School Functioning in Adolescents With Chronic Pain: The Role of
Depressive Symptoms in School Impairment
Deirdre E. Logan, PHD, Laura E. Simons, PHD, and Karen J. Kaczynski, PHD
Children’s Hospital Boston, Harvard University Medical School
Objective To explore associations between depressive symptoms and school functioning, including
school attendance, academic performance, self-perceived academic competence, and teacher-rated school
adjustment among predominantly Caucasian and female adolescent chronic pain patients. Methods A total
of 217 clinically referred adolescents (aged 12–17 years) and their parents completed measures of pain
characteristics, depression, and school functioning. Additional data were collected from school records and
teacher reports. Results Depressive symptoms strongly correlated with school functioning indicators. In linear
regression analyses, higher levels of depressive symptoms predicted more school impairment. A model testing
whether depressive symptoms mediated the association between current pain intensity and parent perceptions
of the interference of pain on school functioning was supported by the data. Conclusions Depressive
symptoms play a key role in influencing the extent of school impairment in adolescents with chronic pain.
Interventions to alleviate depressive symptoms may enhance treatments designed to improve school
functioning in this population.
Key words
adolescents; depression; pain; school functioning.
Pediatric chronic pain is a pervasive problem, affecting an
estimated 15–20% of children (APS, 2001; Goodman &
McGrath, 1991). A subset of these children experience
pain-related functional disability that can severely decrease
quality of life in many areas (Gauntlett-Gilbert & Eccleston,
2007; Perquin et al., 2000). In a recent large communitybased study of school-age children in Germany, 30% of
children in the age group of 4–18 years reported pain
lasting 6 months or longer, with half of these children
endorsing significant functional impairment due to pain
(Roth-Isigkeit, Thyen, Stoven, Schwarzenberger, &
Schmucker, 2005). Other recent studies have shown
that pain characteristics, such as duration and intensity,
cannot fully explain the functional outcomes of chronic
pain conditions in children and adolescents (Eccleston
et al., 2004; Kashikar-Zuck, Goldschneider, Powers,
Vaught, & Hershey, 2001; Logan & Scharff, 2005).
Given the vast individual differences in subjective pain
experiences and functional outcomes, a driving question
in pain research is why some individuals with chronic pain
function adequately, whereas others with comparable
subjective pain are significantly disabled by it. In accordance with the biopsychosocial model of pain (Engel,
1977; Gatchel, Peng, Peters, Fuchs, & Turk, 2007), a
number of factors that appear to account in part for individual differences in functional disability in the face of
chronic pain have been studied, including child coping
styles (Compas, Connor-Smith, Saltzman, Thomsen, &
Wadsworth, 2001; Simons, Claar, & Logan, 2008), psychological distress (Campo, Comer, Jansen-Mcwilliams,
Gardner, & Kelleher, 2002; Kashikar-Zuck et al., 2001),
family pain history (Schanberg et al., 2001), and parental
responses to pain (Claar, Simons, & Logan, 2008;
Peterson & Palermo, 2004). Depressive symptomatology
has emerged as a particularly powerful contributor to
pain-related functional disability.
The frequent co-occurrence of chronic pain and psychological distress—particularly depression—has been
established in both adults (Bair, Robinson, Katon, &
Kroenke, 2003; Campbell, Clauw, & Keefe, 2003) and
children (Campo et al., 2002; Hoff, Palermo, Schluchter,
Zebracki, & Drotar, 2006; Hunfeld et al., 2001;
All correspondence concerning this article should be addressed to Deirdre E. Logan, PhD, Pain Treatment Service,
Children’s Hospital Boston, 333 Longwood Avenue Suite 549, Boston, MA 02115.
E-mail: [email protected]
Journal of Pediatric Psychology 34(8) pp. 882–892, 2009
doi:10.1093/jpepsy/jsn143
Advance Access publication January 30, 2009
Journal of Pediatric Psychology vol. 34 no. 8 ß The Author 2009. Published by Oxford University Press on behalf of the Society of Pediatric Psychology.
All rights reserved. For permissions, please e-mail: [email protected]
Depression, Pain, and School Impairment
Kashikar-Zuck et al., 2001). Although a relatively small
proportion of pediatric chronic pain patients meet full
diagnostic criteria for a depressive disorder, subclinical
depressive symptomatology is common in this population.
Youth whose pain is comorbid with depression exhibit
more functional impairment than those without comorbid
psychological distress (e.g., Palermo, 2000). In a pediatric
musculoskeletal pain sample, Kashikar-Zuck and colleagues (2001) found levels of depressive symptoms to
be strongly associated with global functional disability,
accounting for 45% of the variance in disability (compared
to 13% accounted for by pain ratings). A similar pattern
of findings was reported among children with chronic or
recurrent abdominal pain (Claar & Walker, 2006).
Although prospective studies are rare, some recent work
shows depression to be an important influence on disability in children—particularly girls—over time (El-Metwally,
Salminen, Auvinen, Kautiainen, & Mikkelsson, 2004; Hoff
et al., 2006; Mulvaney, Lambert, Garber, & Walker, 2006).
Taken together, existing research supports the view that
youth with comorbid pain and depressive symptoms are
at elevated risk for global functional disability and that
depression may magnify the detrimental effects of pain
on functioning.
Existing studies view functional disability broadly,
typically defined by scores on the Functional Disability
Index (FDI; Walker & Greene, 1991) or similar checklists
that do not explore specific domains of functioning in
depth. Although few studies to date conceptualize different domains of functional disability in pediatric pain
patients, there is some evidence to suggest that functional disability is a complex construct, with different
domains of functioning influenced by different variables
(Gauntlett-Gilbert & Eccleston, 2007). Because school is
the primary venue for success or failure in the developmental tasks of childhood and adolescence, this specific
domain of functioning merits more detailed scrutiny
(Eccleston, Crombez, Scotford, Clinch, & Connell, 2004;
Walker & Johnson, 2004), Of particular interest is how
pain and depressive symptoms may combine to influence
school functioning.
Frequent school absence is well documented among
adolescents with chronic pain (Chan, Piira, & Betts, 2005;
Eccleston et al., 2004; Walker & Johnson, 2004). Studies
of laboratory-induced pain show that lower pain thresholds
in healthy children predict school absence (Tsao, Glover,
Bursch, Ifekwunigwe, & Zeltzer, 2002; Tsao & Zeltzer,
2003), suggesting links between pain perception and
school attendance. Beyond elevated absence rates, adolescents with chronic pain identify many other school difficulties, including declines in academic performance,
more school-related stress and poor academic competence
(Chan et al., 2005; Claar, Walker, & Smith, 1999; Logan,
Simons, Stein, & Chastain, 2008; Torsheim & Wold,
2001). Yet these domains of functioning have been less
well studied in the chronic pain population despite their
demonstrated links to school success and positive adjustment in the developmental and school psychology
literatures (Cole et al., 2001; Eccles et al., 2006).
Building on previous work that expanded the scope of
school functioning beyond attendance data and revealed
significant school impairment among adolescents with
chroinc pain (Logan et al., 2008), the current study
focuses specifically on the influence of depressive symptoms on the pathway from chronic pain to school impairment. Specific hypotheses are that (1) higher levels of
depressive symptoms will be associated with greater
impairment in school functioning, as measured by a
range of indicators (school absences, academic performance, self-perceived academic competence, and teacher
ratings of academic adjustment); and (2) depressive symptoms will mediate the relation between pain characteristics
and school functioning.
Methods
Participants
Participants are consecutively recruited adolescents (aged
12–17 years) who presented for initial evaluation of
chronic or recurrent functional pain symptoms at a tertiary
pediatric chronic pain outpatient clinic within a large
urban children’s hospital, between October 2005 and
June 2007. As part of this clinic visit they received
medical, psychological, and physical therapy evaluations.
Study inclusion criteria were minimum pain duration of
3 months at time of enrollment, pain frequency of at least
once per week or 5 days per month, and current enrollment in middle school or high school. Patients were
excluded from study participation if they had (1) specific
organic etiology of pain (e.g., cancer pain, juvenile rheumatoid arthritis), (2) severe cognitive impairment by history, or (3) inability to speak sufficient English to complete
questionnaire measures.
Of 289 eligible participants, 217 participants made up
the final sample (7.6% declined, 14.6% signed consents
but failed to fully complete or return questionnaire packets, and 1.8% were deleted from analyses due to substantial missing data). The final sample was primarily female
(80.2%) and Caucasian (92.9%), reflecting the characteristics of the patients seen in this clinic setting and similar to sample characteristics reported in other studies
based in tertiary-care pediatric chronic pain clinics
883
884
Logan, Simons, and Kaczynski
(Eccleston et al., 2004; Guite, Logan, Sherry, & Rose,
2007). Participants had a mean age of 14.7 years
(SD ¼ 1.6). The responding parent was the adolescent’s
mother for 92% of the sample. Overall, parents in the
sample were highly educated, with a median education
level of college graduate for both mothers (education
level ranged from ‘‘some high school’’ to ‘‘professional/
graduate school’’) and fathers (education level ranged
from ‘‘first grade to eighth grade’’ to ‘‘professional/
graduate school’’). Approximately 30% of the sample presented with neuropathic pain complaints; another 30% had
musculoskeletal or joint pain problems; about 25% complained of headaches; and approximately 10% had
abdominal pain, gynecologic, or genitourinary complaints.
A more detailed description of participants’ primary pain
complaints, along with additional characteristics of the
sample and differences between participants and nonparticipants, is reported elsewhere (Logan et al., 2008).
Procedure
Human subjects’ approval was granted by the hospital IRB
(institutional review board) prior to commencing data collection. Data were collected during the school year only,
so that current-year school data could be obtained for all
participants. Potentially eligible families were identified in
advance of clinic appointments, and letters were mailed to
their homes to explain the purpose of the study. Families
wishing not to be approached about the study during their
clinic visit were invited to send back a stamped postcard
indicating this. At the time of clinic visit, the study coordinator met with the family to explain the study in detail and
obtain informed consent from both adolescent and parent
participants. A separate consent form was used for families
to grant permission for the research team to contact
the adolescent’s school to obtain study information.
Consenting families then completed self-report measures.
Measures
Pain characteristics
Time since pain onset was recorded in months since
onset of current pain problem, by parent report. Pain
intensity over the past week was measured at a single
time point on a standard 11-point numeric rating scale
with anchors of 0 ¼ ‘‘no pain’’ and 10 ¼ ‘‘worst pain possible.’’ Adolescents provided ratings for both their current
and average pain.
Depressive symptoms
The Children’s Depression Inventory (CDI; Kovacs,
1985) self-report form was used to obtain adolescents’ ratings of current depressive symptomatology
(expressed continuously). The CDI contains 27 selfreport items rated on a 3-point scale from 0 to 2,
summed to obtain a total score. Higher scores indicate
higher levels of depressive symptoms. The CDI has been
found to have adequate reliability and validity (Kovacs,
1985). Alpha reliability in this sample was .85.
School functioning
To yield a comprehensive view of school functioning, multiple indicators of school functioning were obtained from
multiple reporters. These included the following:
1. School attendance: Official attendance records were
collected from schools to track school attendance for
the month prior to the date on which families
completed questionnaires. This information was
used to create a variable indicating percentage of
days on which the adolescent was absent, arrived
late, and/or was dismissed early. Parents and
adolescents also provided subjective ratings of the
perceived effects of pain on school attendance.
This was measured with a single item worded, ‘‘How
much has pain interfered with your attendance at
school?’’ Responses were recorded on 10-cm visual
analogue scale (VAS) lines with anchors of 0 ¼ ‘‘did
not interfere at all’’ to 10 ¼ ‘‘interfered extremely.’’
2. Academic performance: Parents reported the average
grades their adolescent received in the year prior to
onset of the pain problem and grades currently
received, using a multiple-choice format with
options including ‘‘mostly A’s,’’ ‘‘A’s and B’s,’’
‘‘mostly B’s,’’ ‘‘B’s and C’s,’’ ‘‘mostly C’s,’’
‘‘C’s and D’s,’’ ‘‘mostly D’s,’’ ‘‘D’s and F’s,’’ and
‘‘mostly F’s.’’ Responses were used to derive an
indicator of change in grades since onset of pain
based on the extent of decline or improvement
(i.e., the number of categories that performance fell
or improved from prior to pain onset to current
time, coded so that higher scores indicate larger
declines in performance since pain onset). Despite
possible subjectivity, parent report was considered
the best source of information on grades because
current teachers were not necessarily aware of
students’ grades in previous school years. Parents
and adolescents also provided subjective ratings of
the perceived effects of pain on school performance
by responding to the question, ‘‘How much has
pain interfered with your performance at school?’’
Responses were recorded on a 10-cm VAS with
anchor points identical to the item assessing effects
of pain on school attendance.
Depression, Pain, and School Impairment
3. Self-perceived academic competence: The six-item
Scholastic subscale from the Harter Self-Perception
Profile for Adolescents (SPPA; Harter, 1988) was
used to assess adolescents’ perceptions of their own
academic competence. Most participants were
administered the measure in its traditional format,
wherein adolescents first determine which of two
statements is more like them and then rate how true
the statement is (i.e., ‘‘sort of true for me’’ vs.
‘‘really true for me’’). Participants recruited in the
final 9 months of the study (N ¼ 84) received the
measure in an alternative format that has been
shown to be equivalent to the original but easier
to complete (Wichstrom, 1995). In this format,
participants were given the initial statement from
each of the items and asked to rate how true that
statement is for them (from ‘‘not at all true’’ to
‘‘really true’’). Difference in scores between those
who completed the traditional or alternative formats
in this sample was nonsignificant, F(1, 179) ¼ 0.04,
p ¼ .85. Both formats use a 4-point scale with item
scores averaged to yield the subscale score. Higher
scores indicate higher self-perceived competence
(i.e., less impairment). Cronbach’s alpha scores for
the current sample are .79 for the original version
and .82 for the revised format version.
4. Teacher perceptions of school adjustment: Adolescents
were asked to designate a current teacher who
knows them well. This teacher completed the
Walker–McConnell scales of social competence and
school adjustment, adolescent version (WMS;
Walker & McConnell, 1995). The WMS is a 53-item
scale assessing teacher perceptions of how
frequently students display various academic and
social skills in the classroom setting. It includes the
domains of self-control (e.g., ‘‘copes effectively with
being upset’’), peer relations (e.g., ‘‘interacts with
a number of different peers’’), school adjustment,
(e.g., ‘‘is personally well organized’’), and empathy
(e.g., ‘‘is sensitive to the needs of others’’). Items
are answered on a 5-point scale from 1 ¼ ‘‘never’’
to 5 ¼ ‘‘frequently,’’ with higher scores indicating
better adjustment. Total scale score is used for
data analyses. For the current sample, Cronbach’s
alpha for the total scale score ¼ .95.
Data Analyses
For purposes of data reduction and streamlining of
analyses, composite scores were generated by averaging
the two pain interference questions (i.e., school attendance, academic performance) for each respondent.
These scores are referred to as perceptions of pain interference, adolescent, and parent report. Higher scores indicate greater interference from pain. All other school
functioning indicators were examined separately. Data
were analyzed using SPSS (Version 14.0 for Windows).
For preliminary analyses, means and standard deviations
for variables of interest were examined along with bivariate
associations among participant and family demographics,
pain characteristics, and depressive symptoms via Pearson
product moment correlations. Subsequent analyses
entailed multivariate hierarchical linear regression analyses.
All analyses were run on the available sample with pairwise
deletions for missing data.
Results
Descriptive and Preliminary Analyses
The means and standard deviations for key variables are
presented in Table I. Regarding depressive symptoms, the
majority of the sample scored within the normal range on
the CDI, whereas 12.6% of the sample scored in the ‘‘at
-risk’’ range (i.e., CDI total T score ¼ 60–69) and 7.4%
scored in the ‘‘clinically significant’’ range (i.e., CDI total
T score 70). Level of depressive symptoms was associated with participant age, with older adolescents reporting higher levels of symptoms (r ¼ .16, p < .05); age was,
therefore, included in subsequent multivariate analyses.
Associations between demographic and pain characteristics and school functioning indicators have been reported
elsewhere (Logan et al., 2008); parental education levels
were significantly associated with several parent-reported
indicators and one teacher-reported indicator of school
functioning (i.e., lower education levels were associated
Table I. Means and Standard Deviations for Variables
Variable
Pain duration (months)
Range
3–132
Mean (SD)
22.5 (25.6)
Average pain rating
0.5–10
Current pain rating
0–10
5.6 (2.8)
34–86
50.9 (10.5)
CDI total T score
Percent of full þ Partial days missed
6.0 (2.0)
0–100
33.5 (32)
WMS total score
94–265
206.23 (40.4)
Harter SPPA scholastic competence scale
1.2–4.0
3.1 (0.6)
0–10
0–10
4.9 (3.2)
4.9 (3.5)
Perceptions of pain interference (adolescent)
Perceptions of pain interference (parent)
Children’s Depression Inventory: CDI; SPPA: Harter Self-Perception Profile for
Adolescents; WMS: Walker–McConnell Scales of Social Competence and School
Adjustment; normative sample mean for the WMS total scale score ¼ 176.32
(SD ¼ 39.6); normative sample mean for the SPPA scholastic competence
scale ¼ 2.8 (0.5). Because normative samples for each measure are balanced by
gender and the study sample is 80% female, comparisons between normative and
study samples are greatly limited.
885
886
Logan, Simons, and Kaczynski
with reports of poorer functioning). Due to a strong intercorrelation between paternal and maternal levels of education, only paternal education—which correlated with
more outcomes of interest than maternal education—was
included as a covariate in the relevant multivariate analyses. Associations between gender and depressive symptoms and pain characteristics were explored but found to
be nonsignificant.
Bivariate Pearson correlation analyses were conducted
to examine associations among the primary constructs of
interest, namely, level of depressive symptoms, pain characteristics, and the various indicators of school functioning
(see Table II). Higher levels of depressive symptoms were
strongly associated with most school impairment indicators, including school absences, adolescent and parent
perceptions of the interference of pain on attendance and
performance, and adolescent reports of lower scholastic
competence. The correlation between depressive symptoms and teacher reports of poorer school adjustment
was in the expected direction but not statistically significant. Notably, ratings of average pain intensity and time
of pain onset were not significantly correlated with depressive symptoms nor with any school impairment variables.
Current pain intensity was only weakly correlated (p < .05)
with depressive symptoms and parent perceptions of pain
interference scores.
linear regression analyses were conducted using depressive
symptoms as the primary independent variable and including age and paternal education level to control for the
effects of these demographic characteristics. Separate hierarchical regression equations were generated for the following dependent variables representing school impairment:
(1) percent of full and partial school days missed (school
report), (2) teacher ratings of school adjustment, (3) adolescent perceptions of pain interference, (4) adolescent
perceptions of scholastic competence, (5) parent perceptions of pain interference, and (6) parent report of grade
changes since pain onset. Age and (when relevant) paternal
education were entered on the first step of the equations,
and CDI score was entered on the second step.
Complete results of these analyses, including effect
sizes and confidence intervals, are presented in
Table III. The regression analyses indicate that levels of
depressive symptoms significantly predicted schoolreported absences, teacher report of school adjustment,
adolescent and parent perceptions of pain interference
with school, adolescents’ perceived scholastic competence,
and parent report of grade changes since pain onset.
Tests of Mediation
To test the second hypothesis that depressive symptoms
would mediate the relation between pain characteristics
and school functioning, we followed the procedures outlined in Baron and Kenney (1986) and further clarified in
Holmbeck (2002), supplemented with bootstrapping
methods to apply the Sobel test (Preacher & Hayes,
2004) to determine statistical significance of mediation.
Regression Analyses
To more fully examine our first hypothesis that higher
levels of depressive symptoms would be associated with
greater impairment in school functioning, multivariate
Table II. Bivariate Pearson Correlation Matrix
Variable
N
1
–
2
3
4
5
6
7
8
9
10
Pain characteristics
1. Time since pain onset
184
2. Average pain intensity
197
3. Current pain intensity
197
.04
.12
–
.36**
–
.05
.14
.07
.15
.02
.05
.17
.14
.09
.13
.08
.09
.08
.05
.17
.05
.19
.13
.01
.02
.05
.23*
.32**
.33**
.19*
.43**
Depressive symptoms
4. CDI total score
206
–
.14
School functioning
5. Percentage of full and partial days missed
160
6. Parent PPI
187
7. Adolescent PPI
174
8. Parent report of grade change
175
9. Adolescent perceived scholastic comp.
10. Teacher WMS
181
120
–
.68**
–
.60**
.22*
.17
.28*
.81**
.41**
.21*
.34**
–
.39**
–
.27**
.36**
.18
.47**
–
.18
–
PPI: perception of pain interference score; WMS: Walker McConnell Scales, School Adjustment subscale score. Correlations are two-tailed. To adjust for multiple comparisons,
only p values < .01 are indicated as significant.
*p < .01; **p < .001.
Depression, Pain, and School Impairment
Table III. Regression Analyses: Relations Between Depressive Symptoms and School Functioning
Variables
Beta
95% CI for beta
0.17
0.72
.01
.04
3.14 to 3.47
3.88 to 2.43
1.48
.33
0.76 to 2.20
b
t
r2 [CI] change
Cohen’s f
.01
.10***
0.12
I. Outcome: Days missed (partial þ full)
Step 1: Age
Step 2: Age
Depressive symptoms
0.10
0.45
4.06***
II. Outcome: Teacher WMS score
Step 1: Age
0.00
.01
0.10 to 0.10
0.05
Paternal education
0.12
.17
0.02 to 0.25
1.67
Step 2: Age
Paternal education
0.01
0.10
.03
.15
0.09 to 0.11
0.03 to 0.24
0.28
1.53
0.02
.19
0.04 to 0.01
1.92*
Depressive symptoms
.01
.04*
0.06
III. Outcome: Adolescent PPI
Step 1: Age
0.10
.02
0.55 to 0.75
0.30
Step 2: Age
0.07
.02
0.69 to 0.55
0.22
0.34
.33
0.19 to 0.48
Step 1: Age
0.01
.01
0.06 to 0.07
0.15
Step 2: Age
0.03
.07
0.03 to 0.08
0.96
0.04
.44
0.05 to 0.03
9.40***
Depressive symptoms
.01
.10***
0.11
4.47***
IV. Outcome: SPPA schol. comp.
Depressive symptoms
.01
.18***
0.23
V. Outcome: Parent PPI
Step 1: Age
Paternal education
Step 2: Age
Paternal education
Depressive symptoms
0.50
.11
0.18 to 1.18
0.96
.16
1.89 to 0.04
1.45
.03*
02.06*
0.32
.07
0.34 to 0.98
0.95
0.86
.14
1.75 to 0.02
1.92
0.31
.30
0.16 to 0.46
.11***
0.14
4.05***
VI. Outcome: gGrade change (parent rep.)
Step 1: Age
0.04
.04
0.17 to 0.10
0.52
Step 2: Age
0.06
.07
0.20 to 0.07
0.93
0.04
.20
0.01 to 0.08
Depressive symptoms
.001
.03**
0.04
2.64**
SPPA Schol. Comp: Scholastic Competence Subscale of Self-Perception Profile for Adolescents; PPI: Perception of Pain Interference Score. Cohen’s f2 effect sizes of 0.02, 0.15,
and 0.35 are considered small, medium, and large, respectively (Cohen, 1988).
*p < .05; **p < .01; ***p < .001.
The tested mediation models were guided by the results of
the bivariate correlation analyses, although weaker correlations (p < .05 within the context of multiple comparisons)
were considered worth further examination. Specifically,
because neither average pain ratings nor pain duration
correlated significantly with any of the school functioning
outcome variables, the tests of mediation focused on examining whether depressive symptoms mediated associations
between current pain severity and school impairment.
Separate mediational models were tested for each of the
school functioning–outcome variables: (1) percent of
full and partial school days missed (school report),
(2) teacher ratings of school adjustment, (3) adolescent
perceptions of pain interference, (4) adolescent perceptions of scholastic competence, (5) parent perceptions of
pain interference, and (6) parent report of grade changes
since pain onset.
Following conventional methods of assessing mediation (Baron & Kenny, 1986), it was first established that
the independent variable, current pain intensity, significantly predicted CDI scores (b ¼ .17, t ¼ 2.25, p < .05).
The regression analyses reported in Table III demonstrate
that CDI scores significantly predicted all six school functioning outcomes, satisfying an additional criteria for
establishing depressive symptoms as a mediator of the
association between pain intensity and school functioning.
Next, current pain intensity had to be established as a
significant predictor of the school functioning–outcome
variables. Current pain intensity did not significantly
predict school attendance, teacher ratings of school
adjustment, adolescent perceptions of pain interference,
adolescent perceptions of scholastic competence, or
parent-reported grade change. Only the regression of
parent perceptions of pain interference on current pain
887
888
Logan, Simons, and Kaczynski
Table IV. Depressive Symptoms as a Mediator of the Association Between Current Pain Intensity and Parent Perceptions
of Pain Interference
Variables
b
Beta
95% CI for beta
T
r2 change
Cohen’s f
Outcome: parent PPI
Step 1
Age
Paternal education
.57
.13
.65
.10
.16 to 1.30
1.66 to .37
1.54
1.26
.03
.04*
.03
.06***
.12
Step 2
Age
Paternal education
Current pain
Step 3
Age
Paternal education
.58
.13
.52
.08
1.53 to .48
.14 to 1.30
.50
.20
.10 to .89
2.47*
.42
.09
.29 to 1.12
1.17
.54
.09
1.52 to .43
Current pain
.42
.17
.03 to .80
Depressive symptoms
.26
.26
.10 to .42
1.60
1.03
1.11
2.11*
3.14**
*p < .05; **p < .01; ***p < .001.
intensity was statistically significant (b ¼ .44, t ¼ 2.29,
p < .05). The final requirement to show mediation is a
reduction in the strength of the association between the
independent variable (current pain intensity) and the
dependent variable (parent perceptions of pain interference scores) in the presence of the proposed mediator,
depressive symptoms. These results, reported in
Table IV, support a partial mediation effect. The more conservative Sobel test, assessing statistical significance of the
extent of mediation (i.e., the magnitude of the indirect
effect, accounting for error and variance) also supported
the hypothesized mediation effect (Sobel z ¼ 1.93,
p < .05).
Discussion
Building from previous work with these data documenting
broadly defined school impairment in adolescents with
chronic pain (Logan et al., 2008), the aim of the present
study was to examine associations between depressive
symptoms and school functioning in the context of chronic
pain. In this clinical sample of adolescents with complex
and varied chronic pain conditions, depressive symptoms
were associated with school impairment as measured by
multiple indicators (e.g., school-reported attendance,
parental and adolescent perceptions of academic performance and competence). This finding is consistent with
previous research demonstrating the influence of depression on functional outcomes in children and adolescents
with chronic pain (Claar & Walker, 2006; GauntlettGilbert & Eccleston, 2007; Kashikar-Zuck et al., 2001).
This study extends the literature by examining the association between depressive symptoms and school
functioning specifically. This is an important domain of
functioning that has only recently begun to receive
increased attention in pediatric chronic pain research.
Although an association between depressive symptoms and school impairment may not be surprising in
itself, the finding is more striking when combined with
the fact that pain characteristics were not associated with
many of the school impairment indicators assessed in this
study. In other words, among a group of adolescents with
chronic pain syndromes, those reporting pain of higheraverage intensity or longer duration did not necessarily
have poorer school functioning. Instead, level of depressive
symptoms was a more consistent indicator of school
impairment.
Notably, the study sample was generally characterized
by moderate levels of depressive symptoms, with only 20%
of the sample falling in the at-risk or clinically significant
range of depressive symptomatology. These rates are not
dramatically higher than the rates of depression in the general adolescent population (e.g., Orvaschel, Beeferman, &
Kabacoff, 1997). Some evidence suggests that adolescents
presenting for evaluation and treatment of chronic pain
may minimize psychological distress on self-report (e.g.,
Logan & Claar, 2008); alternatively, these relatively
normal rates of depressive symptoms may indicate resilience in this population. As has been found in the adolescent depression literature (e.g., Orvaschel et al., 1997),
depressive symptoms increased with age in this predominantly female sample. It has been previously observed
that adolescent pain patients may infrequently evidence
diagnosable levels of depression, but that depressive symptoms, even at subclinical levels, may be meaningful
because they indicate poorer emotional coping resources
Depression, Pain, and School Impairment
(Eccleston et al., 2004). Given the findings in this and
previous studies that depressive symptoms are associated
with functional abilities in the face of chronic pain, even
subclinical levels of depressive symptoms deserve attention
and represent a potentially powerful target of treatment
aimed at functional restoration.
Our second hypothesis, that depressive symptoms
would mediate associations between pain characteristics
and school functioning, was partially supported by the
data. The relationship between current pain ratings and
parent perceptions of the interference of pain on school
functioning showed evidence of a mediating role for
depressive symptoms, but this meditational pathway was
not found between other pain-characteristic school functioning variable pairs. The lack of broader support for the
hypothesized mediation models was primarily due to the
fact that pain characteristics were only weakly correlated
with school functioning outcomes. Although this pattern
was contrary to our expectations, it is consistent with
recent findings by Gauntlett-Gilbert and Eccleston
(2007), who found that although pain intensity was closely
related to global functional disability it bore no significant
relation to school attendance. Synthesizing these findings
with our own, we posit that in the context of complex
chronic pain conditions in adolescence, specific pain characteristics such as pain intensity may correlate with overall
disability (especially physical disability, the primary construct measured by common checklists such as the FDI),
but may have less influence on specific domains such as
school functioning, which are likely to be more strongly
associated with psychosocial phenomena.
In recent years, there has been increased emphasis
both in the research literature and in clinical practice on
placing a primary focus on functional restoration in
patients with pediatric chronic pain (e.g., GauntlettGilbert & Eccleston, 2007; Palermo, 2000). An earlier
published study based on the current data (Logan et al.,
2008) lent additional support to the tenet that functional
disability is a multifaceted construct and highlighted the
importance of looking more closely at school functioning
in particular when investigating functional abilities of adolescents with chronic pain. The conclusion that school
functioning is significantly impaired in many adolescent
chronic pain patients raised questions about what psychological factors might influence the extent of school impairment across individuals. The current study begins to
address these questions and shows that symptoms of
depression likely influence school functioning. In light
of the growing evidence for the effectiveness of cognitive
behavior-oriented treatments for pediatric chronic pain
(Degotardi et al., 2006; Eccleston, Malleson, Clinch,
Connell, & Sourbut, 2003; Kashikar-Zuck et al., 2001),
it is useful to consider how treatments targeting depressive
symptoms and negative thinking may improve school functioning outcomes for this population.
Several limitations should be considered when evaluating this research. First, results from this demographically
homogenous sample, referred through several levels of care
to a tertiary chronic pain clinic, may not generalize to other
groups. The sample consists predominantly of female,
Caucasian adolescents from highly educated families, limiting our ability to apply results beyond this group. For
example, we were unable to examine possible gender differences in the relationship between depression and school
impairment, given the low numbers of boys included in
our sample. The specific demographics of the sample, such
as the overrepresentation of females, also limit comparisons to samples used to establish normative information
for some of the measures included in the study and may
explain relatively high scores on ratings of academic competence. However, these characteristics are reflective of the
majority of patients presenting to tertiary care pediatric
chronic pain programs. More work needs to be done to
determine why so few boys and relatively few members of
ethnic minorities are treated in these settings. Second, the
study was cross-sectional and correlational in design. As
such, it is not possible to determine causal links between
the variables of interest. Although we have conceptualized
depressive symptoms as instrumental in determining the
extent of school impairment (and used this conceptualization in developing the tested mediational models), it is
possible that in the presence of chronic pain, adolescents
may experience pain-related school impairment, which
then gives rise to depressive symptoms. Studies with
longitudinal approaches that follow the trajectories of
pain, depression, and school impairment over time are
necessary to verify the causal relations among these
constructs.
A number of measurement issues bear consideration.
Some study measures differed in the time frame assessed
for recent symptoms (e.g., ‘‘the last two weeks’’ vs. ‘‘the
previous month of school’’); this may account for some
discrepancies across the information collected. Pain characteristics assessed in this study were limited to time since
pain onset and pain intensity, which was assessed by
numeric rating scales at a single time point. Previous studies suggest that pain diaries that measure both intensity
and frequency of pain over a period of days to weeks provide a more robust indicator of pain severity (e.g., Gil et al.,
2000). This is particularly important when assessing
recurrent pain conditions where pain waxes and wanes.
Including multiple reporters (i.e., parents, teachers)
889
890
Logan, Simons, and Kaczynski
in the assessment of adolescent depression would
strengthen future studies.
Measurement limitations associated with the selected
school functioning indicators are discussed in a previous
publication (Logan et al., 2008). Briefly, there are limits
inherent in relying on adolescents’ and parents’ subjective
and retrospective views of school functioning. For example, parent reports of adolescents’ grades over a potentially
lengthy time period (i.e., prior to pain onset) are subject to
recall bias; confirming grades with school records was not
possible, given the time span involved for some participants and the possibility of participants having attended
multiple schools during this time. Several of the measures
of school functioning were single-item, unvalidated measures, so caution must be used in interpreting the findings.
Furthermore, the lack of normative data for most of the
school functioning indicators precludes comparisons of
adolescents with chronic pain to healthy peers. Study participants nominated the teacher who completed the WMS;
this ensured that teacher respondents knew the participant
well, but this came at the cost of standardizing respondents and may have led to inflated ratings. Nonetheless,
the scope of school functioning measures included in this
study expands on what has typically been assessed, thereby
increasing our knowledge of how adolescents with chronic
pain are functioning in the school context. More research
is needed to enable accurate and comprehensive assessment of school impairment in youth with chronic pain.
Future research should focus on longitudinal assessments
of school functioning and should strive to incorporate both
subjective and objective methods of assessment, including
measures such as standardized test scores with comparisons to normative results.
It is also important to note that some of the effect sizes
for our findings were relatively small. In particular, analyses with less than medium effect size (particularly the
analysis of teacher-reported school adjustment and
parent-reported grade change where effect sizes are smallest) may not indicate clinically significant findings. The
focus of the study was exclusively on depressive symptoms
and school function, but the variables examined herein are
likely part of a larger, more complex biopsychosocial
framework that accounts for the interplay between pain
characteristics, demographic qualities, and psychosocial
variables and their impact on functional outcomes of the
chronic pain experience. One important area not fully
addressed in this study is the role of the parent–child
relationship; future studies should incorporate parent–
child relationship variables, such as parental responses to
pain, to understand how these may affect school functioning outcomes in the presence of chronic pain and
depressive symptoms. Another important direction for
future research is to compare the pattern of associations
between depressive symptoms and school functioning
found in this sample of adolescents with chronic pain to
patterns among other chronically ill adolescents and
among healthy peers. Only such comparison studies can
clarify whether these patterns are unique to adolescents
with chronic pain syndromes or whether similar patterns
exist outside the context of chronic pain.
In conclusion, study results contribute to an emerging
literature base on specific domains of pain-related functional disability. They lend further evidence to the finding
that when school outcomes are the focus of pain-related
functional disability assessment, psychosocial variables
such as depressive symptoms (even symptomatology
below the level of psychiatric diagnosis) are more closely
linked to functional outcomes than are specific pain characteristics such as intensity or duration. Levels of depressive symptoms appear to explain, at least in part, why some
youth with chronic pain function adequately in the school
setting, whereas others are severely disabled. These results
set the stage for future research to better understand how
depression may interact with chronic pain to influence
school functioning, and how these factors can inform functional rehabilitative approaches to pediatric chronic pain
treatment.
Acknowledgments
The authors thank Michelle Stein, Laura Chastain, and
Laura Blackwell for research assistance, and the adolescents and parents who participated in the study.
Funding
Klingenstein Third Generation Foundation (to D. E. L.).
Conflict of interest: None declared.
Received May 28, 2008; revisions received December 12,
2008; accepted December 14, 2008
References
American Pain Society. (2001). Pediatric chronic pain.
Glenview, IL: American Pain Society.
Bair, M. J., Robinson, R. L., Katon, W., & Kroenke, K.
(2003). Depression and pain comorbidity: A literature
review. Archives of Internal Medicine, 163, 2433–2445.
Baron, R. M., & Kenny, D. A. (1986). The moderatormediator variable distinction in social psychological
Depression, Pain, and School Impairment
research: Conceptual, strategic, and statistical
considerations. Journal of Personality and Social
Psychology, 51, 1173–1182.
Campbell, L. C., Clauw, D. J., & Keefe, F. J. (2003).
Persistent pain and depression: A biopsychosocial
perspective. Biological Psychiatry, 54, 399–409.
Campo, J. V., Comer, D. M., Jansen-McWilliams, L.,
Gardner, W., & Kelleher, K. J. (2002). Recurrent pain,
emotional distress, and health service use in childhood. Journal of Pediatrics, 141, 76–83.
Chan, E., Piira, T., & Betts, G. (2005). The school
functioning of children with chronic and recurrent
pain. Pediatric Pain Letter, 7, 11–16.
Claar, R. L., Simons, L. E., & Logan, D. E. (2008). Parental
response to children’s pain: The moderating impact of
children’s emotional distress on symptoms and
disability. Pain, 138, 172–179.
Claar, R. L., & Walker, L. S. (2006). Functional assessment
of pediatric pain patients: Psychometric properties of
the functional disability inventory. Pain, 121, 77–84.
Claar, R. L., Walker, L. S., & Smith, C. A. (1999).
Functional disability in adolescents and young adults
with symptoms of irritable bowel syndrome: The role
of academic, social, and athletic competence. Journal
of Pediatric Psychology, 24, 271–280.
Cohen, J. (1988). Statistical power analysis for the behavioral
sciences (2nd ed.). Hillsdale, NJ: Lawrence Earlbaum
Associates.
Cole, D. A., Maxwell, S. E., Martin, J. M., Peeke, L. G.,
Seroczynski, A. D., Tram, J. M., et al. (2001).
The development of multiple domains of child and
adolescent self-concept: A cohort sequential
longitudinal design. Child Development, 72,
1723–1746.
Compas, B. E., Connor-Smith, J. K., Saltzman, H.,
Thomsen, A. H., & Wadsworth, M. E. (2001).
Coping with stress during childhood and
adolescence: Problems, progress, and potential in
theory and research. Psychological Bulletin, 127,
87–127.
Degotardi, P. J., Klass, E. S., Rosenberg, B. S., Fox, D. G.,
Gallelli, K. A., & Gottlieb, B. S. (2006). Development
and evaluation of a cognitive-behavioral intervention
for juvenile fibromyalgia. Journal of Pediatric
Psychology, 31, 714–723.
Eccles, J. S., Roeser, R., Vida, M., Fredricks, J., &
Wigfield, A. (2006). Motivational and achievement
pathways through middle childhood. In L. Balter &
C. S. Tamis-LeMonda (Eds.), Child psychology: A
handbook of contemporary issues (2nd ed.,
pp. 325–355). New York: Psychology Press.
Eccleston, C., Crombez, G., Scotford, A., Clinch, J.,
& Connell, H. (2004). Adolescent chronic pain:
Patterns and predictors of emotional distress in
adolescents with chronic pain and their parents. Pain,
108, 221–229.
Eccleston, C., Malleson, P. N., Clinch, J., Connell, H.,
& Sourbut, C. (2003). Chronic pain in adolescents:
Evaluation of a programme of interdisciplinary
cognitive behaviour therapy. Archives of Disease in
Childhood, 88, 881–885.
El-Metwally, A., Salminen, J. J., Auvinen, A.,
Kautiainen, H., & Mikkelsson, M. (2004).
Prognosis of non-specific musculoskeletal pain in
preadolescents: A prospective 4-year follow-up study
till adolescence. Pain, 110, 550–559.
Engel, G. L. (1977). The need for a new medical model: A
challenge for biomedicine. Science, 196, 129–136.
Gatchel, R., Peng, Y., Peters, M., Fuchs, P., & Turk, D.
(2007). The biopsychosocial approach to chronic pain:
Scientific advances and future directions. Psychological
Bulletin, 133, 581–624.
Gauntlett-Gilbert, J., & Eccleston, C. (2007). Disability in
adolescents with chronic pain: Patterns and predictors
across different domains of functioning. Pain, 131,
132–141.
Gil, K. M., Porter, L., Ready, J., Workman, E., Sedway, J.,
& Anthony, K. K. (2000). Pain in children and
adolescents with sickle cell disease: An analysis
of daily pain diaries. Children’s Health Care, 29,
225–241.
Goodman, J. E., & McGrath, P. J. (1991). The epidemiology of pain in children and adolescents: A review.
Pain, 46, 247–264.
Guite, J. W., Logan, D. E., Sherry, D. D., & Rose, J. B.
(2007). Adolescent self-perception: Associations with
chronic musculoskeletal pain and functional disability.
Journal of Pain, 8, 379–386.
Harter, S. (1988). Manual for the self-perception profile for
adolescents. Denver: University of Colorado.
Hoff, A. L., Palermo, T. M., Schluchter, M., Zebracki, K.,
& Drotar, D. (2006). Longitudinal relationships of
depressive symptoms to pain intensity and functional
disability among children with disease-related pain.
Journal of Pediatric Psychology, 31, 1046–1056.
Holmbeck, G. N. (2002). Post-hoc probing of significant
moderational and mediational effects in studies of
pediatric populations. Journal of Pediatric Psychology,
27, 87–96.
Hunfeld, J. A., Perquin, C. W., Duivenvoorden, H. J.,
Hazebroek-Kampschreur, A. A., Passchier, J., van
Suijlekom-Smit, L. W., et al. (2001). Chronic pain
891
892
Logan, Simons, and Kaczynski
and its impact on quality of life in adolescents and
their families. Journal of Pediatric Psychology, 26,
145–153.
Kashikar-Zuck, S., Goldschneider, K. R., Powers, S. W.,
Vaught, M. H., & Hershey, A. D. (2001). Depression
and functional disability in chronic pediatric pain.
Clinical Journal of Pain, 17, 341–349.
Kovacs, M. (1985). The Children’s Depression,
Inventory (CDI). Psychopharmacology Bulletin, 21,
995–998.
Logan, D. E., & Claar, R. L. (2008). Social desirability
response bias and self-report of psychological
distress in pediatric chronic pain patients. Pain, 136,
366–372.
Logan, D. E., & Scharff, L. (2005). Relationships between
family and parent characteristics and functional abilities in children with recurrent pain syndromes:
An investigation of moderating effects on the pathway
from pain to disability. Journal of Pediatric Psychology,
30, 698–707.
Logan, D. E., Simons, L. E., Stein, M., & Chastain, L.
(2008). School impairment in adolescents with
chronic pain. Journal of Pain, 9, 407–416.
Mulvaney, S., Lambert, E. W., Garber, J., & Walker, L. S.
(2006). Trajectories of symptoms and impairment for
pediatric patients with functional abdominal pain:
A 5-year longitudinal study. Journal of the American
Academy of Child and Adolescent Psychiatry, 45,
737–744.
Orvaschel, H., Beeferman, D., & Kabacoff, R. (1997).
Depression, self-esteem, sex, and age in a child and
adolescent clinical sample. Journal of Child Clinical
Psychology, 26, 285–289.
Palermo, T. M. (2000). Impact of recurrent and chronic
pain on child and family daily functioning: A critical
review of the literature. Journal of Developmental and
Behavioral Pediatrics, 21, 58–69.
Perquin, C. W., Hazebroek-Kampschreur, A. A.,
Hunfeld, J. A., Bohnen, A. M., van Suijlekom-Smit, L.
W., Passchier, J., et al. (2000). Pain in children
and adolescents: A common experience. Pain, 87,
51–58.
Peterson, C. C., & Palermo, T. M. (2004). Parental
reinforcement of recurrent pain: the moderating
impact of child depression and anxiety on
functional disability. Journal of Pediatric Psychology, 29,
331–341.
Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS
procedures for estimating indirect effects in simple
mediation models. Behavior Research Methods for
Instrumental Computing, 36, 717–731.
Roth-Isigkeit, A., Thyen, U., Stoven, H.,
Schwarzenberger, J., & Schmucker, P. (2005).
Pain among children and adolescents: Restrictions in
daily living and triggering factors. Pediatrics, 115,
e152–e162.
Schanberg, L. E., Anthony, K. K., Gil, K. M., Lefebvre, J. C.,
Kredich, D. W., & Macharoni, L. M. (2001). Family
pain history predicts child health status in children
with chronic rheumatic disease. Pediatrics, 108, E47.
Simons, L. E., Claar, R. L., & Logan, D. E. (2008).
Chronic pain in adolescence: Parental responses,
adolescent coping, and their impact on adolescent’s
pain behaviors. Journal of Pediatric Psychology, 33,
894–904.
Torsheim, T., & Wold, B. (2001). School-related stress,
school support, and somatic complaints. Journal of
Adolescent Research, 16, 293–303.
Tsao, J. C., Glover, D. A., Bursch, B., Ifekwunigwe, M.,
& Zeltzer, L. K. (2002). Laboratory pain reactivity and
gender: Relationship to school nurse visits and school
absences. Journal of Developmental and Behavioral
Pediatrics, 23, 217–224.
Tsao, J. C., & Zeltzer, L. K. (2003). Sex differences in painrelated symptoms and self-initiated school nurse visits
among pre-adolescents. Journal of Pain and Symptom
Management, 25, 472–480.
Walker, H. M., & McConnell, S. R. (1995). WalkerMcConnell scale of social competence and school adjustment, adolescent version. Belmont, CA: Wadsworth
Publishing.
Walker, L., & Johnson, W. (2004). Recurrent abdominal
pain and functional gastrointestinal disorders in the
school setting. In R. T. Brown (Ed.), Handbook of
pediatric psychology in the school setting. New York:
Lawrence Erlbaum.
Walker, L. S., & Greene, J. W. (1991). The functional
disability inventory: Measuring a neglected dimension
of child health status. Journal of Pediatric Psychology,
16, 39–58.
Wichstrom, L. (1995). Harter’s Self-Perception Profile for
Adolescents: Reliability, validity, and evaluation of the
question format. Journal of Personality Assessment,
65(1), 100–116.