Abstract
Background:
Although adult and late-onset DM1 phenotypes DM1 present distinct lower limb weaknesses portraits, resulting physical limitations have never been described separately for each phenotype.
Objective:
To characterize the lower limb weaknesses and physical limitations among the DM1 adult and late-onset phenotypes separately and to document the contribution of weaknesses on mobility to optimize the management of this population.
Methods:
The strength of four muscle groups among 198 participants was quantified. Participants were categorized according to the severity of their muscular involvement using the Muscular Impairment Rating Scale (MIRS). Physical limitations were assessed using the Timed up-and-go (TUG), Berg Balance Scale (BBS) and 10 meters comfortable walking speed (10MWT). Multiple linear regressions were performed to identify the contribution of each muscle group to the mobility tests scores.
Results:
Late-onset demonstrated less weakness and physical limitations (p < 0.001 – 0.002) than the adult phenotype, but 21.9–47.5% of participants with this phenotype showed mobility scores below reference values. Physical limitations were observed in the first two MIRS grades (37.5–42.1% of the participants) for the TUG and 10MWT. Ankle dorsiflexors and knee extensors were the two muscle groups that showed the strongest relationships with mobility scores.
Conclusion:
Although less impaired, the late-onset phenotype shows significant lower limb muscle weakness associated with physical limitations. The surprising presence of quantitative lower limb muscle weakness in the first two MIRS grades needs to be considered when using this scale. Both ankle dorsiflexors and knee extensors appear to be good indicators of physical limitations in DM1.
INTRODUCTION
Myotonic Dystrophy Type 1 (DM1) is a multisystemic neuromuscular disease mainly characterised by muscle weakness and myotonia [1]. In the adult phenotype, symptoms generally appear in the second or third decade of life [2] while for the late-onset phenotype an older age at onset (>40 years) [3] and usually a less severe muscle weakness are observed [4]. Muscular impairment gravity is frequently assessed using the Muscular Impairment Rating Scale (MIRS) [5].
Muscle weakness leads to physical limitations [6–19] at one time or another in the disease course. Some lower limb muscle groups [6, 10] have also been previously described individually as having a significant impact on physical limitations in this population.
However, although muscle weakness has been reported to be different in DM1 adult and late-onset phenotypes [20], and although DM1 present functional limitations related to premature ageing [21] no distinct lower limb functional profile according to each phenotype has yet been identified. The use of known outcomes with sound metrological properties such as the TUG and BBS would allow characterizing other impairments in regard to mobility and balance in both phenotypes and permit to compare the results with other populations. And indeed, recent findings have shown that these two outcomes, the TUG and BBS, are showing promising responsiveness with DM1 patients [22].
In addition, the MIRS is largely used in clinic and in research, but very few studies have described the functional portrait associated to each grade [5]. Therefore, prognostic and specific interventions as well as prevention targets are difficult to establish. Furthermore, the respective impact of each lower limb muscle group weakness on functional mobility has yet to be documented. Indeed, considering upcoming therapeutic trials in DM1, the identification of impairments of specific muscle groups that would be involved in significant physical limitations is of major concern.
The objectives of this study were to describe and compare, between adult and late-onset phenotypes, lower limb muscle strength impairments and physical limitations, and how they arise according to the Muscular Impairment Rating Scale (MIRS) grades, and compare to reference values. This study also aimed at estimating the extent to which the strength of key lower limb muscle groups can explain the performance to the Timed up-and-go (TUG), Berg Balance Scale (BBS), and 10 meter comfortable walking speed (10MWT).
METHODS
Participants
This study was part of a larger research project, which aimed at identifying determinants of social participation and quality of life in DM1 (Mathieu et al., MOP-49556). A randomized sample was drawn from the Neuromuscular Clinic registry of the Centre Intégré Universitaire de Santé et de Services Sociaux (CIUSSS) du Saguenay–Lac-St-Jean, Québec, Canada. To be included in the study, participants had to be over 18 years old, have a diagnosis of DM1 (adult or late-onset phenotype) confirmed by molecular testing, and provide informed consent. Participants with the late-onset phenotype were described as presenting with 2 of the 3 following criteria: CTG repeats under 200; MIRS score of 1 or 2, and; onset of symptoms after 40 years of age. All other potential participants were classified as having the adult phenotype. Since predictive DNA testing is a routine component in the assessment of our population, this 2 of 3 criteria categorization was used to include in the late-onset phenotype the young (<40 years of age) asymptomatic patients with a small CTG expansion. This is in accordance with the clinical presentation of De Antonio et al. [23]. CTG count was calculated using a blood sample collected the day of assessment [24]. People with congenital or infantile phenotype and those presenting with other diseases that might lead to permanent physical limitations were excluded. The Ethics Review Board of the CIUSSS du Saguenay–Lac-St-Jean approved the study (2001-032). The authors have no conflict of interest to report.
Testing procedures
The same physiotherapist performed the assessment of all participants. For the same participant, all assessments were completed during the same session using a predetermined order of tests: MIRS scale, quantitative muscle testing (QMT), TUG, BBS, and 10MWT. The age, sex and body mass index were recorded for each participant.
Muscle strength
The participant’s muscular impairment level was assessed using the MIRS (Table e-1) [5]. The maximum isometric muscle strength of the hip flexors, knee extensors and flexors and ankle dorsiflexors, for both sides, was assessed using a make test with a handheld dynamometer (Microfet – 2, Hoggan Health Industries, Salt Lake City, UT). A detailed description of the methodology used was described in a previous paper [20]. Hip flexors (hip and knee at 90°) and ankle dorsiflexors (hip and knee at 0°) were assessed supine. Knee flexors and extensors were assessed seated (hip and knee at 90°). For each muscle group, 2 trials of maximal isometric contractions were completed, and the mean of the 2 trials was used for the analysis. When more than a 10% difference was noted between the peak forces of the first 2 trials, a third one was then performed, and the mean of the 2 closest peak values was used for the final analysis.
Global walking abilities, balance and walking speed
Basic mobility and balance were assessed using the TUG [25] following a standardized procedure. This test measures the time needed to stand up from a standard armchair, walk three meters at comfortable and safe pace, turn, walk back to the chair and sit down. Three trials were performed and the mean was calculated. In DM1, the TUG has shown a good intrarater reliability (ICC: 0.83) [26].
The BBS [27] was used to assess static and dynamic balance. This scale contains 14 items, rated from 0 (failure) to 4 (success), for a maximum score of 56, and it includes simple and more complex tasks. The BBS is a valid balance scale for older adults [27] and it allows to discriminate fallers from non-fallers [28]. The BBS was added in the course of the study after reviewing the protocol.
Comfortable walking speed was assessed on a flat 10 meters surface, following a standardized procedure [29]. Three trials were performed and the mean was calculated. The 10MWT shows excellent intrarater reliability in the DM1 population (ICC: 0.94) [6].
Those three mobility tests were performed in a calm environment in a quiet space free from visual and auditory distractions. No physical assistance or verbal stimulation was given. For the 10MWT and the TUG, the participants wore their regular footwear and were allowed to use their usual walking aid and orthotics if appropriate. The BBS was performed bare feet and without any walking aids.
Data analysis
Normality of the data was verified with Kolmogorov-Smirnov test and the Levene’s test was used to verify the assumption of equality of variances between groups. Statistical analyses were performed using IBM SPSS Statistics for Windows version 20.0 (IBM Corp, Armonk, New York).
Descriptive statistics of maximal isometric muscle strength and mobility scores (median, minimum-maximum, mean, standard deviation and 99% confidence interval, depending on the data distribution normality) were calculated, and t-tests (or Mann Whitney U-tests for non-symmetric data distribution) were used to compare adult and late-onset phenotypes, male and female, and sides for the strength measures.
The number of participants in each MIRS grade was calculated. Spearman correlation coefficient analysis was performed to estimate the degree of relationship between muscle strength measures, mobility scores, and the MIRS grades.
DM1 mobility scores were compared to reference values (described by age groups for the TUG [30] and for each sex and by age groups for the BBS [31] and the 10MWT [32]), to cut off scores associated to an increased risk of falling (TUG [33], BBS [34]) and to the speed needed to safely cross a street by walking [35]. The reference values used to compare the TUG [30] and BBS [31] scores were developed with people aged of 60 and over. Proportions of participants with mobility scores below the reference values and cut off scores were calculated. We used t-test to compare the strength results between participants with mobility scores below and above reference values for the three mobility tests.
Linear regressions models, using the enter method, were built to estimate the influence of the maximal isometric strength of each muscle group (independent variables) on the scores of each mobility test (dependant variables). To determine which variables (independent variables and co-founding factors) should be included in the regression models, Pearson (or Spearman for asymmetric distribution or categorical variables) correlation coefficients were used to estimate the strength of the relationships between dependent (mobility scores) and independent variables (QMT scores). Only QMT scores and co-founding factors presenting with a significant correlation (p < 0.05) were included in the final multiple regression analysis models (using sequential stepwise regression). The regression models were adjusted for the age, sex and body mass index (BMI), as these variables are considered to be confounding factors that may influence variations in the mobility scores [30–32, 36]. DM1 presenting a large spectrum of levels of impairments, each regression was performed with a bootstrap to consider the presence of outliers due to the disease presentation. Bootstrap also allows inclusion of variables presenting a non-normal distribution in the linear regressions without prior conversion. Multiple correlation coefficients, adjusted multiple correlation coefficients, beta standardized coefficients and their 95% confidence interval, and p-values were reported. Considering the number of p calculated for the linear regressions, the level of significance was set at p <0.017. Kaiser-Meyer-Olkin test for sampling (KMO) was used to determine the representational quality of our sample.
RESULTS
Forty participants with a late-onset phenotype and 158 with the adult phenotype completed the assessments (Fig. e-1). Two people were excluded due to medical conditions that would have influenced their performance to the tests at the time of the study (recent ankle fracture and a largely uncontrolled diabetes). Participants’ characteristics are presented in Table e-2. Participants with a late-onset phenotype were older than the participants with the adult phenotype. No significant difference in male-female proportion between the two phenotypes was observed (late-onset: 24 women and 16 men; adult: 96 women, 62 men).
Muscle strength
Significant lower strength values were found for the adult phenotype compared to the late-onset one for all lower limb muscle groups (Table 1). A gender effect, with men being stronger than women (p≤0.001), was observed in all muscle groups, except for the ankle dorsiflexors (p = 0.69) (data not shown). Differences in favour of the dominant side were observed for hip flexors, knee flexors and ankle dorsiflexors, differences that were all of very small magnitudes of 7.2, 4.7 and 3.0 N, respectively (p = <0.001 – 0.04) (date not shown). A general significant increase of muscle weakness was observed from MIRS grade 1 to 5 for all muscle groups with significant correlations (Table 2).
Lower limb muscle strength (N) and comparison between phenotypes
Lower limb muscle strength (Newtons) relationship with MIRS grade
*Spearman coefficient significant at 0.01.
Global walking abilities, balance and walking speed
Participants showed significantly lower scores on the three mobility tests for the adult phenotype compared to the late-onset phenotype (Table 3). Both phenotypes included a significant proportion of individuals performing below reference values for TUG, BBS and the 10MWT (21.9 – 47.5% for late-onset and 58.5 – 84.7% for adult phenotype) (Table 3). There was no significant difference between male and female participants (p = 0.205 – 0.532) (data not shown).
Mobility tests performance and corresponding percentages of participants with scores below reference values according to each phenotype
*Percentages of participants with scores below reference values. **Variables presenting with a non-normal distribution. TUG: Timed up & Go; BBS: Berg Balance Scale; 10MWT: 10 Meter Walk Test.
When comparing scores obtained to the mobility tests according to each MIRS grade, a general significant progression of physical limitations was observed from grade 1 to 5 with moderate correlations (Table 4). However, between 31.6 and 42.1% of participants showed scores below reference values for the TUG and the 10MWT in the first (no muscular impairment at manual muscle testing (MMT)) and second (minimal signs of muscle impairment and isolated digit flexors weakness at MMT) MIRS grades. Concerning the BBS, 41% of the whole sample had the maximum score (63% for late-onset participants, 36% for adult participants). From the MIRS grade 4 (mild to moderate proximal weakness at MMT), there was an important increase in the proportion of participants with a TUG score greater than 10 seconds (from 25.6% at grade 3 to 64.0% at grade 4) and a BBS score lower than 45 (from 0% at grade 3 to 17.3% at grade 4), cut offs associated to an increased risk of falling [33, 34] (data not shown). There was also an increase in participants proportion with a comfortable walking speed lower than 1.0 m/s, which is the minimal speed considered required to safely cross a street in Canada [35], passing from 10.3% at grade 3 to 46.7% at grade 4 (data not shown).
Mobility tests performance and corresponding percentages of participants with scores below reference values according to each MIRS grade
*Variables presenting with a non-normal distribution. †Spearman coefficient significant at 0.01. ††Percentages of participants with scores below reference values. TUG: Timed up & Go; BBS: Berg Balance Scale; 10MWT: 10 Meter Walk Test.
Relationships between lower limb muscle strength impairments and physical limitations
The mean of the maximal muscle strength of all tested muscle groups was significantly different between participants with mobility scores below and above reference values (p < 0.001) for all mobility tests performed (data not shown).
Significant correlations were found between strength results of the four muscle groups and the mobility tests (Table e-3). The four muscle strength variables were then entered as independent variables in each of the regression models. Our sample presented an adequate representational quality (KMO of the independent variable matrix = 0.737 (p < 0.001), KMO of the independent and dependant variables matrix = 0.803 (p < 0.001)). The models explained 42 to 62% of the variance observed in the mobility scores (Table 5). For the TUG, the knee extensors muscle group was the independant variable that explains the highest percentage of variations observed, although the knee flexors and the ankle dorsiflexors also had a high significant beta coefficient and the opposite was observed for the BBS. With regards to the 10MWT, the ankle dorsiflexors proved to be the independant variable that explains the highest percentage of variations observed, although the knee extensors and flexors also had a significant beta coefficient.
Regression model data quantifying of the relative influence of the maximal isometric strength on the mobility test scores
TUG: Timed up & Go; BBS: Berg Balance Scale; 10MWT: 10 Meter Walk Test.
DISCUSSION
Muscle strength impairments and physical limitations according to phenotypes
This is the first study to explore the causal link between lower limb muscle strength impairments and physical limitations from such a large cohort of participants with adult and late-onset DM1 phenotypes separately, moreover using quantitative muscle strength values. Not surprisingly, both strength impairments and physical limitations are more important in the adult phenotype. However, we are the first to show that even among the late-onset phenotype, significant physical limitations were also observed (21.9 – 47.5% of participants with mobility scores below reference values for the TUG and the 10MWT, except for BBS), although this DM1 phenotype is usually described as being fairly physically functional [4]. Moreover, the comparisons of the scores of our participants with reference values developed for people aged of 60 and over for the TUG [30] and the BBS [31] instead of matching controls may have underestimated the proportion of participants presenting with physical limitations. Our results suggest then that people with late-onset phenotype may have significant muscle strength impairments and physical limitations. As regular physical activity is an important factor for general health associated with better quality of life [37], and as moderate intensity strength training have been demonstrated to be safe in DM1 [38], and as preliminary results suggest that DM1 skeletal muscles show positive molecular response to training [39], strengthening of late–onset phenotype could be considered early on in their rehabilitation to prevent strength loss and physical limitations.
Although there is a great variability in the profile of physical limitations previously reported [6–12, 17], our cohort, when compared to previous studies, seems to have stronger lower limb muscles and higher functional mobility overall. However, the interpretation of these comparisons should be made with caution as the majority of studies do not adequately describe their cohort and their protocol, and multiple phenotypes are often included and analysed all together. In our study, as clearly exposed in Tables 1 and 3, pooling the strength impairments and physical limitations results of both phenotypes would have led to misinterpretations and wrong conclusions for each phenotype. Adult and late-onset phenotypes present with two distinct profiles of physical limitations and should be separately presented and analyzed, especially in descriptive studies.
Muscle strength impairments and physical limitations according to the MIRS
Interestingly, even for the two first MIRS grades that typically refer to individuals with DM1 being fairly functional, significant quantitative lower limb muscle impairments and physical limitations were also found in 31.6 to 42.1% of our participants classified in those two grades, except for the BBS. The weaknesses observed using QMT in that significant proportion of our participants, in conjunction with the many others multisystem impairments present in DM1, seems to have lead to physical limitations. These results highlight the greater sensitivity of QMT over MMT. The MIRS was originally designed to be used both as a research and clinical tool to provide data that would be quickly and readily available on the body regions impaired from a muscular perspective in patients with DM1. The MIRS has also been widely used internationally to establish relationships between muscular impairments and impairments of other body functions and structures, notably of the cognitive, cardiac, endocrine, and respiratory systems. The MIRS was shown to be reliable, but as it is rated using the MMT scale, it unfortunately lacks sensitivity, particularly when assessing strong muscle groups.
As with the late-onset phenotype, in clinical follow-up, recommendation of maintaining an adequate physical activity level needs to be addressed as soon as the first MIRS grades. Furthermore, in our study, considering the important increase of risk of falling in participants from MIRS grade 4-5 according to their Berg scores, in addition for them to presenting a walking speed that indicate physical limitations for safely crossing a street, the implementation of measures to ensure that the required interventions (e.g. further investigation of risk of falling) are offered to them at this point would be beneficial to optimize their functional mobility.
The scores obtained to the BBS by the DM1 participants of our cohort are not consistent with the issue of falls typically reported in DM1 [7, 12]. As the BBS was designed to assess central balance, the BBS tasks may not be sufficiently challenging to highlight the spectrum of balance impairments in the DM1 population. Our cohort may also present a larger proportion of less impaired participants than in the other studies.
Relationships between lower limb muscle strength impairments and physical limitations
Significant weaknesses were observed proximally and distally in our participants [20]. However, some muscle groups better explain the variations observed in scores of each mobility test. The variations observed in the TUG score are mainly explained by the knee extensor strength, and in a lesser proportion by the knee flexor and ankle dorsiflexor strength. It is not surprising to observe a predominant influence of the knee extensors on the TUG score considering their main action of adjusting the distance of the trunk relative to the floor [40], a movement that takes place during the sit-to-stand transfer, which is a dominant component of the TUG. Otherwise, in accordance with previous publications in DM1 [6, 41], the ankle dorsiflexors muscle strength is the most important variable that explains the highest percentage of variance observed in both balance score and walking speed. Ankle dorsiflexors are a very active muscle group during the gait cycle [42], and this muscle group plays a major role in the avoidance of foot drop and therefore helps to avoid falls due to tripping. Moreover, the ankle movement strategies are the most commonly used in situations in which the balance perturbation is small and the surface is firm [43]. Therefore, it is consistent to observe that weakness in ankle dorsiflexors is more contributory to the scores of mobility tests such as the ones performed in this study.
The regression models proposed explain about 50% of the variance of mobility scores, suggesting that other clinical determinants are also implicated in the physical limitations observed in DM1. In addition to the lower limb strength, numerous factors such as reaction time, vision, pain, health status and cognitive function have been reported to influence functional testing such as the TUG performance [44–46]. Although TUG is an easy test to perform, the interpretation of the results may be difficult. In DM1, the multisystemic symptoms such as fatigue, apathy, attention deficit, daytime sleepiness, myotonia, cataracts, executive dysfunctions and visuo-spatial disabilities may all influence the functional performance [47, 48]. In the next future, regression models including all these factors should be used to determine the real contribution of lower limb strength in the functional impairments of patients with MIRS grade 1 and 2 or patients with a late-onset phenotype. Also, the inclusion of other muscle groups in the regression models, such as plantar flexors (for the walking performances) and hip stabilizers (for proximal stability, static and dynamic balance, and transfer), could have contributed to explain a larger proportion of the variance in the mobility tests.
This study is providing some of the answers with regards to the potential influence of lower limb muscle weakness on physical limitations in adult and late-onset phenotypes of DM1. In light of future clinical trials, the present study highlights the fact that pooling patients as well as phenotypes in DM1 may not allow to clearly demonstrate muscle impairments changes overtime because of the large between subjects’ variability. Separating phenotypes seems then to be a more suitable way to assess the efficacy of a therapy. Longitudinal data will be helpful to assess the effect size and allow determining the optimal population characteristics for clinical trials.
The current findings are a first step towards a better understanding of the physical limitations profile in both adult and late-onset phenotypes. In DM1, a better knowledge of the all the factors that may influence physical limitations profile, and of the key muscle groups impacting on physical capacities, would permit a more proper timing of interventions in the clinical follow-up of all phenotypes. More researchs are required to better manage the DM1 patients and implement sound interventions and tailored therapeutic trial.
Footnotes
ACKNOWLEDGMENTS
The authors would like to thank the participants and their families. The corresponding author received bursaries from the Corporation de recherche et d’action sur les maladies héréditaires, the Faculty of Medicine and Health Sciences of the Université de Sherbrooke, the Go Foundation, and the Ordre professionnel de la physiothérapie du Québec, as well as the financial, professional and technical support of the Groupe de recherche interdisciplinaire sur les maladies neuromusculaires. Cynthia Gagnon holds a FRQ-S career grant (Chercheur-Junior 2). Luc J. Hébert has been the chair of the Muscle Testing and Training Special Interest Group of the past two Outcome Measures in Myotonic Dystrophy Type 1 sessions.
Description of each grade of the Muscular Impairment Rating Scale, demographic data, correlations between lower limb muscle strength and mobility scores and the flow diagram of the recruitment strategy are available online as supplementary material.
