Fatigue  sleep difficulty  depression  lack of

concentration  difficulty completing tasks 

restlessness

Quality of life for children with functional abdominal pain:

a comparison study of patients' and parents' perceptions.
Youssef NN1, Murphy TG, Langseder AL, Rosh JR.
Author information
Abstract

OBJECTIVE:

Children with chronic abdominal pain of nonorganic origin, termed functional abdominal pain (FAP), experience school absences and social withdrawal

and report impaired physical ability. The aim of this study was to assess patients' and parents' perceptions of health-related quality of life (QoL) for

children with FAP.

METHODS:

Between October 2002 and November 2003, 209 children (including 125 girls; age: 11.2 +/- 3.5 years) and 209 parents were recruited from a pediatric

referral center. At the time of their initial evaluations, participants completed a validated, health-related QoL instrument (Pediatric Quality of Life

Inventory), which was scored on a scale of 0 (poor) through 100 (best). Children with FAP (n = 65) and their families were compared with control

groups of healthy children (n = 46) and children with histologically proven inflammatory bowel disease (IBD) (n = 42) or gastroesophageal reflux

disease (GERD) (n = 56).

RESULTS:

Children with FAP had self-reported QoL scores (score: 78) that were similar to those for children with GERD (score: 80) or IBD (score: 84). Children

with FAP had lower QoL scores than did healthy children (score: 88). Parents of children with FAP reported lower QoL scores, compared with their

children's scores (scores: 70 vs 78).

CONCLUSIONS:

Children with FAP reported lower QoL, compared with their healthy peers, and had the same QoL scores as did children with IBD or GERD. Parents'

perceptions of QoL for children with FAP were lower than their children's self-reported scores. These findings highlight the clinical significance of FAP

and may provide insight into one facet of the disease's biopsychosocial etiology.

Pain, fatigue and health-related quality of life in children and

adolescents with chronic pain
Jeffrey Ira Gold, Ph.D.,  Nicole E. Mahrer, B.A., Joyce Yee, M.D., and Tonya M. Palermo, Ph.D.

Author information Copyright and License information Disclaimer

The publisher's final edited version of this article is available at Clin J Pain

See other articles in PMC that cite the published article.

Abstract
Go to:

INTRODUCTION

Chronic or recurrent pain is a common occurrence among children and adolescents, affecting as much as 25% of the pediatric
population.1, 2 Frequent complaints include abdominal pain, headache, and musculoskeletal pain. 3 Children and adolescents with chronic
pain frequently report disturbances in sleep and eating habits, reduced participation in social activities or hobbies, and school
absence,4, 5 which affects their overall sense of well-being.

One method of assessing the impact of a chronic condition on the daily lives of children is to measure health-related quality of life
(HRQOL). A comprehensive and multidimensional construct, HRQOL addresses an individual’s subjective perception of his or her
functioning across multiple domains, including physical, emotional, social, and school functioning. Children with chronic pain report
significantly lower HRQOL than healthy peers 6–8 and children with other chronic conditions. 9,10 For example, a recent study conducted
by Varni et al. found that children with fibromyalgia self-reported severely impaired physical and psychosocial functioning, and their
HRQOL was more impaired than HRQOL in pediatric patients with cancer and other rheumatologic diseases. 9 Among adolescents with
chronic pain, the greater the intensity and frequency of pain, the lower the self-reported quality of life. 7 Studies examining children with
the specific conditions of spina bifida and juvenile arthritis have also found that pain is an important predictor of quality of life. 11, 12

Fatigue is another physical symptom affecting a large portion of the pediatric population. Fatigue has been defined as an overwhelming sense of
tiredness, lack of energy, and feeling of exhaustion.13 Approximately one third of adolescents experience substantial fatigue four or more times a
week.14 In pediatric pain populations, there has been limited description of fatigue with the exception of studies of children with cancer and
rheumatologic conditions.19, 20

Like pain, fatigue is a critical contributor to the impact of chronic illness on multiple domains of functioning. Among survivors of childhood
cancer, for example, an inverse relationship exists between symptoms of fatigue and a patient’s positive perception of HRQOL. 15 Notably,
survivors who did not report fatigue were comparable to healthy controls on measures of functioning. Another study found that children with
chronic idiopathic musculoskeletal pain, though similar on measures of pain intensity and disability, reported greater fatigue and lower levels of
psychosocial functioning than children with juvenile chronic arthritis. 16

While it is increasingly recognized that some children and adolescents with chronic conditions report considerable disruptions to daily life and
marked declines in quality of life, the complex relationships between chronic pain, fatigue, and HRQOL are less well understood. Berrin and
colleagues conducted a recent study of children with cerebral palsy (CP) specifically examining this complex relationship between chronic pain,
fatigue and HRQOL. The study examined the roles of pain and fatigue as mediators between diagnostic subtypes of CP and school
functioning.17 The study also examined fatigue as a potential mediator between pain and school functioning. Results revealed an indirect
relationship between diagnosis and school functioning that was mediated by pain and fatigue. Fatigue partially mediated the relationship between
pain and school functioning. This study highlights the importance of pain and fatigue as potential points of intervention to possibly improve
HRQOL in youth with CP. Although Berrin and colleagues focused on youth with CP, the relationships among fatigue, pain, and aspects of
HRQOL are relevant to all youth with persistent pain. Building on the limited work that has been done in this area, we sought to test a similar
mediation model in a broader sample of youth with chronic pain.

The present study examined fatigue and its relationships with chronic pain, and HRQOL among a heterogeneous sample of children and
adolescents with chronic pain conditions seeking outpatient pain management services at two urban children’s hospitals. Based on findings of
Berrin, et al., 200717, a similar mediation model was proposed where fatigue was hypothesized to mediate the effects of pain on children’s
HRQOL, specifically their school functioning. In addition, similar to previous research on proxy reports of HRQOL 18, we hypothesized a low to
moderate correlation between child self report and parent proxy report of HRQOL and fatigue.

Go to:

MATERIALS AND METHODS

Participants

Eighty participants were recruited from The Childrens Hospital of Los Angeles and Oregon Health and Science University as part of larger
studies at each institution. The combined sample included 80 children and adolescents between the ages of 8 and 18 (M = 13.89, SD = 2.57)
72.5% female, and their caregivers (see Table 1). The majority of caregiver respondents were mothers (71%). Racial backgrounds included
Caucasian (67.5%) and Hispanic/Latino (21.3%). On average, participants reported moderate to severe pain intensity (M = 6.6, SD =1.6 on 10 cm
visual analog scale). Primary pain complaints were headaches (n = 32), arm/leg pain (n = 26), abdominal pain (n = 15), and back pain (n = 7).
Fifty percent of participants reported pain in more than one location. Twenty percent of participants had diagnoses of a comorbid medical
condition (e.g., juvenile rheumatoid arthritis, pancreatitis). Participants from OHSU were older than participants recruited from CHLA (t = −2.21,
p < .05). There were no differences in gender, ethnicity, pain intensity or pain location between participants at the two pain centers.

Table 1

Characteristics of chronic pain sample

Variable M (SD)

Age 13.89 2.57

Average Pain 6.6 1.6

Variable % n

Gender

Female 73 58

Male  27 22

Race

Caucasian  68 54

Hispanic/Latino  21 17
Variable M (SD)

Other  11 19

Pain Locations

Headaches  45 36

Arm/Leg  41 33

Abdominal  41 33

Back  40 32

Chest  10 8

Widespread 39 31
musculoskeletal 

Comorbid Medical Conditions

 Variable M (SD)

Yes  20 16

No  80 64

Open in a separate window

Procedures

IRB approval was obtained to conduct this study at both sites. Participants at both sites included patients currently receiving treatment at an
outpatient pediatric chronic pain clinic. Participants at CHLA were approached during their initial pain management appointment with the clinical
psychologist and the physician. Consenting participants were asked to fill out questionnaires at this initial appointment. Medical charts were later
reviewed to obtain pain diagnoses.

Eligible participants were English speaking children and adolescents between the ages of 8 and 18 who had a caregiver present. If more than one
caregiver was present, the primary caregiver (parent who spent the most time with the child) was asked to complete the measures. Participants
were excluded if they had a developmental disability, cognitive, or neurological deficit that would prevent them from comprehending and
completing the self-report assessment questionnaires (e.g., mental retardation, organic brain dysfunction). Seventy-eight children and their
caregivers were approached to participate before obtaining the 54 completed protocols. Twenty participants returned incomplete protocols, two
withdrew from the study, and one participant refused. This resulted in a CHLA participation rate of 69%. At OHSU, participants were recruited
during pain clinic appointments using the same exclusion criteria. Thirty children were approached for participation and 26 children and parents
agreed to participate (87% participation rate).

Measures

Sociodemographics
Caregivers completed a background questionnaire to assess for patient age, sex, race, and ethnicity.

Pain
A pain questionnaire was completed by the children to assess their pain over the past month. The questionnaire included assessment of pain
intensity and location. Pain intensity was measured using a 10 cm Visual Analog Scale ranging from 0 indicating “no pain” to 10 indicating
“worst pain imaginable”. Location of pain was recorded by markings on a validated body outline displaying an anterior and posterior view of the
body. Comorbid medical diagnoses were obtained from children’s medical records.

Fatigue
Children’s level of fatigue was assessed by self-report and caregiver proxy report using the PedsQL ™ Multidimensional Fatigue scale.19 Previous
research has demonstrated excellent reliability and validity of the measure, and individual comparisons have distinguished between healthy
children and children with fatigue secondary to rheumatologic disease.20 The PedsQL™ Multidimensional Fatigue scale includes assessment of
general fatigue (6 items; e.g., “I feel too tired to do things that I like to do”), sleep/rest fatigue (6 items; e.g., “I rest a lot”), and cognitive fatigue
(6 items; “It is hard for me to remember what people tell me”). Questions are asked about the prior month. Higher scores indicate fewer
difficulties related to fatigue.
Health Related Quality of Life
Children’s health related quality of life was assessed by self report and caregiver-proxy report using the PedsQL ™ 4.0 Generic Core Scale, which
has well established reliability and validity in children with both acute and chronic health conditions.6, 17, 21 The PedsQL™ 4.0 Generic Core Scale
has four subscales: physical functioning, emotional functioning, social functioning, and school functioning. All scales demonstrate high
reliability.22 Each scale uses a Likert 5-point scale to ask the child or caregiver how much of a problem each item has been over the past month (0
= never a problem, 1 = almost never a problem, 2 = sometimes a problem, 3 = often a problem, 4 = almost always a problem). 23 Raw scores are
then transformed into standard scores ranging from 0 to 100. Higher HRQOL scores indicate better health-related quality of life. Varni and
colleagues have established at-risk cutoff scores indicating poor HRQOL for the PedsQL ™ 4.0 Core Scales. These scores were determined by
approximating one standard deviation below mean scores for a normative sample of 5,972 healthy children aged 5–18 and 10,070 caregivers of
children aged 2–18.21, 24 For the purposes of this study, the authors used two scales, the school functioning subscale and the total HRQOL
summary score.

Data Analyses

Data analyses were conducted using SPSS v15.0.25 Descriptive statistics were computed to examine mean scores for child self-report and
caregiver proxy- report for the fatigue and HRQOL measures. Bivariate correlations among pain, fatigue, and HRQOL scores were examined
with Pearson and Spearman Rho correlations. Due to the multiple comparisons, a Bonferroni correction was applied reducing the significant p
value to p < .001. Agreement between child self-report and caregiver-proxy measures were assessed using intraclass correlations. A series of
independent sample t-tests were conducted to examine differences in fatigue and HRQOL scores by gender, age, and ethnicity. Multiple
regression analyses were used to test a model that fatigue would mediate the relationship between pain and HRQOL in children, specifically
school functioning and total HRQOL scores. Analyses for estimating indirect effects in mediation models described by Preacher and Hayes 26 were
followed.

Go to:

RESULTS

Caregiver-child comparisons

Descriptive statistics on the PedsQL and Multidimensional Fatigue Module for the sample are summarized in Table 2. On the child-reported
PedsQL, children were at risk for impaired total HRQOL (M = 59.04), physical (M = 51.51), emotional (M = 56.98), and school functioning (M =
55.89). Mean scores on these HRQOL subscales were lower than mean scores of children with disease-related chronic illness from previous
studies27 and lower than children with other chronic pain conditions.9 Of particular note, children were 2 standard deviations below the population
sample mean on total HRQOL and 3 below the population mean on physical functioning. In addition, the study sample scored in the at-risk range
on all measures of fatigue including general fatigue, sleep/rest fatigue, cognitive fatigue, and total fatigue. Children reported less fatigue, better
school functioning, and greater quality of life than did their caregivers via caregiver proxy-reports. There were significant differences between
child and caregiver proxy-reports on measures of emotional functioning, social functioning, school functioning, and general fatigue (Table 2). As
hypothesized, however, there were moderate levels of agreement between child and caregiver reports on all measures (Table 2).

Table 2

Scale descriptive statistics, cross-informant correlations, and differences for PedQL Generic score and Multidimensional Fatigue Module
 Child Caregiver Cross-informant Difference

N Mean (SD) Mean (SD) r t

Pain  80 6.15 (1.88) N/A N/A N/A

Total HRQOL 79 59.04 57.19 (18.74) .59** .94

(20.19)

Physical 79 51.51 47.75 (25.57) .56** 1.26

functioning  (26.77)

Emotional 79 56.98 51.13 (22.38) .56** 2.45*

functioning  (23.29)

Social functioning  79 75.97 65.95 (25.72) .50** 3.89**

(21.08)

School functioning  76 55.89 49.34 (25.20) .53** 2.17*

(25.23)
 Child Caregiver Cross-informant Difference

N Mean (SD) Mean (SD) r t

N = 80 N = 79

Fatigue Total 55.63 53.40 (20.55) .66** 1.29

(22.80)

General Fatigue  53.44 48.52 (23.41) .68** 2.18*

(27.34)

Cognitive Fatigue  61.38 62.53 (26.41) .62** −.40

(27.28)

Rest Fatigue  52.04 62.53 (26.41) .70** 1.40

 Child Caregiver Cross-informant Difference

N Mean (SD) Mean (SD) r t

(23.22)

Open in a separate window

p < .05;
*

**
p < .01

There were no significant differences in fatigue or HRQOL scores with regard to age, gender, or ethnicity.

As expected, significant relationships were found between the majority of HRQOL and fatigue variables according to both child- and caregiver–
report (Table 3).

Table 3

Bivariate Correlations among Pain, HRQOL, and Fatigue according to Child-and Caregiver-report.

CAREGIVER

Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

CHIL 1. Co-morbid -- −.090 −.07 −.103 −.182 −.08 .097 −.21 −.036 −.26 −.195
D Medical Condition 0 6 2 2

2. Pain −.09 -- −.11 −.336 .008 −.22 −.300 −.17 −.279 −.25 −.286
 CAREGIVER

Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

0 1 3 7 8

3. Physical −.19 −.176 -- .326 .388* .552* .525* .533* .443* .332 .518*
2

4. Emotional −.09 −.280 .625* -- .477* .358* .361* .535* .361* .479* .541*
1

5. Social −.17 −.251 .579* .487* -- .308 .464* .375* .396* .380* .458*
1

6. School −.05 −.107 .588* .510* .502* -- .524* .560* .535* .433* .612*
5

7. Total HRQOL −.16 −.234 .907* .793* .751* .782* -- .410* .530* .400* .535*
8

8. General −.27 −.238 .706* .514* .535* .557* .722* -- .561* .705* .890*
4
 CAREGIVER

Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

9. Cognitive −.21 −.226 .572* .508* .544* .693* .704* .728* -- .423* .804*
9

10. Rest −.23 −.251 .630* .442* .416* .513* .633* .680* .556* -- .831*
5

11. Total Fatigue −.27 −.271 .724* .558* .572* .673* .784* .921* .878* .832* --
6

Open in a separate window

Note: p < .001

*

Fatigue as a mediator of the relationship between pain and HRQOL

Total HRQOL
Fatigue was tested as a mediator between pain intensity and total HRQOL assessed by child self-report (see Table 4). In step 1, pain was
significantly associated with HRQOL reported by children (β = −.23, p = .03). In step 2, pain was significantly associated with fatigue (β = −.27,
p = .02). In step 3, fatigue was significantly associated with HRQOL after controlling for pain intensity (β =.78, p < .001). In step 4, the
relationship between pain and total HRQOL assessed by self-report was not found to be significant after controlling for fatigue (β = −.01, p > .
05), suggesting a mediating effect of fatigue. The Sobel test found this mediated effect to be statistically significant, z = −2.49, p = .01.

Table 4

Mediation analyses for total HRQOL by child and caregiver report

 Predictor Outcome Child Report Caregiver Report

β p β p

Pain Intensity Total −.23 .03 −.30 .01

HRQOL

Pain Intensity Fatigue −.27 .02 −.29 .01

Pain Intensity Fatigue Total .78 .00 .49 .00

HRQOL

Fatigue Pain Intensity Total −.01 .85 −.15 .16

HRQOL

Sobel’s Test z = −2.49, p = .01 z = −2.38, p = .02

Fatigue was also tested as a mediating variable between pain and HRQOL assessed by caregiver proxy-report. In Step 1 for the caregiver proxy
report, pain was significantly associated with HRQOL reported by caregivers (β = −.30, p = .01). In step 2, pain was significantly associated with
fatigue (β = −.29, p = .01). In step 3, fatigue was significantly associated with HRQOL after controlling for pain intensity (β =.49, p < .001). In
step 4, after controlling for fatigue, pain was not significantly associated with HRQOL (β = −.15, p > .05), suggesting a mediating effect of
fatigue. The Sobel test found this mediated effect to be statistically significant, z = −2.38, p = .02.

School Functioning
Fatigue was not found to be a mediator between pain intensity and school functioning for child self-report (see Table 5). No significant
relationship between pain and self-reported school functioning was found.

Table 5

Mediation analyses for school HRQOL by child and caregiver report

 Predictor Outcome Child Report Caregiver Report

β p β p

Pain Intensity School −.11 .35 −.22 .05

HRQOL

Pain Intensity Fatigue −.27 .02 −.29 .01

Pain Intensity Fatigue School .70 .00 .60 .00

HRQOL

Fatigue Pain Intensity School .09 .31 −.05 .61

HRQOL

Sobel’s Test z = −2.43, p = . z = −2.35, p = .02

02

Fatigue was tested as a mediating variable between pain and school functioning assessed by caregiver proxy-report. In Step 1 for the caregiver
proxy report, pain was significantly associated with school functioning (β = −.22, p = .05). In step 2, pain was significantly associated with
fatigue (β = −.29, p = .01). In step 3, fatigue was significantly associated with school functioning after controlling for pain intensity (β = .60, p < .
001). In step 4, the relationship between pain and school functioning was not found to be significant after controlling for fatigue (β = −.05, p > .
05), suggesting mediating effect of fatigue. The Sobel test found this mediated effect to be statistically significant, z = −2.35, p = .02.

Go to:

DISCUSSION

In this study, we examined the relationships among pain, fatigue, and HRQOL in a heterogeneous sample of children receiving treatment for
chronic pain. Although previous studies have emphasized associations between either pain or fatigue and HRQOL, the integration of these
constructs has received limited attention. Similar to recent findings in a population of children with pain related to CP17, we also found fatigue to
be a mediator of the relationship between pain and overall HRQOL in a mixed population of children with chronic pain. This effect was
demonstrated in both self- and caregiver proxy-reports of HRQOL. However, unlike the Berrin model, fatigue did not demonstrate the same
mediation effects for the relationship between pain and school functioning across all informants. Rather, an indirect relationship between pain and
school functioning mediated by fatigue was found only for caregiver proxy-report. These findings suggest that fatigue may help to explain the
association between chronic pain and overall quality of life.

In addition, the current findings document that many children with chronic pain have significant problems with comorbid fatigue, highlighting the
importance of inquiries about this set of symptoms. Both caregiver- and child-reports demonstrate that children have moderate to severe
impairments in total fatigue, general fatigue, cognitive fatigue, and sleep/rest fatigue compared to the population-based normative sample 19, as
well as children with cancer or rheumatologic conditions.20 In general, children with these impairments feel that they need to rest during the day,
that their attention and memory are impaired, and that they are too tired to participate in activities.

Consistent with previous studies of children with specific chronic pain conditions, this study also provides further evidence that children with a
variety of chronic pain conditions are at-risk for poor overall quality of life. 6–10 Scores on the HRQOL scales were lower than those reported by
Varni and colleagues20, 27 in populations of children with other chronic health conditions and the scores were also well below the normative scores
for healthy children.

In line with previous cross-informant findings on the PedsQL, caregivers in our sample reported lower functioning than their children. 17, 28 One
reason for this discrepancy between self and caregiver-proxy reports of functioning may be that caregivers rely solely on behavioral and visual
cues to assess the functioning of their child.28 The child, however, also has access to internal emotional cues contributing to their assessment of
how they feel they are functioning relative to others. Additionally, the children may have adapted to and accepted their chronic pain while their
caregivers may have not.17 However, despite caregivers reporting lower functioning than their children, the cross-informant correlations revealed
moderate to high associations, suggesting that proxy-report methods are important measures of child functioning

Study findings should be interpreted in light of several limitations. We were limited to cross-sectional subjective data in analyzing the
relationships among pain, fatigue, and HRQOL. There is some overlap present among different dimensions of HRQOL that were examined in the
present study given that school functioning goes into the overall HRQOL score. Nonetheless we found different patterns of mediation among
these outcomes suggesting some independence of these variables. Prospective studies are needed to more fully explore fatigue as a potential
mechanism in the pain-HRQOL relationship and to identify the temporal sequence by which problems with pain, fatigue, and HRQOL develop
over time. In addition, a variety of factors that may impact fatigue were not examined in the present study including sleep problems, depressive
symptoms, and medication effects. For example, sleep patterns of the participants were not assessed and therefore, sleep restriction or sleep
disturbance as potential causes of fatigue could not be determined. Effects of pain medications on fatigue also could not be assessed in our sample
and remains an important area of inquiry. Another limitation is that the sample was limited to treatment-seeking patients at two chronic pain
clinics. Although the combination of two study sites enhances generalizability, the sample size was too small to fully examine group differences
on individual factors such as pain diagnosis. The addition of a control or comparison group of youth without chronic pain would have
strengthened the interpretation of findings. Additionally, though multiple raters were used to assess child functioning, teacher-report may enhance
the understanding of school and social functioning in future studies. Also, recall bias could be a potential limitation associated with asking
participants to recall their pain from the past month. More accurate pain ratings might be obtained by asking participants to prospectively record
pain. Strengths associated with the current study include use of well standardized measures to examine comorbid fatigue and replication of the
mediation model of fatigue in the relationship between chronic pain and HRQOL/school functioning, as previously noted in children with CP. 17

The study findings have several clinical implications for the assessment and treatment of children with chronic pain. Because fatigue is a
significant problem for many children, assessment and management of fatigue should be a standard part of care for children with chronic pain.
Clinicians could benefit from the use of standard self-administered questionnaires to assess fatigue routinely in their clinics. Ultimately,
practitioners could conduct comprehensive evaluations to determine best clinical practice for further treatment. Due to the variety of potential
etiologies associated with fatigue, such as the direct effects of chronic daily pain, poor sleep hygiene, and medication effects, many clinical
interventions could be offered. Clinical treatments may include cognitive-behavioral interventions, medications to help decrease fatigue, and
exercise programs, all of which require evaluation in the pediatric population.

Further research is required to better understand the role of fatigue in children with chronic pain in order to identify strategies to reduce its
impact. Because fatigue may have a large impact on children’s physical function and their ability to participate in age appropriate activities, the
methods used to cope with fatigue may be particularly salient to identify.

Impact of recurrent and chronic pain on child and family daily functioning: A
critical review of the literature.

The author reviewed the current status of research on the impact of recurrent and chronic pain on everyday functioning of children and families and
organized the research findings around the specific life contexts (e.g., school, peers) that may be affected by pain. Although findings demonstrate that
many different aspects of child and family life are affected by pain, the prevalence and severity of children's functional limitations associated with pain
remain unknown. Few treatment studies for pediatric recurrent and chronic pain have focused on enhancing children's functioning. It has been shown,
however, that functional outcomes can be improved by cognitive-behavioral interventions. Recommendations for research on functional outcomes and
implications for clinical practice are discussed. (PsycINFO Database Record (c) 2016 APA, all rights reserved)
 Sleep Disturbances in School-age Children
with Chronic Pain 
Anna C. Long, PhD, Vidhya Krishnamurthy, PhD, Tonya M. Palermo, PhD

Journal of Pediatric Psychology, Volume 33, Issue 3, April 2008, Pages 258–
268, https://doi.org/10.1093/jpepsy/jsm129

Published:

13 December 2007

 Article history

 PDF
  Split View
 Cite
 Permissions Icon Permissions
  
Share
 

Abstract
Objectives To examine associations between pain, functional outcomes, and sleep disturbances in children with chronic

pain, specifically juvenile idiopathic arthritis (JIA), sickle cell disease (SCD), and headache (HA). Sleep disturbances were

tested as a risk factor for increased functional disability and decreased health-related quality of life

(HRQOL). Methods One hundred children (JIA n = 30, SCD n = 26, HA n = 44; 8–12 years; 56% female) and their

caregivers participated. Children completed questionnaires regarding pain, depression, and functional disability. Caregivers

completed questionnaires regarding sociodemographics, child sleep habits, functional disability, and

HRQOL. Results Levels of overall sleep disturbances were above the clinical cutoff for 53% of children with chronic pain.

Sleep disturbances predicted lower physical HRQOL and higher functional disability, according to parent

report. Conclusions Sleep disturbances are common and associated with daytime functioning in school-age children with

chronic pain, suggesting that assessment and treatment of sleep problems is clinically relevant.
children, chronic pain, quality of life, sleep, sleep problems

Topic:

 arthritis, juvenile rheumatoid

 depressive disorders
 pain
 sleep disorders
 sleep
 chronic pain
 functional disability
 health-related quality of life
 parent report measures
Issue Section:

 Articles
Introduction

Sleep is noted to be essential for children's physical and mental

well-being (Baiardini, Braido, Cauglia, & Canonica, 2006). A number
of sleep problems are relatively common in middle childhood. In a
large sample of US children in grades 1–4, parents reported that
39% had a hard time getting out of bed, 35% had restless sleep, and
34% woke up in a bad mood (Liu, Liu, Owens, & Kaplan, 2005).
Additional sleep problems such as getting too little sleep, teeth
grinding in sleep, difficulty falling asleep, talking in sleep, taking a
long time to be alert in the morning, awakening once or more during
the night, and being tired during the day were observed in 20–25%
of this sample (Liu et al., 2005). Sleep disturbances collectively refer
to impairment of the ability to initiate or maintain sleep, and can be
measured by parent or child self-report, and by objective measures
(e.g., actigraphy, polysomnography). Sleep disturbances in healthy
children have been associated with behavior problems, decreased
cognitive performance, academic problems, and impairment in daily
living (Sadeh, Gruber, & Raviv, 2002). Sleep disturbances have also
been associated with children's quality of life, negatively impacting
children's physical and emotional well-being (Smaldone, Honig, &
Byrne, 2007).

Sleep disturbances are common among children with chronic or

recurrent pain conditions (Roth-Isigkeit, Thyen, Stoven,
Schwarzenberger, & Schmucker, 2005). Recent research findings
have noted a high prevalence of sleep disturbances in children
diagnosed with juvenile idiopathic arthritis (JIA) (Sawyer et
al., 2004), headache (HA) (Miller, Palermo, Powers, Scher, &
Hershey, 2003), and sickle cell disease (SCD). For instance, Miller
and colleagues found that children with migraine HAs had more
sleep disturbances in the areas of sleep onset delay, night wakings,
parasomnias, sleep anxiety, and sleep-disordered breathing
compared to a normative sample. Further, a bi-directional theory
has been proposed between pain and sleep disturbances, with pain
interfering with sleep, and sleep disturbances exacerbating pain
symptoms and complicating its management (Lewin & Dahl, 1999).
There is increasing support for this model among children with
chronic pain. For instance, self-report of sleep problems has been
shown to correlate with pain intensity ratings among children with
JIA (Bloom et al., 2002). Among children with SCD, high daily pain
has been shown to relate to poor sleep quality that night, which is in
turn predictive of high pain levels the following day (Valrie, Gil,
Redding-Lallinger, & Daeschner, 2007b). Negative mood also plays a
role in this relationship, partially mediating the effects of pain and
sleep on each other (Valrie, Gil, Redding-Lallinger, &
Daeschner, 2007a).
Less is known about the relationship between sleep and daily
functioning in children with chronic pain. Functional outcomes that
have been examined in pediatric chronic pain include functional
disability, which is defined as restriction in performing daily
activities in school, home, recreation, and social interaction, and
health-related quality of life (HRQOL), defined as how a person
perceives their health to impact physical, psychological, and social
well-being. In one sample of children and adolescents (6–18) with
migraine, parent report of sleep disturbances was related to parent
report of daily functioning, specifically child adaptability and
activities of daily living (Heng & Wirrell, 2006). One previous study
(Palermo & Kiska, 2005) has examined the relationship between
sleep disturbances and daily functioning in adolescents with chronic
pain conditions (HAs, JIA, and SCD), but less is known about these
associations in younger children with chronic pain conditions.
Palermo and Kiska (2005) found that functional disability and HRQOL
were strongly correlated with sleep disturbances, such that more
severe sleep disturbances were associated with greater functional
disability and lower HRQOL. In particular, strong associations were
found between functional disability and HRQOL and two sleep
variables, daytime sleepiness and sleep/wake problems. This study
also demonstrated a significant relationship between mood and
sleep in adolescents with chronic pain.

It is unknown whether these same relationships are present in

middle childhood. In this study, we examine associations between
pain, sleep problems and HRQOL, functional disability, and mood in
middle childhood, extending previous research that has been
conducted with adolescents. Middle childhood is an important
developmental period to examine, as both pain and sleep problems
increase in adolescence (LeResche, Mancl, Drangsholt, Saunders, &
Korff, 2005), and adolescents with chronic pain may be particularly
sensitive to sleep disturbances (Meltzer, Logan, & Mindell, 2005).
Knowledge of the relationship between pain, sleep disturbances, and
functional outcomes in middle childhood may inform developmental
models of chronic pain and sleep in children and adolescents. This
knowledge may also help to develop preventative interventions
within a developmental context.

In this cross-sectional study, we describe and compare sleep

disturbances in children with three chronic or recurrent pain
conditions: JIA, SCD, and chronic HA. JIA is a chronic condition that
consists of three subtypes: systemic, polyarticular, and
pauciarticular. Most children with JIA report mild to moderate levels
of pain, however, in 25–30% of children with JIA pain is described as
moderate to severe and can have a significant impact on children's
participation in activities (Rapoff, McGrath, & Lindsley, 2003). SCD is
a genetic blood disorder that primarily impacts individuals of African
and Mediterranean descent. Pain is a common consequence of SCD
and occurs with variable frequency and severity; on average
children report between five and seven episodes per year with each
episode lasting ∼1–3 days (Graumlich et al., 2001). HA is the most
prevalent type of chronic pain in the general pediatric population,
being identified by as many as 18.9% of youth (Perquin et al., 2000).
In the current study, HAs are categorized according to one of three
subtypes: migraine, tension, or mixed.

The goal of the current study was to examine sleep patterns and
problems in a sample of children with several different pain-related
conditions, and to examine broad associations between sleep
disturbances and functional disability and HRQOL outcomes. Based
on previous research findings, we hypothesized that across pain
conditions, more intense and frequent pain, as well as higher levels
of depression symptoms would be associated with increased sleep
disturbances. We also hypothesized that sleep disturbances would
be broadly associated with functional limitations and HRQOL
outcomes across pain conditions, such that increased sleep
disturbances would be related to increased functional limitations
and reduced HRQOL.

Method

Participants

Participants included 100 children between the ages of 8 and 12

(M = 10.22, SD = 1.37), 56% females and 44% males, and their
caregivers. The ethnicity of the children in the sample was African
American (40.4%), Caucasian (57.7%), and other ethnic background
(1.9%). Children and their caregivers were recruited as part of a
larger longitudinal study on chronic and recurrent pain at a Midwest
tertiary care children's hospital, where they were established
patients in the pediatric neurology, rheumatology, and hematology
clinics. The study was approved by the Institutional Review Board of
the study site. Children were diagnosed with either recurrent HAs
(n = 44; 43% females), JIA (n = 30; 80% females), or SCD (n = 26;
50% females). A higher proportion of female children were in the JIA
group, χ2(2, N = 100) = 10.02, p = .002), reflecting the gender
distribution of this disease. Age was similar in the three pain groups.
The ethnic distribution of the groups was significantly different, with
100% of the SCD group, 30% of the HA group, and 10% of the JIA
group reporting African-American ethnicity, χ2(2, N = 98) =
49.98, p < .001. Family income was lower in the SCD group χ2(4, N =
97) = 22.26, p=.002). Demographic characteristics of children in the
SCD group were similar to demographics of other published samples
(Schatz, 2004). Table I presents demographic and illness-related
information.
Table I.

Sociodemographic and Illness-related Characteristics

Total (n = 100)

Characteristic n

Gender (female)  56 

Racial background (minority)  41 

Family income:   

    <$10,000  13 

$10,000–29,000      23 

$30,000–49,000      19 

$50,000–69,000      18 

    >$70,000  24 

Health condition   

Headache      44 

Migraine          23 
 Total (n = 100)

Characteristic n

Tension-type          9 

Mixed headache          12 

JIA      30 

Pauciarticular          21 

Polyarticular          6 

Systemic          3 

SCD      26 

HbSS disease          22 

Sickle beta +
thalassemia          3 

HbSC disease          1 

Open in new tab

Children were recruited from four clinical sites (three suburban, one
inner-city) during specialty care visits for evaluation and treatment
of their condition. Visits were routine, such that patients were not
seen on an urgent care basis or due to an acute increase in pain or
illness intensity. Ninety percent of the families who were
approached about participating in the study agreed. Informed
consent was obtained from caregivers and assent from children.
Children and their caregivers then completed the questionnaires
and interviews while in the clinic. Families were compensated for
participating in the study with gift cards to local stores. Data from
this sample have also been used in previously published work (e.g.,
Lewandowski, Palermo, & Peterson, 2006; Palermo, Witherspoon,
Valenzuela, & Drotar, 2004).

Measures

Sociodemographics

Caregivers completed a questionnaire to provide demographic

information including the child's age, gender, ethnicity, and family
income level. Caregivers estimated annual family income in $10,000
brackets, which were coded from 1 = <$10,000 through 8 =
>$70,000.

Pain Characteristics

Children and parents reported on pain frequency and intensity over

the previous 4 weeks. For pain frequency, they reported on how
often the child experienced pain on a six-point scale with response
options ranging from “less than once a month” to “daily.” The Faces
Pain Scale (FPS), which has well-established reliability and validity,
was utilized to assess average pain intensity (Bieri, Reeve,
Champion, Addicoat, & Ziegler, 1990). The FPS consists of a series
of seven faces displaying various intensities of pain with anchors on
each end of the scale, “no pain” to “worst pain.” At the time data
collection began, the revised version of the Faces Pain Scale—
Revised (FPS-R) was not yet published (Hicks, von Baeyer, Spafford,
van Korlaar, & Goodenough, 2001). Previous research has used
these measures in assessing characteristics of chronic pain in
children (Peterson & Palermo, 2004).

Depression

Children also completed the Revised Children's Anxiety and

Depression Scale (RCADS), a self-report measure of anxiety and
depressive symptoms corresponding to several disorders in the
Diagnostic and Statistical Manual of Mental Disorders, fourth edition
(American Psychiatric Association, 2000). Children rated each item
on a 4-point scale ranging from “never” to “always,” with higher
scores indicating greater frequency. T-scores were calculated on the
basis of the child's gender and grade in school. The Major
Depressive Disorder (MDD) subscale was used in this study. The
measure has good internal consistency (α = .76 for the MDD
subscale) and its test–retest reliability over 1 week was
demonstrated to be adequate (Chorpita, Yim, Moffitt, Umemoto, &
Francis, 2000). Validity has been demonstrated through
relationships with other anxiety and depression measures.

Sleep Problems

Caregivers completed the Children's Sleep Habits Questionnaire

(CSHQ; Owens, Spirito, & McGuinn, 2000), a 45-item questionnaire
that assessed the sleep behaviors of school-aged children over a
typical week. It yields a total score (“total sleep disturbance”) and
eight subscale scores based on the primary presenting clinical sleep
problems in this age group—bedtime resistance, sleep onset delay,
sleep duration, sleep anxiety, night wakings, daytime sleepiness,
parasomnias, and sleep-disordered breathing. Caregivers rated the
frequency of each sleep behavior on a 3-point scale ranging from
“usually” (5–7 times per week) to “rarely” (0–1 time per week).
Higher scores indicate greater sleep disturbance, and a score of 41
has been established as the clinical cutoff. The measure also
assesses bed times and wake times, and has adequate internal
consistency, test–retest reliability, and validity (Owens et al., 2000).
The CSHQ has been used in previous studies with children with
recurrent and chronic pain (Bloom et al., 2002; Heng &
Wirrell, 2006). Reliability of the total sleep disturbance scale in the
current sample was α = .86.

Functional Disability

The Functional Disability Inventory (FDI) assesses a child's

limitations in completing tasks associated with daily living in the
domains of school, home, recreation, and social interaction. Sample
items include attending school, walking up stairs, reading, and doing
homework. Children and parents report the child's difficulty in
completing each task on a 5-point scale ranging from 0—“no
trouble” to 5—”impossible.” Sum scores range from 0 to 60, with
higher scores indicating greater functional disability. The validity
and reliability of this measure in assessing functional disability has
been well-established (Claar & Walker, 2006; Walker &
Greene, 1991).

Health-related Quality of Life

To assess the children's HRQOL, caregivers were administered the

Child Health Questionnaire parent form (CHQ-PF50). This measure
assesses a child's physical, emotional, and social functioning and
well-being and provides summary scores along two dimensions—
physical health and psychosocial health. Each dimension is
measured with multiple items, with higher scores indicating better
HRQOL. The Physical and Psychosocial summary scores have
adequate reliability and validity (Landgraf et al., 1998) and have
been used previously in children with chronic and recurrent pain
conditions (Houlihan, O’Donnell, Conaway, & Stevenson, 2004;
Oliveira et al., 2007).

Analyses

Descriptive statistics, frequencies and means, of sociodemographic

characteristics and key study variables were calculated. Mean
differences by health condition (HA, JIA, and SCD) were computed
using a one-way analysis of variance (ANOVA) for the pain, sleep,
depression, functional disability, and HRQOL variables. Scheffe's test
was used for post hoc comparisons to control for multiple
comparisons, and effect sizes were examined using partial eta-
squared. The proportion of children in this sample exceeding the
clinical cut-off score (raw score ≥41) for total sleep disturbances on
the CSHQ (Owens et al., 2000) was compared to published norms
using a chi-squared test. This normative community sample
consisted of 469, 4 to 10-year-old elementary school enrolled in a
suburban school district in New England (Owens et al., 2000).

Pearson product moment correlations were conducted to examine

the bivariate associations between study variables. Finally,
hierarchical multiple regressions were conducted to determine
whether sleep disturbances would be independently associated with
functional disability and HRQOL outcomes after controlling for
demographics, pain characteristics, and depression. Health
condition group membership was also accounted for by entering two
dummy-coded variables—whether or not the child was in the HA
group and whether or not the child was in the JIA group.

Results

Pain Characteristics, Depression and Functional Outcomes

Descriptive statistics for pain characteristics, depression, and

functional outcomes are presented in Table II. Significant group
differences by diagnostic group were found for pain frequency and
pain intensity. Post hoc comparisons indicated that children in the
HA group reported significantly higher pain frequency than children
with SCD, and children with SCD and HA reported significantly
higher pain intensity than children with JIA. These effect sizes were
medium. There were no significant group differences on the
children's report of depression or functional disability. Children
reported somewhat higher levels of functional disability than their
parents reported, t(80) = 1.71, p=.09. No group differences were
observed in caregiver report of functional disability or HRQOL
outcomes (Table II).
Table II.

Group Comparisons on Pain, Mood, and Daily Functioning

Omn
Total ibus
sampl parti
JIA n  SCD  Headac e n = al
= n = he n = 100  eta-
30 M  26 M  44 M ( M (S squa
(SD) (SD) SD) D) red

Child-          
report
 Omn
Total ibus
sampl parti
JIA n  SCD  Headac e n = al
= n = he n = 100  eta-
30 M  26 M  44 M ( M (S squa
(SD) (SD) SD) D) red

pain 

Pain 2.45
freque 3.30 (1.87) 4.04 3.45
ncy  (2.09)  b
  (1.51)b  (1.88)  .11 

Pain 3.33 4.52

intensi (1.73) (1.89) 4.70 4.23
ty  b, c
  b
  (1.32)c  (1.69)  .13 

Parent-
report
of child
pain           

Pain 2.17
freque 3.13 (1.66) 4.02 3.29
ncy  (1.74)  b
  (1.55)b  (1.78)  .18 

Pain 3.67 4.67

intensi (1.35) (1.93) 5.21 4.60
ty  b, c
  b
  (1.26)c  (1.20)  .17 

Child-
report
mood           
 Omn
Total ibus
sampl parti
JIA n  SCD  Headac e n = al
= n = he n = 100  eta-
30 M  26 M  44 M ( M (S squa
(SD) (SD) SD) D) red

Depres
sion 48.59 47.61 49.20
(RCAD (12.0 (12.4 50.54 (11.71
S)a  4)  9)  (11.12)  )  .01 

Child-
report
functio
ning           

Functio
nal
Disabili 12.62 10.79 13.85
ty (10.3 (10.3 16.80 (11.10
(FDI)  8)  6)  (11.70)  )  .05 

Parent-
report
functio
ning           

Functio
nal
Disabili 10.72 10.80 11.30
ty (12.9 (11.2 12.04 (11.87
(FDI)  8)  1)  (11.68)  )  .00 

Physic 39.48 39.48 41.36 40.25 .01 

 Omn
Total ibus
sampl parti
JIA n  SCD  Headac e n = al
= n = he n = 100  eta-
30 M  26 M  44 M ( M (S squa
(SD) (SD) SD) D) red

al
HRQOL (14.9 (12.8 (13.74
a
  5)  0)  (13.68)  ) 

Psycho
social 48.66 49.49 48.13
HRQOL (12.0 (10.0 46.84 (10.34
a
  8)  1)  (9.20)  )  .01 

T-scores.
a

Matching superscripts b and c indicate significant post hoc

comparisons using Scheffe's test.
Open in new tab

Sleep Patterns and Problems

According to caregiver responses on the CSHQ, the children's mean

bedtime was 9:13 PM and their mean wake time was 7:03 AM on
school days. The average sleep time was 9.41 hr, which is within the
range of typical sleep requirements in this age group (Mindell &
Owens, 2003). Bedtime, wake times, and sleep time were similar
across the groups of children with JIA, SCD, and HA (Table III).

Table III.

Sleep Times and Sleep Problems Across Groups

 Total
chron
ic Omn
pain ibus
sampl parti
JIA n  SCD  Headac e n = al
CSHQ = n = he n = 100  eta-
sleep 30 M  26 M  44 M ( M (S squa
times (SD) (SD) SD) D) red

Weekd
ay 9 : 04 9 : 10 9 : 13
bedtim (0 : (0 : 9 : 21 (0 :
e  26)  32)  (0 : 40)  35)   

Weekd
ay 6 : 58 7 : 01 7 : 03
wake (0 : (0 : 7 : 06 (0 :
time  32)  31)  (0 : 56)  44)   

Total
sleep
time in 9.38 9.72 9.25 9.41
hours  (1.06)  (1.53)  (1.02)  (1.18)  .02 

CSHQ
subscal
es           

Bedtim
e
resista 7.20 7.50 7.77 7.53
nce  (2.04)  (1.94)  (2.38)  (2.17)  .01 

Sleep 3.70 3.88 4.18 3.96 .02 

duratio
 Total
chron
ic Omn
pain ibus
sampl parti
JIA n  SCD  Headac e n = al
CSHQ = n = he n = 100  eta-
sleep 30 M  26 M  44 M ( M (S squa
times (SD) (SD) SD) D) red

n  (1.23)  (1.21)  (1.51)  (1.36) 

Paraso 8.57 8.46 8.48 8.50

mnias  (1.77)  (1.07)  (1.69)  (1.57)  .00 

Sleep-
disorde
red 3.38 4.42
breathi (0.73) (1.65) 3.27 3.61
ng  a
  a, b
  (0.50)b  (1.09)  .20 

Night
waking 3.72 3.77 3.75 3.75
s  (1.31)  (1.45)  (1.20)  (1.29)  .00 

Daytim
e
sleepin 13.46 13.19 12.14 12.80
ess  (3.70)  (3.61)  (2.57)  (3.23)  .04 

Sleep 5.10 4.96 5.25 5.13

anxiety  (1.88)  (1.71)  (1.75)  (1.77)  .00 

Sleep 1.53 1.46 1.59 1.54 .00 

onset (0.78)  (0.65)  (0.73)  (0.72) 
 Total
chron
ic Omn
pain ibus
sampl parti
JIA n  SCD  Headac e n = al
CSHQ = n = he n = 100  eta-
sleep 30 M  26 M  44 M ( M (S squa
times (SD) (SD) SD) D) red

delay 

Total sl
eep
disturb 44.36 45.61 43.88 44.46
ance  (8.75)  (8.02)  (7.96)  (8.14)  .01 

Matching superscripts a and b indicate significant post hoc

comparisons using Scheffe's test, both p < .001.
Open in new tab

Table III shows the average sleep pattern and sleep problem scores
for children with each health condition and for the combined group
of children with chronic pain. Few significant differences in sleep
problems were observed between the different health condition
groups. One significant group difference was observed on the sleep-
disordered breathing scale; children with SCD obtained higher
scores than children with HAs or JIA, and this effect size was large.
The mean total sleep problems score for participants in this study
(M = 44.46) was above the clinical cutoff score of 41 determined in
development of the CSHQ (Owens et al., 2000). The proportion of
children with scores above 41 was not significantly different
between the three pain groups, with 58% of children with HA, 48%
of children with JIA, and 60% of children with SCD being above the
clinical cutoff, χ2(2, N = 100) = 3.72, p=.19). Overall, a significantly
higher proportion of children in this sample (53%) were above the
clinical cutoff compared to 23% in the community sample used in
CSHQ development, χ2(2, N = 100) = 50.82, p < .001. Children with
chronic pain had a mean daytime sleepiness score of 12.80 (SD =
3.23), which is similar to the mean score of the clinical sleep sample
of 11.99 (SD = 3.39) (reported in Owens et al., 2000), suggesting
that children with chronic pain demonstrate clinical levels of
daytime sleepiness.

Associations Among Study Variables

Bivariate correlations are presented in Table IV. As hypothesized,

total sleep disturbances were associated with HRQOL, such that
more sleep problems correlated moderately with lower HRQOL
outcomes. Sleep disturbances were positively correlated with
parent, but not child report of functional disability. Contrary to our
hypotheses, sleep disturbances were not significantly correlated
with child-report or parent-report pain intensity or pain frequency.
Additionally, higher levels of sleep disturbances were not correlated
with higher levels of depressive symptoms in this child sample, in
contrast to findings of this relationship in adolescents (Palermo &
Kiska, 2005). Higher levels of sleep problems were also related to
lower family income.

Table IV.

Bivariate Correlations Among Key Study Variables

 P
a P
i ai Sl Ps
n n ee Ph yc F
In i fr D p ys. h. DI F
A co n e e pr HR HR pa DI
g m t q p o QO QO re ch
e eb .b .b . b. L L nt ild

− −
1. . . − −
Gend 0 1 .0 .1 . −. −. −.
era  5  .05  5  8  5  01  11  .03  09  15 

. − −
2. 0 .0 .0 −. −. . −. −.
Age    .04  7  7  3  17  02  25*  18  16 

−
3. . . − −. . −.
Inco 0 0 .1 37 39* 24 .
meb      8  1  2  **  *  .18  **  09 

4.
Pain .
inten 3 . −. −. .
sity: 1 1 . 31* 31* . 49
childb        **  9  17  *  *  13  ** 

5.
Pain .
freq. 2 −. . .
: 7 . 35* −. 22 39
childb          **  15  *  25*  *  ** 

6.           . −. −. . .
 P
a P
i ai Sl Ps
n n ee Ph yc F
In i fr D p ys. h. DI F
A co n e e pr HR HR pa DI
g m t q p o QO QO re ch
e eb .b .b . b. L L nt ild

Depr
essio
n
RCA 49* 61
DS  06  20  *  11  ** 

7.
Total
sleep −. .
probl 51* −. 38 .
ems              *  27*  **  08 

8.
Physi
cal . −. −.
HRQ 42* 46 33
OL                *  **  ** 

9.
Psyc
hoso
cial −. −.
HRQ 49 46
OL                  **  ** 

10.                   .
FDI: 27
Pare
 P
a P
i ai Sl Ps
n n ee Ph yc F
In i fr D p ys. h. DI F
A co n e e pr HR HR pa DI
g m t q p o QO QO re ch
e eb .b .b . b. L L nt ild

nt-
repo
rt  * 

a
Coded 0 = male, 1 = female.
b
Spearman's rho.
*p < .05; **p < .01.
Open in new tab

As hypothesized, there were significant moderate associations

between pain and HRQOL domains such that higher child-report pain
frequency and intensity were associated with lower Physical and
Psychosocial HRQOL. Higher child-report pain frequency and
intensity were also associated with higher levels of child-reported
functional disability, while pain frequency but not intensity was
related to parent-report functional disability. While depression was
not correlated with sleep disturbances, higher depression scores
were strongly associated with higher levels of child but not parent
report of functional disability, and lower levels of parent-report
Psychosocial HRQOL.

Sleep Disturbance as a Predictor of HRQOL and Functional Disability

Hierarchical multiple regressions were conducted to determine

whether variance in HRQOL and functional disability outcomes was
predicted by sociodemographic variables, pain variables,
depression, and sleep disturbance. Child and family demographic
and disease information was entered into the first step. Child-report
pain intensity and frequency was entered into the second step and
depressive symptoms in the third. Total sleep disturbance was
entered in the last step to test the hypothesis that sleep disturbance
is independently associated with HRQOL and functional disability
after controlling for the effects of demographic variables, disease
information, pain variables, and depression. Parallel analyses were
conducted using parent report of pain intensity and frequency.

As shown in Table V, this multivariate model accounted for 48% of

the variance in Physical HRQOL (F8,58 = 5.75, p < .000), and 45% of
the variance in Psychosocial HRQOL (F8,58 = 5.13, p < .000). Total
sleep disturbance was a significant predictor of Physical HRQOL (β =
−.36, p=.003), accounting for 10% of the variance, but not
Psychosocial HRQOL (β = −.20, p=.10). Other variables that were
significant predictors of Physical HRQOL include family income (β = .
29, p=.02) and child report of pain frequency (β = −.25, p=.03). For
Psychosocial HRQOL, depression was the primary predictor (β =
−.44, p < .01), though age was also a strong predictor (β = .
28, p=.02). Overall, higher Physical HRQOL was associated with
higher family income, lower pain frequency, and fewer sleep
disturbances; and higher Psychosocial HRQOL was associated with
older age and fewer depressive symptoms.
Table V.

Sleep Disturbances as a Predictor of Parent-report HRQOL

 HRQOL:
HRQOL: Physical Psychosocial

β β
at at
F- fin fin
va al F- al
lu st val ste
ΔR2 e ep ΔR2 ue p

4.3
Step 7* 3.69
1  .24**  *    .21**  **   

Head
ache
group 
    −.1
      .08      3 

JIA
group 
    −.
      13      .02 

Child
age 
    −. .
      06      28* 

Famil
y
incom .
e    29
       *      .10 
 HRQOL:
HRQOL: Physical Psychosocial

β β
at at
F- fin fin
va al F- al
lu st val ste
ΔR2 e ep ΔR2 ue p

5.4
Step 4*
2  .13**  *    .06  2.00   

Pain
intens
ity   −. −.1
        13      1 

Pain
frequ
ency  −.
    28
      *      .00 

Step 13.0
3  .00  .02    .15***  7***   

Depre
ssion  −.4
    −. 4**
      07      * 

Step .10**  9.9   .03  2.83   

4  3*
 HRQOL:
HRQOL: Physical Psychosocial

β β
at at
F- fin fin
va al F- al
lu st val ste
ΔR2 e ep ΔR2 ue p

* 

Total
sleep
distur
bance −.
s    36 −.2
       **      0 

Total R2 Total R2
  :.48***      :.45***     

*p < .05; **p < .01; ***p < .05.

Open in new tab

As shown in Table VI, the multivariate model predicted 26% of the

variance in parent report of functional disability (F8,58 = 2.20, p=.04),
and 55% of the variance in child report of functional disability
(F8,58 = 7.72, p < .000). Total sleep disturbance was a strong
predictor of parent report of functional disability (β = .34, p = .02),
accounting for 9% of the variance, but not child report of functional
disability (β = −.10, p=.33). Additional predictors of child report of
functional disability included age (β = −.27, p=.01), child report of
pain intensity (β = .32, p=.01), and depression (β = .51, p < .000).
Analyses conducted with parent-report pain intensity and frequency
yielded the same pattern of significant results.
Table VI.

Sleep Disturbances as a Predictor of Parent and Child Report

Functional Disability
FDI: Parent
Report FDI: Child Report

β
a
t
fi β
n at
F- al fin
va st F- al
lu e valu ste
ΔR2 e p ΔR2 e p

Step 2.
1  .14  20    .13  2.10   

Heada
che
group  .
    0
      1      .16 

JIA
group  −
    .0
      4      .14 
 FDI: Parent
Report FDI: Child Report

β
a
t
fi β
n at
F- al fin
va st F- al
lu e valu ste
ΔR2 e p ΔR2 e p

Child −
age   .1 −.2
        0      7* 

Family
incom −
e    .1 −.1
       5      3 

Step . 6.11
2  .03  85    .17**  **   

Pain
intensi − .
ty    .0 32*
       4      * 

Pain
freque .
ncy   1 −.0
        6      3 
 FDI: Parent
Report FDI: Child Report

β
a
t
fi β
n at
F- al fin
va st F- al
lu e valu ste
ΔR2 e p ΔR2 e p

Step . 27.4
3  .00  01    .24***  4***   

Depre
ssion  . .
    0 51*
      4      ** 

6.
Step 24
4  .09*  *    .01  .95   

Total
sleep
distur
bance .
s    3 −.1
       4*      0 

Total R Total R2:.
  2
:.26*      55***     

*p < .05; **p < .01; ***p < .05.

 Open in new tab

Discussion
Findings from the present study extend previous research on sleep,
mood, and HRQOL with adolescents to children in middle childhood.
Similar to previous research (Bloom et al., 2002; Miller et al., 2003),
this study demonstrates that the majority of children with chronic
pain conditions, in particular HAs, JIA, and SCD, are above the
clinical cutoff for total sleep disturbances, a rate higher than
normative community populations (Owens et al., 2000). It should be
noted that the Owens et al. (2000) sample is significantly younger
(M = 7.6 years) than the current sample [M = 10.2 years; t(567) =
16.09, p < .001], thus interpretation of comparisons should be made
cautiously. On the majority of the sleep variables measured by the
CSHQ, there were similar sleep patterns and behaviors observed
between children with JIA, SCD, and HA. One notable exception was
the finding of significant group difference on sleep-disordered
breathing; similar to previous studies, children with SCD were found
to have the highest rate of symptoms of sleep-disordered breathing
(Bandla & Splaingard, 2004). Higher rates of sleep-disordered
breathing among children with SCD are likely due to disease
characteristics, including lower levels of blood oxygen and nocturnal
hypoxia (Setty, Stuart, Dampier, Brodecki, & Allen, 2003). In
addition, independent of disease, urban African-American children
with lower family income have been found to be at higher risk for
sleep apnea (Spilsbury et al., 2006).

Clinical levels of daytime sleepiness were observed in this sample,

which is deserving of future investigation. Similar levels of daytime
sleepiness have been reported in a sample of 6 to 18-year-old
children with migraine, which were significantly higher levels than
case control siblings without migraine (Heng & Wirrell, 2006). There
is some overlap between sleepiness and fatigue in adults with
chronic pain (Merkelbach & Schulz, 2006), but less is known about
the intersection of these constructs in childhood. It is possible that
daytime sleepiness scores may be reflecting higher daytime fatigue
levels in the children with chronic pain in this study. Fatigue is a
multidimensional construct that has been described in several
pediatric clinical populations and has a profound effect on daily
functioning and HRQOL (Meeske, Patel, Palmer, Nelson, &
Parow, 2006) but the relationship among fatigue, pain, sleep, and
daily functioning in children is unknown. Future research might
investigate daytime sleepiness and fatigue in the context of sleep
duration in order to investigate whether children with chronic pain
experience sleepiness when they achieve adequate sleep duration.
Ecological momentary assessment might also be a useful approach
for researchers to investigate fatigue and daytime sleepiness in
children's natural environments. This methodology could capture
sleepiness and fatigue without retrospective reporter bias, and may
help answer questions such as whether children with chronic pain
show different patterns of daytime sleepiness or fatigue.

Family income emerged as a significant correlate of sleep problems

and predictive of Physical HRQOL in this sample, suggesting that
family resources and related factors are important to consider in
future research. Previous studies investigating sleep habits have
demonstrated that minority children living in urban settings
frequently sleep <9 hr (Spilsbury et al., 2004), suggesting that a
number of sociodemographic and contextual factors contribute to
sleep habits. The immediate home and neighborhood environment
is affected by family resources, and also has the potential to
interfere with sleep. For instance, passive television exposure in the
home predicts sleep disturbances in young children (Paavonen,
Pennonen, Roine, Valkonen, & Lahikainen, 2006). Neighborhood
disadvantage has been shown to predict sleep apnea in school-age
children above the effects of individual family SES, ethnicity,
premature birth, asthma, and obesity (Spilsbury et al., 2006). As
these authors point out, higher levels of environmental
contaminants in disadvantaged neighborhoods may contribute to
airway-related sleep problems, and high environmental noise and
stress may lead to sleep fragmentation and disruption. Previous
studies have shown that family wealth is related to school-age
children's physical HRQOL, and is related to all dimensions of HRQOL
in adolescents, including physical, psychological, and social
functioning (von Rueden, Gosch, Rajmil, Bisegger, & Ravens-
Sieberer, 2006). Future research that investigates the complex
relations between family resources, home and neighborhood
environments, sleep disturbances, and related functioning in
children and adolescents may help create targeted sleep
interventions.

The present study also further delineated the relationship between

sleep disturbances and functional disability and HRQOL in school-
age children with chronic pain conditions. Similar to findings in an
adolescent sample (Palermo & Kiska, 2005), sleep disturbances
were strongly related to parent report of Physical HRQOL and
functional disability. Sleep disturbances were, however, not
predictive of parent report of Psychosocial HRQOL or child report of
functional disability. It should be noted that while this pattern of
findings was observed in this combined sample of children with
three different chronic pain conditions, results cannot be
generalized to each pain condition. In fact, the relations between
sleep disturbances and HRQOL and functional disability outcomes
may differ or be of differing magnitudes among children
experiencing different chronic pain conditions. Future research is
needed to further specify the role of sleep disturbance in predicting
outcomes within a variety of chronic pain conditions. In this study,
children reported higher levels of functional disability than was
reported by their parents. Similar levels were reported by
adolescents (Palermo & Kiska, 2005). Additionally, child and
adolescent reports of functional disability are strongly associated
with their reports of pain and depression, suggesting that child-
report of functional disability may be more useful for assessing
perceptions of disability in pediatric chronic pain. In the previous
study of adolescents (Palermo & Kiska, 2005), self-report depression
symptoms were strongly correlated with self-report daytime
sleepiness and sleep/wake problems, and independently contributed
to sleep/wake problems above the contributions of pain
characteristics and functional disability. In this sample of 8- to 12-
year-olds, self-report of depression symptoms was not significantly
correlated with parent-report of sleep disturbances, suggesting that
the association between sleep problems and depression symptoms
may be stronger during adolescence. This finding may also be due
to a lack of children's self-report of sleep problems, which is a
limitation of the study. Future work should include child-report of
sleep problems.

There are a number of limitations to this study that should be

considered when interpreting the findings. First, the children were
all recruited from a clinic-based setting, as such, these results may
not be generalizable to other populations including children with
chronic pain conditions who do not seek treatment for their pain
problem. Detailed information about children's medication usage
was not available, and many medications are known to influence
sleep. Additionally, it is possible that variations in pain and health
status at the time of the routine specialty care visit may have
influenced child and parent report of sleep disturbances or
functional outcomes. It is also possible that unreported or
undiagnosed psychiatric problems may contribute to sleep problems
in this sample. Another limitation is that assessment of sleep
disturbances was made through caregiver report questionnaires
only. The results may change if child report or objective measures of
sleep disturbance, such as actigraphy or polysomnography, were
used instead.

Additionally, our study was limited by using caregiver report

measures to assess the majority of outcome variables, though some
instances of child report was used. As these measures share method
variance, it might have increased the likelihood of finding significant
relationships. The lack of observed association between parent-
report sleep disturbances and child-report functional disability is
difficult to interpret, as this association was observed within
reporter. Last, the cross-sectional design of this study prevents any
conclusions being drawn concerning direction of effects. Future
research is needed to determine the direction of the relationship
between chronic pain, sleep disturbances, and functional outcomes
such as HRQOL and disability in a variety of pediatric populations
with chronic pain. Longitudinal studies examining changes in sleep,
mood, and functional outcomes over time might further delineate
the relationship among them. Future research might also include an
age- and sex-matched healthy comparison group, in order to more
clearly establish the nature of sleep disturbances in children with
chronic pain. The addition of objective measurement tools, more
thorough examination of possible comorbid psychiatric diagnoses,
and the examination of the role of medications are also important
future directions.

Overall, these findings offer some additional support for the

importance of assessing sleep disturbances in children with chronic
pain, both in research and in clinical settings. Parents are more
likely to be involved in establishing and monitoring sleep habits for
school-age children than they are for adolescents, thus assessment
and intervention with parents is likely to lead to positive changes in
sleep habits for school-age children. Assessment of sleep
disturbances in school-age children can be accomplished via clinical
interviews or sleep logs with parents and children or by using
screening tools such as the CSHQ (Owens et al., 2000). These tools
can provide valuable information about sleep duration and sleep
habits. Additionally, interventions targeting sleep problems in
children with chronic pain may lead to improvements in functioning
and HRQOL. In school-age children, these interventions might
include psychoeducation about the need for adequate sleep and
sleep habit recommendations. Sleep interventions might help
improve daytime functioning for children who are experiencing
chronic pain. Finally, implementing sleep interventions in middle
childhood may help avert the development of sleep problems in
adolescence, which may have increased negative functional
consequences in adolescents with chronic pain.

Emotional Functioning

The high rate of comorbidity of mental health conditions along with CP requires assessment for the presence of anxiety
and mood disorders. It must be emphasized that high comorbidity does not mean that one causes the other.
Psychological distress is both a potential contributing factor and a potential outcome of living with CP. Each condition
often involves separate interventions, which require referral to appropriate mental health services.

Pediatricians should be particularly alert to suicide ideation/attempts and comorbid depression in this at-risk population,
and they should ascertain the suicidality of depressed adolescents (ie, whether and how often these adolescents think
about suicide and whether they have ever attempted suicide). If suicidal ideation or recent suicidal behavior is present in
a depressed teenager, he or she should be immediately referred to the appropriate mental health services.

Cognitions
 Young people and their families hold beliefs about the pain that they experience such as what causes it, how long it will
last, whether it is curable, what effects it will have in their lives, what treatments might be relevant, and whether it is
understood and believed as “real” by clinicians. Beliefs are powerful and can influence not only pain perception but also
treatment adherence and treatment response, and pediatricians may have to actively challenge erroneous beliefs and
provide accurate pain education.115
Specific cognitions that need to be assessed include pain catastrophizing and coping. Pain catastrophizing, currently
conceptualized as both related to a personality-based, dispositional construct and a response that varies in different
situations,116 is characterized by a negative mind-set, magnification, and rumination about pain. 117 Catastrophizing in
children is distinct to anxiety and has been identified as a significant predictor of pain, functional disability, and health-
related quality of life in children and adolescents with CP 118 and persistent pain and central sensitization into young
adulthood.119
CP presents a range of stressors and challenges for young people and their families; pediatricians therefore need to
assess how the young person copes with each of these factors and not just pain. Coping can be viewed as a collection of
purposeful, volitional efforts that are mobilized under stress and directed at the regulation of aspects of the self (ie,
emotion, cognition, behavior, physiology), known as secondary control, and interactions with others and the environment,
known as primary control .120,121 Relinquished control refers to the absence of any coping attempt. 122 In general, the
degree to which a coping strategy leads to better or worse emotional and behavioral adjustment depends in part on the
match between the demands of the stressor and the goals and nature of the coping response. 123 However, overall, studies
have shown that secondary control coping (eg, acceptance, cognitive reappraisal, distraction) is associated with lower
levels of somatic complaints and symptoms of anxiety and depression, 66,124–126 whereas passive coping (eg, behavioral
disengagement, self-isolation, catastrophizing) is related to poorer adjustment. 119,127,128

Social Domain

Complex transactional processes and individual factors mediate children’s and parents’ emotional, cognitive, and
behavioral responses to pain, ultimately influencing the child’s overall functioning. 129 The 5 characteristics of family
functioning commonly assessed in family systems theories 130 relevant in the assessment of the family of a child with CP
are organization, cohesion, communication, affective environment, and problem solving. 131 Poorly functioning families can
be those that are highly disorganized, with unclear communication and high expressions of conflict or negative affect
that only become more disrupted when faced with a stressor. Poorly functioning families can also be characterized by
being overly restrictive and ordered, limiting the adaptability of the system to deal with stressors, and limiting individual
members’ ability to express their emotions or modify maladaptive roles.
In terms of individual factors, parental cognitive responses to pain, such as parental pain catastrophizing or exaggerated
negative pain appraisals, have been found to influence both parents’ emotional reactions to pain and child functional
disability. In addition, higher levels of parents’ catastrophic thinking regarding their children’s CP are associated with a
greater tendency to restrict their children’s pain-inducing activities and a greater tendency to prioritize attempts to
control their children’s pain.132
Behaviorally, parental protective responses to children’s pain behavior (eg, increasing attention to pain symptoms,
excusing the child from responsibilities) have been linked to poorer functional outcomes, serving as the proximal link
between parents’ internal reactions (eg, cognitions, emotional distress) and child outcomes. 133–135 Conversely, young
people of parents with greater levels of psychological flexibility tend to report less physical disability, fewer depressive
symptoms, and greater levels of acceptance of their own pain. 136
Anxiety and mood disorders are prevalent among mothers of children with CP conditions, 137,138 but similar data about
fathers are currently lacking. Mothers of children with functional abdominal pain are 4.9 times more likely to have a
lifetime history of depressive disorders and 4.8 times more likely to have a lifetime history of anxiety disorders compared
with mothers of healthy children.137 Maternal depression is a risk factor for the socioemotional and cognitive
development of children. Depressed mothers generally exhibit less attentiveness and responsiveness to their children’s
needs, and they are also poor models for negative mood regulation and problem solving. The pediatricians’ role in
maternal depression is one of screening, followed by guidance for additional evaluation and treatment.
School functioning is often negatively affected in young people with CP, 139 particularly when there is comorbidity with
depressive symptoms.103 Assessment of school functioning should include a number of dimensions, 140,141 namely: (1)
school attendance, clarifying if the absence is due to pain or some other reason; (2) cognitive 142 and
emotional143 engagement (eg, self-regulation, studying habits and enjoyment, belonging, and attitudes toward every
aspect of school); (3) academic performance (eg, grades across subjects, national standardized test scores, classroom
participation); (4) self and teacher perceptions of academic competence 139; (5) participation in school activities (eg, clubs,
school trips); and (6) social functioning in the school setting (eg, social activities, interaction with peers). The limited
number of existing studies suggests that young people with CP may have fewer friends, are more isolated, and may be
subjected to increased rates of victimization by peers compared with children and adolescents without pain. 1