Introduction

Fibromyalgia (FM) syndrome is one of the most common musculoskeletal disorders with unknown cause involving adults especially women aged 20–55 [1]. The primary diagnosis is based on chronic generalized pain with tenderness during exam, fatigue, sleep disorder, headache and mood/cognitive problems [1, 2]. The American College of Rheumatology (ACR) has defined the fibromyalgia as generalized pains for at least three months with 11/18 tender points during physical examination [3, 4]. Although combining these two criteria can detect 81–88% of cases, other complaints such as fatigue, sleep disorder, morning stiffness, paresthesia and psychosis are also common during the disease [3].

According to ACR criteria, different studies reported the prevalence of fibromyalgia between 0.2 to 5% [5]. Other studies indicate the prevalence among men and women as of 0.2–3.9% and 0.7–13%, respectively [6]. The heterogeneity in different countries can be due to different factors such as different methodologies and even real differences between countries [7]. This syndrome has been observed in most countries with various climates among different ethnic groups [8]. Prevalence of FM among women is 8–9 folds greater than that among men and is increased with age so that 7.4% of women aged 70–79 years are suffered from this problem [1]. According to the available evidences, the pain and disability is as high as that in rheumatoid arthritis which affects the personal activity and recreations leading to occupational problems [1, 9].

According to the results of primary searches, different studies have been carried out with regard to the prevalence of FM. Combining the results of these primary studies provides reliable evidences for policymaking [10,11,12]. This study aims to estimate the total prevalence of fibromyalgia in the world using meta-analysis method.

Methods

Search strategy

During 1–25 December 2016, two independent researchers searched PubMed, Web of Sciences, Ovid, Ebsco, Scopus and Cochrane library databanks and Google scholar search engine using the following Keywords to find relevant evidences written in English from each time to 30 November 2015. A sample of search strategy in PubMed is follow: ((((((((((((((((((((fibromyalgia) OR fibromyalgia-fibromyositis syndrome) OR fibromyalgia-fibromyositis syndromes) OR syndrome, fibromyalgia-fibromyositis) OR syndromes, fibromyalgia-fibromyositis) OR fibromyositis-fibromyalgia syndrome) OR fibromyositis fibromyalgia syndrome) OR fibromyositis-fibromyalgia syndromes) OR syndrome, fibromyositis-fibromyalgia) OR syndromes, fibromyositis-fibromyalgia) OR fibromyalgia, secondary) OR fibromyalgia, secondary) OR secondary fibromyalgia) OR secondary fibromyalgia) OR fibromyalgia, primary) OR fibromyalgia, primary) OR primary fibromyalgia) OR primary fibromyalgia) AND prevalence) OR epidemiology) OR frequency.

Any disagreement was managed by a third researcher. They also investigated the references of the identified papers to increase the search sensitivity.

Inclusion/exclusion criteria

All English-written articles with enough quality scores reporting the prevalence of fibromyalgia as well as sample size of the study were included in the meta-analysis. On the other hand, studies did not report the above-mentioned information, those with low-quality score, articles had not been written in English and abstracts represented in congresses without full text, were excluded from the meta-analysis.

Selection of the studies

During the investigation of the primary identified studies by two independent researchers, first, duplicates were identified and excluded. Then, irrelevant articles were removed after investigating the titles, abstracts and full texts, respectively. In addition, the investigators reviewed the results of these studies to find and omit any repeated results. A third researcher was selected to manage probable disagreements between the above-mentioned study selection processes.

Quality assessment

The above selected studies were quality assessed according to the STROBE checklist including 22 methodological questions [13]. Minimum and maximum scores achieved by each study were zero and 44, respectively. These studies were classified into low quality (less than 15.5), moderate quality (15.5–29.5) and high quality (30–44). Low-quality studies were excluded from the meta-analysis.

Data extraction

Title, first author name, date and country in which the study was conducted, prevalence of fibromyalgia, sample size of the study according to gender, mean age and age range of the participants and diagnostic methods were extracted from each study. These information were entered into the excel spreadsheets.

Statistical analysis

Data analysis was performed using Stata software. Standard error of prevalence for each study was calculated according to binomial distribution formula. The heterogeneity between the results of primary studies was presented based on I-square and Cochrane (Q) indices. P value less than 0.1 was considered significant heterogeneity. According to suggestion of Higgins and et al. [14], the I-squared results were classified into low heterogeneity (less than 25%), moderate heterogeneity (25–75%) and high heterogeneity (more than 75%). According to the results of heterogeneity between the primary studies, random- or fixed-effect model was applied for combining the results. The effect of the suspected factors for heterogeneity was assessed using meta-regression models and subgroup analysis. Pooled and primary prevalences as well as 95% confidence intervals were illustrated by forest plots. Considering the prevalence as the main indicator in the current meta-analysis, no investigation was done for publication bias.

Results

Out of 61,346 studies identified during the primary search, 1742 papers were remained after restricting the search strategy. Of them, 1040 duplicate papers were excluded. Review of titles and abstracts, 571 papers were identified irrelevant. Full text review revealed 49 irrelevant articles. Moreover, five new papers identified during the reference review. Finally, considering inclusion/exclusion criteria and quality assessment, 22 articles were omitted and 65 papers [1,2,3, 5, 7, 15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74] were identified eligible for meta-analysis (Fig. 1).

Fig. 1
figure 1

Literature search and review flowchart for selection of primary studies

Full size image

Among 3,609,810 people investigated in the 65 included studies in the systematic review, 81 cases of evidence were identified the prevalence of fibromyalgia among general population and specific groups. According to the results of 44 cases of evidence (Table 1), prevalence of fibromyalgia among 3,500,756 people of general population was estimated as of 1.78% (95% confidence interval: 1.65–1.92). Subgroup analysis of Prevalence of fibromyalgia in the general population presented by World Health Organization (WHO) regions. In EURO region was reported in 24 studies varied from 0.29% in the study carried out by Sauer [64] in 2010 in Germany to 11.10% in the study conducted by Okumus [47] in Turkey (2012). The total prevalence of fibromyalgia in the EURO region was estimated as of 2.64% (95% confidence interval: 2.10–3.18) (Fig. 2).

Table 1 Characteristics of the primary studies regarding prevalence of fibromyalgia in general population according to different variables
Full size table
Fig. 2
figure 2

Primary and pooled prevalences of fibromyalgia in general population and WHO regions

Full size image

Prevalence of fibromyalgia in the general population of AMRO region was reported in 16 studies from 0.16% in Andary [17] study conducted in the USA (2004) to 6.36% in the study conducted by Vincent [69] in the USA (2013). The total prevalence of fibromyalgia in the AMRO general population was estimated as of 2.41% (95% confidence interval: 1.69–3.13) (Fig. 2).

According to the results of two studies reporting the fibromyalgia prevalence in the WPRO general population, this disease was observed in 1.50 and 1.70% of the population. Using fixed-effect model, the total prevalence was estimated as of 1.62% (95% confidence interval: 1.00–2.24).

Prevalence of fibromyalgia in EMRO general population was reported in two studies as of 8.30 and 0.69%. The pooled estimate of fibromyalgia prevalence in this region was 4.43% (−3.00 to 11.86) (Fig. 2). It is should be mentioned that list of countries by WHO regions presented in a supplemental file.

In 26 studies, prevalence of fibromyalgia was estimated based on ACR criteria as of 2.32% (95% confidence interval: 1.85–2.79). The total prevalences according to diagnostic methods of ACR, Fibromyalgia Impact Questionnaire (FIQ), international classification of disease (ICD 9/10), interview, London Fibromyalgia Epidemiology Study Screening Questionnaire (LFESSQ), physical examination and questionnaire self-made) are represented in Table 2 and Fig. 3.

Table 2 Prevalence of fibromyalgia according to gender and WHO regions
Full size table
Fig. 3
figure 3

Primary and pooled prevalences of fibromyalgia in general population according to diagnostic criteria

Full size image

Prevalence of fibromyalgia in women was reported in nine studies varied from 20% in Vincent study [69] to 10.50% in Forseth study [34]. The total prevalence of fibromyalgia in women was estimated as of 3.98% (95% confidence interval: 2.80–5.20). Prevalence of fibromyalgia among men was reported in seven studies from zero in Lindell [45] and Assumpaco [18] studies to 4.80% in Vincent study [69]. The total prevalence of fibromyalgia in men was estimated as of 0.01 (0.04–0.06).

Using meta-regression models showed that publication date (β = −0.008, p = 0.884), diagnostic method (β = −0.002, p = 0.992) and WHO region (β = −0.197, p = 0.354) did not influence the heterogeneity between the results.

Prevalence of fibromyalgia in specific populations was reported in 37 studies including 109,054 subjects (Table 3). Prevalence of fibromyalgia among patients referring to the internal and rheumatology departments and clinics was reported by four primary studies varied from 14% in the study carried out by Branco [3] to 41% in the study conducted by Brill [21]. Combining the results of these studies, the total prevalence of fibromyalgia in patients visited in the internal and rheumatology departments was estimated as of 15.20% (95% confidence interval: 13.60–16.90). According to the results of three studies, prevalence of fibromyalgia among patients suffering from IBS was reported from 12.90% in Alexander study [28] to 31.60% in the Sperber study [65]. The total prevalence of fibromyalgia among IBS patients was estimated as of 12.90% (95% confidence interval: 12.70–13.10).

Table 3 Characteristics of primary studies regarding prevalence of fibromyalgia in specific populations
Full size table

Prevalence of fibromyalgia among patients under hemodialysis was assessed in four studies reported from 3.90% in the study carried out by Claudio [29] to 12.20% in the study conducted by Samimagham [63]. The total prevalence of disease according to the results of these four studies was estimated as of 6.30% (95% confidence interval: 4.60–7.90). Patients with diabetes mellitus were investigated for fibromyalgia in three studies. The point prevalences varied from 9% in the study carried out by Patucchi [58] to 23.30% in the study performed by Wolak [71]. The total prevalence of fibromyalgia in these studies was estimated as of 14.80% (11.10–18.40).

Fibromyalgia was investigated in the other subpopulations such as HTLV-1-infected patients (one study), patients with vasovagal syncope (one study), full-term pregnant women (one study), Individuals exposed to the combination of physical injury and extreme stress (one study), multiple sclerosis (MS) patients (one study), cardiovascular patients (one study), subjects with chronic hepatitis B infection (one study), SLE (lupus erythematous) patients (one study), women in premenopausal and post menopausal phases referring to gynecology departments (two studies), patients with spondylitis ankylosans (one study), patients with Behcet syndrome (one study), patients with cervical radiculopathy (one study), patients with chronic disabling musculoskeletal disorders (one study), patients with idiopathic chronic urticaria (one study), patients with hereditary hemochromatosis (one study), patients with hypothalamic/hypophysis dysfunction(one study), patients with psoriasis (one study) and women with hyperprolactinemia (one study). The details of these studies are represented in Table 3. The prevalence of fibromyalgia among specific subgroups varied from 3.90% in patients undergoing hemodialysis in Claudio study [29] to 80% among patients with Behcet syndrome in the study carried out by Melikoglu [51].

Discussion

Results of our meta-analysis showed that 1.78% of the general population especially women are suffering from fibromyalgia. Although prevalence of fibromyalgia was higher in the EMRO region, the confidence intervals showed no significant difference between the WHO regions regarding the prevalence of fibromyalgia. It should be noted that most of the primary studies used ACR as the diagnostic method. Subgroup meta-analysis based on the subgroups showed prevalences of fibromyalgia as of 15.20% in patients referred to rheumatology clinic, 12.90% among IBS patients, 6.30% in hemodialysis patients and 14.80% in patients with type 2 diabetes mellitus. In addition, 80% of patients with Behcet syndrome developed fibromyalgia.

In the study conducted by Perrot et al. in 2011, prevalence of fibromyalgia among 3081 French adults aged over 18 was reported as of 1.60%) [59]. Lindell et al. conducted a study among 147 Swedish people and reported the prevalence of fibromyalgia and chronic generalized pains as of 1.30 and 4.20%, respectively [45]. These two studies were carried out in the general population without difference in the prevalences.

According to a cross-sectional study conducted by Buskila et al. among 522 patients hospitalized in the internal department, fibromyalgia diagnosed by ACR criteria in 15% of patients 91% of which were women. They found that prevalence of fibromyalgia among women was associated with age [24]. It seems that patients referring to the internal and rheumatology departments and clinics need more investigation and consideration regarding this disorder.

According to the results of a case control study conducted by Cole et al. prevalence of fibromyalgia which was diagnosed by physical examination was significantly higher among people with IBS (12.90%) compared to those without (4.70%) [28]. Therefore, fibromyalgia should be taken into consideration in patients suffering from IBS.

Kucuksen et al. conducted a study among 118 patients with episodic migraine. According to the ACR criteria, prevalence of fibromyalgia was the same among migraine patients with and without aura [43]. There is no enough evidence for this lack of relationship, and further studies are required to investigate the precise association between episodic migraine and fibromyalgia.

During a cross-sectional study among 221 patients undergoing hemodialysis, Leblebici et al. showed similar distributions of age, gender and hemodialysis duration between subjects with and without fibromyalgia. However, significant differences were observed between these two groups regarding educational level, sleeping, fatigue and cognitive symptoms [44]. Therefore, it seems that fibromyalgia is more common among hemodialysis patients.

Mohammad et al. detected 106 (57%) patients suffering from fibromyalgia among 185 chronic hepatitis C patients with mean age 48.70 years [54]. Buskila et al. found that 15 (71%) of women with hyperprolactinemia as well as 2 (4.50%) of women with normal serum prolactin had fibromyalgia detected by ACR criteria [23]. These results indicate that fibromyalgia is more common among patients with chronic hepatitis C, hyperprolactinemia and other disorders shown in the results section.

Diagnostic criteria for fibromyalgia in the primary studies were ACR, LFESSQ and FIQ. ACR is a standard questionnaire which was applied more than the other criteria. According to these criteria, chronic generalized pain more than 3 months together with tenderness in eleven of eighteen points are in accordance with fibromyalgia. Disseminated pain including pain in right and left hemilaterals, upper and lower back, axial skeleton such as cervical, dorsal, thoracic and lumbar. Diagnosis of fibromyalgia will be ruled out by any concurrent disorder [1].

Another screening method for diagnosis of fibromyalgia is LFESSQ which is successfully applied for screening of fibromyalgia in both general and specific populations. This questionnaire includes pain and fatigue as diagnostic criteria [3]. In addition, FIQ was another diagnostic method including 10 questions and is applied as an effectiveness index for treatment [20].

The current meta-analysis is prone to some limitations such as great heterogeneity between the primary results. However, we combined the results using random effect model and subgroup analysis. On the other hand, in most of the primary studies, results were not presented based on gender. Using different diagnostic methods and language bias are other limitations for our study.

Combining various prevalences of fibromyalgia in general population and different subpopulations reported in primary studies was the main strength of this meta-analysis which can be a suitable opportunity for researchers and policymakers.

Conclusion

Our study showed that prevalence of fibromyalgia in general population is considerably lower than that among populations with specific disorders. It also more common among women. In contrast to the general population, it is important to investigate this disorder among people with specific diseases. In addition, future studies should apply more accurate diagnostic criteria and represent the exact sampling tools and characteristics of the study population.