Introduction
Fibromyalgia (FM) is a chronic disorder defined by chronic, widespread pain, including axial pain and the presence of multiple tender points on physical examination [
1]. Like many chronic illnesses, the symptoms of FM extend far beyond the defining criteria. In addition to pain, most patients also have other clinical signs and symptoms, such as fatigue, disrupted sleep, impaired cognition and poor physical fitness [
2]. Moreover, a recent survey of 2,596 persons with FM reported balance problems as one of the top 10 most debilitating symptoms, with a reported prevalence of 45% [
3]. However, the relationship between the multiple clinical variables and postural control in FM is not known.
We recently reported that FM patients compared to matched controls had significantly impaired postural control in multiple subsystems in a standardized clinical examination using the Balance Evaluation Systems Test (BESTest) [
4,
5]. The FM patients also scored more poorly on balance confidence and reported six times as many falls over the past six months as healthy controls (HCs). Overall, FM severity correlated significantly with balance scores using the BESTest and self-reported balance confidence measures. It was particularly notable that FM patients were unable to maintain gait speed under cognitive distraction [
5]. Although that study highlighted how common measurable balance deficits are in FM, it did not objectively quantify the physiological basis for these balance problems. Toward this end, Russek and Fulk [
6] examined two subsystems, sensory organization and limits of stability, with dynamic posturography. They reported that up to 34% of FM subjects scored below the fifth percentile for population norms on sensory organization conditions. While this study provided support for the common occurrence of balance problems in FM, it did investigate the potential role that common FM clinical variables may have in dysfunctional postural control.
It is possible that balance disorders in FM may be associated with specific clinical and demographic findings, such as increasing age, obesity, reduced muscle strength and impaired cognition, sensory or motor deficits, or lower-extremity myofascial trigger points (MTPs). In support of examining lower-extremity MTPs, the gastrocnemius muscle is the agonist for control of forward postural sway and must contract up to 80% of maximum in response to large perturbations. The anterior tibialis muscle is the agonist for control of backward postural sway. It is postulated that lower-limb MTPs in these muscles may result in pain-related activity and lead to suboptimal muscular coordination. Further study of FM is needed to identify the relative contribution of neural, muscular and anthropomorphic challenges to postural stability in patients with FM and to develop specific balance interventions to remediate these impairments. Additionally, many of the medications used by people with FM are associated with side effects of postural instability. For example, opioids, tricyclics, hypnotics, benzodiazepines and cardiac medications can be associated with falls in the elderly [
7]. More recent evidence suggests that side effects such as dizziness may be self-limiting with newer classes of FM medications such as anticonvulsants [
8]. However, no FM study has quantified the role of medications on postural control in an effort to differentiate balance dysfunction related to the disorder versus balance dysfunction related to medications.
The purpose of the current study was to determine whether people with FM, compared to age-matched HCs, have differences on a broad array of objective balance tests (termed "dynamic posturography"), including sensory, motor and limits of stability. The relationships of balance deficits with poor strength, proprioception and lower-extremity MTPs, FM symptoms, dyscognition, balance confidence and medication use were analyzed. Last, we wanted to replicate self-reported fall data that we published in 2009 [
5]. By doing so, we have sought to establish the functional relevance of the dynamic posturography results by relating it to falls.
Discussion
Balance or postural stability is a complex task that involves the rapid and dynamic integration of multiple sensory, motor and cognitive inputs to execute appropriate neuromuscular activity needed to maintain balance [
27]. In recent surveys, persons with FM reported balance problems as one of the top 10 most debilitating symptoms, with a reported prevalence of 45% to 68% [
28]. On the basis of mounting self-report data on postural instability, we have attempted to objectively measure why postural stability could be problematic in FM. We propose that FM likely affects dynamic balance control because of altered somatosensory inputs to the CNS. Somatosensory input from muscle spindles, Golgi tendon organs and superficial and deep cutaneous afferents are the primary sensory inputs used for postural orientation in space and for automatic postural response [
27,
29,
30]. FM is defined by abnormal perception of pain with light somatosensory stimulation. It is possible that multiple pain processing dysfunctions in FM may lead to poor balance. Moreover, fall prevention requires rapid and often multiple automatic corrections coordinated by the CNS. There is evidence about multiple processing abnormalities in the CNS, including cognitive dysfunction being linked to postural instability [
27].
This study produced some novel findings. Compared to HCs, FM subjects have (1) consistent sensory deficits on dynamic posturography despite a normal clinical neurological examination; (2) poorer scores on all balance- and FM-related questionnaires, less strength, more pain areas and higher total myalgia scores but not anxiety and depression scores; (3) gastrocnemius and anterior tibialis MTPs were found in 76% to 84% of FM subjects, leading to the conjecture that active MTPs in these leg muscles may affect balance and falls as patients attempt to maintain postural stability through activation of the anterior and posterior leg muscles [
31]; (4) postural stability is best predicted by FM severity (FIQR), cognitive impairment (MASQ) and BMI; the use of opioids and/or benzodiazepines, the total number medication tablets consumed per day, and muscle strength or pain scores did not predict objective measures of balance; (5) motor tests, including limits of stability, were largely normal with the exception of longer latencies to backward perturbations and shorter end point excursion on limits of stability; and (6) these data confirm our earlier report of significantly more self-reported falls in FM patients compared to HCs.
To the best of our knowledge, computerized dynamic posturography in FM has been reported only once before. Russek and Fulk [
6] tested 32 female FM patients (mean age ± SD: 52 ± 14 years) with the same NeuroCom sensory organization test with limits of stability. Their SOT data are consistent with the results of our study. We found slightly lower SOT Comprehensive scores (60 vs. 65) with additional impairments in the Visual and Vestibular sections. Although we found a significant difference in somatosensory scores between groups (0.94 vs. 0.91), Russek and Fulk did not. However, the absolute values derived from both FM studies clearly indicate more impairment in visual and vestibular control. Bayazit
et al. [
32] suggested that women with FM have neural brainstem disintegration based on an abnormal auditory brainstem response. This notion supports both the current study results and Russek and Fulk's findings of vestibular impairments. Another area of agreement is the finding that the less taxing sensory component tasks were normal, with deficits worsening as test condition complexity increased.
No evidence of malingering was displayed in the objective data patterns obtained during posturography in the FM subjects. Other groups have identified malingerers as those who have substandard scores in condition 1 or exhibit large intertrial variation, particularly in conditions 1 and 2 compared to conditions 5 and 6 [
33,
34]. Malingering is also more likely when sway is less with eyes open in a sway-referenced visual surrounding (conditions 3 and 6) than with eyes closed (conditions 2 and 5, respectively). These data may inform those who continue to describe FM as a functional somatic syndrome and doubt the veracity of FM patients' complaints [
35].
Interestingly, the balance confidence scores in Russek and Fulk's study [
6] were slightly lower than those in our sample (70.1 vs. 81.2). The percentage of subjects scoring extremely low for their age was also greater in the Russek and Fulk study. Therefore, they reported worse overall balance confidence as well as a higher percentage score in the group consistent with nursing home patients or elderly patients receiving home care. Nonetheless, in both studies, balance confidence, as measured by the ABC questionnaire, correlated with SOT composite scores, suggesting that people with FM are well aware of their objective balance deficits. In both studies, SOT composite scores and ABC scores were correlated with disease severity (FIQ or FIQR), perhaps indicating that FM symptom severity and poor physical function are also related to postural control and balance confidence. Nonetheless, the differences in the objective (dynamic posturography) and subjective (ABC questionnaire) balance scores are not fully explained by the data. Perhaps Russek and Fulk's subjects were more distressed, since they were recruited from a support group and our subjects were recruited from a rheumatology practice. Alternatively, the difference may be due to random variation, self-selection or small sample size. Another major difference in the two studies is that we studied subjects' medical histories, physical examinations, cognition and medications, which required us to have a control group. Russek and Fulk used well-validated but computer-generated, age-matched norms for comparison, thus limiting their ability to explore the relationship between critical FM clinical variables and postural control.
The relevance of the frequent occurrence of gastrocnemius and anterior tibialis MTPs in the FM subjects is essentially conjectural at this time. More detailed analyses of lower-body MTPs in the current study was confounded by group. It may be relevant that Bazzichi
et al. [
36] reported surface electromyography (EMG) responses in 100 women with FM and 50 HCs while evaluating the anterior tibialis and distal part of the vastus medialis during isometric contraction. They found that FM patients had a significantly impaired fatigue index and a decrement of normalized median electrical frequency in the anterior tibialis and distal part of the vastus medialis during isometric contraction. Similarly, Leveille
et al. [
37] prospectively followed > 700 community-dwelling adults > 70 years old and found that chronic musculoskeletal pain predicted falls and poor balance. In a follow-up regarding the same cohort, Eggermont
et al. [
38] recently reported that mobility problems are highly predicted by tender point counts. Further study is needed to determine whether pain at MTPs, in muscles needed to lift and then push off the forefoot during ambulation, could aggravate the risk for falls in persons with FM.
Despite minimal information about consistent areas of balance deficits, some researchers have found improvements in Romberg's test (eyes opened, closed or ratio) or one-legged stance in response to multimodal exercise programs with a significant strength training element for patients with FM [
39‐
41]. Trials of magnetic pulse therapy or vibratory exercise have also demonstrated balance improvements in FM patients, although the mechanisms of action are not well understood [
42,
43]. Most recently, clinical trial data indicated one-legged stance improvement with Tai Chi and yoga exercises with mindfulness modified for FM [
44,
45]. These interventions, although not specific to the type of balance deficits we report here, are critical because poor balance and fear of falling with six other variables best predicted lower functional ability in a regression model used to study 1,735 women with FM [
46].
With regard to falls, there are no interventional data to reduce falls in FM patients. However, Rutledge
et al. [
47] have recently taken a key step toward that goal by reporting the first prospective falls data in FM patients. Their six-month descriptive longitudinal study employed standardized fall diaries with all reported falls followed up by telephone administration of the Fall Interview Guide. Eight-eight women with FM (median age 57 years, age range 21 to 69 years) reported 37 falls and 193 near-falls during the study. Interview data indicated that intrinsic factors, such as dizziness or feeling off-balance, were associated with almost all falls or near-falls. Extrinsic factors such as uneven surfaces, wet or slippery surfaces and objects in the pathway were less commonly associated with falls or near-falls. These data support earlier retrospective reports indicating that falls are common in people with FM. Further investigation is needed to identify other intrinsic fall risk factors that may be unique to FM patients, as interventions to reduce falls in the elderly in general may not be sufficient to prevent falls in this patient population.
The current study may further inform therapeutic goals for a fall prevention program in FM. Because integration of information from all sensory systems appears to be impaired in this population, it is important for clinicians such as physical therapists and exercise instructors to include sensory challenges in their exercise sessions. For example, practicing a task without the use of vision, practicing standing tasks on foam or walking while slightly turning the head encourages patients to adapt their reliance on sensory inputs for balance. Similarly, because cognition was found to correlate strongly with balance control in this population, practicing dual cognitive tasks while walking and balancing may be of value. On the basis of our findings, the clinician could also consider the number of comorbidities, cognitive training and careful monitoring of the effects of medication. Bearing in mind the possible relevance of active lower-limb MTPs to postural imbalance in some patients, future researchers could assess the efficacy of specific MTP therapies [
2]. Another element of possible importance in balance therapy for FM patients is how one reacts to being pushed off balance. We found that FM subjects scored lower on the midrange perturbations. When the postural perturbation was in the midrange, the FM subjects had difficulty recovering their balance, so balance reactions such as these should be incorporated into physical therapy or exercise rehabilitation in this patient population.
A limitation in fully explaining the findings regarding balance and falls in the current study is that gait was not evaluated. However, other researchers have reported gait abnormalities in FM patients. Heredia Jiménez
et al. [
48] reported that 55 women with FM compared to 44 controls exhibited significant differences in gait velocity, stride length, cadence, single- and double-support ratio, stance and swing phase ratio using the GAITRite System (CIR Systems, Inc., Havertown, PA, USA). Furthermore, each of these decrements correlated with the total FIQ score, such that persons with greater impairment from FM had poorer gait performance. Similarly, Auvinet
et al. [
49] used computerized gait analysis of 14 middle-aged women with FM and 14 matched controls and found similarly altered results in stable walking. Also, Pierrynowski
et al. [
50] reported that women with FM walk with different muscle recruitment patterns compared to controls. It is not known whether these patterns develop to minimize pain during ambulation. Indeed, Graven-Neilsen
et al. [
51] suggested that increased activity of antagonistic muscle and decreased activity of agonistic muscle are consistent with Lund
et al.'s [
52] pain adaptation model. Graven-Neilsen
et al.'s hypothesis was based on a study in which hypertonic saline was infused into either the anterior tibialis or gastrocnemius muscle and monitored by EMG during treadmill walking [
53]. Taken as a whole, these studies suggest the need to further explore the relationship between lower-extremity MTPs in FM patients and postural stability and falls.
A novel finding in this study was the strong link between balance and cognition. For the past decade, researchers have amassed substantial cross-sectional data documenting cognitive dysfunction in FM patients, first suggesting that FM patients performed similarly to HCs who were 20 years older [
54]. Multiple studies in both FM and widespread pain reveal self-reports of concentration, alertness and memory problems as well as challenges in completing demanding cognitive tasks [
55‐
58]. Cross-sectional data are emerging that indicate that physical performance predicts attention and speed of cognitive processing in FM [
58]. The relationship between poor fitness and impaired cognition in FM patients is plausible and requires further investigation [
59‐
62].
Limitations
As this was a pilot study, the sample size was small (
n = 25 FM patients vs. 27 HCs). As it was exploratory, it had multiple dependent variables. However, the likelihood of reporting spurious findings was minimized by using the Bonferroni correction and setting the
P value at < 0.01. In fact, most
P values were < 0.001. One area where sample size may have been a factor was in our regression model regarding medications in FM patients. As few patients (
n = 8 of 25) were taking opioids, hypnotics or benzodiazepines, a larger study is needed to confirm the lack of association between balance and CNS-acting medications in FM patients. Alternatively, FM subjects could be tested while on and off CNS-acting medications to clarify the role of these agents in postural stability. However, in a three-year prospective study of 1,002 community-dwelling women ages 65 years and older with chronic widespread pain, an alarming percentage of women fell each subsequent year (39%, 36% and 39%, respectively). These researchers reported a protective effect of analgesic medication use in reducing fall rates and concluded that preventing falls in patients with widespread pain requires a multifactorial treatment approach that includes the use of pain medications [
63].
Another possible limitation of our study is that we relied on self-reports of normal vision on the basis of a visit to an optometrist or ophthalmologist in the past year. Future researchers could measure FM patients' visual acuity directly. As expected, the HCs had more education and higher incomes. The HCs were largely academic employees. These variables were not generally associated with postural stability. Moreover, when adjusting for these differences, significance levels did not change. The examiners were not blinded to group, but this potential limitation was minimized because of the objective nature of computerized posturography and the use of self-reports rather than examiner-administered questionnaires.
Another limitation of the study was the retrospective nature of fall reporting and the lack of inquiries regarding the consequences of falls (for example, fracture, further limitation in physical activity due to fear of falling). Moreover, we did not question subjects regarding intrinsic versus extrinsic variables that may have contributed to each fall. For example, falls related to attempts to rapidly get to a bathroom because of irritable bowel, bladder and/or stress urinary incontinence is plausible in this population. The finding that 15 of 25 subjects fell three or more times indicates that falls were commonly experienced by most FM subjects and that scores were likely not led by outliers. It is unlikely that adjustment to new medication was related to falls in this study, as subjects were required to be on a stable dose of medications for at least the past three months. It is also unlikely that advancing age was related to falls, as most of our sample was middle-aged (mean age 50.8 years), with the oldest participant being 59 years of age. We purposely recruited a younger sample to minimize falls and balance perturbations related to age. Twice we have demonstrated a significantly greater number of falls in FM patients compared to age-matched controls. In another study of falls in 70 community-dwelling people with FM ages 50 years and older, Jones
et al. [
46] found the following variables best predicted falls: gait velocity, cognitive performance, number of cardiovascular drugs, total number of drugs, age, uncorrected vision, perceived postural instability, clinical balance testing and lower-body strength. However, only 42 of 70 persons self-reported falls in the past year, which was considerably lower than the data reported in the current study or by Jones
et al
[
5]. Jones
et al
[
46] did not quantify FM severity in their study, and thus it is not known whether FM severity was a potential predictor of falls. Prospective validation of self-reported falls and their negative consequences is needed.
Another area for further study of postural stability in FM patients is the role of BMI. BMI was higher in FM compared to HCs, and this difference remained in a regression model that predicted sensory balance deficits. However, the literature is mixed in implicating overweight versus morbid obesity with regard to falls or postural stability [
64‐
68]. It is not known whether higher body weight impairs balance control. It is likely that muscle strength does not increase in proportion to total weight in obese individuals; therefore, obese persons may not be able to generate adequate force required to quickly regain postural control. This study was not designed to determine whether obesity was confounded by poor physical fitness. We did find, however, that BMI did not correlate with scores on the FIQ-R Physical subscale (
r = -0.0304,
P = 0.8852). Future studies could explore this potential relationship by determining aerobic capacity or other standardized laboratory measures of fitness.
The computerized dynamic posturography results of the middle-aged adults with FM in this study yielded scores that are comparable to those of healthy persons in their eighth decade of life, based on normative computer-generated data. Thus, the balance deficits and fall frequency reported here are clinically significant. The main targets of physical and rehabilitative medicine at this point are the relief of symptoms and an improvement in the activities of daily living for patients with FM. Balance training is generally not included in FM treatment, because it is not known what aspect of balance, if any, is involved in FM. We believe that balance impairment is an important component of FM and is currently not being assessed in the clinic or being treated by therapists. Balance impairment can negatively affect patients' quality of life by increasing the risk of falling, fall-related injuries and fear of falling [
69]. By identifying and quantifying consistent balance abnormalities in FM patients, we hope to improve the effectiveness and specificity of treatments as well as to more effectively assess outcomes of treatments in relieving balance difficulty. More practically, balance training and fall prevention as an aspect of rehabilitation are not currently the standard of care.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
KDJ participated in all phases of the study, including study design; oversight of MS students; data collection, entry, analyses and interpretation; and final manuscript preparation. LAK participated in balance-related contributions to study design, analyses, interpretation and final manuscript preparation. SDM performed all statistical analyses and final manuscript preparation. RMB participated in all phases of the study, including study design, analyses, interpretation and final manuscript preparation and provided funding from the Fibromyalgia Information Foundation. FBH participated in balance-related contributions to study design, analyses, interpretation and final manuscript preparation and provided access to the Clinical Research Dynamic Posturography System (NeuroCom International, Inc.) and National Institute on Aging funding to support the physical therapist who collected posturography data. All authors read and approved the final manuscript.