y
“5:.
Ya.

w
a

.rrsH
. .. :w , r .,
29):...1 .f .2“.
t. .93 3......
2.. 4:.
<

.6.

 

 

 

 

TEA-lit

THESIS

2004
51/9“ L??2 3 ?

LIBRARY
Michigan State
University

This is to certify that the

dissertation entitled

ATTACHMENT, SOCIAL SUPPORT COMPETENCIES, AND
VULNERABILITY TO AFFECTIVE SYMPTOMATOLOGY AMONG
ADOLESCENT OFFSPRING OF BIPOLAR DISORDER PARENTS

presented by

Tracy Lynne Smith Simko

has been accepted towards fulfillment
of the requirements for

Ph.D. Counseling Psychology

degree in

 

 

Major professor

Date 12/12/03

 

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDue.p65-p.15

ATTACHMENT, SOCIAL SUPPORT COMPETENCIES, AND VULNERABILITY TO
AFFECTIVE SYMPTOMATOLOGY AMONG ADOLESCENT OFFSPRING OF
BIPOLAR DISORDERED PARENTS
By

Tracy Lynne Smith Simko

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Counseling and Educational Psychology

2003

ABSTRACT
ATTACHMENT, SOCIAL SUPPORT COMPETENCIES, AND VULNERABILITY To
AFFECTIVE SYMPTOMATOLOGY AMONG ADOLESCENT OFFSPRING OF
BIPOLAR DISORDERED PARENTS
By

Tracy Lynne Smith Simko

This study proposed and examined a process by which quality of attachment to
parents influences vulnerability for affective symptomatology in offspring of parents with
bipolar disorder. A path model was tested wherein three social support competencies -
perceived social support, social self-efficacy, and fear of negative evaluation - mediate
this relationship. I pursued the idea that secure attachment to parents, in contrast to
insecure attachment, guides at-risk offspring to employ constructive assistance-seeking
strategies and to evaluate social support in ways that reduce their risk for affective
symptomatology. A review of the literature on the process in which secure attachment
promotes social support competence that guides the at-n'sk adolescent to successfully
access, recognize, and utilize support from important others during times of stress
supports this idea. Self-report measures of attachment bonds to father and mother,
perceived social support, social self-efficacy, fear of negative evaluation, and affective
symptomatology were completed by adolescents who have a parent diagnosed with
bipolar disorder. Results did not support the proposed mediational model; however, other
relationships among attachment bonds, social support competency measures, and
affective symptomatology were demonstrated. Implications for how findings advance
current understandings about protection from risk in this population are discussed.

Limitations of the study are noted and directions for fiIture research are addressed.

ACKNOWLEDGEMENTS

I would like to acknowledge and wholeheartedly thank those who were
instrumental in helping me complete this work. The idea and motivation for this project
emerged from my cherished experiences with families of the Rush Presbyterian St.
Luke’s Parent Support Group. I am indebted to them for all they taught me about
parenting with strength, honesty, humor, and compassion. Dr. Patricia Meaden provided
mentorship, enthusiasm, and a realistic perspective that were crucial early in the process.
My dear friend, Anne Mauricio, consistently inspired me with her strong work ethic,
positive attitude, support, and genuine, empathic understanding. Ellie Derr also deserves
thanks for the encouragement she gave me at an important point in the process. My
dissertation chair, Dr. Fred Lopez, is responsible for significantly enhancing the quality
of this work due to his theoretical expertise, thoroughness and continued commitment to
this project. To my family I owe the most. Owen and Connor, lights of my life, always
able to lift my spirit so simply; you fill me with a sense of pride that far exceeds any
other accomplishment in my life. Greg’s remarkably healthy approach to partnership and
parenting always made my working on this project easier on our family; we are blessed
by his selflessness, commitment, and affection. Lastly, thanks, Mom, for your lifelong
support, the pride that you convey, and the unconditional love that has enabled me to

pursue what I value.

iii

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. vii
LIST OF FIGURES ............................................................................... viii
INTRODUCTION .............................................................................. 1
Bipolar Disorder ........................................................................ 2
Parental Bipolar Disorder and Offspring Vulnerability ........................... 3
Genetics ..................................................................... 3
Parenting deficits associated with bipolar disorder ........................ 4
Impact of parental BPD on offspring ...................................... 5
Aim of the current study ........................................................... 7
LITERATURE REVIEW ....................................................................... 10
Attachment theory: A Framework for Examining Vulnerability to Affective
Symptomatology .................................................................. 10
The Theory .................................................................. 10
Internal Working Models .................................................... 13
Attachment Bonds, Parental BPD, and Resilience ....................... 16
Perceived Social Support ............................................................ 20
Fear of Negative Evaluation ......................................................... 23
Social Self Efficacy ................................................................... 24
Path Model Summary ................................................................ 26
SUMMARY OF THE LITERATURE ....................................................... 27
HYPOTHESES ................................................................................ 3O
METHODOLOGY .............................................................................. 32
Participants ............................................................................ 32
Parent participants ........................................................... 32
Adolescent participants ..................................................... 33
Procedure .............................................................................. 34
Measures .............................................................................. 37
Demographic Questionnaire ......................................................... 37
Adolescent Psychopathology Scale ....................................... 37
Inventory of Parent and Peer Attachment ................................. 38
Social Provisions Scale ...................................................... 4O
Adolescent Social Self Efficacy Scale ..................................... 41
Fear of Negative Evaluation Scale ......................................... 42
Design .................................................................................. 43
Analysis .............................................................................. 43

iv

RESULTS 45

oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

Scale Descriptives ................................................................... 45
Preliminary Analyses ............................................................... 47
Demographic variables .................................................... 47
Interrcorrelations Among Demographic Variables and Scales ................. 49
Results of Primary Analyses ....................................................... 51
Relations of adolescents’ attachment bonds with each parent .......... 51
Regression Analysis: Contributions of Parental Attachment Bonds to the
Prediction of Adolescent Affective Psychopathology ............................ 51
Regression Analyses: Contributions of Parental Attachment Bonds to the
Prediction of Social Competence .................................................. 54
Attachment Bonds and Perceived Social Support ....................... 55
Attachment Bonds and Social Self-Efficacy ............................... 55
Attachment Bonds and Fear of Negative Evaluation .................... 55

Regression Analysis: Contributions of Perceived Social Support, Social Self-
Efficacy, and Fear of Negative Evaluation to the Prediction of Adolescent

Affective Psychopathology .......................................................... 58
Post Hoc Analyses ...................................................................... 63
DISCUSSION ..................................................................................... 69
Overview .............................................................................. 69
Correlational Analyses ............................................................... 71
Group Comparisons of Secure Parental Attachment ............................... 72

Parental Attachment Bonds Predicting Adolescent Affective Psychopathology. .75
Social Support Competency Predicting Adolescent Affective Psychopathology. 78

Implications for Intervention and Prevention ......................................... 80
Limitations ................................................................................ 82
Recommendations for Future Research ............................................. 84
APPENDICES
A. Recruiting Notification .............................................................. 87
B. Reminder Card for Materials Not Received ................................... 88
C. Parent Demographic Questionnaire .............................................. 89
D. Adolescent Demographic Questionnaire ....................................... 91
E. Adolescent Psychopathology Scale ............................................. 93
F. Inventory of Parent and Peer Attachment ...................................... 100
G. Perceived Social Support Scale ................................................. 103
H. Social Self Efficacy Scale ...................................................... 105
1. Fear of Negative Evaluation Scale .............................................. 106

REFERENCES .............................................................................. 107

LIST OF TABLES

Table 1: N’s, Means, Standard Deviatons, and Ranges for Key Model Variables

Table 2: Results from ANOVA’S and Independent Sammes T Tests Testing

Significant Relationships Between Demographic Variables and APS Scores ......

Table 3: Intercorrelations Among Selected Demographic Variables and Key

Model Variables ..............................................................................

Table 4: Summary of Hierarchical Regression Analysis for Attachment

Bonds Predicting APS Scores ...............................................................

Table 5: Summary of Hierarchical Regression Analysis for Attachment

Bonds Predicting Social Competence Scores ...........................................

Table 6: Summary of Hierarchical Regression Analysis for Perceived

Social Support Predicting APS Scores ..................................................

Table 7: Summary of Hierarchical Regression Analysis for Social

Self-Efficacy Predicting APS Scores ...................................................

Table 8: Summary of Hierarchical Regression Analysis for Fear of
Negative Evaluation Predicting APS Scores ..........................................

Table 9: Means, Standard Deviations, and Intercorrelations for the
Components of Parental Attachment Bonds (Trust, Communication,

and Alienation) and Key Model Variables .............................................

Table 10: Summary of Hierarchical Regression Analysis Trust,
Communication, and Alienation Predicting APS Scores ...........................

46

48

53

S9

59

6O

67

Figure 1:

Figure 2:

Figure 3:

Figure 4:

LIST OF FIGURES

A model that describes the relationship between quality of
attachment to parents, PSS, SSE, FNE, and affective
symptomatology for adolescents of bipolar parents .....................

Interaction for Regression Analysis: Attachment Bond to
Diagnosed Parent & Attachment Bond to Non-Diagnosed Parent
as Predictor Variables and APS Scores as Criterion Variable .........

Path Diagram Depicting the Proposed Mediational Model
of the Relation Between Parental Attachment Bonds and

Adolescent Affective Psychopathology with Standardized Coefficients

Interaction Between Attachment Alienation Associated with
the Diagnosed Parent and Attachment Alienation Associated
with the Non-Diagnosed Parent in Predicting APS Scores ...............

viii

29

54

62

69

INTRODUCTION

It has been well established that being raised by a parent with an affective
disorder places a child at risk for a number of psychiatric and psychosocial problems (see
reviews by Beardslee, Versage, & Gladstone, 1998; Downey & Coyne, 1990; Goodman
& Gotlib, 1999). In the last two decades, heightened interest in the children of affectively
disordered parents has resulted in an extensive empirical literature base documenting
particularly higher rates of mood disorders and impaired familial and peer relationships
(Cytryn, McKnew, Bartko, Larnour, & Hamovit, 1982; Hammen, Burge, Burney, &
Adrian, 1990).

Within the large body of literature on risks for children raised by parents with
affective disorders there has been an overwhelming focus on children of depressed
parents. Relatively few studies examine children who have experienced both parental
depression and parental mania despite an alarming prevalence of manic-depression or
bipolar disorder among adults and adolescentsl. In a comprehensive study of mortality
and disability from diseases, injuries, and risk factors conducted in 1990 and projected to
2020, bipolar disorder was ranked as the sixth leading cause of disability across physical
and psychiatric disorders (Murray & Lopez, 1996). When bipolar parents and their
offspring are studied, the data are most often collapsed with data from children of
depressed parents with far fewer offspring of bipolar disordered parents included. While
the disorders often share some symptoms of depression, their etiology, course, and
treatment often differ. Nevertheless, a modest but provocative literature on the offspring

of bipolar disordered parents exists and indicates that the disorder represents unique

 

' The terms bipolar disorder and manic depression are used interchangeably in the literature.

challenges for offspring. The current study seeks to examine offspring characteristics and
coping processes that may render some offspring more or less vulnerable to affective
disturbance.
Bipolar Disorder

Bipolar I disorder (BPD), a mood disorder, is typically a lifelong, episodic
condition. The central feature of the disorder is that two kinds of mood episodes of
differing polarity —major depressive and manic episodes—reoccur and are interspersed
over time with periods of well-being. Bipolar depression is typically manifested by the
follOwing symptoms: sleep disturbance, lethargy, inertia and guilt associated with inertia,
decreased interest and motivation, a pessimistic view of self, others, and the future,
indecisiveness, diminished sense of competence and heightened sense of hopelessness,
crying, and social withdrawal. During a depressive episode, oversensitivity to rejection
or criticism is common. Persons with bipolar disorder often have difliculty distinguishing
between real rejection or criticism and their own cognitive distortions - sometimes able
to recognize intellectually that others’ behavior was not intended to be rejecting or
critical, but unable to inhibit or reduce their strong emotional reaction (e.g., sadness,
guilt, embarrassment, or anger) to that behavior. Social withdrawal is common during
depressive episodes. In contrast, the predominant mood of mania may be characterized
as elevated, expansive, and irritable, and angry outbursts can occur. Mania is expressed
through hyperactivity, flight of ideas, aggressiveness, and inflated self-esteem or
grandiosity that often results in a proliferation of new ideas and projects and excessive or
obsessive involvement in activities. While manic, individuals can flee unexpectedly (i.e.,

take spontaneous trips, disappear and reappear without explanation). Sleep disruption is

common and can cause a bipolar to feel exhausted but unable to slow down their thoughts
or physical energy and drive enough to fall asleep. The lack of sleep can initially lead to
more time for adventurous activity. However, unable to maintain the positive benefits of
little sleep, irritability and paranoia take the place of euphoria; behavior becomes
increasingly disorganized and thinking becomes confused. Problem-solving abilities are
substantially diminished during episodes of both depression and mania.

Parentaflipolar Disorder and Offspring Vulnegbjliy

Genetics. A genetic transmission hypothesis is implicit in most studies of
offspring vulnerability to mood disorders. In bipolar disorder, genetics establishes risk.
Trait markers of genetic liability are being actively researched but have yet to be
identified. So, while direct tests are not yet available, family studies, twin studies, and
adoption studies provided an indirect means of testing the genetic hypothesis.

For a person who has a parent with BPD, the risk of developing bipolar disorder
at some time in his or her life is approximately 12% and the risk of developing recurrent
major depressive disorder is also 12% (Rush, Cain, Raese, Stewart, Waller & Debus,
1991; Todd et al., 1996). Thus, there is an approximately 24% chance of a serious
recurrent mood disorder of some kind in each offspring of a parent with BPD compared
to approximately 1% in non-bipolar families (Lewinshon, Klein, & Seeley, 1995;
Weissman, Leaf, & Tischler, 1988). When other affective disorders (e.g., cyclothymia,
dysthymia) and subclinical levels of depression are considered, rates increase to as great
as 63% (Cytryn et al., 1982; Gershon et al., 1985 ; Grigoroiu-Serbanescu, Christodorescu,

Jipescu, Totoescu, Marinescu, & Ardelean, 1989; McKnew, Cytryn, Efron, Gershon, &

Bunney, 1979). Davenport, Adland, Gold, & Goodwin (1979) reported that 10 of 36
children in families with BPD in two generations or more developed bipolar disorder.

Although the biological transmission of affective illness across generations is
supported, Beardslee et al. (1998) noted that even strong proponents of the genetic
hypotheses maintain that the environment and its interaction with genetic predisposition
must not be overlooked. There is little question that parental depression negatively affects
the social-emotional environment within which children develop and that this, in turn,
increases a child’s risk for developing unipolar depression (see review by Goodman &
Gotlib, 1999). Regarding bipolar disorder, genetic vulnerability is undeniably salient in
its etiolog but does not account fully for individual differences in the expression and
maintenance of disturbance. For both unipolar depression and bipolar disorder, social and
environmental factors are thought to explain the portion of variance not accounted for by
genetic transmission and are widely considered to evoke or protect against genetic
vulnerability.

Parenting deficits associated with BPD. Anthony (1975) was the first to describe
impairments in parenting associated with BPD that may place the developing child at risk
for depression or other social-emotional problems. Anthony described problems such as:
a) taking but refirsing to give, b) having little awareness of others and consequently little
ability to empathize with them, and c) having an impoverished capacity to foster stable,
secure attachment behaviors and emotional bonds with the child because of the parent’s
own dependency needs. Davenport et al. (1979) was the first to describe features of
families with bipolar (manic-depressive) parents. These authors described several factors

that influence the maintenance of pathology through generations, including the following:

avoidance of affect by family members (emotions are threatening), absence of intimate
relationships apart from the family, displaced parental low self-esteem, and fears and
defensive behavior related to illness heritability. In a follow up study, Davenport et al.
(1984) reported that bipolar mothers were more overprotective but less attentive to their
children’ 5 health needs and reported more negative affect toward the child. The authors
reported that these mothers were disorganized, less active with their children, and more
unhappy, tense, and ineffective. Bipolar parents also reported lower scores than controls
in the areas of family interaction and social adjustment. Zahn-Waxler et al. (1984)
hypothesized that problems in parent-child bonding occur in families where BPD exists
because of a lack of parental consistency and stability.

Impact of pmenteflPD on offspring, The consequences for children exposed to
parental BPD were studied primarily during the 1970’s and 1980’s. Problems of affect
regulation and interpersonal relatedness dominate the small but thought-provoking
literature. Anthony (1975) argued that children in these conditions suffer inconsistency in
their emotional upbringing. He suggested that mood dysregulation may be a likely result
if the child becomes enmeshed in the ever-shifting moods of the BPD parent. He also
noted that problems of identification with the ill parent can arise, and there may be a
tendency for offspring to become “magic helpers” constantly offering assistance to
parents to make up for self-felt deficiencies.

Mayo, O’Connell, and O’Brien (1979) studied 22 offspring of BPD parents and
found that 10 developed separation-anxiety. Twelve nonsymptomatic high-risk children
were preoccupied with the health of their families. Cytryn et al. (1982) reviewed clinical

and experimental studies of offspring of parents with affective illness, including those

with parents with BPD, spanning infancy, childhood, and early adolescence. These
authors found a clear tendency toward early disturbances in the areas of affect regulation
and social interaction. Gaensbauer et al. (1984) studied social and affective development
in infants with a bipolar parent. These infants demonstrated disturbances in the quality of
their attachments and problems regulating their emotions adaptively. Zahn—Waxler et al.
(1984) also examined the social and emotional functioning of young offspring of manic-
depressive parents. They reported difficulties in maintaining friendly social interactions,
in sharing, and in helping playmates. The children studied had specific difficulty
modulating hostile impulses. The authors suggested that experiencing the ill parent’s
emotional extremes in sadness, excitability, and irritability makes such disregulation of
affect likely for offspring. Furthermore, the authors noted that the interpersonal and
emotional problems of the offspring were similar to those of their parents. These children
were followed and reported on four years later; patterns of maladjustment that had been
identified earlier were evident at the four-year follow up. The children had more
difficulty than controls with insight, empathy, establishing relationships, and regulating
affect. Radke-Yarrow et al. (1985) studied the children of 14 BPD mothers and found
that they had twice the level of insecure attachment than children of controls. The authors
noted consistencies between this study and others that evidence both genetic and
environmental influences on offspring development for children of parents with BPD.
They, too, attributed offspring difficulties to “contrasting extremes and alterations” in the
behavior of the bipolar parent.

Taken together, this literature suggests that experiencing the emotional

inconsistency and shifting availability of a parent with BPD increases a child’s risk of

certain adverse consequences. The children studied had trouble modulating their own
mood and had specific problems with insight into their own behavior, sharing, regulating
hostility, and demonstrating empathy. These deficiencies were linked to early difficulties
establishing positive parent and peer relationships.

Aim of the Current Study

The focus of the research described above was primarily on young children but
the findings have implications for social-emotional development beyond the early
childhood years. That is, fiom a developmental perspective, early close relationships
frame a child’s subsequent transactions with his or her social world. Childhood
difficulties establishing helpful, close bonds with principal caretakers have been linked to
adolescent problems with self-esteem and depression, as well as to the use of less
adaptive (problem-solving) coping strategies (e. g., seeking outside help in times of need)
during adolescence (Armsden & Greenberg, 1987; Greenberg, Seigal, & Leitch, 1983).
Such problems are thought to develop in large part because when one has early difficulty
establishing close bonds, he or she tends to form sub sequent relationships that are not
supportive and are easily disrupted (Carlson & Sroufe, 1995).

Unfortunately, empirical research investigating the adolescent functioning of
those young children involved in the initial offspring studies described above has not
been reported. Furthermore, attempts to understand the impact of parental BPD on
adolescent psychological functioning have not progressed beyond documenting rates of
offspring psychopathologyz. Although provocative, studies that simply correlate parental
BPD with adolescent disturbance leave some important questions unanswered. For

example, what enables some at-risk adolescents to fare better than others? Many children

who grow up with mentally ill parents, including those with parents who have BPD,
adapt successfirlly despite significant adversity. Furthermore, are there certain person-
characteristics that serve to buffer individuals from the negative consequences of parental
BPD? The answers to these questions are undoubtedly complex and multifaceted.
Nevertheless, research that directly addresses risk and protective factors for adolescent
offspring is needed so that interventions to reduce psychopathology can be developed.
Even in the extensive literature base on children of depressed mothers, explanations of
why some children are able to withstand significant parental emotional disturbance
without developing symptoms themselves remain understudied (Goodman & Gotlib,
1999). prrogress in treating and preventing adolescent mood disturbance related to
parental BPD is to be made, a move to more sophisticated, process-oriented explanations
of resilience is necessary. One important next step is to explore individual differences
that may reduce vulnerability. The current study represents such a step.

The present study used a process-oriented conceptual model - John Bowlby’s
attachment theory (1969, 1980, 1988) - to investigate a set of characteristics and
associated coping strategies that may render some adolescent offspring more resilient
than others. Attachment researchers have demonstrated that for adolescents, parental
attachment relationships explain a moderate amount of variability in outcome measures
of emotional adjustment (Armsden & Greenberg, 1987; Greenberg, Siegel, & Leitch,
1984). Despite a growing consensus that parental attachment plays a significant role in
reducing vulnerability to affective psychopathology, virtually no empirical data clarifying

the explanatory mechanisms underlying this relationship are available. To address this

 

2 Mood disorders are the most prevalent form of psychopathology for this population.

limitation, the current study investigates a path model for understanding offspring

resilience using a sample of adolescents who have experienced parental bipolar disorder.

LITERATURE REVIEW

Attachment Theory: A Framework for Examining Vulnerability to Affective
Smmcrmfalcgy

Attachment theory has been used as a framework for explaining adaptational risks
for children exposed to parental illness, in particular risks for affective disturbance
(Bowlby, 1969, 1973, 1980; Cummings & Cicchetti, 1990). In the past decade the theory
“has gained increasing theoretical and research attention for its heuristic value in
explaining how parental closeness can serve as a protective buffer and source of security
throughout adolescence” (Kenny, Moilanen, Lomax, & Brabeck, 1993, p. 409). Before
elaborating on links between attachment bonds and affective symptomatology, the
relationship between attachment and social-emotional development is reviewed.

The Theog

Attachment is generally defined as an important enduring emotional bond of
substantial intensity. Bowlby’s (1969) attachment theory emphasizes bonds that develop
between an infant and one or more caregivers as the first manifestation of the attachment
system — a behavioral system that serves to protect the child and promote survival.
Recently the theory has been used to guide research that describes emotional and
psychological disturbances that can result at any age from the actual or threatened
disruption of attachment bonds.

The essence of the theory is that human beings exhibit greater social and

emotional adjustment when they have confidence in the accessibility and responsiveness

10

of a trusted other or attachment figure’ . Two major dimensions of attachment have been
suggested (Parkes & Stevenson-Hinde, 1982): a) the cognitive-affective dimension,
defined as the underlying quality of affect toward attachment figures, and b) the
behavioral dimension, defined as the utilization of these figures for support in times of
stress and need.

The core assumption of attachment theory is that maintaining relatedness to others
is a universal and fundamental survival need. Rooted in infancy, the attachment system is
activated under conditions of “felt distress”; caregiver responses under these conditions
determine not only the infants’ physical safety but also his or her inner sense of security.
AS the child develops, organized patterns of behavior emerge to regulate proximity to
attachment figures and maintain attachment bonds. As cognitive capacities increase,
experiences with caregivers are organized into “internal working models” (IWMs) of the
self and the attachment figure. These models guide expectations of self and others in
relationships and contribute to one’s developing sense of relational competence.
Consequently, IWM’S influence one’s selection of coping strategies for dealing with
stress, family stress in particular".

Mary Main (1990) described the attachment system as “context sensitive”; that is,
different organizations and types of behavior are required to maintain proximity to
caregivers depending on the anticipated availability of the caregiver. Sensitive,
consistent patterns of caregiver responsiveness early in life tend to produce securely

attached infants. With a foundation of a “secure base” in others, securely attached

 

3 Attachment figures are most often parents, therefore the current study focuses primarily on parent-
adolescent attachment relationships and secondarily on peer-adolescent attachment relationships.

11

children move out into the world, pursue new experiences and relationships, and feel
confident they can return to find attachment figures supportive if they are needed
(Bowlby, 1988). With secure attachment as a base, children spend more and more time
exploring the environment and tolerating separation from caregivers with less distress
(Ainsworth, 1989). When they do feel threatened, securely attached children
communicate with attachment figures in ways that promote responsiveness (Cassidy,
1994). In turn, fears and stress are sufficiently alleviated so that exploration of the
environment continues. Repeated positive experiences of seeking and receiving
assistance from attachment figures when distressed instills confidence or “felt security”
in the accessibility of one or more trustworthy others who can be relied upon to help
regulate future distress. Over time, this confidence becomes internalized - securely
attached children incorporate lessons learned fiom the reliable, sensitive caregiving of
attachment figures and develop the capacity to effectively regulate their own emotions.
As they become increasingly skilled socially, children learn to balance interpersonal
dependence and independence. Having used their primary caregiver as a “secure base,”
securely attached infants demonstrate more enthusiasm, positive affect, and persistence in
problem-solving situations as toddlers (Matas, Arend, & Sroufe, 197 8) and respond more
flexibly, persistently, and resourcefully in preschool (Sroufe, 1983). During middle
childhood and adolescence, individuals with secure attachment histories show less
negative reactions to periods of high family stress (Pianta, Egeland, & Sroufe, 1990;

Sroufe, Levy, & Carlson,1998).

 

" Stress results from an appraisal that the demands of a situation exwed one’s resources. Here, it
specifically refers to threats to the survival of the attachment bond, and, ultimately, to the survival of the
individual.

12

In contrast, an insecure attachment orientation develops when caregivers are
unresponsive, inconsistent, interfering, or when they insufficiently or inaccurately read
and respond to the child’s signals. In these environments, the child is more likely to
develop a sense of self as unworthy of care and/or a model of important others as
uncaring or undependable. Emotional reactions to insensitive caregiving can be intense
(e. g., anger, rage, despair). Behavioral reactions range from detachment to
hypervigilence toward the signals of the attachment figure, depending upon the child’s
subjective experience of others. On one hand, when children experience attachment
figures as consistently inaccessible, motivation to continue help seeking diminishes.
Warm behaviors that promote emotional intimacy fade over time. As hopelessness and
emotional detachment increase the likelihood that the child will effectively communicate
future needs to helpful others diminishes. On the other hand, if children experience
caregivers as unpredictable, the child may become hypervigilant to the attachment
behaviors (or lack there of) of important others to capture whatever attention may be
available. Despite a strong desire for close relationships, these individuals are ambivalent
about the trustworthiness of important others. A tendency to worry excessively about
being evaluated and rejected makes it more difficult for the insecurely attached to
establish healthy, supportive relationships. Interestingly, the tendency to inhibit or
exaggerate emotions (Izard, Haynes, Chisholm, & Baak, 1991) may mirror the affective
system of the attachment figure.

Intml working models. Although attachment is initially behavioral (i.e., parent-
child interaction), models of self and other develop attachment experiences gradually

become internalized cognitively by the child. Though models are progressively

l3

constructed and are modifiable in response to significant changes in circumstances,
mental models grow increasingly firm over time as our ability to seek, shape, and
interpret experience converges with a tendency to view situations in ways that conform
with prior experience. IWMs may therefore be conceptualized as “interpersonal
schemas” (Safran, 1990) that guide how individuals interpret stressfirl interpersonal
events. As a young child experiences repeated relational success via coordinated
emotional communication with attachment figures (e.g., “If I show Mom that I am
frightened, she will help me feel less so”) working models of self and important others
gradually begin to serve as a lens through which one views the relational behavior of
important others. This lens focuses attention, filters information, and guides responses to
stress, often operating at an unconscious level (Bowlby, 1988). Accordingly, as the
young child develops, he or she becomes a more active force in the parent-child
relationship. He or she begins choosing which relationships to seek out and the manner
in which he or She participates in these, depending upon the degree to which they meet
emotional needs. Children with negative models of self and/or others spend valuable time
focused on their interaction with attachment figures to reduce distress. This distraction
impedes opportunities to build relational skills. Dysfunctional personal schema are,
consequently, more likely. For example, repeated unsatisfying experiences with
caretakers can lead to the self-view that one is not capable of establishing firlfilling
relationships (Bowlby, 1980). Despite longing for helpful, trusting interpersonal
connections, the insecurely attached child anticipates and becomes hypersensitive to
inconsistency and rejection. This makes him or her more likely to interpret subsequent

interpersonal stress as threatening and respond accordingly. Unfortunately, these children

14

not only expect rejection from others but actually are often rejected by them (Sroufe,
1983)

Conversely, securely attached infants are likely to be less aggressive, and more
cooperative, sympathetic, and competent in play when they are older (Ainsworth, 1983).
Securely attached school—age children are more competent in their relationships with
peers and adults, less fearfirl of strangers, and less prone to behavior problems (e. g.,
social withdrawal and anxiety) and dependent upon adults (Erickson, Sroufe, Egeland, &
Byron, 1985; Sroufe, 1983). During adolescence, while attachment behavior is often
directed toward non-parental figures, the ability to develop new supportive relationships
continues to be significantly influenced by beliefs about one’s relational competence
brought forward from early parent-child relationships (Weiss, 1982). Although
adolescence may be marked by periods during which parent accessibility is not necessary
for adolescent “felt security,” confidence in parental commitment remains crucial. During
this time the actual presence of the attachment figure may become less important than the
perceived adequacy of attachment relationships. In a study of late adolescence, internal
representations of the quality of early childhood attachments significantly related to
affect regulation and representations of self and others (Kobak & Sceery, 1988).

Securely attached adolescents are rated by peers as being more ego-resilient, less anxious,
and less hostile than less securely attached adolescents.

Hence, by adolescence, secure attachment can be conceptualized as a constructive
interpersonal schema characterized by: a) a positive, coherent view of self that allows the
adolescent to encapsulate stress and prevent it from spreading to the entire self-structure;

b) an optimistic attitude toward others and problems such that stressfirl interpersonal

15

events are appraised in benign terms, enabling more constructive coping strategies to be
employed; and, c) cognitive flexibility which permits one to revise schemas with new
information and adapt to environmental change. Such a positive cognitive schema makes
depression less likely according to cognitive models of depression (Beck, 1967).
Empirical research indeed demonstrates that secure attachment during adolescence
reduces risk for depression and other forms of psychopathology in general adolescent
populations (Armsden, McCauley, Greenbery, Burke, & Mitchell, 1990; Kobak, Sudler,
& Gamble, 1991) and, more specifically, for adolescents exposed to parental depression
(Downey & Coyne, 1990).

Attachment Bonds Parental BPD. and Resilience

 

Imagine the experience of a child whose bipolar-disordered parent fluctuates
between periods of over-involvement to times of marked withdrawal from them. During
a manic episode, the parent may become significantly more dramatic and grandiose about
his or her self-presentation, obsessive about normally inconsequential matters, and
significantly more intrusive and controlling of others in his or her immediate social space.
Mania may then give way to depression, wherein the parent becomes self-absorbed,
disinterested in the activity or behavior of the child, and overwhelmed by feelings of
insecurity and beliefs that they are incapable of helping the child. Under these
circumstances, how the child construes the situation and copes with this relationship
instability should significantly impact his or her own vulnerability to affective
disturbance Attachment security acts as an inner resource to facilitate successful coping.
A child with a secure attachment history is more likely to deal with distress by

acknowledging it and then will appraise stress, and the efforts required to reduce it, as

16

manageable. He or she is more likely to enact active, constructive coping strategies, such
as turning to others for support. Because the securely attached child carries an
underlying sense of trust in self and others he or she is less likely to engage in self-blame
or become hostile towards attachment figures. He or she is also more likely to view the
self as resourcefirl and others as helpful. In addition, individuals with secure attachments
have a higher tolerance for unpredictability, disorder, and ambiguity. This would be
particularly beneficial for children exposed to parental BPD — cognitive flexibility should
render one less distressed by unpredictable emotional episodes than individuals with
more rigid cognitive schemata. For instance, for those who experience dramatic parental
mood shifts, rules for coping behavior that hold in one affective state don’t necessarily
hold in the other (Pound, 1991). In this way, a secure attachment history would enable
some offspring to endure parental mood swings and to be flexible in responding to
changing circumstances. In a study of psychiatrically hospitalized adolescents, including
those with affective disorders, Rosenstein and Horowitz (1996) indeed found that
attachment organization was “overwhelmingly insecure.” These authors suggested that
adolescent symptomatology was related to a tendency to focus on the ill parent, inhibited
exploratory behavior, and the subsequent failure to develop effective affect regulation
strategies.

In contrast, the insecurely-attached child, more sensitive to both rejection and
inconsistency and less tolerant of negative affect, is more likely to feel threatened by such
parental behavior. More specifically, attachment theory holds that the child who has been
unable to establish a sense of attachment security would react to such relational

disappointment in one of two general ways -- by either becoming preoccupied with his or

17

her own feelings or by fiercely turning away from them. In the first case, he or she might
ruminate on the negative thoughts and feelings generated by the parent’s behavior and
focus on memories of similar experiences. Plagued by a sense of helplessness and
alienation, distress may be overwhelming and may spread to other areas of life (e. g.,
school performance, peer relationships). Intense emotional distress may continue even
after the actual threat has terminated (e.g., during a period in which the BPD parent is
functioning well). This child is more likely to cope by intensifying efforts to get the BPD
parent to comfort him or her during periods of the illness when the parent is incapable of
doing so rather than accessing more usefirl support-providers (e.g., the undiagnosed
parent). In the second case, the insecurely attached child might exhibit a more defensive
style, emphasizing the untrustworthiness of significant others. This child would likely
prioritize maintaining distance from potentially helpfirl others and would minimize overt
expressions of distress; both of which make it less likely that needed support is obtained.
Furthermore, such defensiveness is typically accompanied by an underlying sense of
anxiety. In the long run, this makes it more difficult for the defensive (insecurely
attached) child to mitigate pain associated with parental affective episodes.

However, the negative effect of such variable parental emotionality would
conceivably be less if other attachment figures are available or when the child’s
environment otherwise provides a sense of stability. Offspring are more likely to develop
a sense of trust in others and, in turn, a sense of self as resourceful if structures and
resources in the home are efl’ective in: a) alleviating stress for family members and the ill
parent; and, b) for providing stability for the children. Given such circumstances, the risk

for developing affective psychopathology should be less than when the environment is

18

compromised by parental psychopathology. Secure attachment enables one to
realistically evaluate attachment-related experiences and develop relational competence
so that during periods of parental unavailability, alternative sources of support are sought
and utilized to alleviate distress. In contrast, when instability is prominent and other
attachment figures are unavailable, offspring vulnerability to psychopathology should be
highest because there are fewer opportunities to develop relational competence (e. g.,
accurately reading behaviors, particularly emotional, of others) that might buffer them
from firture stress. Insecure attachment strategies may seem adaptive when sensitive
caregiving is compromised; however, the long-terrn consequences (e. g., inability to
develop close, positive relationships in the future) are a diminished sense of personal
resourcefirlness/control that would likely exacerbate risk for affective psychopathology.
A specific example demonstrates how secure attachment, in contrast to insecure
attachment, might protect a child from the negative effects of parental BPD. Envision a
manic parent over-reacting to an expression of sadness from a child who misses a friend
that has recemly moved to another town. The manic parent might dramatize the sadness,
ridicule the child to others, propose unreasonable solutions for seeing the missed child,
and/or become irritable, dismissing the child’s feelings. Alternatively, during an episode
of depression the BPD parent might respond by intense crying, verbally obsess about
additional losses in their own or the child’s life, or express feelings of helplessness and
withdraw. Ifthe child has been able to achieve a sense of security within their attachment
relationships, he or she is equipped with a sense of social competence and corresponding

tendency to effectively access, recognize, and utilize social support; this renders him or

19

her less likely to be disappointed by the parent’s reaction and less likely to internalize the
response as rejection.

Although the adolescent children of BPD parents have not been empirically
studied, young offspring have been. These studies (described earlier in this review)
demonstrate that secure attachment indeed buffers young offspring from developing
affective symptomatology. These studies, and those in the more general adolescent
literature suggest that adolescent offspring of BPD parents are at greater risk for affective
disturbance if they have a history of insecure attachment relationships. However, the
correlational nature of the existing data do little to explain the relationship between
attachment and affective disturbance for at-risk children. Empirical identification of the
mechanisms linking attachment and affective symptomatology in this population would
strengthen the theoretical model of attachment and depression, would increase
understandings of parental BPD and offspring vulnerability to affective symptoms,
(Cummings & Cicchetti, 1990) and may extend to other models linking parental behavior
and offspring resilience.

For the current investigation, a comprehensive literature review revealed three
potential mediators of the relationship between attachment and affective
symptomatology: l) perceived social support, 2) social self-efficacy, and 3) fear of
negative evaluation. Additionally, there is support for examining the moderating role that
the non-diagnosed parent may play in reducing vulnerability.

Perceived Social Support
Perceived social support (PSS) refers to a generalized appraisal that individuals

develop in which they believe that they are cared for and valued, that significant others

20

are available to them in times of need, and that they are satisfied with the relationships
they have (Procidano & Heller, 1983). Perceptions are initially developed through early
close relationship experiences (Sarason, Pierce, & Sarason, 1990); parental attitudes
towards a child and the behavior associated with them gradually become incorporated by
the child into generalized expectations about the supportiveness of others and one’s own
support worthiness. As such, PSS tends to be relatively consistent across time even when
individual social circumstances change (Sarason et al., 1986; Sarason, Sarason, &
Shearin, 1986). Children whose parental relationships have been supportive and well
matched to the child’s needs develop optimistic expectations about the willingness of
others to provide support. In contrast, children whose needs for support are not met in
such a way develop a general view of others as unwilling or unable of meeting their
needs for support (Coble, Gantt, & Mallinckrodt, 1996). In this way parental support
influences cognitive schemata regarding the self (e.g., I am worthy of support and valued
by others) and the availability of supportive relationships (e.g., that others wish to be of
help). Persons are more likely to develop confidence in their own abilities to manage
relational distress if they believe that social support exists. In an empirical study of the
relationship between adolescent PSS and models of self and other, Blain, Thompson, and
Whiffen (1993) demonstrated that secure attachment was associated with high levels of
perceived social support from family and fiiends.

PSS is also associated with vulnerability to affective disturbance. Research
demonstrates that an inverse relationship exists between PSS and depression and between
PSS and manic-depression; higher levels of support correlate with lower levels of

symptomatology (O’Connell, Mayo, Eng, Jones, Gabel, 1985; Stefos, Bauwens, Staner,

21

Pardoen, & Mendlewics, 1996; Johnson, Winett, Beyer, Greenhouse, & Miller, 1999).
Moreover, while the link between PSS and vulnerability to affective symptomatology is
not clearly understood, one explanation may be that perceptions of social support
influence physiological processes (e.g., cardiovascular functioning, immune functioning)
(Uchino, Cacioppo, & Kiecolt-Glaser, 1996); this is particularly intriguing given the
biological foundation of bipolar disorder.

For children of affectively ill parents, Beardslee and Podorefsky (1988)
maintained that interpersonal relationships can protect offspring from developing their
own symptoms. A deep commitment to relationships in particular was linked to
resiliency. One study specifically exarrrined the role of social support in outcomes for
children raised by a parent with BPD. Pellegrini et al. (1986) compared the personal and
social resources in children of patients with BPD to those of controls. They found that
psychiatric well-being is related to the quality of personal and social resources one has;
nondisordered probands demonstrated a strikingly superior profile of personal resources
in comparison to disordered probands. Caution is warranted in generalizing from one
study, and the retrospective nature of this research prohibits causal interpretations.
Nevertheless, the authors’ arguments for additional research investigating links between
PSS and vulnerability to affective disorders for at-risk adolescents are compelling.

Several studies indicate that a positive relationship with the non-ill parent in
particular may buffer the offspring from the negative impact of the ill-parent’s affective
disturbance. Rutter (1987) specifically emphasized the protective role of the non-
disturbed parent in providing a source of support and consistency for offspring of

afl'ectively ill parents. This is especially relevant for the current study, as parental BPD is

22

characterized by mood swings that are typically inconsistent and unpredictable. Similarly,
Fisher, Kokes, Cole, Perkins, and Wynne (1987) reviewed the literature on competent
children at risk and found that perceptions of support from the non-ill parent
characterized well-functioning children of disturbed parents.

To summarize the link between attachment and PSS, confidence in the availability
of others during times of need, coupled with the ability to accurately read the signals of
others should lead to higher levels of PSS for the securely attached. Children with
insecure attachment histories usually have lost confidence that important others are
accessible and willing to offer support in times of need (Bowlby, 1973). Furthermore,
when those with insecure attachment histories do feel threatened and experience
associated negative emotions, they communicate this in ways that are less likely to
promote responsiveness fi‘om others (Cassidy & Kobak, 1988; Kobak & Sceery, 1988).
They also tend to direct heightened distress-related emotions (e.g., desperation, anger,
hostility) toward attachment figures, thereby further diminishing the likelihood that
helpful others will be available when needed. This cycle should increase vulnerability to
affective disturbance as it inhibits environmental exploration and reduces self-confidence
in building new potentially supportive relationships (Ainsworth, Blehar, Waters, & Wall,
1978)

Mf Negative Evm

Included in the notion that working models guide beliefs and expectations about
social relationships, one’s belief about how they are perceived by others is correlated
with the model of self (Sarason, Peirce, Shearin, Sarason, Waltz, & Poppe, 1990) and

theoretically linked with parental attachment and risk for affective disturbance. That is,

23

securely attached individuals value close relationships and hold optimistic beliefs about
how others perceive them. They are less Sensitive to rejection than those who are
insecurely attached and more likely to have emotional needs met through stable, trusting
relationships. Because their own views of self and appraisals of how other’s view them
are more positive, objective, and flexible, persons with secure attachment histories are
better equipped to manage relational stress and environmental change (Rosenstein &
Horowitz, 1996). Thus, the higher an adolescents’ sense of secure parental attachment,
the less preoccupied or debilitated he or she should be by the relational stress and
environmental change that so often characterizes families with a psychologically-
distressed parent.

The often unpredictable parental mood swings and erratic behavior of a bipolar
parent make it more likely that offspring will be subjected to negativity and family stress
within the home (Hodgins, Faucher, Zarac, & Ellenbogen, 2002) and shame and stigma
outside the home (Lefley, 1989; Oestman & Kjellin, 2002). As high anxiety about others’
negative evaluations increases risk for depression (O’Connor, Berry, Weiss, and Gilbert,
2002) the adolescent who has been able to develop a strong sense of parental attachment
and thus a stronger sense of self will have less apprehension about others’ evaluations.
This, in turn, should serve to reduce his or her own vulnerability to affective
psychopathology.

Social Self-Efficacy (SSE)

One’s belief in his or her ability to effect social interaction, or social self-efiicacy

(SSE), is also linked theoretically and empirically to social competence, attachment and

risk for affective disturbance. It has been suggested that deficits (real or imagined) in the

24

ability to affect others’ emotional states contributes to a diminished sense of social
competence, namely social anxiety (V ertue, 2003). However, attachment theorists assert
that a secure relationship with one or both parents throughout childhood contributes to
the development of a positive schemata about one’s own social-effectiveness (SSE)
(Mallinckrodt, 1992). Coordinated communication and a history of relational successes
are incorporated by the child into beliefs about his or her sense of socially competency.
The more optimistic one is about his or her ability to impact the outcome of stressful
events, the less vulnerable one is to using dysfunctional cognitive processes and, in turn,
to developing depression (Beck, 1967). Similarly, Dozier, Stoval and Albus (1999) noted
that depression may be linked to a “sense of uncontrollability [for stressful events] on the
part of the child (p. 500).” This is particularly relevant for offspring of BPD parents,
where intermittent and largely unpredictable mood swings characterize the environment.
Among early to middle adolescents, parent attachment was found to be associated with
lesser hopelessness and a less extemally-oriented locus of control and with greater self-
management (coping) skills (Armsden et al., 1987). Cognitive models of depression
would benefit from research that elucidates the environmental and familial precursors of
the dysfirnctional cognitive processes described in current models of depression (e. g.,
Beck) (Coble, Gantt, & Mallinckrodt, 1996). In one empirical study, Ehrenberg and Cox
(1991) found an inverse relationship between social self-efficacy beliefs and depression;
however, studies that investigate the underlying mechanism between SSE and
psychopathology are rare.

Taken together, PSS, SSE, and FNE are conceptualized in the current study as

social support competencies that guide individuals to organize his or her social world in

25

such a way that risk for affective symptomatology is either increased or reduced.
“Competency” refers to both a set of interpersonal relationship skills and experiences and
the beliefs [about the relational self and others] that prompt an individual to employ those
skills to recruit social support.
Path Model Summary

As described in this review, the attachment system is organized around
maintaining proximity to caretakers and, later, to those emotionally close to us. As we
develop, affectional bonds that have been established with caregivers widen and change
as we enter into other relationships. By adolescence, secure attachment is typified by
emotional closeness with several others and trust in their availability and responsiveness.
Securely attached adolescents are able to display positive emotions that enhance social
interaction and social competence. They are also able to tolerate negative affect while
remaining engaged with others (Sroufe, Schork, Frosso, Keroski, & LaFrenier, 1985). It
is the argument of the current study that social support competence (high levels of PSS
and SSE and low levels of FNE) leads one to engage socially in a manner that should
render some at-risk adolescents more resilient to the negative impact of their parent’s
bipolar illness than those who are less socially competent. PSS, SSE, and FNE are
conceptualized here as related but unique elements of social support competence. Each
competency is included in the path model under investigation because of its potential to
explain variation in the relationship between parental attachment bonds and offspring
affective symptomatology. In sum, one’ s motivation to seek healthy relationships and
success in doing so is influenced by these aspects of social support competence that

develop from repeated satisfying relationship experiences. The more an individual

26

perceives that support is available when needed and the more he or she believes they can
effect interpersonal interaction, the more likely they are to successfully access and utilize
support from others when distressed (e.g., during periods of extreme parental mood
fluctuation). Similarly, the more optimistic persons are about their own social status, the
less vulnerable they should be to developing anxiety or depression when faced with
relational stress.

Summary of the Literature

For offspring of parents with bipolar disorder, the risk for developing affective
disturbance is high. Although genetics establishes risk, psychosocial factors impact
whether that risk is heightened or attenuated. However, research that addresses just how
psychosocial factors influence risk is rare. The development and evaluation of
mediational models that demonstrate how these variables exert their influence are needed
so that more effective preventative and therapeutic interventions can be developed.
Hypothetical links among the variables should be tested, with the most promising models
subjected to further experimentation and analysis.

In the current study, theoretical and empirical evidence is presented that supports
an investigation of the relationship between quality of parental attachment bonds, aspects
of social support competence, and offspring vulnerability for affective symptomatology.
This support comes fi'om three main sources: the broad developmental psychopathology
literature, an extensive literature on offspring of depressed parents, and the small but
persuasive literature on children of BPD parents. Across these bodies of literature, secure
attachment, in contrast to insecure attachment, is associated with better outcomes for

offspring of parents with affective disturbance. Extant theoretical and empirical research

27

suggests that parental attachment quality affects the development and deployment of
various social support competencies that may function as the means by which affective
symptoms are either inhibited or amplified. Rice, Cunningham, and Young (1997) found
that the association of parental attachment bonds to emotional adjustment was indeed
mediated by social competence in a large sample (N= 630) of Black and White
adolescents. Under the stressful conditions associated with having a BPD parent,
attachment organization presumably guides the adolescent’s expectations of availability
of social support (perceived social support) and of their own resourcefulness in accessing
and utilizing such support (social self—efficacy). Additionally, secure attachment
decreases pessimism and anxiety about how others view the self (fear of negative
evaluation). Therefore, in cases where a parent with BPD has been able to call upon
supportive others or supportive others have otherwise been available to meet his or her
needs for safety and comfort during affective episodes, the child is more likely to develop
a sense of others as trustworthy and a sense of self as worthy of care. Trust that
attachment figures understand and respect one’s needs and that they can be relied upon
for help should lead the individual to feel more confident in their own relational
capability. This, in turn, renders the individual more likely to seek others under future
conditions of stress and to accurately read signs of support from others. In this way,
attachment orientation guides the adolescent to actively turn to or avoid others during
stressful times. In contrast, adolescents with insecure attachments fear others’ evaluations
and are less likely to effectively seek or establish support from needed others during
times of stress. These individuals are more likely to exhibit anger toward or emotional

detachment from attachment figures when attachment bonds are threatened. This, in

28

combination with their more negativistic attitudes and compromised ability to recognize
support when it is available renders them more susceptible to affective symptomatology.
Figure 1 demonstrates the relationship between quality of attachment to parents,

social support competence, and affective symptomatology.

Figure l. A model that describes the relationship between quality of
attachment toJarents, PSS, SSE, FNE, and affective symptomatology
for adolescents of bipolar parents.

Attachment
Bond to
BPD $
Parent
V

Affective
/ \ —> Sympto-
matology

 

Trust

  
   
 

  

Communi-
cation

 

 

 

 

 

Alienation

 

Trust

\ ) Attachment /
t 1 Bond to
Communi- L—y
cation Non-BPD @ T
D

 

 

 

 

Parent

: </'l___

Alienation

 

 

 

 

 

 

r r l

29

HYPOTHESES
fldolescecnt AtteLhmMrarlity differences between attachment to BPD parent and
attaphment to non-BPD parent.

1) In comparison to the relationship between the adolescent and BPD parent, the
relationship between the adolescent and the non-diagnosed parent should reflect a
higher degree of secure parent-adolescent attachment.(i.e., higher scores on trust
and communication and lower scores on alienation).

II. Relations of parental attachment quality to affective aymptomatpjgy

1) Indexes of secure parent-adolescent attachment (i.e., higher scores on trust and
communication and lower scores on alienation for each parent) will be
significantly related to lower levels of affective symptomatology.

2) Indexes of secure attachment between the adolescent and the non-diagnosed
parent will moderate the relationship between insecure attachment with the
diagnosed parent and affective symptomatology.

III. Relations of paaentaLaLt_achment quality to social support competencies

3) Indexes of secure parent-adolescent attachment (i.e., higher scores on trust and
communication and lower scores on alienation for each parent) will be
significantly related to social support competence (higher levels of perceived
social support and social self-efficacy, and lower levels of fear of negative

evaluation).

30

IV. Relations of social support competence to affective symptomataltgy
4) Perceived social support, social self-efficacy, and fear of negative evaluation
(reverse scored) will each make significant and unique contributions to the
prediction of affective symptomatology.
V. Mediatipnal Model
5) Social support competencies will significantly mediate the relation between

attachment to parents and affective symptomatology.

31

METHODOLOGY
Participants

Participants were 96 adolescents and 86 parent respondents. A discrepancy
between the number of adolescent and parent respondents exists because siblings were
permitted in the study, therefore some parent respondents provided information for
multiple children. Sixty-three percent of the parent respondents were mothers of the
enrolled adolescents and 37% were fathers of these adolescents. All respondents were
recruited from the Depressive and Manic-Depressive Association, from the National
Alliance of the Mentally 111, or from participating psychiatrists’ offices.

Twelve percent of represented families reported an annual household income of
less than $20,000, 32% reported an annual household income of $20,000-$40,000.
Seventeen percent reported an annual household income of 840,000-60,000. Twenty
percent report an annual household income of 860,000-880,000. Eight percent reported
an annual household income of $80,000-$100,000. Ten percent reported an annual
household income of over $100,000.

Parent participants

The age of the diagnosed parent ranged from 34 to 56 years old, with a mean of
45.27 (SD = 5.75). Seven percent of the sample reported that the diagnosed parent had
attended or completed high school, 62% had attended or received a degree from a college
or university, and 31% had pursued graduate studies. Regarding diagnostic/treatment
history, age of onset ranged fi'om 2-45 years old, with a mean of 22.78 years (SD = 8.97).
Age of diagnosis ranged from 18-53 years old, with a mean of 33.62 years (SD = 10.95).

The approximate number of affective episodes in the lifetime of the diagnosed parent

32

ranged fi’om 1-100 with a mean of 19.03 (SD = 22.17). F iffy-three percent reported that
the diagnosed parent had not been hospitalized because of their bipolar disorder. Thirty-
one percent reported they had been hospitalized for their disorder between 1-3 times, and
6% had been hospitalized four or more times. Ninety-seven percent of the parent
participants reported that the diagnosed parent had taken medication for bipolar disorder.
Ninety-six percent have taken an antidepressant, 28% have taken antianxiety medication,
and 21% have taken antipsychotic medications. Ninety percent of the parent respondents
reported that the adolescent has not taken medication for an affective disorder. Of those
adolescents who have taken medication, 8% reported that they had taken an
antidepressant, 3% reported that an antianxiety medication was used, and 1% reported
that antipsychotic medications had been taken. Parents reporting on the adolescents
provided the following information about the diagnoses and treatment of their
adolescents. Only 7% reported that their adolescent had beendiagnosed with an affective
illness. Of these, the age of onset for the adolescent ranged from 5-17 with median ages
of 10 and 11 reported. Three adolescents had been diagnosed at age 10, three at age 12,
and one at age 13. The approximate number of affective episodes in the adolescent’s
lifetime ranged from two to ten - one reported two episodes, one reported four episodes,
two reported five episodes, and two reported ten. Three adolescents have been
hospitalized from 1-3 times, three have been hospitalized four or more times.

Adolescent participants

Adolescent participants were 49% males and 51% females, ranging in age from
12 to 17 years old, with a mean age of 14.62 (SD = 1.37). Eighty-six percent of the

adolescents in this sample were Caucasian, 3% were Hispanic, 2% were American

33

Indian, and 1% were African American. Six percent selected “other” as their ethnicity.
Fifty percent reported that they had received individual counseling services. Nineteen
percent reported they had received between one and five sessions, 11% reported they had
received between 6 and 15 sessions, 5% reported they had received between 16 and 30
sessions, and 11% reported they had received 30 or more sessions. Fourteen percent
reported that they had received group counseling services. Thirteen reported they had
received between one and five sessions, one reported they had received between 16 and
30. Forty percent reported that they had received family counseling services. Twenty-
five reported they had received between one and five sessions, 7 reported they had
received between 6 and 30 sessions, 1 reported receiving between 16 and 30 sessions, and
4 reported they had received 30 or more sessions. Only 2 adolescents enrolled in the
study have been hospitalized for psychiatric or psychological services. Sixteen percent
reported that they have taken psychotropic medication. Six have taken antidepressant
medication; of these, 4 had also taken anti-anxiety medication. Six percent of respondents
reported that they had been arrested for a misdemeanor crime. No respondents reported
that they had been arrested for a felony crime. Ten percent of the sample reported that
they have used illicit drugs.

from;

Three methods of recruiting subjects were used. For the first method, the
principal investigator visited local chapters of the DMDA and NAMI associations to
inform members of the current research study. Association members were told that the
research involved a study of resiliency in adolescent offspring of BPD parents and were

given information regarding requirements for the study (e.g., subject requirements, time

34

requn’ements, risks and benefits, etc). Persons who indicated interest were asked if they
could be contacted via telephone by the principal investigator for enrollment in the study.
For the second method of recruitment, advertisements with the contact information of the
principal investigator were placed on DMDA and NAMI websites and in association
newsletters (see Appendix A). For the third method of recruitment, notices of the study
and contact information of the investigator (see Appendix A) were distributed to Rush
Medical Center psychiatrists in private practice who were informed of the study and who
agreed to recruit potential subjects. The response cards left in the psychiatrist’s offices
were colored coded. Each office was assigned one color. When the interested individual
contacted the investigator, she asked the individual which color card they obtained so that
the investigator could more accurately describe the subject pool (e.g., all subjects came
fi'om one particular office) while still maintaining some degree of privacy for the
subjects. Ifboth parent and child verbally agreed to participate in the study, consent
forms were mailed to the parent and adolescent. All participants were required to sign an
informed consent before data gathering materials were distributed. The parent also signed
a consent permitting the adolescent to be interviewed. The adolescent also signed an
assent agreeing to participate. When the principal investigator received the completed
consent forms, subjects were given code numbers and then a packet of questionnaires
with code numbers on each was mailed to the adolescent. Participants were given contact
information for the principal investigator and an opportunity to ask questions about the
study with answers provided when possible. Packets distributed to adolescents contained
instructions and all measures. To maintain confidentiality, instructions were written on

each packet that the adolescent information was to remain confidential -- the adolescent

35

was to complete the information alone. A sticker was provided in the packet.
Adolescents were instructed to place and seal the sticker on the envelope when materials
were completed and to Sign the sticker so that confidentiality was assured.

Parent and child were assigned subject numbers. Adolescents were assigned
consecutive subject numbers that began with 100. The parent was assigned a subject
number that began with 200, but that corresponded to the adolescent’s number. For
example, the first adolescent was assigned #100, his/her parent was assigned #200. The
next adolescent was assigned #101, and his/her parent #201. Parental demographic
information, and adolescent data was gathered together as one set and placed in an
envelope together. Parental and adolescent consent forms were kept separate from the
data to protect confidentiality. Adolescent participants received $25 after completing and
returning all materials. Parent participants did not receive financial compensation for
their participation; this was decided upon to reduce the chance that an individual without
the diagnosis would enroll in the study simply to receive the cash incentive. A check was
mailed to the adolescent upon completion of materials. If materials were not returned to
the investigator within two weeks of when they were provided, a reminder card was sent
(see Appendix B). If materials remain outstanding two weeks after the reminder card
was sent, the investigator contacted the participants by phone. Approximately 20
families were dropped after some initial contact because they were either no longer
interested in participating or they provided insufficient information required for inclusion

in the study.

36

Measures

 

Demographic Ouestionnaa’ra. Demographic information was collected from the
parent via telephone interview (see Appendix C). The first question asked was whether
the adolescent was adopted, as adoption is considered an exclusion criteria because it
eliminates the otherwise established genetic vulnerability. The parent participant was
asked the following information regarding the diagnosed parent: date of birth, highest
level of education received, household income, number of others in the adolescent’s
immediate or extended family diagnosed with an affective disorder, parental diagnostic
and treatment history, and an allowance was given for a brief description of the parent’s
illness. Demographic information collected from the adolescent included: date of birth,
gender, grade level in school, ethnicity, and treatment and delinquency history (see
Appendix D).

Adolescent Psychopathology—Sade (APS: Reynolds. 1998). The APS is a
multidimensional self-report measure of adolescent psychopathology and social-
emotional problems and competencies designed for use with adolescents age 12 to 19. It
consists of 20 Clinical Disorders scales which have a high level of content validity with
respect to DSM-IV disorder symptom specifications and were designed to assess
empirically the severity of symptoms associated with specific DSM—IV clinical disorders.
The development of the APS was based on clinical samples of over 500 adolescents from
31 impatient and outpatient treatment settings in 21 states and a non-clinical sample of
over 2800 adolescents. Three APS Clinical Disorders scales were used in the current
study: Major Depression, Dysthymic disorder, and Mania. The Major Depression scale

consists of 29 items reflecting DSM-IV symptoms, with items rated on the basis of their

37

occurrence over the past 2 weeks. The Dysthymic Disorder scale consists of 16 items
rated as to their occurrence over the past six months. The Mania scale consists of 13
items that evaluate symptoms of a Manic Episode. The factor structure of these scales
correspond to the DSM-IV diagnostic criteria for these disorders and provides evidence
for construct validity. Reynolds (1993) reported an internal consistency reliability of .95
for both the normal and the clinical samples for the Major Depression scale. The
Dysthymia Disorder scale demonstrated internal consistency reliability coefficients of .89
and .88 for the clinical and non-clinical samples, respectively. The Mania scale
demonstrated internal consistency reliability coefficients of .81 and .80 for the normal
and clinical samples, respectively. Each scale represents a type of affective
symptomatology.

Scores on the APS are provided in the form of T-scores, with a mean of 50 and a
standard deviation of 10, with clinical levels variable across APS scales to accommodate
differences in base rates across disorders.

Inventory of Pm & Peer Attachmana (IPPA; Armsden. 1986; Armsden &
Greenberga1987). This measure was designed to assess adolescent’s perceptions of the
positive and negative cognitive-affective dimension of attachment relationships with their
parents and close fiiends — particularly how well these figures serve as sources of
psychological security. Three broad dimensions are assessed: degree of mutual trust;
quality of communication; and extent of anger and alienation. Trust (e. g., “When I am
angry about something, my mother/father tries to be understanding”) refers to confidence
or felt security that attachment figures are available and responsive to the adolescent’s

attachment needs. Communication (e.g., “I tell my mother/father about my problems and

38

troubles”) refers to the quality of verbal communication with attachment figures.
Alienation (e. g., “I don’t get much attention from my mother/father”) refers to the
amount of anger toward or emotional detachment from attachment figures. This scale
was included because frequent and intense anger and/or detachment represent responses
to actual or threatened disruption of an insecure attachment bond. The IPPA is a self-
report questionnaire and consists of 25 identical items in each of the mother, father, and
peer sections, yielding 3 attachment scores. Items are rated on a 5-point Likert-type scale
ranging from 1 (never true) to 5 (always true). Respondents are asked to rate attachment
to each parent up to age 12. This measure shows high convergent validity with the
original IPPA where Armsden (1986) reported alpha coefficients for the three scales from
.87 to .92. Armsden and Greenberg (1987) reported three-week test-retest reliabilities of
between .86 and .93.

The parent portion of the IPPA was designed to assess adolescent’s perceived
attachment to each parent by estimating the extent to which adolescents perceive that
each parent understands and is sensitive to the adolescents’ emotional concerns.
Participants or respondents are asked to think how well each item describes each parent’s
behavior while the subject was growing up, especially before age 12.

Parent attachment scores of 12-18 year-olds are moderately correlated with the
degree of positive family coping (communication among family members and relatives
concerning problems) (Lewis, Woods, & Ellison, 1987). In a sample of 10-to 16-year old
psychiatric patients, less secure parent attachment was related to clinical diagnosis of
depression, parent rating of the adolescent’s depressive symptoms, and to patient’s self-

reported level of depression (Armsden, McCauley, Greenberg, Burke, & Mitchell, 1991).

39

ScO‘eS 0n the IPPA have also been moderately associated with positiveness and stability
of self-esteem, life-satisfaction, and affective status (depression, anxiety,
resentment/alienation, covert anger, and loneliness) among late adolescents (Armsden &
Greenberg, 1987; Armsden, 1986). Scores on the IPPA were not found to be significantly
related to socio-economic status among a sample of 400 18- to 20-year-olds. In the same
study, negligible but significant positive correlations were obtained between attachment
and parents’ education levels (Armsden, 1986).

Social Provisions Scale ( SPS: Russell & Cutrona 1981; This measure of global

 

perceived social support was designed to assess the six functions of social relationships
proposed by Weiss (1974). These functions, termed “provisions” by Weiss, include the
following: (a) reliable alliance, assurance that one can count on others for tangible
assistance; (b) reassurance of worth, acknowledgement of one’s competence and skill; (c)
attachment, a sense of emotional closeness and security; (d) social integration, a sense of
belonging to a group of people who share common interests and recreational activities (e)
guidance, advice and information; and (t) opportunity for nurturance, a sense of
responsibility for the well-being of another person. The measure asks respondents to rate
the degree to which their relationships with others are currently supplying each of the
provisions. Each provision is assessed by four items, two that describe the presence and
two that describe the absence of the provision. Respondents indicate on a 4-point Likert-
type scale (1 = not at all true; 4 = completely true) the extent to which each statement
describes their current social relationships. For scoring purposes, the negative items are
reversed and summed together with the positive items to form a score for each social

provision. Summing the six individual provisions scores also forms a total social

40

provisions score. Internal consistency for the total scale score is high, ranging from .85 to
.92 across a variety of populations. Alpha coefficients for the individual subscales range
from .64 to .76. Reliability for the scale ranges from .87 to .91 across a range of samples
(Cutrona & Russell, 1987). Several studies support the validity of the SPS (Cutrona &
Russell, 1987; Russell, Cutrona, & Yurko, 1984). Among first year college students, the
six social provisions in combination accounted for 66% of the variance in scores on the
UCLA Loneliness Scale (Cutrona, 1982). Discriminant validity of the SPS has been
demonstrated against relevant measures of mood (e.g., depression), personality (e.g., self-
esteem), and social desirability (Russell & Cutrona, 1985).

To minimize the shared theoretical and method variance with the IPPA,
instructions for the SPS were slightly modified. The respondent was asked to only report
the level of social support received from “friends, family members other than parents,
and other adult members of the community (e. g., teachers, coaches, etc.).

Adolescent SociLSelf-Efficacy Scale (A-SSE; Connolly, 1989). This scale
contains 25 items which describe commonly occurring social events reported to be
problematic for teenagers (Ford, 1982; Furnham & Argyle, 1981). These include 5 items
describing social assertiveness (e.g., “Stand up for yourself when another kid in your
class makes fish of you”); 5 items describing performance in public situations (e. g.,
“Work on a project with a student you don’t know very well”); 5 items describing
participation in social groups or parties (e.g., “Put yourself in a new and different social
situation”); 7 items describing aspects of fiiendship and intimacy (e. g., Start a
conversation with a boy or girl you don’t know very well”); and 3 items describing

giving or receiving help (e.g., “Ask another student for help when you need it”). The

41

student is asked to rate each item on a 7-point scale ranging from “impossible to do” to
“extremely easy to do”. Total scores can range from 25 to 175. Connolly (1989) reported
reliability and validity information. Alpha coefficients of internal consistency are
reported at .90, .92, and .95 for two samples of high school students and one sample of
emotionally disturbed adolescents, respectively. Two-week test-retest reliability was
reported at .84. SSE was significantly correlated with components of the self-concept
including perceived social acceptance, general self worth, cognitive and physical
competence, and self-esteem. Construct validity was further supported by relations to
ratings of social adjustment.

Paar of Negative Eval_lartion Scale (FNE)-Revised (Watson & Friend. 1969;
Lem, 1983 1.

This scale, a brief version of the original FNE scale, was designed to assess one’s
fear of negative evaluation defined as “apprehension about other’s evaluations, distress
over their negative evaluations, and the expectation that others would evaluate oneself
negatively.” The content assesses concerns regarding negative evaluations as opposed to
the expectation that one was negatively evaluated. The original scale consists of 30 true-
false items that are summed, with higher scores indicating higher FNE. The revised
version utilized in the current study consists of 12 items and correlates with the original
version (.96). For the Brief FNE Scale, item-total correlations range from .43 to .75, and
Cronbach’s alpha coefficient is .90. Internal consistency is high, as reported at .92 and
.94 in two samples. One-month test-retest reliability is .7 5. Evidence for convergent
validity exists; FNE scores correlate moderately with other measures of apprehension in

social situations (e.g., SAD Scale, r =.51; Interaction Anxiousness Scale, r = .32).

42

1.3.6.5189

This study employed a correlational design. There was no manipulation of
variables or group assignment via random or nonrandom methods. Path analysis was
used here to test a theory of the possible causal order among the selected variables under
study. The study was designed to provide estimates of the magnitude of the hypothesized
effects.

One hundred and one subjects were enrolled but data was excluded from the
analyses for five subjects because they did not have scores on one or more of the
variables. Ninety-six adolescent participants and 86 parent participants were enrolled so
that satisfactory power to detect a medium effect size for the five-predictor model (or =
.05) examined here could be reached.

Analyses

A 2 x 2 analysis of variance was used to test for significant gender differences
between selected predictor variables and the criterion variable, APS scores. This
ANOVA tested whether there were adolescent gender differences in APS scores, whether
the gender of the diagnosed parent was related to APS scores, and whether the gender of
adolescent and the gender of the parent interacted to predict APS scores.
Intercorrelations of key variables were also computed. Then, a paired samples t test was
conducted to test for hypothesized mean differences among variables. Finally, a series of
linear regressions were conducted to test whether indexes of attachment bonds to the
diagnosed parent, attachment bonds to the non-diagnosed parent, perceived social
support, social self efficacy, and fear of negative evaluation made independent,

significant contributions to the prediction of adolescent affective psychopathology.

43

Adolescent affective psychopathology was a composite variable created by aggregating
scores and weighting three scales of the APS: dysthymia, mania, and depression. The
significance of interaction effects, after controlling for main effects, in predicting
symptomatology was tested. Data met the regression assumptions regarding normality
and homoscedasticity. Some degree of multicollinearity did exist, though it appeared to

be modest

RESULTS

In this section, data analyses are detailed and findings presented. First, scale
descriptives are provided. Relationships between demographic variables and adolescent
affective psychopathology scores were then tested. Demographic variables that were
found to be significantly associated with APS scores are briefly described.
Intercorrelations among these selected demographic variables and key model variables
are presented. Next, results from a comparison of means between the two predictor
variables - attachment bonds to the diagnosed parent and attachment bonds to the non-
diagnosed parent - are provided. Then, findings from a series of linear regressions testing
for main effects and interaction effects among predictor and criterion variables are
presented. These regressions tested whether attachment bonds to the diagnosed parent,
attachment bond to the non-diagnosed parent, perceived social support, social self-
efficacy, and fear of negative evaluation made Significant and unique contributions to the
prediction of adolescent affective psychopathology. Mediator and moderator hypotheses
were also tested and results provided. A path diagram is included that contains
standardized path coefficients depicting relationships among the variables. Lastly, several
follow-up post-hoc comparisons are presented.
Saale Descriptives

The means, standard deviations, and ranges of the key model variables are
presented in Table 1. There was a substantial range of scores for each scale; this is to be
expected given that individuals representing a wide range of affective symptomatolgy
were included in this study. Comparing means to normative groups for the scales used in

the current study is complicated by the fact that individuals for the current study were not

45

selected according to clinical versus nonclinical characteristics in the manner that

normative groups are chosen. Nevertheless, it is noteworthy that for this sample,

symptomatology is slightly positively skewed; however, this is to be expected given the

nature of the at-risk sample used.

Table 1

N’s Means Standard Deviations, and Ranges for Key Model mables

 

 

 

Variable N M SD Range or
1. Attachment bond 96 90.96 16.58 46-116 .05
to diagnosed parent
2. Attachment bond to 96 85.18 21.65 25-122 .05
non-diagnosed parent
3. PSS 96 79.94 7.68 62-94 .05
4. SSE 96 125.46 23.24 80-164 .05
5. FNE 96 37.05 8.13 16-50 .05
6. APS 96 2.28 .38 1.30-2.77 .05

 

Note. PSS = Perceived Social Support; SSE = Social Self-Efficacy, FNE: Fear of

Negative Evaluation; APS = Adolescent Affective Psychopathology.

46

Beliminary Analyses

Demogzaphic vambles. Analysis of variance and independent samples t test
procedures were used to explore whether demographic group differences among
adolescents that were associated with adolescent affective psychopathology. Results
showed that four demographic variables were significantly related to APS scores: Gender
of the diagnosed parent, parent education level, number of individual therapy sessions an
adolescent participated in, and adolescent psychotropic medication use. Results of the
analysis of variance and independent samples t-test procedures are presented in Table 2.

Analysis of variance showed no significant gender differences among adolescents
in this study on the measure of adolescent affective psychopathology (APS scores), 1 (96)
= 2.22, p > .05. However, a two-way analysis of variance showed that gender of the
diagnosed parent was significantly related to APS scores (F (4,85) = 12.11, p < .001).
Adolescents who had a diagnosed father had higher APS scores (M= 2.47, SD = .24)
than those adolescents with a diagnosed mother (M = 2.193, SD = .41). The interaction
term (gender of the adolescent * gender of the parent) was not significant (F (1,93) =
1.443, p > .05. Thus, when the adolescent’s father is the parent who has been diagnosed
with bipolar disorder, adolescent psychopathology scores are statistically higher
regardless of the gender of the child.

One-way analysis of variance showed that group differences exist regarding level
of parent education, F (2,89) = 5.72, p < .01. Tukey’s post hoc comparisons showed that
adolescents whose parents had attended college (M = 2.30) or graduate training (M =
2.39), scored higher on the APS than those whose parents had completed high school

only (M = 1.88). This comparison was significant at the p < .05 level.

47

One-way analysis of variance showed group differences on APS scores between

five or fewer sessions.

adolescents who reported that they had received different amounts of individual
cormseling, F (3,84) = 3.381, p < .05. Adolescents who have received 16 or more

individual counseling sessions scored lower on the APS than those who have received

Findings from an independent samples t-test procedure showed that adolescent

APS than those who had not.

Table 2
Reaults from ANOVAsJand Indgpendent Samples T Tests Testing Sigm'ficant

psychotropic medication use was significantly associated with APS scores, t (96) =

-2.862, p < .01. Adolescents who had taken psychotropic medication scored lower on the

Relationships Between Demoggphic Variables and APS Scores“

 

 

 

 

Demographic H M 8_D
Variable
Gender of the
Diagnosed Parent
Male 36 2.47 .24
Female 60 2.20 .41
Parent Education
High School 6 1.88 .45
College 57 2.30 .35
College + 29 2.40 .29
# Adolescent Individual
Therapy Sessions
47 2.35 .34
1-5 17 2.36 .42
6-15 10 2.26 .23
16 + 14 2.01 .46
Adolescent
Psychotropic
Medication use
Yes 15 2.03 .26
No 73 2.34 .38

 

 

 

 

48

Note. Only demographic variables reaching statistical significance are presented.

lnterrcorrelations AmonaDemographic Variables and Scales.

A correlation matrix demonstrating relationships among the demographic
variables that were associated with this study’s outcome variable and the key model
variables is presented in Table 3. It was expected that indexes of attachment bonds to
parents would be significantly and positively correlated with each measure of social
support (after fear of negative evaluation scores were reversed) and with the outcome
measure. Only partial support for these relationships was found. Only indexes of
attachment bonds to the diagnosed parent were significantly correlated to measures of
social support competence, namely perceived social support and social self-efficacy.
Unexpectedly, parental attachment bonds were not significantly correlated with APS
scores. As expected, all measures of social support competence were significantly and
positively correlated with each other. Of all the predictor variables studied, only fear of

negative evaluation was correlated with APS scores.

49

_o. .v as.» dmo. V Q a

com: 33 5:862: 6: cc 5523 2365 3.58 2: :2: seem emcee Etc-:6

£3 8: 222362: 39:55me 508225 .wEBmcsoo use 5:325 .3 296. $sz 9.28:2: 35% 5:3: :23 Joint? 325ch 803 .3233
wczcmcaoo 3:239: .8 398:: one 5:83» :8ch .xowEmEoEExm :wE 3:365 38% :wE E; .8732; 3035250 325852 3.58 mm<
Swo—ofimaofixma egos-ca, Euomc_oc< u mad. ”cones—gm 02$ch Lo Lao-a umZu .zumoc-Lm-tcm Eoom u mmm ”toaasm Euom c3680“. u mmm clog-Iona

 

-- E25. nomocwaa .Lo 3980 .o_

2.- -- 33 :28232
uaobozoxmmcccomflocac .o

5.- sum. -- mcofimcm wE—cmcsou
.3636:— Co cons—:2 .w

 

Lem.- rremr 2.- l. 22.833. 23:23 <-

remmr E .- .1.er rem-o... ma< .o

3.- tom. mo. :2. 2mm. -- mZ-L .m-

ioe. oo. 5. 1.5m.- _o. .13. -- mmm .v

o_. crwmr imr rmmr rmm. rewm. Iwm. -- mmn. .m
Eccmq ucmocch-co:

tam.- N- _ .- 3.- 5. oo. o _ .- w _ .- mo. 2 econ EcEsuSE .m
ESE ocmocwflc o.

_o.- _o. .m. rmmr vor 3.- .13. tom. 2. -- econ Ecfizomc< ._
2 a w e c m e m N _ smelt;

 

833:; E52 GM Em 333:; DEG-EmoEcQ cocoa—cm wcoE< macaw—ctoagfi

m 2ch

50

Results of Primg Analyses

Relations of adolescents' attachment bonds with each parent. To test the first

 

hypothesis that in comparison to the relationship between the adolescent and the BPD
parent, the relationship between the adolescent and the non-diagnosed parent should
reflect a higher degree of secure parent-adolescent attachment, a paired-sample t test was
conducted and sample means compared. Results showed that the hypothesis was not
supported. The mean score for the index of attachment bond between the adolescent and
his or her diagnosed parent was 90.96 (SD = 16.58), whereas the mean score for his or
her non-diagnosed parent was 85.17 (SD = 21.65), t(96) = 2.262, p = .026. Adolescents
in this sample reported more secure attachment bonds with the BPD parent than with the
non-diagnosed parent.
Ragression Aaalysis: Contributions of Parental Attachment Bonds to the Prediction of
Adolescent Affective Psychopathology

A hierarchical regression was conducted to test the hypothesis that after
controlling for selected demographic variables, indexes of attachment bonds to the
diagnosed parent and to the non-diagnosed parent would each make independent and
significant contributions to the prediction of adolescent affective symptomatology. In
addition, the hierarchical regression tested the hypothesis that, after controlling for main
effects, the interaction of indexes of attachment bonds with each parent would make
unique, significant contributions to the prediction of APS scores. The demographic
variables: Gender of the diagnosed parent, parent education, number of adolescent
individual counseling sessions, and adolescent psychotropic medication use were entered

as covariates in step one. At step two, indexes of attachment bond to the diagnosed

51

parent and attachment bond to the non-diagnosed parent were regressed on the measure
of adolescent affective psychopathology. At step three, the interaction effect of these
respective bonds was tested. As recommended by Aiken and West (1991), measures of
attachment were centered prior to creating the interaction terms and testing their
significance to reduce multicolinearity effects. Results are presented in Table 4.

The covariate block entered at step one was a significant predictor of APS scores;
the covariates collectively accounted for 23% of the variance in APS scores (132 = .232, p
< .001). Unexpectedly, the addition of parental attachment bonds at the second step was
not significant in adding to the prediction of APS scores (AR2 = .047, p > .05). The
interaction term added in step three did significantly and positively contribute to the
prediction of APS scores (AR2 = .063, p < .01), explaining an additional 6% ofthe
variance in APS scores. As hypothesized, results showed that attachment bonds to the
non-diagnosed parent moderated the relationship between attachment bonds to the
diagnosed parent and APS scores, t(96) = -2.726, p < .01. The significant interaction
effect was plotted as advised by Aiken and West (1991) to help interpret the interaction
(see Figure 2). Data met regression assumptions regarding normality, linearity, and
homoscedasticity. This interaction demonstrated that high indexes of attachment to the
diagnosed parent are associated with lower levels of symptomatolgy and that this
relationship is more pronounced when strong attachment bonds with the non-diagnosed

parent also exist.

52

Table 4
Summary of Hierarchical llagression Analysis for Attachment Bonds PredictingAPS

 

 

 

scores (N = 96).
Variable B S_E B 13
Step 1
Gender of Diagnosed Parent -.24 .08 -.31**
Parent Education .12 .07 .18
Number of Individual -.06 .03 -.23*
Counseling Sessions
Adolescent Psychotropic -. 13 . 13 -. 1 1
Medication use
Step 2
Attachment bond to -.00 .00 .18
diagnosed parent
Attachment bond to -.00 .00 -.20
non-diagnosed parent
Step 3
Interaction -.00 .00 -.28**
(attachment bond to
diagnosed parent *
attachment bond to
non-diagnosed parent)

 

Note. Adjusted R’ =. 194 for Step 1 (p < .001); A Adjusted RZ= .030 for Step 2 (us); A
Adjusted 32 = .059 for Step 3 (p < .01).

* p < .05; ** p <01

53

F_igure 2: Interaction for Regression Angaysis: Attachment Bond to
Diagnosed Parent & Attachment Bond to Non-Diagnosed Parent as
Predictor Variables and APS Scores as Criterion Variable

1.5 ‘1

 

0.5 - \ _+_ Low x2

0— "*a‘ -~~thx2

 

 

 

-0.5 —

 

 

Low x1 High x1

Note. X1 = Attachment Bond to Diagnosed Parent, X2 = Attachment Bond to Non-

Diagnosed Parent

Regression Analyses: Contributions of Parental Attachment Bonds to the Prediction of
Social Support Competence

It was hypothesized that after controlling for selected demographic variables,
indexes of secure parent-adolescent attachment would be Significantly related to each
measure of social support competence. To test this hypothesis, three separate regression
analyses were conducted. For each regression, main effects and interaction effects were
tested using a three-step regression procedure. At the first step of each regression,
selected demographic variables were entered as covariates. Attachment measures were
entered at step two. Each measure of social support competence (PSS, SSE, and FNE)

represented the dependent variable, respectively. Prior to testing the interaction term,

54

indexes of attachment to the diagnosed parent and to the non-diagnosed parent were
appropriately “center ” as advised by Aiken & West (1991). Results of these regression
analyses are presented in Table 5.

Attachment Bonds and Perceived Social Support. The covariate block alone
contributed 27% to the overall variance in PSS scores (R2 = .275, p > .001). While
attachment bond to the diagnosed parent, t(96) = 3.44, p < .01, was significantly and
positively predictive of PS S, no significant main effect was found for attachment bond to
the non-diagnosed parent, t_(96) = - 1.08, p > .05. When entered as a block, attachment
variables explained an additional 9% of the variance in PSS scores beyond that explained
by the covariates alone, F (2,79) = 5.9, p < .01. The entry of the interaction term at the
third step was not significant in enhancing the prediction of perceived social support.

A_ttachment Bonds and Social Self-Efflcaazy. The covariate block alone
contributed almost 18% of the overall variance in SSE scores (32 = .177, p <.01).
Parental attachment bonds made significant individual contributions to the variance in
SSE scores accounting for an additional 12% of the variance in SSE scores (AR2 = .126,
p 5 .001). However, while attachment bond to the diagnosed parent, t(96) = 3.70, p <
.001, was significantly and positively predictive of SSE, no significant main effect was
found for attachment bond to the non-diagnosed parent, t_(96) = -1.613, p > .05. The

inclusion of the interaction term was not significant in enhancing the prediction of SSE.

 

A_ttachment Bonds and Fear of Negative Evaluation. The covariate block did

make a significant contribution to the overall explained variance in FNE scores

55

(R2 = .123, p < .05), explaining 12% of the overall variance. However, after covariates

were controlled, the additional predictors did not make significant, unique contributions

to FNE.

Table 5

Summary of Hierarchical Regression Analysis for Attachment Bonds Predicting Socia_l

Competence scores (N = 96).

 

 

Variable 3 Si: 3 B
Perceived Social Support

Step 1
Gender of Diagnosed Parent -2.1 1.63 -. 13
Parent Education -5.30 1.47 -.37**
Number of Individual -.77 .56 -. 13
Counseling Sessions
Adolescent. Medication Use -12.59 2.75 -.49**

Step 2
Attachment bond to .15 .04 .33**
diagnosed parent
Attachment bond to -.03 .04 -.11
non-diagnosed parent

Step 3
Interaction .OO .00 . 1 57
(attachment bond to
diagnosed parent *
attachment bond to

non-diagnosed parent)

56

(table continues)

 

 

Variable .3. SE B B
Social Self Efficacy

Step 1
Gender of Diagnosed Parent 15.18 5.05 .32**
Parent Education -9.04 4.56 -.21
Number of Individual .48 1.73 .02
Counseling Sessions
Adolescent Medication Use 4.52 8.5 .06

Step 2
Attachment bond to .52 .14 .38***
diagnosed parent
Attachment bond to -.17 .11 -.17
non-diagnosed parent

Step 3
Interaction -.OO .00 -.07
(attachment bond to
diagnosed parent *
attachment bond to
non-diagnosed parent)

Fear of Negative Evaluation

Step 1
Gender of Diagnosed Parent -1.31 1.92 -.07
Parent Education -3.25 1.73 -.21
Number of Individual -.25 .66 -.04
Counseling Sessions
Adolescent Medication Use 5.79 3.24 .21

Step 2
Attachment bond to -.O4 .05 -.09
diagnosed parent

(table continues)

 

Variable

loo
If?
It:
a

 

Attachment bond to -.07 .04 -.21
non-diagnosed parent

Step 3
Interaction -.00 .00 -. 13
(attachment bond to
diagnosed parent *
attachment bond to
non-diagnosed parent)

 

Nata. PSS: Adjusted R2 =.239 for Step 1(p < .001); A Adjusted 32: .08 for Step 2 (p
<.Ol); A Adjusted 32 = .012 for Step 3 (ns). SSE: Adjusted 32 =. 136 for Step 1(p < .01);
A Adjusted l_{2 = .114 for Step 2 (p <.001); A Adjusted 32 = .005 for Step 3 (ns). FNE:
Adjusted 132 = .080 for Step 1 (p < .05); A Adjusted R’ = .00 for Step 2 (ns); A Adjusted
32 = .005 for Step 3 (ns).

p < .05; ** p < .01; *** p_< .001

Rggression Analysis: Contributions of Perceived Social Support, Social Self-Efficacy,
aad Fear of Negative Evalaation to the Prediction of Adolescenit Affective
Psychopathology.

Tables 6, 7 and 8 present the results of these analyses. As hypothesized,
regression analyses demonstrated that after controlling for demographic variables,
perceived social support was significantly and positively predictive of adolescent
affective psychopathology, t(96) = 2.80, p < .01. Results also supported the hypothesis
that fear of negative evaluation (reverse coded) was significantly and positively

predictive of APS scores, t(96) = 5.06, p < .001. Additionally, social self-efficacy was

58

significantly and positively predictive of adolescent affective symptomatology, t(96) =
1.90, p < .05.
Table 6

Summary of Hierarchical Regession Analysis for Perceived chal Sapport Predicting

APS scores (N = 96).

 

 

Variable B SE_B B
Perceived Social Support
Step 1
Gender of Diagnosed Parent -.24 .07 -.31***
Parent Education . 12 .07 .17
Number of Individual -.06 .02 -.23**
Counseling Sessions
Adolescent Medication Use -.13 .13 -.11
Step 2
PSS .01 .00 .33**

 

Note. *p < .05; ** p < .01; *** p_< .001

Table 7

Summary of Hierarchical Reggession Analysis for Social Self Efiicac_y Predicting APS

 

 

 

scores (N = 96;
Variable E SE B [3
Social Self Efficacy
Step 1
Gender of Diagnosed Parent -.24 .07 -31“:

59

Parent Education

Number of Individual
Counseling Sessions

Adolescent Medication Use

Step 2

SSE

.12

-.06

-.13

.00

.07

.02

.13

.00

.17

-.23**

-.11

.22“

 

Note. *p < .05; ** p < .01; *** p_< .001

Table 8

Summary of Hierarchical Regression Analysis for Fear of Negative Eva_lu_ation Predicting

APS scores (N = 96).

 

 

 

Variable B SE_B 13
Fear of Negative Evaluation
Step 1
Gender of Diagnosed Parent -.24 .07 -31***
Parent Education . 12 .07 .17
Number of Individual -.O6 .02 -.23**
Counseling Sessions
Adolescent. Medication Use -. 13 . 13 -.1 1
Step 2
FNE .02 .00 .48***

 

60

Note. PSS: Adjusted R’ =.194 for Step 1(p < .001); A Adjusted R9: .074 for Step 2 (p
<.01). SSE: Adjusted 32 = .194 for Step 1(p < .001); A Adjusted 32: .035 for Step 2 (p

<.05); FNE: Adjusted 32 = .194. for Step 1 (p < .001); A Adjusted 32 = .213 for Step 2 (p
< .001).

* p < .05; "p < .01; *** p_< .001

Figure 3 presents a summary of the obtained coefficients representing relationships

among all key model variables.

61

figure 3: Path Diang Presenting a Summary of the Obtained Standardized
Coefficients for Tested Relationships AmongAll Proposed Model Variables.

(-..23)

 

 

 

   

 

V
A

(48mins)

 

 

(.12)

Nata. X1 = Attachment bond to diagnosed parent; X2 = attachment bond to non-
diagnosed parent; PSS = perceived social support; SSE = social self efficacy; FNE = fear
of negative evaluation, APS = adolescent psychopathology scale.

Standardized coefficients were used.

*p < .05; **p< .01; ***p < .001

Post Hoc Analyses

The goal of the current study was to propose and test a mediational model for
explaining the relationship between parental attachment bonds and adolescent affective
symptomatology. In order to examine mediating effects, however, a predictive
relationship between parental attachment bonds and APS scores was required.
Surprisingly, parental attachment bonds did not predict APS scores for this study sample.
Given this pivotal finding, an analysis of the proposed mediator model as initially
proposed was unwarranted. Instead, post hoc analyses were conducted to explore
whether the attachment construct as measured in this study was masking a relationship
between the variables. Additionally, post hoc analyses were conducted to further test
model variable relationships.

One potential reason for the inability of the attachment measures to predict APS
scores in this study involves the measurement of parental attachment bonds. The parental
attachment bond construct utilized in this study contains three subscale components:
trust, communication, and alienation. In the analyses presented in this chapter to this
point, these subscales were summed to obtain a total parental attachment bond score. This
aggregate score was used to test the study hypotheses. However, given that a relationship
between attachment bonds and affective symptomatology has been well documented, it
seemed worthwhile to explore the possibility that a relationship in fact existed but that
using an aggregate score masked the relationship. Scale descriptives and intercorTelations

among these components of attachment and key model variables are presented in Table 9.

63

_o. V a... .mo. V a...
6.8m 30353263; E8333: u mm< Hogs—gm uZEwoZ :0 Sum n mZm SomoEm toaqsm Ecom u mmm "tomazm Eoom “630on
n mmn. 3:23 cowgwflnéo: on: 5.3 @2383: c2232: E68553 n mx 523:2? 3:88 nomocwflu of 5:.» 33688 5:22.: E0858»:

_x 52222 5:28 nomocwmfiéo: of 5:: 3860mm: :ocmoEsEEoQ u Nx :ozmo_::EEoU 2:23 38:36 05 .23 33888 :2825888

n .x :ocmoEzEEoU .883 nomo:wm€-:o: 2: 5:5 3:38me .25 u NX 53:. 3:28 :30:me 05 5:5 @8383: 3?: n Ca 53,—. .3qu

o_r

eewa

cor

esON£

......

o;

*§—V.

asmm.

w_.

0—.

o_.

vo.

_ox

: ‘I III, II..lrrOr I) to

o;

sewn.

avm.

oo.

oo.

*eoo.

mo.

0;

N_.

0—.

mo.

o;

{*mrN.

no.

iiVm‘!

Nor

in!

*in

o;

aevv.

co.

*eov.

no.

G;

aahv.

*aoo.

*aom.

o;

«cow. 0;

**_o.

to
1 N

eemw.

o;

meg.

«;.o

o_.m

vogw

oNa

__.o_

was:

am

.19!

I..-tier, all -91-I..... ‘ 'ri .

wan”

wogkn

Qth_

voAK.

vmd;

owé;

vw4A”

wmwmm

mvfmm

madam

co

co

co

co

mo

00

co

00

mo

co

us. I.. ‘11.. li.‘n'lv

£2 .3:

M77.
mmm

mm:
Q 8:222
_x 5:822
Nx :ozmggEEoU

_x :o_8o_::EEoU
NXHQCH

_Vn.m=+H

22E;

4 .3 r ivrvir II?

.0

:6

hi

mozmcm> Eco—2 GM cc: €039.22 6::
623358800 $2.5 mucom EoEcowz< .958: go mEocanoU 05 :8 228—28885 we: .mcoumgom Emwafim £882

a 2an

64

Post hoc analyses were conducted to explore relationships among the attachment
components and APS scores. First, post hoc regression analyses were conducted to
determine if the components of the parental attachment made unique and significant
contributions to the prediction of APS scores when measured independently. To
investigate this possibility, APS scores were regressed on each attachment component
separately. To remain consistent with the model as initially proposed, each component
was assessed in relation to the diagnosed parent as well as the non-diagnosed parent.
Thus, six separate regression analyses were required using APS scores as the dependent
variable in each. Main effects and interaction effects for each regression were tested
using a three-step regression procedure. For each regression, the covariates were entered
in the first step. At step two, trust, alienation, and communication were entered,
respectively, as a block. The interaction term was included at step three. Again, as
recommended by Aiken and West (1991), measures of trust, communication, and
alienation were appropriately “centered” prior to conducting the analyses.

Results of the three regressions are presented in Table 10. As already described
in earlier analyses, the covariate block alone contributed 23% to the overall variance in
APS scores (R2 = .232, p <.001). When trust was entered at step two, it also made a
significant contribution to the variance in APS scores, accounting for an additional 5% of
the variance in APS scores (AR2 = .05, p < .05). However, while trust associated with
attachment bond to the diagnosed parent, t(96) = -2.53 3, p < .05, was significantly and
positively predictive of APS scores, no significant main effect was found for trust

associated with attachment bond to the non-diagnosed parent, t (96) = 1.268, p > .05. The

65

inclusion of the interaction term was not significant in enhancing the prediction of APS
scores.

No main effects or interaction effects were found when APS scores were
regressed on communication scores. Given the particularly strong correlational
relationships between trust and communication scores for the diagnosed parent (r = .91,
p < .01) and for the non-diagnosed parent (r = .90, p < .01), post hoc partial correlations
were computed to firrther examine the relationship between communication and APS
scores. However, after controlling for trust scores associated with the diagnosed parent,
partial correlation coefficients indicated that communication scores were no longer
statistically related to APS scores (r = .3 5, p > .05). After controlling for trust scores
associated with the non-diagnosed parent, communication scores were also statistically
unrelated to APS scores (r = .22, p > .05). These findings fail to support the notion that
quality of communication with parents is linearly related to risk for affective
symptomatology.

Results indicated that after controlling for covariates, alienation did not make a
unique significant contribution to the prediction of APS scores. However, the inclusion
of the interaction term for alienation was significant, t (96) = 2.512, p < .05, in enhancing
the prediction of APS scores, which precludes the interpretation of main effects. The
interaction term contributed an additional 6% to the overall variance explained. The
significant interaction effect was plotted to help interpret the interaction (see Figure 4).

Data met regression assumptions regarding normality, linearity and homoscedasticity,

66

Table 10

Summary of Hierarchical Remssion Analysis for Trust. Communication, and Alienation

Predicting APS scores (N = 96;

 

 

Variable E SE B B
TRUST
Step 1
Gender of Diagnosed Parent -.24 .07 -.32**
Parent Education .12 .07 .18
Number of Individual -.05 .07 -.22*
Counseling Sessions
Adolescent Medication Use -.13 .13 .10
Step 2
Trust: Attachment bond -.01 .00 -.29*
to diagnosed parent
Trust: Attachment bond to .00 .00 .15
non-diagnosed parent
Step 3
Trust Interaction .00 .00 .13
(Trust: Attachment bond to
Diagnosed parent * Trust: Attachment
Bond to non-diagnosed parent)
COMMUNICATION
Step 1

Demographic Variables

67

Step 2

Communication: Attachment -.01
bond to diagnosed parent
Communication: Attachment bond 00
to non—diagnosed parent

Step 3

Communication Interaction .00
(Communication: attachment bond

to diagnosed parent *

Communication: attachment bond

to non-diagnosed parent)

ALIENATION

Step 1
Demographic Variables

Step 2
Alienation: Attachment bond .00
to diagnosed parent

Alienation: Attachment bond to .00
non-diagnosed parent

Step 3
Alienation Interaction .00

(Alienation: attachment bond to
diagnosed parent * Alienation:
attachment bond to non-diagnosed parent)

.00

.00

.00

.00

.00

.00

-.26*

.11

.06

.01

-.05

.72*

 

Note. Trust: Adjusted R2 =. 199 for Step 1(p < .001); A Adjusted Ll": .043 for Step 2 (p

<.05); A Adjusted R” = .008 for Step 3 (ns). Communication: Adjusted R’ =.194 for Step

1(p < .001); A Adjusted E = .036 for Step 2 (ns); A Adjusted 32 = .009 for Step 3 (ns).

Alienation: Adjusted 32 = .194 for Step 1 (p < .001); A Adjusted 32 = .068 for Step 2

(ns); A Adjusted 32 = .052 for Step 3 (ns).
* p < .05; ** p < .01; *** p_< .001

68

Figure 4: Interaction Between Attachment Alienation Associated the
Diagnosed Parent and Attachment Alienation Associated with the
Non-Diagnosed Parent in Predicting APS Scores
2.6
2.5 -

2.4 -

 

 

2.3 " 0 Low x2
2.2 - flgh x2

 

 

 

APS Scores

2.1-
2..

 

 

 

1.9 .
Low x1 High x1

Alienation

Three patterns are suggested by the significant interaction: 1) low alienation
associated with the diagnosed parent is related to lower symptomatology when lower
alienation scores associated with the non-diagnosed parent are reported; 2) in cases where
alienation associated with the non-diagnosed parent is high but low levels of alienation
associated with the diagnosed parent exist, slightly lower symptomatology levels occur;
and 3) the greatest effect of alienation on symptomatology occurs when high levels of

alienation are associated with both the diagnosed parent and the non-diagnosed parent.

69

DISCUSSION

mm

This chapter initially summarizes and interprets this study’s key findings. The
ways in which the findings of this study advance research into risk for affective
psychopathology among offspring of bipolar disordered parents are then highlighted.
Where hypotheses of this study were not supported, possible explanations that draw upon
existing research and theory are offered. Implications of this study’s findings for
therapeutic intervention and prevention are also discussed. Finally, methodological
limitations are noted and recommendations for future research are addressed.

The purpose of this study was to advance understanding of risk for affective
psychopathology among adolescent children of bipolar parents. Because specific
mechanisms through which risk is heightened or attenuated are not well understood, the
focus of this study was to identify and test a potentially useful mediational model. The
model tested was premised on extant genetic and psychosocial literature indicating that
both the quality of one’s relationship with parents influences one’s sense of social
competency; this, in turn may buffer the negative effects of parental affective disturbance
on offspring. Attachment theory served as the lens through which this topic was
addressed because of its heuristic utility in explaining the processes linking relational
competency and outcomes for at-risk offspring. Attachment security has repeatedly been
associated with better outcomes for children of affectively disturbed parents.

The present study tested a path model which posited that quality of attachment to
parents influences one’s perception of available social support, self-efficacy, and fear of

being evaluated by others. These elements of social competency, in turn, impact risk for

70

affective symptomatology according to the model proposed. Hypotheses were tested
regarding specific relationships among parental attachment bonds, perceived social
support, social self-efficacy, fear of negative evaluation, and adolescent affective
symptomatology. Next, a regression model indicating that attachment bonds were
unique, significant predictors of adolescent affective psychopathology was tested.
Consistent with the notion that attachment security promotes a healthy sense of self and
others, it was also hypothesized the data would reveal that secure attachment with the
non—diagnosed parent would moderate the relationship between insecure attachment with
the diagnosed parent and affective symptomatology. Furthermore, it was anticipated that
perceived social support, social self-efficacy, and fear of negative evaluation would
predict adolescent affective psychopathology and mediate the relationship between
parental attachment bonds and affective disturbance. These hypotheses were motivated
by research findings which suggested that elucidating the specific coping mechanisms
offspring engage in when dealing with an afl‘ectively ill parent is an important step
toward reducing risk and improving intervention programs for this at-risk population

(Goodman & Gotlib, 1999; Hodgins, F aucher, Zarac, & Ellenbogen, 2002).

Correlational Analyses
The path model presented in the current study was based, in part, upon the

assumption that parental attachment bonds would be significantly correlated with APS
scores; however, the findings did not support this assumption. Additionally, correlational
analyses provided only partial support for anticipated relationships among indexes of
parental attachment, social support competency, and adolescent affective
symptomatology. Indexes of attachment bonds to the diagnosed parent were significantly

and positively correlated with perceived social support and social self-efficacy but were,

71

unexpectedly, unrelated to fear of negative evaluation. It was anticipated that indexes of
parental attachment bonds with the non-diagnosed parent would likewise be correlated
with each measure of social competency; however, findings again did not demonstrate
correlational relationships among these variables. Indexes of perceived social support,
social self-efficacy, and fear of negative evaluation measure distinct but related
constructs; therefore, it is not surprising that they were all significantly and positively
correlated with each other. Further, while perceived social support and fear of negative
evaluation were correlated with affective symptomatology as anticipated, social-self
efficacy was not. A more detailed discussion of findings related to the relationship
between social self-efficacy and adolescent risk follows.

Group Comparisons of Secure Parental Attachment_

Contrary to hypotheses, adolescents in this study reported stronger attachment
bonds with the diagnosed parent than with the non-diagnosed parent. This appears to be
the first study that directly compares parental attachment between these two groups.
Related studies compare parental attachments of clinical versus non-clinical groups rather
than differentiating between the diagnosed parent and the non-diagnosed parent. In
research on attachment theory and parental affective disturbance, the focus has been
primarily on mothers-child relations. However, the risk literature has documented the
positive role of a supportive adult in increasing the resilience of children at risk for
psychopathology (Garmezy, 1985; Werner & Smith, 1980). Including fathers in the
sample of diagnosed parents and attending to the attachment relationship with the non-
diagnosed spouse is somewhat unique and may serve to broaden and diversify

understandings of resiliency and risk.

72

The expectation that attachment bonds with the non-diagnosed parent would be
stronger than those with the ill parent were based on extant research with depressed and
bipolar parents which indicate that affective disturbance compromises parenting abilities.
It has been reported that affective illness reduces a parent’s dependablity, availablity, and
consistency, and that mood and energy dysregulation makes it less likely the parent can
generate the sustained, effortfirl behavior required to cultivate secure attachment bonds
(Anthony, 1975; Zahn-Waxler, 1984). While it should be noted that stronger attachment
bonds with the ill parent found in this study are only relative to the attachment bonds
with the non-ill parent in this study, these findings are in line with the view that
assumptions should not be made about attachment relationships based on diagnosis alone
(Seifer, Sameroff, Dickstein, Keitner, & Miller, 1996).

Methodological explanations for the difference in attachment bonds with each
parent found here are also possible. That is, in this study sixty percent of the parents in
the diagnosed-parent category were mothers. Most research findings have shown that
adolescents describe themselves as closer to mothers than fathers (Pipp, Shaver, Jennings,
Lamborn, & Fischer, 1985). Furthermore, there are indications that children who have an
affectively ill parent may be more empathic and prosocial towards a diagnosed mother
than a diagnosed father. It was reported that children of parents with depressed mood
were more able to recognize their mothers’ unhappiness and describe their own feelings
about their mothers’ moods (Solantaus-Simula, Punamaeki, & Beardslee, 2002). The
authors suggested that differences such as a heightened awareness of, and dialogue about,
their mothers’ feelings and their own feelings about the mother’s moods, in combination

with the mothers’ tendency to more promptly validate their child’s responses might

73

reflect more emotionally close relationships with mothers. Similarly, in a study using a
cross-sectional sample of adolescents investigating perceptions of attachments with
parents, adolescents reported seeking their mothers more than their fathers in support
seeking situations and reported a higher quality of affect toward their mothers than
fathers (Paterson, Field, & Pryor, 1994). The authors suggested that there may be more
to the higher utilization of mothers than simply the fact that she is an easier person to
communicate with. Fathers appeared to be perceived as less emotionally involved and
thus not as equipped as mothers to offer appropriate comforting.

Additionally, bipolar parents who have spouses or who are in close contact with
family members represent only a portion of bipolar parents. Many parents with BPD are
unmarried, divorced, or alienated from their families due to the strain the symptoms of
the illness can place on relationships. It has been reported that a combination of parental
affective illness and divorce is associated with poor outcome in the offspring of
affectively ill adults (Beardslee et al., 1993). The very nature of the two-parent
household required for inclusion in this study suggests that these diagnosed parents may
have higher levels of support than others. Given that support from the non-ill parent has
been linked to better outcomes for persons with affective disorders (Prince & Jacobson,
1995), it is possible that the diagnosed parents in this study were less severely impacted
by their illness than those represented in other studies in this area, making them more
able to foster strong attachment relationships with their adolescent offspring.

Finally, research has also documented that more severe parental symptomatology
may support an understanding of the parent as “ill” (Rutter, 1990). Offspring may be

better able to understand and accept mood variability when it is pronounced. Similarly,

74

offspring may find less severe symptomatology more complicated and confirsing. While
some information was obtained on the nature of the illness and treatment, it is beyond the
scope of this study to be able to paint a clear picture of whether the parent could be
classified as severely affected.
Mal Attachment Bonds Predicting Adolescent Affective Psychopatholgy

The data did not support the key assumption for this study that parental
attachment bonds would make significant and unique contributions to the prediction of
adolescent affective psychopathology. This result is particularly surprising given the
extent of empirical and theoretical support for a relationship between these variables
(Allen, 1999; Armsden & Greenberg, 1987; Engels, F inkenauer, Meeus, & Dekovic,
2001; Papina & Roggmann, 1992). The present study was designed in accordance with
longitudinal research which demonstrates that insecure attachment predicts
psychopathology (Kenny et. al., 1998; Papini & Roggman, 1992). Research also
suggests that insecure attachment can be interpreted as a sequela of more longstanding
psychiatric problems (e. g., depression) (Allen, Hauser, & Borman-Spurell, 1996). In the
present study, long-term psychiatric difficulties were not assessed. However, indexes of
adolescent affective symptomatology and information on their psychiatric history were
obtained. Interestingly, overall the participants had slightly lower levels of
symptomatology than might be expected. It should be noted that a slightly restricted
range was found among APS scores, making it more difficult to detect a statistically
significant effect. Furthermore, to date, published IPPA-R scores for non-clinical
samples of adolescents are not available. However, a comparison of how parental

attachment scores fiom the current study compare to those from a non-clinical sample

75

could serve as a basis for developing more sophisticated research questions concerning
the nature of the relationship between attachment and risk for psychopathology.

Another explanation for the unsubstantiated hypothesis regarding attachment
bonds and APS scores is the idea that the conceptualization of the parent-child
relationship may change with recovery for both the parent and the child (Cummings &
Cicchetti, 1990). That is, perhaps the higher functioning sample represented here
participated because they were able to do so — because they were “recovering.”

However, at this time it is unclear where study participants rank in terms of recovery.

The lack of a significant association between parental attachment bonds and APS
scores may be a result of the study design. Initially, the three components of parental
attachment bonds (e. g., trust, communication, and alienation) were aggregated to indicate
the relative degree of attachment security with each parent. However, measuring a
multidimensional construct in aggregate can mask important information about their
unique contributions. Indeed, when the components of attachment were treated as focal
in post hoc analyses, relationships with APS scores were found. Specifically, trust
associated with the diagnosed parent predicted adolescent affective symptomatology, as
did the interaction between alienation associated with the diagnosed parent and alienation
associated with the non-diagnosed parent. These findings are consistent with research and
theory that stresses the multidimensional character of attachment and those which
emphasize the importance of mutual understanding and respect as well as feelings of
anger and isolation (Armsden & Greenberg, 1987; Parkes & Stevenson-Hinde, 1982).
However, the absence of a relationship between the communication subscale and

affective symptomatology warrants comment. For the current study, communication and

76

trust scores were very highly correlated, which suggests that measurement errors may
have occured. However, it is curious that when the relationship between communication
and APS scores was tested after controlling for the effects of trust, no relationship
emerged. Perhaps communication and trust both tap a construct unrelated to affective
symptomatology. Furthermore, communication and APS scores may have a curvilinear
relationship. Additional research may help clarify how quality of communication with
parents - so often idiosyncratic during adolescence — impacts risk for offspring who
manage a parent who vacillates between withdrawal and excessive, tangential, or unclear
communication.

Lastly, while direct relationships between composite attachment scores and
affective symptomatology were not found, findings did support the hypothesis that
attachment to the non-diagnosed parent moderated the relationship between attachment
with the diagnosed parent and APS scores. Interestingly, high levels of parental
attachment with the diagnosed parent were associated with lower levels of
symptomatology, and this relationship was found to be more pronounced when strong
attachment bonds with the non-diagnosed parent also existed. This finding is in line with
Bowlby’s theory and the wide body of subsequent research that emphasizes the protective
firnction of strong parental attachment bonds during adolescence.

Parental Attachment Bonas Predictingimial Support Competency

Overall, attachment bonds with the diagnosed parent were stronger predictors of
social support competence than were attachment bonds with the non-diagnosed parent in
this study. Attachment bonds with the diagnosed parent predicted perceived social

support and social self-efficacy as expected, but did not predict fear of negative

77

evaluation. Surprisingly, attachment bonds with the non-diagnosed parent did not predict
any measure of social support competency.

In a closely related study, the mediational role of social skills and relational
competence on the relationship between parental attachment and adolescent’s emotional
adjustment was tested (Engels et. al., 2001). In that study, 12-14 year old participants
were analyzed separately from 15-18 year old participants. For the 12—14 year old age
group, no effects of parental attachment on social skills and relational competence
emerged. However, in the 15-18 year old age group, parental attachment was moderately
related to social skills and this, in turn, affected non-familial relational competence.
These findings suggest that knowledge of the underlying developmental processes is
critical to understanding risk for the development of psychopathology in offspring. In the
current study, age differentiations were not made, which may partially explain why some
of the expected relationships among parental attachment bonds and social competency
measures were not found.

Moreover, while attachment bonds and social competency are linked empirically
and theoretically as reviewed earlier, findings from the current study are also in line with
those which emphasize the multidimensional nature of the social competency construct;
this study was likely affected by the measurement challenges this construct presents
(Engels et. al., 2001; Sarason & Sarason, 1985).

Social Support Competency Predicting AdolescenL Affective Pasachcmathology

As expected, perceived social support, social self-efficacy, and fear of negative

evaluation all predicted adolescent affective symptomatology in this study. These

findings are consistent with previous research demonstrating that higher levels of social

78

competency may reduce risk for affective psychopathology (Stefos, Bauwens, Staner,
Pardoen, & Mendlewics, 1996; Johnson, Winett, Beyer, Greenhouse, & Miller, 1999).
Findings are compatible with the notion that “competency” refers to both a set of
interpersonal relationship skills and experiences and the beliefs [about the self and others]
that prompt one to employ those skills to recruit social support. Furthermore, these
findings specify elements of social support competence that guide the adolescent to
organize his or her social world in such a way that risk for affective disturbance. It is
noteworthy that FNE scores predicted symptomatology but were surprisingly unrelated to
parental attachment bonds in this study. This finding is in line with the developmental
notion that peer relationships become increasingly influential during adolescence, and
their influence, relative to that of parents, may fluctuate according to developmental
transitions. For the adolescent offspring of a bipolar parent, his/her perception of the
negative evaluations of others may impact their affective state, but it may be that their
primary sense of how others are negatively evaluating them is linked more directly to
peer attachment bonds than parental attachment bonds.

Data also lend support to conclusions drawn in the single study reviewed earlier
that directly assessed personal and social resources among offspring of bipolar parents
(Pellegrini et. al., 1986). Findings from that study linked psychiatric well-being with a
high quality of social resources, sense of control, and high self-esteem. While the social
competency construct measured in that study was operationalized differently than in the
present study, conclusions about the importance of social support both within and beyond

the family environment for children affected by parental BPD are analogous.

79

IanLcations for IL’tervention and Prevention

Clinical and preventive approaches that target offspring of parents with bipolar
disorder have been slow to develop relative to advancements in the understanding of risk.
While programs exist for offspring who present for psychiatric treatment, preventive
interventions implemented prior to the onset of offspring illness are rare. Nevertheless,
there are indications that interventions designed to reduce risk prior to the onset of
symptoms are effective (Beardslee, Versage, Velde, Swatling, & Hoke, 2002; Rutter,
1990). Such interventions largely rely upon family-based psychoeducational strategies to
modify risk factors and promote offspring resiliency. Risk factors targeted include
dysfunctional parenting practices, family communication problems, misunderstanding
about the illness, and self-blame and guilt (Beardslee et. al., 1997; Chang & Steiner,
2003). There are indications that programs aimed at developing the adolescents’ skill at
accessing and utilizing social support (Beardslee et. al., 1997; Garmezy & Masten, 1990)
promote resiliency.

Data from the current study may inform clinicians and researchers involved in
intervention and prevention programs in two main ways. First, the unanticipated finding
that adolescents were more closely attached to the diagnosed parent than to the non-
diagnosed parent has implications for family-based interventions. Clinicians should be
cautious of making assumptions that diagnostic classification alone implies dysfirnctional
attachment. Rather, clinicians and researchers should closely examine the specific
qualities of the parent-child relationships within the family (e.g., degree of mutual

understanding and respect, interpersonal connectedness) that may serve to buffer the

80

negative effects of the illness. This finding, coupled with the finding that a high level of
attachment with each parent is associated with lower levels of symptomatology supports
the notion that family interventions aimed at strengthening parent-child and marital
relationships are worthwhile. Failure to take into account the potentially protective
function of these relationships and incorporate this knowledge into treatment and
preventive interventions would hinder progress in this area.

Second, findings that link social‘competency to affective symptomatology provide
useful information on the specific qualities that clinicians and other professionals could
target when designing psychosocial interventions. That is, results suggest that it would
be beneficial to focus on the adolescent’s ability to accurately perceive support when it is
available, to effectively utilize relational support, and to develop optimistic attitudes and
realistic appraisals of other’s opinions about the self.

It is noteworthy that the number of individual counseling sessions the adolescent
reported in this study was significantly and negatively associated with adolescent
affective symptomatology. Participation in sixteen or more individual counseling
sessions was associated with lower symptomatology than participation in five or fewer
sessions. A literature review revealed no studies that address the role that length of
counseling involvement plays for this specific population. Thus, it appears that data from
this study may provide a starting point for future investigations of the benefit of long-
term counseling versus more short-term participation for offspring of affectively-
diagnosed parents.

Lastly, the children in the current study appear to represent a healthier subset of

at-risk offspring than was expected given statistics on the prevalence of disturbance

81

among offspring of affectively disordered parents. It has been noted that recognizing
resilient qualities may serve as one fruitful model for the development of preventive-
intervention programs (Beardslee & Wheelock, 1991). Data from the current study could
be used to inform such future research.

Limitatipns

This study contains several noteworthy limitations. First, interpretations are
limited because of the current impossibility of statistically controlling for genetic
heritability; in this study risk was assumed because the child is both biologically related
to the diagnosed parent and grew up in an environment where parental BPD existed. At
this time there is no known way of establishing whether an offspring carries a specific
genetic predisposition (e. g., carries a gene) for affective disturbance. Research is
underway to isolate genetic material; until such evidence becomes available, current
researchers in this area emphasize psychosocial factors when establishing risk (Hodgins,
F aucher, Zarac, & Ellenbogen, 2002). Similarly, factors related to parental BPD, such as
the timing and severity of episodes, treatment success, and life events undoubtedly
impact the effect of the illness on offspring, as do numerous offspring characteristics like
temperament, negative life events, and personality variables.

Sample and instrument problems exist as well. Findings are based upon self-
report instruments, as has most of the research of this type. Assessments of how people
see themselves may not be as beneficial as methodologies without self-interpretation
(e. g., experimental models, objective accounts). However, regarding the assessment of
parental attachment bonds, it has also been noted that the use of a self-report instrument,

rather than an observational procedure, could tap not only behavioral elements of

82

adolescents’ support seeking, but also the cognitive expectations of attachment figures
(Bretherton, 1985). The use of self-report in the current study reflects the view that
attachment represents aspects of a relationship from the point of view of one person in
the dyad, in this case, the adolescent (Hinde, 1982).

There could also be respondent bias in terms of those who choose to participate in
the study and those who returned questionnaires. For example, depressed persons have a
lack of energy and often avoid social contact, which may make them less likely
respondents. Thus, the sample may be biased with an overrepresentation of individuals
who experience fewer depressive symptoms. Further, it has been established that during
periods of mania, in contrast to periods of depression, individuals exhibit less insight into
their own behaviors. Hence, interpretations of results do not preclude the possibility that
some of the responses to questionnaires may have been a consequence of the adolescent’s
psychopathology. 1

Research also suggests that certain qualities of the parent’s disorder impact
offspring risk (e. g., age of onset and diagnosis, required hospitalizations, medication) by
way of their effect on parenting (Sameroff, 1987). Contradictory evidence suggests that
clinical variables do not necessarily significantly predict outcomes for the diagnosed
parent or offspring (Stefos, Staner, Pardoen, & Mendlewicz, 1996), and by extension,
parenting. Findings fi'om the present study are more in line with the latter argument. That
is, clinical variables associated with the parent’s illness were not associated with
adolescent affective symptoms. However, the effect of these illness characteristics on
attachment relationships with either parent is unclear in the current study. For example,

while parents reported information on the amount of medication they received, the scope

83

of this study does not allow for a determination of the effect of that medication on the
illness or on the quality of the parent-child attachment relationship.

Additionally, there are indications that when a family is selected for inclusion in a
research study of this kind where one parent has an affective disorder, there are likely to
be other disorders that will have an impact on the child and on family functioning
(Beardslee & Wheelock, 1997). Factors such as comorbid disorders in the diagnosed
parent, marital discord, or disorders in the non-diagnosed parent were not controlled for
and could have affected results in this study.

Recommendations for Future Researc_h

The findings of unexpected relationships in this study and the design limitations
noted that may explain these findings provide direction for future research. That is, while
the need for process-oriented models that elucidate mechanisms of risk for offspring
remains strong, the complexity of factors influencing offspring risk makes testing such
path models difficult at this time. Simplified models that investigate dyadic mechanisms
that reduce or elevate risk would be beneficial in the effort to develop more elaborate
path models. Nevertheless, findings from the current study do provide some specific
direction for research on this topic.

The finding that attachment bonds with the diagnosed parent were stronger than
those with the non—diagnosed parent warrants additional study. Specifically, determining
the prevalence of attachment classifications (e.g., secure versus insecure) among these
groups would enable researchers to use established attachment theory principles to
interpret the underlying processes at work. Classifications would also permit

comparisons among data and would firrther equip clinicians to develop theoretically and

84

empirically-based interventions given significant advances made recently in the
application of attachment theory principles (Cassidy & Shaver, 1999; Kobak, 1999;
Liddle & Schwartz, 2002). Furthermore, Bowlby’ s idea that securely attached
individuals are more likely to offer support when they sense the attachment figure needs
it could be tested and applied to preventive-interventions if attachment classifications
were available.

Indexes of parental attachment bonds and adolescent symptomatology in the
current study indicate that researchers should examine the extent to which or level at
which a diagnosed parent engages in the negative risk behavior thought to characterize
risk and not merely whether the parent is depressed or manic. Findings fi'om the current
study are in line with those which advocate for multidimensional indicators of risk
(Goodman & Gotlib, 1999; Rutter, 1987; Sameroff et al., 1987).

Findings from the current study also suggest that longitudinal, developmentally
informed studies are needed. The current study grouped together children whose ages
vary widely which likely masked important developmental influences on the relationships
tested. Conclusions drawn from studies of children in one developmental stage cannot be
presumed to generalize to another period in children’s lives. Findings from the current
study should alert researchers to the importance of attending to the developmental issues
of various ages and avoid grouping children or adolescents in this way.

The design of the current study did allow for various levels of symptomatology
among the adolescents to be represented. This design was motivated by the fact that a
significant number of children manifest resiliency in the face of the stress of having a

parent diagnosed with bipolar disorder. These offspring provide examples of how to

85

survive stressful relational circumstances. It is as important to be able to understand and
recognize risk and resiliency as it is to recognize pathology. This idea is in line with
researchers who argue that it is important to understand that non-ill, resilient children
exist in samples of high-risk children and to be able to characterize and understand their
resiliency is as important as it is to recognize impairment (Richmond & Beardslee, 1988).
Finally, although path-analytic techniques cannot rule out alternative causal

explanations or reciprocal effects, testing mediating effects offers potential for
elucidating important information about links between the model variables. Research
aimed at elucidating the adaptive function of attachment bonds as a means of facilitating
the development of social competence and preventing affective psychopathology are

crucial to diminishing offspring risk.

86

APPENDIX A
RECRUITING NOTIFICATION

CALL FOR RESEARCH PARTICIPANTS

Description of the study: This study investigates resiliency in adolescent offspring of
parents diagnosed with bipolar disorder.

Participants needed: Parents and their adolescent(s)
Requirements for participation:

1) You must be the biological parent of an
adolescent (age 12-17) and you must have been
diagnosed with bipolar disorder at some time
during the adolescent’s lifetime,

0R

2) You must be the spouse of a parent who has been
diagnosed with bipolar disorder and you must
have lived with the adolescent for the majority of
the adolescent’s lifetime. (The diagnosed parent
must be the biological parent of the adolescent.)

If you would like more information on this study or you might be interested
in participating, please contact the principal investigator, Tracy Simko, at
7 7 3-665-0003 or by email at tracys@ix.netcom.com.

87

APPENDIX B
REMINDER CARD FOR MATERIALS NOT RECEIVED

Date

 

Dear Research Participant,

Several weeks ago you consented to participate in a research study on parental bipolar
disorder. Your materials have not yet been received. If you have already returned the
materials, please disregard this notice. Ifyou have not yet returned the materials and
would like assistance or more time to complete them, please contact the principal
investigator, Tracy Simko, at 773-665-0003. Your participation in the study is valuable;
it will contribute to the understanding of bipolar disorder and positive parent-child

relationships. Thank you very much for your time.

Tracy Simko, M.A.
Principal Investigator
773-665-0003
2123 N. Magnolia Ave.
Chicago, IL 60614
tracys@ix.netcom.com

88

APPENDIX C
PARENT DEMOGRAPHIC QUESTIONNAIRE

Code

Was the adolescent adopted? Yes No
What is your relationship to the adolescent?
Annual household income (optional): < $20,000
$20,000 - $40,000
$40,000 - $60,000
$60,000 - $80,000
$80,000 — $100, 000
100,000 +

 

How many others in the adolescent’s family (immediate or extended) have been

diagnosed with an affective disorder (e. g., depression, bipolar disorder)?

 

The following questions refer to the diagnosed parent:
Date of birth

Highest level of education received:

 

 

Diagnostic/treatment history:
Age of onset
Age of diagnosis
Approximate number of affective episodes in lifetime
Treatment history:
Number of hospitalizations
0 __
1-3
4 + _

89

Parent Demographic Questionnaire pg 2

Medication
Have you taken medication for bipolar disorder? Y
N
If yes,

Antidepressant _
Dates

Antianxiety _
Dates

Antipsychotic __

Dates
Adolescent diagnostic/treatment history:
Age of onset
Age of diagnosis
Approximate number of affective episodes in lifetime
Treatment history:
Number of hospitalizations
0 __
1-3 _
4 + __
Medication
My child has taken medication for an affective
disorder _

Antidepressant __
Dates

Antianxiety __
Dates

Antipsychotic _

Dates
My child has not taken medication for an affective disorder _

90

l
I _ ,,.__ .._.._ a u" a, Ari

APPENDIX D
ADOLESCENT DEMOGRAPHIC QUESTIONNAIRE

 

 

Code

Date of Birth: Sex: Male Female
Current Grade Level:
Ethnicity: Afiican—American

American Indian

Asian

Caucasian

Hispanic

Other, please specific

 

Have you ever received any of the following counseling services?

Individual Yes No

If Yes, approximately how many?
1-5 6-1 5 16-30 3 0+

Group Yes No

If Yes, approximately how many?
1-5 6-15 16-30 30+

Family Yes No

If Yes, approximately how many?
1-5 6-15 16-30 30+

Have you ever been hospitalized for psychiatric/psychological services? Yes No

if Yes, approximately how many times?
1-2 3-10 10+

Have you ever taken psychotropic medication? Yes No

If Yes,
Antidepressant __
Dates

Antianxiety _
Dates

Antipsychotic _
Dates

91

 

Adolescent Demographic Questionnaire pg. 2

Have you been arrested for a misdemeanor crime? Yes No
Have you been arrested for a felony crime? Yes No
Have you used illicit drugs? Yes No

92

Parent Demographic

93

Parent Demographic Page 2

94

_--.......o .i-'.h
. u

Adol Demographic

95

 

Adol Demo page 2

96

aps

97

 

APS

98

APS

99

h I..“ .l-.‘ ill .t-n. 1‘"

IPPA

 

100

IPPA

101

PSS

102

PSS

 

103

 

S
S
P

 

104

SSE

105

 

106

REFERENCES

 

Aiken, L. & West, S.G. (1991 ). Multiple regression: Testing and interpreting
interactions. Thousand Oaks, CA: Sage Publications, Inc.

Ainsworth, M.D., Blehar, M.C., Waters, E., & Wall, 8. (1978). Patterns of
attachment: A psychological study of the strange situation. Hillsdale, NJ: Erlbaum.

Ainsworth, M. S. (1989). Attachments beyond infancy. American Psychologist.
44, 709-716.

Allen, J. P. & Land, D. Attachment in adolescence. (1999). In J. Cassidy & P.R.
Shaver (Eds), Hancfiook of attachment: Theory, research, and clinical applications 07p.
319-335). New York: Guilford Press.

Allen, J., Hauser, S. & Bonnan-Spurrell, E. (1996). Attachment theory as a
framework for understanding sequelae of severe adolescent psychopathology: An 11 year
follow-up study. Journal of Consulting & Clinical Psycholog, 64, 254-263.

Anthony, El, (1975). The influence of a manic-depressive environment on the
developing child. In B]. Anthony & T. Benedek (Eds), Depression and human
existence. Boston: Little, Brown.

Armsden, G.C., & Greenberg, MT. (1987). The Inventory of Parent and Peer
Attachment: Individual differences and their relationship to psychological well-being in
adolescence. Joumal of Youth and Adolescence, 16, 427-454.

Armsden, G.C., McCauley, E., Greenberg, M.T., Burke, P.M., & Mitchell, J .R.,
(1990). Parent and peer attachment in early adolescent depression. Journal of Abnormal
Child Psychology, 18(6), 683-697.

Armsden, G.G. (1986). Attachment to parents and peers in late adolescence:
Relationships to affective status, self-esteem and coping with loss, threat and challenge.
Dissertation Abstracts International, 4 7(4-B), 1751-1752

Armsden, G., McCauley, E., Greenberg, M., Burke, P., (1990). Parent and peer
attachment in early adolescent depression. Journal of A bnormal Child Psychology, 18(6),
683-697.

Baron, RM., & Kenny, DA. (1986). The moderator-mediator variable
distinction in social psychological research: Conceptual, strategic, and statistical
considerations. Journal of Personality and Social Psychology, 51(6), 1 173-1 182.

 

107

 

~‘ A.._v.'

Beardslee, W.R., Versage, E.M., & Gladstone, TR. (1998). Children of
affectively ill parents: A review of the past 10 years. Journal of the American Academy
of Child and Adolescent Psychiatry, 3 7(1 1), 1 134-1141.

1‘“:

Beardslee, W. R. & Podorefsky, D. (1988). Resilient adolescents whose parents
have serious affective and other psychiatric disorders: Importance of self-understanding
and relationships. American Journal of Psychiatry, 145(1), 63-69.

Beardslee, W., Wright, W., Salt, P., & Drezner, K. (1997). Examination of
children’s responses to two preventive intervention strategies over time. Journal of the
American Academy of Child & Adolescent Psychiatry, 36(2), 196-204.

Beardslee, W. & Wheetlock, I. (1994). Children of parents with affective
disorders: Empirical findings and clinical implications. In W. Reynolds & H.F. Johnston
(Eds). Handbook of depression in children cmd adolescents. Issues in clinical child
psychology, (pp. 463-479).

Beck, A. (1967). Depression: Clinical, experimental, and theoretical aspects.
New York: Harper & Row.

Blain, M.D., Thompson, J.M., & Whiffen, VB. (1993). Attachment and
perceived social support in late adolescence: The interaction between working models of
self and others. Journal of Adolescent Research, 8(2), 226-241.

Bowlby, J. (1969). Attachment and loss: Vol 1. Attachment, New York: Basic
Books.

Bowlby, J. (1973). Attachment and loss: Vol 2. Separation, New York: Basic
Books.

Bowlby, J (1980). Attachment and loss: Vol 3. Loss: Sadness and depression.
New York: Basic Books.

Bowlby, J. (1982). Attachment and loss: Vol. 1. Attachment, (2nd ed.). New
York: Basic Books.

Bowlby, J. (198 8). A secure base: Parent-child attachment and healthy human
development. New York: Basic Books.

Bretherton, E. (1985). Attachment theory: Retrospect and prospect.
Monographs of the Society for Research in Child Development, 5 0(1-2), 3-35.

Carlson, E. A, & Sroufe, L. (1995). Contribution of attachment theory to
developmental psychopathology. In D. Cicchetti & D. Cohen (Eds), Developmental
psychopathology, Vol. 1: Yheory and methods. Wiley series on personality processes. (pp.
581-617). New Yorlg NY,: John Wiley & Sons

108

Cassidy, J. (1994). Emotion regulation: Influences on attachment relationships.
In N. Fox (Ed), The development of emotion regulation: Biological and behavioral
considerations. Monographs for the Society for Research in Child Development, 59, (l -
2), 228-249.

Cassidy, J., & Shaver, P. R. (1999). Handbook of attachment: Theory, research,
and clinical applications. Hillsdale, NJ: Erlbaum.

Cassidy, J. & Kobak, R. R. (1988). Avoidance and its relation to other defensive
processes. In J. Belsky, & T. Neworski (Eds), Clinical implications of attachment.
Hillsdale, NJ: Erlbaum.

Chang, K., Steiner, PL, Dienes, K., Adleman, N., & Ketter, T. (2003). Bipolar
offspring: A window into bipolar disorder evolution. Biological Psychiatry, 53(11), 945-
951.

Coble, H.M., Gantt, D., & Mallinckrodt, B. (1996). Attachment and social
competencies. In GR Pierce et al, (Eds), Handbook of social support and the family
(pp. 141-172). New York: Plenum.

Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155-159.

Connolly, J .A. (1989). Social self efficacy in adolescence: Relations with self
concept, social adjustment and mental health. Canadian Journal of Behavioral Science,
21, 259-269.

Cummings, E.M. & Cicchetti, D. (1990). Toward a transactional model of
relations between attachment and depression. In M.T. Greenberg, D. Cicchetti, & E.M.
Cummings (Eds), Attachment in the preschool years: Theory, research, and intervention
(pp. 339-3 72), University of Chicago Press.

Cutrona, CE. (1982). Transition to college: Loneliness and the process of social
adjustment. In LA. Peplau & D. Perlman (Eds), Loneliness: A sourcebook of current
theory, research, and therapy (pp. 291-3 09). New York: Wiley.

Cutrona, C.E., & Russel, D.W. (1987). The provisions of social relationships and
adaptation to stress. In W.H. Hones & D. Perlman (Eds), Advances in personal
relationships (pp. 37-67). Greenwich, CT: Jai Press.

Cytryn, L., McKnew, D.H., Bartko, J.J., Lamour, M., & Hamovit, J. (1982).
Offspring of parents with affective disorders. H. Journal of the American Academy of
Child Psychiatry, 21, 389-391.

Davenport, Y.B., Adland, M.L., Gold, P.W., & Goodwin, F .K. (1979). Manic-

109

 

 

depressive illness: Psychodynamic features of multigenerational families. American
Journal of Orthopsychiatry, 49, 24-35.

Davenport, Y.B., Zahn-Waxler, C., Adland, M.L., & Mayfield, A. (1984). Early
child-rearing practices in families with a manic-depressive parent. American Journal of
Psychiatry, 141, 230-235.

Downey, G., & Coyne, J .C. (1990). Children of depressed parents: An
integrative review. Psychological Bulletin, 108(1), 50-76.

Dozier, M., & Kobak, R., (1993). Psychophysiology in attachment interviews:
Converging evidence for deactivating strategies. Annual Progress in Child Psychiatry
and Child Development, 80-92.

Dozier, M., Stovall, K.C., & Albus, KB. (1999). Attachment and

psychopathology in adulthood, In J. Cassidy & P.A. Shaver (Eds), Hancfiook of
Attachment (pp. 497-519).

Ehrenberg, M., & Cox., A. (1991). The relationship between self efficacy and
adolescents. Adolescence, 102, 361-374.

Engels, R., Finkenauer, C., Meeus, W., & Dekovic, M. (2001). Parental
attachment and adolescents’ emotional adjustment: The associations with social skills and
relational competence. Journal of Counseling Psychology, 48(4), 428-439.

Erickson, M. F., Sroufe, L. A., Egeland, B., & Byron, T. (1985). The relationship
between quality of attachment and behavior problems in preschool in a high-risk sample.
Monographs of the Society for Research in Child Development, 50(1-2),147-166

Fisher, L., Kokes, R.F., Cole, R.E., Perkins, P. M., & Wynne, LC. (1987).
Competent children at risk: A study of well-functioning offspring of disturbed parents. In
A.E. James, E. Cohler & J. Bertram (Eds), The invulnerable child. The Guilford
psychiatry series. (pp. 211-228). NY: Guilford Press.

Ford, ME. (1982). Social cognition and social competency in adolescence.
Developmental Psychology, 18, 323-341.

Furnham, A., & Argyle, M. (1981). Responses of four groups to difficult social
situations. In M. Argyle, A. F umham, & J. Graham (Eds), Social situations. (pp.336-
344). Cambridge University Press.

Gaensbauer, T.J., Harmon, R.J., Cytryn, L., & McKnew, DH. (1984). Social and

affective development in infants with a manic-depressive parent. American Journal of
Psychiatry, 141, 223-229.

110

 

 

 

Garmezy, N. (1989). The role of competence in the study of children and
adolescents under stress. In B. Schneider & G. Attili, (Eds), Social competence in
developmental perspective. NATO Advanced Science Institutes series. Series D:
Behavioural and social sciences, 51, (pp. 25-39). Boston: Kluwer Academic Publishers.

Garmezy, N., & Masten, A. (1990). The adaptation of children to a stressful
world: Mastery of fear. In E.L. Arnold (Ed), Chilflrood stress. Wiley series in child and
adolescent mental health, (pp. 460-47 3). New York: Wiley.

Garmezy, N. & Masten, A. (1991). The protective role of competence indicators
in children at risk. In M. E. Cummings and A. L. Greene (Eds), Life-span developmental
psychology: Perspectives on stress and coping, 15, 151-174. New York: Wiley.

Gershon, E., McKnew, D., Cytryn, L., Harnmovit, J ., Schreiber, J ., Hibbs, E., &
Pellegrini, D. (1985). Diagnosis in school-age children of bipolar affective disorder
parents and normal controls. Journal of Aflective Disorders, 8(3), 283-291.

Goodman, S.H., & Gotlib, 1.]. (1999). Risk for psychopathology in children of
depressed mothers. Psychological Review, 106, 45 8-490.

Greenberg, M.T., Seigal, J .M., & Leitch, CJ. (1983). The nature and importance
of attachment relationships to parents and peers during adolescence. Journal of Youth
and Adolescence, 12, 373-386.

Grigoroiu-Serbanescu, M., Christodorescu, D., Jipescu, I., Totoescu, A.,
Marinescu, E., & Ardelean, V. (1989). Psychopathology in children aged 10-17 of
bipolar parents: Psychopathology rate and correlates of the severity of the
psychopathology, Journal of Affective Disorders, 16, 167-179.

Hammen, C., Burge, D., Burney, E., & Adrian, C. (1990). Longitudinal study of
diagnosis in children of women with unipolar and bipolar affective disorders. Archives of
General Psychiatry, 47, 1112-1 117.

Hodgins, S., Faucher, B., Zarac, A., & Ellenbogen, M. (2002). Children of
parents with bipolar disorder: A population at high risk for major affective disorders.
Child and Adolescent Psychiatric Clinics of North America, 11(3), 533-554.

Izard, C. E, Haynes, O., Chisholm, G., & Baak, K. (1991). Emotional
determinants of infant-mother attachment. Child Development, 62(5), 906-917.

Johnson, S.L., Winett, C.A., Meyer, B., Greenhouse, W.J., & Miller, I. (1999).
Social support and the course of bipolar disorder. Journal of A bnormal Psychology, 108
(4), 558-566.

Kenny, M. (1987). The extent and firnction of parental attachment among first-
year college students. Journal of Youth and Adolescence, 16, 17-29.

111

 

I“.
r
l
l:

 

Kenny, M., Moilanen, D., Lomax, R, & Brabeck, M. (1993). Contributions of
parental attachments to view of self and depressive symptoms among early adolescents.
Journal of Early Adolescence, 13(4), 408-430.

Kobak, R (1999). The emotional dynamics of disruptions in attachment
relationships: Implications for theory, research, and clinical intervention, In J. Cassidy &
P. Shaver (Eds), Handbook of attachment: Theory, research, and clinical applications,

(up. 21.43)

Kobak, R R, & Sceery, A. (1988). Attachment in late adolescence: Working
models, affect regulation, and representations of self and others Child Development, 59,
135-146.

Kobak, R, Sudler, N., & Gamble, W. (1991). Attachment and depressive
symptoms during adolescence: A developmental pathways analysis. Development &
Psychopathology. Vol 3(4), 461-474.

Leary, M.R (1983). A brief version of the Fear of Negative Evaluation Scale.
Personality and Social Psychology Bulletin, 9, 371 -375.

Lafley, HP. (1989). Family burden and family stigma in major mental illness.
American Psychologist, 44(3 ), 556-560.

Lewinsohn, R, Klein, D., & Seeley, J. (1995). Bipolar disorders in a community
sample of older adolescents: prevelance, phenomenology, comorbidity, and course.
Journal of American Academic of Child Adolescent Psychiatry, 34, 454-463.

Lewis, F.M., Woods, N.F., & Ellison, B. (1983-1986). The Family Impact Study:
The impact of cancer on the family. Grant R01-NU-01000, University of Washington,
Division of Nursing, Public Health Service.

Liddle, H., & Schwartz, S. (2002). Attachment and family therapy: The clinical
utility of adolescent-family attachment research, Family Process. 41 (3 ), 455-476.

Main, M. (1990). Cross-cultural studies of attachment organization: Recent
studies, changing methodologies, and the concept of conditional strategies. Human
Development, 33(1), 48-61.

Mallinckrodt, B. (1992). Childhood emotional bonds with parents, development
of adult social competencies, and availability of social support. Journal of Counseling
Psychology, 39(4), 453-461.

Matas, L., Arend, R., & Sroufe, A. (1978). Continuity of adaptation in the

second year: The relationship between quality of attachment and later competence. Child
Development. 49(3), 547-556.

112

“A. D‘A'r- ' /

Mayo, J.A., O'Connell, R.A.., & O'Brien, J .D. (1979). Families of manic-
depressive patients: effect of treatment. American Journal of Psychiatry, 136(12), 153 5-
1539.

McKnew, D.H., Cytryn, L., Efron, A.M., Gershon, E.S., & Bunney, W.E. (1979).
Offspring of patients with affective disorders. British Journal of Psychiatry, 134, 148-
1 52.

Murray, C.J.L., & Lopez, AD. (1996) The global burden of disease: A
comprehensive assessment of mortality and disability fiom diseases, injuries, and risk
factors in 1990 and projected to 2020. Boston: Harvard University Press.

O’Connell, R.A., Mayo, J .A., Eng, L.K., Jones, J .S., & Gabel, RM. (1985).
Social support and long-term lithium outcome. British Journal of Psychiatry, 14 7, 272-
275.

O’Connor, L., Berry, J ., Weiss, J ., & Gilbert, P. (2002). Guilt, fear, submission,
and empathy in depression. Journal of Aflective Disorders, 71(1-3), 19-27.

Oestman, M, & Kjellin, L. (2002). Stigma by association: Psychological factors
in relatives of people with mental illness. British Journal of Psychiatry, 181(6), 494-498.

Papini, D., & Roggrnan, L.. (1992). Adolescent perceived attachment to parents
in relation to competence, depression, and anxiety: A longitudinal study. Journal of Early
Adolescence, 12(4), 420-440.

Parkes, C. M., & Stevenson-Hinde, J. (1982). The place of attachment in human
behavior. New York: Basic Books.

Patterson, J., Field, J ., and Pryor, J. (1994). Adolescents’ perceptions of their
attachment relationships with their mothers, fathers, and fiiends Journal of Youth and
Adolescence, 23, 579-600.

Pellegrini, D., Kosisky, S., Nackman, D., Cytryn, L., McKnew, D.H., Gershon,
E., Hamovit, J., & Cammuso, K. (1986). Personal and social resources in children of
patients with bipolar affective disorder and children of normal control subjects. Annual
Progress in Child Psychiatry & Child Development, 494-507.

Pianta , R, Egeland, B., & Sroufe, L. (1990). Maternal stress in children’s
development: Predictions of school outcomes and identification of protective factors. In
J Rolf, A Masten, D. Cicchetti, K. Neuchterlen, & S. Weintraub (Eds), Risk and
protective factors in the development of psychopathology (pp. 215-23 5). New York:
Cambridge University Press.

113

Pipp, S., Shaver, P., Jennings, S., Lambom, S., & Fischer, K. (1985).
Adolescents’ theories about the development of their relationships with parents. Journal
of Personal & Social Psychology, 48, 991-10001.

Pound, A. (1996). Parental affective disorder and childhood disturbance. In M.
Goepfert & J. Webster (Eds), Parental psychiatric disorder: Distressed parents and their
families. (pp. 201-218). Cambridge, England UK: Cambridge University Press.

Prince, 8., & Jacobson, N. (1995). A review and evaluation of marital and family
therapies for affective disorders. Journal of Marital Family Therapy, 21, 377-401.

Procidano, M.E., & Heller, K. (1983). Measures of perceived social support from
fiiends and family: Three validation studies. American Journal of Community
Psychology, 11, 1-23.

Radke-Yarrow, M., Cummings, E.M., Kucynski, L., & Chapman, M. (1985).
Patterns of attachment in two- and three-year olds in normal families and families with
parental depression. Child Development, 56, 884-889.

Reynolds, W.M. (1998). Adolescent Psychopathology Scale: Psychometric and
technical manual. Odessa, FL: Psychological Assessment Resources.

Reynolds, W. M. (1993). Self-report methodology. In T. Ollendick & M. Ersen
(Eds). Handbook of child and adolescent assessment: General psychology series, 1 6 7
(pp. 98-123). Boston: Allyn & Bacon, Inc.

Reynolds, W.M. (1992). Introduction to the nature and study of internalizing
disorders in children and adolescents. In W.M. Reynolds (Ed), Internalizing disorders in
children and adolescents (pp. 1-18). Wiley Publications: US.

Rice, K.G., Cunningham, T.I., & Mitchell, B.Y. (1997). Attachment to parents,
social competence, and emotional well-being: A comparison of Black and White late
adolescents. Journal of Counseling Psychology, 44(1), 89-101.

Richmond, J ., & Beardslee, W. (1988). Resiliency: Research and practical
implications for pediatricians. Journal of Developmental & Behavioral Pediatrics, 9(3),
1 5 7-163 .

Rosenstein, D.S., & Horowitz, HA. (1996). Adolescent attachment and
psychopathology. Journal of Consulting and Clinical Psychology, 64(2), 244-253.

Rush, A.J., Cain, J.W., Raese, J., Stewart, RS., Waller, D.A., & Debus, JD.

(1991). Neurobiological bases for psychiatric disorders. In RN. Rosenberg (Ed),
Comprehensive neurology (pp. 555-603.) New York: Raven Press.

114

 

Russell, D., Cutrona, E., & Yurko, K., (1984). Social and emotional loneliness:
an examination of Weiss’s typology of loneliness. Journal of Personality and Social
Psychology, 46(6), 1313-1321.

Rutter, M. (1990). Some focus and process considerations regarding effects of
parental depression on children. Developmental Psychology, 26, 60-67.

Rutter, M. (1987). Psychosocial resilience and protective mechanisms.
American Journal of Orthopsychiatry, 5 7, 316-331.

Safran, J. (1990). Towards a refinement of cognitive therapy in light of
interpersonal theory: 1. Theory. Clinical Psychology Review, 10, 87-106.

Sarason, 1., Pierce, G., & Sarason, B. (1990). Social support and interactional
processes: A triadic hypothesis. Journal of Social & Personal Relationships. 7, 495-506.

Sarason, I; Sarason, B. Shearin, E. (1986). Social support as an individual
difference variable: Its stability, origins, and relational aspects. Journal of Personality &
Social Psychology. 50, 845-855.

Sarason, B.R., Pierce, GR, Shearin, E.N., Sarason, I.G., Waltz, J., & Poppe, L.
(1990). Perceived social support and working models of self and actual others. Journal
of Personality and Social Psychology, 60, 273-287.

Seifer, R, Sameroff, A, Dickstein, S., Keitner, G., & Miller, I. (1996). Parental
psychopathology, multiple contextual risks, and one-year outcomes in children. Journal
of Clinical Child Psychology, 25(4), 423 -43 5.

Solantaus-Sirnula, T., Punamaeki, R, & Beardslee, W. (2002). Children’s
responses to low parental mood. Balancing between active empathy, overinvolvement,
indifference and avoidance. Journal of the American Academy of Child & Adolescent
Psychiatry, 41(3), 2787-286.

Sroufe, A., Schork, E., Motti, F., Lawroski, N., & LaFreniere, P. (1985). The
role of affect in social competence. In C. Izard & J. Kagan (Eds), Emotions, cognition,
and behavior, (pp. 289-319).

Sroufe, LA. (1983). Infant-caregiver attachment and patterns of adaptation in
preschool: The roots of maladaptation and competence. In M. Perlmutter (Ed),
Minnesota symposium on child psychology, 16, (pp. 41-83). Hillsdale, NJ: Erlbaum.

Sroufe, L., Levy, A.K., & Carlson, E. (1998). Resilience in adolescence:

Changing the odds. Paper presented at the biennial meeting of the Society for Research
on Adolescence, San Diego, CA.

115

 

Stefos, G., Bauwens, F., Staner, L., Pardoen, D., & Mendlewics, J. (1996).
Psychosocial predictors of major affective recurrences in bipolar disorder: a 4-year
longitudinal study of patients on prophylactic treatment. Acta Psychiatrica
Scandinaviaca, 93, 420-426.

 

Todd, R D., Reich, W., Petti, T.A., Joshi, P., DePaulo, J .R., Numberger, J., &
Reich, T. (1996). Psychiatric diagnoses in the child and adolescent members of extended
families identified through adult bipolar affective disorder probands. Journal of
American Academy of Child and Adolescent Psychiatry, 35(5), 664-671.

Uchino, B.N., Cacioppo, J .T., & Kiecolt-Glaser, J. (1996). The relationship
between social support and physiological processes: A review with emphasis on
underlying mechanisms and implications for health. Psychological Bulletin, 119, 488-
531.

Vertue, RN. (2003). From adaptive emotion to dysfirnction: An attachment
perspective on social anxiety disorder. Personality and Social Psychology Review, 7(2),
170-191.

Watson, D., & Friend, R (1969). Measurement of social-evaluative anxiety.
Journal of Consulting and Clinical Psychology, 33, 448-45 7.

Werner, E., & Smith, R. (1980). An epidemiologic perspective on some
antecendents and consequences of childhood mental health problems and learning
disabilities. Annual Progress in Child Psychiatry & Child Development, 133-147.

Weiss, RS. (1982). Attachment in adult life. C.M. Parkes & J. Stevenson-Hinde
(Eds), The place of attachment in human behavior, New York: Basic Books.

Weiss, RS. (1974). The provisions of social relationships. In Z. Rubin (Ed),
Doing unto others (pp. 17-26). Englewood Cliffs, NJ: Printice-Hall.

Weissman, M., Leaf, P., & Tischler, G. (1988). Affective disorders in five
United States communities. Psychological Medicine, 18, 141-153.

Winemiller, D.R., Mitchell, M., Sutliff, J ., & Cline, DJ. (1993). Measurement
strategies in social support: A descriptive review of the literature. Journal of Clinical
Psychology, 49(5), 638-648.

Zahn-Waxler, C., Cummings, E.M., McKnew, D.H., & Radke—Yarrow, M.

(1984). Altruism, aggression, and social interaction in young children with a manic-
depressive parent. Child Development, 55, 112-122.

116

   

lllllllllllllrilll