Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Review - Intervention

Mind and body therapy for fibromyalgia

Authors' declarations of interest

Version published: 09 April 2015 Version history

https://doi.org/10.1002/14651858.CD001980.pub3
Collapse all Expand all

Background

Mind‐body interventions are based on the holistic principle that mind, body and behaviour are all interconnected. Mind‐body interventions incorporate strategies that are thought to improve psychological and physical well‐being, aim to allow patients to take an active role in their treatment, and promote people's ability to cope. Mind‐body interventions are widely used by people with fibromyalgia to help manage their symptoms and improve well‐being. Examples of mind‐body therapies include psychological therapies, biofeedback, mindfulness, movement therapies and relaxation strategies.

Objectives

To review the benefits and harms of mind‐body therapies in comparison to standard care and attention placebo control groups for adults with fibromyalgia, post‐intervention and at three and six month follow‐up.

Search methods

Electronic searches of the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (Ovid), EMBASE (Ovid), PsycINFO (Ovid), AMED (EBSCO) and CINAHL (Ovid) were conducted up to 30 October 2013. Searches of reference lists were conducted and authors in the field were contacted to identify additional relevant articles.

Selection criteria

All relevant randomised controlled trials (RCTs) of mind‐body interventions for adults with fibromyalgia were included.

Data collection and analysis

Two authors independently selected studies, extracted the data and assessed trials for low, unclear or high risk of bias. Any discrepancy was resolved through discussion and consensus. Continuous outcomes were analysed using mean difference (MD) where the same outcome measure and scoring method was used and standardised mean difference (SMD) where different outcome measures were used. For binary data standard estimation of the risk ratio (RR) and its 95% confidence interval (CI) was used.

Main results

Seventy‐four papers describing 61 trials were identified, with 4234 predominantly female participants. The nature of fibromyalgia varied from mild to severe across the study populations. Twenty‐six studies were classified as having a low risk of bias for all domains assessed. The findings of mind‐body therapies compared with usual care were prioritised.

There is low quality evidence that in comparison to usual care controls psychological therapies have favourable effects on physical functioning (SMD ‐0.4, 95% CI ‐0.6 to ‐0.3, ‐7.5% absolute change, 2 point shift on a 0 to 100 scale), pain (SMD ‐0.3, 95% CI ‐0.5 to ‐0.2, ‐3.5% absolute change, 2 point shift on a 0 to 100 scale) and mood (SMD ‐0.5, 95% CI ‐0.6 to ‐0.3, ‐4.8% absolute change, 3 point shift on a 20 to 80 scale). There is very low quality evidence of more withdrawals in the psychological therapy group in comparison to usual care controls (RR 1.38, 95% CI 1.12 to 1.69, 6% absolute risk difference). There is lack of evidence of a difference between the number of adverse events in the psychological therapy and control groups (RR 0.38, 95% CI 0.06 to 2.50, 4% absolute risk difference).

There was very low quality evidence that biofeedback in comparison to usual care controls had an effect on physical functioning (SMD ‐0.1, 95% CI ‐0.4 to 0.3, ‐1.2% absolute change, 1 point shift on a 0 to 100 scale), pain (SMD ‐2.6, 95% CI ‐91.3 to 86.1, ‐2.6% absolute change) and mood (SMD 0.1, 95% CI ‐0.3 to 0.5, 1.9% absolute change, less than 1 point shift on a 0 to 90 scale) post‐intervention. In view of the quality of evidence we cannot be certain that biofeedback has a little or no effect on these outcomes. There was very low quality evidence that biofeedback led to more withdrawals from the study (RR 4.08, 95% CI 1.43 to 11.62, 20% absolute risk difference). No adverse events were reported.

There was no advantage observed for mindfulness in comparison to usual care for physical functioning (SMD ‐0.3, 95% CI ‐0.6 to 0.1, ‐4.8% absolute change, 4 point shift on a scale 0 to 100), pain (SMD ‐0.1, CI ‐0.4 to 0.3, ‐1.3% absolute change, less than 1 point shift on a 0 to 10 scale), mood (SMD ‐0.2, 95% CI ‐0.5 to 0.0, ‐3.7% absolute change, 2 point shift on a 20 to 80 scale) or withdrawals (RR 1.07, 95% CI 0.67 to 1.72, 2% absolute risk difference) between the two groups post‐intervention. However, the quality of the evidence was very low for pain and moderate for mood and number of withdrawals. No studies reported any adverse events.

Very low quality evidence revealed that movement therapies in comparison to usual care controls improved pain (MD ‐2.3, CI ‐4.2 to ‐0.4, ‐23% absolute change) and mood (MD ‐9.8, 95% CI ‐18.5 to ‐1.2, ‐16.4% absolute change) post‐intervention. There was no advantage for physical functioning (SMD ‐0.2, 95% CI ‐0.5 to 0.2, ‐3.4% absolute change, 2 point shift on a 0 to 100 scale), participant withdrawals (RR 1.95, 95% CI 1.13 to 3.38, 11% absolute difference) or adverse events (RR 4.62, 95% CI 0.23 to 93.92, 4% absolute risk difference) between the two groups, however rare adverse events may include worsening of pain.

Low quality evidence revealed that relaxation based therapies in comparison to usual care controls showed an advantage for physical functioning (MD ‐8.3, 95% CI ‐10.1 to ‐6.5, ‐10.4% absolute change) and pain (SMD ‐1.0, 95% CI ‐1.6 to ‐0.5, ‐3.5% absolute change, 2 point shift on a 0 to 78 scale) but not for mood (SMD ‐4.4, CI ‐14.5 to 5.6, ‐7.4% absolute change) post‐intervention. There was no difference between the groups for number of withdrawals (RR 4.40, 95% CI 0.59 to 33.07, 31% absolute risk difference) and no adverse events were reported.

Authors' conclusions

Psychological interventions therapies may be effective in improving physical functioning, pain and low mood for adults with fibromyalgia in comparison to usual care controls but the quality of the evidence is low. Further research on the outcomes of therapies is needed to determine if positive effects identified post‐intervention are sustained. The effectiveness of biofeedback, mindfulness, movement therapies and relaxation based therapies remains unclear as the quality of the evidence was very low or low. The small number of trials and inconsistency in the use of outcome measures across the trials restricted the analysis.

PICOs

Population
Intervention
Comparison
Outcome

The PICO model is widely used and taught in evidence-based health care as a strategy for formulating questions and search strategies and for characterizing clinical studies or meta-analyses. PICO stands for four different potential components of a clinical question: Patient, Population or Problem; Intervention; Comparison; Outcome.

See more on using PICO in the Cochrane Handbook.

Interventions focusing on the link between the mind and body for adults with fibromyalgia

Research question

What are the effects of mind and body therapy for fibromyalgia on pain, physical function, mood and side effects?

What problems does fibromyalgia cause?
People with fibromyalgia have chronic, widespread body pain, and often have fatigue (feeling tired), stiffness, depression and problems sleeping.

What are mind‐body interventions?

Mind‐body interventions include treatments such as biofeedback, mindfulness, movement therapies, psychological therapy and relaxation therapies. Biofeedback is when you are connected to electrical sensors that help you receive information about your body to make subtle changes in your body, such as relaxing. Mindfulness means having awareness of thoughts, feelings and bodily sensations. All mind‐body therapies make the link between thoughts, behaviour and feelings to help people to cope with their symptoms.

Study characteristics

We conducted a review of the effect of mind‐body therapies for adults with fibromyalgia. After searching for all relevant studies until October 2013, we found 61 studies including 4234 adults.

‐ Many studies only included female participants, but some males were included in a few studies.
‐ Participants had mild to severe fibromyalgia.
‐ Mind‐body interventions were compared to 'usual care', such as medication use. Secondary analysis also compared findings in comparison to an 'attention control therapy' which involved receiving information for the same amount of time as the mind‐body therapy.

Key results at the end of treatment

‐ Low quality evidence revealed that psychological therapies improved physical functioning, pain, mood and side effects compared to usual care. More people withdrew from the psychological therapy group compared to usual care.

‐ There was little or no difference in physical functioning, pain and mood between people receiving biofeedback and usual care but this may have happened by chance. More people withdrew from the biofeedback than the usual care group. No studies reported any side effects.

‐ There was little or no difference in physical functioning, pain, mood and the number of withdrawals between people receiving mindfulness therapy and usual care. No studies reported any adverse events.

‐ We are uncertain whether movement therapies improve physical functioning, pain, mood, side effects or the number of people who withdrew from the treatment. There were improvements in pain and mood for people receiving movement therapies but the quality of the evidence was very low. More people withdrew and two participants reported experiencing increased pain in the intervention group.

‐ We are uncertain whether relaxation therapies improve physical functioning and pain compared to usual care because the quality of evidence was very low. There was little or no difference in mood and withdrawal from treatment between people receiving relaxation therapies and those receiving usual care. No adverse events were reported.

Best estimates of what happens at the end of treatment in people with fibromyalgia when they use mind‐body therapies

The main findings on the use of psychological therapies are summarised below.

‐ Physical functioning after 1 to 25 weeks (higher scores mean greater limitations)

People who used psychological therapies rated their physical functioning as 2 points lower on a scale of 0 to 100 compared to those who received usual care (7.5% absolute improvement).

‐ Pain after 3 to 14 weeks (higher scores mean worse or more severe pain)

People who used psychological therapies rated their pain as 2 points lower on a scale of 0 to 100 compared to those who received usual care (3.5% absolute improvement).

‐ Mood (higher scores mean worse or more severe pain)

People who used psychological therapies rated their mood as 3 points lower on a scale of 20 to 80 compared to those who received usual care (4.8% absolute improvement).

‐ Withdrawing from the treatment for any reason

A total of 204 out of 1000 people withdrew from psychological therapies compared with 148 out of 1000 from usual care (6% absolute improvement).

‐ Side effects

Nineteen people out of 1000 who received psychological therapies experienced a side effect compared with 51 out of 1000 who had usual care (4% absolute improvement). This may have happened by chance.

We do not have precise information about side effects and complications of mind‐body therapies. Rare adverse events may include worsening of pain.

Authors' conclusions

Implications for practice

Mind‐body interventions are becoming increasingly incorporated into treatment programmes for fibromyalgia (FM). The findings of this review indicate that psychological therapies may improve physical functioning, pain and mood following treatment but highlight that the quality of evidence is low. The observed effects were not sustained at six months follow‐up. There was wide variation in intervention mode of delivery and there were an insufficient number of trials to enable calculation of the effect of different modes of delivery on outcomes. Psychological therapies were delivered over a period of between one and 25 weeks (mean 11 weeks), with greater effects observed with longer duration. There was low reporting on the presence or absence of adverse events, however equivalent rates of dropout between the treatment and control groups indicate that the risks to people receiving psychological interventions are low.

There is insufficient evidence to determine the use of biofeedback, mindfulness, movement therapies or relaxation based therapies for adults with FM.

Implications for research

Evidence

The evidence for outcomes in this review was limited by the low or very low quality of the trials identified. To enable recommendations on the use of mind‐body therapies for adults with fibromyalgia to be determined, robust randomised controlled trials are needed. Following the findings of this review, trials should take into account the need to accurately report on randomisation procedures and allocation concealment processes and to clearly report data including measures of variance for all outcomes at all time points assessed. There is no need for further low quality trials (trials with high risk of bias) in this area.

Population

As fibromyalgia predominantly affects females, many studies only included female participants in order to reduce heterogeneity, however this limits the applicability of the findings for males with fibromyalgia. Subgroup analyses could help explore the impact of gender on treatment effect and future studies should consider inclusion of male and female participants. This study only presents data for adults over 18 years and future reviews are needed to review the evidence of mind‐body therapies for children.

Intervention

There was wide heterogeneity in the interventions delivered within the groups specified, this was particularly evident within the sensitivity analyses of the data. For example, psychological therapies encompassed written therapies, educational based approaches as well as specific therapeutic techniques such as the Resseguier approach. There was also variability in the mode of intervention delivery with therapies being self‐administered or delivered by a therapists on an individual basis or within a group. Future reviews would benefit from having a narrower focus to ensure that the effective elements of the specific components of mind‐body therapies can be identified.

Comparison

Trials used a combination of usual care and attention control groups for comparison. Greater differences between groups were observed when the intervention was compared to usual care, suggesting that therapeutic attention or a placebo effect may have been observed.

Outcomes

It was often difficult to determine why participants withdrew from the included trials and at which time point. Higher numbers of withdrawals from the intervention group can indicate difficulties with the feasibility of the intervention or mode of delivery and should be reported; although this was not found to be the case for studies in our analyses where this information could be determined. CONSORT diagrams outlining reasons for withdrawal between groups over time are a useful way of presenting this information for future studies.

Few studies described any adverse events experienced by participants and this information is critical for ensuring the safety and feasibility of interventions in clinical practice, and should be reported. Declarations that no adverse events were experienced, if this was the case, would also facilitate interpretation of the results.

A wide range of outcomes and outcome measures were reported between trials. As recommended by Choy 2009 and Bernardy 2013, a core set of outcome measures that should be assessed and reported across clinical trials needs to be established by consensus to facilitate pooling of trial data and the comparison of study findings. As recommended by Choy 2009, the Fibromyalgia Impact Questionnaire should be considered to be the primary outcome measures for all fibromyalgia randomised controlled trials, with secondary outcomes of self‐reported pain, fatigue and sleep.

Time

This review presents data identified up to October 2013, and further updates will be required as new evidence emerges. More trials with follow‐up at three and six months are needed to determine if the effects of mind‐body interventions are sustained.

Summary of findings

Open in table viewer
Summary of findings for the main comparison. Psychological therapies compared to usual care for fibromyalgia

Psychological therapies compared to usual care for fibromyalgia

Patient or population: patients with fibromyalgia
Settings: outpatients
Intervention: psychological therapies
Comparison: usual care

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of p
(studies)

Quality of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Usual care

Psychological therapies

Functioning as assessed post‐intervention
Fibromyalgia Impact Questionnaire. Scale from: 0 to 100
Follow‐up: 1 to 5 weeks

The mean functioning as assessed post‐intervention in the control groups was
6.77

The mean functioning as assessed post‐intervention in the intervention groups was
0.43 standard deviations lower
(0.57 to 0.28 lower)

733
(10 studies)

⊕⊕⊝⊝
low1,2

SMD ‐0.4 (95% CI ‐0.6 to ‐0.3).

Absolute change ‐7.5% (95% CI ‐9.9 to ‐4.9), 2 point shift on a scale of 0‐100.

Relative improvement ‐10.8% (95% CI ‐5.8 to ‐14.3)

NNT 5 (95% CI 4 to 7)

Pain as assessed post‐intervention
100 point visual analog scale. Scale from: 0 to 100

Follow‐up: 3 to 14 weeks

The mean pain as assessed post‐intervention in the control groups was
7.48

The mean pain as assessed post‐intervention in the intervention groups was
0.33 standard deviations lower
(0.52 to 0.15 lower)

453
(9 studies)

⊕⊕⊝⊝
low3,4

SMD ‐0.3 (95% CI ‐0.5 to ‐0.2)

Absolute change ‐3.5% (95% CI ‐5.4 to ‐1.6), 2 point shift on a scale of 0‐100 Relative improvement ‐5.3% (95% CI ‐7.0 to ‐8.3)

NNT 6 (95% CI 4 to 14)

Mood as assessed post‐intervention
State Trait Anxiety Inventory ‐ State Scale. Scale from: 20 to 80
Follow‐up: 1 to 25 weeks

The mean mood as assessed post‐intervention in the control groups was
7.8

The mean mood as assessed post‐intervention in the intervention groups was
0.45 standard deviations lower
(0.64 to 0.26 lower)

492
(8 studies)

⊕⊕⊝⊝
low5,6

SMD ‐0.5 (95% CI ‐0.6 to ‐0.3). Absolute change ‐4.8 (95% CI ‐6.8 to ‐2.8), 3 point shift on a scale of 20‐80 Relative improvement ‐10.8% (95% CI ‐2.5 to ‐6.3)

NNT 5 (95% CI 3 to 8)

All cause attrition post‐intervention
Number of people withdrawing from the study before completing the intervention
Follow‐up: 1 to 25 weeks

Study population

RR 1.38
(1.12 to 1.69)

1687
(22 studies)

⊕⊝⊝⊝
very low7,8

Absolute risk difference 6% (95% CI 0.0 to 0.1)

Relative per cent change 38% (95% CI 12 to 69)

NNTH 18 (95% CI 10 to 55)

148 per 1000

204 per 1000
(165 to 249)

Adverse events post‐intervention
Number of people reporting an adverse event before completing the intervention
Follow‐up: 4 to 6 weeks

Study population

RR 0.38
(0.06 to 2.5)

126
(2 studies)

⊕⊕⊝⊝
low9,10

Absolute risk difference 4% (95% CI ‐0.1 to 0.0)

Relative per cent change 62% (95% CI ‐94 to 150)

Not statistically significant

51 per 1000

19 per 1000
(3 to 127)

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1Downgraded one level due to risk of bias: For some studies allocation concealment was unclear and there was a high risk of selective reporting in one study
2Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery between studies
3Downgraded one level due to risk of bias: For some studies allocation concealment was unclear and there was a high risk of selective reporting in one study
4Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery between studies
5Downgraded one level due to risk of bias: For some studies allocation concealment, blinding of participants and selective reporting were unclear
6Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery between studies
7Downgraded two levels due to risk of bias: Two studies were classified as having a high risk of outcome data and 3 studies were classified as having a high risk of selective reporting bias. Some studies were classified as having an unclear risk of sequence generation, allocation concealment, blinding of outcome assessors and outcome data.
8Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery between studies
9Downgraded one level due to risk of bias: Some studies were classified as having an unclear risk of sequence generation, allocation concealment and one study was classified as having a high risk of selective reporting
10Downgraded one level due to imprecision: There were less than 200 participants in the analysis

Open in table viewer
Summary of findings 2. Biofeedback compared to usual care for fibromyalgia

Biofeedback compared to usual care for fibromyalgia

Patient or population: patients with fibromyalgia
Settings: outpatients
Intervention: biofeedback
Comparison: usual care

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(studies)

Quality of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Usual care

Biofeedback

Functioning as assessed post‐intervention
Fibromyalgia Impact Questionnaire Revised. Scale from: 0 to 100
Follow‐up: 8 to 24 weeks

The mean functioning as assessed post‐intervention in the control groups was
17.16

The mean functioning as assessed post‐intervention in the intervention groups was
0.06 standard deviations lower
(0.44 lower to 0.33 higher)

106
(2 studies)

⊕⊝⊝⊝
very low1,2,3

SMD ‐0.1 (95% CI ‐0.4 to 0.3)

Absolute change ‐1.2% (95% CI ‐8.8 to 6.6) Relative improvement 2.2% (95% CI ‐16.3 to12.2)

Not statistically significant

Pain as assessed post‐intervention
100 point visual analog scale. Scale from: 0 to 100
Follow‐up: mean 8 weeks

The mean pain as assessed post‐intervention in the control groups was
1.3

The mean pain as assessed post‐intervention in the intervention groups was
2.6 lower
(91.29 lower to 86.09 higher)

65
(1 study)

⊕⊝⊝⊝
very low4,5

MD ‐2.6 (95% CI ‐91.3 to 86.1)

Absolute change ‐2.6% (95% CI ‐91.0 to 86.0)

Relative improvement ‐4.0% (95% CI ‐1.0 to1.0)

Not statistically significant

Mood as assessed post‐intervention
The Symptom Checklist‐90 Revised. Scale from: 0 to 90
Follow‐up: 8 to 24 weeks

The mean mood as assessed post‐intervention in the control groups was
7.3

The mean mood as assessed post‐intervention in the intervention groups was
0.13 standard deviations higher
(0.26 lower to 0.52 higher)

104
(2 studies)

⊕⊝⊝⊝
very low6,7,8

SMD 0.1 (95% CI ‐0.3 to 0.5)

Absolute change 1.9% (95% CI ‐3.7 to 7.4)

Relative improvement 3.6% (95% CI ‐7.2 to 14.5)

Not statistically significant

All cause attrition post‐intervention
Number of people withdrawing from the study before completing the intervention
Follow‐up: 4 to 24 weeks

Study population

RR 4.08
(1.43 to 11.62)

125
(3 studies)

⊕⊝⊝⊝
very low9,10,11

Absolute risk difference 20% (95% CI 0.8 to 0.3)

Relative per cent change 308% (95% CI 43 to 1062)

NNTH 7 (95% CI 3 to 41)

63 per 1000

259 per 1000
(91 to 738)

Adverse events post‐intervention ‐ not reported

See comment

See comment

Not estimable

‐

See comment

Not estimable

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1Downgraded one level due to risk of bias: Random sequence generation and allocation concealment was unclear for one study
2Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery
3Downgraded one level due to imprecision: There were less than 200 participants in this analysis
4Downgraded one level due to risk of bias: Random sequence generation and allocation concealment was unclear for one study
5Downgraded one level due to imprecision: There were less than 100 participants in the analysis
6Downgraded one level due to imprecision: Downgraded one level due to risk of bias: Random sequence generation and allocation concealment was unclear for one study
7Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery
8Downgraded one level due to imprecision: There were less than 200 participants in the analysis
9Downgraded one level due to risk of bias: Random sequence generation and allocation concealment was unclear for one study
10Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery
11Downgraded one level due to imprecision: There were less than 200 participants in the analysis

Open in table viewer
Summary of findings 3. Mindfulness compared to usual care for fibromyalgia

Mindfulness compared to usual care for fibromyalgia

Patient or population: patients with fibromyalgia
Settings: outpatients
Intervention: mindfulness
Comparison: usual care

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(studies)

Quality of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Usual care

Mindfulness

Functioning as assessed post‐intervention
Fibromyalgia Impact Questionnaire. Scale from: 0 to 100
Follow‐up: mean 8 weeks

The mean functioning as assessed post‐intervention in the control groups was
17.22

The mean functioning as assessed post‐intervention in the intervention groups was
0.26 standard deviations lower
(0.6 lower to 0.09 higher)

128
(2 studies)

⊕⊕⊝⊝
low1

SMD ‐0.3 (95% CI ‐0.6 to 0.1)

Absolute change ‐4.8% (95% CI ‐11.2 to 1.7%)

Relative improvement ‐8.5% (95% CI ‐19.3 to 3.5)

Not statistically significant

Pain as assessed post‐intervention
Visual analog scale 0 to 100. Scale from: 0 to 10.
Follow‐up: mean 8 weeks

The mean pain as assessed post‐intervention in the control groups was
0.21

The mean pain as assessed post‐intervention in the intervention groups was
0.09 standard deviations lower
(0.44 lower to 0.26 higher)

128
(2 studies)

⊕⊕⊝⊝
low2,3

SMD ‐0.09 (95% CI ‐0.4 to 0.3)

Absolute change ‐1.28% (95% CI ‐6.2 to 3.7)

Relative improvement ‐2.3% (95% CI ‐11.1 to 6.6)

Not statistically significant

Mood as assessed post‐intervention
State Trait Anxiety Inventory State Scale. Scale from: 0 to 60.
Follow‐up: mean 8 weeks

The mean mood as assessed post‐intervention in the control groups was
10.28

The mean mood as assessed post‐intervention in the intervention groups was
0.24 standard deviations lower
(0.51 lower to 0.03 higher)

218
(3 studies)

⊕⊕⊕⊝
moderate4

SMD ‐0.24 (95% CI ‐0.5 to 0.0)

Absolute change ‐3.7% (95% CI ‐7.9 to 0.5)

Relative improvement ‐8.7% (95% CI ‐18.5 to 1.2)

Not statistically significant

All cause attrition post‐intervention
Number of people withdrawing from the study before completing the intervention
Follow‐up: mean 8 weeks

Study population

RR 1.07
(0.67 to 1.72)

195
(3 studies)

⊕⊕⊕⊝
moderate5

Absolute risk difference 2% (95% CI ‐0.10 to 0.14)

Relative per cent change 98% (95% CI ‐90 to ‐86)

Not statistically significant

223 per 1000

239 per 1000
(150 to 384)

Adverse events post‐intervention ‐ not reported

See comment

See comment

Not estimable

‐

See comment

Not reported

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1Downgraded two levels due to imprecision: There were less than 200 participants in the analysis
2Downgraded one level due to risk of bias: One study was classified as having a high risk of blinding of the outcome assessors
3Downgraded one level due to imprecision: There were less than 200 participants in the analysis
4Downgraded one level due to risk of bias: One study was classified as having a high risk of blinding of the outcome assessors
5Downgraded one level due to risk of bias: One study was classified as having a high risk of blinding of the outcome assessors with one study classified as having an unclear risk of sequence generation, allocation concealment and blinding of the outcome assessors

Open in table viewer
Summary of findings 4. Movement therapies compared to usual care for fibromyalgia

Movement therapies compared to usual care for fibromyalgia

Patient or population: patients with fibromyalgia
Settings: outpatients
Intervention: movement therapies
Comparison: usual care

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(studies)

Quality of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Usual care

Movement therapies

Functioning as assessed post‐intervention
Fibromyalgia Impact Questionnaire ‐ Revised. Scale from: 0 to 100.
Follow‐up: 8 to 14 weeks

The mean functioning as assessed post‐intervention in the control groups was
13.3

The mean functioning as assessed post‐intervention in the intervention groups was
0.19 standard deviations lower
(0.53 lower to 0.15 higher)

143
(4 studies)

⊕⊝⊝⊝
very low1,2,3

SMD ‐0.19 (95% CI ‐0.5 to 0.2).

Absolute change ‐3.4% (95% CI ‐9.4 to 2.7) 2 point change on 0 to 100 scale

Relative improvement ‐6.8% (95% CI ‐19.1 to 5.5)

Not statistically significant

Pain as assessed post‐intervention
10 point visual analog scale. Scale from: 0 to 10
Follow‐up: mean 8 weeks

The mean pain as assessed post‐intervention in the control groups was
‐0.37

The mean pain as assessed post‐intervention in the intervention groups was
2.3 lower
(4.19 to 0.41 lower)

28
(1 study)

⊕⊝⊝⊝
very low4,5

MD ‐2.3 (95% CI ‐4.2 to ‐0.4)

Absolute change ‐23.0% (95% CI ‐42.0 to ‐4.0)

Relative improvement ‐3.0% (95% CI ‐6 to ‐0.6)

NNT 3 (95% CI 2 to 41)

Mood as assessed post‐intervention
Center for Epidemiologic Studies Depression Scale. Scale from: 0 to 60
Follow‐up: mean 8 weeks

The mean mood as assessed post‐intervention in the control groups was
0.41

The mean mood as assessed post‐intervention in the intervention groups was
9.84 lower
(18.51 to 1.17 lower)

29
(1 study)

⊕⊝⊝⊝
very low6,7

MD ‐9.8 (95% CI ‐18.5 to ‐1.2)

Absolute change ‐16.4% (95% CI ‐31.0 to ‐2.0)

Relative improvement ‐0.7% (95% CI ‐1.3 to ‐0.1)

NNT 3 (95% CI 2 to 34)

All cause attrition post‐intervention
Number of people withdrawing from the study before completing the intervention
Follow‐up: 8 to 24 weeks

Study population

RR 1.95
(1.13 to 3.38)

240
(5 studies)

⊕⊝⊝⊝
very low8,9

Absolute risk difference 11% (95% CI 0.0 to 0.2)

Relative per cent change 95% (95% CI 13 to 238)

NNTH 13 (95% CI 5 to 105)

106 per 1000

206 per 1000
(119 to 357)

Adverse events post‐intervention
Number of people reporting an adverse event before completing the intervention
Follow‐up: 8 to 24 weeks

Study population

RR 4.62
(0.23 to 93.72)

98
(1 study)

⊕⊝⊝⊝
very low11,12,13

Absolute risk difference 4% (95% CI ‐0.0 to 0.1)

Relative per cent change 362% (95% CI ‐77 to 9272)

Not statistically significant

0 per 1000

40 per 100010
(0 to 0)

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1Downgraded one level due to risk of bias: One study was classified as having a high risk of allocation concealment and blinding of outcome assessors
2Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery
3Downgraded one level due to imprecision: There were less than 200 participants in the analysis
4Downgraded one level due to risk of bias: One study was classified as having a high risk of allocation concealment and blinding of outcome assessors
5Downgraded one level due to imprecision: There were less than 100 participants in the analysis
6Downgraded one level due to risk of bias: One study was classified as having a high risk of allocation concealment and blinding of outcome assessors
7Downgraded one level due to imprecision: There were less than 200 participants in the analysis
8Downgraded one level due to risk of bias: One study was classified as having a high risk of allocation concealment and blinding of outcome assessors and one study had a high risk of selective reporting
9Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery
10 Absolute effect calculated from risk difference

11Downgraded one level due to risk of bias: One study was classified as having a high risk of selective reporting and unclear sequence generation and allocation concealment
12Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery
13Downgraded one level due to imprecision: There were less than 200 participants in the analysis

Open in table viewer
Summary of findings 5. Relaxation compared to usual care for fibromyalgia

Relaxation compared to usual care for fibromyalgia

Patient or population: patients with fibromyalgia
Settings: outpatients
Intervention: relaxation
Comparison: usual care

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of p
(studies)

Quality of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Usual care

Relaxation

Functioning as assessed post‐intervention
Fibromyalgia Impact Questionnaire. Scale from: 0 to 80
Follow‐up: 6 to 10 weeks

The mean functioning as assessed post‐intervention in the control groups was
3.16

The mean functioning as assessed post‐intervention in the intervention groups was
1.63 standard deviations lower
(10.14 to 6.53 lower)

67
(2 studies)

⊕⊝⊝⊝
very low1,2

MD ‐8.3 (95% CI ‐10.1 to ‐6.5).

Absolute change ‐10.4% (95% CI ‐13.0 to ‐8.0), 5 point shift on 0 to 80 scale

Relative improvement ‐20.0% (95% CI ‐0.2 to ‐0.2)

NNT 2 (95% CI 1 to 2)

Pain as assessed post‐intervention
Short Form ‐ McGill Pain Questionnaire Total Score. Scale from: 0 to 78
Follow‐up: 6 to 10 weeks

The mean pain as assessed post‐intervention in the control groups was
1.86

The mean pain as assessed post‐intervention in the intervention groups was
1.02 standard deviations lower
(1.55 to 0.5 lower)

67
(2 studies)

⊕⊝⊝⊝
very low3,4

SMD ‐1.0 (95% CI ‐1.6 to ‐0.5).

Absolute change ‐3.5% (95% CI ‐5.3 to ‐1.7), 2 point shift on a scale of 0 to 8

Relative improvement ‐9.5% (95% CI ‐14.5 to ‐4.8)

NNT 2 (95% CI 1 to 4)

Mood as assessed post‐intervention
Center for Epidemiologic Disease Depression Scale. Scale from: 0 to 60
Follow‐up: mean 6 weeks

The mean mood as assessed post‐intervention in the control groups was
‐1.9

The mean mood as assessed post‐intervention in the intervention groups was
4.44 lower
(14.46 lower to 5.58 higher)

19
(1 study)

⊕⊝⊝⊝
very low5,6

MD ‐4.4 (95% CI ‐14.5 to 5.6)

Absolute change ‐7.4% (95% CI ‐24 to 9)

Relative improvement ‐27% (95% CI ‐0.9 to ‐0.3)

Not statistically significant

All cause attrition post‐intervention
Number of people withdrawing from the study before completing the intervention
Follow‐up: mean 6 weeks

Study population

RR 4.4
(0.59 to 33.07)

21
(1 study)

⊕⊝⊝⊝
very low7,8

Absolute risk difference 31% (95% CI ‐0.0 to 0.7)

Relative per cent change 340% (95% CI ‐41 to 3207)

Not statistically significant

91 per 1000

400 per 1000
(54 to 1000)

Adverse events post‐intervention ‐ not reported

See comment

See comment

Not estimable

‐

See comment

Not estimable

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1Downgraded one level due to risk of bias: One study was classified as having an unclear risk of blinding of outcome assessors
2Downgraded one level due to imprecision: There were less than 100 participants in the analysis
3Downgraded one level due to risk of bias: One study was classified as having an unclear risk of blinding of outcome assessors
4Downgraded one level due to imprecision: There were less than 100 participants in the analysis
5Downgraded one level due to risk of bias: One study was classified as having an unclear risk of blinding of outcome assessors
6Downgraded one level due to imprecision: There were less than 100 participants in the analysis
7Downgraded one level due to risk of bias: One study was classified as having an unclear risk of blinding of outcome assessors
8Downgraded one level due to imprecision: There were less than 100 participants in the analysis

Background

Description of the condition

Fibromyalgia (FM) is a complex, chronic condition, which is characterised by widespread persistent pain, fatigue, cognitive impairment and sleep disturbances that make it difficult for people to engage in everyday activities (Arnold 2011; Bennett 2007; Wolfe 1990). Fibromyalgia has been associated with high individual and societal healthcare costs (Berger 2007; Sicras‐Mainar 2009) with many patients reporting reduced physical functioning and poor quality of life (Burckhardt 1991). The term fibromyalgia (FM) is used in this review in accordance with Cochrane convention.

Estimates suggest that FM affects between 2% to 5% of the general population (Branco 2010; Wolfe 1995). There is a higher prevalence in females (female:male ratio of 9 to 10:1) (MacFarlane 2002; Wolfe 1990; Yunus 2001), with prevalence rising to 8% in women between 55 and 64 years of age (White 1999). Emerging evidence suggests that the condition is linked to dysregulation of the central and sympathetic nervous systems (Mease 2005) that results from neurochemical imbalances leading to both an amplification of pain signals and reduced ability to inhibit the pain response (Ceko 2011; Clauw 2011).

A diagnosis of fibromyalgia is usually based on the exclusion of other potential causes of symptoms and through clinical evaluation. The American College of Rheumatology (ACR) criteria stipulate that pain must be distributed across the four quadrants of the body (that is pain above the waist, below the waist, on the left and right sides of the body) and in the axial skeleton, with tenderness in 11 or more of the 18 specific sites known as tender points during digital palpation (using 4 kg pressure) or dolorimetry (Wolfe 1990). There has been considerable debate regarding the diagnostic accuracy of FM as the ACR criteria have proven problematic, with no objective standardised test. Changes to the criteria have recently been proposed that do not require tender point examination and include a severity rating scale for fibromyalgia symptoms (Wolfe 2011). The revised criteria show potential in refining the diagnostic criteria for FM; however, as the criteria remain preliminary and further evidence of the validity, acceptance reliability and consistent implementation of the new criteria is required, this review classified FM based on the ACR criteria that have been widely implemented since 1990 (Wolfe 2010; Wolfe 2011; Wolfe 2011b).

Description of the intervention

Non‐pharmacological interventions have received increasing attention for helping patients to manage the demands of complex conditions such as FM. Indeed, it has been revealed that people with neurological conditions use complementary therapies more than other therapeutic approaches (Wells 2010). Mind‐body therapies have been defined as focusing on the interactions among the brain, mind, body and behaviour. The aim of mind‐body therapy is to enhance the capacity for self‐knowledge, self‐care and to provide tools that can improve coping, mood and quality of life (NCCAM 2005Appendix 1; Wahbeh 2008). Mind‐body interventions are considered to be a type of approach that falls under the umbrella of complimentary and alternative medicine, which also includes manipulative therapies and herbal products. The National Center for Complementary and Alternative Medicine (NCCAM 2005) describes mind‐body interventions as treatment approaches that are based on the holistic principle that mind, body and behaviour are all interconnected, incorporate strategies that are thought to improve psychological and physical well‐being, and aim to allow patients to take an active role in their treatment and to promote people's ability to cope. Mind and body interventions include a range of treatments (NCCAM 2012). Examples of mind and body therapies include biofeedback (use of technology to give audio or visual feedback on physiological processes such as heart rate to assist people in being able to gain more control over their bodies); mindfulness (a way of looking at the world in a non‐judgemental manner); movement therapies (use of physical movement to stimulate mental clarity, such as yoga, tai chi, qi‐gong); psychological therapies (use of techniques to help people become aware of their own thoughts and behaviours, such as written emotional disclosure and cognitive behaviour therapy); and relaxation strategies (techniques to help calm the mind and relax the body, such as breathing techniques, visual imagery, guided imagery, progressive muscle relaxation).

How the intervention might work

FM is a complex condition and psychological, social and lifestyle factors have all been found to play an important role in the symptom experience (Bergman 2005; Nicassio 2002; Theadom 2008). Interventions that aim to improve well‐being, self esteem, coping ability and reduce stress may therefore improve physical symptoms and quality of life for people with FM. The relevance of mind‐body interventions to FM is also supported by emerging evidence of the interactions between the central nervous, endocrine, immune, and peripheral autonomic nervous systems, suggesting "a mechanism by which mind–body medicine could influence physical health" (Vitetta 2005).

Why it is important to do this review

Symptom‐specific medication has been the primary method of treatment for FM with many patients prescribed tricyclic antidepressants (TCAs), selective serotonin uptake inhibitors (SSRIs), simple analgesics and serotonin norepinephrine reuptake inhibitors (SNRIs), which have demonstrated efficacy for reducing pain (Dworkin 2003; Hauser 2013; Moore 2009). Medications previously used in the treatment of epilepsy such as gabapentin and pregabalin are now more widely used for FM, however many people report side effects and continue to experience symptoms despite using the medication (Moore 2009; Moore 2011). Additionally, a recent review of guidelines on the management of FM (Hauser 2010) highlights the need for a multidimensional approach including a combination of non‐pharmacological and pharmacological therapies.

A review on psychological therapies for the management of chronic pain (excluding headache) in adults revealed that psychological therapies had weak effects in improving pain but that cognitive behaviour therapy and behaviour therapy improved low mood with some evidence of improvements being maintained at six months, in comparison to usual care and attention controls. Whilst this review included participants with FM, the impact of interventions may vary between different pain populations. Previous Cochrane reviews have explored the evidence for the use of exercise and resistance training for FM and found that supervised aerobic exercise and resistance training have beneficial effects on pain and physical functioning (Busch 2007; Busch 2013). A recent review has also found that cognitive behaviour therapy shows a small benefit in comparison to control in reducing pain, negative mood and disability in people with FM (Bernardy 2013).

There is evidence that mind‐body therapies are more effective in comparison to waiting list or placebo control groups on self efficacy and quality of life outcomes for FM (Hadhazy 2000). Since the publication of Hadhazy's review in 2000, a wealth of studies have since been published in this area. The present review aims to provide evidence of the efficacy of mind‐body therapies for adults with FM.

Objectives

To review the benefits and harms of mind‐body therapies in comparison to standard care and attention placebo control groups for adults with fibromyalgia (FM), post‐intervention and at three and six month follow‐up.

Methods

Criteria for considering studies for this review

Types of studies

All randomised controlled trials (RCTs) that aimed to explore the benefits or harm for people diagnosed with FM who received a mind‐body intervention in comparison to usual care or a treatment that was not thought to have therapeutic effects but was delivered by an equivalent therapist and for the same amount of time as the mind‐body therapy group (known as an attention control) were included in the review. Case studies, clinical observations and quasi‐randomised controlled trials were excluded from the review in order to minimise bias.

Types of participants

All persons 18 years of age or older with a clinical diagnosis of FM (as defined by the ACR 1990 criteria) (Wolfe 1990). If people with FM were recruited into a trial in addition to participants with other medical conditions, the study was only included if the data for people with FM were available separately.

Types of interventions

Interventions incorporating at least one type of mind‐body therapy were included. Based on the definition of mind‐body interventions proposed by the Centre for Complementary and Alternative Medicine (NCCAM 2005), six criteria were established to determine whether an intervention met the definition of a mind and body intervention for this review.

The criteria specified that the intervention must: 1) be based on the principle that the mind and body are interconnected; 2) aim to increase self knowledge; 3) aim to increase people’s ability to self‐manage their health and consequences of ill‐health; 4) actively engage and involve the participant in the intervention delivery; and 5) provide tools to improve coping and self‐management of the condition. As mind and body interventions are often incorporated with other techniques the sixth criterion, that 6) at least 80% of the total intervention delivery must include components meeting the aforementioned five principles, was added to prevent the findings from trials including only a small mind‐body component influencing the results.

Due to the wide diversity of available mind‐body therapies, interventions were categorised into broad groups to enable comparison.

  • Psychological therapies (including cognitive behaviour therapy (CBT), psychoanalytic and humanistic approaches).

  • Biofeedback (providing immediate feedback on bodily functions, such as muscle tension, to raise the patient's awareness and enable the possibility of conscious control of those functions.

  • Mindfulness meditation therapies (being aware of the present moment in a non‐judgemental and accepting way).

  • Movement therapies (e.g. yoga, tai chi, qi‐gong).

  • Relaxation based therapies (e.g. breathing techniques, visual imagery, guided imagery, progressive muscle relaxation).

Interventions delivered in all settings including community, primary care or hospital were included in the review to facilitate the generalisability of the review findings. Exercise based interventions for FM have been subject to their own Cochrane review (Busch 2007) and were therefore not included. Only movement therapies that met the definition of a mind‐body therapy were included in the review. Interventions delivered to a participant manually by a therapist (such as massage, acupuncture, physiotherapy) were not included within the review as participants are not actively engaged in the treatment, a key criterion of mind‐body interventions according to the NCCAM 2005 definition of mind‐body therapy.

Eligible comparative interventions included both usual care, which involved the treatment that people would usually receive (such as medication), or wait‐list conditions or attention control interventions involving participants receiving similar levels of contact with researchers or therapists in a similar format as the experimental intervention (such as sham therapy or peer group support).

Types of outcome measures

Major outcomes

The five major outcomes for this review were:

  • self‐reported physical functioning (ability to complete everyday tasks e.g. scores on the Fibromyalgia Impact Questionnaire (Bennett 2009));

  • self‐reported levels of pain (e.g. pain intensity numerical rating scale). A 30% or two point reduction in a 10 point numerical rating scale has been reported to be a relevant clinical outcome in evaluating trials in chronic pain (Farrar 2001);

  • mood, encompassing both anxiety and depression (e.g. Hospital Anxiety and Depression Scale (Zigmond 1983));

  • participant withdrawals;

  • adverse events (e.g. increased pain).

Data on all outcome measures assessed post‐intervention and at three and six month follow‐up were extracted for the review.

Minor outcomes

Minor outcomes were assessed post‐intervention and at three and six month follow‐up. These included:

  • fatigue (e.g. scores on the Multidimensional Assessment of Fatigue scale (Smets 1995));

  • sleep (e.g. Pissburgh Sleep Quality Index (Buysse 1989));

  • self efficacy (perceived ability to manage their overall health e.g. Chronic Pain Self‐Efficacy Scale (Anderson 1995));

  • tender point score (measured by dolorimetry or digital palpitation);

  • quality of life (e.g. Short Form Medical Outcome Study (Hays 1993)).

Search methods for identification of studies

Electronic searches

The electronic searches were conducted by the Trial Search Co‐ordinator of the Musculoskeletal Group: Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library (2013, Issue 10), MEDLINE (Ovid) (1950 to October 2013), EMBASE (Ovid) (1974 to October 2013), PsycINFO (Ovid) (1806 to October 2013), AMED (EBSCO, Allied and Complementary Medicine) (1985 to October 2013) and CINAHL (Ovid, 1982 to 2008; EBSCO, 2008 to October 2013). The search strategy is shown in Appendix 2.

Searching other resources

The reference lists of relevant articles were searched for additional relevant trials. Authors were also contacted to identify any other unpublished or published studies. The lists of identified articles were then combined and duplicate references were deleted.

Data collection and analysis

Selection of studies

Three review authors (AT, AK, SM) independently assessed all citations and identified abstracts of relevance to the review against the core inclusion criteria using a pre‐designed study selection form. Core criteria included being a RCT, inclusion of participants with FM, diagnosis based on the ACR criteria (Wolfe 1990), included participants aged over 18 years, inclusion of an intervention likely to meet the mind‐body criteria (for example articles on exercise, massage, use of treatment devices or medication or supplements were excluded) and availability of an abstract describing the trial. Full text articles were acquired for any citations meeting the core inclusion criteria for the review or where additional information was required to determine eligibility.

All full text articles were then re‐assessed against the core inclusion criteria and against the additional six inclusion criteria defining what constituted a mind‐body intervention for this review (see criteria described in type of interventions). Reasons for inclusion or exclusion were recorded in an electronic spreadsheet and on the hard copy data extraction form. The results were compared between the review authors and any disagreement resolved through discussion and consensus. Where resolution was not possible through discussion, the full review team was consulted until a consensus decision was reached. Where information was not available in the full article, trial authors were contacted for further details to clarify eligibility for the review.

Data extraction and management

Three review authors (AT, AK, SM) were involved in extracting data from the included trials, with two review authors allocated to each trial to independently extract the data. Any disputes were resolved through discussion. The data were extracted using a hard copy data extraction standard form designed specifically for this review. The data extraction form recorded information on the type of intervention (such as length of programme, therapeutic components and therapist details), setting, study procedures (such as blinding of outcome assessors and treatment allocation), details of participants and outcome measure data. Where the information needed was insufficient or incomplete, multiple attempts were made to contact the trial authors. Data were extracted from graphs if this could be accurately measured with 100% agreement by two independent researchers.The data extracted from the included trials was entered into RevMan 5.

Endpoint versus change data

Continuous data collected from self‐report questionnaires were extracted if the measure explicitly aimed to assess one of the primary or secondary outcomes and was used in its standardised form. Endpoint scores were extracted from the trial articles. Group means were used throughout the analysis (Higgins 2011).

Skewed data

Data collected using questionnaires to measure clinical and psychological outcomes often does not reveal a normal distribution. To avoid the influence of skewed data on the analyses, data were only analysed if: 1) both means and standard deviations could be derived from the data provided in the article or provided by the trial authors; and 2) if the standard deviation was less than half the mean (Altman 1996).

Assessment of risk of bias in included studies

The studies included in the review were assessed for possible risk of bias using the Cochrane Collaboration tool for assessing the risk of bias (Higgins 2011). The methods of each study were assessed independently by two review authors (AT and MC) to ascertain if the procedures applied in the study were adequate. Any disagreement identified between the review authors was resolved through discussion or through the involvement of a third review author.

These components of trials forming the risk of bias assessment included:

1) sequence generation (e.g. was the sequence generation process truly random);

2) concealment of treatment allocation;

3) blinding of the outcome assessor;

4) completeness of outcome data (e.g. participant attrition rates post‐intervention and withdrawal rates between groups);

5) selective reporting bias (e.g. were all pre‐specified outcomes reported).

Other risks of bias such as design‐specific risks were not considered in this review, which only included randomised controlled trials. No studies reported early stopping. For each component the trials were classified as low risk of bias, high risk of bias or unclear (if there was insufficient information provided in the article to make a decision). If information on the procedures used within the trial were unclear, the authors of the article were contacted to yield the necessary information. If the necessary information could not be retrieved, the potential risk was classified as unclear. To assess the direction and magnitude of the risk of bias and the possible impact this may have on the findings, sensitivity analysis was conducted. The ’blinding of participants’ was not applied in this review as it would be extremely difficult to blind people delivering the intervention or participants in accordance with other Cochrane reviews (Bernardy 2013; Williams 2012). As mind‐body interventions require the participant to actively participate in the treatment, it was considered that it was not possible to blind the participant to their treatment allocation. However, it was considered to be both feasible and desirable to randomise participants to their treatment condition, so evidence of randomisation was an important criterion for inclusion in this review. Blinding of the outcome assessors was considered as part of the risk of bias assessment of the included studies.

Measures of treatment effect

For continuous data, the weighted mean difference in endpoint scores between groups (using the same version and scoring method for outcome measure on each of the outcome domains) was calculated with the 95% confidence interval (CI). Standardised mean differences (SMD) were used for continuous outcome data measuring the same outcome variable but using different: 1) scales or subscales; 2) versions published in different languages; or 3) scored using a different approach, due to the likelihood that there would be differences in measurement between the outcome measures (Puhan 2006). For binary data, standard estimation of the risk ratio (RR) and its 95% CI were used. The P < 0.05 significance level and 95% CIs were used as the conventional significance level (Higgins 2011). All outcome data were transformed, if necessary, before analysis to ensure that high scores on each measure reflected poorer health outcomes (by subtracting the mean from the maximum score on the measure). Numbers of withdrawals between the groups post‐intervention and adverse events reported were described in terms of frequencies.

Unit of analysis issues

Cross‐over trials

Cross‐over trials were excluded from this review as there is no evidence to suggest a suitable duration of a washout period following a mind‐body intervention and it is likely that some components (such as increased knowledge) may be sustained or retained over time.

Multiple treatment arms

Where a given trial presented relevant control data for more than one group (for example if a treatment group had both a usual care and an attention placebo comparison group), each set of data were used for the respective separate analyses. If an additional treatment group that met the criteria was presented, this was included separately in the analysis (as long as the control group data were not used more than once in a given comparison).

Dealing with missing data

Missing outcome data not reported

Where possible, trial authors were contacted to request any data of potential relevance to the review that was not presented in the article. For example, requests were made when the trial authors reported that an outcome measurement was conducted at follow‐up but the data were not presented, or if means or standard deviations were not able to be derived from the information provided. For studies where standard deviations were not available for outcome data but CIs were provided, the lower CI was used in addition to the mean to calculate the variance, using the Revman calculator.

Attrition

As high rates of attrition can influence the credibility of outcome data and observation of any treatment effect, any studies with attrition rates higher than 40% (calculated as the number of participants at follow‐up divided by the number of participants randomised x 100) were not included in the analyses but were included in the attrition analyses. This decision was based on evidence that overall completion rates of between 50% to 80% are considered to be acceptable (Altman 2000; Fewtrell 2008). Four studies were found to have high (above 40%) attrition rates (Astin 2003; Brattberg 2008; Edinger 2005; Vlaeyen 1996) and were excluded from the analysis.

Assessment of heterogeneity

Statistical heterogeneity

The statistical heterogeneity of trials was assessed using the I2 statistic, calculated using RevMan 5. A cut‐off point of I2 > 50% and a P value of < 0.10 from the Mantel‐Haenszel Chi2 test were used to determine if statistically significant heterogeneity was found between the trials (Higgins 2011).

Visual inspection of the graphs

All graphs were inspected by the review team to investigate the possibility of heterogeneity. Where differences in the findings were evident, the methodology of the studies included in the analysis were reviewed for potential reasons for heterogenous findings for example clinical heterogeneity or influence of different subtypes of therapy.

Assessment of reporting biases

Funnel plots were not reported due to the low numbers of trials included in the analyses (< 10), which may prevent adequate detection of publication bias (Lau 2006). Approaches to reduce publication bias such as searching for unpublished studies and setting clear inclusion and appraisal criteria were implemented to reduce the impact of possible publication bias on the review findings, however the possibility of publication bias remained. The risk of publication bias was considered in the grading of evidence in the summary of findings tables.

Data synthesis

In the absence of statistical heterogeneity a fixed‐effect model of meta‐analysis was used for combining data. If heterogeneity was found, a sensitivity analysis was completed.

Main comparisons

The main comparisons were conducted at the post‐intervention time point in this review.

  • Psychological therapies versus usual care.

  • Psychological therapies versus attention control.

  • Biofeedback versus usual care.

  • Biofeedback versus attention control.

  • Mindfulness meditation therapies versus usual care.

  • Mindfulness meditation therapies versus attention control.

  • Movement therapies versus usual care.

  • Movement therapies versus attention control.

  • Relaxation based therapies versus usual care.

  • Relaxation based therapies versus attention control.

It was evident that some interventions applied more than one mind‐body approach within the intervention, so interventions were categorised based on the primary focus or the largest component of the intervention, or both. In one study (Astin 2003) both mindfulness and a movement therapy were applied equally within the intervention and so the data were described but not included in the analyses as the primary focus could not be determined.

Subgroup analysis and investigation of heterogeneity

Subanalyses of longer‐term outcomes of mind‐body interventions including the short‐term (one to three months post‐intervention, where data closest to three months were used) and the medium‐term (three to six months post‐intervention, where data closest to six months were used) were calculated where outcome data were available.

Heterogeniety was investigated if there was observed inconsistency in the findings resulting from the main analyses and subanalyses. If heterogeneity was observed, firstly the accuracy of data entry was checked. Secondly any outliers were specifically investigated to determine if there was a possible explanation for the different findings for example different mode, duration or type of intervention, or risk of bias. Sensitivity analyses were planned to explore the effect of heterogeneity on the findings, where possible.

Sensitivity analysis

A cut‐off point of I2 > 50% and a P value of < 0.10 from the Mantel‐Haenszel Chi2 test were used to determine if statistically significant heterogeneity was found between the trials (Higgins 2011). Sensitivity analyses were completed to explore any potential effect of the intervention content or duration, and inclusion of studies classified as having a high risk of bias.

Grading of evidence and summary of findings tables

The data are presented in the summary of findings tables (Higgins 2011), conducted using GRADEpro software. The primary outcomes of self‐reported functioning and pain were included in the summary of findings tables. Data on adverse events were used only for the groups included in the analysis. Studies were downgraded based on assessments of risk of bias, inconsistency (for example differences in treatment duration), indirectness (for example if no males were included in the analysis to enhance generalisability to the fibromyalgia population), imprecision (studies were downgraded ‐1 if there were < 200 participants in the analysis and ‐2 if < 100 participants in the analysis).

Results

Description of studies

See: 'Characteristics of included studies'; 'Characteristics of excluded studies'; 'Characteristics of studies awaiting classification'.

Results of the search

The search elicited 2083 citations, with 2009 citations excluded as the studies did not meet the inclusion criteria for this review (Figure 1).

Figure 1
Study flow diagram.

Study flow diagram.

Included studies

There were 61 distinct trials identified from 74 publications, each of which met the inclusion criteria for the review (see 'Characteristics of included studies' table). Studies were conducted across 13 countries including; USA (22 studies), Spain (11 studies), Sweden (8 studies), Germany (4 studies), Canada (3 studies), Netherlands (3 studies), Norway (3 studies), Turkey (2 studies), Brazil (1 study), France (1 study), Italy (1 study), India (1 study), UK (1 study).

Interventions

The types of mind‐body interventions encompassed by the identified articles included in this review were classified into different mind‐body therapy categories.

The overall length of treatment ranged between 1 day to 25 weeks. The average treatment duration was 17 hours. Mind‐body interventions were implemented in a range of settings, with over half of the studies (34 studies, 55.7%) conducted in a healthcare setting such as in a hospital or primary care clinic. Thirteen studies (21.3%) were conducted in a community setting such as in the person's home, with 7 studies (11.5%) conducted in a university or academic research centre. For 7 studies the type of setting where the intervention was delivered was not clear.

Most interventions (44.3%) were facilitated by a healthcare professional and 27.9% by a trained specialist in the particular therapy. Just over half (54.1%) of the studies included in this review involved only female participants. The mode of delivery of the intervention varied between trials with 54.1% of interventions delivered within a group based format, 37.7% delivered on an individual basis, and 6.6% using both a group and individual format for different elements of the intervention. For 1.6% of the interventions the mode of administration was unable to be clearly determined from the intervention description.

One study reported findings on multiple treatment arms (Thieme 2006) which included both a cognitive behaviour therapy intervention group and an operant behavioural therapy experimental group. Given there were no other studies in the review which also looked at the effectiveness of operant behavioural therapy and because both experimental groups would be in the same analysis (psychological therapies), and only one control group was available, only the cognitive behavioural group and control group (attention placebo) were included in the analyses.

Excluded studies

There were two studies that met the inclusion criteria of the review but that were still ongoing at the time of data extraction. These have be specified in the list of ongoing studies and should be included in future updates of this review (Garcia‐Campayo 2009; Miles 2010). Articles that met the inclusion criteria of the review but were excluded based on the six mind‐body intervention criteria for inclusion are outlined in the table 'Characteristics of excluded studies', with reasons for exclusion described.

Risk of bias in included studies

Allocation

All trials included in this review were described as randomised controlled trials or it was stated that a random component of participant allocation to the treatment group had been implemented. As shown in Figure 2, one study utilising a randomisation approach was classified as having a high risk of bias in accordance with the recommendations by Higgins 2011. As shown in Figure 3, a number of studies (13.0%) were classified as having an unclear risk of selection bias as insufficient details of the randomisation procedure were provided. Two studies (Connais 2009; Holmer 2004) were classified as having a high risk for allocation concealment as participants were alternately allocated to treatment groups and the researchers may have been able to foresee treatment allocation.

Figure 2
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figure 3
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Blinding

Due to the nature of delivering mind and body interventions, where it would be clear to participants which group they were in, it was not expected that the included studies would be able to be double blinded (blinding of participants). With regard to blinding of outcome assessors most studies were rated as having a low risk of bias. Six trials (Calandre 2009; Fontaine 2010; Holmer 2004; Lera 2009; Parra‐Delgado 2013; Picard 2013) were classified as having a high risk of detection bias as the outcome assessors were not blind to treatment allocation.

Incomplete outcome data

Additional data or clarification of study procedures were obtained from the authors of 36 studies included in this review (Altan 2009; Astin 2003; Babu 2007; Bakker 1995; Baumuller 2009; Brattberg 2008; Burckhardt 1994; Calandre 2009; Carson 2010; Carson 2012; Castel 2007; Castel 2009; Connais 2009; de Souza 2008; Falcao 2008; Fontaine 2010; Gillis 2006; Holmer 2004; Hsu 2010; Kayiran 2010; Lera 2009; Lynch 2012; Maddali‐Bongh 2010; Mannerkorpi 2004; Martinez‐Valero 2008; Menzies 2006; Miro 2011; Oliver 2001; Oneva‐Zafra 2010; Parra‐Delgado 2013; Scheidt 2013; Soares 2002; Stuifbergen 2010; Wang 2010; Wigers 1996; Williams 2010).

Most included studies (N = 48, 78.7%) were rated as having a low risk of attrition bias. Six studies (Bakker 1995; Burckhardt 1994; Connais 2009; Garcia 2006; Lynch 2012; Williams 2002) were rated as having a high risk of attrition bias since we were unable to extract the means and standard deviations from the information provided, precluding inclusion in the meta‐analysis. Three studies (Astin 2003; Edinger 2005; Vlaeyen 1996) were classified as having a high risk of attrition bias as they reported attrition rates over 40%. One study (Brattberg 2008) was classified as being at high risk of bias as a large number of participants (40%) did not undertake or complete the intervention sessions but completed the outcome assessments. Details of reasons for attrition were often not provided.

Selective reporting

Forty‐seven (77.0%) studies were classified as having a low risk of reporting bias as data on the outcome measures of relevance to the review were provided. Seven studies (Bojner‐Horwitz 2003; Jones 2012; Liu 2012; Luciano 2011; Soares 2002; Williams 2002; Woolfolk 2012) were classified as having a high risk of reporting bias as data were not reported on specified outcome measures. It was not always clear from the reports whether measures were planned on being used as outcome measures or that their purpose was solely to provide baseline information or to act as covariates in the analysis.

Effects of interventions

See: Summary of findings for the main comparison Psychological therapies compared to usual care for fibromyalgia; Summary of findings 2 Biofeedback compared to usual care for fibromyalgia; Summary of findings 3 Mindfulness compared to usual care for fibromyalgia; Summary of findings 4 Movement therapies compared to usual care for fibromyalgia; Summary of findings 5 Relaxation compared to usual care for fibromyalgia

The primary outcome assessment time point was post‐intervention (up to one month following intervention delivery). This would provide the greatest opportunity to determine if any treatment effect was evident as any effects were most likely to be at their strongest immediately following a mind‐body intervention. Outcomes were also assessed in the short term (within one to three months post‐intervention) and medium term (greater than three to six months post‐intervention). If an outcome assessment was made at three months this was classified as a short‐term outcome.

Comparison 1. Psychological therapies versus usual care

There were 18 studies with data available for this comparison. Data were unable to be extracted from eight trials (Burckhardt 1994; Connais 2009; Edinger 2005; Garcia 2006; Martinez‐Valero 2008; Vlaeyen 1996; Williams 2002; Woolfolk 2012) exploring psychological therapies in comparison to usual care or were unable to be incorporated due to very high attrition rates (> 40%). Two studies (Falcao 2008; Soares 2002) revealed standard deviations that were more than half the mean on a specific outcome measure, indicating that the mean was unlikely to accurately reflect the centre‐point of the distribution for that variable (Altman 1996). As skewed data is less likely to be problematic if the data set is large, the sample sizes of these two studies were considered. As both studies had sample sizes of less than 100 participants it was decided to exclude the data from the analyses for variables where the standard deviation was more than half the mean. The data from a trial were included in the analyses for variables where this was not the case.

Major outcomes
1.1 Self‐reported physical functioning

Ten trials explored psychological therapies in comparison to usual care on physical functioning outcomes (Alda 2011; Castel 2009; Castel 2012; Falcao 2008; Hamnes 2012; Luciano 2011; Maddali‐Bongh 2010; Scheidt 2013; Soares 2002; Wicksell 2013). There was an advantage for psychological therapies observed post‐intervention (N = 733, SMD ‐0.43, 95% CI ‐0.57 to ‐0.28, Analysis 1.1), at 3 month follow‐up (N = 148, SMD ‐0.54, 95% CI ‐0.87 to ‐0.21, Analysis 1.2) and at 6 month follow‐up (N = 112, MD ‐3.66, 95% CI ‐7.29 to ‐0.03, Analysis 1.3).

1.2 Self‐reported pain

Data from nine trials (Alda 2011; Castel 2009; Castel 2012; de Souza 2008; Hsu 2010; Jensen 2012; Maddali‐Bongh 2010; Soares 2002; Wigers 1996) revealed a difference between groups receiving psychological therapy and usual care that favoured psychological therapy post‐intervention (N = 453, SMD ‐0.33, 95% CI ‐0.52 to ‐0.15, Analysis 1.4). The advantage for psychological therapies over usual care was not observed at 3 month follow‐up (Falcao 2008; Castel 2012) (N = 115, MD ‐0.85, 95% CI ‐1.76 to ‐0.06, Analysis 1.5) but was observed at 6 months (N = 371, SMD ‐0.51, 95% CI ‐0.72 to ‐0.30, Analysis 1.6).

1.3 Mood

There was an advantage for psychological therapies in comparison to usual care post‐intervention, evident in eight trials (Alda 2011; Castel 2012; Falcao 2008; Hamnes 2012; Jensen 2012; Scheidt 2013; Wicksell 2013; Wigers 1996) (N = 492, SMD ‐0.45, 95% CI ‐0.64 to ‐0.26, Analysis 1.7). There was high heterogeneity between studies; removing the study by Castel 2012, which had a longer intervention delivery, reduced the I2 value to 7%. The advantage of psychological therapies post‐intervention remained (SMD ‐0.29, 95% CI ‐0.48 to ‐0.10). The advantage of psychological therapies was still evident at 3 months (N = 182, SMD ‐1.15, 95% CI ‐1.50 to ‐0.80). There was high heterogeneity observed. Removing the study by Castel 2012 reduced the I2 value to 0%. The advantage for psychology over usual care remained. At 6 months there was no advantage of psychological therapies over usual care (N = 213, SMD ‐0.17, 95% CI ‐0.44 to 0.10, Analysis 1.9).

1.4 Participant withdrawals

The RR of withdrawing from the study was statistically higher in the psychological therapy group in comparison to the control group (RR 1.38, 95% CI 1.12 to 1.69, Analysis 1.10).

1.5 Adverse events

There was no difference between the number of adverse events in the psychological therapy and control groups (RR 0.38, 95% CI 0.06 to 2.50, Analysis 1.11). Only one study reported one person experiencing a worsening of symptoms in the psychological therapy group but it was not clear if this was directly related to the intervention or not (Vlaeyen 1996).

Minor outcomes
1.6 Fatigue

Only two studies presented data on fatigue following intervention delivery (Hsu 2010; Williams 2010). There was no advantage for psychological therapies in comparison to usual care at post‐intervention (N = 82, SMD ‐0.09, 95% CI ‐0.53 to 0.34, Analysis 1.12) nor at 6 month follow‐up (N = 160, SMD ‐0.07, 95% CI ‐0.38 to 0.24, Analysis 1.13). No follow‐up data were available for the 3 month follow‐up time point. Moderate heterogeneity was observed in the findings post‐intervention; neither study included in the analysis had a high risk of bias and the heterogeneity may have been reflective of the different psychological interventions included in the analysis, with one trial implementing a self‐awareness intervention and the other a stress management intervention (Analysis 2.3).

1.7 Sleep

Data on sleep outcomes were presented by five trials (Castel 2012; Hsu 2010; Maddali‐Bongh 2010; Soares 2002; Wigers 1996). There was an advantage observed for psychological therapies in comparison to usual care for sleep post‐intervention (N = 222, SMD ‐0.52, 95% CI ‐0.80 to ‐0.25, Analysis 1.19). High heterogeneity was observed. Removing the study by Castel 2012, which was delivered over a much longer duration than the other trials, reduced the heterogeneity however the advantage of psychological therapies over usual care was no longer observed (N = 158, SMD ‐0.18, 95% CI ‐0.50 to 0.13). At 3 month follow‐up, one study revealed an advantage for psychological therapy over usual care (N = 64, MD ‐11.30, 95% CI ‐15.44 to ‐7.16). At 6 month follow‐up three studies provided data for analysis. No advantage for psychological therapies was observed (N = 224 , SMD ‐0.15, 95% CI ‐0.42 to 0.12, Analysis 1.21). High heterogeneity was observed within the data. Removing the study with a longer intervention duration (Castel 2012) reduced the heterogeneity, however there remained no advantage of psychological therapies over usual care (N = 160, SMD = 0.22, 95% CI ‐0.09 to 0.53, Analysis 2.4).

1.8 Self efficacy

There were four trials that assessed self efficacy as an outcome (Brattberg 2008; Hamnes 2012; Soares 2002; Wicksell 2013). No advantage was found for psychological therapy in comparison to usual care post‐intervention (N = 255, SMD ‐0.25, 95% CI ‐0.50 to ‐0.00, Analysis 1.14). One study (Wicksell 2013) conducted a 3 month follow‐up and found that no difference between groups was observed (N = 23, MD ‐15.10, 95% CI ‐44.95 to 14.75).

1.9 Tender points

No data were able to be extracted from trials assessing tender point count post‐intervention or at 3 month follow‐up. One trial (Hsu 2010) presented data at 6 month follow‐up. There was no advantage for psychological therapies over usual care at 6 months (N = 42, MD 0.38, 95% CI ‐0.12 to 0.88, Analysis 1.15).

1.10 Quality of life

Six trials presented data on quality of life post‐intervention (Brattberg 2008; Falcao 2008; Hsu 2010; Maddali‐Bongh 2010; Scheidt 2013; Wicksell 2013). There was no difference between groups on quality of life post‐intervention (N = 276, SMD ‐0.19, 95% CI ‐0.44 to 0.06, Analysis 1.16). Moderate heterogeneity was observed. Removing the study by Scheidt 2013 (that had a longer intervention delivery period) reduced the I2 value to 28% and an advantage of psychological therapies was observed (Analysis 1.18). At 3 months only one study provided data and the advantage for psychological therapies remained (N = 33, MD ‐15.16 95% CI ‐21.90 to ‐8.30). At 6 months the advantage for psychological therapies was no longer evident (N = 42, MD ‐2.50, 95% CI ‐7.95 to 2.95).

Comparison 2. Psychological therapies versus attention control

There were seven studies with data available for this comparison (Fontaine 2010; Gillis 2006; Langford 2009; Lera 2009; Miro 2011; Stuifbergen 2010; Thieme 2006).

Major outcomes
2.1 Self‐reported physical functioning

Seven studies reported data on functioning as an outcome (Fontaine 2010; Gillis 2006; Langford 2009; Lera 2009; Miro 2011; Stuifbergen 2010; Thieme 2006). There was no advantage of psychological therapy in comparison to an attention control post‐intervention (N = 561, SMD ‐0.10, 95% CI ‐0.27 to 0.07, Analysis 3.1) or in the short term (3 months) (N = 447, SMD 0.02, 95% CI ‐0.17 to 0.20, Analysis 3.2) or medium term (6 month follow‐up) (N = 326, SMD 0.00, 95% CI ‐0.22 to 0.23, Analysis 3.3). Moderate heterogeneity was observed within the findings for functioning post‐intervention; removing the study at high risk of bias for blinding of outcome assessors reduced the heterogeneity (Analysis 4.1). High heterogeneity was also observed at 6 month follow‐up, and a review of the forest plot indicated that the findings by Thieme 2006 were outliers. This may reflect that risk of bias was categorised as unclear as there was insufficient information in the article to determine level of risk, which may be reflective of trial quality. Removing the findings by Thieme 2006 reduced the heterogeneity; there remained no difference between groups at 6 month follow‐up (Analysis 4.2).

2.2 Self‐reported pain

An advantage was found when psychological therapy was compared to an attention control post‐intervention (N = 324, SMD ‐0.28, 95% CI ‐0.51 to ‐0.06, Analysis 3.4), however this advantage was not sustained in the short term (N = 115, SMD 0.13, 95% CI ‐0.24 to 0.50, Analysis 3.5), or medium term (N = 60, MD ‐0.34, 95% CI ‐0.89 to 0.21, Analysis 3.6). Moderate heterogeneity was observed within the data for pain outcomes post‐intervention. On review of the type of interventions incorporated within the analysis it became apparent that three trials implemented CBT and one trial implemented written emotional disclosure as an intervention. Removing the written emotional disclosure intervention (Gillis 2006) from the analysis reduced the heterogeneity. An advantage for psychological therapy remained consistent following the sensitivity analysis (Analysis 4.3).

2.3 Mood

No advantage was observed for psychological therapy post‐intervention (N = 330, SMD ‐0.12, 95% CI ‐0.33 to 0.10, Analysis 3.7) and at 3 month follow‐up (N = 115, SMD 0.24, 95% CI ‐0.13 to 0.61, Analysis 3.8). No follow‐up data were available at 6 month follow‐up to explore the medium‐term outcomes.

2.4 Participant withdrawals

The RR of withdrawing from the study for any reason was statistically higher in the control group than the psychological therapies group (RR 0.68, 95% CI 0.54 to 0.87, Analysis 3.9).

2.5 Adverse events

No studies reported data on any adverse events observed.

Minor outcomes
2.6 Fatigue

Only two studies reported data on fatigue (Fontaine 2010; Gillis 2006). No advantage was observed for psychological therapy post‐intervention (N = 153, SMD ‐0.12, 95% CI ‐0.44 to 0.20, Analysis 3.10) or at 3 month follow‐up (N = 69, MD ‐0.18, CI ‐0.73 to 0.37, Analysis 3.11). No studies reported data at 6 month follow‐up.

2.7 Sleep

No differences were observed in group outcomes for sleep when assessed post‐intervention (N = 109, SMD ‐0.12, 95% CI ‐0.50 to 0.25, Analysis 3.19) and at 3 month follow‐up (N = 69, MD 0.01, 95% CI ‐0.45 to 0.47, Analysis 3.20). No data were available for analysis at 6 month follow‐up. Moderate heterogeneity was observed in the findings for sleep outcomes. The two trials utilised different interventions, which may explain the heterogeneity, with one trial implementing CBT and one trial implementing a written emotional disclosure intervention (Analysis 4.4).

2.8 Self efficacy

Only Langford 2009 reported outcomes with regard to self efficacy post‐intervention. No advantage was observed for psychological therapy in comparison to attention control (N = 105, MD 0.48, 95% CI ‐0.27 to 1.23, Analysis 3.12). Further data were available at the 3 month follow‐up time point (Hammond 2006; Langford 2009), however no differences in group outcomes were observed (N = 151, SMD ‐0.27, 95% CI ‐0.59 to 0.05, Analysis 3.13). One trial reported outcomes at 6 month follow‐up with no differences between the groups observed (N = 36, MD 0.01, 95% CI ‐1.31 to 1.33, Analysis 3.14).

2.9 Tender points

There was no advantage observed for psychological therapies in comparison to attention control with regards to the tender point count post‐intervention (N = 150, MD ‐0.80, 95% CI ‐1.62 to 0.02, Analysis 3.15). No short or medium‐term follow‐up data were available.

2.10 Quality of life

Three trials reported data on quality of life (Langford 2009; Lera 2009; Stuifbergen 2010). No advantage was observed for psychological therapies at any endpoint: post‐intervention (N = 308, SMD ‐0.13, 95% CI ‐0.35 to ‐0.10, Analysis 3.16), 3 month follow‐up (N = 218, SMD ‐0.05, 95% CI ‐0.31 to 0.22, Analysis 3.17), 6 month follow‐up (N = 171, SMD ‐0.04, 95% CI ‐0.34 to 0.26, Analysis 3.18).

Comparison 3. Biofeedback versus usual care

There were two studies with data available for this comparison (Baumuller 2009; Van Santen 2002).

Major outcomes
3.1 Self‐reported physical functioning

Two studies provided data post‐intervention. No advantage was observed when biofeedback was compared to usual care post‐intervention (N = 106, SMD ‐0.06, 95% CI ‐0.44 to 0.33, Analysis 5.1). Only one study provided data at 3 month follow‐up (Baumuller 2009) revealing no advantage for biofeedback in the short to medium term (N = 36, MD ‐0.41, 95% CI ‐8.88 to 8.06, Analysis 5.2).

3.2 Self‐reported pain

Only one study provided data on pain post‐intervention (Van Santen 2002). It was revealed that there was no effect of biofeedback on pain in comparison to usual care (N = 65, MD ‐2.60, 95% CI ‐91.29 to 86.09, Analysis 5.3).

3.3 Mood

There was no overall effect favouring biofeedback when compared to usual care post‐intervention (N = 104, SMD 0.13, 95% CI ‐0.26 to 0.52, Analysis 5.4) and at 3 month follow‐up (N = 36, MD 4.61, 95% CI ‐0.16 to 9.38, Analysis 5.5).

3.4 Participant withdrawals

The RR of withdrawing from the study for any reason was statistically lower in the control group than the intervention group (RR 4.08, 95% CI 1.43 to 11.62, Analysis 5.6).

3.5 Adverse events

No studies reported data on any adverse events observed.

Minor outcomes
3.6 Fatigue

No data were available for analysis

3.7 Sleep

No data were available for analysis.

3.8 Self efficacy

No data were available for analysis.

3.9 Tender points

No effect in favour of biofeedback was observed post‐intervention (N = 101, MD ‐0.92, 95% CI ‐2.29 to 0.45, Analysis 5.7) or at 3 month follow‐up (N = 36, MD ‐0.09, 95% CI 0.09 to 0.62, Analysis 5.8).

3.10 Quality of life

Only one study (Baumuller 2009) presented data on quality of life as an outcome, as assessed by the German version of the Short Form 36 (SF36). There was an overall effect for the vitality domain post‐intervention (N = 36, MD ‐13.43, 95% CI ‐24.06 to ‐2.80, Analysis 5.13) but this was not sustained at 3 month follow‐up (Analysis 5.22).

There was no overall effect for biofeedback over usual care post‐intervention on seven out of the eight outcome domains (Analysis 5.9; Analysis 5.10; Analysis 5.11; Analysis 5.12; Analysis 5.14; Analysis 5.15; Analysis 5.16), nor at 3 month follow‐up (Analysis 5.17; Analysis 5.18; Analysis 5.19; Analysis 5.21Analysis 5.20; Analysis 5.23; Analysis 5.24). There was an overall effect for the vitality domain post‐intervention (N = 36, MD ‐13.43, 95% CI ‐24.06 to ‐2.80, Analysis 5.13) but this was not sustained at 3 month follow‐up (Analysis 5.22).

Comparison 4. Biofeedback versus attention control

Only one study presented data using an attention placebo control group (Babu 2007). Outcome assessments were only completed post‐intervention for this study therefore three and six month data were not available.

Major outcomes
4.1 Self‐reported physical functioning

There was a difference in Fibromyalgia Impact Questionnaire scores between biofeedback and sham attention control post‐intervention favouring biofeedback (N = 30, MD 13.60, 95% CI 1.05 to 26.13, Analysis 6.1).

4.2 Self‐reported pain

An advantage was observed for biofeedback in comparison to sham attention control post‐intervention on a 100 point visual analog scale for pain post‐intervention (N = 30, MD 2.66, 95% CI 1.21 to 5.71, Analysis 6.2).

4.3 Mood

No data were available for analysis.

4.4 Participant withdrawals

The RR of withdrawing from the study for any reason did not differ between the biofeedback group and control group (RR 3.46, 95% CI 0.44 to 27.19, Analysis 6.3).

4.5 Adverse events

No studies reported data on any adverse events observed.

Minor outcomes
4.6 Fatigue

No data were available for analysis.

4.7 Sleep

No data were available for analysis.

4.8 Self efficacy

No data were available for analysis.

4.9 Tender points

Data available for the tender point count post‐intervention revealed an advantage for biofeedback over a sham attention control group (N = 30, MD 2.93, 95% CI 0.15 to 5.71, Analysis 6.4).

4.10 Quality of life

No data were available for analysis.

Comparison 5. Mindfulness meditation therapies versus usual care

Only three studies reported data that could be extracted for this analysis (Parra‐Delgado 2013; Schmidt 2011; Sephton 2007).

Major outcomes
5.1 Self‐reported physical functioning

Two studies reported data relating to self‐reported physical functioning following a mindfulness intervention in comparison to a usual care control group (Parra‐Delgado 2013; Schmidt 2011). There were no differences between the mindfulness and the wait‐list control groups post‐intervention (N = 128, SMD ‐0.26, 95% CI ‐0.60 to 0.09, Analysis 7.1) or at short‐term follow‐up (N = 103, MD ‐0.06, 95% CI ‐1.78 to 0.66, Analysis 7.2). No statistical or clinical heterogeneity was observed in this comparison.

5.2 Self‐reported pain

Two studies reported data on pain as an outcome measure (Parra‐Delgado 2013; Schmidt 2011). There was no advantage of mindfulness in comparison to a wait‐list control group post‐intervention (N = 128, SMD ‐0.09, 95% CI ‐0.44 to 0.26, Analysis 7.3) and at short‐term follow‐up (N = 103, MD ‐0.28, 95% CI ‐2.37 to 1.81, Analysis 7.4). No statistical or clinical heterogeneity was observed in this comparison.

5.3 Mood

Three trials reported data relating to mood as an outcome measure (Parra‐Delgado 2013; Schmidt 2011; Sephton 2007). There was no difference between the mindfulness and wait‐list control groups post‐intervention (N = 218, SMD ‐0.24, 95% CI ‐0.51 to 0.03, Analysis 7.5) and at short‐term follow‐up (N = 193, SMD ‐0.21, 95% CI ‐0.50 to 0.07, Analysis 7.6). There was a moderate level of heterogeneity observed at the 3 month follow‐up for mood. On review of the included studies it became evident that one study (Parra‐Delgado 2013) was classified as having a high risk of bias for outcome assessment. Following removal of this study there remained no difference between participants receiving the mindfulness intervention and controls (Analysis 4.3).

5.4 Participant withdrawals

There was no difference in participant withdrawals between the intervention and control groups (RR 1.07, CI 0.67 to 1.72, Analysis 7.7), however as the CI included one there was uncertainty in the estimate.

5.5 Adverse events

No studies reported data on any adverse events observed.

Minor outcomes
5.6 Fatigue

No data were available for this analysis.

5.7 Sleep

Only one study (Schmidt 2011) reported on sleep as an outcome (assessed using the Pittsburgh Sleep Quality Index). There was no advantage of mindfulness in comparison to usual care post‐intervention (N = 97, MD ‐0.64, 95% CI ‐2.27 to 0.99, Analysis 7.8) or at short‐term follow‐up (N = 103, MD ‐0.36, 95% CI ‐1.91 to 1.19, Analysis 7.9).

5.8 Self efficacy

No data were available for this analysis.

5.9 Tender points

No data were available for this analysis.

5.10 Quality of life

No data were available for this analysis.

Comparison 6. Mindfulness meditation therapies versus attention control

There was no data available for this comparison.

Comparison 7. Movement therapies versus usual care

Data were available for four studies exploring a movement therapy in comparison to a usual care control group (Carson 2010; Carson 2012; Holmer 2004; Mannerkorpi 2004).

Major outcomes
7.1 Self‐reported physical functioning

Four studies providing data on functioning post‐intervention (Carson 2010; Carson 2012; Holmer 2004; Mannerkorpi 2004). There was an advantage for movement therapies over usual care (N = 124, SMD ‐0.19, 95% CI ‐0.5 to ‐0.2, Analysis 9.1). However, high statistical heterogeneity was observed. To explore reasons for this heterogeneity two sensitivity analyses were completed, one to explore the effect of removing one trial classified as having a high risk of bias (Holmer 2004) and the other to explore potential differences between movement therapy types by removing the two studies looking at qi‐gong interventions (as the other three studies looked at the effect of yoga). Despite completing these two sensitivity analyses high heterogeneity remained and no other reasons for the heterogeneity were observed. In the short and medium‐term follow‐ups, there remained an advantage for movement therapies over usual care (N = 143, MD ‐0.65, 95% CI ‐1.08 to ‐0.22, P < 0.01 at 3 months; MD ‐11.21, 95% CI ‐19.13 to ‐3.29, P < 0.01 at 6 month follow‐up).

7.2 Self‐reported pain

One study (Holmer 2004) reported data on pain post‐intervention. There was an advantage for movement therapies in comparison to usual care (N = 28, MD ‐2.30, 95% CI ‐4.19 to ‐0.41, Analysis 9.2).

7.3 Mood

Data were available for one study that assessed mood post‐intervention (Holmer 2004). An overall effect was observed for movement therapy over usual care (N = 29, MD ‐9.84, 95% CI ‐18.51 to ‐1.17, P = 0.03, Analysis 9.3).

7.4 Participant withdrawals

The RR of withdrawing from the study for any reason was statistically lower in the control group than the movement therapy group (RR 1.95, 95% CI 1.13 to 3.38, Analysis 9.4).

7.5 Adverse events

There was no difference between adverse events in the movement therapy and control group (RR 4.62, 95% CI 0.23 to 93.92, Analysis 9.5). The CIs were large, which may reflect that only one study (Lynch 2012) reported on adverse events that occurred. In this study two people reported experiencing increased pain in the intervention group.

Minor outcomes
7.6 Fatigue

The data presented on fatigue (Holmer 2004), as assessed using the Multidimensional Assessment of Fatigue scale, revealed that there was an advantage of movement therapy in comparison to usual care (N = 29, MD ‐10.80, 95% CI ‐18.57 to ‐1.17, Analysis 9.6).

7.7 Sleep

One study reported data assessing sleep post‐intervention (Holmer 2004). Sleep was assessed using the Pittsburgh Sleep Quality Index. There was a difference observed in sleep quality post‐intervention, with those receiving a movement therapy intervention revealing improved outcomes (N = 29, MD ‐4.68, 95% CI ‐8.14 to ‐1.22, Analysis 9.8).

7.8 Self efficacy

No data were available.

7.9 Tender points

Data were only available for two studies which incorporated a tender point count as an outcome measure (Carson 2010; Carson 2012). There was no advantage observed for movement therapy over usual care (N = 93, SMD 0.18, 95% CI ‐0.25 to 0.60, Analysis 9.7). High heterogeneity was observed (91%) but no reasons for the heterogeneity were apparent in terms of intervention delivery or study quality.

7.10 Quality of life

No data were available.

Comparison 8. Movement therapies versus attention control

There were three studies with data available for this comparison. One study (Wang 2010) revealed standard deviations that were more than half the mean on a specific outcome measure, indicating that the mean was unlikely to accurately reflect the centre‐point of the distribution for that variable (Altman 1996). Due to the low number of studies that could be analysed for each outcome domain, it was not possible to complete a sensitivity analyses with and without these data, as planned. As skewed data were less likely to be problematic if the data set was large, and the study had a sample size of less than 100 participants, it was decided to exclude the data from the analyses for variables where the standard deviation was more than half the mean (but the data from the trial were included in the analyses for variables where this was not the case).

Major outcomes
8.1 Self‐reported physical functioning

There was an advantage observed from three studies (Altan 2009; Calandre 2009; Wang 2010) for movement therapy over an attention control group on functioning post‐intervention (N = 191, SMD ‐0.65, 95% CI ‐0.94 to ‐0.35, Analysis 12.1) and this was sustained at 3 month follow‐up (N = 189, SMD ‐0.53, 95% CI ‐0.82 to ‐0.23, Analysis 12.2). Removing the data from one trial with inadequate blinding of outcome assessors (Calandre 2009) reduced the heterogeneity and the advantage for movement therapy remained (Analysis 13.1; Analysis 13.2).

8.2 Self‐reported pain

Three studies assessed pain as an outcome using a 10 point visual analog scale (Altan 2009; Calandre 2009; Wang 2010). An advantage was revealed for movement therapy over attention control post‐intervention (N = 172, MD ‐1.45, 95% CI ‐2.08 to ‐0.81, Analysis 12.3) that was sustained at 3 month follow‐up (N = 165, MD ‐1.19, 95% CI ‐1.87 to ‐0.52, Analysis 12.4). When one trial (Calandre 2009) showing inadequate blinding of outcome assessors was removed from the analysis statistical heterogeneity was reduced and an advantage for movement therapies remained (Analysis 13.3; Analysis 13.4).

8.3 Mood

Two studies presented data on mood as an outcome (Calandre 2009; Wang 2010). There was a difference in mood scores between movement therapy and the attention control groups favouring movement therapy post‐intervention (N = 141, SMD ‐0.49, 95% CI ‐0.83 to ‐0.15, Analysis 12.5). The group difference remained evident at 3 months (N = 140, SMD ‐0.35, 95% CI ‐0.69 to ‐0.01, Analysis 12.6). After removing one trial from the analysis due to the high risk of bias identified (due to inadequate blinding of outcome assessors) differences in group outcomes remained, favouring movement therapy (Analysis 13.5; Analysis 13.6).

8.4 Participant withdrawals

There was no difference between the rates of participant withdrawals for the movement therapy and control groups (RR 1.16, 95% CI 0.65 to 2.09, Analysis 12.7).

8.5 Adverse events

There was no difference between the number of adverse events in the movement therapy and control groups (RR 7.00, 95% CI 0.37 to 131.17, Analysis 12.8). The CIs were large, which may reflect that only one study (Calandre 2009) reported on adverse events. In this study three people experienced adverse events in the intervention group including one person who reported being hypersensitive to chlorine (as the intervention was conducted in a pool) and two participants who reported increased pain.

Minor outcomes
8.5 Fatigue

No studies reported data that could be used in this analysis. Calandre 2009 reported data on the fatigue questions of the Fibromyalgia Impact Questionnaire, but as these data was included in the total score for self‐reported physical functioning variable the data were not presented here.

8.7 Sleep

The Pittsburgh Sleep Quality Index was used to assess sleep quality in two trials (Calandre 2009; Wang 2010). There was an advantage for movement therapy in comparison to attention control for sleep post‐intervention (N = 141, MD ‐1.88, 95% CI ‐3.27 to ‐0.48, Analysis 12.15), but this was not evident at 3 month follow‐up (N = 140, MD ‐1.35, 95% CI ‐2.77 to 0.07, Analysis 12.16).

8.8 Self efficacy

One trial presented data on self efficacy for movement therapy versus an attention control group (Wang 2010). There was an advantage observed for movement therapies post‐intervention (N = 60, MD ‐45.20, 95% CI ‐46.14 to ‐44.22, Analysis 12.9) and this was sustained at 3 month follow‐up (N = 59, MD 1.20, 95% CI 0.15 to 2.25, Analysis 12.10).

8.9 Tender points

There was no advantage revealed for movement therapies in the short (N =130, MD 0.09, 95% CI ‐1.16 to 1.33, Analysis 12.11) or medium term (N = 130, MD ‐0.39, 95% CI ‐1.63 to 0.85, Analysis 12.12) across two trials (Altan 2009; Calandre 2009).

8.10 Quality of life

Two studies presented data from quality of life assessments (Altan 2009; Wang 2010). An advantage was observed for movement therapies in comparison to attention control post‐intervention (N = 109, SMD ‐0.70, 95% CI ‐1.09 to ‐0.31, Analysis 12.13) and at 3 month follow‐up (N = 108, SMD ‐0.52, 95% CI ‐0.91 to ‐0.14, Analysis 12.14).

Comparison 9. Relaxation based therapies versus usual care

There were two studies with data available for this comparison (Menzies 2006; Riedel 2012).

Major outcomes
9.1 Self‐reported physical functioning

Two trials (Menzies 2006; Riedel 2012) presented data relating to functioning post‐intervention, which was assessed using the Fibromyalgia Impact Questionnaire. There was an advantage for relaxation in comparison to usual care (N = 67, MD ‐8.34, 95% CI ‐10.14 to ‐6.53, Analysis 14.1). No follow‐up data were available to determine short and medium‐term effectiveness and no statistical heterogeneity was observed.

9.2 Self‐reported pain

Menzies 2006 and Riedel 2012 reported on pain as an outcome following a relaxation intervention in comparison to usual care. There was an advantage observed for relaxation post‐intervention (N = 67, SMD 1.02, 95% CI ‐1.55 to 0.50, Analysis 14.2). No follow‐up data were available. Statistical heterogeneity was observed between the two studies but no major methodological reasons were identified.

9.3 Mood

Riedel 2012 presented data on depression following intervention delivery as assessed by the Center for Epidemiological Studies Depression Scale. There were no differences observed between the experimental and control groups post‐intervention (N = 19, MD ‐4.44, 95% CI ‐14.46 to 5.58).

9.4 Participant withdrawals

There was no difference between participant withdrawal rates for the relaxation and control groups (RR 4.40, 95% CI 0.59 to 33.07, Analysis 14.4).

9.5 Adverse events

No studies reported data on any adverse events observed.

Minor outcomes
9.6 Fatigue

Only one study (Riedel 2012) presented data on fatigue post‐intervention. There were no differences observed between relaxation and control participants (N = 19, MD ‐0.82, 95% CI ‐2.91 to 1.27).

9.7 Sleep

The study by Riedel 2012 presented information on sleep quality as assessed by the Pittsburgh Sleep Quality Index. There were no differences observed between the experimental and control groups post‐intervention (N = 19, MD 1.03, 95% CI ‐2.23 to 4.29).

9.8 Self efficacy

There was an advantage observed for relaxation over usual care on self efficacy as assessed by two studies post‐intervention (Menzies 2006; Riedel 2012) (N = 67, SMD ‐1.54, 95% CI ‐2.13 to ‐0.95, Analysis 14.5). No follow‐up data were available.

9.9 Tender points

No data were available for this analysis.

9.10 Quality of life

No data were available for this analysis.

Comparison 10. Relaxation based therapies versus attention control

There was only one study with data available for this comparison (Fors 2000).

Major outcomes
10.1 Self‐reported physical functioning

No data were available for this analysis.

10.2 Self‐reported pain

One trial presented data on pain, which was assessed by a 100 point visual analog scale (Fors 2000). There was an advantage identified for the relaxation group in comparison to an education control group (N = 39, MD ‐23.17, 95% CI ‐36.73 to ‐9.61, Analysis 15.1). No follow‐up assessment data were available.

10.3 Mood

The data presented by Fors 2000, which assessed mood using a 100 point visual analog scale for anxiety, found an improvement with the use of relaxation based therapies in comparison to an education control (N = 39, MD ‐32.10, 95% CI ‐46.35 to ‐17.85, Analysis 15.2). No follow‐up data were available.

10.4 Participant withdrawals

As no participants were reported to have withdrawn from the Fors 2000 study estimates could not be derived for this outcome.

10.5 Adverse events

No studies reported data on any adverse events observed.

Minor outcomes
10.6 Fatigue

No data were available for this analysis.

10.7 Sleep

No data were available for this analysis.

10.8 Self efficacy

No data were available for this analysis.

10.9 Tender points

No data were available for this analysis.

10.10 Quality of life

No data were available for this analysis.

Discussion

Summary of main results

Moderate, low or very low quality of evidence from 61 trials (including a total of 4234 participants) was analysed. Mind‐body interventions were analysed based on the type of intervention including biofeedback, movement therapies, psychological therapies, relaxation based therapies and mindfulness.

There was no advantage observed for biofeedback in comparison to usual care controls and no studies reported any adverse events, however the quality of the evidence was very low so we cannot be certain if there is any effect or not. There was also no advantage observed for mindfulness in comparison with usual care, There was no difference in withdrawals between groups. Adverse events were not reported.

There was no advantage observed for mindfulness in comparison to usual care for physical functioning, pain or mood post‐intervention. However the quality of the evidence was very low. There was uncertainty as to whether there were statistical differences in withdrawals between the two groups. No studies reported any adverse events.

There were improved outcomes for movement therapies over usual care and attention controls for physical functioning, pain and mood post‐intervention. However the risk of increased pain reported by one trial suggests caution is needed in interpreting the results and we cannot be certain of any effect due to the very low quality of evidence.

Results for the main analyses on the use of psychological therapies in comparison to usual care controls revealed low quality evidence from 10 trials (733 participants) suggesting that psychological therapies provide a small improvement in physical functioning, pain and mood at the end of treatment. Low quality evidence revealed that improvements in physical functioning and mood were sustained at three month follow‐up and at six month follow up for physical functioning. There was very low quality evidence for the secondary outcomes resulting from psychological therapies.

Relaxation based therapies showed an advantage over usual care for physical functioning and pain outcomes post‐intervention; for pain however the quality of the evidence was very low. No differences in withdrawals or adverse events were reported for relaxation based therapies.

The small number of studies that provided short to medium‐term (three to six month follow‐up) data in this review is a concern and limited evidence was available to determine the short to medium‐term impact of mind‐body interventions for adults with FM.

Overall completeness and applicability of evidence

Overall completeness

In the search strategy we included efforts to identify and include unpublished data to reduce the possible impact of publication bias. Whilst some unpublished data have been included in the review, we cannot rule out the possibility that negative study results may not have been published or identified for inclusion by this review.

The applicability of the evidence included in this review is considered to be strong for a number of reasons. Firstly, the review includes a number of mind‐body interventions delivered across a range of contexts including hospital settings, primary care centres and in the community. Secondly, many samples included both male as well as female adults with FM, which is important as although FM predominantly affects women it can affect men; however, due to the low numbers no study performed a subgroup analysis for male participants. Thirdly, many trials included participants with additional co‐morbidities (not serious or life‐threatening), such as depression, which commonly occur in adults with FM.

Due to the diversity of symptoms experienced by people with FM, this review analysed the evidence on a wide range of outcomes including the major outcomes of self‐reported physical functioning, pain, withdrawals and adverse events, and minor outcomes such as fatigue and self efficacy. Outcomes such as walk time, self confidence, use of medication and healthcare visits may be important but we were not able to incorporate them within the scope of this review as there is a limit to the number of outcomes that can reliably be studied within the context of a Cochrane review.

This review aimed to quantitatively summarise the effects of mind‐body interventions for FM. Whilst this review was targeted at one specific population (adults with FM), the findings may have relevance to other populations where complex symptomology presents. Within the context of current practice, many chronic pain programmes already implement components of mind‐body therapies, such as the use of guided imagery. Mind‐body interventions that require specialist expertise to deliver, such as tai chi or cognitive behaviour therapy (CBT), may be more challenging to incorporate into practice without additional resources.

Quality of the evidence

The evidence presented in this review was extracted from trials published in academic journals and was requested from the trial authors. The overall quality of the evidence was moderate, low or very low (see summary of findings tables). Trial quality was reduced by unclear details or high risk of allocation concealment, non‐blinding of outcome assessors or risk of bias from selective reporting; however, sensitivity analyses revealed that the findings were not influenced by the removal of studies with a high risk of bias from the analyses. The sample size of the included trials was often small and even in the meta‐analyses participant numbers were as low as N = 19, increasing to N = 733. Few studies reported information on any adverse events arising from the interventions. It is important to record any instances where there is a decrease in health or well‐being, such as increased levels of pain or fatigue, that may have been exacerbated by the intervention.

Potential biases in the review process

Despite efforts to reduce the impact of publication bias in the review, the possibility remains that some studies (with positive or negative findings) may not have been identified by the search. Where further clarification of study methodology could not be obtained from the authors, there is the possibility that the risk of bias for the studies may have been overestimated. Whilst contacting authors for additional information assisted in the accuracy of the information reported in most cases, this may have introduced a 'response bias' into the risk of bias assessment. Some values needed to be imputed for missing data (such as variability estimates when the lower confidence interval was used) using the Revman calculator. Data were unable to be extracted accurately for several trials that presented their findings graphically, thus limiting the generalisability of the findings. The small number of trials included in some analyses further reduces the robustness of these findings.

In many cases determining the assessment time point was difficult, as it was not always clear if the timeframe was post‐randomisation or intervention delivery and the window within which assessments were completed was rarely documented.

Agreements and disagreements with other studies or reviews

One previous systematic review of mind‐body therapies for adults with FM was found (Hadhazy 2000). The review included 13 trials of 802 participants and searched the literature until 1999. The findings of the review support the findings of this current review suggesting that there is some limited evidence of the effectiveness of mind‐body therapies in comparison to placebo or attention control for self‐reported pain and physical functioning.

Reviews examining different types of mind‐body interventions include a Cochrane review of cognitive behaviour therapy (CBT) for adults and children with FM, conducted until August 2013 (Bernardy 2013). The results from this current review were more negative than for the review conducted by Bernardy 2013, which revealed small effects for pain, mood and disability that were sustained at follow‐up. By including CBTs in comparisons of treatment effect with other psychological therapies may have obscured unique effects of different types of interventions. The inclusion of children in the other review may also have increased the observed treatment effect.

Another systematic review explored the effectiveness of qi‐gong interventions for adults and children with FM (Chan 2012). The review completed a search of studies until February 2011 and revealed that it was too early to draw conclusions as to the efficacy of qi‐gong and that further robust RCTs were warranted. The current review supports these findings that there is currently insufficient evidence on movement therapies to draw any firm conclusions.

The authors of a Cochrane review on psychological therapies for chronic pain (excluding headache) searched the literature until September 2011 (Williams 2012). This review revealed that people receiving psychological therapies experienced small improvements in functioning, pain and mood when compared to usual care but not when compared to attention control participants. These findings are comparable to the findings in the current review.

Other systematic reviews in this field have been published outside of The Cochrane Library. Findings revealed some inconsistencies which may be due to differences in the inclusion criteria set and outcomes domains explored. In a review of mindfulness based relaxation studies for FM (Lauche 2013) the authors revealed that mindfulness group participants showed reductions in pain and improved quality of life post‐intervention in comparison to controls. In contrast, this review found no difference between the experimental and control groups on any of the major outcomes including pain and quality of life. This disparity in the findings may reflect the inclusion of non‐randomised trials in the review by Lauche 2013. A review of guided imagery for FM that was conducted by Bernardy 2011 revealed similar findings to the relaxation analysis conducted as part of this Cochrane review, where both studies included in the relaxation analysis used a guided imagery intervention. Both reviews revealed that participants receiving guided imagery showed reductions in pain but not quality of life post‐intervention. This Cochrane review also identified that participants receiving biofeedback demonstrated improved physical functioning post‐intervention compared to controls. In a review of biofeedback for people with FM conducted by Glombiewski 2013, it was revealed that participants receiving biofeedback reported reductions in pain post‐intervention in comparison to controls. This Cochrane review was unable to detect a difference between participants receiving biofeedback and controls. This may be due to the inclusion of trials using additional interventions such as cognitive strategies or exercise. In contrast, for inclusion in this review biofeedback needed to be the primary focus of the intervention (constituting at least 80% of the intervention) to ensure any effects detected were due to the biofeedback rather than inclusion of other treatments. A consistent finding between these three reviews published outside of Cochrane and the current review is that the quality of the available evidence in this area is poor.

Study flow diagram.
Figures and Tables -
Figure 1

Study flow diagram.

Navigate to figure in ReviewOpen in new tab

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Figures and Tables -
Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Navigate to figure in ReviewOpen in new tab

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Figures and Tables -
Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 1 Functioning as assessed post‐intervention.
Figures and Tables -
Analysis 1.1

Comparison 1 Psychological therapies versus usual care, Outcome 1 Functioning as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 2 Functioning as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 1.2

Comparison 1 Psychological therapies versus usual care, Outcome 2 Functioning as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 3 Functioning as assessed at 6 month follow‐up.
Figures and Tables -
Analysis 1.3

Comparison 1 Psychological therapies versus usual care, Outcome 3 Functioning as assessed at 6 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 4 Pain as assessed post‐intervention.
Figures and Tables -
Analysis 1.4

Comparison 1 Psychological therapies versus usual care, Outcome 4 Pain as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 5 Pain as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 1.5

Comparison 1 Psychological therapies versus usual care, Outcome 5 Pain as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 6 Pain as assessed at 6 month follow‐up.
Figures and Tables -
Analysis 1.6

Comparison 1 Psychological therapies versus usual care, Outcome 6 Pain as assessed at 6 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 7 Mood as assessed post‐intervention.
Figures and Tables -
Analysis 1.7

Comparison 1 Psychological therapies versus usual care, Outcome 7 Mood as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 8 Mood as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 1.8

Comparison 1 Psychological therapies versus usual care, Outcome 8 Mood as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 9 Mood as assessed at 6 month follow‐up.
Figures and Tables -
Analysis 1.9

Comparison 1 Psychological therapies versus usual care, Outcome 9 Mood as assessed at 6 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 10 All cause attrition post‐intervention.
Figures and Tables -
Analysis 1.10

Comparison 1 Psychological therapies versus usual care, Outcome 10 All cause attrition post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 11 Adverse events post‐intervention.
Figures and Tables -
Analysis 1.11

Comparison 1 Psychological therapies versus usual care, Outcome 11 Adverse events post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 12 Fatigue as assessed post‐intervention.
Figures and Tables -
Analysis 1.12

Comparison 1 Psychological therapies versus usual care, Outcome 12 Fatigue as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 13 Fatigue as assessed at 6 months post‐intervention.
Figures and Tables -
Analysis 1.13

Comparison 1 Psychological therapies versus usual care, Outcome 13 Fatigue as assessed at 6 months post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 14 Self‐efficacy as assessed post‐intervention.
Figures and Tables -
Analysis 1.14

Comparison 1 Psychological therapies versus usual care, Outcome 14 Self‐efficacy as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 15 Tender point count as assessed at 6 month follow‐up.
Figures and Tables -
Analysis 1.15

Comparison 1 Psychological therapies versus usual care, Outcome 15 Tender point count as assessed at 6 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 16 Quality of life as assessed post‐intervention.
Figures and Tables -
Analysis 1.16

Comparison 1 Psychological therapies versus usual care, Outcome 16 Quality of life as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 17 Quality of life as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 1.17

Comparison 1 Psychological therapies versus usual care, Outcome 17 Quality of life as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 18 Quality of life as assessed at 6 month follow‐up.
Figures and Tables -
Analysis 1.18

Comparison 1 Psychological therapies versus usual care, Outcome 18 Quality of life as assessed at 6 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 19 Sleep as assessed post‐intervention.
Figures and Tables -
Analysis 1.19

Comparison 1 Psychological therapies versus usual care, Outcome 19 Sleep as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 20 Sleep as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 1.20

Comparison 1 Psychological therapies versus usual care, Outcome 20 Sleep as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 21 Sleep as assessed at 6 month follow‐up.
Figures and Tables -
Analysis 1.21

Comparison 1 Psychological therapies versus usual care, Outcome 21 Sleep as assessed at 6 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 1 Psychological therapies versus usual care, Outcome 22 Self‐efficacy as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 1.22

Comparison 1 Psychological therapies versus usual care, Outcome 22 Self‐efficacy as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 2 Psychological therapies versus usual care sensitivity analyses, Outcome 1 Mood as assessed post‐intervention.
Figures and Tables -
Analysis 2.1

Comparison 2 Psychological therapies versus usual care sensitivity analyses, Outcome 1 Mood as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 2 Psychological therapies versus usual care sensitivity analyses, Outcome 2 Mood as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 2.2

Comparison 2 Psychological therapies versus usual care sensitivity analyses, Outcome 2 Mood as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 2 Psychological therapies versus usual care sensitivity analyses, Outcome 3 Fatigue as assessed post‐intervention.
Figures and Tables -
Analysis 2.3

Comparison 2 Psychological therapies versus usual care sensitivity analyses, Outcome 3 Fatigue as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 2 Psychological therapies versus usual care sensitivity analyses, Outcome 4 Sleep as assessed post‐intervention.
Figures and Tables -
Analysis 2.4

Comparison 2 Psychological therapies versus usual care sensitivity analyses, Outcome 4 Sleep as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 2 Psychological therapies versus usual care sensitivity analyses, Outcome 5 Sleep as assessed at 6 month follow‐up.
Figures and Tables -
Analysis 2.5

Comparison 2 Psychological therapies versus usual care sensitivity analyses, Outcome 5 Sleep as assessed at 6 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 1 Functioning as assessed post‐intervention.
Figures and Tables -
Analysis 3.1

Comparison 3 Psychological therapies versus attention control, Outcome 1 Functioning as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 2 Functioning as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 3.2

Comparison 3 Psychological therapies versus attention control, Outcome 2 Functioning as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 3 Functioning as assessed at 6 month follow‐up.
Figures and Tables -
Analysis 3.3

Comparison 3 Psychological therapies versus attention control, Outcome 3 Functioning as assessed at 6 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 4 Pain as assessed post‐intervention.
Figures and Tables -
Analysis 3.4

Comparison 3 Psychological therapies versus attention control, Outcome 4 Pain as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 5 Pain as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 3.5

Comparison 3 Psychological therapies versus attention control, Outcome 5 Pain as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 6 Pain as assessed at 6 month follow‐up.
Figures and Tables -
Analysis 3.6

Comparison 3 Psychological therapies versus attention control, Outcome 6 Pain as assessed at 6 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 7 Mood as assessed post‐intervention.
Figures and Tables -
Analysis 3.7

Comparison 3 Psychological therapies versus attention control, Outcome 7 Mood as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 8 Mood as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 3.8

Comparison 3 Psychological therapies versus attention control, Outcome 8 Mood as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 9 All cause attrition post‐intervention.
Figures and Tables -
Analysis 3.9

Comparison 3 Psychological therapies versus attention control, Outcome 9 All cause attrition post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 10 Fatigue as assessed post‐intervention.
Figures and Tables -
Analysis 3.10

Comparison 3 Psychological therapies versus attention control, Outcome 10 Fatigue as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 11 Fatigue as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 3.11

Comparison 3 Psychological therapies versus attention control, Outcome 11 Fatigue as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 12 Self‐efficacy as assessed post‐intervention.
Figures and Tables -
Analysis 3.12

Comparison 3 Psychological therapies versus attention control, Outcome 12 Self‐efficacy as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 13 Self efficacy as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 3.13

Comparison 3 Psychological therapies versus attention control, Outcome 13 Self efficacy as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 14 Self‐efficacy as assessed at 6 month follow‐up.
Figures and Tables -
Analysis 3.14

Comparison 3 Psychological therapies versus attention control, Outcome 14 Self‐efficacy as assessed at 6 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 15 Tender point score as assessed post‐intervention.
Figures and Tables -
Analysis 3.15

Comparison 3 Psychological therapies versus attention control, Outcome 15 Tender point score as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 16 Quality of life as assessed post‐intervention.
Figures and Tables -
Analysis 3.16

Comparison 3 Psychological therapies versus attention control, Outcome 16 Quality of life as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 17 Quality of life as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 3.17

Comparison 3 Psychological therapies versus attention control, Outcome 17 Quality of life as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 18 Quality of life as assessed at 6 month follow‐up.
Figures and Tables -
Analysis 3.18

Comparison 3 Psychological therapies versus attention control, Outcome 18 Quality of life as assessed at 6 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 19 Sleep as assessed post‐intervention.
Figures and Tables -
Analysis 3.19

Comparison 3 Psychological therapies versus attention control, Outcome 19 Sleep as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 3 Psychological therapies versus attention control, Outcome 20 Sleep as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 3.20

Comparison 3 Psychological therapies versus attention control, Outcome 20 Sleep as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 4 Psychological therapies versus attention control sensitivity analyses, Outcome 1 Functioning as assessed post‐intervention.
Figures and Tables -
Analysis 4.1

Comparison 4 Psychological therapies versus attention control sensitivity analyses, Outcome 1 Functioning as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 4 Psychological therapies versus attention control sensitivity analyses, Outcome 2 Functioning as assessed at 6 month follow‐up.
Figures and Tables -
Analysis 4.2

Comparison 4 Psychological therapies versus attention control sensitivity analyses, Outcome 2 Functioning as assessed at 6 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 4 Psychological therapies versus attention control sensitivity analyses, Outcome 3 Pain as assessed post‐intervention.
Figures and Tables -
Analysis 4.3

Comparison 4 Psychological therapies versus attention control sensitivity analyses, Outcome 3 Pain as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 4 Psychological therapies versus attention control sensitivity analyses, Outcome 4 Sleep as assessed post‐intervention.
Figures and Tables -
Analysis 4.4

Comparison 4 Psychological therapies versus attention control sensitivity analyses, Outcome 4 Sleep as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 1 Functioning as assessed post‐intervention.
Figures and Tables -
Analysis 5.1

Comparison 5 Biofeedback versus usual care, Outcome 1 Functioning as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 2 Functioning as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 5.2

Comparison 5 Biofeedback versus usual care, Outcome 2 Functioning as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 3 Pain as assessed post‐intervention.
Figures and Tables -
Analysis 5.3

Comparison 5 Biofeedback versus usual care, Outcome 3 Pain as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 4 Mood as assessed post‐intervention.
Figures and Tables -
Analysis 5.4

Comparison 5 Biofeedback versus usual care, Outcome 4 Mood as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 5 Mood as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 5.5

Comparison 5 Biofeedback versus usual care, Outcome 5 Mood as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 6 All cause attrition post‐intervention.
Figures and Tables -
Analysis 5.6

Comparison 5 Biofeedback versus usual care, Outcome 6 All cause attrition post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 7 Tender point score as assessed post‐intervention.
Figures and Tables -
Analysis 5.7

Comparison 5 Biofeedback versus usual care, Outcome 7 Tender point score as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 8 Tender point score as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 5.8

Comparison 5 Biofeedback versus usual care, Outcome 8 Tender point score as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 9 Quality of life (Physical functioning) as assessed post‐intervention.
Figures and Tables -
Analysis 5.9

Comparison 5 Biofeedback versus usual care, Outcome 9 Quality of life (Physical functioning) as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 10 Quality of life (Role‐Physical) as assessed post‐intervention.
Figures and Tables -
Analysis 5.10

Comparison 5 Biofeedback versus usual care, Outcome 10 Quality of life (Role‐Physical) as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 11 Quality of life (Bodily Pain) as assessed post‐intervention.
Figures and Tables -
Analysis 5.11

Comparison 5 Biofeedback versus usual care, Outcome 11 Quality of life (Bodily Pain) as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 12 Quality of life (General Health) as assessed post‐intervention.
Figures and Tables -
Analysis 5.12

Comparison 5 Biofeedback versus usual care, Outcome 12 Quality of life (General Health) as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 13 Quality of life (Vitality) as assessed post‐intervention.
Figures and Tables -
Analysis 5.13

Comparison 5 Biofeedback versus usual care, Outcome 13 Quality of life (Vitality) as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 14 Quality of life (Social Functioning) as assessed post‐intervention.
Figures and Tables -
Analysis 5.14

Comparison 5 Biofeedback versus usual care, Outcome 14 Quality of life (Social Functioning) as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 15 Quality of life (Role‐Emotional) as assessed post‐intervention.
Figures and Tables -
Analysis 5.15

Comparison 5 Biofeedback versus usual care, Outcome 15 Quality of life (Role‐Emotional) as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 16 Quality of life (Mental Health) as assessed post‐intervention.
Figures and Tables -
Analysis 5.16

Comparison 5 Biofeedback versus usual care, Outcome 16 Quality of life (Mental Health) as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 17 Quality of life (Physical functioning) as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 5.17

Comparison 5 Biofeedback versus usual care, Outcome 17 Quality of life (Physical functioning) as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 18 Quality of life (Role‐Physical) as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 5.18

Comparison 5 Biofeedback versus usual care, Outcome 18 Quality of life (Role‐Physical) as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 19 Quality of life (Bodily Pain) as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 5.19

Comparison 5 Biofeedback versus usual care, Outcome 19 Quality of life (Bodily Pain) as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 20 Quality of life (Social Functioning) as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 5.20

Comparison 5 Biofeedback versus usual care, Outcome 20 Quality of life (Social Functioning) as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 21 Quality of life (General Health) as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 5.21

Comparison 5 Biofeedback versus usual care, Outcome 21 Quality of life (General Health) as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 22 Quality of life (Vitality) as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 5.22

Comparison 5 Biofeedback versus usual care, Outcome 22 Quality of life (Vitality) as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 23 Quality of life (Role‐Emotional) as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 5.23

Comparison 5 Biofeedback versus usual care, Outcome 23 Quality of life (Role‐Emotional) as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 5 Biofeedback versus usual care, Outcome 24 Quality of life (Mental Health) as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 5.24

Comparison 5 Biofeedback versus usual care, Outcome 24 Quality of life (Mental Health) as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 6 Biofeedback versus attention control, Outcome 1 Functioning as assessed post‐intervention.
Figures and Tables -
Analysis 6.1

Comparison 6 Biofeedback versus attention control, Outcome 1 Functioning as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 6 Biofeedback versus attention control, Outcome 2 Pain as assessed post‐intervention.
Figures and Tables -
Analysis 6.2

Comparison 6 Biofeedback versus attention control, Outcome 2 Pain as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 6 Biofeedback versus attention control, Outcome 3 All cause attrition post‐intervention.
Figures and Tables -
Analysis 6.3

Comparison 6 Biofeedback versus attention control, Outcome 3 All cause attrition post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 6 Biofeedback versus attention control, Outcome 4 Tender point score as assessed post‐intervention.
Figures and Tables -
Analysis 6.4

Comparison 6 Biofeedback versus attention control, Outcome 4 Tender point score as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 7 Mindfulness versus usual care, Outcome 1 Functioning as assessed post‐intervention.
Figures and Tables -
Analysis 7.1

Comparison 7 Mindfulness versus usual care, Outcome 1 Functioning as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 7 Mindfulness versus usual care, Outcome 2 Functioning assessed at 3 month follow‐up.
Figures and Tables -
Analysis 7.2

Comparison 7 Mindfulness versus usual care, Outcome 2 Functioning assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 7 Mindfulness versus usual care, Outcome 3 Pain as assessed post‐intervention.
Figures and Tables -
Analysis 7.3

Comparison 7 Mindfulness versus usual care, Outcome 3 Pain as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 7 Mindfulness versus usual care, Outcome 4 Pain as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 7.4

Comparison 7 Mindfulness versus usual care, Outcome 4 Pain as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 7 Mindfulness versus usual care, Outcome 5 Mood as assessed post‐intervention.
Figures and Tables -
Analysis 7.5

Comparison 7 Mindfulness versus usual care, Outcome 5 Mood as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 7 Mindfulness versus usual care, Outcome 6 Mood as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 7.6

Comparison 7 Mindfulness versus usual care, Outcome 6 Mood as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 7 Mindfulness versus usual care, Outcome 7 All cause attrition post‐intervention.
Figures and Tables -
Analysis 7.7

Comparison 7 Mindfulness versus usual care, Outcome 7 All cause attrition post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 7 Mindfulness versus usual care, Outcome 8 Sleep as assessed post‐intervention.
Figures and Tables -
Analysis 7.8

Comparison 7 Mindfulness versus usual care, Outcome 8 Sleep as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 7 Mindfulness versus usual care, Outcome 9 Sleep as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 7.9

Comparison 7 Mindfulness versus usual care, Outcome 9 Sleep as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 8 Mindfulness versus usual care ‐ sensitivity analyses, Outcome 1 Mood as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 8.1

Comparison 8 Mindfulness versus usual care ‐ sensitivity analyses, Outcome 1 Mood as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 9 Movement therapies versus usual care, Outcome 1 Functioning as assessed post‐intervention.
Figures and Tables -
Analysis 9.1

Comparison 9 Movement therapies versus usual care, Outcome 1 Functioning as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 9 Movement therapies versus usual care, Outcome 2 Pain as assessed post‐intervention.
Figures and Tables -
Analysis 9.2

Comparison 9 Movement therapies versus usual care, Outcome 2 Pain as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 9 Movement therapies versus usual care, Outcome 3 Mood as assessed post‐intervention.
Figures and Tables -
Analysis 9.3

Comparison 9 Movement therapies versus usual care, Outcome 3 Mood as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 9 Movement therapies versus usual care, Outcome 4 All cause attrition post‐intervention.
Figures and Tables -
Analysis 9.4

Comparison 9 Movement therapies versus usual care, Outcome 4 All cause attrition post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 9 Movement therapies versus usual care, Outcome 5 Adverse events post‐intervention.
Figures and Tables -
Analysis 9.5

Comparison 9 Movement therapies versus usual care, Outcome 5 Adverse events post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 9 Movement therapies versus usual care, Outcome 6 Fatigue as assessed post‐intervention.
Figures and Tables -
Analysis 9.6

Comparison 9 Movement therapies versus usual care, Outcome 6 Fatigue as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 9 Movement therapies versus usual care, Outcome 7 Tender point count as assessed post‐intervention.
Figures and Tables -
Analysis 9.7

Comparison 9 Movement therapies versus usual care, Outcome 7 Tender point count as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 9 Movement therapies versus usual care, Outcome 8 Sleep as assessed post‐intervention.
Figures and Tables -
Analysis 9.8

Comparison 9 Movement therapies versus usual care, Outcome 8 Sleep as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 10 Movement therapies versus usual care ‐ sensitivity analyses intervention type, Outcome 1 Functioning as assessed post‐intervention.
Figures and Tables -
Analysis 10.1

Comparison 10 Movement therapies versus usual care ‐ sensitivity analyses intervention type, Outcome 1 Functioning as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 11 Movement therapies versus usual care ‐ sensitivity analyses quality, Outcome 1 Functioning as assessed post‐intervention.
Figures and Tables -
Analysis 11.1

Comparison 11 Movement therapies versus usual care ‐ sensitivity analyses quality, Outcome 1 Functioning as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 1 Functioning as assessed post‐intervention.
Figures and Tables -
Analysis 12.1

Comparison 12 Movement therapies versus attention control, Outcome 1 Functioning as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 2 Functioning as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 12.2

Comparison 12 Movement therapies versus attention control, Outcome 2 Functioning as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 3 Pain as assessed by a 10‐point VAS scale post‐intervention.
Figures and Tables -
Analysis 12.3

Comparison 12 Movement therapies versus attention control, Outcome 3 Pain as assessed by a 10‐point VAS scale post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 4 Pain as assessed by a 10‐point VAS scale at 3 month follow‐up.
Figures and Tables -
Analysis 12.4

Comparison 12 Movement therapies versus attention control, Outcome 4 Pain as assessed by a 10‐point VAS scale at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 5 Mood as assessed post‐intervention.
Figures and Tables -
Analysis 12.5

Comparison 12 Movement therapies versus attention control, Outcome 5 Mood as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 6 Mood as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 12.6

Comparison 12 Movement therapies versus attention control, Outcome 6 Mood as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 7 All cause attrition post‐intervention.
Figures and Tables -
Analysis 12.7

Comparison 12 Movement therapies versus attention control, Outcome 7 All cause attrition post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 8 Adverse events post‐intervention.
Figures and Tables -
Analysis 12.8

Comparison 12 Movement therapies versus attention control, Outcome 8 Adverse events post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 9 Self‐efficacy as assessed post‐intervention.
Figures and Tables -
Analysis 12.9

Comparison 12 Movement therapies versus attention control, Outcome 9 Self‐efficacy as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 10 Self‐efficacy as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 12.10

Comparison 12 Movement therapies versus attention control, Outcome 10 Self‐efficacy as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 11 Tender points as assessed post‐intervention.
Figures and Tables -
Analysis 12.11

Comparison 12 Movement therapies versus attention control, Outcome 11 Tender points as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 12 Tender points as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 12.12

Comparison 12 Movement therapies versus attention control, Outcome 12 Tender points as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 13 Quality of life as assessed post‐intervention.
Figures and Tables -
Analysis 12.13

Comparison 12 Movement therapies versus attention control, Outcome 13 Quality of life as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 14 Quality of life as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 12.14

Comparison 12 Movement therapies versus attention control, Outcome 14 Quality of life as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 15 Sleep quality as assessed by the Pittsburgh Sleep Quality Index post‐intervention.
Figures and Tables -
Analysis 12.15

Comparison 12 Movement therapies versus attention control, Outcome 15 Sleep quality as assessed by the Pittsburgh Sleep Quality Index post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 12 Movement therapies versus attention control, Outcome 16 Sleep quality as assessed by the Pittsburgh Sleep Quality Index at 3 month follow‐up.
Figures and Tables -
Analysis 12.16

Comparison 12 Movement therapies versus attention control, Outcome 16 Sleep quality as assessed by the Pittsburgh Sleep Quality Index at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 13 Movement therapies versus attention control ‐ sensitivity analyses, Outcome 1 Functioning as assessed post‐intervention.
Figures and Tables -
Analysis 13.1

Comparison 13 Movement therapies versus attention control ‐ sensitivity analyses, Outcome 1 Functioning as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 13 Movement therapies versus attention control ‐ sensitivity analyses, Outcome 2 Functioning as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 13.2

Comparison 13 Movement therapies versus attention control ‐ sensitivity analyses, Outcome 2 Functioning as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 13 Movement therapies versus attention control ‐ sensitivity analyses, Outcome 3 Pain as assessed by a 10‐point VAS scale post‐intervention.
Figures and Tables -
Analysis 13.3

Comparison 13 Movement therapies versus attention control ‐ sensitivity analyses, Outcome 3 Pain as assessed by a 10‐point VAS scale post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 13 Movement therapies versus attention control ‐ sensitivity analyses, Outcome 4 Pain as assessed by a 10‐point VAS scale at 3 month follow‐up.
Figures and Tables -
Analysis 13.4

Comparison 13 Movement therapies versus attention control ‐ sensitivity analyses, Outcome 4 Pain as assessed by a 10‐point VAS scale at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 13 Movement therapies versus attention control ‐ sensitivity analyses, Outcome 5 Mood as assessed post‐intervention.
Figures and Tables -
Analysis 13.5

Comparison 13 Movement therapies versus attention control ‐ sensitivity analyses, Outcome 5 Mood as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 13 Movement therapies versus attention control ‐ sensitivity analyses, Outcome 6 Mood as assessed at 3 month follow‐up.
Figures and Tables -
Analysis 13.6

Comparison 13 Movement therapies versus attention control ‐ sensitivity analyses, Outcome 6 Mood as assessed at 3 month follow‐up.

Navigate to figure in ReviewOpen in new tab

Comparison 14 Relaxation versus usual care, Outcome 1 Functioning as assessed post‐intervention.
Figures and Tables -
Analysis 14.1

Comparison 14 Relaxation versus usual care, Outcome 1 Functioning as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 14 Relaxation versus usual care, Outcome 2 Pain as assessed post‐intervention.
Figures and Tables -
Analysis 14.2

Comparison 14 Relaxation versus usual care, Outcome 2 Pain as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 14 Relaxation versus usual care, Outcome 3 Mood as assessed post‐intervention.
Figures and Tables -
Analysis 14.3

Comparison 14 Relaxation versus usual care, Outcome 3 Mood as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 14 Relaxation versus usual care, Outcome 4 All cause attrition post‐intervention.
Figures and Tables -
Analysis 14.4

Comparison 14 Relaxation versus usual care, Outcome 4 All cause attrition post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 14 Relaxation versus usual care, Outcome 5 Self‐efficacy as assessed post‐intervention.
Figures and Tables -
Analysis 14.5

Comparison 14 Relaxation versus usual care, Outcome 5 Self‐efficacy as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 14 Relaxation versus usual care, Outcome 6 Fatigue as assessed post‐intervention.
Figures and Tables -
Analysis 14.6

Comparison 14 Relaxation versus usual care, Outcome 6 Fatigue as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 14 Relaxation versus usual care, Outcome 7 Sleep as assessed post‐intervention.
Figures and Tables -
Analysis 14.7

Comparison 14 Relaxation versus usual care, Outcome 7 Sleep as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 15 Relaxation versus attention control, Outcome 1 Pain as assessed post‐intervention.
Figures and Tables -
Analysis 15.1

Comparison 15 Relaxation versus attention control, Outcome 1 Pain as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab

Comparison 15 Relaxation versus attention control, Outcome 2 Mood as assessed post‐intervention.
Figures and Tables -
Analysis 15.2

Comparison 15 Relaxation versus attention control, Outcome 2 Mood as assessed post‐intervention.

Navigate to figure in ReviewOpen in new tab
Summary of findings for the main comparison. Psychological therapies compared to usual care for fibromyalgia

Psychological therapies compared to usual care for fibromyalgia

Patient or population: patients with fibromyalgia
Settings: outpatients
Intervention: psychological therapies
Comparison: usual care

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of p
(studies)

Quality of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Usual care

Psychological therapies

Functioning as assessed post‐intervention
Fibromyalgia Impact Questionnaire. Scale from: 0 to 100
Follow‐up: 1 to 5 weeks

The mean functioning as assessed post‐intervention in the control groups was
6.77

The mean functioning as assessed post‐intervention in the intervention groups was
0.43 standard deviations lower
(0.57 to 0.28 lower)

733
(10 studies)

⊕⊕⊝⊝
low1,2

SMD ‐0.4 (95% CI ‐0.6 to ‐0.3).

Absolute change ‐7.5% (95% CI ‐9.9 to ‐4.9), 2 point shift on a scale of 0‐100.

Relative improvement ‐10.8% (95% CI ‐5.8 to ‐14.3)

NNT 5 (95% CI 4 to 7)

Pain as assessed post‐intervention
100 point visual analog scale. Scale from: 0 to 100

Follow‐up: 3 to 14 weeks

The mean pain as assessed post‐intervention in the control groups was
7.48

The mean pain as assessed post‐intervention in the intervention groups was
0.33 standard deviations lower
(0.52 to 0.15 lower)

453
(9 studies)

⊕⊕⊝⊝
low3,4

SMD ‐0.3 (95% CI ‐0.5 to ‐0.2)

Absolute change ‐3.5% (95% CI ‐5.4 to ‐1.6), 2 point shift on a scale of 0‐100 Relative improvement ‐5.3% (95% CI ‐7.0 to ‐8.3)

NNT 6 (95% CI 4 to 14)

Mood as assessed post‐intervention
State Trait Anxiety Inventory ‐ State Scale. Scale from: 20 to 80
Follow‐up: 1 to 25 weeks

The mean mood as assessed post‐intervention in the control groups was
7.8

The mean mood as assessed post‐intervention in the intervention groups was
0.45 standard deviations lower
(0.64 to 0.26 lower)

492
(8 studies)

⊕⊕⊝⊝
low5,6

SMD ‐0.5 (95% CI ‐0.6 to ‐0.3). Absolute change ‐4.8 (95% CI ‐6.8 to ‐2.8), 3 point shift on a scale of 20‐80 Relative improvement ‐10.8% (95% CI ‐2.5 to ‐6.3)

NNT 5 (95% CI 3 to 8)

All cause attrition post‐intervention
Number of people withdrawing from the study before completing the intervention
Follow‐up: 1 to 25 weeks

Study population

RR 1.38
(1.12 to 1.69)

1687
(22 studies)

⊕⊝⊝⊝
very low7,8

Absolute risk difference 6% (95% CI 0.0 to 0.1)

Relative per cent change 38% (95% CI 12 to 69)

NNTH 18 (95% CI 10 to 55)

148 per 1000

204 per 1000
(165 to 249)

Adverse events post‐intervention
Number of people reporting an adverse event before completing the intervention
Follow‐up: 4 to 6 weeks

Study population

RR 0.38
(0.06 to 2.5)

126
(2 studies)

⊕⊕⊝⊝
low9,10

Absolute risk difference 4% (95% CI ‐0.1 to 0.0)

Relative per cent change 62% (95% CI ‐94 to 150)

Not statistically significant

51 per 1000

19 per 1000
(3 to 127)

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1Downgraded one level due to risk of bias: For some studies allocation concealment was unclear and there was a high risk of selective reporting in one study
2Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery between studies
3Downgraded one level due to risk of bias: For some studies allocation concealment was unclear and there was a high risk of selective reporting in one study
4Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery between studies
5Downgraded one level due to risk of bias: For some studies allocation concealment, blinding of participants and selective reporting were unclear
6Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery between studies
7Downgraded two levels due to risk of bias: Two studies were classified as having a high risk of outcome data and 3 studies were classified as having a high risk of selective reporting bias. Some studies were classified as having an unclear risk of sequence generation, allocation concealment, blinding of outcome assessors and outcome data.
8Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery between studies
9Downgraded one level due to risk of bias: Some studies were classified as having an unclear risk of sequence generation, allocation concealment and one study was classified as having a high risk of selective reporting
10Downgraded one level due to imprecision: There were less than 200 participants in the analysis

Figures and Tables -
Summary of findings for the main comparison. Psychological therapies compared to usual care for fibromyalgia
Navigate to table in Review
Summary of findings 2. Biofeedback compared to usual care for fibromyalgia

Biofeedback compared to usual care for fibromyalgia

Patient or population: patients with fibromyalgia
Settings: outpatients
Intervention: biofeedback
Comparison: usual care

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(studies)

Quality of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Usual care

Biofeedback

Functioning as assessed post‐intervention
Fibromyalgia Impact Questionnaire Revised. Scale from: 0 to 100
Follow‐up: 8 to 24 weeks

The mean functioning as assessed post‐intervention in the control groups was
17.16

The mean functioning as assessed post‐intervention in the intervention groups was
0.06 standard deviations lower
(0.44 lower to 0.33 higher)

106
(2 studies)

⊕⊝⊝⊝
very low1,2,3

SMD ‐0.1 (95% CI ‐0.4 to 0.3)

Absolute change ‐1.2% (95% CI ‐8.8 to 6.6) Relative improvement 2.2% (95% CI ‐16.3 to12.2)

Not statistically significant

Pain as assessed post‐intervention
100 point visual analog scale. Scale from: 0 to 100
Follow‐up: mean 8 weeks

The mean pain as assessed post‐intervention in the control groups was
1.3

The mean pain as assessed post‐intervention in the intervention groups was
2.6 lower
(91.29 lower to 86.09 higher)

65
(1 study)

⊕⊝⊝⊝
very low4,5

MD ‐2.6 (95% CI ‐91.3 to 86.1)

Absolute change ‐2.6% (95% CI ‐91.0 to 86.0)

Relative improvement ‐4.0% (95% CI ‐1.0 to1.0)

Not statistically significant

Mood as assessed post‐intervention
The Symptom Checklist‐90 Revised. Scale from: 0 to 90
Follow‐up: 8 to 24 weeks

The mean mood as assessed post‐intervention in the control groups was
7.3

The mean mood as assessed post‐intervention in the intervention groups was
0.13 standard deviations higher
(0.26 lower to 0.52 higher)

104
(2 studies)

⊕⊝⊝⊝
very low6,7,8

SMD 0.1 (95% CI ‐0.3 to 0.5)

Absolute change 1.9% (95% CI ‐3.7 to 7.4)

Relative improvement 3.6% (95% CI ‐7.2 to 14.5)

Not statistically significant

All cause attrition post‐intervention
Number of people withdrawing from the study before completing the intervention
Follow‐up: 4 to 24 weeks

Study population

RR 4.08
(1.43 to 11.62)

125
(3 studies)

⊕⊝⊝⊝
very low9,10,11

Absolute risk difference 20% (95% CI 0.8 to 0.3)

Relative per cent change 308% (95% CI 43 to 1062)

NNTH 7 (95% CI 3 to 41)

63 per 1000

259 per 1000
(91 to 738)

Adverse events post‐intervention ‐ not reported

See comment

See comment

Not estimable

‐

See comment

Not estimable

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1Downgraded one level due to risk of bias: Random sequence generation and allocation concealment was unclear for one study
2Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery
3Downgraded one level due to imprecision: There were less than 200 participants in this analysis
4Downgraded one level due to risk of bias: Random sequence generation and allocation concealment was unclear for one study
5Downgraded one level due to imprecision: There were less than 100 participants in the analysis
6Downgraded one level due to imprecision: Downgraded one level due to risk of bias: Random sequence generation and allocation concealment was unclear for one study
7Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery
8Downgraded one level due to imprecision: There were less than 200 participants in the analysis
9Downgraded one level due to risk of bias: Random sequence generation and allocation concealment was unclear for one study
10Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery
11Downgraded one level due to imprecision: There were less than 200 participants in the analysis

Figures and Tables -
Summary of findings 2. Biofeedback compared to usual care for fibromyalgia
Navigate to table in Review
Summary of findings 3. Mindfulness compared to usual care for fibromyalgia

Mindfulness compared to usual care for fibromyalgia

Patient or population: patients with fibromyalgia
Settings: outpatients
Intervention: mindfulness
Comparison: usual care

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(studies)

Quality of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Usual care

Mindfulness

Functioning as assessed post‐intervention
Fibromyalgia Impact Questionnaire. Scale from: 0 to 100
Follow‐up: mean 8 weeks

The mean functioning as assessed post‐intervention in the control groups was
17.22

The mean functioning as assessed post‐intervention in the intervention groups was
0.26 standard deviations lower
(0.6 lower to 0.09 higher)

128
(2 studies)

⊕⊕⊝⊝
low1

SMD ‐0.3 (95% CI ‐0.6 to 0.1)

Absolute change ‐4.8% (95% CI ‐11.2 to 1.7%)

Relative improvement ‐8.5% (95% CI ‐19.3 to 3.5)

Not statistically significant

Pain as assessed post‐intervention
Visual analog scale 0 to 100. Scale from: 0 to 10.
Follow‐up: mean 8 weeks

The mean pain as assessed post‐intervention in the control groups was
0.21

The mean pain as assessed post‐intervention in the intervention groups was
0.09 standard deviations lower
(0.44 lower to 0.26 higher)

128
(2 studies)

⊕⊕⊝⊝
low2,3

SMD ‐0.09 (95% CI ‐0.4 to 0.3)

Absolute change ‐1.28% (95% CI ‐6.2 to 3.7)

Relative improvement ‐2.3% (95% CI ‐11.1 to 6.6)

Not statistically significant

Mood as assessed post‐intervention
State Trait Anxiety Inventory State Scale. Scale from: 0 to 60.
Follow‐up: mean 8 weeks

The mean mood as assessed post‐intervention in the control groups was
10.28

The mean mood as assessed post‐intervention in the intervention groups was
0.24 standard deviations lower
(0.51 lower to 0.03 higher)

218
(3 studies)

⊕⊕⊕⊝
moderate4

SMD ‐0.24 (95% CI ‐0.5 to 0.0)

Absolute change ‐3.7% (95% CI ‐7.9 to 0.5)

Relative improvement ‐8.7% (95% CI ‐18.5 to 1.2)

Not statistically significant

All cause attrition post‐intervention
Number of people withdrawing from the study before completing the intervention
Follow‐up: mean 8 weeks

Study population

RR 1.07
(0.67 to 1.72)

195
(3 studies)

⊕⊕⊕⊝
moderate5

Absolute risk difference 2% (95% CI ‐0.10 to 0.14)

Relative per cent change 98% (95% CI ‐90 to ‐86)

Not statistically significant

223 per 1000

239 per 1000
(150 to 384)

Adverse events post‐intervention ‐ not reported

See comment

See comment

Not estimable

‐

See comment

Not reported

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1Downgraded two levels due to imprecision: There were less than 200 participants in the analysis
2Downgraded one level due to risk of bias: One study was classified as having a high risk of blinding of the outcome assessors
3Downgraded one level due to imprecision: There were less than 200 participants in the analysis
4Downgraded one level due to risk of bias: One study was classified as having a high risk of blinding of the outcome assessors
5Downgraded one level due to risk of bias: One study was classified as having a high risk of blinding of the outcome assessors with one study classified as having an unclear risk of sequence generation, allocation concealment and blinding of the outcome assessors

Figures and Tables -
Summary of findings 3. Mindfulness compared to usual care for fibromyalgia
Navigate to table in Review
Summary of findings 4. Movement therapies compared to usual care for fibromyalgia

Movement therapies compared to usual care for fibromyalgia

Patient or population: patients with fibromyalgia
Settings: outpatients
Intervention: movement therapies
Comparison: usual care

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of participants
(studies)

Quality of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Usual care

Movement therapies

Functioning as assessed post‐intervention
Fibromyalgia Impact Questionnaire ‐ Revised. Scale from: 0 to 100.
Follow‐up: 8 to 14 weeks

The mean functioning as assessed post‐intervention in the control groups was
13.3

The mean functioning as assessed post‐intervention in the intervention groups was
0.19 standard deviations lower
(0.53 lower to 0.15 higher)

143
(4 studies)

⊕⊝⊝⊝
very low1,2,3

SMD ‐0.19 (95% CI ‐0.5 to 0.2).

Absolute change ‐3.4% (95% CI ‐9.4 to 2.7) 2 point change on 0 to 100 scale

Relative improvement ‐6.8% (95% CI ‐19.1 to 5.5)

Not statistically significant

Pain as assessed post‐intervention
10 point visual analog scale. Scale from: 0 to 10
Follow‐up: mean 8 weeks

The mean pain as assessed post‐intervention in the control groups was
‐0.37

The mean pain as assessed post‐intervention in the intervention groups was
2.3 lower
(4.19 to 0.41 lower)

28
(1 study)

⊕⊝⊝⊝
very low4,5

MD ‐2.3 (95% CI ‐4.2 to ‐0.4)

Absolute change ‐23.0% (95% CI ‐42.0 to ‐4.0)

Relative improvement ‐3.0% (95% CI ‐6 to ‐0.6)

NNT 3 (95% CI 2 to 41)

Mood as assessed post‐intervention
Center for Epidemiologic Studies Depression Scale. Scale from: 0 to 60
Follow‐up: mean 8 weeks

The mean mood as assessed post‐intervention in the control groups was
0.41

The mean mood as assessed post‐intervention in the intervention groups was
9.84 lower
(18.51 to 1.17 lower)

29
(1 study)

⊕⊝⊝⊝
very low6,7

MD ‐9.8 (95% CI ‐18.5 to ‐1.2)

Absolute change ‐16.4% (95% CI ‐31.0 to ‐2.0)

Relative improvement ‐0.7% (95% CI ‐1.3 to ‐0.1)

NNT 3 (95% CI 2 to 34)

All cause attrition post‐intervention
Number of people withdrawing from the study before completing the intervention
Follow‐up: 8 to 24 weeks

Study population

RR 1.95
(1.13 to 3.38)

240
(5 studies)

⊕⊝⊝⊝
very low8,9

Absolute risk difference 11% (95% CI 0.0 to 0.2)

Relative per cent change 95% (95% CI 13 to 238)

NNTH 13 (95% CI 5 to 105)

106 per 1000

206 per 1000
(119 to 357)

Adverse events post‐intervention
Number of people reporting an adverse event before completing the intervention
Follow‐up: 8 to 24 weeks

Study population

RR 4.62
(0.23 to 93.72)

98
(1 study)

⊕⊝⊝⊝
very low11,12,13

Absolute risk difference 4% (95% CI ‐0.0 to 0.1)

Relative per cent change 362% (95% CI ‐77 to 9272)

Not statistically significant

0 per 1000

40 per 100010
(0 to 0)

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1Downgraded one level due to risk of bias: One study was classified as having a high risk of allocation concealment and blinding of outcome assessors
2Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery
3Downgraded one level due to imprecision: There were less than 200 participants in the analysis
4Downgraded one level due to risk of bias: One study was classified as having a high risk of allocation concealment and blinding of outcome assessors
5Downgraded one level due to imprecision: There were less than 100 participants in the analysis
6Downgraded one level due to risk of bias: One study was classified as having a high risk of allocation concealment and blinding of outcome assessors
7Downgraded one level due to imprecision: There were less than 200 participants in the analysis
8Downgraded one level due to risk of bias: One study was classified as having a high risk of allocation concealment and blinding of outcome assessors and one study had a high risk of selective reporting
9Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery
10 Absolute effect calculated from risk difference

11Downgraded one level due to risk of bias: One study was classified as having a high risk of selective reporting and unclear sequence generation and allocation concealment
12Downgraded one level due to inconsistency: There was diversity in the duration of intervention delivery
13Downgraded one level due to imprecision: There were less than 200 participants in the analysis

Figures and Tables -
Summary of findings 4. Movement therapies compared to usual care for fibromyalgia
Navigate to table in Review
Summary of findings 5. Relaxation compared to usual care for fibromyalgia

Relaxation compared to usual care for fibromyalgia

Patient or population: patients with fibromyalgia
Settings: outpatients
Intervention: relaxation
Comparison: usual care

Outcomes

Illustrative comparative risks* (95% CI)

Relative effect
(95% CI)

No of p
(studies)

Quality of the evidence
(GRADE)

Comments

Assumed risk

Corresponding risk

Usual care

Relaxation

Functioning as assessed post‐intervention
Fibromyalgia Impact Questionnaire. Scale from: 0 to 80
Follow‐up: 6 to 10 weeks

The mean functioning as assessed post‐intervention in the control groups was
3.16

The mean functioning as assessed post‐intervention in the intervention groups was
1.63 standard deviations lower
(10.14 to 6.53 lower)

67
(2 studies)

⊕⊝⊝⊝
very low1,2

MD ‐8.3 (95% CI ‐10.1 to ‐6.5).

Absolute change ‐10.4% (95% CI ‐13.0 to ‐8.0), 5 point shift on 0 to 80 scale

Relative improvement ‐20.0% (95% CI ‐0.2 to ‐0.2)

NNT 2 (95% CI 1 to 2)

Pain as assessed post‐intervention
Short Form ‐ McGill Pain Questionnaire Total Score. Scale from: 0 to 78
Follow‐up: 6 to 10 weeks

The mean pain as assessed post‐intervention in the control groups was
1.86

The mean pain as assessed post‐intervention in the intervention groups was
1.02 standard deviations lower
(1.55 to 0.5 lower)

67
(2 studies)

⊕⊝⊝⊝
very low3,4

SMD ‐1.0 (95% CI ‐1.6 to ‐0.5).

Absolute change ‐3.5% (95% CI ‐5.3 to ‐1.7), 2 point shift on a scale of 0 to 8

Relative improvement ‐9.5% (95% CI ‐14.5 to ‐4.8)

NNT 2 (95% CI 1 to 4)

Mood as assessed post‐intervention
Center for Epidemiologic Disease Depression Scale. Scale from: 0 to 60
Follow‐up: mean 6 weeks

The mean mood as assessed post‐intervention in the control groups was
‐1.9

The mean mood as assessed post‐intervention in the intervention groups was
4.44 lower
(14.46 lower to 5.58 higher)

19
(1 study)

⊕⊝⊝⊝
very low5,6

MD ‐4.4 (95% CI ‐14.5 to 5.6)

Absolute change ‐7.4% (95% CI ‐24 to 9)

Relative improvement ‐27% (95% CI ‐0.9 to ‐0.3)

Not statistically significant

All cause attrition post‐intervention
Number of people withdrawing from the study before completing the intervention
Follow‐up: mean 6 weeks

Study population

RR 4.4
(0.59 to 33.07)

21
(1 study)

⊕⊝⊝⊝
very low7,8

Absolute risk difference 31% (95% CI ‐0.0 to 0.7)

Relative per cent change 340% (95% CI ‐41 to 3207)

Not statistically significant

91 per 1000

400 per 1000
(54 to 1000)

Adverse events post‐intervention ‐ not reported

See comment

See comment

Not estimable

‐

See comment

Not estimable

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio

GRADE Working Group grades of evidence
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

1Downgraded one level due to risk of bias: One study was classified as having an unclear risk of blinding of outcome assessors
2Downgraded one level due to imprecision: There were less than 100 participants in the analysis
3Downgraded one level due to risk of bias: One study was classified as having an unclear risk of blinding of outcome assessors
4Downgraded one level due to imprecision: There were less than 100 participants in the analysis
5Downgraded one level due to risk of bias: One study was classified as having an unclear risk of blinding of outcome assessors
6Downgraded one level due to imprecision: There were less than 100 participants in the analysis
7Downgraded one level due to risk of bias: One study was classified as having an unclear risk of blinding of outcome assessors
8Downgraded one level due to imprecision: There were less than 100 participants in the analysis

Figures and Tables -
Summary of findings 5. Relaxation compared to usual care for fibromyalgia
Navigate to table in Review
Comparison 1. Psychological therapies versus usual care

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Functioning as assessed post‐intervention Show forest plot

10

733

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.43 [‐0.57, ‐0.28]

2 Functioning as assessed at 3 month follow‐up Show forest plot

3

148

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.54 [‐0.87, ‐0.21]

3 Functioning as assessed at 6 month follow‐up Show forest plot

1

112

Mean Difference (IV, Fixed, 95% CI)

‐3.66 [‐7.29, ‐0.03]

4 Pain as assessed post‐intervention Show forest plot

9

453

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.33 [‐0.52, ‐0.15]

5 Pain as assessed at 3 month follow‐up Show forest plot

2

115

Mean Difference (IV, Fixed, 95% CI)

‐0.85 [‐1.76, 0.06]

6 Pain as assessed at 6 month follow‐up Show forest plot

5

371

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.51 [‐0.72, ‐0.30]

7 Mood as assessed post‐intervention Show forest plot

8

492

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.45 [‐0.64, ‐0.26]

8 Mood as assessed at 3 month follow‐up Show forest plot

4

182

Std. Mean Difference (IV, Fixed, 95% CI)

‐1.15 [‐1.50, ‐0.80]

9 Mood as assessed at 6 month follow‐up Show forest plot

2

213

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.17 [‐0.44, 0.10]

10 All cause attrition post‐intervention Show forest plot

22

1687

Risk Ratio (M‐H, Fixed, 95% CI)

1.38 [1.12, 1.69]

11 Adverse events post‐intervention Show forest plot

2

126

Risk Ratio (M‐H, Fixed, 95% CI)

0.38 [0.06, 2.50]

12 Fatigue as assessed post‐intervention Show forest plot

2

82

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.09 [‐0.53, 0.34]

13 Fatigue as assessed at 6 months post‐intervention Show forest plot

2

160

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.07 [‐0.38, 0.24]

14 Self‐efficacy as assessed post‐intervention Show forest plot

4

255

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.25 [‐0.50, ‐0.00]

15 Tender point count as assessed at 6 month follow‐up Show forest plot

1

42

Mean Difference (IV, Fixed, 95% CI)

‐0.38 [‐0.88, 0.12]

16 Quality of life as assessed post‐intervention Show forest plot

6

276

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.19 [‐0.44, 0.06]

17 Quality of life as assessed at 3 month follow‐up Show forest plot

1

33

Mean Difference (IV, Fixed, 95% CI)

‐15.10 [‐21.90, ‐8.30]

18 Quality of life as assessed at 6 month follow‐up Show forest plot

1

42

Mean Difference (IV, Fixed, 95% CI)

‐2.50 [‐7.95, 2.95]

19 Sleep as assessed post‐intervention Show forest plot

5

222

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.52 [‐0.80, ‐0.25]

20 Sleep as assessed at 3 month follow‐up Show forest plot

1

64

Mean Difference (IV, Fixed, 95% CI)

‐11.30 [‐15.44, ‐7.16]

21 Sleep as assessed at 6 month follow‐up Show forest plot

3

224

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.15 [‐0.42, 0.12]

22 Self‐efficacy as assessed at 3 month follow‐up Show forest plot

1

33

Mean Difference (IV, Fixed, 95% CI)

‐15.10 [‐44.95, 14.75]
Figures and Tables -
Comparison 1. Psychological therapies versus usual care
Navigate to table in Review
Comparison 2. Psychological therapies versus usual care sensitivity analyses

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Mood as assessed post‐intervention Show forest plot

7

428

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.29 [‐0.48, ‐0.10]

2 Mood as assessed at 3 month follow‐up Show forest plot

3

118

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.63 [‐1.00, ‐0.26]

3 Fatigue as assessed post‐intervention Show forest plot

1

42

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.42 [‐1.04, 0.19]

4 Sleep as assessed post‐intervention Show forest plot

4

158

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.18 [‐0.50, 0.13]

5 Sleep as assessed at 6 month follow‐up Show forest plot

2

160

Std. Mean Difference (IV, Fixed, 95% CI)

0.22 [‐0.09, 0.53]
Figures and Tables -
Comparison 2. Psychological therapies versus usual care sensitivity analyses
Navigate to table in Review
Comparison 3. Psychological therapies versus attention control

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Functioning as assessed post‐intervention Show forest plot

7

561

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.10 [‐0.27, 0.07]

2 Functioning as assessed at 3 month follow‐up Show forest plot

4

447

Std. Mean Difference (IV, Fixed, 95% CI)

0.02 [‐0.17, 0.20]

3 Functioning as assessed at 6 month follow‐up Show forest plot

3

326

Std. Mean Difference (IV, Fixed, 95% CI)

0.00 [‐0.22, 0.23]

4 Pain as assessed post‐intervention Show forest plot

5

324

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.28 [‐0.51, ‐0.06]

5 Pain as assessed at 3 month follow‐up Show forest plot

2

115

Std. Mean Difference (IV, Fixed, 95% CI)

0.13 [‐0.24, 0.50]

6 Pain as assessed at 6 month follow‐up Show forest plot

1

60

Mean Difference (IV, Fixed, 95% CI)

‐0.34 [‐0.89, 0.21]

7 Mood as assessed post‐intervention Show forest plot

5

330

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.12 [‐0.33, 0.10]

8 Mood as assessed at 3 month follow‐up Show forest plot

2

115

Std. Mean Difference (IV, Fixed, 95% CI)

0.24 [‐0.13, 0.61]

9 All cause attrition post‐intervention Show forest plot

8

669

Risk Ratio (M‐H, Fixed, 95% CI)

0.68 [0.54, 0.87]

10 Fatigue as assessed post‐intervention Show forest plot

2

153

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.12 [‐0.44, 0.20]

11 Fatigue as assessed at 3 month follow‐up Show forest plot

1

69

Mean Difference (IV, Fixed, 95% CI)

‐0.18 [‐0.73, 0.37]

12 Self‐efficacy as assessed post‐intervention Show forest plot

1

105

Mean Difference (IV, Fixed, 95% CI)

0.48 [‐0.27, 1.23]

13 Self efficacy as assessed at 3 month follow‐up Show forest plot

2

151

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.27 [‐0.59, 0.05]

14 Self‐efficacy as assessed at 6 month follow‐up Show forest plot

1

32

Mean Difference (IV, Fixed, 95% CI)

0.01 [‐1.31, 1.33]

15 Tender point score as assessed post‐intervention Show forest plot

2

150

Mean Difference (IV, Fixed, 95% CI)

‐0.80 [‐1.62, 0.02]

16 Quality of life as assessed post‐intervention Show forest plot

3

308

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.13 [‐0.35, 0.10]

17 Quality of life as assessed at 3 month follow‐up Show forest plot

2

218

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.05 [‐0.31, 0.22]

18 Quality of life as assessed at 6 month follow‐up Show forest plot

1

171

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.04 [‐0.34, 0.26]

19 Sleep as assessed post‐intervention Show forest plot

2

109

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.12 [‐0.50, 0.25]

20 Sleep as assessed at 3 month follow‐up Show forest plot

1

69

Mean Difference (IV, Fixed, 95% CI)

0.01 [‐0.45, 0.47]
Figures and Tables -
Comparison 3. Psychological therapies versus attention control
Navigate to table in Review
Comparison 4. Psychological therapies versus attention control sensitivity analyses

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Functioning as assessed post‐intervention Show forest plot

6

390

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.25 [‐0.45, ‐0.05]

2 Functioning as assessed at 6 month follow‐up Show forest plot

2

266

Std. Mean Difference (IV, Fixed, 95% CI)

0.14 [‐0.10, 0.38]

3 Pain as assessed post‐intervention Show forest plot

4

255

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.40 [‐0.66, ‐0.15]

4 Sleep as assessed post‐intervention Show forest plot

1

40

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.51 [‐1.14, 0.12]
Figures and Tables -
Comparison 4. Psychological therapies versus attention control sensitivity analyses
Navigate to table in Review
Comparison 5. Biofeedback versus usual care

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Functioning as assessed post‐intervention Show forest plot

2

106

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.06 [‐0.44, 0.33]

2 Functioning as assessed at 3 month follow‐up Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐0.41 [‐8.88, 8.06]

3 Pain as assessed post‐intervention Show forest plot

1

65

Mean Difference (IV, Fixed, 95% CI)

‐2.60 [‐91.29, 86.09]

4 Mood as assessed post‐intervention Show forest plot

2

104

Std. Mean Difference (IV, Fixed, 95% CI)

0.13 [‐0.26, 0.52]

5 Mood as assessed at 3 month follow‐up Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

4.61 [‐0.16, 9.38]

6 All cause attrition post‐intervention Show forest plot

3

125

Risk Ratio (M‐H, Fixed, 95% CI)

4.08 [1.43, 11.62]

7 Tender point score as assessed post‐intervention Show forest plot

2

101

Mean Difference (IV, Fixed, 95% CI)

‐0.92 [‐2.29, 0.45]

8 Tender point score as assessed at 3 month follow‐up Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐0.09 [‐0.85, 0.67]

9 Quality of life (Physical functioning) as assessed post‐intervention Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐4.92 [‐19.30, 9.46]

10 Quality of life (Role‐Physical) as assessed post‐intervention Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐19.27 [‐41.03, 2.49]

11 Quality of life (Bodily Pain) as assessed post‐intervention Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

6.27 [‐4.49, 17.03]

12 Quality of life (General Health) as assessed post‐intervention Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐8.14 [‐20.47, 4.19]

13 Quality of life (Vitality) as assessed post‐intervention Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐13.43 [‐24.06, ‐2.80]

14 Quality of life (Social Functioning) as assessed post‐intervention Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐10.42 [‐26.61, 5.77]

15 Quality of life (Role‐Emotional) as assessed post‐intervention Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐9.49 [‐39.26, 20.28]

16 Quality of life (Mental Health) as assessed post‐intervention Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐9.32 [‐22.93, 4.29]

17 Quality of life (Physical functioning) as assessed at 3 month follow‐up Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

0.0 [‐13.73, 13.73]

18 Quality of life (Role‐Physical) as assessed at 3 month follow‐up Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐5.21 [‐24.28, 13.86]

19 Quality of life (Bodily Pain) as assessed at 3 month follow‐up Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

0.71 [‐8.15, 9.57]

20 Quality of life (Social Functioning) as assessed at 3 month follow‐up Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐7.43 [‐24.21, 9.35]

21 Quality of life (General Health) as assessed at 3 month follow‐up Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐0.94 [‐12.33, 10.45]

22 Quality of life (Vitality) as assessed at 3 month follow‐up Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐10.17 [‐20.57, 0.23]

23 Quality of life (Role‐Emotional) as assessed at 3 month follow‐up Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐23.84 [‐53.57, 5.89]

24 Quality of life (Mental Health) as assessed at 3 month follow‐up Show forest plot

1

36

Mean Difference (IV, Fixed, 95% CI)

‐6.44 [‐18.27, 5.39]
Figures and Tables -
Comparison 5. Biofeedback versus usual care
Navigate to table in Review
Comparison 6. Biofeedback versus attention control

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Functioning as assessed post‐intervention Show forest plot

1

30

Mean Difference (IV, Fixed, 95% CI)

13.60 [1.05, 26.15]

2 Pain as assessed post‐intervention Show forest plot

1

30

Mean Difference (IV, Fixed, 95% CI)

2.66 [1.21, 4.11]

3 All cause attrition post‐intervention Show forest plot

2

74

Risk Ratio (M‐H, Fixed, 95% CI)

3.46 [0.44, 27.19]

4 Tender point score as assessed post‐intervention Show forest plot

1

30

Mean Difference (IV, Fixed, 95% CI)

2.93 [0.15, 5.71]
Figures and Tables -
Comparison 6. Biofeedback versus attention control
Navigate to table in Review
Comparison 7. Mindfulness versus usual care

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Functioning as assessed post‐intervention Show forest plot

2

128

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.26 [‐0.60, 0.09]

2 Functioning assessed at 3 month follow‐up Show forest plot

1

103

Mean Difference (IV, Fixed, 95% CI)

‐0.06 [‐0.78, 0.66]

3 Pain as assessed post‐intervention Show forest plot

2

128

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.09 [‐0.44, 0.26]

4 Pain as assessed at 3 month follow‐up Show forest plot

1

103

Mean Difference (IV, Fixed, 95% CI)

‐0.28 [‐2.37, 1.81]

5 Mood as assessed post‐intervention Show forest plot

3

218

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.24 [‐0.51, 0.03]

6 Mood as assessed at 3 month follow‐up Show forest plot

2

193

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.21 [‐0.50, 0.07]

7 All cause attrition post‐intervention Show forest plot

3

195

Risk Ratio (M‐H, Fixed, 95% CI)

1.07 [0.67, 1.72]

8 Sleep as assessed post‐intervention Show forest plot

1

97

Mean Difference (IV, Fixed, 95% CI)

‐0.64 [‐2.27, 0.99]

9 Sleep as assessed at 3 month follow‐up Show forest plot

2

134

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.24 [‐0.59, 0.10]
Figures and Tables -
Comparison 7. Mindfulness versus usual care
Navigate to table in Review
Comparison 8. Mindfulness versus usual care ‐ sensitivity analyses

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Mood as assessed at 3 month follow‐up Show forest plot

1

90

Mean Difference (IV, Fixed, 95% CI)

‐1.5 [‐4.77, 1.77]
Figures and Tables -
Comparison 8. Mindfulness versus usual care ‐ sensitivity analyses
Navigate to table in Review
Comparison 9. Movement therapies versus usual care

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Functioning as assessed post‐intervention Show forest plot

4

143

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.19 [‐0.53, 0.15]

2 Pain as assessed post‐intervention Show forest plot

1

28

Mean Difference (IV, Fixed, 95% CI)

‐2.3 [‐4.19, ‐0.41]

3 Mood as assessed post‐intervention Show forest plot

1

29

Mean Difference (IV, Fixed, 95% CI)

‐9.84 [‐18.51, ‐1.17]

4 All cause attrition post‐intervention Show forest plot

5

240

Risk Ratio (M‐H, Fixed, 95% CI)

1.95 [1.13, 3.38]

5 Adverse events post‐intervention Show forest plot

1

98

Risk Ratio (M‐H, Fixed, 95% CI)

4.62 [0.23, 93.72]

6 Fatigue as assessed post‐intervention Show forest plot

1

29

Mean Difference (IV, Fixed, 95% CI)

‐10.8 [‐18.57, ‐3.03]

7 Tender point count as assessed post‐intervention Show forest plot

2

93

Std. Mean Difference (IV, Fixed, 95% CI)

0.18 [‐0.25, 0.60]

8 Sleep as assessed post‐intervention Show forest plot

1

29

Mean Difference (IV, Fixed, 95% CI)

‐4.68 [‐8.14, ‐1.22]
Figures and Tables -
Comparison 9. Movement therapies versus usual care
Navigate to table in Review
Comparison 10. Movement therapies versus usual care ‐ sensitivity analyses intervention type

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Functioning as assessed post‐intervention Show forest plot

3

121

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.26 [‐0.63, 0.11]
Figures and Tables -
Comparison 10. Movement therapies versus usual care ‐ sensitivity analyses intervention type
Navigate to table in Review
Comparison 11. Movement therapies versus usual care ‐ sensitivity analyses quality

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Functioning as assessed post‐intervention Show forest plot

3

115

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.06 [‐0.44, 0.31]
Figures and Tables -
Comparison 11. Movement therapies versus usual care ‐ sensitivity analyses quality
Navigate to table in Review
Comparison 12. Movement therapies versus attention control

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Functioning as assessed post‐intervention Show forest plot

3

191

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.65 [‐0.94, ‐0.35]

2 Functioning as assessed at 3 month follow‐up Show forest plot

3

189

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.53 [‐0.82, ‐0.23]

3 Pain as assessed by a 10‐point VAS scale post‐intervention Show forest plot

3

172

Mean Difference (IV, Fixed, 95% CI)

‐1.45 [‐2.08, ‐0.81]

4 Pain as assessed by a 10‐point VAS scale at 3 month follow‐up Show forest plot

3

165

Mean Difference (IV, Fixed, 95% CI)

‐1.19 [‐1.87, ‐0.52]

5 Mood as assessed post‐intervention Show forest plot

2

141

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.49 [‐0.83, ‐0.15]

6 Mood as assessed at 3 month follow‐up Show forest plot

2

140

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.35 [‐0.69, ‐0.01]

7 All cause attrition post‐intervention Show forest plot

5

279

Risk Ratio (M‐H, Fixed, 95% CI)

1.16 [0.65, 2.09]

8 Adverse events post‐intervention Show forest plot

1

78

Risk Ratio (M‐H, Fixed, 95% CI)

7.0 [0.37, 131.17]

9 Self‐efficacy as assessed post‐intervention Show forest plot

1

60

Mean Difference (IV, Fixed, 95% CI)

‐1.60 [‐2.54, ‐0.66]

10 Self‐efficacy as assessed at 3 month follow‐up Show forest plot

1

59

Mean Difference (IV, Fixed, 95% CI)

‐1.20 [‐2.25, ‐0.15]

11 Tender points as assessed post‐intervention Show forest plot

2

130

Mean Difference (IV, Fixed, 95% CI)

0.09 [‐1.16, 1.33]

12 Tender points as assessed at 3 month follow‐up Show forest plot

2

130

Mean Difference (IV, Fixed, 95% CI)

‐0.39 [‐1.63, 0.85]

13 Quality of life as assessed post‐intervention Show forest plot

2

109

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.70 [‐1.09, ‐0.31]

14 Quality of life as assessed at 3 month follow‐up Show forest plot

2

108

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.52 [‐0.91, ‐0.14]

15 Sleep quality as assessed by the Pittsburgh Sleep Quality Index post‐intervention Show forest plot

2

141

Mean Difference (IV, Fixed, 95% CI)

‐1.88 [‐3.27, ‐0.48]

16 Sleep quality as assessed by the Pittsburgh Sleep Quality Index at 3 month follow‐up Show forest plot

2

140

Mean Difference (IV, Fixed, 95% CI)

‐1.35 [‐2.77, 0.07]
Figures and Tables -
Comparison 12. Movement therapies versus attention control
Navigate to table in Review
Comparison 13. Movement therapies versus attention control ‐ sensitivity analyses

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Functioning as assessed post‐intervention Show forest plot

2

110

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.99 [‐1.39, ‐0.59]

2 Functioning as assessed at 3 month follow‐up Show forest plot

2

108

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.83 [‐1.23, ‐0.43]

3 Pain as assessed by a 10‐point VAS scale post‐intervention Show forest plot

2

110

Mean Difference (IV, Fixed, 95% CI)

‐2.18 [‐2.96, ‐1.40]

4 Pain as assessed by a 10‐point VAS scale at 3 month follow‐up Show forest plot

2

108

Mean Difference (IV, Fixed, 95% CI)

‐1.94 [‐2.77, ‐1.10]

5 Mood as assessed post‐intervention Show forest plot

1

60

Std. Mean Difference (IV, Fixed, 95% CI)

‐1.28 [‐1.84, ‐0.72]

6 Mood as assessed at 3 month follow‐up Show forest plot

1

59

Std. Mean Difference (IV, Fixed, 95% CI)

‐0.96 [‐1.51, ‐0.42]
Figures and Tables -
Comparison 13. Movement therapies versus attention control ‐ sensitivity analyses
Navigate to table in Review
Comparison 14. Relaxation versus usual care

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Functioning as assessed post‐intervention Show forest plot

2

67

Mean Difference (IV, Fixed, 95% CI)

‐8.33 [‐10.14, ‐6.53]

2 Pain as assessed post‐intervention Show forest plot

2

67

Std. Mean Difference (IV, Fixed, 95% CI)

‐1.02 [‐1.55, ‐0.50]

3 Mood as assessed post‐intervention Show forest plot

1

19

Mean Difference (IV, Fixed, 95% CI)

‐4.44 [‐14.46, 5.58]

4 All cause attrition post‐intervention Show forest plot

1

21

Risk Ratio (M‐H, Fixed, 95% CI)

4.4 [0.59, 33.07]

5 Self‐efficacy as assessed post‐intervention Show forest plot

2

67

Std. Mean Difference (IV, Fixed, 95% CI)

‐1.54 [‐2.13, ‐0.95]

6 Fatigue as assessed post‐intervention Show forest plot

1

19

Mean Difference (IV, Fixed, 95% CI)

‐0.82 [‐2.91, 1.27]

7 Sleep as assessed post‐intervention Show forest plot

1

19

Mean Difference (IV, Fixed, 95% CI)

1.03 [‐2.23, 4.29]
Figures and Tables -
Comparison 14. Relaxation versus usual care
Navigate to table in Review
Comparison 15. Relaxation versus attention control

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Pain as assessed post‐intervention Show forest plot

1

39

Mean Difference (IV, Fixed, 95% CI)

‐23.17 [‐36.73, ‐9.61]

2 Mood as assessed post‐intervention Show forest plot

1

39

Mean Difference (IV, Fixed, 95% CI)

‐32.1 [‐46.35, ‐17.85]
Figures and Tables -
Comparison 15. Relaxation versus attention control
Navigate to table in Review