Effect of amitriptyline vs. physiotherapy in management of fibromyalgia syndrome: What predicts a clinical benefit?MN Joshi, R Joshi, AP Jain
Department of Medicine, MGIMS, Sewagram, Wardha, Maharashtra, India
Context : Fibromyalgia is a chronic disabling condition, and physicians treat it using a number of different treatment modalities. It is not known if one or more of such modalities are better than the others. We compared the efficacy of physiotherapy and amitriptyline in disability reduction in patients of fibromyalgia syndrome in a rural tertiary care hospital in Central India. Design : Open-label alternate patient treatment allocation. Materials and Methods : A six-month follow-up was done to assess the benefit of amitriptyline and physiotherapy for disability reduction in patients with fibromyalgia syndrome. Primary outcome measure was improvement in fibromyalgia impact questionnaire (FIQ) score. Statistical Analysis Used : Predictors of benefit were determined using multivariate logistic regression. Results : A total of 175 outpatients were assigned to either amitriptyline (n=87) or structured physiotherapy (n=88) treatments. There was a significant but similar (P=0.82) improvement in disability in both groups. High FIQ score at baseline and low socioeconomic status scores were significant predictors of benefit. Conclusions : Therapy with amitriptyline or physiotherapy is equally effective in improving outcome in patients of fibromyalgia over a period of six months.
Keywords: Amitriptyline, benefit, fibromyalgia, fibromyalgia impact questionnaire score, physiotherapy
Fibromyalgia syndrome is a non-articular rheumatic syndrome disorder comprising chronic widespread pain associated with variable symptoms.  It is a fairly common clinical entity, which is still poorly understood. Fibromyalgia is a major cause of morbidity and economic loss in the community and has a significant impact on patient's quality of life. Numerous studies assessing the resource utilization in rheumatic patients have found that patients of fibromyalgia have a multifold increased risk of lower wellbeing with a substantial cost burden. ,, The diagnosis of fibromyalgia is clinical, based on the American College of Rheumatology (ACR) criteria.  However, the pathogenesis, and clinical profile of fibromyalgia are poorly understood, and it frequently remains undiagnosed. 
A number of treatment options for fibromyalgia have been recommended, which include antidepressants, analgesics, exercise, cognitive behavioral therapy and other non-pharmacological interventions. ,, Of these options, antidepressants, analgesics and exercise have been studied in well-designed randomized control trials. Based on these trials, it has been found that there is a strong evidence favoring the use of amitryptiline and other tricyclic antidepressants. , Recently, the US FDA has approved the use of pregabalin,  duloxetine  and milnacipran  for patients of fibromyalgia.  Exercise in the form of cardiovascular training, strength training, aerobics, flexibility training, all have led to improvement in patients of fibromyalgia. ,
Most of these studies have been done in the developed countries, and there is paucity of data about the burden of fibromyalgia, and performance of treatment options in developing countries. This is further complicated by the fact that most of the drugs advocated for treatment are expensive, need to be consumed for prolonged periods and hence are likely to be unaffordable in rural central India. Non-pharmacological modalities provide a reasonable alternative in this scenario. We conducted this study with an aim to compare the performance of physiotherapy and amitriptyline for disability reduction in patients with fibromyalgia. We also aimed to determine which clinical features at baseline would predict benefit with either therapy.
The study was conducted at our institute, tertiary care hospital located in rural Central India. All outpatients presenting to the hospital are screened by doctors in general outpatient department (GOPD), where after a brief history and examination, appropriate referrals to specialists are advised. For the purpose of the study, doctors in GOPD were requested to advise patients with chronic muscular pain, of three months or greater in duration, to meet the study investigators.
The investigators screened all patients who were referred to them for inclusion in the study. The inclusion criteria being age between 18 and 60 years, symptoms of chronic muscular pain of at least 12 weeks' duration, and fulfillment of ACR criteria (presence of tenderness in at least 11/18 trigger points) for the diagnosis of fibromyalgia syndrome (FMS).  Patients were excluded if they were pregnant or lactating, or had history of trauma, fractures, fever, malignancy, chronic renal or hepatic disorders, alcohol abuse, cerebrovascular or a neurological abnormality. Patients were allowed to continue their previous medications and exercise regimens, if any. The study was approved by the institutional ethics committee and participants were enrolled after obtaining a written informed consent.
The patients included in the study were administered a standardized questionnaire to record socio-demographic characteristics, symptom characteristics and history of previous visits to a healthcare provider. The socioeconomic score (SES) was determined using a validated questionnaire.  Briefly, the questionnaire consists of following: (1) household profile, (2) material possession profile, (3) educational profile, (4) occupational profile, (5) economic profile (6) possessed land or house cost profile and (7) social profile. Based on the average score in each of the above domains, SES was categorized into five grades: Grade 1: 57-70 points; Grade 2: 42-57 points; Grade 3: 27-42 points; Grade 4: 12- 27 points; and Grade 5: 0-12 points. Individuals in Grades 1 to 3 were considered as having a higher SES.
Fibromyalgia impact questionnaire (FIQ) was used to determine a baseline score.  For the purpose of the study this questionnaire was translated into the local language (Marathi), and back-translated to English to ensure its accuracy. Minor modifications were done in the questionnaire to make it suitable for our setting (for example, the use of washer or dryer for laundry was modified to washing the laundry by themselves). Assessment of psychiatric co-morbidity was done using the Brief Psychiatric Rating Scale (BPRS), which is an 18-item instrument, with each item having seven severity grades.  A score of 32 or more indicates presence of psychopathology. 
The study investigator (MJ) was trained in the administration of these questionnaire-based instruments, and pilot tested them on a subset of patients before initiating the study. In addition erythrocyte sedimentation rate (ESR) and hemoglobin levels were obtained in all patients to determine the presence of anemia and collagen vascular diseases. We designed an open-label alternate patient treatment allocation strategy to observe and compare the performance of physiotherapy and amitriptyline in these patients. This observational study design was chosen as a pilot, and the study was open-labeled as it is difficult to conceal the treatments received. A trained physiotherapist conducted a uniform structured physical training and aerobic session for one group of patients. The patients were advised to perform the exercises daily twice for at least 10 min. There was a step-up pattern of exercise regimen followed by relaxation, stretching and strengthening techniques. The methodology is described in Box 1. The second group of patients received antidepressant in the form of open-label amitriptyline 25 mg once daily at bedtime. The dose of the drug was increased to 50 mg if no benefit was seen. No patient required further escalation of doses. All patients were also offered pharmacologic treatment with 50 mg tramadol in thrice daily doses and as required. The treatment recommendations were based on standard recommendations on the management of fibromyalgia. 
All patients were followed up at monthly intervals for a period of six months. At each visit the study investigators determined the course of the disease and compliance with therapy. If patients adhered to [Additional file 1] the prescribed therapies more than or equal to 2/3 days, they were considered to be compliant. The FIQ score was assessed at the end of the six monthly visits. Reduction in FIQ score from the baseline was used as the principle outcome measure.
All data was double-entered using MS Excel. We performed a descriptive statistical analysis of the baseline data. The analyst (RJ) blinded to the treatment allocation compared the outcome measure (FIQ score at six months and reductions in FIQ score over six months) for the two groups using student's t-test (significance level 0.05). In the post hoc analysis we defined benefit as reduction in FIQ score of 2 standard deviations or more from the mean reduction over six months. We performed a univariate and a multivariate logistic regression to determine the demographic and clinical variables which predicted benefit. All statistical analysis was done using STATA version 9.0 (STATA Corp, College Road, Texas, USA).
Between 1 March 2006 and 1 December 2007 a total of 324 patients were screened by the investigators and 175 (mean age: 38.8 6 9.3 with, 95% females) were included in the study. Amongst them, 19 patients (five in the amitriptyline group, and 14 in the physiotherapy group) were lost to follow-up and were not included in further analysis [Figure 1]. Hence the study sample consisted of 156 patients (82 in amitriptyline and 74 in physiotherapy groups).
Only one patient in the study fulfilled criteria for mild psychopathology, and the range of BPRS scores obtained was from 1 to 12. The range of ESR and hemoglobin levels was from 8 to 110 and 5.5 to 14, respectively. The baseline characteristics of the study patients in these two groups are described in [Table 1]. There was no statistically significant difference between the two groups in terms of baseline characteristics.
On follow-up, 65 (79.2%) patients in the amitriptyline group and 66 (89.2%) patients in the physiotherapy group reported adherence to the prescribed therapies for more than 75% days of the follow-up period. Of the 180 days of follow-up, the individuals in the amitriptyline and physiotherapy treatment groups adhered to treatment on an average for 135 (75%) and 141 (78%) days, respectively. Participants consumed narcotic analgesic (tramadol) on an average for 35 (19.4%) and 34 (18.8%) days in the amitriptyline and physiotherapy groups respectively; duration was similar in both groups. None of the patients had any major side-effects to warrant discontinuation of the therapy.
There was a significant improvement in the total FIQ scores at six months in both the treatment groups [Figure 2]. The mean reduction in FIQ score in both treatment groups being 12.37 (95% CI 10.71-14.04) in the amitriptyline group vs. 12.12 (95% CI 10.47-13.77) in the physiotherapy group (P = 0.82). There was no difference in reduction of FIQ scores in the two groups.
The mean reduction in total FIQ score obtained in all patients combined was 12.2 ± SD 7.3 points. A total of 12 (14%) patients in the amitriptyline group, and seven (9%) patients in the physiotherapy group fulfilled the criteria for benefit (reduction in FIQ score by more than 22, which is more than 2 SD above mean). On multivariate analysis, patients who had a lower socioeconomic status and more disability (FIQ > 50) had significant benefit from either therapy [Table 2].
The results of our study suggest that both amitriptyline and physiotherapy strategies significantly reduced disability due to fibromyalgia, as seen in a rural-based tertiary care hospital in central India. Both the strategies were equally effective in reduction of total FIQ score. Low socioeconomic status and high disability at baseline (as measured by FIQ score) were the only parameters that could predict benefit.
Our study has evaluated the effect of two different therapeutic modalities in the treatment of fibromyalgia. No similar study has been undertaken in the developing world. Our study had a few limitations, firstly a hierarchical screening method was used and this could have perhaps resulted in the small number of participants who were referred to the study investigator. As only clinical history was used for exclusion of patients, it could have potentially resulted in over-inclusion. However, in a resource-limited setting it was not feasible to perform laboratory investigations in all patients who were screened. Secondly, one-thirds of our study population was educated up to primary level only, and this could have limited the understanding of questionnaire-based scores. To overcome this we preferred interviewed questionnaire rather than using a self-administered questionnaire. Thirdly, given the nature of interventions, allocation could not be concealed. Lack of allocation concealment raises a high risk of bias. As fibromyalgia is a chronic illness it is likely that study duration was not long enough to demonstrate any significant benefit. The improvement in both the groups could be linked to Hawthorn effect or regression towards the mean. It does not mean that both treatments are effective. This could be a major limitation.
There could be a possibility that the improvement observed in both the groups could, to some extent, be attributed to the use of tramadol. Keeping a control group receiving only tramadol could have helped to assess this possibility. Lastly, the rate of loss to follow-up was relatively important and patients lost to follow-up were excluded in the statistical analyses. Despite tracing them to their residential addresses we could not obtain cooperation and decided to exclude them from the final analysis.
We did not find any co-morbid psychiatric illness based on the questionnaire used. It is likely that we underestimated the prevalence of depression, anxiety which are common associations of the disease. There could have been cultural issues which could have prevented expression of psycho-social concerns. Future studies are required to bring out the psychiatric associations in these patients.
We reviewed similar studies which were done in past to compare the effects of amitriptyline and physiotherapy. We found that a study evaluating the combined benefit of amitriptyline and exercise  showed that only the combination of amitriptyline and exercise had significant improvements at three months.  Recently, Salek et al. found no significant differences between exercise and anti-depressant group vs. amitriptyline only groups in a randomized trial.  None of the studies have evaluated predictors which explain benefit.
Our study indicates that amitriptyline and physiotherapy are equally beneficial in patients of fibromyalgia. Our study is also unique in the respect that we have studied factors which could help to predict response to either therapy. We found that patients with low socioeconomic scale had better response to therapy. This could be due to the fact that these patients have poor access to health facility and feel psychologically better on receiving care. We also found that patients with high disability score had better outcomes, probably due to higher quantum of relief expected.
We conclude that amitriptyline and physiotherapy are equally beneficial in improving outcome in patients of fibromyalgia, as seen in the study population in a rural-based tertiary care hospital in central India. Considering the cost of drug therapy we recommend use of physiotherapy over amitriptyline in resource-limited settings.
[Figure 1], [Figure 2]
[Table 1], [Table 2]