Journal of Pediatric Psychology, Vol. 28, No. 1, 2003, pp. 29-39
© 2003 Society of Pediatric Psychology
Psychological Adjustment of Children and Adolescents With Chronic Arthritis: A Meta-analytic Review
1 Northwestern University Feinberg School of Medicine, 2 University of Illinois at Chicago, Institute for Juvenile Research
All correspondence should be sent to John V. Lavigne, Department of Child Psychiatry #10, Children's Memorial Hospital, 2300 Children's Plaza, Chicago, Illinois 60614. E-mail: jlavigne{at}childrensmemorial.org.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAsterisk indicates studies... |
|---|
Objective To review studies of psychological adjustment among children and adolescents with chronic arthritis to determine whether they are at more risk for development of adjustment problems than controls. Methods We used meta-analytic techniques to review 21 studies reporting overall adjustment problems, internalizing symptoms, externalizing symptoms, or self-concept among youths with arthritis. Results Youths with arthritis displayed increased risk for overall adjustment problems and internalizing symptoms, but not for externalizing symptoms or poor self-concept. Risk was greater in studies making comparisons to study controls rather than to norms and in studies including mixed disease samples (arthritis plus other rheumatic diseases) rather than samples of youths with arthritis only. Conclusions Results suggest the importance of assessing for internalizing problems among youths with chronic arthritis. Future research may benefit from inclusion of child self-report of adjustment problems, diagnostic specificity in reporting results, and use of adjustment measures without somatic items.
Key words: chronic arthritis; juvenile rheumatoid arthritis; pediatric chronic illness; psychological adjustment; meta-analysis.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAsterisk indicates studies... |
|---|
Chronic arthritis in childhood is a heterogeneous group of diseases characterized by joint inflammation and stiffness. There are two major classification systems used for the diagnosis and classification of children with chronic arthritis. American College of Rheumatology (ACR; Brewer, 1977) criteria for juvenile rheumatoid arthritis (JRA) define three subtypes of JRA: pauciarticular (arthritis in four or fewer joints), polyarticular (arthritis in five or more joints), and systemic onset (arthritis in one or more joints in association with a daily fever for at least 2 weeks). European League Against Rheumatism (EULAR, 1977
) criteria for juvenile chronic arthritis (JCA) include the
subtypes of JRA defined by the ACR critieria, as well as several conditions
that fall under the category of spondyloarthropathies (arthritic conditions
that affect entheses, insertions of tendons and ligaments into bone, including
juvenile ankylosing spondylitis, psoriatic arthritis, and arthropathy
associated inflammatory bowel disease). The prevalence of chronic arthritis is
approximately 95 to 130 per 100,000 children
(Oen & Cheang, 1996;
Peterson, Mason, Nelson, O'Fallon, &
Gabriel, 1996; Towner, Michet,
O'Fallon, & Nelson, 1983). Chronic arthritis is two to three
times more common in girls than in boys, with a more balanced gender
distribution seen when children with spondyloarthopathies are included
(Cassidy & Petty, 1995;
Kewman, Warschausky, & Engel,
1995).
There is no known cure for chronic arthritis. Treatment is aimed at
controlling pain; improving range of motion, muscle strength, and function;
and facilitating normal nutrition, growth, and development
(Cassidy & Petty, 1995).
The approach to disease management is multidisciplinary, incorporating
pharmacotherapy, physical and occupational therapy, orthopedic surgery, and
psychosocial services (Cassidy &
Petty, 1995).
Children with chronic arthritis and their families face many challenges
associated with this disease and its treatment. In addition to pain and
physical limitations, children may also experience altered body image, anxiety
around teasing and social acceptance
(Barlow, Harrison, & Shaw,
1998; Konkol et al.,
1989; Taylor, Passo, &
Champion, 1987), and fears about the course of the disease and
their future (Barlow et al.,
1998; Degotardi, Revenson,
& Ilowite 1999; Konkol et
al., 1989). During periods of symptom exacerbation, children may
have difficulty performing daily classroom activities
(Ansell & Chamberlain,
1998; Lovell et al.,
1990; Taylor et al.,
1987). Treatment requirements and medical appointments may create
financial burdens and place limitations on family activities
(Barlow et al., 1998;
Degotardi et al., 1999).
In recognition of the stressors and burdens associated with chronic
arthritis, a considerable number of studies have been conducted to assess the
impact of the illness on children's psychological adjustment. However, a
consensus about the psychological impact of arthritis has not been reached.
Some authors report higher rates of adjustment problems (e.g., behavioral and
emotional problems, low self-esteem) among children with chronic arthritis
than among healthy controls or compared to published norms
(Billings, Moos, Miller, & Gottlieb,
1987; McAnarney, Pless,
Satterwhite, & Friedman, 1974;
Timko, Stovel, Moos, & Miller,
1992), whereas others report no significant differences
(Garstein, Short, Vannatta, & Noll,
1999; Harris, Newcomb, &
Gewanter, 1991; Huygen, Kuis,
& Sinnema, 2000; Noll et
al., 2000; Ungerer, Chaitow,
& Champion, 1988).
There have been several helpful reviews of research on the psychological
adjustment of children and adolescents with chronic arthritis
(Miller, 1993;
Quirk & Young, 1990;
Varni & Jay, 1984), but
conclusions as to whether this population is at increased risk for adjustment
problems have been inconsistent. Moreover, prior reviews have not included all
relevant published research reports for the time periods covered, and criteria
for inclusion in the reviews have not always been presented. Lavigne and
Faier-Routman (1992) provide a
meta-analytic review of children's adjustment to a wide range of pediatric
physical disorders. Results specific to chronic arthritis, however, were based
on only two studies, the most recent published in 1988. Given that the most
recent review of research on psychological adjustment to chronic arthritis was
published in 1993, and several studies have been published since then, our
ability to examine the psychological adjustment of children with this disorder
is now considerably improved.
The goal of this study is to provide a systematic review of the current
empirical body of literature on psychological adjustment in children and
adolescents with chronic arthritis to determine whether they are at higher
risk for the development of psychological adjustment problems than are healthy
control children. The challenge in such a review is to integrate seemingly
disparate research findings meaningfully without introducing subjective bias
in the interpretation of results. Meta-analysis provides one means of
accomplishing this task. Meta-analytic techniques allow for statistical
analysis of results obtained from multiple studies, such that results from
individual studies, rather than from individual subjects, serve as data points
for analysis (Glass, McGaw, & Smith,
1981; Wolf, 1986). Findings across studies can then be
quantitatively integrated without subjective bias in interpreting results, and
statistical tests can be used to determine whether differences in
methodological or sample characteristics influence findings. Therefore, we
used meta-analytic techniques to provide a quantitative review of studies that
have assessed whether children and adolescents with chronic arthritis differ
from controls on specified measures of psychological adjustment.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAsterisk indicates studies... |
|---|
Identification of Studies
Potential studies for this meta-analytic review of adjustment to chronic arthritis were obtained via literature searches of the topics of JRA, JCA, juvenile rheumatic disease, and juvenile arthritis in PsychInfo and Medline from the earliest years available in these databases (1887 and 1966, respectively) through July 2001. Studies were also obtained from review articles and chapters on adjustment to chronic arthritis, as well as from references contained within individual articles. Dissertation Abstracts International was not searched, and studies presented at a conference or meeting but not published in a journal or a text were excluded.
Criteria for inclusion in the meta-analysis were as follows: (a) studies were original, published research reported in English; (b) participants were children or adolescents (age 20 and under) diagnosed with chronic arthritis; (c) studies included a broad-band questionnaire yielding an estimate of overall psychological adjustment problems, a questionnaire describing internalizing or externalizing symptom frequency, or a child-rated measure of self-esteem of self-concept; and (d) studies presented data in a manner that permitted calculation of effect sizes through comparison of a chronic arthritis group with a study control group or with normative data. In cases in which adjustment data for the same sample of participants was presented in more than one report, the sample was counted only once in the analyses.
Sample Sizes and Patient Demographics
Based on the above criteria, 21 studies were selected for inclusion in the
meta-analysis (indicated by an asterisk in the references). Sample sizes
ranged from 8 to 272 chronic arthritis subjects, with a median sample size of
39 participants. Across studies, participants ranged in age from 1 to 20
years, with the majority of samples spanning a wide age range (e.g., early
childhood through late adolescence). Approximately 50% to 85% of participants
within each study were female, consistent with the higher observed prevalence
of chronic arthritis among girls than among boys. Only two studies, however,
reported results separately for boys and for girls. (See
Table I for detailed
information regarding sample size and gender breakdown for the 21 studies
included in the meta-analysis.) For studies that provided information
regarding participants' ethnic backgrounds (n = 8), the majority (66%
to 100%) were Caucasian.
|
Medical Conditions
Five studies included participants with diseases other than chronic
arthritis. Of these five samples, 8% to 22% included children with other
diseases, generally nonarthritic rheumatic diseases, including systemic lupus
erythematosus, dermatomyositis, vasculitis, scleroderma, fibromyalgia, and
mixed connective tissue disease. One of the five studies also included a JRA
observation group (patients who had symptoms of JRA but did not meet
diagnostic criteria) that included 13% of the sample. Across all of the
studies included in the meta-analysis, children with diseases other than
chronic arthritis constituted only 6% of the total number of participants.
(See Table I for information
regarding participants' diagnoses in each of the studies.)
Control Groups
Eleven studies recruited control groups of children without chronic
arthritis. In eight of these studies, potential control participants were
excluded if they had a chronic medical condition. For nine studies, control
participants were drawn from the communities or classrooms of children with
arthritis. These control children were generally matched with children with
arthritis on characteristics such as age, gender, and ethnicity. Control
groups were generally similar to arthritis groups on characteristics such as
socioeconomic status and number of parents living in the home. For two
studies, control groups included siblings closest in age to the children with
arthritis.
Ten studies in the meta-analysis did not recruit control groups. One of these studies provided normative data on the Child Behavior Checklist (CBCL) for purposes of comparison with participants with chronic arthritis, three provided normative data for measures of self-concept, and the remaining six required comparison with published normative data at the time of this review.
Measurement of Psychological Adjustment
In total, 9 studies presented data on overall psychological adjustment
problems (e.g., internalizing and externalizing symptoms), 13 on the frequency
of internalizing symptoms (e.g., anxiety, depression, social withdrawal), 13
on the frequency of externalizing symptoms (e.g, hyperactivity, oppositional
behavior, aggression), and 9 on self-concept. The outcome measure used in the
majority of studies of internalizing or externalizing symptoms (n =
13) was the CBCL (Achenbach &
Edelbrock, 1983), a parent-report measure of emotional and
behavior problems. Two studies used the Youth Self-Report (YSR; Achenback
& Edelbrock, 1987), the child self-report version of the CBCL. Among
studies of self-concept, the most common outcome measure (n = 4) was
the global self-worth sub-scale of the Self-Perception Profile for Children
(Harter, 1985). Studies of
self-concept yielded the greatest number of comparisons of arthritis and
control groups based on child self-report. There was not a sufficient number
of studies specifically measuring depression (n = 5) or anxiety
(n = 1) to include these studies in the analysis.
Computation and Analysis of Effect Sizes
Methodology for the meta-analysis followed that of Lavigne and
Faier-Routman's (1992)
meta-analytic review of children's adjustment to pediatric physical disorders.
For studies in which control groups were recruited, effect sizes were
calculated by comparing results for chronic arthritis groups and study
controls. For studies without control groups, effect sizes were calculated
based on comparisons with normative data provided by the authors or available
published normative data when the authors did not provide this data.
Effect sizes were calculated using the procedures described by Glass,
McGraw, and Smith (1981) and Wolf (1986). Nineteen of the 21 studies reported
means and standard deviations on outcome measures, from which effect sizes
could be calculated by comparing arthritis groups to control groups or to
normative data. For these studies, the effect size index calculated was
d, the difference between the arthritis and control group means,
divided by the standard deviation of the control group. Cohen
(1977) has provided rough
guidelines for "small," "medium," and
"large" effect sizes in the behavioral sciences (for the effect
size d: .20, .50, and .80, respectively).
One study reported no significant difference between the chronic arthritis
group and the control group on a measure of self-concept, but did not report
information necessary to calculate an effect size. For this study, the
conservative strategy of assigning an effect size of zero was used, indicating
exactly no difference between the two groups. Another study reported the
percentages of children in the arthritis and control groups who exceeded a
cutoff score on a measure of behavioral and emotional problems and who
endorsed negative self-comments on a measure of self-concept. For this study,
the effect sizes were determined by consulting a table using probit
transformation methods to convert differences in proportions to effect size
(Glass et al., 1981).
For studies that reported outcome data separately for subgroups of the sample (e.g., reported outcomes separately for boys and girls or for different age groups) but did not report adjustment data for the sample as a whole (n = 4), we created a composite effect size for the entire study by averaging the effect sizes calculated for each subgroup of the sample. Similarly, for one study that reported data separately for mother and father ratings on outcome measures, we created composite effect sizes by averaging the effect sizes for both parents' ratings. The decision to combine mother and father ratings was made because parent ratings consisted of a combination of mother and father reports in the majority of studies. A composite effect size was also created for one study that reported data separately for parent and child ratings to avoid artificially inflating the number of effects beyond the number of independent studies (Wolf, 1986). For this study, parent ratings were also examined separately from child ratings.
For studies that reported internalizing and externalizing symptom frequency but did not report data on overall psychological adjustment problems (n = 6), we created a composite effect size for overall psychological functioning for that sample by averaging a study's effect sizes for internalizing and externalizing problems. Using this strategy, we could calculate an effect size for overall psychological adjustment problems for 15 of the studies included in the meta-analysis.
To determine whether the total group of children and adolescents with chronic arthritis differed from controls on outcome measures, we averaged individual effect sizes for each of the studies in the meta-analysis to create an average effect size for each outcome measure of interest (e.g., overall psychological adjustment problems [based on 15 studies], internalizing symptoms [based on 13 studies], externalizing symptoms [based on 13 studies], and self-concept [based on 9 studies]).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAsterisk indicates studies... |
|---|
Table I lists the effect sizes obtained for each of the studies included in the meta-analysis, for overall psychological adjustment problems, internalizing symptom frequency, externalizing symptom frequency, and self-concept.
Preliminary Analyses
There were no significant relationships between study sample size or year
of study publication and effect sizes for overall adjustment problems,
internalizing symptoms, externalizing symptoms, or self-concept.
Overall Adjustment Problems and Internalizing/Externalizing
Symptoms
The differences in overall psychological adjustment problems, internalizing
symptom frequency, and externalizing symptom frequency between arthritis
groups and study-recruited and normative control groups are summarized in
Table II using mean effect
sizes across studies. A higher level of adjustment problems in the arthritis
group is indicated by a positive mean effect size. Statistical significance
was determined by calculating 95% confidence intervals (CIs) for the mean
effect sizes and determining if the means differed from an effect size of
.00.
|
The first analyses combined data from both parent report and the two studies that obtained child report of psychological data. As seen in Table II, the effect sizes for overall adjustment problems and for internalizing symptoms differed significantly from zero, indicating higher levels of these problems for the total group of children and adolescents with arthritis than for controls. There was no difference in the level of externalizing symptom frequency between arthritis and control groups. The effect size for internalizing symptoms was significantly larger than the effect size for externalizing symptoms, t(24) = 3.14, p = .00.
Because there were only two studies in which outcome data were based on
child self-report (Degotardi et al.,
1999, in which only child-report data were available for overall,
internalizing, and externalizing problems, and
Brace, Smith, McCauley, & Sherry,
2000, which, included both child and parent report for
internalizing problems), we recalculated mean study-level effect sizes
excluding child ratings. Comparing average effect sizes based only on parent
report, we found that results remained the same. Arthritis groups had
significantly higher levels of overall psychological adjustment problems
(d = .31; CI = .17, .46; p = .00) and internalizing problems
(d = .53; CI = .34, .72, p = .00) than did control groups.
There was no difference in the level of externalizing symptom frequency
between arthritis and control groups.
Comparisons of Study Controls Versus Norms. There may be
differences in the effect sizes across studies when illness groups are
compared to normative data versus study-recruited control groups
(Lavigne & Faier-Routman,
1992). If adjustment measures are sensitive to demographic
characteristics and these characteristics vary between study and normative
samples, group differences may be inappropriately attributed to the effect of
the illness (Lavigne & Faier-Routman,
1992). To determine whether the type of control group used
influenced results, we calculated mean effect sizes for overall psychological
adjustment separately for the eight studies that compared arthritis groups to
study-recruited controls and the seven studies that compared arthritis groups
to normative data. For studies using study-recruited controls, arthritis
groups had significantly higher levels of overall psychological adjustment
problems than did controls (d = .37; CI = .20, .54; p =
.00). For studies comparing adjustment of children with arthritis to normative
data, however, there was no significant difference in levels of psychological
adjustment problems between arthritis groups and controls. The difference
between effect sizes for studies using control participants versus studies
using normative data was not significant.
Comparisons of Samples With Chronic Arthritis Only Versus Mixed
Rheumatic Disease Groups. Children with rheumatic diseases other than
chronic arthritis may experience more fatigue, significant disease in other
organs, or risk of mortality than children with arthritis
(Billings et al., 1987;
Cassidy & Petty, 1995).
Therefore, we calculated mean effect sizes and confidence intervals separately
for the five studies with mixed disease groups and for the ten studies
including only participants with arthritis to determine whether the inclusion
of mixed samples may have influenced overall results. For both studies
including only participants with arthritis and studies with mixed samples, the
disease samples had significantly higher levels of overall psychological
adjustment problems than controls (for arthritis only, d = .17; CI =
.04, .29; p = .02; for mixed samples, d = .55; CI = .35,
.75; p = .01). Effect sizes for studies with mixed disease samples,
however, were significantly larger than effect sizes for studies including
only children and adolescents with chronic arthritis, t(13) = 3.35,
p = .01, indicating higher levels of overall adjustment problems
among children in the mixed disease groups than among children in the
arthritis only groups.
Self-Concept
The difference in child-reported self-concept between arthritis groups and
study-recruited and normative control groups is reported in
Table II. More positive
self-concept in the arthritis group is indicated by a positive mean effect
size. As seen in Table II,
based on calculations of 95% CIs, the effect size for self-concept did not
differ significantly from zero, indicating no difference between arthritis and
control groups.
Comparisons of Study Controls Versus Norms. Mean effect sizes for self-concept were calculated separately for the five studies that compared arthritis groups to study-recruited controls and the four studies that compared arthritis groups to normative data. Because the number of studies is small, results are suggestive of differences between the types of controls. For studies using recruited controls, the mean effect size was negative, suggesting a trend toward more negative self-concept among children with arthritis than controls (d = -.26; CI = -.55, .03), although for studies using normative data, the mean effect size was positive, suggesting a trend toward more positive self-concept among children with arthritis than controls (d = .46; CI = -.13, 1.05). The number of studies in each group was small, however, and the standard error large, so effect sizes did not differ from zero. Effect sizes for studies using recruited controls were, however, significantly lower than effect sizes for studies using normative data, t(7) = 2.36, p = .05, indicating a tendency for comparisons with normative data on self-concept measures to underestimate self-concept problems of children with arthritis.
Comparisons of Samples With Chronic Arthritis Only Versus Mixed Rheumatic Disease Groups. Mean effect sizes and confidence intervals were calculated separately for the four studies with mixed disease groups and for the five studies including only participants with arthritis. Effect sizes did not differ significantly from zero for either comparison. The difference between effect sizes for studies using mixed samples and studies using only participants with arthritis was not significant, suggesting that the inclusion of mixed disease samples did not influence study findings on measures of self-concept.
Calculation of the Fail-Safe N
Published studies may be a biased sample of the studies that are actually
carried out, since nonsignificant findings are rarely published
(Rosenthal & Rosnow, 1991;
Wolf, 1986). The "fail-safe N" is an estimate of the
number of additional studies averaging nonsignificant results that would be
necessary to reverse the conclusion drawn from a meta-analysis (e.g., bring
the average effect size below a particular criterion value; Cooper, 1979;
Rosenthal & Rosnow, 1991).
Calculation of the fail-safe N was conducted using guidelines
provided by Orwin (1983) for
outcome measures that showed significant differences between children with
arthritis and controls. Seven additional studies averaging nonsignificant
results would be required to bring the average effect size for total
adjustment problems down to d = .20, a small effect size based on the
guidelines provided by Cohen
(1977). Eighteen studies would
be required to bring the average effect size for internalizing problems down
to d = .20. This would constitute a 47% increase in the existing
number of studies of overall problems and a 138% increase in the existing
number of studies of internalizing problems meeting criteria for inclusion in
the meta-analysis.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAsterisk indicates studies... |
|---|
Results of the meta-analysis indicate that children and adolescents with chronic arthritis are at increased risk for psychological adjustment problems in comparison with controls, according to parent report. Specifically, children with arthritis appear to be at risk for the development of overall adjustment problems and internalizing symptoms, but not for externalizing problems. Youths with arthritis were not found to be at greater risk for poor self-concept or self-esteem problems than controls were. For the purposes of this meta-analysis, effect sizes indicate the degree to which, in standard deviation units, the average participant with chronic arthritis differs from the average control child. The average arthritis child was .30 standard deviation above the average control child on overall adjustment problems, .47 standard deviation above the average control child on internalizing problems (effect sizes between the small and medium range for behavioral research), and only .04 and .06 standard deviation above the average control child on externalizing problems and self-concept, respectively.
It is of interest that children and adolescents with chronic arthritis
appear to be more likely to exhibit internalizing than externalizing symptoms.
The pattern of greater risk for internalizing problems is consistent with
findings from epidemiological research and literature reviews of psychological
adjustment in pediatric chronic illness
(Gortmaker, Walker, Weitzman, & Sobol,
1990; Lavigne &
Faier-Routman, 1992; Thompson
& Gustafson, 1996). It is not entirely clear, however, why
children with arthritis display more internalizing than externalizing
problems. One possibility is that the physical limitations associated with
this disease constrain "acting out" behaviors, such as getting
into fights. The gender distribution of children within studies may have
influenced results. Although depression occurs at approximately the same rate
in childhood, by adolescence the female-to-male ratio is approximately 2:1
(Birmaher et al., 1996).
Because all but one of the studies in the meta-analysis included adolescents
in their samples, and the majority of children with arthritis were girls,
these factors may have contributed toward the higher rate of internalizing
problems in this population.
Participants with arthritis had significantly higher rates of overall
psychological adjustment problems when compared to study-recruited controls,
but not when compared to normative data. In their meta-analytic review of
psychological adjustment to pediatric physical disorders, Lavigne and
Faier-Routman (1992) found
that the more careful the matching of chronic illness and control samples on
demographic characteristics, the greater the effect sizes for comparisons
between the two groups on adjustment measures. As studies using controls in
this meta-analysis generally used careful matching procedures, such procedures
may have contributed to the larger effect sizes for comparisons using
study-recruited controls.
Studies including children and adolescents with rheumatic diseases other
than chronic arthritis demonstrated significantly larger differences from
control samples on overall adjustment problems than did studies with pure
arthritis samples. This difference may be related to greater risk for
complications in other rheumatic diseases than in chronic arthritis, including
risk of mortality and more significant involvement of other organ systems such
as the heart, lungs, liver, and central and peripheral nervous systems
(Cassidy & Petty, 1995),
and requirement for more complex medical care. These findings suggest the
importance of greater specificity in future research (e.g., presenting results
for diagnostic groups separately, as well as for the total sample in studies
using mixed rheumatic disease populations).
Although youths with arthritis were not found to be at greater risk for
poor self-concept than controls, the number of studies was small. Nonetheless,
examination of the impact of the type of control group used suggests that
researchers should use caution in drawing conclusions about the self-concept
of children with arthritis based solely on comparison with normative data, as
this practice may underestimate the impact of arthritis on self-esteem. Of
note, several of the studies in the meta-analysis reported data on
domain-specific assessments of competency (e.g., scholastic competence,
athletic competence, physical appearance, social acceptance, etc.) in addition
to total self-concept/self-esteem scores or assessments of global self-worth.
Although these were too few in number to include in the meta-analysis, future
research with this population may benefit from examination of the ways in
which children's assessments of specific competencies relate to their global
sense of self-worth, an issue addressed only in one study to date
(Aasland & Diseth,
1999).
Several potential limitations of this review should be addressed. As noted earlier, published research tends to be biased in favor of significant findings, since nonsignificant findings are rarely published (Wolf, 1986). Calculation of the fail-safe N, however, indicates that 18 additional studies with nonsignificant results would be necessary to alter the findings of increased risk for internalizing problems among children and adolescents with chronic arthritis. It is unlikely that an additional 18 studies, all nonsignificant, on this topic remain unpublished.
Although statistical differences were found between levels of adjustment problems in arthritis and control samples, suggesting the importance of assessing for adjustment problems in this population, the results of the meta-analysis do not directly address the issue of clinical significance. It is not known, for example, how many children with chronic arthritis experience problems that would meet criteria for a psychiatric diagnosis and require clinical intervention. Examination of the mean T scores and standard deviations on the CBCL and YSR for studies included in the meta-analysis, however, can provide information regarding the percentage of children and adolescents who scored above clinical cutoffs on these measures. Assuming that T scores were normally distributed in the study samples, across studies, 13% to 45% (median = 20.5%) of youths with arthritis scored above clinical cutoffs for internalizing problems.
This meta-analysis relied heavily on studies using the CBCL as a measure of
child psychological adjustment. Concerns have been raised regarding the use of
the CBCL in research with children with a chronic illness
(Perrin, Stein, & Drotar,
1991). The CBCL and YSR include eight items that refer to physical
symptoms "without medical cause," the majority of which are scored
on the somatic complaints subscale and contribute toward the internalizing
problems score. Because the designation "without medical cause" is
open to subjective interpretation by the respondent, endorsement of
illness-related symptoms (e.g., aches/pains or fatigue for youths with
arthritis) could artificially inflate internalizing problem scores
(Perrin et al., 1991).
It is unclear to what extent endorsement of somatic items on the CBCL may
have elevated internalizing scores in the studies included in the
meta-analysis. For three of these studies, authors examined the impact of
dropping physical symptom items from raw scores on the CBCL. Vandvik
(1990) and Noll et al.
(2000) found that when somatic
components were subtracted from internalizing raw scores, differences between
arthritis and control group scores were no longer significant. Daltroy et al.
(1992), however, found that
adjusted and unadjusted scores were extremely highly correlated with one
another (r =.98), and that removal of somatic items reduced the
percentage of their sample falling in the clinical range for overall problems
only slightly, from 18% to 15%. Unfortunately, none of these studies reported
adjusted means and standard deviations on the CBCL following the removal of
physical symptom items from the internalizing scale. Therefore, it was not
possible to determine whether the removal of physical symptom items would
change the magnitude of the effect size for the difference between scores of
children with arthritis versus controls and thus influence the results of this
meta-analysis. Future research may benefit from use of measures with
demonstrated validity in populations of children with chronic illness. One
example is the Personal Adjustment and Roles Skills Scale III
(Stein & Jessop, 1990;
Walker, Stein, Perrin, & Jessop,
1990), which does not include items about physical symptoms that
could bias results.
Other potential concerns regarding the use of the CBCL with children with
chronic illness may be less relevant to the results of this study. Caution has
been advised in making inferences about the frequency of behavioral problems
based on comparison to CBCL norms instead of to recruited control groups. The
CBCL standardization sample excluded children with recent mental health
referrals; thus, comparison to norms could exaggerate differences from a
study's sample of interest (Perrin et al.,
1991). It is unlikely that use of normative data artificially
inflated effect sizes for this meta-analysis, however, because greater risk
for adjustment problems was found in studies using recruited controls than for
studies using comparison to norms. Additionally, concerns have been raised
regarding the limited ability of the CBCL to detect subtle adjustment problems
among youths with chronic illness (Perrin
et al., 1991). In this study, however, clinically relevant
differences from controls were of greater concern than variations within the
normal range.
Another limitation of the meta-analysis is the potential impact of rater
bias on results. Assessment of internalizing and externalizing symptoms in the
studies included in the meta-analysis was based almost exclusively on parent
report. For the one study that included both child and parent report of
adjustment problems, no analyses were provided comparing problem reports by
rater. Of note, a handful of studies of adjustment to chronic arthritis have
included child self-report in the form of depression and anxiety symptom
rating scales, but as noted earlier, not a sufficient number to include in the
meta-analysis. It is unclear to what extent inclusion of child report of
adjustment problems would change the nature of the results of the
meta-analysis, however. Studies comparing parent and child reports of
psychiatric symptoms and problem behaviors have found that children and
adolescents tend to report more internalizing problems than their parents
(Edelbrock, Costello, Dulcan, Conover,
& Kala, 1986; Herjanic
& Reich, 1982; Rey,
Schrader, & Morris-Yates, 1992). If this population follows
the pattern found by others, the average effect size for the difference
between youths with chronic arthritis and controls on internalizing problems
likely would actually increase based on child report. Clearly, more routine
inclusion of the child's perspective in future research with this population
will be necessary for a more complete picture of adjustment to this
illness.
One of the strengths of this review is its use of a meta-analytic approach to integrate the seemingly inconsistent results among studies investigating risk for adjustment problems in children and adolescents with chronic arthritis. A recent review of meta-analytic studies in PsychInfo shows that more than 200 have been published overall, with the technique applied to both basic questions about the nature of disorders and to the effects of treatments for different problems. The procedure is applied much less commonly to address problems in the clinical child and pediatric psychology areas, with no more than 20 papers reported in PsychInfo. Although comparisons between arthritis samples and controls may not reach statistical significance in individual studies, this quantitative integration of findings across studies allows for a clearer view of the direction and magnitude of differences between these two groups. Particularly, the results of this study suggest the importance of assessment for internalizing problems among children and adolescents with chronic arthritis.
Given the results of this meta-analysis, an important direction for future research may be to move beyond the broad focus on internalizing problems toward the examination of specific domains of psychopathology or adjustment that may account for higher scores on parent-report measures of internalizing symptoms among children and adolescents with chronic arthritis. For example, it may be useful to look more carefully at symptoms of anxiety and depression, or at other psychosocial domains, such as peer relations, withdrawal, or dependency. A clearer understanding of the specific areas in which these youths face challenges will be important for improving screening and intervention efforts.
Whereas the results of the meta-analysis suggest that children and
adolescents with chronic illness, as a group, face greater risk for the
development of internalizing problems, there are significant individual
differences in response to the life stress of chronic illness
(Wallander & Thompson,
1995). Examination of factors associated with resilience or poor
adjustment will be important in promoting positive outcomes for these
children. At the time the literature review was conducted for this study, we
reviewed a number of studies examining correlates of adjustment among youths
with chronic arthritis. Although there was not a sufficient number of studies
examining any specific risk or resistance factor to conduct a meta-analysis of
correlates of adjustment, several areas appear promising for further
investigation, including the child's social support and social involvement
(Timko, Stovel, Baumgartner, & Moos,
1995; Ungerer et al.,
1988), family coping strategies
(Degotardi et al., 1999), and
child perceived burden of illness (Ennett
et al., 1991). As more research is conducted in this area, a
quantitative review of findings may be helpful in integrating results.
|
Acknowledgments |
|---|
We thank Sara Knight, PhD, for her thoughtful comments on a draft of this manuscript.
|
Notes |
|---|
This paper was reviewed and accepted during the term of the previous editor, Anne. E. Kazak, PhD, ABPP.
Received October 12, 2001; accepted March 14, 2002
|
Asterisk indicates studies included in the meta-analysis |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAsterisk indicates studies... |
|---|
*Aasland, A., & Diseth, T. H. (1999). Can the Harter Self-Perception Profile for Adolescents (SPPA) be used as an indicator of psychosocial outcome in adolescents with chronic physical disorders? European Child and Adolescent Psychiatry, 8, 78-85.[CrossRef][ISI][Medline]
Achenbach, T. M., & Edelbrock, C. S. (1983). Manual for the Child Behavior Checklist and revised profile. Burlington: University of Vermont, Department of Psychiatry.
Achenbach, T. M., & Edelbrock, C. S. (1987). Manual for the Youth Self-Report and profile. Burlington: University of Vermont, Department of Psychiatry.
Ansell, B. M., & Chamberlain, M. A. (1998). Children with chronic arthritis: The management of transition to adulthood. Bailliere's Clincial Rheumatology, 12, 363-374.
*Baldaim, E. M., Holt, P. J. L., Conway, S. C., &
Morton, M. J. S. (1995). The association between physical
function and psychological problems in children with juvenile chronic
arthritis. British Journal of Rheumatology,
34, 470-477.
Barlow, J., Harrison, K., & Shaw, K. (1998). The experience of parenting in the context of juvenile chronic arthritis. Clinical Child Psychology and Psychiatry, 3, 445-463.[Abstract]
Billings, A. G., Moos, R. H., Miller, J. J., & Gottlieb, J. E. (1987). Psychosocial adaptation in juvenile rheumatic disease: A controlled evaluation. Health Psychology, 6, 343-359.[CrossRef][ISI][Medline]
Birmaher, B., Ryan, N. D., Williamson, D. E., Brent, D. A., Kaufman, J., Dahl, R. E., Perel, J. P., & Nelson, B. (1996). Childhood and adolescent depression: A review of the past 10 years. Part I. Journal of the American Academy of Child and Adolescent Psychiatry, 35, 1427-1439.[CrossRef][ISI][Medline]
*Brace, M. J., Smith, M. S., McCauley, E., & Sherry, D. D. (2000). Family reinforcement of illness behavior: A comparison of adolescents with chronic fatigue syndrome, juvenile arthritis, and healthy controls. Developmental and Behavioral Pediatrics, 21, 332-339.
Brewer, E. J., Bass, J., Baum, J., Cassidy, J. T., Fink, C., Jacobs, J., Hanson, V., Levinson, J. E., Schaller, J., & Stillman, J. S. (1977). Current proposed revision of JRA criteria. Arthritis & Rheumatism, 20, 195-199.[Medline]
Cassidy, J. T., & Petty, R. E. (1995). Textbook of pediatric rheumatology (3rd ed.). Philadelphia: W. B. Saunders.
Cohen, J. (1977). Statistical power analysis for the behavioral sciences (Rev. ed.). New York: Academic press.
Cooper, H. M. (1984). The integrative research review: A systematic approach. Beverly Hills, CA: Sage.
*Daltroy, L. H., Larson, M. G., Pless, I. B., Rogers, M.
P., & Liang, M. H. (1992). Psychosocial adjustment in
juvenile arthritis. Journal of Pediatric Psychology,
17, 277-289.
*Daniels, D., Moos, R. H., Billings, A. G., & Miller, J. J. (1987). Psychosocial risk and resistance factors among children with chronic illness, healthy siblings, and healthy controls. Journal of Abnormal Child Psychology, 15, 295-308.[CrossRef][ISI][Medline]
*Degotardi, P. J., Revenson, T. A., & Ilowite, N. T. (1999). Family-level coping in juvenile rheumatoid arthritis: Assessing the utility of a quantitative family interview. Arthritis Care and Research, 12, 314-324.
Edelbrock, C., Costello, A. J., Dulcan, M. K., Conover, N. C., & Kala, R. (1986). Parent-child agreement of child psychiatric symptoms assessed via structured interview. Journal of Child Psychology and Psychiatry, 27, 181-190.[ISI][Medline]
Ennett, S. T., DeVellis, B. M., Earp, J., Kredich, D., Warren, R.,
& Wilhelm, C. (1991). Disease experience and psychosocial
adjustment in children with juvenile rheumatoid arthritis: Children's versus
mothers' reports. Journal of Pediatric Psychology,
16, 557-568.
European League Against Rheumatism (EULAR). (1977). Nomenclature and classification of arthritis in children. Basel: National Zeitung AG.
*Flato, B., Sorskaar, D., Vinje, O., Lien, G., Aasland, A., Moum, T., & Oystein, F. (1998). Measuring disability in early juvenile rheumatoid arthritis: Evaluation of a Norweigan version of the Childhood Health Assessment Questionnaire. Journal of Rheumatology, 25, 1851-1858.
*Frank, R. G., Hagglund, J. J., Schopp, L. H., Thayer, J. F., Vieth, A. Z., Cassidy, J. T., Goldstein, D. E., Beck, N. C., Clay, D. L., Hewett, J. E., Johnson, J. C., Chaney, J. M., & Kashani, J. H. (1998). Disease and family contributors to adaptation in juvenile rheumatoid arthritis and juvenile diabetes. Arthritis Care and Research, 11, 166-176.
*Garstein, M. A., Short, A. D., Vannatta, K. V., & Noll, R. B. (1999). Psychosocial adjustment of children with chronic illness: An evaluation of three models. Developmental and Behavioral Pediatrics, 20, 157-163.
Glass, G. V., McGaw, B., & Smith, M. L. (1981). Meta-analysis in social research. Beverly Hills, CA: Sage.
Gortmaker, S. L., Walker, D. K., Weitzman, M., & Sobol, A. M.
(1990). Chronic conditions, socioeconomic risks, and behavioral
problems in children and adolescents. Pediatrics,
85, 267-276.
*Hagglund, K. J., Vieth, A. Z., Sadler, C. S., Johnson, J. C., & Hewett, J. E. (2000). Caregiver personality characteristics and adaptation to juvenile rheumatic disease. Rehabilitation Psychology, 45, 242-259.[CrossRef]
*Harris, J. A., Newcomb, A. F., & Gewanter, H. L. (1991). Psychosocial effects of juvenile rheumatic disease: The family and peer systems as a context for coping. Arthritis Care and Research, 4, 123-130.
Harter, S. (1985). Manual for the Self-Perception Profile for Children. Denver, CO: University of Denver.
Herjanic, B., & Reich, W. (1982). Development of a structured psychiatric interview for children: Agreement between child and parent on individual symptoms. Journal of Abnormal Child Psychology, 10, 307-324.[CrossRef][ISI][Medline]
*Huygen, A. C. G., Kuis, W., & Sinnema, G.
(2000). Psychological, behavioural, and social adjustment in
children and adolescents with juvenile chronic arthritis. Annals of
the Rheumatic Diseases, 59,
276-282.
Kewman, D. G., Warschausky, S. A., & Engel, L. (1995). Juvenile rheumatoid arthritis and neuromuscular conditions: Scoliosis, spinal cord injury, and muscular dystrophy. In M. C. Roberts (Ed.), Handbook of pediatric psychology (2nd ed., pp. 384-402). New York: Guilford Press.
Konkol, L., Lineberry, J., Gottlieb, J., Shelby, P. E., Miller, J. J., & Lorig, K. (1989). Impact of juvenile arthritis on families: An educational assessment. Arthritis Care and Research, 2, 40-48.
Lavigne, J. V., & Faier-Routman, J. (1992).
Psychological adjustment to pediatric physical disorders: A meta-analytic
review. Journal of Pediatric Psychology,
17, 133-157.
*Lavigne, J. V., Ross, C. K., Berry, S. L., Hayford, J. R., & Pachman, L. M. (1992). Evaluation of a psychological treatment package for treating pain in juvenile rheumatoid arthritis. Arthritis Care and Research, 5, 101-110.
Lovell, D .J., Athreya, B., Emery, H. M., Gibbas, D. L., Levinson, J. E., Lindsley, C. B., Spencer, C. H., & White, P. H. (1990). School attendance and patterns, special services and special needs in pediatric patients with rheumatic diseases. Arthritis Care and Research, 3, 196-203.
*McAnarney, E. R., Pless, I. B., Satterwhite, B., &
Friedman, S. B. (1974). Psychological problems of children with
chronic juvenile arthritis. Pediatrics,
53, 523-528.
Miller, J. J. (1993). Psychosocial factors related to rheumatic diseases in childhood. Journal of Rheumatology, 20, 1-11.
*Noll, R. B., Kozlowski, K., Gerhardt, C., Vannatta, K., Taylor, J., & Passo, M. (2000). Social, emotional, and behavioral functioning of children with juvenile rheumatoid arthritis. Arthritis and Rheumatism, 43, 1387-1396.[CrossRef][ISI][Medline]
Oen, K. G., & Cheang, M. (1996). Epidemiology of chronic arthritis in childhood. Seminars in Arthritis and Rheumatism, 26, 575-591.[CrossRef][ISI][Medline]
Orwin, R. G. (1983). A fail-safe N for effect size. Journal of Educational Statistics, 8, 157-159.
Perrin, E. C., Stein, R. E. K., & Drotar, D.
(1991). Cautions in using the Child Behavior Checklist:
Observations based on research about children with a chronic illness.
Journal of Pediatric Psychology,
16, 411-421.
Peterson, L. S., Mason, T., Nelson, A. M., O'Fallon, W. M., & Gabriel, S. E. (1996). Juvenile rheumatoid arthritis in Rochester, Minnesota 1960-1993: Is the epidemiology changing? Arthritis & Rheumatism, 39, 1385-1390.[ISI][Medline]
Quirk, M. E., & Young, M. H. (1990). The impact of JRA on children, adolescents, and their families: Current research and implications for future studies. Arthritis Care and Research, 3, 36-43.
Rey, J. M., Schrader, E., & Morris-Yates, A. (1992). Parent-child agreement of children's behaviours reported by the child behaviour checklist (CBCL). Journal of Adolescence, 15, 219-230.[CrossRef][ISI][Medline]
Rosenthal, R., & Rosnow, R. L. (1991). Essentials of behavioral research: Methods and data analysis. New York: McGraw-Hill.
Stein, R. E. K., & Jessop, D. J. (1990). Manual for Personal Adjustment and Role Skills Scale III (PARS III). Bronx, NY: Albert Einstein College of Medicine.
Taylor, J., Passo, M. H., & Champion, V. L. (1987). School problems and teacher responsibilities in juvenile rheumatoid arthritis. Journal of School Health, 57, 186-190.
Thompson, R. J., & Gustafson, K. E. (1996). Adaptation to chronic childhood illness. Washington, DC: American Psychological Association.
Timko, C., Stovel, K. W., Baumgartner, M., & Moos, R. H. (1995). Acute and chronic stressors, social resources, and functioning among adolescents with juvenile rheumatoid disease. Journal of Research on Adolescence, 5, 361-385.
*Timko, C., Stovel, K. W., Moos, R. H., & Miller, J. J. (1992). Adaptation to juvenile rheumatic disease: A controlled evaluation of functional disability with a one-year follow-up. Health Psychology, 11, 67-76.[CrossRef][ISI][Medline]
Towner, S. R., Michet, C .J., O'Fallon, W. M., & Nelson, A. M. (1983). The epidemiology of juvenile arthritis in Rochester, Minnesota 1960-1979. Arthritis & Rheumatism, 26, 1208-1213.[ISI][Medline]
*Ungerer, J. A., Chaitow, J., & Champion, G. D.
(1988). Psychosocial functioning in children and young adults
with juvenile arthritis. Pediatrics,
81, 195-202.
*Van der Net, J., Kuis, W., Prakken, A. B. J., Lukkassen, I., Sinnema, G., Hutter, A. P., Brackel, H. J. L., de Wilde, E. J., & Helders, P. J. M. (1997). Correlates of disablement in systemic onset juvenile chronic arthritis. Scandinavian Journal of Rheumatology, 26, 188-196.[ISI][Medline]
*Vandvik, I. H. (1990). Mental health and psychosocial functioning in children with recent onset of rheumatic disease. Journal of Child Psychology and Psychiatry, 31, 961-971.[ISI][Medline]
Varni, J. W., & Jay, S. M. (1984). Biobehavioral factors in juvenile rheumatoid arthritis: Implications for research and practice. Clinical Psychology Review, 4, 543-560.[CrossRef]
Walker, D. K., Stein, R. E. K., Perrin, E. C., & Jessop, D. J. (1990). Assessing psychosocial adjustment of children with chronic illnesses: A review of the technical properties of PARS III. Journal of Developmental and Behavioral Pediatrics, 11, 116-121.[ISI][Medline]
Wallander, J. L., & Thompson, R. J. (1995). Psychosocial adjustment of children with chronic physical conditions. In M. C. Roberts (Ed.), Handbook of pediatric psychology (2nd ed., pp. 124-141). New York: Guilford Press.
*Wallander, J. L., Varni, J. W., Babani, L., Banis, H.
T., & Wilcox, K. T. (1988). Children with chronic physical
disorders: Maternal reports of their psychological adjustment.
Journal of Pediatric Psychology,
13, 197-212.
*Weiss, K. A., Schiaffino, K. M., & Ilowite, N. (2001). Predictors of sibling relationship characteristics in youth with juvenile chronic arthritis. Children's Health Care, 30, 67-77.
Wolf, F. M. (1996). Meta-analysis: Quantitative methods for research synthesis. Beverly Hills, CA: Sage.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. Ding, A. Hall, K. Jacobs, and J. David Psychological functioning of children and adolescents with juvenile idiopathic arthritis is related to physical disability but not to disease status Rheumatology, May 1, 2008; 47(5): 660 - 664. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Zebracki, K. Holzman, K. J. Bitter, K. Feehan, and M. L. Miller Brief Report: Use of Complementary and Alternative Medicine and Psychological Functioning in Latino Children with Juvenile Idiopathic Arthritis or Arthralgia J. Pediatr. Psychol., September 1, 2007; 32(8): 1006 - 1010. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. A. Karsdorp, W. Everaerd, M. Kindt, and B. J.M. Mulder Psychological and Cognitive Functioning in Children and Adolescents with Congenital Heart Disease: A Meta-Analysis J. Pediatr. Psychol., June 1, 2007; 32(5): 527 - 541. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Riddle, C. N. Ryser, A. A. Morton, J. D. Sampson, R. H. Browne, M. G. Punaro, and R. J. Gatchel The Impact on Health-Related Quality of Life from Non-Steroidal Anti-Inflammatory Drugs, Methotrexate, or Steroids in Treatment for Juvenile Idiopathic Arthritis J. Pediatr. Psychol., April 1, 2006; 31(3): 262 - 271. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Rodenburg, G. J. Stams, A. M. Meijer, A. P. Aldenkamp, and M. Dekovic Psychopathology in Children with Epilepsy: A Meta-Analysis J. Pediatr. Psychol., September 1, 2005; 30(6): 453 - 468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. S. LeBovidge, J. V. Lavigne, and M. L. Miller Adjustment to Chronic Arthritis of Childhood: The Roles of Illness-Related Stress and Attitude Toward Illness J. Pediatr. Psychol., April 1, 2005; 30(3): 273 - 286. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Olsson, M. F. Boyce, J. W. Toumbourou, and S. M. Sawyer The Role of Peer Support in Facilitating Psychosocial Adjustment to Chronic Illness in Adolescence Clinical Child Psychology and Psychiatry, January 1, 2005; 10(1): 78 - 87. [Abstract] [PDF]
|
|
|
|
 |
|
 R. A. Boothroyd and M. I. Armstrong Comorbidity and Unmet Service Needs Among Medicaid-Enrolled Children With Identified Disabilities Journal of Emotional and Behavioral Disorders, January 1, 2005; 13(1): 43 - 51. [Abstract] [PDF]
|
|
|
|
|
|
M. J. Sandstrom and L. E. Schanberg Brief Report: Peer Rejection, Social Behavior, and Psychological Adjustment in Children with Juvenile Rheumatic Disease J. Pediatr. Psychol., January 1, 2004; 29(1): 29 - 34. [Abstract] [Full Text] [PDF]
|