Rheumatol Int (2012) 32:2285–2292

DOI 10.1007/s00296-011-1958-z

O R I G I N A L A R T I CL E

EVectiveness of physical activity in reducing pain in patients

with Wbromyalgia: a blinded randomized clinical trial
Andrea Harumi Kayo · Maria Stella Peccin ·
Carla Munhoz Sanches · Virgínia Fernandes Moça Trevisani

Received: 17 December 2010 / Accepted: 8 May 2011 / Published online: 19 May 2011
 Springer-Verlag 2011

Abstract The purpose of this study was to evaluate and from the control group (P = 0.01). At the end of the treat-
compare the eVectiveness of muscle-strengthening exercises ment, 80% of subjects in the control group took pain
(MS) and a walking program (WA) in reducing pain in medication, but only 46.7% in the WA and 41.4% in the
patients with Wbromyalgia. Ninety women, 30–55 years of MS groups. Mean FIQ total scores were lower for the WA
age, diagnosed with Wbromyalgia according to the Ameri- and MS groups (P = 0.96) compared with the control group
can College of Rheumatology 1990 criteria, were random- (P < 0.01). Patients in the WA and MS groups reported
ized into 3 groups: WA Group, MS Group, and control higher scores (better health status) than controls in almost
group. Pain (visual analog scale) was evaluated as the all SF-36 subscales. MS was as eVective as WA in reducing
primary outcome. Physical functioning (Fibromyalgia pain regarding all study variables; however, symptoms
Impact Questionnaire, FIQ), health-related quality of life management during the follow-up period was more
(Short-Form 36 Health Survey, SF-36), and use of medication eYcient in the WA group.
were evaluated as secondary outcomes. Assessments were
performed at baseline, 8, 16, and 28 weeks. Intention-to-treat Keywords Fibromyalgia · Pain · Aerobic exercise ·
and eYcacy analyses were conducted. Sixty-eight patients Resistance training · Clinical trial
completed the treatment protocol. All 3 groups showed
improvement after the 16-week treatment compared to
baseline. At the 28-week follow-up, pain reduction was Introduction
similar for the WA and MS groups (P = 0.39), but diVerent
Fibromyalgia is a syndrome characterized by chronic wide-
spread body pain, which results in reduced physical func-
tion and frequent use of health-care services [1].
A. H. Kayo · C. M. Sanches
Universidade Federal de São Paulo (UNIFESP), São Paulo, Fibromyalgia symptoms aVect every aspect of the patient’s
SP, Brazil life, including work capacity and productivity, family life,
and leisure activities [1, 2]. Pain is the most common com-
M. S. Peccin
plaint of patients [2–4]. Fibromyalgia is strongly associated
UNIFESP, Santos, SP, Brazil
with a sedentary life style and below-normal cardiopulmo-
V. F. M. Trevisani nary capacity [1, 5], resulting in low functional capacity,
Division of Rheumatology, reduced ability to sustain intense exercise, and low levels of
Santo Amaro University (UNISA), São Paulo, SP, Brazil
physical Wtness [6].
V. F. M. Trevisani Because of the absence of apparent pathological
Division of Emergency Medicine and Evidence-Based Medicine, changes, the treatment of Wbromyalgia is directed toward
UNIFESP, São Paulo, SP, Brazil the management of symptoms, and physical exercises are
frequently recommended [1, 5, 7, 8]. Since the Wrst study
M. S. Peccin (&)
Rua Pedro de Toledo, 598, São Paulo, SP CEP 04039-001, Brazil published in 1988 [9], several other studies have also inves-
e-mail: [email protected] tigated the eVects of physical exercises on Wbromyalgia.

123
2286 Rheumatol Int (2012) 32:2285–2292

However, these studies have used diVerent combinations of was organized by an investigator (MSP), and concealed in
exercise modalities [10–13], making it diYcult to deter- opaque sealed envelopes. The patients started the physical
mine which modality is in fact eVective. There are few exercise program (WA or MS) under supervision of the
evidences that strengthening exercises, as a single modality principal investigator (AHK) at the physiotherapy unit.
[1, 14–17], improve Wbromyalgia symptoms. At present, Dropouts were deWned as patients who missed more than 9
aerobic exercises have been the most recommended type of (20%) treatment sessions, missed 3 consecutive sessions,
treatment, with walking being the modality of choice [18, 19]. and/or had their evaluations scheduled 10 days after the
In this context, the aims of this study were to investigate regular appointment. Participants who were taking tricyclic
the eVectiveness of single modalities of physical exercise in antidepressants to control Wbromyalgia symptoms were
reducing pain in patients with Wbromyalgia, and assess self- asked to discontinue the medication 30 days prior to the
reported physical function, using the Fibromyalgia Impact study (washout period), and advised to take acetaminophen
Questionnaire (FIQ), health-related quality of life, using the (750 mg) tablets, if needed. The patients were assessed for
Medical Outcomes Study 36-Item Short Form Health Sur- pain at monthly clinic visits, and medication was restarted
vey (SF-36), and use of medication in the treatment of if pain persisted or worsened. In order to ensure homogene-
Wbromyalgia. In order to test our hypothesis, a controlled ity of the sample, the study groups consisted exclusively of
comparison of the eVectiveness of muscle-strengthening women in a speciWc age group, so that age-related musculo-
exercises (MS) and walking program (WA), as single skeletal comorbidities and/or other muscle conditions
modalities, was carried out. would neither interfere with results nor with pain percep-
tion and pain threshold.

Materials and methods Procedures

This trial was approved by the Research Ethics Committee In the initial examination (T0), all patients were assessed
of the Universidade Federal of São Paulo (UNIFESP), for pain using a visual analog scale (VAS) and tenderness
Brazil, and registered with ClinicalTrials.gov, number were examined by digital palpation. Following, all patients
NCT00498264. The study was performed in accordance completed self-administered questionnaires (SF-36 and
with the ethical standards of the 1964 Declaration of FIQ) and were allocated randomly to the study groups. The
Helsinki and its succeeding revisions. patients were also examined after 8 (T8) and 16 (T16)
In a period of 2 years (January 2005 to December 2007), weeks of treatment, and 12 weeks after the end of the study
312 patients diagnosed with Wbromyalgia have attended the protocol (28-week follow-up, T28). Therefore, data were
Rheumatology Services at the Interlagos Specialty Outpa- collected at 4 time points: T0, T8, T16, and T28 for a total
tient Clinic, São Paulo, Brazil. Of these, 90 women, who of 7 months of follow-up (including the treatment period).
met the criteria for Wbromyalgia, according to the American The exercise routine for the WA and MS groups consisted
College of Rheumatology 1990 criteria for the classiWca- of physical activity for about 60 min, 3 times per week for
tion of Wbromyalgia [2], were invited to participate in the 16 weeks, and followed the American College of Sports
study. Written informed consent was obtained from all Medicine recommendations [20]. All 48 exercise sessions
patients prior to their inclusion in the study. were supervised by a physical therapist, who was not
All patients were clinically examined by the same rheu- involved in the evaluation process. Patients in the control
matologist (CSM), who was blinded to group assignment group were assessed at 6 time points during the study
throughout the study. Inclusion criteria were women aged period (T0, T4, T8, T12, T16, and T28).
between 30 and 55 years, who agreed to participate in an For the WA group, walking was performed either out-
exercise program 3 times per week for 16 weeks, and to doors or indoors in a gymnasium, depending on the
discontinue medication for Wbromyalgia 4 weeks before the weather. Each session consisted of a warm-up period,
start of the study (washout period), and who had at least stretching (5–10 min), conditioning stimulus, and a cool
4 years of schooling. Women with any contraindication to down period (5 min). During the exercise session, heart rate
exercise on the basis of prior clinical rheumatological (HR) was monitored using a heart rate monitor (Polar A1,
examination, and those involved in cases of medical litiga- Polar Electro Inc., Finland) [5, 9, 19]. Every 4 weeks, walk-
tion were excluded from the study. ing duration was increased (25–30 min to 50 min), as well
After the initial evaluation, the participants were allo- as the intensity of the conditioning stimulus [began at 40–
cated randomly to 1 of 3 groups: walking program (WA) 50% and progressed to 60–70% of the heart rate reserve
group, muscle-strengthening exercises (MS) group, and (HRR)] by week 16. The HR target was calculated using
control group. The allocation sequence was based on a ran- the Karvonen or % heart rate reserve method [18, 20]. In
dom number list (GraphPad Statmate version 1.0), which case of pain while walking, the participants were advised to

123
Rheumatol Int (2012) 32:2285–2292 2287

slow down for a short period of time to relieve pain, and try the Statistical Analysis System (SAS) version 9.0. The
again to reach the target HR. analysis included all 90 participants (intention-to-treat anal-
Participants in the MS group followed an exercise proto- ysis) using the last observed response for patients with
col consisting of 11 free active exercises, using free weights missing data (carry forward) [27]. EYcacy (as-treated)
and body weight performed in the standing, sitting, and lying analysis was also performed. The signiWcance level was set
positions to improve the muscle strength of the upper and at 0.05. Categorical variables were presented as absolute
lower limbs and trunk muscles. On average, the exercise load and relative frequencies (%), while quantitative variables
and intensity were increased every 2 weeks, according to the were expressed as mean, standard deviation (SD), median,
patient’s tolerance and by following the Borg Scale [21]. In Wrst and third quartiles, and minimum and maximum val-
the Wrst 2 weeks, the exercises were performed without load. ues. Box plots were used to represent the distribution of
This was an adaption period in which the participants learned quantitative variables with 95% conWdence intervals for
how to do the exercises and how to maintain posture. Exer- mean values. Data from the 3 groups were compared using
cises were performed for 3 sets of 10 repetitions with rest Pearson’s chi-square test or Fisher’s exact test for categori-
intervals of 1 min between sets. From week 3 to week 16, the cal variables. Analysis of variance (ANOVA) was used for
number of repetitions was increased to 15, and after week 5, the statistical analysis of quantitative variables normally
load was included to the exercises, which were performed distributed, and the non-parametric Kruskal–Wallis test
within a safe range of motion [15, 22]. was used for non-normally distributed variables. Two-fac-
tor ANOVA (group versus time) with repeated measures
Clinical outcomes was used to test group eVect (diVerences between groups),
time eVect (diVerences between time points), and group-
Pain was evaluated as a primary outcome and assessed time interaction (diVerences between groups over time).
using a VAS graded from 0 (no pain) to 10 (worst possible VAS scores were analyzed using two-factor non-parametric
pain) [23]. The criterion for improvement in pain was a ANOVA.
reduction in pain intensity of 40% in the WA and MS
groups compared with 20% in the control group. Physical
function was evaluated as a secondary outcome using the Results
FIQ [24], which is a 10-item, self-administered instrument
that measures physical functioning and other Wbromyalgia Of the 90 patients initially enrolled, 2 in the WA group, 7 in
outcomes. FIQ total scores range from 0 to 100, with higher the MS group, and 2 in the control group left the study
scores indicating greater disability. Health-related quality during the treatment period. In addition, 5 in the WA group,
of life was evaluated using the SF-36 [25], which contains 1 in the MS group, and 5 in the control group did not attend
8 subscales: physical functioning, role physical, bodily the 28-week follow-up assessment. A total of 79 patients
pain, general health, vitality, social functioning, role emo- completed the 16-week treatment protocols, and 68 were
tional, and mental health. Scores on each dimension range evaluated in the 28-week follow-up. The reasons given by
from 0 to 100, with 0 corresponding to the worst health patients for not completing the study, group distribution,
status and 100 to the best health status. and other relevant information are shown in Fig. 1. In the
case of a patient not completing the study, the data col-
Sample size lected in the last evaluation of the patient were imputed for
the missing data, even if only baseline values were avail-
In order to calculate the sample size, the patient’s estima- able. Although some patients in the MS group left the study
tion of pain (VAS) was chosen as the primary outcome at an earlier stage, there were no signiWcant diVerences
measure. A 40% reduction in pain intensity was considered between Kaplan–Meier curves (log-rank test, P = 0.82).
to be the minimum clinically signiWcant diVerence for the Also, there were no signiWcant diVerences in clinical data
present trial [26]. Setting the signiWcance level at 5% and between groups at baseline, except that duration of disease
the power of the sample at 90%, the sample size of 27 was longer for patients in the control group than for those in
patients in each group would be required to detect this the other 2 groups (P = 0.04) (Table 1).
diVerence. Therefore, 30 patients were allocated randomly Regardless of treatment group, the 22 patients who did not
to each group. complete the study had signiWcantly lower mean SF-36 scores
for social functioning (P < 0.01) and vitality (P = 0.03), and
Statistical analysis higher mean FIQ scores (P = 0.04) than the 68 patients who
completed the study (Table 2). No complications or adverse
Data analysis was carried out using the Statistical Package eVects were observed during the study period among patients
for Social Sciences (SPSS) version 11.0 for Windows and who completed the treatment protocols.

123
2288 Rheumatol Int (2012) 32:2285–2292

Fig. 1 Study design and study

Xow chart

Table 1 Clinical characteristics of participants after randomization

Variables Groups P-value

WA (n = 30) MS (n = 30) Control (n = 30)

Age (years) 47.7 (5.3) 46.7 (6.3) 46.1 (6.4) 0.573a

2
BMI (kg/m ) 26.3 (4.5) 26.2 (4.8) 26.7 (4.9) 0.896a
Duration of disease (years) 4.0 (3.1) 4.7 (5.7) 5.4 (3.5) 0.039*b
Tender points 36.57 (9.60) 40.53 (8.82) 39.47 (10.29) 0.258
VAS scores—baseline 8.62 (1.61) 8.67 (1.63) 8.37 (1.45) 0.490
FIQ total score—baseline 63.06 (14.74) 67.32 (16.54) 63.84 (16.70) 0.549
SF-36 subscales
Physical functioning 39.33 (18.56) 37.33 (17.60) 35.83 (17.27) 0.748
Role physical 24.17 (36.25) 17.50 (30.19) 18.33 (28.57) 0.678
Bodily pain 32.13 (16.34) 26.60 (14.62) 31.40 (13.07) 0.292
General health 43.33 (19.05) 40.30 (21.03) 45.57 (17.41) 0.569
Vitality 37.67 (17.99) 31.55 (17.68) 34.40 (19.00) 0.439
Social functioning 57.67 (24.48) 46.00 (29.92) 50.23 (26.39) 0.244
Role emotional 44.40 (40.48) 29.93 (37.51) 27.77 (32.92) 0.174
Mental health 51.33 (24.28) 46.00 (22.77) 46.00 (22.23) 0.589
Data are expressed as mean (SD)
* Statistical signiWcance, a ANOVA, b Kruskall–Wallis test, WA Walking program, MS Muscle-strengthening exercises, SD standard deviation,
BMI body mass index, VAS visual analog scale of pain, FIQ Fibromyalgia impact questionnaire, SF-36 MOS Short-form 36-item health survey

123
Rheumatol Int (2012) 32:2285–2292 2289

Table 2 Comparison of the characteristics of patients who did and did not complete the study
Variables Completed the study Did not complete P-value

Mean (SD) SE Mean (SD) SE

Age (years) 47.1 (6.1) 0.735 45.9 (5.8) 1.230 0.431

Duration of disease (years) 4.6 (3.8) 0.467 4.9 (5.5) 1.168 0.772
Tender points—baseline 39.8 (9.5) 1.153 35.8 (9.6) 2.040 0.089
VAS scores—baseline 8.6 (1.4) 0.173 8.4 (1.9) 0.409 0.742
FIQ scores—baseline 62.8 (15.7) 1.909 70.7 (15.4) 3.292 0.044*
SF-36 scores—baseline
Physical functioning 36.5 (18.1) 2.190 40.7 (16.4) 3.502 0.334
Role physical 22.4 (32.1) 3.890 12.5 (29.6) 6.317 0.202
Bodily pain 31.3 (14.7) 1.779 26.3 (14.8) 3.160 0.175
General health 44.1 (18.2) 2.213 39.7 (21.7) 4.635 0.349
Vitality 36.8 (19.2) 2.335 27.2 (11.9) 2.602 0.034*
Social functioning 56.1 (26.3) 3.185 36.5 (24.9) 5.311 0.003*
Role emotional 36.7 (37.9) 4.593 25.7 (35.5) 7.575 0.231
Mental health 49.4 (2.6) 2.745 42.7 (23.8) 5.085 0.238
* Statistical signiWcance, SD standard deviation, SE standard error, VAS visual analog scale of pain, FIQ Fibromyalgia impact questionnaire,
SF-36 MOS Short-form 36-item health survey

Fig. 2 Changes in pain intensity

over time (VAS scores) in the 3
groups. Intention-to-treat analy-
sis (left) and eYcacy (as-treated)
analysis (right)

Visual analog scale WA group. After the 8th week until the last evaluation, pain
intensity remained stable in the control (P = 0.56) and WA
Although the curves obtained by the intention-to-treat anal- (P = 0.71) groups. On the other hand, reduction in pain
ysis were not completely parallel, group-time interaction intensity was also observed from weeks 8 to 16 (P < 0.01)
was not statistically signiWcant (P = 0.15). A signiWcant in the MS group, but pain increased by week 28 (P < 0.01),
group eVect was observed (P < 0.01). VAS scores were sig- reaching an intensity similar to that reported at T8
niWcantly higher in the control group than in both the WA (P = 0.49). There were no diVerences in pain intensity
group (P = 0.01) and MS group (P = 0.03). There were no between groups at T0 and T8. At week 16, pain intensity
signiWcant diVerences in VAS scores between the WA and was signiWcantly more severe in the control group than in
MS groups (P = 0.19). A signiWcant time eVect was also the other 2 groups (P = 0.01); however, no signiWcant
observed (P < 0.01). There was a signiWcant reduction in diVerence in pain intensity (P = 0.39) was observed
pain only during the Wrst 8 weeks of treatment (P < 0.01), between the WA and MS groups (Fig. 2).
although a non-signiWcant trend toward reduced pain was
observed in the MS group (P = 0.54). Fibromyalgia impact questionnaire
EYcacy analysis revealed a signiWcant group-time inter-
action (P = 0.01), indicating diVerent treatment responses Intention-to-treat analysis showed no signiWcant group-
over time. There was a signiWcant reduction in pain inten- time interaction (P = 0.13) regarding FIQ scores. However,
sity in the Wrst 8 weeks of study in the 3 study groups a signiWcant group eVect was observed; FIQ scores were
(P < 0.01), but pain reduction was more pronounced in the higher in the control group than in the WA (P < 0.01) and

123
2290 Rheumatol Int (2012) 32:2285–2292

Fig. 3 FIQ scores over time in

the 3 groups. Intention-to-treat
analysis (left) and eYcacy
(as-treated) analysis (right)

Fig. 4 SF-36 bodily pain scores over time in the 3 groups. Intention-to-treat analysis (left) and eYcacy (as-treated) analysis (right)

MS (P = 0.02) groups, and lower in the WA group than in functioning, and mental health (P < 0.05) in the WA and
the MS group (P < 0.01). With regard to the time eVect, MS groups, and for physical functioning (P < 0.05) and
there were signiWcant diVerences in FIQ scores only role physical only in the WA group. At the 28-week follow-
between baseline and T8 (P < 0.01); no diVerences were up, there was also an increase in SF-36 scores for bodily
found after T8 (P = 0.18). pain and general health (P < 0.05) in the WA and MS
EYcacy analysis revealed a signiWcant reduction in FIQ groups, and for physical functioning and role physical
scores in the Wrst 8 weeks in the WA (P < 0.01) and MS (P < 0.05) only in the WA group (Fig. 4).
(P < 0.01) groups. In the WA group, FIQ scores remained sta-
ble until the 28-week follow-up (T28). There was also a reduc- Use of medication
tion in FIQ scores in the MS group between weeks 8 and 16
(P = 0.04); however, an increase in FIQ scores was observed From a total of 90 participants, 51 (56.6%) restarted medi-
at week 28 (P = 0.08). There were no signiWcant diVerences in cation (tricyclic antidepressants) to control Wbromyalgia
FIQ scores between groups at T8, T16, and T28 (Fig. 3). symptoms, of which 14 (46.7%) were in the WA group,
13 (41.4%) in the MS group, and 24 (80%) in the control
36-item short form health survey group. The analysis of the time medication was restarted
showed an early need for resumption of medication in the
Intention-to-treat analysis revealed a signiWcant increase in control group compared with the other 2 groups (log-rank
SF-36 scores for bodily pain (P < 0.01), general health, and test, P = 0.01; chi-square test, P < 0.01).
vitality (P < 0.05) in the WA and MS groups, for physical
functioning and mental health (P < 0.05) only in the WA
group, and for social functioning (P < 0.05) only in the MS Discussion
group. At the 28-week follow-up, there was also an
increase in SF-36 scores for bodily pain and vitality Our results revealed that muscle-strengthening exercises
(P < 0.05) in the WA and MS groups, and for physical were as eVective as the walking program regarding pain
functioning, role emotional and mental health (P < 0.05) relief, reduction in the impact of Wbromyalgia symptoms,
only in the WA group. and improvement of health-related quality of life. MS
EYcacy analysis showed a signiWcant increase in SF-36 showed to be a secure modality of physical activity and a
scores for bodily pain (P < 0.01), vitality (WA group, therapeutic option to be recommended for patients with
P < 0.05; MS group, P < 0.01), general health, social Wbromyalgia.

123
Rheumatol Int (2012) 32:2285–2292 2291

Medication alone is not generally suYcient for the treat- which is in agreement with the Wndings of other studies [10,
ment of Wbromyalgia, because the restoration of physical 18]. However, 28-week follow-up results showed a recur-
and emotional functioning, as well as quality-of-life rence of Wbromyalgia symptoms in the MS group.
improvement are also important factors to be considered. In In the WA and MS groups, but not in the control group,
this way, physical exercise programs have been used for there was a within-group increase in SF-36 scores over time
decades as an integral part of the treatment of Wbromyalgia for all subscales, with statistical signiWcance for bodily
[1, 5, 8]. In order to test the actual eVectiveness of exercises pain, general health, social functioning, and mental health,
in relieving pain in patients with Wbromyalgia, all medica- indicating an improvement in the quality of life of patients
tion was stopped for a 30-day washout period before the in the WA and MS groups, similar to the Wndings of Kingsley
supervised exercise program started. et al. [17]. However, the short study period of 8 weeks and
Few studies have examined the impact of single modali- lack of follow-up data in the latter study did not allow the
ties of physical exercise in patients with Wbromyalgia, mak- comparison of our results with those of Kingsley et al. [17],
ing it diYcult to determine which modality is in fact especially regarding the late eVects of the proposed
eVective in improving the symptoms. Only one study was exercise programs. The improvement in health-related
found in the literature comparing the eVects of aerobic quality of life was also observed at the 28-week follow-up
training with those of a program of muscle strengthening assessment (T28). Patients in the WA group reported higher
carried out as isolated interventions [17]. But it is diYcult SF-36 scores (improvement) in role physical, social func-
to compare our Wndings with those reported in the litera- tioning, and role emotional at T28, probably because of the
ture, because the study protocols are diVerent. adoption of physical activity and social interaction among
When comparing our results with those of previous the participants [22] during the study period, which may
studies on the eVects of MS on patients with Wbromyalgia, have contributed to these positive results.
we observed that the size of our sample was larger, as well Patients attended the exercise program in small groups,
as the follow-up period used in the study (28 weeks includ- enabling proper supervision. Care was taken in the design of
ing the treatment period). Moreover, most of the other the study protocol [18]; all patients underwent clinical exami-
studies involved programs using exercise machines [14, 16] nation at monthly intervals; and medication for Wbromyalgia
and strength protocols with relative load, based on a dose- was restarted if necessary. Patients in the WA and MS groups
escalating percentage of 1 repetition maximum (1RM). On took less medication, which is consistent with the Wndings of
the other hand, few studies had used free weights or resis- other study, which had reported that the aerobic exercise group
tance bands in strengthening exercises [15, 17]. took fewer analgesics than did the controls [12].
As expected, this study showed that both treatments Patients in both intervention groups were encouraged to
were eVective in reducing pain during the supervised exer- continue the exercise program after the end of the 16-week
cise program, which is consistent with the Wndings of other supervised treatment. However, those in the MS group had
studies [15–17, 19]. However, by 3-month post-interven- diYculties in incorporating the exercise protocol into the
tion, 28-week follow-up results showed no persistent daily routine. The reasons given by the patients were lack
reduction in pain [28]. During the treatment period, the of equipment at home and diYculty in remembering the
mean percentage of pain relief was 40.81% in the WA exercises. The absence of supervision may give rise to fears
group, 45.38% in the MS group, and 22.71% in the control that an inappropriate exercise could induce pain [28].
group, according to intention-to-treat analysis. No statisti- Deconditioning eVects were more pronounced in the MS
cally signiWcant diVerences in pain relief were found group and may be linked to poor adherence to the exercise
between the WA and MS groups, which is consistent with program after the end o the supervised treatment, leading to
the Wndings of other studies [15, 17], although there was a a more rapid loss of adaptation [20].
non-signiWcant trend toward pain improvement in the MS On the other hand, patients in the WA group did not
group between weeks 8 and 16. These Wndings indicate that have these concerns. For these patients, walking became a
the improvement in Wbromyalgia symptoms due to exer- habit and part of their daily routine.
cises was probably related to the combat of eVects of It was expected that the reasons given by the patients for
deconditioning [29]. not completing the study, especially by those in the MS
Apparently, the MS group had a better response to the group, who left the study in the early stages, would indicate a
progressive increase in workload, which was part of the worsening of pain or fear of exercise-induced pain. However,
diVerent exercise protocols. Both intervention (WA and the reasons for leaving the study included starting a job or the
MS) groups showed a non-signiWcant trend toward lower need to care for a family member, which are the reasons
FIQ total scores (improvement in physical functioning) related to the low socioeconomic status of the participants.
throughout the 16-week treatment. Although not signiW- Our study has some limitations. We initially intended to
cant, this trend was not observed in the control group, use a larger study population, but because the exercise

123
2292 Rheumatol Int (2012) 32:2285–2292

sessions were held 3 times per week, some patients decided 11. van Santen M, Bolwijn P, Landewe R, Verstappen F, Bakker C,
not to participate. The interventions involved simple exercise Hidding A et al (2002) High or low intensity aerobic Wtness train-
ing in Wbromyalgia: does it matter? J Rheumatol 29:582–587
programs, and all patients were instructed during the 12. Jones KD, Burckhardt CS, Deodhar AA, Perrin NA, Hanson GC,
16 weeks of training to continue the exercise program at Bennett RM (2008) A six-month randomized controlled trial of
home after the end of the study. However, the patients found exercise and pyridostigmine in the treatment of Wbromyalgia.
it diYcult to enroll in exercise programs in the community Arthritis Rheum 58:612–622. doi:10.1002/art.23203
13. Martin L, Nutting A, MacIntosh BR, Edworthy SM, Butterwick D,
and were not motivated to continue the programs at home. Cook J (1996) An exercise program in the treatment of Wbromyalgia.
Another limitation of the study is that we did not assess J Rheumatol 23:1050–1053
changes in strength or cardiorespiratory capacity of the par- 14. Hakkinen A, Hakkinen K, Hannonen P, Alen M (2001) Strength
ticipants over time to evaluate whether exercise intensity and training induced adaptations in neuromuscular function of
premenopausal women with Wbromyalgia: comparison with healthy
response to physical training were associated with improve- women. Ann Rheum Dis 60:21–26. doi:10.1136/ard.60.1.21
ment in Wbromyalgia symptoms and pain relief. 15. Jones KD, Burckhardt CS, Clark SR, Bennett RM, Potempa KM
In conclusion, there is as yet no consensus on which is (2002) A randomized controlled trial of muscle strengthening versus
the most eVective exercise intervention to reduce pain. Our Xexibility training in Wbromyalgia. J Rheumatol 29:1041–1048
16. Kingsley JD, Panton LB, Toole T, Sirithienthad P, Mathis R, McMil-
results revealed that both exercise modalities (WA and MS) lan V (2005) The eVects of a 12-week strength-training program on
provided better pain relief in patients with Wbromyalgia strength and functionality in women with Wbromyalgia. Arch Phys
than medication alone or conventional treatment, which is Med Rehabil 86:1713–1721. doi:10.1016/j.apmr.2005.04.014
in agreement with other studies. Individual preferences and 17. Bircan C, Karasel SA, Akgün B, El O, Alper S (2008) EVects of
muscle strengthening versus aerobic exercise program in Wbromy-
the general condition of the patient should be considered algia. Rheumatol Int 28:527–532. doi:10.1007/s00296-007-0484-5
when prescribing an exercise program. 18. Schachter CL, Busch AJ, Peloso P, Sheppard MS (2003) EVects of
short versus long bouts of aerobic exercise in sedentary women with
ConXict of interest The authors declare that they have no conXict of Wbromyalgia: a randomized controlled trial. Phys Ther 83:340–358
interest. 19. Valim V, Oliveira L, Suda A, Silva L, de Assis M, Barros NT et al
(2003) Aerobic Wtness eVects in Wbromyalgia. J Rheumatol 30:1060–
1069
20. Pollock ML, Gaesser GA, Butcher JD, Després JP, Dishman RK,
References Franklin BA, Garber CE (1998) ACSM position stand: the recom-
mended quantity and quality of exercise for developing and main-
1. Busch AJ, Barber KA, Overend TJ, Peloso PMJ, Schachter CL taining cardiorespiratory and muscular Wtness, and Xexibility in
(2007) Exercise for treating Wbromyalgia syndrome. Cochrane healthy adults. Med Sci Sports Exerc 30:975–991
Database Syst Rev 4:CD003786 21. Borg GAV (1982) Psychophysical bases of perceived exertion.
2. Wolfe F, Smythe HA, Yunus MB, Bennett RM, Bombardier C, Med Sci Sport Exerc 14:377–381
Goldenberg DL et al (1990) The American College of Rheumatol- 22. Jones KD, Clark SR (2002) Individualizing the exercise prescrip-
ogy 1990 criteria for the classiWcation of Wbromyalgia. Report of tion for persons with Wbromyalgia. Rheum Dis Clin North Am
the multicenter criteria committee. Arthritis Rheum 33:160–172. 28:419–436
doi:10.1002/art.1780330203 23. Marques AP, Assumpção A, Matsutani LA, Pereira CAB, Lage L
3. Russel IJ, Orr MD, Littman B, Vipraio GA, Alboukrek D, (2008) Pain in Wbromyalgia and discrimination power of the
Michalek JE et al (1994) Elevated cerebrospinal Xuid levels of instruments: visual analog scale, dolorimetry and the McGill pain
substance P in patients with the Wbromyalgia syndrome. Arthritis questionnaire. Acta Reumatol Port 33:345–351
Rheum 37:1593–1601. doi:10.1002/art.1780371106 24. Burckhardt CS, Clarck SR, Bennett RM (1991) The Wbromyalgia
4. Roizenblatt S, Moldofsky H, Benedito-Silva AA, TuWk S (2001) impact questionnaire: development and validation. J Rheumatol
Alpha sleep characteristics in Wbromyalgia. Arthritis Rheum 18:728–733
44:222–230. doi:10.1002/1529-0131(200101)44:1<222:AID- 25. Ciconelli RM, Ferraz MB, Santos W, Meinão I, Quaresma MR
ANR29>3.0.CO;2-K (1999) Tradução para a língua portuguesa e validação do ques-
5. Mannerkorpi K, Henriksson C (2007) Non-pharmacological tionário genérico de avaliação de qualidade de vida SF-36 (Brasil
treatment of chronic widespread musculoskeletal pain. Best SF-36). Rev Bras Reumatolol 39:143–150
Pract Res Clin Rheumatol 21:513–534. doi:10.1016/j.berh. 26. Farrar JT, Young J-PJ, LaMoreaux L, Werth JL, Poole RM (2001)
2007.04.001 Clinical importance of changes in chronic pain intensity measured
6. Clark SR (1994) Prescribing exercise for Wbromyalgia patients. on an 11-point numerical pain rating scale. Pain 94:149–158.
Arthritis Care Res 7:221–225. doi:10.1002/art.1790070410 doi:10.1016/S0304-3959(01)00349-9
7. Gowans SE, deHueck A (2004) EVectiveness of exercise in man- 27. Hollis S, Campbell F (1999) What is meant by intention to treat
agement of Wbromyalgia. Curr Opin Rheumatol 16:138–142 analysis? Survey of published randomised controlled trials. BMJ
8. Rooks SD (2007) Fibromyalgia treatment update. Curr Opin 319:670–674
Rheumatol 19:111–117. doi:10.1097/BOR.0b013e328040bffa 28. Ramsay C, Moreland J, Ho M, Joyce S, Walker S, Pullar T (2000)
9. McCain GA, Bell DA, Mai FM, Halliday PD (1988) A controlled An observer-blinded comparison of supervised and unsupervised
study of the eVects of a supervised cardiovascular Wtness training aerobic exercise regimens in Wbromyalgia. Rheumatology 39:501–
program on the manifestations of primary Wbromyalgia. Arthritis 505. doi:10.1093/rheumatology/39.5.501
Rheum 31:1135–1141. doi:10.1002/art.1780310908 29. Simms RW, Roy SH, Hrovat M, Anderson JJ, Skrinar G, LePoole
10. Richards SC, Scott DL (2002) Prescribed exercise in people with SR et al (1994) Lack of association between Wbromyalgia
Wbromyalgia: parallel group randomised controlled trial. BMJ syndrome and abnormalities in muscle energy metabolism. Arthri-
325:185–189. doi:10.1136/bmj.325.7357.185 tis Rheum 37:794–800. doi:10.1002/art.1780370603

123