Journal of Pediatric Psychology Advance Access originally published online on August 8, 2007
Journal of Pediatric Psychology 2008 33(3):317-322; doi:10.1093/jpepsy/jsm058
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Brief Report: Daily Mood as a Mediator or Moderator of the Pain–Sleep Relationship in Children with Sickle Cell Disease
1University of North Carolina at Chapel Hill and 2East Carolina University
All correspondence concerning this article should be addressed to Cecelia R. Valrie, PhD, Department of Psychology, 320 Rawl Building, Greenville, NC 27858. E-mail: valriec{at}ecu.edu
 |
Abstract |
|---|
 Top Abstract MethodResultsDiscussionReferences |
|---|
Objective To investigate mood as a mediator or moderator of the pain-sleep relationship in children with sickle cell disease (SCD). Method Children with SCD (n = 20; aged 8–12 years) completed daily diaries assessing mood, sleep, and pain for up to 2 months. Data was analyzed using multilevel modeling. Results Results indicate that negative mood partially mediates the relationship between high daily pain and poor sleep that night as well the relationship between poor sleep and high daily pain the following day. The impact of poor sleep on high pain the following day was weakened at increasing levels of positive mood. Conclusion Research is needed to fully explore the ways positive and negative mood may relate to pain and sleep characteristics. This information may be beneficial for developing more effective pain management and sleep interventions.
Key words: diary; mood; pain; sickle cell disease; sleep.
Children with a chronic illness are at increased risk for sleep difficulties (National Center on Sleep Disorders Research, 2003
). Recent research with children with sickle cell disease (SCD), a genetic disorder affecting about 1 in 600 African Americans (Steinberg, 1999), has shown a link between pain and poor sleep (Jacob et al., 2006; Shapiro et al., 1995; Valrie, Gil, Redding-Lallinger, & Daeschner, 2007, in press). However, there is a lack of research examining factors that might explain or influence the pain–sleep relationship.
The current study examines mood as a possible mediating or moderating factor on the pain–sleep relationship in children with SCD. A recent review by Pressman and Cohen (2005) concluded that positive mood is related to better health, including less pain in individuals who experience pain. Specifically, dispositional positive affect has been linked to less pain in patients with cancer, fibromyalgia, and rheumatoid arthritis. More recent research has focused on the role of fluctuations in common daily mood states, versus dispositional mood states, in adjusting to SCD pain. Findings from a daily diary study with adolescents with SCD indicates that low positive mood and high negative mood are independently related to high same day SCD pain and adjustment to the pain experience (e.g., health care use and activity reduction) (Gil et al., 2003). In addition, a body of research has established a link between poor sleep and mood fluctuations (Buysse, Germain, Nofzinger, & Kupfer, 2006).
The sample used for the current study consisted of children who completed daily diaries as part of a larger investigation (Valrie et al., 2007, in press). In prior studies, we found that the pain–sleep relationship may be cyclic; high daily pain was related to poor sleep quality that night and poor sleep quality was related to high daily pain the following day. The present study extends prior research by analyzing the contribution of daily mood to the pain–sleep relationship. We hypothesized that mood would mediate the pain–sleep relationship, such that high daily pain would lead to negative mood and in turn, negative mood would lead to poor sleep quality at night. We also hypothesized that poor sleep quality would lead to negative mood the next day and in turn, negative mood would lead to high daily pain. Moreover, we tested whether mood moderated the pain–sleep relationship. Specifically, we hypothesized that the relationships between high daily pain and poor sleep quality that night and between poor sleep quality and high daily pain the following day would be weakened at increasing levels of positive mood.
|
Method |
|---|
|
TopAbstractMethodResultsDiscussionReferences |
|---|
Participants and Procedures
Participants and their guardians were recruited during regular out-patient visits at two university clinics located in the Southeast. The Institutional Review Boards at both universities approved the study. A health care provider screened potential children for medical contraindications (e.g., neurological impairment, comorbid medical conditions). Only children who had at least one pain episode (pain of at least 20 min attributed to SCD by the child) in the past year were eligible. Data on pain episodes were obtained from the medical file or from a screening question asked of the children or their guardians. The sample was 20 children (13 girls, 7 boys) aged 8–12 years (M = 10.1 years, SD = 1.07 years). The mean level of maternal education was 12.7 years (SD = 2.10 years, range = 7.5–16 years). Of the 20 children, 14 (70%) had sickle cell anemia, 5 (25%) had hemoglobin SC, and 1 (5%) had sickle thalassemia. For analyses, SCD type was coded as 1 = sickle cell anemia and 0 = other SCD types.
After a baseline interview, the children were asked to complete daily diaries which included an assessment of mood, sleep, and pain for 8 weeks. The children were instructed to complete the section on sleep in the morning and the sections on mood and pain in the evening. The children's guardians were also asked to assist their children with completing the diaries if needed. The children were given stamped, addressed envelopes, and asked to mail back the diaries on a weekly basis. The children completed a total of 712 days out of a possible 1120 days (20 children x 7 days/week for 8 weeks), representing a completion rate of 64%. The mean number of diaries completed per child was 36 diaries (SD = 19 diaries, range = 7–56 diaries).
Daily Diary of Mood, Sleep, and Pain
Mood
Mood was assessed by the Facial Affective Scale (FAS) (McGrath, de Veber, & Hearn, 1985; McGrath et al., 1996). The FAS consists of nine faces that vary in levels of affect expressed from a very upset face to a very happy smiling face. The measure was selected because it is simple to complete by young children and easily adapted to a daily rating format. At the end of each day, children were instructed to "mark the face which looks like how you felt deep down inside today—not just how your face looked but how you felt you really felt inside." The numeric values for this scale have been determined experimentally and transformed to a 0 to 1 scale, where 1 = maximum negative affect and 0 = maximum positive affect. The FAS was initially developed to assess affective response to pain and the reliability and validity of the scale in that context is well-researched (McGrath et al., 1985). However, several studies (Sandstrom, Cillessen, & Eisenhower, 2003; Schanberg, Gil, Anthony, Yow, & Rochon, 2005; Schanberg et al., 2000) have utilized the scale as a more general mood measure, and support for the reliability and validity of the scale in this context is accumulating. For example, Sandstrom et al. (2003) found mood to be related to internalizing behavior problems in 9 to 11-year-old children as measured by a composite score based upon parental reports on the Child Behavior Checklist (Achenbach & Edelbrock, 1983) and self-report measures of depression, social anxiety, and self-perception.
Sleep
The following sleep items were based on a diary used with children with juvenile rheumatoid arthritis (JRD) (Schanberg et al., 2000). To assess subjective sleep quality, children were asked to rate on a 100 mm horizontal visual analogue scale (VAS), ranging from "did not sleep well" (0 mm) to "slept very well" (100 mm), how well they slept the night before. "Slept very well" indicated that the children felt well-rested and refreshed after the night of sleep and "Did not sleep well" indicated that the children did not. Research (Valrie et al., 2007, in press) has provided support for the reliability and validity of the measure in children with SCD.
Pain
Children were asked to report whether they experienced any SCD pain during the day, and then to rate the average level of their pain by placing a vertical mark on a 100 mm horizontal VAS with ratings ranging from "hurting a whole lot" (100 mm) to "not hurting at all" (0 mm). Research (Gil et al., 2003; Valrie et al., 2007, in press) has provided support for the reliability and validity of the VAS in assessing SCD pain in children and adolescents.
Data Analyses and Management
Multilevel models for analyzing temporal data (Schwartz & Stone, 1998) were used to examine the relations among daily reports of sleep, SCD pain, and mood. Multilevel models facilitate the modeling of variance at the day-to-day within-person level, while controlling for differences in person-level factors (e.g., age, gender). In these models, data are not aggregated, preserving the richness of daily variation and avoiding heteroskedasticity. Other advantages include that these models (a) handle unequal numbers of observations per participant, (b) control for missing data and serial dependency without needing to impute values for missing data, and (c) utilize observations versus individuals as the unit of analysis, leading to an expanded number of degrees of freedom and increased power. Although there are no well-validated methods for estimating power at the observation level, research by Curran & Muthén (1999) indicates that a large number of repeated measures per participant (as is common in longitudinal daily diary studies) compensates for a smaller sample size.
|
Results |
|---|
|
TopAbstractMethodResultsDiscussionReferences |
|---|
Descriptive Information
Across all of the diaries, the average mood rating was.24 (SD = .26, range = .04–.97). The average sleep quality rating was 84.41 (SD = 25.91, range = 0–100), 16 of the 20 children experienced pain during the course of the study with pain reported on 22.24% of diary days, and the average pain severity on pain days was 48.73 on the 100 point scale (SD = 27.20, range = 2–100). Negative mood was related to high pain severity (r = .59, p < .01) and low sleep quality that night and the previous night (r = –.19, p <.01 and r = –.26, p < .01, respectively).
Mood as a Mediator of the Pain–Sleep Quality Relationship
The procedure detailed in Pituch, Whittaker, & Stapleton (2005) for examining lower level mediation in a multilevel modeling framework was used. This consists of a series of multilevel random effects models to estimate the relations between the predictor and the mediator (herein referred to as Path A) and the mediator and the outcome variable when the predictor is controlled (herein referred to as Path B). Thus, a series of multilevel random effects models were calculated to examine whether mood acted as a mediator of (a) the influence of pain on sleep quality that night and (b) the influence of sleep quality on pain the next day. Level 2 covariates in each model consisted of age, gender, level of maternal education, SCD type, and aggregated person-means for the independent diary variables used in the models. The Sobel (1982) procedure was then used to statistically investigate the effect of the proposed mediator on the predictor–outcome relationship. Research indicates that this procedure is appropriate for investigating mediation in a multilevel modeling framework and displays suitable power and Type 1 error rates to do so effectively (Pituch et al., 2005).
To estimate Path A of the mediation models, two multilevel models predicting mood using (a) pain and (b) sleep quality were calculated (Table I). High pain severity was related to negative mood (β = 0.15, p < .01) and low sleep quality was related to negative mood (β = –0.06, p <.01). To estimate Path B, two multilevel models (a) predicting sleep quality using pain and mood the previous day and (b) predicting pain using sleep quality the night before and mood that day were calculated. The first model indicated that high pain severity (β = –2.55, p = .04) and negative mood (β = –2.82, p = .01) were related to low sleep quality, and the Sobel test indicated that mood (z = –2.50, p = .01) was a significant mediator of the influence of pain on sleep quality. Similarly, the second model indicated that low sleep quality the previous night (β = –3.90, p < .01) and negative mood that day (β = 11.72, p < .01) were related to high pain severity, and the Sobel test indicated that mood (z = –6.45, p < .01) was a significant mediator of the influence of sleep quality on pain. Taken together, negative mood appears to partially mediate the relationship between high pain and poor sleep quality that night and the relationship between poor sleep and high pain the next day.
|
Mood as a Moderator of the Pain–Sleep Quality Relationship
To investigate whether mood moderated the pain–sleep relationships, the interaction between mood and pain was entered into the model predicting sleep quality and the interaction between mood and sleep quality was entered into the model predicting pain (Table I). In the model predicting sleep quality, the interaction between mood and pain was not predictive of sleep quality (β = –0.05, p = .95). In the model predicting pain, the interaction between mood and sleep quality was related to pain (β = –2.05, p < .01), indicating that the relationship between sleep quality and pain was influenced by mood. The online interactive calculator for probing interactions developed by Preacher, Curran, & Bauer (2006
) was used to calculate simple slopes describing the relationship between sleep quality and pain at varying levels of positive mood. Specifically, negative mood was designated as 1SD above the mean, average daily mood was the mean, and positive mood was 1SD below the mean. At negative mood, the simple slope was –4.78 (z = –6.50, p < .01), at average daily mood, the simple slope was –2.72 (z = –3.49, p < .01), and positive mood, the simple slope was –0.66 (z = –0.56, p = .57). This pattern of findings indicates that as positive mood increases the relationship between low sleep quality and pain the following day decreases. |
Discussion |
|---|
|
TopAbstractMethodResultsDiscussionReferences |
|---|
The current study adds to the literature by demonstrating that mood appears to partially mediate the pain–sleep relationship. Negative mood partially mediated the relationship between high daily pain and poor sleep quality that night as well the following day. These findings are consistent with research indicating that high SCD pain and low sleep quality are related to low positive mood and high negative mood (Buysse et al., 2006
; Gil et al., 2003). However, given that mood only partially explained the pain–sleep relationship, future research should focus on examining other possible mechanisms. For example, pain may persist into the night and/or dis-regulate sleep mechanisms. Also, sleep has been linked to cytokines, molecules associated with regulating immune system functioning, and there is evidence indicating that cytokines may play a role in increasing the pain experience (Motivala & Irwin, 2007). Results also indicated that mood moderates the pain–sleep relationship. Specifically, the impact of poor sleep on high pain the following day was weakened at increasing levels of positive mood. These findings indicate that there may be benefit to incorporating methods to promote positive mood when designing interventions to lessen the impact of poor sleep on the pain experience.
A limitation of the current study is that the FAS scale used to assess daily mood was initially developed as a measure of affective response to pain (McGrath et al., 1985), and has only been applied in the context of general mood measurement in a few prior studies. Although the instructions used in the present study did not ask the children to describe their feelings about their pain, the strength of the pain–mood relationship seen in the current study could possibly be due to the overlap in the constructs of pain intensity and affective response to pain (as well as the fact that both measures were completed at the same time). These concerns can be offset somewhat by the fact that the relationship reported between negative mood and high pain is consistent with previous research utilizing general mood measures in other populations (Pressman & Cohen, 2005). However, future research is needed to specifically address the reliability and validity of the FAS measure as a general mood measure. Another limitation of the FAS was that it does not allow for the separate examination of positive mood and negative mood. This is an important distinction given that prior research has demonstrated that positive and negative mood independently influence pain and adjustment to pain in adolescents with SCD (Gil et al., 2003). Overall, these limitations indicate that future research should incorporate additional general mood measures that assess both positive and negative mood.
Additional limitations include the self-report nature of the measures used, specifically the mood and sleep measures. Future research should incorporate parental measures, possibly concurrently completed parent diaries. Also, self-reported perception of sleep quality is only one aspect of sleep. Future research should incorporate more objective sleep assessments, such as sleep actigraphies, and should examine other aspects of sleep such as sleep efficiency and fragmentation. Other potential limitations include missing data and variability in the number of diaries completed. Since these are common issues when collecting daily diary data, we used strategies recommended by Butz & Alexander (1991) and detailed in an earlier publication (Valrie et al., 2007) to improve diary completion. Moreover, missing data and variability was handled statistically through the use of multilevel models. Notably, a possible limitation of one strategy we used (parental assistance) is that parents’ may have inadvertently influenced their children's responses on the diaries. Other limitations include the small sample size and restricted age range, which may make it difficult to generalize the results to the general population of children with SCD. Regarding the sample size, the multilevel models analyzed a large number of repeated measures per participant and this in part compensates for having a small sample (Curran & Muthén, 1999). Moreover, the number of participants and observations was comparable to other similar diary studies (Gil et al., 2003).
Taken together, findings indicate that mood plays varied roles in describing the pain–sleep relationship in children with SCD. Future research is needed to fully explore the ways positive and negative mood may relate to pain and sleep characteristics in this and other pain populations. Lastly, research examining factors that influence or explain the pain–sleep relationship may be beneficial for developing more effective pain and sleep management interventions.
Conflict of interest: None declared.
Received April 29, 2007; revision received June 15, 2007; accepted July 10, 2007
|
References |
|---|
|
TopAbstractMethodResultsDiscussionReferences |
|---|
Achenbach TM, Edelbrock C. The classification of child psychopathology: A review and analysis of empirical efforts. Psychological Bulletin (1983) 58:1275–1301.
Butz AM, Alexander C. Use of health diaries with children. Nursing Research (1991) 40:59–61.[Web of Science][Medline]
Buysse DJ, Germain A, Nofzinger EA, Kupfer DJ. Mood disorders and sleep. In: The American Psychiatric Publishing Textbook of Mood Disorders.—Stein DJ, Kupfer DJ, Schatzberg AF, eds. (2006) Washington, DC: American Psychiatric Publishing, Inc. 717–737.
Curran PJ, Muthén BO. The application of latent curve analysis to testing developmental theories in intervention research. American Journal of Community Psychology (1999) 27:567–595.[CrossRef][Web of Science][Medline]
Gil KM, Carson JW, Porter LS, Ready J, Valrie C, Redding-Lallinger R, et al. Daily stress and mood and their association with pain, health-care use, and school activity in adolescents with sickle cell disease. Journal of Pediatric Psychology (2003) 28:363–373.
Jacob E, Miashowski C, Savendra M, Beyer JE, Treadwell M, Styles L. Changes in sleep, food intake, and activity levels during acute painful episodes in children with sickle cell disease. Journal of Pediatric Nursing (2006) 21:23–34.[CrossRef][Medline]
McGrath PA, de Veber LL, Hearn MT. Multidimensional pain assessment in children. In: Advances in Pain Research and Therapy.—Fields HL, Dubner R, Cervero F, eds. (1985) New York: Raven Press. 387–393.
McGrath PA, Seifert CE, Speechley KN, Booth JC, Stitt L, Gibson MC. A new analogue scale for assessing children: An initial validation study. Pain (1996) 64:435–443.[CrossRef][Web of Science][Medline]
Motivala SJ, Irwin MR. Sleep and immunity: Cytokine pathways linking sleep and health outcomes. Current Directions in Psychological Science (2007) 16:21–25.[CrossRef][Web of Science]
National Center on Sleep Disorders Research. 2003 National Sleep Disorders Research Plan. (2003) Retrieved April 22, 2007, from http://www.nhlbi.nih.gov/health/prof/sleep/res_plan/index.html.
Pituch KA, Whittaker TA, Stapleton LM. A comparison of methods to test for mediation in multisite experiments. Multivariate Behavioral Research (2005) 40:1–23.[CrossRef][Web of Science]
Preacher KJ, Curran PJ, Bauer DJ. Computational tools for probing interaction effects in multiple linear regression, multilevel modeling, and latent curve analysis. Journal of Educational and Behavioral Statistics (2006) 31:437–448. Retrieved June 8, 2007 from http://www.psych.ku.edu/preacher/interact/index.html.
Pressman SD, Cohen S. Does positive affect influence health? Psychological Bulletin (2005) 131:925–971.[CrossRef][Web of Science][Medline]
Sandstrom MJ, Cillessen AHN, Eisenhower A. Children's appraisal of peer rejection experiences: Impact on social and emotional adjustment. Social Development (2003) 12:530–550.[CrossRef][Web of Science]
Schanberg LE, Gil KM, Anthony KK, Yow E, Rochon J. Pain, stiffness, and fatigue in juvenile polyarticular arthritis. Arthritis and Rheumatism (2005) 52:1196–1204.[CrossRef][Web of Science][Medline]
Schanberg LE, Sandstrom MJ, Starr K, Gil KM, Lefebvre JC, Keefe FJ, et al. The relationship of daily mood and stressful events to symptoms in juvenile rheumatic disease. Arthritis Care and Research (2000) 13:33–41.[CrossRef][Web of Science]
Shapiro BS, Dinges DF, Orne EC, Bauer N, Reilly LB, Whitehouse WG, et al. Home management of sickle cell-related pain in children and adolescents: Natural history and impact on school attendance. Pain (1995) 61:139–144.[CrossRef][Web of Science][Medline]
Schwartz JE, Stone AA. Strategies for analyzing ecological momentary assessment data. Health Psychology (1998) 17:6–16.[CrossRef][Web of Science][Medline]
Sobel ME. Asymptotic confidence intervals for indirect effects in structural equation models. In: Sociological Methodology 1982.—Leinhardt S, ed. (1982) Washington, DC: American Sociological Association. 290–312.
Steinberg MH. Management of sickle cell disease. The New England Journal of Medicine (1999) 340:1021–1030.
Valrie CR, Gil KM, Redding-Lallinger R, Daeschner C. Sleep in children with sickle cell disease: An analysis of daily diaries utilizing multilevel models. Journal of Pediatric Psychology (2007) 32:857–861.
Valrie CR, Gil KM, Redding-Lallinger R, Daeschner C. The influence of pain and stress on sleep in children with sickle cell disease. Children's Health Care. (in press).
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||