1 011' i: as. s A 1'1“ rung-111611 21:.1'1 1.11.191
. O , . O I .
1.:1111 |:.'1'-1 -1111 -'1.111 11°11 .1. 111. ‘sru‘. ‘ :11

 

 

Source of Variation Sum of DF Mean F Sig.
Squares Square of F
Between Groups 7661.40 1 7661.40 6.37 .03
Within Groups 12036.60 10 1203 .66
Total 19698.00 11
176::1 31.1511 L‘ 1:.111‘. 1111;, 1 911‘ r: s k. s A .1 I1.‘ 11a1r-111a1
:.1-.1'1 111111 :.1.-111 Us.°1' .1‘13'1211C1°°11 1.

111 . ' ‘ s “ac-m ' :11 hwy-1111111 :1:.1 ‘ 1 Dr. '1 ’ um .111 ‘ 3 1.1.1 -

numb

 

HI IMPULSIVE LOW IMPULSIVE
AGGRESSIVE AGGRESSIVE Total
=2 N=10 N=12
Daughter’s 5-HT M=150.50 M=218.30 M=184.40

(ng/ml) SD=1 3 .44 SD=36.29 SD=24. 87

56

ANOVA table and adjusted means. Given these results, combining 5-HT scores across

gender was deemed appropriate; it also would increase power in analyses.

 

Next, an AN OVA with mother’s rating of child’s current behavior as the independent
variable and child’s current 5-HT as the dependent variable was conducted. Results
indicated a significant main effect, with children rated as high impulsive aggressive having
lower 5-HT levels than children classified as low impulsive aggressive, see Table 10.
Second, a similar AN OVA with child’s 5-HT as the dependent variable was performed
with both child’s age and season as the covariates. While neither of the covariates was
significant, the main effect of child classification remained highly significant, see Appendix
C, Table C6 for AN OVA table and adjusted means. Table 11 provides a description of
the sociodemographics of these children classified by mother’s rating of current behavior.

Importantly, this is the first study to show an inverse relationship between childhood
impulsive aggression and whole blood S-HT. These results indicate that the inverse
relationship between serotonergic function and impulsive aggression found in adult
samples can also be found in a high risk sample of children who exhibit similar behavioral
characteristics. The lack of previous findings is partially attributable to the dearth of
studies which examine the relationship between the childhood behavioral characteristics
of impulsive aggression and serotonergic function in children. However, the lack of such
findings and the inconsistencies in the literature regarding this phenomena may also result
fiom methodological differences and inadequacies in the existing studies (Pine et al.,

1996). It has been suggested that studies of children are hampered by the use of small

57

TablelO
° ’
111111 :.1‘ 1 011' i: as. s A .1 '1' Han-1111111 21:.1'1 111.111‘
:‘1.:11111" .-1'1:‘1.:1111°11 .1. 111.‘s°°1'-r..‘ ‘ :11I1'
. 0

 

 

 

Source of Variation Sum of DF Mean F Sig.
Squares Square of F
Between Groups 22654.03 1 22654.03 6.45 .02
Within Groups 12036.60 43 1203.66
Total 19698.00 44
112:1 .:11.:1|I‘r:111 :11 ’ 1 ”1": as. s A 1I1' 1111131111111
111.1‘1 1111‘ 1.. .1.=.111 11 1‘1 3‘1.-1111°11 .1. 111. '
s°°1‘-:~.. ‘ :1.1 '1' 9161111111111 :1:_1‘ 1 11" 1111- {1,1,1 -.I 11 11
HI MULSIVE LOW NPULSIVE
AGGRESSIVE AGGRESSIVE Total
=8 N=37 N=45
(6 Boys, 2 Girls) (27 Boys, 10 Girls)
Child’s 5-HT M=l7l.13 M=229.8l M=200.47

(ng/ml) SD=62.13 SD=58.67 SD=60.40

58

Table 11

C C O O O O . - .. —
1 11111111911:11 1 .1°1:..11 .1.111.":.°°1*..“ 11.411; ‘ 11 -

  

 

High Impulsive Low Impulsive

 

 

 

Aggressive Aggressive

N=8; 6 boys, 2 girls N=37; 27 boys, 10 girls

M SD M SD F
Child’s Age 11.26 1.53 10.31 1.65 .61
Dad’s Years
ofEducation 13.00 1.73 15.03 2.72 3.57
Mom’s Years
of Education 12.88 2.47 14.19 2.33 2.22
Dad’s Degree' .29 .76 1.08 1.50 1.86
Mom’s Degree' .38 .74 .76 .93 .86
Dad’s Socioeconomic
Status” 25.33 6.07 43.69 24.98 3.67
Mom’s Socioeconomic
Statusb 35.01 19.95 32.30 14.20 .01
Family Income 4L071 20,044 53,722 25,594 1.52
‘See text for coding formula

”Duncan TSE12 (Stevens & Featherman, 1981)

Note: one boy did not have a rating from mom, hence number of boys =33 not 34.
Analyses with fathers variables include a family where the father had committed suicide.
The son was rated low impulsive aggression and the girl was rated high impulsive
aggression. Hence the analyses of paternal demographic variables for both high and low
groups have an N of 7 and 36 respectively.

59
samples; typically 30 subjects or less. Additionally, few of the existing studies have taken
seasonal rhythms on serotonergic measures into account.

- -,,~-. :Next, the

 

relationship between child serotonergic function and impulsive aggression was tested by a
series of Pearson product moment correlations. The first set of analyses examined the
relationship between current child 5-HT and child behavior problem ratings at Time 1.
Since 7 of the 13 girls in the sample were enrolled in the parent study at time periods
following Time 1, T1 ratings and classifications were available for only 6 girls (total
sample N=40). Additionally, it should be noted that one mother’s rating of her son at Time
1 was missing, hence only 39 subjects are available for analyses using mother’s Time 1
ratings. A one-tailed PPMC was run between CBCL total behavior problem raw scores,
total behavior problem T-scores, classification as high or low impulsive aggressive, and
current 5-HT. No significant correlations between childs’ S-HT and Time 1 ratings of
child were observed, see Table 12.

Although under only special circumstances would a covariate analysis render a non-
significant zero order relationship significant, nonetheless, for sake of completeness, an
examination of possible efl‘ects of covariates was conducted through a correlational
analysis, controlling for childs’ age and season. As expected, none of these correlations
were significant, see Appendix C, Table C 7.

The next set of analyses examined the association between current child 5-HT and

current child behavior problem ratings. First, a one-tailed PPMC was conducted between

60
Table 12

. O ‘.- -' I , ' ’ O - . .. G . -
11 :.111 11114-11 11 .1-11-1I 1°11 .11 11 1.-.- 1.111 1111-111

 

 

1 2 3 4
1.Childs’ 5-HT 1.00 -.12 -.06 -.01
(us/ml)
2.TBP Raw Score 1.00 .96" .75“
3.TBP T-Score 1.00 .82"
4. Impulsive 1 .00
Aggression
Classification
‘p<.05, ”p<.005

One-tailed Pearson Product Moment Correlations
=39 for those analyses using mother’s TBP raw score, mom’s TBP T-score, and mom’s

classification, due to one missing rating of a son by his mother.

61

current CBCL total behavior problem raw scores, total behavior problem T-scores,
classification of child as high or low impulsive aggressive, and current S-HT. As predicted,
results indicated significant negative correlations between child’s’ 5-HT and all three
ratings of child behavior by mother, see Table 13. To examine the possible effects of
covariates, the same correlations were re-run, controlling for both child’s age and season
of blood draw. As expected, all ratings of children remained significantly negatively
correlated with childs’ 5-HT. See Appendix C Table C8 for complete listing of

correlations.

62

Table 13

I.
-
.
-
-
O
.
-
¥
.

O . O , .
11':.11‘-'-11.1r.‘-11 11 -1I 1'11 :11 _1-11

 

 

1 2 3 4

1.Childs’ 5-HT 1.00 -.33* -.34* -.36*
(us/ml)

2.Mom’s TBP 1.00 .96" .69"

Raw Score

3.Mom’s TBP 1.00 .61"
T-Score

4.Mom’s Class. 1.00
Of Child

°p<.05, ”p<.005

One-tailed Pearson Product Moment Correlations

63
Hypothesis 3
Intrafamilial Correlations in Serotonergic Function

To examine the relationship between parental 5-HT and child 5-HT, a correlation
matrix was performed between father’s 5-HT, mother’s 5-HT, son’s 5-HT, and daughter’s
S-HT, see Table 14. Results indicated that son’s 5-HT was not significantly correlated
with father’s 5-HT nor mother’s 5-HT. To examine the potential efl‘ects of covariates a
correlation between son’s 5-HT, father’s 5-HT, and mother’s S-HT was performed,
adding son’s age and season of son’s blood draw as covariates. Results indicated that
son’s 5-HT was not significantly correlated with fathers’ 5-HT (N=32, r = .07, p = .35) or
mothers’ 5-HT (N=29, r = .23, p = .12).

Results from the first correlational analysis above found that daughters’ 5-HT was
significantly correlated with father’s 5-HT, but was not significantly correlated with
mother’s S-HT, see Table 14. To examine the potential effects of covariates a correlation
was performed between daughter’s 5-HT, father’s 5-HT, and mother’s 5-HT, adding
daughter’s age and season of daughters’ blood draw as covariates. Daughters’ 5-HT
remained significantly correlated with fathers’ 5-HT (N=12, r = .61, p = .03), but was not
significantly correlated with mothers’ S-HT (N=11, r = .05, p = .45).

While no formal predictions were made in regard to the relationship between S-HT
in brothers and sisters, exploratory analyses were performed. Results from the original
correlation between all family members indicated that brother’s S-HT was significantly
correlated with sister’s 5-HT, see Table 14. To examine the potential effects of covariates

a correlation was performed between brother’s 5-HT and sister’s 5-HT, adding brother’s

Table 14

IE'I'ICI"S 'E'

 

 

1 2 3 4
1. Son’s 5-HT 1.00 .59“ .20 .10
(ng/ml) (12) (29) (32)
2. Daughter’s 1.00 .08 .61*
S-HT (11) (12)
(Hg/ml)
3. Mother’s 1.00 .20
5-HT (31)
(us/ml)
4. Father’s 1.00
5-HT
(us/ml)
'p<.05, ”p<.005

One-tailed Pearson Product Moment Correlations
( )=number of subjects

65

age, sister’s age, and season of blood draw as covariates. Brother’s 5-HT was not
significantly correlated with sister’s 5-HT (N=12, [ = .51, p = .08). The non-significant
findings in this covariate analysis are most likely a product of low power resulting from
the use of three covariates and small sample size. Overall, the data suggest that there is a

positive relationship between 5-HT in brothers and sisters.

DISCUSSION

A large literature spanning the last 20 years describes an inverse relationship
between impulsive aggression and a variety of measures of serotonergic function in adult
alcoholics and clinical samples of non-alcoholics (Asberg et al, 1987; Buydens-Branchey
at al., 1989; Coccaro, 1989; Coccaro, Kavoussi, & Lesser, 1992; Linnoila, 1990;
Virkkunen & Linnoila et al., 1983; Virkkunen et al., 1987; Roy & Linnoila, 1988).
However, studies among children and adolescents with these behavioral characteristics
have only recently been undertaken and are rare. Additionally, these few studies have
been restricted almost exclusively to male clinical samples and the results have been
inconsistent (Birmaher at al., 1990; Halperin et al., 1994; Kruesi et al, 1990; Kruesi et al.,
1992; Pliszka et al., 1988; Stofl’ et al., 1987; Weizrnan et al., 1988). The relationship is of
particular interest as a phenotypic marker of biological vulnerability to impulsive
aggression, which itselfhas been hypothesized to be a risk factor for the development of
antisocial alcoholism.

The current study adds valuable information to the research literature through two
unique aspects of the study’s design. First, this study was designed to contribute further
data regarding the role of childhood serotonergic function in the development of impulsive
aggressiveness through a mixed gender sample of children . Secondly, this study was
designed to contribute information regarding serotonergic firnctioning in not only two
samples of alcoholics (AALs and NAALs), but also in healthy community controls, a

group that has been largely neglected.

66

67
Childhood Impulsive Aggression

The hypothesized relationship between serotonergic dysfunction and childhood
impulsive aggression was largely supported by the present findings. Although results are
not entirely uniform, in instances where the most powerfirl tests (large N) were run and
where the phenotypic classification of impulsive aggression was most proximal to the
assay of serotonergic firnction (i.e., ratings of current child functioning), results
unambiguously show the hypothesized negative relationship. Thus, results showed that
children currently rated as high impulsive aggressive had a significantly lower level of
whole blood S-HT than children currently rated as low impulsive aggressive, but no
parallel finding was found when using Time 1 classifications of children. However, results
fiom the analysis using Time 1 ratings of children are hard to interpret given the low
number of children in the control group who were classified as high impulsive aggressive.

Relatedly, while a negative and substantial correlation was found between children’s
current 5-HT and current rating of impulsive aggression as measured by CBCL Total
Behavior Problem Raw Scores, Total Behavior Problem T-Scores, and child classification,
no parallel relationship was found for Time 1 ratings of children.

Importantly, the present findings give firrther support to the majority of available
studies on children and adolescents which suggest that decreased serotonergic function is
positively related to impulsive aggression, (Brown et al., 1986; Kruesi et al., 1990;
Kruesi, et al., 1992; Birmaher at al., 1990; Stofl‘et al., 1987). However, the lack of an
observed relationship between current serotonergic function and developmentally earlier

(i.e., Time 1) impulsive aggression make it dificult to be unequivocally confident about

68

the initial model. Further research is necessary to adequately test the hypothesized role
serotonergic dysfunction may play in the developmental trajectory of childhood irnpulsivity
and conduct problems, adolescent aggressiveness and alcohol use, and later alcoholism.
Intrafamilial Correlations of Serotonergic Function

The present findings also provided only partial support for the predicted positive
relationship between serotonergic fimction in parents and their ofl‘spring. The only parent-
child dyad to exhibit a statistically significant, positive relationship in serotonergic function
was fathers and their daughters. Importantly, there was no relationship between father’s
5-HT and son’s 5-HT. Given that the design of the study chose fathers and sons who
shared behavioral dimensions related to serotonergic function, this finding is in direct
contrast to the expected relationship.

Additionally, a statistically significant, positive relationship in serotonergic function
was found between brothers and sisters. While the relationships between fathers and their
daughters and brother and sisters were robust, the lack of farniliality for whole blood 5-HT
between mothers and their children and between fathers and their sons is in contrast to
data fiom studies of autism and attention-deficit hyperactivity disorder (Abramson et al.
1989; Cook et al., 1990; Cook, Stein, Ellison, Unis, & Leventhal, 1995; Kuperrnan,
Beeglrly, Burns, & Tsai, 1985; Leventhal, Cook, Morford, Ravitz, & Freedman, 1990).
Perhaps a more sophisticated analytical approach may be necessary to accurately measure

intrafarnilial correlations of whole blood 5-HT.

69
Adult Impulsive Aggression

The current study chose to examine serotonergic firnction in families headed by
antisocial alcoholic and non-antisocial alcoholic men; populations that have demonstrated
decreased serotonergic function. This population is of particular interest since parental
alcoholism has been shown to be a risk factor for the development of behavior problems
and later alcoholism in their offspring (Cotton, 1979; West & Prinz, 1987). Consequently,
this design allowed us to examine associations between childhood impulsive aggression
and serotonergic firnction in children at risk for later problems. Additionally, given that
much of the research suggests that behavioral disinhibition or impulsive aggression is the
behavioral correlate associated with decreased serotonergic function, we chose to exanrine
populations marked by higher and lower impulsive aggression (AALs and NAALs
respectively).

While it was hypothesized that impulsive aggressive alcoholics (AALs) would
exhibit decreased serotonergic function in comparison to non-antisocial alcoholics
(N AALs), results did not support this hypothesis. Relatedly, results failed to show
significant differences in serotonergic function between alcoholics (AAL + NAALs) and
non-alcoholic controls (N C8).

The lack of expected findings for decreased serotonergic firnction in alcoholics may
be related to the fact that this study classified fathers as alcoholic or non-alcoholic through
measures of behavior exhibited nearly 6.5 years prior to the current study (N = 38,

M = 6.59, SD = 1.13). While this sampling design was utilized in an attempt to maximize

the identification of a longstanding pattern of alcoholism, alcoholism diagnosis has been

70

shown to be an epiphenomenal process that ebbs and flows over time (Zucker et al.,
1997). Consequently, it would be likely that alcoholism diagnosis and impulsive
aggression have changed for some of the men in the current sample over the intervening
years since classification at Time 1. Hence, more current indices of alcoholism may be
necessary to accurately assess the relationship between alcoholism and serotonergic
firnction.

Results fiom correlational analyses also failed to support the predicted negative
relationship between fathers’ serotonergic function and impulsive aggression as measured
by the total score on the Antisocial Behavior Checklist (Zucker & Noll, 1980) both at
Time 1 and the current time period. These results are in contrast to the large literature
documenting a negative relationship between impulsive aggression and serotonergic
fiinction (Coccaro, 1989; Asberg et al., 1987; Roy & Linnoila, 1988; Linnoila et al., 1983;
Linnoila et al., 1985). It is possible that the construct of impulsive aggression
hypothesized to be most closely associated with behavioral disinhibition or the propensity
to irnpulsively and violently react to irritable or noxious stimuli is more accurately
captured by measures other than the Antisocial Behavior Checklist.

Methodological Limitations

Given that the likelihood of a Type 1 error was increased as a result of the large
number of statistical tests performed, the reader is alerted to interpret the statistically
significant results with caution. While all significant results reported herein are judged
worthy of attention and interpretation, they should be subject to replication. In relation to

the subject pool, it should be noted that since this sample is restricted to non-Hispanic

71

Caucasian families, the results may not be generalizable to other ethnic groups and races.
Additionally, given that the study design attempted to maximize the presence of
alcoholism in all but the control fathers, results may not generalize to non-alcoholic

Given the literature documenting an inverse relationship between impulsive
aggression and serotonergic fimction in adults, and the lack of such results in the present
worlg this study may have been restricted by the use of Antisocial Behavior Checklist
scores as a measure of impulsive aggression. While this instrument measures a variety of
aggressive behaviors (gang fighting, hitting a teacher, beating up another person, teasing
or killing an animal) as a whole, it may not adequately target the dimension of irritable,
impulsive aggression hypothesized to result fi'om behavioral disinhibition. Other measures
including the Buss-Durkee Hostility Inventory Assault Scale (Buss & Durkee, 195 7),
Aggression score from the Life History Aggression interview (Brown et al., 1979), total
score from the self report Barratt Impulsiveness Scale (BIS-l l; Barratt, 1985), and DSM-
IV antisocial personality diagnoses may be useful in accurately measuring impulsive
aggression.

Similarly, firture investigations may want to replicate the present findings from the
analyses with children by using irnpulsivity and aggression measures such as the Brown-
Goodwin Assessment for Life History of Aggression (Brown et al., 1979; Brown et al.,
1982), the aggressiveness scale from the CBCL (Achenbach & Edelbrock, 1983), the
Iowa Conners Aggression Factor from the Conners Rating Scales (Conners, 1990), and

DSM-IV Conduct Disorder diagnoses.

72

While this study attempted to capitalize on the earliest behavioral manifestation of
impulsive aggression in children, results indicate that this construct may be
epiphenomerul, not static, and change considerably over time. Relatedly, the current
study may have been confounded by the use of Time 1 measures of alcohol problems in
adults which in many cases may not have accurately reflected current alcohol diagnosis.
Importantly, this study may have been limited by the small sample of participants which
may have produced a lack of power for some analyses.

Another limitation of this study is the use of only one measure of serotonergic
function, whole blood platelet S-HT to characterize central serotonergic function. While
future studies may want to replicate the current findings with whole blood platelet S-HT, it
will also be important to gather other indices of central serotonergic function, including
postsynaptic measures.

Implications

One interpretation that can be made of these data is that the observed negative
relationship between current childhood impulsive aggression and serotonergic fimction
may indicate the presence of a phenotypic marker of biological vulnerability to impulsive
aggression. While the expected relationship between current serotonergic fimction and
Time 1 impulsive aggression was not found, this may be related to problems in the design
of the study. The potential for the observed relationship to represent a marker for later
development of antisocial alcoholism or other behavioral difiiculties will need to be

assessed by continued study of this population through time.

73

Given that the present data did not find an overall relationship between father’s
and son’s 5-HT, an alternative interpretation that needs to be explored is that the expected
relationship is not present. Clearly, additional research needs to be conducted to determine
this. “Within the context of the present study, this would entail increasing the number of
children available for the Time 1 high impulsive aggressive group with non-alcoholic
control fathers as well as increasing the overall sample size.

These findings suggest that it may be possible to identify children at risk for the
future development of problem behaviors. There is consensus that attending to behavioral
characteristics in children is important, especially in relation to impulsive aggressiveness.
Relatedly, parental ratings are one method of obtaining such information Importantly, the
current study has shown that maternal ratings of current child behavior are related to a
biological substratum within the child. This emphasizes the clinical importance of
attending to maternal ratings of children’s behavior and provides construct validity for
maternal ratings. In addition, early identification of at risk individuals makes it possible to
clinically intervene early in the life course, possibly disrupting the developmental
trajectory before such problems become more stable and problematic.

Future Directions

The results of the current study indicate that current serotonergic fimction is
negatively related to current impulsive aggression in male and female children and
differentiates between high and low impulsive aggressive children in a sample at risk for
the firture development of alcohol problems. These results are consistent with adult data

and suggest that childhood serotonergic dysfunction is related to impulsive aggressive

74

behavior which is hypothesized to be a result of behavioral disinhibition. Future studies
are needed to replicate the present findings and to continue to explore the relationship
between childhood serotonergic dysfirnction as it relates to the later development of
aggressiveness, antisociality, and alcoholism in AAL, NAAL, and NC families. Future
studies may also examine the relationships between social and environmental factors, such
as harsh parenting environments and childhood serotonergic function. Additionally, the
relationship between serotonergic function and other psychiatric problems such as
depression should be explored to determine whether these relationships parallel those

found in adults.

APPENDICES

APPENDIX A

Q1.

Q2.

Q3.

Q5.

PARENT HEALTH HISTORY
FORM BBS (8/95)

Within the past 3 years, have you had a regular physician or a clinic you usually
attend?

1 ............ YES
2 ............ NO [GO TO Q2]

Do you remember the date of your last physical examination within the past 3 years?

 

 

 

 

 

 

l ............ YES
2 ............ NO [GO TO Q3]
/ /
12 ear 0 last 31' at
What is your current height? ft. ins.
What is your current weight? lbs.

Have you been on a medication program for hyperactivity for a period of time, such
as ritalin or other medication in the past year?

1 ............ YES
2 ............ NO [GO TO Q6]

 

AGE STARTED: AGE ENDED:

 

 

 

Have you been on a medication program for any other long term or chronic condition
in the past year (such as asthma, cystic fibrosis, etc.)?

1 ............ YES
2 ............ NO [GO TO Q7]

 

MEDICATION AGE STARTED TO AGE ENDED REASON (CONDITION)

 

 

 

 

 

 

 

 

 

75

76

The next set of questions deal with other medications you have taken in the past month. If
you have taken the medication, write in the brand of medication taken, TOTAL number of
days you have taken the medication in the past month (estimate, if necessary), and the
reason(s) for taking them.

Q7. Have you taken any pain or fever relievers (aspirin, Tylenol, etc.) in the past month?

1 ............ YES
2 ............ NO [GO TO Q8]

 

 

 

 

. MEDICAHON DAYS REASON (MESS/CONDITION)

 

Q8. Have you taken any cough medicine (Robitussin, Pediacare, T riaminic, etc.) .in the
past month?

1 ............ YES
2 ............ NO [GO TO Q9]

 

 

 

 

MEDICATYON DAYS REASON (MESS/CONDITION)

 

Q9. Have you taken any decongestants/nasal spray (Sudafed, Dimetapp, Actzfed, etc. ) in
the past month?

2 ............ NO [GO TO Q10]

 

 

 

 

' MEDICATION DAYS REASON (MESS/CONDITION)

 

Q10. Have you taken any Antihistamines (Chlortrimaton, Acnfed, Benadryl, etc.) in the
past month?

1 ............ YES
2 ............ NO [GO TO Q11]

 

 

 

MEDICAUON DAYS REASON (HLNESS/CONDITIONL

77

Q11. Have you taken any multisympton cold remedies (Nyquil, Corididin, Catylenol, etc.)
in the past month? '

2 ............ NO [GO TO Q12]

 

 

 

 

MEDICATION DAYS REASON @LNESS/CONDIT ION)

 

 

Q12. Have you taken any antibiotics (Penicillin, Amaxicillin, etc.) in the past month?

1 ............ YES
2 ............ NO [GO TO Q13]

 

 

 

MQICAHON DAYS REASON (HLNESS/CONDITION)

 

Q13. Have you taken any asthma medication in the past month?

1 ............ YES
2 ............ NO [GO TO Q14]

 

 

 

 

@101 non DAYS REASON {H.LNESS/CONDITION]

 

Q14. Have you taken any allergy medication in the past month?

1 ............ YES
2 ............ NO [GO TO Q15]

 

 

 

. MEDICATION DAYS REASON (HLNESS/CONDmON) l

Q15. , Have you taken any vitamins/dietary supplements in the past month?

 

l ............ YES
2 ............ NO [GO TO Q16]

l MEDICATION DAYS REASON (WSS/COQHYOM i

 

 

 

78

Q16. Have you taken any neuroleptic medications (Thorazine, Mellan'l, Stelazine, Prolixin,
Haldol, Clozaril, Risperidol, etc.) in the past month?

2 ............ NO [GO TO Q17]

 

 

 

 

. MEDICATION DAYS REASON (MESS/CONDITION)

 

Q17. Have you taken any tricyclic antidepressants (Prozac, Zoloft, Desyrel, Etfexor, etc) in
the past month?

2 ............ NO [GO TO Q18]

 

 

 

 

MEDICATION DAYS REASON (HLNESS/CONDITION)

 

Q18. Have you taken any monoamine oxidase inhibitors (Nardil, Parnate, etc.) in the past
month?

2 ............ NO [GO TO Q19]

 

 

 

 

‘ MEICATION DAYS REASON (ILLNESS/CONDITION)

Q19. Have you taken lithium in the past month?

 

1 ............ YES
2 ............ N 0 [GO TO Q20]

 

 

 

MEDICATION DAYS REASON (HLNESS/CONDITIOM l

Q20. Have you taken any anticonvulsant medication (Depacote, Tegretol, etc.) in the past '
month?

2 ............ NO [GO TO Q21]

 

 

 

MEDICATION DAYS REASON (ILLNESS/CONDITION) ,

 

Q21.

Q22 .

Q23.

Q24 .

79

Have you taken any anti-anxiety medication (Buspar, Valium, Librium, Dalmane, etc.)
in the past month?

1 ............ YES
2 ............ NO [GO TO Q22]

 

 

 

MEDICATION DAYS REASON ILLNESS/CONDITION)

Have you taken any stimulant medication (Ritalin, Cylert, Dexedrine, etc.) in the past
month?

1 ............ YES
2 ............ NO [GO TO Q23]

 

 

 

MEDICATION DAYS REASON (ILLNESS/CONDIT ION)
Have you taken any anti-hypertensive medication (Inderal, Catapres, T enex, etc.) in
the past month?

 

1 ............ YES
2 ............ NO [GO TO Q24]

 

 

 

 

MEDICATION DAYS REASON (ILLNESS/CONDITION)

 

Have you taken any other medications in the past month?

1 ............ YES
2 ............ NO [GO TO 25]

 

 

 

 

MEDICATION DAYS REASON (HLNESS/CONDIT ION)

 

 

 

 

 

80

Q25. Have you been hospitalized in the past 3 years?

 

l ............ YES
2 ............ NO [GO TO 26]
AGE IN .. REASON: OPERATION OR ILLNESS # OF DAYS

 

 

 

 

99.ng

 

 

 

 

Q26. Have you been in an accident resulting in injury serious enough to require immediate
medical treatment (e.g., broken bones, concussion, stitches, burns,
poisonings/accidental ingestion, etc.) in the past 3 years?

 

l ............ YES
2 ............ NO [GO TO Q27]
AGE IN .. REASON: INJURY OR ACCIDENT # OF DAYS

 

 

 

 

99.ng

 

 

 

 

Cigarette Use

Q27. Have you ever smoked cigarettes?
_ Never __ Once or twice _ Occasionally, but not regularly
_ Regularly in the past _ Regularly now

Q28. Over the past 4 weeks, on average how many days did you smoke cigarettes?

days

Q29.

Q30.

81

Over the past 4 weeks, on days you did smoke, how many cigarettes did you usually

smoke?
cigarettes

When was the last time you smoked any cigarettes?
date of last cigarette or number of days since last cigarette

 

YOUR RECENT ALCOHOL USE

31.

32.

33.

34.

35.

36.

Over the past 4 weeks, on the average, how many days a week did you have a drink?
days a week

Over the past 4 weeks, on a day when you did drink, how many drinks did you
usually have in 24 hours? (One drink is a 12 oz. can of beer, a 4 02. glass of wine, or

a single shot of 80 proof) ‘
_ drinks per 24 hours

Over the past 4 weeks what is the most number of days per week you had a drink?
days

Over the past 4 weeks, what is the most number of drinks you’ve had in 24 hours?
drinks

Over the past 4 weeks, how long was the longest period you were abstinent (went
without drinking)?

_ days

When was that? (date) to (date)

Were your drinking patterns over the last 4 weeks fairly typical for you?

_No, it was less than usual _Yes, it was fairly typical

_No, it was more than usual

82

YOUR SPOUSE’S RECENT ALCOHOL USE

37.

38.

39.

40.

41.

42.

Over the past 4 weeks, on the average, how many days a week did your spouse have

a drink?
_ days a week

Over the past 4 weeks, on a day when you spouse drank, how many drinks did
he/she usually have in 24 hours? (One drink is a 12 oz. can of beer, a 4 oz. glass of
wine, or a single shot of 80 proof)

_ drinks per 24 hours

Over the past 4 weeks what is the most mnnber of days per week your spouse had a
drink?
__ drinks

Over the past 4 weeks, what is the most number of drinks your spouse had in 24

hours?
_ drinks

Over the past 4 weeks, how long was the longest period your spouse was abstinent

(went without drinking)?
days When was that? (date) to (date)

Were your spouse’s drinking patterns over the last 4 weeks fairly typical for him/her?

_No, it was less than usual _Yes, it was fairly typical
_No, it was more than usual

APPENDIX B

83

Table B1

c ' ’
nun :_.- 0 010‘ r: as. a r r Ir' rc'vmmc'n 21:.- ' o :_.-.,
s .0” u ' I H' u‘ a .0” 0 tor-5 )1. A .I 'I'.A‘vlh’t"-il

 

Source of Variation Sum of DF Mean F Sig.
Squares Square of F
Between Groups 440.77 1 440.77 .28 .60
“fithin Groups 57526.60 36 1597.96
Total 57967.37 37
UV! .HH'UA‘ALH int, 0 010' A: at. a r _l Iu' H‘vlh‘vil"!
z..,i-o_ .g‘.) 5 ,0..." ., Igor: _u‘ 5 ,0” o tor-.3 ,oro

 

 

ALCOHOLIC NON-
(AAL+NAAL) ALCOHOLIC TOTAL
N=26 N=12 N=38
Father’s S-HT M=157.58 M=150.25 M=155.26

(ng/ml) SD=40.83 SD=37.96 SD=39.58

84

How—o mm

 

2-"... 75.... "L... ... J. I. .. n... .rfl
_ N w A m a a m 0

~95 Boo .: .oo .oa .No .33. .Nm Lu ..3
Ptooso— O N m. foo um: .83.. .6 .8 .8 33 Lo
w.><m. Deva UREA _.oo .32. .8 .oo .8 JG .8
A.><m. Unawaogmmmos 800 .NA :3... .NN .. Hm -.~o
amaze—”mam o x m. Boo .mo: .09.... JB .. 8
m.><m. Umvcm mac—8a foo .mw: .23... )8
q.><m. 05.508ng foo -.~o now
marmo an Econ 68$ Boo -. ;
Pmommoa om Econ 632 :5

 

85

 

TableB3
Ian. I o I In Mr II: . a. a r I II “ms-nun In I
I ’ 5 .III II I II a; La; \ II‘ "var-nun a!‘. I
:t‘ .I‘I AIII‘imn 'I' war :II I :I.:‘0 =.I'-r‘a er:.°‘\_uw
Sum of Mean Sig
Source of Variation Squares DF Square F of F
Covariates 9988.072 1 9988.072 7.147 .011
DSMOKEZ 9988.072 1 9988.072 7.147 .011
Main Effects 465.469 2 232.734 .167 .847
RISK 465.469 2 232.734 .167 .847
Explained 10453.541 3 3484.514 2.493 .077
Residual 47513.827 34 1397.466
Total 57967.368 37 1566.686
5' 1%.! u I 3H ’ I In 3310' an. s A.l"'|“'1h"il..'i' :I.:.o'
I I‘ . .0001 I II 5. 85.5 \ JI‘D‘WII'I :-I=.I‘I
2.1;: .I ‘I MM : 0.0.. -.I H II I' I I. 0 2.1%: a 6m.“ L ..ll I161
AAL NAAL NC TOTAL
N=l3 N=l3 N=12 N=33
Father’s S-HT M=151.6l M=153.29 M=l60.54 M=155.26

(Hg/ml)

APPENDIX C

86

 

 

 

 

 

 

 

 

Table Cl
,IIIIJI :.I' I .U'l’vl 2. I.:IIII II: II -I°II .I.
III '.'°I' ' . ‘Il ’ a.III II .:I = II' It. L66
al.. I A. I l'_.'l'l"| :IzI I II 1‘! 'l' I°II :II Ir
WW
Sum of Mean
Source of Variation Squares DF Square F
Covariates 25247.488 2 12623.744 3.627
BSERAGE 13560.928 1 13560.928 3.896
BSEASONZ 5388.533 1 5388.533 1.548
Main Effects 996.149 3 332.050 .095
MBOYGRPl 2.206 1 2.206 .001
RISK 941.697 2 470.849 .135
2-Way Interactions 36355.325 2 18177.662 5.222
MBOYGRPl RISK 36355.325 2 18177.662 5.222
Explained 62598.962 7 8942.709 2.569
Residual 90504.803 26 3480.954
Total 153103.765 33 4639.508
;I. 313; “gm :II 5‘ I .UII' I .III I II II -.°I I
III . .“I ' I I I‘ . .III II I II —.‘.a, \.5
It. I II II ' Ir‘II‘I II I II ,I‘vm -II I°II II II
WWW
AAL NAAL NC
LOW Impulsive M=237.93 M=248.87 M=l92.64 M=226.48
Aggressive N=6 N=7 =8 =21
High Impulsive M=207.56 M=207.41 M=327.7O M=247.55
Aggressive N=5 N=6 N=2 N=13
M=222.75 M=228.14 M=260.17
N=ll N=l3 N=10

\

Sig
of F
.041

.059
.225

.962
.980
.874

.012
.012

.037

87

 

Table C2
IIII; I LI ‘ I .ILO 2.H II I I 9 a I I II' II H'rfil' 'I : I=.I I
\III‘ 'I III II I‘I 3 I:II CI°II .I. III ' .“I'
.A‘I‘IC'vjl z“) I II‘ II." Alt. {I‘m -.I I° II I ugh; I ;° II
Season
Sum of Mean Sig
Source of Variation Squares DF Square F of F
Covariates 25983.863 2 12991.931 3.343 .049
BSERAGE 15238.407 1 15238.40? 3.921 .057
BSEASONZ 4065.209 1 4065.209 1.046 .315
Main Effects 12706.553 1 12706.553 3.269 .081
MBGRPCUR 12706.553 1 12706.553 3.269 .081
Explained 38690.416 3 12896.805 3.318 .034
Residual 112711.100 29 3886.590
Total 151401.515 32 4731.297
4' ‘19.. U‘JI‘ II I‘ I III 2,61II ;\0 , I I 'I' IO‘ij‘IHC'val :I_:,I‘I
u I Ia’ I.=III I II .l'l 3 I: [I ‘!.'l I .I. III, ' ‘ a°'z‘ '
.A'OGjQI II I D 0 0|, I BI All. 5 0‘0” -.' I‘ I! 0 2.2 0 fi‘ .0
Season
HIIMPULSIVE LOW IMPULSIVE
AGGRESSIVE AGGRESSIVE TOTAL
N=6 N=27 N=33
Son’s S-HT M=179.58 M=232.50 M=206.04

(118/ ml)

Table C3

Source of Variation
Covariates

GSERAGE

GSEASONZ

Main Effects

MGGRPCUR
Explained
Residual
Total
6' 'v.I Ulil‘ :II ’
U. IA" 1 H A'III

a°u‘ 0 0'“ 0

Sum of
Squares

659.914
443.914
448.778

11704.098
11704.098

12364.012

7333.988

19698.000

88

11

II‘ II fi'xfil‘ HI
'.:[I ‘Q.'l|

.I an IIII ' 3 Iim

Mean
Square

329.957
443.914
448.778

11704.098
11704.098

4121.337

916.749

1790.727

A J IV

'.:[I I; ‘II

III I ' 3 I.I.I

O
I I D
.|J
O

F
.360
.484
.490

12.767
12.767

4.496

I. 31%;“. '8'

1H.

2'.
Sig
of F
.708
.506
.504

.007
.007

.040

 

HI IMPULSIVE LOW IMPULSIVE
AGGRESSIVE AGGRESSIVE TOTAL
N=2 N=10 N=12
Daughter’s 5-HT M=138. 19 M=230.61 M=184.40

(fig/ml)

89

 

 

 

Table C4
9 1"” ‘ 'vi 0 '
Source of Variation Sum of DF Mean F Sig.
Squares Square of F
Between Groups 2802.39 1 2802.39 .73 .40
Mthin Groups 173208.84 45 3849.09
Total 17601 1.23 46
u‘.=I. .:II.: I .A’-Ui.!0I. :II ’ I .1!” I: II. I A .I II' Irvin-rum
it?) ‘ I HIV.) :II .I‘ .A'u'lO‘I 2'2) ' I 1.0, .I an A‘IO ‘ 3 I,I,I -.I
(numb
BOYS GIRLS TOTAL
N=34 N=13 Nfl7
S-HT (ng/ml) M=222.65 M=205.38 M=214.02

SD=68.1 1 SD=40.93 SD=54.52

9O

TableCS
_uuza 21‘ o no" :uu: ; L0 a A .1 Ir uwumcm .1... ' 0
mm Daron“ 21:) - o ' ’ 1 am An ‘ = 0.0.. "I I! u :u
Sum of Mean Sig
Source of Variation Squares DF Square F of F
Covariates 10547.573 2 5273.787 1.398 .258
SERAGE 1852.247 1 1852.247 .491 .487
SEASON 5979.173 1 5979.173 1.585 .215
Main Effects 3225.463 1 3225.463 .855 .360
GENDER 3225.463 1 3225.463 .855 .360
Explained 13773.036 3 4591.012 1.217 .315
Residual 162238.198 43 3772.981
Total 176011.234 46 3826.331
5 " a! x; :zu‘ :I 0 \." 0 'III ‘ :.n,o {2 a L 0 a r 0 In' uc a~.auc an : f;.u ‘ I
LaoO' D an ac 'u : (z) ‘ o u. c’ .I an A“. ‘ 3 13.0 «-.I u° nu zu-
BOYS GIRLS TOTAL
N=34 N=l3 N47

 

5-HT (ng/ml) M=223 .28 M=204.75 M=214.02

91

 

TableC6
JHIIZQ 2_I' I My ‘2 ;\0 ; AI I' UI‘I‘vfiO'v}! 3|:j‘0 “In;
I.:.III II .l'vll :1.“ II'II .I. ll'.‘é"l' °
.A-.'l.-' =I:.I‘I II" ..I‘-II Ah":hh‘ "I I! II :II I :Iz“I a"
andfisasnn
Sum of Mean Sig
Source of Variation Squares DF Square F of F
Covariates 11747.428 2 5873.714 1.765 .184
SERAGE 2127.029 1 2127.029 .639 .429
SEASON 6493.749 1 6493.749 1.951 .170
Main Effects 25401.231 1 25401.231 7.632 .009
MGRPCUR 25401.231 1 25401.231 7.632 .009
Explained 37148.659 3 12382.886 3.721 .019
Residual 136457.919 41 3328.242
Total 173606.578 44 3945.604
aI ".'U°.3,I‘1l'\, I .n‘ A: at. a A .I II” II‘var‘vIII‘vJI :I:.I‘I
UII' .3...l_.3.010 I°I’ .I‘I 31:“ II'II .I. _III..‘6"J":~..
.'0'l"-l| =I:.I‘I II" .I'vll A‘.'-:.A.A. -II I'll :II I :Iz.‘-.~‘I .3’
andjmn
HIMULSIVE LOWHVIPULSIVE
AGGRESSIVE AGGRESSIVE TOTAL
N=8 N=37 N=45
(6 Boys, 2 Girls) (27 Boys, 10 Girls)
Child’s S-HT M=l 67. 80 M=233. 15 M=200.48

(fig/ml)

92

TableC7
II 'II 3311.46“ I I I I -II I II II II I - I II BII vm
{:HII IIII I I I I‘. II a II I IAIAI I). I\— 0 II =

 

1.Childs’5-HT 1.00 -.10 -.05 .00

 

2.TBP Raw Score 1.00 .96" .74"

3.TBP T-Score 1.00 .75"

4.Classification l .00
'p<.05, "p<,005

One-tailed Pearson Product Moment Correlations
N=39 for those analyses using mother’s TBP raw score, mom’s TBP T-score, and mom’s

classification, due to one missing rating of a son by his mother.

93

TableC8

. ' ._ -. ° ’ . , . I. ' I -
II 1..l, 3.11.30 II ,_I II -II l‘ II .II _IvII . I_:II II II
{34' 0H! I’ I IUI.;°‘:II 'AS_JIO :OAOAI [1.1. 05“ ' II =

 

1 2 3 4
1.Childs’5-HT 1.00 -.34* -.36** -.39**

2.Mom’s TBP 1.00 .96“ .70M
Raw Score

3.Mom’s TBP 1.00 .61"
T-Score

4.Mom’s Class. 1.00
Of Child

‘p<.05, ”p<.005

One-tailed Pearson Product Moment Correlations

LIST OF REFERENCES

LIST OF REFERENCES

Abramson, RK. ,,Wright H. H. ,,Carpenter R, Brennan, W. ,Lumpuy, 0. ,,Cole
E. ,& Young, S. R. (1989). Elevated blood serotonin in autistic probands and their first-

degree relatives WWW 397-407

Achenbach, T M & Edelbrock, C (1983) W
W. Burlington: Department of Psychiatry, University of

Vermont.

Anderson, G. M. ,Young, J. G, Cohen, D. J., & Schlicht, K. R (1981) Liquid
chromatographic determination of serotonin and tryptophan 1n whole blood and plasma.

WWI. 775- 776

Asberg, M. ,Traskman, L ,& Thoren, P (1976). 5 -HIAA 1n the cerebrospinal
fluid: A biochemical suicide predictor? W 1193-1197.

Asberg, M. ,Schalling, D. ,Traskman-Bendz, L. ,& Wagner A (1987). The
psychobiology of suicide, irnpulsivity,and related phenomena. In H. Meltzer (Ed. ),
- : ' ° (pp. 655-688). New york. Raven

 

Ballenger, J. C. ,Goodwin, F. K. ,Major, L. F. ,& Brown, GL. (1979). Alcohol and
central serotonin metabolism 1n man. W 224-227.

Bailly, D. ,Vignau, J. ,Lauth, B. ,,Racadot N. ,,Beuscart R. ,Servant,D., &
Parquet, P. J. ( 1990). Platelet serotonin decrease 1n alcoholic patients. Wanna
Scandinayicadl. 68-72

Banki, C. M. (1978). 5-Hydroxytryptamine content of the whole blood 1n
psychiatric illness and alcoholism. MW 232-238.

Banki, C. M & Arato, M. (1983). Amine metabolites and neuroendocn'ne
responses related to depression and suicide. W; 223 -232.

Barber, B.K., Olsen, J.E., & Shagle, SC. (1993). Associations between parental
psychological and behavioral control and youth internalized and externalized
behaviors. CW2. 1120-1136.

Barrat, E. S. (1985). Factor analysis of some psychometric measures of
impulsiveness and anxiety. BMW 547-554.

94

95

Bear, M.F., Connors, B.W., & Paradiso, MA. (1996). Neuroscience: exploring
the brain, Baltimore, Maryland: W1lliams & W1lkins.

Beck, C. M ,Borg, S. H,olmstedt, B. ,,Kvande H., & Skroder, R (1980).
Concentration of serotonin metabolites 1n the cerebrospinal fluid from alcoholics before

and dufing disulfimm therapy Actflhannacoloaicutloricolmfl. 305-307

Berzenyi, P. ,,Galateo E. ,Valzelli, L. (1983). Fluoxetine activity on muricial
aggression induced 1n rats by p-chlorophynylalanine. WW 333-338

Birmaher, B., Stanley, M. ,Greenhill, L. ,Twomey, J. ,Gavrilescu, A, &
Rabinovich, H. (1990). Platelet unipramine binding 1n children and adolescents with
impulsive behavior. ,: , . . °
22, (6), 914-918.

 

Boismare, F., Lhuintre, J .P., Daoust, M., Moore, N., Saligaut, C., & Hillemand, B.
(1987). Platelet affinity for serotonin is increased in alcoholics and former alcoholics: A
biological marker for dependence? WM 155-159.

Bokii, I. V. ,Kiseleva, I. P. ,Lapin, I. P. ,,Prakhe I B ,& Rybakova, T. G. (1984).
Changes 1n dexamethasone test indices and reduction 1n the rate of serotonin absorption by

thrombocytes In chronic alcoholics WWW
Korsakmjfi. 234-237

Branchey, L., Branchey, M., Shaw 8., & Lieber, CS. (1984). W
12, 219-226.

Brewerton, T. D ,Fl,ament M R, Rapoport, J L ,& Murphy, D. L (1993).
Seasonal effects on platelet 5-HT content in patients with OCD and controls. Amhiyesgf

W409

Brown, G.L, Ebert, M. ,,Goyer P. ,Jimerson, D. ,Klein, W. ,Bunney, W.E. Jr., &
Goodwin, F. K. ( 1982). Aggression, suicide and serotonin. Relationship to CSF

monoamine metabolites. Amencanloumalnffisrchimllz 741-746

Brown, G.L. & Goodwin, PK. (1984). Diagnostic, clinical and personality

characteristics of aggressive men with low 5-HIAA. WWW S408-
S409.

Brown, G..L ,Goodwin,FK., Ballenger, J. ,Go,yer P. ,&Major, L. (1979).
Aggression 1n humans correlates with CSF metabolites. PM], 131-139.

96

Brown, G.L., Klein, W.J., Goyer, P.F., Minichiello, M.D., Kruesi, M.J.P., &
Goodwin, F. K. ( 1986). Relationship of childhood characteristics to cerebrospinal fluid 5-
hydroxyindoleacetic acid 1n aggressive adults. In C. Shagass, RC. Josiassen, W. H

Bridger, K..J Weiss, D Stofl‘, & G..M Simpson (Eds), Warmer) 177-
179). Amsterdam, Netherlands: Elsevier.

Brown, G. L. & Linnoila, MI. (1990). CSF serotonin metabolite (5-HIAA) studies

in depression, impulsivity, and violence W (suppl. 4), 31-
41

Buss, A. H & Durkee, A. (1957). An inventory for assessing difi‘erent kinds of
hostility WWWL 343-348

Buydens-Branchey, L., Branchey, M H. ,,Noumair D. ,& Lieber, C. S. (1 989). Age
of alcoholism onset. II. Relationship to susceptabiity to serotonin precursor availability.

WW 231-236

Chamberlain, B. G. ,Ervin, F. R, Pihl, RO. ,& Young, S. N. (1987). The effect of
raising or lowering tryptophan levels on aggression in vervet monkey. manamkm

W503 510

Cloninger, R, Bohman, M., & Sigvardsson, S. (1981). Inheritance of alcohol

abuse: Cross-fostering analysis of adopted men. MW 861-
868.

Cloninger, C R. ,Sigvardsson, S. ,von Knorring, A-L., & Bohman, M. (1988). The
swedish studies of the adopted children of alcoholics. A reply to Littrell.lnuma1_qf

Mesonflcoholfl. 500-509

Cloninger, C. R, Von Knorring, L. ,& Oreland, L. (1985). Pentametric
distribution of platelet monoamine oxidase activity. 2W1; 133-143.

Coccaro, E. F. (.1989) Central serotonin and impulsive aggression. mm
W (suppl 8) 52-62

Coccaro, E. F. ,Kavoussi, RJ., & Lesser, J. (1992). Self- and other- directed

human aggression: The role of the central serotonergic system. WW
Emmphannacolmn. (supp! No 6) 70-83

Coccaro, E.F., & Murphys, BL. (1990). Serotonin in Major Psychiatric Disorders.
Washington DC: American Psychiatric Press.

 

97

Coccaro, E.F., Siever, L.J. Klar, H.M., Maurer, G. Cochrane, K. ,C.ooper,TB.,
Mohs, RC., & David, K. (1989). Serotonergic studies 1n patients with affective and

personality disorders WW 587-599

Conners, CK. (1990). W Toronto: Multi-Health
Systems, Inc.

Cook, E. H, Leventhal, B. L. ,,Heller W. ,Metz, J., Wainwright, M. ,& Freedman,
D. X. (1990). Autistic children and their first-degree relatives. Relationships between

serotonin and norepinephrine levels and intellignece. WWW
W2 268-274

Cook, E.H., Stein, MA, Ellison, T., Unis, A.S., & Leventhal, BL. (1995).
Attention deficit hyperactivity disorder and whole blood serotonin levels: Effects of
oomol'biditY- BMW 13-20-

C°°P°F 1R, Bloom, F-E., & Rath, RH. (1986). In; bimhemical basis of
mahannacolm (51h ed). New York: Oxford University pres,

Copenhaver, J. H. ,Schalock, RL. ,& Carver, M. J., (1978). Parachloro-D, L-

phenylalanine induced filicidal behavior 1n the female rat. PW
Beharior..8. 263-270

Cotton, N. (1979). The familial incidence of alcoholism: A review. W
SmdieunAlcohoLAQ. 89-116.

Daoust, M. ,Chretien, P. ,Moo,re N. ,Saligaut, C. ,Lhuintre, J. P. ,& Boismare, F.
(1985). Isolation and striatal (3H) serotonin uptake: Role 1n the voluntary intake of

$113301 by rats WWW“ 205-203

Daoust, M., Lhuintre, J .P., Ernouf, D., Legrand, E., Breton, P., & Boucly, P.
(1991). Ethanol intake and 3H—Paroxetine Binding. W 1977-1983.

Daoust, M., Saligaut, C., Chadelaud, M., Chretien, R, Moore, N., & Boismare, F.
(1984). Attenuation by antidepressant drugs of alcohol intake in rats. AlgohoLl, 379-3 83.

Da Prada, M., Cesura, AM., Launay, J.M., & Richards, J .G. ( 1988). Platelets as a
model for neurones? W 115-126.

Eichelman, B. (1979). Role of biogenic amines in aggresive behavior. In M.

Sandler (ed). Embanhammlmoimm (pool-93)- New York, NY: Raven
Press.

98

E11is,.,DA Bingham,C.,R Zucker, RA. ,&Fitzgerald,H..E (1996).:[mlqgigal
O_i‘-fl,.w,l'-j" I l‘ I:I:II II:I2I1II1.IIIII III : ,III :II

Manuscript submitted for publication.

Feighner, J, Robins, E., Guze, S. ,Woodrufi‘, R, Winokur, G, &Munoz,R.

(1972). Diagnostic criteria for use in psychiatric research. Wm
26, 57-63.

Fitzgerald, H.E., Davies, W.H., Zucker, RA, & Klinger, MT (1993).
Developmental systems theory and substance abuse: A conceptual and methodolgogical
framework for analyzing patterns of variation in families. In L. L'Abate (Ed. ) Handhmkgf

. .z. I 0 O i (pp350-372). New York. erey.

 

Fitzgerald, H.E., Sullivan, L.A., Ham, H.P., Zucker, RA, Bruckel, S., &
Schneider, AM. (1993). Predictors of behavior problems in three-year-old sons of
alcoholics: Early evidence for the onset of risk. WW 110-123.

Fitzgerald, H.E., Zucker, RA, Maguin, E., & Reider, BE. (1994). Time spent
with child and parental agreement about preschool children's behavior. Wand
11.49th1611le 336-338.

Fuller, RW. (1992). Basic advances in serotonin pharmacology. Joumalgffllinical
Hashim; (suppl 10136-45.

Gitlin, M1 (1990). Ihensrchothemmstlaarfidelonsflhophamaeoloax New
York, New York: The Free Press.

Goodwin, D. W. ,Schulsinger, F. ,Herrnansen, L, Guze, S. B. ,& Winokur, G.

(1975). Alcoholism and the hyperactive child syndrome. WW
215359.269. 349-3 53

Greenberg, AS, & Coleman, M (197 6) Depressed 5-hydroxyindole levels

associated with hyperactive and aggressive behavior. MW;
33 1-3 36.

Guze, S. B. ,Goodwin, D. W., & Crane, J. B. ,.(1969) Criminality and psychiatric
disorders WM 583- 591

Halperin, J .M., Shanna, V., Siever, L.J., Schwartz, S.T., Matier, K., Womell, G.,
& Newcom, J .H. (1994). Serotonergic function in aggressive and nonaggressive boys with
attention deficit hyperactivity disorder. WW (2), 243-248.

99

Irwin, M., Schuckit, M ,& Smith, T. L. (1990). Clinical importance of age at onset

in type 1 and type 2 Primary alcoholics Archiresnffieneralllsrchiatmdl. 320-324
Jones, M C. (1968). Personality correlates and antecedents of drinking patterns in adult

males W242

Kantak, K., Hegstrand, L., & Eichelman, B. (1981). Facilitation of shock-induced
fighting following intraventricular 5, 7-dihydroxytryptamine and 6-hydroxy DOPA
thonhannamloul‘l. 157-160.

Katz, R (1980). Role of serotonergic mechanisms 1n animal models of predation.
. - . . = . . z . 219-231

 

Kent, T..A ,Campbell,J..,L Pazdernik,T..,E Hunter, R, Gunn,W..,H &Goodwin,
D. W, (1985). Blood platelet uptake of serotonin in men alcoholics. WM

1519611356. 357-359

Kiianamaa, J. (1976). Alcohol intake in the rat afier lowering brain 5-
hydroxytryptamine content by electrolytic rnidbrain raphe lesions, 5, 6-
dihydroxytryptamine, or p-clorophenylalanine. MW 203.

Kremer, H...PH ,Goekoop,J..,G &VanKempen,G...MJ (1990). Clinicaluseof

the determination of serotonin in whole blood. MW
19, 83 -87.

Kruesi, M. J. P, Swedo, S. E. ,Harnburger, S. D. ,Potter, W. Z. ,& Rapoport, J .L
(1988). Concentration gradient of CSF monoamine metabolites 1n children and

adolescents. WW5. 507-514

Kruesi, M.J.P., Rapoport, J .L., Hamburger, S., Hibbs, E., Potter, W.Z., Lenane,
M., & Brown, GL. (1990). Cerebrospinal fluid monoamine metabolites, aggression, and
irnpulsivity in disruptive behavior disorders of children and adolescents. Amhiyesgf
WM 419-426.

Kruesi, M.J.P., Hibbs, E.D., Zahn, T.P., Keysor, C.S., Hamburger, S.D., Bartko,
J .J ., Rapoport, J .L. (1992). A 2-year prospective follow-up study of children and
adolescents with disruptive behavior disorders: Prediction by cerebrospinal fluid 5-
hydroxyindoleacetic acid, homovanillic acid, and automnomic measures? Archivesgf

WW2. 429-435.

Kuperman, S. ,Beeghly, J. H, Burns, T. L. ,Tsai, L. Y. (1985). Serotonin
relationships of autistic probands and their first-degree relatives. Wm

WWW 186-190

100

Leventhal, B. L. ,Cook, EH. ,Morford, M ,Ravitz, A, Freedman, D. X. (1990).
Relationships of whole blood serotonin and plasma norepinephrine within families of

autistic childreln WWW 499-511

Levitan, I.B., & Kaczmarek, LR (1997). The neuron: Cell and molecular biology.
New York, New York; Oxford University Press.

Li T. K. ,Lumeng, L, McBride, W. J ,Murphy, J. M F roehlrch J. C., & Morzorati,
S. (1989). Pharmacology of alcohol preference 1n rodents. W
81165133111353. 73- 86

Lidberg, L. Mod1n,I,Ore1and, L. ,Tuck, J. R, & Gillner, A (1985). Platelet
MAO oxidase activity and psychopathy. WW 339-343.

Lidberg, L. ,Tuck, J. R, Asberg, M. ,Scalla-Tomba, G. P. ,& Bertilson, L. (1985).
Homicide, suicide, and CSF S-HIAA MW 230-236.

Linnoila, M. DeJong, J, & Virkunnen, M. (1989). Family history of alcoholism 1n
violent ofl‘enders and impulsive fire setters. NW 613-616.

Linnoila, M., Virkkunen, M., Sheinin, M., Nuutila, A, Rimon, R, & Goodwin,
FR (1983). Low CSF SHIAA concentration differentiates impulsive fiom non-impulsive
violent behavior. W 2609-2614.

Loukas, A. ,Piejak, L. A, Mun, E-Y, Bingharn, C. R, Fitzgerald, H E. ,& Zucker,
R.,A(1997 April) . . - - . ~
WW5 Poster session presented at the
annual meeting of the Society for the Research on Child Development, Washington, DC.

 

Marks, P.C., O’Brien, M. & Paxinos, G. (1977). 5, 7-DHT-induced muricide:
Inhibition as a result of exposure of rats to mice. BMW 383-388.

Martin, M.B., Owen, C.M., & Morihisa, J.M. (1987). An overview of
neurotransmitters and neuroreceptors. In RE. Hales, & S.C. Yudofsky (Eds), 1mm

33119125325111 (1313-55-85).

McBride, W. J. ,Murphy, J. M. ,Lumeng, L. ,& Li, T. K. (1989). Serotonin and
ethanol preference WWW 187-209

McCord, W., McCord, J., & Gudeman, J. (1960). W (pp. 23-
44). Stanford: Stanford University Press.

101

Moss, HE. (1987). Serotonergic activity and disinhibitory psychopathy in
alcoholism. WW2}, 353-361.

Murphy, BL, (1990). Peripheral indices of central serotonin firnction in humans.
AnnalsaneflorkAMennLotSciences. 282-295-

Murphy, J. M McBride, W. J. Lumeng, L. ,& Li, T. -K. (1982). Regional brain
levels of monoamines in alcohol-preferring and non-preferring lines of rats. W
213W 145- 149

Murphy, J.M., Waller, M.B., Gatto, G.J., McBride, W.J., Lumeng, L., & Li, TR
(1985). Monoamine uptake inhibitors attenuate ethanol intake in alcohol-preferring (P)
rats. AlcthLZ, 349-352.

Naranjo, C. A, Sellers, E. M., & Larwin, M O. (1986). Modulation of ETOH
intake by 5-HT Uptake inhibitor WW 16-22

Naranjo, C.A, Sellers, E.M., Roach, C.A., Woodley, D.V., Sanchez-Craig, M., &
Sykora, K. ( 1984). Zirneldine-induced variations in alcohol intake by nondepressed heavy
drinkers. Clinicallhannamlmandlhemticali. 374-381.

Naranjo, C.A, Sellers, E.M., Sullivan, J.T., Woodley, D.V., Kadlec, K., & Sykora,
K. (1987). The serotonin uptake inhibitor citalopram attenuates ethanol intake. Clinical
Phamamlosunflheraperficail. 266-274.

Neiman, J. ,Beving, H., & Malmgren, R, (1987). Platelet uptake of serotonin (5-
HT) during ethanol withdrawal 111 male alcoholics. WW 803-809.

Oreland, L. (1983). Why do alcoholics have low platelet MAO activity? In E.

Shaskan&R..W Meyer(Edsu)Broloai3l_andLoL29netic_f59tors_in913holism(Research
monograph 9, pp. 119-130). Washington, DC: U. S. Government Printing Office.

Oreland, L., Gottfiies, C.G., Kiianmaa, K., Wiberg, A, & melad, B. (1983). The
activity of monoamine oxidase-A and -B in brains from chronic alcoholics. [numamf
Nemllransmissim 73-83.

Oreland, L., von Knorring, L., von Knorring, A-L., & Bohman, M. (1985). Studies
on the connection between alcoholism and low platelet monoamine oxidase activity. In S.

Parvez, E. Burns, Y. Burrow, & H. Parvez (Eds). WW (pp.83-
117). Utrecht: VNU Science Press.

102

Patterson, G. R (1992). Developmental changes 1n antisocial behavior. In R D.

Peters, 1U. McMahon& V..L Quinsey(Eds.,) Aggressionandmnlenmhrmmhontme
lifespan, New York, NY: Sage.

Pine, D.S, Wasserman, G.A, Coplan, J ., Fried, J .A., Huang, Y-Y., Kassir, S,
Greenhill, L., Shafi’er, D. ,& Parsons, B. (1996). Platelet serotonin 2A (5-HT2A) receptor
characteristics and parenting factors for boys at risk for delinquency. A preliminary report.

Amerimlnumalnfifisrchiatnufil. 538- 544

Pletscher, A (1987). The S-Hydroxytryptamine system of blood platelets:
Physiology and pathorihysiology WW 177-188

Pliszka, S. R, Rogeness, G.,,A Renner, P. ,Sherman, J., & Broussard, T. (1988).

Plasma neurochemistry m juvenile offenders WW6
WW 588-594

Prior, M,Sma1t, D. ,Sanson, A, Pedlow, R, & Oberklaid, F. (1992). Transient
versus stable behavior problems m a normative sample: Infancy to school age. ImmaLQf

2.9613192511911619qu 423443

Rausch, J .L., Monteiro, M.G., & Schuckit, MA (1991). Platelet serotonin uptake
in men with family histories of alcoholism. WA, (2), 83-86.

Regier, DA, Farmer, M.E. ,Raw, D.S ,Loc,ke B..Z ,..Keith,SJ, Judd,L..,L &
Goodwin, F K. (1990). Comorbidity of mental disorders with alcohol and other drug
. , . . : ..; (19), 2511-2518.

 

Riddle, M.A, Anderson, G.M., McIntosh, S., Harcherik, D.F., Shaywitz, B.A, &
Cohen, DJ. (1986). Cerebrospinal fluid monoamine precursor and metabolite levels in
children treated for leukemia: Age and sex effects and individual variablity. Biologim

W21. 69-83-

Ritvo, E.R, Yuwiler, A, Geller, E. ,Plotkin, S. ,Mason, A. ,,Saegar K. (1971).
Maturational changes 1n blood serotonin levels and platelet counts. Bimhemical

Medicinal 90-96

Robins, L..N ,.(1966) - . , '
WW3. Baltimore Maryland Williams &

 

Robins, L., Helzer, J., Croughan, J. & Ratclifl’e, K. (1981). The NIMH Diagnostic
Interview Schedule: It's history, characteristics, and validity. AmhiymffienmL
12511911151388. 381-3 89.

 

103

Rockman, G.E., Amit, Z., Carr, G., Brown, Z.W., & Ogren, S.-O. (1979).
Attenuation of ethanol intake by 5-hydroxytryptamine uptake blockage in laboratory rats.
1. Involvement of brain 5-hydroxytryptamine 1n the mediation of the positive reinforcement
: . : z. . 245.

 

Rosenberg. CM- (1969) Young alcoholics. Bnfislrloumalnfifisrchramlli
181-188.

Roy, A, Adinofi‘, B., & Linnoila, M. (1988). Acting out hostility in normal
volunteers: Negative correlation with levels of SHIAA on C SF. MW
187-194.

Rydelius, P. A (1983). Alcohol abusing teenage boys: Testing a hypothesis on

alcohol abuse and personality factors using a personality inventory. W
83313111368. 381 -385

Rydin, E. ,Schalling, D. ,& Asberg, M. (1982). Rorschach ratings in depressed and
suicidal patients with low levels of 5-hydroxyindoleacetic acid 1n cerebrospinal fluid.

P.5119h111mLR953L9hJ. 229-243

Schalling, D & Asberg, M. (1985). Biological and psychological correlates of
impulsiveness and monotony avoidance. In J. Strelau, F. Farley, & A Gale (Eds), The

Biologicalflamifiersonahmmfiehamrmp 181- 194) WashingtonDC:
Hemisphere Publishing.

Schalling D. ,Asberg, M. ,Edman, G, & Levander, S. (1984). Irnpulsiviity,
nonconfomrity, and sensation seeking as related to biological markers for vulnerability

WW, (Suppl. 1), 746-747

Schalling, D. ,Edman, G., & Asberg, M (1983) Impulsive cognitive style and
inability to tolerate boredom: Psychobiological studies of temperamental vulnerability. In
M. Zuckerman(Ed), ' . : . .
(123-245). Hillsdale, NJ: Lawrence Erlbaum.

 

Schuckit, M. A, Klein, J. K. ,Twitchell, G., & Smith, T. (1994). Personality test

scores as predictors of alcoholism almost a decade later. Ammjcanlgnmalflmhm
(7,) 1038-1042.

Selzer, M. L. (1975). A self-administered short Michigan Alcoholism Screening
Test (SMAST) W86. 117-126

 

104

Sewell, RG. ,Gallus, J. A, Gault, F. O., & Cleary, J. P. (1982). P-
Chlorophenylalanine efl‘ects on shock-induced attack and pressing responses in rats.

WW1. 945-950

Sher. K-J-(1991).Chfldren_o£alcoholi3.A9663lappraisalnfth3n1_and
research, Chicago: University of Chicago Press.

Sher, K.J., Walitzer, K.S., Wood, R, & Brent, EB. (1991). Characteristics of
children of alcoholics: Putative risk factors, substance use and abuse, and

psychopathology. ImmalanhnonnalPsrcholoame, (4), 427-448.

Soubrie, P. (1986). Reconciling the role of central serotonin neurons in human and
animal behavior. WW2. 319-364

Stevens, G., & Featherman, D.L. (1981). A revised socioeconomic index of
oooumtional status. 831W 364-395.

Stofi’, D.M., Pollock, L., Vitiello, B., Behar, D., & Bridger, W.H. (1987).
Reduction of (3H)-imipramine binding sites on platelets of conduct-disordered children.
NmpmhonhannacnlmL 55-62.

Tarter, RE. (1982). Psychosocial history, minimal brain dysfunction, and

difi‘erential drinking patterns of male alcoholics. W88, 867-
873.

Tarter, RE. ,Hegedus, AM ,Goldstein, G., Shelly, C. ,& Alterman, A1. (1984).
Adolescent sons of alcoholics: Neuropsychological and personality characteristics.

WWW 216-222.

Tollefson, GD. (1989). Serotonin and alcohol: Interrelationships.
25119331316322. (Sappl- 1). 37-48.

Traskrnan, L. ,Asberg, M. ,Bertilsson, L., & Sjostrand, L. (1981). Monoamine
metabolites 1n CSF and suicidal behavior. W18, 631-636.

Valzelli, L. (1969). Aggressive behavior induced by isolation. In S. Garattini, &
E. B. Sigg (Eds. ), WW9: (pp. 70-76). New York. John Wiley & Sons Inc.

Valzelli, L. (1971). Further aspects of the exploratory behavior in aggressive mice.
MW 91-94.

Valzelli, L. (1981). Psychobiology of Aggression and violence. New Yorlg NY:
Raven Press.

105

Valzelli,. L. (1984). Reflections on experimental and human pathology of

 

van Praag, H. M. ( 1982). Depression, suicide, and the metabolites of serotonin in
the brain W935. 275-290

van Praag, HM. (1983). CSF 5-HIAA and suicide in non-depressed
schizophrenics. 11331.11. 977-978.

van Praag, H. M. (1986). Biological suicide research. Outcome and limitations.
WW 1305-1323

Virkkunen, M., DeJong, J ., Goodwin, F.K., & Linnoila, M. (1989). Relationship of
psychobiological variables to recidivism in violent ofl‘enders and impulsive fire setters. A
follow-Up study. WWW 600-603.

Virkkunen, M & Linnoila, M. (1990). Serotonin m early onset, male alcoholics
with violent behaviour. AnnalsnggdmnePZZ, 327-331.

Virkkunen, M., Nuutila, A, Goodwin, F ., & Linnoila, M. (1987). CSF monoamine
metabolites in male arsonists. W55. 241-247.

Vogt, M (1982). Some fimctional aspects of central serotonergic neurons. In N.

Osborne (Ed). WWW (pp 299-315) Chichester: John
Wiley.

Von Knorring, A-L. B,ohman, M., Von Knorring, L. ,& Oreland, L. (1985).
Platelet MAO activity as a biological marker 1n subgroups of alcoholism. W

83313111312. 51-58

Von Knoning, L., Oreland, L., & melad, B. (1984). Personality traits related to
monoamine oxidase (MAO) in platelets. W12, 11-26.

Von Knoning, L. ,von Knorring, A-L ,Smigan, L. ,Lindberg, U., & Edholm, M.

(1987). Personality traits in subtypes of alcoholics. W58,
523- 527.

Waller, M B. ,Murphy, J. M, McBride, W. J. Lumeng, L. ,& Li, T-K. (1985).
Studies on the reinforcing properties of ethanol 1n alcohol-preferring (P) rats. Almhnl;

WW2. 207

106

Waldbilllig, R]. (1979). The role of the dorsal and median raphe in the inhibiition
of muricide. W 341-346.

Weizrnan, A, Bernhout, E. ,Weitz, R, Tyano, S., & Rehavi, M. (1988)
Irniprarnine binding to platelets of children with attention deficit disorder with

hyperactivity 91919213939322.491496

West, M. O. & Prinz, R. J. (1987).Parenta1 alcoholism and childhood
psychopathology 25119331313331.1122. 204-218

Yen, C. Y ,Stangler, R. L. ,& Millman, N. (1959) Ataractrc suppression of
isolation-induced aggressive behavior. . -. : . : . .
179-185.

 

Zucker, RA (1987). The four alcoholisms: A developmental account of the

etiologic Process In P C Rivers (Ed ) WWW 131331.113
W35, 27-83. Lincoln, NE: University of Nebraska Press.

Zucker, RA (1991). Scaling the developmental momentum of alcoholic process
via the Lifetime Alcohol Problems Score. AlthLandflthfism (suppl, No. 1), 505-
510.

Zucker, R. A ,& Barron, F. H. (1973). Parental behaviors associated with problem
drinking and antisocial behavior among adolescent males In M. E. Chafetz (Ed. ),

101'1."'_|°0 1' . ,H,: u .010 ll 01‘ 11".1'51301-1‘11’01

C O O. C
D u
o .010 .0. ' 11‘ i .010,.H \'.‘.1. 1013.010 11 .15 0 0' H. -1' 0915..

mpulafigim (pp.276-296). Rockville, MD, National Institute on Alcohol Abuse and
Alcoholism.

Zucker, RA, Ellis, D.A., Fitzgerald, H.E., Bingharn, CR, & Sanford, KR
(1997). Other evidence for at least two alcoholisms, II: Life course variation in

antisociality and heterogeneity of alcoholic outcome. MW
8, 831-848.

Zucker, RA, & Fitzgerald, HE, (1991). Early developmental factors and risk for
alcohol problems. W312. 18-22.

Zucker, RA, & Fitzgerald, HE. (1991). Risk and coping in children of alcoholics:
Years 6 through 10 of the Michigan State University Longitudinal Study. Unpublished
grant application to the National Institue of Alcohol Abuse and Alcoholism (AA 070065).
Department of Psychology, Michigan State University, East Lansing, MI.

 

107

Zucker, RA, Fitzgerald, H.E., & Moses, RD. (1995). Emergence of alcohol
problems and the several alcoholisms: A developmental perspective on etiologic theory
and life course trajectory. In D. Cicchetti & D. J. Cohen (Eds), DexelepmenteL

mmmmmmmmmmmm (pp. 677-711) New York: Wiley.

Zucker, RA, & Gomberg, E. S. (1986). Etiology of alcoholism reconsidered. The
case for a biopsychosocial process. WW 783-793

Zucker RA & N011, RB, (1980). Ih9.AntismiaLB9ha1119LCh99klist
Unpublished instrument, Michigan State University, Department of PSychology, East
Lansing, MI.

Zucker, RA, Noll, R.B., & Fitzgerald, HE. (1986). Risk and coping in children

of alcoholics. WW, Department of Psychology, Michigan State
University, East Lansing, MI.

Zucker, RA, Noll, RB. ,Ham,H, Fitzgerald, H.E. &,Sullivan,L.S. (1992).
6‘."r‘..1'6111.1.*1-\ 3‘111‘1-111111’15 :'1-l.1 1".~ ‘1 \'.5-1 \ -11 1 1_\

Studies. Unpublished manuscript, Michigan State University, East Lansing, MI.

MICHIGAN STATE UNIV. LIBRARIES
llHIWWIIWI111111111111111“11111111111111
31293015654399