Falls are common and can lead to costly sequelae in people with multiple sclerosis (PwMS) (Planche et al., 2016). Approximately 56 % of PwMS fall every 6 months, consistent with population-level estimates (53-59 %), with 30 % of falls resulting in injury (Abou et al., 2024b) and 37 % being recurrent fallers (Coote et al., 2020). Seeking fall prevention interventions is crucial for PwMS.

To develop fall prevention paradigms, it is critical to understand what factors are related to falls. Based on self-reported fall data in everyday life (Abou et al., 2022; Cameron et al., 2013), previous studies identified fall risk factors in PwMS, including a history of falls, fear of falling, walking speed, disability level, cognition, balance, and fatigue (Coote et al., 2020). Although these studies provide valuable insights into fall risk in real-world contexts, some inherent limitations are associated with self-reported data. For example, cognitive dysfunction among PwMS may lead to recall bias and inaccurate fall information (Nilsagård et al., 2009). Falls are a multifactorial issue resulting from interactions among personal (e.g., poor balance, fatigue, pain, fear of falling), environmental (e.g., surface conditions), and activity-related (e.g., multitasking) factors (Abou et al., 2024a). The uncontrolled and heterogeneous causes of falls could result in biases in identifying the mechanisms of falls in PwMS. Therefore, exposing PwMS to a standardized external postural perturbation could be a better option for understanding the mechanisms of falls in PwMS. Lab-based treadmill-induced slips or trips have been widely used to examine fall mechanisms and assess fall risk in various populations, including healthy (Roeles et al., 2018) and clinical (Joshi et al., 2018) populations. Yet, no study has applied treadmill-based postural perturbations to PwMS to assess their fall risk.

Treadmill-induced perturbations have also been used as a fall prevention training platform in several populations (Brown et al., 2023; Yang et al., 2018). Based on motor learning theory, perturbation training involves exposing trainees to repeated controlled postural disturbances, which stimulate the central nervous system (CNS), allowing them to adapt to the repeated postural destabilization and promote the development of fall-resistant motor skills (McCrum et al., 2017). Previous studies have reported that different populations, including healthy older adults (Pai et al., 2014), people with stroke (Mansfield et al., 2017), lower back pain (Sung and Danial, 2018), and Parkinson’s disease (Ribeiro de Souza et al., 2023), can quickly adapt to repeated slips or trips.

Given the reliance of motor learning on CNS and the MS-induced CNS damage, the neuroplasticity and motor learning capability could be compromised in PwMS. For example, a past study, which exposed 13 PwMS to five repeated treadmill gait-slips (Yang et al., 2019), showed a promising motor learning ability across the slips, as evidenced by the substantially reduced fall rate from 92.3 % on the first slip to 30.8 % on the fifth slip. However, four PwMS still fell on the fifth trial and seemed not to respond to the perturbation training. It is meaningful to examine if all PwMS can adapt to a different type of postural perturbation and what factors could determine their ability of motor learning performance.

The purpose of this study was two-fold: 1) to identify what factors could differentiate fallers from non-fallers on the first standing-slip, and 2) to determine what factors distinguish responders from non-responders to five repeated standing-slips among ambulatory PwMS. Given their possible relationship with disease progression, the following factors were considered: age (Coote et al., 2020), disability level (Nilsagård et al., 2015), disease duration (Cameron and Nilsagard, 2018), gait speed (Jawad et al., 2023), leg muscle strength (Pau et al., 2021), fall history (Gunn et al., 2013), and cognition level (Bilgin et al., 2023). We hypothesized that the factors of interest would be significantly worse 1) among fallers than non-fallers on the first standing-slip and 2) among non-responders than responders across the five slips. The first purpose would provide insight into the mechanisms of falls for PwMS, and the second may help us identify responders and non-responders to perturbation training. Overall, this study could facilitate fall prevention efforts in this population.