Article Text


Acupuncture for fibromyalgia in primary care: a randomised controlled trial
  1. Jorge Vas1,2,
  2. Koldo Santos-Rey1,
  3. Reyes Navarro-Pablo3,
  4. Manuela Modesto1,
  5. Inmaculada Aguilar1,
  6. M Ángeles Campos4,
  7. José Francisco Aguilar-Velasco5,
  8. Milagrosa Romero4,
  9. Patricia Párraga5,
  10. Vanesa Hervás3,
  11. Olalla Santamaría3,
  12. Carmen Márquez-Zurita1,
  13. Francisco Rivas-Ruiz2,6
  1. 1Pain Treatment Unit, Doña Mercedes Primary Health Centre, Dos Hermanas, Spain
  2. 2Red de Investigación en Servicios de Salud en Enfermedades Crónicas (REDISSEC), Andalucía, Spain
  3. 3Mental Health Services, Virgen de Valme University Hospital, Sevilla, Spain
  4. 4Primary Health Centre Las Cabezas de San Juan, Las Cabezas de San Juan, Spain
  5. 5San Francisco Primary Health Centre, Moron, Spain
  6. 6Research Unit, Agencia Sanitaria Costa del Sol, Marbella, Spain
  1. Correspondence to Dr Jorge Vas, Pain Treatment Unit, Doña Mercedes Primary Health Centre, Dos Hermanas, Sevilla 41700, Spain; jorgef.vas.sspa{at}


Objectives To evaluate the efficacy of an individualised acupuncture protocol for patients with fibromyalgia.

Methods Randomised controlled multicentre trial, blinded to participants and to data analysts. Conducted in three primary care centres in southern Spain. A total of 164 participants aged over 17 years and diagnosed with fibromyalgia were enrolled in this trial; 153 participants completed the study. Participants were randomly assigned to either the real intervention (individualised acupuncture, IA) or the sham intervention (sham acupuncture, SA). In both the IA and SA groups, one session per week (lasting 20 min) was provided, in addition to usual pharmacological treatment. The primary outcome was change in pain intensity at 10 weeks.

Results Intention-to-treat analysis revealed that the decrease in pain intensity at 10 weeks was greater (p=0.001) in the IA group (−41.0%, 95% CI −47.2% to −34.8%) than in the SA group (−27.1%, 95% CI −33.2% to −20.9%). During the follow-up period, significant differences (p<0.01) in favour of the IA group persisted at 12 months (IA: −19.9%, 95% CI −24.6% to −15.1%; vs SA: −6.2%, 95% CI −11.2% to −1.2%).

Conclusions Individualised acupuncture treatment in primary care in patients with fibromyalgia proved efficacious in terms of pain relief, compared with placebo treatment. The effect persisted at 1 year, and its side effects were mild and infrequent. Therefore, the use of individualised acupuncture in patients with fibromyalgia is recommended.

Trial registration number ISRCTN60217348.


Statistics from


Fibromyalgia affects 2–5% of the population1 ,2 and has a high economic and social impact.3 ,4 It is classified as a functional chronic disease, and is manifested clinically by chronic widespread pain associated with fatigue, sleep disorder and/or depression, among other symptoms.5

At present, there is no laboratory test or imaging technique to help establish a diagnosis of fibromyalgia, and no structural abnormalities have been observed in the musculoskeletal system of patients with fibromyalgia that could justify its diagnosis.6 The American College of Rheumatology (ACR) proposed new diagnostic criteria in 2010, examined in a 31-point questionnaire.7 These criteria were subsequently modified in 2011,8 and replaced the original criteria published in 1990.9

Recent systematic reviews support the use of pregabalin to treat pain, sleep problems, fatigue and anxiety in patients with fibromyalgia.10 ,11 Positive results have also been reported in studies of serotonin and norepinephrine reuptake inhibitors12 ,13 and of tramadol.14 However, the benefits of drug therapy are questionable.15

Ninety-one per cent of patients with fibromyalgia have tried some form of complementary and alternative medicine.16 Although some treatment guidelines recommend prioritising non-pharmacological interventions,17 and recent systematic reviews suggest that both cognitive behavioural therapy18 and approaches such as massage, hydrotherapy and acupuncture may be beneficial in the treatment of fibromyalgia,19 evidence of benefit is limited.15

One in five patients with fibromyalgia resorts to acupuncture within 2 years of diagnosis.20 The results of a recent systematic review concluded that there is low to moderate evidence for the use of acupuncture to treat fibromyalgia pain and stiffness, compared to standard treatments.21 Moreover, most randomised controlled trials performed to date have not included personalised treatment protocols, which is in contrast to usual clinical practice.22 Furthermore, acupuncture appears to be beneficial in the treatment of depression, whether or not it is associated with chronic pain.23 ,24

Taking the above considerations into account, we designed a randomised controlled trial to determine the clinical efficacy of traditional acupuncture as a treatment for patients with fibromyalgia, using a personalised point-selection algorithm based on the characteristics of each patient. In this respect, we sought to reproduce standard clinical practice. We aimed to evaluate the progression of the disease using a specific measurement system, following the OMERACT (Outcome Measures in Rheumatoid Arthritis Clinical Trials) recommendations,25 and to assess levels of depression and pain sensation, quality of life, duration of short-term disability, use of analgesic and antidepressant medication, and the potential effect of factors considered to be predictors of poor prognosis.26


This study was performed in accordance with the trial protocol, which has been published elsewhere.27

Study design and participants

In this controlled multicentre prospective trial, all participants were aged over 17 years and had been diagnosed with fibromyalgia according to ACR criteria.9 They had not received acupuncture before and were referred by their general practitioner (GP) at one of the three primary care centres included in the trial (forming part of the Andalusian Public Health System) to the physician leading the study at the health centre. The participants were randomly assigned to the actual intervention group (individualised acupuncture, IA) or to the sham intervention group (sham acupuncture, SA). Participants in both groups also received pharmacological treatment prescribed by their GP. All participants gave written informed consent. Participants were excluded if they had chronic pain in relation to any process other than fibromyalgia, were using anticoagulants or opiates, were pregnant or a nursing mother, or were involved in occupational litigation for reasons involving fibromyalgia. The recruitment of participants began in October 2010 and concluded in December 2012.

Randomisation and masking

After the participants had given informed consent, a secure centralised telephone and fax randomisation service was used to allocate them to one of the two arms of the study. The randomisation sequence was generated and administered at an independent clinical trials unit (Marbella, Spain) to ensure that the allocation was concealed. Randomisation was on a 1:1 basis, stratified by centre. Given the nature of the interventions conducted, the healthcare staff involved were not blind to allocation, but all of the evaluations were recorded by research nurses who were blind to the treatment groups. These same nurses provided any assistance required with the self-administered questionnaires. The specialist psychiatrists applying the Hamilton Depression Rating hetero-evaluation Scale (HDRS) were also blind to the allocation, as were the staff responsible for performing the statistical analysis.


The healthcare personnel who applied the different acupuncture treatments at each centre involved in the study had received at least 300 h training in the field and had over 3 years’ practical experience.

The participants in the study received nine 20 min sessions of acupuncture (real or simulated) at a frequency of one per week, in addition to their usual drug treatment. All participants were diagnosed according to the principles of Traditional Chinese Medicine (TCM) by the medical expert at the healthcare centre.

The true acupuncture group received IA based on their TCM diagnosis and according to a previously established algorithm.27 The SA group received an acupuncture simulation on the dorsal and lumbar regions, in which guide tubes for the same type of needle as used in the real acupuncture group were applied to the body surface, but after removal of the needles.27 The time devoted to the treatment and evaluation of participants was identical in both groups.


The data required to evaluate the efficacy of the treatment were collected by interviews, self-applied questionnaires, and physical measurements at baseline (T0) and at 10 weeks (T1), 6 months (T2), and 12 months (T3) after the intervention had begun, in both groups.

The primary outcome was change in pain intensity, measured on a visual analogue scale (PainVAS 0–100 mm), at 10 weeks. Secondary outcomes included change in level of depression, measured on the HDRS,28 at 10 weeks and at 6 months; change in health-related quality of life as measured by the two summary components (physical and mental) of the Short Form 12 (SF-12) health questionnaire,29 at 10 weeks, 6 and 12 months; change in the Fibromyalgia Impact Questionnaire (FIQ),30 at 10 weeks, 6 and 12 months; change in PainVAS at 6 and 12 months; and change in the pain threshold and number of tender points, at 10 weeks, 6 and 12 months. In addition, we measured the improvement perceived by the participant31 at 10 weeks, and changes in use of medication at 10 weeks, 6 and 12 months. A cut-off point of HDRS ≥21 was established to evaluate the effect on participants with medium or severe depression.32 In order to evaluate the influence of family structure, a semi-structured interview, incorporating the genogram technique, was performed.27

Statistical plan

The χ2 test (for binary and ordinal variables) and the Mann-Whitney U test (for continuous variables, after confirming skewed distributions using the Kolmogorov-Smirnov test) were used to assess the differences between the groups, for both independent and outcome variables. In the latter case, mean differences in the relative changes at T1, T2 and T3 compared to baseline (T0) between the treatment groups were calculated and compared, both per protocol and by intention-to-treat (ITT). For ITT, missing values were imputed using simple linear regression with baseline values. In addition, the effect size of the ITT analysis was calculated using Cohen's d,33 based on standardised mean differences. Finally, a stepwise forward multiple linear regression model was constructed, taking the main outcome variable (relative change in pain intensity T0-T1), and adjusting for age, baseline disease severity (PainVAS, FIQ and HDRS), pain threshold, and tolerance levels. The level of statistical significance applied was p<0.01.


Of the 164 participants included in the study, 162 (98.8%), all female, constituted the final baseline sample. Two withdrew before the start of treatment, one due to comorbidity (acute myocardial infarction) and the other due to a change of address. Thus, the study comprised 80 participants in the IA group and 82 in the SA group (figure 1).

Figure 1

Flow chart of the trial.

There were no differences between groups in the distribution of any demographic and baseline clinical variables (table 1). The most frequently diagnosed TCM syndromes were Liver Qi stagnation (44.4%), followed by Spleen and Kidney Yang deficiency (21%) and Yin deficiency (13.6%). In 85.8% of the participants, a secondary associated syndrome was also diagnosed. There were no differences between the groups regarding the distribution of the primary diagnosed syndromes or the presence of a second syndrome (see online supplementary file 1). A total of 114 participants (70.4%) responded to the ‘family structure’ questionnaire item. There were no significant differences between the groups for any of the variables related to family structure or relationships, according to the results obtained using the genogram technique (see online supplementary file 2).

Table 1

Descriptive analysis of baseline clinical and sociodemographic variables for the two treatment groups

The crude analysis of the main study variable (change in PainVAS T0-T1), both per protocol (see online supplementary file 3) and by ITT (p=0.001), revealed a relative pain intensity reduction that was greater in the IA group (−41.2%) than in the SA group (−27.1%). The differences in favour of the IA group persisted in the follow-up period, relative to baseline, and at T3 (ITT: IA −19.9% vs SA −6.1%; p≤0.001). For the FIQ scale, in the three evaluations made, there were differences in favour of the IA group, with ITT values of relative change at T1, T2 and T3 of −35.9%, −25% and −22.2%, respectively, for the IA group, and of −24.5%, −11.2% and −4.9%, respectively, for the SA group. On the physical scale of the SF12, and also with respect to the pressure pain threshold, there was a relative improvement in favour of the IA group at all time points assessed. Finally, the reduction in the number of tender points was higher in the IA group for both the T0-T1 and the T0-T3 comparisons. Of a total of 20 tests, and after calculating the Cohen’s d values, 12 medium-strength effect sizes and one strong effect size were observed (table 2).

Table 2

Evaluation of outcome measures

The per protocol analysis of the various items included in the FIQ (fatigue, night rest, stiffness, anxiety and depression) revealed differences in favour of the IA group. The improvement observed at 10 weeks after starting treatment (T1) was significant for the following items: fatigue (IA −32.8 vs SA −19.0, p<0.001), anxiety (IA −31.0 vs SA −17.8, p=0.01), and depression (IA −33.0 vs SA−18.5, p=0.01). During follow-up, statistically significant differences in favour of the IA group were observed for all items mentioned, both at 6 months (T2) and 12 months (T3) (table 3).

Table 3

Relative change (with respect to baseline measurement) in FIQ items concerning fatigue, night rest, stiffness and depression

In the adjusted regression analysis of the primary dependent variable (relative difference in pain intensity T0-T1), after adjusting for the number of tender points (T0) and the total FIQ score (T0), participants in the IA group presented a reduction in pain intensity of −14.4 versus the SA group (table 4).

Table 4

Multivariate linear regression model evaluating difference in relative pain intensity (PainVAS T0-T1)

In the assessment of the blinding process, over 75% of the participants in both groups were confident that the treatment received was real acupuncture. Finally, in the four questions concerning treatment expectations, the scores exceeded 8.5 points, and there were no significant differences between the two groups (see online supplementary file 4).

No serious adverse events arising from the study interventions were reported in either group. The following mild adverse events occurred during the 1431 sessions received by the total number of participants: aggravation of the symptoms of fibromyalgia, after 37 sessions (2.6%); headache, after 7 sessions (0.5%). The following adverse events were exclusive to the IA group (702 sessions): post-acupuncture pain, after 12 sessions (1.4%), post-acupuncture bruising, after 18 sessions (2.6%); and post-acupuncture vagal symptoms, after 5 sessions (0.7%). Online supplementary files 5 and 6 describe the impact of treatment on use of medication and fitness for work.


The results of this study show that the application of an individualised acupuncture protocol significantly reduced perceptions of pain intensity, and at the same time improved the functional capacity and quality of life of patients with fibromyalgia, compared to placebo treatment. Both methods were safe, as indicated by the low number of adverse events recorded.

Perception of pain intensity

We observed a significant improvement in all indicators related to pain perception (PainVAS, tender points, pressure pain threshold) in the intervention group compared to the control group, both at the end of treatment and during the follow-up period. Although both real and sham acupuncture were effective at improving the indicators, the effect of real acupuncture was significantly greater, both at the end of treatment and during the 1-year follow-up period. According to Cohen's categories, the effect of acupuncture compared to placebo on pain intensity, number of tender points, and pressure pain threshold was of medium strength. This effect size was attenuated at 6 months (T2) for pain perception, at which point it became a low-intensity effect. It is noteworthy that at 12 months (T3) this effect persisted, and even increased in the case of the PainVAS variable. We should view the T3 results with caution, as there was an increased use of antidepressants in both groups that may account for the increased size of effect on the PainVAS variable, among others, at this stage.

According to the IMMPACT (Initiative on Methods, Measurement and Pain Assessment in Clinical Trials) recommendations concerning the interpretation of clinical trials for the treatment of chronic pain,34 the clinical improvement secondary to the reduction of pain intensity experienced by the participants given real acupuncture treatment was moderately important.

Functionality and quality of life/sense of wellbeing

Both the functionality (FIQ) and the quality of life/sensation of wellbeing (SF12) of the participants in the real acupuncture group were significantly better than in the control group at the end of the treatment period. The effect size was greater with respect to the physical aspect of quality of life (medium intensity) than for the mental aspect (low intensity), although this effect persisted during the follow-up period. With respect to functionality, the effect size was, in fact, greater at the end of the follow-up period than at previous time points. This may be due to an increase in the amount of antidepressant medication taken by participants at T3, as mentioned above, which may have accounted for the improvement in the FIQ scores.


The individualised protocol applied in this study did not prove to be efficacious in alleviating the level of depression of the participants. Although the HDRS scores decreased in both groups by the end of the treatment period, the difference was not statistically significant and the minimum HDRS reduction of 25% considered to constitute a response was not reached.35 However, at T3, a significant decrease in the Hamilton test scores was observed, and a low-intensity effect size was obtained when comparing the two groups. As commented on above, this result is probably due to the increased quantity of antidepressants taken.

Comparison with previous studies of acupuncture

In this study, we observed reductions of 41% and 36% on the pain intensity scale and FIQ scale, respectively, at the end of the treatment period. These values are higher than those reported for the same variables in previous studies.36–39 In our study, the strength of the effect size observed at the end of the treatment period was notably greater (0.5 vs 0.14) than that reported in a recent systematic review that took into account major clinical trials of acupuncture treatment for fibromyalgia.21

The observation herein that the positive effects of acupuncture on pain reduction, functional capacity, and quality of life were maintained both at 6 and 12 months is in contrast to previous similar studies.36–38 One fundamental difference between the present study and earlier research is the implementation of individualised treatment that was reassessed throughout the intervention period. This may explain the larger size and/or longer duration of the effect, although we cannot be certain about this as there was no non-individualised acupuncture comparator group in the present study.

Comparison with previous studies of pharmacological treatment

When comparing the results of this study with those reported by systematic reviews of previous clinical trials of various pharmacological treatments for fibromyalgia, we observe, firstly, that the effect size obtained herein was greater than that of trials evaluating drugs considered first-line treatment for fibromyalgia, such as pregabalin (0.5 vs 0.28)10 or duloxetine (0.5 vs 0.23).40 Moreover, the acupuncture treatment described in this study was associated with a lower rate of adverse effects than drug treatments used in clinical practice.

Strengths and limitations

The main tangible advantage of our study is that we carried out a clinical trial with blinding of the participants and data analysts and incorporated individualised acupuncture treatment protocols for the treatment of fibromyalgia. This treatment produced an improvement in the participants’ condition, reflected by a reduction in pain intensity and enhanced functional capacity and quality of life after the intervention and during the follow-up period. Such an outcome has not been reported by previous studies following the application of standardised treatments; therefore, our results suggest that applying individualised treatment algorithms when starting a course of acupuncture may be important.

The present study has several limitations. Firstly, the diagnostic criteria used were not the latest ones proposed by the ACR,7 but those published in 1990,9 because the latter were the most current criteria at the time when the study was being prepared. Our observation that, at the end of the follow-up period, participants were using a significantly higher quantity of antidepressant medication means that the results obtained for that time point should be interpreted with caution. In addition, under baseline conditions, a higher percentage of participants in the IA group were taking higher doses of pain medication relative to the SA group, which could have magnified the improvement observed in pain scores in this group. Notably, however, this minor imbalance (the significance level of which was above our predefined threshold of p<0.01) was a chance result of the randomisation and is therefore not likely to indicate bias.

Moreover, although at the end of the follow-up period there had been a greater number of withdrawals in the real versus sham acupuncture groups (7 vs 2), the data obtained from the ITT analysis did not vary greatly from those obtained in subgroup analysis. Although variables such as fatigue, sleep disorders, anxiety or stiffness (symptoms that frequently accompany fibromyalgia) were not studied specifically, the effect on these variables was determined indirectly by taking into account the results for the different FIQ items. In addition, it would have been desirable to obtain a larger number of measurements, in view of the chronic nature of the pain and its possible fluctuations. However, methodologically, this was not possible.


In this study, individualised acupuncture treatment of fibromyalgia in primary care was shown to be efficacious in providing pain relief and enhancing the quality of life of patients. The effect persisted at 1 year, and adverse effects were mild and infrequent. Furthermore, the effect of individualised acupuncture was greater than that of sham acupuncture. The conflicting results between this and previous studies, which have reported negative results, may be accounted for by differences in the application of the technique (individualised vs standardised).


View Abstract

Review history and Supplementary material

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

  • Press release

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors JV conceived the study, designed the study protocol, and sought funding and ethical approval. MM, IA, KS-R, RN-P, MAC, JFAV, MR, PP, VH and OS acquired the data. JV, MM, KS-R, JFAV and MAC designed the individualised acupuncture algorithm for the study. JV, KS-R, FR-R and CMZ analysed and interpreted the data. FR-R provided statistical expertise on this paper and analysed and interpreted the data. JV, KS-R and FR-R drafted the manuscript. MM, IA, RN-P, MAC and JFAV advised on the preparation of the manuscript. All authors read, edited and approved the final version of the manuscript. All authors had full access to all of the data in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis. JV and FR-R are the guarantors.

  • Funding This paper presents independent research funded by the Spanish Ministry of Health and Consumer Affairs (Carlos III Health Institute, project number PI10/00675) and by the Andalusian Public Health System (project number PI0436/09). The funding agencies had no influence on the design of the study, the analysis, or the writing of the paper.

  • Competing interests None declared.

  • Patient consent Obtained.

  • Ethics approval The study was reviewed and approved by the medical ethical committee of the Andalusian Committee for Clinical Trials, following approval by the Research Committee at the Valme Hospital (Sevilla, Spain). All participants gave written informed consent.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.