Accepted Manuscript The effect of exercise training on fitness in multiple sclerosis: A meta-analysis M.E. Platta, I. Ensari, R.W. Motl, L.A. Pilutti PII:
S0003-9993(16)00091-5
DOI:
10.1016/j.apmr.2016.01.023
Reference:
YAPMR 56446
To appear in:
ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION
Received Date: 26 October 2015 Revised Date:
6 January 2016
Accepted Date: 28 January 2016
Please cite this article as: Platta ME, Ensari I, Motl RW, Pilutti LA, The effect of exercise training on fitness in multiple sclerosis: A meta-analysis, ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION (2016), doi: 10.1016/j.apmr.2016.01.023. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
RUNNING HEAD: Fitness in multiple sclerosis
RI PT
Title: The effect of exercise training on fitness in multiple sclerosis: A meta-analysis Authors: M.E. Platta, I. Ensari, R.W. Motl, L.A. Pilutti
SC
From the Department of Kinesiology and Community Health, University of Illinois Urbana-
M AN U
Champaign, Urbana, IL, USA
Acknowledgements:
This research received no specific funding from any agency.
TE D
Disclosure Statement:
Dr. Motl reports a financial relationship with EMD Serono, Acorda Therapeutics, and Biogen
EP
Idec, outside the submitted work. The other authors have no disclosures.
Corresponding author: Lara A Pilutti, PhD, Department of Kinesiology and Community
AC C
Health, University of Illinois Urbana-Champaign, 213 Freer Hall, Urbana, IL 61801, USA. Phone: 217-333-6126. Fax: 217-244-7322. Email:
[email protected] For correspondence and reprints, contact Lara A. Pilutti, Ph.D., Department of Kinesiology and Community Health, University of Illinois, 213 Freer Hall, Urbana, IL 61801, email:
[email protected].
ACCEPTED MANUSCRIPT Fitness in multiple sclerosis 1 The effect of exercise training on fitness in multiple sclerosis: A meta-analysis
1
Objectives: To provide a quantitative synthesis of randomized controlled trials (RCT) examining
3
the effect of exercise training on muscular and cardiorespiratory fitness in persons with MS.
4
Data Sources: Three electronic databases, Pubmed, Google Scholar and Web of Science were
5
searched for all relevant articles published up until October 2014.
6
Study Selection: Keywords included “exercise” OR “aerobic” OR “strength” OR “resistance
7
training” OR “cardiorespiratory” AND “multiple sclerosis”. Trials examining the effect of
8
exercise training on muscular and/or cardiorespiratory fitness parameters were included.
9
Data Extraction: The initial search yielded 1501 articles of which 62 were reviewed in detail
10
and 20 RCTs met the inclusion criteria and provided enough data to compute effect sizes (ESs;
11
Cohen d). The meta-analyses was conducted using Comprehensive Meta-Analysis software and a
12
random-effects model was used to compute the overall or mean ES per fitness parameter.
13
Data synthesis: The mean ES was 0.27 (SE=0.05, 95% CI=0.17-0.38, z=5.05, p< .001) for
14
muscular fitness outcomes and 0.47 (SE=0.09, 95% CI=0.30-0.65, z=5.40, p< .001) for
15
cardiorespiratory fitness outcomes. The weighted mean ES was not heterogeneous for muscular
16
(Q=11.09, df=13, p=.60, I²=0.00) or cardiorespiratory (Q=7.83, df=9, p=0.55, I²=0.00) fitness
17
outcomes.
18
Conclusions: The cumulative evidence supports that exercise training is associated with
19
changes in muscular (small in magnitude) and cardiorespiratory (moderate in magnitude) fitness
20
outcomes in persons with MS. Such an indication of magnitude is important for clinical research
21
and practice by providing an evidence-based estimate of the actual benefit that exercise training
22
confers on physiological fitness.
23
Keywords: exercise; fitness; muscular; cardiorespiratory; multiple sclerosis
AC C
EP
TE D
M AN U
SC
RI PT
2
ACCEPTED MANUSCRIPT
Abbreviations
25
1-RM = 1-repetition maximum
26
ES = effect size
27
MS = multiple sclerosis
28
Nm = Newton meters
29
RCT = randomized controlled trial
30
VO2peak = peak aerobic capacity
31
Wp = peak power output
M AN U
32 33 34 35
40 41 42 43 44 45 46
EP
39
AC C
38
TE D
36 37
SC
24
RI PT
Fitness in multiple sclerosis 2
ACCEPTED MANUSCRIPT Fitness in multiple sclerosis 3 47
Introduction Multiple sclerosis (MS) is a neurological disease characterized by inflammatory attacks
49
within the central nervous system that result in axonal demyelination and transection as well as
50
neurodegeneration.1 The pathophysiological processes associated with the disease result in
51
impairments in various domains including walking, cognition, and symptomatic changes such as
52
depression and fatigue. The consequences of MS further compromise participation in physical
53
activity and this results in significant detraining (i.e., physiological deconditioning), particularly
54
in the domains of muscular and cardiorespiratory fitness.2 Muscular fitness reflects the body’s
55
ability for generating and maintaining muscular force via skeletal muscle contraction and is
56
commonly measured as muscular strength (e.g., peak force) and endurance (e.g., maximum
57
repetitions). Cardiorespiratory fitness reflects the body’s ability for delivering, extracting, and
58
using oxygen for prolonged endurance exercise and is typically measured as peak aerobic
59
capacity (VO2peak), peak power output (Wp), and anaerobic or lactate threshold (l/mmol). The
60
maintenance of muscular and cardiorespiratory fitness is critical for preserving and restoring
61
physical functioning and participation among persons with MS.3
SC
M AN U
TE D
There is evidence that both muscular and cardiorespiratory fitness are impaired in persons
EP
62
RI PT
48
with MS. For example, one study of 15 participants with MS and 15 healthy controls
64
demonstrated that persons with MS had significantly lower peak torque production in the
65
nondominant knee extensors, dominant knee flexors, and nondominant knee flexors compared to
66
healthy controls.4 This has been replicated in other studies.5,6 Similarly, cardiorespiratory fitness,
67
based on VO2peak, was 28% higher in a sample of 25 healthy controls compared to 25 participants
68
with MS,7 and such deficits have been replicated in other studies.6,8,9 The detraining and loss of
69
fitness in persons with MS combined with the importance of physiological fitness for functional
AC C
63
ACCEPTED MANUSCRIPT Fitness in multiple sclerosis 4 70
and symptomatic outcomes underscores the importance of developing and evaluating the
71
efficacy of interventions that target indices of muscular and cardiorespiratory fitness.
72
Exercise training probably reflects the intervention modality with the best likelihood and relevance for improving physiological fitness in persons with MS. There is evidence for the
74
benefits of resistance and aerobic training for improving physiological fitness in persons with
75
MS.10–12 For example, one systematic review reported that 8-20 weeks of moderate intensity
76
(i.e., 10-12 repetitions @ 70-80% 1-repetition maximum (1-RM)) resistance training performed
77
2 to 3 times per week increased muscular strength in persons with MS; there were insufficient
78
data for demonstrating a change in muscular endurance.13 That review further reported that
79
moderate intensity (i.e., 60-80% Wp or 60% VO2peak) aerobic exercise training, 2 to 3 times a
80
week for 30 to 60 minutes increased cardiorespiratory fitness (VO2peak) in persons with MS.13
SC
M AN U
81
RI PT
73
Those data are promising, but have important limitations that can be overcome in a metaanalysis. The studies often suffer from small sample sizes resulting in non-significant effects
83
resulting in the possible interpretation of no consistent or definitive effect of exercise training on
84
fitness. The studies and literature review do not provide an indication of the magnitude of
85
improvement in fitness outcomes following exercise training that can be provided through meta-
86
analysis. Such an indication of magnitude is important for clinical research and practice by
87
providing an evidence-based estimate of the actual benefit that exercise training may confer on
88
physiological fitness. We further note that there is likely variability in the effect of exercise
89
training on fitness adaptations, and this can be quantified using moderator analysis in meta-
90
analysis. This will be important for indicating the characteristics of exercise or the sample that
91
maximize the improvement in fitness outcomes.
AC C
EP
TE D
82
ACCEPTED MANUSCRIPT Fitness in multiple sclerosis 5 92
We undertook a meta-analysis for providing a quantitative assessment of the effect of exercise training on muscular and cardiorespiratory fitness outcomes in persons with MS. To do
94
this, we conducted a systematic and comprehensive meta-analysis of published randomized
95
controlled trials (RCTs) involving exercise training and MS that included outcomes of muscular
96
and cardiorespiratory fitness. We hypothesized that exercise training would be associated with
97
improvements in muscular and cardiorespiratory fitness outcomes and that these improvements
98
would be small to moderate in magnitude based on previous meta-analyses of exercise training
99
on other outcomes (e.g., walking, fatigue, and quality of life) in MS.14–17
SC
RI PT
93
101
M AN U
100
Methods
This meta-analysis was conducted consistent with the Meta-analysis of Observational
103
Studies in Epidemiology (MOOSE) framework,18 and a visual description of our step-by-step
104
methods is provided in Figure 1. We conducted a search PubMed and Web of Science using the
105
keywords “exercise” AND “aerobic” OR “strength” OR “resistance” OR “resistance training”
106
OR “cardiorespiratory” OR “fitness” AND “multiple sclerosis”. We included RCTs that were
107
published up to October 2014. We initially retrieved 1,501 articles of which 1,439 were excluded
108
for the following reasons: duplicates (n=880), not primary research articles (i.e., reviews, meta-
109
analyses, abstracts) (n=230), no inclusion of exercise training (n=163) or persons with MS
110
(n=104). We further excluded studies that were non-RCTs (n=44) or did not include fitness
111
measures (n=18). We further searched the bibliographies of the retrieved articles and reviewed a
112
total of 62 articles in detail. Of these articles, there were 42 without an exercise training
113
manipulation, appropriate control condition or randomization, or objective and usable fitness
114
outcome data, and these studies were excluded. If the papers did not include sufficient
AC C
EP
TE D
102
ACCEPTED MANUSCRIPT Fitness in multiple sclerosis 6 information for calculating effect sizes (ES), we contacted the authors for further information.
116
We included RCTs (i.e., studies that compared exercise training vs. no-treatment control and
117
stated assigning participants to conditions randomly) that administered measures of
118
cardiorespiratory or muscular fitness (e.g., VO2peak, isokinetic strength) pre-post intervention. We
119
excluded RCTs that did not include persons with MS, an objective fitness outcome, or an
120
appropriate exercise training intervention. Exercise was defined as a planned, structured, and
121
repetitive form of physical activity conducted over an extended period of time with the goal of
122
improving health-related fitness (i.e., cardiorespiratory, muscular, motor, metabolic, or
123
morphological components).19 This resulted in a total of 20 RCTs from 21 publications that were
124
included in the meta-analysis (Table 1).7,10,12,20–37
SC
M AN U
125
RI PT
115
We computed ESs expressed as Cohen’s d.38 To do this, we computed the mean change from before to after exercise training and subtracted the mean change of the control condition.
127
The resulting difference in mean change between the conditions was then divided by the baseline
128
standard deviation (SD) pooled between conditions. The ESs were calculated so that a positive
129
ES indicated an improvement in fitness levels after exercise training, whereas a negative ES
130
indicated a worsening of fitness levels in the exercise condition compared with control. Separate
131
ESs were calculated per dependent variable (i.e., some studies had multiple fitness outcomes) as
132
well as per type of exercise (i.e., some studies had more than one type of exercise training
133
modality or condition). The overall analysis took place using a single ES per study (i.e., an
134
average ES when there was more than one ES computed by the software). This was necessary as
135
multiple ESs from the same study are not independent.39 The lack of independence can bias the
136
standard error (SE) for judging the significance of the overall ES and multiple ESs from one
137
study bias the overall ES disproportionately compared with the studies that have a single ES.39
AC C
EP
TE D
126
ACCEPTED MANUSCRIPT Fitness in multiple sclerosis 7 138
The ESs along with the associated standard errors (SEs) were entered into the Comprehensive Meta-analysis Software,a Version 2.0. The analyses were conducted separately
140
for the muscular and cardiorespiratory fitness outcomes. We used a random-effects model for
141
computing the overall or mean ES (for each fitness outcome) as this model assumes that the
142
samples come from populations with different ESs and that the true effect differs between
143
studies.39 We further computed a 95% confidence interval (CI) around the mean ES. An overall
144
Q value and I2 value were calculated to test for homogeneity of variance among ESs. The Q
145
value is a measure of variance among the ESs and a statistically significant (p