Association between musculoskeletal function and postural balance in patients with long ‐ lasting dizziness. A cross ‐ sectional study

Background and purpose: Reduced balance and musculoskeletal pain are frequently reported among patients with long ‐ lasting dizziness. However, the association between musculoskeletal function and postural sway among these patients has not been examined. The objective of this study was to examine if there is an association between aspects of musculoskeletal function and postural balance in patients with long ‐ lasting dizziness. Methods: This was a cross ‐ sectional study, using data of 105 outpatients with long ‐ lasting dizziness. Aspects of musculoskeletal function was assessed by examining body flexibility, grip strength, preferred and fast walking speed, in addition to musculoskeletal pain. Musculoskeletal pain was evaluated using the Subjective Health Complaints questionnaire. Postural balance was assessed by path length of postural sway by using a balance platform on both firm and soft surfaces, with eyes open and closed. The association between musculoskeletal function and postural sway was assessed using linear regression analyses. Results: When adjusting for age and gender we found that on a firm surface, there was an association between increased musculoskeletal pain and increased postural sway measured with eyes open ( p = 0.038). In addition, there


| INTRODUCTION
Dizziness is one of the most frequently reported symptoms in primary care and affects adults of all ages (Neuhauser et al., 2008).
Dizziness is often followed by other symptoms such as psychiatric comorbidity and musculoskeletal pain (Kvåle et al., 2008;Lahmann et al., 2015;Malmström et al., 2020) and reduced function such as balance deficits (Herssens et al., 2020;Söhsten et al., 2016;Wilhelmsen & Kvåle, 2014) and decreased gait speed (Marchetti et al., 2008). However, little is known about whether or how aspects of musculoskeletal function and pain affect balance in this patient group. Previously, gait function has been linked to falls and balance in patients with vestibular disorders (Whitney et al., 2000) and grip strength has shown association to postural control in older adults (Strandkvist et al., 2021;Wiśniowska-Szurlej et al., 2019).
Persons with persistent dizziness may be prone to musculoskeletal pain and reduced function (Kvåle et al., 2008) as dizziness often is aggravated by head movements. The patients may attempt to minimize head motion during movement leading to avoidance behavior and "en bloc" movement of the head and trunk (locking the head on the trunk) causing a more rigid movement pattern.
This may lead to muscular stiffness, pain, and reduced musculoskeletal function. To keep and maintain balance, we are dependent on a fine-tuned interaction between the visual, vestibular, and somatosensory systems (Kristjansson & Treleaven, 2009). Studies have implied that neck pain may disrupt afferent information from the cervical spine and thus influence balance (Knapstad et al., 2019;Treleaven, 2017). However, this relationship is not well explored in with musculoskeletal pain in general. It is possible to theorize that musculoskeletal ailments would influence balance, either through pain which may disturb the afferent proprioceptive information from the somatosensory system (Kristjansson & Treleaven, 2009) or through loss of function due to avoidance behavior. Compared to healthy controls, increased postural sway, a commonly used measure of postural balance, has been documented in persons with vestibular disorders (Allum et al., 2001), general musculoskeletal pain (Lihavainen et al., 2010), non-specific low back pain (Ruhe et al., 2011), idiopathic neck pain and whiplashassociated disorders (Silva & Cruz, 2013). Maintaining balance is critical to independence in daily life (Shumway-Cook & Woollacott, 2017, p. 153), and both dizziness and increased postural sway are strong predictors of falls, which can lead to further disability (Kollén et al., 2017). Knowledge about the potential associations between aspects of musculoskeletal function, pain and balance in a population prone to balanced disturbance is therefore valuable for practitioners treating this group. The aim of this study was to investigate possible associations between aspects of musculoskeletal function, such as gait, strength, flexibility and pain, and postural sway in persons with long-lasting dizziness.

| Study design
This cross-sectional study was carried out as part of the LODIP study (Kristiansen et al., 2019), a project examining the effect of vestibular rehabilitation combined with cognitive therapy in persons with longlasting dizziness in primary care. The study was conducted at the Western Norway University of Applied Sciences (HVL), Bergen, Norway. The present study analyzed data from baseline assessments.

| Recruitment and sampling
Participants were recruited through general practitioners, physiotherapists, and otorhinolaryngology clinic, newspaper ads, and on social media from the region in and around the municipality of Bergen, Norway, between 2016 and 2019. In total, 229 potential participants were screened by telephone interview, and 86 of these were excluded. After the initial screening, 143 persons were invited to further assessment, and another 36 persons were excluded. In total, 107 participants who fulfilled the following criteria were included: age 18-70 years, acute onset of dizziness, symptoms lasting for more than three months, and dizziness aggravated by head movements.
Exclusion criteria were non-vestibular causes of dizziness (e.g., neurological disorders), diseases with fluctuating vestibular function, patients who had had treatment for benign paroxysmal positional vertigo within one month of testing, diseases where fast head movements were contraindicated, severe/terminal pathology, inadequate verbal and written Norwegian language proficiency, or inability to attend test locations. The participants signed an informed consent and completed baseline assessments at HVL. Three experienced physiotherapists familiar with the testing procedures conducted the assessment.

| Sample size
This study uses the same study population as the main study (Kristiansen et al., 2019). The main study was designed as a randomized controlled trial comparing two intervention groups, with a power sample calculation estimating at least 96 participants. To account for possible dropouts the goal was to include 125 participants; however, 107 participants were finally included, still within the sample size calculation (Kristiansen et al., 2019). As the sample size of this study was based on already collected data, a post hoc sample size calculation was not recommended (Lydersen, 2015).

| Musculoskeletal pain
To assess musculoskeletal pain during the last 30 days, the subscale musculoskeletal pain (SHCmusc) from the Subjective Health Complaints (SHC) questionnaire was used. SHCmusc contains eight items related to muscle pain (headache, pain in the neck, upper and lower back, shoulders, arms, migraine, and legs). Severity of pain was scored on an ordinal scale ranging from zero (no complaints) to three (serious complaints), giving a sum score from 0 to 24 points. The structural validity and reliability of the scale were found to be acceptable for use among the general population (Eriksen et al., 1999).

| Body flexibility
The Global Physiotherapy Examination (GPE) was used to evaluate bodily aberrations, and the subdomain bodily flexibility (GPEflex), which includes testing of lumbo-sacral flexion, head-nod flexion, shoulder retraction, and elbow drop, was used. Each item was scored on a 15-step numbered scale ranging from −2.3 to +2.3. Optimal flexibility is defined as zero, while a higher score indicates less flexibility. The total sum score of the four tests was reported and included in the analysis. The GPEflex has demonstrated to be reliable and valid in patients with long-lasting musculoskeletal pain, and also, to discriminate between those with and without musculoskeletal pain (Kvåle et al., 2003).

| Grip strength
To assess musculoskeletal strength, the grip strength-test was administered using a handheld dynamometer (Mie medical research Myometer). The mean grip strength of two attempts in each hand was measured in kg. Grip strength is a reliable and valid test for musculoskeletal strength (Abizanda et al., 2012).

| Preferred and fast walking speed
The participants walked a 6-m pathway and were asked to walk at their preferred speed. To avoid the acceleration and deceleration phase, a 1-m start-up and slow-down at the start and end of the pathway were set. A stopwatch was used to measure the time in seconds. The mean preferred walking speed of two attempts was calculated in meters per second (m/s). The same protocol was used for fast walking, except that the participants were instructed to walk as fast as possible. The walking test protocols used in this study are considered to be valid and reliable (Hall & Herdman, 2006;Heitzman, 2013;Muhaidat et al., 2014).

| Dizziness symptoms
The Vertigo Symptom Scale-short form (VSS-sf) was used to assess self-perceived severity of dizziness. The scale consists of 15 items assessing symptoms during the past month. Each item is scored according to the frequency of symptoms on a five-point ordinal scale from 0 (never) to 4 (very often/almost every day), providing a total score ranging from 0 to 60 (Wilhelmsen & Kvåle, 2014

| Statistical analysis
Distribution of variables was examined using histograms and scatter plots. Descriptive data are presented as frequencies (percentages), means and standard deviations (SD), or median and quartiles as HAUKANES ET AL.
-3 of 9 appropriate. Linear regression models were used to test the associations between postural sway (dependent variable) and aspects of musculoskeletal function and pain (independent variables). The different measures of musculoskeletal function (walking speed, grip strength, body flexibility, and pain) were analyzed separately in each condition of sway (EOfirm, ECfirm, EOsoft, and ECsoft), in univariate and adjusted models. Age and gender were used as adjustment variables as balance and muscular function can be influenced by both (Gribble et al., 2012). Since the four postural sway variables were significantly skewed, they were log-transformed so that the residuals approached normality. To facilitate interpretation of the regression coefficients, they were back-transformed after the regression. Analyses were two-tailed with a p-level of less than 0.05. Statistical analyses were performed using IBM SPSS Statistics version 26 (IBM Corp.).

| RESULTS
Two participants were excluded from further analyses due to extreme scores in the four postural sway tests. Thus, the analysis included 105 participants, of which 79 (75%) were women and mean age was 49 years (SD = 12.9). Mean score of VSS-sf was 21. Duration of dizziness ranged from 3 to 482 months, with a median of 72 months.
Postural sway EOfirm showed the lowest values of sway with a median path length of 137 mm, while ECsoft showed the highest values of sway with a median path length of 844 mm (Table 1).
The adjusted models on a firm surface showed a statistically significant association between SHCmusc and postural sway in the EO condition, and a 1-point increase on the SHCmusc resulted in a 2% increase in sway. An inverse significant association was found between GPEflex and postural sway in the EO test (Table 2). A 1-point increase in the GPEflex resulted in a 6% reduction in sway with EO and a 5% reduction in sway with EC in the adjusted model.
In the adjusted models on a soft surface, there was a significant association between grip strength and postural sway in the EC condition, with a 1-point increase (kg) in grip strength resulted in a 2% reduction in sway, and finally, a significant association between fast walking speed and postural sway in the EO condition, with a 1-point increase (m/s) in walking speed was associated with a 52% reduction sway (Table 3).

| Main findings
The present study has demonstrated associations between musculoskeletal function, pain and postural sway in persons with longlasting dizziness. Analyses with adjustments for age and gender showed that increased musculoskeletal pain, decreased fast walking speed, and decreased grip strength all were associated with increased sway. Surprisingly, decreased body flexibility was associated with decreased sway.
We found that increased self-reported musculoskeletal pain was associated with increased postural sway (EOfirm). These findings are in line with studies reporting that geriatric patients with general pain have increased postural sway (Lihavainen et al., 2010), and similar associations were found with localized pain in patients with low back pain (Ruhe et al., 2011). Since balance is dependent on somatosensory input from the musculoskeletal system (Lihavainen et al., 2010), pain may cause disturbances (Brumagne et al., 2000), leading to a greater reliance on visual and vestibular information to maintain postural control.
Postural sway may therefore increase when sensory information is reduced. Surprisingly, no association was found between musculoskeletal pain and postural sway when visual input was removed, and somatosensory information disturbed (ECsoft). Previous studies have found associations between pain intensity in the neck and postural sway (Knapstad et al., 2019;Ruhe et al., 2013), and moderate to severe musculoskeletal pain has been found to associate with increased postural sway compared to patients with mild or no pain (Lihavainen et al., 2010). The mean SHCmusc score T A B L E 1 Clinical characteristics of the 105 participants with long-lasting dizziness  -5 of 9 in our participants was 7 (score ranges from 0 to 24), while the mean SHCmusc in the general population is reported to be 4.7 (Ihlebaek et al., 2002). This indicates that the participants in our study were only mildly affected by pain and that higher levels of pain could demonstrate associations also when vision and sensory input were disturbed. Even though visual dependency is common in persons with vestibular disturbances (Maire et al., 2017), the same is not established in patients with somatosensory disturbances, making the interpretation challenging. Another factor is that pain levels are a difficult construct to measure, highly influ-  (Fujimoto et al., 2009). In addition, grip strength has been associated with reduced balance (Alonso et al., 2018;Singh et al., 2015), supporting the results from our study.
Among patients with dizziness, restricted body flexibility has been described compared to healthy persons (Kvåle et al., 2008).
Therefore, an association between reduced body flexibility and increased postural sway was expected, as flexibility is deemed beneficial for balance strategies. On the contrary, we found that reduced body flexibility was associated with decreased postural sway (EOfirm). An explanation for this unexpected finding may be that reduced flexibility acts as a compensatory mechanism to reduce sway on firm surfaces. Also, the total score was in this study was 3.5, which is somehow higher than what Kvåle et al. (2003) found among healthy participants (3), but lower compared to a previous study in patients with dizziness (4.4) and patients with dizziness and additional neck pain (4.8) (Knapstad et al., 2020).
In the fast walking condition, we found an association between walking speed and postural sway (EOsoft), indicating that a reduced ability to walk at a more rapid speed was associated with increased postural sway. This association is interesting, as walking speed can be used to detect persons at a higher risk of major health-related events (Cesari et al., 2005). It has been stated that persons reduce their walking speed to cope with increased risk of falling (England & Granata, 2007). In addition, patients with dizziness may be more reluctant to move at higher speeds due to a pre-existing sensory deficit, which may explain the association with increased postural sway.
Last, it could be argued that the included population was only mildly affected by musculoskeletal pain, with scores in the lower aspect of the normal range within their age group, in for instant walking speed (Fritz & Lusardi, 2009) and grip strength (Massy-Westropp et al., 2011). This may have influenced the results, and perhaps higher disability would have resulted in a stronger relationship with postural sway. However, the aim of this paper was merely to examine if an association existed between the different aspects of musculoskeletal function and postural sway, and not if their physical function.

| Strengths and limitations
A strength of our study is the use of reliable and valid measurement instruments, in addition to the fact that the participants are likely to be representative of persons with long-lasting dizziness in terms of age and gender (Kvåle et al., 2008). In addition, the VSS-sf was high (mean score 21), indicating a high severity of dizziness at inclusion.
This and the large number of participants (n = 105), increases the external validity of the study. Despite this, there may be a selection bias as only the most motivated and least affected patients tend to volunteer for research studies (Rothwell, 2005). A participant pool of persons with reduced function and higher pain levels may have resulted in a stronger association between musculoskeletal function and postural sway.
It could be argued that walking speed is not a representative measure of musculoskeletal function. However, walking speed is a vital sign for overall health where a well-functioning musculoskeletal system is indeed important (Fritz & Lusardi, 2009). It is also considered as a strong predictor of ADL disability in communitydwelling older adults (Vermeulen et al., 2011). As neurological disorders were excluded from this study, we believe that gait speed is an important aspect of musculoskeletal function in the study population.
Four tests on a balance platform were used to examine postural sway as we wanted to include aspects of excessive reliance on vision and a possible reduction in proprioceptive input (Ruhe et al., 2011). In individuals with vestibular deficits, the Balance Evaluation Systems Test (Horak et al., 2009) has been described as valid to assess balance. There are disadvantages to using postural sway alone as a measure of balance, as it does not capture the myriad aspects of balance, and a combination of several balance tests may be preferable (Horak et al., 2009 Finally, the associations were weak, and the low explanatory power (R 2 ) underscores that musculoskeletal function may have a limited effect on postural sway in persons with long-lasting dizziness.
However, the aim of this study was merely to examine whether an association existed.

| IMPLICATION FOR PHYSIOTHERAPY PRACTICE
We found small but significant associations between postural sway and musculoskeletal function among patients with long-lasting dizziness. The results imply that reduced musculoskeletal function is to a certain degree associated with reduced balance, in persons with long-lasting dizziness. For clinicians treating these patients, the relationship may be of importance since balance is affected in patients with dizziness, and musculoskeletal issues may add to these balance issues. However, the detected associations were small, and the clinical significance is uncertain. Since altered postural control has a multitude of possible causes, these findings need to be corroborated in future studies.