Is Irritable Bowel Syndrome Considered as Comorbidity in Clinical Trials of Physical Therapy Interventions in Fibromyalgia? A Scoping Review

Evidence supports the presence of comorbid conditions, e.g., irritable bowel syndrome (IBS), in individuals with fibromyalgia (FM). Physical therapy plays an essential role in the treatment of FM; however, it is not currently known whether the IBS comorbidity is considered in the selection criteria for clinical trials evaluating physiotherapy in FM. Thus, the aim of the review was to identify whether the presence of IBS was considered in the selection criteria for study subjects for those clinical trials that have been highly cited or published in the high-impact journals investigating the effects of physical therapy in FM. A literature search in the Web of Science database for clinical trials that were highly cited or published in high-impact journals, i.e., first second quartile (Q1) of any category of the Journal Citation Report (JCR), investigating the effects of physical therapy in FM was conducted. The methodological quality of the selected trials was assessed with the Physiotherapy Evidence Database (PEDro) scale. Authors, affiliations, number of citations, objectives, sex/gender, age, and eligibility criteria of each article were extracted and analyzed independently by two authors. From a total of the 412 identified articles, 20 and 61 clinical trials were included according to the citation criterion or JCR criterion, respectively. The PEDro score ranged from 2 to 8 (mean: 5.9, SD: 0.1). The comorbidity between FM and IBS was not considered within the eligibility criteria of the participants in any of the clinical trials. The improvement of the eligibility criteria is required in clinical trials on physical therapy that include FM patients to avoid selection bias.


Introduction
According to the American College of Rheumatology (ACR), fibromyalgia (FM) is a chronic condition associated with widespread pain, fatigue, sleep problems and tenderness [1]. With a worldwide prevalence of 2-3% [2], its prevalence in Spain is 2.4% [3]. Although the diagnostic criteria have evolved from the exclusive presence of generalized pain and pain on palpation at specific locations [4] to the inclusion of questionnaires on pain perception and distress [5], its prevalence has increased as diagnostic criteria evolved [6]. Nonetheless, this pain condition is under-, over-, or misdiagnosed [7]. Regarding the distribution by sex, FM is more frequent in women than in men (female: male ratio 9:1) [8]. Although FM is considered a noninflammatory generalized musculoskeletal pain condition associated with fatigue and sleep disturbances, many patients also exhibit cognitive dysfunction (e.g., brain fog), mood disorders, or intestinal comorbidities, e.g., irritable bowel syndrome (IBS) [9].
The etiology and pathophysiology of FM are not well understood [2]. Current theories describe a potential disorder of nociceptive pain processing and central sensitization as the main pathophysiological mechanism [10]. Current evidence shows that individuals with FM usually exhibit several medical comorbidities, e.g., diabetes mellitus, IBS, or mood disorders [11]. Clinicians routinely overlook somatic symptoms, however, if comorbidities are taken into account and treated appropriately, the severity of the symptoms in patients with FM would be reduced [12].
Specifically, IBS is present in 46.2% of people with FM [13], whereas FM occurs in 12.9% to 31.6% of patients with IBS [14]. Both conditions share a wide variety of symptoms such as fatigue, insomnia, anxiety or depression, as well as a predilection for the female gender. As in FM, the pathophysiology of IBS is not understood. There is a hypothesis suggesting a dysregulation of the brain−gut axis resulting in a state of generalized hyperalgesia (also known as central sensitivity disorders) leading to increased excitability of central nociceptive pathways and inhibition of descending pain modulation [15]. This hypothesis could explain the comorbidity presentation between FM and IBS and also the presence of common symptoms.
This similarity between both syndromes should create the need to more exhaustively define the selection of participants with FM in clinical trials [12,13]. Treatment of FM is clearly multifactorial, but it seems that physical therapy has shown to be beneficial for these patients. Several meta-analyses support that physical therapy interventions are effective for reducing symptoms and for improving health-related quality of life in individuals with FM, exercise probably being the most effective [16][17][18]. The main target of physical therapy interventions is the musculoskeletal system, the presence of a comorbid visceral disorder, e.g., IBS, could lead to the necessity of different therapeutic strategies.
Meta-analyses represent the highest level of evidence supporting or refuting a therapeutic intervention, but they are based on randomized clinical trials. The influence of relationships between FM with psychological and somatic-visceral conditions could lead to uncertainty regarding the etiology of symptoms and may influence those clinical outcomes. If medical comorbidities are ignored in the selection criteria of clinical trials, this may limit the potential benefits of physical therapy, which mainly targets musculoskeletal symptomatology. In fact, it has already been identified that the comorbid relationship of IBS with temporomandibular disorders [19,20] is underreported when designing clinical trials evaluating the effectiveness of physical therapy [21]. It is, therefore, reasonable to consider that the relationship of FM with IBS could also influence clinical outcomes. No study has previously investigated this topic in clinical trials including individuals with FM. The objective of this scoping review was to identify whether the presence of IBS has been considered in the participating selection criteria of "relevant" clinical trials evaluating the effects of physical therapy in individuals with FM.

Methods
This scoping review was conducted following the methodological framework suggested by Arksey and O'Malley [22] consisting of: 1, identify the research question; 2, identify relevant studies; 3, study selection; 4, data extraction; 5, compiling, summarizing, and reporting results. Additionally, it also adheres to the adjusted items for systematic reviews and meta-analysis extension for scoping reviews (Prisma-ScR) and it has been registered in The Open Science Framework Registry (https://osf.io/ns35d (accessed on 20 December 2020)).

Research Question
The research question of the current scoping review was: has the presence of visceral disorders, e.g., IBS, been considered within the selection criteria of "relevant" clinical trials investigating physical therapy interventions in individuals with FM?

Identifying Relevant Studies
The literature search was conducted in the Web of Science (WOS) database from the inception of the database to 22 December 2020. The search was conducted by two different assessors and was limited to high-quality clinical trials including humans. No language restriction was applied. A combination of the following terms employing Boolean operators was used for the search: "fibromyalgia" AND, "physical therapy" OR "physiotherapy" OR "exercise" OR "manual therapy".

Study Selection
In this review, the PCC (Population, Concept and Context) mnemonic rule was used to define the inclusion criteria.
Population: Men or women diagnosed with FM according to 1900 ACR diagnostic criteria [4], 2010 ACR diagnostic criteria [23] or modified 2010 ACR classification criteria [5]. Alternative diagnostic criteria for FM were also accepted if properly described in the paper.
Concept: Randomized clinical trials evaluating any type of physical therapy intervention, alone or in combination with others, in individuals with FM.
Context: Articles that met one of the following criteria were considered "relevant": 1, most cited clinical trials from the selection; or, 2, clinical trials published in high impact journals, e.g., of the first quartile (Q1) of any category of the Journal Citation Reports (JCR) evaluated in the year of publication of the study, according to WOS (JCR criterion−impact factor).
The selected studies were identified independently by two investigators considering the title and abstract. For potentially eligible articles, the full text of the article was read. If both researchers disagreed on the inclusion/exclusion of any of the articles, a third researcher decided whether to include/exclude it. All data were saved and managed through Microsoft Office.

Data Extraction
From selected studies, the following information was extracted following a standardized form: number of authors affiliated with a clinical institution (e.g., hospital, private clinic, or health center), number of authors affiliated with a nonclinical institution (e.g., university or research center), total number of citations in WOS, PEDro score, study objectives, sample size, characteristics of participants (sex distribution, mean age), and inclusion and exclusion criteria for participants selection.

Methodological Quality
The Physiotherapy Evidence Database (PEDro) scale score was used to determine the methodological quality of clinical trials. This scale contains 11 items, which can be scored as absent (0) or present (1), except for the first item that refers to the external validity of the study. Each item is scored from 0 to 10 points. The PEDro scale is a valid, reliable and widely used tool for rating the methodological quality of clinical trials [24]. Studies were considered to be of high quality if they obtained at least 5 points on this scale. The score for each article was extracted from the PEDro database, except for those that were not evaluated by the database. In this case, they were evaluated by two researchers following the guidelines established by the PEDro scale [25].

Study Selection
The search identified a total of 412 articles. Of these, 316 were excluded because they did not meet any of the criteria. Of the remaining 96, the full text was accessed, after which 15 articles were excluded. Finally, 20 trials according to the citation criteria and another 61 published in Q1 JCR journals were included. Eight of the selected articles met both criteria; accordingly, a total of 73 different clinical trials were included ( Figure 1).

Study Characteristics
Data extracted from the most cited trials are shown in Table 1, data from trials published in Q1 of any category of the JCR are summarized in Table 2, and those fulfilling both criteria are described in Table 3. The total number of citations from those trials included in Tables 1-3 was 3502 (mean citations 175.1, SD: 53.7 per trial). It should be noted that three clinical trials were based on the same sample population [26][27][28]; therefore, the first published one was considered since the inclusion/exclusion criteria are the same [27].              Authors from non-clinical institutions were part of all trials except one [92], whereas authors from clinical institutions were present in 45.1% of the studies (n = 32/73). The total number of participants in the clinical trials were 6688 (279 men and 6409 women). Sixty-one percent of the trials (n = 45/73) just recruited women, and the remaining studies presented a higher number of women than men in their sample. One study failed to specify the sex [51], and another did not detail either sex or mean age [78]. Another trial did not specify the mean age [36], which ranged from 38 to 59 years for the trial.

Methodological Quality
According to the PEDro scale, the methodological quality scores ranged from 2 to 8 points (mean: 5.9, SD: 0.1) out of a maximum of 10 points. From the 20 articles included by citation criteria (mean: 5.6, SD: 0.1), all were rated ≥5 points, except one with a score of 4 [35]. In addition, 53 out of 61 of those published in Q1 JCR journal achieved ≥5 points (mean score: 6, SD: 0.14), while the remaining 8 trials received <5 points. There were four clinical trials without evaluation in the PEDro database. According to researchers evaluation, three obtained a good methodological quality [48,56,63].

Inclusion Criteria in Clinical Trials
After examining the inclusion criteria of all trials, none considered comorbidity from a visceral origin. Only two trials included the absence of concomitant somatic disorders but without any other specification [48,84]. Generally, the common criterion for most clinical trials was "Patients diagnosed with fibromyalgia according to ACR criteria". Three clinical trials presented diagnostic criteria that differed from the ACR, such as the criteria proposed by Yunus [91], Smythe [95], or both authors [40]. In addition, five studies did not specify which criteria they used for diagnosis of FM [34,51,52,56,62].

Findings
The current scoping review has observed that the presence of IBS, a common medical comorbidity, is not consistently considered for the selection of participants in highly cited or published in high-impact journal clinical trials investigating the effects of physical therapy in FM, which could lead to a selection bias.
All clinical trials included individuals with a diagnosis of FM according to ACR criteria, except Buckelew et al. [91], McCain et al. [95] and Wigers et al. [40] that used other criteria. In addition, five studies [34,51,52,56,62] failed to specify the criteria by which the diagnosis of FM was made. The exclusion criteria were generally more heterogeneous, including psychological/psychiatric diseases, neurological diseases, rheumatic diseases, medication, or diabetes mellitus, among the most common. In general, the presence of visceral pathology was not summarized as an exclusion criterion. Four studies considered the possible comorbid conditions that may worsen FM symptomatology without specifying any particular visceral pathology [50,64,89,90]. In addition to IBS, it is important to highlight that other pathologies of visceral origin are also highly comorbid in FMS, and, again, they were not considered in the included trials. This should prompt us to consider the current diagnosis of FM, since it is mainly based on the presence of pain symptoms, considering these comorbidities into this complex spectrum could help to improve the quality of life and management of these patients. Interestingly, albeit the high comorbidity between IBS and FM [12,99], no clinical trial included in this scoping review commented anything on this relationship.

Why IBS Can Be Relevant for FMS Clinical Outcomes?
Current hypotheses support that IBS ad FMS share common underlying mechanisms leading to increased excitability of central nociceptive pathways [15]. In fact, the presence of previous IBS has been found to be the strongest predictor for new-onset FM development [100]. The presence of comorbid visceral conditions, e.g., IBS, in a musculoskeletal pain condition such as FM, should be considered in clinical practice since visceral pain enhances sensitization [101]. An exacerbation of the symptoms when two comorbid conditions exist is labeled as functional somatic syndrome [102], a situation which should be carefully explored and considered in the management of chronic pain conditions exhibiting manifestations at different levels such as those occurring in individuals with FM. In such a scenario, early recognition of comorbid syndromes of different etiology, but exhibiting a common mechanism, may identify subgroups of patients with different etiologies and different needs of treatment [103].
Comorbid visceral conditions should not be ignored when a physical therapy intervention is tested, as potentially occurred in the identified trials in the current scoping review. We do not know if considering comorbid visceral conditions in people with musculoskeletal pain conditions participating in physical therapy clinical trials could lead to potentially different clinical outcomes. This is relevant, considering that visceral pain shares several features with musculoskeletal pain but clearly requires different therapeutic strategies. For instance, the current understanding of the neurosciences is continuously evolving for better adaptation of exercise programs in individuals with nociplastic pain, a category where FM could be included [104]. It is possible that individuals with FM and comorbid IBS need different exercise programs to those with other comorbidities or without IBS. Future clinical trials should investigate the effects of multimodal therapeutic approaches considering the presence of these visceral comorbidities, e.g., IBS, in patients with a primary musculoskeletal complaint, e.g., FM.

Strengths and Limitations
Findings from the current scoping review should be considered according to its strengths and limitations. Strengths include a comprehensive literature search, methodological data extraction, and the inclusion of highly cited and published clinical trials in high-impact journals investigating physical therapy for FM. Among the limitations, first, the search was conducted on a single database, the WOS, because it is the only database presenting the index classification by JCR. Second, physical therapy interventions were heterogeneous, ranging from manual therapy to exercise alone or combined with physical agents. Third, 95% of the patients included in the clinical trials were female; nonetheless, this is related to the fact that FM is more prevalent in females and also that a greater frequency of comorbidity in pain syndromes is present in females [105].

Conclusions
This scoping review found that highly cited clinical trials or those published in high impact journals investigating the effects of physical therapy interventions in individuals with FM did not consider the presence of comorbid IBS in their eligibility criteria. In turn, other pathologies were sometimes considered, mostly linked to exercise, e.g., cardiac or kidney diseases. Current results highlight that the presence of intestinal pathology is underestimated when treating a musculoskeletal pain condition. Based on our results, stricter inclusion and exclusion criteria would be required in clinical trials involving patients with FM to avoid possible subject selection biases.