Longitudinal modeling of upper extremity function in multiple sclerosis: Associations for clinical and sociodemographic factors

Abstract

Background:

Upper extremity impairment is a common disabling consequence of multiple sclerosis (MS). Little is known about factors influencing upper extremity functioning in MS.

Objective:

Identify associations with longitudinal changes in upper extremity performance.

Methods:

Using a real-world, retrospective cohort study (4140 persons with MS and 14,544 observations), multivariable repeated-measures analyses were completed for nine-hole peg test (9HPT) dominant and non-dominant hand times and their difference (asymmetry), with clinical, demographic, and social determinants as covariates. Interactions with time (disease duration) were examined.

Results:

Cross-sectionally, dominant, and non-dominant hand times were strongly correlated (ρ = 0.78), and moderately correlated with asymmetry, walking speeds, and depressive symptoms. Longitudinally, older age, males, Hispanics, Blacks, ever smokers, public insurance beneficiaries, greater lower-limb impairment, progressive disease, and greater depressive symptoms were independently associated with worsening dominant and non-dominant times. Public insurance beneficiaries, greater lower-limb impairment, progressive disease, and greater depressive symptoms were associated with greater asymmetry. Males accrued dominant hand dysfunction more rapidly and had greater asymmetry than females over time, while those with increasing lower-limb impairment more rapidly experienced asymmetry over time.

Discussion:

These findings identify high-risk groups and inform future research to guide strategies for preserving upper extremity function in multiple sclerosis.

Keywords

Multiple sclerosis nine-hole peg test 9HPT dexterity hand function

Introduction

Upper extremity impairment is a common and disabling consequence of multiple sclerosis (MS) limiting personal and instrumental activities of daily functioning such as feeding, dressing, housework, and transportation.¹ It also impacts employment and overall quality of life.^1–3 In a large North American Research Committee on MS (NARCOMS) cross-sectional study, hand dysfunction emerged early with only 40% of persons with MS (PwMS) reporting no impairment in the first year, and declining to 31%, 27%, and 19% in years 5, 10, and 15, respectively.⁴ Moderate or severe impairment was reported in 8% in year 1, then in 12%, 18%, and 22% for years 5, 10, and 15, respectively.⁴ Importantly, hand dysfunction may occur even in PwMS with preserved Expanded Disability Status Scale (EDSS).⁵ These observations underscore the need to identify correlates of upper extremity dysfunction, considering both disease duration and impairment in other domains.

While little research has examined upper extremity symptoms and deficits in MS, multiple factors influencing other MS-related impairments may have overlapping and distinct effects. Neurological mechanisms link upper and lower extremity function—including shared corticospinal tract involvement and cerebellar or brainstem lesions—which can impair coordination, strength, and motor control.^6,7 Thus, factors known to influence lower extremity function may also affect upper extremity function. Social determinants of health (i.e. individual- and neighborhood-level socioeconomic conditions) and modifiable behavioral/lifestyle factors (i.e. smoking and obesity) have also been linked to longitudinal changes in depressive symptoms, cognition, brain atrophy, and disability in PwMS.^8–11 These factors exert broad systemic effects on inflammation and possibly on neurodegeneration, as well as access to care and rehabilitation, making them biologically and clinically relevant to upper extremity outcomes in MS.^12–15 Emerging neuroimaging evidence reinforces this rationale, showing that both structural damage (e.g. regional atrophy and demyelination) and functional reorganization within sensorimotor networks are linked to upper-limb motor performance in PwMS.¹⁶ Collectively, these findings suggest that upper- and lower-limb function share overlapping neural substrates, and that disruptions in these pathways may manifest asymmetrically, underscoring the importance of investigating clinical and social correlates of hand impairment alongside established measures of disability.

The nine-hole peg test (9HPT) is a widely used, reliable, and sensitive tool for measuring manual dexterity in PwMS,¹⁷ and it is a key component of the MS Functional Composite.¹⁸ 9HPT correlates with greater disability, longer disease duration, EDSS, Timed 25-Foot Walk (T25FW), low-contrast letter acuity, Symbol Digit Modalities Test, and depressive symptom severity.^19,20 Upper-limb asymmetry, defined as the absolute difference between dominant and non-dominant 9HPT times, has been reported at high prevalence among those with severe disability and correlates with EDSS.^21,22 Asymmetry may also have functional implications, as greater reliance on the less impaired limb may foster compensatory mechanisms that reduce efficiency, increase fatigue, and impair bimanual tasks. Thus, monitoring inter-limb differences may offer complementary information to unilateral measures and may provide additional insight into disability progression.

Most existing MS studies of 9HPT are cross-sectional with limited scope (e.g. reporting correlations with sex and disease duration).^23–25 Only one longitudinal study that examined dominant and non-dominant 9HPT impairment in ~650 young clinical trial participants, explored associations with time, age (18–25 vs 26–35 years), and sex.²⁶ Consequently, important knowledge gaps remain regarding the clinical and demographic factors influencing longitudinal 9HPT performance and asymmetry in diverse, real-world MS populations spanning the age spectrum. These relationships should also be evaluated for independent from lower-limb impairment and depressive symptoms, and whether they change over time. Therefore, our objectives were to: (1) identify demographic, clinical, and social determinants of health associated with longitudinal variation in 9HPT times and asymmetry, after accounting for lower-limb impairment, depressive symptoms, and disease course in PwMS and (2) assess whether any of these associations vary over time by testing interactions with disease duration.

Methods

Study population

We conducted a retrospective cohort study using real-world electronic health record (EHR) information from PwMS receiving care at the Cleveland Clinic’s Mellen Center for Multiple Sclerosis Treatment and Research (Cleveland, OH) and the Lou Ruvo Center for Brain Health (Las Vegas, NV)—tertiary MS referral centers with ~20,000 clinical encounters and ~1000 new patients per year. In 2007, the Knowledge Program © (KP) has electronically captured patient-reported and provider-entered outcomes, linked to the EHR.²⁷ Patients were included if they were US residents (eFigure 1) and had ⩾2 clinical encounters ⩾6 months apart between January 2008 and June 2016 with 9HPT recorded.¹⁰ This study was approved by the institutional review boards of the Cleveland Clinic, Case Western Reserve University, and The MetroHealth System.

Dependent variables of interest

The 9HPT is a provider-administered task that objectively measures upper extremity function with exceptional reliability (>95%) in PwMS.¹⁷ The average of two trials, time to place nine pegs in nine empty holes in a block followed by the removal of pegs, was recorded separately for self-reported dominant and non-dominant hands. The three outcomes were repeated measures of 9HPT times for dominant and non-dominant hands and asymmetry (absolute difference between hand times).

Independent variables of interest

We assessed clinical and demographic covariates, including disease duration (years), age (years), sex, ethnicity (Hispanic/Latinx, Not Hispanic), and race (White, Black, Other). We included area deprivation index (ADI; state-level ranking of census block socioeconomic disadvantage) of baseline residence.²⁸ Time-varying covariates were smoking status (never, current, former), insurance payer (private, self-pay, Medicare, and Medicaid as proxies for socioeconomic status (SES) in United States where self-pay, Medicare, and Medicaid are indicative of financial hardship, disability, or limited income-based coverage), marital status, assistive device use (none, unilateral, bilateral), T25FW (seconds; eFigure 2(a)), and depressive symptom severity (Patient Health Questionnaire 9; PHQ-9; eFigure 2(b)). Body mass index (BMI) was included if measured within 3 months of a 9HPT measure. EDSS was unavailable as it is not a part of routine care.

Statistical analysis

Descriptive statistics were reported as means (standard deviation (SD)) for continuous variables and percentages for categorical variables. At baseline, pairwise relationships between measures were evaluated using Spearman’s rank correlation, with 95% confidence intervals (CIs) and p-values derived from nonparametric bootstrapping (1000 repetitions).

Longitudinal associations were evaluated using generalized estimating equations with an identity link function and robust sandwich estimator using STATA v17.0 (StataCorp, TX; xtgee function). Given the unequal time intervals between encounters, we examined the quasi-likelihood for independence model criterion (QIC) for autoregressive (AR1) and exchangeable working correlation structures for the baseline models described below, for patients with ⩾2 encounters (a requirement for AR1). QIC was consistently lower for the exchangeable working correlation structure and, therefore, it was selected. Two longitudinal baseline models were generated per outcome: (1) all covariates except BMI (N = 4140 patients, 14,544 observations) and (2) including BMI (N = 2540 patients and 5090 observations). Fixed covariates were modeled as time-invariant, while smoking, insurance payer, marital status, assistive device use, T25FW, PHQ-9, and BMI were modeled as time-varying. Indicator variables retained observations with missing categorical data. Disease duration (years since diagnosis) was the primary time variable; interactions between disease duration and each covariate were tested to assess whether relationships changed over time, and significant interactions were retained in final models. Scatter, contour, and local polynomial smoothing plots with 95% CIs guided interpretations. Statistical significance was defined as two-sided α = 0.05.

Results

Patients averaged 3.5 encounters (SD = 2.8) over an average of 3.5 years (SD = 2.1; maximum 7.6 years). Baseline attributes are presented in Table 1 (N = 4140): mean age was 46.6 years (SD = 11.5), average disease duration 11.1 years (SD = 9.3), 74% were female, 95% non-Hispanic, 87% White, 11% Black, 45% never smokers, and 56% privately insured. Most patients had relapsing MS (77%) and did not require assistance ambulating (82%). Consistent with global trends, 90.4% were right-handed. On average, 9HPT dominant speeds were faster than non-dominant speeds (24.6 seconds (SD = 10.5) vs 26.6 seconds (SD = 12.3); p < 0.0001), and mean asymmetry was 4.7 seconds (SD = 10.1). Sixty-one percent had a BMI measure within 3 months of a 9HPT measure; these patients had less missing data for smoking, were less reliant on assisted devices, and less likely to have progressive disease (eTable 1).

Table 1.

Baseline attributes of the study population.

Attributes at baseline	Mean (SD)/%
N	4140
Number of encounters	14,544
Age (years)	46.6 (11.5)
Age at onset (years)	35.6 (10.1)
Disease duration (years)	11.1 (9.3)
Sex
Female	73.9%
Male	26.1%
Ethnicity
Non-Hispanic	95.1%
Hispanic/Latinx	1.1%
Missing	3.7%
Race
White	86.7%
Black	11.2%
Other	1.0%
Missing	1.2%
Tobacco exposure
None	44.5%
Current smoker	17.2%
Former smoker	27.7%
Missing	10.6%
Insurance
Private	56.0%
Self-pay	22.7%
Medicaid	5.4%
Medicare	13.7%
Missing	2.3%
Marital status
Married	65.9%
Single	23.0%
Divorced	7.3%
Widowed	2.2%
Missing	1.5%
ADI State rank	4.4 (2.7)
ADI National rank	57.6 (23.3)
Assisted device
None	81.7%
Unilateral	8.9%
Bilateral	8.5%
Wheelchair-bound	0%
Missing	0.9%
Clinical course
Relapsing	76.5%
Progressive	20.3
Missing	3.3%
Hand dominance
Right	90.4%
Left	9.6%
Dominant hand 9HPT (second)	24.6 (10.5)
Non-dominant hand 9HPT (second)	26.6 (12.3)
Averaged dominant and non-dominant 9 HPT (second)	25.6 (10.1)
Asymmetry (second)	4.7 (10.1)
Timed 25-Foot Walk (second)	7.7 (8.5)
PHQ-9	6.3 (5.8)
BMI (kg/m²)	29.0 (7.1); N = 1196^a

A subset of patients also had BMI measurements available within 3 months of baseline visit.

At baseline, dominant and non-dominant 9HPT times were strongly correlated (ρ = 0.78; 95% CI: 0.76, 0.79), both were similarly correlated with T25FW (ρ = 0.56–0.57) and PHQ-9 (ρ = 0.26–0.27) (Table 2). Asymmetry significantly correlated with all measures, with a stronger correlation with non-dominant (ρ = 0.51; 95% CI: 0.48, 0.53) than dominant (ρ = 0.28; 95% CI: 0.25, 0.31) hand times, suggesting the prior is a greater driver of asymmetry.

Table 2.

Spearman’s rank correlations (and 95% confidence intervals) among clinical measures at baseline.^a

Measure	Dominant hand time	Non-dominant hand time	Absolute asymmetry	T25FW
Non-dominant hand time	0.78 (0.76, 0.79)	—
Absolute asymmetry	0.28 (0.25, 0.31)	0.51 (0.48, 0.53)	—
T25FW	0.56 (0.54, 0.59)	0.57 (0.55, 0.59)	0.31 (0.28, 0.34)	—
PHQ-9	0.27 (0.24, 0.30)	0.26 (0.23, 0.29)	0.09 (0.06, 0.12)	0.32 (0.29, 0.35)

All correlations were statistically significant (p < 5 × 10⁻⁸).

Multivariable longitudinal associations for the three outcomes (Table 3; N = 4140 patients; 14,544 observations) showed that older age, male sex, Hispanic/Latinx ethnicity, Blacks race, ever smoking, public insurance or self-pay beneficiaries, assistive device users, progressive disease, slower T25FW times, and greater depressive symptom severity (PHQ-9) were independently associated with greater dominant hand impairment (longer/slower 9HPT times). A significant sex-disease duration interaction showed that males slowed an additional 1.5 second per 10 years (p = 0.0002) in dominant 9HPT times (Table 3, Figure 1(a)), beyond the average male–female difference of 1.6 second slower (p = 0.00078) dominant 9HPT times for males. In the raw data, treating all observations as independent, males had steeper slopes (eFigure 3(a)). However, after adjusting for covariates, the positive unadjusted slope in females (eFigure 3(a)) was no longer present (Figure 1(a)).

Table 3.

Longitudinal multivariable associations for the three hand impairment outcomes.

Predictor	Dominant hand time		Non-dominant hand time		Absolute asymmetry
Predictor	β (95% CI)	p	β (95% CI)	p	β (95% CI)	p
Age (year)	0.04 (0.02, 0.07)	0.00036	0.05 (0.02, 0.09)	0.0028	0.02 (−0.01, 0.04)	0.20
Disease duration (year)	0.01 (−0.02, 0.05)	0.51	0.06 (0.02, 0.10)	0.0022	−0.03 (−0.07, 0.01)	0.09
Sex
Female	Ref		Ref		Ref
Male	1.61 (0.81, 2.42)	0.000078	3.45 (2.60, 4.30)	<0.00001	0.66 (−0.16, 1.48)	0.12
Interaction
Male × disease duration	0.15 (0.07, 0.22)	0.0002	—		0.10 (0.05, 0.15)	0.00030
Ethnicity
Not Hispanic	Ref		Ref		Ref
Hispanic/Latinx	1.85 (0.10, 3.6)	0.038	2.99 (−0.12, 6.11)	0.06	0.69 (−1.57, 2.95)	0.55
Race
White	Ref		Ref		Ref
Black	1.93 (1.05, 2.81)	0.000017	2.26 (1.10, 3.40)	0.00011	0.04 (−0.74, 0.83)	0.91
Other	−0.84 (−2.71, 1.04)	0.38	−1.13 (−3.19, 0.93)	0.28	−1.36 (−3.77, 1.04)	0.27
Tobacco exposure
None	Ref		Ref		Ref
Current smoker	1.04 (0.40, 1.69)	0.0015	1.49 (0.92, 2.35)	0.00078	0.21 (−0.43, 0.86)	0.52
Former smoker	0.59 (0.03, 1.16)	0.040	1.10 (0.41, 1.78)	0.0017	0.16 (−0.37, 0.70)	0.55
Insurance
Private	Ref		Ref		Ref
Self-pay	0.31 (0.04, 0.57)	0.023	0.18 (−0.08, 0.44)	0.17	0.24 (−0.08, 0.57)	0.14
Medicare	0.81 (0.12, 1.50)	0.022	1.48 (0.37, 2.60)	0.0091	0.80 (0.00, 1.61)	0.05
Medicaid	1.35 (0.69, 2.01)	0.000056	1.33 (0.59, 2.07)	0.00043	0.68 (0.17, 1.20)	0.0092
Marital status
Married	Ref		Ref		Ref
Single	0.46 (−0.12, 1.05)	0.12	0.31 (−0.42, 1.04)	0.40	0.04 (−0.54, 0.62)	0.89
Divorced	0.49 (−0.50, 1.47)	0.33	0.50 (−0.87, 1.87)	0.48	0.03 (−0.83, 0.90)	0.94
Widowed	−1.12 (−2.45, 0.21)	0.10	−0.68 (−2.17, 0.81)	0.37	−1.20 (−2.65, 0.25)	0.10
ADI state rank	0.06 (−0.05, 0.16)	0.30	0.15 (0.02, 0.27)	0.02	0.02 (−0.07, 0.11)	0.66
Assistive device
None	Ref		Ref		Ref
Unilateral	2.01 (1.33, 2.68)	<0.00001	2.56 (1.71, 3.42)	<0.00001	1.13 (0.60, 1.65)	0.00003
Bilateral	4.30 (3.24, 5.36)	<0.00001	5.04 (3.61, 6.48)	<0.00001	2.91 (2.26, 3.57)	<0.00001
Wheelchair-bound	3.46 (3.28, 3.64)	<0.00001	3.57 (3.42, 3.72)	<0.00001	−0.39 (−13.05, 12.28)	0.95
Clinical course
Relapsing	Ref		Ref		Ref
Progressive	3.77 (2.86, 4.67)	<0.00001	4.06 (2.97, 5.15)	<0.00001	2.85 (2.23, 3.46)	<0.00001
T25FW (second)	0.20 (0.11, 0.28)	<0.00001	0.22 (0.14, −.29)	<0.00001	0.08 (0.04, 0.12)	0.00025
Interaction T25FW × disease duration	—		—		0.005 (0.003, 0.007)	0.00001
PHQ-9	0.15 (0.12, 0.19)	<0.00001	0.16 (0.11, 0.20)	<0.00001	0.04 (0.01, 0.07)	0.019
Intercept	16.30 (15.17, 17.43)	<0.00001	16.33 (14.82, 17.84)	<0.00001	1.28 (0.07, 2.49)	0.038

Two-sided p-values <0.05 are bolded.

For non-dominant times, associations were similar (Table 3). Greater ADI was only significantly associated with worsening non-dominant 9HPT, and Hispanics showed a trend toward slower times (p = 0.06). The effect sizes for all attributes were larger in the non-dominant hand model compared with the dominant hand model even with comparable intercepts. There was no evidence for any interactions with disease duration. For non-dominant hand times, the raw data (eFigure 3(b)) and model-predicted estimates (Figure 1(b)) the slopes for males and females only differed by the intercept.

For asymmetry, public insurance or self-pay beneficiaries, assistive device use, progressive disease, slower T25FW, and greater PHQ-9 were significantly associated with worsening (Table 3). Sex and T25FW interacted with disease duration. In males, asymmetry increased by 1 second per decade (Table 3 and Figure 2), driven by worsening non-dominant hand times while dominant hand performance plateaued after 75 seconds (eFigure 4). The T25FW, a 1 second increase in T25FW corresponded to an additional 0.05 second increase in asymmetry per decade (Table 3 and Figure 3).

Including BMI as a covariate (N = 2540 patients and 5090 observations) did not change results, and BMI was not associated with 9HPT outcomes (data not shown).

Discussion

This is the first repeated-measures longitudinal study examining dominant and non-dominant 9HPT times, as well as asymmetry, conducted in a large, diverse, real-world MS patient population. We observed independent associations between multiple clinical, demographic, and social determinants of health and worsening dominant 9HPT times, with more pronounced effects on non-dominant hands. Males exhibited a faster rate of decline in dominant hand function with increasing disease duration. While fewer attributes were associated with increases in asymmetry, two key interactions emerged: males and individuals with slower walking speeds experienced more pronounced discordance between hand times as a function of disease duration.

In general, these 9HPT and asymmetry associations parallel longitudinal findings for T25FW, PHQ-9, perceived global disability, and cognitive impairment, except for the null findings for BMI.^8–10,29 Sexual dimorphism in MS risk, presentation, and progression is well documented,³⁰ and our results for sex are consistent with a secondary analysis of clinical trials data,²⁶ and we also demonstrated greater asymmetry in males. Not only do males on average have worse upper extremity functioning, they more rapidly accrued dominant hand dysfunction and greater discordance between hands over the disease course evidenced by statistical interactions. These findings are biologically plausible: upper- and lower-limb function share overlapping neural substrates, and lesions affecting the corticospinal tracts, cerebellum, or brainstem can impair coordination and motor control across both limbs.^6,7 Neuroimaging studies further support this rationale, showing that structural damage and functional reorganization within sensorimotor networks are directly linked to upper- and lower-limb performance in PwMS.^6,16,31 In parallel, the observed effects of social determinants and modifiable lifestyle factors align with their systemic influence on inflammation, neurodegeneration, and access to rehabilitation. Together, these biological and contextual mechanisms provide a coherent framework for why both clinical and social factors shape the trajectory of upper extremity dysfunction in MS.

Disease duration, assistive device use, slower walking speed, and progressive disease were independently associated with worse dominant and non-dominant hand times and greater asymmetry over time, as expected given our understanding of MS. Older age was associated with slower 9HPT times but not asymmetry, suggesting symmetric decline in upper extremity functioning with aging. Depressive symptom severity negatively impacted all three outcomes, paralleling longitudinal associations for T25FW and perceived global disability.⁸ While these relationships may not reflect causal relationships, we hypothesize they may reflect shared biological processes (i.e. diminished repair capacity and accelerated age-related decline).

We observed greater hand impairment for Hispanic/Latinx and Blacks compared with non-Hispanics and Whites. A prior cross-sectional study observed Black PwMS had 2.1 second slower 9HPT (averaged across hands) compared with White PwMS.³² We observed similar magnitudes longitudinally for the dominant and non-dominant hand. However, the underlying race difference is likely larger than what we report as potential mediators (factors downstream of race in causal paths to upper extremity impairment, for example, insurance payer, ADI) were included as covariates. In post hoc analyses of our models excluding insurance payer and ADI, the difference between Black versus White PwMS increased by 15% (2.22 seconds; 95% CI = 1.34, 3.09) and 26% (2.84s; 95% CI = 1.73, 3.94) for dominant and non-dominant 9HPT times, respectively. For Hispanic/Latinx versus non-Hispanic PwMS, dominant and non-dominant 9HPT times increased by 3% (1.91s; 95% CI = 0.14, 3.69) and 5% (3.14s; 95% = CI 0.004, 6.28), respectively. The null results for race and ethnicity in the asymmetry model were unchanged. These post hoc analyses suggest social inequities may mediate portions of the ethnoracial differences in upper extremity dysfunction. However, the relative sample sizes of these ethnoracial subgroups were modest and may be subject to sampling variability. Thorough mediation and moderation analyses are warranted to generate clearer insights.

Two of the modifiable behavioral and lifestyle factors (tobacco smoke exposure, obesity), social determinants of health (individual SES, care access reflected by insurance payer, neighborhood-level SES reflected by ADI), and other social attributes (marital status) investigated had profound associations with upper extremity function. Smoking has been consistently associated with worsening MS outcomes, including disease progression, walking speed, perceived disability, depressive symptom, processing speed, and gray matter loss, with greater deficits in current smokers.^8,9,33 Our results builds on a cross-sectional study where active smokers with MS self-reported greater hand impairment.³⁴ Here, current and former smokers had significant worsen of 9HPT times, with greater deficits experienced by active smokers independent of disease course and lower-limb functioning. Public insurance beneficiaries also had greater deficits across the three outcomes, even with adjustment for covariates often considered as confounders (i.e. age and race/ethnicity). Since patients were receiving care at a tertiary referral facility, these associations more likely reflect adverse SES conditions rather than variation in care continuity/access, and parallels associations observed for other MS outcomes.^8,9,32 A noteworthy finding was that ADI was only associated with worsening non-dominant 9HPT times and not dominant 9HPT times. We speculate this finding may reflect compensatory prioritization of dominant hand function for PwMS living in low-resourced settings. This aligns with growing evidence that adverse neighborhood SES independently contributes to worse MS outcomes.⁸

There are several strengths to this study, including the opportunity to simultaneously examine independent relationships for multiple clinical, demographic, and social determinants of health in a large, diverse, real-world MS patient population. We were able to examine three outcomes, as well as interactions with time. A primary limitation is that this study was comprised of patients seen within a single tertiary healthcare enterprise, although they were residents spanning the US with a preponderance of Ohio residents (eFigure 1). Second, the modest sample sizes for ethnoracial minorities indicate that those findings should be interpreted with caution. Third, BMI data were only available for a subset of encounters for patients less reliant on assisted devices and who had a lower prevalence of progressive disease. The absence of detailed disease-modifying treatment information is another limitation. DMT use, type, and treatment history may influence the rate decline, and accounting for these factors could refine the models. However, as our cohort was drawn from long-term patients managed at tertiary MS centers, it is likely that the majority were receiving guideline-based care; nonetheless, variation in DMT efficacy, timing of initiation, and access, potentially influenced by insurance payer, may partially underlie some observed associations. Similarly, physical activity is known to affect functional outcomes in MS, yet it was not available and may have led to residual confounding. Another limitation is that this study was restricted to those with 9HPT for both sides; thus, generalizability to those with full impairment in one arm may be limited. Nonetheless, recent analyses of progressive MS trials reinforce the importance of the 9HPT, showing that it remains one of the few outcomes capable of detecting meaningful change even in non-ambulatory patients and that alternative scoring approaches (e.g. either-hand 9HPT) can increase sensitivity to worsening.³⁵ These complementary findings support the broader validity of 9HPT-based measures as indicators of upper-limb disability across the full spectrum of MS.

In summary, we identified multiple prominent and independent associations for longitudinal changes in upper extremity performance in MS, which may inform prognostication efforts and highlight the multifaceted nature of MS disability accrual and progression. These findings contribute to the growing literature on sex dimorphisms, ethnoracial differences, and highlight the need for quantifying health disparities and understanding the broader impact of social determinants of health in MS. Importantly, future research should also investigate other aspects of hand functioning, such as grip strength, as well as the concordance with patient-reported outcomes, which are likely to capture different components of upper extremity function that are most relevant to daily activities. Also, the influence of various disease-modifying treatments needs to be examined. Finally, it is essential to characterize the drivers of upper extremity performance in vulnerable populations, including structural barriers to care, access to rehabilitation services, and biological contributors to differential disease progression. Identifying potentially modifiable risk factors, such as active smoking, may help guide equitable care strategies aimed at preserving upper extremity function in MS patients.

Supplemental Material

sj-docx-1-msj-10.1177_13524585251384060 – Supplemental material for Longitudinal modeling of upper extremity function in multiple sclerosis: Associations for clinical and sociodemographic factors

Supplemental material, sj-docx-1-msj-10.1177_13524585251384060 for Longitudinal modeling of upper extremity function in multiple sclerosis: Associations for clinical and sociodemographic factors by Farren BS Briggs, Jiayue Yang, Karlo Toljan, Devon S Conway, Carrie M Hersh, Marisa P McGinley, Robert A Bermel, Deborah M Miller, Alessandro S De Nadai, Douglas D Gunzler and Daniel Ontaneda in Multiple Sclerosis Journal

Footnotes

Acknowledgements

We thank Mr. Scott Husak for his assistance with data curation and management.

Author Contributions

FB was responsible for conception and design of the study; all authors provided critical feedback on the study design. DO, DC, CH, MG, RB, and DM contributed to data acquisition. FB, AD, and DG guided the statistical analysis plan. JY conducted exploratory analyses. FB conducted final analyses. JY and KT drafted portions of the manuscript; FB completed the first draft and generated all figures. All authors guided interpretations and edited and approved the final draft. The authors have full access to the data used in the analyses, and they have the right to publish all data, separate and apart from the guidance of any sponsor. FB takes full responsibility for the data, the analyses and interpretation, and the conduct of the research.

Data Availability

Data are available from the Cleveland Clinic to qualified investigators.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Farren BS Briggs

Jiayue Yang

Karlo Toljan

Devon S Conway

Carrie M Hersh

Marisa P McGinley

Robert A Bermel

Deborah M Miller

Alessandro S De Nadai

Daniel Ontaneda

Supplemental material

Supplemental material for this article is available online.

References

Kierkegaard

Einarsson

Gottberg

, et al. The relationship between walking, manual dexterity, cognition and activity/participation in persons with multiple sclerosis. Mult Scler 2012; 18(5): 639–646.

Julian

Vella

Vollmer

, et al. Employment in multiple sclerosis. Exiting and re-entering the work force. J Neurol 2008; 255(9): 1354–1360.

Bertoni

Lamers

Chen

, et al. Unilateral and bilateral upper limb dysfunction at body functions, activity and participation levels in people with multiple sclerosis. Mult Scler 2015; 21(12): 1566–1574.

Kister

Bacon

Chamot

, et al. Natural history of multiple sclerosis symptoms. Int J MS Care 2013; 15(3): 146–158.

Krieger

Antoine

Sumowski

JF.

EDSS 0 is not normal: Multiple sclerosis disease burden below the clinical threshold. Mult Scler 2022; 28(14): 2299–2303.

Strik

Cofré Lizama

Shanahan

, et al. Axonal loss in major sensorimotor tracts is associated with impaired motor performance in minimally disabled multiple sclerosis patients. Brain Commun 2021; 3(2): fcab032.

Madsen

MAJ

Wiggermann

Marques

MFM

, et al. Linking lesions in sensorimotor cortex to contralateral hand function in multiple sclerosis: A 7 T MRI study. Brain 2022; 145(10): 3522–3535.

Briggs

FBS

Thompson

Conway

DS.

Prognostic factors of disability in relapsing remitting multiple sclerosis. Mult Scler Relat Disord 2019; 30: 9–16.

Alshehri

Cohen

Ontaneda

, et al. The impact of cigarette smoking on cognitive processing speed and brain atrophy in multiple sclerosis. Mult Scler 2023; 29(7): 846–855.

10.

Gunzler

De Nadai

Miller

, et al. Long-term trajectories of ambulatory impairment in multiple sclerosis. Mult Scler 2023; 29(10): 1282–1295.

11.

Briggs

FBS

Trapl

Mateen

, et al. Common social and health disparities contribute to racial differences in ambulatory impairment in multiple sclerosis. Int J MS Care 2024; 26(1): 36–40.

12.

Towfighi

Berger

Corley

AMS

, et al. Recommendations on social determinants of health in neurologic disease. Neurology 2023; 101(7 suppl 1): S17–S26.

13.

Baumer

Pita

Baez

, et al. By what molecular mechanisms do social determinants impact cardiometabolic risk? Clin Sci 2023; 137(6): 469–494.

14.

Neto

Fernandes

Barateiro

The complex relationship between obesity and neurodegenerative diseases: An updated review. Front Cell Neurosci 2023; 17: 1294420.

15.

Gomez-Mejiba

Zhai

Akram

, et al. Inhalation of environmental stressors & chronic inflammation: Autoimmunity and neurodegeneration. Mutat Res 2009; 674(1–2): 62–72.

16.

Zhong

Nantes

Holmes

, et al. Abnormal functional connectivity and cortical integrity influence dominant hand motor disability in multiple sclerosis: A multimodal analysis. Hum Brain Mapp 2016; 37(12): 4262–4275.

17.

Feys

Lamers

Francis

, et al. The Nine-hole peg test as a manual dexterity performance measure for multiple sclerosis. Mult Scler 2017; 23(5): 711–720.

18.

Rudick

Cutter

Reingold

The multiple sclerosis functional composite: A new clinical outcome measure for multiple sclerosis trials. Mult Scler 2002; 8(5): 359–365.

19.

Goldman

LaRocca

Rudick

, et al. Evaluation of multiple sclerosis disability outcome measures using pooled clinical trial data. Neurology 2019; 93(21): e1921–e1931.

20.

Alonso

Eizaguirre

Cohen

, et al. Upper limb dexterity in patients with multiple sclerosis: An important and underrated morbidity. Int J MS Care 2021; 23(2): 79–84.

21.

Solaro

Grange

Di Giovanni

, et al. Nine hole peg test asymmetry in refining upper limb assessment in multiple sclerosis. Mult Scler Relat Disord 2020; 45: 102422.

22.

Farrell

3rd Motl

Learmonth

, et al. Persons with multiple sclerosis exhibit strength asymmetries in both upper and lower extremities. Physiotherapy 2021; 111: 83–91.

23.

Solaro

Cattaneo

Brichetto

, et al. Clinical correlates of 9-hole peg test in a large population of people with multiple sclerosis. Mult Scler Relat Disord 2019; 30: 1–8.

24.

Solaro

Di Giovanni

Grange

, et al. Box and block test, hand grip strength and nine-hole peg test: Correlations between three upper limb objective measures in multiple sclerosis. Eur J Neurol 2020; 27(12): 2523–2530.

25.

Solaro

Di Giovanni

Grange

, et al. Correlation between patient-reported manual ability and three objective measures of upper limb function in people with multiple sclerosis. Eur J Neurol 2023; 30(1): 172–178.

26.

Kuspinar

Mayo

, et al. Longitudinal analysis of disability outcomes among young people with MS. Mult Scler Relat Disord 2021; 52: 102966.

27.

Katzan

Speck

Dopler

, et al. The knowledge program: An innovative, comprehensive electronic data capture system and warehouse. AMIA Annu Symp Proc 2011; 2011: 683–692.

28.

Maroko

Doan

Arno

, et al. Integrating social determinants of health with treatment and prevention: A new tool to assess local area deprivation. Prev Chronic Dis 2016; 13: E128.

29.

Briggs

Gunzler

Miller

, et al. Mobility trajectories in multiple sclerosis: A comparative study of timed 25-foot walk and a patient-reported outcome measure. Mult Scler 2024; 30(11–12): 1479–1489.

30.

Gilli

DiSano

Pachner

AR.

SeXX matters in multiple sclerosis. Front Neurol 2020; 11: 616.

31.

Strik

Shanahan

van der Walt

, et al. Functional correlates of motor control impairments in multiple sclerosis: A 7 Tesla task functional MRI study. Hum Brain Mapp 2021; 42(8): 2569–2582.

32.

Gray-Roncal

Fitzgerald

Ryerson

, et al. Association of disease severity and socioeconomic status in black and white Americans with multiple sclerosis. Neurology 2021; 97(9): e881–e889.

33.

Manouchehrinia

Tench

Maxted

, et al. Tobacco smoking and disability progression in multiple sclerosis: United Kingdom cohort study. Brain 2013; 136(pt 7): 2298–2304.

34.

Briggs

Gunzler

Ontaneda

, et al. Smokers with MS have greater decrements in quality of life and disability than non-smokers. Mult Scler 2017; 23(13): 1772–1781.

35.

Koch

Repovic

Mostert

, et al. The nine hole peg test as an outcome measure in progressive MS trials. Mult Scler Relat Disord 2023; 69: 104433.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.62 MB