Temporomandibular Disorders and Fibromyalgia Prevalence: A Systematic Review and Meta-Analysis.

The results from the database search is shown in Fig 1 . All titles and abstracts were initially screened, and eligible articles were investigated further according to the eligibility criteria and their relevance to the review question. This systematic review included 19 studies in total. Studies were grouped into two data groups: the prevalence of CWP/FMS in people with TMDs (9 studies), and the prevalence of TMDs in people with CWP/FMS (10 studies). We present characteristics of studies individually by research design, country of survey, number, source, sex, and mean age of participants, diagnostic criteria and method used for FMS/CWP and TMD, and the prevalence found. Tables 3 11 , 42 – 49 and 4 22 – 25 , 50 – 55 describe the characteristics and results of the selected investigations of CWP/FMS prevalence in TMD groups and TMD prevalence in FMS groups, respectively. Flowchart showing study selection process. Characteristics of Included studies on Prevalence of CWP and FMS in TMD patients (n = 9) AWMF = Arbeitsgemeinschaft der Wissenschafylichen Medizinischen Fachgesellschaften (Association of the Medical Scientific Societies of Germany); CCD = craniocervical dysfunction; CMD = craniomandibular dysfunction; CWP = chronic widespread pain; MMP = masticatory myofascial pain. Characteristics of Included studies on Prevalence of TMD in FMS Patients (n = 10) FBS = failed back syndrome. In summary, of the 19 studies retrieved, 10 were case-control design, 11 , 23 – 25 , 42 , 43 , 50 – 53 6 were cross-sectional, 22 , 44 – 46 , 54 , 55 and 3 were cohort studies. 47 – 49 Two of the included studies were population-based samples, while the remainder (17 studies) used clinic-based sampling. The sex breakdown of study samples varied considerably, with some studies examining women only. 24 , 43 , 45 , 46 , 50 , 52 , 54 Most of the studies were conducted in adults (aged 18 to 75 years), except for one study of Brazilian adolescents (aged 12 to 13 years). 53 TMDs and FMS assessment were conducted in different ways across retrieved studies: for TMDs, 2 studies relied on self-reports, 4 studies used questionnaires, and 13 studies performed formal clinical examinations; for FMS, 2 studies relied on self-reports, 2 used questionnaires, and 15 performed formal clinical evaluations. Meta-analyses were subsequently administered by pooling studies according to the eligibility criteria described above. Quality assessment of included studies was scored and shown in Table 5 for FMS in TMD patients and Table 6 for TMDs in FMS patients, while the WoE is reported in Table 7 . All included studies achieved ≥ 5 stars (out of 9 stars), which represents moderate to high quality. Most articles obtained moderate to high WoE, indicating that the included articles are suitable to answer our question about the prevalence of TMDs or FMS. Quality Assessment (Newcastle-Ottawa Scale) of Studies on Prevalence of CWP/FMS in TMD Patients (n = 9) Rating guidelines are shown in Appendix Table 1 . Quality Assessment (Newcastle-Ottawa Scale) of Studies on Prevalence of TMD in FMS Patients (n = 10) Rating guidelines are shown in Appendix Table 1 . Weight of Evidence Ratings for Outcome Studies Three studies 43 , 47 , 48 considered the prevalence of CWP in a TMD population. The approach to classifying patients with widespread pain varied markedly across these studies, precluding the possibility of formal data pooling. Two studies with clinical samples reported prevalence estimates greater than 70% (72.7% and 75.6%). 43 , 47 Both studies, however, adopted generous criteria to determine widespread pain (≥ 1 body pain sites in last 6 months or “generalized pain complaints”). Velly et al 48 employed more conservative criteria (a “yes” response to questionnaire item “Do you experience widespread bodily pain [on both your right and left sides as well as above and below the waist]?” in a community-based sample and reported a modest prevalence rate of 30.6%. The estimated pooled proportion (95% CI) of FMS in TMD patients reported from four studies 42 , 46 , 48 , 49 was 32.7% (4.5% to 71.0%; Appendix Fig 1 ). Large heterogeneity was observed across studies (I 2 > 99.3%; P < .001). Studies used different clinical assessment criteria and varying TMD populations (eg, nonpainful/painful TMD). The highest proportions were in a study of patients with painful disorders of the masticatory muscles and TMJs lasting beyond 6 months (63.2%, 59.1% to 67.3%) and from a smaller clinical study of TMD patients referred to a physiatrist for the evaluation of possible FMS (52.4%, 29.8% to 74.3%), for which prevalence was especially high in patients with masticatory myofascial pain (9/11 or 81.8%). In contrast, the lowest proportion was derived from a population-based study of people with TMDs that included those with clinically nonsignificant TMD pain (10.5%, 8.1% to 13.3%). The pooled prevalence estimate (95% CI) for TMDs in people experiencing FMS 23 , 25 , 46 , 50 , 51 , 54 , 55 was 76.8% (69.5% to 83.3%; Appendix Fig 2 ). Medium heterogeneity among seven studies was found (I 2 = 51.1%). The two lowest prevalence estimates were observed in studies where TMD was determined via clinical examination guided by diagnostic criteria or responses to structured questionnaire items derived from diagnostic criteria (rather than formal clinical diagnostic assessment). When data pooling was done without these studies, the pooled prevalence estimate increased slightly to 81.4% (75.5% to 86.3%) with little evidence for cross-study heterogeneity (Q(6) = 1.4; P = .842; I 2 = 0.0%). A subgroup meta-analysis was additionally conducted according to TMD subtype: muscle, disc displacement, and inflammatory-degenerative disorders ( Appendix Fig 3 ). Pooled estimates (95% CI) indicated more than 60% (63.0%, 47.7% to 77.3%) of FMS patients presented with myogenous TMDs, while 41.8% (21.9% to 63.2%) presented with inflammatory degenerative disorders. Almost a quarter (24.2%, 19.4% to 39.5%) of FMS patients also had disc displacement disorder ( Appendix Fig 3b ). Studies concerning the prevalence of muscle and inflammatory degenerative disorders in FMS were notably heterogenous (I 2 > 85%; P < .001), but less heterogeneity was found across studies of comorbid disc displacement disorder (I 2 = 37.6%). Only one study in each subgroup meta-analysis relied on clinical examination guided by diagnostic criteria (rather than formal diagnosis) to determine TMD subtype classification. When this study was excluded from analyses, prevalence estimates remained largely the same (myogenous TMDs = 66.7%, 49.3% to 81.9%; disc displacement disorder = 25.7%, 20.5% to 31.5%; and inflammatory-degenerative disorders = 40.7%, 17.8% to 66.0%) and did not serve to decrease high levels of heterogeneity in muscle (I 2 = 87.3%) and inflammatory-degenerative disorders (I 2 = 94.0%).

The search strategy and protocol were registered and are available in the PROSPERO database. 35 The PICO model 36 and PRISMA guidelines 37 were used for data synthesis and reporting with meta-analysis. The search was conducted in the following electronic bibliographic databases: PubMed, CINAHL, Web of Science, MEDLINE, PsycINFO, Scopus, Embase, and EBM Review Cochrane (published up to April 2020). Additional literature searches in Google Scholar, OpenGrey, and the reference lists of downloaded articles were also performed. We used search keywords for TMDs with the following terms: temporomandibular disorder, jaw joint pain, orofacial pain, facial pain, myofascial, aching jaw, mandibular dysfunction, masticatory system disorder, and oro-mandibular disorder; and combined with “AND” followed by fibromyalgia terms: fibromyalgia, and chronic widespread pain. The search was conducted during April to May 2020. Inclusion criteria followed the PECOS process (population, exposure, comparison, outcome, study design). The review considered studies with no restriction of participants regarding age, sex, or other characteristics (population). Informal and formal standardized diagnostic or defined clinician-based criteria were described in included studies (exposure, comparison); for example, in TMDs (painful or nonpainful), formal diagnosis was made using the RDC/TMD, DC/TMD, and AAOP criteria, and for FMS, the ACR criteria were used. Studies included one or more measurements of prevalence rates of TMDs in CWP/FMS and/or of CWP/FMS in TMDs (outcome). Observational study designs, such as prospective, case-control, cohort, and cross-sectional, were included (study design). All retrieved articles are accessible, published in the English language, and without time limitations. Studies were selected on the basis of the previously mentioned criteria and the presence of the proportion of TMD patients with comorbid FMS and/or vice versa. Association measurements between the TMD group and FMS group (prevalence rate, odds ratio) were additionally collected. One reviewer (P.Y.) screened initially, then four reviewers (P.Y., J.S., P.C., P.R.) independently assessed full articles for inclusion in the reviews. In case of difficulties and disagreements, the reviewers discussed and resolved before achieving consensus. The following information was extracted from the included studies: author and year of publication, study design, sample size and source of the sample, location of study, sample demographics, method of diagnosis of TMDs and FMS, and outcomes. A meta-analysis was undertaken by pooling the prevalence rates from relevant studies. Meta-analyses included only studies with adult populations whereby formal diagnosis of FMS had been made according to ACR criteria (determined via clinical evaluation or previous diagnosis) and formal diagnosis of TMDs had been made via clinical evaluation according to standardized criteria such as the RDC/TMD or AAOP, or via clinical assessments and/or use of structural questionnaires guided by diagnostic criteria. Studies in which FMS and/or TMDs were ascribed via self-report of symptoms or symptom history or by examination of TMD signs were excluded from data pooling. Similarly, as CWP was typically determined via participant self-report in the included studies (and the criteria varied greatly), it was not possible to perform a meta-analysis of prevalence rates relating to CWP. Fixed- or random-effects meta-analyses were conducted using Freeman-Tukey transformations to calculate weighted summary proportions. 38 Prevalence estimates were presented with 95% CI, and Cochran Q and I 2 statistics were calculated to indicate the presence of heterogeneity. Random-effects modeling was applied where there was high heterogeneity across included studies (I 2 > 50%). 39 Forest plots were created for all estimates. Analyses were performed using SPSS (version 26.0, IBM) and MedCalc Statistical Software. Rating for study quality was assessed using the Newcastle-Ottawa Scale. 40 The checklist for quality criteria is shown in Appendix Table 1 . The scoring of all criteria is based on the Newcastle-Ottawa guideline. The measure of sample size was adjusted by rating one star if the number of participants was ≥ 100 per group, which we considered an appropriate number for representative samples. To specifically analyze each study’s appropriateness with respect to the review question on prevalence, we additionally appraised each article based on Gough’s Weight of Evidence Framework (WoE), 41 with the aim to evaluate, in particular, whether an individual study was suitable to answer the review question rather than the study question in general. The study quality and WoE were performed by the same group of reviewers.

The current systematic review included a total of 19 articles—9 studies on the prevalence of CWP or FMS in TMD patients, and 10 studies on the proportion of TMDs in FMS patients. As noted in previous reviews in this area, 32 , 34 the various criteria guidelines or protocols used in diagnosing TMDs, CWP, and FMS and their subjective dependence on patient symptoms and clinician assessment, as well as their differences in application over time (eg, ACR criteria revisions), contribute to the heterogeneity of the pooled studies. More specifically, the various classification systems for TMDs create a field of diagnostic confusion. There is uncertainty, overlap, and many different terminologies that refer to similar entities. A unified consensus would minimize confusion for physicians and patients. When clinicians use the same criteria and taxonomy, clinical questions and experiences can be more easily translated into relevant research questions. Four studies included measurements of the prevalence of FMS in TMDs. 42 , 46 , 48 , 49 Although there is large heterogeneity in the results of the meta-analysis, the observed pooled prevalence (32%) supports the hypothesis of increased risk of (comorbid) FMS in people experiencing (painful) TMDs. Heterogeneity across studies likely reflects differences in sample composition. Participants with TMDs in Velly et al 48 were recruited from the community and did not present with painful TMDs. Conversely, patients in the clinical studies of Leblebici et al 46 and Losert-Bruggner et al 49 were seeking or undergoing treatment to alleviate persistent painful TMD symptoms. The distinction is likely to be important with respect to rates of comorbid FMS (or CWP) and accounting for discrepancies with the lower (pooled) prevalence of 14% reported in the recent review performed by Kleykamp et al. 34 For example, Nguyen et al found that coexisting pain beyond orofacial areas (eg, pain in the neck or abdomen) was more frequently observed in patients with chronic TMD pain compared to acute TMD symptoms. 56 A recent study using voxel-based morphometry reported that, relative to controls, TMD patients drawn from clinic-based samples showed smaller gray matter volume in the anterior medial cingulate cortex reaching into the medial prefrontal cortex (a marker of vulnerability to CPS development), whereas no significant differences between controls and participants with TMD symptoms recruited from the community were observed. 57 Three studies 34 , 38 , 39 explored widespread bodily pain (CWP) in TMD populations; prevalence estimates in individual studies tended to be higher than in studies of co-occurring TMD and FMS, a likely consequence of the assessed CWP populations falling below the threshold of an FMS diagnosis. Patients are often preliminarily diagnosed with CWP before receiving a diagnosis of FMS by exclusion of other possible contributing conditions, such as inflammatory rheumatic diseases, nonrheumatic musculoskeletal conditions (hypermobile joints), nonrheumatic medical conditions (infections, malignancy, thyroid disease), neurologic diseases (Parkinson disease), spinal stenosis, myopathy, mental health disorders, and medication-induced pain conditions (opioids, chemotherapy). 58 Aside from the sample differences noted above, the discrepancy in rates across included studies probably reflects divergent CWP classification methods. One study with a lower prevalence 48 asked a specific question about experiencing widespread bodily pain on both the right and left sides, as well as above and below the waist, to assess CWP. The other two studies that observed > 70% CWP comorbidity used less conservative criteria (≥ 1 body pain site[s] in the past 6 months or “generalized pain complaints”). 43 , 47 Although further work is needed using established widespread pain criteria to better quantify the TMD and CWP association, it appears that a not insignificant number of TMD patients present with pain outside the orofacial region. To the extent that TMD patients with widespread pain present with more psychologic distress and respond less favorably to conventional TMD treatment, 59 routine consideration of the presence of widespread bodily pain is an important indicator of treatment strategy with a view to prevent an increase in the number of pain sites and severity of pain at affected sites. In line with previous reviews concerning the association between TMDs and FMS, 29 , 30 the estimated pooled prevalence from this meta-analysis suggested that three-quarters (76%) of patients with FMS have TMDs and that studies were largely consistent. Analysis on TMD subtype revealed that FMS patients more commonly presented with a myogenic disorder of the masticatory system (63%) than inflammatory-degenerative disorders of the TMJ (42%) or disc displacement disorders (24%), although some pooled estimates had wide CIs (because of the moderate to high heterogeneity), and only the CIs of prevalence rates for myogenic disorders of the masticatory systems and disc displacement orders did not overlap. Nevertheless, this finding coincided with a previous study suggesting that TMD signs reported by FMS patients were most often tenderness of the masticatory muscles (93.3%) and the TMJ (83.3%), while a smaller percentage of FMS patients had joint sounds (63.3%). 22 Both TMDs and FMS are COPCs that share similar inflammatory and hyperalgesic features or symptoms of the facial and cervical musculoskeletal structures 24 , 60 but are nevertheless discrete conditions. FMS patients have a lower pain threshold, frequently experience fatigue, and have a lower muscle burden than TMD patients. However, FMS groups show a high prevalence of TMDs and pain in sites upon palpation of the head and neck area, frequently complain of fatigue in the orofacial region, and experience pain with jaw movements. 23 There are no clear etiologies or pathogenesis on the development of coexisting TMDs and FMS. Multiple risk factors such as trauma, oral-facial parafunctional habits, and connective tissue diseases contribute to TMD development, while FMS is related to dysfunction of the central nervous system, genetics, and hormone and metabolite imbalance. 61 There is a high frequency of psychophysiologic and psychiatric disorders in patients with TMDs and patients with FMS, including sleep disturbances, depressive and anxiety disorders, oral ulcers, and other COPCs. 33 , 34 , 61 Two recent reviews indicated that the prevalence of other COPCs (eg, chronic back pain, chronic stomach pain, chronic migraine headache, and IBS) ranged from 39% to 76% in FMS patients and from 19% to 66% in individuals with TMDs depending on the specific condition. 33 , 34 These coexisting conditions contribute to the complexity of TMDs and likely increase the overall pain burden associated with this group of conditions. Our review suggests that about one-third of people with TMDs have comorbid FMS, although there is wide variation according to sample composition, whereas more than three-quarters of the FMS population have comorbid TMDs, with much less variation in prevalence across the latter set of studies. The higher prevalence of TMDs in FMS than of FMS in TMDs is consistent with some previous studies 62 , 63 and likely reflects the neuromuscular impairment and central sensitization that characterizes FMS, which could lead to the temporomandibular musculoskeletal system’s failure to adapt to continuous stress and loading. 64 Furthermore, the overlapping neuroinflammatory pathophysiology in TMDs and CWP leads to the argument that the subgroup of hypersensitive TMD patients could transit to CWP. 64 This is supported by evidence that individuals with signs of painful TMDs are at higher risk of developing central sensitization than pain-free adolescents. 65 The precise relationship between timing of onset of coexisting TMDs and FMS remains unclear. One study 45 suggested that TMDs and FMS frequently occurred within the same year; if not, facial pain preceded widespread pain in most patients. Hoffmann et al 11 also reported that 4% of TMD patients experienced FMS before developing TMDs, and the proportion of FMS increased significantly to 21% after the onset of TMDs. In addition, other studies have suggested that CWP and FMS had a partial influence on the occurrence of clinically significant TMD pain and its persistence. 48 John et al reported that CWP predicted the occurrence of dysfunctional TMD pain in women (but not in men) as well as its persistence. 47 Furthermore, previous studies have found that the presence of multiple painful areas elsewhere in the body may increase the risk of onset of TMD pain within the next 3 years 66 , 67 and that the level of facial pain is positively correlated with that of general body pain. 68 While research findings concerning whether the emergence of TMDs precedes, coincides, or follows the development of CWP or FMS remains inconsistent, the evidence for an elevated risk of overlapping conditions firmly indicates clinicians should be wary of the possible coexistence and exacerbation of TMD pain when CWP or FMS has developed. From this view, investigation of other (bodily) clinical pain features reported by the patient should better enable more comprehensive TMD patient management. Harper et al 69 recently reported that higher levels of FMS symptoms were associated with greater pain at rest and higher perceived functional limitation of the jaw in TMD patients, suggesting that treatments aimed at decreasing central pain sensitization and reducing spontaneous pain may also contribute to TMD symptom relief. More generally, combined management of the contributing factors to TMDs and FMS may improve patients’ oral and general quality of life, with pain reduction and improved temporomandibular system physical functions. 70 This review included a variety of study designs, involving those with and without control groups, but the difference in prevalence estimates between patient and control samples with respect to comorbid conditions were not examined in the meta-analysis. In addition, our systematic review included only English-language publications. Thus, further reviews without language restriction will obtain more data and reduce systematic bias. As noted, data from this review were pooled across studies that used diagnostic criteria that differed according to classification systems and revisions over time, which can lead to disparity in prevalence rates. 59 The recently published ICOP-2, endorsed by most of the leading orofacial pain institutions, will help to reduce diagnostic ambiguity and improve consistency across TMD studies in future reviews. Study samples often had a female dominance, likely reflecting that, for example, 80% to 90% of populations with the FMS condition are female. 71 But this potentially limits the representativeness of the FMS and CWP populations included in the review. Further, the TMD samples in most studies were individuals with painful TMDs who sought treatment, presumably to relieve TMD pain. However, TMDs are a collective condition associated with pain and/or dysfunction, and therefore the inclusion of individuals with painful TMDs does not reflect the broader population with TMDs and these patients may be more vulnerable to the development of chronic widespread pain or FMS. Accordingly, analyses investigating the pooled prevalence rate of FMS in people with TMDs incorporated studies with both clinical and population-based samples, yielding large heterogeneity in outcomes. While subgroup analyses based on sample composition could be considered, it is of limited value when the number of studies included in the meta-analyses is too small to perform meaningful analyses. In any case, some caution is warranted when interpreting these findings. Finally, this systematic review was conducted in 2020. By the time of publication, there were additional empirical studies examining the coexistence of TMDs and FMS/CWP, although these tended to reaffirm findings from the present review. For instance, a recent Swedish study reported that almost 30% (47 of 161) of patients referred to orofacial pain clinics fulfilled the ACR (2016) criteria for FMS, although the rate was much higher in the subgroup of patients with myofascial pain with referral (45.7%) than those with myalgia (12.5%). 72 In a US study of clinic patients who presented for treatment, 17 of 89 (19.1%) patients with TMDs screened positive for FMS. 69 Another recent Swedish study, but with a community-based sample, found the overlap between widespread pain (≥ 7 pain sites identified from a full-body pain diagram) and any myofascial orofacial pain diagnosis was 57.3%. 73 Altogether, these recent studies support the findings of this review, suggesting that individuals with TMDs, particularly those seeking treatment for painful myogenous TMDs, have an elevated risk of experiencing concurrent widespread pain and/or having FMS.

This systematic review found a high co-occurrence of TMDs and FMS. Pooled prevalence estimates indicate that about one-third of TMD patients have FMS, whereas more than three-quarters of the FMS population have comorbid TMDs with higher rates of myogenous TMDs than disc displacement disorders. The variability in TMD sample composition across studies investigating comorbid FMS yielded marked heterogeneity in the corresponding meta-analysis, complicating the interpretation of overall prevalence. Experience of CWP was also common in people with TMDs, with estimates across individual studies ranging from 30% to 76%, although the criteria used to classify CWP varied greatly. These findings suggest a need for clinicians to consider the overlap between TMDs and CWP/FMS when treating affected populations, and, where appropriate, to consider multidisciplinary approaches to care. TMDs are prevalent in FMS patients, affecting 3 out of 4 individuals. Most FMS patients present with a myogenous TMD condition. Patients seeking treatment for painful TMDs appear more likely to have CWP or FMS.