ADVERTISEMENT

Issue Navigator

Volume 14 No. 11
Earn CME
Accepted Papers





Scientific Investigations

Sleep Discrepancy in Patients With Comorbid Fibromyalgia and Insomnia: Demographic, Behavioral, and Clinical Correlates

Wai Sze Chan, PhD1; Meredith P. Levsen, PhD2; Svyatoslav Puyat, MA2; Michael E. Robinson, PhD3; Roland Staud, MD4; Richard B. Berry, MD, FAASM4; Christina S. McCrae, PhD2
1Department of Psychiatry, Geisel School of Medicine, Dartmouth College, New Hampshire; 2Department of Psychiatry, University of Missouri, Columbia, Missouri; 3Department of Clinical and Health Psychology, University of Florida, Gainesville, Florida; 4Department of Medicine, University of Florida, Gainesville, Florida

ABSTRACT

Study Objectives:

Individuals with primary insomnia often have poorer self-reported sleep than objectively measured sleep, a phenomenon termed negative sleep discrepancy. Recent studies suggest that this phenomenon might differ depending on comorbidities. This study examined sleep discrepancy, its night-to-night variability, and its correlates in comorbid insomnia and fibromyalgia.

Methods:

Sleep diaries and actigraphy data were obtained from 223 adults with fibromyalgia and insomnia (age = 51.53 [standard deviation = 11.90] years; 93% women) for 14 days. Sleep discrepancy was calculated by subtracting diary from actigraphy estimates of sleep onset latency (SOL-D), wake after sleep onset (WASO-D), and total sleep time (TST-D) for each night. Night-to-night variability in sleep discrepancy was calculated by taking the within-individual standard deviations over 14 days. Participants completed measures of mood, pain, fatigue, sleep/pain medications, nap duration, and caffeine consumption.

Results:

Average sleep discrepancies across 14 days were small for all sleep parameters (< 10 minutes). There was no consistent positive or negative discrepancy. However, sleep discrepancy for any single night was large, with average absolute discrepancies greater than 30 minutes for all sleep parameters. Greater morning pain was associated with larger previous-night WASO-D, although diary and actigraphy estimates of WASO remained fairly concordant. Taking prescribed pain medications, primarily opioids, was associated with greater night-to-night variability in WASO-D and TST-D.

Conclusions:

Unlike patients with primary insomnia, patients with comorbid fibromyalgia do not exhibit consistent negative sleep discrepancy; however, there are both substantial positive and negative discrepancies in all sleep parameters at the daily level. Future research is needed to investigate the clinical significance and implications of high night-to-night variability of sleep discrepancy, and the role of prescribed opioid medications in sleep perception.

Citation:

Chan WS, Levsen MP, Puyat S, Robinson ME, Staud R, Berry RB, McCrae CS. Sleep discrepancy in patients with comorbid fibromyalgia and insomnia: demographic, behavioral, and clinical correlates. J Clin Sleep Med. 2018;14(11):1911–1919.


BRIEF SUMMARY

Current Knowledge/Study Rationale: Some patients with insomnia perceive their sleep more negatively than it is when objectively measured, a phenomenon linked to maladaptive beliefs about sleep and the maintenance of insomnia. Recent research has shown that patients with insomnia and comorbid medical conditions do not seem to have consistent negative sleep discrepancy.

Study Impact: Findings of this study show that adults with comorbid insomnia and fibromyalgia did not have consistent negative sleep discrepancy. Sleep discrepancy varied to a large extent from night to night, but the averages of diary and actigraphy estimates over 14 days were fairly concordant. Taking prescribed opioid medications was associated with greater night-to-night variability in the discrepancies of wake after sleep onset and total sleep time.

INTRODUCTION

Chronic insomnia is diagnosed by self-reported complaints of difficulty initiating sleep, difficulty maintaining sleep, or awakening too early for at least 3 nights per week for at least 3 months. These complaints are accompanied by significant distress or impairments in at least one area of daytime functioning.1,2 Given that insomnia is diagnosed by self-reported sleep complaints, sleep diaries are recommended for the assessment of insomnia.1,2 Research has found that diary-assessed sleep and actigraphy-assessed sleep, although correlated, are largely discrepant among older adults and individuals with chronic insomnia.36 Rather than an issue of the accuracy of assessments or reporting errors, recent research suggests that sleep discrepancy is a clinically meaningful feature of insomnia and may affect the maintenance and treatment of insomnia.7

Prior reviews have found that actigraphic estimates of sleep onset latency (SOL) and wake after sleep onset (WASO) were consistently smaller than diary estimates whereas actigraphic estimates of total sleep time (TST) were larger than diary values in individuals who have chronic insomnia.8,9 In other words, individuals with insomnia tend to experience greater sleep difficulties than objectively measured. This pattern of sleep discrepancy, also called negative sleep discrepancy, is a characteristic in some individuals with chronic insomnia and has important clinical implications. Cognitive interventions that reduce sleep discrepancy have been found to significantly improve symptoms of insomnia.10 Negative sleep discrepancy may represent an important mediator or a proximal outcome for recovery from insomnia. Harvey and Tang reviewed possible mechanisms underlying sleep discrepancy and suggested that (1) elevated cortical arousal may lead to memory bias for negative sleep experience and thus exacerbate self-reported report of poor sleep; and that (2) elevated physiological arousal prior to sleep onset may lead to transient wakening at night that is not captured by actigraphy but is perceived as wake.7 Consistent with Harvey and Tang's framework, greater sleep discrepancy has been found to be associated with higher levels of depression and anxiety in a number of populations including the elderly,11 individuals with insomnia,12 children with anxiety,13 and adults with depression.14 Additionally, situational heightened attention to negative experience, such as feelings of fatigue and poor sleep quality, have been found to be associated with greater sleep discrepancy.15,16

The prevalence of chronic insomnia is elevated in patients with chronic pain or fibromyalgia. Approximately 50% of patients with chronic pain or fibromyalgia also have chronic insomnia, compared to 10% to 15% among individuals who do not have chronic pain or fibromyalgia.17,18 Prior studies have found mixed results regarding sleep discrepancy in patients with fibromyalgia and chronic pain conditions. Wilson et al. found that patients with chronic musculoskeletal pain reported similar WASO and TST but longer SOL compared to actigraphic estimates.19 Similarly, Lunde et al. found similar diary and actigraphic estimates of TST but diary estimates of SOL were larger than actigraphy estimates of SOL in older adults with chronic pain.20 Okifuji and Hare found significant discrepancy in TST in patients with comorbid insomnia and fibromyalgia; however, their nightly analysis of diary and actigraphy data over 7 nights showed that sleep discrepancy varied greatly from night to night, from positive discrepancy (actigraphy > diary) to negative discrepancy (diary > actigraphy).21 In addition, Okifuji et al. did not report data on discrepancy in other sleep parameters such as SOL and WASO.

Few studies have examined the predictors and correlates of sleep discrepancy in patients with comorbid fibromyalgia and insomnia.21 Identifying factors that characterize patients with fibromyalgia whose diary estimates differ substantially from actigraphic measurement of their sleep will improve the understanding of sleep discrepancy and the subsequent implications in the assessment and treatment of insomnia in this population. Additionally, given that sleep discrepancy is a variable characteristic with high intraindividual night-to-night variability,22 research that evaluates both average and night-to-night variability in sleep discrepancy is important for more complete understanding of this characteristic.

The Current Study

In summary, the current study aimed to evaluate sleep discrepancy and its night-to-night variability in patients with comorbid fibromyalgia and insomnia. This study also aimed to examine a set of demographic, clinical, and behavioral correlates of sleep discrepancy and sleep discrepancy variability. It was hypothesized that correlates of higher cognitive arousal, including (1) higher levels of depressive and anxiety symptoms, (2) dysfunctional beliefs about sleep, and (3) feelings of fatigue, and higher physiological arousal related to (4) clinical pain would be associated with greater negative sleep discrepancy. Further, we examined the potential associations of behavioral factors, including daytime napping, caffeine consumption, and pain and sleep medication use with sleep discrepancy. Finally, we examined whether these clinical and behavioral factors were also associated with night-to-night variability in sleep discrepancy.

METHODS

Participants

This study analyzed baseline data of a sample recruited for a clinical trial (NCT02001077).23 Adults with fibromyalgia and insomnia were recruited from Gainesville, Florida and surrounding areas through radio, newspaper, and television advertisements. The University of Florida's Institutional Review Board (UF IRB-01) approved the protocol, and all participants signed informed consent forms prior to participating. A total of 223 participants completed an in-person baseline assessment to determine eligibility for the study. Their baseline data were used for the current study. Demographic variables are presented in presented in Table 1.

Demographic variables.

jcsm.14.11.1911.t01.jpg

table icon
Table 1

Demographic variables.

(more ...)

Measures

Sleep Diaries

Participants were instructed to complete a sleep diary for 14 days. The sleep diary provided the following variables: (1) SOL—time from initial lights-out until sleep onset, (2) WASO—time spent awake after initial sleep onset until last awakening; and (3) TST—time spent asleep.

Actigraphy

Participants wore an actigraph, the Actiwatch 2 (Philips Respironics, Bend, Oregon, United States), on their nondominant wrist for 24 hours per day for the 14 days coincident to completing the sleep diaries. Bedtime and time out of bed in the morning were based on sleep diary entries as recommended in the software manual. They were used to compute the total time in bed for subsequent calculations of SOL, WASO, and TST. If there was a greater than 30-minute discrepancy between diary and actigraphy values of bedtime and time out of bed, a three-step procedure was used to determine the bedtime and time out of bed: (1) if diary times fell within 15 minutes of a sharp decrease/increase of activity or light (50% change of activity), the diary times were used; (2) if diary times did not fall within 15 minutes of a sharp change of activity or light, but actigraphy times did, the actigraphy times were used; (3) if both diary and actigraphy times did not fall within 15 minutes of a sharp change of activity or light, the times at which there was a sharp change of activity or light, closest to the diary times, were used.

Depressive and Anxiety Symptoms

The Beck Depression Inventory-Second Edition (BDI-II)24 was used to measure depressive symptoms. The BDI-II is a 21-item self-report inventory. Each item is rated on a 3-point scale from 0 (not at all) to 3 (very much so). Typically, respondents answer for the previous week, but the previous 2 weeks were used in this study to match the 2-week activity recording period for each assessment. Total scores range from 0 to 63, with higher scores indicating greater levels of depressive symptoms. The BDI has good internal consistency (Cronbach alpha = .93) in the current study.

The State-Trait Anxiety Inventory-Form Y1 (STAI-Y1)25 was used to measure anxiety symptoms. The STAI-Y1 asks respondents to rate how true 20 self-descriptive statements (eg, “I feel calm”) are on a 4-point scale ranging from 1 (not at all) to 4 (very much so). Participants made their ratings based on how they generally felt over the preceding 2 weeks in order to match the 2-week activity recording period for each assessment. Total scores range from 20 to 80, with higher scores indicating greater maladjustment. The STAI has good internal consistency (Cronbach alpha = .95) in the current study.

Feelings of Fatigue

One item was used to measure usual level of fatigue. Participants rated their usual levels of fatigue in the past week on an analog scale from 0 (none) to 100 (worst imaginable).

Dysfunctional Beliefs and Attitudes about Sleep

The Dysfunctional Beliefs and Attitudes about Sleep (DBAS) consists of 30 questions intended to measure 5 dimensions: misconceptions about the causes of insomnia, misattributions or amplification of its consequences, unrealistic expectations, control and predictability of sleep, and faulty beliefs about sleep-promoting practices.26 Participants responded to each question on an 11-point Likert scale (0 = strongly disagree, 10 = strongly disagree). The DBAS has good internal consistency (Cronbach alpha = .83) in the current study.

Clinical Pain

Daily pain level was measured using an item on the sleep diary. Participants provided morning (AM pain) and bedtime (PM pain) ratings of current clinical pain intensity for 14 days. Ratings were made using visual analog scales (VAS) with anchors of “no pain sensation” and “most intense pain imaginable” for pain intensity. Additionally, the McGill Pain Questionnaire (MPQ)27 was used to measure sensory, affective, and evaluative dimensions of participants' pain experiences. The MPQ has good internal consistency (Cronbach alpha = .85) in the current study. The Pain Disability Inventory (PDI)28 was used to measure pain-related disabilities that are common to patients with FM. The PDI is a seven-item measure of the degree to which chronic pain interferes with patients' functioning in the following areas of life: family/home responsibilities, recreation, social activity, occupation, sexual behavior, self-care, and life-support activity. An 11-point scale ranging from 0 (no disability) to 10 (total disability) is used to indicate the amount of disability experienced in each of the domains. The PDI has good internal consistency (Cronbach alpha = .90) in the current study.

Daytime Napping and Caffeine Consumption

In the sleep diary, participants were asked to report the duration of nap (if they did nap) each day and the number of caffeinated drinks they consumed each day.

Pain and Sleep Medications

In the sleep diary, participants indicated whether they used any medication for pain and sleep each day. Three variables were derived from the responses—(1) using prescription medications for pain, (2) using over-the-counter (OTC) medications for pain, etc., and (3) using any medications for sleep. Each variable was coded as 1 or 0 indicating usage or nonusage for each day.

Data Analysis

Calculation of Sleep Discrepancy and Sleep Discrepancy Variability

Sleep discrepancy was calculated for each day for each participant by subtracting the diary value of sleep parameters (SOL, WASO, TST) from the actigraphic value for the same parameter. The daily sleep discrepancy averages for the 2-week assessment period (SOL-D, WASO-D, TST-D) were then calculated for each individual and were used as the dependent variable in multiple regression analyses (described in the next paragraph). Night-to-night sleep discrepancy variability was calculated by taking the within-individual standard deviation (SD) of daily sleep discrepancy for each participant across the assessment period (SOL-DV, WASO-DV, TST-DV). These variables were then used as another dependent variable in the multiple regressions. Because larger SD in sleep discrepancy can be a mathematical artifact of larger means of daily sleep discrepancy, all models predicting night-to-night variability in sleep discrepancy included the mean values of absolute sleep discrepancy as covariates.

Multiple Regressions

Multiple regression models were used to examine the demographic, clinical, and behavioral predictors of (1) average sleep discrepancy and (2) night-to-night variability in sleep discrepancy. We assessed demographics variables, clinical variables, and behavioral variables in three separate models for each outcome. Demographics predictors were as follows: sex (0 = male, 1 = female), age, white ethnicity (0 = no, 1 = yes), Hispanic race (0 = no, 1 = yes), marital status (0 = not married, 1 = married), level of education (coded 1 to 7; treated as a continuous variable), body mass index, number of other medical conditions, duration of sleep problems (months), and time since initial fibromyalgia diagnosis (months). Clinical predictors included BDI-II, STAIY1, fatigue, DBAS, MPQ, PDI, and averages of daily previous evening pain rating (PM pain) and subsequent morning pain rating (AM pain). Behavioral predictors included the following variables: average caffeine consumption, average nap duration, taking prescribed opioid medications, taking nonopioid pain medications, and taking sleep medications. Medication variables were coded as 0 or 1 indicating not taking the medication or taking the medication at least once in the assessment period.

Missing data in these models were handled using the Full Information Maximum Likelihood estimation implemented in the R package Lavaan. All models were analyzed using R version 3.3.1. Bonferroni-corrected P values were used in each of these models to correct for the inflated type II error due to testing multiple variables.

Mixed Models

Mixed modeling was used to investigate the daily associations of daily assessed clinical and behavioral variables (ie, AM pain, PM pain, daily prescribed medication use, daily OTC medication use, daily sleep medication use, daily caffeine consumption, and daily nap duration) with daily sleep discrepancy. The daily variables were centered within individuals, and the between-individual variables (14-day averages of the daily variables) were grand-mean centered.29 For each model, we assumed an unstructured covariance structure and used the Satterthwaite degrees of freedom estimation method. We adopted the optimal random structure, including all possible random effects in the initial model and then removing the random effects that did not contribute to significant improvements of model fit.30 To estimate effect sizes, we calculated the partial R2 for the mixed models.31 This coefficient signifies the amount of variance explained by each of the dependent variables.

All available data from the 234 participants were used in the mixed models. All multilevel models (MLMs) were analyzed in R version 3.3.1 using the lme4 package.32 Bonferroni-corrected P values were used in each of these models to correct for the inflated type II error due to testing multiple variables.

RESULTS

Descriptives

The diary estimates of SOL, WASO, and TST were 50.71, 42.51, and 400.34 minutes respectively, compared to 43.25, 51.06, and 395.41 minutes measured by actigraphy. The average sleep discrepancy was small; there were less than 10 minutes of discrepancy for all of the sleep parameters (see Table 2), indicating that the averages of diary estimates of sleep parameters over 2 weeks were fairly concordant with the averages of actigraphy estimates. However, most participants both overestimated and underestimated sleep parameters by self-report at the daily level; there was substantial positive or negative discrepancy on most days (see Figure 1). The absolute differences between the two measures, on average, were large (44.15 minutes for SOL, 32.09 minutes for WASO, 83.02 minutes for TST). Diary and actigraphy estimates were mildly correlated with 1-night polysomnography estimates (see Table 3).

Descriptive data of sleep discrepancy and clinical and behavioral variables.

jcsm.14.11.1911.t02.jpg

table icon
Table 2

Descriptive data of sleep discrepancy and clinical and behavioral variables.

(more ...)

Bland-Altman plots.

Bland-Altman plots for discrepancy in sleep onset latency (SOL-D; A), wake after sleep onset (WASO-D; B), and total sleep time (TST-D; C). All numbers shown are in minutes. Lowercase letters following SOL, WASO, and TST indicate: a = actigraphy estimates, d = diary estimates. SOL-D, WASO-D, and TST-D were not consistently positive or negative; substantial positive and negative discrepancies were observed. SOL-D and WASO-D became greater as the estimates of SOL and WASO increased.

jcsm.14.11.1911a.jpg

jcsm.14.11.1911a.jpg
Figure 1

Bland-Altman plots.

(more ...)

Correlations between sleep diary and actigraphy estimates and polysomnography estimates of sleep.

jcsm.14.11.1911.t03.jpg

table icon
Table 3

Correlations between sleep diary and actigraphy estimates and polysomnography estimates of sleep.

(more ...)

We used Bland-Altman plots to illustrate the pattern and direction of sleep discrepancy by plotting daily sleep discrepancy against the average value of the diary and actigraphy estimates (the approximation of the true values). As shown in Figure 1, there was not a consistent pattern of positive or negative discrepancy across nights; the direction of the differences was about equally likely to be either positive or negative. There were trends indicating that SOL-D and WASO-D became larger as the estimates of SOL and WASO increased. This means that the discrepancy was greater when individuals had poorer sleep.

Correlates of Sleep Discrepancy

None of the demographic or behavioral variables were significantly associated with SOL-D, WASO-D, or TST-D at the adjusted significance levels (see Table 4).

Behavioral correlates of sleep discrepancy.

jcsm.14.11.1911.t04.jpg

table icon
Table 4

Behavioral correlates of sleep discrepancy.

(more ...)

Table 5 presents the mixed-model results of clinical correlates of daily and average sleep discrepancy. Within-subject AM pain was positively associated with WASO-D (F1, 87.61 = 15.70, P < .001, R2partial = .15). Consistent with our hypothesis, on days when individuals had greater morning pain, their diary estimates of their previous-night WASO tended to be larger, relative to the actigraphy estimates (see Figure 2). However, given that, on average, diary estimates of WASO were smaller than actigraphy estimates of WASO, the increase in diary estimates as morning pain increased did not increase the magnitude of the discrepancy. The discrepancy remained minimal as morning pain increased. None of the other clinical correlates were significantly associated with SOL-D or TST-D at the P < .005 Bonferroni-corrected level.

Clinical correlates of sleep discrepancy.

jcsm.14.11.1911.t05.jpg

table icon
Table 5

Clinical correlates of sleep discrepancy.

(more ...)

Association between morning pain and wake after sleep onset discrepancy (WASO-D).

As morning pain increased, diary estimates of WASO increased relative to the actigraphy estimates of WASO.

jcsm.14.11.1911b.jpg

jcsm.14.11.1911b.jpg
Figure 2

Association between morning pain and wake after sleep onset discrepancy (WASO-D).

(more ...)

Correlates of Sleep Discrepancy Variability

Within-individual night-to-night sleep discrepancy variability was large, exceeding 30 minutes (see Table 2). None of the demographic or clinical variables were significantly associated with SOL-DV, WASO-DV, or TST-DV at the adjusted significance levels (Table 6).

Clinical correlates of sleep discrepancy variability.

jcsm.14.11.1911.t06.jpg

table icon
Table 6

Clinical correlates of sleep discrepancy variability.

(more ...)

Table 7 presents the results of behavioral correlates predicting SOL-DV, WASO-DV, and TST-DV. Taking prescribed pain medications and the number of caffeinated beverages consumed were significantly associated with WASO-DV (P = .001). Individuals who took prescribed pain medications had greater WASO-DV (mean = 51.86 minutes, SD = 34.36) than those who did not (mean = 36.69 minutes, SD = 19.80). Consuming more caffeinated beverages was associated with smaller WASO-DV (P = .019), although by just about 2 minutes for each caffeinated beverage per day. Taking prescribed pain medications was also significantly associated with TST-DV (P = .008). Individuals who took prescribed pain medications had greater TST-DV (mean = 103.11 minutes, SD = 53.85) than those who did not (mean = 75.86 minutes, SD = 42.58).

Behavioral correlates of sleep discrepancy variability.

jcsm.14.11.1911.t07.jpg

table icon
Table 7

Behavioral correlates of sleep discrepancy variability.

(more ...)

DISCUSSION

The current study examined discrepancies in diary versus actigraphy estimates of multiple sleep parameters, ie, SOLD, WASO-D, TST-D, in patients with comorbid fibromyalgia and chronic insomnia. Although the 2-week averages of sleep diary and actigraphy estimates were fairly concordant, the data showed that participants both overestimated and underestimated their sleep parameters at the daily level. There was not a consistent pattern of negative or positive sleep discrepancy. This finding converges with a prior study in patients with fibromyalgia—Okifuzi and Hare found that the number of nights where there was positive discrepancy was about the same as the number of nights where there was negative discrepancy.21 Okifuzi and Hare found that the average absolute discrepancy in TST across a week was 73 minutes, fairly close to our finding of 83 minutes across 2 weeks. This finding is, however, contrary to several prior studies in which negative discrepancy was found in individuals with primary insomnia or comorbid insomnia8,9 and in patients with insomnia and chronic pain conditions.19,20

The discrepancy in these findings might be due to methodological differences such as the analysis of daily data versus aggregated data across days and the length of the assessment periods. Studies that used aggregated data across days, such as that by Lunde et al.,20 were not able to capture the high night-to-night variability in sleep discrepancy. In several prior studies, actigraphy and diary were collected for 1 to a few nights8,9,19 as opposed to the 2-week assessment in the current study. As shown in prior studies and the current study, there is high night-to-night variability in sleep discrepancy.22 The concordance between the two measures increases drastically as the assessment period increases.33

Alternatively, the lack of consistent positive or negative diary-actigraphy discrepancy might be due to population-specific characteristics. Negative sleep discrepancy is also less consistently observed in patients with comorbid insomnia and cancer.3436 Actigraphy's underestimation of SOL has been suggested to result from its inability to detect wakefulness when the sleeper exhibits very little movement while being awake.8 Individuals with fibromyalgia or some comorbid medical illnesses might exhibit more movements on some nights due to pain or pain medications.19 Hence, actigraphy might be more capable of detecting pain-related wake time on those nights, which may reduce the diary-actigraphy discrepancy.

Greater daily morning pain was associated with larger diary estimates of previous night WASO, relative to actigraphy estimates. Because participants completed the sleep diary in the morning, greater morning pain might draw one's attention to negative experiences when participants were evaluating their sleep in the previous night. Therefore, participants with greater pain in the morning might report longer wake time in the previous night. Nonetheless, because actigraphy estimates of WASO were, on average, larger than diary estimates of WASO, these two estimates actually became more concordant as morning pain increased. Therefore, although clinical pain seems to exacerbate self-reported poor sleep, it did not increase the discrepancy between diary and actigraphy estimates of WASO.

Night-to-Night Variability in Sleep Discrepancy and Its Correlates

The high night-to-night variability in sleep discrepancy is consistent with previous studies in older adults and adults with FM.21,22 It is, however, unclear what causes the variability in sleep discrepancy. Diary and actigraphy might be subject to different sources of influence and measurement errors. For instance, diary estimates might be more influenced by feelings and distress at the time of assessments whereas actigraphy is influenced by the levels of movement and perhaps physiological arousal during sleep. Although nonsignificant, taking sleep medications, caffeine consumption, and nap duration were associated with WASO-D and that fatigue and clinical pain were associated SOL-D and TST-D at trend levels (see Table 4 and Table 5). These influences could vary day by day and subsequently contribute to daily fluctuation of sleep discrepancy.

The current study found that individuals who took prescribed pain medications at least once during the assessment period had greater WASO-DV and TST-DV than those who did not by 15 and 25 minutes. The type of prescribed pain medications used in this sample was primarily opioids (only one report of nonopioid prescribed pain medication use). Prior studies have shown that prescription opioid medication usage is associated with changes in sleep architecture, including reduced deep sleep and increased light sleep, but no changes in WASO or TST as assessed by polysomnography.37 With further exploratory analyses of our data, we found that taking prescribed pain medications was significantly and positively associated with actigraphy-assessed WASO (r = .21, P = .002) but not diary-assessed WASO (r = .07, P = .31), suggesting that actigraphy might register bodily movements during light sleep as wakefulness, which is not experienced or recalled on self-reports.

Because it appears that individuals who took prescribed pain medications had greater average actigraphy-assessed WASO but not diary-assessed WASO, there might be more room for night-to-night fluctuations in actigraphy-assessed WASO but not in diary-assessed WASO, leading to greater night-to-night variability in the discrepancy between actigraphy and diary estimates of WASO. Nonetheless, these speculations need to be examined in future research before conclusions can be drawn about the mechanisms by which taking prescribed pain medications is linked to higher sleep discrepancy variability.

Limitations

Interpretations of the current findings should take into consideration of the following limitations. The sample is primarily white women, which might limit the generalizability of the current findings to other populations. Paper diaries were used. Although participants were asked to complete the diaries daily, there were no timestamps to prove that the diaries were completed accordingly. We were not able to examine discrepancies between sleep diaries and actigraphy to polysomnography, because polysomnography was only collected for a single night during the 14-day assessment period. Measures of nap duration and caffeine consumption were self-reported. There might be unsystematic overestimations or underestimations of these variables. Finally, this study is cross-sectional. Although we identified some significant correlates of sleep discrepancy, we could not determine whether these correlates play a causal role in sleep discrepancy. Future studies should be conducted to better understand how these factors influence sleep discrepancy.

CONCLUSIONS

The current findings showed that the average sleep diary and actigraphy estimates of SOL, WASO, and TST were fairly concordant over a 2-week period of assessment in adults with fibromyalgia and insomnia. However, there were substantial positive and negative discrepancies in sleep parameters (either actigraphy > diary or diary > actigraphy) at the daily level. Unlike adults with primary insomnia, patients with comorbid insomnia and fibromyalgia did not exhibit consistent negative sleep discrepancy (ie, there was not a consistent pattern of diary overestimation of SOL or WASO or diary underestimation of TST compared to actigraphy estimates at the daily level). Clinical pain at the time of assessment elevated self-reported WASO, although both actigraphy and diary estimates of WASO remained fairly concordant. Night-to-night variability in sleep discrepancy was high. Taking prescribed opioid medications was associated with greater night-to-night variability in WASO and TST. Future studies are needed to investigate sleep discrepancy as a variable characteristic from night to night and the significance and implication of night-to-night sleep discrepancy variability in the etiology and treatment of insomnia.

DISCLOSURE STATEMENT

All authors have read and approved the manuscript. This research was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (R01AR055160; PI: McCrae). The authors report no conflicts of interest.

ABBREVIATIONS

BDI

Beck Depression Inventory

BMI

body mass index

DBAS

Dysfunctional Beliefs and Attitudes about Sleep

OTC

over the counter

PDI

Pain Disability Inventory

SOL

sleep onset latency

SOL-D

discrepancy of SOL

SOL-DV

within-individual variability in the discrepancy of SOL

STAI-Y1

The State-Trait Anxiety Inventory Form Y1

TST

total sleep time

TST-D

discrepancy of TST

TST-DV

within-individual variability in the discrepancy of TST

VAS

visual analog scale

WASO

wake after sleep onset

WASO-D

discrepancy of WASO

WASO-DV

within-individual variability in the discrepancy of WASO

REFERENCES

1 

Edinger JD, Bonnet MH, Bootzin RR, et al. Derivation of research diagnostic criteria for insomnia: report of an American Academy of Sleep Medicine Work Group. Sleep. 2006;27(8):1567–1596. [PubMed]

2 

American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. Washington, DC: American Psychiatric Association Publishing; 2013.

3 

Landry GJ, Best JR, Liu-Ambrose T. Measuring sleep quality in older adults: a comparison using subjective and objective methods. Front Aging Neurosci. 2015;7:166. [PubMed Central][PubMed]

4 

Kay DB, Buysse DJ, Germain A, Hall M, Monk TH. Subjective-objective sleep discrepancy among older adults: associations with insomnia diagnosis and insomnia treatment. J Sleep Res. 2015;24(1):32–39. [PubMed]

5 

Williams JM, Kay DB, Rowe M, McCrae CS. Sleep discrepancy, sleep complaint, and poor sleep among older adults. J Gerontol B Psychol Sci Soc Sci. 2003;68(5):712–720

6 

Means MK, Edinger JD, Glenn DM, Fins AI. Accuracy of sleep perceptions among insomnia sufferers and normal sleepers. Sleep Med. 2003;4(4):285–296. [PubMed]

7 

Harvey AG, Tang NKY. (Mis)perception of sleep in insomnia: a puzzle and a resolution. Psychol Bull. 2012;138(1):77–101. [PubMed]

8 

Sadeh A. The role and validity of actigraphy in sleep medicine: an update. Sleep Med Rev. 2011;15(4):259–267. [PubMed]

9 

Lichstein KL, Stone KC, Donaldson J, et al. Actigraphy validation with insomnia. Sleep. 2006;29(2):232–239. [PubMed]

10 

Tang NKY, Harvey AG. Altering misperception of sleep in insomnia: behavioral experiment versus verbal feedback. J Consult Clin Psychol. 2006;74(4):767–776. [PubMed]

11 

Baillet M, Cosin C, Schweitzer P, et al. Mood influences the concordance of subjective and objective measures of sleep duration in older adults. Front Aging Neurosci. 2016;8:181. [PubMed Central][PubMed]

12 

Fernandez-Mendoza J, Calhoun SL, Bixler EO, et al. Sleep misperception and chronic insomnia in the general population: role of objective sleep duration and psychological profiles. Psychosom Med. 2011;73(1):88–97. [PubMed]

13 

Alfano CA, Patriquin MA, De Los Reyes A. Subjective - objective sleep comparisons and discrepancies among clinically-anxious and healthy children. J Abnorm Child Psychol. 2015;43(7):1343–1353. [PubMed]

14 

Tsuchiyama K, Nagayama H, Kudo K, Kojima K, Yamada K. Discrepancy between subjective and objective sleep in patients with depression. Psychiatry Clin Neurosci. 2003;57(3):259–264. [PubMed]

15 

Jackowska M, Dockray S, Hendrickx H, Steptoe A. Psychosocial factors and sleep efficiency: discrepancies between subjective and objective evaluations of sleep. Psychosom Med. 2011;73(9):810–816. [PubMed]

16 

Watson NF, Kapur V, Arguelles LM, et al. Comparison of subjective and objective measures of insomnia in monozygotic twins discordant for chronic fatigue syndrome. Sleep. 2003;26(3):324–328. [PubMed]

17 

Taylor DJ, Mallory LJ, Lichstein KL, Durrence HH, Riedel BW, Bush AJ. Comorbidity of chronic insomnia with medical problems. Sleep. 2007;30(2):213–218. [PubMed]

18 

White KP, Speechley M, Harth M, Ostbye T. The London Fibromyalgia Epidemiology Study: comparing the demographic and clinical characteristics in 100 random community cases of fibromyalgia versus controls. J Rheumatol. 1999;26(7):1577–1585. [PubMed]

19 

Wilson KG, Watson ST, Currie SR. Daily diary and ambulatory activity monitoring of sleep in patients with insomnia associated with chronic musculoskeletal pain. Pain. 1998;75(1):75–84. [PubMed]

20 

Lunde L-H, Pallesen S, Krangnes L, Nordhus IH. Characteristics of sleep in older persons with chronic pain: a study based on actigraphy and self-reporting. Clin J Pain. 2010;26(2):132–137. [PubMed]

21 

Okifuji A, Hare BD. Nightly analyses of subjective and objective (actigraphy) measures of sleep in fibromyalgia syndrome: What accounts for the discrepancy? Clin J Pain. 2011;27(4):289–296. [PubMed Central][PubMed]

22 

Kay DB, Dzierzewski JM, Rowe M, McCrae CS. Greater night-tonight variability in sleep discrepancy among older adults with a sleep complaint compared to noncomplaining older adults. Behav Sleep Med. 2013;11(2):76–90. [PubMed]

23 

McCrae CS, Craggs J, Vatthauer K, et al. Neuroplasticity in comorbid chronic pain and chronic insomnia: impact of improved sleep on central sensitization [abstract]. Sleep. 2014;37(abstract suppl.):A176–A177.

24 

Beck AT, Steer RA, Brown GK. Beck Depression Inventory-II. San Antonio, TX: The Psychological Corporation; 1996.

25 

Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, Jacobs GA. Manual for the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press; 1983.

26 

Morin CM, Stone J, Trinkle D, Mercer J, Remsberg S. Dysfunctional beliefs and attitudes about sleep among older adults with and without insomnia complaints. Psychol Aging. 1993;8(3):463–467. [PubMed]

27 

Melzack R. The McGill Pain Questionnaire: major properties and scoring methods. Pain. 1975;1(3):277–299. [PubMed]

28 

Pollard CA. Preliminary validity study of the pain disability index. Percept Mot Skills. 1984;59(3):974. [PubMed]

29 

Enders CK, Tofighi D. Centering predictor variables in cross-sectional multilevel models: a new look at an old issue. Psychol Methods. 2007;12(2):121–138. [PubMed]

30 

Matuschek H, Kliegl R, Vasishth S, Baayen H, Bates D. Balancing Type I error and power in linear mixed models. J Mem Lang. 2017;94:305–315.

31 

Edwards LJ, Muller KE, Wolfinger RD, Qaqish BF, Schabenberger O. An R2 statistic for fixed effects in the linear mixed model. Stat Med. 2008;27(29):6137–6157. [PubMed Central][PubMed]

32 

Bates D, Mächler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67(1):1–48.

33 

Van Someren EJ. Improving actigraphic sleep estimates in insomnia and dementia: how many nights? J Sleep Res. 2007;16(3):269–275. [PubMed]

34 

Moore CM, Schmiege SJ, Matthews EE. Actigraphy and sleep diary measurements in breast cancer survivors: discrepancy in selected sleep parameters. Behav Sleep Med. 2015;13(6):472–490. [PubMed]

35 

Dean GE, Redeker NS, Wang Y, Rogers AE, Dickerson SS, Steinbrenner LM. Sleep, mood, and quality of life in patients receiving treatment for lung cancer. Oncol Nurs Forum. 2014;40(5):441–451.

36 

Wang SY, Chang HJ, Lin CC. Sleep disturbances among patients with non-small cell lung cancer in Taiwan: congruence between sleep log and actigraphy. Cancer Nurs. 2010;33(1):E11–E17. [PubMed]

37 

Dimsdale JE, Norman D, DeJardin D, Wallace MS. The effect of opioids on sleep architecture. J Clin Sleep Med. 2007;3(1):33–36. [PubMed]