Content

Issues

Research

Permissions

Community Content

Videos and Multimedia

IJMSC Awards

Cover Art

For Authors

Information for Authors

Manuscript Submission

Patient Consent Form

About IJMSC

Aim and Scope

Ethics

Editorial Board

Peer Review

Contact Us

Home

Advertising

IJMSC

1rem

Authors

Validity and Reliability of Four Clinical Gait Measures in Patients with Multiple Sclerosis

Test Battery for Evaluating Fatigue in Multiple Sclerosis Patients

 A wearing-off effect has been reported in people with multiple sclerosis (MS) receiving ocrelizumab and other monoclonal antibodies. We sought to describe the real-world occurrence of wearing-off effect among people with MS in Canada treated with ocrelizumab.

 Canadian self-reports coded to the Medical Dictionary for Regulatory Activities preferred term 

) between November 5, 2008, and March 15, 2023, were identified from the Roche Global Safety Database. Cases were then linked to patient-level data from the Canadian Roche Patient Support Program (COMPASS). Only people with MS who had received 3 or more doses of ocrelizumab were included.

 A total of 10,055 people with MS in Canada received 3 or more doses of ocrelizumab within COMPASS. 

was reported 140 times among 138 unique cases (1.3%) over an average of 7.4 infusions. Discontinuation of ocrelizumab was generally uncommon. In the overall COMPASS population, 11.4% (1143 of 10,055) discontinued ocrelizumab. Among those who reported 

, 15.9% (22 of 138) discontinued while the majority (116 of 138; 84.1%) continued ocrelizumab treatment. Of individuals reporting the adverse event of interest, those with relapsing-remitting MS more frequently remained on therapy (98 of 112; 87.5%) compared with those with primary progressive MS (18 of 26; 69.2%).

 A wearing-off effect was infrequently reported and did not limit treatment among people with MS in the real-world COMPASS patient support program.

 The authors thank all people with MS enrolled in Roche Canada COMPASS. The medical writing support provided by Stevie Kenyon of Placencia Holdings (Hamilton, Ontario, Canada) is also appreciated.

In the past 3 years, Sarah A. Morrow, MD, has served on advisory boards/had speaking engagements for Biogen Idec, Bristol Myers Squibb/Celgene, EMD Serono, Greenwich Biosciences, Novartis, Roche, Sanofi Genzyme, and Teva Neurosciences; has received investigator-initiated grant funds from Biogen Idec, Novartis, Roche, and Sanofi Genzyme; and has acted as site primary investigator for multicenter trials funded by Bristol Myers Squibb/Celgene, EMD Serono, Novartis, roche, and Sanofi Genzyme. Victor Gitman, PharmD; Sofija Spasojevic, PhD; Jessica Rassi, MSc; Noemi Pasquarelli, PhD; and Kocho Fitovski, MD, are employees and shareholders of Hoffmann-La Roche/F. Hoffmann-La Roche. Galina Vorobeychik, MD, has received research support, educational grants, and/or honoraria from Alexion, Berlex, Biogen Idec, Bristol Myers Squibb/Celgene, EMD Serono, Novartis, Roche, Sanofi Genzyme, and Teva Neurosciences.

Hoffmann-La Roche (Mississauga, Ontario, Canada) sponsored this research and writing services for the preparation of the manuscript. The company, in collaboration with independent authors, was also involved in the study design, collection, analysis, and interpretation of data; writing the report; and the decision to submit the article for publication.

Presented as an abstract/poster at the European Committee for Treatment and Research in Multiple Sclerosis 2022; October 26 to 28, 2022; Amsterdam, Netherlands.

Multiple sclerosis (MS) is a chronic, neurologic, inflammatory disease affecting approximately 90,000 Canadians.

 MS can affect various parts of the central nervous system, leading to a range of possible symptoms (eg, fatigue, lack of coordination, sensory symptoms, vision problems, bladder problems, cognitive impairment, mood changes). In Canada, approximately 90% of MS cases are relapsing-remitting (RRMS),

 characterized by reversible episodes of neurological dysfunction and disability. However, damage from these events can accumulate over time, with resulting lesions in the brain and spinal cord.

 In 10% of Canadian people with MS, the disease is progressive from onset (primary progressive MS [PPMS]).

Although there is no cure for MS, current therapies support people with MS in reducing disease activity, delaying progression, and gaining better symptom control.

 Among a broad class of disease-modifying therapies (DMTs) are the higher-efficacy monoclonal antibodies (mAbs) alemtuzumab, natalizumab, and 3 anti-CD20s: ocrelizumab, ofatumumab, and ublituximab.

 Each has a unique structure, mechanism of action, dosing regimen, and safety profile.

 Ocrelizumab, a humanized mAb that selectively targets CD20-expressing B cells, is approved in Canada for the treatment of RRMS and PPMS.

 After initial dosing of a split, twice-administered 300-mg infusion, ocrelizumab is administered at a single fixed dose (600 mg) every 6 months.

Natalizumab, ocrelizumab, and ofatumumab have been linked with reports of a therapeutic phenomenon called 

 A wearing-off effect is described as the experience of MS symptoms, such as fatigue, in the days or weeks prior to the next scheduled mAb infusion and a disappearance of such symptoms following receipt of the treatment.

 It has been reported at varying frequencies (16%-67%), which may be a result of different case definitions and/or assessment methods.

 The wearing-off effect is often referred to as a phenomenon for which research has failed to elucidate either the true cause or consequence of the effect.

 This is notably different from disease reactivation or rebound, which has been reported and studied extensively following natalizumab and fingolimod discontinuation.

 For natalizumab, findings from a study of the wearing-off effect associated the event with lower α

-integrin receptor occupancy on circulating leukocytes, suggesting a decreased treatment effect.

 For ocrelizumab, research suggests that the reported wearing-off phenomenon is not related to dose interval, B-cell count (remains depleted), symptom burden, or clinical efficacy but may be correlated with higher body mass index (BMI).

Considering the differences in reported frequency as well as within patient variability of wearing-off effect reports among people with MS treated with ocrelizumab, further investigation was deemed worthwhile. In Canada, the provision of ocrelizumab for MS through a national patient support program presented an opportunity to further study wearing-off effect events in a large population of people with MS treated with ocrelizumab in a real-world setting. The primary objective of this study was to determine the frequency with which the wearing-off effect occurs during real-world treatment with ocrelizumab and, secondarily, to characterize how the event affected the course of treatment, if at all.

This is a retrospective, observational study of Canadian people with MS treated with ocrelizumab enrolled in the Canadian Roche Patient Support Program (COMPASS) and reporting wearing-off effect (within the Roche Global Safety Database).

COMPASS is a patient support program in Canada, sponsored by the manufacturer of ocrelizumab, that supports the provision of the drug to people with MS. All people with MS in Canada receiving ocrelizumab would at some point in their care be enrolled in COMPASS. The Roche Global Safety Database is a centralized repository for reports of adverse events (AEs) received for products manufactured by F. Hoffmann-La Roche Ltd.

AEs were voluntarily reported, meaning event information was volunteered in an unprompted manner. 

) was the Medical Dictionary for Regulatory Activities v24.1 (MedDRA) highest-level preferred term used in the analysis to capture wearing-off events. Aligned with routine pharmacovigilance processes, medical coders selected the most appropriate lowest-level term from the MedDRA that accurately reflected the reported verbatim information. Medical judgment was applied to assign a MedDRA term that represented the reported concept when the verbatim term was not an exact match to 

 Examples of verbatim terms that were coded to the preferred term (

. Individual case safety reports coded to the MedDRA preferred term 

were extracted from the Roche Global Safety Database (collected from November 5, 2008, to March 15, 2023) and linked with data from the COMPASS program using unique patient identification numbers or initials and date of birth (

There was no imputation of data; cases with missing identifying information were excluded from the analysis. Reports not linked to the COMPASS program (eg, those from clinical trials) were excluded from this study. Only adults (aged ≥ 18 years) with MS who had received at least 3 doses of ocrelizumab were included in the analysis of TRS. This ensured those included in the analysis had received 2 initial doses (300 mg each) and, approximately 6 months later, a third dose (600 mg; full dose), allowing for at least one 6-month dosing interval in which TRS might occur. TRS events are herein referred to as 

in keeping with the common literature, as terms synonymous with reports of wearing-off events were coded to TRS as per MedDRA.

The primary variables of interest for this study included reports of wearing-off effect, indication for treatment, therapy status (collected by COMPASS; eg, 

[within the 6-month dosing interval and intends to receive the next ocrelizumab infusion] or 

[exceeded the 6-month dosing interval or does not intend to receive the next ocrelizumab infusion]), and reason for treatment discontinuation (eg, unknown reason, physician recommendation, AE with program drug, financial reason, switched to another therapy, or other). The Roche Global Safety Database served as the primary source for the following data: AE information (seriousness and outcome), sex at birth, and age. The COMPASS program provided baseline demographics, including data related to treatment history, current therapy status, ocrelizumab dosage, dosing interval treatment indication, and supplementary information on sex and age if unavailable in the AE report. Individuals were stratified by indication for ocrelizumab therapy (ie, RRMS or PPMS). Standard follow-up conducted by case managers in the COMPASS program includes a call after the individual is medically cleared by their neurologist, a call 5 business days prior to the scheduled infusion, a call 2 business days after the infusion, and a 4-month check-in call after the last infusion (ie, 2 months before the next).

This study involved the secondary use of personal data; hence, informed consent was not feasible nor required. A waiver of consent was granted by Advarra (May 3, 2022). All people with MS participating in the COMPASS program consented to sharing their personal information (including personal health information) with Roche and the possibility that this information would be used in an aggregated manner for publication. Still, precautions were taken to protect participant anonymity, including aggregating and anonymizing/pseudonymizing final data and ensuring linkage back to the individual participant was impossible or tightly controlled. The study protocol was reviewed and approved by the Advarra Institutional Review Board (approval number: MOD01320274).

Analyses were conducted using Python 3.8.8. For continuous variables, measures of central tendency (medians and means) and dispersion (SD and range) are reported. The COMPASS database provided a total number of individuals exposed to ocrelizumab (a denominator), and the Roche Global Safety Database, after exclusions, provided the number of individuals affected (the numerator). This enabled the reporting of frequencies and percentages for each categorical variable. All data are reported using descriptive statistics.

As of March 15, 2023, a total of 10,567 people with MS had received at least 1 administration of ocrelizumab through the COMPASS program, with 10,055 having received at least 1 full dose. Of those, 138 unique cases and 140 reports of wearing-off effect (2 individuals reported wearing-off effect twice) were identified in the safety database and linked to COMPASS, representing 1.3% of the estimated total exposed population (Figure S1). Reports were primarily received by patients or other non–health care professionals (see 

Characteristics of the study and source cohorts are outlined in 

Patients reporting a wearing-off effect received an average of 7.4 infusions, with a mean time between full doses of 6.2 months. Generally, people with MS treated with ocrelizumab reporting wearing-off effect had similar disease and treatment characteristics as those of the total exposed population (Table 1). The wearing-off effect occurred in people with PPMS and RRMS in proportions similar to the overall study makeup (Table 1). The majority of individuals with PPMS were treatment naive prior to receiving ocrelizumab. Mean ocrelizumab dosing interval was similar between the groups, with the majority receiving ocrelizumab every 6 months with no variation (IQR, 0; Table 1). Duration of therapy and number of infusions received were higher in the group reporting wearing-off effect (Table 1).

All reports of wearing-off effect were nonserious (

Overall, 11.4% of the total exposed population and 15.9% of people with MS reporting wearing-off effect discontinued treatment (

The discontinuation rate was higher in those with wearing-off effect and PPMS vs RRMS (30.8% vs 12.5%, respectively; Table 3). Reasons for discontinuation in people with MS reporting wearing-off effect were similar to those of the overall COMPASS population (Table 3).

A comparison of people who reported wearing-off effect and continued vs discontinued ocrelizumab is outlined in 

The groups had a similar ocrelizumab dosing interval (every 6 months). People who reported wearing-off effect and discontinued ocrelizumab were more likely to be older, were more likely to be men, and had a lower overall duration of treatment than those who continued ocrelizumab.

AEs reported in conjunction with the wearing-off effect are detailed in 

We investigated reports of a wearing-off effect in people with MS in Canada by combining events from the global safety database with patient-level data from the national COMPASS database. Given that all people with MS receiving ocrelizumab outside of clinical trials in Canada receive it via COMPASS, the data set allowed for a reliable estimate of total Canadian exposure. This database, coupled with the Roche Global Safety Database, provided a useful means of analyzing the wearing-off effect phenomenon in Canada. We observed that wearing-off effects coded to 

were reported by 1.3% of the studied population and were generally not treatment limiting, with 15.9% of people with MS who reported wearing-off effect discontinuing ocrelizumab.

Studies of wearing-off effect published to date have ranged from case series

 to prospective, phase 4, nonrandomized, multicenter trials (N = 376).

 When surveyed, people with MS treated with natalizumab have reported a wearing-off effect at a frequency of 37% to 67%.

 Similarly, a wearing-off effect has been reported in 51% to 61% of people with MS receiving ocrelizumab.

 These prospective studies related to ocrelizumab (88-100 participants) used a wearing-off questionnaire and noted the potential for overreporting due to enhanced recall and reporting bias.

In contrast, our study presents the largest cohort of people with MS yet studied for the wearing-off effect (N = 10,055) using spontaneous AE reporting rather than topic-specific questionnaires. Spontaneous AE reporting has been shown to underrepresent the prevalence of an AE compared with active monitoring (or true incidence).

 Although in some cases, rare AEs may be subject to less underreporting,

 in the case of wearing off, underreporting may be amplified given its subjective nature and relative nonseriousness. Interestingly, a retrospective chart review from British Columbia, Canada, found a wearing-off phenomenon recorded in 16% (11 of 69) of patient charts.

 Our data show that the spontaneous reporting of wearing off in a large real-world population is lower than that found in prospective studies. We suspect the true incidence of wearing off lies somewhere between the rate of spontaneous reports seen in our study and the frequency of reports solicited in prospective studies.

In the prospective study by Toorop and colleagues, treatment satisfaction did not differ between those reporting wearing off and those not reporting it

; however, in the multi-mAB study noted previously, wearing off was associated with reduced treatment satisfaction scores, and 20% of participants noted they had considered switching to another DMT.

 Uniquely, our data report on treatment status and reasons for ocrelizumab discontinuation. Importantly, wearing-off events were not reported as the reason for ocrelizumab discontinuation, and most people with MS affected remained on treatment (84.1%). Furthermore, wearing-off events were confirmed to be nonserious in nature for all in our study. The majority received ocrelizumab at 6-month intervals, and overall duration of therapy was similar in the total COMPASS population as it was among those who reported wearing off and discontinued treatment.

The potential cause of wearing off remains unknown. Findings from studies related to ocrelizumab used at extended dosing intervals have shown a maintenance of clinical efficacy, even with altered B-cell repopulation.

 In our study, a median dosing interval of 6 months was observed in both people with MS reporting wearing-off effect as well as the total exposed population, which is in line with the recommended dosing interval of every 6 months.

Given this consistency between groups, we were unable to observe any impact of dosing interval on wearing off.

Another theory to explain wearing off that has been explored is a decreased pharmacological effect. In the SymBOLS study (NCT04855617), no correlation was found between serum neurofilament light chain (sNfL) or ocrelizumab serum levels and symptom burden, suggesting wearing off is not a result of subclinical disease activity.

 found wearing-off reports were associated with higher levels of CD8 and CD3/CD27 lymphocytes (ie, reduced immunomodulation) and increased plasma NfL levels (ie, neuroaxonal damage); however, these were not correlated with clinical or radiological disease activity nor ocrelizumab serum levels, so a causal effect was not established. Results from a cross-sectional study by Meca-Lallana et al found no correlation between sNfL, CD19+, baseline Expanded Disability Status Scale score, age, progression, or relapse in people with MS treated with ocrelizumab experiencing a wearing-off effect.

 Finally, Toorop and colleagues observed a correlation between higher BMI and a wearing-off effect that did not appear correlated with B-cell count.

They speculated that higher rates of clinical disability experienced by people with MS and obesity may be responsible for greater MS symptom fluctuation.

The prevailing hypothesis, based on current evidence, is that the wearing-off phenomenon may be a subjectively perceived lack of effect or disease-related symptom fluctuations misattributed to treatment.

The low rates of unprompted reports observed in our study support this. In CASTING, a phase 3b study (NCT02861014), the patient-reported outcome tool SymptoMScreen, which includes elements such as fatigue, cognitive function, and sensory symptoms—all in common with wearing-off effect symptoms—was applied.

 People with MS who previously had a suboptimal response to 1 or 2 DMTs had significant improvement in SymptoMScreen scores, which were observed 24 weeks into ocrelizumab treatment and maintained over 96 weeks.

 Furthermore, results from the SymBOLS study of 103 people with MS treated with ocrelizumab failed to show correlation between wearing off and meaningful symptom changes throughout the infusion cycle as measured by Quality of Life in Neurological Disorders short forms, SymptoMScreen, and the Work Productivity and Activity Impairment Questionnaire at the beginning, middle, and end of the ocrelizumab treatment cycle.

 Importantly, the significant intra- and interindividual symptom variability associated with the MS disease course (regardless of treatment)

 is likely to affect the ability of prospective studies to objectively and consistently capture wearing-off phenomenon within a group over time. Results from a recent study of wearing-off effect with 4 mAbs found wearing off to be associated with symptoms of depression, which may support the theory that it is a psychological rather than biological phenomenon.

The present work is characterized by some limitations. Inherent to the nature of safety databases, the extracted data require careful interpretation due to the varying quality of AE reports, which could in some instances have missing, nonverifiable, and/or incomplete information. This also precluded the analysis of comorbidities (such as BMI and laboratory parameters) that may overlap with wearing-off effect. However, the COMPASS data supplement many other data points potentially missing from AE reports.

The calculation of population-based AE rates was not feasible because voluntary AE reporting is unprompted and, particularly for nonserious events, prone to be underreported.

 Given the unsolicited and voluntary nature of the information retrieval, there is a possibility that some wearing-off symptoms were not reported, potentially leading to an underestimation of the true frequency as discussed previously. Furthermore, although the preferred term 

captured a broad range of synonymous terms, other common MS symptoms (eg, fatigue) reported independently of the analyzed preferred term may not have been coded in the Roche Global Safety Database as 

, further contributing to the potential underestimation of the frequency of wearing off.

Despite the above considerations, our data provide insights into a phenomenon that is difficult to objectively measure. Wearing-off event analysis can vary by the definition of the reported event, use of a biological/symptomatic surrogate, timing/interval of data collection, study design and resulting biases, and individual awareness of the effect. These are all factors at play when considering the results of this study in the context of the limited literature on the topic.

Although these findings support the provision of ocrelizumab according to the recommended dosing schedule, with no observed impact due to wearing-off effect, an opportunity to better understand the phenomenon remains. In addition to routine pharmacovigilance assessment for lack of efficacy (including wearing-off effect), longitudinal studies designed to minimize reporting bias and monitor symptom recrudescence and biological markers over the course of treatment may be insightful. Two ongoing phase 3 studies investigating ocrelizumab administered at higher doses may provide additional insights into symptom control under other clinical circumstances.

In this large and representative population of people with MS in Canada treated with ocrelizumab, reports of wearing-off effect were infrequent (< 2%). Observed symptoms and treatment patterns in people with MS reporting wearing-off effect were nonserious and support the continuation of therapy according to label recommendations.

Unsolicited reports of the wearing-off effect related to ocrelizumab were infrequent (< 2%).

Wearing-Off Effect Reports in People With Multiple Sclerosis Receiving Ocrelizumab

In the absence of gait automaticity, the need to allocate cognitive resources to walking has implications for everyday mobility and community participation. The purpose of this study was to examine the effects of different attentionally demanding conditions on gait metrics in people with multiple sclerosis (MS).

Twenty-one individuals (11 with MS, 10 without MS) wore wireless inertial sensors and completed 1-minute walking trials under 4 different attentional conditions: no-attentional-instruction, focus on gait quality, attention to a narrative discourse task, and attentional switching. Walking trials were performed at a self-selected and fast-paced walking speed. A 3-way linear mixed model including conditions, groups, speeds, and interactions was used to evaluate the effects on gait speed and stride length variability.

Participants with MS demonstrated slower gait speed and greater gait variability in all attentional conditions than participants without MS. During self-paced walking, participants with MS showed the greatest decline in mobility metrics during the attentional switch condition. When walking fast, significant differences were observed for both the divided attention and attentional switching conditions, with divided attention having the greatest effect on gait speed and attentional switching resulting in the greatest decrease in gait variability.

Difficult walking conditions, such as attentional switching, elicit greater gait deficits in people with MS. Evaluating attentional switching may be useful for mobility assessment in people with MS, and incorporating this attentional technique in future mobility training may improve stability and prevent falls.

The authors declare no conflicts of interest for
this study.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

This work, in part, was presented previously as 2 posters at the North American Federation of Adapted Physical Activity Conference; October 11-13, 2022; St. Catharines, Ontario, Canada.

Well-controlled and coordinated walking is essential for safe and successful community mobility, with automaticity—defined as the ability to walk with minimal reliance on attention-demanding cognitive resources—serving as a critical component of efficient gait.

 Automaticity allows individuals to walk without the need for continuous attentional monitoring, preserving cognitive capacity for navigating complex and dynamic environments.

When automaticity is compromised, walking becomes increasingly dependent on cognitive control, resulting in competition for attentional resources and contributing to gait difficulties and impaired performance of concurrent tasks.

 This is particularly relevant for individuals with multiple sclerosis (MS), who often exhibit reduced walking automaticity

 and report needing to consciously direct attention to their walking in order to maintain stability.

 Such compensatory strategies are concerning, as attentional difficulties have been shown to negatively impact walking

 and are associated with increased fall risk in this population.

 Accordingly, understanding how attentional deficits influence gait has important real-world implications.

While research has reported that between 20% and 50% of adults with MS have specific attentional difficulties,

much of the attentional research regarding gait has focused on walking while dividing attention (ie, dual tasking).

 Yet, to walk safely and successfully, individuals not only have to divide attention but must also be able to flexibly prioritize and switch attention in accordance with the demands of everyday conditions and circumstances.

 Attentional switching denotes the ability to flexibly shift back and forth between multiple tasks or mental sets.

 Recent research has sought to examine outcomes associated with flexible and adaptive attentional control in individuals with MS through task-switching paradigms. Migliore et al

 found that participants with MS demonstrated greater deficits in attentional switching compared with the control group when completing a computer-based cued-switching task while seated. However, research by Shiri and colleagues

 found no relationship between cognitive flexibility and static balance in study participants when attentional switching was assessed via common neuropsychological measures. Notwithstanding the methodological differences and inconsistencies across these 2 studies, no study to date has specifically investigated whether impaired attentional switching influences walking in people with MS.

During walking, competition for attentional resources is influenced not only by the timing of concurrent tasks but also by their complexity—that is, tasks that demand greater cognitive load or involve more advanced mental processes.

 Individuals with MS tend to have a reduced capacity to recruit additional cognitive resources in response to increasing task complexity.

 Moreover, gait impairments can often be exacerbated at faster walking speeds,

 highlighting the importance of understanding how individuals with MS manage attentional switching across varying gait speeds. This is particularly important given the demands of real-world ambulation, where navigating busy or unpredictable environments often requires rapid shifts in attention and adaptive motor responses.

Given the importance of cognitive resources for optimal gait performance, isolating the effects of cognitive flexibility from other cognitive processes and under differential cognitive loads may further the development of novel therapeutic interventions designed to enhance attentional switching and facilitate gait automaticity in people with MS. As such, the purposes of this study were to compare the effects of divided attention and attentional switching on gait in adults with and without MS and to explore the differential effects of these attentional conditions when walking at self-selected or fast-paced speeds.

A convenience sample of 11 individuals with MS and 10 without MS (NoMS) was recruited. An email invitation was sent to individuals with MS listed in the authors’ research database who had participated in previous research. Additionally, a flyer was posted in the multiple sclerosis center of the local medical center. Those with MS had to self-report (1) physician-diagnosed MS and (2) minimal to moderate disability (a score of 1-4 on the Patient-Determined Disease Steps [PDDS] scale).

 Individuals with MS were excluded if they had (1) a severe exacerbation of symptoms requiring medical intervention in the previous month or (2) any health condition or diagnosis besides MS that could influence balance or ambulation. Participants without MS were matched by age (±3 years) and sex, and also had to report no health condition or diagnoses affecting balance or ambulation. This study was approved by the Institutional Review Board of the University of Vermont, and all participants provided written consent prior to data collection.

Demographic data included age, sex, years post diagnosis, and MS phenotype. For additional insight into the sample, cognitive functioning was assessed using the Symbol Digit Modalities Test (SDMT) and the Montreal Cognitive Assessment (MoCA). The SDMT, a reliable and valid measure of cognitive impairment in MS,

 evaluates constructs of attention, perceptual processing, working memory, and psychomotor speed through a timed task of matching abstract symbols to numbers. The MoCA is a sensitive and useful cognitive screening tool used to detect early and mild cognitive impairment in people with MS.

 It assesses function across the cognitive domains of attention, executive functions, memory, language, visuo-constructional skills, conceptual thinking, calculations, and orientation.

Mobility impairment was measured with the 12-item Multiple Sclerosis Walking Scale, a reliable and valid instrument,

 whereby participants self-report how much their MS impacts their ability to walk under various circumstances. The Falls Efficacy Scale–International

 was used to assess participant concerns about falling during 16 activities of daily living.

Measurements of continuous walking were derived using APDM Opals, small wireless inertial motion sensors, and its related Mobility Lab software (Clario). Sensors were placed on a participant’s lumbar spine, sternum, and both feet. Data from the Opal sensors were collected wirelessly at 128 Hz. Measurements included gait speed (m/s) and stride length variability, which was expressed as the coefficient of variation (CV [%]) and calculated by stride length SD/stride length mean to represent relative gait variability.

 These metrics have been identified as sensitive measures of dual-task mobility impairment in individuals with MS.

Participants completed eight 1-minute trials walking between 2 lines 25 feet apart. The 8 walking bouts consisted of 4 different attentional conditions at both a self-paced (SP) and a fast-paced (FP) walking speed. For the no-attentional-instruction (NI) condition, participants were simply asked to walk back and forth. For the internal-focus (IF) condition, participants were instructed to focus their attention on their gait throughout the walking bout, a task selected based on research highlighting the impact of internally focused attention on daily walking in individuals with MS.

 In the divided-attention (DA) condition, participants were asked to walk while simultaneously completing a narrative discourse task, which required them to verbally produce a detailed and coherent story in response to an open-ended prompt (eg, “trip to New York City”) while walking, or a similarly structured alternate form (eg, “moving to a new city”).

 This task has been shown to be a cognitively demanding task of natural language formulation, planning skills, and discourse production requiring substantial attentional resources.

 In the attentional-switching (AS) condition, participants were instructed to switch their attention every 15 seconds between the IF and DA tasks throughout the 1-minute bout, allowing for an assessment of their ability to flexibly shift attention between internally and externally oriented tasks—an ability that mirrors the dynamic attentional demands of real-world environments (

). All walking trials and discourse topics were randomized across participants to minimize order effects. Participants were instructed to rest, as needed, between trials to minimize the impact of fatigue from both the mobility and cognitive components of the testing paradigm.

Descriptive statistics were performed to examine group differences with independent 

 test, Wilcoxon rank-sum test, or Fisher exact test after inspection of normality. To evaluate the effects of attentional conditions and walking speeds on gait metrics (gait speed and stride length variability), we employed a 3-way linear mixed model including attentional conditions (NI, IF, DA, AS), groups (MS, NoMS), speeds (SP, FP), and interaction terms (group × condition, group × speed, condition × speed, group × condition × speed) as fixed effects with participants as random effects to allow for the entry point to vary across participants. A linear mixed model was selected, as it accounts for the heterogeneity and correlation of repeated measures under different conditions and retains participants with partially missing data. Effect sizes and 95% CIs of the linear mixed model were calculated using fixed-effects regression coefficients. All analyses were performed using Stata 17 (StataCorp), and statistical significance was set at 

There were no significant differences in age (

 = .99) between participants with and without MS. Participants with MS had greater concerns about falling (

 < .001) than those without MS and lower cognitive functioning than the NoMS group, as indicated by SDMT (

 < .001) scores. The MS group reported a median PDDS score of 3 (ie, gait disability) and average years post diagnosis of 24.1 ± 12.8 (

The 3-way linear mixed model found significant group effects (

 < .05) in both gait metrics. Participants with MS showed significantly lower gait speed and greater stride length variability than participants without MS in all attentional conditions, regardless of speed instructions, except for stride length variability in the fast-paced internal-focus condition (

We also observed significant speed main effects (

 < .05) in both groups. Multiple post hoc contrast comparisons indicated that participants significantly increased their gait speed when instructed to walk at a faster speed. However, no significant differences in gait speed were found between self-paced and fast-paced walking in the divided-attention condition for participants with MS. Unlike gait speed, stride length variability did not differ among walking speeds across attentional conditions (Table 2).

Results revealed different patterns of attentional condition main effects based on speed instruction and group. In self-paced walking, there were significant attentional condition main effects on gait speed in the NoMS group, but not in participants with MS (

). No significant gait speed differences were observed across any of the attentional conditions in participants with MS. The NoMS group, on the other hand, showed the greatest decline in gait speed in the attentional-switching condition compared to the no-attentional-instruction condition (B = –0.13 m/s; 95% CI, –0.20 to –0.06; 

 < .001) and the divided-attention condition (B = –0.09 m/s; 95% CI, –0.17 to –0.01; 

 = .024). Compared to gait speed, however, we observed that the attentional-switching condition resulted in the greatest stride length variability in both groups at the self-selected pace. Participants with MS demonstrated significantly greater gait variability in the attentional-switching condition than in the no-attentional-instruction condition (B = 1.28%; 95% CI, 0.18-2.37; 

 = .022) and the internal-focus conditions (B = 1.67%; 95% CI, 0.03-3.31; 

 = .043), whereas the NoMS group had greater variability in the attentional-switching condition compared to the divided-attention condition (B = 1.23%; 95% CI, 0.08-2.38; 

In fast-paced walking, both groups demonstrated the slowest gait speed in the divided-attention condition (

). For the MS group, the divided-attention condition (B = –0.20 m/s; 95% CI, –0.28 to –0.12; 

 < .001) resulted in slower walking speed than in the attentional-switching condition (B = –0.16 m/s; 95% CI, –0.23 to –0.08; 

 < .001) when compared to the no-attentional-instruction condition. For the NoMS group, both the divided-attention and attentional-switching conditions significantly differed from the other conditions and one another (B = –0.08 m/s; 95% CI, –0.16 to –0.001; 

 = .047), although the divided-attention condition had the greatest effect on speed. For the MS group, while the divided-attention condition elicited greater stride length variability than the no-attentional-instruction (B = 1.47%; 95% CI, 0.30-2.63; 

 = .013) and the internal-focus conditions (B = 1.41%; 95% CI, 0.23-2.59; 

 = .019) during fast-paced walking, attentional switching had the greater impact on this outcome than did the divided-attention condition. Consistent with the MS group, the attentional-switching condition also revealed the greatest stride length variability across the attentional tasks for the NoMS group, with significant differences observed between the attentional switching and the no-attentional-instruction conditions (B = 1.69%; 95% CI, 0.50-2.87; 

 = .004) as well as the attentional-switching and divided-attention conditions (B = 1.61%; 95% CI, 0.46-2.75; 

Despite significant group and condition main effects, there were no significant group × condition interaction effects (

 .05) in either gait metric, as both groups showed similar patterns across attentional conditions. This insignificant finding indicated that the participants who did and did not have MS responded similarly to the varying attentional contexts.

Multiple contrasts analyses for the significant condition × speed interaction effect on gait speed (MS: 

 = .005) revealed significant differences between the no-attentional-instruction and the divided-attention conditions in both groups (

). Specifically, participants with and without MS maintained gait speed in the divided-attention condition compared to the no-attentional-instruction condition while self-paced walking. However, participants in both groups had decreased gait speed while fast-paced walking in the divided-attention condition compared to the no-attentional-instruction condition, contributing to the significant interaction effects. There was no significant condition × speed interaction effect for stride length variability.

Significant group × speed interactions and multiple contrast comparisons in gait speed (NI: 

 < .001) showed that the magnitude of gait speed differences between groups increased when walking at a fast pace compared to the self-selected pace under all attentional conditions (

). There were no significant group × speed interaction effects for stride length variability. Additionally, no significant group × condition × speed interaction effects were seen in either gait metric.

Moving through daily life requires the ability to flexibly switch attention between walking and other cognitive tasks or across varied environmental situations. For individuals with MS, walking requires greater attentional resources,

 and this can present a considerable challenge and an increased fall risk. The results of this study indicate that alterations in gait speed and gait variability are sensitive to varying attentional demands; the greatest decrements in gait quality occurred when participants walked faster and engaged in attentional switching. These findings extend prior research highlighting the critical role of attention in mobility among individuals with MS.

 Importantly, our results offer novel insights into mobility decline specifically related to attentional switching—an ecologically valid cognitive demand that reflects real-world challenges in everyday ambulation.

 participants with MS had slower gait speed and greater stride length variability compared to their counterparts without MS. As has been observed previously,

 participants with MS demonstrated twice as much variability in their gait as the participants without MS. Our MS sample also had slower cognitive processing speed, known to correlate with greater gait variability.

 Gait speed differences between groups were greater at a fast pace than at a self-selected pace, aligning with prior findings that fast walking reveals more pronounced deficits in individuals with MS.

 Collectively, participants with MS showed worse gait deficits than those without MS, regardless of attentional condition, and greater gait decrements were elicited in their fast-paced walking in this population.

When examining between-group differences, results were significant for all conditions except for internal focus while walking fast. While both groups showed variability, the underlying cause may differ. Whereas the gait variability observed in the MS group could be attributable to neuromotor and cognitive deficits, the variability seen in the NoMS group may have been a function of redirecting unnecessary attentional resources inward to their walking. Research findings show that internal focus on one’s movement pattern, thus bringing it under more conscious control, can interfere with already automatized movement control processes.

 Given that motor function was not directly assessed in the adults without MS, it is unclear if the observed variability in gait was indeed a function of attentional focus or of preexisting motor deficits. Future research is needed to more clearly elucidate these differences.

Although participants with MS exhibited worse gait speed and stride length variability than participants without MS, both groups showed relatively similar patterns of reductions in gait in response to attentional condition, resulting in insignificant group x condition interactions. Our finding is supported by recent evidence that suggests that there is only a minimal difference in cognitive-motor interference between individuals with and without MS.

 This further indicates that attentional conditions, particularly DA and AS, required effort from both groups and led to similar levels of cognitive-motor interference on gait. However, this should be interpreted with caution, given that this study did not assess cognitive performance during the DA and AS tasks to determine if there was a greater cost to the cognitive task in individuals with MS.

When examining within-group differences across the attentional conditions at a self-selected pace, the MS group exhibited no significant differences in gait speed. While speculative, previous research supports the idea that these participants may have employed increased cortical activity needed to maintain gait speed under the various attentional conditions or that the cost of dividing or switching attention may have negatively impacted the unmeasured cognitive task rather than the observed motor task.

 However, for participants without MS, when walking at a self-selected speed, the slowest gait speed occurred when engaging in attentional switching. It has been shown that as individuals age and mobility becomes more dependent on executive control, attentional switching produces high cognitive-motor interference and is a detriment to gait speed.

Walking while directing attention to other cognitively demanding tasks can negatively impact gait stability in people with multiple sclerosis.

Machiko R. Tomita, PhD

Articles

Latest Updated Articles