Table 1 Study and intervention characteristics
Author Country, Recruitment dates Setting Study design | Inclusion Number randomised (intervention; control) Age % female Common treatment | Intervention/ control Number of sessions/ duration Therapist Fidelity | Follow up Outcomes Losses to follow up Risk of bias issues Methodological quality scorea Key results | |
|---|---|---|---|---|
1. Psychological interventions | ||||
1.a. Cognitive behavioral therapy focusing on insomnia compared with control | ||||
Abbasi et al. 2016 [39] Iran, 2014 Setting not specified Parallel group RCT | Multiple sclerosis, at least 6 months from time of diagnosis, score of 5 or more for PSQI quality of sleep 72 (36; 36) Mean age 34 years 100% female | CBT targeting sleep quality 8 weekly 90 min sessions Psychiatric nurse Fidelity not reported Group sessions talking about feelings and experiences. Treatment with “common drugs” 3 sessions Therapist not described | Before, immediately and 1 month after treatment PSQI. No pain, HRQoL, psychological health measures. Adverse events not reported. Overall loss to follow up 8% Unclear risk of bias as selected outcomes reported and methodological detail of trial limited. 19/28 Mean score of sleep quality of patients in the intervention group had a significant difference at 3 stages of before, immediately and 1 month after the intervention. | |
Currie et al. 2000 [40] Canada, dates not stated Group Parallel group RCT | Chronic musculoskeletal pain (excluding fibromyalgia) and sleep difficulties, age < 60 years 60 (32; 28) Mean age 45 years 55% female | CBT. Sleep and insomnia education, behavioural therapy, relaxation training, cognitive restructuring, sleep hygiene education. Coping with chronic pain sleep problems manual. 7 × 2 h group sessions held once a week for 7 weeks Clinical psychology doctoral students or interns with training in CBT. Structured manual. Regular supervision Waiting list controls | Baseline, end of treatment (7 weeks), 3 months Sleep diaries. Actigraphy measured nocturnal activity levels. PSQI. Multidimensional Pain Inventory Pain Severity scale, no HRQoL measure, BDI. Adverse events not reported. Overall loss to follow up 15% at 3 months Unclear risk of bias due to a waiting list control group 18/28 Authors concluded that short-term CBT improves sleep in people with insomnia secondary to chronic pain. | |
Edinger et al. 2005 [41] USA, date not stated Clinic Parallel group RCT | Fibromyalgia (ACR diagnosis) aged 21–65 with insomnia 47 (18 CBT; 11 control; 18 sleep hygiene) Mean age 49 years 96% female | CBT for insomnia 6 weekly individual sessions. First 45–60 min, then 15–30 min Clinical psychologists Fidelity not reported Usual care Sleep hygiene | Baseline, end of treatment week 3, post-intervention, 6 months Insomnia symptoms questionnaire. McGill Pain questionnaire, no general HRQoL, POMS, SF-36 MH. Adverse events not reported. Overall loss to follow up 45% at 6 months Unclear risk of bias concern due to limited reporting. High risk at 6 months due to high loss to follow up and in comparisons including control due to uneven randomisation. Small study 19/28 The CBT group achieved nearly a 50% reduction in their nocturnal wake time by study completion compared with 3.5% in usual care and 20% in sleep hygiene groups. 57% of CBT group met strict subjective sleep improvement criteria at end of treatment compared with 0% in usual care and 17% in sleep hygiene groups. | |
Jungquist et al. 2010 [42] USA, dates not stated Community Parallel group RCT | Chronic spinal pain with insomnia. Age 25+ years 28 (19;9) Mean age 49 years 78% female | CBT for insomnia 8 weekly sessions (30–90 min) Trained masters prepared nurse therapist Fidelity assessed with sleep diaries and actigraphy Appointment with nurse therapist. Reviewed sleep/pain diaries and BDI items above a scale score of 0 for prior week. Discussed rationale that life stress and depression likely to contribute to insomnia and pain. No directed form of therapy provided. 8 weekly sessions, 45–90 min Nurse therapist | Baseline, post-treatment (8 weeks) Sleep diaries, actigraphy, insomnia severity index, Epworth sleepiness scale. Pain disability index, multidimensional pain inventory. No HRQoL or anxiety, BDI Adverse events collected on checklist of symptoms but not reported in article. Overall loss to follow up 25% Low risk of bias. 23/28 Significant improvements were found in sleep as well as in the extent to which pain interfered with daily functioning. | |
McCrae et al. 2019 [43] USA, 2009–2013 Clinic Parallel group RCT | Fibromyalgia (ACR criteria) and insomnia aged 18 years or older 113 (39; 37; 37) Mean age 53 years 97% female | CBT for insomnia. Sleep education, sleep hygiene and stimulus control, relaxation, sleep restriction, cognitive therapy – monitoring automatic thoughts, cognitive therapy – challenging/ replacing dysfunctional thoughts, cognitive therapy – practical recommendations, review of skills and long-term maintenance 8 weekly 50 min sessions Predoctoral students in clinical psychology with training and weekly supervision All treatment sessions audiotaped. Half were randomly selected for scoring by another interventionist, and 25% of the scored tapes were double-scored for reliability by the lead supervising clinical psychologist. Interventionists encouraged adherence and emphasised importance of regular home practice, which was monitored by daily practice logs CBT for pain. Pain education and diaphragmatic breathing, progressive muscle relaxation, activity-rest cycle and autogenic relaxation, visual imagery, cognitive therapy – monitoring automatic thoughts, cognitive therapy – challenging /replacing dysfunctional thoughts, cognitive therapy – balanced thinking, review of skills and long-term maintenance 8 weekly 50 min sessions Predoctoral students in clinical psychology with training and weekly supervision Fidelity as with CBT for insomnia Waiting list control | Baseline, post-treatment, 6 months Self-report sleep diary, dysfunctional beliefs and attitudes about sleep, actigraph, ambulatory polysomnography. McGill Pain questionnaire, VAS pain, Pain Disability Index. No HRQoL, STAI, BDI. Adverse events reported. Overall study loss to follow up 35% Unclear risk of bias due to concern for high loss to follow up 24/28 CBT for insomnia and CBT for pain led to improvements in self-reported sleep and this was sustained at 6 months after CBT for insomnia. No differences in pain compared with control group | |
Pigeon et al. 2012 [44] USA, date not reported Setting not reported Parallel group feasibility/ pilot RCT | Chronic non-malignant pain in the spine, shoulders, hips or limbs (unrelated to autoimmune disease or fibromyalgia) and insomnia 21 (6;6;5;4) Mean age 51 years 67% female | CBT for insomnia 10 individual weekly sessions 2 CBT psychologists Sessions videotaped and rated using treatment fidelity instrument created for study CBT for insomnia and pain 20 individual weekly sessions (1 each of CBT modalities) 2 CBT psychologists Sessions videotaped and rated using treatment fidelity instrument created for study CBT for pain 10 individual weekly sessions 2 CBT psychologists Sessions videotaped and rated using treatment fidelity instrument created for study Waiting list control | Baseline, 10 weeks (end of intervention) Insomnia severity scale, Epworth sleepiness scale, sleep diary. Multidimensional Pain Inventory, no HRQoL. Adverse events not reported No losses to follow up Unclear risk of bias for wait list controls. Pilot/ feasibility study 20/28 Authors concluded that a combined CBT for insomnia and pain intervention was feasible to deliver. | |
Smith et al. 2015 [45] USA, 2008–2013 Clinic Parallel group RCT | Knee osteoarthritis and insomnia 100 (50; 50) Mean age 59 years 79% female | CBT for insomnia including sleep restriction therapy, stimulus control therapy, cognitive therapy for insomnia, and sleep hygiene education. 8 × 45 minute weekly sessions Postdoctoral clinical psychologists, doctoral psychology candidates or faculty with experience in behavioural medicine. Structured checklist, supervision of intervention providers, sessions taped, and random sample assessed. Behavioral desensitization – manual adapted to match the CBT insomnia protocol for session number and duration. 8 × 45 minute weekly sessions Postdoctoral clinical psychologists, doctoral psychology candidates or faculty with experience in behavioural medicine. Structured checklist, supervision of intervention providers, sessions taped, and random sample assessed. | Baseline, mid-treatment, post-treatment, 3 months, 6 months Sleep diaries, polysomnography, actigraphy measures, Insomnia Severity Index. WOMAC pain (also VAS pain). No HRQoL or psychological health measures. No aAdverse events reported. Overall loss to follow up 27% Low risk of bias 22/28 In people with knee osteoarthritis, CBT for insomnia reduced sleep maintenance insomnia and clinical pain compared with active control. | |
Smitherman et al. 2016 [46] USA, 2011–2013 Clinic Parallel group RCT | Chronic migraine and ICSD-3 criteria for insomnia 32 (16; 16) Mean age 31 years 30.8 (12.9) 90% female | CBT for insomnia with 4 instructions in stimulus control and 1 in sleep restriction. 3 × 30 min sessions, with 2 weeks between each session. Graduate-level therapists with backgrounds in CBT and behavioural medicine Therapist fidelity to treatment protocols was assessed at each treatment session via therapist self-ratings Lifestyle modification with 5 skills taught and practiced at home including dinner at consistent time each night and consistent liquid intake. 3 × 30 min sessions, with 2 weeks between each session. Graduate-level therapists with backgrounds in CBT and behavioural medicine Therapist fidelity to treatment protocols was assessed at each treatment session via therapist self-ratings | Baseline, 2 and 6 weeks after completing treatment Actiwatch II, PSQI, Epworth Sleepiness Scale. Headache severity (0–10), no HRQoL, GAD-7 anxiety, PHQ-9 depression. Adverse events not reported. Overall loss to follow up 22% Low risk of bias, analysis was ITT 24/28 Significant group differences favoring CBT for insomnia in PSQI scores, total sleep time and sleep efficiency. No difference in Epworth Sleepiness Scale or change in headache severity between groups. | |
Tang et al. 2012 [47] UK, date not stated Clinic Parallel group pilot RCT | Chronic pain (non-malignant, not neurological) and clinical insomnia. Age 18–65 years 24 (12;12) Mean age 48.5 years 90% female | Hybrid of CBT for insomnia combined with interventions designed to target cognitive-behavioural processes maintaining chronic pain – sleep psychoeducation, stimulus control therapy, sleep restriction therapy, cognitive therapy, individual formulation, goal setting, reducing pain catastrophizing, reversing mental defeat 4 weekly 2 h individual sessions Clinician and health psychologist Intervention according to checklist which specified the content of each session. All sessions were video recorded. A sample of the recordings (20%) was independently reviewed for treatment fidelity. Wait list control. Monitoring group kept a pain and sleep diary for 4 weeks and then received the hybrid CBT intervention | Baseline, after treatment (4 weeks) Insomnia Severity Index, Anxiety and Preoccupation about sleep questionnaire, dysfunctional beliefs and attitudes about sleep questionnaire, pain-specific sleep beliefs, pre-sleep arousal scale, sleep onset latency, wake after sleep onset, total sleep time, sleep efficiency. Brief Pain Inventory. No HRQoL. HADS anxiety and depression. Adverse events not reported. Overall loss to follow up 17% Unclear risk of bias for wait list controls 19/28 Described as pilot RCT but authors reported that compared with symptom monitoring, the hybrid intervention was associated with greater improvement in sleep (as measured with the Insomnia Severity Index and sleep diary) at post-treatment. | |
Vitiello et al. 2009 [48] USA, 2001–2003 Academic medical centre Parallel group RCT | Osteoarthritis with moderate pain and insomnia. Age 55 or older 51 (23; 28) Mean age 69 years 88% female | CBT for insomnia. Stimulus control, sleep restriction, cognitive restructuring, relaxation training, sleep-hygiene education. 8 weekly 2 h classes Two clinical psychologists Fidelity not reported Attention-control stress management and wellness. Skill-training on mind-body relationship, reduction of stress and anxiety, effective communication and assertiveness, problem solving and goal setting, nutrition and exercise for individuals with chronic conditions. 8 weekly 2 h classes Physician, psychologist, nutritionist, exercise physiologist | Baseline (before treatment), post-treatment. One year not considered as included cross-over patients. Total sleep time, naps, sleep latency, wake after sleep onset, sleep efficiency. SF-36 bodily pain, no HRQoL measure, GDS. Adverse events not reported Losses to follow up not reported Unclear risk of bias due to limited reporting 21/28 CBT for insomnia improved immediate self-reported sleep and pain in older patients with osteoarthritis and comorbid insomnia compared with an attention control. | |
1b. Brief cognitive behavioural therapy and sleep hygiene compared with wait list controls | ||||
Berry et al. 2015 [49] Canada, dates not stated Outpatient clinic Parallel group RCT | Chronic non-cancer pain including neuropathic pain, musculoskeletal pain, complex regional pain syndrome, joint pain, visceral pain and headaches, age 18–80 years 132 (65;67) Mean 49 years 61% female | Education with CBT component. Education session incorporating sleep hygiene and cognitive behavioral strategies 15 min session once. Subsequently participants contacted weekly for 4 weeks by telephone to address questions or concerns about completing diaries and to ensure they were engaging in treatment strategies. Trained research assistant Consistency of intervention delivery monitored No treatment but offered after study completion | Baseline, week 1, week 2, week 3 during intervention. Week 4 at end of intervention Daily sleep diary. No pain, HRQoL or psychological measures. Adverse events not reported Overall 36% lost to follow up 20/28 High risk of bias due to large loss to follow up No differences between groups except sleep latency improved in Education group | |
1c. Cognitive behavioral therapy focusing on insomnia compared with sleep hygiene | ||||
Edinger et al. 2005 [41] Details above | ||||
Martinez et al. 2014 [50] Spain, dates not reported Clinic Parallel group RCT | Women with fibromyalgia, aged 25–60 years 64 (32;32) Mean age 48 years 100% female | CBT for insomnia 6 group 1.5 h sessions, 1 per week. 5–6 participants per group. Female therapists with experience of pain management and sleep disorders. Fidelity not reported Sleep hygiene 6 group 1.5 h sessions, 1 per week. 5–6 participants per group. Female therapists with experience of pain management and sleep disorders. Fidelity not reported | Baseline, post-treatment, 3 months, 6 months PSQI. McGill pain questionnaire (Spanish), FIQ, SCL-90-R. Adverse events not reported Overall loss to follow up 27% Low risk of bias 19/28 Authors reported that patients who received CBT for insomnia reported significant, positive and sustained changes in subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency and sleep disturbances. Patients in the sleep hygiene group only reported a significant improvement in subjective sleep quality and a trend towards improvement in sleep efficiency. | |
Miro et al. 2011 [51] Spain, dates not reported Group Parallel group RCT | Women with fibromyalgia and insomnia 44 (22;22) Mean age 46 years 100% female | CBT for insomnia 6 weekly group sessions lasting 90 mins Female CBT experts with experience in fibromyalgia Fidelity not reported Sleep hygiene 6 weekly group sessions lasting 90 mins Female CBT experts with experience in fibromyalgia Fidelity not reported | Baseline, 1 week post-intervention PSQI. McGill pain questionnaire (Spanish). Adverse events not reported Overall loss to follow up 9% Low risk of bias 23/28 Authors reported that compared with the sleep hygiene group, patients receiving CBT for insomnia had improved sleep quality. No difference in pain between groups | |
Sanchez et al. 2012 [52] Spain, dates not reported Clinic Parallel group RCT | Fibromyalgia with chronic insomnia. Age 25–60 years 26 (13;13) Mean age 47 years 100% female | CBT for insomnia 6 weekly group sessions of 90 mins each Female CBT experts with experience in fibromyalgia Fidelity not reported Sleep hygiene 6 weekly group sessions of 90 mins each Female CBT experts with experience in fibromyalgia Fidelity not reported | Baseline, 6 weeks (post-treatment) Polysomnographic parameters. No pain outcome. Adverse events not reported. Losses to follow up not reported Low risk of bias 22/28 Authors conclude that use of CBT for insomnia in fibromyalgia patients can significantly improve objective sleep parameters | |
1d. Cognitive behavioural therapy focusing on insomnia and pain compared with control | ||||
Castel et al. 2012 [53] Spain, dates not stated Clinic Parallel group RCT | Fibromyalgia, age 18–65 years 93 Mean age 50 years 97% female | CBT for pain and insomnia 14 weekly sessions, 120 mins each, all group except session 2 Not described who delivered No information on fidelity CBT for pain and insomnia plus hypnosis 14 weekly sessions 120 mins each, group except session 2. Session 2 received hypnosis training. Hypnosis exercises performed at end of each session instead of autogenic. Audio compact disc for home practice with analgesic self-hypnosis exercises. Not described who delivered. No information on fidelity Standard care | Baseline, 1 week post-treatment, 3 months, 6 months Medical outcomes study sleep scale. NRS pain, FIQ, HADS total. Adverse events not reported Overall loss to follow up 24% Unclear risk of bias due to limited reporting 17/28 CBT for pain and insomnia showed greater improvements in sleep than standard care. Adding hypnosis led to increased benefit. | |
Lami et al. 2018 [54] Spain, dates not reported Clinic, group Parallel group RCT | Women with fibromyalgia and insomnia aged 25–65 years 126 (42;42;42) Mean age 50 years 100% female | CBT for insomnia and pain 9 weekly 90 min group (5–7 people) sessions Therapists with high level professional training and experience in chronic pain and sleep disorders Fidelity assessed with video recordings and regular meetings CBT for pain 9 weekly 90 min group (5–7 people) sessions Therapists with high level professional training and experience in chronic pain and sleep disorders Fidelity assessed with video recordings and regular meetings Control group receiving usual care | Baseline (pre-treatment), 1 week after completion of the intervention, 3 months post-treatment PSQI. McGill Pain Questionnaire Short Form, FIQ, SCL-90-R. Adverse events not reported Large losses to follow-up (28%) at post-treatment timepoint High risk of bias due to large loss to follow up 20/28 Authors report significant improvement in sleep variables in CBT for pain and insomnia compared with CBT for pain and control. CBT groups had improvements in pain outcomes compared with control | |
Pigeon et al. 2012 [44] Details above | ||||
Vitiello et al. 2013 [55] USA, 2009–2011 Classroom or primary care clinic Cluster RCT | Osteoarthritis and insomnia aged 60+ 367 (122;122;123) Mean age 73 years 75% female | CBT for pain and insomnia 6 weekly 90-min sessions 2 female mental health professionals (MSc level family counselor and PhD psychologist) Authors report ongoing treatment fidelity monitoring CBT for pain 6 weekly 90-min sessions 2 female mental health professionals (MSc level family counselor and PhD psychologist) Authors report ongoing treatment fidelity monitoring Education only control 6 weekly 90-min sessions 2 female mental health professionals (MSc level family counselor and PhD psychologist) Authors report ongoing treatment fidelity monitoring | Baseline, 2 months (post-treatment), 9 months after baseline, 18 months Insomnia severity index, sleep efficiency via actigraphy. Graded Chronic pain Scale, no specific HRQoL (AIMS2 symptom subscale reflects pain). Adverse events not reported Overall loss to follow up 13% Low risk of bias 24/28 At 9 months CBT for pain and insomnia reduced insomnia severity compared with CBT for pain and control. No significant differences between CBT groups and control in sleep or pain outcomes at 18 months. | |
1e. Cognitive behavioural therapy focusing on insomnia and painversuscognitive behavioural therapy for pain | ||||
Lami et al. 2018 [54] Details above | ||||
McCrae et al. 2019 [43] Details above | ||||
Pigeon et al. 2012 [44] Details above | ||||
Vitiello et al. 2013 [55] Details above | ||||
1 f. Cognitive behavioural therapy focusing on insomnia and pain alone or with additional hypnosis | ||||
Castel et al. 2012 [53] Details above | ||||
1 g. Acceptance and Commitment Therapy based stress management compared with control 1 h. Acceptance and Commitment Therapy based stress management compared with exercise | ||||
Wiklund et al. 2018 [56] Sweden, dates not reported Clinic/conference room at hospital/ training facility Parallel group RCT | Adults (18–60 years) with chronic (> 3 months) benign neck, low back, and/or generalised pain 299 (99 ACT;100 control; 100 Exercise) Mean age 54 years % female not reported | Acceptance and commitment therapy (ACT) based stress management 7 weekly 2-h sessions Specialist ACT psychologists 61% completed 4 or more sessions Group-based exercise. Graded exercises: endurance, coordination, balance, functional strength, and movement training. After 4 weeks: increased aerobic training and decreased range of movement exercises. Individualised exercise: graded strength exercises for back, neck, abdomen, shoulders, and arms. Home exercise. One hour, twice a week for 8 weeks. Physiotherapist and physician. 65% completed 9 or more sessions Control group with group discussion of persistent pain and participants’ experiences moderated by medical or psychology student, research nurse, or licensed psychologist. 7 weekly 2-h meetings. 60% completed 4 or more sessions | Baseline, immediately post-intervention,6 and 12 months Insomnia Severity Index. Pain NRS for average pain intensity over past 7 days, no HRQoL measure, HADS. Adverse events not reported. Overall loss to follow up 25% at 6 months 20/28 Unclear risk of bias due to selective reporting. Unclear risk of bias due to losses to follow up at 6 months No benefit for ACT-based stress management over control or exercise therapy | |
1i. Mindfulness compared with control | ||||
Cash et al. 2015 [57] USA, date not specified Clinic, group Parallel group RCT | Women volunteers aged 18+ with physician diagnosed fibromyalgia 91 (51;40) Mean age not reported 100% female | Mindfulness-based stress reduction group. Formal and informal mindfulness practices including attention-focusing, sitting meditation and yoga positions taught to encourage relaxed and focused movement. Home practice guided by a workbook and audiotapes. Half-day meditation retreat Experienced, trained instructor 1 weekly 2.5-h group session for 8 weeks. Home 45 min per day, 6 days a week No fidelity relating to delivery reported Waiting list controls | Baseline, after 8 week programme, 2 months after completion of programme Stanford Sleep Questionnaire. VAS pain. Adverse events not reported Overall loss to follow up 25% Unclear risk of bias, authors reported ITT with analysis including baseline data for losses to follow up. Wait list controls 20/28 Authors reported that the mindfulness intervention was associated with significant and maintained reductions of sleep problems and symptom severity compared with controls | |
Van Gordon et al. 2017 [58] UK, 2012–2014 Community Parallel group RCT | Fibromyalgia. Age 18–65 years 148 (74;74) Mean age 47 years 83% female | Meditation awareness training – Second-generation mindfulness-based intervention. Taught component; facilitated group discussion, guided meditation and/or mindfulness exercises. One-to-one support sessions and compact disc of guided meditations to facilitate daily self-practice. 8 weekly 2 h 25 participant workshops plus CD of guided meditation for self-practice. 2 one-to-one support sessions. Instructors attended a 3-year supervised training programme. At least 15 min of each weekly session were observed and discussions held with the programme facilitator on a weekly basis. Group CBT with education focus. Taught presentation, facilitated group discussion component, guided discovery educational exercises with the same number and duration of breaks as the target intervention. Identical to the intervention condition on all non-specific factors such as overall course length, individual session duration, group and one-to-one discussion component, group size, and inclusion of an at-home practice element Programme duration and intensity, deliverer, and fidelity assessment as with Meditation awareness training intervention. | Baseline, post-treatment, 6 month follow-up PSQI. Short Form-McGill Pain Questionnaire, DASS. Adverse events not reported Losses to follow-up at final assessment 43% High risk of bias due to large loss to follow up and retrospective registration 24/28 Authors report that Meditation awareness training participants demonstrated significant and sustained improvements over control group participants in sleep quality and pain perception. | |
1j. Relaxation compared with control | ||||
Soares and Grossi 2002 [59] Sweden, dates not reported Community Parallel group RCT | Women with fibromyalgia 53 (18;18;17) Mean age 45 years 100% female | Behavioural intervention. Individual sessions focused on applied relaxation. Group sessions covering symptoms, stress, behaviour patterns and self-management 5 individual sessions (1 h each) and 15 group sessions (2 h each:3–5 patients in each group) over 10-weeks Experienced CBT therapist and pain physician Sessions observed Educational intervention with a sleep hygiene focus 2 individual sessions (2 h each) and 15 group sessions (2 h each/3–5 patients per group) over 10 weeks Experienced physiotherapist and occupational therapist Sessions observed Waiting list control | Baseline (before treatment), after treatment and 6 months (only post-treatment for wait list controls) Karolinska Sleep Questionnaire.McGill Pain Questionnaire, The Pain Questionnaire, pain VAS, FIQ. No psychological measures. Adverse events not reported No losses to follow up post-treatment. Overall 23% lost to follow up at 6 months Unclear risk of bias for relaxation/ education comparison due to wait list control 18/28 Behavioural intervention associated with long-term improvement in sleep quality. No other differences between groups at long-term follow up | |
2. Sleep hygiene compared with control | ||||
Edinger et al. 2005 [41] Details above | ||||
Soares and Grossi 2002 [59] Details above | ||||
3. Exercise | ||||
3a. Group exercise compared with control | ||||
Arcos-Carmona et al. 2011 [60] Spain, 2008–2009 Clinic Parallel group RCT | Fibromyalgia (ACR diagnosis) in previous 2 years, age 30–60 years 56 (28; 28) Mean age 44 years 100% female | Aerobic exercise plus relaxation. 30 min graded aerobic exercise in pool followed by 30 mins progressive relaxation Twice a week for 10 weeks Not stated who delivered Fidelity not reported Sham magnetotherapy. Patients lying down with the screen covered to avoid detecting that the machine was disconnected. 10 min cervical spine, 10 min lumbar. Twice a week for 10 weeks Not stated who delivered Fidelity not reported | Baseline, at end of 10 week intervention PSQI. SF36 bodily pain. Adverse events not reported Overall 5% lost to follow up Unclear risk of bias due to limited methodological reporting 20/28 The authors report that the combination of aerobic exercise and progressive relaxation improved night rest | |
Durcanet al. 2014 [61] Ireland, date not stated Home Parallel group RCT | Rheumatoid arthritis 80 (42;38) Mean age 60 years 64% female | Home-based cardiovascular exercise, resistance training, flexibility and neuromotor conditioning. 12 weeks Assessment by doctor and senior physiotherapist every 3 weeks Fidelity not described Advice on benefits of exercise | Baseline and post-treatment (12 weeks) PSQI. VAS pain. Adverse events not reported Overall loss to follow up 3% Low risk of bias 23/28 Significant improvement in sleep quality and pain in exercise group but not in control group | |
Eadie et al. 2013 [62] Ireland, 2008–2010 Outpatient Parallel group feasibility RCT | Chronic or recurrent non-specific low back pain. Age 18–70 years 60 (20 Home-based walking; 20 Control; 20 Supervised exercise) Mean age 45 years 62% female | Home-based walking programme 8 weeks. 30 min moderate-intensity physical activity, 5 days/ week by week 5 Physiotherapist 77.5% patient adherence Supervised exercise class Once per week for 8 weeks Physiotherapist 50% patient adherence Control of usual physiotherapy. 1-to-1 advice, manual therapy and exercise Number and duration at discretion of treating physiotherapist Physiotherapist Fidelity not described (Supervised exercise class) | Baseline, 3 and 6 months (equivalent to post-treatment and 3 months) PSQI, Insomnia Severity Index, Pittsburgh Sleep Diary, actigraphy. NRS pain, SF-36 MH, HADS. Adverse events documented reported to the trial coordinator Overall 30% lost to follow up at 6 months Unclear risk of bias. High risk of bias at 6 months follow up due to high loss to follow up 21/28 Feasibility study only | |
Freburger et al. 2010 [63] USA, dates not stated Community Parallel group RCT | Self-report arthritis and activity limitations, 18 years or over 321 (166; 155) Mean age 70 years 88% female | Exercise programme. Low-to-moderate-intensity physical activity program: People With Arthritis Can Exercise (PACE)/ Arthritis Foundation Exercise Program. Land-based exercise, health education on arthritis self-management and exercise. Activities to promote social interaction, movement, balance, and body awareness. Relaxation techniques. I hour, twice weekly for 8 weeks PACE trained instructor Fidelity not assessed Waiting list control | Baseline, 8 weeks Jenkins sleep scale. Pain not reported. Adverse events not reported Overall 15% lost to follow up Unclear risk of bias due to wait list control 21/28 No sustained benefit for improved sleep beyond the 8 week, end of programme assessment | |
Wiklund et al. 2018 [56] Details above | ||||
3b. Home-based walking programme versus supervised exercise | ||||
Eadie et al. 2013 [62] Details above | ||||
3c. Moderate aerobic exercise compared with low intensity home-based exercise | ||||
Al-Sharman et al. 2019 [64] Jordan, 2015–2018 Clinic and control at home Parallel group pilot RCT | Multiple sclerosis with poor sleep quality, age > 18 years 40 (20;20) Mean age 35 years 77% female | Moderate-intensity aerobic exercise programme. Supervised moderate-intensity aerobic exercise using a recumbent stepper machine. Upper and lower body stretching exercises before and after each exercise session 40 min per session, 3 times a week for 6 weeks Not stated who delivered Weekly exercise logs Low-intensity home-exercise programme. DVD and exercise manual. DVD showed warm up and cool down activities, flexibility, strength, balance and endurance exercises. Also, relaxation, stretching, and breathing techniques. 50–60 min per session, 3 sessions a week for 6 weeks Participants asked to demonstrate exercises with examiner. Weekly exercise logs | Baseline, end of intervention (6 weeks) PSQI, Insomnia Severity Index and Actigraph in subsample of patients. Pain not reported. Adverse events not reported Overall loss to follow up 25% High risk of bias due to high and uneven loss to follow up and limited reporting of methods 19/28 Pilot study only | |
3d. Aquatic biodance compared with stretching exercises | ||||
Lopez-Rodriguez et al. 2013 [65] Spain, 2011 Clinic Parallel group RCT | Fibromyalgia (ACR diagnosis) in previous 2 years, age 18–68 years. 76 (38;38) Mean age 55 years 100% female | Aquatic biodance for 1 h in pool heated to 29C. 10 min flexibility/ breathing exercises, 40 min creative dance movement to music involving upper/lower limbs, 10 min gentle exercise Twice weekly 1 h session for 12 weeks Not stated who delivered Fidelity not reported Stretching exercises including neck, trunk, quadriceps and calves. Twice weekly 1 h sessions for 12 weeks | Baseline, post-treatment PSQI. McGill Pain questionnaire, FIQ, STAI, CES-D. Adverse events not reported Overall 22% lost to follow up Unclear risk of bias, concern for loss to follow up but ITT 22/28 Intervention associated with improvements in sleep quality and pain | |
3e. Tai Ji Quan | ||||
Lu et al. 2017 [66] China, 2013 Community Parallel group RCT | Knee osteoarthritis aged 60–70 years 46 (23;23) Mean age 65 years 100% female | Tai Ji Quan: 8 forms adapted for use in people with osteoarthritis 60 min session 3 times weekly for 24 weeks Two instructors with training and academic specialisation in Tai Ji Quan Specialist monitored fidelity of delivery on a weekly basis Education focusing on wellness and health promotion 60 min class twice per week for 24 weeks Additional 10–15 min weekly check-in phone call from research staff to monitor activity levels, changes in knee pain, and medication usage Fidelity not reported | Baseline (before intervention), at end of study (after intervention), 24 weeks PSQI. WOMAC pain, SF-36 MH. Adverse events not reported Overall 13% lost to follow up at 24 weeks Low risk of bias 23/28 Significant improvement in sleep measures and pain in Tai Ji Quan group compared with controls | |
Maddali Bongi et al. 2016 [67] Italy, dates not stated Community and home Parallel group RCT | Women with fibromyalgia (ACR criteria) 44 (22;22) Mean age 52 years 100% female | Tai Ji Quan: breathing exercises, concentration, postural maintenance, rebalancing, and precise movement. Twice weekly 60 min sessions for 16 weeks. Daily DVD led home exercises in two 15 min sessions Not reported who delivered Fidelity not described Education about the disease, symptoms, management and coping Twice weekly 60 min sessions for 16 weeks Fidelity not described | Baseline, 16 weeks (end of treatment) PSQI. Widespread pain index, HADS. Adverse events not reported No losses to follow up reported (6 withdrew before study) Unclear risk of bias due to limited reporting 20/28 Improvement in some sleep parameters and widespread pain in Tai Ji Quan group but not in education group | |
4. Physical therapy | ||||
4.1. Hydrotherapy | ||||
Calandre et al. 2009 [68] Spain, dates not stated Clinic Parallel group RCT | Fibromyalgia aged 18+ years 81 (39;42) Mean age 50 years 90% female | Hydrotherapy exercise with stretching 18 sessions of 1 h, 3 times per week for 6 weeks Physiotherapist Fidelity not reported Hydrotherapy exercise with Tai Chi 18 sessions of 1 h, 3 times per week for 6 weeks Physiotherapist Fidelity not reported | Baseline, end of treatment (6 weeks), 4 weeks after end of treatment, 12 weeks after end of treatment PSQI. VAS pain, FIQ, STAI, Bdi. Adverse reactions listed Overall loss to follow up 30% High risk of bias, large loss to follow up 19/28 No differences were found between groups | |
Vitorino et al. 2006 [69] Brazil, dates not reported Clinic Parallel group RCT | Women with fibromyalgia 50 (25;25) Mean age 48 years 100% female | Hydrotherapy with warm-up, stretching, aerobic exercises and relaxation 60 mins, 3 times per week for 3 weeks Physiotherapist Fidelity not described Conventional physiotherapy with infra-red lamp, stretching, aerobic exercise and relaxation 60 mins, 3 times per week for 3 weeks Physiotherapist Fidelity not described | Baseline (pre-treatment), post-treatment. Sleep logs completed for 21 days before and after treatment Total sleep time, total nap time. SF-36 bodily pain, SF-36 MH. Adverse events not reported Overall loss to follow up 6% Low risk of bias 21/28 Sleep quality showed grater improvement in hydrotherapy group but no difference in improvement in pain outcome | |
4.2. Massage or Manual therapy compared with relaxation or control | ||||
Field et al. 2007 [70] USA, dates not stated Group clinic and home-based Parallel group RCT | Chronic lower back pain 30 total Mean age 41 years 47% female | Massage therapy 30 mins twice per week for 5 weeks Trained massage therapists Fidelity not reported Relaxation therapy 30 mins twice per week for 5 weeks Home-based Patient log | Baseline, 5 weeks (post-treatment) Verran and Snyder-Halperin Sleep scale. VAS pain, POMS, STAI. Adverse events not reported Losses to follow up not reported Unclear risk of bias for limited reporting. Baseline difference in sleep disturbance 20/28 Authors report sleep disturbance and pain less in massage therapy group | |
Castro-Sanchez et al. 2014 [71] Spain, dates not stated Clinic Parallel group RCT | Fibromyalgia, age 18–70 years 89 (45;44) Mean age 54 years 53% female | Manual therapy 5 weekly sessions of 45 mins each Specialist physiotherapist Fidelity not described No treatment | Baseline, post-intervention (48 h after end of 5 week intervention) PSQI. McGill Pain Questionnaire. Adverse events not reported No losses to follow up Unclear risk of bias due to lack of blinding 23/28 Authors concluded that manual therapy was effective in improving sleep and pain | |
4.3. Physical therapy programme | ||||
Külcü et al. 2009 [72] Turkey, 2006–2007 Rehabilitation clinic Parallel group RCT | Primary fibromyalgia age 18–55 years 60 (40;20) Mean age 37 year 95% female | Physical therapy programme with hot pack, ultrasound, TENS and low power laser 15 sessions Fidelity not described No physical therapy | Baseline and at end of intervention Insomnia Severity Index. VAS pain. Adverse events not reported No losses to follow up High risk of bias due to limited reporting and lack of blinding 19/28 Sleep and pain improved in physical therapy group compared with controls | |
4.4. Pompage and stretching and aerobic exercise compared with stretching and aerobic exercise | ||||
Correia Moretti et al. 2016 [73] Brazil, 2011–2013 Clinic Parallel group RCT | Fibromyalgia, aged 18–60 years 23 (13;10) Mean age 45 years 100% female | Pompage. Global, lymph, trapezius, torso, lumbar, and quad pompage plus stretching and aerobic exercise Twice per week for 12 weeks Not specified who delivered Fidelity not reported Stretching and aerobic exercise Twice per week for 12 weeks Not specified who delivered Fidelity not reported | Baseline, 6 weeks, 12 weeks (post-treatment) Sleep inventory. McGill Pain Questionnaire, no HRQoL or psychological health measures. Adverse events not reported Overall 35% lost to follow up High risk of bias for high and uneven loss to follow up 22/28 Authors report no benefit for pompage for sleep and limited benefit regarding pain | |
5. Miscellaneous interventions | ||||
5.1. Acupressure | ||||
Yeh et al. 2016 [74] USA, dates not reported Clinic/ office Parallel group RCT | Chronic low back pain 61 (30;31) Mean age 63 years 67% female | Auricular point acupressure. Vaccaria seeds placed on active ear points corresponding to low back pain and alleviation of stress and pain. Participants told to press the seeds on each ear at least 3 times a day for 3 min and whenever they experienced pain. Seed removed after 5 days Four weekly clinic visits to place seeds Not specified who delivered Participants completed treatment diaries Sham auricular point acupressure. Vaccaria seeds taped to stomach, mouth, duodenum, and eye acupoints of ears Four weekly clinic visits to place seeds Not specified who delivered Participants completed treatment diaries | Baseline, during each of the 4 treatments, end of intervention, and 1 month after the last treatment PSQI, daily sleep diary. BPI short form, no HRQoL or psychological health measures. Adverse events not reported Overall 25% lost to follow up High risk of bias for high and uneven loss to follow up 20/28 Authors reported that auricular point acupressure led to improvement in several sleep parameters and pain | |
Murphy et al. 2019 [75] USA, 2013–2016 Clinic Parallel group RCT | Chronic nonspecific low back pain and fatigue. Aged 18+ years 67 (22;22;23) Mean age 50 years 63% female | Self-administered relaxing acupressure at 9 points. Pressure applied to each point in circular motion 27–30 min per day for 6 weeks, each point for 3 min Trained acupressure educators Standardised training of educators, proper enactment of intervention and methods to track adherence Self-administered stimulating acupressure at 10 points. Pressure applied to each point in circular motion 27–30 min per day for 6 weeks, each point for 3 min Trained acupressure educators Standardised training of educators, proper enactment of intervention and methods to track adherence Usual care. Weekly calls regarding health status to ensure similar levels of researcher contact | Baseline, post-treatment (6 weeks) PSQI. Brief Pain Inventory. Adverse events reported Overall loss to follow up 18% Unclear risk of bias due to limited reporting 23/28 Authors report no improvement in sleep quality between acupressure groups and compared with control. Pain reduced in acupressure groups but no change in control | |
5.2. Bright light therapy | ||||
Pearl et al. 1996 [76] Canada, 1992–1993 Home Crossover RCT | Fibromyalgia. Aged 21–65 years 19 randomised but crossover protocol completed by 14 people Mean age 38 years 100% female | Bright light therapy delivered by a visor with Krypton incandescent bulbs with a mean of 4750 (SD 2337) lux 4 weeks of 1 condition, break week, and 4 weeks alternative treatment Administered at home Fidelity not reported No light condition. Visor system fitted with opaque filter (exposed Kodak film) 4 weeks of 1 condition, break week, and 4 weeks alternative treatment Administered at home Fidelity not reported | Pre-light, light week 4, pre no light, no light week 4 Daily post sleep questionnaire, VAS sleep quality, daily sleep diary. Daily NRS for pain in 10 body regions, FIQ anxiety and depression. Adverse events not reported Overall 26% did not complete crossover protocol High risk of bias due to large loss to follow up and limited reporting of methods 21/28 Authors reported no significant differences between light and no light conditions on sleep or pain | |
5.3. Foot reflexology | ||||
Bakir et al. 2018 [77] Turkey, 2015 Clinic Parallel group RCT | Rheumatoid arthritis aged 18+ years 68 (34;34) Mean age 50 years 77% female | Foot reflexology 60 min repeated once a week for 6 weeks Certified researcher Treatment fidelity not reported Routine polyclinic monitoring and information | Baseline, 1 week, 6 weeks (pain recorded weekly) PSQI. VAS pain. No HRQoL or psychological health. Adverse events not reported 12% lost to follow up Unclear risk of binding due to lack of blinding 20/28 Authors report that sleep and pain improved in the foot reflexology group | |
5.4. Transcranial stimulation | ||||
Harvey et al. 2017 [78] Canada, dates not stated Laboratory Parallel group feasibility RCT | Stable musculoskeletal pain and insomnia, age 60+ years 16 (8;8) Mean age 71 years 81% female | Transcranial direct current stimulation (tDCS) applied over the primary motor cortex (2 mA, 20 min) 5 daily sessions of 20 min given in afternoon or evening Investigator No information on treatment fidelity Sham transcranial direct current stimulation applied over the primary motor cortex 5 daily sessions of 20 min given in afternoon or evening Investigator No information on treatment fidelity | Baseline, day 12 (post-treatment), and day 19 (7 days post-treatment). PSQI. Pain and sleep log books completed each day at home, actigraph days 1–19 (only available for 4 participants). VAS pain. No HRQoL or psychological health. Adverse events not reported 13% lost to follow up Feasibility study, unclear risk of bias due to limited reporting 21/28 Study provides guidelines for future studies. Authors report no difference in sleep parameters between groups | |
5.5. Mattress | ||||
Colbert et al. 1999 [79] USA, 1997 Home Parallel group RCT | Fibromyalgia (ACR criteria) 30 (15; 15) Mean age 50 years 100% female | Magnetic mattress pad strength 1100 G which delivered 200–600 G to the skin surface. Delivered to patients with instructions on placement. Each night for 16 weeks. No information on fidelity Sham mattress pad, Delivered to patients with instructions on placement. Each night for 16 weeks. No information on fidelity | Baseline, 16 weeks VAS (sleep, fatigue, tiredness on waking, total sleep time). Pain VAS, FIQ-ADL, no psychological measure. Diary of adverse reactions. Low risk of bias Overall 17% lost to follow up. 23/28 Patients sleeping on the intervention magnetic mattress pad had improvement in reported sleep (p < 0.01) and decrease in pain (p < 0.05) with no adverse events related to the magnetic mattress pad. | |
Minetto et al. 2018 [80] Italy, dates not reported Home Parallel group pilot RCT | Chronic lower back pain 38 (28; 10) Median 58 years 58% female Both groups had usual rehabilitation exercises to improve back strength, balance and mobility. 1 h daily, 3 days per week for 2 months | Mattress overlay. Aiartex overlay with suspensory monofilaments to support a person lying in bed. No supervision. Each night for 2 months. All patients “compliant with its use” Usual rehabilitation only | Baseline, end of intervention at 2 months PSQI. VAS pain. No HRQoL or psychological health. Adverse events not reported. Losses to follow up not reported High risk of bias due to randomization procedure, non-blinding and reporting. 18/28 Authors report that the mattress overlay was associated with better and clinically meaningful sleep and pain compared with controls. | |