Introduction

Frailty is a syndrome of diminished reserve and resilience arising from the accumulation of deficits across multiple physiologic systems that leads to vulnerability to adverse health outcomes [1]. The estimated prevalence of frailty in middle-aged (40–65 years of age) people with MS (pwMS) is two to four times higher than in older adults (≥65 years of age) without MS [2]. In addition, frailty is associated with reduced quality of life [3], adverse clinical outcomes [2,4], and common MS-related signs and symptoms in pwMS, including impaired mobility, muscle weakness, fatigue, mental health issues, and cognitive dysfunction [5–9]. Overall, these findings strongly suggest that minimizing frailty could have a positive impact on the health of pwMS. Therefore, there is a critical need to identify and evaluate strategies for reducing frailty in the MS community.

Exercise-based rehabilitation can improve physiological function and hallmarks of frailty such as walking speed, muscle mass, and fatigue in pwMS [10]. Therefore, it may represent a viable strategy to counteract frailty in this population. In a recent study, we provided initial evidence that a 6-week treadmill training (TT) program with and without virtual reality (VR) augmentation resulted in a reduction of frailty levels (pooled mean difference in frailty index: 0.024; 95% CI: 0.010–0.038) in 83 pwMS [11]. Compared to TT alone, however, participants randomized to TT augmented by VR (TT + VR) had a greater improvement in the cognitive frailty index. Importantly, TT + VR targets both motor and cognitive function, which are both affected by MS [12,13] and impaired in frail individuals [14–16]. Although our previous investigation had several merits, it also had some drawbacks. Specifically, frailty was not one of the inclusion criteria, and, due to the secondary analysis research design, the intervention was not conceived to treat frailty as the primary objective. In addition, it did not include a resistance training (RT) component, which is considered an essential active ingredient of exercise programs aiming to reduce frailty in older adults [17–19].

Progressive RT promotes anabolic effects leading to gains in muscle strength and reducing the risk of geriatric problems, such as frailty, in both older adults and people with chronic diseases [20,21]. Studies involving RT in the context of MS have shown that it is safe, tolerable, and that it improves muscle strength as well as MS-related symptoms [22–28]. Previous research in other clinical populations suggests that combining RT with other exercise modalities may represent the best strategy for counteracting frailty [29]. In this respect, combining our previous intervention (TT + VR) with progressive RT may have a synergistic effect on MS-related frailty, as the TT + VR component would target motor and cognitive function while the RT component would counteract the muscle weakness component of frailty. To date, however, it is not known whether this type of multimodal exercise training (MET) is feasible in frail pwMS.

The current study aimed to address this knowledge gap and move the field forward by examining the feasibility of a MET program consisting of motor-cognitive rehabilitation and progressive RT to reduce frailty in frail pwMS. We hypothesized the retention rate, and the proportion of completed training sessions would be more than 80%. Additional feasibility outcomes comprised participant recruitment, participant safety, the appropriateness of data collection and training procedures, evaluation of resources, and measures of user engagement. The effects of MET on exploratory frailty-related outcomes were also examined.

Materials and methods

Results

Feasibility outcomes

Participant recruitment.

Recruitment was conducted between December 5, 2023, and December 12, 2024. Fig 1 details the participant flow diagram throughout the study. Sixteen participants (8.9% of patients approached for consent; observed recruitment rate = 0.33 participants/week), corresponding to 67% of the a priori recruitment target (n = 24), completed the baseline assessment and were successfully randomized to MET (n = 8) or to the waitlist control group (n = 8). The baseline characteristics of the study participants are summarized in Table 1.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Baseline participant characteristics: results are expressed as frequencies and percentages for categorical variables and mean ± standard deviation for continuous variables.

https://doi.org/10.1371/journal.pone.0347063.t001

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. CONSORT flowchart.

Legend: MET: multimodal exercise training.

https://doi.org/10.1371/journal.pone.0347063.g001

Participant retention.

Fourteen participants, eight in the intervention group and six in the control group, completed the study (overall retention rate = 87.5%). Two participants in the control group dropped out from the study due to medical complications unrelated to the study (n = 1) or loss to follow-up (n = 1). Both participants had relapsing–remitting MS, identified as Black, were over 50 years of age, and represented one male and one female.

Participant safety.

In terms of safety, no adverse events were recorded throughout the study and the median training-related pain (NPRS) score in the MET group was 0.0 (interquartile range = 0.0). In addition, none of the training sessions were stopped because of patient-reported pain or discomfort.

Exploratory outcomes

The within-group changes in exploratory outcomes, as well as the between-group differences in baseline to 6-week change, are fully summarized in Table 2. Figs 2 and 3 display the boxplots of the baseline to 6-week changes in the primary and secondary exploratory outcomes, respectively, as a function of group.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Exploratory outcomes. Results are summarized as mean (standard deviation).

https://doi.org/10.1371/journal.pone.0347063.t002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. 6-week changes in the primary exploratory outcome (frailty) as a function of group (boxplots).

Legend: EFIP: evaluative frailty index for physical activity; MET: multimodal exercise training. Note: the dotted line (zero) represents no change in outcome from baseline to 6 weeks.

https://doi.org/10.1371/journal.pone.0347063.g002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. 6-week changes in the secondary exploratory outcomes (quality of life, fatigue, and physiological fall-risk) as a function of group (boxplots).

Legend: MET: multimodal exercise training; MFIS: modified fatigue impact scale; MHC: mental health composite score of the MSQoL-54 questionnaire; PHC: physical health composite score of the MSQoL-54 questionnaire; PPA: physiological profile assessment. Note: the dotted line (zero) represents no change in outcome from baseline to 6 weeks.

https://doi.org/10.1371/journal.pone.0347063.g003

Discussion

This study aimed to examine the feasibility of a novel MET program to reduce frailty in frail pwMS. The intervention satisfied the main criteria for feasibility, as the retention rate and proportion of completed training sessions were 87.5% and 97.2%, respectively. Other feasibility aspects, such as the absence of adverse events and training-related pain, the time spent in training, and the positive feedback provided by participants on the SPFQ indicate that MET was safe, acceptable, and well-received by frail pwMS.

A closer examination of the feasibility data suggests that, overall, our site had adequate resources for the conduct of the study. For instance, the achieved study timeline was aligned with the proposed timeline, and we did not have missing data in the exploratory outcomes. However, the recruitment rate (0.33 enrolled participants/week) was relatively low for an RCT [50]. Only 8.9% of approached individuals ultimately consented to participate, highlighting challenges for recruitment in future trials. The relatively low consent rate likely reflects the combination of the substantial time commitment associated with a supervised, facility-based, thrice-weekly intervention and the stringent inclusion criteria. For instance, we only included individuals who were between 40 and 65 years of age, meeting objective diagnostic criteria for frailty, and able to walk without bilateral support—a hard-to-reach population. Including frail pwMS was one of the most innovative aspects of our study, as the MET program was purposefully designed to counteract frailty in the ambulatory segment of the MS population. The upper age limit (65 years) was selected based on the inclusion criteria from our previous trial and to maximize safety [11,34]. The findings from this study indicated that MET was highly safe and feasible in the included population. Therefore, as a future direction, broadening the inclusion criteria to pwMS over 65 years would be a logical step. Indeed, frailty is associated with higher age in MS, [5] and frail older adults living with MS are a group of people likely to benefit from MET. Other strategies to maximize recruitment in future trials may include increasing flexibility in scheduling training sessions and engaging clinicians more directly in the recruitment process. Moreover, findings from the current investigation raise the possibility that a modified version of our MET program may represent a viable strategy to minimize frailty in the non-ambulatory segment of the MS population. As shown by previous research, upwards of 90% of pwMS with limited ambulation ability (PDDS score of 7) are frail [51]. Therefore, reducing the burden of frailty in this group of people is probably an important yet overlooked therapeutic goal. Several adaptations to promote MET among non-ambulatory pwMS (and ambulatory individuals with substantial mobility impairments), such as utilizing a recumbent stationary bike and RT, are possible. Further research to develop MET protocols to reduce frailty in non-ambulatory pwMS is warranted. In addition, from an implementation perspective, the recruitment challenges observed in this pilot study underscore the importance of optimizing accessibility and scalability of MET interventions for frail pwMS. While the current protocol required in-person, supervised training using specialized equipment, future implementation efforts may benefit from adapting MET components to community-based or clinical rehabilitation settings, or from leveraging tele-rehabilitation and lower-cost technologies where appropriate. Understanding and addressing barriers related to transportation, fatigue, and perceived burden will be critical for successful translation of MET into routine clinical practice should its efficacy be confirmed in larger trials.

The exploratory outcomes were promising, as all participants who were randomized to MET exhibited a reduction in EFIP scores after the intervention (range: −0.02 to −0.14, mean = −0.08, SD = 0.05). Previous research has suggested that reductions in frailty index measures greater than 0.03 represent a clinically meaningful change in frailty levels in older adults [52,53]. Interestingly, the mean change in EFIP found in the current study was greater than these values. However, a direct comparison with previous studies is not possible, as different frailty index measures were used, and the thresholds for clinically meaningful change are provided for contextual interpretation only. Further research is needed to establish what represents a clinically meaningful change in EFIP for frail pwMS. Importantly, participants randomized to MET also showed improvements in quality of life (both PHC and MHC of the MSQoL-54) and fatigue (total MFIS score), which are often seen as corollary measures of frailty. The mean pre-post changes in MSQoL-54 and MFIS exhibited by the experimental group were also consistent with a clinically meaningful change [54,55]. This provides evidence that our MET program satisfied the principle of pleiotropic benefit, showing potential to improve multiple health aspects impacted by MS. We should also note that the training duration for this study was relatively short, and it is quite possible that prolonging MET beyond six weeks may have resulted in a greater impact on the measured outcomes. Relatedly, no meaningful between-group changes in physiological fall-risk (PPA z-score) were observed over the 6-week intervention. This may reflect the short duration of training, suggesting that longer or higher-intensity MET programs may be needed to impact fall-risk in frail pwMS. Finally, the study findings also provide preliminary information that may inform the planning of a future trial. Although formal power calculations were not a primary objective of this feasibility study, the observed between-group differences and corresponding standardized effect size estimates for key outcomes such as EFIP, MSQoL-54, and MFIS may serve as an initial reference for sample size considerations. However, these estimates should be interpreted cautiously due to the small sample size and ideally supplemented with pragmatic considerations related to recruitment and retention when designing the next trial.

Study limitations

The current investigation is not without limitations. First, the study findings should be interpreted with caution due to the small sample size, which may have affected the estimation accuracy of the 95% confidence intervals for the pre-post changes in outcomes. Second, the relatively small sample size raises the possibility that our inclusion and exclusion criteria may have been too stringent, as discussed previously. Relatedly, owing to the selected and hard-to-reach nature of the study population (particularly with respect to age range, ambulatory status, and objective frailty criteria), the generalizability of the findings to broader MS populations may be limited. Further refinements of the inclusion criteria may be needed prior to the next trial. In addition, we used a passive waitlist control group. This methodological approach was chosen to generate proof of concept feasibility data in this early phase of research. However, the lack of an active control group should be acknowledged as a limitation in terms of our ability to generate more robust preliminary efficacy data. Future trials will incorporate an active control group to allow for more robust evaluation of intervention-specific effects. We should also acknowledge that staff time and intervention-related costs were not formally evaluated in this study. Future trials should incorporate detailed assessments of personnel requirements and costs to better inform scalability and implementation of MET interventions. In addition, while feasibility thresholds for retention and adherence were pre-specified (>80%), formal go/no-go progression criteria were not defined for other feasibility outcomes, such as recruitment, safety, or user engagement. This should also be considered a study limitation. Moreover, deviations from the original analytical plan, including reclassification of pre-specified primary and secondary outcomes as exploratory and the use of descriptive rather than inferential statistics, were required by recruitment challenges and should be considered when interpreting the findings.

Conclusions

This was the first study to examine the feasibility of MET to reduce frailty in frail pwMS. The findings indicated that MET was safe, feasible, and well-received by the study participants. In addition, our site demonstrated adequate resources for the conduct of the study, and refinements to the inclusion criteria to maximize recruitment for the next trial were identified. The promising results support the viability of MET for future larger-scale trials targeting frailty reduction in this population.

Supporting information

Acknowledgments

We would like to thank all the participants who took part in this study. We would also like to thank Emilia Potts for the help provided throughout the study.