Journal of Pediatric Psychology, Vol. 25, No. 2, 2000, pp. 105-115
© 2000 Society of Pediatric Psychology
Identity in Adolescent Survivors of Childhood Cancer
1 Children's Hospital of Alabama, Birmingham, 2 Department of Pediatrics, Medical University of South Carolina, 3 Georgia State University, 4 Emory University School of Medicine, 5 Kennedy-Krieger Institute, Johns Hopkins University, 6 Children's Hospital Medical Center, Cincinnati, 7 University of North Carolina at Charlotte
All correspondence should be sent to Ronald T. Brown, Department of Pediatrics, Division of Genetics and Child Development, Medical University of South Carolina, Children's Hospital, 171 Ashley Avenue, Charleston, South Carolina 29425. E-mail: brownron{at}musc.edu .
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionReferences |
|---|
Objective: To investigate identity formation among adolescent survivors of childhood cancer. Family functioning, perceived emotional support from family and peers, life stress, and anxiety produced by the cancer experience also were examined as they influenced identity development.
Method: Participants were 52 adolescent survivors and their mothers recruited from a medical center and 42 healthy adolescent counterparts and their mothers recruited from the community.
Results: A greater frequency of survivors than their healthy peers was found within the foreclosed identity status. Factors associated with the foreclosed identity status included the cancer diagnosis, symptoms of posttraumatic stress disorder (PTSD), and family functioning characterized by greater levels of conflict.
Conclusions: Data were interpreted to suggest that the foreclosed identity status may serve a protective function in assisting survivors to cope with the stressors of the cancer experience.
Key words: cancer; adolescent; identity formation; late effects.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Recent advances in the treatment of childhood cancer have resulted in significant increases in the survival rate of these youths; consequently, the late effects of the disease have garnered burgeoning interest. Medical and psychological late effects are defined as occurring after successful completion of medical therapy, usually 2 or more years from the time of diagnosis (Mulhern, 1994
). The
medical and psychological late effects are often at odds with the tasks of
adolescence and influence survivors' adaptation, impeding mastery of such
typical developmental tasks as identity formation
(Erikson, 1968) and
independence and autonomy from the family. Adolescent identity formation
requires an individuation process in which one differentiates self from parent
without becoming totally disconnected from the family of origin
(Grotevant & Cooper, 1985).
This task is frequently fraught with difficulty and may be intensified as
physical and cognitive changes are added to cancer survival sequelae
(Madan-Swain & Brown,
1991).
Evidence suggests that a chronic illness may negatively influence identity
formation (Hauser et al.,
1992). Gavaghan and Roach
(1987) found fewer adolescents
diagnosed with cancer in the identity achievement category, the highest status
of identity development, than their nondiseased counterparts. However, many of
these participants were in the active phase of treatment, so an understanding
of identity formation in survivors is still needed. Those variables that
influence identity formation, including adaptation to the disease, family
functioning, and peer support, have not been examined in survivors of
childhood cancer.
Related to the process of identity formation among adolescent cancer
survivors is the literature examining adaptation to the disease process
(Kazak, 1994). Kazak and
colleagues (Kazak et al.,
1997) conceptualized cancer as a traumatic event because survivors
repeatedly re-experience certain psychological aspects of both the illness and
its associated medical therapy. Core symptoms of posttraumatic stress disorder
(PTSD) (American Psychiatric Association,
1994) may result, including avoidance of the illness experience
(e.g., hospital, clinic), intrusion of the cancer experience during the course
of daily activities and in response to minor illnesses (e.g., colds, aches,
pain) (Kazak et al., 1997),
and hypervigilance (e.g., extreme vigilance to any sounds within the
environment). These symptoms may persist well after termination of treatment.
This tendency to escape reminders of the cancer experience and vest emotional
energy in avoidance behavior may cause survivors to harbor a sense of a
foreshortened future (American Psychiatric
Association, 1994). Thus, survivors may not consider and explore
alternative choices. Moreover, some research has suggested that the anxiety
and life stressors associated with PTSD symptoms negatively influence
cognition (Tramontana & Hooper,
1997), and this is apt to restrict higher levels of identity
formation.
During adolescence, most families reorganize boundaries that accommodate
the teenagers' flexible movement into and out of the family system
(Grotevant & Cooper, 1985).
In fact, some research suggests that maintenance of low levels of flexibility
and autonomy among family members may thwart typical adolescent development.
For example, as Hauser and colleagues
(1992) noted, increasing
restrictiveness in the family system in response to a chronic illness may
impede successful identity formation. Some research shows increased rigidity
in families coping with cancer
(Madan-Swain et al., 1994),
but other investigations have not observed this phenomenon
(Rait et al., 1992).
The contribution of peers to identify development in typical adolescence
has been well documented (Feldman &
Elliott, 1990). Adolescents seeking support frequently vacillate
between their families and their peers
(LaGreca et al., 1995). For
adolescents who have experienced a chronic illness, socialization experiences
with peers, believed to be important in identity formation, may have been
limited. However, the role of peers in the task of identity formation among
youths coping with chronic illness has not been examined.
This study is one of the first late-effect investigations in the cancer
literature examining issues pertaining to identity development. It addresses
prior methodological limitations such as a narrow definition of identity
(Stern, Norman, & Zevon,
1991) or a heterogeneous sample of youths both on and off therapy
(Gavaghan & Roach, 1987) by
broadening the focus of identity formation and including only survivors of
childhood cancer in the sample. We predicted more cancer survivors within the
foreclosed identity status group than healthy controls. We favored the idea
that high social desirability may be associated with an enduring tendency to
internalize the value system of significant adults in their lives, or
prematurely foreclose, or decide on a career
(Stern et al., 1991). Based
upon the data provided by Gavaghan and Roach
(1987), we posited that fewer
survivors than healthy counterparts would reach the identity achievement
status. Finally, an additional intention of our investigation was to examine
the contribution of individual factors including life stress, anxiety
regarding the cancer experience, family functioning, and perceived emotional
support from family and friends to the foreclosed identity status.
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Participants
Childhood cancer survivors were 12- to 23-year-olds from a major university-affiliated medical center and had completed treatment at least 1 year prior to participating in this investigation. Survivors and their parents were consecutively recruited at routine follow-up appointments during a 9-month period. Because of the severe cognitive impairments of children with brain tumors, which may have impeded their ability to complete the assessment instruments, this group of cancer survivors was excluded from this investigation.
Patients were sent a letter describing our plans to assess the goals and identity formation of adolescent and young adult survivors of childhood cancer. If they were interested in participating, parents were requested either to call or return a stamped self-addressed postcard enclosed with the letter. Of the 75 families who were consecutively referred to the late effects program, 52 (69%) agreed to participate. Results of t tests for continuous variables and chisquares for frequency cross-tabulation data showed that the 23 nonparticipants did not differ from the participants on chronological age, number of months off treatment, type of cancer, severity of disease, or severity of medical late effects. The most common reasons for refusal to participate included not having enough time to complete the measures during the visit or questionnaires that were not currently relevant.
In addition to cancer survivors, a healthy comparison group was recruited from youth clubs and activity centers, technical schools, and private and public universities. The investigators administered a brief medical history interview to determine the existence of prior medical conditions. Four adolescents were excluded from participation because of a previously diagnosed medical condition in their immediate family (e.g., a congenital heart defect). Efforts were made to ensure that the healthy comparison group was similar to the cancer survivor group in age, gender, ethnicity, socioeconomic background, and economic dependence on the family. Table I describes the study participants.
|
Cancer Survivors
The sample consisted of 52 survivors and their mothers from a total of 52
families. The patients ranged in age from 12 to 23 years and had completed
treatment on average 5 years prior to participation in this investigation
(M = 5 years 9 months, SD = 3.34, range = 1 year to 14 years
4 months) (see Table I for
survivor characteristics). The majority of subjects (n = 27, 51.9%)
had survived leukemia (ALL n = 20, AML n = 3, t-cell
leukemia n = 3, infant leukemia n = 1). The remainder of the
group had survived various other types of
cancer.1,2,3
A large percentage of the survivors (n = 36, 69.2%) were from intact
two-parent families, and 92.3% (n = 48) were supported by and
financially dependent on their parents. Only 5.8% (n = 3) of the
survivors were married. Of the survivors, 76.9% (n = 40) were
residing with their families at the time of data collection, 17.3% (n
= 9) were residing in dormitories while attending college, and 5.8%
(n = 3) were living independently in apartments or single-family
homes.
Healthy Comparison Group
This sample consisted of 42 adolescents or young adults (18 males, 24
females) and their mothers from a total of 42 families. The participants
ranged in age from 12 to 23 years.
Most of the healthy comparison group (n = 32, 76.2%) were from intact two-parent families, and 95.2% (n = 40) were financially dependent. Only 4.8% (n = 2) were married. Most (n = 32, 76.2%) were residing with their families at the time of data collection, 11.9% (n = 5) were residing in dormitories while attending college, and 11.9% (n = 5) were living independently.
Procedure
Informed consent and study data were obtained from the adolescents and
their parents at the time of their scheduled off-therapy clinic visit. The
healthy comparison group was administered an assessment package identical to
that administered to the pediatric cancer survivors and their parents.
Participants were compensated $10.
Measures
Disease Severity
Additional data were obtained on all survivors from their medical records
after their participation in the study. A disease severity index was computed
based on information from medical records and independent ratings provided by
both the survivor's primary oncologist and another oncologist (initials R.V.)
who reviewed each of the survivors' medical records. Items selected for
inclusion on the disease severity index were obtained from the literature
(e.g., Kazak, 1994) and were
delineated for review by two pediatric oncologists. Both oncologists reviewed
all of the items, and those retained were based on consensus.
The disease severity index was based on three components including treatment history (i.e., whether the child was treated with chemotherapy, radiation, or both and/or transplantation), medical and psychological late effects (i.e., growth failure, cardiac impairment, sterility, second malignancy, skeletal malformations such as scoliosis, asymmetry, and demonstrated developmental late effects including learning disabilities and attention deficit hyperactivity disorder) and the physician's subjective assessment of the survivors' disease (i.e., histology, spread, and other biologic measures) and treatment variables (diagnosis, age at diagnosis, length of treatment, types of cytotoxic therapies, complications, time since completion of medical treatment, and number of relapses). Subjects were rated only on the subjective component, and the scores ranged from 4 (mild) to 12 (severe). The reliability of the two raters was.90 (Pearson correlation coefficient). When the frequency of late effects was calculated, a severity score was provided and employed in the analyses. Because of the high reliability, the rating provided by one oncologist (R.V.) was used to quantify each survivor's disease severity.
Extended Objective Measure of Ego Identity Status-2 (EOMEIS-2)
This self-report scale (Adams, Bennion,
& Huh, 1989) consists of 64 items reflecting the presence or
absence of crisis and commitment in both ideological and interpersonal
identity content domains. The EOMEIS-2 classifies individuals into one of four
identity states: diffusion, foreclosure, moratorium, and achievement. Identity
achievement refers to those youths who have experienced a period of
questioning and exploration and consequently have formed commitments to a
personal set of values and beliefs (exploration, commitment). Youths
classified as being in moratorium are not committed to personal goals, values,
or beliefs, but are in a period of active exploration and reflection
(exploration, no commitment). Youths classified in the foreclosure state are
committed to goals, values, and beliefs without having experienced a period of
questioning or reflection (commitment, no exploration). Finally, youths
classified in the identity diffused state are not committed to goals, values,
or beliefs and are unconcerned about their present state of indecision (no
exploration, no commitment).
EOMEIS-2 raw-scale scores for the identity states are derived by summing
responses to the appropriate items (1 = strongly disagree through 6 = strongly
agree). Scoring guidelines are provided for treating identity scores both as a
categorical and continuous variable. The EOMEIS-2 has demonstrated adequate
reliability and concurrent validity
(Bennion, 1988;
Bennion & Adams, 1986).
Internal consistency coefficients were computed for each status (ideological
domain: diffused 
=.50, foreclosure =.83, moratorium
=.68, and identity achievement =.67; interpersonal domain: diffused
=.68, foreclosure =.83, moratorium =.57, and identity
achievement =.70).
Correlates of Identity Formation
The variables were derived theoretically, and the measures selected had
been used extensively in the literature and shown to have adequate reliability
and validity data. Broad-band scores were employed for each measure.
Adolescent Reports
Marlowe-Crowne Social Desirability Scale. The Marlowe-Crowne
(Crowne & Marlowe, 1960), a
measure of social desirability, was included because self-report studies are
plagued by response bias. The Marlowe-Crowne has been used successfully both
with college students and with adolescent cancer survivors
(Madan-Swain et al., 1994) and
internal consistency coefficient for the present sample was.78.
Posttraumatic Stress Disorder Reaction Index (PTSD Reaction
Index). The adolescent and young adult participants were asked to
complete the PTSD Reaction Index (Pynoos,
Frederick, Nader, & Arroyo, 1987), a self-report measure of 20
possible symptoms of PTSD. Internal consistency coefficient computed for this
sample was.84.
Survivors were instructed to respond based on their cancer experience while the healthy comparison group was instructed to identify a significant life stressor, note the date, and refer to it when completing the measure. While it may have been more parsimonious to allow both groups to select any significant life stressor, it would have limited an examination of the cancer experience for survivors per se.
Adolescent Inventory of Life Events and Changes (A-FILE). The
adolescent and young adult subjects were requested to complete the A-FILE
(McCubbin, Patterson, Bauan, & Harris,
1996), a 50-item self-report measure designed to record normative
and non-normative life events. The measure yields two scores and internal
consistency coefficients computed for the recent and past events scale were as
follows: A-FILE Recent Events =.67, A-FILE Past Events =.66.
Married participants did not complete this measure.
The Perceived Social Support-Family and Perceived Social Support-Friend
(PSS-Fa & PSS-Fr). The PSS-Fa and PSS-Fr
(Procidano & Heller, 1983)
were used to assess adolescent and young adults' perceived emotional support
from family and friends. These measures have been used with adolescents
(LaGreca et al., 1995).
Internal consistency coefficients were.82) for the PSS-Fa and.73 for the
PSS-Fr.
Maternal and Family Functioning
Posttraumatic Stress Reaction Index (PTSD Index). Mothers of all
the participating subjects completed the 20-item PTSD Reaction Index
(Pynoos et al., 1987), a
measure similar to the adolescent and young-adult version. The mothers of
childhood cancer survivors responded based on the child's cancer treatment
while comparison families responded to an identified significant life
stressor. The internal consistency coefficient for the study sample
was.71.
Family Inventory of Life Events and Change (FILE). Mothers of all
subjects completed the FILE (McCubbin,
Patterson, & Wilson, 1996), a 71-item self-report instrument
designed to record both normative and nonnormative life events and changes
experienced by a family in the past year. In our investigation the internal
consistency coefficient for recent and past events was FILE Present Events
=.71, FILE Past Events =.69.
The Family Environment Scale (FES). All study participants and
their mothers completed the FES based on their family of origin
(Moos & Moos, 1981). Three
higher-order FES factors have been delineated and replicated with families of
children with chronic illnesses
(Kronenberger & Thompson,
1990). These dimensions have been demonstrated to reflect
supportive, conflicted, and controlling aspects of families. Internal
consistency coefficients were computed and found to be acceptable for
supportiveness ( =.61) and conflicted ( =.65), but not for the
controlling factor ( =.33). For this reason, we did not use the
controlling factor in any further analyses.
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
We compared the survivors and the healthy comparison group on all demographic variables. Continuous variables were analyzed with t tests and categorical variables were analyzed by chi-square tests of association. We found no significant demographic differences between the groups (see Table I).
Results are organized in three sections. The first group of analyses
focused on identity formation and examined differences in identity status
between the survivors and the healthy comparison group. In accordance with the
scoring criteria outlined by Adams et al.,
(1989), identity formation was
examined both as a categorical variable (diffused, foreclosed, moratorium, or
identity achievement) and as a continuous variable. We used a categorical
analysis so that frequency data could be directly compared with the method
used by Gavaghan and Roach
(1987). We used log-linear
analyses to examine identity as a categorical variable. Second, in order to
take into account individual variability within each identity status and to
explore the association between overall identity status and intraindividual
and family variables, we also treated identity as a continuous variable. A
multivariate analysis of covariance (MANCOVA) was used to examine the identity
states as continuous variables. Finally, analyses were conducted to examine
the associations of the foreclosed identity status with intraindividual
factors, family functioning, and social support.
Identity Formation
We hypothesized that survivors would be less likely to reach the identity
achievement status and more likely to fall into the foreclosed identity status
than the healthy comparison group members. These predictions reflected
differences in the interpersonal and ideological domains.
Categorical Examination of Identity Status
We used log-linear analyses to examine the association of group membership
(survivor versus healthy comparison group), identity status, and chronological
age (young adolescent group, 17 years 11 months, and younger versus older
adolescent group, 18 years and older). Two separate hierarchical log-linear
analyses were conducted to determine goodness of fit, one for the
interpersonal and one for the ideological identity domain data.
Table II shows observed and
expected counts. The saturated models in both analyses were reduced using
backward selection until all nonsignificant terms were eliminated. A
dependency between identity status and age was selected for both the
interpersonal, 
2 (8) = 7.76, p =.46, and the ideological,
2 (8) = 9.15, p =.33, domains, and group membership did not
enter the model. These statistically nonsignificant likelihood-ratio
chi-square values indicated no significant deviation from the age by identity
status model selected (i.e., this model of dependency fit the frequency data,
and group membership did not make an overall contribution). The null
hypothesis indicates that the data fit the hypothesized model, while the
alternative hypothesis states that the data do not fit the hypothesized model.
Thus, failing to reject the null hypothesis supports the model. For both the
survivors and their healthy counterparts, a larger frequency of individuals
within the diffused status was found during early and middle adolescence and a
corresponding increase in the identity achievement status was found during
late adolescence.
|
Consistent with stage theory recognizing a developmental continuum of
identity stages and prior research with childhood cancer survivors
(Stern et al., 1991), we
hypothesized a priori that a higher frequency of survivors would attain the
foreclosed status than their healthy counterparts. To examine any difference
in the frequency of individuals within the foreclosed status relative to the
other three identity statuses, we conducted a further analysis by collapsing
diffusion, moratorium, and achievement into a single category and comparing it
to the foreclosed status. The log-linear analyses for the interpersonal domain
revealed that the data fit the model proposed between group membership and the
collapsed identity states, as well as a main effect for age, 2
(3) =.36, p =.95. The results of the log-linear analyses computed for
the ideological domain yielded a similar interaction between group membership
and the collapsed identity states and a main effect for age, 2
(3) = 1.98, p =.58. Again, both for survivors and healthy controls,
there was an age-dependent progression of identity development. Because of the
statistical nonsignificance obtained from the model, goodness of fit was
demonstrated. The interaction accounted for the greater frequency of survivors
(n = 19) in the foreclosed status relative to the healthy comparison
group (n = 7).
Because we also hypothesized a priori that fewer survivors than healthy
peers would attain the identity achieved status, we conducted a further
analysis by collapsing the diffused, foreclosed, and moratorium states into a
single category and comparing it to the identity achieved status. A dependency
between identity status and age was selected for both the interpersonal,
2 (4) = 4.72, p =.32, and ideological,
2 (4) = 1.85, p =.76, domains and group membership
did not enter the final model. Thus, the model failed to confirm our
hypothesis that group membership was associated with identity.
Examination of Identity as a Continuous Variable
Finally, based on the notion that identity follows a hierarchical
progression, we used the identity measures as continuous variables to examine
overall group differences in the magnitude of the four identity states.
Identity scores were obtained for each subject based on the scoring rules
outlined by Adams et al.
(1989). Scores from each of the
four states (diffused, foreclosed, moratorium, achieved) were entered as
dependent variables into a MANCOVA using the Marlowe-Crowne
(Crown & Marlowe, 1960) as
a covariate because of the potential influence of social desirability on the
various dependent measures. This mitigated the potential impact of this
response set on the measures employed. The results of this MANCOVA indicated a
significant difference for group membership (survivors, healthy controls),
F(4, 88) = 3.5, p <.01. Subsequent univariate analyses of
covariance (ANCOVAs) revealed that the survivors and the healthy comparison
groups differed only on the foreclosed status, F(1, 91) = 12.8,
p <.001, with the survivors endorsing significantly greater levels
of foreclosure (M = 52.8, SD = 12.7) than their healthy
counterparts (M = 42.8, SD = 11.0). Finally, the analyses
were repeated using diagnosis as an independent variable (leukemia, all other
cancer diagnoses, and healthy controls). We obtained no significant results as
a function of cancer type.
Correlates of Identity Formation
We hypothesized that individual factors, family functioning, and perceived
emotional support from family and peers would influence the foreclosed
identity status and contribute significantly to the variance within this
status. A series of Pearson product-moment correlation coefficients was
performed among the continuous foreclosed identity type and each of the parent
and child variables. Because of the numerous correlations performed, we
performed a Bonferroni correction procedure when we examined associations. The
level applied to each comparison in order to maintain a family-wise
of.05 was.002. Results of the correlation coefficients performed
between the foreclosed identity status and the adolescent dependent measures
revealed significant associations between the foreclosure status and the FES
conflict (r =.33, p <.001), indicating that family
ratings of more conflict were associated with higher foreclosure scores. After
Bonferroni corrections were employed to maintain a family-wise significance
level of.05, for the survivors,.01 was used for each comparison. Significant
associations also were found between age off therapy (r =.29,
p <.01), disease severity index (r =.32, p
<.01), number of years off therapy (r =.29, p <.01),
length of treatment (r =.33, p <.01), and foreclosure,
indicating that greater foreclosure scores were positively associated with the
aforementioned disease variables (i.e., longer length of time off therapy,
greater severity of disease, length of treatment). No other significant
associations were obtained.
Results of the correlation coefficients computed for the dependent measures and the identity types for maternal variables revealed a significant association between FES conflict (r =.29, p <.002) and the foreclosed status. Consistent with the adolescent family data, mothers who rated their families as more conflicted had adolescents who endorsed greater foreclosure. We obtained no other significant associations for the mother data.
Based on the correlation coefficients obtained and current empirical
findings in the cancer literature (e.g., PTSD findings by
Kazak et al., 1997), we
constructed a regression equation for the foreclosed identity raw scores as
the dependent variable. The variables selected in the regression equation
included chronological age, group membership (cancer vs. no disease),
adolescent PTSD symptoms, and maternal FES conflict. Although the bivariate
association between the PTSD measure and the foreclosure status was not
significant, the adolescent PTSD measure was entered into the equation due to
theoretical assumptions. Because both the strength and the direction of the
relationships between maternal and adolescent ratings on the FES conflict
factor (r =.29, mother; r =.33, adolescent; both ps
<.002) was similar, only the maternal FES conflict factor was used in the
regression model.
The results of a hierarchical regression model yielded a significant overall model, R2 =.31, F(4, 89) = 10.2, p <.01. This model allowed for an examination of the uniform effects of group membership, PTSD symptoms, and family functioning, once age was controlled. The order of entry was controlled, and chronological age was entered first into the model and accounted for 7% of the variance in the foreclosure status, R2 =.07, F(1, 92) = 7.12, p <.01. Subsequently, when group membership was entered, it accounted for an additional 16% of the variance, R2 change =.16, F(2, 91) = 18.50, p <.001. In addition, when adolescent and young adult PTSD symptoms were entered, they accounted for an additional 5% of the variance, R2 change =.05, F(3, 90) = 6.25, p <.02. Finally, the FES Parent Conflict factor was entered on the fourth step and accounted for an additional 3% of the variance, R2 change =.03, F(4, 89) = 4.63, p <.04.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
We sought to examine identity formation in survivors of childhood cancer. We hypothesized that a greater number of survivors would be found in the foreclosed identity status. This is consistent with the data provided by Stern and associates (1991
), who
found that these youths prematurely foreclose on career decisions. Individuals
in the foreclosed status have not yet actively questioned alternative points
of view but have made a strong commitment to one position, which they will
defend. This commitment is typically an extension of the values and
expectations of significant adults in their lives and suggests that survivors
may be more reluctant to risk exploration of alternative value systems.
Because of the cancer experience, survivors may have had to foreclose on their
options for a variety of reasons, including limited cognitive and physical
functioning or because of the specific realities of their treatment
outcome.
In further support of our findings, recent research by Phipps and
Srivastava (1997) has provided
data to suggest a tendency for patients diagnosed with cancer to exhibit a
"repressive style of adaptation" (or, more specifically, an
avoidant coping style), a tendency to minimize distress, and repression of
dysphoric affect. They suggested that the etiology of such a style may
originate in the stressors of the illness and its associated therapies. Their
data and our results suggest that adopting the views of significant others
(i.e., foreclosed status) might serve a protective function in assisting these
adolescents to cope with the stress of the cancer experience and the sequelae
of medical and psychological late effects. It is noteworthy that the disease
variables were significantly correlated with the foreclosed status, suggesting
that the cancer experience is strongly associated with survivors' membership
within this category. By committing and internalizing the belief system of
significant adults, the survivor does not have to deal with the uncertainty
and ambiguity of exploring future alternatives. This speculation warrants
future investigation.
We found no difference between survivors and their healthy counterparts in the attainment of the identity achievement status. This result is in contrast to the data provided by Gavaghan and Roach, who studied children who were still undergoing treatment for cancer, while our participants were all off-therapy. Methodologically, Gavaghan and Roach used an interview format focusing on specific identity domains (career, marriage, religion, and social relationships), whereas our investigation used an objective self-report measure of identity that yielded global scores.
After controlling for the expected contribution of chronological age in the regression equation, the hypothesized processes underlying the finding of survivors within the foreclosed status, and the association among stressful events, we confirmed PTSD symptoms related to the cancer experience and family functioning. These findings provide additional information regarding the association between adolescent PTSD symptoms and family functioning and the foreclosed identity status. One possible explanation is that PTSD symptoms may increase the structured, restrictive style of the families. The restrictiveness of these family patterns may limit the change that is central to the adolescents' development of identity. In essence, families with childhood cancer likely are more rule-bound and role-governed and thus less able to accommodate to the developmental tasks of adolescence.
We found no significant associations for the contribution of peer emotional support to the process of identity formation. Moreover, each group endorsed peer support that was within the normative range, suggesting that survivors believe that their needs for support, information, and feedback are fulfilled by their social network. Whether peer networks or social supports are essential factors in the identity formation of adolescent survivors to the cancer experience is an area destined for future research.
The contribution of our investigation must be considered within its limitations. First, the study used a cross-sectional design, which does not permit evaluation of patterns of development over time. Future studies will need to incorporate longitudinal designs. Further, our design controlled only for the effect of a chronic disease on identity rather than allowing for specific control of the cancer experience. To examine specifically for the cancer experience, future research will need to include both a healthy and chronically ill comparison group. Second, the association of family functioning within the foreclosed status, though intriguing, cannot be considered causal because of the correlational nature of the investigation. Third, the use of the objective measurement strategy did not allow exploration of identity within specific content domains. Future studies will need to incorporate additional measurement strategies, such as interviews that allow for an examination of specific content domains. Fourth, inherent in this investigation is the notion of a bipolar stage theory in which the adolescent is thought to move from one clear descriptive level (i.e., identity status) to another. Such a stage theory does not allow for an explanation of adolescents' behavior as they move within stages, a pattern accounted for by Piaget's notion of decolage. Within identity formation, it is possible for individuals to exhibit both horizontal and vertical decolage. Thus, individuals might be classified into different identity states depending on the specific identity content domain. Unfortunately, our assessment instrument precluded an examination of identity across content domains. Also, characteristic of many clinical studies, the sample size might have limited power and obscured significant effects that otherwise may have been revealed with a larger sample. Sixth, the heterogeneity in the age range of the participants, time off treatment, and type of cancer may have obscured possible associations among the variables. However, age effects were controlled statistically and no differences were obtained for cancer diagnoses.
How does the experience of having survived cancer alter the trajectory of identity formation? Further study is required to determine whether survivors will remain foreclosed, or whether they will follow the same developmental path of identity, progressing, later in life, to the higher states of moratorium and identity achievement. Such research would have clinical implications for medical treatment teams working with adolescents and young adults undergoing cancer treatment, as well as with survivors of childhood cancer.
|
Acknowledgments |
|---|
This article is based on Avi Madan-Swain's dissertation. Madan-Swain was awarded the 1998 Student Research Award from the Society of Pediatric Psychology for this research project.
|
Notes |
|---|
1 51.9 (n = 27) of this sample had survived acute lymphocytic leukemia (ALL). Among the adolescents who had survived leukemia, 11.5% (n = 6) also received cranial radiation as part of their medical therapy. Only 5.8% (n = 3) of the ALL survivors received a bone marrow transplant; 3.8% (n = 2) survived Ewing's sarcoma, 5.8% (n = 3) Hodgkin's, 5.8% (n = 3) Wilm's tumor, 11.5% (n = 6) osteogenic sarcoma, 3.8% (n = 2) rhabdomyosarcoma, 1.9% (n = 1) histiocytosis, and 1.9% (n = 1) large cell lymphoma.
2 The mean age at diagnosis was 9 years 5 months (SD = 4.88, range 1
year 1 month to 18 years 6 months) and the mean age off therapy was 11 years 5
months (SD = 4.56, range = 3 years 1 month to 19 years 6 months).
3 Of the survivors, 82.7% (n = 43) remained in complete and
continuous remission after completion of their medical treatment, 11.5%
(n = 6) experienced one relapse, and 5.8% (n = 3)
experienced two relapses. When comparing the adolescents who relapsed with
those who remained in complete and continuous remission across the four
identity states, we found no differences.
Received April 1, 1998; revision received August 14, 1998; revision received November 30, 1998; accepted February 2, 1999
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Adams, G., Bennion, L., & Huh, K. (1989). Objective measure of ego identity status: A reference manual. Guelph, Canada: University of Guelph.
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders. 4th ed. Washington, DC:AAA.
Bennion, L. (1988). Measuring adolescent ego identity status: A comparison of the semi-structured interview and the EOMEIS. Unpublished master's thesis, Utah State University, Logan, Utah.
Bennion, L., & Adams, G. (1986). A revision of the extended version of the objective measure of ego-identity status: An identity instrument for use with late adolescents. Journal of Adolescent Research, 1, 183-198.[Abstract]
Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of psychopathology. Consulting and Clinical Psychology, 24, 349-354.
Erikson, E. (1968). Identity, youth, and crisis. New York: Norton.
Feldman, S. S., & Elliott, G. R. (1990). At the threshold: The developing adolescent. Cambridge, MA: Harvard University Press.
Gavaghan, M. P., & Roach, J. E. (1987). Ego
identity development of adolescents with cancer. Journal of
Pediatric Psychology, 12,
203-213.
Grotevant, H., & Cooper, C. (1985). Patterns of interaction in family relationships and the development of identity exploration in adolescence. Child Development, 56, 415-428.[Web of Science][Medline]
Hauser, S. T., Jacobson, A. M., Milley, J., Wertlieb, D., Wolfsdorf, J., Herskowitz, R. D., Lavori, P., & Bliss, R. L. (1992). Ego development paths and adjustment to diabetes: Longitudinal studies of preadolescents and adolescents with insulin dependent diabetes mellitus. In E. J. Susman, L. V. Feagans, & W. J. Ray (Eds.), Emotion, cognition, health, and development in children and adolescents (pp. 133-152). Hillsdale, NJ: Erlbaum.
Kazak, A. E. (1994). Implications of survival: Pediatric oncology patients and their families. In D. Bearison & R. Mulhern (Eds.), Pediatric psycho-oncology: Psychological perspectives on children with cancer (pp. 171-192). New York: Oxford.
Kazak, A. E., Meeske, K., Penati, B., Barakat, L. P., Christakis, D., Meadows, A. T., Casey, R., & Stuber, M. L. (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, 120-129.[Web of Science][Medline]
Kronenberger, W., & Thompson, R. J., Jr. (1990). Dimensions of family functioning in families with chronically ill children: A higher order factor analysis of the family environment scale. Journal of Clinical Child Psychology, 19, 380-388.[Web of Science]
LaGreca, A., Auslander, W. F., Greco, P., Spetter, D., Fisher, E.
B., & Santiago, J. V. (1995). I get by with a little help
from my family and friends: Adolescents' support for diabetes care.
Journal of Pediatric Psychology,
20, 449-476.
Madan-Swain, A., & Brown, R. T. (1991). Cognitive and psychosocial sequelae for children with acute lymphocytic leukemia and their families. Clinical Psychology Review, 11, 267-294.
Madan-Swain, A., Brown, R. T., Sexson, S. B., Baldwin, K., Pais,
R., & Ragab, A. (1994). Adolescents who have survived cancer:
Psychosocial and familial adaptation. Psychosomatics,
35, 453-459.
McCubbin, H. I., Patterson, J., Bauan, E., & Harris, L. (1996). Adolescent family inventory of life events (A-FILE). In H. I. McCubbin, A. I. Thompson, & M. A. McCubbin (Eds.), Family assessment: Resiliency, coping, and adaptation: Inventories for research and practice. Madison: University of Wisconsin System.
McCubbin, H. I., Patterson, J., & Wilson, L. (1996). Family inventory of life events and changes (FILE). In H. I. McCubbin, A. I., Thompson, & M. A. McCubbin (Eds.), Family assessment: Resiliency, coping, and adaptation: Inventories for research and practice. Madison: University of Wisconsin System.
Moos, R. H., & Moos, B. S. (1981). Family Environment Scale Manual. Palo Alto, CA: Consulting Psychologists Press.
Mulhern, R. (1994). Neuropsychological late effects. In D. Bearison & R. Mulhern (Eds.), Pediatric psycho-oncology: Psychological perspectives on children with cancer (pp. 99-121). New York: Oxford.
Phipps, S., & Srivastava, D. K. (1997). Repressive adaptation in children with cancer. Health Psychology, 16, 521-528.[Web of Science][Medline]
Procidano, M., & Heller, K. (1983). Measures of perceived social support from friends and from family: Three validation studies. American Journal of Community Psychology, 11, 1-24.[Web of Science][Medline]
Pynoos, R., Frederick, C., Nader, K., & Arroyo, W.
(1987). Life threat and posttraumatic stress in school age
children. Archives of General Psychiatry,
44,
1057-1063.
Rait, D., Ostroff, J. S., Smith, K., Cella, D. F., Tan, C., & Lesko, L. M. (1992). Lives in a balance: Perceived family functioning and the psychosocial adjustment of adolescent cancer survivors. Family Process, 31, 383-397.[Web of Science][Medline]
Stern, M., Norman, S. L., & Zevon, M. A. (1991). Career development of adolescent cancer patients: A comparative analysis. Journal of Counseling Psychology, 38, 431-439.[Web of Science]
Tramontana, M. G., & Hooper, S. (1997). Neuropsychology of child psychopathology. In C. R. Reynolds & E. Fletcher-Janzen (Eds.), Handbook of clinical child neuropsychology (2nd ed.). New York: Plenum.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  V. Hendricks-Ferguson Hope and Spiritual Well-Being in Adolescents With Cancer West J Nurs Res, April 1, 2008; 30(3): 385 - 401. [Abstract] [PDF]
|
|
|
|
 |
|
 S. S. Hawkins and J. Radcliffe Current Measures of PTSD for Children and Adolescents J. Pediatr. Psychol., May 1, 2006; 31(4): 420 - 430. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Parry and M. A. Chesler Thematic Evidence of Psychosocial Thriving in Childhood Cancer Survivors Qual Health Res, October 1, 2005; 15(8): 1055 - 1073. [Abstract] [PDF]
|
|
|
|
|
|
A. F. Patenaude and M. J. Kupst Psychosocial Functioning in Pediatric Cancer J. Pediatr. Psychol., January 1, 2005; 30(1): 9 - 27. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. F. Last, M. A. Grootenhuis, and C. Eiser International Comparison of Contributions to Psychosocial Research on Survivors of Childhood Cancer: Past and Future Considerations J. Pediatr. Psychol., January 1, 2005; 30(1): 99 - 113. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Haase The Adolescent Resilience Model as a Guide to Interventions Journal of Pediatric Oncology Nursing, September 1, 2004; 21(5): 289 - 299. [Abstract] [PDF]
|
|
|
|
|
|
R. T. Brown Society of Pediatric Psychology Presidential Address: Toward a Social Ecology of Pediatric Psychology J. Pediatr. Psychol., March 1, 2002; 27(2): 191 - 201. [Full Text] [PDF]
|
|
|
|
|
|
S. B. Johnson and L. J. Meltzer Disentangling the Effects of Current Age, Onset Age, and Disease Duration: Parent and Child Attitudes Toward Diabetes as an Exemplar J. Pediatr. Psychol., January 1, 2002; 27(1): 77 - 86. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||