University of Southern California Dissertations and Theses

The impact of post-traumatic stress and post-traumatic growth on young adult cancer survivors

(USC Thesis Other)

The impact of post-traumatic stress and post-traumatic growth on young adult cancer survivors

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

THE IMPACT OF POST-TRAUMATIC STRESS AND POST-TRAUMATIC
GROWTH ON YOUNG ADULT CANCER SURVIVORS

by

Jaehee Yi

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

May 2011

Copyright 2011 Jaehee Yi

ii

DEDICATION
This dissertation is dedicated to cancer patients, survivors, and their family members who inspire
me with their great human power.

iii

ACKNOWLEDGEMENTS
My personal interest in the power of human beings to bounce back and grow when facing
life's adversities may have started in my childhood when I witnessed how strong my poor family
and neighbors were in confronting their life difficulties and trying to improve their lives and their
children’s lives. It became my professional interest and positive growth phenomena have always
been an empowering and exciting subject for me, so I thought it would be a topic that would
never tire me throughout the dissertation process. At one moment in the process, however, I faced
strong doubts about anything positive coming out of a terrible experience when my father got
diagnosed with cancer, went through rigorous treatment, and finally died. It was heartbreaking for
me to try to concentrate on examining posttraumatic growth of cancer survivors when he lost
weight at a remarkable speed, could not swallow anything, and suffered extreme pain. Amidst the
agony, I felt like the whole concept of posttraumatic growth was not true.
After witnessing my father’s gracious last wishes and death, and going through my own
changes in many aspects of life afterwards, I myself experienced posttraumatic growth and
realized that the subject that I am trying to understand and write about in this dissertation is
indeed real and precious. Such realization itself is a blessing for my scholarly life because I can
now sincerely value what I do. What I am doing is not only about doing research, publishing
articles, and earning a Ph.D, but also about hearing and telling real people’s true stories.
I dedicate this dissertation to all people who believe in our power to go through and grow
out of the difficulties in life that we may have right now. I thank my dissertation committee, Drs.
Palinkas, Chi, Knight, and Chou, for providing me with great support through the process. I
especially thank Dr. Palinkas for taking on the task of chairing my dissertation committee during
my transition time and giving me his warmest encouragement during the difficult time in my life.
I thank Dr. Brad Zebrack for generously letting me to use his data and continuing to mentor me

iv

long distance. Without my friends who did not complain (much) about my frequent whining, the
writing time would have been less joyful. I say my sincere thank-you to all my friends. My
deepest gratitude is reserved for my family and JY who were concerned more about my health
than my dissertation with their never-changing love for me.
Last but not least, I thank myself for surviving the doctoral program and growing through
the passage of this great life milestone.

v

TABLE OF CONTENTS

DEDICATION ................................................................................................................................. ii
ACKNOWLEDGEMENTS ............................................................................................................ iii
LIST OF TABLES ......................................................................................................................... vii
LIST OF FIGURES ...................................................................................................................... viii
ABSTRACT .................................................................................................................................... ix
CHAPTER 1. INTRODUCTION .................................................................................................... 1
CHAPTER 2. CURVILINEAR RELATIONSHIP .......................................................................... 4
INTRODUCTION ....................................................................................................................... 4
METHODS .................................................................................................................................. 7
Sample and procedure .............................................................................................................. 7
Measures .................................................................................................................................. 9
Analyses ................................................................................................................................. 10
RESULTS .................................................................................................................................. 11
Descriptive analysis ............................................................................................................... 11
Curve estimation analyses ...................................................................................................... 13
DISCUSSION ............................................................................................................................ 17
CHAPTER 3. POST-TRAUMATIC PROFILES .......................................................................... 22
INTRODUCTION ..................................................................................................................... 22
METHODS ................................................................................................................................ 25
Sample and procedure ............................................................................................................ 25
Measures ................................................................................................................................ 27
Analysis ................................................................................................................................. 30
RESULTS .................................................................................................................................. 33
Descriptive statistics .............................................................................................................. 33
Latent profile analysis results ................................................................................................ 35
Characteristics of the survivor profiles .................................................................................. 38
DISCUSSION ............................................................................................................................ 40

vi

CHAPTER 4. BUFFERING ROLES OF POST-TRAUMATIC GROWTH ................................ 46
INTRODUCTION ..................................................................................................................... 46
METHODS ................................................................................................................................ 49
Sample and procedure ............................................................................................................ 49
Measures ................................................................................................................................ 51
Analysis ................................................................................................................................. 53
RESULTS .................................................................................................................................. 57
Descriptive statistics .............................................................................................................. 57
Moderation effects ................................................................................................................. 59
DISCUSSION ............................................................................................................................ 64
CHAPTER 5. CONCLUSION ....................................................................................................... 67
REFERENCES .............................................................................................................................. 74

vii

LIST OF TABLES
Table 1: Participants characteristics 12
Table 2. Curve estimation regression analysis results 15
Table 3: Participants characteristics 34
Table 4: Model fit indexes for the 2-, 3-, 4-, and 5-class solutions 36
Table 5: Overall sample means and means of profiles in the 3-cluster model 37
Table 6: Posttraumatic profile differences in correlates 40
Table 7: Sample characteristics by PTS levels 58
Table 8: Comparison of models 59
Table 9: Coefficients in 3 models 61

viii

LIST OF FIGURES
Figure 1: Graphical presentation of the relationships between PTS and PTG 16
Figure 2: Graphic presentation of Model 3 56
Figure 3: Moderating effects of PTG 63
Figure 4: Conceptual model of quadratic interaction effects of PTG 63

ix

ABSTRACT
This dissertation examines the interrelationships between Post-Traumatic Stress (PTS)
and Post-Traumatic Growth (PTG), the consequences of their coexistence in some survivors, and
the buffering roles of PTG on distress in young adult cancer survivors. Specifically, three
independent papers comprising the dissertation aim (1) to test curvilinear relationships between
PTS and PTG; (2) to classify latent profiles of the survivors based on their PTS and PTG levels
and investigate the characteristics of the survivor profiles; and (3) to test the moderating effect of
PTG on the negative relationship between PTS and psychological distress.
Participants were recruited from four medical centers that treat pediatric oncology
patients in Southern California and Michigan. Inclusion criteria included current age 18-39; age at
diagnosis 0-21 years; and disease-free status. Volunteer participants answered the mailed survey
questionnaires. Of the 618 participants who responded and met the eligibility criteria, 593
survivors who had both PTG and PTS scores were used for this dissertation.
In the first paper, the overall relationships between PTG and PTS were not statistically
significant. The two phenomena seemed to coexist, but independently from each other. In the
second paper using latent profile analyses, three clusters of survivors were found based on PTS
and PTG levels: “Thrivers” who grew the most with their traumatic distress at medium level;
“Sufferers” who suffered the highest level of distress with the level of growth at medium; and
“Recoverers” at the lowest levels in both PTS and PTG. There was no difference among the
clusters in demographic variables such as gender, employment, income, education, and current
age. Thrivers comprised the most recent survivors with the highest level of social support and the
greatest mental health, whereas Sufferers had the worst physical health. In the third paper, PTG
was found to buffer the negative impact of PTS on psychological distress and the moderating
effect was greater as the PTG level increased.

x

The study findings provide important empirical evidence to the complex experiences of
cancer survivorship in young adulthood. The positive roles of PTG should continue to be
examined to explore its mechanism and to develop growth-oriented programs and services for
survivors.

1

CHAPTER 1. INTRODUCTION
Thanks to the remarkable advance in treatment models, about 78% of the children who
are diagnosed with cancer survive. However, survival often comes at a price with complications,
disabilities, or adverse outcomes that are the result of the disease process, treatment, or both
(Institute of Medicine, 2003). Although the lingering impact of cancer on surviving children is
evident, there is little understanding of how long-term survivorship evolves as they grow into the
later life stages and face different challenges. Young adulthood is one of the most turbulent life
stages for anyone, and how childhood cancer survivors experience their young adulthood is far
from known.
Cancer was relatively recently recognized as one of the trauma categories in the research.
In 1994, the Diagnostic Statistical Manual of Mental Disorders, 4
th
edition (American
Psychological Association, 1994) broadened its taxonomy of Post-Traumatic Stress Disorder and
included being “diagnosed with a life-threatening illness or learning that one’s child” has such an
illness as a “traumatic” event. Henceforth, childhood cancer survivorship has been studied from
the theoretical models of post-traumatic experiences.
Given its short history, cancer survivorship research from the perspective of trauma has at
least three research gaps. First, the interrelationships between Post-Traumatic Stress (PTS) and
Post-Traumatic Growth (PTG) are far from known. The two post-trauma phenomena have been
considered as opposite constructs, so the consequences of their relationships were not examined.
Second, the multidimensional nature of cancer survivorship has not been adequately studied. By
focusing either on PTS or on PTG, many studies have failed to capture the reality that some
survivors might have both positive and negative posttraumatic experiences. Third, the roles of

2

PTG were neglected in the research. Because the focus of research has been put more on the
negative impact of cancer, the potential roles of PTG and clinical interventions to promote the
growth phenomena have been inadequately studied.
This dissertation extends the previous research on cancer survivorship in young adulthood by
examining the interrelationships between posttraumatic growth and stress, consequences of these
phenomena on the quality of life among young adult survivors, and the roles of PTG on
alleviating distress. Specifically, this dissertation explores the following research questions in
three distinct papers:
1) Is the relationship between PTS and PTG curvilinear? Does PTG increase as PTS
increases up to a certain point, from which PTG decreases again?
2) What are the latent profiles of survivors based on PTS and PTG? Can survivors be
clustered based on different combinations of PTS and PTG levels? What are the
characteristics of such cluster?
3) Does PTG moderate the relationship between PTS and psychological distress? Does PTG
buffer the negative impact of PTS on psychological distress?
The three papers comprising this dissertation use different theoretical frameworks to examine
each of the questions posed above. This section provides a brief overview of these theoretical
frameworks as well as research questions addressed in each paper.
In Chapter 2 (paper 1), I use PTG theories (R. Tedeschi & Calhoun, 1995) and stress theories
(R. Hill, 1949) to hypothesize curvilinear relationships between PTS and PTG. The more
traumatic distress cancer survivors have, the more growth they are expected to experience up to a
certain level of distress, because one can grow after experiencing an event only when its impact is
severe enough to shake his or her world views (R. Tedeschi & Calhoun, 1995) and transform the

3

person (Parry & Chesler, 2005). Beyond the tipping point, however, more distress might decrease
positive growth experiences among cancer survivors, because extreme distress would hamper
functioning and adaptation according to stress theories (R. Hill, 1949).
Chapter 3 (paper 2) builds upon Chapter 2 by further exploring how PTS and PTG together
happen in young adult cancer survivors. Using the model-based Latent Profile Analysis, this
study classifies the survivors based on the levels of the two constructs and examines what
demographic, medical, socioeconomic, and health characteristics of each cluster.
Chapter 4 (paper 3) tests the moderating effect of PTG in the relationship between PTS and
psychological distress. The moderation hypothesis is based on the theory that cancer survivors
might grow in the process of deliberately reflecting on the trauma events (R. Tedeschi & Calhoun,
1995) and reconstructing the shattered existing assumptions prior to the trauma (R. Janoff-
Bulman, 1992), and the PTG might improve psychological well-being with the reconstructed
hopes and expectations (Calman, 1984).
Finally, Chapter 5 summarizes the main findings. It concludes with study limitations,
implications, and future research suggestions.

4

CHAPTER 2. CURVILINEAR RELATIONSHIP
INTRODUCTION
In 1994, the Diagnostic Statistical Manual of Mental Disorders, 4
th
edition (American
Psychiatric Association, 1994) broadened its taxonomy of Post-Traumatic Stress Disorder
(PTSD), and included being “diagnosed with a life-threatening illness or learning that one’s child”
(American Psychiatric Association, 1994, p. 426) has such an illness as a “traumatic” event.
Cancer diagnosis had been always considered a shocking event to the individual patient and his or
her family. However, the official inclusion of cancer as a possible cause of PTSD marks a
transition in research and practice for cancer-affected people, because it recognizes that the
impact of cancer might continue to linger, even after treatment ends and throughout the
survivorship. Henceforth, cancer survivorship has been studied from the theoretical models of
post-traumatic experiences.
Deficit models have been widely used, focusing on the negative impact of cancer in
survivorship. Cancer survivors are likely to have experienced many stressors related to diagnosis,
prognosis, invasive treatments, side effects, and risk of recurrence (Bruce, 2006). Cancer-specific
PTSD symptoms include the intrusion of unwanted memories, such as nightmares or flashbacks;
the avoidance of reminders of the events, such as doctors or hospital; an increased startle response
to sudden noise; and the constant monitoring for danger (Foa, Cashman, Jaycox, & Perry, 1997).
Current or lifetime rates of cancer-related PTSD range from 2% to 35% (Green, Rowland,
Krupnick, & et al., 1998; Shelby, Golden-Kreutz, & Andersen, 2008). Many survivors report one
or more Post-Traumatic Stress Symptoms (PTSS) (Jim & Jacobsen, 2008), which provides a
continuous measure of post-traumatic stress reactions and risk of PTSD diagnosis.

5

Adding complexity to the post-cancer experience, survivors also report being positively
influenced by cancer. From ancient literary works such as poetries, plays, and essays, it is not rare
to find human beings change in positive ways after near-death or traumatic experiences. Scholarly
attention, however, began to be paid to such phenomena in empirical studies only recently (L. G.
Calhoun & R. G. Tedeschi, 2006a). Positive changes commonly reported among cancer survivors
include better social relationships, greater personal resources such as religious satisfaction, and
better coping skills (Jim & Jacobsen, 2008). These changes are referred to as Post-Traumatic
Growth (PTG), which emphasizes that they occur after an event that is major enough to shake
fundamental life values and world views of the survivors (Zoellner & Maercker, 2006). Stanton,
Bower, and Low (2006) find in their review that the prevalence of positive life changes and
personal growth after a cancer diagnosis ranges from 53% to 95% in cancer survivors.
Using a trauma framework provides a new angle for understanding cancer survivorship,
and it is encouraging that both its positive and negative aspects have been empirically examined.
The gap in the existing literature, however, is that Post-Traumatic Stress (PTS) and PTG have
been studied separately. We only know that a large number of cancer survivors suffer from PTS
and also a large number of survivors experience PTG, but we do not know if the same survivors
have both PTS and PTG or if some survivors are negatively affected by cancer while others are
positively affected. Anecdotally, the same cancer survivors seem to experience these seemingly
contrasting phenomena at the same time; they say, “Cancer was a terrible thing, but it was the
best thing that had happened in life.”
In research, especially when given the opportunity to describe their survivorship
experiences without particular focus on either its positive or negative aspects, survivors seem to
readily report their complex experiences. For example, in a study of young adult cancer survivors

6

using the Photovoice methodology – a participant-driven qualitative research methodology which
empowers participants to determine, through group consensus, salient themes of their experiences
and take photos about the themes--the survivors chose to talk about their simultaneous
experiences of both the positive and negative impact of cancer (Yi & Zebrack, 2010). Such
complex cancer survivorship experiences are difficult to capture in quantitative studies, but the
first step toward understanding the complex human experiences after cancer would be to examine
the relationships between PTS and PTG in cancer survivors.
Levine et al. (2008) identified four theoretically possible relationships between PTS and
PTG. First, because it disturbs human functioning and quality of life, PTS may be inversely
associated with PTG (Johnson et al., 2007). Second, PTG is theorized as an outcome of suffering
from a certain level of PTS that can lead to a change in the survivor’s value system and world
views; hence, PTS may be positively associated with PTG (L. Calhoun & R. Tedeschi, 2006; R.
Tedeschi & Calhoun, 1996). Third, PTS and PTG are not related but may coexist independently
(Linley & Joseph, 2004). Fourth, the relationship between PTG and PTS may be curvilinear
(inverted U), whereby low levels of distress are associated with minimal growth, moderate to
high distress are associated with maximal growth, and extreme distress is associated with poor
adaptation (Levine, et al., 2008; Schnurr, Rosenberg, & Friedman, 1993; Solomon & Dekel, 2007;
R. Tedeschi & Calhoun, 1996). Although PTS and PTG can be experienced simultaneously, there
may be an optimal level of distress that promotes growth. Beyond that level, a person is
overwhelmed by distress, and adaptation and growth are impeded (Butler et al., 2005).
Based on a combination of PTG theories and stress theories, the present study
hypothesizes a curvilinear relationship between PTS and PTG among young adult cancer
survivors. The more traumatic distress cancer survivors have, the more growth they are expected

7

to experience up to a certain level of distress, because one can grow after experiencing an event
only when its impact is severe enough to shake his or her world views (R. Tedeschi & Calhoun,
1995) and transform the person (Parry & Chesler, 2005) according to PTG theories. Beyond the
tipping point, more distress might decrease positive growth experiences among cancer survivors,
because extreme distress would hamper functioning and adaptation according to stress theories (R.
Hill, 1949). Do PTS and PTG coexist in a curvilinear interrelationship among young adult cancer
survivors? Or, do they coexist, but independently? Answers to these questions are essential for
our understanding about the multifaceted nature of cancer survivorship.
METHODS
Sample and procedure
Subjects were recruited from four medical centers that treat pediatric oncology patients
(University of California Los Angeles Mattel Children’s Hospital, the City of Hope National
Medical Center, Los Angeles Children’s Hospital, and University of Michigan Mott Children’s
Hospital). When considering the demographics of the catchment areas for Children's Hospital
Los Angeles, City of Hope National Medical Center and University of Michigan, there are some
demographic differences. Almost 50% of the patient population seen at CHLA is Hispanic/Latino.
City of Hope has a demographic that reflects the nation as a whole, as patients come from all over
the country. University of Michigan draws from across the state, a predominantly Caucasian
population, but representative of urban and rural areas. This strategy of including several
treatment centers throughout the Los Angeles region and in Michigan enhances the ability to
achieve greater variation in patient characteristics and experiences, thereby increasing the
generalizability of the findings. Inclusion criteria included current age 18-39; age at diagnosis 0-

8

21 years (with diagnosis and treatment conducted within a pediatric setting); and disease-free
status. Exclusion criteria included currently receiving treatment, relapse or second cancer.
Each of the respective participating institutions sent out an introductory letter on
letterhead describing the project, outlining the eligibility criteria, and inviting survivors to
participate. Enclosed was a postage-paid response form for potential participants to return
indicating whether they accepted or declined participation and, if they declined, their reasons for
non-participation. Survivors who indicated that they were willing to participate were mailed the
survey. Included was an informed consent form reiterating the purpose of the study, describing
the survey content and the length of time it would take to complete, and explaining the voluntary
and confidential nature of the survey. Participants were asked to sign the informed consent and to
complete and return it with their surveys in a prepaid envelope within one week.
A log was kept indicating when surveys were mailed and returned. Returned surveys
were reviewed for completeness. Missing data or items needing clarification were noted and a
staff research assistant contacted the respondents, if possible, to obtain the missing data. Those
failing to return their surveys within two weeks were telephoned to determine the status of the
survey and whether they intended to complete the survey. Those who did not wish to participate
were asked their reasons for non-participation, which were recorded. For those still willing to
participate, an additional week was given for completion of the survey and they were re-contacted
in two weeks if the survey had not been received. A third and final opportunity was given for
completion of the survey, and those who did not return the survey were considered non-
responders. A comprehensive survey was sent out to 1,500 survivors and responses were obtained
from 666 individuals (44%).

9

Of the 666 individuals, 45 did not meet the inclusion criteria in age and age at diagnosis, so
they were removed for analysis. After removing these cases, the total sample size was 618. 593
survivors who had both PTG and PTS scores were used for this paper, which examined the
relationships between the two constructs.
Measures
The Post-Traumatic Stress Diagnostic Scale (PDS). The PDS was developed and
validated by Foa (1997) to provide a brief but reliable self-report measure of PTSD for use in
both clinical and research settings. Using a four-point scale, respondents rate 17 items
representing the cardinal symptoms of PTSD experienced in the past 30 days, such as re-
experiencing, avoidance, and hyperarousal. The symptoms severity score is the sum of the
responses of the individual items and ranges from 0 to 51. The cut offs for symptom severity
rating are 0 for no rating, 1–10 for mild, for 11–20 moderate, 21–35 for moderate to severe and
36 and higher for severe symptom levels.
The PDS has high face validity because items directly reflect the experience of PTSD
with high internal consistency (coefficient alpha of 0.92). Test–retest reliability was also highly
satisfactory for a diagnosis of PTSD over a 2- to 3-week period (kappa 0.74). Test–retest using
the symptom severity scores yielded a highly significant correlation (0.83). Analysis also revealed
an 82% agreement between diagnosis using the PDS and the Structured Clinical Interview for
DSM (Foa, et al., 1997). In the sample of the present study, internal consistency of the total PDS
scores was high (α=.88).
The scale was validated on samples aged 18–65 and has been used in a wide range of
clinical and research contexts with a high degree of confidence when use of a structured clinical
interview is impractical (Foa, et al., 1997). The PDS has been used in prospective treatment

10

studies helping establish a role for cognitive behavioral therapy in those with established PTSD
(Duffy, Gillespie, & Clark, 2007). Recently, the PDS has been employed in diagnosing PTSD of
the patients in emergency services (Haslam & Mallon, 2003).
The Post-Traumatic Growth Inventory (PTGI). The 21-item scale measures the degree
of the positive changes experienced in the aftermath of a traumatic event. Each item is rated using
a 6-point Likert scale, with values ranging from 0 (I did not experience this change as a result of
my crisis) to 5 (I experienced this change to a very great degree as a result of my crisis). The
PTGI consists of five subscales: Relating to Others (seven items), New Possibilities (five items),
Personal Strength (four items), Spiritual Change (two items), and Appreciation of Life (three
items). Internal consistency for the total score and subscales of the PTGI has been reported as
satisfactory (α coefficient for the total scale=.90, Relating to Others =.85, New Possibilities=.84,
Personal Strength=.72, Spiritual Change=.85, and Appreciation of Life=.67), and the test–retest
reliability (.71) over 2 months has also been reported based on the sample of university students
in the original study (R. Tedeschi & Calhoun, 1996). In addition, the concurrent, discriminant,
and construct validity of the PTGI were examined by assessing the correlations among the PTGI,
social desirability, and personality variables, and by comparing those who had experienced severe
trauma and those who had not.
In the sample of the present study, internal consistency of the total PTGI (α=.96);
Relating to Others (α=.86); New Possibilities (α=.83); Personal Strength (α=.79); Spiritual
Change (α=.60); and Appreciation of Life (α=.81) were high.
Analyses
First, descriptive analyses were conducted to examine the characteristics of major
demographic, medical, PTS, and PTG variables in the sample. Second, to test a curvilinear

11

relationship between PTS and PTG, hierarchical multiple regression analyses were conducted.
The PDS scores were first mean-centered and then squared to create the quadratic term. In
hierarchical regression, PTGI scores (the total scores and the five subscale scores) were then
regressed onto the linear PDS effect in Step 1, and the quadratic PDS effect in Step 2, to examine
whether the quadratic effects exist even beyond the linear terms. The Statistical Package for the
Social Sciences (SPSS version 16.0) was used for the analyses. The alpha level was set at p=0.05.
RESULTS
Descriptive analysis
On average, the participating survivors were diagnosed about 16 years ago at about 11
years of age, and the average age at the time of study enrollment was around 27 years old. There
were slightly more women (n=314, 53.1%) than men (n=277, 46.9%). Most of the participants
were non-Hispanic White (n=379, 64.5%) or Latino (n=136, 23.1%). Descriptive statistics of the
sample characteristics are listed in Table 1.

12

Table 1
Participants characteristics (N=593)
Variables n %
Sex Female 314 53.1
Male 277 46.9

Race
a
White 379 64.5
Non-White 209 35.5

Income $25,000 or below 204 35.4
Above $25,000 373 64.6

Education High school and below 112 18.9
Some college and A.A. degree 275 46.4
College graduate or above 199 33.6

Employment Status
b
Unemployed 75 12.6
Employed 512 86.3

Type of cancer Hematological 371 62.6
CNS or brain tumor 71 12.0
Solid or soft tissue tumors 151 25.5
Age at diagnosis (0-21) M=11.22 SD=6.01
Age (18-39) M=26.94 SD=5.43
Years since diagnosis (2-37) M=15.71 SD=6.99
Percentage indicates valid percentage.
a
White/Caucasian (n=379, 64.5%); Black or African American (n=26, 4.4%); Asian/Pacific islander (n=38, 6.5%);
Hispanic/Latino (n=136, 23.1%); American Indian/Alaskan native (n=7, 1.2%); and other (n=1, 0.2%).
b
Employed includes full-time, part-time, homemaker, and student; unemployed includes disability, unemployed, and
unable to work.

13

In this sample, the mean of the PDS scores was 7.16 (SD=7.43) in the range of 0-43. 96
(16.2%) individuals reported no PTS symptoms; 348 (58.7%) cases reported mild (1-10 in PDS
scores) PTS symptoms; 112 (18.9%) reported moderate (11-20) symptoms; 34 cases (5.7%)
reported moderate to severe (21-35) PTS symptoms; and 3 (.5%) cases reported severe (36 and
higher) PTS symptoms.
The large majority (88.5% ,n= 524) of the survivors reported some degree of positive
change as reflected by a mean score above 1 (higher than very little influence of cancer on growth)
on the 6-point scale PTGI. Posttraumatic growth scores were generally moderate, with a mean
PTGI score of 2.74 (SD=1.20, range=0-5). The highest scores were found for Appreciation of
Life (M=3.31, SD=1.39) and Personal Strength (M=3.09, SD=1.33).
Curve estimation analyses
Linear and quadratic relations between PTS and PTG were tested in hierarchical
regression analyses where PTGI scores (the total scores and the five subscale scores) were
regressed onto the linear PDS effect in Model 1, and the linear and quadratic PDS effects in
Model 2.

14

For the Personal Strength subscale, both Model 1 (F=5.068, df=1, 586, p=.025) and
Model 2 (F=5.712, df=2, 585, p=.011) were significant. In Model 1, the linear term was
significant (B=.017, t=2.251, p=.025). In Model 2, both the linear (B=.034, t=3.373, p=.001) and
quadratic (B =-.002, t=-2.51230, p=.012) relationships were significant. For the Appreciation of
Life subscale, both Models 1 (F=13.449, df=1, 586, p=.00) and 2 (F=7.394, df=2, 585, p=.001)
were significant. In Model 1, the linear term of PDS was significant (B=.026, t=3.33, p=.001). In
Model 2, the linear term (B=.037, t=3.43, p=.001) was significant, but its quadratic term was not
significant. The other subscales of PTGI and the total PTGI were not significantly related to PDS.
Table 2 delineates the significant results of the hierarchical regression analyses.

15

Table 2
Curve estimation regression analysis results
Model 1 Model 2
B T B t
Personal Strength
Step 1
Linear PTS 0.017 2.251* 0.034 3.373**
Step 2

Quadratic
PTS -.002 -2.512*
R
2
0.009 0.019
∆ R
2
0.011*
Appreciation of Life
Step 1
Linear PTS 0.028 3.667** 0.037 3.43**
Step 2

Quadratic
PTS 0.00 -1.154
R
2
0.022 0.025
∆ R
2
0.002
* p<0.05, ** p<0.01

16

Figure 1 graphically presents the relationships between PDS and PTGI scores. Only the
subscale of Personal Strength had a significant curvilinear relationship with PTS, while the
relationships between PTS and the other posttraumatic growth scores were not statistically
significant despite their curvilinear trends.

Figure 1
Graphical presentation of the relationships between PTS and PTG

0
0.5
1
1.5
2
2.5
3
3.5
No PTS Mild Moderate Moderate
to Severe
Severe
PTGI Average
Relating to Others
New Possibilities
Personal Strength
Spiritual Change
Appreciation of Life

17

DISCUSSION
This study examined the PTS/PTG interrelationships with the question, “Do PTS and
PTG coexist in a curvilinear interrelationship among young adult cancer survivors? Or, do they
coexist, but independently?” The findings were mixed: the linear relationship between
Appreciation of Life (a PTG subscale) and PTS; the quadratic relationship between Personal
Strength (a PTG subscale) and PTS; and non-significant relationships between the other three
subscales of PTG and PTS. These mixed results and predominantly non-significant relationships
are different from the hypothesized quadratic relationships.
These findings support a small, but growing body of literature suggesting that growth and
symptom severity may be independent of one another (M. Cordova, Cunningham, Carlson, &
Andrykowski, 2001; Lehman, Davis, & Delongis, 1993; Videka-Sherman, 1982; Wild & Paivio,
2003). Individuals may report some gains as a result of their trauma while still experiencing
significant distress. In other words, both growth outcomes and psychological distress can co-exist.
Survivors might well be expected to experience posttraumatic symptoms about the cancer
diagnosis and the possibility of recurrence, as well as sadness about the late effects, or even their
lives. Concurrently, however, they might experience a significant enhancement of bonds with
close family members and friends, have a greater appreciation for the small things in daily life,
and become more peaceful about fundamental existential questions about life’s meaning and
purpose. In their conceptual overview of growth, Tedeschi and Calhoun (2004) assert that there
may not be a direct relationship between these constructs, but that this lack of relationship is not a
limitation of the growth construct, as growth is simply not the same as a decrease in distress or an
increase in well-being.

18

Although statistically not significant, the trend of the relationships had curvilinear shapes,
as graphically depicted in Figure 2. By looking at the differences of this sample from the other
studies where curvilinear relationships were detected to be significant, the factors that might
influence these different results could be theorized.
In the existing literature, four studies report significant curvilinear relationships between
PTS and PTG among the people directly or indirectly affected by the 9/11 bombing (Butler, et al.,
2005); ex-prisoners of the 1973 Yom Kippur War (Solomon & Dekel, 2007); survivors of various
forms of terrorism (Levine, et al., 2008); and survivors of physical assaults (Kleim & Ehlers,
2009). There are at least four distinct characteristics in this study sample in comparison with the
previous four studies.
First, the type of trauma is unique in this sample. No study has examined the
relationships between the two constructs among cancer survivors, although it is not clear how the
type of trauma affects the results. Distinct differences of cancer from other kinds of traumas
include multiple stressors, the internal nature of the crisis, future-focused fears, difficulty of
establishing the termination of the trauma (Sumalla, Ochoa, & Blanco, 2009). Some acute
traumatic events can be characterized by their discrete nature. For example, a traffic accident or
an assault is a traumatic event that occurs in a discrete and usually brief period of time, and then
is over, although the effects can be long lasting. On the other hand, oncological illnesses represent
an ongoing trauma that can last for a very long time, perhaps for as long as the person lives. Even
when a person completes treatment and is in remission, the potential exists for additional trauma
to arise in the future, stemming from the original diagnosis. The traumatic stressors may be
associated with the diagnosis of cancer, its severity and prognosis, the aggressiveness of
treatments, alternations in body image, a decrease in the level of functional autonomy and/or role

19

alternations. Thus, in cancer, it is usually very difficult to identify the exact stressor or group of
stressors that precipitate posttraumatic growth. These differences might lead to different
relationships between PTS and PTG among cancer survivors from survivors of other traumas.
Second, participants of this sample are young adults (M=26.95). Participants of the three
studies are middle-aged between 34 and 53 at average while those of one study are children at
grades 7-9.
Third, participants of this sample are relatively long-term survivors (15.74 years since
cancer diagnosis). Time since trauma exposure is much shorter in two studies (17-468 days). The
prevalence of PTSD symptoms seems to decline considerably for the majority of survivors within
3 months postdiagnosis or following treatment completion.(Manuel, Roth, Keefe, & Brantley,
1987; Mundy et al., 2000). Thus, the PTS/PTG relationships might look different between the
acute and long-term phases.
Fourth, PTG levels are higher in this sample. The mean PTGI scores are between 32.76
and 56.8 in the previous studies while the mean PTGI in this sample is 57.75.
Temporal factors such as age and time since trauma exposure might be valuable in
examining the PTS/PTG relationships. Longitudinal studies examining such relationships along
the passage of the time since trauma would be important to understand the possible change in the
patterns of the relationships. Testing the same relationships in other samples of cancer survivors
would help prove whether the two posttraumatic phenomena are curvilinearly related, as in the
survivorship of other traumas, or they are not related or related in different ways in the case of
cancer.
Whether PTG is a unitary or multidimensional construct is another research question
beyond the scope of this study. Total growth scores tend to have higher internal consistency

20

reliability than the individual subscales in spite of comprising a variety of domains of growth (R.
Tedeschi & Calhoun, 1996). However, diverse patterns of associations between the subscales of
PTG and PTS in this study might imply that the subscales are better than the total PTG construct
in examining their relationships with PTS.
Some limitations of this study need to be considered in interpreting the findings. The
sample of this study might not represent the general young adult cancer survivors, because
participant recruitment was volunteer-based and those who function better and are socially active
are more likely to participate in research studies. Lower social economic status (M. J. Cordova et
al., 1995), lower education (M. J. Cordova, et al., 1995; Jacobsen et al., 1998), and poor social
functioning (Kelly et al., 1995; Smith, Redd, DuHamel, Vickberg, & Ricketts, 1999) are
predictive of PTSD. Lower level of PTS in this sample might be relevant to some sample
characteristics of this study such as higher education level, employment, and income.
This study employed a cross-sectional methodology that limits the causal inferences
about the PTS/PTG relationships and precludes shedding light on the possible change of such
relationships along the passage of time since diagnosis in survivorship. Although the present
study defined PTS as a predictor and PTG as an outcome based on the theory that a certain level
of PTS is necessary to produce PTG, cross-sectional studies cannot distinguish cause and effect.
We cannot tell whether moderate levels of PTS promoted the perceived personal strength or the
perceived personal strength helped reduce the distress level to a moderate level. Future studies
should tease apart and explain the relationship between PTS and PTG with longitudinal study.
The present findings can still be clinically applied for young adult cancer survivors. It
should be recognized that cancer survivors with higher levels of PTS might also experience
positive growth experiences from cancer; it might be useful to provide opportunities for the

21

survivors to discover the meanings of their previous life adversities. Also, we should not neglect
to remember that the survivors full of personal strength and can-do spirit might also have
posttraumatic distress at quite high levels. They might be the group that needs extra attention
because their distress levels might be close to the tipping point. As one of the first studies
examining the curvilinear relationships between PTS and PTG with cancer survivors based on
PTG theories, this study proved a quantitatively possible coexistence of the positive and negative
impact of cancer, laying the foundation for understanding complex post-cancer experiences. It is
recommended that future studies examine demographic, psychosocial, and medical factors that
are associated with such complex experiences.

22

CHAPTER 3. POST-TRAUMATIC PROFILES
INTRODUCTION
Since the 1970s, cancer survival rates have continued to increase, reaching higher than 78%
(Ries et al., 2003). Research and practical attention has shifted from survival alone to quality of
life in survivorship. We hope that the child not only survives cancer, but also adapts to their
circumstances and experiences a healthy development even after treatment. Such hope has led to
numerous studies on correlates of the negative and positive impact of cancer. These studies
attempt to identify areas that need focused attention and modifiable factors that could make a
difference in the lives of cancer survivors.
Posttraumatic theories and perspectives have been applied to understanding cancer
survivorship since the inclusion of cancer as a type of trauma in 1994 (American Psychiatric
Association, 1994). Negative and positive aspects of cancer were often examined with the
constructs of Post-Traumatic Stress (PTS) and Post-Traumatic Growth (PTG). By investigating
what factors are associated with the different psychosocial outcomes among survivors,
researchers have been interested in finding who among cancer survivors are likely to experience
either PTS, or PTG, under the implied assumption that only one of the two can happen in each
individual. Thus, in most studies in this field, either PTS or PTG was used as a dependent
variable and the significance of its relationships with correlates were tested. Such approach is
referred to as a variable-centered approach in which the aim is to predict outcomes, and relate
independent and dependent variables (Lubke & Muthen, 2005; Muthen, 2001; Pastor, Barron,
Miller, & Davis, 2007)

23

Cancer survivors are likely to have experienced many stressors related to diagnosis,
prognosis, invasive treatments, side effects, and risk of recurrence (Bruce, 2006). Cancer-specific
Post-Traumatic Stress Disorder (PTSD) symptoms include the intrusion of unwanted memories,
such as nightmares or flashbacks; the avoidance of reminders of the events, such as doctors or
hospital; an increased startle response to sudden noise; and the constant monitoring for danger
(Foa, et al., 1997). Bruce (2006) reviewed 24 articles on PTSD among childhood cancer
survivors and found that incidences of current cancer-related PTSD ranged from 4.7% to 21% in
this group. Consistent with the trend in other traumas, higher rates of cancer-related PTSD
correlated with female gender and reduced social support. But, unlike in other traumas, objective
trauma features such as cancer treatment type and intensity were not related with PTSD. The
passage of time since the cancer diagnosis failed to correlate with PTSD, which is also
inconsistent with the findings in other traumas in which a longer passage of time since trauma
exposure is associated with less PTSD. PTSD is generally related with morbidity (Hobbie et al.,
2000), poor quality of life, and depression (Schwartz & Drotar, 2006) in the cancer survivor
population.
At the other end of the posttraumatic research spectrum is research focusing on Post-
Traumatic Growth (PTG) among cancer survivors. Some cancer survivors report that they have
grown beyond their previous state after experiencing cancer (R. Tedeschi & Calhoun, 1995). The
last 15 to 20 years have seen PTG research increasing. Stanton, Bower, and Low (Stanton, et al.,
2006) find in their recent review that the prevalence of positive life changes and personal growth
after a cancer diagnosis ranges from 53% to 95% in cancer survivors. Studies on correlates of
PTG have produced inconsistent and divergent findings. No conclusive relationships have been
found between PTG and sociodemographic factors such as ethnicity, age, and gender.

24

Counterintuitively, PTG is generally not associated with positive physical and mental health
outcomes (Bower et al., 2005; Fromm, Andrykowski, & Hunt, 1996; S. Manne et al., 2004;
Schulz & Mohamen, 2004).
As reported in many aforementioned studies based on variable-centered approaches,
there are inconsistent findings on correlates of Post-Traumatic Stress (PTS) and PTG. If PTS and
PTG are simply positioned at the different ends of the continuum of quality of life among cancer
survivors, namely one at the positive end and the other at the negative end, quality of life
correlates would behave simply in opposite directions in their relationships with these two
seemingly contrasting aspects of survivorship, but they do not. Such mixed findings might be
explained by curvilinear relationships between PTS and PTG.
Childhood cancer survivors seem be affected by the complex interactions of two forces,
namely PTS and PTG. PTG may be achieved only when the trauma is at a certain level that could
lead to a change in identity (R. Tedeschi & Calhoun, 1995), so the perception of growth from
traumatic experience does not necessarily suggest the absence of negative effects, and both
negative effects and positive effects are widely reported in the same person (Aldwin, 2007;
Joseph, Williams, & Yule, 1993; R. Tedeschi & Calhoun, 1996). On the other hand, according to
stress and coping theories, too much or piled-up stress (R. Hill, 1949) might lead to dysfunctional
health outcomes in the trauma survivors. Thus, possible curvilinear relationships between PTS
and PTG can theoretically be established. Although limited, there is empirical evidence
supporting this theory in survivors of different kinds of traumas (Butler, et al., 2005; Kleim &
Ehlers, 2009; Levine, et al., 2008; Solomon & Dekel, 2007). Although curvilinear relationships
between PTG and PTSD were not significant, the two phenomena seemed to coexist in some
young adult cancer survivors, as reported in Chapter 2 of this dissertation. Survivors might have

25

complex combinations of PTS and PTG. For example, some might have high PTS and high PTG,
while others have high PTS and low PTG. As the prevalence of PTSD (4.7% to 21%) and PTG
(53% to 95%) among childhood cancer survivors imply, some survivors might have both stress
and growth experiences together.
The present study utilizes a person-centered analysis for posttraumatic phenomena
among young adult cancer survivors. The idea of looking at PTS and PTG in the same individual
is not new to the field from the theoretical perspective. However, there is no such empirical study
in the existing literature. Person-centered analyses, such as latent profile analysis, identify the
types of survivors with similar patterns of posttraumatic impact profiles, examine configurations
of PTS and PTG factors within the person, and portray trauma phenomena in a holistic fashion
that complement and extend the traditional variable-centered research (Pastor, et al., 2007). The
typological approach focuses on relations among individuals with the goal to sort individuals into
groups whose members are similar to each other and different from those in other groups
(Muthén & Muthén, 2006). While variable-centered models cluster variables, person-centered
approaches cluster individuals. By identifying these clusters of survivors, it is possible to find
correlates and outcomes of each profile of individuals, instead of the variables.
METHODS
Sample and procedure
Subjects were recruited from four medical centers that treat pediatric oncology patients
(University of California Los Angeles Mattel Children’s Hospital, the City of Hope National
Medical Center, Los Angeles Children’s Hospital, University of Michigan Mott Children’s
Hospital). When considering the demographics of the catchment areas for Children's Hospital Los

26

Angeles, City of Hope National Medical Center and University of Michigan, there are some
demographic differences. Almost 50% of the patient population seen at CHLA is Hispanic/Latino.
City of Hope has a demographic that reflects the nation as a whole, as patients come from all over
the country. University of Michigan draws from across the state, a predominantly Caucasian
population, but representative of urban and rural areas. This strategy of including several
treatment centers throughout the Los Angeles region and in Michigan enhances the ability to
achieve greater variation in patient characteristics and experiences, thereby increasing the
generalizability of the findings. Inclusion criteria included current age 18-39; age at diagnosis 0-
21 years (with diagnosis and treatment conducted within a pediatric setting); and disease-free
status. Exclusion criteria included currently receiving treatment, relapse or second cancer.
Each of the respective participating institutions sent out an introductory letter on
letterhead describing the project, outlining the eligibility criteria, and inviting survivors to
participate. Enclosed was a postage-paid response form for potential participants to return
indicating whether they accepted or declined participation and, if they declined, their reasons for
non-participation. Survivors who indicated that they were willing to participate were mailed the
survey. Included was an informed consent form reiterating the purpose of the study, describing
the survey content and the length of time it would take to complete, and explaining the voluntary
and confidential nature of the survey. Participants were asked to sign the informed consent and to
complete and return it with their surveys in a prepaid envelope within one week.
A log was kept indicating when surveys were mailed and returned. Returned surveys
were reviewed for completeness. Missing data or items needing clarification were noted and a
staff research assistant contacted the respondents, if possible, to obtain the missing data. Those
failing to return their surveys within two weeks were telephoned to determine the status of the

27

survey and whether they intended to complete the survey. Those who did not wish to participate
were asked their reasons for non-participation, which were recorded. For those still willing to
participate, an additional week was given for completion of the survey and they were re-contacted
in two weeks if the survey had not been received. A third and final opportunity was given for
completion of the survey, and those who did not return the survey were considered non-
responders. A comprehensive survey was sent out to 1,500 survivors and responses were obtained
from 666 individuals (44%).
Of the 666 individuals, 48 did not meet the inclusion criteria in age and age at diagnosis, so
they were removed for analysis. After removing these cases, the total sample size was 618. For
this study, 593 who had both PTSD and PTG scores were included.
Measures
The Post-Traumatic Stress Diagnostic Scale (PDS). The PDS was developed and
validated by Foa (1997) to provide a brief but reliable self-report measure of PTSD for use in
both clinical and research settings. Using a four-point scale, respondents rate 17 items
representing the cardinal symptoms of PTSD experienced in the past 30 days, such as re-
experiencing, avoidance, and hyperarousal. The symptoms severity score is the sum of the
responses of the individual items and ranges from 0 to 51. The cut offs for symptom severity
rating are 0 for no symptoms, 1–10 for mild, for 11–20 moderate, 21–35 for moderate to severe
and 36 and higher for severe symptom levels.
The PDS has high face validity because items directly reflect the experience of PTSD
with high internal consistency (coefficient alpha of 0.92). Test–retest reliability was also highly
satisfactory for a diagnosis of PTSD over a 2- to 3-week period (kappa 0.74). Test–retest using
the symptom severity scores yielded a highly significant correlation (0.83). Analysis also revealed

28

an 82% agreement between diagnosis using the PDS and the Structured Clinical Interview for
DSM (Foa, et al., 1997). The scale was validated on samples aged 18–65 and has been used in a
wide range of clinical and research contexts with a high degree of confidence when use of a
structured clinical interview is impractical (Foa, et al., 1997). The PDS has been used in
prospective treatment studies helping establish a role for cognitive behavioral therapy in those
with established PTSD (Duffy, et al., 2007). Recently, the PDS has been employed in diagnosing
PTSD of the patients in emergency services (Haslam & Mallon, 2003).
Total PDS scores were used for the analyses of the present study and the internal
consistency of the total PDS scores was high (α=.88).
The Post-Traumatic Growth Inventory (PTGI). The 21-item scale measures the degree
of the positive changes experienced in the aftermath of a traumatic event. Each item is rated using
a 6-point Likert scale, with values ranging from 0 (I did not experience this change as a result of
my crisis) to 5 (I experienced this change to a very great degree as a result of my crisis). The
PTGI consists of five subscales: Relating to Others (seven items), New Possibilities (five items),
Personal Strength (four items), Spiritual Change (two items), and Appreciation of Life (three
items). Internal consistency for the total score and subscales of the PTGI has been reported as
satisfactory (α coefficient for the total scale=.90, Relating to Others =.85, New Possibilities=.84,
Personal Strength=.72, Spiritual Change=.85, and Appreciation of Life=.67), and the test–retest
reliability (.71) over 2 months has also been reported based on the sample of university students
in the original study (R. Tedeschi & Calhoun, 1996). In addition, the concurrent, discriminant,
and construct validity of the PTGI were examined by assessing the correlations among the PTGI,
social desirability, and personality variables, and by comparing those who had experienced severe
trauma and those who had not.

29

For the analyses of the present study, the five subscale scores of PTG were used and
internal consistency of Relating to Others (α=.86); New Possibilities (α=.83); Personal Strength
(α=.79); Spiritual Change (α=.60); and Appreciation of Life (α=.81) were high.
The Medical Outcomes Study Social Support Survey (MOS SSS). The MOS SSS
(Sherbourne, 1991) was developed for chronically ill patients in a medical outcomes study. The
MOS SSS is a 19-item self-administered questionnaire comprising four subscales to assess
functional dimensions of social support. Eighteen items are used to form the subscales:
emotional/informational (8 items), affectionate (3 items), tangible (4 items), and positive social
interaction (3 items). Subscale scores or the total index score can be used. Emphasis is on the
perceived availability of support if needed. This instrument assesses the types of support, but not
the sources of support. Respondents are asked to indicate on a Likert-type scale from 1 (none of
the time) to 5 (all of the time) how often the type of support is available if needed. Examples of
items are “Someone you can count on to listen to you when you need to talk” and “Someone to
help with daily chores if you were sick.”
For the analyses of the present study, total scores of the scale were used and its internal
consistency was high (α=.96)
SF-36, Version 2. Also known as the MOS SF-36, this 36-item survey is a widely-used
and well-validated instrument that assesses several aspects of quality of life including physical,
social and psychological functioning. This self-report measure contains eight individual subscales
that represent three general areas of health-related quality of life: physical, emotional, and social
well-being. The subscales include physical functioning, role function-physical (assessing role
limitations caused by physical factors), bodily pain, social functioning, mental health, role

30

function-emotion (assessing role limitations caused by emotional factors), vitality, and general
health. General population norms are available for the SF-36.
The General Health Scale of the SF-36 is often used as a global rating of health status,
with some evidence that it more frequently tracks with specific health problems, physical
functioning, and health behaviors and tracks less strongly with aspects of mental health. The SF-
36 can also be scored as two summary scales – one for physical health and a second for mental
health. The data for these summary scales are presented as t-scores, with a normal healthy
population mean score set at 50 and a score of 60 or 40 representing 1 standard deviation above
or below the mean, respectively. These scales are called the SF-36 Physical and Mental
Component Summary scales. For the analyses of the present study, the two summary scales were
used and the internal consistency of the Physical Component Summary (α=.79) and Mental
Component Summary (α=.88) were high.
Other demographic and medical variables were used for analyses, such as age at
diagnosis, time since diagnosis, education, employment, and income.
Analysis
Analyses were conducted in two steps: (1) to identify the latent profiles of the cancer
survivors based on their posttraumatic responses, PTS and PTG; and then (2) to examine the
characteristics of the classified survivors. In Step 1, the Latent Profile Analysis (LPA), MPlus
(version 5.21 for Windows) was used. In Step 2, the Statistical Package for the Social Sciences
(SPSS version 16.0) was used.
In Step 1, LPA was used to classify the survivors into clusters based on the total PDS
scores and the 5 subscale scores of the PTG. LPA is a type of latent variable mixture model
(Pastor, et al., 2007). The term, latent variable, is here referring to the latent categorical variable

31

of cluster membership. This latent categorical variable has K number of clusters. A person’s
value on this latent variable is thought to cause the individual’s levels on the observed cluster
indicators, which in this sample would be the PTS and PTG subscale scores. The term mixture is
referring to the notion that the data are conceived as being sampled from a population with a mix
of distributions, one for each cluster. The latent variable mixture model used with only
continuous cluster indicators is often called LPA (Lanza, Flaherty, & Collins, 2003). LPA first
utilizes all observations that are associated with the dependent variables, PTS and five subscales
of PTG, and performs maximum likelihood estimation (Little & Rubin, 1987).
Three models, including 2-cluster, 3-cluster, and 4-cluster models, were tested in LPA.
The 2-cluster model was conceived based on the theory that PTS and PTG are linearly related,
thus survivors in one cluster have the lower PTS and PTG levels and those in the other cluster
have the higher PTS and PTG levels. The 3- and 4-cluster models were conceived based on the
theory that PTS and PTG are curvilinearly related. In these models, survivors with the lowest
levels of PTS and PTG belong to one cluster; one (in the 3-cluster model) or two (in the 4-cluster
model) clusters are located along the increasing PTS levels (e.g. those with mild PTS and PTG
and those with moderate PTS and PTG); and the last cluster consists of the survivors with the
highest PTS and the lower PTG.
Plausibility of the 2-, 3-, and 4 cluster models was examined to determine the best model
fit for the data according to both statistical and interpretive perspectives. Model fit was evaluated
using the Lo-Mendell-Rubin Adjusted Likelihood Ratio Test (LMRT) which is a statistical
indicator of the number of classes that best fit the data (Lo, Mendell, & Rubin, 2001). The LMRT
statistically compares the fit of a target model (e.g., a 3-class model) to a model that specifies on
a fewer clusters (e.g., a 2-cluster model). P values less than .05 indicate that the “higher cluster”

32

model fits better (e.g., 3-cluster better than 2-cluster). P-values greater than .05 indicate that the
“lower cluster” model fits better. Both the Akaike Information Criteriion (AIC; Akaike, 1974) )
and the sample size-adjusted Bayesian Information Criterion (BIC; Schwarz, 1978) were also
examined to choose the most optimal cluster model. Optimal model fit is defined by lower AIC
and BIC values (Kline, 2005). Finally, the Entropy criterion was also examined. Entropy is an
index that determines the accuracy of classifying people into their respective profiles or clusters,
with higher values indicating that this model fits better. Each individual’s probability of cluster
membership can be estimated so that the person may be classified into the most appropriate
cluster (A. L. Hill, Degnan, Calkins, & Keane, 2006). Sample size of each cluster, uniqueness of
the cluster profiles for each model, and theoretical interpretability were also considered in
deciding upon the final model.
After the optimal model was determined in Step 1, LPA, Step 2 of the analysis was
conducted to examine the characteristics of the survivor profiles. A series of analyses were
conducted, such as Analysis of Variance (ANOVA) and chi square analyses, on the relationships
between the cluster membership and demographic (gender, race, age, income, education,
employment), cancer-related (time since cancer diagnosis, cancer type, age at diagnosis), coping
resources (social support), and health outcome (physical and mental health) variables. The alpha
level was set at p=0.05.

33

RESULTS
Descriptive statistics
On average, the participating survivors were diagnosed about 16 years ago at about 11
years of age, and the average age at the time of study enrollment was around 27 years old. There
were slightly more women (n=314, 53.1%) than men (n=277, 46.9%). Most of the participants
were non-Hispanic White (n=379, 64.5%) or Latino (n=136, 23.1%). Descriptive statistics of the
sample characteristics are listed in Table 3.

34

Table 3
Participants characteristics (N=593)
Variables n %
Sex Female 314 53.1
Male 277 46.9

Race
a
White 379 64.5
Non-White 209 35.5

Income $25,000 or below 204 35.4
Above $25,000 373 64.6

Education High school and below 112 18.9
Some college and A.A. degree 275 46.4
College graduate or above 199 33.6

Employment Status
b
Unemployed 75 12.6
Employed 512 86.3

Type of cancer Hematological 371 62.6
CNS or brain tumor 71 12.0
Solid or soft tissue tumors 151 25.5
Age at diagnosis (0-21) M=11.22 SD=6.01
Age (18-39) M=26.94 SD=5.43
Years since diagnosis (2-37) M=15.71 SD=6.99
Percentage indicates valid percentage.
a
White/Caucasian (n=379, 64.5%); Black or African American (n=26, 4.4%); Asian/Pacific islander (n=38, 6.5%);
Hispanic/Latino (n=136, 23.1%); American Indian/Alaskan native (n=7, 1.2%); and other (n=1, 0.2%).
b
Employed includes full-time, part-time, homemaker, and student; unemployed includes disability, unemployed, and
unable to work.

35

Latent profile analysis results
Statistical model fits were compared across the different models, using fit indexes such as
LMRT, AIC, BIC, and Entropy. Substantive interpretability was also considered by classifying
the survivors into the 2, 3, and 4 clusters and examining the unique characteristics of each cluster
based on the hypothesized theories on PTS and PTG levels. The 2-cluster model was better than
the 1-cluster model, evidenced by the significance of the LMRT value. The 2-cluster model
confirmed the theory that the relationships between PTS and PTG are linear: One cluster with the
lower PTS (M=6.73, SD=7.63) and PTG (M=1.56, SD=.77) and the other cluster with the higher
PTS (M=7.46, SD=7.29) and PTG (M=3.56, SD=.64). However, the 3-cluster model was
considered better than the 2-cluster model due to lower AIC and BIC values, a higher Entropy
value, and a significant LMRT value.
The 4-class solution was considered better than the 3-class solution in lower AIC and
BIC values, and a significant LMRT value. But, the 3-class solution was considered better than
the 4-class solution with a higher Entropy value. Substantively, the 4-class solution was not
interpretable due to the lack of unique characteristics of each cluster: Cluster 1 with PTS (M=6.69,
SD=6.69) and PTG (M=4.14, SD=.39); Cluster 2 with PTS (M=7.26, SD=8.11) and PTG (M=.67,
SD=.422); Cluster 3 with PTS (M=3.62, SD=3.19) and PTG (M=1.96, SD=.35); Cluster 4 with
PTS (M=9.38, SD=8.49) and PTG (M=2.96, SD=.38). Therefore, the 3-cluster model was
determined to be the best model, considering both statistical fit and substantive interpretation.
Table 4 contains that AIC, BIC, LMRT, and Entropy values for the latent profile analyses
conducted.

36

Table 4
Model fit indexes for the 2-, 3-, 4-, and 5-class solutions
Fit index 2-class 3-class 4-class
AIC 12546.959 11696.973 11409.915
BIC 12577.221 11742.973 11471.651
LMRT 1665.066, p= .0020 865.558, p=.0016 309.331, p= .0137
Entropy .885 .915 .892
AIC, Akaike Informaiton Criterion; BIC, sample size-adjusted Bayesian Information Criterion;
LMRT, Lo-Mendell-Rubin Test.

In the 3-cluster model, Cluster 1 was composed of 199 survivors (33.6%), Class 2 of 273
survivors (46%), and Class 3 of 121 survivors (20.4%). Cluster 1 reflected individuals who grew
the most although they still have negative posttraumatic symptoms. They have the highest mental
functioning in the sample. Accordingly, this cluster was referred to as Thrivers. Cluster 2
indicated survivors who suffer the highest level of distress although they report some growth.
They have the poorest physical functioning in the sample, so this cluster was referred to as
Sufferers. Cluster 3 are the survivors who have the least impact of trauma both in positive and
negative senses. They bounce back to the state prior to the trauma exposure. Thus, this profile
was referred to as Recoverers. Table 5 shows the overall means and cluster means of PTS and
PTG subscales in the 3-cluster model.

37

Table 5
Overall sample means and means of profiles in the 3-cluster model
Post-Traumatic
response items
Sample mean
(n=593) (SD)
Cluster 1:
Thrivers (n=199)
(SD)
Cluster2:
Sufferers
(n=273) (SD)
Cluster3:
Recoverers
(n=121) (SD)
Relating to
Others
2.73 (1.30) 4.05 (.56) 2.58 (.67) .88 (.65)
New Possibilities 2.22 (1.35) 3.53 (.85) 2.00 (.88) .55 (.56)
Personal Strength 3.09 (1.33) 4.25 (.59) 3.10 (.79) 1.16 (.91)
Spiritual Change 2.51 (1.75) 3.92 (1.16) 2.29 (1.49) .68 (1.04)
Appreciation of
Life
3.31 (1.39) 4.43 (.60) 3.34 (.94) 1.39 (1.07)
Post-Traumatic
Stress
7.16 (7.43) 6.62 (6.60) 8.10 (7.95) 5.91 (7.31)

38

Characteristics of the survivor profiles
Demographic
Chisquare and ANOVA tests were conducted between the cluster membership and
demographic variables, such as gender, employment, income, education, and age. No significant
relationship was found in any of the relationships. It seems these three groups do not have
differences in the demographic variables.
Cancer-related
Chisquare and ANOVA tests were conducted between the cluster membership and
cancer-related variables such as age at diagnosis, years since diagnosis, and cancer type. Cancer
type was not related to the class membership, but age at diagnosis (F (2, 590) =10.30, p=.00) and
years since diagnosis (F (2, 590) =5.49, p=.004) were significantly related. As post-hoc analyses,
Tukey tests were conducted. Thrivers were the most recent survivors with the mean number of
years since diagnosis at 14.77 (SD=6.93); Sufferers at15.65 (SD=7.05); and Recoverers were the
most long-term survivors (M=17.41, SD=6.69). Thrivers were significantly more likely to be
recent survivors of cancer than those in Recoverers (p=.003).
Recoverers were diagnosed with cancer at the youngest ages (M=9.07, SD=6.29);
Sufferers at (M=11.61, SD=5.98); and Thrivers at the oldest ages (M= 12.00, SD=5.60). The
differences of the mean ages at diagnosis were significant between Recoverers and both Thrivers
(p=.00) and Sufferers (p=.00).
Social support
ANOVA tests were conducted between the cluster membership and social support and
the result was significant (F (2, 583) =10.88, p=.00). Mean social support scores were 4.36

39

(SD=.70) in Thrivers; 4.05 (SD=.83) in Sufferers; and 3.99 (SD=.92) in Recoverers. Tukey tests
were conducted as post-hoc analyses. Thrivers had significantly higher social support than those
in Recoverers (p=.00), and also Sufferers (p=.00).
Physical and mental health
ANOVA tests conducted between cluster membership and physical/mental functioning
revealed significant differences across the clusters both in physical health (F (2, 565) = 3.60,
p=.028) and in mental health (F (2, 565) =3.68, p=.026). Physical health scores were highest in
Recoverers (M=54.02, SD=7.05); medium in Thrivers (M=51.90, SD=8.40); and lowest in
Sufferers (M=51.44, SD=9.55). Sufferers had significantly lower physical health than Recoverers
(p=.022).
Mental health scores were highest in Thrivers (M= 48.27, SD=11.37); medium in
Recoverers (M=47.52, SD=11.88); and lowest in Sufferers (M=45.34, SD=11.17). Tukey tests as
post-hoc analyses revealed that survivors in Thrivers had significantly higher mental functioning
than those in Sufferers (p=.026).

40

Table 6
Posttraumatic profile differences in correlates
Correlates Thrivers (M) (SD) Sufferers (M) (SD) Recoverers (M) (SD)
Age at diagnosis 12.00 (5.60) 11.61 (5.98) 9.07 (6.29)
a

Years since diagnosis 14.77 (6.93)
b
15.65 (7.05) 17.41 (6.69)
Social support 4.36 (.70)
c
4.05 (.83) 3.99 (.92)
Physical functioning 51.90 (8.40) 51.44 (9.55)
d
54.02 (7.05)
Mental functioning 48.27 (11.37)
e
45.34 (11.17) 47.52 (11.88)
a
p=.00 (Recoverers were diagnosed with cancer at significantly younger ages than Thrivers and
Sufferers).
b
p=.003 (Thrivers were significantly more recent survivors of cancer than Recoverers).
c
p=.00 (Thrivers had significantly higher social support than Sufferers and Recoverers).
d
p=.022 (Sufferers had significantly lower physical functioning than Recoverers).
e
p=.026 (Thrivers had significantly higher mental functioning than Sufferers).
DISCUSSION
The primary goal of the current study was to identify a typology of posttraumatic impact
in young adult cancer survivors. Second, the emergent posttraumatic impact typologies classified
young adult cancer survivors based on their levels of PTS and PTG and related these clusters to
demographic, medical, social support, and physical and mental health characteristics. Overall, the
current study found three posttraumatic impact profiles that represented a large sample of young
adult cancer survivors. Thrivers reflected survivors who grew the most with their traumatic
distress at a medium level. Sufferers indicated survivors who suffered the highest level of distress
with the level of growth at medium. Recoverers were at the lowest levels in both PTS and PTG.
The typological profiles reflect the conceptualizations of PTG theories that low levels of
distress are associated with minimal growth, moderate to high distress are associated with

41

maximal growth, and extreme distress is associated with poor adaptation (Levine, et al., 2008;
Schnurr, et al., 1993; Solomon & Dekel, 2007; R. Tedeschi & Calhoun, 1996). PTG subscales did
not behave differently within the same cluster, although Personal Strength and Appreciation of
Life subscales were higher than the other subscale scores in all the clusters.
In the current study, there was no difference across the cluster in demographic variables
such as gender, employment, income, education, and current age. The absence of a significant
relationship between gender and cluster membership was particularly interesting because gender
was generally found as an important variable in PTS. Women tend to be more susceptible than
men to distress (Fife, Kennedy, & Robinson, 1994), and being female is found to be a risk factor
for higher PTS in some studies (Langeveld, Grootenhuis, Voute, & de Haan, 2004; M. Stuber et
al., 1997). On the other hand, women also tend to be more open to the discovery of positive
meaning than men (Fife, et al., 1994). However, Stanton, Bower, and Low (2006) conclude from
their review that the existing literature strongly suggests male and female cancer patients do not
differ in the extent to which PTG is experienced. Based on the findings of this study, although
women are thought to be susceptible to distress symptoms, their combined posttraumatic impact
of PTS and PTG do not seem to be different from that of men.
Age at diagnosis has not been examined as a correlate of PTG in young adult survivors of
childhood cancer. There are mixed findings about the relationships between current age and PTS:
Some studies fail to support age difference in PTS (A. Kazak et al., 1997; Landolt, Vollrath, Ribi,
Gnehm, & Sennhauser, 2003; Pelcovitz et al., 1996).; but Hobbie and others (2000) find in a
study that young adult survivors (mean age=25) have higher PTSD levels than child and
adolescent survivors (ages 8-18). In the current study, age at diagnosis was a significant factor
that differentiated the survivor profiles: Recoverers were diagnosed with cancer at the youngest

42

ages; Sufferers at medium; and Thrivers at the oldest ages, although the difference between
Thrivers and Sufferers was not statistically significant.
In the general trauma research, PTS gradually disappears in the ensuing months
following the trauma (A. Ehlers & Clark, 2000; Kessler, Sonnega, Bromet, Hughes, & Nelson,
1995; Perrin, Smith, & Yule, 2000). But, in childhood cancer survivors, the passage of time since
cancer diagnosis fails to reliably correlate with cancer-related PTS (Brown, Madan-Swain, &
Lambert, 2003; A. Kazak, et al., 1997; Langeveld, et al., 2004; Pelcovitz, et al., 1996). In regards
to PTG, the review by Stanton, Bower, and Low (2006) concludes that the relationship between
PTG and the passage of time since diagnosis may be more positive and stronger in the one or two
years following diagnosis and treatment than the relationship after several years of survivorship
(Stanton, et al., 2006).
In the current study, in which PTS and PTG were considered together, Thrivers were the
most recent survivors and Recoverers were the most long-term survivors, with Sufferers in
between these other two groups. It can be theorized that, along the passage of time in cancer
survivorship, survivors might feel high growth at first despite or due to some perceived trauma,
and then the PTG level might decrease while the PTS level is maintained to some extent.
Although PTS is found to decrease with time in survivors of other traumas, it might be different
in cancer survivorship because its distinct characteristics such as fears of recurrence of the cancer
and difficulty of establishing the termination of the trauma (Sumalla, et al., 2009).
Social support was found to be a significant factor affecting posttraumatic profiles among
young adult cancer survivors. Such importance of social support in survivorship is consistent with
the existing literature. PTS is related to poor general social support (Bruce, 2006), less family
support (Brown, et al., 2003), less family satisfaction and communication (A. Kazak, et al., 1997),

43

perceived social constraint and lack of social network (S Manne, Duhamel, & Redd, 2000), and
living alone (Lee & Santacroce, 2007). Correlations are found between PTG and social support,
especially in specific social areas, such as marital support and contact with a role model who had
experienced PTG (Weiss, 2004), and talking about cancer with others (M. Cordova, et al., 2001).
In the present study, Thrivers had the greatest social support and Recoverers had the lowest social
support. It is theorized that young adult survivors, despite some perceived trauma, seems to feel
high growth from their cancer experience when they have social support available.
Physical health was the greatest in Recoverers, medium in Thrivers Group; and lowest in
Sufferers. The important role of PTS in physical health is consistent with the existing literature.
PTSD is generally related with morbidity (Hobbie, et al., 2000) and poor quality of life (Schwartz
& Drotar, 2006) in the cancer survivor population. The point that Sufferers has a medium level of
PTG is important because previous studies found it counterintuitive that PTG did not correlate
with physical health (Sears, Stanton, & Danoff-Burg, 2003) or was related with poor quality of
life (Tomich & Helgeson, 2004). Survivors, despite some perceived growth, might still have poor
physical health because growth tends to come together with traumatic experiences. PTS seems to
affect physical health at a greater level than PTG does.
Mental health was also a significant correlate of the cluster membership in this study with
the greatest level in Thrivers, medium in Recovers, and lowest in Sufferers. Some studies in the
existing literature did not find significant relationships between PTG and mental health (M.
Cordova, et al., 2001; Fromm, et al., 1996; Sears, et al., 2003). But, when the current study
compared the mental health difference across the survivor profiles based on both PTG and PTS,
mental health was found to be an important correlate and Thrivers, despite their medium level of

44

PTS, had significantly higher mental health than Sufferers. PTG seems to affect mental health
more than PTS does.
Limitations of this study need to be considered in interpreting its findings. Although the
differences of physical and mental health across the survivor profiles were statistically significant,
they were not large enough to be clinically meaningful. Perhaps because the sample is composed
of long-term survivors, the impact of trauma might have dissipated over the time of the long
survivorship. The passage of time since diagnosis is an important factor for PTS and PTG, so a
sample with more diversity in PTS levels and the length of survivorship might have greater
quality of life differences across survivor profiles, enough to be clinically meaningful.
Longitudinal studies that follow survivors through long term survivorship since the termination of
treatment would be valuable in understanding whether and how PTS and PTG levels change over
time. Although LPA is a model-based analysis, the survivor profiles in this study emerged from
this particular sample. Thus, the findings might be limited in generalization to other young adult
cancer survivors.

45

This study was the first in an effort to classify young adult cancer survivors into profiles
considering both PTS and PTG together. It was hypothesized that the two posttraumatic
phenomena interact with each other, affecting individual survivors at the same time, maybe in
different levels and/or different aspects. This theory was confirmed with the finding that the two
phenomena have curvilinear relationships and some survivors had both PTS and PTG. Age at
diagnosis, length of survivorship, social support were important correlates and PTS affected
physical health and PTG affected mental health. The study findings are meaningful in that it
found the complex cancer survivorship experience, which has been qualitatively reported, in a
quantitative manner. The multi-faceted impact of cancer is not new to the field of cancer
survivorship studies, but it is meaningful that it was found in this study through quantitative
analyses.

46

CHAPTER 4. BUFFERING ROLES OF POST-TRAUMATIC
GROWTH
INTRODUCTION
Post-Traumatic Stress Disorder (PTSD) is a clinical anxiety disorder that occurs
following an intensely threatening, traumatic event and childhood cancer experience is considered
to be potentially traumatic enough to result in a diagnosis of PTSD (American Psychiatric
Association, 1994). A PTSD diagnosis requires the following criteria: criterion A (exposure to
events threatening life or bodily integrity of self or loved one); criterion B (at least 1
reexperiencing symptom); criterion C (at least 3 avoidance symptoms); criteria D (at least 2
arousal symptoms); criterion E (more than 30 days after the event); and criterion F (significant
distress or functional impairment). A recent study (M. L. Stuber et al., 2010) finds that young
adult cancer survivors have 4 times higher rates of PTSD than a comparison group of siblings and
that the prevalence of PTSD (9%) in the young adult survivors is higher than in other age groups.
There seem to be substantial rates of sub-clinical Post-Traumatic Stress Symptoms
(PTSS), although not fully diagnosed, among cancer survivors. Kazak and others (2001) found
that 50% of survivors fulfilled the re-experiencing criteria and 29% fulfilled the arousal criteria
according to the PTSD diagnosis criteria. In another study by Erickson and Steiner (2001), 78%
of the sample met at least one symptom of reexperiencing, arousal, or avoidance at a functionally
significant level. The high prevalence of Post-Traumatic Stress (PTS) among childhood cancer
survivors is alarming because PTS is found to be related to lower quality of life, depression
(Schwartz & Drotar, 2006), and morbidity (Hobbie, et al., 2000). It is important to understand the

47

mechanism of how PTS is related to psychological distress and what can alleviate the negative
effects.
As a theory of explaining how traumas affect people’s adjustment and quality of life,
Janoff-Bulman (1992) conceptualizes that traumas may shatter deeply held and probably
unexamined assumptions about how the survivors believe the world and themselves to be, destroy
their protective illusions, and force survivors to confront their own vulnerability and fragility (R.
Janoff-Bulman, & Frieze, J. H., 1983). Amidst the chaos of vulnerability, survivors might fall
into mental defeat, which is defined as the “perceived loss of all autonomy, a state of giving up in
one’s own mind all efforts to retain one’s identity as a human being with a will of one’s own (A.
Ehlers, Maercher, A., & Boos, A., 2000).” Such a psychological state might be manifested as
depression.
Recovering from the sense of defeat and disorientation is essential for the survivor to
adjust and function well. Janoff-Bulman (1992) theorizes that survivors recognize shattered past
assumptions and reconstruct new assumptions, and that the process might produce psychological
pains and symptoms of PTSD, such as reexperiencing the traumatic experiences. Calhoun and
Tedeschi (2006b) also emphasizes the importance of cognitive and reflective rumination, which
tend to repair, restructure, or rebuild the survivor’s way of understanding the world, despite the
accompanying psychological pains. With such deliberate reflection, survivors can not only
merely survive the traumatic event, but also recognize some other possibilities that become Post-
Traumatic Growth (PTG), a phenomenon that allows survivors to grow in many aspects of life
after experiencing a traumatic event (L. Calhoun & R. Tedeschi, 2006).
Reconstructing new assumptions in the survivors’ world is important to improve their
quality life in the new reality. Quality of life is a broad and elusive concept, but as suggested by

48

Calman (1984), it might be how satisfied we are about our lives and can be measured by
examining the difference between our hopes/expectations and our present experiences. In this
sense, quality of life is very personal experience and the person’s perception is important
(Calman, 1984). According to this theory, how new hope and expectations are constructed and
how present experiences are perceived determine the survivor’s quality of life. PTG might affect
the process and the outcome by helping to modify their expectations and to reinterpret the current
status, thus narrowing the gap between their perceived current status and expectation.
PTG might be a potential buffer against the adverse impact of PTS on psychological
distress. PTG may also reflect cognitive adaptation in response to a cancer diagnosis that can
alter the global meaning of the cancer experience (Helgeson, Reynolds, & Tomich, 2006) and
positively affect psychological well-being. But, empirically, there are mixed findings in the
relationship between PTG and adjustment outcomes. On the other hand, Maercker and Zoellner
(2006), in their review, state that all longitudinal studies to date show positive relationships
between perceived growth and adjustment. In other studies (Cadell, Regehr, & Hemsworth, 2003;
Tomich & Helgeson, 2004), negative relationships were found between PTG and measures of
general well-being and distress. These mixed and counterintuitive findings might be because
experiences of trauma may indeed produce growth, but also tend to generate severe pain. As
Chapter 2 in this dissertation found, the two post-traumatic constructs, PTS and PTG, seem to
coexist in some survivors. Due to such complex inter-relationships between PTS and PTG and
their potential coexistence in some survivors, the roles of PTG might have been undetected in
some studies.
Although limited, there is empirical evidence that supports the moderating effect of PTG
in the relationship between PTS and quality of life. Morrill, Brewer, O’Neill, Lillie, Dees, Carey,

49

and Rimer (2008) found a significant moderating effect of PTG, whereby higher PTS is related
with a lower quality of life, and the magnitude of this relationship is smaller among the group
with the higher PTG than among the group with the lower PTG. Although the study takes
meaningful initiative and produces novel findings, its limitations include a relatively small
(N=161) and homogeneous sample, especially in that it includes only women breast cancer
survivors between the ages of 36 and 87 with a mean of 59 years. Also, the study acknowledges
that measurement errors are not addressed.
The present study aims to further examine the roles of PTG in moderating the negative
impact of PTS on psychological distress among young adult cancer survivors. It hypothesizes that
although PTS level is related to distress, the impact decreases as PTG level increases.
METHODS
Sample and procedure
Subjects were recruited from four medical centers that treat pediatric oncology patients
(University of California Los Angeles Mattel Children’s Hospital, the City of Hope National
Medical Center, Los Angeles Children’s Hospital, and University of Michigan Mott Children’s
Hospital). When considering the demographics of the catchment areas for Children's Hospital
Los Angeles, City of Hope National Medical Center and University of Michigan, there are some
demographic differences. Almost 50% of the patient population seen at CHLA is Hispanic/Latino.
City of Hope has a demographic that reflects the nation as a whole, as patients come from all over
the country. University of Michigan draws from across the state, a predominantly Caucasian
population, but representative of urban and rural areas. This strategy of including several
treatment centers throughout the Los Angeles region and in Michigan enhances the ability to

50

achieve greater variation in patient characteristics and experiences, thereby increasing the
generalizability of the findings. Inclusion criteria included current age 18-39; age at diagnosis 0-
21 years (with diagnosis and treatment conducted within a pediatric setting); and disease-free
status. Exclusion criteria included currently receiving treatment, relapse or second cancer.
Each of the respective participating institutions sent out an introductory letter on
letterhead describing the project, outlining the eligibility criteria, and inviting survivors to
participate. Enclosed was a postage-paid response form for potential participants to return
indicating whether they accepted or declined participation and, if they declined, their reasons for
non-participation. Survivors who indicated that they were willing to participate were mailed the
survey. Included was an informed consent form reiterating the purpose of the study, describing
the survey content and the length of time it would take to complete, and explaining the voluntary
and confidential nature of the survey. Participants were asked to sign the informed consent and to
complete and return it with their surveys in a prepaid envelope within one week.
A log was kept indicating when surveys were mailed and returned. Returned surveys
were reviewed for completeness. Missing data or items needing clarification were noted and a
staff research assistant contacted the respondents, if possible, to obtain the missing data. Those
failing to return their surveys within two weeks were telephoned to determine the status of the
survey and whether they intended to complete the survey. Those who did not wish to participate
were asked their reasons for non-participation, which were recorded. For those still willing to
participate, an additional week was given for completion of the survey and they were re-contacted
in two weeks if the survey had not been received. A third and final opportunity was given for
completion of the survey, and those who did not return the survey were considered non-

51

responders. A comprehensive survey was sent out to 1,500 survivors and responses were obtained
from 666 individuals (44%).
Of the 666 individuals, 48 did not meet the inclusion criteria in age and age at diagnosis, so
they were removed for analysis. After removing these cases, the total sample size was 618. For
this study, 593 who had both PTSD and PTG scores were included.
Measures
The Post-Traumatic Stress Diagnostic Scale (PDS). The PDS was developed and
validated by Foa (1997) to provide a brief but reliable self-report measure of PTSD for use in
both clinical and research settings. Using a four-point scale, respondents rate 17 items
representing the cardinal symptoms of PTSD experienced in the past 30 days, such as
reexperiencing, avoidance, and hyperarousal. The PDS has high face validity because items
directly reflect the experience of PTSD with high internal consistency (coefficient alpha of 0.92).
Test–retest reliability was also highly satisfactory for a diagnosis of PTSD over a 2- to 3-week
period (kappa 0.74). Test–retest using the symptom severity scores yielded a highly significant
correlation (0.83). Analysis also revealed an 82% agreement between diagnosis using the PDS
and the Structured Clinical Interview for DSM (Foa, et al., 1997). The scale was validated on
samples aged 18–65 and has been used in a wide range of clinical and research contexts with a
high degree of confidence when use of a structured clinical interview is impractical (Foa, et al.,
1997). The PDS has been used in prospective treatment studies helping establish a role for
cognitive behavioral therapy in those with established PTSD (Duffy, et al., 2007). Recently, the
PDS has been employed in diagnosing PTSD of the patients in emergency services (Haslam &
Mallon, 2003).

52

Based on the DSM-IV diagnosis criteria, a PTSD diagnosis requires at least 1 symptom
in the reexperiencing items, at least 3 symptoms in the avoidance items, and at least 2 symptoms
in the arousal items (Foa, et al., 1997; M. L. Stuber, et al., 2010). In this study, the survivors were
categorized into two groups, depending on whether one meets the PTSD diagnosis criteria B, C,
and D. Those who meet all the three criteria were categorized as the Higher PTS group (1) and
those who do not as the Lower PTS group (0). The dichotomous variable was used as an
independent variable in the analyses.
The Post-Traumatic Growth Inventory (PTGI). The 21-item scale measures the
degree of the positive changes experienced in the aftermath of a traumatic event. Each item is
rated using a 6-point Likert scale, with values ranging from 0 (I did not experience this change as
a result of my crisis) to 5 (I experienced this change to a very great degree as a result of my
crisis). The PTGI consists of five subscales: Relating to Others (seven items), New Possibilities
(five items), Personal Strength (four items), Spiritual Change (two items), and Appreciation of
Life (three items). Internal consistency for the total score and subscales of the PTGI has been
reported as satisfactory (α coefficient for the total scale=.90, Relating to Others (RO)=.85, New
Possibilities (NO) =.84, Personal Strength (PS) =.72, Spiritual Change (SC) =.85, and
Appreciation of Life (AL) =.67), and the test–retest reliability (.71) over 2 months has also been
reported based on the sample of university students in the original study (R. Tedeschi & Calhoun,
1996). In addition, the concurrent, discriminant, and construct validity of the PTGI were
examined by assessing the correlations among the PTGI, social desirability, and personality
variables, and by comparing those who had experienced severe trauma and those who had not.

53

For the analyses of the present study, the 5 subscale scores (RO, NP, PS, SC, and AL)
were used as indicators of a latent variable, PTG. The latent variable was used as a moderating
variable in the analyses.
Brief Symptom Inventory (BSI). Psychological distress was assessed using the BSI,
an 18-item self-reported measure of psychological symptoms with three sub-scales: somatization
(6 items), depression (six items), and anxiety (6 items) (Derogatis, 2000). This instrument, which
has widespread use in oncology settings, was selected as the best measure for the normative
population of adult oncology patients. In this tool, symptoms are rated on a five-point Likert scale,
with values ranging from 0 (not at all) to 4 (extremely) for each symptom experienced over the
past seven days. For this study, subscale scores of depression, somatization, and anxiety were
used as indicators for a latent variable, psychological distress. The latent variable was used as a
dependent variable in the analyses.
Analysis
First, descriptive analyses were conducted to examine the characteristics of major
demographic, medical, PTS, PTG, and distress variables in the sample. Then, to examine the roles
of PTG in the impact of PTS on psychological distress, Mplus 5 was used. The maximum-
likelihood (ML) estimation was used to treat the missing data. This method partitions the cases
into subsets with the same patterns of missing observations, so all available statistical information
is extracted from each subset, and all cases are retained in the analysis (Kline, 2005).
There are multiple methods to test moderation effects in Structural Equation Modeling
and this study used the method embedded in Mplus which is based on the Latent Moderated
Structural Equations method introduced in the paper by Klein and Moosbrugger (2000).
Interaction terms and quadratic interaction terms between latent variables can be set up by using

54

the XWITH command in Mplus. Three alternative hypothesized models were tested: (Model 1)
PTS and/or PTG are only directly related to psychological distress without the moderating effect
of PTG; (Model 2) PTG has a linear moderation effect on the relationship between PTS and
psychological distress; and (Model 3) PTG has a quadratic moderation effect on the relationship
between PTS and psychological distress.
Model 1, the no interaction model, includes an observed exogenous variable, PTS; a
latent exogenous variable, PTG; a latent endogenous variable, DISTRESS, and a disturbance term.
The latent exogenous variable, PTG, is measured by five observed indicators (RO, NP, PS, SC,
and AL), and the endogenous variable, DISTRESS, is measured by three observed indicators
(DEP, SOM, and ANX).
In Model 2, the path between the endogenous variable, DISTRESS, and a latent linear
interaction term, INT, was added to Model 1. INT is an interaction term of an observed
exogenous variable, PTS; and a latent exogenous variable, PTG. If the linear interaction term is
significant, having higher PTS may be related to more distress for low-PTG survivors, but it may
not be related to more distress or may be related to less distress for high-PTG survivors (Baron &
Kenny, 1986).
In Model 3, the paths between the endogenous variable, DISTRESS, and the latent
quadratic term, PTG
2
; and the latent quadratic interaction term, QUADINT were further added to
Model 2. Model 3 includes direct main effects, a linear moderation effect, and a quadratic
interaction effect. While the linear interaction represents a gradual, steady change in the effect of
the independent variable on the dependent variable as the moderator changes, the quadratic
interaction represents an increase in the effect of the independent variable on the dependent
variable as the moderator changes (Baron & Kenny, 1986). More specifically, if the quadratic

55

interaction is significant in this model, it can be interpreted that having higher PTS may be related
to more distress for all low-PTG survivors, but as PTG level increases, the impact of PTS loses
its impact on distress. Model 3 are graphically presented in Figure 2. Models 2 and 3 are the
hypothesized models based on the theory that the adverse impact of PTS is buffered by PTG.

56

Figure 2
Graphic presentation of Model 3

The three models were evaluated by examining several criteria. Both the Akaike
Information Criterion (AIC; Akaike, 1974) ) and the sample size-adjusted Bayesian Information
Criterion (BIC; Schwarz, 1978) were examined to choose the better-fitting model. Optimal model
fit is defined by lower AIC and BIC values (Kline, 2005). Log likelihood values were used to

57

compute chisquare difference statistics
1
(Satorra & Saris, 1985; UCLA: Academic Technology
Services), which tested the statistical significance of the improvement in fit as paths were added
(Kline, 2005; Moosbrugger, Schermelleh-Engel, Kelava, & Kelin, 2009). The moderation effects
of PTG were graphically presented to clearly show the relationships. Although this study used
latent variables, total scores of PTG and BSI were used for graphical simplicity.
RESULTS
Descriptive statistics
On average, the participating survivors were diagnosed about 16 years ago at about 11
years of age, and the average age at the time of study enrollment was around 27 years old. There
were slightly more women (n=314, 53.1%) than men (n=277, 46.9%). Most of the participants
were non-Hispanic White (n=379, 64.5%) or Latino (n=136, 23.1%). Descriptive statistics of the
sample characteristics in subgroups with higher or lower PTS are listed in Table 7.

1
For the Mplus Maximum Likelihood Robust estimation, there is an additional test for nested models. This
test compares the log-likelihoods for the null and alternative models rather than the chi-squared values. A
list of the information needed is as follows. L0 = log-likelihood for the null model; L1 = log-likelihood for
the alternative model; c0 = scaling correction factor for the null model; c1 = scaling correction factor for
the alternative model; p0 = number of parameters estimated in the null model; p1 = number of
parameters estimated in the alternative model. From this information the tests statistic TRd can be
calculated, along with the degrees of freedom: cd = (p0*c0-p1*c1)/(p0-p1); TRd = -2*(L0-L1)/cd; df = p1-
p0 (Satorra & Saris, 1985; UCLA: Academic Technology Services)

58

Table 7
Sample characteristics by PTS levels
All (n=593) Higher PTS
(n=245)
Lower PTS
(n=348)
Variables n % n % n %
Sex Female 314 53.1 129 52.9 185 53.3
Male 277 46.9 115 47.1 162 46.7
Race White 379 64.5 154 63.4 225 65.2
Non-White 209 35.5 89 36.6 120 34.8
Income $25,000 or below 204 35.4 91 38.2 113 33.3
Above $25,000 373 64.6 147 61.8 226 66.7
Education High school and below 112 18.9 50 20.6 62 18.1
Some college and A.A. degree 275 46.4 112 46.1 163 47.5
College graduate or above 199 33.6 81 33.3 118 34.4
Employment Unemployed 75 12.6 42 17.1 33 9.6
Status Employed 512 86.3 202 82.8 310 90.4
Type of cancer Hematological 371 62.6 158 64.5 213 61.2
CNS or brain tumor 71 12.0 24 9.8 47 13.5
Solid or soft tissue tumors 151 25.5 63 25.7 88 25.3
Mean SD Mean SD Mean SD
Age at diagnosis (0-21) 11.22 6.01 11.25 6.19 11.20 5.89
Age (18-39) 26.94 5.43 26.74 5.20 27.09 5.59
Years since diagnosis (2-37) 15.71 6.99 15.53 7.35 15.85 6.74

59

Moderation effects
Table 8 presents the model comparison results. Model 1 tests no interaction effect; Model
2 tests the linear interaction effect; and Model 3 tests the quadratic interaction effect. With the
lowest AIC and BIC fit indices, Model 3 was the best fitting model among the three. The log
likelihood difference tests indicate that Model 3 was significantly better than both Model 1 and
Model 2 at p <.001.

Table 8
Comparison of models
Model Log
Likelihood
Scaling
Correction
Factor
df AIC
a
BIC
b
Δ χ
2
/ Δdf
c

Model 1 -4626.394 1.262 26 9304.788 9336.261 -
Model 2 -4625.433 1.260 27 9304.867 9337.550 1.59/1
Model 3 -4611.733 1.224 29 9281.466 9316.571 15.44/2**
d

32.77/3**
e

a
AIC: Akaike Informaiton Criterion
b
BIC: sample size-adjusted Bayesian Information Criterion
c
Chisquare difference statistics were used to measure the significance of the difference between
two SEM models for the same data, one of which is a nested subset of the other.
d
Chisquare difference statistics between Model 3 and Model 2
e
Chisquare difference statistics between Model 3 and Model 1

60

In Model 3, the best-fitting model, the quadratic interaction (B=-21.33, p=.00), the linear
interaction (B=-3.80, p=.024), and the main effect of PTS (B=4.20, p=.00) were significant,
whereas the main effect of PTG was not significantly related to distress. The main effect of PTS
on distress was significant in all three models, and its unstandardized coefficient was the highest
in Model 3 (B=4.2), compared to Model 1 (B=2.76) and Model 2 (B=2.76). The main effect of
PTG on distress was not significant in all three models. The linear interaction term was not
significant in Model 2 when the quadratic interaction term was not included. But, when the
quadratic term was added in Model 3, both the linear interaction (B=-3.8, p<.05) and the
quadratic interaction (B=-21.33, p<.001) were significant. Coefficient of all the paths are
described in Table 9.

61

Table 9
Coefficients in 3 models
Variable Model1 Model 2 Model 3
Factor loadings
PTG→RO 1.00 (.00)
PTG→NP .99 (.02)**
PTG→PS .99 (.02)**
PTG→SC 2.57 (.06)**
PTG→AL 2.32 (.07)**
DISTRESS→DEP 1.00 (.00)
DISTRESS→SOM .68 (.05)**
DISTRESS→ANX .88 (.05)**
Direct Relations
PTS→DISTRESS 2.76 (.42)** 2.76 (.42)** 4.20 (.60)**
PTG→DISTRESS -.85 (.82) .13 (.74) .07 (.78)
INT→ DISTRESS -2.12 (1.68) -3.80 (1.69)*
PTG2→ DISTRESS -.34 (2.60)
QUADINT → DISTRESS -21.33 (5.65)**
Unstandardized and standardized (in parentheses) coefficients are presented.
*p<.05; **p<.001.

62

Latent interactions cannot be graphically represented, but observed variables were used in
Figure 3 as an attempt to visualize the significant interactions, PTG average scores were used for
grouping the survivors into three levels: The Low PTG group (n=104) is 1 standard deviation (1.2)
or more below the mean (2.74); the High PTG group (n=385) is 1 standard deviation or more
above the mean; the Mid PTG group (n=104) is in-between. In the Y axis, the total BSI scores
were used. In the X axis, the dichotomous variable, PTS Level, with two values of low PTS and
high PTS was used. Overall, the High PTS survivors were higher in distress and the Low PTS
survivors were lower in distress. But, the slopes of the lines decrease, which implies a decreasing
impact of PTS on distress, as the PTG level increases – a linear interaction. The quadratic
interaction is impossible to be graphically presented, but it may be conceptually depicted as in
Figure 4. As the level of PTG increases, its interaction effect seems to increase.

63

Figure 3
Moderating effects of PTG

Figure 4
Conceptual model of quadratic interaction effects of PTG

PTG
Interaction
effect

64

DISCUSSION
This study examined the potential positive roles of PTG in the relationship between PTS
and distress using Structural Equation Modeling methods. The significant study findings are
meaningful both in practice and in theory development. As one of the first studies that found the
moderating effects of PTG in cancer survivors, the study adds to the empirical support for the
utilitarian values of PTG in survivorship. The moderating roles of PTG were found in older
women breast cancer survivors (Morrill, et al., 2008). The present study reconfirmed the positive
roles in young adult survivors of childhood cancer.
Adolescence and young adulthood are stages of critical developmental transitions from
childhood to adulthood. They face life issues related to developing their self identities, seeking
independence, and establishing relationships (Eiser & Kuperberg, 2007). Going through these
confusing and critical times, young adult survivors of childhood cancer also experience diagnosis,
treatment, and survivorship traumas. The higher PTSD prevalence in young adult survivors than
in other age groups (M. L. Stuber, et al., 2010) might be due to one or more of the following: the
dual burdens of trying to manage the heavy developmental milestones and cancer traumas are
extreme; the assumptions (R. Janoff-Bulman, 1992) of a protective and safe world are shattered;
and the experience of a vulnerable self affects the identity developing process of survivors.
Finding the possibility in the current study that PTG may moderate the negative impact of PTS on
distress in young adulthood provides potential intervention ideas to help young adult survivors
improve their psychological well-being.
There are doubts about the construct of PTG and its usefulness for trauma survivors’
quality of life. At the center of the discussion about the construct of PTG is whether PTG really
happens and the individual grows in the actual sense; or whether PTG is an illusion intentionally

65

used as a coping strategy by the survivor to deal with the current adversity (Sumalla, et al., 2009).
In Constructivist philosophy, reality is constructed based on the individual’s perceptions and
worldviews, as articulated by Thomas (1928, p. 572): “What is defined or perceived by people as
real is real in its consequences.” Theoretically, then, perceptional reframing, or the world viewed
from the PTG mind, is a reality. Moreover, the findings of this study add empirical evidence to
theoretical reasoning by providing the real and positive roles of PTG in reducing the negative
impact of PTS on distress.
The relationships between PTG and psychological well-being and other aspects of quality
of life have been undiscovered or negatively reported, which contributed to the suspected lack of
a utilitarian role of PTG (L. G. Calhoun & R. G. Tedeschi, 2006b). This might result from
methodological problems in research. Unlike mixed and counterintuitive findings in cross-
sectional studies (Cadell, et al., 2003; Tomich & Helgeson, 2004), most longitudinal studies find
positive relationships between PTG and quality of life outcomes (Zoellner & Maercker, 2006).
The stronger impact of PTG in longitudinal studies might imply that PTG creates “psychological
preparedness” for the survivor to adjust to the reality and prepare for other life adversities (L. G.
Calhoun & R. G. Tedeschi, 2006b). The present study examined a quadratic interaction because it
is encouraged to test whether an interaction is of a different functional form such as a quadratic
moderation effect (Baron & Kenny, 1986; Osborne, 2008). Interaction effects, both linear and
quadratic, were significant when a quadratic interaction was added in the model. The findings
might be due to the curvilinear relationship between PTS and PTG, as found in Chapter 1 of this
dissertation. Different and advanced methodological approaches might help discover the hidden
relationships between PTG and adjustment variables in cancer survivorship in future studies.

66

Due to the cross-sectional design of the study, causal relations among the study variables
cannot be inferred. It is possible that psychological distress affects PTG and/or PTS, or that other
variables that were not included are the causes for the studied constructs. Replication of the
moderating effects with other samples and longitudinal studies would establish causality and
promote generalizability of the findings. Since we are all interested in finding how to help cancer
survivors enjoy happy lives, it is encouraging to find a potential moderator of distress, PTG. The
next question is how to promote it. To answer this question, future studies should aim to
understand the mechanism of PTG; for example, what demographic, medical, and sociocultural
variables are related to PTG level, and how PTG affects well-being (e.g. if other variables, such
as optimism and self-esteem, mediate the relationship.) Intervention studies examining the
effects of potentially PTG-promoting programs, such as narrative writing and social support
groups, would contribute to providing services for cancer survivors.

67

CHAPTER 5. CONCLUSION
The dissertation aimed to examine the interrelationships between Post-Traumatic Stress
(PTS) and Post-Traumatic Growth (PTG), consequences of these phenomena in young adult
cancer survivorship, and the roles of PTG on distress.
First, the study findings support the idea that PTG may coexist with PTS in cancer
survivors. To date, research has yielded mixed results regarding the relationship between PTG
and PTS. The findings from Study 1 in this dissertation were also mixed, but dominantly not-
significant relationships between PTG and PTS. PTG seem to be an independent construct from
PTS, but one that coexists with PTS. Although it could be argued that the severity of distress
endorsed by the young adult cancer survivors in the present study limited possible growth
outcomes, these survivors did report, on average, moderate levels of growth as a result of their
trauma. The overall presence of some levels of symptomatology (i.e., 41.3% of the sample met
criteria B, C, D of the DSM –IV PTSD diagnosis) and moderate reports of growth in the whole
sample provide further evidence that growth and symptom severity may not be directly related,
but do coexist.
A majority of the existing literature focuses either on PTS or on PTG, based on the
assumption that these two posttraumatic phenomena are at the opposite ends of the quality of life
continuum; those with high PTS adjust poorly and those with high PTG adjust well. Future
research needs to change this research framework in order to incorporate these two constructs
concurrently in research. Attempting to address this research gap, Study 2 of this dissertation
aims to understand the posttraumatic profiles of the young adult cancer survivors by clustering
them based on combinations of PTS and PTG. The Thrivers and Sufferers profiles add evidence
to the coexistence of PTS and PTG. The Thrivers grew the most from cancer and their mental

68

health was the highest in the sample, but they still had some levels of posttraumatic stress. The
Sufferers experienced the highest levels of posttraumatic symptoms and their physical health was
worst in the sample, but they still had some levels of posttraumatic growth. These findings, based
on data-driven analysis methods, need to be replicated in other samples of cancer survivors.
If PTS and PTG coexist, do they affect each other in a particular way? That was the
question that I originally posed when I hypothesized their quadratic relationships. Although this
hypothesis was not proven by Study 1 of this dissertation, the possible quadratic relations
between the two constructs are worth further examination in future research. Although not
statistically significant, the relationships between the two phenomena showed quadratic patterns.
Moreover, the three profiles produced in Study 2 indicate that the level of PTS might affect the
level of PTG in the survivors; the PTS levels were lowest in Recoverers, medium in Thrivers, and
highest in Sufferers, while the PTG levels were lowest in Recoverers, highest in Thrivers, and
medium in Sufferers. More study on this topic is needed before definitive conclusions can be
drawn.
Clinically, the possible coexistence of positive and negative posttraumatic phenomena
has important implications. Clinicians should consider both aspects when seeing cancer survivors;
those who report experiencing growth from cancer might have hidden posttraumatic stress
symptoms and those who manifest distress might have potential to discover growth experiences,
as found in the Thrivers and Sufferers in Study 2 of this dissertation.
Second, the study findings provide insights into the question, “What good is PTG?”
Because PTG is theorized to coexist with, although potentially independently from PTS, the
utilitarian purpose and meaning of PTG has been doubted. If PTG may not necessarily be
accompanied by greater well-being and less distress, it may appear to be neither helpful nor

69

useful. Calhoun and Tedeschi (2006b) emphasize, though, that PTG may still help the survivors
live, at least from their perspectives, “fuller, richer, and perhaps more meaningful lives.” Since
perceptions easily become the targets of suspicion, PTG has been the topic of discussion about
whether it is real or illusory (Sumalla, et al., 2009).
Study 2’s findings in this dissertation indicate the possible influence of PTG despite its
coexisting PTS. Although the Thrivers had some levels of posttraumatic stress symptoms, they
had significantly better mental health than either Recoverers or Sufferers. The findings from
Study 3 reconfirm the positive roles of PTG by showing the moderating and buffering effect of
PTG on the negative association between PTS and psychological distress. The moderating effect
seems to increase as the PTG levels increases. In other words, PTG seems to serve the function of
reducing stress and improving the quality of life in young adult cancer survivors.
These findings strongly support the utility of PTG. The significant buffering effect of
PTG evidenced in this dissertation has important research and clinical implications. In studies
examining the relationships between PTG and health outcomes, PTG has often been found not to
be associated with positive physical and mental health outcomes (Bower, et al., 2005; Fromm, et
al., 1996; S. Manne, et al., 2004). This is perplexing, because PTG is supposed to be linked to
good outcomes. The lack of relationship between PTG and positive outcomes might result more
from methodological issues than from its often-alleged illusory nature. By examining survivor
profiles based on both constructs in Study 2, the positive effect of PTG on mental health stood out.
PTG was also found to buffer the negative influence of PTS on psychological distress in Study 3.
To capture their complex interrelationships including, but potentially not limited to, independent
but concurrent, and interactive relationships, advanced and innovative methodological and
theoretical approaches may be warranted.

70

Clinically, positive changes may be used as foundations for further therapeutic work,
providing hope that the trauma can be overcome (Calhoun & Tedeschi, 1999). The facilitation of
PTG may be considered a legitimate therapeutic aim (Linley & Joseph, 2002). Such therapeutic
inclusion of the PTG notion needs caution, nonetheless, because cancer patients often complain
of the "prison of positive thinking" as they are encouraged by others to look at the bright side or
to keep a good attitude (Spiegel, 1993; Spiegel & Classen, 2000). With this caveat, it might make
a difference if clinicians appreciate the concept of PTG and its potentially positive roles in the
quality of life of cancer survivors, so they validate the growth experiences shared by survivors
and witness them discover the meanings of traumatic experiences as expert companions.
Future research should continue to examine the mechanism and consequences of PTG
and investigate how to promote PTG. Some of the research questions could include what
demographic, medical, sociocultural variables are related to PTG level, and how PTG affects
well-being (e.g. if other variables, such as optimism and self-esteem, mediate the relationship.)
Intervention studies examining the effects of potentially PTG-promoting programs, such as
narrative writing and social support groups, would contribute to providing services for cancer
survivors.
Third, the study findings add to the knowledge about young adult cancer survivors, a
unique and understudied subset of survivors who have experienced a childhood trauma. As found
in Study 1 and Study 3, the coexistence of PTS and PTG and the buffering roles of PTG on PTS
and psychological distress in this sample of young adult cancer survivors are consistent with the
existing studies with other age groups. The findings from Study 2 need further research attention;
age at diagnosis and the number of years since cancer diagnosis were found to be significant in

71

differentiating the cluster memberships among Thrivers, Recovers, and Sufferers. These two
factors are highly correlated, so which has more influence cannot be known in the present study.
Age at trauma exposure has been neglected in trauma survivorship research. Yet,
according to Janoff-Bulman (1992)’s posttrauma theory, trauma shatters one’s existing
assumptive world and the process of reconstructing a new set of assumptions leads to PTG
experiences. From this framework, it is important to take into account the age of trauma exposure
because the characteristics of the shattered assumptive world might depend on the affected
individual's developmental stage. Childhood and adolescence are formative years when one
makes trials and errors to develop self identities (Erikson, 1963). The existing assumptive worlds
of children might be different from those of adults in terms of the levels of their flexibility and
clarity. Because children have not yet formed rigid assumptive worlds or value systems prior to
trauma, their responses to the shattering trauma could be more flexible and adaptable than those
of adults (L. G. Calhoun & R. G. Tedeschi, 2006b). Study 2, in fact, proves that PTS and PTG
levels are the lowest in Recoverers (mean age at diagnosis is about 9), who were significantly
younger than the other two groups (mean ages at diagnosis are about 12). Recoverers seem to be
the least affected by trauma in the sample. If this theory is correct, young adult survivors seem to
be more vulnerable to traumatic distress and also more likely to experience growth after cancer if
they are diagnosed as adolescents rather than if they are diagnosed as children.
The influence of age at diagnosis on trauma responses, however, is not definitive because
it may be entangled with that of time since trauma exposure. In Study 2, Recoverers who have the
lowest level of PTS and PTG in this sample are not only those who were diagnosed at the
youngest age, but were also the longest survivors. Theoretically, the impact of trauma might have
dissipated over the time. The existing literature does not have consistent findings about the

72

association between the passage of time since cancer diagnosis and PTS/PTG (Bruce, 2006).
Future research would benefit from teasing out the impact of age at diagnosis and time since
diagnosis, particularly among young adult survivors of childhood cancer.
What is clear and encouraging from the findings from this dissertation, however, is that
these young adult survivors reported moderate levels of PTG on average. Given the higher
prevalence of PTG in young adult cancer survivors when compared to all other age groups, the
potential buffering roles of PTG should be utilized. It is important to develop ways to encourage
PTG in young adult survivors who went through trauma in childhood. As suggested by Calhoun
and Tedeschi (2006b), when it is hard to remember being a different person, or when that
different self was so immature that the trauma disrupted the entire developmental process,
questions such as “Can you imagine the kind of person you might have become if this hadn’t
happened to you?” would help the survivor discover or construct growth experiences that are
useful to them. How to promote PTG while not re-traumatizing survivors, and an exploration of
the question if this is even possible should be an important topics of research in theory and
practice.

73

Some of the shortcomings of this study need to be considered for understanding the
results. First, the sample of this study might not represent general young adult cancer survivors,
because participant recruitment was volunteer-based; those who function better and are socially
active are more likely to participate in research studies. Second, this study employed a cross-
sectional methodology that limits the causal inferences about the PTS/PTG relationships and their
relationships with outcome variables such as physical health, mental health, and psychological
distress. Longitudinal studies that follow survivors through long-term survivorship after the
termination of treatment would be valuable in understanding whether, and how, PTS and PTG
levels change over time, and in finding the direction of the relationships among the constructs.

74

REFERENCES

Akaike, H. (1974). A new look at the statistical model identification. IEEE Transactions
on Automatic Control, 19, 716-723.
Aldwin, C. (2007). Stress, coping, and development: An integrative perspective: The
Guilford Press.
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental
disorders (4 ed.). Washington, DC: American Psychiatric Association.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in
social psychological research: Conceptual, strategic, and statistical considerations.
Journal of Personality and Social Psychology, 51(6), 1173-1182.
Bower, J. E., Meyerowitz, B. E., Desmond, K. A., Bernaards, C. A., Rowland, J. H., &
Ganz, P. A. (2005). Perceptioins of positive meaning and vulnerability following
breast cancer: Predictors and outcomes among long-term breast cancer survivors.
Annals of Behavioral Medicine, 29, 236-245.
Brown, R., Madan-Swain, A., & Lambert, R. (2003). Posttraumatic stress symptoms in
adolescent survivors of childhood cancer and their mothers. Journal of Traumatic
Stress, 16(4), 309-318.
Bruce, M. (2006). A systematic and conceptual review of posttraumatic stress in
childhood cancer survivors and their parents. Clinical Psychology Review, 26(3),
233-256.
Butler, L., Blasey, C., Garlan, R., McCaslin, S., Azarow, J., Chen, X., . . . Hastings, T.
(2005). Posttraumatic growth following the terrorist attacks of September 11,
2001: Cognitive, coping, and trauma symptom predictors in an internet
convenience sample. Traumatology, 11(4), 247.

75

Cadell, S., Regehr, C., & Hemsworth, D. (2003). Factors contributing to posttraumatic
growth; A proposed structural equation model. American Journal of
Orthopsychiatry, 73, 279-287.
Calhoun, L., & Tedeschi, R. (2006). The foundations of posttraumatic growth: An
expanded framework. Handbook of posttraumatic growth: Research and practice,
1–23.
Calhoun, L. G., & Tedeschi, R. G. (2006a). The foundations of posttraumatic growth: An
expanded framework. Mahwah, NJ: Erlbaum.
Calhoun, L. G., & Tedeschi, R. G. (Eds.). (2006b). Handbook of posttraumatic growth:
research and practice. Mahwah, NJ: Lawrence Erlbaum.
Calman, K. C. (1984). Quality of life in cancer patients - an hypothesis. journal of
medical ethics, 10, 124-127.
Cordova, M., Cunningham, L., Carlson, C., & Andrykowski, M. (2001). Posttraumatic
growth following breast cancer: A controlled comparison study. Health
Psychology, 20(3), 176-185.
Cordova, M. J., Andrykowski, M. A., Red, W. H., Kenady, D. E., McGrath, P. C., &
Sloan, D. A. (1995). Frequency and correlates of posttraumatic-stress-disorder-
like symptoms after treatment for breast cancer. Journal of Consulting and
Clinical Pschology, 63, 981-986.
Derogatis, L. R. (2000). Brief Symptoms Inventory (BSI) 18 administration, scoring, and
procedures manual. Minneapolis, MS: National Computer Systems.
Duffy, M., Gillespie, K., & Clark, D. (2007). Post-traumatic stress disorder in the context
of terrorism and other civil conflict in Northern Ireland: randomised controlled
trial. British Medical Journal, 334(7604), 1147.
Ehlers, A., & Clark, D. (2000). A cognitive model of posttraumatic stress disorder.
Behaviour research and therapy, 38(4), 319-345.

76

Ehlers, A., Maercher, A., & Boos, A. (2000). Posttraumatic stress disorder following
political imprisonment: the role of mentaql defeat, alienation, and perceived
permanent change. Journal of Abnormal Psychology, 107, 508-519.
Eiser, C., & Kuperberg, A. (2007). Psychological support. In W. B. Bleyer & R. D. Barr
(Eds.), Cancer in adolescents and young adults (pp. 365-373). Berlin Heidelberg,
Ner York: Springer Verlag.
Erickson, S., & Steiner, H. (2001). Trauma and personality correlates in long term
pediatric cancer survivors. Child Psychiatry and Human Development, 31, 195-
213.
Erikson, E. (1963). Childhood and Society. New York: Norton.
Fife, B., Kennedy, V., & Robinson, L. (1994). Gender and Adjustment to Cancer.
Journal of Psychosocial Oncology, 12(1), 1-21.
Foa, E., Cashman, L., Jaycox, L., & Perry, K. (1997). The validation of a self-report
measure of posttraumatic stress disorder: The Posttraumatic Diagnostic Scale.
Psychological Assessment, 9(4), 445-451.
Fromm, K., Andrykowski, M. A., & Hunt, J. (1996). Positive and negative psychosocial
sequelae of bone marrow transplanation: Implications for quality of life
assessment. Journal of Behavioral Medicine, 19, 221-240.
Green, B. L., Rowland, J. H., Krupnick, J. L., & et al. (1998). Prevalence of
posttraumatic stress disorder in women with breast cancer. Psychosomatics, 39,
102-111.
Haslam, C., & Mallon, K. (2003). A preliminary investigation of post-traumatic stress
symptoms among firefighters. Work & Stress, 17(3), 277-285.
Helgeson, V. S., Reynolds, K. A., & Tomich, P. L. (2006). A metaanalytic review of
benefit finding and growth. Journal of Consulting and Clinical Psychology, 74,
797-816.

77

Hill, A. L., Degnan, K. A., Calkins, S. D., & Keane, S. P. (2006). Profiles of
externalizing behavior problems for boys and girls across preschool: The roles of
emotion regulation and inattention. Developmental Psychology, 42, 913-928.
Hill, R. (1949). Families under stress: Adjustment to the crises of war, separation and
return. New York: Harper & Brothers, Publishers.
Hobbie, W., Stuber, M., Meeske, K., Wissler, K., Rourke, M., Ruccione, K., . . . Kazak,
A. (2000). Symptoms of posttraumatic stress in young adult survivors of
childhood cancer. Journal of Clinical Oncology, 18(24), 4060.
Jacobsen, P. B., Widowns, M. R., Hann, D. M., Andrykowski, M. A., Kronish, L. E., &
Fields, K. K. (1998). Posttraumatic stress disorder symptoms after bone marrow
transplantation for breast cancer. Psychosomatic Medicine, 60, 366-371.
Janoff-Bulman, R. (1992). Shattered assumptions: towards a new psychology of trauma.
New York: Free Press.
Janoff-Bulman, R., & Frieze, J. H. (1983). A theoretical perspective for understanding
reactions to victimization. Journal of Social Issues, 39, 1-17.
Jim, H., & Jacobsen, P. (2008). Posttraumatic stress and posttraumatic growth in cancer
survivorship: a review. The Cancer Journal, 14(6), 414.
Johnson, R., Hobfoll, S., Hall, B., Canetti-Nisim, D., Galea, S., & Palmieri, P. (2007).
Posttraumatic growth: Action and reaction. Applied Psychology, 56(3), 428-436.
Joseph, S., Williams, R., & Yule, W. (1993). Changes in outlook following disaster: The
preliminary development of a measure to assess positive and negative responses.
Journal of Traumatic Stress, 6(2), 271-279.
Kazak, A., Barakat, L., Meeske, K., Christakis, D., Meadows, A., Casey, R., . . . Stuber,
M. (1997). Posttraumatic stress, family functioning, and social support in
survivors of childhood leukemia and their mothers and fathers. Journal of
Consulting and Clinical Psychology, 65(1), 120-129.

78

Kazak, A. E., Barakat, L., Alderfer, M., Rourke, M., Meeske, K. A., Gallagher, P., & et al.
(2001). Posttraumatic stress in survivors of childhood cancer and mothers:
Development and validation of the Impact of Traumatic Stressors Interview
Schedule (ITSIS). Journal of Clinical Psychology in Medical Settings, 8, 307-323.
Kelly, B., Raphael, B., Smithers, M., Swanson, C., Reid, C., McLeod, R., . . . Walpole, E.
(1995). Psychological responses to malignant melanoma. An investigation of
traumatic stress reactions to life-threatening illness. General Hospital Psychiatry,
17, 126-134.
Kessler, R., Sonnega, A., Bromet, E., Hughes, M., & Nelson, C. (1995). Posttraumatic
stress disorder in the National Comorbidity Survey. Archives of general
psychiatry, 52(12), 1048.
Kleim, B., & Ehlers, A. (2009). Evidence for a curvilinear relationship between
posttraumatic growth and posttrauma depression and PTSD in assault survivors.
Journal of Traumatic Stress, 22(1), 45-52.
Klein, A., & Moosbrugger, H. (2000). Maximum likelihood estimation of latent
interaction effects with the LMS method. Psychometrika, 65(4), 457-474.
Kline, R. (2005). Principles and practice of structural equation modeling: The Guilford
Press.
Landolt, M., Vollrath, M., Ribi, K., Gnehm, H., & Sennhauser, F. (2003). Incidence and
associations of parental and child posttraumatic stress symptoms in pediatric
patients. Journal of Child Psychology and Psychiatry, 44(8), 1199-1207.
Langeveld, N., Grootenhuis, M., Voute, P., & de Haan, R. (2004). Posttraumatic stress
symptoms in adult survivors of childhood cancer. Pediatric blood & cancer, 42(7),
604-610.
Lanza, S. T., Flaherty, B. P., & Collins, L. M. (2003). Latent class and latent transition
analysis. In J. A. Schinka & W. E. Velicer (Eds.), Handbook of psychology:
Research methods in psychology (pp. 663-685). New York, NY: Wiley.

79

Lee, Y., & Santacroce, S. (2007). Posttraumatic stress in long-term young adult survivors
of childhood cancer: A questionnaire survey. International journal of nursing
studies, 44(8), 1406-1417.
Lehman, D. R., Davis, C. G., & Delongis, A. (1993). Positive and negative life changes
following bereavement and their relations to adjustment. Journal of Social and
Clinical Psychology, 12, 90-112.
Levine, S., Laufer, A., Hamama-Raz, Y., Stein, E., & Solomon, Z. (2008). Posttraumatic
growth in adolescence: Examining its components and relationship with PTSD.
Journal of Traumatic Stress, 21(5), 492-496.
Linley, P., & Joseph, S. (2004). Positive change following trauma and adversity: A
review. Journal of Traumatic Stress, 17(1), 11-21.
Little, R. J., & Rubin, D. B. (1987). Statistical analysis with missing data. New York:
Wiley.
Lo, Y., Mendell, N., & Rubin, D. B. (2001). Testing the number of components in a
normal mixture. Biometrika, 88, 767-778.
Lubke, G. H., & Muthen, B. (2005). Investigating population heterogeneity with factor
mixture models. Psychological Methods, 10, 21-39.
Manne, S., Duhamel, K., & Redd, W. (2000). Association of psychological vulnerability
factors to post-traumatic stress symptomatology in mothers of pediatric cancer
survivors. Psycho-Oncology, 9(5), 372-384.
Manne, S., Ostroff, J., Winkel, G., Goldstein, L., Fox, K., & Grana, G. (2004).
Posttraumatic growth after breast cancer: Patient, partner, and couple perspectives.
Psychosomatic Medine, 66, 442-454.
Manuel, G. M., Roth, S., Keefe, F. J., & Brantley, B. A. (1987). Coping with cancer.
Journal of Human Stress, 13, 149-158.

80

Moosbrugger, H., Schermelleh-Engel, K., Kelava, A., & Kelin, A. G. (2009).
Comparisoon of alternative estimation approaches. In T. Teo & M. S. Khine
(Eds.), Structural Equation Modelling in Educational Research: Concepts and
Applications. Rotterdam, NL: Sense Publishers.
Morrill, E., Brewer, N., O'Neill, S., Lillie, S., Dees, E., Carey, L., & Rimer, B. (2008).
The interaction of post-traumatic growth and post-traumatic stress symptoms in
predicting depressive symptoms and quality of life. Psycho-Oncology, 17(9), 948-
953.
Mundy, E. A., Blanchard, E. B., Cirenza, E., Gargiulo, J., Maloy, B., & Blanchard, C. G.
(2000). Posttraumatic stress disorder in breast cancer patients following
autologous bone marrow transplantation or conventional cancer treatments.
Behaviour Research and Therapy, 38, 1015-1027.
Muthen, B. O. (2001). Latent variable mixture modeling. In G. Marcoulides & R.
Schumacker (Eds.), New developments and techniques in structural equation
modeling (pp. 1-33). Mahwah, NJ: Lawrence Erlbaum Associates, Inc.
Muthén, L., & Muthén, B. (2006). Mplus: Version 4.0. Los Angeles: Muthén & Muthén.
Osborne, J. W. (2008). Best Practices in Quantitative Methods. Thousand Oaks, CA:
Sage Publishing.
Parry, C., & Chesler, M. A. (2005). Thematic evidence of psychosocial thriving in
childhood cancer survivors. Qualitative Health Research, 15(8), 1055-1073.
Pastor, D. A., Barron, K. E., Miller, B. J., & Davis, S. L. (2007). A latent profile analysis
of college students' achievement goal orientation. Contemporary Educational
Psychology, 32, 8-47.
Pelcovitz, D., Goldenberg, B., Kaplan, S., Weinblatt, M., Mandel, F., Meyers, B., &
Vinciguerra, V. (1996). Posttraumatic stress disorder in mothers of pediatric
cancer survivors. Psychosomatics, 37(2), 116.

81

Perrin, S., Smith, P., & Yule, W. (2000). Practitioner review: The assessment and
treatment of post-traumatic stress disorder in children and adolescents. Journal of
Child Psychology and Psychiatry, 41(3), 277-289.
Ries, L. A. G., Eisner, M. P., Kosary, C. L., Hankey, B. F., Miller, B. A., Clegg, L., &
Edwards, B. K. (2003). SEER Cancer Statistics Review, 1975-2000, from
http://seer.cancer.gov/csr/1975_2000/
Satorra, A., & Saris, W. E. (1985). Power of the likelihood ratio test in covariance
structure analysis. Psychometrika, 50(1), 83-90.
Schnurr, P., Rosenberg, S., & Friedman, M. (1993). Change in MMPI scores from
college to adulthood as a function of military service. Journal of Abnormal
Psychology, 102(2), 288-296.
Schulz, U., & Mohamen, N. E. (2004). Turning the tide: Benefit finding after cancer
surgery. Social Science and Medicine, 59, 653-662.
Schwartz, L., & Drotar, D. (2006). Posttraumatic stress and related impairment in
survivors of childhood cancer in early adulthood compared to healthy peers.
Journal of Pediatric Psychology, 31(4), 356.
Schwarz, G. (1978). Estimating the dimension of a model. Annals of Statistics, 6, 461-
464.
Sears, S., Stanton, A., & Danoff-Burg, S. (2003). The yellow brick road and the emerald
city: Benefit finding, positive reappraisal coping, and posttraumatic growth in
women with early-stage breast cancer. HEALTH PSYCHOLOGY-HILLSDALE-,
22(5), 487-497.
Shelby, R. A., Golden-Kreutz, D. M., & Andersen, B. L. (2008). PTSD diagnoses,
subsyndromatl symptoms, and comorbidities contribute to impairments for breast
cancer survivors. Journal of Traumatic Stress, 21, 165-172.

82

Sherbourne, C. (1991). The MOS social support survey. Social Science & Medicine,
32(6), 705-714.
Smith, M. Y., Redd, W., DuHamel, K., Vickberg, S. J., & Ricketts, P. (1999). Validation
of the PTSD-Civilian Version in survivors of bone marrow transplanation.
Journal of Traumatic Stress, 12, 485-499.
Solomon, Z., & Dekel, R. (2007). Posttraumatic stress disorder and posttraumatic growth
among Israeli ex-POWs. Journal of Traumatic Stress, 20(3), 303-312.
Stanton, A., Bower, J., & Low, C. (2006). Posttraumatic growth after cancer. Handbook
of posttraumatic growth: Research and practice, 138–175.
Stuber, M., Kazak, A., Meeske, K., Barakat, L., Guthrie, D., Garnier, H., . . . Meadows, A.
(1997). Predictors of posttraumatic stress symptoms in childhood cancer survivors.
Pediatrics, 100(6), 958.
Stuber, M. L., Meeske, K. A., Krull, K. R., Leisenring, W., Stratton, K., Kazak, A. E., . . .
Zeltzer, L. K. (2010). Prevalence and predictors of posttraumatic stress disorder in
adult survivors of childhood cancer. Pediatrics, 123, 1124-1134.
Sumalla, E., Ochoa, C., & Blanco, I. (2009). Posttraumatic growth in cancer: Reality or
illusion? Clinical Psychology Review, 29(1), 24-33.
Tedeschi, R., & Calhoun, L. (1995). Trauma and transformation: Growing in the
aftermath of suffering: Sage Publications Thousand Oaks, CA.
Tedeschi, R., & Calhoun, L. (1996). The Posttraumatic Growth Inventory: Measuring the
positive legacy of trauma. Journal of Traumatic Stress, 9(3), 455-471.
Tedeschi, R. G., & Calhoun, L. G. (2004). Posttraumatic growth: Conceptual foundations
and empirical evidence. Psychological Inquiry, 15, 1-18.
Thomas, W. I., & Thomas, D. (1928). The Child in America. New York: Knopf.

83

Tomich, P. L., & Helgeson, V. S. (2004). Is finding something good in the bad always
good? Benefit finding among women with breast cancer. Health Psychology, 23,
16-23.
UCLA: Academic Technology Services, S. C. G. Difference testing using the
loglikehood Retrieved December 2, 2010, from
http://www.ats.ucla.edu/stat/mplus/faq/s_b_chi2.htm
Videka-Sherman, L. (1982). Effects of participation in a self-help group for bereaved
parents: Compassionate friends. Prevention in Human Services, 1, 69-77.
Weiss, T. (2004). Correlates of posttraumatic growth in married breast cancer survivors.
Journal of Social and Clinical Psychology, 23(5), 733-746.
Wild, N. D., & Paivio, S. C. (2003). Psychological adjustment, coping, and emotion
regulation as predictors of posttraumatic growth. Journal of Aggression,
Maltreatment, & Trauma, 8, 97-122.
Yi, J., & Zebrack, B. (2010). Self portraints of families with young adult cancer survivors:
Using photovoice. Journal of Psychosocial Oncology, 28(3), 219-243.
Zoellner, T., & Maercker, A. (2006). Posttraumatic growth in clinical psychology--A
critical review and introduction of a two component model. Clinical Psychology
Review, 26(5), 626-653.

Abstract (if available)

Abstract This dissertation examines the interrelationships between Post-Traumatic Stress (PTS) and Post-Traumatic Growth (PTG), the consequences of their coexistence in some survivors, and the buffering roles of PTG on distress in young adult cancer survivors. Specifically, three independent papers comprising the dissertation aim (1) to test curvilinear relationships between PTS and PTG

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The effects of physical functioning and public stigma on psychological distress as mediated by cognitive and social factors among Korean survivors of childhood cancer

PDF

Multilevel influences of care engagement and long-term survival among childhood, adolescent, and young adult cancer survivors

PDF

Post-traumatic growth among high-risk youth: predictors, impact of stressful life events, and relationship with changes in substance use behaviors

PDF

Prospective associations of stress, compulsive internet use, and posttraumatic growth among emerging adults

PDF

Family relationships and their influence on health outcomes over time among older adults in rural China

PDF

Intergenerational support between grandparents and grandchildren in rural China and its effect on the psychological well-being of older adults

PDF

TREAT SAW: an integrated model of MDMA-assisted therapy and Save A Warrior’s evidence-based practices for holistic healing of complex post-traumatic stress (C-PTS) in California’s returning veterans

PDF

A multilevel model of work-family conflict in a global context: a comparative study across 24 countries

PDF

Youth interrupted: stopping the cycle of institutionalization for traumatized youth

PDF

Life transitions, leisure activity engagement, and cognition among older adults

PDF

Direct and indirect predictors of traumatic stress and distress in orphaned survivors of the 1994 Rwandan Tutsi genocide

PDF

Burning out: how California firefighters quell post-traumatic stress and suicides

PDF

The effects of time and space on health status of older adults in China

PDF

The role of dyadic and triadic factors on psychosocial wellbeing and healthcare interactions among childhood cancer survivors, parents, and medical providers

PDF

A comparison of three different sources of data in assessing the adolescent and young adults cancer survivors

PDF

Impacts of caregiving on wellbeing among older adults and their spousal caregivers in the United States

PDF

#BrainCareSavesLives

PDF

Childhood cancer survivorship: parental factors associated with survivor's follow-up care behavior and mental health

PDF

Examining the longitudinal influence of the physical and social environments on social isolation and cognitive health: contextualizing the role of technology

PDF

Survival trends and related outcomes of survivors of childhood and young adult cancer

Asset Metadata

Creator Yi, Jaehee (author)

Core Title The impact of post-traumatic stress and post-traumatic growth on young adult cancer survivors

School School of Social Work

Degree Doctor of Philosophy

Degree Program Social Work

Publication Date 05/05/2011

Defense Date 03/15/2011

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag cancer survivorship,childhood cancer survivorship,OAI-PMH Harvest,possttraumatic growth,young adult cancer survivors

Place Name California (states), Los Angeles (city or populated place), USA (countries)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Palinkas, Lawrence A. (committee chair), Chi, Iris (committee member), Chou, Chih-Ping (committee member), Knight, Bob G. (committee member)

Creator Email jaehee.yi@gmail.com,jaeheeyi@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3863

Unique identifier UC1392031

Identifier etd-Yi-4444 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-443654 (legacy record id),usctheses-m3863 (legacy record id)

Legacy Identifier etd-Yi-4444.pdf

Dmrecord 443654

Document Type Dissertation

Rights Yi, Jaehee

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu