Practice Points
- Brief cognitive assessments, such as the Symbol Digit Modalities Test and Trail Making Test part A, may provide additional context when evaluating balance performance in individuals with multiple sclerosis.
- Evaluating concern about falling using tools such as the Falls Efficacy Scale-International could help identify individuals who may experience greater postural instability under challenging conditions.
- Cognitive and psychological factors may contribute to balance performance and could be considered when interpreting fall risk in clinical assessments.
Multiple sclerosis (MS) is a neurodegenerative disease characterized by axon demyelination in the central nervous system.1 Balance difficulties are among the most disabling symptoms of MS, affecting approximately 75% of individuals with MS.2,3 Balance impairments are a key contributor to fall risk in MS,4,5 with 56% of individuals with MS experiencing a fall in any 3 months and 37% classified as frequent fallers.6 Falls in MS can lead to physical injury,7 activity restriction,8 and diminished quality of life.9 Importantly, these experiences often lead to increased concern about falling, reported by up to 60% of individuals with MS,10 and have been associated with motor impairment.11,12 Together, these findings underscore the need for a deeper understanding of balance dysfunction in people with MS.
In addition to motor impairments, cognitive deficits are common in MS and have been identified as significant risk factors for falls.13-16 In individuals with MS, declines in cognitive domains such as information processing, attention, and memory,17-19 which can impair everyday functioning and quality of life,20 have been noted. Prior research data have demonstrated that adding a cognitive task during postural assessments reduces balance performance in individuals with MS.21-23 In addition, greater cognitive impairment has been linked to higher fall frequency and accelerated mobility decline.13,15
Although these findings support an interaction between cognitive and balance function in people with MS, most studies have relied on simple balance assessments and global cognitive scores, limiting insight into how specific cognitive domains contribute to postural control. Although prior work has linked cognitive domains such as processing speed, attention, verbal function, and memory to static balance performance, many study designs have relied on relatively simple or low-demand balance assessments, such as quiet standing under eyes-open conditions or clinical balance tests that do not systematically manipulate sensory input or base of support.24 Conversely, studies that have employed more comprehensive or complex postural sway assessments, such as instrumented posturography involving sensory manipulation (eg, eyes closed, unstable surfaces) or dual-task paradigms, have often relied on global cognitive measures, limiting insight into domain-specific contributions.25 Thus, few studies have simultaneously examined the role of discrete cognitive domains in the context of complex or challenging balance tasks. Although previous research findings have established links between cognitive processes such as processing speed, immediate memory recall, visuospatial memory, and attention as well as more complex motor tasks such as backward walking and stair navigation,26-28 their specific associations with the complexity of standing balance assessments in MS remain poorly understood. Taken together, these gaps highlight the need for studies that simultaneously employ domain-specific cognitive measures and complex balance tasks to better characterize the interplay between cognition and postural control in MS.
Therefore, this study examines the associations between domain-specific cognitive functioning and standing balance in complex, sensory-conflicting balance conditions. Based on previous findings,16,24,29 we hypothesized that processing speed, visuospatial memory, and attention would be associated with balance performance and that stronger relationships would be observed with more complex balance tasks. Moreover, given the established link between concern about falling and impaired balance under increasingly complex conditions, we aimed to examine how concern about falling influences balance control across varying levels of postural challenge.30 By elucidating how distinct cognitive domains and fall-related concerns relate to postural control under different task demands, this study aims to deepen our understanding of cognitive-balance interactions in MS and inform more targeted fall-risk assessment and intervention strategies.
Methods
Participants
All study procedures were approved by the Wayne State University Institutional Review Board, and all participants provided informed consent before participation. Participants were recruited using convenience sampling methods through the Wayne State University MS Center, the local chapter of the National Multiple Sclerosis Society, and a laboratory registry of individuals with MS who had previously expressed interest in research participation. Participants were eligible to participate in the study if they were 18 years or older, were diagnosed with MS using the McDonald criteria,31 and reported a Patient-Determined Disease Steps (PDDS) score of 6 or less.32,33 Participants were excluded if they had experienced an MS relapse or exacerbation of symptoms within the past 30 days, had a comorbid neurological disorder or another condition that would impact their motor or cognitive function, or were unable to comprehend and follow study commands.
Procedures
All balance and cognitive assessments were completed in a single testing session. Via REDCap, participants completed a series of surveys, which included demographic information (eg, age, sex, height, weight) and clinical history, such as disease severity (PDDS) and symptom duration. Cognitive and balance assessments were administered as well as the Falls Efficacy Scale-International (FES-I), a validated measure of concern about falling during everyday activities.34 To ensure safety, participants wore a gait belt and were continuously supervised by trained research staff during all balance assessments. To minimize fatigue, participants were offered seated rest breaks between balance trials and conditions, and rest duration was individualized based on participant need.
Balance Assessments
Participants completed 4 different balance conditions: eyes open with feet apart (EOFA), eyes open with feet together (EOFT), eyes closed with feet apart (ECFA), and eyes closed with feet together (ECFT). Each condition was performed twice for 30 seconds, and sway measures were averaged across trials. During testing, participants stood with arms crossed over their chest, facing a fixed visual target at eye level. For feet-apart trials, feet were placed shoulder-width apart; for feet-together trials, feet were placed as close together as possible. These balance conditions were designed to progressively challenge postural control by manipulating visual input and base of support, 2 key contributors to balance regulation.35 Prior studies have used similar modifications to assess balance in older adults and individuals with neurological disorders.36,37
Cognitive Assessments
Participants completed a battery of standardized cognitive assessments. The Symbol Digit Modalities Test (SDMT) assesses information processing speed by asking participants to match symbols with numbers for 90 seconds. The SDMT was presented orally, and the number of correct responses within the time frame was used in the analysis.38 The Brief Visuospatial Memory Test-Revised (BVMT-R) tested visuospatial memory.39 Participants viewed a 2×3 matrix of 6 shapes for 10 seconds, which was then removed. They were then asked to recall and draw the shapes in their exact layout as accurately as possible. Immediate recall was calculated as the number of correct responses across 3 trials. The California Verbal Learning Test (CVLT) measures auditory-verbal learning and memory.40 Participants listened to and freely recalled a list of 16 items read aloud; this was repeated 5 times. The total number of correct responses across the 5 trials was used in the analysis. Finally, the Trail Making Test (TMT) parts A and B assessed attention and cognitive flexibility.41 For TMT-A, participants connected numbers in ascending numerical order from 1 to 25; for TMT-B, participants connected numbers 1 to 12 and letters from A to L in ascending numerical and alphabetical order. The time taken to complete each assessment was used in the analysis. These domains were selected based on their relevance to MS-related cognitive decline and prior associations with complex motor tasks such as backward walking and stair navigation.17-19,26,27
Data Collection
Participants were equipped with wireless Opal inertial sensors (128 Hz; APDM), and quantitative balance measures were obtained via Mobility Lab’s instrumented test of postural sway (ISway) protocol.42 The ISway algorithm automatically generates sway measures in the time and frequency domains using the triaxial accelerometer from the lumbar sensor. Additional protocol details, as well as the validity and sensitivity of this approach, are described by Mancini et al.42 Sway velocity was defined as the mean resultant sway velocity across the 30-second recording, calculated automatically by the Mobility Lab ISway algorithm from lumbar triaxial accelerometry data. This metric reflects the combined anterior-posterior and mediolateral components of postural sway.43 All participants were able to stand independently without assistive devices during testing. No trials were terminated early, and all participants completed the full 30-second duration across conditions. For safety purposes, participants wore a gait belt and were supervised by trained research staff; however, no physical assistance was provided during data collection.
Statistical Analysis
All survey, balance, and cognitive outcomes are presented as mean (SD) unless otherwise noted. To assess normality of data, skewness and kurtosis values were examined. Given the nonnormality of all balance assessments (EOFA, EOFT, ECFA, ECFT), Spearman ρ was used to examine the correlations between sway velocity under all 4 balance conditions and cognitive outcomes and concern about falling. Next, a linear regression with backward selection was used to identify variables that significantly explain variance in balance performance under all 4 balance conditions. Only variables demonstrating significant associations with sway velocity within each condition were entered into the corresponding regression model to support parsimonious estimation. Variables were removed one at a time until all included variables were deemed significant predictors of the dependent variable (P < .05). All linear regression models were corrected for age, sex, and PDDS. Height was examined as a potential anthropometric confound by testing Spearman correlations with sway velocity across all balance conditions. Height was not significantly associated with sway velocity in any condition (all P > .15); therefore, height was not included as a covariate in subsequent regression models. To assess potential multicollinearity among predictors, variance inflation factors (VIF) and tolerance values were examined for all final regression models. VIF values ranged from 1.02 to 1.69, indicating no evidence of multicollinearity. All data analyses were performed in SPSS version 29.0 (IBM Corp).
Results
A convenience sample of 54 participants was included in the study. Table 1 provides an overview of the demographic and clinical characteristics of the sample. Participants had an average (± SD) age of 50.96 (±11.56) years, ranging from 31 to 74 years. Ambulation disability, as interpreted by PDDS, ranged from 0 to 6, with a median of 2, indicating mild to moderate walking impairment.
Table 2 presents the results of the correlation analyses for all balance conditions. In the eyes-open conditions, feet-apart (EOFA) balance showed a significant negative correlation with the SDMT (ρ = –.29; P = .04) whereas no variable tested had a significant correlation with feet together (EOFT; all P > .05). In the eyes-closed conditions, feet apart (ECFA) had a significant positive association with TMT-A (ρ = .42; P < .01), TMT-B (ρ = .43; P < .01), PDDS (ρ = .37; P = .01), and FES-I (ρ = .38; P = .01) and a significant negative association with SDMT (ρ = –.43; P < .01) and BVMT-R (ρ = –.32; P = .02). For the feet-together condition (ECFT), a significant positive association was observed with PDDS (ρ = .27; P = .05) and FES-I (ρ = .39; P < .01) whereas a significant negative association was observed with SDMT (ρ = –.38; P < .01). No significant associations were observed between sway velocity and verbal memory (CVLT) in any balance condition (all P > .05). Visuospatial memory (BVMT-R) and executive function (TMT-B) were associated with sway only in the ECFA condition and did not demonstrate consistent relationships across other balance tasks.
Results of backward linear regression for EOFA, ECFA, and ECFT analyses after correction for age, sex, and PDDS are presented in Table 3. As no significant correlations were observed for EOFT, no regression analysis was conducted. For EOFA, SDMT remained independently associated with sway velocity (β = −.34; t = −2.27; P = .03). For ECFA, the final model included TMT-A (β = .37; t = 2.73; P = .01) and FES-I (β = .29; t = 2.12; P = .04), with TMT-A demonstrating the strongest association based on standardized β coefficients. For ECFT, FES-I demonstrated a strong independent association with sway velocity (β = .51; t = 3.43; P < .01) after correcting for age, sex, and PDDS.
Discussion
This study aimed to examine how domain-specific cognitive functioning relates to standing balance performance under increasingly challenging, sensory-conflicting conditions in individuals with MS. Although findings from prior studies have shown associations between global cognition and balance,24 few have systematically assessed how distinct cognitive domains relate to static balance, particularly under conditions that manipulate base of support and visual input. Our findings demonstrate that cognitive performance, particularly in processing speed (SDMT) and attention (TMT), is associated with postural control and that the pattern of independent associations varies depending on the level of sensory and postural challenge. Notably, concern about falling also emerged as an independent predictor of balance performance under the most demanding conditions. Examination of standardized β coefficients further suggested a shift in the strongest independent associations across increasing task demands, with processing speed showing the strongest association under low-demand conditions, attention under moderate sensory challenge, and concern about falling under the most demanding postural condition. These results underscore the importance of integrating both cognitive and psychological screening into fall-risk assessments for individuals with MS.
The relation between cognition and balance performance in individuals with MS has been well-documented, but often in studies using less challenging tasks or global cognitive composites. In line with recent work by Kalron et al, which demonstrated associations between static posturography and executive, verbal, and memory domains in individuals with MS,24 our findings further support the presence of domain-specific cognition-balance relationships. However, the pattern of domains observed in the current study differed somewhat, with processing speed and attention emerging as the most consistent contributors. These differences likely reflect variations in task design and sample characteristics. Whereas prior studies examined static balance without systematically manipulating sensory demands, our progressively challenging conditions (altering visual input and base of support) may have more selectively taxed attentional control and sensory integration processes. In this context, attention appeared to play a more prominent role, particularly under eyes-closed conditions where reliance on nonvisual sensory systems increases. Consistent with this interpretation, sway during the least complex condition (EOFA) was associated with processing speed (SDMT), suggesting that efficient cognitive processing supports postural stability when sensory input is fully available. As task demands increased, particularly with removal of visual input, attention-based performance (TMT-A) became more strongly associated with sway. These findings extend the work of Perrochon et al, who also observed stronger cognition-balance relations under unstable surface conditions, particularly involving working memory and executive function.44 Our study builds on that work by using systematically manipulated, progressively challenging static balance conditions, allowing for clearer insight into how specific cognitive domains dominate as task demands increase.
Notably, we did not observe significant associations between standing balance performance and memory measures (CVLT, BVMT-R), nor did executive function as assessed by TMT-B emerge as an independent predictor. The absence of memory associations aligns with meta-analytic findings in older adults without MS, demonstrating limited relationships between episodic or visuospatial memory and static balance, with stronger effects observed for dynamic or dual-task conditions.29 Although executive functions have been linked to balance in findings from prior studies, these associations appear more robust for dynamic or cognitively loaded motor tasks than for quiet standing. TMT-B primarily indexes set shifting, which may be less engaged during constrained static postural tasks. In MS, impairments in processing speed and complex attention are more prominent at the group level than executive dysfunction,45 and task-switching research suggests greater difficulty maintaining attentional sets than switching.46 Together, these findings suggest that static balance in MS may rely more heavily on attentional capacity and processing efficiency than on memory or executive set-shifting.
This pattern also highlights potential differences in task-specific cognitive control between static and dynamic motor behaviors. Static standing primarily requires sustained attention and sensory integration to maintain postural stability within a relatively constrained motor framework. In contrast, dynamic mobility tasks such as gait and backward walking demand continuous motor adaptation, anticipatory control, and flexible cognitive engagement. Consistent with this distinction, prior findings in studies of people with MS have shown that executive functions, visuospatial memory, and higher-order cognitive control processes are more strongly associated with dynamic tasks, including backward walking and gait adaptability, than with quiet standing.26,27 For example, recent data have shown that attention, visuospatial memory, and executive function are linked to backward walking performance in individuals with MS.26 As a more complex gait task, backward walking imposes greater cognitive and sensory integration demands compared with forward walking.47,48 A recent systematic review identified processing speed as uniquely associated with laboratory-based mobility measures,49 whereas another study highlighted the relationship between processing speed, fear of falling, and real-world mobility outcomes.50 Additionally, processing speed and immediate memory recall were found to be associated with backward walking speed modulation, a key marker of gait adaptability and complexity.27 Together, these findings underscore the critical role of cognitive resources in maintaining balance as motor tasks become more complex, whether dynamic (eg, backward walking) or static (eg, eyes closed with a narrow base).
An additional and important finding from this study was the contribution of concern about falling, particularly under eyes-closed conditions. Concern about falling was a significant predictor of sway in both eyes-closed conditions, even after adjusting for ambulation disability (PDDS). Neuroimaging findings complement these results, showing that heightened concern about falling is associated with lower volumes in visual processing brain regions.51 When visual input is removed, reliance on proprioceptive and vestibular systems increases, and heightened concern or anxiety may further impair balance performance. This aligns with theoretical models demonstrating that postural threat increases sway amplitude due to anxiety-induced shifts in attentional focus and motor stiffening.52,53 Recent work offers further insight, proposing that low perceived control over balance, such as in situations without visual feedback, amplifies threat perception that disrupts automatic balance control.54 Moreover, concern about falling can trigger defensive responses such as rigidity and hypervigilance, which may destabilize posture in complex sensory contexts.54,55 Our findings support this idea in people with MS, suggesting that concern-related attentional shifts or fear-driven rigidity may partially explain postural instability in challenging sensory contexts.
Our data are consistent with recent MS-specific research, indicating that concern about falling and avoidance behavior are closely linked to dynamic balance dysfunction, mobility limitations, and cognitive load.12,56,57 A recent scoping review further emphasized the multifactorial nature of fall risk in people with MS, as psychological concerns interact with physical and cognitive deficits.10 Similarly, data from a systematic review support this association, describing a vicious cycle in which concern about falling contributes to worsening gait, balance deficits, fatigue, and cognitive impairments, each of which, in turn, increases fall risk.4 Collectively, these results suggest a multifactorial destabilization mechanism in people with MS, where cognitive processing demands, sensory integration deficits, and psychological factors, such as concern about falling, interact in complex ways to influence balance performance.
In addition, balance impairments in MS are multifactorial and may be influenced by musculoskeletal and sensorimotor factors, including muscle weakness, spasticity, altered lower-limb biomechanics, and impaired proprioception.5,58-60 These physical contributors are well-established determinants of postural instability and likely play an important role in static balance performance, particularly under conditions that increase reliance on somatosensory input.35 Although PDDS was included as a measure of ambulatory disability, we did not directly assess specific neuromuscular or proprioceptive impairments. The present study was designed to isolate domain-specific cognitive and psychological contributions to balance performance; however, these findings should be interpreted within the broader context of multifactorial balance dysfunction in MS. Future studies incorporating comprehensive biomechanical, sensory, and cognitive assessments will be important to more fully characterize the interactive contributions of motor, sensory, and cognitive systems to postural control.
This study should be interpreted in light of several limitations. First, the sample size was modest, which may limit the generalizability of findings. Further, although MS subtype was recorded, the sample was predominantly relapsing-remitting, with limited representation of progressive phenotypes. Future studies with larger and more balanced subtype representation are needed to determine whether MS disease course moderates cognitive-balance associations. Second, the cross-sectional design precludes any conclusions about causality or directionality between cognitive function, concern about falling, and balance performance. Third, only static balance assessments were included; dynamic balance and dual-task paradigms, which may elicit stronger cognitive-motor interactions, were not examined. Additionally, balance conditions were administered in a fixed order that progressed from less to more challenging tasks. Although this approach prioritized safety and standardized task exposure, it may have introduced potential order effects. Future studies employing randomized condition order may help further isolate task-specific cognitive-balance associations. Finally, participants were relatively homogeneous in age and disability status (PDDS score ≤ 6), which may limit insight into how these associations manifest in older adults with MS or those with more advanced functional impairments, as both may experience additional age-related changes in postural control and cognition.
Conclusions
Findings from this study provide novel evidence that specific cognitive domains, particularly processing speed and attention, are associated with postural control in individuals with MS and that these associations become more pronounced under sensory-conflicting balance conditions. In addition, concern about falling independently contributed to sway performance in more challenging tasks, highlighting the interplay between psychological and cognitive factors in balance regulation. Clinically, these findings suggest that incorporating brief cognitive assessments (eg, SDMT, TMT) and measures of concern about falling (eg, FES-I) may aid in identifying individuals with MS who are at increased risk of balance impairment, particularly under more challenging conditions. Although further research is needed, rehabilitation strategies that address cognitive, psychological, and motor factors in tandem may offer added value for fall-risk management in MS.
