Content uploaded by Rene Streber
Author content
All content in this area was uploaded by Rene Streber on Jun 15, 2016
Content may be subject to copyright.
REVIEW ARTICLE (META-ANALYSIS)
Systematic Review of Correlates and Determinants of
Physical Activity in Persons With Multiple Sclerosis
Rene
´Streber, Dipl. Sport Sci,
a
Stefan Peters, Dipl. Sport Sci,
a,b
Klaus Pfeifer, PhD
a
From the
a
Institute of Sport Science and Sport, Division Exercise and Health, Department Psychology and Sport Science, Friedrich-Alexander-
University Erlangen-Nu¨rnberg, Erlangen; and
b
Department of Medical Psychology and Psychotherapy, Medical Sociology and Rehabilitation
Sciences, University of Wu¨rzburg, Wu¨rzburg, Germany.
Abstract
Objective: To review the current evidence regarding correlates and determinants of physical activity (PA) in persons with multiple sclerosis (pwMS).
Data Sources: PubMed and Scopus (1980 to January 2015) and reference lists of eligible studies.
Study Selection: Eligible studies include adults with multiple sclerosis; have a cross-sectional or prospective observational design; or examine
the effect of a theory-based intervention trial on PA, including a mediation analysis. Eligible studies also apply a quantitative assessment of PA
and correlates or proposed mediators and are published in English or German language.
Data Extraction: Two reviewers independently evaluated the risk of bias, extracted data, and categorized variables according to the International
Classification of Functioning, Disability and Health.
Data Synthesis: Consistency and the direction of associations were evaluated with a semiquantitative approach. Fifty-six publications with data
from observational studies and 2 interventional studies provided evidence for 86 different variables. Consistent correlates of PA were the disability
level, walking limitations in particular, PA-related self-efficacy, self-regulation constructs, employment status, and educational level. One
interventional study provided evidence for a causal relation between self-regulation and PA. However, 59 of the 86 investigated variables in
observational studies are based on 1 or 2 study findings, and most results stem from cross-sectional designs.
Conclusions: Beside the importance of the general disability level and walking limitations, the results highlight the importance of personal factors
(eg, PA-related self-efficacy, self-regulatory constructs, sociodemographic factors). Limitations and implications of the current review are dis-
cussed. Research that is more rigorous is needed to better understand what affects PA in pwMS.
Archives of Physical Medicine and Rehabilitation 2016;97:633-45
ª2016 by the American Congress of Rehabilitation Medicine
Multiple sclerosis (MS) is a chronic, immune-mediated disease of
the central nervous system that is characterized by inflammation,
demyelination, axonal degeneration, and gliosis.
1-3
Usually MS is
diagnosed between the ages of 20 and 50 years, with women being
affected 2 to 3 times more often than men.
4
The clinical disease
courses can be categorized into relapsing-remitting multiple
sclerosis (RRMS), secondary-progressive MS, and primary-
progressive MS.
1
The progressive nature of MS is characterized
by an increase in neurologic impairments,
5,6
activity limitations,
and participation restrictions.
7-9
MS has a substantial affect on the
quality of life of persons with multiple sclerosis (pwMS).
10,11
There is established evidence that highlights the multiple
health benefits (eg, physical fitness, fatigue, quality of life,
walking)
12-16
and the safety of physical activity (PA) for pwMS.
17
Regular PA may even have positive long-term effects on the
progression of the disability.
18
However, PA levels among pwMS
are low in comparison with the general population.
19-21
Nearly
60% of pwMS are engaging in insufficient PA. Those with MS
were 2.5 times more likely to report insufficient PA and 2.3 times
less likely to report sufficient PA levels compared with healthy
controls.
21
To benefit from the positive health effects of PA and to
avoid the effects of physical inactivity (eg, risk factor for car-
diovascular or metabolic diseases),
22
long-term maintenance of
regular PA is of the utmost importance.
23
To understand what affects PA behavior in pwMS, it is
crucial to identify factors that are associated with PA levels
(eg, sociodemographic, psychological, environmental).
19,24,25
Presented in part to the Mobility and Education Group of the European Network Rehabilitation
in Multiple Sclerosis, June 7-8, 2013, Limerick, Ireland.
Disclosures: none.
0003-9993/16/$36 - see front matter ª2016 by the American Congress of Rehabilitation Medicine
http://dx.doi.org/10.1016/j.apmr.2015.11.020
Archives of Physical Medicine and Rehabilitation
journal homepage: www.archives-pmr.org
Archives of Physical Medicine and Rehabilitation 2016;97:633-45
A former narrative review on determinants of PA in pwMS
identified elements of the physical and social environment,
disability level, MS-related symptoms, perceived barriers, and
self-efficacy as correlates of PA in pwMS. Major limitations of
this review were the number of existing studies and that most
studies had a cross-sectional design.
26
Research on determinants
of PA in pwMS has emerged in recent years,
27,28
and the question
has been rated as one of the top 10 research questions regarding
PA and MS.
29
However, a systematic overview regarding the
current state of evidence in combination with a critical quality
appraisal of the evidence is lacking. This knowledge is essential
for the development and advancement of interventions that aim to
increase PA participation among pwMS.
24,25,30
Variables that are
modifiable and causally related to PA behavior would constitute
potential targets to increase PA levels. Additionally, characteristics
of people, such as biologic or sociodemographic factors (eg, sex,
age), might inform decisions on whom to target with PA
interventions.
24,30
When evaluating the evidence of the causal relation between a
variable and PA, it is necessary to distinguish between correlates and
determinants.
30
A correlate describes a statistical association be-
tween a measured variable and PA.
30
Evidence is mainly derived
from cross-sectional and prospective observational studies.
30-32
Results from correlational research designs have limitations in
terms of drawing inferences about the causal relation.
30-32
In
contrast, determinants describe causal factors. The relation is more
likely to be causal if variations in these variables are followed sys-
tematically by variations in PA behavior.
30
Results derived from
experimental studies provide stronger evidence for a causal relation
than do results from observational studies.
30-32
In experimental intervention studies, mediation analysis is a
statistical method to investigate the proposed causal relation be-
tween a variable and PA.
31-33
The assumption in a mediating
variable framework is a causal pathway from intervention activ-
ities that target change to intervening causal variables (ie, medi-
ator) that in turn lead to systematic changes in the desired
behavior.
30,33,34
A pure examination of an intervention effect on
PA behavior often makes it difficult to draw definite conclusions
about the hypothesized mechanisms that cause variations in PA
behavior. Other mechanisms than those postulated can lead to
changes in behavior.
31
In this context, the mediation model is
informed by 2 different theoretical links. How a mediating vari-
able is expected to be causally related to PA behavior is referred to
as conceptual theory, and how the intervention will change the
targeted mediators is referred to as an action theory link.
33,35,36
The conceptual theory test provides evidence for the causal rela-
tion between a variable and PA. For example, if there is an
intervention effect on the mediator without a significant relation
between the mediator and PA change, then there is evidence that
the mediator is not causally related to the outcome.
33
To evaluate
whether and to what extent the intervention effect on PA behavior
is mediated by the hypothesized variables and to interpret the
results of a mediation analysis properly, both tests need to be
considered.
The aim of this review is to systematically review the existing
literature to identify correlates and determinants of PA behavior in
pwMS. Quantitative associations from studies with an observa-
tional or experimental intervention design, including a mediation
analysis, will be evaluated.
Methods
This systematic review is structured according to the Preferred
Reporting Items for Systematic Reviews and Meta-Analyses
37
(workflow in supplemental appendix 1, available online only at
http://www.archives-pmr.org/).
Eligibility criteria
Eligible studies included adults (18y) diagnosed with MS, had
a cross-sectional or prospective observational design, or exam-
ined the effect of a theory-based intervention with a quasi or
experimental design on PA or exercise behavior and on the
proposed mediators of PA change and examined the causal
relation between the hypothesized mediators and PA. Eligible
studies applied a quantitative assessment of PA, potential corre-
lates, or proposed mediators and were published in English or
German language.
PA is understood as “any bodily movement produced by
skeletal muscles that results in energy expenditure.”
38(p126)
Eligible PA assessments for this review had to include a health-
enhancing PA domain (eg, exercise).
39
Studies were excluded if
they exclusively measured domestic or occupational activities, had
intervention adherence (eg, attendance), or primarily aimed to
validate a PA measure in pwMS.
Literature search strategy
Scopus and PubMed were systematically searched from 1980 to
January 31, 2015, with the following text words or, if possible,
Medical Subject Heading terms: multiple sclerosis,exercise,
physical activity,correlat*,facilitat*,barrier*, and impediment*
(supplemental appendix 2, available online only at http://www.
archives-pmr.org/). Reference lists of included articles and
search results from a systematic overview of PA research in
pwMS
27
were scanned for additional relevant articles. Two re-
viewers initially checked eligibility based on the article titles.
Subsequently, abstracts and, if necessary, the full texts were read
critically and discussed for clarification.
Risk of bias assessment
Two reviewers independently assessed the risk of bias (RoB) of
included articles. The tool for observational studies comprises
5 items that are scored with yes (1) or no (0).
40,41
A summary score is aggregated and interpreted as low (5
points), moderate (3e4) or high (0e2) (supplemental appendix 3,
available online only at http://www.archives-pmr.org/). For inter-
ventional studies, a modified 11-item tool was used, also scored
with yes (1) or no (0).
35,42,43
The aggregated summary score is
interpreted as low (9e11), moderate (5e8), or high RoB (0e4)
(supplemental appendix 4, available online only at http://www.
archives-pmr.org/).
List of abbreviations:
ICF International Classification of Functioning,Disability and
Health
MS multiple sclerosis
PA physical activity
pwMS persons with multiple sclerosis
RoB risk of bias
RRMS relapsing-remitting multiple sclerosis
634 R. Streber et al
www.archives-pmr.org
Data extraction
Data were extracted by 1 reviewer to predefined data extraction
forms, separately for observational (supplemental appendix 5,
available online only at http://www.archives-pmr.org/) and inter-
vention studies (supplemental appendix 6, available online only at
http://www.archives-pmr.org/). The second reviewer controlled
the extracted information with the original information and
adjusted mismatches accordingly.
Data regarding the study population, theoretical framework,
study design, examined correlates, PA measure, and results for an
association between the examined correlates and PAwere extracted.
For intervention studies, additional information regarding inter-
vention and control conditions and the results from the examined
statistical mediation model was extracted.
Categorization of variables
Examined variables from included studies were extracted
verbatim. For a systematic organization, the International Clas-
sification of Functioning, Disability and Health (ICF)
44
is regar-
ded as an adequate framework.
45-47
The evidence summary table
(supplemental appendix 7, available online only at http://www.
archives-pmr.org/) is subdivided according to the ICF compo-
nent functioning/disability and the ICF component contextual
factors (environmental and personal).
The functioning part was subdivided into a specific functioning
variables category and a generic functioning variables category.
The specific functioning variables include particular impairments,
activity limitations, and participation restrictions (eg, fatigue,
walking). However, some of the included studies applied multi-
dimensional measurements that assessed multiple ICF categories
of single or multiple functioning components and provided a
composite score. Therefore, a generic functioning variables cate-
gory was created. This category includes 4 different subcategories:
overall functioning level, impairment level, activity limitation, and
participation restriction level.
The environmental component was subdivided into physical/
natural and social environmental factors.
44
Regarding personal
factors, numerous health behavior theories were applied in the
included studies. Such theories contain similar constructs but use
different terminology.
48
Therefore, personal factors were divided
according to a structured overview of 5 different health behavior
theories
48
in the following manner: (1) self-efficacy, (2) attitudinal
beliefs and knowledge, (3) risk-related beliefs and emotional re-
sponses, (4) normative and norm-related beliefs and activities, and
(5) intention/commitment/planning. Sociodemographic/biologic
factors are an additional category. Where possible, constructs
with conceptual overlaps were combined in the same categori-
cal variable.
Data synthesis
For observational studies, evidence synthesis is based on a semi-
quantitative evaluation.
49
Evidence for an association between a
potential correlate with PA is described by its statistical signifi-
cance, direction, and consistency. In each included study, the re-
sults of the examined associations between a potential correlate
with PA were coded as significant, positive/inverse (þ/)oras
nonsignificant (0). Statistical significance and direction were
evaluated by the final statistical model used in the included
studies. For example, if bivariate correlations and a structural
equation model were concurrently available in a study, only the
results of the latter were extracted. Statistical models that were
adjusted for relevant covariates were used to evaluate the associ-
ation. The total scores of the independent variable measure were
considered to evaluate the association with PA unless subscales
were examined exclusively.
50,51
Consistency of results for the same variable across individual
studies was evaluated by a summary score. This score was coded
as 0 if a variable is rarely or not associated with PA, ? if the
association is inconsistent, and þ/if the variable is consis-
tently positively/inversely associated with PA (details in
supplemental appendix 7).
For intervention studies, the evidence synthesis is similar to the
approach for PA correlates and has also been applied in previous
reviews of PA interventions on mediators.
35,42
Results
Study selection
Fifty-six publication articles with an observational design and 2
intervention studies were included (fig 1). In 8 cases, findings
from an independent study were reported in >1 publication (see
fig 1). Another publication presented results from a secondary
analysis on a combined data set of various preceding in-
vestigations of 1 research group.
20
Study characteristics and RoB
The results from the 56 publications about observational studies are
based on 37 independent study populations with sample sizes from
11 to 3260 subjects (table 1, see supplemental appendix 5). Twenty-
eight studies (47 publications) were conducted in the United States.
Thirty-four studies (53 publications) included both sexes, whereas
3 studies included only women.
51-53
The mean age of subjects
ranged from 43 to 61 years. The mean disease duration of subjects
ranged from 8 to 24 years. Twenty-four observational studies
simultaneously investigated persons with progressive types of MS
and RRMS (33 publications), 3 studies focused exclusively on
persons with RRMS (12 publications), no study included only
persons with progressive types of MS, and 10 studies (11 publi-
cations) did not report any information regarding the disease
course. A variety of applied disability measures and statistical de-
scriptions regarding the functioning status of pwMS were reported.
The functioning status across studies ranged from essentially
normative to severe impairments. According to central tendency
values (mean or median), most study participants were fully
ambulatory up to 500m (corresponding to an Expanded Disability
Status Scale score 4.5 or a Patient Determined Disease Steps
score 3). One study focused exclusively on persons with a severe
disability level (Expanded Disability Status Scale range, 6e8).
54
In terms of RoB (see supplemental appendix 3) and study
characteristics (see supplemental appendix 5) of observational
studies, none of them exhibited a low RoB. In 37 articles, the RoB
was rated as moderate, and in 19 articles it was rated as high
(table 2). Forty (69%) publications
20,50,51,54-90
reported results
from a cross-sectional design, 9 (16%) publications reported re-
sults from a prospective design,
52,53,91-97
and 7 (14%) articles
reported prospective as well as cross-sectional findings.
98-104
Both
intervention studies additionally reported cross-sectional baseline
Physical activity in multiple sclerosis 635
www.archives-pmr.org
associations between proposed mediators and PA.
55,105,106
The
duration of follow-up in articles reporting prospective study de-
signs ranged from 1 month to 5 years. A follow-up of <6 months
was reported in 2 publications, a follow-up between 6 and 12
months was reported in 9 articles, and 5 publications reported a
follow-up of >1 year.
A theoretical framework was reported by approximately
50% of all articles with data from observational studies (see
table 2,supplemental appendix 5). Specifically, 17 articles re-
ported a health behavior theory,
51,52,55-59,64,71,73,76,77,86,91,93,95,98
9 publications reported functioning-/disability-related the-
ories,
63,67,68,70,75,80,94,99,101
and 1 publication applied an integrated
model of PA behavior and its relation with functioning and
disability
66
(see table 2). The degree to which the variables
explored were consistent with the theoretical models used is
presented in table 2.
Both included interventional studies applied an experimental
study design. RoB was rated as moderate
105,106
(see table 2,
supplemental appendix 4). Both studies were conducted in the
United States, had a sample size of 50 or 54 persons, included both
sexes, and were designed for pwMS being ambulatory with or
without minimal assistance, whereas only 1 study explicitly spec-
ified the inclusion of inactive pwMS
105,106
(see table 1,
supplemental appendix 6). One study compared group health ed-
ucation classes, including cognitive-behavioral strategies guided
by the theory of planned behavior and the social cognitive theory,
with an individually tailored physical rehabilitation program over 7
weeks with an 8-week follow-up.
105
The other intervention study
examined the effects of a 12-week Internet-based PA promotion
program based on the social cognitive theory
106
(see table 2).
Results from the observational studies
Overall, 86 different variables have been examined: 2 varia-
bles according to disease characteristics, 21 functioning-related
variables, 43 personal variables, and 20 environmental variables.
Table 3 presents an evidence summary, and supplemental
appendix 7 presents the summary and consistency of findings
across observational studies for every examined variable.
Thirty-one of 86 (36%) variables were coded as consistently
positively or inversely associated with PA, 14 (16%) were coded
as inconsistently associated with PA, and 41 (48%) were coded as
Duplicates removed
Total n=200
Cases of multiple manuscripts reporting
results from the same study population
Case References Case References
1 50, 76, 77, 78 5 67, 95, 97,
100, 102,
104
2 65, 75, 80 6 58, 59
3 73, 74 7 55, 106
4 72,94, 99,103 8 56, 91
Records identified through
electronic database search
PubMed n=268
Scopus n=413
Total n=681
Full-text manuscripts screened for
eligibility
by electronic search n=109
by manual search n=52
Total n=161
Records screened at title level
Total n=481
Records excluded
Total n=320
Full-text articles excluded (n=103)
Excluded observational studies (n=60)
PA not dependent variable (n=38)
Inadequate study design (n=1)
Evaluation of psychometric properties (n=12)
Qualitative research methods (n=7)
No full-text (nI=1)
Language (n=1)
Excluded Intervention studies (n=36)
PA not outcome (n=19)
No mediation analysis (n=11)
Qualitative research methods (n=3)
Study protocol (n=3)
Others (n=7)
Review (n=6)
Conceptual Paper (n=1)
Manuscripts of
observational studies n=56
intervention studies n=2
Total included n=58
Fig 1 Literature search process and reasons for exclusion. Literature search included studies from 1980 to January 31, 2015.
636 R. Streber et al
www.archives-pmr.org
not associated with PA (see supplemental appendix 7). Forty-two
of 86 (49%) variables were investigated once, 17 (20%) variables
were investigated twice, 5 (6%) variables were examined 3 times,
7 (8%) variables were examined 4 times, and 15 (17%) variables
were investigated >4 times.
Forty-seven articles
20,51,54-70,72-81,83-88,91,92,94-97,99-104
included at
least 1 functioning variable, 34 publications
20,51-59,63,64,66,69-71,73,
76-79,85-87,89-91,93-95,97,98,101,104
included at least 1 personal factor, and
7articles
50,51,66,76,82,95,101
included at least 1 environmental
factor. Five publications included variables of all ICF
components,
51,66,76,95,101
22 articles included variables of at least 2
components,
20,54-59,63,64,69,70,73,77-79,85-87,91,94,97,104
and29publica-
tions exclusively focused on variables from 1 ICF component.
Specifically, 20 articles investigated variables of functioning
components,
60-62,65,67,68,72,74,75,80,81,83,84,88,92,96,99,100,102,103
7 articles
exclusively focused on personal factors,
52,57,71,89,90,93,98
and 2
studies exclusively included environmental factors.
50,82
Functioning and health condition
Five generic functioning variables, 16 specific functioning variables,
and 2 disease-related variables were identified (see table 3,
supplemental appendix 7). Consistent inverse associations with PA
were found for the frequently examined variables (eg, overall func-
tioning/disability level,
57-60,62,66,74,81,83,85,86,96,97,101,103
impairment
level,
67,72-75,77,79,80,92,99
activity limitations level,
51,55,56,80,91,94,95,99
walking limitations
20,53,62,64, 75,78,79,83,85 ,96,102,104
). Also consistently
and inversely associated were different types of symptom clus-
ters,
68,94
apathy,
54
balance impairments,
62
and participa-
tion restrictions.
76,80,99
Positively associated were cognitive
functions
54,64,65
and information processing speed.
84,88
Inconsistent
findings were observed for overall fatigue,
54,60-62,64,66,67,69,75,100,104
depression levels,
54,66,67,75, 100,104
thermosensitivity,
69,78
and time
since diagnosis.
20,78,85-87,97,101
Not associated with PA were
pain,
67,70,75
sleep functions,
60
overall mental health level,
53
learning
and memory functions,
84
and type of MS.
20,66,69,78, 79,87
Personal level
Twenty-seven psychological variables from different health
behavior theories and 16 sociodemographic and biologic variables
were identified (see table 3,supplemental appendix 7).
Regarding self-efficacy, PA-related self-efficacy beliefs (eg,
barrier, exercise, maintenance, relapse) were the most frequently
examined and were consistently positively associated with
PA.
51,52,55-59,63,64,66,71,73,76,77,79,91,93-95,98,104,105
Also positively
associated with PA were self-efficacy related to household
chores,
63
falls-related efficacy,
53,62
and general self-efficacy.
54
Illness/symptom self-efficacy
63
and leisure time self-efficacy
63
were not associated with PA.
Regarding attitudinal beliefs and knowledge, perceived bar-
riers
57,59,63,66,71
and self-identity
105
were positively associated
with PA, whereas enjoyment of PA
66,76
and overall positive
outcome expectations
52,55,57,59,66,71,93
were inconsistently associ-
ated with PA. No association was evident for knowledge of ex-
ercise health effects
66
or for negative,
52
physical,
56,86,91,95
self-
evaluative,
56,86,91,95
and social outcome expectations.
56,86,91,95
The 2 variables of normative and norm-related beliefs
57,66
and the 3 variables of risk-related beliefs and emotional
responses
57-59,63,66
were not associated with PA.
Table 1 Summary of study population characteristics of included studies
Characteristic Observational Studies Intervention Studies Total
No. of independent study populations 37 (100) 2 (100) 39 (100)
Country
Australia 1 (3) 0 (0) 1 (3)
Canada 1 (3) 0 (0) 1 (3)
Germany 2 (5) 0 (0) 1 (5)
Netherland 2 (5) 0 (0) 2 (5)
New Zealand 1 (3) 0 (0) 1 (3)
Belgium 1 (3) 0 (0) 1 (3)
Sweden 1 (3) 0 (0) 1 (3)
United States 28 (76) 2 (100) 23 (85)
Sex
Women only 3 (8) 0 (0) 3 (8)
Men and women 34 (92) 2 (100) 36 (92)
Disability measures
EDSS 14 (38) 0 (0) 14 (36)
PDDS/DS 7 (19) 1 (50) 8 (21)
Ambulation status 6 (16) 0 (0) 6 (15)
Miscellaneous 8 (22) 1 (50) 9 (23)
Not reported 2 (5) 0 (0) 2 (5)
Disease courses
RRMS 3 (8) 1 (50) 4 (10)
PPMS/SPMS 0 (0) 0 (0) 0 (0)
Mixed*24 (65) 0 (0) 24 (62)
Not reported 10 (27) 1 (50) 11 (28)
NOTE. Values are n (%).
Abbreviations: DS, disease steps; EDSS, Expanded Disability Status Scale; PDDS, Patient Determined Disease Steps; PPMS, primary-progressive multiple
sclerosis; SPMS, secondary-progressive multiple sclerosis.
* Studies recruited persons with RRMS and persons with PPMS/SPMS.
Physical activity in multiple sclerosis 637
www.archives-pmr.org
Regarding intention/commitment/planning variables, exercise
goal setting,
55,56,91,95
action and coping planning,
58,59
behavioral
and cognitive processes of change,
52,64
and intentions
59
were consistently positively associated with PA, but stage of
change
66
and expectations
105
were not. Higher PA levels were
consistently observed for pwMS who were employed
(vs unemployed)
20,78,85,87
and had a higher education
level.
20,70,78,86,87
Few studies reported that receiving regular
physical therapy
85
and living alone
70
were associated with higher
levels of PA, whereas receiving a disability pension
66
was asso-
ciated with lower levels of PA. Age,
66,70,78,85,86,97,101
parental
status,
66,78
history of falls,
53,89
and body mass index
64,70
showed
inconsistent associations with PA. Sex,
20,57,66,69,70,78,85-87,97,101
income,
20,78
marital status,
70,78,86
race,
20,70,78,85,87
medical co-
morbidity,
66,70
receiving any kind of disease-modifying pharma-
cologic treatment,
85
and being a member in a patient
organization
66
were not related to PA.
Environmental level
Four studies addressed 15 variables of the physical environ-
ment,
50,51,82,101
and 6 studies addressed 5 variables of the social
environment
50,51,66,76,95,105
(see table 3,supplemental appendix 7).
Proximity to transit stops was positively related to PA,
50
whereas
type of residency
82,101
showed inconsistent findings. General so-
cial support had an indirect effect via exercise self-efficacy on
PA
76
; however, exercise-related social support was rarely related
with PA.
66,95,105
No other physical or social environmental vari-
able was associated with PA.
Results from the intervention studies
Seven of all 9 examined mediators were personal factors; 1 study
included a functioning variable and 1 study including a social
environmental factor (see supplemental appendix 6).
105,106
PA-
related self-efficacy was investigated twice, whereas the remain-
ing variables were examined once.
The 2 included intervention studies showed significant positive
changes of PA
105,106
(see supplemental appendix 6). However,
Plow et al
105
could detect within-group differences but could not
detect between-group differences for PA change. In terms of ac-
tion theory, exercise goal setting
106
and self-identity
105
signifi-
cantly increased, whereas expectations
105
significantly decreased
after the intervention. Inconsistent results were found for PA-
related self-efficacy.
105,106
Both studies reported a decrease in
Table 2 Summary of study characteristics of included articles
Characteristic Observational Studies Intervention Studies Total
Completeness of
Applied Theory
(completely/partly)*
No. of included articles 56 (100) 2 (100) 58 (100)
Theoretical framework
HAPA 2 (4) 0 (0) 2 (3) 2/0
Health promotion model 1 (2) 0 (0) 1 (2) 0/1
HBM 1 (2) 0 (0) 1 (2) 1/0
ICF 1 (2) 0 (0) 1 (2) 0/1
Nagi disablement model extended by
Verbrugge and Jette
3 (5) 0 (0) 3 (5) 2/1
PAD model 1 (2) 0 (0) 1 (2) 1/0
SCT 11 (20) 2 (100) 13 (22) 5/8
Theory of unpleasant symptoms 4 (7) 0 (0) 4 (7) 4/0
TTM 2 (4) 0 (0) 2 (3) 2/0
Miscellaneous 1 (2) 0 (0) 1 (2) NA
Not reported 29 (52) 0 (0) 29 (50) NA
PA measure
Self-report 29 (52) 2 (100) 31 (53) NA
Objective 16 (29) 0 (0) 16 (28) NA
Objective and self-report 11 (20) 0 (0) 11 (19) NA
RoB
High 19 (34) 0 (0) 19 (33) NA
Moderate 37 (66) 1 (50) 38 (66) NA
Low 0 (0) 1 (50) 1 (2) NA
Study design
Cross-sectional 40 (71) NA 40 (69) NA
Prospective 9 (16) NA 9 (16) NA
Cross-sectional and prospective 7 (13) NA 7 (12) NA
RCT NA 2 (100) 2 (3) NA
NOTE. Values are n (%) or as otherwise indicated.
Abbreviations: HAPA, health action process approach; HBM, health belief model; NA, not applicable; PAD, physical activity for people with a disability
model; RCT, randomized controlled trial; SCT, social cognitive theory; TTM, transtheoretical model of behavior change.
* Degree to which the explored variables were consistent with the theoretical models used: completely, studies measured and tested the theory in its
entirety; partly, studies focused on selected constructs of the referred theoretical model.
638 R. Streber et al
www.archives-pmr.org
Table 3 Evidence summary table of examined variables across included observational studies
Consistency*
Functioning- and
Disease-Related Variables Personal Variables Environmental Variables
4 studies consistently supported the same
positive/inverse association with PA
Functioning/disability level
Impairment level
Activity limitations
Walking limitations
PA-related self-efficacy
Self-regulation and planning (including planning,
goal setting, behavioral and cognitive strategies)
Employment status
Education level
NA
Variable is consistently positively/inversely
associated with PA but was less frequently
examined
Particpation restrictions
Symptom cluster
Mental fatigue
Apathy
Balance
Cognitive functions
Information processing speed
Perceived barriers
Household and general self-efficacy
Falls-related efficacy
Self-identity
PA intention
Receive physical therapy
Disability pension
Living alone
Proximity to transit stops
Social support (general)
Frequently examined variable with
considerable lack of consistency in findings
Time since diagnosis
Fatigue
Positive outcome expectations
Age
NA
Results for this association are inconsistent
but these variables were less frequently
examined
Thermosensitivity
Depression
Muscle strength
Enjoyment of PA
History of falls
Body mass index
Parental status
Type of residence
4 studies consistently showed no evidence
for an association for this variable with PA
Type of MS Sex
Race
NA
Variable is rarely or not associated with PA
but was less frequently examined
Mental health
Physical fatigue
Pain
Learning and memory
Sleep functions
Negative outcome expectations
Self-evaluative outcome expectations
Physical outcome expectations
Social outcome expectations
Leisure self-efficacy
Illness- and symptom-related self-efficacy
Knowledge of exercise effects
Cues to action
Motivation to adhere to normative beliefs
Risk perception
Perceived seriousness of unhealthy consequences
of inactivity
Cognitive health beliefs and illness behaviors
Exercise stage of change
Expectation
Receive disease-modifying pharmacologic
treatment
Medical comorbidity
Marital status
Income
Exercise-related social support
Residential density
Presence of bicycle facilities
Land use/diversity
Presence of low-cost recreation service
Access to services
Access to amenities
Presence of heavy traffic
Safety from traffic
Street connectivity
Presence of sidewalks
Neighborhood aesthetics
Neighborhood satisfaction
No. of motor vehicles at home
Normative expectations/beliefs of family
members and friends
Presence of active others
Safety from crime
Abbreviation: NA, not applicable.
* Consistency of results for the same variable across individual studies was evaluated by a summary score (see supplemental appendix 7).
Physical activity in multiple sclerosis 639
www.archives-pmr.org
self-efficacy; however, this was significant in only 1 study. No
action theory link was observed for activity limitations; self-
evaluative, physical, or social outcome expectations; and
exercise-related social support. Although for the latter, an almost
significant reduction was reported. In terms of the conceptual
theory, improvements in goal setting were associated with an in-
crease in PA behavior in a single and multiple mediator anal-
ysis.
106
The relation of increased expectations was only supported
by the single but not the multiple mediator analysis.
105
No support
was found for the different types of examined outcome expecta-
tions, PA-related self-efficacy, activity limitations, self-identity, or
exercise-related social support. Evidence for a mediated inter-
vention effect on PA was only observed for goal setting.
Discussion
Recent research highlights the beneficial effects of PA,
12,13,107
re-
ports that physical inactivity is common,
19-21,108
and highlights the
need for a systematic evaluation of PA behavior correlates in
pwMS.
27
This review systematically evaluates the evidence for
relations between disease-related, functioning, personal, and envi-
ronmental variables on PA behavior in pwMS.
Observational studies revealed consistent associations between
PA and disability level, (particularly walking limitations), PA-
related self-efficacy, and self-regulation constructs. A person’s
education level and employment status were the most consistent
sociodemographic correlates of PA.
Beside consistency of associations, several studies provided
evidence for the independent effect of each identified factor.
For example, studies that incorporated disability measures,
self-efficacy, and self-regulatory constructs consistently reported
their independent effects on PA.
56,58,59,64,91,95
Dlugonski et al
87
reported independent effects of employment and education con-
trolling for other demographic and disability-related factors.
Employment but not education was a significant predictor in another
study that simultaneously evaluated these 2 factors.
78
Nevertheless,
several other studies that accounted for disability- or person-related
factors supported the independent effect of employment
20,85
or
education.
20,70,86
Considering that most included studies relied on observa-
tional study designs, these identified variables are rather statis-
tical associations (ie, correlates) than determinants of PA in a
strict sense. One intervention study provided evidence for a
causal relation between self-regulation and PA, which is
congruent with the synthesized results of the observational
studies included in the current review. This is also supported by a
systematic review of mediators of PA behavior change in
nonclinical populations
42
and aligns with a broad range of the-
ories/models applied in PA research.
109,110
This supports self-
regulatory constructs as a determinant and potential target to
change PA in pwMS.
In line with other populations with long-term neurologic con-
ditions,
45,111
the general disability is a major factor affecting PA
levels in pwMS; however, these variables are many-faceted and do
not provide a concrete target for PA interventions.
68
No other spe-
cific factor than walking limitations provided consistent evidence
across several studies. However, there are at present too few results
for most of these functioning variables to draw a definite conclusion.
Additionally, several factors (eg, fatigue, depression) showed
heterogeneous results. This may be explained because of meth-
odologic differences (eg, statistical analysis, multidimensional vs
unidimensional questionnaires, study designs) or different sample
characteristics. Furthermore, pwMS face a broad range of multiple
impairments
7-9
; it might be speculated that the presence of multiple,
specific impairments (as a symptom cluster) provides a better un-
derstanding of PA variations compared with the examination of
single impairments. Currently, only a few examinations have been
done in this direction.
68,94
The importance of PA-related self-efficacy and self-regulation
skills is in line with social-cognitive theories of health
behavior,
112-114
with empirical evidence in apparently healthy
adults
25
and other chronic conditions.
45,111,115-118
Beliefs of
beneficial or adverse effects as a consequence of PA are central
constructs in social-cognitive theories.
119
However, the inconsis-
tent associations between outcome expectations with PA are in
line with results across various systematic reviews of PA correlates
in various populations.
25,120
None of the subtypes of outcome
expectations except perceived barriers were consistently associ-
ated with PA. Reasons might be the heterogonous methodology
and the sample characteristics of included studies. Williams
et al
119
recommended to explore the direct and indirect relation of
outcome expectations with PA, consider potential moderators, and
broaden the concept. Finally, PA-related self-efficacy, self-
regulation, and perception of barriers are potential targets for PA
interventions.
In contrast with healthy persons,
25
in this review, age and sex
were inconsistently associated with PA among pwMS. These
differences are not explained because of simple methodologic
differences of the included studies. Based on the results of the
included studies, we suggest that PA levels rather depend on an
individual’s disability level and personal characteristics (eg, self-
efficacy) than on age or sex per se. PA is also affected by
cohort or period effects (eg, generational differences in PA so-
cialization).
121,122
For example, health has become an important
PA motive, and in particular, older adults tend to start exercising
for health reasons.
121,122
In the current review, few environmental variables were found
to be associated with PA of pwMS. Only proximity to transit stops
was associated with PA,
50
and inconsistent associations were
evident for type of residency
82,101
and social support.
66,76,95,105
In
healthy adults, social support, transportation, recreation facilities/
locations, and aesthetic environmental variables were consistent
correlates for total PA.
25
Environmental aspects (eg, lack of
accessible or convenient recreation/exercise facilities,
54,57,63,71
lack of information and assistance from fitness or health care
professionals according to PA,
54,63
financial issues
54
) appear as
perceived barriers to exercise in pwMS. Future research needs to
explore the role of environmental factors more in depth.
Study limitations
Reliance on cross-sectional designs, application of self-report PA
measures, and convenience samples are main limitations across
the studies. Persons with a relapsing-remitting disease course and
mild to moderate disability levels were predominantly recruited.
Most studies were conducted in the United States. Therefore,
research should expand investigations to persons with progressive
types of MS or severe limitations, consider cultural differences,
apply long-term longitudinal studies, and increase the use of
objective PA measures.
At the variable level, the lack of a credible number of findings
per variable and the inconsistent results regarding several exam-
ined variables are major limitations. There is a need to clarify the
640 R. Streber et al
www.archives-pmr.org
role of variables with heterogeneous results and of variables with
few study findings considering possible mediating, moderating, or
confounding factors.
Various theoretical approaches have been chosen to understand
PA participation in pwMS. One major shortcoming for explaining
PA variation is the focus on single variables (eg, self-efficacy)
rather than validating complete theories. Testing the complete
theory is seen as a critical step in this respect and a basis to adapt
or augment them.
123
Long-term longitudinal studies of change and experimental
studies are needed. Experiments would demonstrate the strongest
evidence for a causal relation of a theory’s constructs with PA.
However, only 2 interventional studies with a mediation analysis
were included.
105,106
Interestingly, most of the examined variables
were not altered by the intervention, and in some cases, the
intervention effects on proposed determinants were in the opposite
direction to those expected.
105,106
One reason might be action
theory failure. The contrary findings do not disprove the meaning
of the investigated constructs as potential determinants of PA
behavior in pwMS but imply a need for a systematic examination
of how to affect potential mechanisms of PA behavior change.
124-
126
Furthermore, none of the available but excluded theory-based
PA intervention programs for pwMS examined the proposed
causal mechanisms of change.
127-134
To develop, advance, and
disseminate effective and efficient PA promotion programs for
pwMS, the causal pathway from intervention through mediators to
PA behavior change needs to be understood.
31,34
Most of the research focused on social-cognitive theory. The
utility of other models should be further explored: models that
have been studied to a limited extent (eg, health action process
approach
112
), are rather motivation-focused (eg, the self-
determination theory
135
), or were recently developed in the PA
domain
109,110
(eg, integrated behavior change model for PA,
114
physical activity maintenance theory
136
). Additionally, subtypes
of constructs (eg, affective components of outcome expecta-
tions,
41,42
subtypes of PA-related self-efficacy
112
) should be
further explored. Relapse self-efficacy may be important, espe-
cially considering the progressive nature of MS and the occur-
rence of relapses.
59,112
(Socio-)ecological models of health
behavior
137,138
and disability and health
44
could help to broaden
the perspective from the individual factors to social and physical
environment or to policy factors. This might also help to deepen
the understanding of the influence of physical and social envi-
ronmental factors and identify resources and opportunities that act
as facilitators for PA participation among pwMS.
The strengths of this review are the consideration of results
from observational and interventional studies, the RoB assess-
ment, the application of a systematic approach to quantify the
consistency of results, and the usage of the ICF to systemize re-
sults. However, it cannot be ruled out that the results are subject to
publication and selection bias. Another limitation is that the
applied evidence synthesis approach does not take into account the
strength of relations. Hence, conclusions can only be made about
the consistency of results across studies. The results are limited to
findings acquired with quantitative research methods. Qualitative
research methods are also helpful to deepen the understanding of
PA behavior in pwMS.
139,140
Practical implications
This review provides information that can be used in clinical care
and practice by health professionals and in the development of
programs to increase PA participation among pwMS.
24,25,30
It
might also inform decisions on whom to target.
Subgroups at risk for being inactive are persons with a higher
disability level, especially those with severe walking limitations,
low self-regulatory skills, low self-efficacy, a high perception of
barriers, and who are unemployed or with a low level of educa-
tion. These groups constitute potential target groups for PA in-
terventions. The diversity of factors and possible combinations
imply the need of tailored, person-oriented PA programs instead of
one-size-fits-all approaches.
To strengthen PA-related self-efficacy and self-regulatory
skills
141
as potential targets for PA interventions, certain
behavior change techniques
126
have been found to be efficacious
in adults with and without diseases.
110,124,125,142
For example, the
provision of specific instructions on where, when, and how to
perform health-enhancing PA and reinforcing effort or progress
toward behavior and action planning were recently identified as
effective techniques to increase self-efficacy and PA in healthy
adults.
124
Enabling pwMS to set, monitor, and evaluate goals; to
cope with barriers; and to plan actions in terms of what, when,
where, and how are examples of how to promote self-regulatory
skills.
110,143
It should be an important target to improve, maintain, or
restore an existing functioning level, especially walking limita-
tions. Exercise itself has been proven to be an effective treatment
to improve walking capacities of pwMS with mild to moderate
disability levels.
13,15,144
Furthermore, considering that gait and
balance impairments have been identified even in the absence of
a clinically meaningful disability level,
145-147
it can be argued
that PA promotion should start very early after diagnosis.
28
Beyond, providing information and opportunities and teaching
skills on how to adapt PA to the individual’s needs and limita-
tions could help pwMS to stay or become more physi-
cally active.
110
Most important, taken into account the health benefits of PA
and the PA prevalence among pwMS, results of this review
support the suggestion that PA behavior change interventions
should be integrated into the care of persons
110
and over the
course of the disease.
28
Additionally, such information implies
the need to develop low-cost or affordable programs and to in-
crease the accessibility of training facilities in the community
(eg, fitness centers). Home-based or especially Internet-based
programs could be a promising intervention modality to pro-
mote PA outside of clinical care.
106
Different models of financial
support as facilitators for people with disabilities or low socio-
economic status should be discussed (eg, for member fee,
transportation services).
Conclusions
This review expands on previous research by systematic evalua-
tion of the correlates and determinants of PA among pwMS. The
research highlights the importance of personal factors (eg, PA-
related self-efficacy and self-regulatory constructs) and func-
tioning variables (eg, walking limitations). However, current
research has several limitations. Research that is more rigorous is
needed to better understand what affects PA in pwMS. The results
of this review can build the basis for further research regarding
correlates and determinants of PA and might inform decision-
making in the development and advancement of PA programs
for pwMS.
Physical activity in multiple sclerosis 641
www.archives-pmr.org
Keywords
Health behavior; Health promotion; Motor activity; Multiple
sclerosis; Rehabilitation
Corresponding author
Klaus Pfeifer, PhD, Institute of Sport Science and Sport, Friedrich-
Alexander-University Erlangen-Nu
¨rnberg, Gebbertstrasse 123b,
91058 Erlangen, Germany. E-mail address: Klaus.pfeifer@fau.de.
Acknowledgment
We thank Ulrike Hartz, MA, for her support regarding data
extraction and risk of bias assessment.
References
1. Compston A, Coles A. Multiple sclerosis. Lancet 2008;372:1502-17.
2. Confavreux C, Vukusic S, Moreau T, Adeleine P. Relapses and
progression of disability in multiple sclerosis. N Engl J Med 2000;
343:1430-8.
3. Trapp BD, Nave K. Multiple sclerosis: an immune or neurodegen-
erative disorder? Annu Rev Neurosci 2008;31:247-69.
4. Alonso A, Hernan MA. Temporal trends in the incidence of multiple
sclerosis: a systematic review. Neurology 2008;71:129-35.
5. Degenhardt A, Ramagopalan SV, Scalfari A, Ebers GC. Clinical
prognostic factors in multiple sclerosis: a natural history review. Nat
Rev Neurol 2009;5:672-82.
6. Tremlett H, Zhao Y, Rieckmann P, Hutchinson M. New perspectives in
the natural history of multiple sclerosis. Neurology 2010;74:2004-15.
7. Berno S, Coenen M, Leib A, Cieza A, Kesselring J. Validation of the
Comprehensive International Classification of Functioning,
Disability, and Health Core Set for multiple sclerosis from the
perspective of physicians. J Neurol 2012;259:1713-26.
8. Conrad A, Coenen M, Schmalz H, Kesselring J, Cieza A. Validation
of the comprehensive ICF core set for multiple sclerosis from the
perspective of physical therapists. Phys Ther 2012;92:799-820.
9. Coenen M, Cieza A, Freeman J, et al. The development of ICF Core
Sets for multiple sclerosis: results of the International Consensus
Conference. J Neurol 2011;258:1477-88.
10. Benito-Leo
´n J, Morales JM, Rivera-Navarro J, Mitchell AJ. A review
about the impact of multiple sclerosis on health-related quality of
life. Disabil Rehabil 2003;25:1291-303.
11. Mitchell AJ, Benito-Leo
´nJ,Gonza
´lez JM, Rivera-Navarro J. Quality of
life and its assessment in multiple sclerosis: integrating physical and
psychological components of wellbeing. Lancet Neurol 2005;4:556-66.
12. Latimer-Cheung AE, Martin Ginis KA, Hicks AL, et al. Develop-
ment of evidence-informed physical activity guidelines for adults
with multiple sclerosis. Arch Phys Med Rehabil 2013;94:1829-36.
13. Latimer-Cheung AE, Pilutti LA, Hicks AL, et al. Effects of exercise
training on fitness, mobility, fatigue, and health-related quality of life
among adults with multiple sclerosis: a systematic review to inform
guideline development. Arch Phys Med Rehabil 2013;94:1800-28.
14. Kuspinar A, Rodriguez AM, Mayo NE. The effects of clinical in-
terventions on health-related quality of life in multiple sclerosis: a
meta-analysis. Mult Scler 2012;18:1686-704.
15. Pearson M, Dieberg G, Smart N. Exercise as a therapy for
improvement of walking ability in adults with multiple sclerosis: a
meta-analysis. Arch Phys Med Rehabil 2015;96:1339-48.e7.
16. Pilutti LA, Greenlee TA, Motl RW, Nickrent MS, Petruzzello SJ.
Effects of exercise training on fatigue in multiple sclerosis: a meta-
analysis. Psychosom Med 2013;75:575-80.
17. Pilutti LA, Platta ME, Motl RW, Latimer-Cheung AE. The safety of
exercise training in multiple sclerosis: a systematic review. J Neurol
Sci 2014;343:3-7.
18. Dalgas U, Stenager E. Exercise and disease progression in multiple
sclerosis: can exercise slow down the progression of multiple scle-
rosis? Ther Adv Neurol Disord 2012;5:81-95.
19. Motl RW, McAuley E, Snook EM. Physical activity and multiple
sclerosis: a meta-analysis. Mult Scler 2005;11:459-63.
20. Klaren RE, Motl RW, Dlugonski D, Sandroff BM, Pilutti LA.
Objectively quantified physical activity in persons with multiple
sclerosis. Arch Phys Med Rehabil 2013;94:2342-8.
21. Motl RW, McAuley E, Sandroff BM, Hubbard EA. Descriptive
epidemiology of physical activity rates in multiple sclerosis. Acta
Neurol Scand 2015;131:422-5.
22. Lee IM, Shiroma EJ, Lobelo F, Puska P, Blair SN, Katzmarzyk PT.
Effect of physical inactivity on major non-communicable diseases
worldwide: an analysis of burden of disease and life expectancy.
Lancet 2012;380:219-29.
23. Garber CE, Blissmer B, Deschenes MR, et al. American College of
Sports Medicine position stand. Quantity and quality of exercise for
developing and maintaining cardiorespiratory, musculoskeletal, and
neuromotor fitness in apparently healthy adults: guidance for pre-
scribing exercise. Med Sci Sports Exerc 2011;43:1334-59.
24. Sallis JF, Owen N, Fotheringham MJ. Behavioral epidemiology: a
systematic framework to classify phases of research on health pro-
motion and disease prevention. Ann Behav Med 2000;22:294-8.
25. Bauman AE, Reis RS, Sallis JF, et al. Correlates of physical activity:
why are some people physically active and others not? Lancet 2012;
380:258-71.
26. Motl RW. Physical activity and its measurement and determinants in
multiple sclerosis. Minerva Med 2008;99:157-65.
27. Dixon-Ibarra A, Vanderbom K, Dugala A, Driver S. Systematic
framework to evaluate the status of physical activity research for
persons with multiple sclerosis. Disabil Health J 2014;7:151-6.
28. Ellis T, Motl RW. Physical activity behavior change in persons with
neurologic disorders: overview and examples from Parkinson disease
and multiple sclerosis. J Neurol Phys Ther 2013;37:85-90.
29. Motl RW, Learmonth YC, Pilutti LA, Gappmaier E, Coote S. Top 10
research questions related to physical activity and multiple sclerosis.
Res Q Exerc Sport 2015;86:117-29.
30. Bauman AE, Sallis JF, Dzewaltowski DA, Owen N. Toward a better
understanding of the influences on physical activity: the role of de-
terminants, correlates, causal variables, mediators, moderators, and
confounders. Am J Prev Med 2002;23(2 Suppl 1):5-14.
31. Noar SM, Mehrotra P. Toward a new methodological paradigm for
testing theories of health behavior and health behavior change. Pa-
tient Educ and Couns 2011;82:468-74.
32. Weinstein ND. Misleading tests of health behavior theories. Ann
Behav Med 2007;33:1-10.
33. MacKinnon DP, Luecken LJ. Statistical analysis for identifying
mediating variables in public health dentistry interventions. J Public
Health Dent 2011;71(Suppl 1):S37-46.
34. Baranowski T, Anderson C, Carmack C. Mediating variable frame-
work in physical activity interventions: how are we doing? How
might we do better? Am J Prev Med 1998;15:266-97.
35. Cerin E, Barnett A, Baranowski T. Testing theories of dietary
behavior change in youth using the mediating variable model with
intervention programs. J Nutr Educ Behav 2009;41:309-18.
36. Chen H. Theory-driven evaluations. Newbury Park: Sage; 1990.
37. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for
reporting systematic reviews and meta-analyses of studies that
evaluate health care interventions: explanation and elaboration. J
Clin Epidemiol 2009;62:e1-34.
38. Caspersen CJ, Powell KE, Christenson G. Physical activity, exercise
and physical fitness: definitions and distinctions for health-related
research. Public Health Rep 1985;100:126-31.
39. Centers for Disease Control and Prevention (CDC). Physical activity
guidelines for Americans. Okla Nurse 2008;53:25.
642 R. Streber et al
www.archives-pmr.org
40. Nasuti G, Rhodes RE. Affective judgment and physical activity
in youth: review and meta-analyses. Ann Behav Med 2013;45:
357-76.
41. Rhodes R, Fiala B, Conner M. A review and meta-analysis of af-
fective judgments and physical activity in adult populations. Ann
Behav Med 2009;38:180-204.
42. Rhodes RE, Pfaeffli LA. Mediators of physical activity behaviour
change among adult non-clinical populations: a review update. Int J
Behav Nutr Phys Act 2010;7:37.
43. Lubans DR, Foster C, Biddle, Stuart JH. A review of mediators of
behavior in interventions to promote physical activity among chil-
dren and adolescents. Prev Med 2008;47:463-70.
44. World Health Organization. International Classification of Func-
tioning, Disability and Health: ICF. Geneva: World Health Organi-
zation; 2001.
45. Fekete C, Rauch A. Correlates and determinants of physical activity
in persons with spinal cord injury: a review using the International
Classification of Functioning, Disability and Health as reference
framework. Disabil Health J 2012;5:140-50.
46. Rimmer JH. Use of the ICF in identifying factors that impact
participation in physical activity/rehabilitation among people with
disabilities. Disabil Rehabil 2006;28:1087-95.
47. Nieuwenhuijsen ER, Zemper E, Miner KR, Epstein M. Health
behavior change models and theories: contributions to rehabilitation.
Disabil Rehabil 2006;28:245-56.
48. Noar SM, Zimmerman RS. Health Behavior Theory and cumulative
knowledge regarding health behaviors: are we moving in the right
direction? Health Educ Res 2005;20:275-90.
49. Sallis JF, Prochaska JJ, Taylor WC. A review of correlates of phys-
ical activity of children and adolescents. Med Sci Sports Exerc 2000;
32:963-75.
50. Doerksen SE, Motl RW, McAuley E. Environmental correlates of
physical activity in multiple sclerosis: a cross-sectional study. Int J
Behav Nutr Phys Act 2007;4:49.
51. Morris KS, McAuley E, Motl RW. Self-efficacy and environmental
correlates of physical activity among older women and women with
multiple sclerosis. Health Educ Res 2008;23:744-52.
52. Levy SS, Li K, Cardinal BJ, Maddalozzo GF. Transitional shifts in
exercise behavior among women with multiple sclerosis. Disabil
Health J 2009;2:216-23.
53. Kasser SL, Jacobs JV, Littenberg B, Foley JT, Cardinal BJ,
Maddalozzo GF. Exploring physical activity in women with multiple
sclerosis: associations with fear of falling and underlying impair-
ments. Am J Phys Med Rehabil 2014;93:461-9.
54. Vanner EA, Block P, Christodoulou CC, Horowitz BP, Krupp LB.
Pilot study exploring quality of life and barriers to leisure-time
physical activity in persons with moderate to severe multiple scle-
rosis. Disabil Health J 2008;1:58-65.
55. Dlugonski D, Wo
´jcicki TR, McAuley E, Motl RW. Social cognitive
correlates of physical activity in inactive adults with multiple scle-
rosis. Int J Rehabil Res 2011;34:115-20.
56. Suh Y, Weikert M, Dlugonski D, Sandroff B, Motl RW. Social
cognitive correlates of physical activity: findings from a cross-
sectional study of adults with relapsing-remitting multiple scle-
rosis. J Phys Act Health 2011;8:626-35.
57. Kasser SL, Kosma M. Health beliefs and physical activity behavior
in adults with multiple sclerosis. Disabil Health J 2012;5:254-60.
58. Chiu C, Fitzgerald SD, Strand DM, Muller V, Brooks J, Chan F.
Motivational and volitional variables associated with stages of
change for exercise in multiple sclerosis: a multiple discriminant
analysis. Rehabil Couns Bull 2012;56:23-33.
59. Chiu C, Lynch RT, Chan F, Berven NL. The Health Action Process
Approach as a motivational model for physical activity self-
management for people with multiple sclerosis: a path analysis.
Rehabil Psychol 2011;56:171-81.
60. Merkelbach S, Schulz H, Ko
¨lmel HW, et al. Fatigue, sleepiness, and
physical activity in patients with multiple sclerosis. J Neurol 2011;
258:74-9.
61. Rietberg MB, van Wegen EE, Uitdehaag BM, Kwakkel G. The as-
sociation between perceived fatigue and actual level of physical ac-
tivity in multiple sclerosis. Mult Scler 2011;17:1231-7.
62. Cavanaugh JT, Gappmaier VO, Dibble LE, Gappmaier E. Ambula-
tory activity in individuals with multiple sclerosis. J Neurol Phys
Ther 2011;35:26-33.
63. Kayes NM, McPherson KM, Schluter P, Taylor D, Leete M, Kolt GS.
Exploring the facilitators and barriers to engagement in physical ac-
tivity for people with multiple sclerosis. Disabil Rehabil 2011;33:
1043-53.
64. Plow MA, Finlayson M, Cho C. Correlates of stages of change for
physical activity in adults with multiple sclerosis. Res Nurs Health
2011;34:378-88.
65. Prakash RS, Snook EM, Kramer AF, Motl RW. Correlation of
physical activity with perceived cognitive deficits in relapsing-
remitting multiple sclerosis. Int J MS Care 2010;12:1-4.
66. Beckerman H, de Groot V, Scholten MA, Kempen JC, Lankhorst GJ.
Physical activity behavior of people with multiple sclerosis: under-
standing how they can become more physically active. Phys Ther
2010;90:1001-13.
67. Motl RW, McAuley E, Wynn D, Suh Y, Weikert M, Dlugonski D.
Symptoms and physical activity among adults with relapsing-
remitting multiple sclerosis. J Nerv Ment Dis 2010;198:213-9.
68. Motl RW, Weikert M, Suh Y, Dlugonski D. Symptom cluster and
physical activity in relapsing-remitting multiple sclerosis. Res Nurs
Health 2010;33:398-412.
69. Fjeldstad C, Brittain DR, Fjeldstad AS, Pardo G. Fatigue and thermo
sensitivity affect physical activity in multiple sclerosis. J Appl Res
2010;10:99-106.
70. Turner AP, Kivlahan DR, Haselkorn JK. Exercise and quality of life
among people with multiple sclerosis: looking beyond physical
functioning to mental health and participation in life. Arch Phys Med
Rehabil 2009;90:420-8.
71. Stroud N, Minahan C, Sabapathy S. The perceived benefits and
barriers to exercise participation in persons with multiple sclerosis.
Disabil Rehabil 2009;31:2216-22.
72. Motl RW, Snook EM, Schapiro RT. Neurological impairment as
confounder or moderater of association between symptoms and
physical activity in multiple sclerosis. Int J MS Care 2008;10:
99-105.
73. Snook EM, Motl RW. Physical activity behaviors in individuals with
multiple sclerosis: roles of overall and specific symptoms, and self-
efficacy. J Pain Symptom Manage 2008;36:46-53.
74. Motl RW, Snook EM, Wynn DR, Vollmer T. Physical activity cor-
relates with neurological impairment and disability in multiple
sclerosis. J Nerv Ment Dis 2008;196:492-5.
75. Motl RW, Snook EM, Schapiro RT. Symptoms and physical activity
behavior in individuals with multiple sclerosis. Res Nurs Health
2008;31:466-75.
76. Motl RW, Snook EM, McAuley E, Scott JA, Douglass ML. Corre-
lates of physical activity among individuals with multiple sclerosis.
Ann Behav Med 2006;32:154-61.
77. Motl RW, Snook EM, McAuley E, Gliottoni RC. Symptoms, self-
efficacy, and physical activity among individuals with multiple
sclerosis. Res Nurs Health 2006;29:597-606.
78. Motl RW, Snook EM, McAuley E, Scott JA, Hinkle ML. De-
mographic correlates of physical activity in individuals with multiple
sclerosis. Disabil Rehabil 2007;29:1301-4.
79. Motl RW, Snook EM, Wynn D. Physical activity behavior in in-
dividuals with secondary progressive multiple sclerosis. Int J MS
Care 2007;9:139-42.
80. Motl RW, Snook EM, McAuley E, Scott JA, Gliottoni RC. Are
physical activity and symptoms correlates of functional limitations
and disability in multiple sclerosis? Rehabil Psychol 2007;52:463-9.
81. Gulick EE, Goodman S. Physical activity among people with mul-
tiple sclerosis. Int J MS Care 2006;8:121-9.
82. Stuifbergen AK. Barriers and health behaviors of rural and urban
persons with MS. Am J Health Behav 1999;23:415-25.
Physical activity in multiple sclerosis 643
www.archives-pmr.org
83. Gijbels D, Alders G, van Hoof E, et al. Predicting habitual walking
performance in multiple sclerosis: relevance of capacity and self-
report measures. Mult Scler 2010;16:618-26.
84. Motl RW, Gappmaier E, Nelson K, Benedict Ralph HB. Physical
activity and cognitive function in multiple sclerosis. J Sport Exerc
Psychol 2011;33:734-41.
85. Kohn CG, Coleman CI, Michael White C, Sidovar MF, Sobieraj DM.
Mobility, walking and physical activity in persons with multiple
sclerosis. Curr Med Res Opin 2014;30:1857-62.
86. Morrison JD, Stuifbergen AK. Outcome expectations and physical
activity in persons with longstanding multiple sclerosis. J Neurosci
Nurs 2014;46:171-9.
87. Dlugonski D, Pilutti LA, Sandroff BM, Suh Y, Balantrapu S,
Motl RW. Steps per day among persons with multiple sclerosis:
variation by demographic, clinical, and device characteristics. Arch
Phys Med Rehabil 2013;94:1534-9.
88. Sandroff BM, Dlugonski D, Pilutti LA, Pula JH, Benedict RH,
Motl RW. Physical activity is associated with cognitive processing
speed in persons with multiple sclerosis. Mult Scler Relat Disord
2014;3:123-8.
89. Sosnoff JJ, Sandroff BM, Pula JH, Morrison SM, Motl RW. Falls and
physical activity in persons with multiple sclerosis. Mult Scler Int
2012;2012:315620.
90. Anens E, Emtner M, Zetterberg L, Hellstro
¨m K. Physical activity in
subjects with multiple sclerosis with focus on gender differences: a
survey. BMC Neurol 2014;14:47.
91. Suh Y, Weikert M, Dlugonski D, Balantrapu S, Motl RW. Social
cognitive variables as correlates of physical activity in persons with
multiple sclerosis: findings from a longitudinal, observational study.
Behav Med 2011;37:87-94.
92. Motl RW, Arnett PA, Smith MM, Barwick FH, Ahlstrom B,
Stover EJ. Worsening of symptoms is associated with lower physical
activity levels in individuals with multiple sclerosis. Mult Scler 2008;
14:140-2.
93. Ferrier S, Dunlop N, Blanchard C. The role of outcome expectations
and self-efficacy in explaining physical activity behaviors of in-
dividuals with multiple sclerosis. Behav Med 2010;36:7-11.
94. Motl RW, McAuley E. Symptom cluster as a predictor of physical
activity in multiple sclerosis: preliminary evidence. J Pain Symptom
Manage 2009;38:270-80.
95. Suh Y, Joshi I, Olsen C, Motl RW. Social cognitive predictors of
physical activity in relapsing-remitting multiple sclerosis. Int J Behav
Med 2014;21:891-8.
96. Shammas L, Zentek T, von Haaren B, Schlesinger S, Hey S, Rashid A.
Home-based system for physical activity monitoring in patients with
multiple sclerosis (Pilot study). Biomed Eng Online 2014;13:10.
97. Motl RW, Mullen S, Suh Y, McAuley E. Does physical activity
change over 24 months in persons with relapsing-remitting multiple
sclerosis? Health Psychol 2013;33:326-31.
98. Motl RW, McAuley E, Doerksen S, Hu L, Morris KS. Preliminary
evidence that self-efficacy predicts physical activity in multiple
sclerosis. Int J Rehabil Res 2009;32:260-3.
99. Motl RW, McAuley E. Longitudinal analysis of physical activity and
symptoms as predictors of change in functional limitations and
disability in multiple sclerosis. Rehabil Psychol 2009;54:204-10.
100. Motl RW, McAuley E, Wynn D, Suh Y, Weikert M. Effects of change
in fatigue and depression on physical activity over time in relapsing-
remitting multiple sclerosis. Psychol Health Med 2011;16:1-11.
101. Stuifbergen AK, Blozis SA, Harrison TC, Becker HA. Exercise,
functional limitations, and quality of life: a longitudinal study of per-
sons with multiple sclerosis. Arch Phys Med Rehabil 2006;87:935-43.
102. Motl RW, McAuley E, Wynn D, Vollmer T. Lifestyle physical ac-
tivity and walking impairment over time in relapsing-remitting
multiple sclerosis: results from a panel study. Am J Phys Med
Rehabil 2011;90:372-9.
103. Motl RW, McAuley E. Association between change in physical ac-
tivity and short-term disability progression in multiple sclerosis. J
Rehabil Med 2011;43:305-10.
104. Motl RW, McAuley E, Sandroff BM. Longitudinal change in phys-
ical activity and its correlates in relapsing-remitting multiple scle-
rosis. Phys Ther 2013;93:1037-48.
105. Plow MA, Mathiowetz V, Resnik L. Multiple sclerosis: impact of
physical activity on psychosocial constructs. Am J Health Behav
2008;32:614-26.
106. Motl RW, Dlugonski D, Wo
´jcicki TR, McAuley E, Mohr DC.
Internet intervention for increasing physical activity in persons with
multiple sclerosis. Mult Scler 2011;17:116-28.
107. Turner AP, Hartoonian N, Maynard C, Leipertz SL, Haselkorn JK.
Smoking and physical activity: examining health behaviors and 15-
year mortality among individuals with multiple sclerosis. Arch Phys
Med Rehabil 2015;96:402-9.
108. Sandroff BM, Dlugonski D, Weikert M, Suh Y, Balantrapu S,
Motl RW. Physical activity and multiple sclerosis: new insights
regarding inactivity. Acta Neurol Scand 2012;126:256-62.
109. Rhodes RE, Yao CA. Models accounting for intention-behavior discor-
dance in the physical activity domain: a user’s guide, content overview,
and review of current evidence. Int J Behav Nutr Phys Act 2015;12:9.
110. Geidl W, Semrau J, Pfeifer K. Health behaviour change theories:
contributions to an ICF-based behavioural exercise therapy for in-
dividuals with chronic diseases. Disabil Rehabil 2014;36:2091-100.
111. Mulligan HF, Hale LA, Whitehead L, Baxter GD. Barriers to phys-
ical activity for people with long-term neurological conditions: a
review study. Adapt Phys Activ Q 2012;29:243-65.
112. Schwarzer R. Modeling health behavior change: how to predict and
modify the adoption and maintenance of health behaviors. Appl
Psychol 2008;57:1-29.
113. Bandura A. Health promotion by social cognitive means. Health
Educ Behav 2004;31:143-64.
114. Hagger MS, Chatzisarantis Nikos LD. An integrated behavior change
model for physical activity. Exerc Sport Sci Rev 2014;42:62-9.
115. Luszczynska A. An implementation intentions intervention, the use
of a planning strategy, and physical activity after myocardial
infarction. Soc Sci Med 2006;62:900-8.
116. Scholz U, Schu
¨z B, Ziegelmann JP, Lippke S, Schwarzer R. Beyond
behavioural intentions: planning mediates between intentions and
physical activity. Br J Health Psychol 2008;13:479-94.
117. Sniehotta FF, Scholz U, Schwarzer R. Bridging the intentione
behaviour gap: planning, self-efficacy, and action control in the
adoption and maintenance of physical exercise. Psychol Health 2005;
20:143-60.
118. Ellis T, Cavanaugh JT, Earhart GM, et al. Factors associated with
exercise behavior in people with Parkinson disease. Phys Ther 2011;
91:1838-48.
119. Williams DM, Anderson ES, Winett RA. A review of the outcome
expectancy construct in physical activity research. Ann Behav Med
2005;29:70-9.
120. Young MD, Plotnikoff RC, Collins CE, Callister R, Morgan PJ.
Social cognitive theory and physical activity: a systematic review and
meta-analysis. Obes Rev 2014;15:983-95.
121. Klein T, Becker S. Age and exercise: a theoretical and empirical
analysis of the effect of age and generation on physical activity.
Journal of Public Health 2012;20:11-21.
122. Breuer C. Cohort effects in physical inactivity: a neglected category
and its health economical implications. Journal of Public Health
2005;13:189-95.
123. Rhodes RE, Nigg CR. Advancing physical activity theory: a review
and future directions. Exerc Sport Sci Rev 2011;39:113-9.
124. Williams SL, French DP. What are the most effective intervention
techniques for changing physical activity self-efficacy and physical
activity behaviour-and are they the same? Health Educ Res 2011;26:
308-22.
125. Olander EK, Fletcher H, Williams S, Atkinson L, Turner A,
French DP. What are the most effective techniques in changing
obese individuals’ physical activity self-efficacy and behaviour: a
systematic review and meta-analysis. Int J Behav Nutr Phys Act
2013;10:29.
644 R. Streber et al
www.archives-pmr.org
126. Michie S, Ashford S, Sniehotta FF, Dombrowski SU, Bishop A,
French DP. A refined taxonomy of behaviour change techniques to help
people change their physical activity and healthy eating behaviours: the
CALO-RE taxonomy. Psychol Health 2011;26:1479-98.
127. Saxton JM, Carter A, Daley AJ, et al. Pragmatic exercise intervention
for people with multiple sclerosis (ExIMS trial): study protocol for a
randomised controlled trial. Contemp Clin Trials 2013;34:205-11.
128. Pilutti LA, Dlugonski D, Sandroff BM, Klaren R, Motl RW. Ran-
domized controlled trial of a behavioral intervention targeting
symptoms and physical activity in multiple sclerosis. Mult Scler
2014;20:594-601.
129. Kersten S, Mahli M, Drosselmeyer J, et al. A pilot study of an
exercise-based patient education program in people with multiple
sclerosis. Mult Scler Int 2014:306878.
130. Carter A, Daley A, Humphreys L, et al. Pragmatic intervention for
increasing self-directed exercise behaviour and improving important
health outcomes in people with multiple sclerosis: a randomised
controlled trial. Mult Scler 2014;20:1112-22.
131. Hale LA, Mulligan HF, Treharne GJ, Smith CM. The feasibility and
short-term benefits of Blue Prescription: a novel intervention to
enable physical activity for people with multiple sclerosis. Disabil
Rehabil 2013;35:1213-20.
132. Dlugonski D, Motl RW, Mohr DC, Sandroff BM. Internet-delivered
behavioral intervention to increase physical activity in persons with
multiple sclerosis: sustainability and secondary outcomes. Psychol
Health Med 2012;17:636-51.
133. Dlugonski D, Motl RW, McAuley E. Increasing physical activity in
multiple sclerosis: replicating Internet intervention effects using objec-
tive and self-report outcomes. J Rehabil Res Dev 2011;48:1129-36.
134. Plow M, Bethoux F, McDaniel C, McGlynn M, Marcus B. Ran-
domized controlled pilot study of customized pamphlets to promote
physical activity and symptom self-management in women with
multiple sclerosis. Clin Rehabil 2014;28:139-48.
135. Teixeira PJ, Carrac¸a EV, Markland D, Silva MN, Ryan RM. Exercise,
physical activity, and self-determination theory: a systematic review.
Int J Behav Nutr Phys Act 2012;9:78.
136. Nigg CR, Borrelli B, Maddock J, Dishman RK. A theory of physical
activity maintenance. Appl Psychol 2008;57:544-60.
137. Sallis JF, Bauman A, Pratt M. Environmental and policy interventions
to promote physical activity. Am J Prev Med 1998;15:379-97.
138. King AC, Stokols D, Talen E, Brassington GS, Killingsworth R.
Theoretical approaches to the promotion of physical activity: forging
a transdisciplinary paradigm. Am J Prev Med 2002;23(2 Suppl 1):
15-25.
139. Borkoles E, Nicholls AR, Bell K, Butterly R, Polman Remco CJ. The
lived experiences of people diagnosed with multiple sclerosis in
relation to exercise. Psychol Health 2008;23:427-41.
140. Ma
ˆsse LC, Dassa C, Gauvin L, Giles-Corti B, Motl R. Emerging
measurement and statistical methods in physical activity research.
Am J Prev Med 2002;23(2 Suppl 1):44-55.
141. Sniehotta FF. Towards a theory of intentional behaviour change:
plans, planning, and self-regulation. Br J Health Psychol 2009;14:
261-73.
142. Michie S, Abraham C, Whittington C, McAteer J, Gupta S. Effective
techniques in healthy eating and physical activity interventions: a
meta-regression. Health Psychol 2009;28:690-701.
143. Greaves CJ, Sheppard KE, Abraham C, et al. Systematic review of
reviews of intervention components associated with increased
effectiveness in dietary and physical activity interventions. BMC
Public Health 2011;11:119.
144. Snook EM, Motl RW. Effect of exercise training on walking mobility
in multiple sclerosis: a meta-analysis. Neurorehabil Neural Repair
2009;23:108-16.
145. Spain RI, St George RJ, Salarian A, et al. Body-worn motion sensors
detect balance and gait deficits in people with multiple sclerosis who
have normal walking speed. Gait Posture 2012;35:573-8.
146. Givon U, Zeilig G, Achiron A. Gait analysis in multiple sclerosis:
characterization of temporal-spatial parameters using GAITRite
functional ambulation system. Gait Posture 2009;29:138-42.
147. Martin CL, Phillips BA, Kilpatrick TJ, et al. Gait and balance
impairment in early multiple sclerosis in the absence of clinical
disability. Mult Scler 2006;12:620-8.
Physical activity in multiple sclerosis 645
www.archives-pmr.org
Supplemental Appendix 2 Electronic
Literature Search Strategy
PubMed
Search: 1980 to January 31, 2015
(“physical activity” OR exercise OR exercise [MeSH]) AND
(“multiple sclerosis” OR Multiple Sclerosis [MeSH]) AND
(correlat* OR barrier* OR impediment* OR facilitat*)
Scopus
Search: 1980 to January 31, 2015
((“physical activity” OR exercise) AND (“multiple sclerosis”)
AND (correlat* OR barrier* OR impediment* OR facilitat*))
Abbreviation: MeSH, Medical Subject Heading.
Supplemental Appendix 1 Workflow of methodologic steps with reference to the associated appendices, tables, and figures
Methodologic Steps
Appendix, Table,
or Figure
1. Literature search strategy Supplemental Appendix 2
2. Electronic and manual literature search process Figure 1
3. Risk of bias assessment of articles with an observational design (5-item tool) Supplemental Appendix 3
4. Risk of bias assessment of articles with an interventional design (11-item tool) Supplemental Appendix 4
5. Data extraction of articles with an observational study design Supplemental Appendix 5
6. Data extraction of articles with an interventional design Supplemental Appendix 6
7. Evidence summary table of the associations of potential correlates with PA across observational studies Supplemental Appendix 7
8. Summary of study population characteristics, study characteristics, and data synthesis of observational and
intervention studies
Table 1
Table 2
Table 3
645.e1 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 3 RoB assessment of articles with an observational design
First Author (year)
RoB Interpretation Scores for the RoB Assessment Questions*
Total Score
y
RoB-1 RoB-2 RoB-3 RoB-4 RoB-5
Stuifbergen (1999)
82
100012
Motl (2006)
77
010012
Motl (2006)
76
110013
Stuifbergen
z
(2006)
101
100e10 1 2e3
Gulick (2006)
81
100012
Motl (2007)
80
110013
Doerksen (2007)
50
110013
Motl (2007)
78
110013
Motl (2007)
79
100013
Morris (2008)
51
110013
Motl (2008)
75
100012
Motl (2008)
74
110013
Snook (2008)
73
110013
Motl (2008)
72
110013
Motl (2008)
92
001012
Vanner (2008)
54
110002
Motl
z
(2009)
99
110e10 1 3e4
Motl (2009)
94
111014
Motl
z
(2009)
98
110e10 0 2e3
Stroud (2009)
71
100113
Levy (2009)
52
101013
Turner (2009)
70
100113
Ferrier (2010)
93
001012
Fjeldstad (2010)
69
100012
Motl (2010)
67
110013
Motl (2010)
68
110013
Beckerman (2010)
66
100012
Prakash (2010)
65
010012
Gijbels (2010)
83
110002
Chiu (2011)
59
100012
Dlugonski (2011)
55
100012
Plow (2011)
64
100012
Suh (2011)
91
110013
Suh (2011)
56
110013
Motl
z
(2011)
100
110e10 1 3e4
Kayes (2011)
63
100113
Rietberg (2011)
61
110013
Cavanaugh (2011)
62
110002
Merkelbach (2011)
60
110013
Motl (2011)
103
111014
Motl
z
(2011)
102
110e10 1 3e4
Motl (2011)
84
110003
Chiu (2012)
58
100011
Kasser (2012)
57
100012
Sosnoff (2012)
89
110013
Dlugonski (2013)
87
110013
Klaren (2013)
20
110114
Motl (2013)
97
111014
Motl (2013)
104
111014
Anens (2014)
90
100012
Kasser (2014)
53
101013
(continued on next page)
Physical activity in multiple sclerosis 645.e2
www.archives-pmr.org
Supplemental Appendix 3 (continued )
First Author (year)
RoB Interpretation Scores for the RoB Assessment Questions*
Total Score
y
RoB-1 RoB-2 RoB-3 RoB-4 RoB-5
Kohn (2014)
85
100113
Morrison (2014)
86
100012
Sandroff (2014)
88
110013
Shammas (2014)
96
111003
Suh (2014)
95
111014
* RoB assessment items: 1 (did the study include a validated measure for independent variables?); 2 (did the study include a validated measure for
physical activity behavior [at least 1 objective or 2 self-report measures]?); 3 (was the design prospective, and if so, was there low attrition?); 4 (was
the sample representative?); and 5 (was the study powered to detect significant effects?).
y
RoB interpretation: low risk (5), moderate risk (3e4), and high risk (0e2).
z
The same article reported data from a prospective and cross-sectional observational study design; in these cases RoB assessment results differed
because of the different study designs.
Supplemental Appendix 4 RoB assessment of articles with an intervention design
First Author (Year)
RoB Interpretation Scores for the RoB Assessment Questions*
Total Score
y
RoB-1 RoB-2 RoB-3 RoB-4 RoB-5 RoB-6 RoB-7 RoB-8 RoB-9 RoB-10 RoB-11
Plow (2008)
105
1110010110 0 6
Motl (2011)
106
1111010110 0 7
* RoB assessment items: 1 (did the study include a theoretical framework?); 2 (was the study design a randomized controlled trial?); 3 (were the
methods/procedures designed to influence mediator variables?); 4 (did the study include validated measures for proposed mediators and were the
mediator measures reliable [ie, test-retest reliability]?); 5 (did the study include a validated and at least 1 objective measure for physical activity
behavior?); 6 (was the applied PA measure reliable?); 7 (was the study powered to detect significant effects?); 8 (was baseline physical activity
considered in analyses?); 9 (were statistically appropriate/acceptable methods of data analyses used?); 10 (did the study ascertain whether changes in
the mediator precede changes in the PA outcome?); and 11 (did the authors report conducting pilot studies to test mediation?).
y
RoB bias interpretation: low risk (9e11), moderate risk (5e8), and high risk (0e4).
645.e3 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 5 Observational study characteristics and evidence of potential correlates of PA
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Stuifbergen
(1999)
82
USA
C;
NR;
HPLP-II physical activity subscale;
NZ807 (rural 82%, urban 79%);
Age: rural: 4910.7, urban: 47.710.6;
DC: NR;
DL: ISS rural: 19.48.2, urban: 17.48.8;
Years: rural: 11.88.3, urban: 10.67.8
Type of residency (urban vs rural) þ/H
Motl
(2006)
77
USA
C;
SCT;
GLTEQ, Acc;
NZ196 (88%);
Age: 469.8;
DC: mixed (89%);
DL: ambulatory 77% without and 33% with cane;
Years: 9.07.1
Symptoms (no.) /H
Self-efficacy (exercise and barriers) þ/
Motl
(2006)
76
USA
C;
SCT;
GLTEQ, Acc;
NZ196 (88%);
Age: 469.8;
DC: mixed (89%);
DL: ambulatory 77% without and 33% with cane;
Years: 9.07.1
Enjoyment of PA þ/M
Self-efficacy (exercise) þ/
Social support (global) (þ)/
Disability (LL-FDI) ()/
(continued on next page)
Physical activity in multiple sclerosis 645.e4
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Stuifbergen
(2006)
101
USA
CþP (5y);
Nagi and its extension by Verbrugge and Jette disablement process model
and explanatory model of health promoting behavior and quality of life for
persons with chronic disabling conditions;
Physical activity subscale of HPLP-II;
NZ611 (83%);
Attrition: 90% (response rate by years of follow-up 89.6 (1y), 88.6 (2y), 87.4
(3y), 85.3 (5y);
Age: 49.110.2;
DC: mixed (41%);
DL: ISS 18.40.4;
Years: 13.5
Functional limitations (LL-FDI) C: H;
P: MAge 0 0
Sex (women vs men) 0 0
Time since diagnosis 0 0
Residence type (urban vs rural) 0
Gulick
(2006)
81
USA
C;
NR;
YPAS;
NZ123 (90%);
Age: 42.8 for PDDS 0e1, 4.8 for 2e3, 51.5 for 4e6, 58.8 for 7e8);
DC: NR;
DL: PDDS range, 0e8 (44%: 0e1, 24%: 2e3, 25%: 4e6, 7%: 7e8);
Years: 7.9 for PDDS 0e1, 7.2 for 2e3, 11.5 for 4e6, 21 for 7e8
Disability level (PDDS) /H
Motl
(2007)
80
USA
C;
Disablement process model by Nagi and its extension by Verbrugge and Jette;
GLTEQ, Acc;
NZ133 (78%);
Age: 5111;
DC: mixed (62%);
DL: EDSS mean 5.5 (range, 1e8.5);
Years: 129.0
Symptoms (intensity þfrequency) /M
Functional limitations (LL-FDI) /
Disability level (LL-FDI) /
(continued on next page)
645.e5 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Doerksen
(2007)
50
USA
C;
NR;
IPAQ, Ped;
NZ196 (88%);
Age: 469.8;
DC: mixed (89%);
DL: ambulatory 77% without and 33% with a cane;
Years: 9.07.1
Dwelling Density 0 / M
Access to amenities 0 /
Access to transit stops within 10e15min
walking distance
þ/
Presence of side walks 0 /
Presence of bicycle facilities 0 /
Presence of low-cost recreation facilities 0 /
Level of crime 0 /
Presence of heavy traffic 0 /
Presence of active others 0 /
Aesthetics of neighborhood 0 /
No. of motor vehicles at home 0 /
Motl
(2007)
78
USA
C;
NR;
Ped;
NZ196 (88%);
Age: 469.8;
DC: mixed (89%);
DL: ambulatory 77% without and 33% with a cane;
Years: 9.07.1
Age /M
Education level 0 /
Income 0 /
Marital status (single vs married) 0 /
Employment status (employed vs
unemployed)
þ/
Parental status (children vs no children) 0 /
Race (white vs minority group) 0 /
Sex (women vs men) 0 /
Type of MS (progressive vs relapsing-
remitting)
/
Time since diagnosis (y) /
Thermosensitivity 0 /
Use of a cane for ambulation /
Motl
(2007)
79
USA
C;
NR;
IPAQ;
NZ123 (SPMS: 66%, RRMS: 68%); Age: SPMS: 57.3, RRMS: 47.5;
DC: mixed (66%);
DL: ambulatory (with or without cane); Years: SPMS: 16.6, RRMS: 10.0
Symptoms (frequency) 0 / M
Self-efficacy (exercise) þ/
Walking limitations /
Type of MS (progressive vs relapsing-
remitting)
0/
(continued on next page)
Physical activity in multiple sclerosis 645.e6
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Morris
(2008)
51
USA
C;
SCT;
Acc;
NZ173 (100%);
Age: 469.5;
DC: mixed (95%);
DL: LL-FDI function subscale 35.311.2 (range, 15e60);
Years: 8.97.0
Self-efficacy (exercise) þ/M
Functional limitation /
Residential density 0 /
Land use mix/diversity 0 /
Access to services 0 /
Street connectivity NA /
Walking/cycling facilities 0 /
Aesthetics 0 /
Safety from traffic 0 /
Safety from crime 0 /
Neighborhood satisfaction 0 /
Motl
(2008)
75
USA
C;
Theory of unpleasant symptoms;
GLTEQ;
NZ133 (78%);
Age: 5111;
DC: mixed (62%);
DL: EDSS mean 5.5 (range, 1e8);
Years: 129.9
Symptoms (intensity) /H
Depression 0 /
Pain 0 /
Walking limitations /
Fatigue 0 /
(continued on next page)
645.e7 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Motl
(2008)
74
USA
C;
NR;
Acc;
NZ80 (81%);
Age: 4911.4;
DC: mixed (78%);
DL: EDSS median 4.0 (range, 1e6.5);
Years: 10.17.9
Symptoms intensity (SI) overall /M
SI: spinal cord subscale /
SI: brainstem subscale /
SI: cerebellum subscale /
SI: visual subscale /
SI: left hemisphere subscale /
SI: right hemisphere subscale /
SI: nonlocalized subscale /
Neurologic disability overall (PS) /
PS: mobility subscale /
PS: hand function subscale 0 /
PS: vision subscale 0 /
PS: fatigue subscale /
PS: cognition subscale 0 /
PS: bladder/bowel subscale /
PS: sensory subscale /
PS: spasticity subscale /
Disability (EDSS) /
Snook
(2008)
73
USA
C;
SCT;
Acc;
NZ80 (81%);
Age: 4911.4
DC: mixed (78%);
DL: EDSS 3.91.8;
Years: 10.17.9
Self-efficacy (exercise) þ/M
Symptoms (frequency) /
Motor symptoms /
Brainstem symptoms 0 /
Sensory symptoms 0 /
Mental/emotions 0 /
Elimination symptoms 0 /
(continued on next page)
Physical activity in multiple sclerosis 645.e8
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Motl
(2008)
72
USA
C;
NR;
GLTEQ, Acc;
NZ292 baseline (84%);
Age: 4810.3;
DC: mixed (84%);
DL: PDDS median 3 (range, 0e6);
Years: 10.3 (range, 1e35)
Symptoms (frequency and intensity) /M
Motl
(2008)
92
USA
P(3e5y);
NR;
IPAQ;
NZ51 (80%);
Attrition: 0%;
Age: 529.4;
DC: mixed (57%);
DL: EDSS 4.12.2;
Years: 14.88.8
Symptoms / H
Vanner
(2008)
54
USA
C;
NR;
PADS, NLQ;
NZ43 (72%);
Age: 5410.8;
DC: mixed (12%);
DL: EDSS mean 6.5 (range, 6e8; 6e6.5Z79%; 7e7.5Z26%; 8Z5%);
Years: NR
Apathy /H
Self-efficacy (general) þ/
Fatigue 0 /
Cognitive function /
Depression /
(continued on next page)
645.e9 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Motl
(2009)
99
USA
CþP (6mo);
Nagi and its extension by Verbrugge and Jette disablement process model;
GLTEQ, Acc;
NZ292 (84%);
Attrition: 4.5% (n
post
Z276);
Age: 4810.3;
DC: mixed (84%);
DL: PDDS median 3 (range, 0e6);
Years: 10.37.9 (range, 1e35)
Symptoms (frequency and intensity) 0C:M;
P: MFunctional limitations (LL-FDI) 0
Disability level (LL-FDI) () (0)
Motl
(2009)
94
USA
P (6mo);
Theory of unpleasant symptoms;
GLTEQ, Acc;
NZ292 (84%);
Attrition: 0%;
Age: 4810.3;
DC: mixed (84%);
DL: PDDS median 3 (range, 0e6);
Years: 10.37.9 (range, 1e35)
Symptom cluster (pain, depression,
fatigue)
/()M
Functional limitations (LL-FDI) /
Self-efficacy (exercise) / 0
Motl
(2009)
98
USA
CþP (3mo);
SCT;
Acc;
NZ16 (88%);
Attrition: 0%;
Age: 438.7;
DC: RRMS (100%);
DL: ambulatory with minimal assistance;
Years: NR
Self-efficacy (exercise) þ0C:H;
P: MSelf-efficacy (barrier) 0 þ
(continued on next page)
Physical activity in multiple sclerosis 645.e10
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Stroud
(2009)
71
Australia
C;
Health promotion model;
IPAQ;
NZ93 (82%);
Age: 50.010;
DC: mixed (58%);
DL: DSS range 0e6;
Years: 128.0
Perceived benefits þ/M
Perceived barriers /
Self-efficacy (exercise) þ/
Levy
(2009)
52
USA
P (1y);
TTM;
Stage of exercise behavior algorithm;
NZ105 (100%);
Attrition: 18% (n
post
Z86);
Age: 50.9 (range, 29e76);
DC: NR;
DL: EDSS mean 3.18 (range, 0e7; 88%: EDSS<4.5, 12%: EDSS>5.0);
Years: NR
Behavioral process of change / þM
Cognitive process of change / þ
Self-efficacy (barrier) / 0
Perceived pros / 0
Perceived cons / 0
Turner
(2009)
70
USA
C;
ICF;
Single item: frequency of exercise per week;
NZ2995 (13.5%);
Age: 5512.2;
DC: NR;
DL: VR-physical 2328;
Years: NR
Age /M
Sex (women vs men) 0 /
Race (white vs nonwhite) 0 /
Education level þ/
Living alone (yes vs no) þ/
Marital status (single vs married) 0 /
Body mass index þ/
Pain intensity /
Medical comorbidity (yes vs no) 0 /
(continued on next page)
645.e11 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Ferrier
(2010)
93
Canada
P (1mo after baseline assessment);
SCT;
GLTEQ;
NZ76 (76%);
Age: 5110.4;
DC: mixed (51%);
DL: 60.5% uses mobility aid all the time;
Years: 10.38.4
Outcome expectation / þH
Self-efficacy (exercise and barrier) / þ
Fjeldstad
(2010)
69
USA
C;
NR;
GLTEQ;
NZ77 (70%);
Age: 45.21.4;
DC: mixed (79%);
DL: NR;
Years: 7.60.8 (range, 1e30)
Fatigue /H
Thermosensitivity /
Type of MS (progressive vs relapsing-
remitting)
0/
Sex (women vs men) /
Motl
(2010)
67
USA
C;
Theory of unpleasant symptoms;
IPAQ, GLTEQ, Acc;
NZ269 (83%);
Age: 469.6;
DC: RRMS (100%);
DL: PDDS median 2 (range, 0e6);
Years: 8.87.0
Symptoms (severity and frequency) ()/ M
Fatigue /
Pain 0 /
Depression /
Motl
(2010)
68
USA
C;
Theory of unpleasant symptoms;
IPAQ, GLTEQ;
NZ218 (90%);
Age: 4410.0;
DC: RRMS (100%);
DL: median PDDS 1 (range, 0e6);
Years: 8
Symptom cluster I (fatigue, depression,
pain)
/M
Symptom cluster II (fatigue, depression,
pain, sleep quality, perceived cognitive
dysfunction)
/
(continued on next page)
Physical activity in multiple sclerosis 645.e12
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Beckerman
(2010)
66
The Netherlands
C;
PAD model;
SQUASH;
NZ106 (62%),
Age: 439.6,
DC: mixed (83%);
DL: EDSS mean 3 (0e3Z53%, 3.5e6Z38%, 6.5e10Z9%)
Years: NR
Type of MS (progressive vs relapsing-
remitting)
0/H
Fatigue 0 /
Perceived health 0 /
Comorbidity (Cumulative Illness Rating
Scale) (yes vs no)
0/
Disability (EDSS) /
Depression 0 /
Age 0 /
Sex (women vs men) 0 /
Parental status (children vs no children) /
Knowledge of exercise effects 0 /
Disability pension (yes vs no) /
Self-efficacy (exercise and barrier) 0 /
Attitude toward PA/enjoyment of PA 0 /
Perceived benefits of physical activity 0 /
Personal barriers for physical activity 0 /
Environmental barriers 0 /
Exercise stage of change 0 /
Motivation to adhere to the normative
beliefs of family members and friends
0/
Social Support for Exercise Habits Scale 0 /
Normative expectations/beliefs family
members and friends
0/
Member of a patient organization (yes vs
no)
0/
(continued on next page)
645.e13 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Prakash
(2010)
65
USA
C;
NR;
Acc, GLTEQ;
NZ82 (83%);
Age: 47.511.3;
DC: RRMS (100%);
DL: subsample (see Motl 2008
75
)
EDSS median 4.5 (range, 1e7.5);
Years: 10.08.7
Perceived cognitive impairment /H
Gijbels
(2010)
83
Belgium
C;
NR;
Acc;
NZ50 (66%);
Age: 4910;
DC: mixed (46%);
DL: EDSS 4.51.2 (range, 2.0e6.5);
Years: 117
Disability Level (EDSS, Multiple Sclerosis
Impact Scale-29, Medical Outcomes
Study 36-Item Short-Form Health
Survey, activity limitation and
participation restriction)
/H
Leg muscle strength þ/
Walking limitations (2MWT, 6MWT, timed
Up & Go test, T25FW, Rivermead
Mobility Index)
/
Balance limitation þ/
Chiu
(2011)
59
USA
C;
HAPA;
PASC;
NZ195 (87%) (initially 215);
Age: 4710 (range, 19e67);
DC: NR;
DL: ISS 0.910.62;
Years: NR
Disease severity (symptoms and
performance of ADL)
()/ H
Self-efficacy (action) (þ)/
Self-efficacy (maintenance) þ/
Self-efficacy (relapse) þ/
Outcome expectation of
exercise/perception of benefits
(þ)/
Perceived barriers to health promoting
activities of disabled persons
/
Health risk perception 0 /
Action and coping planning þ/
Health behavior intention (þ)/
(continued on next page)
Physical activity in multiple sclerosis 645.e14
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Dlugonski
(2011)
55
USA
C;
SCT;
GLTEQ;
NZ54 (83%);
Age: 469.9;
DC: RRMS (100%);
DL: PDDS median 2 (range, 0e6);
Years: 8.57.2
Functional limitations (LL-FDI) 0 / H
Outcome expectations 0 /
Exercise goal setting þ/
Self-efficacy (exercise) 0 /
Plow
(2011)
64
USA
C;
TTM;
Physical activity stage algorithm;
NZ335 (80%);
Age: 5310.2;
DC: mixed (63%);
DL: ambulatory with and without assistive device;
Years: NR
Self-efficacy (barrier) þ/H
Behavioral processes of change þ/
Cognitive processes of change þ/
Mobility limitations /
Body mass index 0 /
Frequency of fatigue 0 /
Perceived cognitive dysfunction 0 /
Suh
(2011)
91
USA
P (18mo);
SCT;
GLTEQ, IPAQ;
NZ218 baseline (80%),
Attrition: 22.3 (n
post
Z167);
Age: 43.510;
DC: RRMS (100%);
DL: PDDS median 1 (range, 0e6);
Years: 86.9
Self-efficacy (exercise) / þM
Outcome expectations (self-evaluative) / 0
Outcome expectations (social) / 0
Outcome expectations (physical) / 0
Functional limitations (LL-FDI) / 0
Exercise goal setting / þ
Suh
(2011)
56
USA
C;
SCT;
GLTEQ, IPAQ;
NZ218 (90%);
Age: 43.510;
DC: RRMS (100%);
DL: PDDS median 1 (range, 0e6);
Years: 86.9
Self-efficacy (exercise) (þ)/ M
Outcome expectations (self-evaluative) þ/
Outcome expectations (social) 0 /
Outcome expectations (physical) 0 /
Functional limitations (LL-FDI) /
Exercise goal setting þ/
(continued on next page)
645.e15 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Motl
(2011)
100
USA
CþP (6mo);
NR;
GLTEQ, IPAQ;
NZ269 baseline (83%);
Attrition: 2% (n
post
Z263);
Age: 469.6;
DC: RRMS (100%);
DL: PDDS median 2 (range, 0e6);
Years: 13.39.2
Fatigue (severity) C: M,
P: MDepression
Kayes
(2011)
63
New Zealand
C;
Various cognitive behavioral models of pain, disability and fatigue;
PADS-R;
NZ282 (79%);
Age: 52 (range, 23e83);
DC: Mixed (32%);
DL: GNDS 15.99.4 (range, 0e42);
Years: 11.0 (range, 0e50)
Health beliefs and illness behaviors 0 / M
Fatigue, physical 0 /
Fatigue, mental þ/
Perceived barriers /
Illness-related factors (type of MS, leg
use, time since diagnosis)
/
Self-efficacy (exercise) þ/
Self-efficacy (household chores) þ/
Self-efficacy (managing illness) 0 /
Self-efficacy (leisure and recreation) 0 /
Self-efficacy (managing symptoms) 0 /
Rietberg
(2011)
61
The Netherlands
C;
NR;
Portable 24-h activity monitoring; NZ45 (70%);
Age: 48.77.0 (range, 38e64);
DC: mixed (61%);
DL: EDDS median 3.5 (range, 1e6);
Years: 14.39.2 (range, 2e51)
Fatigue Severity Scale /M
MFIS, overall /
MFIS, physical /
MFIS, cognitive 0 /
MFIS, psychosocial 0 /
CIS20R, overall 0 /
CIS20R, subjective experience of
fatigue
0/
CIS20R, reduction in motivation 0 /
CIS20R, reduction in activity 0 /
CIS20R, reduction in concentration 0 /
(continued on next page)
Physical activity in multiple sclerosis 645.e16
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Cavanaugh
(2011)
62
USA
C;
NR;
Step activity monitor; NZ21 (57%);
Age: 57.612.7;
DC: NR;
DL: EDDS range 3.5e7.5
Years: Group EDSS <4.5: 1112, Group EDSS >4.5: 17.37.1
Disability (EDSS) /H
Ambulatory function /
Balance þ/
Balance confidence/falls-related
self-efficacy
þ/
Fatigue 0 /
Merkelbach
(2011)
60
Germany
C;
NR;
Acc;
NZ80 (71%);
Age: 43.29.8;
DC: Mixed (64%);
DL: EDDS 3.11.6;
Years: 86.8
Disability (EDSS) /M
Sleepiness 0 /
Fatigue 0 /
Motl
(2011)
103
USA
CþP (6mo);
NR;
Acc;
NZ292 (84%),
Attrition: 4.5% (n
post
Z276);
Age: 4810.3;
DC: mixed (84%);
DL: PDDS median 3 (range, 0e6);
Years: 10.3 (range, 1e35)
Disability (PDDS) C: M;
P: M
Motl
(2011)
102
USA
CþP (6mo);
NR;
GLTEQ, IPAQ;
NZ269 (83%),
Attrition: 2.3% (n
post
Z263);
Age: 469.6;
DC: RRMS (100%);
DL: PDDS median 2 (range, 0e6);
Years: 8.87.0
Walking limitations C: M;
P: M
(continued on next page)
645.e17 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Motl
(2011)
84
USA
C;
NR;
Acc;
NZ33 (67%);
Age: 5910.
DC: NR;
DL: ambulatory 100%, EDSS median 6 (range, 3.5e7.0);
Years: 17.610.0
Cognitive processing speed
Learning and memory
/M
0/
Chiu
(2012)
58
USA
C;
HAPA;
PASC;
NZ215 (86%);
Age: 4710.2;
DC: NR,
DL: ISS 0.910.62;
Years: NR
Severity (symptoms and performance of
ADL)
/H
Self-efficacy (action, maintenance,
relapse)
þ/
Outcome expectation of exercise/
perception of benefits
þ/
Perceived barriers to health promoting
activities of disabled persons
/
Health/safety risk perception scale þ/
Action planning and coping planning þ/
Kasser
(2012)
57
USA
C;
HBM;
PASIPD;
NZ348 (84%);
Age: 5011;
DC: mixed (75%);
DL: ambulatory, mild to moderate disability level without cognitive
dysfunctions;
Years: 11.810.4
Sex (women vs men) þ/H
Disability level (measure NR) /
Perceived susceptibility to serious health
problems
0/
Perceived seriousness of unhealthy
consequences of inactivity
0/
Perceived benefits of exercise þ/
Perceived barriers to exercise 0 /
Self-efficacy (barriers) þ/
Cues to action 0 /
(continued on next page)
Physical activity in multiple sclerosis 645.e18
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Sosnoff
(2012)
89
USA
C;
NR;
Acc;
NZ75 (67%);
Age: 5212;
DC: mixed (84%);
DL: EDSS median 4.0 (range, 2.0e6.5);
Years: 12.810
No. of falls (fall history) 0 / M
Dlugonski
(2013)
87
USA
C;
NR;
Pedometer and accelerometer
NZ645 (85%);
Age: 4610.6;
DC: mixed (89%);
DL: PDDS median 2.0 (range, 0e6);
Years: 9.37.8
Sex (women vs men) 0 / M
Disability level (PDDS) /
Education level þ/
Type of MS (progressive vs relapsing-
remitting)
0/
Disease duration 0 /
Race (white vs nonwhite) 0 /
Employment status (employed vs
unemployed)
þ/
Klaren
(2013)
20
USA
C;
NR;
Acc;
NZ800 (84%)
Age: 47.310.1
DC: mixed (92%)
DL: PDDS (range, 0e8)
Years: 10.07.5
Walking limitations (assistive device) /M
Disease duration /
Type of MS (progressive vs
relapsing-remitting)
/
Employment status (employed vs
unemployed)
þ/
Education level þ/
Sex (women vs men) 0 /
Income 0 /
Race (white vs nonwhite) 0 /
(continued on next page)
645.e19 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Motl
(2013)
97
USA
P (24mo);
NR;
GLTEQ;
NZ269 (83%),
Attrition at 12mo: 6% (n
post
at 12moZ253), attrition at 24mo: 9%
(n
post
at 24moZ245);
Age: 469.6;
DC: RRMS (100%);
DL: PDDS median 2 (range, 0e6);
Years: 8.8 (range, 7)
Age (45 vs 46e64) / 0 M
Sex (women vs men) /
Duration of MS (0e10 vs >10y) / 0
Disability (PDDS: 0e2vs3) /
Motl
(2013)
104
USA
P (30mo);
NR;
Acc þIPAQ;
NZ269 (83%),
Attrition at 12mo: 6% (n
post
at 12moZ253), attrition at 24mo: 9%
(n
post
at 24moZ245);
Age: 469.6;
DC: RRMS (100%);
DL: PDDS median 2 (range, 0e6);
Years: 8.8 (range, 7)
Walking impairment (MSWS-12) 0M
Disability (PDDS) 0
Depression / 0
Pain / 0
Fatigue / 0
Self-efficacy (exercise) þþ
(continued on next page)
Physical activity in multiple sclerosis 645.e20
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Anens
(2014)
90
Sweden
C;
NR;
PADS-R;
NZ287 (71%);
Age: 51.513.5;
DC: mixed (83%);
DL: PDDS median 2 (range, 0e6);
Years: median 11
Sex (women vs men) þ/H
Kasser
(2014)
53
USA
P;
NR;
Aerobic Centers Longitudinal Study PA Questionnaire;
NZ99 (100%),
Age: 50.58.4;
DC: NR;
DL: EDDS median 3 (range, 0e5.5);
Years: NR
Fear of falling / M
Balance / 0
Gait impairments / 0
Muscle strength (ie, muscle power
asymmetry)
/0
Mental and emotional functioning / 0
No. of falls /
Kohn
(2014)
85
USA
C;
NR;
GLTEQ;
NZ3260 (79%);
Age: mostly 55e64 (exact mean and SD not calculable);
DC: NR;
DL: PDDS median 3 (exact range not reported; 38%: no impairment,
57%: moderately impaired, 5%: severely impaired);
Years: NR
Disability (PDDS) /H
Walking limitations (MSWS-12) /
Age /
Duration of MS /
Sex (women vs men) /
Employment status (employed vs
unemployed)
þ/
Receiving physical therapy (yes vs no) þ/
Disease-modifying pharmacologic
treatment (yes vs no)
0/
Race (white vs nonwhite) þ/
(continued on next page)
645.e21 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Morrison
(2014)
86
USA
C;
SCT;
HPLP-II PA subscale;
NZ369 (85%);
Age: 61.49.3 (range, 33e87);
DC: NR;
DL: NR;
Years: 24.66.6 (range, 15e55)
Outcome expectation (physical) þ/H
Outcome expectation (social) þ/
Outcome expectation (self-evaluative) 0 /
Age 0 /
Disease duration 0 /
Sex (women vs men) 0 /
Marital status (single vs married) 0 /
Education level þ/
Functional limitations/disability (ISS) /
Sandroff
(2014)
88
USA
C;
NR;
Acc;
NZ212 (80%);
Age: 5010.3;
DC: mixed (83%);
DL: T25FW 6.96.7;
Years: 11.48.8
Cognitive processing speed /M
Shammas
(2014)
96
GER
P (1y);
NR;
Acc;
NZ11 (80%);
Age: 5010.3;
DC: mixed (83%);
DL: EDSS range 1e5; T25FW 6.96.7;
Years: 11.48.8
Disability (EDSS) / M
Walking limitation (speed) /
(continued on next page)
Physical activity in multiple sclerosis 645.e22
www.archives-pmr.org
Supplemental Appendix 5 (continued )
First Author
(Year)
Country
Study Design (duration);
Theoretical Framework;
PA Measure;
Sample Size (% female),
Attrition;
Age (y);
Disease Course*(% RRMS);
Disability Level; and Time Since Diagnosis (y) Potential Correlate
Evidence
y
RoB
z
CP
Suh
(2014)
95
USA
P (6wk);
SCT;
Acc þGLTEQ;
NZ68 (83%);
Age: 49.18.8;
DC: RRMS (100%);
DL: PDDS 2.0 (IQR, 0e3);
Years: 12.17.9
Self-efficacy (exercise) / (þ)M
Outcome expectations (self-evaluative) / 0
Outcome expectations (social) / 0
Outcome expectations (physical) / 0
Functional limitations (LL-FDI) /
Social support (physical activity) / 0
Exercise goal setting / þ
Abbreviations: 2MWT, 2-minute walk test; 6MWT, 6-minute walk test; Acc, accelerometer; ADL, activities of daily living; C, cross-sectional design; CIS20R, Checklist Individual Strength; DC, disease course; DL,
disability level; DSS, Disease Steps Score; EDSS, Expanded Disability Status Scale; GLTEQ, Godin Leisure-Time Exercise Questionnaire; GNDS, Guy’s Neurological Disability Scale; H, high risk of bias; HAPA, Health
Action Process Approach; HBM, Health Belief Model; HPLP-II, Health-Promoting Lifestyle Profile II; IPAQ, International Physical Activity Questionnaire; IQR, interquartile range; ISS, Incapacity Status Scale;
LL-FDI, Late-Life Function & Disability Instrument; M, moderate risk of bias; MFIS, Modified Fatigue Impact Scale; MSWS-12, Multiple Sclerosis Walking Scale; NA, not applicable; NLQ, Nottingham Leisure
Questionnaire; n
post
, participants at follow-up; NR, not reported; P, prospective observational study design; PAD model, physical activity for people with a disability model; PADS, physical activity and
disability survey; PADS-R, Physical activity disability survey- revised; PASC, physical activity stage of change instrument; PASIPD, Physical Activity Scale for Individuals with Physical Disabilities; PDDS,
Patient Determined Disease Steps; Ped, pedometer; PS, Performance Scale; SCT, social cognitive theory; SI, Symptom Inventory; SPMS, secondary-progressive multiple sclerosis; SQUASH, Short Questionnaire to
Assess Health-Enhancing Physical Activity; T25FW, Timed 25-Foot Walk; TTM, transtheoretical model of behavior change; VR, Veteran RAND 36-Item Health Survey; YPAS, Yale Physical Activity Survey.
* Type of MS: progressive (only persons with progressive types of MS); RRMS (only persons with relapsing-remitting type of MS); mixed (persons with progressive and relapsing-remitting type of MS); and NR
(no information regarding type of MS).
y
Evidence for an association between with PA: 0 (nonsignificant association); þ, significant positive association (in case of categorical variables, the first listed manifestation is significantly more active
than the comparator); , significant inverse association (in case of categorical variables, the first listed manifestation is significantly less active than the comparator); indirect effects are denoted with
brackets (þ)or(); /, not tested in this individual study.
z
Risk of bias: low, moderate, or high.
645.e23 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 6 Individual intervention study characteristics and evidence for proposed mediators
First Author
(Year)
Country
Study Design (duration);
Sample Size (% female),
Attrition Rate;
Age;
DC (% RRMS);
DL;
Time Since Diagnosis (y);
PA criteria for inclusion at baseline;
PAM
Theoretical Framework;
Intervention Condition: Content
Control Condition
Theoretical Framework;
Content PA Change*Proposed Mediator AT
y
CT
z
ME
x
RoB
k
Plow et al
(2008)
105
USA
RCT (7wk, 8-wk follow-up);
NZ50 (79%);
Attrition: 16% (n
post
Z42);
Age: 48.5;
DC: NR;
DL: ability to walk with or without assistive
device, MSFC score both groups mean 0.04;
Years: NR;
PA: no specific in-/exclusion criteria for
physical activity level;
PAM: subscale of HPLP-II
Adopted social cognitive
perspective (TPB and SCT);
group-based wellness/education
intervention: modules to
promote health and physical
activity including energy
conversation, goal and priority
setting, time/stress
management, increase self-
efficacy and social support plus
16-wk home exercise program
Physical therapy
concept, 4
individualized
physical therapy
sessions, telephone
calls plus 16-wk home
exercise program
Significant increase
in physical
activity across
both groups but
without between-
subject effects
Self-efficacy
(barrier-specific)
0NSM
Self-identity þ0NS
Expectation for regular
PA
0NS
Social support 0 0 NS
Motl et al
(2011)
106
USA
RCT (12wk, no follow-up);
NZ54 (90%);
Age: 46;
DC: RRMS (100%);
DL: PDDS 2;
Years: 7e8;
PA: inactive, not engaging in regular physical
activity (30min accumulated/d) on >2d of
the week during the previous 6mo;
PAM: GLTEQ
SCT; web-based PA promotion
program, modules: goal setting
and feedback, self-monitoring,
outcome expectations, self-
efficacy, barrier management
and social support, information
for maintenance of an active
physical activity lifestyle and
relapse prevention, toll-free
telephone line plus e-mail, chat
session twice per week
Waitlist control Significant large
increase in PA
after 12 weeks
compared with
control group
Exercise goal setting þþMM
Self-efficacy (exercise) 0 0 0
Outcome expectations
(self-evaluative)
000
Outcome expectation
(social)
000
Outcome expectations
(physical)
000
Functional limitations 0 0 0
Abbreviations: AT, action theory test; CT, conceptual theory test; DC, disease course; DL, disability level; GLTEQ, Godin Leisure Time Exercise Questionnaire; HPLP-II, Health-Promoting Lifestyle Profile II; ME,
mediated effect; MSFC, Multiple Sclerosis Functional Composite; n
post
, participants at follow-up; NR, not reported; NS, not stated; PAM, physical activity measure; PDDS, Patient Determined Disease Steps; RCT,
randomized controlled trial, SCT, social cognitive theory; TPB, theory of planned behavior.
* Direct intervention effect on PA (primary outcome).
y
Evidence for action theory test: 0 (nonsignificant change); þ(significant increase); or (significant decrease of proposed mediator).
z
Evidence for conceptual theory test: 0 (nonsignificant association); þ(significant positive association); or (significant negative association with PA).
x
Evidence for mediated effect: 0 (nonsignificant mediation); M (mediator); or S (suppressor).
k
Risk of bias: L (low risk); M (moderate risk); or H (high risk).
Physical activity in multiple sclerosis 645.e24
www.archives-pmr.org
Supplemental Appendix 7 Evidence summary table of the associations of potential correlates with PA across observational studies
Potential Correlates RoB
Relation to PA of Individual Studies*(noted by reference numbers) Consistency
y
Positive Association Nonsignificant Association Inverse Association Summary Code Summary Score
Health condition
Type of MS (progressive vs relapsing)
z
M 79, 87 20, 78 0 2/6 (33%)
H 66, 69
Time since diagnosis M 101
x
, 87, 97
x
20, 78 ?? 3/8 (34%)
H 101, 86 85
Functioning
Generic functioning variables
Overall functioning/disability level
k
M 101
x
, 103, 103
x
60, 74, 74,96, 97
x
e17/17 (100%)
H 101, 57, 58
{
/59
#
, 81,
62, 66, 83, 85, 86
Neurologic impairments
k
M99
x
, 79 99/72
{
,67
#
, 73/74
{
,
75/80
{
e6/8 (75%)
H 77, 92
x
Activity limitations
k
M99
x
,91
x
51, 99, 56, 80, 94
x
,95
x
e6/9 (67%)
H55
Participation restrictions
k
M99
x,#
99
#
,76
#
,80
#
3/4 (75%)
Mental health M 53 0 0/1 (0%)
Specific functioning variables
Body functions/structures
Symptom cluster I** M 68, 94
x,#
2/2 (100%)
Symptom cluster II** M681/1 (100%)
Thermosensitivity M 78 ? 1/2 (50%)
H69
Fatigue overall M 60, 104
x
100
x
, 61, 67/100
{
?? 4/11 (36%)
H 54, 62, 64, 66, 75 69
Physical fatigue M 63 0 0/1 (0%)
Mental fatigue M 63 þ1/1 (100%)
Depression M 104
x
100
x
, 67/100
{
? 3/6 (50%)
H 66, 75 54
Apathy H 54 1/1 (100%)
Pain M 67, 104
x
70 0 1/4 (25%)
H75
(continued on next page)
645.e25 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 7 (continued )
Potential Correlates RoB
Relation to PA of Individual Studies*(noted by reference numbers) Consistency
y
Positive Association Nonsignificant Association Inverse Association Summary Code Summary Score
Balance capacity M 53
x
þ2/3 (67%)
H 62, 83
Muscle strength lower extremities M 53
x
? 1/2 (50%)
H83
Cognitive functions overall M 64 2/3 (67%)
H 54, 65
Information processing speed M 84, 88 2/2 (100%)
Learning and memory M 84 0 0/1 (0%)
Sleep functions/sleepiness M 60 0 0/1 (0%)
Activity/participation
Walking/mobility limitations M 53
x
, 104
x
104, 96, 20, 64,78, 79,
102, 102
x
e12/14 (86%)
H 62, 75, 83, 85
Personal factors
Self-efficacy/perceived control
Physical activity/exercise specific self-efficacy M 51, 98, 98
x
,56
#
, 57, 63, 64, 71,
73, 76
{
/77, 79, 91
x
,93
x
,95
#,x
,
104, 104
x
98, 98
x
, 105, 52
x
,94
x
þþ 17/24 (71%)
H 59/58
{
55, 66
Leisure and recreation self-efficacy M 63 0 0/1 (0%)
Household chores self-efficacy M 63 þ1/1 (100%)
General self-efficacy H 54 þ1/1 (100%)
Illness- and symptom-related self-efficacy M 63, 63 0 0/2 (0%)
Falls-related efficacy H 53
x
,62 þ2/2 (100%)
Attitudinal beliefs and knowledge
Knowledge of exercise effects on health H 66 0 0/1 (0%)
Enjoyment of physical activity M 76 ? 1/2 (50%)
H66
Overall positive outcome expectations/perceived
benefits
M 71, 93
x
52
x
?? 4/7 (57%)
H 57, 59
#
55, 66
Self-evaluative outcome expectations M 56 91
x
,95
x
0 1/4 (25%)
H86
Physical outcome expectations M 56, 91
x
,95
x
0 1/4 (25%)
H86
Social outcome expectations M 56, 91
x
,95
x
0 1/4 (25%)
H86
Negative outcome expectations M 52
x
0 0/1 (0%)
(continued on next page)
Physical activity in multiple sclerosis 645.e26
www.archives-pmr.org
Supplemental Appendix 7 (continued )
Potential Correlates RoB
Relation to PA of Individual Studies*(noted by reference numbers) Consistency
y
Positive Association Nonsignificant Association Inverse Association Summary Code Summary Score
Perceived barriers M 63, 71 3/5 (60%)
H 57, 66 59
Self-identity H 57 þ1/1 (100%)
Normative and norm-related beliefs and activities
Cues to action H 57 0 0/1 (0%)
Motivation to adhere to normative beliefs of family
members amd friends
H 66 0 0/1 (0%)
Risk-related beliefs and emotional responses
Risk perception/perceived susceptibility H 58
{
57, 59
#,{
, 66 0 1/3 (33%)
Perceived seriousness of unhealthy consequences
of inactivity
H 57 0 0/1 (0%)
Cognitive health beliefs and Illness behaviors M 63 0 0/1 (0%)
Intention/commitment/planning
Exercise stage of change H 66 0 0/1 (0%)
Expectation H 105 0 0/1 (0%)
Intention H 59
#
þ1/1 (100%)
Behavioral process of change M 52
x
,64 þ2/2 (100%)
Cognitive process of change M 52
x
,64 þ2/2 (100%)
Exercise goal setting M 56, 91
x
,95
x
þ4/4 (100%)
H55
Action and coping planning H 58
{
/59 þ1/1 (100%)
Sociodemographic and biologic variables
Age M 101
x
,97
x
70, 78, 85 ?? 3/8 (38%)
H 101, 66, 86
Sex (women vs men)
z
M 20, 101
x
, 70, 78 78, 87 97
x
00 3/13 (23%)
H 90 53, 57, 66, 86, 101 69, 85
Education level M 20, 70, 87 78 þþ 4/5 (80%)
H86
Income M 20, 78 0 0/2 (0%)
Marital status (single vs married)
z
M 70, 78 0 0/3 (0%)
H86
Living alone (yes vs no)
z
M70 þ1/1 (100%)
Employment status (employed vs unemployed)
z
M 20, 78, 87 þþ 4/4 (100%)
H85
Disability pension (yes vs no)
z
H661/1 (100%)
(continued on next page)
645.e27 R. Streber et al
www.archives-pmr.org
Supplemental Appendix 7 (continued )
Potential Correlates RoB
Relation to PA of Individual Studies*(noted by reference numbers) Consistency
y
Positive Association Nonsignificant Association Inverse Association Summary Code Summary Score
Parental status/having children to care for
(children vs no children)
z
M 78 ? 1/2 (50%)
H66
Race (white vs minority)
z
M 20, 70, 78, 87 00 1/5 (20%)
H85
Body mass index M 70 64 ? 1/2 (50%)
Medical comorbidity (yes vs no)
z
M 70 0 0/2 (0%)
H66
Member in patient organization (yes vs no)
z
H 66 0 0/1 (0%)
Receive physical therapy/health care service
(yes vs no)
z
H85 þ1/1 (100%)
Disease-modifying pharmacologic treatment (yes
vs no)
z
H 85 0 0/1 (0%)
History/no. of falls M 89 53
x
? 1/2 (50%)
Environmental factors
Physical and natural environmental factors
Type of residency (urban vs rural)
z
M 101
x
? 1/3 (33%)
H 82 101
Residential density/dwelling density M 50, 51 0 0/2 (0%)
Presence of bicycle facilities M 50 0 0/1 (0%)
Land use mix/diversity M 51 0 0/1 (0%)
Presence of low-cost recreation facilities M 50 0 0/1 (0%)
Proximity to transit stops M 50 þ1/1 (100%)
Access to services M 51 0 0/1 (0%)
Access to amenities M 50 0 0/1 (0%)
Presence of heavy traffic M 50 0 0/1 (0%)
Safety from traffic M 51 0 0/1 (0%)
Street connectivity M 51 0 0/1 (0%)
Presence of side walks M 50 0 0/1 (0%)
Neighborhood aesthetics M 50, 51 0 0/2 (0%)
Neighborhood satisfaction M 51 0 0/1 (0%)
No. of motor vehicles at home M 50 0 0/1 (0%)
Social environmental factors
Social support (general) M 76
#
þ1/1 (100%)
Social support (physical activity) M 95
x
0 1/3 (33%)
H 105 66
(continued on next page)
Physical activity in multiple sclerosis 645.e28
www.archives-pmr.org
Supplemental Appendix 7 (continued )
Potential Correlates RoB
Relation to PA of Individual Studies*(noted by reference numbers) Consistency
y
Positive Association Nonsignificant Association Inverse Association Summary Code Summary Score
Normative expectations/beliefs of family members
and friends
H 66 0 0/1 (0%)
Presence of active others M 50 0 0/1 (0%)
Level of/safety from crime M 50, 51 0 0/2 (0%)
Abbreviations: H, high risk of bias; L, low risk of bias; M, moderate risk of bias.
* Significance and direction of findings for an association between a potential correlate and PA from individual studies.
y
Based on the summary score result, consistency for the same variable across individual studies was interpreted as: 0, variable is rarely or not associated with PA (0%e33% of results supported a hy-
pothesized association divided by the total number of extracted results); ?, results for this association are inconsistent (34%e59%); þ/, variable is consistently positively/inversely associated with PA
(60%e100%); þþ,, or 00, if 4 studies consistently supported the same result; ??, frequently examined variable with considerable lack of consistency in results across studies. For interpretation of
results, a summary score was calculated by the number of extracted results supporting an association (þor ) divided by the total number of results in this row (both noted by reference numbers). This ratio is
presented in numbers and percentages: n
associated
/n
total
(%).
z
Interpretation of the relationship of categorical variables with PA: manifestations that are listed firstly in brackets are positive and the opposite is inverse associated with PA.
x
Results from prospective observational studies.
k
Impairments were measured as number, severity, frequency, or a combination of the aforementioned qualities measured with the Multiple Sclerosis-Related Symptom Checklist and the Symptom Inventory;
the overall functioning level includes data measured with the Expanded Disability Status Scale; Incapacity Status Scale; Minimal Record of Disability; Patient Determined Disease Steps; Performance Scale; the
activity limitation level includes data measured with the functional limitations subscale of the Late Life Function & Disability Instrument; and participation restriction level includes data measured with the
disability subscale of the Late Life Function & Disability Instrument.
{
In some cases, examinations of 1 variable with data from the same study population and the same study design were published more than once. To avoid overreporting of specific findings, articles were
checked accordingly. Such findings were summarized and denoted in this evidence summary table. This overlapping result was only counted as one result. If applicable, we referred to the relevant reference
numbers and presented them together (separated through a backslash). Some prospective
98-104
and intervention studies
55,105,106
simultaneously reported results from cross-sectional and longitudinal data
analyses. In such cases, both results from the same variable from the same study sample were separately extracted and included as unique results.
#
Indirect effects examined in path or mediation analysis.
** Symptom clusters: symptom cluster I (fatigue, pain, depression) or symptom cluster II (fatigue, pain, sleep quality, depression, perceived cognitive dysfunction).
645.e29 R. Streber et al
www.archives-pmr.org
