Construct validity of the timed 25-foot walk and robustness of clinical benchmarks in multiple sclerosis: exploratory analyses of TEAMS study

Abstract

Background

The Timed 25-Foot Walk (T25-FW) is a widely used performance measure of ambulatory disability in multiple sclerosis (MS). Fewer studies have examined its construct validity relative to patient-reported outcomes or assessed robustness of previously proposed benchmarks in large cohorts.

Objective

To examine the construct validity of T25-FW using patient-reported health outcomes and to evaluate robustness of previously proposed clinical benchmarks.

Methods

Baseline data from 759 participants with MS enrolled in Tele-Exercise and Multiple Sclerosis (TEAMS) trial were analyzed. Associations between T25-FW performance and Short Form-36 physical functioning and emotional well-being subscales were evaluated using Spearman correlations. Previously proposed clinical benchmarks (<6 s, 6–7.99 s, ≥8 s) were evaluated in relation to patient-reported outcomes. Additional exploratory analyses included median-based anchors, segmented regression, and ROC methods.

Results

T25-FW performance showed a strong association with SF-36 physical functioning (ρ = −0.71), which remained significant after adjustment for demographic factors (ρ = −0.39). In contrast, correlations with SF-36 emotional well-being were negligible. Stratification by established T25-FW clinical benchmarks demonstrated a clear gradient in physical functioning, with highest scores among participants completing the test in <6 s and lowest among those requiring ≥8 s or unable to perform test. Candidate cut points derived from multiple analytical approaches clustered around similar walking-time ranges and demonstrated consistent associations with physical functioning.

Conclusions

The T25-FW demonstrates strong construct validity with patient-reported physical functioning and clear divergent validity from emotional well-being. Findings support robustness of proposed T25-FW benchmark categories and reinforce the utility of T25-FW as measure of ambulatory disability in MS.

Clinical trial registration: ClinicalTrials.gov Identifier: NCT03117881.

Highlights

• •Construct validity of T25-FW evaluated using SF-36 patient-reported outcomes.
• •T25-FW performance strongly associated with patient-reported physical function.
• •Established T25-FW benchmarks validated in a large real-world MS cohort.
• •Multiple analytic approaches tested robustness of T25-FW benchmarks.
• •T25-FW benchmarks correspond to meaningful differences in physical functioning.

1. Introduction

Walking impairment is common, burdensome, and poorly managed in multiple sclerosis (MS) [1]. This has motivated the focus on reliable and valid measurement of walking impairment for MS clinical care and research [2]. The Timed 25-Foot Walk (T25-FW) is one of the best characterized measures of walking impairment based on walking speed [3] and further a component of the Multiple Sclerosis Functional Composite [4]. The T25-FW has excellent reliability (e.g. test–retest ICC ≈ 0.90–0.98) across a wide range of disability levels in people with MS. [5], [6] The T25-FW further has construct validity based on associations with other performance measures of walking, lower extremity function, and disability status in MS, and it demonstrates responsiveness in clinical and rehabilitation contexts with a 20% or larger change often considered clinically meaningful [3], [5], [6], [7]. There further are benchmarks for the interpreting the T25-FW as indicating “mild,” “moderate,” and “severe” walking impairment [6].

There are notable limitations of prior research on the T25-FW in MS. The robust literature on T25-FW psychometrics has largely been conducted in clinical trial samples or large multicenter registries, that may not reflect the demographic, socioeconomic, and clinical heterogeneity of real-world, regionally distinct cohorts. The construct validity of the T25-FW has often been evaluated with clinician-rated disability scales and objective walking performance measures, yet comparatively less evidence has focused on its validity relative to patient-reported outcome measures (PROMs) such as the Patient-Determined Disease Steps (PDDS) and general health-related quality of life instruments like the SF-36 Physical Functioning scale. This gap is important given the emphasis on PROMs in MS [8] and because patient-reported disability and physical functioning may capture aspects of ambulatory impairment that differ from clinician-rated or pure performance measures [3]. The benchmarks for interpreting T25-FW were identified and prospectively validated in relatively small cohorts, and there is a need for larger studies that confirm the clinical utility and relevance across the full range of ambulatory performance [9].

The present study examined the construct validity of the T25-FW relative to both patient-reported disability and self-reported physical functioning in a large Deep South MS cohort, thereby extending prior work with a real-world, diverse population and enhancing understanding of how T25-FW performance correlates with PROMs in MS. We evaluated the construct validity of the T25-FW in this population by examining its associations with two SF-36 subscales, Physical Function and Emotional Well-Being, and the PDDS. Those scales capture patient-reported physical disability and mental health, respectively. We hypothesized that T25-FW performance would correlate strongly with SF-36 physical function and PDDS (convergent validity), but weakly with emotional well-being (divergent validity). We further evaluated the robustness and reproducibility of previously proposed T25-FW clinical benchmarks by examining whether similar cut points emerge in our dataset and whether alternative data-driven thresholds yield consistent associations with patient-reported outcomes. Prior work has suggested thresholds such as <6 s, 6–7.99 s, and ≥ 8 s that reflect increasing levels of functional limitation, though additional validation across independent cohorts is needed.

2. Materials and methods

2.1. Data source

This ancillary study utilized baseline data from participants with MS enrolled in the Tele-Exercise and Multiple Sclerosis (TEAMS) study, a pragmatic therapeutic exercise clinical trial [10]. The trial encompassed 43 physical therapy clinics across Mississippi, Alabama, and Tennessee. Participants were enrolled across the participating clinic sites using multiple recruitment strategies, including clinic- and physician-based outreach, study flyers and mailings, the study website and social media, community events and health fairs, and outreach through MS self-help groups and community organizations [11]. Participants were eligible for participating in the study if they met the following criteria: (1) age 18–70 years; (2) self-reported diagnosis of MS; (3) a PDDS score between 0 and 7, where the PDDS is an ordinal scale ranging from 0 (Normal) to 8 (Bedridden) that measures self-reported disability status; (4) ability to use both arms and legs for exercising while standing or seated, inclusive of individuals with hemiparesis; and (5) physician permission to participate in the study. Exclusion criteria encompassed: (1) visual acuity impairments that precluded exercise with a computer tablet; (2) cardiovascular disease events within the past six months; (3) severe pulmonary disease; (4) renal failure; (5) presence of an active pressure ulcer; (6) pregnancy at the time of the study; (7) receipt of rehabilitation services within the past 30 days; and (8) classification as physically active, defined by a health contribution score of ≥24—calculated from moderate and vigorous activity scores—on the Godin Leisure-Time Exercise Questionnaire [12]. Therapists at each clinic received training from research personnel on administering outcome measures, supplemented with videos for reference, written instructions, and checklists to ensure consistency. The TEAMS trial was prospectively registered at ClinicalTrials.gov (NCT03117881) on April 18, 2017, before participants enrollment.

2.2. Measures and procedures

Participants in the main TEAMS study completed a demographic and health history questionnaire, capturing data on age, sex, race, height, weight, and PDDS score, and then completed the 36-Item Short Form Health Survey (SF-36) and the T25-FW test, and additional performance-based measures and PROMs. These assessments were conducted in clinic by trained occupational or physical therapists at baseline, three, six, and twelve months. For the present ancillary analysis, we restricted the dataset to baseline measurements for the T25-FW and selected PROMs.

The SF-36 [13] is a widely used PRO measure assessing health-related quality of life across eight subscales: physical functioning, role limitations due to physical health, bodily pain, general health perceptions, vitality, social functioning, role limitations due to emotional problems, and emotional well-being. Scores range from 0 to 100, with higher scores indicating better health status. The measurement properties of the SF-36 have also been specifically evaluated in people with MS. Prior studies across a broad range of MS disability have examined its clinical appropriateness, reliability, validity, and responsiveness and have provided support for scale-level scoring, while recommending caution in the interpretation of conventional physical and mental component summary scores [14], [15]. For this study, all subscale scores were reported, with particular focus on the physical functioning and emotional well-being subscales for convergent and divergent validity analyses, respectively.

We chose these domain-specific subscales rather than the Physical and Mental Component Summary scores (PCS, MCS) because summary scores are derived using weighted combinations of multiple domains, including negative cross-domain weighting, whereby mental health scales contribute to the PCS and physical scales contribute to the MCS. This scoring approach can produce inconsistencies between subscale profiles and summary scores and may not accurately reflect underlying physical and mental constructs. Therefore, use of individual subscales is preferred when examining construct-specific associations [16], [17], [18].

The T25-FW assesses mobility by measuring the time (in seconds) it takes a participant to walk 25 ft as quickly and safely as possible. Participants were allowed to use assistive devices if needed, and timing began at movement initiation and ended when the participant crossed the 25-ft mark. Each participant performed two trials, and the average time was recorded [19].

2.3. Statistical analysis

2.3.1. Missing data handling and descriptive analysis

No imputation was performed for missing data, except for participants unable to complete the T25-FW. In such cases, a value of 180 s was assigned based on the MSFC administration guidelines [20]. Additionally, T25-FW values exceeding 180 s were truncated at 180 s, consistent with the established maximum allowable time for test completion [21]. Baseline characteristics of the study population were summarized using descriptive statistics. Continuous variables were reported as means and standard deviations (SD), while categorical variables were summarized as frequencies and percentages.

2.3.2. Construct validity analysis

To evaluate construct validity, including convergent and divergent validity, analyses of variance (ANOVA) and analysis of covariance (ANCOVA) were conducted to compare SF-36 physical functioning and emotional well-being scores across established T25-FW clinical benchmarks [9], providing group-level comparisons of patient-reported outcomes across clinically meaningful categories.

Subsequently, Spearman partial correlations were calculated between T25-FW scores and the aforementioned SF-36 subscales, treating T25-FW as both a continuous and categorical variable based on the benchmarks. A two-step process was employed for partial correlation estimation:

• 1.Linear mixed models were fitted for each outcome measure (T25-FW, SF-36 physical functioning, and SF-36 emotional well-being), with clinic site as a random effect and age, sex, and race as fixed effects.
• 2.Residuals from these models were used to compute partial Spearman correlations between T25-FW and SF-36 subscale scores, both unadjusted and adjusted for demographic variables.

The interpretation of the correlation coefficient, whether raw or partial, is guided by established literature. Generally, correlation coefficients ranging from 0.00 to 0.10 are considered negligible, while those between 0.10 and 0.39 are classified as weak. Coefficients within the range of 0.40 to 0.69 are considered moderate, while those between 0.70 and 0.89 are regarded as strong. Finally, correlation coefficients exceeding 0.90 are considered very strong [22].

2.3.3. Exploratory analysis of benchmark robustness

To further evaluate the robustness of previously proposed T25-FW clinical benchmarks, we conducted exploratory analyses to examine whether data-driven cut points derived from the TEAMS cohort were consistent with established thresholds [9]. SF-36 physical functioning items related to daily activities were used as anchors and categorized into three levels: 0 = no limitation, 1 = some limitation, and 2 = substantial limitation, as detailed in Table 1. Item-level responses were used rather than overall subscale scores to capture specific limitations in daily activities (e.g., walking, stair climbing), providing clinically interpretable anchors for ambulatory performance [9].

Table 1.

Original scores of the physical function anchors for each ternary level.

	No limit (0)	Some limit (1)	Substantial limit (2)
SF-36 items determining daily activity (SF-36 Questions 5, 7, 8, 11, 12)
Lifting or carrying groceries	100	50	0
Climbing one flight of stairs	100	50	0
Bending, kneeling, or stooping	100	50	0
Walking one block	100	50	0
Bathing or dressing yourself	100	50	0

Participants from the TEAMS study were divided into two groups: those receiving the remote Complementary and Alternative Medicine (CAM) intervention (TeleCAM, n = 377) and those receiving in-clinic CAM intervention (DirectCAM, n = 382). Baseline data from TeleCAM group was used to determine new benchmarks, which were then validated using the DirectCAM group.

Three complementary analytical approaches were applied to identify candidate cut points:

• 1.Median T25-FW values across anchor-defined levels of physical functioning.
• 2.Segmented regression analysis to identify potential inflection points in the relationship between T25-FW and SF-36 physical functioning scores.
• 3.Cumulative receiver operating characteristic (ROC) curve analysis, using the Youden Index to identify thresholds that best discriminate between levels of physical functioning [23].

3. Results

Table 2 summarizes the participant demographics and baseline clinical characteristics. A total of 759 participants with MS were recruited, among which nearly 89% were females and about 73% were White. The average age was approximately 50 years old, and over 80% of the participants had a relapsing-remitting type of MS.

Table 2.

Demographic and clinical characteristics of the 759 participants with MS.

Variable	Descriptive statistics (n = 759)
Age, mean years (SD)	50.20 (10.67)
Gender, n female (%)	676 (89.06)
BMI, mean kg/m2 (SD)	30.1 (7.00)
Race⁎, n (%)
White	551 (72.60)
Black	180 (23.72)
American Indian	2 (0.26)
Hispanic/Latino	2 (0.26)
Hawaiian/Pacific Islander	1 (0.13)
Other	8 (1.05)
Multi race	12 (1.58)
MS types⁎, n (%)
Relapsing/remitting	620 (81.69)
Progressive	83 (10.94)
Unknown	56 (7.38)
PDDS, median (IQR)	3.00 (1.00, 4.00)

^⁎Indicates variables with missing data.

The descriptive statistics of T25-FW and SF-36 are shown in Table 3. The average T25-FW tests were finished in 8.26 s with a wide variation between the participants (Fig. 1). About 44% of participants finished the 25-ft walk in less than 6 s, about 23% ranged from either 6–7.99 s or over 8 s, while around 5% of participants were not able to perform the test. Most of the participants (74.97%) did not need to use an assistive device for the walk. The average SF-36 physical functioning score was 47.30 among all participants, and the average score of emotional well-being was 68.30.

Table 3.

Descriptive statistics of T25-FW and SF-36 subdomains.

Variable	N (%)
T25-FW time (sec)
Mean (SD)	8.26 (9.08)
Median (IQR)	6.20 (5.05, 8.35)
T25-FW 4 clinical benchmarks
< 6 s	336 (44.27)
6–7.99 s	178 (23.45)
≥8 s	203 (26.75)
Unable to perform	42 (5.53)
T25-FW device
No Assist	569 (74.97)
Ankle-Foot Orthosis (AFO)	21 (2.77)
Cane	62 (8.17)
Walker/ Rollator	76 (10.01)
Crutches	1 (0.13)
Other	7 (0.92)
SF-36 subdomains	Mean (SD)
Physical functioning	47.30 (29.17)
Emotional well-being	68.30 (20.56)

Fig. 1.

Distribution of T25-FW times with clinical benchmarks.

Stratification by T25-FW clinical benchmarks demonstrated a clear gradient in physical functioning. Participants who completed the T25-FW in <6 s had the highest SF-36 physical functioning scores (67.48; 95% CI: 65.18–69.77), whereas those requiring ≥8 s had substantially lower scores (24.55; 95% CI: 21.58–27.52), and participants unable to perform the test had the lowest scores (9.63; 95% CI: 3.09–16.18). In contrast, SF-36 emotional well-being scores were similar across T25-FW benchmark groups, and the differences were not statistically significant (ANOVA p = 0.136) (Table 4).

Table 4.

Differences in SF-36 physical functioning and emotional well-being scores across T25-FW clinical benchmarks.

T25-FW clinical benchmarks	SF-36 physical functioning		SF-36 emotional well-being
	Estimate	95% CI	Estimate	95% CI
	Crude estimates
< 6 s	67.48	(65.18, 69.77)	68.58	(66.37, 70.78)
6–7.99 s	43.98	(40.82, 47.13)	66.15	(63.12, 69.18)
≥8 s	24.55	(21.58, 27.52)	68.40	(65.54, 71.27)
Unable to perform	9.63	(3.09, 16.18)	74.34	(68.05, 80.63)
ANOVA test (p-value)	117.39 (<0.0001)		1.86 (0.136)
	Estimates accounting for age, gender and race
< 6 s	67.77	(58.65, 76.89)	72.72	(64.10, 81.33)
6–7.99 s	44.45	(35.12, 53.78)	68.45	(59.64, 77.27)
≥ 8 s	25.43	(16.22, 34.65)	69.99	(61.28, 78.70)
Unable to perform	10.44	(−0.62, 21.50)	76.06	(65.62, 86.60)
ANCOVA test (p-value)	98.50 (<0.0001)		2.75 (0.05)

Spearman correlation analysis showed that the raw correlation between continuous T25-FW scores and SF-36 physical functioning was strong (ρ = −0.71, 95% CI: −0.75, −0.67). After adjustment for age, gender, and race, the partial correlation was reduced (ρ = −0.39, 95% CI: −0.45 to −0.33). In contrast, when T25-FW was categorized using clinical benchmarks defined by Goldman et al., the correlation with SF-36 physical functioning was not markedly affected by covariate adjustment (−0.69 vs. -0.65). In accordance with the ANOVA and ANCOVA results shown in Table 4, the SF-36 emotional well-being scores were not correlated with T25-FW scores (Table 5).

Table 5.

Spearman correlation between T25-FW and SF-36 scores.

Variable 1	Variable 2	Raw Spearman correlation coefficient	Partial Spearman correlation coefficient⁎
T25-FW time (sec)	SF-36 physical	−0.71 (−0.75, −0.67)	−0.39 (−0.45, −0.33)
T25-FW clinical benchmarks	SF-36 physical	−0.69 (−0.72, −0.65)	−0.65 (−0.69, −0.61)
T25-FW time (sec)	SF-36 emotional	0.001 (−0.07, 0.07)	0.03 (−0.04, 0.10)
T25-FW clinical benchmarks	SF-36 emotional	0.03 (−0.03, 0.10)	−0.02 (−0.09, 0.05)

^⁎Adjusted for age, gender, race and clinic location.

To assess the robustness of previously proposed T25-FW clinical benchmarks, we evaluated whether similar cut points emerged in our dataset and whether these cut points corresponded to differences in patient-reported physical functioning. Candidate cut points derived from the TeleCAM cohort were approximately 5.23 and 9.95 s, 4.0 and 8.66 s, and 6.95 and 7.35 s across the different analytical approaches (Fig. 2, Fig. 3).

Fig. 2.

Cut points for Timed 25-Foot Walk score based on segmented regression.

Fig. 3.

Cut points for Timed 25-Foot Walk score based on cumulative ROC analysis. Receiver operating characteristic (ROC) curves evaluating the discriminative performance of T25FW thresholds. (3A) ROC curve with an area under the curve (AUC) of 0.881 and optimal cut-off of 6.95 s. (3B) ROC curve with an AUC of 0.837 and optimal cut-off of 7.35 s.

When these cut points were evaluated in the validation cohort (DirectCAM), all benchmark structures demonstrated moderate to strong correlations between T25-FW categories and SF-36 physical functioning scores (Table 6). The magnitude of these correlations ranged from −0.59 to −0.66 for raw correlations and − 0.49 to −0.61 after adjustment, indicating consistent associations between slower walking performance and worse patient-reported physical functioning.

Table 6.

Spearman correlations between T25-FW benchmarks and SF-36 physical functioning and emotional well-being.

Method	Benchmark	Raw Spearman	Partial Spearman
1.SF-36 physical functioning score
Median	<5.23, 5.23–9.94, ≥9.95	−0.65(−0.71, −0.59)	−0.61(−0.67, −0.54)
Segmented regression	<4.00, 4.00–8.65, ≥8.66	−0.59 (−0.65, −0.51)	−0.49 (−0.56, −0.41)
Cumulative ROC	<6.95, 6.95–7.34, ≥7.35	−0.66 (−0.72, −0.60)	−0.60 (−0.66, −0.53)
2.SF-36 emotional well-being
Median	<5.23, 5.23–9.94, ≥9.95	−0.01(−0.12, 0.09)	−0.10(−0.20, −0.003)
Segmented regression	<4.00, 4.00–8.65, ≥8.66	0.06(−0.04, 0.16)	0.01(−0.09, 0.11)
Cumulative ROC	<6.95, 6.95–7.34, ≥7.35	0.07(−0.03, 0.17)	0.01(−0.09, 0.11)

In contrast, correlations between the same T25-FW clinical benchmarks and SF-36 emotional well-being scores were negligible (Table 6), further supporting the divergent validity of the T25-FW with respect to non-mobility health domains. Overall, these findings indicate that the relationship between T25-FW performance and patient-reported physical functioning is stable across multiple benchmark structures and is consistent with the previously proposed clinical benchmarks.

4. Discussion

The objective of this study was to examine the construct validity of the T25-FW using patient-reported outcomes and to evaluate the robustness of previously proposed T25-FW clinical benchmark categories in a large cohort of people with multiple sclerosis. Specifically, we assessed whether T25-FW performance was associated with patient-reported physical functioning and emotional well-being and whether established benchmark categories corresponded to meaningful differences in these outcomes.

Our findings show that T25-FW performance is strongly associated with patient-reported physical functioning but demonstrates little relationship with emotional well-being. Participants with faster walking times reported substantially higher SF-36 physical functioning scores, whereas emotional well-being scores were similar across performance groups. This pattern is consistent with the expected construct validity of the T25-FW: walking speed reflects mobility-related health but is largely independent of emotional or psychological domains [3], [21]. Previous studies have similarly reported that the T25-FW correlates strongly with measures of ambulation and physical functioning while showing weaker associations with non-mobility constructs such as mental health or mood [3]. Thus, our findings reinforce the construct that T25-FW is an effective measure of ambulatory disability (convergent validity) while appropriately diverging from unrelated constructs such as emotional well-being (divergent validity).

Importantly, the previously proposed clinical benchmarks demonstrated clear differences in patient-reported physical functioning within our cohort. Participants completing the test in less than 6 s reported the highest levels of physical functioning, whereas progressively slower walking times were associated with markedly lower scores. These findings align with the benchmark framework originally proposed by Goldman et al. [9], who reported that T25-FW categories of <6 s, 6–7.99 s, and ≥8 s correspond to increasing levels of disability, including occupational change, use of assistive devices, and reduced independence in daily activities. Our results demonstrate a similar pattern: participants in the fastest group (<6 s) reported the highest physical functioning scores, whereas those in the slowest groups reported substantially lower scores. These findings suggest that the previously proposed benchmark categories remain meaningful when applied in a large real-world cohort. A meta-analysis of T25-FW performance further supports this interpretation, showing that walking speed declines substantially as disability severity increases in people with MS. [24]

Adjusting for demographic covariates substantially attenuated the correlation between T25-FW and SF-36 physical functioning scores (from ρ ≈ −0.71 to ρ ≈ −0.39), suggesting that factors such as age, sex, or race account for part of the raw association. Older age and disease progression are known to be associated with slower walking speed and reduced physical functioning, so adjusting for these variables removes some shared variance [5]. Adjustment for demographic covariates may therefore account for part of the shared variance between T25-FW performance and physical functioning. In analyses of baseline associations, such attenuation is expected, as adjustment reduces bias introduced by confounders that are associated with both variables [25], [26]. Nonetheless, the adjusted correlation remained statistically significant and of moderate magnitude, indicating that the relationship between walking performance and patient-reported physical functioning persists beyond these demographic factors. In contrast, when T25-FW performance was categorized into clinical benchmarks, the association with SF-36 physical functioning remained strong and was only minimally affected by covariate adjustment. This likely reflects the distinct nature of these representations, whereby clinically defined benchmark categories capture broader functional strata of disability and are less influenced by extreme values and variability, supporting their stability and clinical interpretability. The handling of non-completers may also contribute to this differential attenuation. In the continuous analysis, assigning 180 s to non-completers places these observations at a common upper boundary that enters the covariate-adjustment model numerically and may influence fitted values and residuals. In contrast, categorical analyses retain non-completers as a distinct category and therefore are less dependent on the exact numerical value assigned to non-completion [27], [28]. This methodological distinction may partly contribute to the greater attenuation observed for the adjusted continuous association, although its specific contribution was not directly examined in the present analysis.

Additional analyses further supported the robustness of these benchmark categories. Although different analytical approaches produced slightly different candidate values, the resulting groupings clustered around similar ranges of walking performance. These analyses were not intended to redefine or optimize existing thresholds but rather to examine their reproducibility. Notably, slower-walking participants (T25-FW ≳6 s) consistently demonstrated markedly lower physical functioning. This pattern is consistent with prior benchmark studies indicating that walking times around 6 s represent a transition point at which people with MS begin to experience meaningful functional limitations, including greater reliance on walking aids and reduced independence in daily activities [9], [29].

Statistically, our segmented regression analysis suggested an inflection in the relationship between T25-FW performance and SF-36 physical functioning at approximately 6 s, and ROC analysis demonstrated good discrimination between patients with higher and lower physical functioning (AUC ≈ 0.8). While the median-based approach produced somewhat wider candidate values, the cut points identified by the regression and ROC analyses clustered within the previously reported benchmark range of approximately 6–8 s. The convergence of these estimates with established clinical benchmark categories therefore provides additional support for the robustness of the T25-FW benchmark framework.

Collectively, these findings place the T25-FW within the broader landscape of disability assessment in MS. The measure behaves similarly to other objective ambulation tests, such as the 6MWT and the Multiple Sclerosis Walking Scale, which correlate strongly with physical functioning but less strongly with unrelated domains of health [1], [5]. Overall, our results reinforce the role of the T25-FW as a robust and clinically meaningful indicator of ambulatory disability in people with MS.

5. Conclusion

In summary, the T25-FW shows excellent convergent validity with patient-reported physical functioning in multiple sclerosis and clear divergent validity from emotional well-being. Faster walking times were associated with substantially better SF-36 physical-function scores, whereas no systematic association was found with the SF-36 emotional subscale. These results reinforce the T25-FW's role as a specific measure of ambulatory capacity. Adjusted analyses indicate this relationship is robust even after accounting for demographic factors. The demonstrated construct validity against SF-36 scales supports continued use of the T25-FW as an ambulation measure, and the identified ∼6-s benchmark provides a practical performance threshold. Ongoing research should validate these cut points longitudinally and explore their utility in diverse MS cohorts. By aligning T25-FW performance with patient-reported outcomes, we can improve the clinical interpretation of gait measures and ultimately enhance MS care.

6. Limitations

This study has limitations. Because participants were originally recruited for an exercise intervention trial, the TEAMS eligibility criteria may have introduced selection effects. In particular, the age restriction, requirement for PDDS scores of 0–7, ability to use the arms and/or legs for exercise while standing or seated, and physician clearance to participate, and exclusion of individuals with selected medical conditions or already meeting physical activity guidelines may limit representation of the broader MS population. Accordingly, the generalizability of the present findings to individuals outside these eligibility criteria should be considered. We relied on a generic instrument (SF-36) rather than an MS-specific QOL scale, although SF-36 is well-validated in MS. The emotional well-being subscale captures only one facet of mental health, and other psychological factors might behave differently. Our choice of SF-36 physical function as the anchor, while meaningful, is patient-reported and may not capture all facets of walking disability (other anchors like EDSS or objective gait measures could yield different cut points). Finally, the partial correlations adjusted only for basic demographics; unmeasured confounders (e.g. cognition, fatigue, comorbidities) could influence the associations.

7. Clinical implications

For clinicians, these results underscore the interpretability of the T25-FW. Its strong alignment with SF-36 physical-function confirms that slowing on the test reflects perceptions of real-world functional decline. The established cutoffs (<6, 6–8, >8 s) can be used to stratify patients: e.g. patients crossing the 6–8 s thresholds are likely experiencing notable loss of independence (consistent with Goldman's findings [9]). Importantly, a patient's T25-FW result should not be taken as an indicator of mood or mental health, since no relation was observed with emotional well-being. Therefore, while gait speed is a robust physical endpoint, comprehensive MS assessment still requires separate evaluation of emotional or psychological domains. Adjusting for age, sex and race had little effect on the categorical validity of T25-FW, suggesting it can be applied broadly across patient subgroups. In rehabilitation settings, the T25-FW can serve as a quick screening tool: significant changes in walking speed would signal the need for interventions targeting mobility and endurance, rather than necessarily indicating changes in mood or mental quality of life.

Reprints

Reprints are not available.

CRediT authorship contribution statement

Salma Aly: Writing – original draft, Visualization, Validation, Supervision, Software, Methodology, Formal analysis, Data curation. Hui-Ju Young: Writing – review & editing, Validation, Supervision, Resources, Project administration, Investigation, Funding acquisition, Conceptualization. Tracy Flemming-Tracy: Writing – review & editing, Validation, Project administration, Investigation, Funding acquisition. Navneet Kaur Baidwan: Writing – review & editing, Methodology, Investigation, Formal analysis, Data curation. James H. Rimmer: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization. Robert Motl: Writing – review & editing, Investigation, Conceptualization. Tapan Mehta: Writing – review & editing, Validation, Resources, Project administration, Investigation, Funding acquisition, Data curation, Conceptualization.

Ethics approval and consent to participate

This study was approved by the University of Alabama at Birmingham Institutional Review Board for Human Use (IRB-161017003). All participants provided written informed consent prior to enrollment.

Presentation

This work has not been previously presented.

Declaration of Generative AI and AI-assisted technologies in the writing process

During the preparation of this work the author(s) used Microsoft copilot to summarize the text to meet word count required. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication.

Funding

This work was supported by the Patient-Centered Outcomes Research Institute (PCORI) [grant number MS-1511-33653]. The funder had no role in the design of the study; collection, analysis, or interpretation of data; writing of the manuscript; or decision to submit the manuscript for publication.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The data that support the findings of this study are not publicly available due to ethical and institutional restrictions. However, the data may be made available upon reasonable request and with permission the Principal Investigator (blinded).