You are on page 1of 12

NeuroRehabilitation 42 (2018) 131–142 131

DOI:10.3233/NRE-172238
IOS Press

How does strength training and balance


training affect gait and fatigue in patients
with Multiple Sclerosis? A study protocol
of a randomized controlled trial
Jacob Callesena,b,∗ , Davide Cattaneoc , John Brincksa and Ulrik Dalgasb
a Department of Physiotherapy, VIA University College, Faculty of Health Science, Aarhus, Denmark
b Department of Public Health, Aarhus University, Section of Sport Science, Aarhus, Denmark
c Larice Lab, Don Gnocchi Foundation, Gait and Balance Rehabilitation Lab, Milan, Italy

Abstract.
INTRODUCTION: Multiple sclerosis (MS) is characterized by a demyelination that results in reduced conductivity in the
somatosensory nervous system, decreased muscle strength, vestibular alteration, and severe fatigue. Progressive resistance
training (PRT) has proven to be a promising intervention showing a positive effect on muscle strength. Another promising
intervention frequently used in neuro-rehabilitation is task specific training where also Balance and Motor Control Training
(BMCT) are incorporated. Interestingly, the principles of BMCT do fundamentally contrast the principles of PRT in terms of
variation in movement pattern, loading and repetitions. Consequently, knowledge of any diverse effect would be of clinical
relevance.
AIM: To evaluate the effects of PRT and BMCT on gait, balance and fatigue in persons with MS.
METHOD: A three-armed multi-center, single-blinded cluster randomized controlled trial with two intervention groups
(1. PRT of the lower extremities. 2. BMCT that challenges gait function) and a control group that receives usual care
while on a waitlist for a combined PRT + BMCT intervention performed after the two interventions groups have completed
their interventions. The interventions last ten weeks with two sessions per week, in groups of 3–6 participants. Number of
participants is 30 per intervention – 90 in total. Primary outcome measures for gait function are the Timed 25 Foot Walk
(T25FW) and the Six Spot Step Test (SSST). Secondary outcomes are fatigue, perceived gait function, temporo-spatial gait
characteristics, balance and strength.
Inclusion criteria are: EDSS 2–6, SSST >8 sec and T25FW >5 sec. Exclusion: Recent attacks and ongoing intensive
rehabilitation.
ANALYSIS: The effects in the three groups are examined in a mixed effects regression analysis with group and time as fixed
effects and center and patient within center as random effects. Spearman or Pearson correlation analysis will be conducted
on baseline data to determine associations between the primary outcomes on gait function and the secondary outcomes on
fatigue, spatial gait parameters, balance and patient reported measures.
TRIAL REGISTRATION: The study is approved by the Regional ethical committee and registered at clinicaltrials.gov,
NCT02870023.

Keywords: Resistance training, balance training, multiple sclerosis, physiotherapy, gait

∗ Address for correspondence: Jacob Callesen, Department of ence, Hedeager 2, 8200 Aarhus N, Denmark. Tel.: +45 87552330;
Physiotherapy, VIA University College, Faculty of Health Sci- E-mail: jacc@via.dk.

1053-8135/18/$35.00 © 2018 – IOS Press and the authors. All rights reserved
132 J. Callesen et al. / How does strength training and balance training affect gait and fatigue in MS?

1. Background physiotherapists in neurorehabilitation (Carr & Shep-


herd, 2010). In this study protocol, motor function is
Multiple sclerosis (MS) is characterized by limited to gait related functions with a particular focus
demyelination of the sensory and motor axons in the put on balance and motor control, why the term Bal-
central nervous system. MS mostly affects women ance and Motor Control Training (BMCT) is applied.
(women:men ratio: 2–4:1) in the second or third There is no universally accepted definition of human
decade of life (Cameron & Wagner, 2011; Nosewor- balance, but balance defined as “the inherent ability
thy, Lucchinetti, Rodriguez, & Weinshenker, 2000), of a person to maintain, achieve or restore a spe-
and approximately 2.3 million people have the dis- cific state of balance and not to fall, with reference
ease world wide (Multiple Sclerosis International to the motor and sensory systems and to the physical
Federation, 2014) The MS disease course is either properties of the person” (Pollock, Durward, Rowe,
characterized by stable phases interrupted by periods & Paul, 2000), is applied in this study.
of disease activity (relapsing-remitting MS), or by Effects obtained from BMCT partly result from
a continuous progressive development (primary and plastic changes in the nervous system (Khan, Amatya,
secondary progressive MS). Galea, Gonzenbach, & Kesselring, 2016). To induce
People with MS experience a wide variety of such effects, repetition of a simple task only has
symptoms including impaired muscle strength and limited efficiency in order to improve performance.
balance, fatigue, impaired cognition, depression and Once a task has been learned to a certain level,
spasticity (Cameron & Wagner, 2011). Of these, further practice of the same task will not be accom-
impaired balance (Boes et al., 2012) and severe panied by further induction of plasticity and little is
fatigue (Claros-Salinas et al., 2013) are described therefore gained by continued practice of the task
as two of the most debilitating symptoms leading (Nielsen, Willerslev-Olsen, Christiansen, Lundbye-
to limitations in activities such as upright posture Jensen, & Lorentzen, 2015). To provide challenges
and gait (Heesen et al., 2008; Penner et al., 2007). that ensures continued learning, training exercises
Generally, pharmacological symptomatic treatment have to progress from simple movement trajectories
has not proven efficient in the treatment of balance to more complex movements, that also incorpo-
problems, fatigue and walking impairments, with the rates goal setting (Nielsen et al., 2015). Moreover, it
exception that Fampridine has beneficial effects on has been shown that shaping and variation of tasks
gait performance in a subgroup of patients (Lee, in combination with feedback on movement qual-
Newell, Ziegler, & Topping, 2008). Consequently, ity is of great importance for the learning outcome
non-pharmacological interventions that effectively (Homberg, 2013). The underlying concept for per-
target these symptoms are warranted. forming BMCT is, therefore, that improved motor
In the last decade progressive resistance training control will optimize the movement strategy, which
(PRT) has proven to be one of the promising interven- further leads to improved gait function (Carling, Fors-
tions in patients with MS showing a consistent and berg, Gunnarsson, & Nilsagård, 2016).
positive effect on muscle strength (Kjolhede, Viss- Regarding the effects of BMCT on fatigue, there
ing, & Dalgas, 2012). However, the effect of PRT are diverging results in the literature, but the literature
on functional outcomes are heterogeneous but with on BMCT for patients with MS is generally of low
promising effects on daily activities such as walking quality with an inadequate description of interven-
and chair rise (Kjolhede et al., 2012; Latimer-Cheung tions (Paltamaa, Sjogren, Peurala, & Heinonen, 2012;
et al., 2013; Motl & Pilutti, 2012). The evidence for Rietberg, Brooks, Uitdehaag, & Kwakkel, 2005),
a beneficial effect of PRT on balance and postural why further studies are warranted.
control is divergent and yet inadequately investi- Interestingly, the principles of task specific training
gated (Kjolhede et al., 2012). Regarding fatigue, do fundamentally contrast the principles of PRT, that
a recent Cochrane review reported that one could normally consist of monotonous movement patterns
expect improvements in MS fatigue after exercise performed under heavy loading for a low number of
interventions, despite methodological flaws in the repetitions (American College of Sports Medicine,
existing literature, but only few studies evaluating 2009). Consequently, studies comparing the effects of
PRT were located (Heine, van de Port, Rietberg, van BMCT and PRT on gait function would add to the cur-
Wegen, & Kwakkel, 2015). rent literature as no studies doing so could be located.
Another promising intervention is task specific Such a comparison would help clarify whether
training of motor function that is widely used by potential effects are overlapping or differentiated
J. Callesen et al. / How does strength training and balance training affect gait and fatigue in MS? 133

and would help guiding future rehabilitation inter- combining PRT and BMCT. Both interventions last
ventions in persons with MS. ten weeks with two sessions per week.
The primary objective of this study is, therefore, The study protocol conforms to the SPIRIT state-
to investigate and compare the effects of 10 weeks of ment and is registered at www.clinicaltrials.gov
PRT to BMCT on gait function, balance and fatigue (NCT02870023). The study will be performed in
in mobility limited persons with MS. agreement with the Declaration of Helsinki and
It is hypothesized that PRT will be superior in is approved by the Regional ethical committee
improving maximal straight gait speed, whereas (Videnskabsetisk Komite, Region Midt; ID-number:
BMCT will have a greater impact on balance, fatigue, 1-10-72-316-15).
and more complex walking tasks that include ele-
ments of balance and coordination.
2.2. Recruitment and eligibility

Patients will be invited via regional MS Clinics


2. Methods/design in the western part of Denmark. Invitations will be
provided at consultations at the clinics or will be sent
2.1. Study design to patients from the registers at the participating MS
clinics. Invitations will also be distributed via social
As depicted in Fig. 1, the study is a three-armed media and network groups established via the Danish
multi-center, single-blinded cluster randomized con- MS Society.
trolled trial with two intervention groups (1: PRT; Patients that reply to an invitation will after-
2: BMCT) and a control group that receives usual wards receive a detailed written information along
care while on a waitlist followed by an intervention with a leaflet from the National ethical committee.

Fig. 1. Illustration of study design.


134 J. Callesen et al. / How does strength training and balance training affect gait and fatigue in MS?

Eligibility according to inclusion criteria 1–4 (see remain blinded to patient exposure and thus serve as
below) are determined by a neurologist from either blinded assessor. Patients will be informed about the
medical records or from a consultation in case of purpose and the procedures regarding blinding. Iden-
insufficient medical record information. tities will be concealed from the primary investigator
Inclusion criteria are: 1) Expanded Disability Sta- until data collection is complete.
tus Scale (EDSS) (Kurtzke, 1983) 2.0–6.0 (≥2.0 in
the functional system “pyramidal function”); 2) Age
>18; 3) Able to walk 100 m; 4) Able to manage own 3. Study flow
transportation to weekly training sessions as well as
to pre- and post-testing; 5) Six Spot Step Test >8 sec; Patients that fulfil the inclusion criteria 1–4 will be
and 6) Timed 25 Foot walk >5 sec. baseline tested by the primary investigator at Section
Exclusion criteria are: 1) Co-morbidity in terms of Sport Science, Dep. Public Health, Aarhus Uni-
of cognitive disorders or alcohol abuse; 2) Attack versity. Final allocation to an intervention depends
within the last eight weeks; 3) Systematic intensive on the result of the baseline scores for the two pri-
rehabilitation/training within the last three months; mary outcome measures (SSST & T25FW). If the
4) Adjustment in medication within 2 months before scores fall below the predefined limit mentioned in
inclusion (applies for medication that has a disease inclusion criteria 5 or 6, patients are not enrolled in
modifying effect and/or affects gait performance and the study. Results from all baseline testing (includ-
spasticity). ing those who are not enrolled in the study) will
Interventions will be carried out at physiotherapy be recorded and used in a secondary cross sectional
clinics located in the community. For a clinic to par- analysis evaluating baseline associations.
ticipate it is required that: 1) the responsible staff in When groups are formed and assigned to an inter-
charge of the interventions are experienced in MS vention, 10 weeks of either BMCT, PRT or Control
rehabilitation; 2) the clinic has a gym facility that intervention will begin. Hereafter a follow-up test
enables 10 meter of straight walking as well as con- identical to the baseline test procedure will be carried
tain sufficient balance accessories and equipment for out. After the intervention period, the control group
PRT of the lower extremities. start their training consisting of combined PRT and
BMCT for 10 weeks followed by a third identical test
2.3. Random allocation procedures round (Fig. 1).

Group allocation will be determined by cluster 3.1. Intervention procedures


randomization with a concealed allocation. The par-
ticipating clinics are randomly assigned to perform 3.1.1. PRT
either PRT, BMCT or Control (and later PRT + Training sessions starts with a 10-minute warm-
BMCT). To ensure an even distribution of patients up on a stationary bicycle or treadmill. All exercises
in the intervention groups, allocation is carried out will be performed in machines, sitting, lying or
whenever three groups of 3–6 patients are enrolled. standing adequately supported (Table 1). Varia-
Three concealed envelopes, each containing the name tion in exercise equipment in terms of brand and
of one intervention, is randomly drawn and allocated types of machines across the different participating
to a group/clinic. As the patients will be assigned physiotherapy clinics is acceptable as long as the
to a clinic located close to their home, the alloca- execution of exercises complies with the description
tion depends on their residential location. A person (Table 1). Training is supervised by a physiothera-
with no further involvement in the study manages the pist to ensure high quality during execution and that
randomized allocation. the intended progressive overload model is followed
An assistant in charge of the test logistics will (Table 2).
assign patients with an ID number and pseudonyms
that will be used during baseline- and follow- 3.1.2. BMCT
up testing. This procedure is carried out to avoid Each training session starts with a 10-minute
that the primary investigator recognizes patients by warm-up on a stationary bicycle or treadmill.
their names, which are used during the recruit- The BMCT intervention consists of stations with
ment procedure, where the primary investigator is exercises that challenges functions related to mobil-
involved. This way the primary investigator can ity. The stations aim at standing, stepping, walking,
J. Callesen et al. / How does strength training and balance training affect gait and fatigue in MS? 135

Table 1
Description of exercises in the Progressive Resistance Training intervention
Exercises Start position End position
Leg press Sitting or supine
– using both legs – Dorsal flexion in ankle – Plantar flexion
– Flexion in knee and hip – Extension in knee and hip
Knee extension Sitting
– Mid-range knee flexion – Full extension
Hip flexion Standing
– Slight hip extension – Full flexion
Hamstring curl Prone
– Slight knee flexion – Full flexion
Hip extension Standing
– Mid-range hip flexion – Full extension

Table 2
Progression model for the PRT intervention
Week Sets Repetitions Load in RM
1&2 3 10 15
3&4 3 12 12
5&6 4 12 12 Fig. 2. Error rate scale with intensities.
7&8 4 10 10
9 & 10 4 8 8
RM = Repetitions Maximum.
When supervising an exercise, an adequate level of
challenge for a simple exercise is determined. After
sit to stand, and vestibular rehabilitation through gaze several successful executions of the exercise, cog-
control. A detailed framework of the balance inter- nitive and/or motor tasks are added when possible
vention is illustrated in appendix 1. – in this study defined as “dual task”. Manipula-
Balance skill is interpreted from the ability to per- tion of sensory conditions by elimination of vision
form the exercises while keeping center of mass and performing exercises on a foam will also be
(COM) in control within the base of support (BOS). applied. By adding motor or cognitive tasks, per-
Multiple failed attempts are interpreted as “an inad- formance quality declines, but it should not exceed
equate balance for the given task”. Reduced speed, an error rate of 40%. In case of a higher error rate,
abnormal fluctuating displacement of COM, exces- the motor challenge is reduced, to lower the error
sive corrective upper limb movement and a high rate.
level of concentration with cessation of talking and Due to differences in patients’ level of function,
head movement is interpreted as “balance being symptoms, and training experience, the perceived
challenged by the given task”. The exercises are challenges from the same exercise will differ between
progressed by altering BOS, loads and speed dur- subjects. Hence, the BMCT framework has been
ing movements, body and/or head turns, and by developed in an attempt to standardize the training
using external tasks (e.g. catching and throwing). protocol in a way that allows the therapist to tailor
The BMCT framework is defined under inspiration and vary the exercises with respect to the patient’s
from already published treatment programs by Coote individual functional level and needs, while ensuring
et al. (Coote, Garrett, Hogan, Larkin, & Saunders, that all patients receive exercises in the same domains
2009), Hebert et al. (Hebert, Corboy, Manago, & and at a similar relative intensity.
Schenkman, 2011) and Nilsagard et al. (Nilsagard,
von Koch, Nilsson, & Forsberg, 2014). 3.2. Control group
An Error Rate is used to describe intensity of
an exercise and thereby the level of the challenge The control group is on a waiting list during the
(Fig. 2). The therapist adds and adjust tasks aiming first 10 weeks where they are encouraged to maintain
at an adequate level of challenge, corresponding to usual care and level of physical activity. Hereafter
an error rate of 20–40%. This means that an exercise they will receive an intervention with one weekly ses-
executed at “moderate” intensity is conducted with sion of BMCT and one weekly session of PRT. Both
approximately 60–80% error free repetitions, cycles sessions will start out with a 10-minute warm-up on a
or time depending on the task/exercise. stationary bicycle or treadmill. BMCT will be carried
136 J. Callesen et al. / How does strength training and balance training affect gait and fatigue in MS?

out in accordance with the described protocol for the explained by the short walking distance, where the
BMCT-group. PRT consists of the same exercises as consequences of fatigue, spasticity, and inadequate
prescribed for the PRT-group but only an intensity of coordination and balance will not manifest clearly
12 RM is applied during the 10 weeks due to fewer (Nieuwenhuis, Van Tongeren, Sorensen, & Ravn-
sessions of PRT. borg, 2006; Schwid et al., 1999).

3.3. Adherence 3.6. Six Spot Step Test

All interventions are carried out in groups of 3–6 The SSST involves walking as fast as possible
patients. To ensure adherence to the predefined proto- across a rectangular course, while kicking five blocks
col, physiotherapists will be instructed by the primary out of their circles marked on the ground. Compared
investigator on how to provide the interventions as to the T25FW, which reflects a certain degree of
well as receive written guidelines describing the monotonous and automatized movement pattern, the
exercises and the applied progression model. The SSST can be seen as a more complex task that requires
therapists will register how training went in terms an ongoing adaptation in terms of orientation, speed
of type and execution of exercises, level of intensity and limb support. As such, the SSST seems to be
and repetitions. The physiotherapist in charge of the a better expression of complex lower-limb function
intervention at the participating physiotherapy clinic than maximal straight-line walking.
registers symptoms of possible adverse events related SSST is superior to T25FW in discriminat-
to training. Any adverse event is then reported to the ing ambulatory function across levels of disability
primary investigator who then reports to the Regional (Nieuwenhuis et al., 2006), though it still calls for
ethical committee. further validation and evaluation of psychometric
properties such as responsiveness. Some perspectives
3.4. Primary outcomes on a clinical relevant change have been presented but
more data is required.
Gait will be measured by the Timed 25 Foot
Walk and the Six Spot Step Test. Gait function 3.7. Secondary outcomes
is a complex phenomenon where the interpretation
highly dependent on what aspect of function is in 3.7.1. Fatigue
focus – e.g. speed, coordination, duration, gener- Fatigue is measured by the Danish version of the
ation of force or task oriented gait. As the two modified fatigue impact scale (MFIS). The MFIS is
interventions under investigation may have differ- a multidimensional scale designed to assess the per-
ent effects on gait function (cf. the hypotheses), ceived impact of fatigue related to different aspects
a design with two primary outcomes was chosen, of daily life. It consists of 21 standardized questions
that include a simple test of maximal gait speed where each item has five levels of response. The scale
(T25FW) and a test of more complex gait that include assesses perceived impact of fatigue on physical, cog-
elements of balance and coordination (SSST). Per- nitive and psychosocial function (Fisk et al., 1994).
spectives on clinical relevant changes are published MFIS is reliable in people with MS (Learmonth et al.,
for T25FW (Baert et al., 2014) and SSST (H. B. 2013)
Jensen, Mamoei, Ravnborg, Dalgas, & Stenager,
2016). 3.7.2. Endurance and self-reporeted gait
function
3.5. Timed 25 Foot Walk The six-minute walk test (6 MW) induces an
endurance component and is therefore more prone to
Patients complete a T25FW (s) on a 10-meter exhaust patients. The 6 MW has been shown to cap-
walkway at fastest safe walking speed (FWS), respec- ture motor fatigue, as the test might affect gait pattern
tively. Timed 25-Foot Walk (T25FW) has been and reveal a change in motor strategies after rehabilita-
validated in MS as one of the three components of the tion (Leone et al., 2016). Outcome is the distance (m)
Multiple Sclerosis Functional Composite (MSFC) walked per minute and total distance covered in 6 min-
(Fischer, Rudick, Cutter, & Reingold, 1999). A limi- utes. Additionally, the distance covered in the first and
tation of the T25FW, is a floor effect seen in patients the sixth minute is registered (Leone et al., 2016). The
with mild disability (Kalkers et al., 2000). This is 6 MW is a feasible, reproducible, and reliable mea-
J. Callesen et al. / How does strength training and balance training affect gait and fatigue in MS? 137

sure (Goldman, Marrie, & Cohen, 2008). To measure ter displacement of COM. Outcome is trajectory path
the subjectively experienced impact of MS on walk- length and ellipse area (Kalron, Nitzani, & Achiron,
ing ability, the 12-item MS Walking Scale (MSWS) 2016).
(Hobart, Riazi, Lamping, Fitzpatrick, & Thompson, Mini-BESTest is a 14 item function test that mea-
2003) is used. The quality of the questionnaire eval- sures dynamic balance (Franchignoni, Horak, Godi,
uated from a graded response model is found to be Nardone, & Giordano, 2010). The mini BESTest is a
good (Mokkink, Galindo-Garre, & Uitdehaag, 2016) reliable measure of balance and it has shown good
The current validity of the scale in terms of psychome- discriminative properties for identifying abnormal
tric properties is primarily based on correlations with postural response (King, Priest, Salarian, Pierce, &
measures of gait speed (T25FW) (Kurtzke, 1983) and Horak, 2012) as well as for determining the risk of
distance (6 MW) (Goldman et al., 2008; Pilutti et al., fall in patients with Parkinson’s (Ross et al., 2016).
2013), despite these measures would expectedly only The “Activities-specific Balance Confidence
partly explain the MSWS score. Scale” (ABC) questionnaire asses balance con-
fidence in 16 different activities from everyday
3.7.3. Temporospatial and kinetic measures life (Powell & Myers, 1995) that requires static,
Walking at fastest and at self-selected speed is dynamic, proactive and reactive balance. It was
recorded by a Qualisys system including seven developed as an instrument to detect risk of falling
ProReflex infrared high-speed cameras. The sub- in elderly people, but has also been used for persons
jects are equipped with 37 retro-reflexive markers with MS. It is validated using an Italian translation
mounted on the skin according to Visual 3D® marker (Cattaneo, Regola, & Meotti, 2006). It is believed
set guidelines. The tests are performed bare-footed that confidence in the bodily abilities is a prerequisite
unless orthosis or other gait assistive devices prevent for bodily actions. Hence, in this study the correlation
this. The system collects kinematic data for a three- between objective and subjective balance-measures
dimensional analysis of gait function and is used for is determined.
analysing: step and stride length, step width, time in
swing and stance, gait speed and hip and knee angle 3.9. Strength
during gait. A thin carpet on the walkway hides the
force plate and patients strike the force plate with 3.9.1. Maximum isometric and isokinetic muscle
one foot. The test is repeated in order to measure strength
foot strike on the opposite side. From the force plate Maximum voluntary contraction (MVC) of the
data, a standard three-dimensional inverse-dynamics knee flexors (KF), knee extensors (KE), dorsal flex-
method is used to quantify the relevant forces and ors (DF) and plantar flexors (PF) are determined by
moments acting on the lower limb joints (Everett & a Cybex Norm dynamometer (Humac Norm, CSMi,
Kell, 2010). Combined with angular joint velocity Stoughton, MA, USA). The individual settings of the
the peak net joint power, normalized to body mass dynamometer are registered and used during the re-
(W/kg) can be calculated. test. Isometric muscle strength in KE and KF is tested
at 70 and 30 degrees flexion, respectively. Isometric
3.8. Balance PF and DF are tested with the knee extended and the
ankle in 0◦ and 25◦ flexion, respectively and isoki-
Static balance is measured using the “(modified) netic (60◦ /sec.) strength is tested in a range of motion
Clinical Test for Sensory Interaction and Balance going from 10◦ dorsal flexion to 25◦ plantar flexion.
(CTSIB)” (Shumway-Cook & Horak, 1986). The The patients maximally extend their knee for
somatosensory, visual, vestibular systems and their approximately 4 seconds followed by a 30 sec. rest.
interactions are essential for postural control and the This sequence is repeated 3 times. The torque signal
test investigates the organization of sensory inputs for is sampled at 100 Hz, and peak torque is determined
postural control. as the highest attained torque during a single attempt.
The CTSIB test assesses the influence of sensory
interaction on postural stability. This is done by com- 3.10. Health related quality of life
promising one or more senses during 30 sec stance in
four conditions: 1) eyes open 2) eyes closed 3) eyes Information on health status is measured by the
open standing on foam 4) eyes closed standing on Short-Form Health Survey (SF-12) (Jenkinson et al.,
foam. CTSIB is performed at a force plate to regis- 1997; Ware, Kosinski, & Keller, 1996). It is a generic
138 J. Callesen et al. / How does strength training and balance training affect gait and fatigue in MS?

questionnaire evaluating physical and mental health into account that observations on different patients at
that was developed to provide a shorter alternative to the same center are typically better correlated than
the SF-36. Scores on both physical and mental health what is seen in patients from different centers. Simi-
is found reliable and valid (Cheak-Zamora, Wyrwich, larly, the stronger correlation among observations on
& McBride, 2009). The questionnaire is generic and the same patient compared to different patients at the
does not specifically target issues related to MS. same center is accounted for.
It will be checked whether the assumption of equal
3.11. Sample size variance and a normal distribution of residuals is met
based on the mixed effect model. In case of violation
Sample size is estimated from an expected change of assumptions the analyses will be carried out on a
in gait speed assessed by Timed 25 foot walk (T25FW) logarithmic scale.
and the Six Spot Step Test (SSST). The effect Descriptive statistics will be used to present and
of each intervention are compared to the control compare characteristics across groups at baseline and
group (effect = 0, Power = 0.80 and level of signif- to determine the distributional properties of the data.
icance = 0.05). As two primary outcomes (T25FW Spearman or Pearson correlation analysis will be
and SSST) are applied the sample size estimation conducted on baseline data from all patients that
is based on four different calculations (two for each underwent baseline testing in order to determine
intervention): associations between the primary outcomes and the
secondary outcomes on fatigue, spatial gait parame-
1. PRT evaluated by T25FW: A study evaluating
ters, balance and self-reported measures.
a comparable PRT intervention found a 12%
Analyses will be carried out as intention-to-treat
increase in gait speed (Dalgas et al., 2009) with
where all patients that are tested at follow-up are
an SD of 8%. With a similar result 7 participants
include regardless of their compliance to the pro-
in each group are required.
tocol. In case of low adherence, a supplementary
2. PRT evaluated by SSST: In an unpublished
per-protocol analysis will be performed, where atten-
study, a group of comparable patients conducted
dance according to the protocol will be defined as
30 weeks of PRT (H. Jensen et al., 2015),
participants having a minimum of 80% attendance
showing an increase of 16.8%, SD = 15%. From
(corresponding to 16 sessions out of 20). Level of
this estimate, 13 participants in each group are
significance is set at 0.05.
needed.
3. BMCT evaluated by T25FW: Studies on elderly
with osteoporosis (Halvarsson, Franzen, &
4. Discussion
Stahle, 2015) are used as no relevant studies
in MS could be located. Twelve weeks of bal-
This RCT will expand our understanding of how
ance training resulted in an 8% increase in gait
different rehabilitation interventions affect gait, bal-
speed, SD = 11% (data from Halvorsson et al.
ance and fatigue and the study will thereby help
(Halvarsson et al., 2015) and personal com-
optimizing future rehabilitation interventions in per-
munication). Based on this 29 participants are
sons with MS. Moreover, not many studies evaluate
required in each group.
different related rehabilitation interventions in a head
4. BMCT evaluated by SSST: No publications
to head design in persons with MS. Expectedly,
usable for the power calculation were found.
BMCT will improve balance (Gandolfi et al., 2015;
In summary, 32 patients in each group (total n = 96) Paltamaa et al., 2012), while PRT will improve mus-
will be sufficient, when assuming a dropout rate of cle strength (Kjolhede et al., 2012). However, neither
10%. the size of the effects, nor whether one interven-
tion is superior to the other is clear, when evaluated
3.12. Statistical analysis by a functional test relying on both strength and
balance.
The effects observed from baseline to follow-up on The amount of high quality literature on PRT is
T25FW, SSST and secondary outcomes in the three substantially larger (Kjolhede et al., 2012) than on
groups is examined in a mixed effects regression anal- BMCT (Paltamaa et al., 2012). This may in part, be
ysis with group and time as fixed effects and center explained by the scientific feasibility of PRT in terms
and patient within center as random effects. This takes of standardization, where intensity and exercises
J. Callesen et al. / How does strength training and balance training affect gait and fatigue in MS? 139

are easily defined and described and where muscle participants are similarly challenged relative to their
strength outcomes can be highly standardised and motor control ability. Likewise in the PRT interven-
easily determined. Despite the extensive use in phys- tion, loads differs across patients depending on their
iotherapy practise, BMCT is harder to fit into a highly strength level.
standardized protocol, as required when causation For the BMCT, therapists select and adjust exer-
is investigated. Moreover, improvement of balance cises aiming at an error rate of approx. 20–40%. It
and motor skills that are transferable to everyday can be argued, that the dependency of the physiother-
tasks require BMCT that is tailored with respect apeutic skill combined with the multi-center design,
to repetition and variation in the applied exercises where interventions are carried out by different thera-
(Carr & Shepherd, 2010). Despite the challenges pists, will pose a threat to the validity of the study. On
when investigating BMCT it may pose elements that the contrary, it can be argued that a BMCT interven-
are complementary to PRT, and addition of BMCT tion, with only one therapist in charge of all training
may therefore help optimize motor rehabilitation sessions, tend to reflect only one person’s handling
in MS. of the prescribed intervention. In the present study,
Execution of gait is the result of a complex the external validity is thought to be high since the
interaction between the environment, physical and treatment is performed in a real clinical setting, is
cognitive abilities and the task (Shumway-Cook & provided by therapists experienced in MS rehabil-
Woollacott, 2017). Hence, manipulation of either sen- itation, is implemented into the framework of the
sory input, cognitive requirements or task-oriented national health care system and that the inclusion and
demands affects motor processing and can poten- exclusion criteria are quite wide.
tially alter movement pattern during gait. In order The SSST and the T25FW are the primary out-
to provide an exercise intervention that transfers into comes that evaluate function capacity. However, from
improved functional ability during complex use in the array of secondary measurements it will be pos-
everyday life, variation in exposure during exercise is sible gain insight into a potential diverse effect of
advised (Nielsen et al., 2015). To address this issue, PRT and BMCT when looking in mu-scle strength
the framework developed for the BMCT interven- and balance, fatigue, gait characteristics and the self-
tion defines training in stations that ensures variation. reported level of function. Additionally, as the effects
Moreover, the BMCT exercises at a given station are are results of rehabilitation conducted at different
altered and adjusted, as the program progresses, but clinics, data allows for stratification, to determine sys-
the functional aim and duration at a given exercise temically differences across the participating centers
station is maintained. that provide the same intervention. This might give
This concept of tailored exercises that varies across rise to future studies of the management of guidelines
patients, calls for a discussion of the reproducibility in clinical practice.
of particularly the BMCT intervention. In this study,
the idea is to standardize the treatment paradigm
rather than the treatment itself. The therapists are Authors’ contributions
provided reliable rules for the content and progres-
sion, but without specifying the exercises. Intensity JC contributed to the conception and design of the
of BMCT is then determined from the level of diffi- study and drafted the manuscript. UD contributed to
culty (the error rate) interpreted from the quality of the conception and design of the study and edited
the execution of a particular exercise. When perform- the manuscript. DC contributed to the conception and
ing BMCT, the level of function is to a greater extend design of the study and edited the manuscript. JB con-
dependent on cognitive processing than during PRT. tributed to the conception and design of the study and
This is primary explained by increased requirements edited the manuscript. All authors read and approved
in relation to fine motor function, divided attention the final manuscript.
and the continuous exposure to new tasks. All though,
muscle strength is still related to performance.
Regarding standardization, two patients with Conflict of interest
different clinical characteristics will receive dif-
ferent BMCT interventions, while subjects with JC, DC and JB declare that they have no con-
similar characteristics will receive a comparable flict of interests associated with this paper. UD
intervention. Despite diverse BMCT exposure, the has received research support, travel grants and/or
140 J. Callesen et al. / How does strength training and balance training affect gait and fatigue in MS?

teaching honorary from Biogen Idec, Merck Serono, Coote, S., Garrett, M., Hogan, N., Larkin, A., & Saunders, J.
Novartis, Bayer Schering and Sanofi Aventis as well (2009). Getting the balance right: A randomised controlled
trial of physiotherapy and exercise interventions for ambula-
as honoraria from serving on scientific advisory
tory people with multiple sclerosis. BMC Neurology, 9(34).
boards of Biogen Idec and Genzyme. doi:10.1186/1471-2377-9-34
Dalgas, U., Stenager, E., Jakobsen, J., Petersen, T., Hansen,
H. J., Knudsen, C., . . . & Ingemann-Hansen, T. (2009).
Resistance training improves muscle strength and functional
Funding capacity in multiple sclerosis. Neurology, 73(18), 1478-1484.
doi:10.1212/WNL.0b013e3181bf98b4
This work was supported by VIA University Col- Everett, T., & Kell, C. (2010). Human movement: An introductory
lege - Faculty of Health Sciences and by the Danish text (6th ed.). Edinburgh: Churchill Livingstone.
Fischer, J. S., Rudick, R. A., Cutter, G. R., & Reingold, S. C. (1999).
foundation “TrygFonden”, grant number: 108916.
The multiple sclerosis functional composite measure (MSFC):
An integrated approach to MS clinical outcome assessment.
National MS society clinical outcomes assessment task force.
Multiple Sclerosis (Houndmills, Basingstoke, England), 5(4),
References
244-250.
Fisk, J. D., Ritvo, P. G., Ross, L., Haase, D. A., Marrie, T. J.,
American College of Sports Medicine. (2009). American college & Schlech, W. F. (1994). Measuring the functional impact of
of sports medicine position stand progression models in resis- fatigue: Initial validation of the fatigue impact scale. Clinical
tance training for healthy adults. Medicine and Science in Infectious Diseases: An Official Publication of the Infectious
Sports and Exercise, 41(3), 687-708. doi:10.1249/MSS.0b013 Diseases Society of America, 18(Suppl 1), S79-S83.
e3181915670 Franchignoni, F., Horak, F., Godi, M., Nardone, A., & Gior-
Baert, I., Freeman, J., Smedal, T., Dalgas, U., Romberg, A., dano, A. (2010). Using psychometric techniques to improve
Kalron, A., . . . & Feys, P. (2014). Responsiveness and clini- the balance evaluation systems test: The mini-BESTest. Jour-
cally meaningful improvement, according to disability level, nal of Rehabilitation Medicine, 42(4), 323-331. doi:10.2340/
of five walking measures after rehabilitation in multiple 16501977-0537
sclerosis: A european multicenter study. Neurorehabilitation Gandolfi, M., Munari, D., Geroin, C., Gajofatto, A., Benedetti, M.
and Neural Repair, 28(7), 621-631. doi:10.1177/1545968314 D., Midiri, A., . . . & Smania, N. (2015). Sensory integration
521010 balance training in patients with multiple sclerosis: A ran-
Boes, M. K., Sosnoff, J. J., Socie, M. J., Sandroff, B. M., Pula, J. H., domized, controlled trial. Multiple Sclerosis (Houndmills, Bas-
& Motl, R. W. (2012). Postural control in multiple sclerosis: ingstoke, England), 21(11), 1453-1462. doi:10.1177/1352458
Effects of disability status and dual task. Journal of the Neu- 514562438
rological Sciences, 315(1-2), 44-48. doi:10.1016/j.jns.2011. Goldman, M. D., Marrie, R. A., & Cohen, J. A. (2008). Evalua-
12.006 tion of the six-minute walk in multiple sclerosis subjects and
Cameron, M. H., & Wagner, J. M. (2011). Gait abnormalities in healthy controls. Multiple Sclerosis (Houndmills, Basingstoke,
multiple sclerosis: Pathogenesis, evaluation, and advances in England), 14(3), 383-390. doi:10.1177/1352458507082607
treatment. Curr Neurol Neurosci Rep, 11(5), 507-515. doi:10. Halvarsson, A., Franzen, E., & Stahle, A. (2015). Balance
1007/s11910-011-0214-y training with multi-task exercises improves fall-related self-
Carling, A., Forsberg, A., Gunnarsson, M., & Nilsagård, Y. (2017). efficacy, gait, balance performance and physical function in
CoDuSe group exercise programme improves balance and older adults with osteoporosis: A randomized controlled trial.
reduces falls in people with multiple sclerosis: A multi-centre, Clinical Rehabilitation, 29(4), 365-375. doi:10.1177/026921
randomized, controlled pilot study. Multiple Sclerosis Journal, 5514544983
23(10), 1394-1404. doi:10.1177/1352458516677591 Hebert, J. R., Corboy, J. R., Manago, M. M., & Schenkman,
Carr, J. H., & Shepherd, R. B. (2010). Neurological rehabilitation: M. (2011). Effects of vestibular rehabilitation on multiple
Optimizing motor performance (2nd ed.). Edinburgh: Churchill sclerosis-related fatigue and upright postural control: A ran-
Livingstone. domized controlled trial. Physical Therapy, 91(8), 1166-1183.
Cattaneo, D., Regola, A., & Meotti, M. (2006). Validity of six doi:10.2522/ptj.20100399
balance disorders scales in persons with multiple sclerosis. Heesen, C., Bohm, J., Reich, C., Kasper, J., Goebel, M., & Gold,
Disability and Rehabilitation, 28(12), 789-795. doi:10.1080/ S. M. (2008). Patient perception of bodily functions in multi-
09638280500404289 ple sclerosis: Gait and visual function are the most valuable.
Cheak-Zamora, N. C., Wyrwich, K. W., & McBride, T. D. (2009). Multiple Sclerosis (Houndmills, Basingstoke, England), 14(7),
Reliability and validity of the SF-12v2 in the medical expen- 988-991.
diture panel survey. Quality of Life Research: An International Heine, M., van de Port, I., Rietberg, M. B., van Wegen, E. E., &
Journal of Quality of Life Aspects of Treatment, Care and Reha- Kwakkel, G. (2015). Exercise therapy for fatigue in multiple
bilitation, 18(6), 727-735. doi:10.1007/s11136-009-9483-1 sclerosis. The Cochrane Database of Systematic Reviews, 9,
Claros-Salinas, D., Dittmer, N., Neumann, M., Sehle, A., Spi- CD009956. doi:10.1002/14651858.CD009956.pub2
teri, S., Willmes, K., . . . & Dettmers, C. (2013). Induction of Hobart, J. C., Riazi, A., Lamping, D. L., Fitzpatrick, R., & Thomp-
cognitive fatigue in MS patients through cognitive and phys- son, A. J. (2003). Measuring the impact of MS on walking
ical load. Neuropsychological Rehabilitation, 23(2), 182-201. ability: The 12-item MS walking scale (MSWS-12). Neurol-
doi:10.1080/09602011.2012.726925 ogy, 60(1), 31-36.
J. Callesen et al. / How does strength training and balance training affect gait and fatigue in MS? 141

Homberg, V. (2013). Neurorehabilitation approaches to facilitate Decline in walking distance induced by the 6-minute walk
motor recovery. Handbook of Clinical Neurology, 110, 161- test. Neurorehabilitation and Neural Repair, 30(4), 373-383.
173. doi:10.1016/B978-0-444-52901-5.00014-9 doi:10.1177/1545968315597070
Jenkinson, C., Layte, R., Jenkinson, D., Lawrence, K., Petersen, S., Mokkink, L. B., Galindo-Garre, F., & Uitdehaag, B. M. (2016).
Paice, C., & Stradling, J. (1997). A shorter form health survey: Evaluation of the multiple sclerosis walking scale-12 (MSWS-
Can the SF-12 replicate results from the SF-36 in longitudinal 12) in a dutch sample: Application of item response
studies? Journal of Public Health Medicine, 19(2), 179-186. theory. Multiple Sclerosis (Houndmills, Basingstoke, Eng-
Jensen, H. B., Nielsen, J. L., Ravnborg, M., Dalgas, U., Aagaard, land), 22(14), 1867-1873.
P., & Stenager, E. (2016). Effect of slow release-fampridine Motl, R. W., & Pilutti, L. A. (2012). The benefits of exercise
on muscle strength, rate of force development, functional training in multiple sclerosis. Nature Reviews Neurology, 8(9),
capacity and cognitive function in an enriched population of 487-497. doi:10.1038/nrneurol.2012.136
MS patients. A randomized, double blind, placebo controlled Multiple Sclerosis International Federation. (2014). Atlas of MS
study. Multiple Sclerosis and Related Disorders, 10, 137-144. 2013: Mapping multiple sclerosis around the world Retrieved
doi:10.1016/j.msard.2016.07.019 from https://www.msif.org/wp-content/uploads/2014/09/Atlas-
Jensen, H. B., Mamoei, S., Ravnborg, M., Dalgas, U., & Ste- of-MS.pdf
nager, E. (2016). Distribution-based estimates of minimum Nielsen, J. B., Willerslev-Olsen, M., Christiansen, L., Lundbye-
clinically important difference in cognition, arm function and Jensen, J., & Lorentzen, J. (2015). Science-based neurorehabil-
lower body function after slow release-fampridine treatment of itation: Recommendations for neurorehabilitation from basic
patients with multiple sclerosis. Multiple Sclerosis and Related science. Journal of Motor Behavior, 47(1), 7-17. doi:10.1080/
Disorders, 7, 58-60. doi:10.1016/j.msard.2016.03.007 00222895.2014.931273
Kalkers, N. F., de Groot, V., Lazeron, R. H., Killestein, J., Ader, Nieuwenhuis, M. M., Van Tongeren, H., Sorensen, P. S., & Ravn-
H. J., Barkhof, F., . . . & Polman, C. H. (2000). MS functional borg, M. (2006). The six spot step test: A new measurement
composite: Relation to disease phenotype and disability strata. for walking ability in multiple sclerosis. Multiple Sclerosis
Neurology, 54(6), 1233-1239. (Houndmills, Basingstoke, England), 12(4), 495-500.
Kalron, A., Nitzani, D., & Achiron, A. (2016). Static posturogra- Nilsagard, Y. E., von Koch, L. K., Nilsson, M., & Forsberg, A.
phy across the EDSS scale in people with multiple sclerosis: S. (2014). Balance exercise program reduced falls in peo-
A cross sectional study. BMC Neurology, 16(70). doi:10.1186/ ple with multiple sclerosis: A single-group, pretest-posttest
s12883-016-0603-6 trial. Archives of Physical Medicine and Rehabilitation, 95(12),
Khan, F., Amatya, B., Galea, M. P., Gonzenbach, R., & Kessel- 2428-2434. doi:10.1016/j.apmr.2014.06.016
ring, J. (2017). Neurorehabilitation: Applied neuroplasticity. Noseworthy, J. H., Lucchinetti, C., Rodriguez, M., & Weinshenker,
Journal of Neurology, 264(3), 603-615. doi:10.1007/s00415- B. G. (2000). Multiple sclerosis. The New England Journal of
016-8307-9 Medicine, 343(13), 938-952.
King, L. A., Priest, K. C., Salarian, A., Pierce, D., & Horak, F. Paltamaa, J., Sjogren, T., Peurala, S. H., & Heinonen, A. (2012).
B. (2012). Comparing the mini-BESTest with the berg bal- Effects of physiotherapy interventions on balance in multiple
ance scale to evaluate balance disorders in parkinson’s disease. sclerosis: A systematic review and meta-analysis of random-
Parkinson’s Disease, 2012, 375419. doi:10.1155/2012/375419 ized controlled trials. Journal of Rehabilitation Medicine,
Kjolhede, T., Vissing, K., & Dalgas, U. (2012). Multiple sclero- 44(10), 811-823. doi:10.2340/16501977-1047
sis and progressive resistance training: A systematic review. Penner, I. K., Bechtel, N., Raselli, C., Stöcklin, M., Opwis, K., Kap-
Multiple Sclerosis, 18(9), 1215-1228. doi:10.1177/13524585 pos, L., & Calabrese, P. (2007). Fatigue in multiple sclerosis:
12437418 Relation to depression, physical impairment, personality and
Kurtzke, J. F. (1983). Rating neurologic impairment in multiple action control. Multiple Sclerosis (Houndmills, Basingstoke,
sclerosis: An expanded disability status scale (EDSS). Neurol- England), 13(9), 1161-1167.
ogy, 33(11), 1444-1452. Pilutti, L. A., Dlugonski, D., Sandroff, B. M., Suh, Y., Pula, J. H.,
Latimer-Cheung, A. E., Pilutti, L. A., Hicks, A. L., Martin Ginis, K. Sosnoff, J. J., & Motl, R. W. (2013). Further validation of mul-
A., Fenuta, A. M., MacKibbon, K. A., & Motl, R. W. (2013). tiple sclerosis walking scale-12 scores based on spatiotemporal
Effects of exercise training on fitness, mobility, fatigue, and gait parameters. Archives of Physical Medicine and Rehabili-
health-related quality of life among adults with multiple scle- tation, 94(3), 575-578. doi:10.1016/j.apmr.2012.08.214
rosis: A systematic review to inform guideline development. Pollock, A. S., Durward, B. R., Rowe, P. J., & Paul, J. P. (2000).
Archives of Physical Medicine and Rehabilitation, 94(9), 1800- What is balance? Clinical Rehabilitation, 14(4), 402-406.
1828.e3. doi:10.1016/j.apmr.2013.04.020 Powell, L. E., & Myers, A. M. (1995). The activities-specific bal-
Learmonth, Y. C., Dlugonski, D., Pilutti, L. A., Sandroff, B. M., ance confidence (ABC) scale. The Journals of Gerontology.
Klaren, R., & Motl, R. W. (2013). Psychometric properties Series A, Biological Sciences and Medical Sciences, 50A(1),
of the fatigue severity scale and the modified fatigue impact M28-34.
scale. Journal of the Neurological Sciences, 331(1-2), 102-107. Rietberg, M. B., Brooks, D., Uitdehaag, B. M., & Kwakkel,
doi:10.1016/j.jns.2013.05.023 G. (2005). Exercise therapy for multiple sclerosis. The
Lee, D., Newell, R., Ziegler, L., & Topping, A. (2008). Treatment Cochrane Database of Systematic Reviews, (1), CD003980.
of fatigue in multiple sclerosis: A systematic review of the doi:10.1002/14651858.CD003980.pub2
literature. International Journal of Nursing Practice, 14(2), Ross, E., Purtill, H., Uszynski, M., Hayes, S., Casey, B., Browne,
81-93. doi:10.1111/j.1440-172X.2008.00670.x C., & Coote, S. (2016). Cohort study comparing the berg bal-
Leone, C., Severijns, D., Dolezalova, V., Baert, I., Dalgas, U., ance scale and the mini-BESTest in people who have multiple
Romberg, A., . . . & Feys, P. (2016). Prevalence of walking- sclerosis and are ambulatory. Physical Therapy, 96(9), 1448-
related motor fatigue in persons with multiple sclerosis: 1455. doi:10.2522/ptj.20150416
142 J. Callesen et al. / How does strength training and balance training affect gait and fatigue in MS?

Schwid, S. R., Thornton, C. A., Pandya, S., Manzur, K. L., Sanjak, Shumway-Cook, A., & Woollacott, M. H. (2017). Motor control:
M., Petrie, M. D., . . . & Goodman, A. D. (1999). Quantitative Translating research into clinical practice (5th ed.). Philadel-
assessment of motor fatigue and strength in MS. Neurology, phia: Wolters Kluwer.
53(4), 743-750. Ware, J., Jr, Kosinski, M., & Keller, S. D. (1996). A 12-item short-
Shumway-Cook, A., & Horak, F. B. (1986). Assessing the influ- form health survey: Construction of scales and preliminary
ence of sensory interaction on balance: Suggestion from the tests of reliability and validity. Medical Care, 34(3), 220-233.
field. Physical Therapy, 66(10), 1548-1550.

You might also like