A Narrative Review of Musculoskeletal Impairments Associated With Nonspecific Chronic Pelvic Pain.

A systematic literature search was performed using PubMed, Cochrane, CINAHL, Embase, and Web of Science databases from 1998 to 2018 to identify studies reporting on the relationship among CPP and musculoskeletal impairments of the hip, pelvis and trunk. Various combinations of the following search terms were used: pelvic pain, pelvic floor muscle, neuromotor control, motor control and musculoskeletal. In anticipation that the literature may be limited, we used broad search terms to reduce the likelihood of missing articles. Study eligibility and exclusion criteria are provided in Table 1 . The search resulted in 2106 articles retrieved. Duplicate articles were removed and the titles and abstracts of the remaining articles were screened by two authors (DP, TS). If the two authors disagreed on the inclusion of an article, a third author was consulted (MHH). Thirty-one articles remained after initial screening. Full text papers were then reviewed and 25 articles were excluded for not meeting the outlined criteria. Six additional articles were located through review of the reference lists of included articles and personal searches. A total of twelve articles were included in the final review and outlined in Table 2 .

Of the 12 articles reviewed in the study, seven studies were cross-sectional cohorts 3 , 9 , 14 – 18 or case-control 19 comparing patients with CPP to asymptomatic controls. One study was cohort study that included patients with only CPP. 20 The remaining studies were treatment studies with either randomized controlled trial (RCT) 10 , 11 or prospective, longitudinal cohorts. 21 We used the Oxford Centre for Evidence-Based Medicine 2011 criteria to rate each article. 22 , 23 Assessing the study design, accounting for limitations noted in the following summary, then adjusting the level based on the how directly each study addressed the specific question of our review, the level of evidence for all studies reviewed was rated low, levels 4 and 5. Patients enrolled in the reviewed studies were given the diagnosis of CPP, however the inclusion and exclusion criteria varied across studies. Sample sizes were small ranging from 12 to 108 patients with CPP. All studies exclusively recruited women. A number of studies reporting CPP in men were excluded due to the diagnosis of chronic prostatitis. Symptom duration for inclusion in each study ranged, from at least 3 months 3 , 14 , 16 , at least 6 months 15 , 17 – 19 or greater than one year. 10 , 11 , 21 Symptom duration was undefined in two studies. 9 , 20 For inclusion, two studies 11 , 21 required the presence of elevated resting pressures of pelvic floor muscles (PFM), indicating increased PFM state, in addition to CPP. The most common exclusion criteria reported were current or history of pelvic cancer, infection or surgery 3 , 10 , 16 , 17 , 20 and current pregnancy. 15 – 18 , 21 Some studies excluded neuromuscular conditions 10 , 17 , 21 and musculoskeletal conditions of the spine, 10 , 14 , 16 , 20 hip 20 and pelvis. 20 Three studies excluded people with other concomitant conditions such as psychiatric disorders, 10 , 17 diabetes 17 , 20 and cardiovascular conditions. 20 Two studies provided no additional detail regarding inclusion and exclusion criteria other the presence of chronic pelvic pain. 9 , 19 Asymptomatic control participants were included in eight of the 12 studies. 3 , 9 , 14 – 19 Primary inclusion criteria for all studies’ control participants was the absence of CPP, however other inclusion and exclusion criteria varied. Four articles matched asymptomatic participants to patients with CPP by age. 9 , 14 , 15 , 17 One study also matched using anthropometrics (height, weight, BMI) and subjective anthropomorphic characteristics such as muscle composition. 17 Although asymptomatic control participants reported no CPP, it is not safe to assume the control participants were free of pelvic floor disorders. Two studies 18 , 19 specifically state that pelvic floor disorder symptoms such as dysmenorrhea, dyspareunia, sexual dysfunction, intestinal symptoms and urinary symptoms were reported by their asymptomatic participants. The remaining studies did not provide information specific to pelvic floor disorders in their asymptomatic population. Parameters associated with pelvic floor muscle (PFM) performance were examined in seven of the 12 articles. 3 , 11 , 14 – 16 , 18 , 21 Five studies used a cross-sectional design to determine PFM differences between patients with CPP and asymptomatic participants. 3 , 14 – 16 , 18 Two studies, one a prospective cohort 21 and one RCT, 11 assessed the effects of botulinum toxin type A (BOTOX) injection on decreasing PFM activity. The resting state in the PFM was assessed using SEMG, 3 , 15 vaginal manometry 11 , 21 or digital palpation. 14 , 15 , 18 Pelvic floor muscle strength was examined by vaginal manometry 11 , 21 or digital palpation. 14 , 15 , 18 Finally, three studies reported on PFM relaxation responses, as recorded via SEMG 3 and digital palpation. 14 , 15 The PFM resting state, assessed by SEMG or manometry was considered elevated in patients with CPP, relative to a study-specific criteria, in five of the six articles that examined this parameter. 3 , 11 , 15 , 18 , 21 Methodological limitations in determining criteria for elevated resting state limit interpretation of these results. In two studies, 3 , 15 SEMG readings with each participant at rest was used to determine whether patients had elevated resting myoelectrical activity compared to controls. Tu et al 3 used a common clinical standard they had previously recommended 3 which classifies patients with SEMG readings > 2 microvolts as having elevated resting PFM activity. Tu et al 3 reported that 56% of patients with CPP and 40% of control participants were classified as having elevated resting PFM activity according to the > 2 microvolt criteria. However, Loving et al 15 reported a mean SEMG below this clinical standard of 1.90 microvolts. Together, these two studies suggest that the 2.0 microvolt cut off for resting state should not be used in isolation to classify PFM resting activity as elevated in patients with CPP. Baseline pressures measured by vaginal manometry have also been used to determining pelvic floor resting state. 11 , 21 Manometry determines the pressure exerted on a closed tube or balloon inserted into the vaginal canal. Though any pressure detected is typically assumed to be exerted by PFMs, manometry measurements can be impacted by intraabdominal pressure not just by PFM activity. 24 These two studies reported varied manometry pressure mean ranges to classify patients as having elevated resting pressures. However, these pressure ranges were not compared to a validated reference range. Mean vaginal pressures in patients with CPP were reported by Jarvis et al, 21 as 57 cm H 2 O (29–82) and Abbot et al, 11 as 44–49 cm H 2 O. 25 Due to no validated reference ranges and no comparison to asymptomatic participants, the clinical utility of using resting vaginal pressures to determine resting PFM state may be limited. The resting state of the PFMs is commonly determined subjectively by a clinician via digital palpation. Two articles reported an elevated resting state assessed via vaginal palpation in patients with CPP compared to asymptomatic participants 15 , 18 and one article reported no difference between patients and controls. 14 Reliability of using palpation to determine resting PFM state was either not reported 14 , 15 or considered poor. 18 These results suggest that resting state of the pelvic floor muscle should not be based solely on muscle palpation in patients with CPP. Two of the studies reporting on resting state were treatment trials focused on reducing PFM resting state as their primary intervention for CPP using BOTOX injections directly into the PFMs. 11 , 21 , 25 We included these studies in the review to determine if treatment targeting a musculoskeletal impairment would assist in understanding the relationship between the musculoskeletal impairments and CPP. In both studies, the use of BOTOX resulted in significant reduction in both non-menstrual pain, and resting vaginal pressures. 11 , 21 However, in the study by Abbot et al, 11 between-group differences were not found when the group receiving BOTOX were compared to the group receiving a saline injection. 11 Jarvis et al 21 did not include a comparison group receiving a placebo or alternate intervention. 25 In summary, elevated PFM resting state has been proposed to be associated with CPP, however significant limitations in study methodology limit our ability to make a conclusive statement about the proposed relationship. Maximal squeeze pressure of the PFMs as measured via vaginal manometry was reported in one study. 11 The primary goal of this study was to determine the effect of treatment on CPP and squeeze pressure, therefore the study was not designed to determine if muscle weakness existed among those enrolled in the studies. Three studies used palpation to estimate PFM strength. 14 – 16 Hellman et al 14 found no differences in muscle strength between those with CPP and asymptomatic controls. However, Loving et al 15 concluded that PFM weakness existed in those with CPP. The reporting of Neville et al 16 cluster analysis does not allow for conclusive statements regarding between-group differences in PFM strength. Summarizing the available articles reporting on PFM muscle strength, no quantitative studies were found that assesses the association of PFM strength and CPP. Using palpation to determine strength has significant limitations given the subjective nature, poor reliability 26 and challenges in blinding the examiner. Based on this review the relationship between PFM strength and CPP is inconclusive. Future investigation focused on determining this relationship is needed and quantitative testing should be used to assess PFM strength. Muscle performance associated with the ability to relax after contraction was measured in three studies. 3 , 14 , 15 Tu et al, 3 defined normal relaxation as the SEMG voltage dropping and maintaining below baseline after a trial of both 10 second and 2 second contractions. Significantly more patients with CPP than controls were unable to relax between contractions for both 10 sec and 2 sec contractions. Using digital palpation, Loving et al, 15 also found that women with CPP had a reduced ability to relax the PFMs. In contrast, Hellman et al 14 reported no group differences in ability to relax the PFMs as determined by digital palpation. Both palpation studies 14 , 15 used a Likert scale, with the examiner judging the ability to relax after muscle contraction as absent, partial or complete. 27 Investigating the ability of the pelvic floor muscle to relax after contraction may be a promising possible parameter for future investigation. However due to the lack of consistency of measurement and the reduced number of studies found for this review, the findings are inconclusive. More research is needed. Postural impairments may suggest maladaptations of muscle length, muscle performance or structural alignment of joint segments. While a common component of a musculoskeletal examination, the usefulness of postural assessment is often debated. Seven of the articles eligible for this review reported findings specific to posture. 3 , 9 , 10 , 16 – 19 Six were either cross-sectional or case control design to determine postural differences between patients with CPP and asymptomatic participants. 3 , 9 , 16 – 19 One study was a RCT designed to assess the effect of treatment on improving musculoskeletal impairments and symptoms associated with CPP. 10 Standing postural assessment was performed in all studies 3 , 9 , 10 , 16 – 19 and sitting postural assessment was also performed in two studies. 9 , 10 Qualitative postural assessment performed by trained examiners was most commonly reported, 3 , 9 , 10 , 16 – 19 with only one study reporting both qualitative and quantitative assessment. 17 The lumbopelvic region was the most commonly assessed, 3 , 9 , 10 , 16 – 19 followed by the thoracic and cervical spine. 9 , 10 , 17 , 19 Three studies also reported on the peripheral joints. 9 , 10 , 19 Comparisons across studies is difficult to perform due to the variation in methods to assess and determine classification of a postural impairment. Most studies reported on impairments in the sagittal plane, such as lumbar lordosis. 3 , 16 – 19 Some studies classified posture as impaired if an impairment were noted in any anatomical plane; therefore a lumbar impairment would be recorded if the patient demonstrated excessive lumbar flexion, extension or lateral bend. 9 , 10 Most reports provided some detail of their qualitative assessment; however the criteria to determine the impairment were ill-defined and often subjective, with categories such as slightly impaired, markedly impaired, hyperlordosis, and asymmetry. Among the studies that compared those with CPP to asymptomatic participants, all studies found at least one postural impairment among patients with CPP. The region of the postural impairment varied; the pelvis was most commonly involved, 3 , 9 , 17 , 18 followed by the thoracic and cervical spine. Interestingly only one of the six cross-sectional studies that included lumbar spine assessment reported significant differences between those with and without CPP. 9 Haugstad et al 10 reported the findings of their RCT to compare an experimental treatment to standard care. Although patients were randomized to two different treatment groups, the authors did not report between-group differences. Instead the authors reported the within-group change in musculoskeletal impairments such as posture, and change in patient function. The experimental treatment in this trial targeted body awareness, balanced posture and controlled movement, awareness of tension and relaxation and functional respiration. Patients with CPP demonstrated an improvement in posture and improvement in symptoms after treatment. Patients who received standard care showed minimal improvement in posture and symptoms. Although this might suggest that improvement in posture may be associated with improvement in symptoms, the authors did not report a direct association between postural changes and patient outcomes. Although each of these studies reports at least one postural impairment in patients with CPP, there are significant limitations, including no blinding of examiners 9 , 10 , 28 and poorly defined criteria for postural categories. Based on the studies reviewed, the evidence to support an association between postural impairments and CPP is inconsistent. Future studies to assess posture could benefit from objective and quantitative methods to assess alignment. Definitions for postural impairment classifications, such as excessive lumbar lordosis or flexion are needed. Assessing postural impairments in a patient population is complicated by the potential to have an impairment in any plane of motion, thereby increasing the likelihood to find an impairment. Movement pattern impairments defined as poor quality of movement and/or the presence of symptoms with movement have been suggested to be relevant to a number of musculoskeletal pain conditions. This suggestion is due to differences in movement patterns between patients and controls 29 – 31 or improved symptoms and function with improved movement patterns after treatment. 32 , 33 Three cross-sectional studies 9 , 16 , 20 and one RCT 10 assessed movement among patients with CPP. Two of those cross-sectional studies included comparisons with control participants. 9 , 16 Most movement patterns assessed focus on the trunk and hips, but two studies 9 , 10 included assessment of limb movements as well. Three of these studies used subjective judgement of a physical therapist to assess movement quality 9 , 10 , 16 and one used patient-reported pain with movement. 20 None of these studies included a validated gold standard for comparison. Haugstad et al 9 reported that women with CPP exhibit poorer quality of gait and functional movement than asymptomatic women as subjectively judged by physical therapists blinded to participant group. In a second study, Haugstad et al, 10 reported that patients’ movement quality scores and symptoms improved with physical therapy treatment, but no tests of associations between movement scores and symptoms were reported. Unfortunately, in both of these studies, 9 , 10 parametric statistics were inappropriately used to test differences or changes in ordinal data. In their study, Merits et al, 20 found that patients with CPP reported pain with active lumbar range of motion in at least one direction. The relevance to the patients’ CPP is unclear as the authors do not specify whether the patients’ reported pain was a reproduction of the patients’ CPP or some other pain such as low back pain. Since this study did not include a control group for comparison, it is not possible to know whether the prevalence of pain with lumbar active range of motion is higher in patients with CPP than asymptomatic controls. Lastly, Neville 16 subjectively judged functional hip weakness via the Trendelenburg test and lower abdominal performance via maintenance of lumbar curvature during unilateral leg extension in supine. They found that qualitative impairments were common among both patients with CPP and controls. However, the reported interrater reliability of these tests are quite low, kappa = 0.37 for the Trendelenburg test and kappa = 0.04 for the abdominal test. 34 Due to methodological and statistical limitations, there is weak evidence that movement pattern quality may be relevant for patients with CPP. Future research would benefit from improved interrater agreement, appropriate statistical methods, and potentially a more objective method of quantifying movement impairments such as video motion capture. Additional musculoskeletal impairments have been assessed in a small number of articles. These include hip joint passive range of motion, 3 , 16 hip muscle flexibility, 3 balance tests 17 and pain provocative testing on the pelvis. 3 , 18 , 20 The small number of articles that have reported each of these constructs, along with the low level of evidence rating for each article, further limit our ability to determine associations.

Given the heterogeneity of disorders included within the medical condition of CPP, there is a clear need to better understand the factors that are associated with the disorder. In this review, we attempted to focus solely on non-specific CPP as it relates to the musculoskeletal system. Based on this limited review, we are unable to make any strong recommendations regarding the relationship between musculoskeletal impairments and CPP. Musculoskeletal impairments are present in those with CPP, however it is unclear if these impairments are more prevalent than among people who are asymptomatic. A number of articles suggests a relationship exists between musculoskeletal impairments and CPP, however the studies are limited by methodological issues including widely varied inclusion and exclusion criteria, small samples, limited reliability and validity of examination items, lack of examiner blinding and questionable statistical analysis. Clearly, a standardized method to describe CPP and associated impairments as they relate to the musculoskeletal system is needed. Further, studies involving rigorous study design to assess clinically meaningful differences in musculoskeletal impairments between patients and controls is needed to further develop an evidence based approach to clinical care. One challenge of this review lies in the medical condition of CPP itself. Diagnosis of CPP is poorly defined and often overlaps with other conditions such as pelvic floor disorders, low back pain and sacroiliac joint pain. Additionally, clinicians and researchers face significant challenges in definitively ruling out organ, nerve, or integumentary pathology that may drive CPP symptoms. Although we limited our review to studies that include patients with non-specific CPP, the patients may also have had sub diagnoses such as dysmenorrhea, dyspareunia, constipation, painful bladder syndrome, and vulvodynia, endometriosis or concomitant disorders, such as incontinence, low back pain or sacroiliac pain. While some studies reported the presence of additional disorders 3 , 11 , 14 , 18 , 19 other studies were less clear. Given the anatomical proximity of the trunk, hips and pelvis to the location of pain among those with CPP, it seems reasonable that a musculoskeletal examination should be included in the patient workup when the diagnosis is unknown. Unfortunately, the evidence related to musculoskeletal examination among those with CPP is limited. A typical musculoskeletal exam includes items that are subjective in nature such as clinician judgement of impaired posture or movement patterns and patient report of pain as a response to palpation or clinical maneuver. Objective methods to quantify muscle and movement impairments are needed. We recommend this exam be performed by healthcare providers with specific experience and training in musculoskeletal disorders including physiatrists and physical therapists. We summarized the studies’ findings by each construct in isolation. For example, we summarized findings related to posture and movement in isolation but did not consider them in co-existing state. However, a musculoskeletal examination that were to include the multiple constructs examined in this review may be a more appropriate approach to better define patterns that associated with CPP. Three investigative groups included multiple constructs within their study design. 3 , 9 , 10 , 16 , 35 Neville et al 16 proposed a classification system using a cluster of examination findings that may assist in differentiating those with and without CPP. A cluster approach holds promise, especially from a clinical evaluation perspective. Clusters of positive examination tests would guide clinicians to include or exclude factors related to CPP and guide a treatment plan. Further study to investigate this approach is needed. There are several limitation of this review. Because the literature related to musculoskeletal impairments in people with low back pain (LBP) and sacroiliac joint (SIJ) pain is not standardized in description and therefore vast, we did not include publication with the primary diagnosis of LBP and SIJ pain. Likely, each disorder has enough literature to warrant an independent review. Instead, our review focused on CPP of unknown etiology. Also, we did not include publication that reported CPP as defined as merely patient report of pain on palpation of the PFMs. Pain on palpation of the PFMs can be attributed to musculoskeletal, visceral, central and peripheral sensitization and biopsychosocial disorders and problems. We were surprised we found no information related to muscle performance of the hip and trunk among patients with CPP. This may be related to our inclusion/exclusion criteria that excluded SIJ pain and urinary incontinence. A number of articles suggest that muscle performance of the trunk is associated with SIJ pain. 36 – 40 Preliminary evidence exists to demonstrate that hip muscle performance, in particular the obturator internus, may be related to PFM performance. 41 This literature is variable in design and descriptive definitions and thus, difficult to compare. Finally, because prostatitis can be a diagnosis of exclusion without definitive evidence of inflammation in the visceral organ as the source of pain, we did not include studies involving prostatitis. Standardized definitions of pain related to prostatitis is needed to better understand the literature associated with this not uncommon disorder. This review of the relationship among musculoskeletal impairments and non-specific CPP identified 12 publications that assessed pelvic floor muscle performance, resting state, strength, activation, posture and movement patterns. The level of evidence for the studies included in this review were low ranging from 4 to 5. Each of these assessments are often used in isolation or in clusters by healthcare providers and researchers. Though the literature is wide and varied and we were unable to make clear conclusions regarding validity and use of clinical tests to identify musculoskeletal impairments among patients with CPP. It is clear these constructs are important to assess and study. Further research is needed with standardized definitions and measurements to better understand the musculoskeletal system as it relates to non-specific CPP.

Chronic pelvic pain (CPP), classically defined as pain of long duration in the region of the pelvis below the umbilicus, is a debilitating condition that has a major impact on quality of life for both women and men. 1 – 3 Worldwide healthcare costs associated with CPP create a significant burden to society and appear to be increasing. 4 – 6 The costs of CPP may be growing because the conditions that cause CPP are commonly multifactorial and the parameters for defining the population are poorly reported. Pain in the pelvic region can arise from both visceral and somatic structures including the gynecologic, urologic, gastrointestinal and musculoskeletal systems. 7 Pain and dysfunction caused somatic structures of the musculoskeletal system are often considered as a diagnosis of exclusion following negative testing and negative response to treatment of the pelvic viscera. 8 Hence, conclusive objective testing to “rule in” somatic pain and dysfunction is not available and the healthcare provider is left to rely on clinical experience and the patient’s history and physical examination findings to determine a diagnosis. Unfortunately, the contributions of the musculoskeletal system to CPP are still not well understood. Postural and movement impairments of the trunk 9 , 10 and pelvis, 3 and impairments of the pelvic floor muscles including elevated resting state, 3 weakness, 11 , 12 impaired activation or delay relaxation, 3 and pain to touch 13 have been reported in individuals with CPP. If these musculoskeletal factors are found to be important to address among patients with CPP, each of these factors would call for a different treatment approach. A better understanding of the musculoskeletal impairments associated with CPP, will provide greater insight in how to clinically recognize the musculoskeletal contributions to this perplexing condition and guide the development of targeted, individualized treatment strategies. The purpose of this narrative review was to present and review the evidence of the relationship of musculoskeletal impairments found in people with non-specific chronic pelvic pain (CPP). The musculoskeletal impairments assessed in this review include pelvic floor muscle: performance, resting state, strength, activation, and posture and movement patterns. The secondary purpose is to highlight the gaps in the literature and provide recommendations for future research directions. We limited our review to non-specific CPP; therefore articles describing conditions with specific diagnoses, such as sacroiliac joint pain, low back pain, lower urinary tract symptoms, pelvic organ prolapse and chronic prostatitis were not included.