A rehabilitation program to increase balance and mobility in ataxia of Charlevoix-Saguenay: An exploratory study

Autosomal recessive spastic ataxia of Charlevoix-Saguenay (ARSACS) is characterized by balance impairment and mobility limitations, which both increase the risk of falling. The objective of this study was to explore the effects of a rehabilitation program aimed at increasing trunk and lower limb motor control on balance and walking abilities, and accomplishment of activities of daily living. In this exploratory study, a group-supervised rehabilitation program was performed three times a week for 8 weeks (two sessions at a rehabilitation gym and one pool session). Outcome measures included the Ottawa Sitting Scale, Berg Balance Scale, modified Activities-specific Balance Confidence Scale, 30-Second Chair Stand Test, 10-Meter Walk Test, Barthel Index, and Scale for the Assessment and Rating of Ataxia. Significant improvements in balance, trunk control, maximal and self-selected walking speed difference, ataxia severity and accomplishment of specific activities of daily living were noted for the whole group at the end of the program. At the individual level, all participants improved beyond the standard error of measurement in at least two outcome measures. Also, most participants reported many perceived improvements related to balance, posture and functional mobility. This study provides encouraging results on the effects of a rehabilitation program for ambulatory people with ARSACS. Group intervention could have a positive impact on their daily lives and improve the health care service offered to this population. Future studies with larger sample sizes including control groups and other forms of ataxia are necessary to validate our results to generalize them.


Introduction
Autosomal recessive spastic ataxia of Charlevoix-Saguenay (ARSACS) is a progressive ataxic disorder [1] caused by mutations in the SACS gene [2] located on chromosome 13q12.12 [3]. Two mutations are mainly present, but over 250 mutations have been identified worldwide  [4]. ARSACS is characterized by impairments in three components: pyramidal (spasticity, muscular weakness), cerebellar (ataxia), and distal neuropathy (amyotrophy). Individuals have decreased upper and lower limb coordination, upper limb dexterity, and balance control [5]. They may also experience severe mobility limitations including walking and sit-to-stand transfer limitations, starting in early adulthood and ultimately leading to participation restrictions [5]. Walking abilities are generally lost in their 40s [3,5]. However, high intersubject variability is observed for most signs and symptoms in terms of severity and progression [5].
To date, there is no cure and only one documented exercise program for ARSACS. The latter was an exploratory study with the objective to improve physical condition using an individualized multimodal program including sports and upper and lower strength training. No effect was observed for physical condition but significant improvement was observed for balance and mobility [6]. In other degenerative ataxias, physical therapy programs have been shown to have positive effects on balance, gait and daily living activities [7,8]. Another study has also demonstrated that patients with cerebellar ataxia can learn new motor strategies and compensate for their deficits [9]. To the best of our knowledge, no studies have ever focused on a rehabilitation program for the ARSACS population.
The objective of this study was to explore the effects of a rehabilitation program aiming to increase trunk and lower limb motor control and at improving walking abilities, balance and accomplishment of daily living activities (ADLs) in people with ARSACS.

Trial design
An exploratory study.

Participants
Ten participants were recruited in April 2019 among a subset of 52 people with ARSACS who have participated in a previous longitudinal study led by the Groupe de recherche interdisciplinaire sur les maladies neuromusculaires, Québec, Canada (phase 1-2013 [5]; phase 2-2015 [10]; phase 3-2017). They were selected using a convenience sampling strategy. Patients meeting the following criteria were included in the study: (1) �18 years old; (2) genetically confirmed ARSACS diagnosis; (3) ability to walk 10 meters (with or without walking aids); (4) ability to provide informed consent. Having other medical diagnosis causing physical limitations was the only exclusion criterion (Fig 1). The Ethics Review Board of the Centre intégré universitaire de santé et de services sociaux du Saguenay-Lac-St-Jean (Quebec, Canada) approved the study and written informed consent was obtained. The protocol was registered at clinicaltrials.gov (NCT05479656).

Intervention
Participants were divided into two groups of five for safety considerations. Two well-trained and experienced physiotherapists (PT) delivered the group rehabilitation program. It involved three training sessions per week (one session [50 minutes] in an adapted pool with 4 levels of water, and two sessions [60 minutes each] at the gym of the rehabilitation center) over 8 weeks, for a total of 24 training sessions. The rehabilitation program focused on the following three domains: 1) postural control, 2) balance, trunk-limbs and multi-joints control and coordination and 3) functional mobility (walking and stairs climbing). Each domain contained a standardized list of exercises with one or more component(s). For each exercise component, different difficulty levels were suggested and standardized. The initial level of difficulty for each exercise was determined by the PT based on individual results of the pre-intervention assessment (T0) related to each domain. To ensure individualized progression throughout the rehabilitation program, the PT ensured high balance challenge for each exercise component based on the participant's performance, fatigue level and safety (e.g. modifying the level of stabilization or facilitation, reducing the base of support, reducing arm supports, moving to limits of sway, shifting weight, increasing the number of sets and repetitions or diminishing the  water level in the pool for the execution of the same exercise). A modality interventions log, including participant adherence, each exercise component with the level of difficulty performed, the number of sets and repetitions (dosage), and the participant's perceived difficulty, was compiled by the PT at each session for each participant. In the first weeks, the PT taught participants the importance of slowing movement and using good methods of transfers from sitting to standing (anterior displacement of the center of gravity, foot placement, etc.), among other things.
Participants were assessed by the two physiotherapists before the rehabilitation program (T0), after 4 weeks (T1) and after 8 weeks (T2) during a 3-hour session at the Hôpital de Jonquière of the Centre intégré universitaire de santé et de services sociaux du Saguenay-Lac-St-Jean (Quebec, Canada). Outcome measures were administered following standard operating procedures (SOP) to ensure the reliability of the results between the two raters and the three assessments. The physiotherapists did not assess the same patient at each measurement time point to decrease recall bias. At T0, a sociodemographic questionnaire was administered to document age, sex, height and weight. The type of walking aid used for indoor mobility was also documented at the beginning (T0) and the end (T2) of the project. The genotype of participants was retrieved from their medical records. Using an open-ended question, participants were asked to indicate the main changes they perceived at the end of the program. The adherence rate is based on the number of sessions executed divided by the total number of sessions planned throughout the study.

Outcomes
Trunk control. Sitting balance and trunk control were assessed with the Ottawa Sitting Scale (OSS) [11]. It includes six items graded from 0 to 4, and each item must be done with the feet supported on the ground and unsupported for a total score of 48.
Balance. The Berg Balance Scale (BBS) was used to assess balance and risk of falling [12]. It includes 14 items graded from 0 to 4, for a maximum score of 56. Its validity has been demonstrated in ARSACS [13]. The participant's level of confidence in their balance was assessed using the modified Activities-specific Balance Confidence (m-ABC) Scale. It contains 16 items related to daily activities, each rated from 0% (no confidence) to 100% (completely confident) [14].
Mobility. Short-distance walking speed was assessed using the 10-Meter Walk Test (10mWT) at the self-selected and maximal speed. One trial was performed and the speed in meter/second was used for analyses. The 10mWT has shown to be valid in the ARSACS population and has excellent interrater reliability (ICC = 0.992) [13]. The ability to perform sit-tostand transfers was assessed using the 30-Second Chair Stand Test (30s-CST). The number of completed sit-to-stands performed without support was noted.
Activities of daily living (ADL). The Barthel Index (BI) [15] was administered and includes 10 items and the maximum score is 100.
Ataxia severity. The severity of cerebellar ataxia was assessed using the Scale for the Assessment and Rating of Ataxia (SARA), which includes eight items: evaluation of gait, stance, sitting, speech, and four tests assessing dynamic limb function for a total score ranging from 0 (no ataxia) to 40 (most severe ataxia). The inter-rater reliability (ICC = 0.98), test-retest reliability (ICC = 0.90) and internal consistency (Cronbach's alpha = 0.94) are all excellent in ataxia populations including ARSACS [16,17].

Sample size
Given the design (exploratory study), the sample size was based on recommendations for pilot studies [18].

Statistical analyses
To be included in the statistical analysis, each participant needed to have an adherence rate of at least 80% (at least 17 sessions out of 24) [19,20]. Descriptive statistics were used: mean, median, SD and ranges for continuous variables, and frequency and percentage for categorical variables. A delta between maximal speed and self-selected speed (Δ 10mWT = maximal speed-self-selected speed) was calculated for T0 and T2. To compare the performance of participants between each assessment (T0, T1 and T2), a Friedman test was used. Post-hoc analyses were done for pairwise comparisons with Bonferroni correction to control for multiple comparisons. Individual change between T2 and T0 was calculated for each outcome measure and compared with the standard error of measurement (SEM), i.e the systematic and random error of a patient's score that is not attributed to true changes in the construct to be measured [21]. When the SEM was not available for the ARSACS population, the ICC found in the literature for another comparable population and the SD from our sample at T0 were used to calculate the SEM of the measure as follows: SD T0 � p (1-ICC) [21]. The following SEM values were used in the analysis: BI (2.7) [22], BBS (3.1) [13], 10mWT (0.026) [13], SARA (1.2) [23], 30s-CST (1.27) [24], OSS (1) [11] and m-ABC (4.7) [25]. As used in other exploratory studies, [26][27][28] p-values <0.1 were a priori selected since they identify clinically relevant for practice improvement, particularly in small studies [29]. Data were analyzed using IBM SPSS Statistics for Mac, Version 20.0 (Armonk, NY: IBM Corp). The open-ended question for participants' perceived changes at the end of the program was coded using thematic analysis.

Results
All participants completed the program and had a mean adherence rate of 95% (83-100%). The mean age was 36.6 years old and 60% were women. Table 1 shows all participant characteristics. A total of three participants walked without a walking aid inside their home at T0, and that number increased to five participants after the program (T2).
For the whole group, significant improvements in balance (BBS) (p = 0.067) and sitting balance (OSS) (p = 0.002) were observed between T0 and T2 and in addition for OSS between T1-T2. The change between T0 and T2 is observed in the delta of the walking speed (p = 0.061). The SARA demonstrated a change for two measurement times (T0-T2 and T1-T2) (p = 0.085). A significant change was present for the BI as well for two measurement times (T0-T1 and T0-T2) (p = 0.069). Table 2 shows the mean results of all outcome measures for T0, T1 and T2.
All participants showed improvement beyond the SEM in at least two outcome measures (see Table 3). The self-selected walking speed is the variable that improved the least during the program, with only two participants having improved outside the SEM. The mean performance of the whole cohort improved outside the SEM for maximal walking speed, trunk control and disease severity. Sitting balance, maximal walking speed, balance confidence and ataxia severity, are the variables for which the greatest number of participants improved outside the SEM (between 5 and 6). Five participants worsened outside the SEM on a total of one to three outcome measures, including disease severity, balance confidence, walking speed and sit-to-stand transfer. Table 4 shows the main improvements perceived by the participants at the end of the 8-week rehabilitation program. The answers were grouped into three categories: 1) Balance and posture, 2) Walking capacity/Functional mobility and 3) Others. Among these, improvement of postural control, change or decrease in the use of walking aids and increase in strength are the outcomes most often reported by participants. The most striking finding is the change in walking aids following the program. They also stated that they perceived an improvement in various activities of daily living that required maintaining balance in a standing position, such as taking a shower or requiring good trunk strength, such as wiping after having a bowel movement. Only one participant reported no perceived changes. Although there were decreases in individual performance on some outcomes assessed at the end of the program, no participant reported deterioration in these abilities as a result of participation in the program.

Discussion
This exploratory study suggests the positive effects of an 8-week rehabilitation program aiming to increase trunk and lower limb motor control on walking abilities, balance and accomplishment of daily living activities in ambulatory people with ARSACS.
Significant improvements in balance, trunk control, ataxia severity and accomplishment of specific activities of daily living were noted at end of the program. In addition to the individual improvements outside of SEM observed for many outcome measures, most participants perceived positive effects and reported mainly changes in balance, trunk control, various activities related to mobility and activities of daily living. A recent study showed that an 8-week training program including physical activities, strength power and aerobic training for people with ARSACS (mean age: 28.1 years) has significant benefits on functional mobility activities such as walking (endurance and speed), transfer from sitting to stand, upper and lower limb muscle strength [6]. The results of this previous study are consistent with our findings.
Participants' results at T0 illustrated the high level of impairments, activity limitations and ataxia severity, which most likely had an impact on the group effects of the rehabilitation program. Only three participants walked without a walking aid inside their home at T0. Two participants used a wheelchair to get around their home, despite being able to walk, due to fear of falling and the need for supervision. The mean age of the participants was 36.6 years, which is close to the age at which walking abilities are lost in this population (38.9 years) [3,5]. A review suggests that interventions aimed at improving walking abilities in hereditary ataxia populations are more likely to be beneficial earlier in the disease course [30]. Even people with the most advanced disease course could benefit from a rehabilitation program with an improvement in functional abilities. The average improvement of the whole cohort, i.e. 1.9 points on the SARA scale, 2.1 on the BBS and 2.5 on the BI, indicates that the participants regained functional capacity performance equivalent to one year or more of the disease progression [10]. The improvements obtained on the SARA scale for ataxia severity were between T0-T2 and T1-T2, which means that they occurred between five and eight weeks after the start of the rehabilitation program. These results are consistent with the review, which suggests that improvements in ataxia minimally occur after four weeks of intervention [30]. All participants were at high risk of falling at T0 with a BBS score lower than 45 [31] and results lower than 69% on the m-ABC-scale (an indicator of recurrent falls in a similar population) [32]. Despite this high level of impairment, the balance measured using the BBS   Table 4. Perceived improvements reported by participants at the end of the rehabilitation program (n = 9).

Improved functions or activities n (%)
Balance and posture significantly improved after the program (T0-T2). Fifty percent of the participants showed an improvement outside the SEM for the balance confidence even if the total group did not reach statistical significance. Change or decrease in the use of walking aids reported by four participants are important findings that also reflect balance improvement. However, despite the observed improvements, all participants remained at high risk of falling at the end of the program as demonstrated by the final BBS score. The most significant improvement following the training program is related to trunk control and seated balance with the OSS. A significant improvement was found between T0-T2 and T1-T2, suggesting that improvements in sitting and standing balance require rehabilitation longer than 4 weeks. The results observed in our study are not in agreement with the review on hereditary ataxia which noted that balance improvement could occur after only three weeks of rehabilitation [30]. At T0, participants had a good static seated balance, but great difficulty maintaining balance when shifting their center of gravity in all directions. A great difficulty in dissociating the trunk movements was also observed in the OSS and this element was greatly improved during the program. Participants perceived positive impacts in carrying out some important activities of their daily life that require the abilities to dissociate the trunk movements, such as wiping after having a bowel movement. The accomplishment of many activities of daily living involves reaching as well as movements of the trunk or upper body to extend the reach.
Decreased trunk control makes it hard to execute sit-to-stand transfers. The performance in mobility activities at each time shows that participants had a lot of difficulties performing the sit-to-stand transfer without using their upper limbs. In fact, their results are similar to those obtained by healthy people over 95 years old [33]. The majority of participants had adaptations in the different rooms of their homes to maintain their autonomy in transfers from sitting to standing. Although the practice of sit-to-stand transfers was part of the program, no significant changes were obtained during the assessment with the 30s-CST test at T1 and T2, and only three participants improved beyond the SEM. This lack of improvement in the number of sit-to-stand transfers achieved in 30 seconds between T0 and T2 could be due to the caution exercised by participants when executing the transfers.
As for the 30s-CST performance, we hypothesize that the lack of significant improvement in the self-selected walking speed was due to the caution exercised by participants, who are very aware of the risk of falling as well as of the importance to have a good walking pattern. Regarding the maximal walking speed, although the change between T0 and T2 is not significant, there is an increase in the maximal average walking speed of the whole group. Additionally, 50% of participants improved their maximal walking speed beyond the SEM. Since the average of the self-selected walking speed did not change, the difference between the maximal and the selfselected walking speed significantly increased between T0 and T2 for the whole group. In other degenerative diseases, such as myotonic dystrophy type 1, this difference between the maximum and self-selected walking speed is a factor that significantly influences the frequency of falls [34]. More precisely, Hammarén et al. [34] showed that an increase in time-difference of 1 s between self-selected and maximal walking speeds, measured over a distance of 10 m, increases the fall risk by 42% in this population. The difference of 1.96 s between the maximal speed and the self-selected speed between T0 (5.98 s) and T2 (7.94 s) supports the importance to develop rehabilitation programs aiming to improve mobility in populations with physical impairments.
Significant improvement in the performance of activities of daily living assessed using the BI was noted between T0-T1 and T0-T2. We hypothesize that the improvements that occurred between T0 and T1 were mainly the result of learning good execution techniques rather than real improvements in functions. As the BI assesses performance in only six basic activities of daily living, it would be interesting for a future study to assess the impact on the performance of a more complete portrait of activities of daily living.
At the end of the training program, deterioration outside the SEM was observed for some participants, mainly in the functional tasks of walking (self-selected [n = 3] and maximal speed [n = 3]) and sit-to-stand transfer (n = 1). People with ARSACS tend to execute their movements very quickly, which impairs the execution quality. In the first weeks of the program, the physiotherapist taught the importance of slowing down the movement speed to increase the execution quality, such as walking and sit-to-stand transfer, to promote good motor learning. It is therefore not surprising that some participants obtained lower scores on tasks that primarily assessed the execution speed, such as the walking speed and the sit-to-stand transfer task. Unfortunately, none of the outcome measures selected were able to validate an improvement in the quality of movement execution in the task. In a future study, it would be interesting to evaluate changes in gait using a layer pressure-sensitive walkway such as the GAITRite.
The most important limitation of this study is the small sample size. To properly document the impact of our rehabilitation program, it would be essential to conduct a study with a larger sample size, including a control group. However, the results of this pilot study help to establish the choice of a primary outcome, as well as the sample size required in future larger studies. In addition, the results of this study and the adherence of the participants support the feasibility and acceptability of rehabilitation programs with this population. As the majority of improvements occurred after the fourth week of intervention, it would be interesting to see the impact of a 12-week program. We also only focused on ambulatory patients and further studies are needed to examine whether people with ARSACS with more severe impairments would also benefit from rehabilitation programs. We also did not evaluate the persistence of gains from this intervention by a follow-up assessment.
The study's findings provide encouraging results on the effects of a rehabilitation program for ambulatory people with ARSACS. The main findings of this study are the significant improvements in trunk control, balance, and the change in walking aids needed for ambulation. Group rehabilitation programs could have a positive impact on the health care services offered to individuals with ARSACS.