Abstract
Objective:
Youth with inflammatory bowel disease (IBD) may be at increased risk for sleep difficulties due to the painful and inflammatory nature of their disease. Moreover, children and adolescents with IBD experience impairment across a variety of psychosocial domains. However, researchers have yet to investigate the complex interplay between sleep, disease-related symptoms, and psychosocial factors in this population. The purpose of this study was to examine sleep patterns, pain, and mood in pediatric IBD.
Method:
A sample of 25 children and adolescents with IBD (Mage = 14.24, range = 10–18 years; 56% male) were recruited from a pediatric gastroenterology clinic. Youth wore an actigraphy watch and completed daily measures of affect and pain over the course of 14 days. Statistical analyses involved repeated measures general estimating equations.
Results:
No significant association for sleep with negative affect was demonstrated. Despite majority of this sample being in disease remission, results revealed that increased sleep onset latency was associated with presence of next day pain and pain was associated with better next night sleep efficiency.
Conclusions:
Findings of the current study suggest youth with IBD experience poor sleep quality, which is significantly related to the pain they experience. Consequently, healthcare providers should screen for and address sleep quality to optimize outcomes in their pediatric patients. Objectively assessing sleep patterns (e.g., actigraphy) may prove useful for pediatric IBD samples; however, additional research is needed to determine actigraphy’s feasibility and efficacy in assessing sleep patterns in real world settings (e.g., pediatric medical clinics).
Implications for Impact Statement
There is increasing awareness of the prevalence of poor sleep quality among children and adolescents with inflammatory bowel disease (IBD). This study suggests that such sleep difficulties are related to abdominal pain and may be present even for youth whose disease is in remission. It is essential to regularly screen for poor sleep quality and target these concerns in the pediatric IBD population.
Inflammatory bowel disease (IBD), consisting of Crohn’s disease (CD) and ulcerative colitis (UC), is characterized by an abnormal immune response and gastrointestinal inflammation (Centers for Disease Control and Prevention, 2014). Both disorders are marked by several symptoms, such as diarrhea, abdominal cramping and pain, rectal bleeding, weight loss, fever, delayed puberty, and osteoporosis (Kim & Ferry, 2004). The course of IBD is unpredictable in nature, with periods of symptom flare-ups and disease quiescence (Banez & Cunningham, 2009). Treatment for IBD includes medication (e.g., aminosalicylates and corticosteroids), maintaining a healthy diet that excludes foods known to trigger symptoms, and potential surgery if complications occur (Crohn’s & Colitis Foundation of America, 2014). Given the increasing incidence of IBD in youth (Sýkora et al., 2018), much research has emerged regarding psychosocial functioning and disease sequelae, such as fatigue, among individuals with IBD. However, few studies have examined the role of sleep in pediatric IBD.
For patients with IBD, a bidirectional relationship exists between sleep and gastrointestinal inflammation, such that inflammation and pain lead to poor sleep quality, which in turn leads to additional inflammation (Qazi & Farraye, 2019; Swanson et al., 2011). Youth with IBD experience poor sleep quality compared to healthy peers (Mählmann et al., 2017). In one study, 54% of participants reported experiencing daytime fatigue and night wakings (Nachmias et al., 2006). More specifically, existing findings suggest that children with IBD are more likely to experience sleep disturbance relative to their healthy peers, specifically daytime sleepiness, nightmares, and frequent night wakings, and this disturbance is negatively associated with their health-related quality of life and disease activity (Benhayon et al., 2013; Mählmann et al., 2017; Manhart et al., 2016; Nachmias et al., 2006; Pirinen et al., 2010). Research is needed to further understand specific factors associated with poor sleep quality.
Youth with IBD experience high rates of internalizing psychological symptoms (i.e., depression and anxiety; Mikocka-Walus et al., 2016). This is sensible, as abdominal pain is a common symptom for youth with IBD, and those who experience pain likewise report greater internalizing symptoms (Palermo et al., 2007). Research with other pediatric chronic pain conditions has demonstrated a positive association between pain and sleep disturbance (Haim et al., 2004; Huntley et al., 2007; Valrie et al., 2013). One study to date has evaluated associations amongst sleep, pain, and internalizing symptoms in pediatric CD (Benhayon et al., 2013), revealing that youth with greater sleep disturbance reported greater depression, anxiety, and pain, suggesting an association not only between pain and sleep, but between sleep and mood symptoms as well. This study had several limitations, however, including exclusive use of subjective measures and a cross-sectional design. These limitations point to a need for future research using objective assessments, a prospective design, and a sample including both pediatric CD and UC.
Taken together, the results of the aforementioned studies suggest that pain and emotional functioning are important factors to consider in the investigation of sleep patterns and disturbances in children and adolescents with IBD. These factors may be considered targets of future interventions designed to promote optimal sleep, as well as health and psychosocial outcomes in this population. Consequently, the purpose of our study was to longitudinally examine the associations of objectively measured sleep patterns with abdominal pain and negative affect among children and adolescents with IBD while accounting for patient age, gender, and day of the week. Based on previous research within samples of IBD and other pediatric pain populations (e.g., Benhayon et al., 2013; Pirinen et al., 2010; Valrie et al., 2013), the following hypotheses were proposed: (a) youth who report pain during the day will have worse sleep quality that night; (b) youth with worse sleep quality will report greater pain the next day; (c) youth who report more negative affect during the day will have worse sleep quality that night; and (d) youth who have worse sleep quality will report more negative affect the next day.
Method
Participants
Patients (ages 10–18 years) with an established diagnosis (i.e., ≥ 3 months; medically stabilized after initial diagnosis) of CD or UC and their primary caregivers were approached within a mid-Atlantic gastroenterology (GI) clinic to participate in a larger study examining associations among sleep, pain, mood, anxiety, and health-related quality of life. This age range was chosen given increased prevalence of sleep disturbance and chronic pain among older children and adolescents (Valrie et al., 2013). This study was conducted from 2014 to 2016 and only during the school year given known differences in pediatric sleep schedules during the summer and on holiday breaks (Szymczak et al., 1993). Of the 51 dyads initially approached, 40 families agreed to participate. Reasons for refusal included lack of interest in the study (n = 4) and being too busy to complete study measures (n = 7). No significant differences in terms of patient age, gender, diagnosis, or race were found between participants and nonparticipants. After accounting for participants who withdrew from the study (n = 4), were lost to follow-up (n = 10), and had incomplete diary data to corroborate actigraphy data (n = 1), the final sample included 25 families. For the current study, only patient data were included. Of the 25 participants, 56% (n = 14) were male and the mean age was 14.24 years (SD = 2.42). Most of the sample was White (88%, n = 22), which is representative of IBD incidence rates nationally (Nguyen et al., 2014), and diagnosed with CD (80%, n = 20). Full demographic and descriptive variables for the dataset at baseline are shown in
Procedure
Participants were recruited during their routine clinic appointment (n = 23) or via a letter sent to families from the clinic’s patient registry (n = 2) as part of a larger study. All primary caregivers and 18-year-old patients completed informed consent while patients aged 10–17 provided written assent. Families who agreed to participate were initially given unrelated baseline questionnaires to be returned via mail prior to our research team mailing out study questionnaires and actigraphy watches. This procedure was employed to help ensure that families would return costly actigraphy equipment. Those that did not return baseline questionnaires were considered lost to follow-up and were not included in the current analysis. Participants completed paper versions of study questionnaires that were mailed back with the actigraph using self-addressed, stamped envelopes. Participants were instructed to complete questionnaires each day, as close to bedtime as possible, to capture information from the entire day for a period of 14 days. A research assistant contacted each family twice a week during the study period to remind them to complete the questionnaires and wear the actigraph daily, and to answer any questions. Families received $2.50 in cash for each day they completed the study measures and actigraphy, as well as $5 in cash for completing all measures and actigraphy during the 14-day study period. Additionally, families who returned the actigraphy equipment in working order were entered into a lottery to receive one of five $50 gift cards. This study received approval from the West Virginia University Institutional Review Board.
Measures
Pain
Each day, patients rated their abdominal pain intensity on an 11-point Likert-type scale with 0 = no pain and 10 = worst pain ever via a daily diary created for the study. Numerical pain rating scales have demonstrated adequate reliability and validity, as well as being sensitive to changes in pain intensity in this age group (Williamson & Hoggart, 2005).
Mood
Daily, patients completed the Positive and Negative Affect Schedule for Children (PANAS-C; Laurent et al., 1999), a 30-item Likert-type questionnaire, to assess positive and negative mood. Higher scores equate to greater negative or positive mood. The PANAS-C has good discriminant and convergent validity with validated, self-report measures of anxiety and depression among children (Laurent et al., 1999). For this study, only the negative affect scores were utilized. Cronbach’s α coefficients were all satisfactory but varied by day across the 14-day study period: 0.72 (on Day 3) to 0.95 (on Day 14).
Sleep
The Actiwatch-64 (Mini Mitter, Bend, Oregon, United States) is a wristwatch-like device worn on the patient’s nondominant wrist to assess sleep/wake patterns using an accelerometer. Actigraphs have high sensitivity for identifying sleep and low specificity for identifying wake after sleep onset (WASO) in comparison to polysomnography and direct observation for pediatric samples (Meltzer et al., 2012). However, actigraphy provides an objective method of assessing sleep/wake patterns that may be more ecologically valid than polysomnography, as well as more reliable and valid than self-report (Meltzer et al., 2012). The Actiwatch was programmed to the 30-s epoch setting to allow for continuous monitoring during the 14-day study period. Sleep onset latency (SOL), total sleep time (TST), sleep efficiency (i.e., TST divided by time in bed), and WASO were used in analyses. Patients also completed diary items regarding timing and duration of night awakenings and waking in the morning, as well as depressed a button on the device to indicate bedtime to corroborate Actiwatch data. Research assistants manually entered time in bed and time out of bed based on this corroborating data, as well as manually scored sleep onset as 2 min of consistent inactivity per actigraphy data.
Disease Activity
Physicians completed either the Pediatric Crohn’s Disease Activity Index (PCDAI; Hyams et al., 1991) or Pediatric Ulcerative Colitis Activity Index (PUCAI; Turner et al., 2007), depending on the child’s diagnosis, to assess the patient’s recent disease activity. This measure was obtained during the appointment that patient was enrolled (for those recruited in clinic) or their most recent clinic appointment (for those recruited via registry letters). Items on both measures are summed to create a total disease activity score, with higher values corresponding to greater disease activity. For the PCDAI, patients can be categorized as having inactive disease, mild disease, or moderate to severe disease. On the PUCAI, patients can be categorized into disease remission, mild disease, moderate disease, and severe disease. Studies have suggested good psychometric validity for both the PCDAI and PUCAI among youth with IBD (Hyams et al., 2005). For the current study, the categories of moderate and severe disease on the PUCAI were combined to enable consistency across the two measures.
Demographic and Medical Information
Patient age, ethnicity, education level, and medical history were collected by primary caregiver report on a study specific form.
Analysis Plan
Statistical Analysis Software (SAS) 9.4 was used for analysis, and Statistical Package for the Social Sciences (SPSS) for some data management. Negative affect was log transformed for model analysis due to strong left skew. Continuous data were described with means and standard deviations, while categorical data were reported with frequencies and valid percentages. Separate linear mixed models were run for continuous measures not violating assumptions. Random effects were included for intercept and time, as they provided the best fitting model via smallest Akaike Information Criteria (AIC). Because there was little variance in pain severity, pain was coded as being present or not present (i.e., binary variable) and was tested with a repeated measure generalized estimating equation (GEE) with binomial link. SOL was distributed as a negative binomial (i.e., a count in minutes where most responses were close to 0) and tested with a repeated measure GEE. To account for differing sleep patterns during the weekend for adolescents versus younger children, a weekend by age interaction term was included as a covariate for all models, along with time (for each day of the diary from 1 to 14), age in years, gender, and weekend day.
Results
Despite reminders to complete data collection and wear the actigraphy watch, not all participants wore their watches daily. Analyses were pulled from all available time points. All participants with at least 7 days of diary and actigraphy data including both weekdays and weekends were retained in the final sample for analyses (N = 25). Twenty-five participants wore their watch on night 1, while only seven wore their watch on night 14. The number of participants wearing their watches each night in between varied from 13 to 25 participants. The average number of nights that participants wore their actigraphy amounted to 13.7 (SD = 1.4). Participants reported on their mood and pain on most days, with 23–25 participants completing measures from day 1 to day 14. The full sample of 25 participants reported on their mood and pain on 10 of 14 days. Participants reported that for all daily observations, pain was experienced on 41.1% of days. See
Generalized Estimating Equation Fixed Effects Results for Negative Affect Variable and Sleep Variables
Note. TST = total sleep time; WASO = wake after sleep onset; SOL = sleep onset latency.
Generalized Estimating Equation Fixed Effects Results for Pain Variable and Next Day SOL Variable
Note. TST = total sleep time; WASO = wake after sleep onset; SOL = sleep onset latency; CI = confidence interval.
Discussion
The current study demonstrated several key findings regarding the association of sleep quality with abdominal pain and mood in pediatric IBD. First, children and adolescents who experienced worse overnight SOL were more likely to report abdominal pain the next day; however, such a finding was not demonstrated with pain and same-night sleep onset. Youth who reported pain during the day were more likely to have better sleep efficiency that night. Sleep patterns were also linked to patient age and gender such that older children had shorter sleep duration and females had shorter SOL. This is consistent with prior research demonstrating that sleep duration decreases with age (Maslowsky & Ozer, 2014). Contrary to hypothesis, no significant associations were demonstrated among any sleep variables and negative affect. Regarding weekday and weekend differences, children had lower TST and greater SOL on weekends. These finding regarding sleep onset fits logically with the notion that staying up later during weekends; however, our finding regarding reduced sleep duration on weekends is contradictory to previous research (Becker et al., 2017).
This current study partially supports previous research among youth with IBD (Benhayon et al., 2013) and chronic pain (Valrie et al., 2013), as longer sleep onset was associated with pain the following day. Our study expands on existing literature by being the first to identify the relation between abdominal pain and sleep quality using objective sleep assessment (i.e., actigraphy) with a combined sample of pediatric CD and UC. Importantly, this association was demonstrated despite most study participants (80%) being rated by their physicians as having inactive disease, consistent with previous research on rates of disease remission in pediatric IBD (Benhayon et al., 2013). Nevertheless, our results are not suggestive of the bidirectional relation between pain and sleep disturbance found in other pediatric pain populations (Valrie et al., 2013). Instead, the presence of pain was associated with greater sleep efficiency that night in our study. This finding may be due to differences in disease processes, as IBD is characterized by gastrointestinal inflammation (Banez & Cunningham, 2009) unlike other pain populations (e.g., sickle cell disease, headache) included in previous literature regarding decreased sleep efficiency (Valrie et al., 2013). It is also possible that youth may be spending time in the bathroom if symptoms are present due to the nature of bowel disturbance and thus, decreasing their time in bed lying awake with pain. Given these findings, future research might focus on objective sleep measurement and prospective pain in other pediatric pain conditions to examine whether consistent patterns emerge.
The lack of a significant association between negative affect and sleep was unexpected given previous research suggesting sleep and aspects of psychological functioning are related to one another (Benhayon et al., 2013; Palermo et al., 2007). However, our sample reported relatively low negative affect on the PANAS-C (i.e., mean negative affect = 21.6) when compared to the measure’s maximum score (i.e., 75). It is possible this low level of negative affect contributed to lack of significant associations with sleep variables and thus, our findings may not generalize to a sample of children with IBD who have higher levels of mood or psychological concerns. Another reason for this finding may be differences in measurement of psychological functioning. Prior studies employed specific measures of depression and/or anxiety (e.g., Revised Child Anxiety and Depression Scale; Palermo et al., 2007), while the current study utilized a more general measure of affect (i.e., PANAS-C). Negative affect is only one of several diagnostic criteria of depression (American Psychiatric Association, 2013). Future research may consider examining other aspects of depression (e.g., anhedonia) to better elucidate the association between depression and sleep in youth with IBD. Moreover, the PANAS-C may not account for the unique psychological issues present within pediatric IBD, including intrusiveness and anxiety about gastrointestinal symptoms (Banez & Cunningham, 2009). A recently developed measure of gastrointestinal-specific anxiety, the Visceral Sensitivity Index (Labus et al., 2004), may better assess the psychological concerns of this population in relation to disrupted sleep and thus, should be considered in future studies. Despite these concerns, the PANAS-C was specifically chosen because its daily rating may be more responsive to factors in the child’s life that vary day-to-day and impact mood.
Our results should be considered in the light of study limitations. First, the current sample included children and adolescents spanning a wide age range (i.e., 10–18 years). Although this age range allows for increased generalizability of results, there are drastic changes regarding sleep patterns that occur across this developmental range. Specifically, decreasing sleep duration, slow wave sleep, and latency to rapid eye movement (REM) sleep have been correlated with increasing age among youth (Ohayon et al., 2004). Such differences in sleep patterns were reflected in this study’s results in that older youth experienced less objective TST. However, the significance of some age-related correlations may have been reduced due to the overall small sample size. Future studies including a larger sample size of youth with IBD should be conducted to replicate these results and determine if additional age-related associations exist within this population. The use of paper-and-pencil measures is another limitation as participants may have been susceptible to mass completion of questionnaires despite instructions to complete daily at bedtime. It is recommended that future research consider use of ecological momentary assessment (EMA) using mobile technology to examine sleep patterns in pediatric samples. EMA has been feasible with children as young as 7 years old, reduces retrospective recall bias, and increases compliance with study instructions (Heron et al., 2017). Additionally, 80% of the sample in this study was diagnosed with CD and were rated as being in a state of disease remission with only seven participants endorsing pain over the course of the study. Thus, results may not be as generalizable to pediatric patients with UC or those experiencing an active IBD flare. Because the goal of treatment for IBD is to attain and maintain disease remission, our results may instead generalize to the large proportion of patients who are in remission (i.e., approximately 48%–50% of patients with IBD; CCFA, 2014). Moreover, our findings suggest that sleep is an important factor to assess even when patients’ IBD is in remission. This study’s lack of assessment of daytime fatigue is another limitation, particularly as previous studies have demonstrated high rates of daytime fatigue in this population (e.g., Nachmias et al., 2006). Finally, although the study’s sample size should provide enough statistical power to detect medium to large effects given previous literature that used similar statistical techniques and examined similar constructs (e.g., Valrie et al., 2007), it is possible the results were impacted by type II error and the study was underpowered to detect some significant associations. Thus, future research using larger sample sizes of youth with IBD, particularly across a range of disease status, may yield additional hypothesized associations among sleep and psychological factors.
In terms of clinical utility, using actigraphy to examine sleep patterns may be one feasible way for GI providers to gauge whether their patients are experiencing sleep dysfunction. Despite its proven utility in pediatric populations (Meltzer et al., 2012), most studies conducted on sleep in pediatric IBD have relied upon sleep diaries or other subjective measures of sleep (Kinnucan et al., 2013). Moreover, although there are many less expensive accelerometer options on the market (e.g., Fitbit©), research has demonstrated such devices may overestimate sleep and can be less sensitive to differences in sleep patterns (e.g., Dickinson et al., 2016; Lee et al., 2017). While it may cost more than administering subjective measures of sleep, actigraphy can provide a wealth of data (e.g., total sleep duration, sleep efficiency, daytime napping) with less demands placed on the care team or patients than undergoing polysomnography. Pediatric psychologists may consider using actigraphy to determine whether behavioral sleep intervention with this population is effective. However, future studies are necessary to determine the efficacy and feasibility of using actigraphy to assess sleep patterns for youth with IBD in real world settings (e.g., pediatric medical clinics).
Behavioral sleep intervention may also present a pathway to improving health outcomes among youth with IBD given the current study’s results and previous literature that has identified associations between inflammatory processes and sleep difficulties (Ranjbaran et al., 2007; Swanson et al., 2011). Such interventions have recently demonstrated efficacy in improving overnight sleep, as well as acceptability and feasibility when implemented in pediatric primary care settings (Williamson et al., 2022). Given the similarities between primary care and outpatient medical specialty clinics (e.g., fast-paced and busy medical setting), it is possible that behavioral sleep interventions could show similarly positive results if implemented within pediatric GI clinics. One specific intervention to consider within pediatric GI clinics is education regarding adaptive sleep hygiene practices (e.g., consistent bedtimes and wake times across the week, limiting electronic use prior to bedtime). Such information could be provided on an educational handout to families with their patient instructions and may not place undue burden on medical providers. Moreover, a referral to a behavioral sleep specialist to provide cognitive-behavioral treatment for insomnia could be considered given this study’s findings of low TST and previous research suggesting efficacy of this approach in adolescent populations (de Bruin et al., 2015).
In conclusion, the current study provided further insight regarding the complex relationship between sleep and pain in pediatric IBD. When participants experienced longer SOL, they were more likely to experience pain the following day, suggesting that SOL may be a potential target for sleep interventions for youth with IBD in the future. Having pain also influenced children’s same-day sleep efficiency positively, indicating that pain during the day may not have a negative impact on that night’s sleep. It is worth noting that the current study had some methodological strengths. This study utilized objective sleep data, which is less susceptible to biases associated with subjective measurement, and a prospective study design that allows for investigation of temporal associations between study variables. Because poor sleep quality and circadian rhythm disruption may be associated with gastrointestinal inflammation (Swanson et al., 2011) and inflammatory processes have been linked to sleep disturbance (Ranjbaran et al., 2007), it is essential for gastroenterologists to screen for sleep quality and consider referral to sleep medicine or behavioral sleep specialists for youth with IBD who screen positive for sleep difficulties. This screening may be especially important among adolescents and those experiencing abdominal pain, even if their IBD is in remission, given current results. It also is necessary for future research to identify additional variables that are amenable to intervention (e.g., depression, GI-specific anxiety) in efforts to promote sleep in this population. Such research may provide evidence on how to foster positive psychosocial functioning and optimal health outcomes among children and adolescents with IBD.