Skip to main content
Free AccessSleep Restriction

Gray Matter Changes Following Cognitive Behavioral Therapy for Patients With Comorbid Fibromyalgia and Insomnia: A Pilot Study

Published Online: by:12


Study Objectives:

Insomnia frequently co-occurs with fibromyalgia, which is associated with gray matter atrophy. We examined the effect of cognitive behavioral therapy for insomnia (CBT-I) and pain (CBT-P) on cortical thickness.


Patients with fibromyalgia and insomnia underwent MRI before and after random assignment to CBT-I (n = 14), CBT-P (n = 16), or waitlist control (WLC; n = 7).


Repeated-measures analyses of variance revealed significant interactions for two regions (left lateral orbitofrontal cortex, left rostral middle frontal, Ps < .05) and trends for four regions (right medial orbitofrontal cortex, right posterior cingulate, left caudal middle frontal, left transverse temporal; Ps < .10). Cortical thickness increased in all regions for CBT-I and decreased in five regions for CBT-P and WLC. Hierarchical regressions revealed that for the CBT-I group, reductions in wake after sleep onset were associated with an increase in cortical thickness.


Our pilot study presents novel evidence suggesting that CBT-I may slow or reverse cortical gray matter atrophy in patients with fibromyalgia and insomnia.

Clinical Trial Registration:

Registry:, Identifier: NCT02001077, Title: Sleep and Pain Interventions in Fibromyalgia (SPIN), URL:


McCrae CS, Mundt JM, Curtis AF, Craggs JG, O'Shea AM, Staud R, Berry RB, Perlstein WM, Robinson ME. Gray matter changes following cognitive behavioral therapy for patients with comorbid fibromyalgia and insomnia: a pilot study. J Clin Sleep Med. 2018;14(9):1595–1603.


Current Knowledge/Study Rationale: Insomnia commonly occurs in patients with fibromyalgia. The extent to which gray matter changes may be affected by treatment for chronic pain and insomnia is unknown.

Study Impact: Results from the current pilot study suggest that cognitive behavioral therapy for insomnia may lead to slowing or reversal of cortical atrophy in neural regions associated with central sensitization of pain.


Insomnia and chronic pain are highly prevalent, occurring co-morbidly in 50% to 80% of cases,1,2 and place an enormous burden on individuals and society in terms of health care costs and decreased productivity as well as diminished quality of life. Sleep and pain exert reciprocal effects on one another, as pain can disrupt sleep, and poor sleep often leads to increased pain.3 Although the interplay between the two conditions is apparent, the extent to which they may actually share causal factors is less well understood. In recent years, attempts to understand the underlying causes of chronic pain have focused largely on investigating the role of the central nervous system, and a sizeable body of research emphasizes the strong contributions of central sensitization to chronic pain.4 Similarly, researchers have identified chronic hyperarousal as a contributing factor in the development and maintenance of insomnia.5 As pain and sleep disturbance are so commonly comorbid, it may be the case that they share a common neurophysiological pathway. This notion finds support in the Cognitive Activation Theory of Stress (CATS), which posits that “subjective health complaints” such as pain, fatigue, and gastrointestinal symptoms may result from a shared underlying psychobiological sensitization.6

Individuals with fibromyalgia (FM) are an appropriate clinical population in which to study the underlying mechanisms of pain and insomnia, as almost all individuals with FM report problems with sleep.7 Sleep disturbance and fatigue are so common among FM sufferers that the American College of Rheumatology's recent revision of FM diagnostic criteria8 included these symptoms as core features of the disorder. Research using magnetic resonance imaging (MRI) has demonstrated that FM is associated with atrophy of cortical gray matter in regions such as the amygdala, anterior cingulate cortex, insula, medial frontal cortex, parahippocampus, pre-frontal cortex, and thalamus.912 It is unclear whether FM may actually be related to increased gray matter in some regions, though some research supports this idea.13 There has been less neuroimaging research regarding insomnia, but there is evidence that insomnia is associated with reduced gray matter in the hippocampus, orbitofrontal cortex, and precuneus,14,15 and that the severity and duration of insomnia may be negatively associated with hippocampal volume.16,17 However, not all studies have found an association between insomnia and hippocampal volume,18,19 or gray matter volume more generally.20 Given that the cause of FM is still unknown, it remains to be determined whether changes in cortical gray matter are associated with its etiology.

Although prior research suggests that FM and—perhaps to a lesser extent—insomnia are associated with changes in brain morphology, the extent to which gray matter changes may be affected by treatment for chronic pain and insomnia is unknown. Based on research with other chronic pain populations, there is reason to believe that treatment of chronic pain can lead to changes in cortical thickness. Notably, a longitudinal study of chronic low back pain sufferers found increases in gray matter after treatment with surgery or injections.21 Reversal of gray matter loss has also been demonstrated among individuals who underwent surgical treatment for hip osteoarthritis.22,23 A recent study provided evidence that behavioral interventions may also lead to increased cortical gray matter among patients with chronic pain.24 In this pilot study, we sought to investigate whether individuals with FM and comorbid insomnia (FMI) showed changes in cortical thickness after receiving cognitive behavioral therapy for pain (CBT-P) or insomnia (CBT-I). Because prior research has most commonly found these conditions (especially FM) to be associated with reductions in gray matter, we hypothesized that individuals who underwent treatment would show increases in cortical thickness compared to a waitlist control (WLC) group. We had no hypothesis as to whether CBT-P or CBT-I would show better outcomes, as only one study has examined the effect of CBT-P on gray matter, and no studies have investigated CBT-I.



This analysis used data from a larger clinical trial (NCT02001077) investigating the efficacy of CBT-I and CBTP for FMI. From 2009 to 2012, individuals with FMI were recruited from the community for the parent study, and a subset of participants also underwent MRI before and after 8 weeks of treatment. The collection of neuroimaging data was supported by a supplement grant that started 6 months after recruitment for the parent study began. The CONSORT diagram for the current analysis (Figure 1) therefore does not include the 234 individuals who inquired about the study during the first 6 months of recruitment, as they did not have the opportunity to participate in the MRI portion of the study. The sample in the current analysis included 37 participants (95% female) with an average age of 55.9 years (standard deviation [SD] = 12.2). Most of the sample was Caucasian (86.5%), with the remaining identifying as either African American (10.8%) or Native American (2.7%). Latino ethnicity was endorsed by 8.1% of the participants. Baseline demographic characteristics for each treatment group are presented in Table 1. The groups did not differ on baseline demographic and clinical characteristics (all Ps > .05).

Figure 1: Study flow.

Details available from the first author upon request. CBT-I = cognitive behavioral therapy for insomnia, CBT-P = cognitive behavioral therapy for pain, WLC = waitlist control.

Table 1 Baseline demographic and clinical characteristics by treatment group.

Table 1

Individuals were eligible to participate if they reported suffering from FM for at least 6 months, and the presence of FM was confirmed by tender point testing, using guidelines established by the American College of Rheumatology (with application of 4-kg force, participants reported pain in at least 11 of 18 tender points, including points in all four body quadrants). The presence of insomnia was determined based on standard research and diagnostic criteria: (1) individual-reported insomnia (sleep onset or awake time during night > 30 minutes) at least 3 nights/wk for more than 6 months; (2) sleep diary-confirmed insomnia (sleep onset or awake time during night > 30 minutes) at least 6 nights during a 2-week baseline period; and (3) daytime dysfunction due to insomnia (mood, cognitive, social, or occupational impairment).25,26 Prescription and over-the-counter sleep medications were allowed provided the participant had been stabilized on the medication for at least 6 months. Pain medications were also allowed.

Individuals were excluded from participation in the parent study for the following reasons: (1) sleep disorder other than insomnia, specifically sleep apnea (apnea-hypopnea index greater than 15 events/h or between 10–15 events/h with oxygen saturation below 88%) or periodic limb movement disorder (periodic limb movements of sleep index greater than 15 events/h); (2) bipolar disorder or seizure disorder (due to potential risk of sleep restriction treatment); (3) significant medical (eg, cancer) or neurological disorder (eg, dementia); (4) severe untreated psychopathology (eg, schizophrenia, substance abuse); (5) cognitive impairment based on Mini-Mental State Examination score below 26; and (6) concurrent participation in other nonpharmacological sleep treatment. Additionally, for the neuroimaging portion of the study, individuals were excluded if they were pregnant or had implanted metal objects or electrical devices.

Study procedures were conducted at the University of Florida (UF). The UF Health Science Center Institutional Review Board (IRB-01) approved the trial protocol (#627-2007). All participants provided written informed consent.


Randomization and Blinding

Following baseline assessments, participants were randomly assigned by computer-generated block randomization to one of three groups: CBT-I, CBT-P, or WLC. Participants were informed of their assignment by the project coordinator. Researchers involved in recruitment and who obtained and assessed outcomes were blinded to assignment, as were the statisticians who undertook the analysis. Because of the nature of the treatment, interventionist and participant blinding was not possible.


CBT-I and CBT-P interventions were offered to participants randomized into these groups at no charge. These programs consisted of 8 weekly 50-minute individualized sessions, and were administered by doctoral students in the clinical psychology program at the UF. Therapists were trained and supervised by a licensed clinical psychologist (C.S.M.). Participants were given take-home workbooks that contained details on treatment techniques and rationales. The CBT-I intervention consisted of sleep education and training regarding how to practice proper sleep hygiene (session 1), training on how to practice stimulus control (session 2), autogenic relaxation (session 3), sleep restriction (session 4), cognitive therapy (sessions 5–7), and review of techniques and long-term maintenance (session 8). The CBT-P intervention involved pain education (session 1), progressive muscle relaxation (session 2), education on the activity-rest cycle, and adaptive techniques for pacing activity and managing rest and autogenic relaxation (session 3), additional activity-rest cycle problem solving and visual imagery relaxation (session 4), cognitive restructuring (sessions 5–7), and review of techniques and long-term maintenance (session 8). The WLC completed “treatment as usual” and were offered either CBT-I or CBT-P treatment at no charge following study completion.


As part of the parent trial, participants completed a baseline assessment that included the following: interview, tender point testing, questionnaires (pertaining to sleep, pain, and mood), single-night ambulatory polysomnography, 2 weeks of actigraphy, 2 weeks of sleep diaries (including pain ratings), and quantitative sensory testing. Participants included in this analysis also underwent MRI as part of the baseline assessment. Participants completed the same assessment and scanning procedures at the conclusion of the 8-week treatment period. The average length of time between scans was 87.0 days (SD = 35.0). Intervention details are available from the first author and will be published in a forthcoming paper. Baseline and posttreatment sleep and pain outcomes for each treatment group are presented in Table 2. There were no significant baseline differences for any variable (all Ps > .05). The group × time interaction was significant for wake after sleep onset (WASO) (F2,34 = 6.23, P = .005, η2p = .27) and trended toward significance for total wake time (F2,34 = 3.18, P = .05, η2p = .16). In both instances, only the CBT-I group demonstrated improvement (decreased wake time during the night, Ps < .001), whereas CBT-P and WLC showed no change (Ps > .05).

Table 2 Baseline and posttreatment sleep and pain outcomes by group.

Table 2

Data Acquisition and Image Preprocessing

Neuroimaging data were collected with a research-dedicated Phillips Achieva 3.0T full body scanner and an eight-channel head coil, from participants who were placed in the head first, supine position. A high-resolution three-dimensional anatomical image was acquired during the scanning session (T1-weighted, 180 slices, 1 mm3 voxels, repetition time = 2000 ms, echo time = 4.13 ms, flip angle = 8°, matrix: 240 × 240 mm, field of view = 240 mm).

The anatomical data were processed with the FreeSurfer image analysis suite version 5.1.0, a program freely available for download ( Technical details about the imaging pipeline are described on the web-site and in published articles referenced therein. The processing steps included: (1) motion correction, (2) skull stripping/ brain extraction, (3) automated warping into standardized space, (4) tissue segmentation, (5) intensity normalization, (6) enhancement of the gray/white matter boundary, (7) automated topology correction, and (8) surface deformation and reconstruction. To extract reliable thickness estimates, images were automatically processed with the longitudinal stream. Once complete, anatomical masks were used to extract the thickness values from multiple brain regions along the cortical ribbon of each hemisphere. These values were then exported and analyzed using SPSS version 24 (IBM Corp, Armonk, New York, United States).

Regions of Interest

Based on previous literature, we examined the effects of treatment on cortical thickness in regions known to be associated with cortical atrophy among individuals with FM and insomnia. We therefore examined the following cortical regions bilaterally (as segmented by FreeSurfer): frontal9,11,12 (caudal middle, rostral middle, and superior), orbitofrontal14 (lateral and medial), cingulate912 (caudal anterior, posterior, and rostral anterior), temporal13 (inferior, middle, superior, and trans-verse), insula,9,10 parahippocampal,6,8,9,11,12,15 and precuneus.11 Fifteen separate regions of interest (ROI) were identified.

Statistical Analysis

To explore patterns of cortical thickness change across groups, separate repeated-measures analyses of variance (ANOVAs) were conducted for each ROI in the left and right hemisphere to assess the effects of treatment (CBT-I versus CBT-P versus WLC) on changes in cortical thickness (baseline and posttreatment). Potential covariates (FM duration, sex, and age) were found to not be significantly related to cortical thickness and were thus not included in the ANOVAs. Given the exploratory nature of the current pilot study, we accepted the false-positive risk, and no familywise error correction was applied. Therefore, ANOVAs were evaluated at an alpha level of P < .05. Statistically significant ANOVAs were followed up by planned pairwise comparisons to determine whether there were significant differences across time points for each treatment groups.


Cortical Thickness Changes By Group

As shown in Table 3, two regions demonstrated a significant group × time interaction: left lateral orbitofrontal cortex, F2,34 = 5.15, P < .05, η2p = .23; and left rostral middle fron -tal cortex F2,34 = 3.67, P < .05, η2p = .18 (for illustrations see Figure 2). Post hoc pairwise comparisons indicated that in both regions, WLC experienced significant thinning (lateral orbitofrontal mean = −0.054, standard error [SE] = 0.024; rostral middle frontal mean = −0.064, SE = .024). Additionally, CBT-I became thicker in the left lateral orbitofrontal cortex (mean = 0.034, SE = 0.017), and CBT-P showed thinning in the left rostral middle frontal (mean = −0.033, SE = 0.016).

Table 3 Repeated-measures analysis of variance group × time interactions for cortical thickness change.

Table 3
Figure 2: Cortical regions with significant or near-significant repeated-measures analysis of variance.

In four additional regions, the group × time interaction approached but did not meet the .05 criterion for significance: right medial orbitofrontal cortex, right posterior cingu-late, left caudal middle frontal, and left transverse temporal (see Figure 2 and Table 3). Post hoc pairwise comparisons were also conducted for these ANOVAs to examine whether within-group changes in thickness were statistically significant (pretreatment and posttreatment means and effect sizes for these changes are shown in Table 4). Although not all changes were significant, there was a general trend of cortical thickening in the CBT-I group and cortical thinning among those assigned to CBT-P and WLC (see Table 4 and Figure 3). Within the CBT-I group, all six regions became thicker (range = 0.012 mm to 0.038 mm), and effect sizes were generally in the medium or large range (η2p range = .02 to .25) although the observed change in thickness was statistically significant in only the left lateral orbitofrontal cortex (t13 = 2.00, P = .05, η2p = .25). In contrast, individuals in the CBT-P and WLC groups showed thinning in all regions (with the exception of a 0.015-mm increase in the right posterior cingulate for CBT-P, η2p = .04), with large effect sizes for almost all regions (η2p range = .06 to .52). Cortical thinning in the left rostral middle frontal region was statistically significant for both CBT-P, t15 = −2.06, P < .05, η 2p = .34, and WLC, t6 = −2.67, P < .05, η2p = .52. For the WLC group, significant changes were also observed in the left caudal middle frontal (t6 = −2.22, P < .05, η2p = .36) and left lateral orbitofrontal (t6 = −2.25, P < .05, η2p = .43).

Figure 3: Effect sizes (η2p) for cortical thickness changes in each region.

Positive values indicate an increase in thickness from pretreatment to posttreatment, and negative values indicate a decrease in thickness. CBT-I = cognitive behavioral therapy for insomnia, CBT-P = cognitive behavioral therapy for pain, WLC = waitlist control.

Table 4 Average thickness by group.

Table 4

Predictors of Cortical Thickness Change

Based on the observed pattern of change across groups, the next analyses examined how behavioral changes observed over the course of treatment were related to changes in cortical thickness. Hierarchical regressions were conducted for the two regions with a significant group × time interaction (left lateral orbitofrontal, left rostral middle frontal) in order to examine whether changes in sleep and pain predicted changes in cortical thickness. For these analyses, we used sleep diaries and ratings of pain intensity, as these measures parallel each other in terms of subjectivity and have the greatest clinical relevance. Given the better outcomes for CBT-I, sleep was hypothesized to have a stronger relationship to cortical thickness change than pain. Residualized change scores were entered into the regressions, with total sleep time (TST; total minutes asleep), WASO (total minutes awake between sleep onset and final awakening), and pain intensity as predictors. Results showed that increased cortical thickness in both regions was predicted by improved WASO but not by TST or pain (see Table 5).

Table 5 Hierarchical regressions predicting change in cortical thickness.

Table 5

We also examined whether use of sleep or pain medications had an effect on changes in cortical thickness. We compared residual change scores for left lateral orbitofrontal and left rostral middle frontal cortices between those who used or did not use sleep and pain medications using independent sample t tests. Left lateral orbitofrontal cortical thickness changes did not differ between those who used or did not use sleep medications, t35 = −0.49, P = .63, or pain medications, t35 = −0.08, P = .94. Similarly, left rostral middle frontal cortical thickness changes did not differ between those who used or did not use sleep medications, t35 = 0.07, P = .95, or pain medications, t35 = 0.02, P = .99.


FM and insomnia have been previously associated with atrophy of cortical gray matter, and our preliminary findings suggest that cognitive behavioral treatment may slow or even reverse this atrophy. In this pilot study, although individuals in a control group showed thinning of the cortex relative to baseline, individuals who received CBT-P showed less atrophy than controls during this same time period, and individuals who received CBT-I actually demonstrated increases in cortical thickness following treatment. In contrast, a previous study of CBT-P demonstrated increased cortical thickness.24 A number of critical differences between that study and ours may account for the lack of replication of the effects of CBT-P. First, given our small sample sizes in each group in this pilot study, it is possible that the study was underpowered to detect cortical thickness changes in this group. Second, and perhaps most importantly, the study populations differed, as Seminowicz et al.24 examined individuals with six types of pain (only 1 of the 13 participants had FM). Additionally, their article makes no mention of the sleep characteristics of the sample, whereas our sample had chronic insomnia. Differences in the intervention type (group versus individual) and length (11 weeks versus 8 weeks) may also have contributed to disparate outcomes. Finally, our study had a shorter average interval between scans (approximately 109 days versus 87 days).

Although we expected the WLC group to fare worse than those in treatment, the emergence of potentially different patterns for individuals who received CBT-P and CBT-I in our pilot study is interesting, unexpected, and warrants follow-up in future larger scale studies. However, although preliminary, our cortical thickness results parallel the results of the larger clinical trial, which indicated that the CBT-I group fared better in terms of behavioral outcomes (manuscript in preparation). Specifically, CBT-I showed greater improvements in WASO, sleep efficiency (SE; percent of total time in bed actually spent sleeping), and sleep quality. However, neither CBT-I nor CBTP showed significantly improved TST or pain intensity. Previous research has suggested TST is a critical sleep parameter, and reduced TST has been associated with worse cardiovascular health,27 increased daytime fatigue,28 and cognitive dys-function.29 Although we did not observe change in this variable in patients with FM following CBT-I, it is important to note that in the context of insomnia, where WASO is a hallmark symptom, CBT-I is particularly effective in decreasing this variable in patients with insomnia.30 Additionally, a number of previous studies examining the efficacy of CBT-I for various chronic pain populations have similarly found improvements in WASO, and SE (as well as SOL), without any accompanying reduction in pain (with the exception of a study involving patients with osteoarthritis, which found improvement on two items of the Short Form-36 related to pain severity and interference).26,3133 It is possible that in patients with FM, WASO is particularly important as an indicator of increased arousal or central sensitization. Thus, the mechanism by which neural changes occur in this group is likely through reduced arousal, reflected by reduced WASO, rather than other sleep parameters such as increased TST. Collectively, these results suggest that the key improvement experienced by patients with chronic pain who receive CBT-I is a gain in terms of sleep consolidation rather than actual amount of sleep, and there is generally no improvement seen in pain. Although our CBT-I group experienced the expected improvements in sleep, pain intensity did not improve significantly in our sample; the improvement in pain intensity reported by our CBT-P group (mean = 5.4 on a 0–100 scale) indicated a small effect size (d = .30). Similarly, a recent meta-analysis of CBT for FM34 concluded that such treatments generally yield only a small improvement in pain intensity at posttreatment (d = .29).

As discussed previously, there is reason to believe that FM and chronic insomnia may—at least to some extent—share an underlying physiological mechanism. The CATS model35 suggests that health complaints such as pain and sleep may stem from chronic arousal, and this notion is also present in etiological conceptualizations of both pain (ie, central sensitization) and insomnia (ie, hyperarousal). Reducing physiological arousal through the use of relaxation was a major focus of both treatment protocols used in this study. Participants in both treatments were taught specific strategies and instructed to practice relaxation twice per day (morning and evening for CBT-P, daytime and bedtime for CBT-I). In addition to relaxation, both treatments included a cognitive therapy component designed to correct maladaptive sleep- or pain-specific cognitions. Finally, both treatments provided behavioral techniques aimed at creating a more stable routine and reducing variability in the targeted symptom—pain (via activity scheduling) or sleep (via sleep restriction and stimulus control). Thus, the treatments had shared components that targeted cognitions, affect, and arousal, but they diverged with regard to behavioral techniques more directly related to modifying sleep or pain. In the end, the CBT-I group demonstrated improved sleep consolidation and increased cortical thickness, whereas the CBT-P group did not have an improvement in either sleep or pain and showed a pattern of continued cortical thinning. It is therefore conceivable that the different patterns of cortical thickness change for the treatment groups may reflect the restorative benefit of consolidated sleep experienced only by the CBT-I group. Particularly, in agreement with a previous study on patients with insomnia,14 the left lateral orbitofrontal cortex may play a key role in restorative sleep. This initial finding of a reversal of cortical thinning suggests that this may be an early indicator of a reversal of central sensitization. These preliminary findings have clinical implications as they suggest that a brief behavioral treatment for insomnia may be associated with cortical plasticity. Furthermore, the patterns of change across specific brain regions could improve our understanding of mechanisms underlying FM and insomnia.


The small sample size for each group in the current pilot study limits the generalizability of our findings. Importantly, however, this study provides the first (to our knowledge) promising preliminary evidence of potential changes in cortical thickness in certain ROIs in patients with comorbid fibromyalgia and insomnia, following behavioral treatment for insomnia. Therefore, future studies with larger sample sizes are needed to improve statistical power and verify these observations in each treatment group, and examine whether the cortical thickening in the CBT-I group is associated with better sleep consolidation. Additionally, the current study design did not allow us to examine the durability of these results in order to investigate whether the improvement seen in the CBT-I group would be maintained over time or revert to pretreatment cortical thickness. Furthermore, it is unknown whether the results of the current study generalize to individuals with chronic pain conditions other than FM.


This pilot study offers novel promising evidence that a behavioral intervention as short as 8 weeks is sufficient to alter central nervous system structure in patients with FMI. Importantly, our preliminary results suggest that CBT-P seemed to merely attenuate cortical atrophy, whereas CBT-I produced increases in cortical thickness. These data serve as a starting point for future work that might lead to larger changes in structure and more meaningful clinically relevant changes in sleep and pain experiences. Further study with larger sample sizes and longer follow-up will be necessary in order to determine the validity and limits of plasticity in this population and investigate the durability of these changes.


Work for this study was performed at the University of Florida. All authors have seen and approved this manuscript. This research was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (R01AR055160 and R01AR055160-S1; McCrae, PI; Robinson, Co-PI). The authors report no conflicts of interest.



analysis of variance


Cognitive Activation Theory of Stress


cognitive behavioral therapy for insomnia


cognitive behavioral therapy for pain




fibromyalgia with comorbid insomnia


magnetic resonance imaging


wake after sleep onset


waitlist control


region of interest


total sleep time


University of Florida


  • 1 Taylor DJ, Mallory LJ, Lichstein KL, Durrence HH, Riedel BW, Bush AJComorbidity of chronic insomnia with medical problems. Sleep; 2007;302:213-218, 17326547.

    CrossrefGoogle Scholar
  • 2 Baker S, McBeth J, Chew-Graham CA, Wilkie RMusculoskeletal pain and co-morbid insomnia in adults; a population study of the prevalence and impact on restricted social participation. BMC Fam Pract; 2017;181:17, 28173767.

    CrossrefGoogle Scholar
  • 3 Roizenblatt S, Neto NSR, Tufik SSleep disorders and fibromyalgia. Curr Pain Headache Rep; 2011;155:347-357, 21594765.

    CrossrefGoogle Scholar
  • 4 Woolf CJCentral sensitization: implications for the diagnosis and treatment of pain. Pain; 2011;1523:S2-S15, 20961685.

    CrossrefGoogle Scholar
  • 5 Merrigan JM, Buysse DJ, Bird JC, Livingston EHInsomnia. JAMA; 2013;3097:733, 23423421.

    CrossrefGoogle Scholar
  • 6 Ursin H, Eriksen HRThe cognitive activation theory of stress. Psychoneuroendocrinology; 2004;295:567-592, 15041082.

    CrossrefGoogle Scholar
  • 7 Theadom A, Cropley M, Humphrey K-LExploring the role of sleep and coping in quality of life in fibromyalgia. J Psychosom Res; 2007;622:145-151, 17270572.

    CrossrefGoogle Scholar
  • 8 Wolfe FNew American College of Rheumatology criteria for fibromyalgia: a twenty-year journey. Arthritis Care Res; 2010;625:583-584.

    CrossrefGoogle Scholar
  • 9 Kuchinad A, Schweinhardt P, Seminowicz DA, Wood PB, Chizh BA, Bushnell MCAccelerated brain gray matter loss in fibromyalgia patients: premature aging of the brain?J Neurosci; 2007;2715:4004-4007, 17428976.

    CrossrefGoogle Scholar
  • 10 Robinson ME, Craggs JG, Price DD, Perlstein WM, Staud RGray matter volumes of pain-related brain areas are decreased in fibromyalgia syndrome. J Pain; 2011;124:436-443, 21146463.

    CrossrefGoogle Scholar
  • 11 Lutz J, Jäger L, de Quervain Det al.White and gray matter abnormalities in the brain of patients with fibromyalgia: A diffusion-tensor and volumetric imaging study. Arthritis Rheum; 2008;5812:3960-3969, 19035484.

    CrossrefGoogle Scholar
  • 12 Burgmer M, Gaubitz M, Konrad Cet al.Decreased gray matter volumes in the cingulo-frontal cortex and the amygdala in patients with fibromyalgia. Psychosom Med; 2009;715:566-573, 19414621.

    CrossrefGoogle Scholar
  • 13 Schmidt-Wilcke T, Luerding R, Weigand Tet al.Striatal grey matter increase in patients suffering from fibromyalgia-a voxel-based morphometry study. Pain; 2007;132:S109-S116, 17587497.

    CrossrefGoogle Scholar
  • 14 Altena E, Vrenken H, Van Der Werf YD, van den Heuvel OA, Van Someren EJReduced orbitofrontal and parietal gray matter in chronic insomnia: a voxel-based morphometric study. Biol Psychiatry; 2010;672:182-185, 19782344.

    CrossrefGoogle Scholar
  • 15 Riemann D, Voderholzer U, Spiegelhalder Ket al.Chronic insomnia and MRI-measured hippocampal volumes: a pilot study. Sleep; 2007;308:955-958, 17702263.

    CrossrefGoogle Scholar
  • 16 Neylan TC, Mueller SG, Wang Zet al.Insomnia severity is associated with a decreased volume of the CA3/dentate gyrus hippocampal subfield. Biol Psychiatry; 2010;685:494-496, 20598672.

    CrossrefGoogle Scholar
  • 17 Noh HJ, Joo EY, Kim STet al.The relationship between hippocampal volume and cognition in patients with chronic primary insomnia. J Clin Neurol; 2012;82:130-138, 22787497.

    CrossrefGoogle Scholar
  • 18 Winkelman JW, Plante DT, Schoerning Let al.Increased rostral anterior cingulate cortex volume in chronic primary insomnia. Sleep; 2013;367:991-998, 23814335.

    CrossrefGoogle Scholar
  • 19 Winkelman JW, Benson KL, Buxton OMet al.Lack of hippocampal volume differences in primary insomnia and good sleeper controls: an MRI volumetric study at 3 Tesla. Sleep Med; 2010;116:576-582, 20466585.

    CrossrefGoogle Scholar
  • 20 Spielberger CD, Gorsuch RL, Lushene Ret al.State-Trait Anxiety Inventory, Form Y. Palo Alto, CA: Consulting Psychologists Press; 1983.

    Google Scholar
  • 21 Seminowicz DA, Wideman TH, Naso Let al.Effective treatment of chronic low back pain in humans reverses abnormal brain anatomy and function. J Neurosci; 2011;3120:7540-7550, 21593339.

    CrossrefGoogle Scholar
  • 22 Gwilym SE, Filippini N, Douaud G, Carr AJ, Tracey IThalamic atrophy associated with painful osteoarthritis of the hip is reversible after arthroplasty: A longitudinal voxel-based morphometric study. Arthritis Rheum; 2010;6210:2930-2940, 20518076.

    CrossrefGoogle Scholar
  • 23 Rodriguez-Raecke R, Niemeier A, Ihle K, Ruether W, May AStructural brain changes in chronic pain reflect probably neither damage nor atrophy. PLoS One; 2013;82:e54475, 23405082.

    CrossrefGoogle Scholar
  • 24 Seminowicz DA, Shpaner M, Keaser MLet al.Cognitive-behavioral therapy increases prefrontal cortex gray matter in patients with chronic pain. J Pain; 2013;1412:1573-1584, 24135432.

    CrossrefGoogle Scholar
  • 25 American Academy of Sleep MedicineInternational Classification of Sleep Disorders: Diagnostic and Coding Manual2nd edWestchester, IL: American Academy of Sleep Medicine; 2005.

    Google Scholar
  • 26 Edinger JD, Bonnet MH, Bootzin RRet al.Derivation of research diagnostic criteria for insomnia: report of an American Academy of Sleep Medicine Work Group. Sleep; 2004;278:1567-1596, 15683149.

    CrossrefGoogle Scholar
  • 27 Cappuccio FP, Cooper D, D'elia L, Strazzullo P, Miller MASleep duration predicts cardiovascular outcomes: a systematic review and meta-analysis of prospective studies. Eur Heart J; 2011;3212:1484-1492, 21300732.

    CrossrefGoogle Scholar
  • 28 Lockley SW, Barger LK, Ayas NTet al.Effects of health care provider work hours and sleep deprivation on safety and performance. Jt Comm J Qual Patient Saf; 2007;3311 Suppl:7-18.

    CrossrefGoogle Scholar
  • 29 Banks S, Dinges DFBehavioral and physiological consequences of sleep restriction. J Clin Sleep Med; 2007;35:519-528, 17803017.

    LinkGoogle Scholar
  • 30 Okajima I, Komada Y, Inoue YA meta-analysis on the treatment effectiveness of cognitive behavioral therapy for primary insomnia. Sleep Biol Rhythm; 2011;91:24-34.

    CrossrefGoogle Scholar
  • 31 Currie SR, Wilson KG, Pontefract AJ, deLaplante LCognitive-behavioral treatment of insomnia secondary to chronic pain. J Consult Clin Psychol; 2000;683:407-416, 10883557.

    CrossrefGoogle Scholar
  • 32 Jungquist CR, O'Brien C, Matteson-Rusby Set al.The efficacy of cognitive-behavioral therapy for insomnia in patients with chronic pain. Sleep Med; 2010;113:302-309, 20133188.

    CrossrefGoogle Scholar
  • 33 Vitiello MV, Rybarczyk B, Von Korff M, Stepanski EJCognitive behavioral therapy for insomnia improves sleep and decreases pain in older adults with co-morbid insomnia and osteoarthritis. J Clin Sleep Med; 2009;54:355-362, 19968014.

    LinkGoogle Scholar
  • 34 Bernardy K, Klose P, Busch AJ, Choy EH, Häuser WCognitive behavioural therapies for fibromyalgia. Cochrane Database Syst Rev; 20139:CD009796, 24018611.

    Google Scholar
  • 35 Eriksen HR, Ursin HSensitization and subjective health complaints. Scand J Psychol; 2002;432:189-196, 12004958.

    CrossrefGoogle Scholar