Introduction

In order to maintain dynamic stability, humans need to successfully adapt their locomotor movements. In doing so, the central nervous system (CNS) must accurately control the location of the centre of mass (COM) relative to its limits of stability (i.e. base-of-support, BOS). To regulate this COM–BOS relationship and maintain dynamic stability, the integration of sensory information within the CNS is required to accurately detect the locations of the COM and BOS (Horak 2006; Winter 1995). Disturbances in the processing or quality of sensory information affecting the CNS’ ability to regulate the COM–BOS relationship may lead to situations of instability as is the case in populations with sensory-related deficits [e.g. ‘individuals with multiple sclerosis’ (IwMS) and ‘community-dwelling older adults’ (OA)]. Although visual and vestibular sensory inputs during locomotion provide immediate optic flow information and positional information of the head in space, it is the somatosensory system that provides the CNS with immediate information regarding the COM–BOS relationship (Perry et al. 2008; Pai and Patton 1997).

Multiple sclerosis (MS) is an autoimmune disease that affects the rate of somatosensory conduction in the CNS (Cameron et al. 2008). Although MS causes a wide range of neurological symptoms, the impaired balance control presents as one of the most prominent symptoms (Kraft and Wessman 1974 as cited in Frzovic et al. 2000). Previous research has demonstrated the subtle changes in postural control in individuals with MS who do not display overt balance impairment according to neurological examination (i.e. IwMS with mild balance impairment). These studies have found that this group of IwMS experience difficulties voluntarily moving their COM towards their BOS limits during standing (Frzovic et al. 2000; Karst et al. 2005; Martin et al. 2006), responding to perturbations (Jackson et al. 1995), and maintaining postural control when their BOS is reduced and vision is removed (Karst et al. 2005; Findling et al. 2011; Cattaneo and Jonsdottir 2009). Likewise, research using clinical tests during situations requiring dynamic control of balance (i.e. gait) have found that IwMS with mild balance impairment display reduced walking speed (Benedetti et al. 1999; Remelius et al. 2008; Thoumie et al. 2005; Sosnoff et al. 2012), shorter step lengths (Martin et al. 2006; Sosnoff et al. 2012), slower cadence (Benedetti et al. 1999; Thoumie et al. 2005), reduced range of motion about their joints (Martin et al. 2006; Benedetti et al. 1999), and increased angular trunk sway (Findling et al. 2011; Fanchamps et al. 2012; Corporaal et al. 2013) compared to controls. These impairments in dynamic stability in IwMS are thought to result from slowed somatosensory conduction and impaired central integration (Cameron et al. 2008), affecting their ability to regulate the dynamically changing COM–BOS relationship.

Although evidence suggests that dynamic stability impairments exist in IwMS who experience mild balance impairment during various clinical balance tasks (i.e. timed up and go, dynamic gait index, twenty-five foot walk and straight walking) (Findling et al. 2011; Benedetti et al. 1999; Remelius et al. 2008; Sosnoff et al. 2012; Fanchamps et al. 2012; Corporaal et al. 2013; Spain et al. 2012), these tasks do so through the evaluation of simple gait kinematics (i.e. trunk roll angle, walking speed, step length, width, and cadence). Although these measures provide insightful analyses of gait kinematics in IwMS, they do not provide information about the integrity of the CNS’ ability to effectively adapt locomotion in response to changes in walking direction (i.e. evaluation of the dynamically changing COM–BOS relationship during steering) in IwMS. Steering is a common daily occurrence which requires the CNS to coordinate and maintain control of the body’s re-orientation in the medial–lateral (M-L) plane (Fuller et al. 2007; Patla et al. 1999; Hollands et al. 2001, 2002). Changes in direction during locomotion which exceed 30° have been shown to require greater balance control (Gillespie et al. 2012) because modifications to normal stepping behaviour are required. The current study used IwMS in order to evaluate the contributions of the somatosensory system’s involvement in the control of locomotion during a steering task with changes in direction larger than 30° to better highlight balance impairment resulting from lesions of the somatosensory system, because evidence displays that lesions of the somatosensory system from the legs tend to increase M-L sway as noted by Horlings et al. (2009) during quiet standing situations.

Balance impairments observed in IwMS (mild balance impairment) from the above-mentioned studies parallel those commonly displayed by OA; however, these studies neglect to directly compare dynamic stability differences between IwMS and OA. In addition to other concomitant factors affecting balance, OA are known to experience multi-sensory-related balance impairments due to natural ageing and may therefore serve as a good comparator group to IwMS. Fuller and colleagues (Fuller et al. 2007) found that OA make re-orientation changes during steering in a similar sequence as young adults; however, they require more steps prior to the turn to initiate these changes. Spain and colleagues (Spain et al. 2012) found that IwMS with mild balance impairment required more time than controls to complete the turn portion of a timed up and go task. These studies suggest that OA and IwMS (mild balance impairment) require more time to coordinate body segments in order to successfully complete a task which challenges dynamic stability requiring individuals to deviate from a normal walking path. These deviations in behaviours from healthy young adults may be related to sensory decline, resulting from natural ageing (Horak 2006; Woollacott et al. 1986; Lin and Woollacott 2002; Maki and McIlroy 1999) or MS (Cameron et al. 2008).

The main objective of the current study was twofold: (1) to determine whether a novel-goal-directed steering task could elicit differences in dynamic stability (i.e. COM–BOS relationship) between IwMS who experience mild balance impairment and ‘healthy age-matched individuals’ (HAMI) and (2) to determine whether these IwMS (mild balance impairment) dynamic stability is different from OA; a novel comparison population which experiences balance impairments as a result of suspected somatosensory loss due to natural ageing. Since IwMS are thought to possess somatosensory-related deficits, we hypothesized that they would have difficulties regulating the relationship between their COM and BOS, which would result in poorer dynamic stability than HAMI during all aspects of the steering task. Conversely, we believed that IwMS would display similar levels of dynamic stability to OA since both groups possess varying deficits in somatosensory integration. IwMS have been shown to display impaired central integration of somatosensory information (Cameron et al. 2008; Cameron and Lord 2010), which is known to provide critical stability information to regulate the COM–BOS relationship. However, despite this somatosensory impairment, IwMS are likely able to up-regulate accurate visual and vestibular sensory information to assist in the regulation of the COM–BOS relationship. Conversely, OA are thought to display suspected somatosensory impairments due to natural ageing. These deficits may lead to improper regulation of the COM–BOS relationship (Perry et al. 2008). Although these age-related changes result in impairments in the sensitivity of global sensory inputs, OA still have the capability of integrating sensory information from all three sources of afferent sensory input. For these reasons, we believe that IwMS will have dynamic stability control levels that closely resemble those of OA rather than their age-matched counterparts.

General methodology

Apparatus

The study was conducted in the Lifespan and Psychomotor Behaviour Lab (LPMB Lab) at Wilfrid Laurier University. Whole body kinematics was calculated using an infrared optoelectric camera recording system (Optotrak Certus, Northern Digitical Inc., Waterloo, ON). The experiment was conducted in a 12 m × 7 m open space. The research project was approved by the university’s Research Ethics Board. Prior to starting the experiment, all participants provided their written informed consent.

Participants

Twelve IwMS with mild balance impairment diagnosed with any type of MS (i.e. relapsing/remitting, secondary progressive, primary progressive, or progressive/relapsing), 12 HAMI and 12 OA volunteered to participate in the study. Prior to the experiment, all participants completed a general screening questionnaire to assess their level of physical activity, history/fear of falling, arthritis/joint problems, neurological/sensory conditions, and visual acuity.

IwMS were recruited from the patient database of a trained Royal College Accredited Neurologist (refer to Table 1 for IwMS characteristics). Participants were screened to meet inclusion criteria set out by the researchers to coincide with criteria previously used in other studies involving similar populations (Findling et al. 2011; Corporaal et al. 2013; Fanchamps et al. 2012; Spain et al. 2012). Participants were included in the study if they could walk 10 m and stand for 10 min unassisted, reported mild balance impairment (i.e. evidence of mild proprioceptive loss in lower extremities as determined via failure of tuning fork test), displayed an overall expanded disability status scale (EDSS) score of ≤3.0 (Kurtzke 1983) and experienced no other confounding neurological or concomitant medical conditions other than MS. EDSS scores for each participant were evaluated by the same neurologist by whom participants were recruited from and performed within 3 days prior to participants completing the experimental protocol. Participants were also required to be medically stable and have no change in medications within the previous 30 days, not used steroids within the previous 60 days, have corrected-to-normal vision and able to produce full extra-ocular movements. Participants were excluded from the study if they experienced peripheral neuropathy, a relapse phase of disease within the previous 60 days, displayed significant pathological balance impairment during the neurological examination.

Table 1 Participant characteristics of individuals with multiple sclerosis with mild disability
Full size table

HAMI served as a comparative group for the first part of the experiment (refer to Table 2 for HAMI characteristics) which focused on quantifying fundamental differences in balance control during a steering task using novel evaluation measures. OA participants served as a comparative group to the IwMS for the second part of the experiment (refer to Table 3 for OA characteristics) which focused on quantifying the existence of balance control difference between IwMS and OA. HAMI and OA were included in the experiment if they self-reported no known neurological conditions, had normal or corrected-to-normal vision and were free of any medical conditions impairing their ability to walk 10 m or stand for 10 min unassisted. Participants were excluded from participation if they experienced any significant sensory disorders which may affect balance (i.e. visual impairment, sensory neuropathy, vestibular dysfunction), cognitive impairments affecting their ability to understand instructions and perform the required tasks, functional limitation in use of limbs, and any other medical conditions affecting ability to adequately perform activities of daily living (Table 4).

Table 2 Participant characteristics of healthy age-matched individuals
Full size table
Table 3 Participant characteristics of older adults
Full size table
Table 4 Mean values of dynamic stability measures represented (± standard deviation)
Full size table

Experimental protocol

The experiment consisted of a steering task (Fig. 1) to assess dynamic stability. The task was separated into three distinct phases [i.e. (1) approach phase (start position to trigger mat), (2) anticipatory postural adjustment (APA) phase (trigger mat to step before turn), and (3) turn phase (first step of turn to end of travel path)], in which certain measures were calculated and separated. All participants performed the task barefoot.

Fig. 1
figure 1

Aerial view of the experimental set-up. The steering task included an 8 m path with 5 possible steering directions following contact with the trigger

Full size image

Participants were instructed to walk straight along a 3 m path towards a pressure sensitive mat (i.e. trigger) at their self-selected pace. Once their foot came into contact with the trigger, one of four lights illuminated which were positioned on the ground at the end of each of the four new travel paths (i.e. 45° to the left or right and 60° to the left or right) and outlined the future direction of travel. Participants were instructed that if no light illuminated, they were to continue walking straight ahead. Once contact with the trigger was made, participants were instructed to take two additional steps (~1.2 m) before turning towards the intended direction and completing the remaining 2.5 m towards the end of the travel path.

In order to familiarize the participants with the walking task, four baseline straight walking trials were completed prior to the start of the experimental trials to ensure participants were comfortable stepping on the trigger with the same foot each time. Participants completed a total of 25 walking trials which were separated into five blocks of trials. Each block consisted of one trial of each of the five different turning conditions presented in a random order.

Participants were allowed to strike the trigger with either foot, but were instructed to remain consistent throughout the study. Since participants contacted the trigger with the same foot each trial and were required to take two steps following contact with the trigger, participants implemented equal stepping strategies (i.e. step narrow or step wide) when turning towards each of the five intended directions. This allowed for comparison of trials across participants, as steering directions were labelled as: step wide 45°, step wide 60°, step narrow 45°, step narrow 60°, or step straight 0°. Turning angles of 45° and 60° were chosen for this study because previous research has shown that changes in direction during locomotion which exceed 30° require greater dynamic stability (Gillespie et al. 2012). Kinematic data were collected at a sampling frequency of 60 Hz.

Data analysis

Gait kinematics

The location of each participant’s anterior–posterior (A-P) and M-L COM was determined using a weighted average of the 17 digitized marker set-up (see (Winter 2004) for adjusted calculation). The A-P COM was used to calculate the approach velocity for each participant by identifying the position of the A-P COM over time. The movement of each participant’s trunk segment was estimated using the five digitized markers located on the trunk (i.e. right and left glenohumeral joints, 12th thoracic vertebrae, and right and left posterior-superior iliac spines) and was used to calculate the M-L trunk (i.e. roll) angle and the variability (i.e. root mean square, RMS) of the trunk roll angle throughout a trial. Participant’s step lengths and widths were determined using the heel markers in the A-P (step length) and M-L (step width) position during the first instant of double support. Cadence was also calculated for each participant during the approach phase by calculating the number of steps taken over a one second period of time.

Onset of segmental re-orientation

Segmental re-orientation onset was calculated, following the illumination of one of the target lights (i.e. APA phase) and identified when participants re-oriented their head and trunks about the vertical axis (i.e. yaw) towards the intended turn direction. The onset of segmental re-orientation was defined as the time in which the head or truck segment exceeded three standard deviations of the normal magnitude during the approach phase. The difference between head and trunk (i.e. trunk-head) segmental rotation onsets were used to determine the turning strategy employed by the participants during the turn (i.e. head first, trunk first or en bloc).

M-L dynamic stability margin (DSM) measurements: minimum distance (DSMmin) and range distance (DSM range)

The M-L DSM was calculated as the M-L distance between the COM and the lateral BOS border (identified using the 5th metatarsal) during single support. The DSMmin was classified as the minimum distance between the M-L COM and the BOS during single support (refer to Perry et al. 2008; Pai and Patton 1997). The DSMmin identifies the point in which the COM has reached its limits of stability (i.e. BOS) and the centre of pressure (COP) propels it away from the BOS and towards the next subsequent step. The DSM range was calculated as the difference between the DSMmin and DSMmax (i.e. maximum distance between M-L COM and BOS during single support).

Temporal dynamic stability measurement (M-L COM velocity)

The M-L COM velocity was calculated as the rate in which the COM was moving in the M-L direction during the first instant in which the foot was in single support. This calculation was performed in order to determine the rate at which the M-L COM was moving towards the lateral BOS.

Statistical analysis

Statistical analyses were separated in order to compare groups of interest and address each of the two objectives independently. Blocks of trials were compared to each other (i.e. block one versus block five) to determine whether fatigue was an influencing factor within or between groups during the dynamic steering task. Stepping strategy was not included as an independent variable for the approach phase gait kinematics analyses because participants were not aware of which direction they would be required to turn during this phase of the task. Analyses during the approach phase consisted of multiple five (block: 1–5) × 2 (group: IwMS and comparison group) repeated measures between group analysis of variance (ANOVA) were performed for step length, step width, cadence, trunk roll RMS angle, and A-P COM velocity (i.e. walking speed).

Separate repeated measures between groups ANOVAs were performed for the difference between trunk and head segmental re-orientation onsets (5 blocks, 4 stepping strategies, 2 groups), DSMmin, DSM range, and M-L COM velocity (5 blocks, 5 stepping strategies, 2 groups). Bonferroni adjustments were made when necessary for all significant main effects and all significance levels were set at p < 0.05.

Results

IwMS versus HAMI

Fatigue did not influence results between IwMS and HAMI as no main effects of block (i.e. block one vs five) or interactions between block and group (i.e. block one vs five within or between IwMS and HAMI) were observed for all dependent measures.

Age

An independent t test revealed that there was no effect of age between IwMS and HAMI.

Gait kinematics

There was no effect of group for step length, step width, or trunk roll RMS angle during the straight walking portion of the task (i.e. p > 0.05). However, results revealed a main effect of group for cadence (F (1, 22) = 4.34, p < 0.05) and walking speed (F (1, 22) = 5.58, p < 0.05). Participants with MS took significantly less steps per second (Fig. 2a) and walked significantly slower (Fig. 2b) than HAMI.

Fig. 2
figure 2

IwMS displayed reduced cadence (a) and slower walking speed (i.e. velocity) (b) than HAMI during the approach to the trigger

Full size image

Segmental re-orientation onset

There was no effect of group in segmental re-orientation onset for all stepping strategies (i.e. p > 0.05). All participants consistently employed the same re-orientation strategy during trials requiring a change in direction. This strategy involved first turning their head followed by their trunk (i.e. onset of head rotation occurred before trunk rotation).

Spatial DSM characteristics (DSMmin and DSM range)

There was no effect of group in average DSMmin (i.e. p > 0.05) (Fig. 3a). Stepping strategy revealed a main effect (F (4, 88) = 29.15, p < 0.001), where the straight walking condition elicited the largest DSMmin (\(\bar{x} = 8. 4 9\pm 2. 2 3\;{\text{cm}}\)) compared with all other turn directions (\(\bar{x} = 6. 3 3\pm 2.0 9\; {\text{cm}}\)).

Fig. 3
figure 3

No differences were observed between IwMS and HAMI for DSMmin throughout the steering task (a). However, IwMS displayed a smaller DSM range than HAMI (b)

Full size image

Results for the average DSM range revealed a main effect of group (F (1, 22) = 7.49, p < 0.05), where IwMS displayed a smaller DSM range than HAMI’s (Fig. 3b). An interaction effect (stepping strategy × group) was observed (F (4, 88) = 5.01, p < 0.05), such that IwMS had a smaller DSM range for all stepping strategies requiring a change in direction. However, both groups displayed similar DSM range values during unchallenging straight walking trials.

Temporal dynamic stability characteristics (M-L COM velocity)

Results for the average M-L COM velocity revealed a significant main effect of group (F (1, 22) = 7.80, p < 0.05), where IwMS moved their COM moved much slower towards their M-L BOS than HAMI (Fig. 4). An interaction (stepping strategy × group) revealed that IwMS moved their COM towards their M-L BOS much slower, regardless of stepping strategy.

Fig. 4
figure 4

IwMS displayed a reduced M-L COM velocity throughout the entire steering task during all steering directions compared with HAMI

Full size image

IwMS versus OA

Fatigue did not influence results between IwMS and OA as no main effects of block (i.e. block one vs five) or interactions between block and group (i.e. block one vs five within or between IwMS and HAMI) were observed for all dependent measures.

Age

An independent t test revealed that the OA were significantly older (t (22) = − 7.98, p < 0.001) than the IwMS (p < 0.01).

Gait kinematics

There was no effect of group for step length, step width, trunk roll RMS angle, cadence, or walking speed during the straight walking portion of the task (i.e. p > 0.05) (Fig. 5).

Fig. 5
figure 5

IwMS displayed similar cadence and walking speed (i.e. velocity) as OA

Full size image

Segmental re-orientation

There was no effect of group in segmental re-orientation onset for all stepping strategies (i.e. p > 0.05). All participants consistently employed the same steering strategy during trials, requiring a change in direction. This strategy involved first turning their head followed by their trunk (i.e. onset of head rotation occurred before trunk rotation).

Spatial stability margin characteristics (DSMmin and DSM range)

There was no effect of group in DSMmin and DSM range regardless of the stepping strategy (i.e. p > 0.05) (Fig. 6).

Fig. 6
figure 6

No differences were observed between IwMS and HAMI for DSMmin (a) or DSM range (b)

Full size image

Temporal stability margin (M-L COM velocity)

Similar to spatial stability margin characteristics, there was no effect of group when comparing their M-L COM velocity regardless of stepping strategy (i.e. p > 0.05) (Fig. 7).

Fig. 7
figure 7

IwMS displayed similar M-L COM velocity compared to HAMI during all steering directions

Full size image

Discussion

The objective of the current study was twofold: (1) to substantiate that differences in dynamic stability exist between IwMS and HAMI during a steering task and using a novel evaluation measure (i.e. the dynamic stability margin); and (2) to determine whether IwMS display different dynamic stability control from OA. The latter objective provided a novel comparison population (i.e. OA) which experiences balance impairments as a result of suspected somatosensory loss due to natural ageing. Findings from the first part of the experiment indicate that IwMS displayed: (1) reduced dynamic stability (via DSM) throughout the steering task; and (2) reduced walking speed and cadence during the straight walking portion of the task. No changes in trunk roll angle, step length, step width, or segmental re-orientation strategy were observed between IwMS and HAMI. Findings from the second part of our experiment revealed that IwMS displayed similar dynamic stability characteristics to those of OA. These latter findings present unique insights which highlight that balance impairment arising from the suspected contribution of somatosensory impairments in ageing is similar to that experienced by individuals with isolated somatosensory impairments observed in IwMS.

Do dynamic stability differences exist between IwMS and HAMI?

The somatosensory system provides immediate information to the CNS regarding dynamic stability (COM–BOS relationship) (Perry et al. 2008; Pai and Patton 1997). In order to evaluate the role of the somatosensory system in controlling dynamic stability, the current study employed a steering task to challenge the COM–BOS relationship. Since changes in direction larger than 30° in the M-L direction require modifications to normal stepping behaviour (Gillespie et al. 2012), and since lower limb somatosensory lesions tend to increase M-L sway during quiet standing (Horlings et al. 2009), the current study set out to assess how IwMS, who are thought to experience somatosensory-related balance impairments, responded to challenging walking conditions aimed to highlight their balance impairments.

The DSMmin provides a good indication of the integrity of the COM–BOS relationship (Perry et al. 2008; Pai and Patton 1997). If step width was similar between HAMI and IwMS, a larger DSMmin would indicate greater stability (i.e. COM further from lateral border of BOS during single support) and a smaller DSMmin would indicate less stability (i.e. COM closer to lateral border of BOS during single support). Although it was expected that IwMS would have difficulty regulating the COM–BOS relationship and would display smaller DSMmin than HAMI, both groups produced similar DSMmin values throughout the steering task (Fig. 3a). However, significant differences in DSMmin for both groups were equally observed during conditions requiring changes in direction, such that straight walking elicited the largest DSMmin (i.e. most stable condition) compared with all other turning directions. This suggests that changes in turning direction of 45° and 60° away from the midline effectively challenged participants’ balance control system and required them to adapt their gait parameters to maintain stability.

Although DSMmin values were not different between groups, IwMS displayed a smaller DSM range (Fig. 3b). This reduced DSM range indicates that IwMS’ COM was contained within an area closer to their lateral BOS limit. The reduced DSM range may suggest a reduced ability to respond to lateral perturbations, since the ankle musculature controlling each step has less time to respond and prevent the COM from travelling beyond the limits of the lateral BOS (Perry et al. 2008; Pai and Patton 1997). Since IwMS consistently employed this reduced DSM range (Fig. 5b), it could be that they were not able to effectively use somatosensory feedback throughout the walking task to adapt their stepping behaviour, but instead may have relied on a proactive, feed-forward stepping strategy.

In the current experiment, IwMS seemingly employed a proactive control strategy in order to maintain dynamic stability by significantly reducing their walking speed in comparison with the HAMI (Fig. 2b). This proactive control strategy of reduced walking speed was observed during the approach to the trigger in anticipation that they (IwMS) may be required to change their direction of travel, following contact with the trigger. The reduction in walking speed was a product of stepping more slowly (i.e. decreased cadence) (Fig. 2a), which also effectively reduced the rate of lateral COM movement towards the dynamically changing BOS (Fig. 4) (Frzovic et al. 2000; Karst et al. 2005; Martin et al. 2006). This reduction in lateral COM velocity in IwMS was observed even after the trigger during all travel path directions (including straight walking trials), suggesting that IwMS employed a proactive control strategy which persisted throughout the entire trial rather than a reactive strategy in response to changes in travel path direction. Therefore, findings from the current study suggest that when somatosensory information is unavailable, individuals employ an appropriate proactive control strategy to modulate COM movements to ensure dynamic stability. As discussed earlier, multiple sclerosis affects the rate of somatosensory conduction within the CNS (Cameron et al. 2008). Therefore, these proactive control strategies produced by the IwMS are likely a result of the slowed relay of somatosensory feedback within the CNS, ultimately requiring these individuals to employ behaviours that will best increase their chances of stability during the task in order to complete the task safely and without falling.

Although differences in somatosensory feedback information are thought to be the primary cause for understanding differences observed in DSM and gait characteristics between IwMS and HAMI, visual processing of information cannot be entirely discounted as a possible alternative for understanding differences in DSM and gait characteristics between IwMS and HAMI. Visual sensory information is primarily involved in proactive control of dynamic stability aiding in route selection and avoidance of obstructions during locomotion (Patla 1997; Lee and Lishman 1977; Cinelli and Patla 2008). Previous research has demonstrated that IwMS display poor postural and dynamic control when vision is removed (Findling et al. 2011; Cattaneo and Jonsdottir 2009; Fanchamps et al. 2012; Corporaal et al. 2013) highlighting the importance this population places on vision to maintain stability. The proactive control strategy experienced by walking slower in IwMS may have therefore served a second purpose; enabling IwMS to spend more time processing visual information from the environment in order to maintain dynamic stability.

Hamilton et al. (2009) found that IwMS walked slower than HAMI when performing a secondary cognitive task, suggesting that IwMS need to devote greater cognitive resources to walking. The current steering task required participants to scan their environment and following contact with the trigger respond correctly by changing their travel path in the direction of the illuminated cue light. In order to be successful at this task, attention and decision making skills were required, making this steering task a cognitively demanding task. Therefore, IwMS may have produced a strategy in which they walked slower to better process visual information about the task and allocate the proper amount of attention on controlling balance during the steering task in order to correctly respond and turn in the intended direction.

The current study showed that following contact with the trigger, all participants employed a similar segmental re-orientation strategy, such that participants first re-oriented their head followed by their trunk towards their new travel path direction. Head first movements illustrate that the light served as an exogenous cue to capture participants’ attention and help set up a more stable locomotor axis (Hollands et al. 2001, 2002; Franconeri et al. 2005). Therefore, both groups were able to use the light cues at the end of each travel path as stable reference points to set up and guide their stability for the remaining portion of the task (Hollands et al. 2001, 2002; Franconeri et al. 2005). Since IwMS rely heavily on vision to maintain balance (Findling et al. 2011; Cattaneo and Jonsdottir 2009; Fanchamps et al. 2012; Corporaal et al. 2013), it is possible that this visual information supplied by the light cue in addition to walking slower to better process this information may have mitigated the potential of instability during the turns.

Do dynamic stability differences exist between IwMS and OA?

Despite these populations experiencing differences in their nature of sensory impairment, current findings demonstrate that IwMS displayed similar dynamic stability characteristics to OA during straight walking and steering. Participants walked at the same speed and produced the same stepping behaviours (i.e. DSMmin, DSM range and M-L COM velocity) when completing the steering task. Since no differences were observed between these groups, this finding suggests that both groups employed similar adaptive strategies. These similarities in strategies used may have been elicited in an effort to compensate for reduced sensory information, resulting in conservative gait behaviours to help complete the task and maintain stability.

The locomotion is an adaptive process which requires strong communication between peripheral sensory inputs and central processing of inputs in order to elicit the appropriate motor outputs required to maintain stability and prevent falling. Specifically somatosensory information is extremely important in this adaptive process since it reports information about the environmental support surface under the feet and how the lower limbs interact with this surface and respond. Inability to communicate information between peripheral somatosensory inputs and the CNS results in impaired control of the dynamically changing COM as noted through unstable stepping behaviours (i.e. reduced DSMmin, increased trunk roll sway, slower walking speed, shorter steps, etc.).

Unfortunately, the current experiment did not compare degree of cutaneous sensation (i.e. filament testing of cutaneous mechanoreceptors) or integrity of the 1A afferents (i.e. tuning fork testing) between groups. As a result, the current experiment successfully indicates that IwMS display similar actions to OA; however, it is unclear which of the sensory system declines affected dynamic stability to a greater extent. It is possible that during a steering task, somatosensory information is most important in controlling dynamic stability because both groups had some somatosensory loss and the task challenged M-L balance control (Horlings et al. 2009). However, both groups were able to use vision as a dominant resource to guide their dynamic stability, compensating for differences that are observed during sensory perturbations (i.e. removal of vision). Both groups display similar conservative gait behaviours as observed in the current experiment and used vision to guide and control locomotion as discussed in experiment one. This suggests that individuals with sensory impairments appear to adopt proactive control strategies (walk slower, reduce lateral COM movement during single support, etc.) in order to successfully complete a steering task and maintain dynamic stability.

Conclusions

The steering task presented a challenging task which effectively perturbed individuals’ dynamic stability when required to change direction. These changes in direction therefore required enhanced on-line control of the DSM in order to successfully complete the task and regulate the distance between the COM and BOS. Although the current results are limited to the fact that level of somatosensory loss was not directly compared between groups, somatosensory loss was objectively confirmed in the pre-screening of the IwMS participants (i.e. failure of tuning fork task and EDSS score ≤3.0). Therefore, results of the current study have the ability to demonstrate that IwMS have the potential to possess poorer dynamic stability compared with HAMI as indicated through differences in DSM characteristics (Fig. 3a, b), which may be due to impairments in their somatosensory processing ability. However, IwMS offset this possible reduction in dynamic stability by implementing a proactive control strategy via reducing their walking speed and limiting the lateral movement of their COM. Furthermore, this study suggests that IwMS may be relying more on visual information from the environment to guide and aid them in successfully completing the task. However, this task cannot definitively answer whether or not vision played an integral role in supporting dynamic stability since non-visual walking behaviours were not assessed. Furthermore, IwMS display similar dynamic stability margin and gait characteristics as OA. This suggests that balance impairment resulting from somatosensory deficits in IwMS is similar to that observed in ageing where there is suspected somatosensory loss. It is further suggested that perhaps both populations may not have been able to effectively use sensory information to regulate their DSM and therefore required control strategies to help them successfully complete the steering task. The experiment extends previous works with IwMS as it provides novel insights into how IwMS respond to changes in walking direction during the gait cycle, a task which is experienced every day. Furthermore, our paradigm reveals new information into how IwMS regulate the biomechanical relationship between their COM–BOS during locomotion.

Since IwMS used vision to help regulate postural and dynamic balance control (Findling et al. 2011; Cattaneo and Jonsdottir 2009; Fanchamps et al. 2012; Corporaal et al. 2013) and clearly used the lights as exogenous cues to set up and align their locomotor axis to increase stability in the current task, the future research should focus on how the DSM in IwMS changes during situations where vision is not available. As well, future studies should address the limitation of the current study and compare level of sensory feedback between groups as well as consider evaluating other, non-sensory factors which may contribute to dynamic stability impairments in IwMS.