Effectiveness of different styles of massage therapy in ﬁbromyalgia:
A systematic review and meta-analysis
Susan Lee King Yuan
, Luciana Akemi Matsutani, Am
elia Pasqual Marques
University of Sao Paulo, School of Medicine, Department of Physical Therapy, Occupational Therapy and Speech Therapy, Rua Cipotanea, 51 eCidade
Universitaria, CEP: 05360-160 Sao Paulo, SP, Brazil
Received 28 January 2014
Received in revised form
21 September 2014
Accepted 29 September 2014
Physical therapy modalities
The systematic review aimed to evaluate the effectiveness of massage in ﬁbromyalgia. An electronic
search was conducted at MEDLINE, SCiELO, EMBASE, ISI, PEDro, SPORTDiscus, CINAHL, Cochrane CEN-
TRAL and LILACS (Jan 1990eMay 2013). Ten randomized and non-randomized controlled trials investi-
gating the effects of massage alone on symptoms and health-related quality of life of adult patients with
ﬁbromyalgia were included. Two reviewers independently screened records, examined full-text reports
for compliance with the eligibility criteria, and extracted data. Meta-analysis (pooled from 145 partici-
pants) shows that myofascial release had large, positive effects on pain and medium effects on anxiety
and depression at the end of treatment, in contrast with placebo; effects on pain and depression were
maintained in the medium and short term, respectively. Narrative analysis suggests that: myofascial
release also improves fatigue, stiffness and quality of life; connective tissue massage improves depression
and quality of life; manual lymphatic drainage is superior to connective tissue massage regarding
stiffness, depression and quality of life; Shiatsu improves pain, pressure pain threshold, fatigue, sleep and
quality of life; and Swedish massage does not improve outcomes. There is moderate evidence that
myofascial release is beneﬁcial for ﬁbromyalgia symptoms. Limited evidence supports the application of
connective tissue massage and Shiatsu. Manual lymphatic drainage may be superior to connective tissue
massage, and Swedish massage may have no effects. Overall, most styles of massage therapy consistently
improved the quality of life of ﬁbromyalgia patients.
©2014 Elsevier Ltd. All rights reserved.
Fibromyalgia is a chronic, widespread pain disorder commonly
associated with fatigue, sleep disturbance, tenderness, stiffness and
mood disturbances. It can have devastating effects on quality of life,
impairing the patient's ability to work and participate in everyday
activities and affecting relationships with family, friends and em-
ployers (Arnold et al., 2011). Fibromyalgia is common in the general
population (0.6e4.4%), with a higher prevalence among women
(Cavalcante et al., 2006; Branco et al., 2010).
While there is strong evidence for the effectiveness of phar-
macological interventions in the management of ﬁbromyalgia,
there is no conclusive evidence of the effectiveness of non-
pharmacological interventions, which are frequently recom-
mended by health professionals and used by patients (Carville et al.,
2008; Hauser et al., 2012).
Massage has been investigated in the management of ﬁbromy-
algia, as described in ﬁve literature reviews. Three reviews
(Baranowsky et al., 2009; Terhorst et al., 2011; Terry et al., 2012)
included a wide range of interventions of which massage was just
one. Baranowsky et al. (2009) performed a systematic review
focused on complementary and alternative medicine. The cate-
gories identiﬁed were manual manipulation, acupuncture, bal-
neotherapy, thermotherapy, magnetic therapy, homeopathy, mind-
body medicine, diet therapy and music therapy. They found only
one trial assessing massage (Brattberg, 1999), and suggested its
possible effect on ﬁbromyalgia pain and quality of life. Several re-
cords relating to massage might have been missed, possibly due to
the extensive search required for the wide range of interventions.
Terhorst et al. (2011) conducted another systematic review on
complementary and alternative therapies for ﬁbromyalgia, and
performed a more comprehensive search. The meta-analysis sug-
gests that massage is not effective for pain, but it should be noted
that trials with high risk of bias and considerable heterogeneity
were included. Additionally, other outcomes besides pain were not
investigated, and the results for massage were reported and dis-
cussed brieﬂy, considering that massage was just one topic among
many others in the review.
*Corresponding author. Tel.: þ55 11 3091 8423; fax: þ55 11 3091 7462.
E-mail address: email@example.com (S.L.K. Yuan).
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/math
1356-689X/©2014 Elsevier Ltd. All rights reserved.
Manual Therapy 20 (2015) 257e264
Kalichman (2010) issued a narrative review focused on massage
therapy and concluded that most of the evidence supports the
assumption that massage is beneﬁcial for patients with ﬁbromy-
algia. The reliability of his ﬁndings is limited due to the lack of a
systematic method for performing the review. The author stressed
the need for further research to establish the effectiveness of
massage. Kong et al. (2011) conducted a systematic review aimed
for the effect of massage on ﬁbromyalgia pain, and suggested that it
might have positive effects on this outcome. They just considered
pain as the outcome of interest, and included trials with high risk of
bias and considerable heterogeneity in the meta-analysis.
Furthermore, the review is poorly described, as only the annual
scientiﬁc meeting abstract is available.
Recently, Terry et al. (2012) performed an overview of system-
atic reviews of complementary and alternative medicine for ﬁbro-
myalgia, summarizing the evidence from multiple interventions
reviews. They suggested that massage therapy may have short-
term beneﬁcial effects. However, their conclusion may be limited,
because it is based on the results of a single narrative review
Accordingly, there is a need for an updated, focused and rigorous
systematic review to identify, appraise and synthesise all available
evidence regarding the effects of massage on ﬁbromyalgia. Recent
studies might assist in settling disagreements stemming from
previous conﬂicting reviews. With the acknowledgement of the
importance of symptoms other than pain by the time of the pub-
lication of the American College of Rheumatology 2010 criteria for
ﬁbromyalgia (Mease, 2005; Wolfe et al., 2010), other relevant
symptoms that were not considered outcomes of interest in pre-
vious systematic reviews (Kong et al., 2011; Terhorst et al., 2011)
now need to be investigated. The aim of this review was to evaluate
the effectiveness of massage alone to improve pain, pressure pain
threshold (PPT), fatigue, stiffness, anxiety, depression, sleep and
health-related quality of life (HRQoL) in adult ﬁbromyalgia patients.
1.1. Protocol and registration
The protocol of this systematic review was registered at Inter-
national Prospective Register of Systematic Reviews (PROSPERO),
1.2. Eligibility criteria
Eligibility criteria were as follows. (1) Types of studies: rando-
mised or non-randomised, controlled clinical trials; control treat-
ment in non-pharmacological trials could be placebo, usual care,
active treatment or waiting list (Boutron et al., 2008). (2) Type of
participants: adults (18 years) with medical diagnosis of ﬁbro-
myalgia. (3) Types of intervention: massage alone for at least one of
the study groups; touch therapies such as Reiki, and massage with
mechanical devices were excluded, per the deﬁnition of massage in
Medical Subject Headings (MeSH): a group of systematic and sci-
entiﬁc manipulations of body tissues best performed with the
hands. (4) Types of outcome: pain, PPT, fatigue, stiffness, state
anxiety, depression, sleep and HRQoL, assessed immediately after
the end of treatment, over short- (1e3 months), medium- (3
monthse1 year) or long-term (>1 year) follow-up (Haraldsson
et al., 2006).
1.3. Data sources
Studies were identiﬁed by searching MEDLINE, SCiELO, EMBASE,
ISI (Web of Knowledge), PEDro, SPORTDiscus, CINAHL, Cochrane
CENTRAL and LILACS (last access: 31 May 2013). Additional records
were identiﬁed from reference lists of other reviews (Kalichman,
2010; Kong et al., 2011) or indicated by the review authors.
1.4. Electronic search strategy
The databases were searched using the terms “massage”and
“ﬁbromyalgia”. The search was limited to records published since
1990, when the American College of Rheumatology classiﬁcation
criteria for ﬁbromyalgia were ﬁrst published (Wolfe et al., 1990).
1.5. Study selection
After removing duplicates, a screening of records was per-
formed by examining titles and abstracts. Full-text reports were
retrieved and examined for compliance with eligibility criteria.
Screening of records and eligibility assessment were performed
independently by two reviewers. Disagreements between re-
viewers were resolved through consensus; if no consensus could be
reached, a third reviewer decided.
1.6. Data collection process
Two reviewers independently extracted data using a form based
on the checklist of the Cochrane Handbook for Systematic Reviews
of Interventions (Higgins and Green, 2011). Due to missing data and
need for clariﬁcation, attempts were made to contact the authors of
nine included studies.
1.7. Risk of bias assessment
Two reviewers independently assessed risk of bias according to
recommendations of the Cochrane Collaboration's tool (Higgins
and Green, 2011). Five domains were used to assess four types of
bias: selection bias (random sequence generation and allocation
concealment), detection bias (blinding of outcome assessment),
attrition bias (incomplete outcome data) and reporting bias (se-
lective reporting). Performance bias was not used because it is not
possible to blind participants and therapists in massage interven-
tion (Boutron et al., 2004). The reviewers assigned a judgment of
low, high or unclear risk of bias for each domain according to
Cochrane Handbook criteria (Higgins and Green, 2011). Summing
up selection, detection and attrition bias, the overall risk of bias in
individual studies was considered low if at least three domains met
the low risk criteria; high if two or more domains met the high risk
criteria; and unclear otherwise. The risk of bias across studies was
assessed with reporting bias.
1.8. Data analysis
Studies were grouped according to massage style. For each
outcome, in each assessment time point, the following compari-
sons were investigated: two different styles of massage; one style of
massage and another type of intervention; or one style of massage
and one inactive treatment.
A meta-analysis of clinically homogeneous studies with low risk
of bias was conducted. Statistical analyses were conducted using
RevMan 5.2 (2012). Heterogeneity was assessed using the chi-
squared test and I
statistic. Values of p0.1 indicated signiﬁcant
heterogeneity. According to I
results, heterogeneity was consid-
ered not important (0e40%), moderate (30e60%), substantial
(50e90%) or considerable (75e100%) (Higgins and Green, 2011). A
ﬁxed-effect model was used when heterogeneity was considered
not important. For moderate, substantial or considerable hetero-
geneity, a random-effects model was applied if no methodological
S.L.K. Yuan et al. / Manual Therapy 20 (2015) 257e264258
or clinical reason could be found to explain the heterogeneity.
Standardised mean differences for continuous outcomes were used
to express the intervention effect in each study, and the summary
effect estimate was calculated as a weighted average. Scale di-
rections were aligned by adding negative values where required.
Values of p0.05 indicated signiﬁcant effects. Precision of the
summary estimate was expressed with 95% conﬁdence intervals.
The summary effect estimates were used to evaluate the effect
sizes. According to Cohen's categories, the magnitude of effects was
considered small (0.1e0.3), medium (0.4e0.6) or large (0.7e1.0)
When meta-analysis was not feasible, only narrative synthesis
was performed, based on reports of between-groups statistical
comparisons. Quality of evidence was assessed using the following
deﬁnitions (van Tulder et al., 2003): strong (consistent ﬁndings
among multiple high-quality randomised trials), moderate
(consistent ﬁndings among multiple low-quality randomised trials
and/or controlled clinical trials and/or one high-quality randomised
trial), limited (one low-quality randomised trial and/or controlled
clinical trial), conﬂicting (inconsistent ﬁndings among multiple,
randomised and/or controlled clinical trials), and no evidence from
trials (no randomised or controlled clinical trial).
2.1. Study selection
A total of 532 titles were identiﬁed through database searches
and other sources; 147 duplicates were removed, 203 articles were
excluded after preliminary screening, and 182 articles were
assessed for eligibility. Of 168 excluded articles, 145 did not report a
randomised or non-randomised controlled trial, four did not cover
the population of interest, and 19 did not include an intervention
that could be characterised as massage. A total of ten studies re-
ported in 14 articles were included (Fig. 1).
2.2. Study characteristics
Table 1 summarises the characteristics of the primary studies.
There were 478 participants with ﬁbromyalgia across the ten
studies; 212 participants in eight studies were assigned to massage
intervention (Brattberg, 1999; Alnigenis et al., 2001; Lund et al.,
2006; Ekici et al., 2009; Castro-Sanchez et al., 2011a, 2011b; Lip-
tan et al., 2013; Yuan et al., 2013), and in the other two, the number
of participants assigned to each group was not speciﬁed (Sunshine
et al., 1996; Field et al., 2002). Sample sizes ranged from 12 to 94
participants, with a median of 39. Mean age ranged from 34.5 to
53.7 years. Of the 389 participants with reported gender, 97.4%
Six types of massage intervention were evaluated: Swedish
massage, connective tissue massage, manual lymphatic drainage,
myofascial release, shiatsu and a combination of different massage
Forty-one measures were used to assess the eight outcomes of
interest: seven measures for pain, four for PPT, three for fatigue,
three for stiffness, four for anxiety, six for depression, six for sleep,
and ﬁve for HRQoL. Sunshine et al. (1996) did not specify the
measures used to assess pain, fatigue, stiffness or sleep.
MEDLINE (n=49) SCiELO (n=0)
EMBASE (n=176) ISI (n=72)
PEDro (n=28) SPORTDiscus (n=66)
CINAHL (n=74) CENTRAL (n=63)
Fig. 1. Flow diagram of study selection.
S.L.K. Yuan et al. / Manual Therapy 20 (2015) 257e264 259
2.3. Risk of bias
Fig. 2 presents the reviewers' risk of bias judgments. Overall risk
of bias in individual studies was considered low in two studies
(Castro-Sanchez et al., 2011a, 2011b), high in ﬁve (Brattberg, 1999;
Alnigenis et al., 2001; Lund et al., 2006; Liptan et al., 2013; Yuan
et al., 2013), and unclear in three (Sunshine et al., 1996; Field
et al., 2002; Ekici et al., 2009). Because a study protocol was not
available for comparison with the published report for the majority
of the trials, it was not possible to assign a judgment of low or high
risk of publication bias, with the exception of two trials that were
considered low risk.
2.4. Synthesis of results
Results of the effectiveness of the different massage styles on
the outcomes of interest, in contrast to control treatment, were
synthesised according to quality of evidence, and are presented in
2.4.1. Myofascial release
Two trials compared massage with placebo and were included
in the meta-analysis for presenting low risk of bias (Castro-Sanchez
et al., 2011a, 2011b). With regard to the pain outcome, the meta-
analysis (pool of 145 participants) showed that myofascial release
had a large effect immediately after treatment (Fig. 3A), a large
effect in short-term follow-up (Fig. 3B), and a small effect in
medium-term follow-up (Fig. 3C). Regarding anxiety, the meta-
analysis showed that myofascial release had a medium effect af-
ter treatment (Fig. 4), and no effects were observed in short-term or
medium-term follow-up. Regarding depression, the meta-analysis
showed that myofascial release had a medium effect after treat-
ment (Fig. 5A), and a medium effect in short-term follow-up
(Fig. 5B); no signiﬁcant effect was observed in medium-term
In one of the trials, fatigue and stiffness were measured, and
there were statistically signiﬁcant reductions in myofascial release
after treatment and in short-term follow-up. Only fatigue main-
tained signiﬁcant differences in the medium term (Castro-Sanchez
Characteristics of included studies.
Study ID and design Participants gender (F:M); age
Interventions (number and
frequency of sessions)
Sunshine et al., 1996
controlled clinical trial.
30:0; 49.8 years (unknown SD). Swedish massage versus TENS
versus sham TENS (10 sessions,
Anxiety (STAI-S); depression
(CES-D); pain, pain threshold,
fatigue, stiffness and sleep
Randomised, controlled clinical
47:1; 48 ±12.4 years
(completed the trial)
CTM (15 sessions, 1.5/week)
versus no intervention/
discussion group (10 sessions,
Pain (VAS); anxiety and
depression (HADS); sleep (10
questions measured in 0e5
scale); HRQoL (FIQ).
Alnigenis et al., 2001
Randomised, controlled clinical
37:0; 46.4 ±8.1 years. Swedish massage (10 sessions
over 24 weeks) versus standard
care (5 physician visits over 28
weeks) versus standard care
plus phone calls (5 visits and 8
calls over 28 weeks).
Pain (AIMS); anxiety (AIMS);
depression (CES-D and AIMS);
HRQoL (quality of well-being
Field et al., 2002
controlled clinical trial.
20 participants (unspeciﬁed
gender); 50.9 years.
Combination of styles of
massage versus guided
progressive relaxation (10
Pain, fatigue and stiffness
(NRS); pain threshold (TP
count); anxiety (STAI-S);
depression (CES-D); sleep
(movement and hours).
Lund et al., 2006
controlled clinical trial.
19:0; 50.7 ±9.7 years. Swedish massage versus guided
progressive relaxation (12
Pain (CPRS-A and NHP pain);
Ekici et al., 2009
controlled clinical trial.
53:0; 38.84 ±6.38 years (MLD),
36.96 ±8.88 years (CTM).
MLD versus CTM (15 sessions,
Pain (VAS); pain threshold;
fatigue, stiffness, anxiety, and
depression (FIQ); sleep (NHP),
HRQoL (NHP and FIQ).
Castro-Sanchez et al., 2011a
Randomised, blinded, placebo-
controlled clinical trial.
83:3 (completed the trial);
53.7 ±11.5 years.
Myofascial release massage
versus sham short-wave and
ultrasound treatment (40
Pain (McGill questionnaire);
pain threshold (participants per
TP); fatigue, stiffness, anxiety,
depression and HRQoL (FIQ).
Castro-Sanchez et al., 2011b
controlled clinical trial.
60:4; 47.8 ±13.9 years. Myofascial release massage
versus sham magnetotherapy
(20 sessions, 1/week).
Pain (VAS); pain threshold
(participants per TP); anxiety
(STAI-S); depression (Beck
depression inventory); sleep
(PSQI); HRQoL (SF-36).
Liptan et al., 2013
Controlled clinical trial
12:0; 34.5 ±5.5 years. Swedish massage versus
myofascial release massage (4
Pain (modiﬁed Nordic
Questionnaire); HRQoL (FIQ-
Yuan et al., 2013
Controlled clinical trial
38:2; 49.1 ±7.9 years. Full-body Shiatsu (16 sessions,
2/week) versus booklet with
Pain (VAS); pain threshold;
anxiety (STAI-S); sleep (PSQI);
fatigue, stiffness, depression
and HRQoL (FIQ).
AIMS: Arthritis Impact Measurement Scales; CES-D: center for epidemiologic studies depression scale; CPRS-A: comprehensive psychopathological rating scale-affective;
CTM: connective tissue massage; F:M: female and male participants proportion; FIQ: ﬁbromyalgia impact questionnaire; HADS: hospital anxiety and depression scale;
HRQoL: health-related quality of life; MLD: manual lymphatic drainage; NHP: Nottingham Health Proﬁle; NRS: numerical rating scale; PSQI: Pittsburgh Sleep Quality Index;
SD: standard deviation; SF-36: 36-Item Short Form Health Survey; STAI-S: state anxiety scale of the state-trait anxiety inventory; TENS: transcutaneous electrical nerve
stimulation; TP: tender point; VAS: visual analogue scale.
S.L.K. Yuan et al. / Manual Therapy 20 (2015) 257e264260
et al., 2011a). One study used the Fibromyalgia Impact Question-
naire (FIQ) to assess HRQoL (Castro-Sanchez et al., 2011a), while the
other used the 36-Item Short Form Health Survey (SF-36) (Castro-
Sanchez et al., 2011b). A meta-analysis was not conducted because
it was not possible to combine the measures from these in-
struments. In the ﬁrst study, myofascial release resulted in a sta-
tistically signiﬁcant improvement of the total FIQ score after
treatment and in short-term follow-up. In the latter study, myo-
fascial release resulted in signiﬁcant improvement of the scores for
physical functioning, role-physical, bodily pain, vitality and social
functioning after treatment, and for physical functioning, role-
physical and bodily pain in the short-term follow-up. There were
no differences in medium-term follow-up in either of the trials.
Both studies assessed tender points with a pressure algometer,
but no data were provided for PPT or total number of positive
tender points (Castro-Sanchez et al., 2011a, 2011b). Similarly, one
trial assessed sleep, using the Pittsburgh Sleep Quality Index, but
total and component scores were not provided (Castro-Sanchez
et al., 2011b). Due to insufﬁcient information, it was not possible
to analyse PPT or sleep.
A quasi-experimental pilot study with high risk of bias
compared myofascial release (n¼8) and Swedish massage (n¼4)
(Liptan et al., 2013). Pain and HRQoL were measured immediately
after treatment. It was not possible to analyse pain due to insufﬁ-
cient information: no values were provided and no statistical
analysis was performed. No differences in HRQoL were found be-
2.4.2. Swedish massage
Four trials included Swedish massage as the intervention for one
study group. One of the studies was mentioned in the previous
subsection (Liptan et al., 2013). No meta-analysis was performed, as
all of the studies presented high risk of bias. Alnigenis et al. (2001)
compared Swedish massage with standard care for pain, anxiety,
depression and HRQoL, measured in short-term follow-up. No
statistically signiﬁcant differences were found between groups.
Sunshine et al. (1996) compared massage with transcutaneous
electrical nerve stimulation and placebo for pain, PPT, fatigue,
stiffness, anxiety, depression and sleep, measured immediately
after treatment. However, it was not possible to analyse these
outcomes, as no between-group statistical comparison was per-
formed. Lund et al. (2006) compared Swedish massage and guided
progressive relaxation. Pain and depression were measured with
the Comprehensive Psychopathological Rating Scale-Affective
(CPRS-A). The Nottingham Health Proﬁle was also used to mea-
sure pain. However, from the data reported, it was only possible to
extract and analyse pain measured with CPRS-A, and no signiﬁcant
differences were found between groups after treatment or in short-
2.4.3. Connective tissue massage and manual lymphatic drainage
Two randomised trials included connective tissue massage as
the intervention in one study group. Brattberg (1999) compared
connective tissue massage and group discussion on pain, anxiety,
depression, sleep and HRQoL, measured immediately after treat-
ment; massage resulted in statistically signiﬁcant improvement in
depression and HRQoL. This trial could not be included in the meta-
analysis because it had a high risk of bias. Ekici et al. (2009) con-
ducted a study, with unclear risk of bias, which compared con-
nective tissue massage with manual lymphatic drainage on pain,
PPT, fatigue, stiffness, anxiety, depression, sleep and HRQoL,
measured immediately after treatment. Connective tissue massage
was inferior to manual lymphatic drainage, with statistically sig-
niﬁcant differences in stiffness, depression and HRQoL when
measured with FIQ, but not with the Nottingham Health Proﬁle.
Fig. 2. Risk of bias summary.
Quality of evidence of the effectiveness of different styles of massage on the out-
comes of interest.
Quality of evidence Styles of massage eoutcomes
Moderate Myofascial release is more effective than
placebo epain, fatigue, stiffness, anxiety,
depression and HRQoL
Limited Similar effects between myofascial release
and Swedish massage eHRQoL
Similar effects between Swedish massage
and standard care epain, anxiety,
Similar effects between Swedish massage
and guided progressive relaxation epain
Connective tissue massage is more effective
than group discussion edepression, HRQoL
Similar effects between connective tissue
massage and group discussion epain,
Manual lymphatic drainage is more
effective than connective tissue massage e
stiffness, depression, HRQoL
Similar effects between manual lymphatic
drainage and connective tissue massage e
pain, pain threshold, fatigue, anxiety, sleep
Shiatsu is more effective than educational
guidance epain, pain threshold, fatigue,
Similar effects between Shiatsu and
educational guidance estiffness, anxiety,
No evidence Myofascial release is more effective than
placebo epain threshold, sleep
Myofascial release is more effective than
Swedish massage epain
Swedish massage is more effective than
TENS or placebo epain, pain threshold,
fatigue, stiffness, anxiety, depression, sleep
Swedish massage is more effective than
guided progressive relaxation edepression
Combination of several styles of massage is
more effective than guided progressive
relaxation epain, pain threshold, fatigue,
stiffness, anxiety, depression, sleep
TENS: transcutaneous electrical nerve stimulation; HRQoL: health-related quality of
S.L.K. Yuan et al. / Manual Therapy 20 (2015) 257e264 261
One controlled clinical trial with a high risk of bias compared
shiatsu with educational guidance (Yuan et al., 2013). Pain, PPT,
fatigue, stiffness, anxiety, depression, sleep and HRQoL were
measured immediately after treatment, and shiatsu resulted in
statistically signiﬁcant improvements in pain, PPT, fatigue, sleep
2.4.5. Combination of several massage styles
One randomised trial with unclear risk of bias compared mas-
sage therapy with a combination of styles (Swedish massage,
shiatsu and Trager massage) and guided progressive relaxation
(Field et al., 2002). Pain, PPT, fatigue, stiffness, anxiety, depression
and sleep were measured, but it was not possible to analyse them,
because no between-group statistical comparison was performed.
The aim of this systematic review was to appraise the effec-
tiveness of massage therapy on pain, tenderness and other impor-
tant ﬁbromyalgia outcomes such as fatigue, stiffness, anxiety,
depression, sleep and HRQoL. There is currently consensus that
clinically meaningful response to treatment should not be deﬁned
by pain alone in ﬁbromyalgia, which is characterised by multiple
symptoms (Mease, 2005). The comprehensive perspective pro-
vided by assessing a set of outcomes contributes signiﬁcantly to the
current body of knowledge, and distinguishes the present work
from previous systematic reviews that adopted pain as the only
outcome of interest (Kong et al., 2011; Terhorst et al., 2011).
The present review collated a large amount of results from
clinical trials and presented the information in an organised, critical
synthesis, which assists the viewing of differences between trials
that investigate the effects of massage therapy in ﬁbromyalgia. In
conducting the data analysis, the ten studies selected were grouped
according to massage style, considering that the philosophical,
mechanical, physiological and psychological characteristics of each
style of massage might determine its therapeutic efﬁcacy and in-
ﬂuence clinical decision-making on the part of both patients and
health care professionals.
Based on the results of two studies, this review presents mod-
erate evidence that myofascial release has beneﬁcial effects on ﬁ-
bromyalgia in terms of pain, fatigue, stiffness, anxiety, depression
and HRQoL (Castro-Sanchez et al., 2011a, 2011b). These two were
the only trials with low risk of bias, and they investigated the same
style of massage in similar comparisons, allowing the undertaking
of a meta-analysis of pain, anxiety and depression. The meta-
analysis shows a large effect on pain after treatment, which
reduced progressively over the short- and medium-term follow-
ups. A medium effect on anxiety was found after treatment, which
disappeared in short-term follow-up, and a medium effect on
depression was found after treatment, which decreased progres-
sively over the short-term follow-up. Insufﬁcient information on
PPT and sleep was reported,and different instruments were used in
the trials to measure HRQoL; thus, none of these outcomes could be
included in the meta-analysis. Future studies should apply stand-
ardised, validated instruments to measure relevant outcomes for
The limited evidence suggests that connective tissue massage
has beneﬁcial, immediate effects on depression and HRQoL in ﬁ-
bromyalgia patients (Brattberg, 1999). The purpose of connective
tissue massage is to produce an autonomic response via cutaneo-
visceral reﬂexes by applying a specialised stroke to connective
tissue reﬂex zones, speciﬁcally in the bony attachments of fascia or
where fascia is superﬁcial (Holey, 2000). Although the focus of
connective tissue massage differs from that of myofascial release,
improvement in some outcomes might be explained by manipu-
lation of the fascia in both styles.
Liptan (2010) hypothesised that fascial dysfunction in ﬁbromy-
algia leads to widespread pain and central sensitisation. From a
physiological point of view, she suggested that massage aimed at
Fig. 3. Meta-analysis for effects of myofascial release, outcome: pain. A) immediately after treatment; B) short-term follow-up; C) medium-term follow-up.
Fig. 4. Meta-analysis for effects of myofascial release immediately after treatment, outcome: anxiety.
S.L.K. Yuan et al. / Manual Therapy 20 (2015) 257e264262
releasing fascial restriction can treat myofascial ﬁbrotic changes by
breaking up excessive collagen adhesions, thereby reducing excess
tension in the fascial system and promoting tissue healing. These
factors could contribute to pain improvement. The author encour-
aged the undertaking of further research that directly compares a
therapy such as myofascial release to a massage that focuses on
muscle relaxation, which would help deﬁne the role of fascia in
producing ﬁbromyalgia pain.
Accordingly, in 2013, Liptan et al. reported a pilot study
comparing myofascial release and Swedish massage. However, due
to low methodological quality and underpowered sample size, no
evidence was found that myofascial release is more effective than
Swedish massage in reducing pain, and there was limited evidence
of a lack of difference in HRQoL between groups. Therefore, ques-
tions regarding the role of fascia in ﬁbromyalgia and the effec-
tiveness of therapies aimed at releasing fascial restrictions remain.
Other styles of massage presented limited or no evidence of
effectiveness in speciﬁc outcomes. Results of two randomised trials
showed that Swedish massage had no positive effect on the out-
comes at any assessment time point when compared to standard
care, with and without phone calls, or guided progressive relaxa-
tion (Alnigenis et al., 2001; Lund et al., 2006). The limited evidence
available suggested that Swedish massage was not beneﬁcial for
ﬁbromyalgia. Methodological problems with the studies raise
questions about the validity of these ﬁndings, which may under-
estimate the true intervention effect; in addition, a meta-analysis
could not be performed to increase the power to detect an effect.
The limited evidence suggested that manual lymphatic drainage
was superior to connective tissue massage in terms of stiffness,
depression and HRQoL at the end of treatment. The authors
hypothesised that shorter sessions or the more intense pressure of
connective tissue massage might have been responsible for their
ﬁndings (Ekici et al., 2009). There is limited evidence of the bene-
ﬁcial, immediate effects of shiatsu on pain, PPT, fatigue, sleep and
HRQoL, showing the potential of combining the beliefs and prac-
tices of Eastern culture with the effects of Western medicine (Yuan
et al., 2013). There was no evidence that a combination of Swedish
massage, shiatsu and Trager massage was more effective than
guided progressive relaxation. The low methodological quality and
absence of a between-groups comparison did not allow a proper
analysis of the study (Field et al., 2002). Further high-quality
randomised trials with larger sample sizes are necessary to deter-
mine the effects of different massage styles on ﬁbromyalgia.
The present review suggests that every style of massage, except
for Swedish massage, displays positive effects on symptoms and
HRQoL of patients with ﬁbromyalgia. Therefore, Swedish massage,
which is commonly practiced in health care, cannot be recom-
mended at the moment. On the other hand, myofascial release
presented the best evidence of effectiveness for multiple outcomes
and could be preferred over other styles. However, patients and
health care professionals should consider additional criteria in
evidence-based clinical decision-making. The therapy must: affect
outcomes that are important to the patient; have large enough,
positive effect sizes; have no or few adverse effects; and be cost-
effective (Centre for Evidence-Based Physiotherapy, 2014).
3.1. Study limitations
The electronic search was limited to the English and Portuguese
languages, excluding potentially relevant trials published in other
languages. Some authors did not respond to contact attempts, and
many who did respond had not keep the database or could not
remember details of the methods. Some data could not be retrieved
for more precise analysis of the studies and risk of bias assessment.
There is moderate evidence that myofascial release has positive
effects on multiple ﬁbromyalgia symptoms, especially pain, anxiety
and depression, for which the effect sizes are clinically relevant.
Effects on pain and depression were observed in the medium and
short terms, respectively. When comparing connective tissue
massage or shiatsu with educational approaches, limited evidence
supports the application of these styles of massage. Manual
lymphatic drainage might be superior to connective tissue massage
in terms of stiffness and depression. Swedish massage may not be
beneﬁcial for ﬁbromyalgia. Overall, most styles of massage therapy
consistently improved the HRQoL of ﬁbromyalgia patients.
Alnigenis MNY, Bradley JD, Wallick J, Emsley CL. Massage therapy in the manage-
ment of ﬁbromyalgia: a pilot study. J Musculoskelet Pain 2001;9:55e67.
Arnold LM, Clauw DJ, McCarberg BH. Improving the recognition and diagnosis of
ﬁbromyalgia. Mayo Clin Proc 2011;86:457e64.
Baranowsky J, Klose P, Musial F, Hauser W, Dobos G, Langhorst J. Qualitative sys-
temic review of randomized controlled trials on complementary and alternative
medicine treatments in ﬁbromyalgia. Rheumatol Int 2009;30:1e21.
Boutron I, Moher D, Altman DG, Schulz KF, Ravaud P. Extending the CONSORT
statement to randomized trials of nonpharmacologic treatment: explanation
and elaboration. Ann Intern Med 2008;148:295e309.
Boutron I, Tubach F, Giraudeau B, Ravaud P. Blinding was judged more difﬁcult to
achieve and maintain in nonpharmacologic than pharmacologic trials. J Clin
Branco JC, Bannwarth B, Failde I, Abello Carbonell J, Blotman F, Spaeth M, et al.
Prevalence of ﬁbromyalgia: a survey in ﬁve European countries. Semin Arthritis
Brattberg G. Connective tissue massage in the treatment of ﬁbromyalgia. Eur J Pain
Carville SF, Arendt-Nielsen S, Bliddal H, Blotman F, Branco JC, Buskila D, et al. EULAR
evidence-based recommendations for the management of ﬁbromyalgia syn-
drome. Ann Rheum Dis 2008;67:536e41.
Castro-Sanchez AM, Mataran-Penarrocha GA, Arroyo-Morales M, Saavedra-
Hernandez M, Fernandez-Sola C, Moreno-Lorenzo C. Effects of myofascial
release techniques on pain, physical function, and postural stability in patients
Fig. 5. Meta-analysis for effects of myofascial release, outcome: depression. A) immediately after treatment; B) short-term follow-up.
S.L.K. Yuan et al. / Manual Therapy 20 (2015) 257e264 263
with ﬁbromyalgia: a randomized controlled trial. Clin Rehabil 2011a;25:
Castro-Sanchez AM, Mataran-Penarrocha GA, Granero-Molina J, Aguilera-
Manrique G, Quesada-Rubio JM, Moreno-Lorenzo C. Beneﬁts of massage-
myofascial release therapy on pain, anxiety, quality of sleep, depression, and
quality of life in patients with ﬁbromyalgia. Evid Based Complement Altern Med
Cavalcante AB, Sauer JF, Chalot SD, Assumpç~
ao A, Lage LV, Matsutani LA, et al. The
prevalence of ﬁbromyalgia: a literature review. Rev Bras Reumatol 2006;46:9.
Centre for Evidence-Based Physiotherapy. Tutorials: Is the therapy clinically useful?
[internet]. Sydney: Physiotherapy Evidence Database (PEDro). [updated 2014
May 05]. Available from http://www.pedro.org.au/english/tutorial/is-the-
Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale, N.J:
L. Erlbaum Associates; 1988.
Ekici G, Bakar Y, Akbayrak T, Yuksel I. Comparison of manual lymph drainage
therapy and connective tissue massage in women with ﬁbromyalgia: a ran-
domized controlled trial. J Manip Physiol Ther 2009;32:127e33.
Field T, Diego M, Cullen C, Hernandez-Reif M, Sunshine W, Douglas S. Fibromyalgia
pain and substance P decrease and sleep improves after massage therapy. J Clin
Haraldsson BG, Gross AR, Myers CD, Ezzo JM, Morien A, Goldsmith C, et al. Massage
for mechanical neck disorders. Cochrane Database Syst Rev 2006:CD004871.
Hauser W, Jung E, Erbsloh-Moller B, Gesmann M, Kuhn-Becker H, Petermann F, et al.
The German ﬁbromyalgia consumer reports ea cross-sectional survey. BMC
Musculoskelet Disord 2012;13:74.
[internet]. In: Higgins JPT, Green S, editors. Cochrane handbook for systematic re-
views of interventions Version 5.1.0. The Cochrane Collaboration; 2011. p. 982.
Available from, www.cochrane-handbook.org.
Holey EA. Connective tissue massage: a bridge between complementary and or-
thodox approaches. J Bodyw Mov Ther 2000;4:9.
Kalichman L. Massage therapy for ﬁbromyalgia symptoms. Rheumatol Int 2010;30:
Kong L, Bannuru RR, Yuan W, Cheng Y, Fang M, McAlindon T, et al. Therapeutic
massage on pain relief for ﬁbromyalgia: a systematic review and meta-analysis.
Arthritis Rheum 2011;63:1917.
Liptan GL. Fascia: a missing link in our understanding of the pathology of ﬁbro-
myalgia. J Bodyw Mov Ther 2010;14:3e12.
Liptan G, Mist S, Wright C, Arzt A, Jones KD. A pilot study of myofascial release
therapy compared to Swedish massage in Fibromyalgia. J Bodyw Mov Ther
Lund I, Lundeberg T, Carleson J, Sonnerfors H, Uhrlin B, Svensson E. Corticotropin
releasing factor in urineea possible biochemical marker of ﬁbromyalgia. Re-
sponses to massage and guided relaxation. Neurosci Lett 2006;403:166e71.
Mease P. Fibromyalgia syndrome: review of clinical presentation, pathogenesis,
outcome measures, and treatment. J Rheumatol Suppl 2005;75:6e21.
Review Manager (RevMan) [Computer program]. Version 5.2. Copenhagen: The
Nordic Cochrane Centre, The Cochrane Collaboration; 2012.
Sunshine W, Field TM, Quintino O, Fierro K, Kuhn C, Burman I, et al. Fibromyalgia
beneﬁts from massage therapy and transcutaneous electrical stimulation. J Clin
Terhorst L, Schneider MJ, Kim KH, Goozdich LM, Stilley CS. Complementary and
alternative medicine in the treatment of pain in ﬁbromyalgia: a systematic
review of randomized controlled trials. J Manip Physiol Ther 2011;34:483e96.
Terry R, Perry R, Ernst E. An overview of systematic reviews of complementary and
alternative medicine for ﬁbromyalgia. Clin Rheumatol 2012;31:55e66.
van Tulder M, Furlan A, Bombardier C, Bouter L. Updated method guidelines for
systematic reviews in the cochrane collaboration back review group. Spine
(Phila Pa 1976) 2003;28:1290e9.
Wolfe F, Clauw DJ, Fitzcharles MA, Goldenberg DL, Katz RS, Mease P, et al. The
American College of Rheumatology preliminary diagnostic criteria for ﬁbro-
myalgia and measurement of symptom severity. Arthritis Care Res (Hoboken)
Wolfe F, Smythe HA, Yunus MB, Bennett RM, Bombardier C, Goldenberg DL, et al.
The american college of rheumatology 1990 criteria for the classiﬁcation of ﬁ-
bromyalgia. Report of the multicenter criteria committee. Arthritis Rheum
Yuan SL, Berssaneti AA, Marques AP. Effects of shiatsu in the management of ﬁ-
bromyalgia symptoms: a controlled pilot study. J Manip Physiol Ther 2013;36:
S.L.K. Yuan et al. / Manual Therapy 20 (2015) 257e264264