Rehabilitation Psychology © 2014 American Psychological Association

2014, Vol. 59, No. 4, 415– 421 0090-5550/14/$12.00 http://dx.doi.org/10.1037/a0037739

Physical Activity and Health-Related Quality of Life Over Time in Adults

With Multiple Sclerosis
Robert W. Motl and Edward McAuley
University of Illinois at Urbana-Champaign

Objective: This prospective panel study examined the relationship between changes in physical activity
and health-related quality of life (HRQOL) across a 6-month period in persons with multiple sclerosis
(MS). Method: Adults with a definite diagnosis of MS completed a battery of questionnaires that
included the Godin Leisure-Time Exercise Questionnaire (Godin & Shephard, 1985) and the 36-item
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Medical Outcomes Study Short-Form Health Survey (SF-36; Ware & Sherbourne, 1992) at baseline (n ⫽
This document is copyrighted by the American Psychological Association or one of its allied publishers.

292) and 6-month follow-up (n ⫽ 276). The data were analyzed using panel analysis in Mplus 6.0.
Results: The panel model represented an acceptable fit for the data (␹2 ⫽ 140.72, df ⫽ 56, standardized
root mean square residual ⫽ 0.06, comparative fit index ⫽ 0.98). The standardized path coefficients were
statistically significant between follow-up physical activity and follow-up Physical Function (␤ ⫽ .12,
p ⬍ .005), Role–Emotional (␤ ⫽ .16, p ⬍ .01), Vitality (␤ ⫽ .13, p ⬍ .001), and Social Function (␤ ⫽
.12, p ⬍ .05). Those who reported a change (increase or decrease) in levels of physical activity over 6
months reported a change (improving or worsening, respectively) in HRQOL on 4 of 8 domains on the
SF-36, independent of disability status, MS clinical course and duration, age, and sex. Conclusion: The
observed pattern of relationships supports the possibility that changing physical activity through an
intervention might yield desirable changes in HRQOL, particularly domains representing the mental
component.

Keywords: physical activity, multiple sclerosis, quality of life, patient-reported outcomes

Impact and Implications
• Multiple sclerosis (MS) is a common neurological disease that often
results in reduced physical activity and health-related quality of life
(HRQOL), yet there is limited research on the association between change
in free-living physical activity and domains of HRQOL.
• This study confirmed both cross-sectional and longitudinal, but not
prospective, associations between physical activity and four of eight do-
mains of HRQOL, even after controlling for covariates, including disability
status.
• Health-care professionals, including clinicians, rehabilitation psychol-
ogists, physical therapists, and exercise specialists, might consider promot-
ing free-living physical activity as a method for improving domains of
HRQOL among those with MS.
Introduction
Multiple sclerosis (MS) is a common neurological disease that
often results in reduced physical activity and health-related quality
This article was published Online First August 25, 2014.
Robert W. Motl and Edward McAuley, Department of Kinesiology and
Community Health, University of Illinois at Urbana-Champaign.
Funded by the United States Department of Health and Human Services,
National Institutes of Health, National Institute of Neurological Diseases
and Stroke (Grant NS054050).
Correspondence concerning this article should be addressed to Robert
W. Motl, PhD, Department of Kinesiology and Community Health, Uni-
versity of Illinois, 233 Freer Hall, Urbana, IL 61801. E-mail: robmotl@
illinois.edu
415
of life (HRQOL). By definition, physical activity is a behavior that
involves bodily movement produced by contraction of skeletal
muscles and results in increased energy expenditure compared
with rest (Bouchard & Shephard, 1994). HRQOL is an umbrella
term (Rejeski & Mihalko, 2001; Stewart & King, 1991) and one
way it can be described involves the physical and mental compo-
nents of health status that include eight domains of Physical
Functioning, Role–Physical, Bodily Pain, General Health, Vitality,
Social Functioning, Role–Emotional, and Mental Health (Ware &
Sherbourne, 1992). Persons with MS are less physically active
than adults from the general population (Motl, McAuley, & Snook,
2005), and recent evidence indicates that free-living physical ac-
tivity decreases over time in this population (Motl, McAuley, &
Sandroff, 2013). Those with MS have lower scores on domains of
HRQOL than nondiseased populations (Benito-León, Morales,
Rivera-Navarro, & Mitchell, 2003; Mitchell, Benito-León,
González, Rivera-Navarro, 2005) and those suffering from inflam-
matory bowel disease, ischemic stroke, and rheumatoid arthritis
(Lankhorst et al., 1996; Naess, Beiske, & Myhr, 2007; Rudick,
Miller, Clough, Gragg, & Farmer, 1992). This paper examined
whether change in free-living physical activity was associated with
change in domains of HRQOL consistent with a hierarchical
model proposed by Stewart and King (1991).
To date, very few researchers have examined whether free-
living physical activity is associated with domains of HRQOL in
persons with MS (Benito-León, 2011), although supervised and
structured exercise training generally improves HRQOL (Motl &
Gosney, 2008). The focus on free-living physical activity, rather
than structured exercise training, is important considering that this
behavior is self-selected rather than prescribed, planned or un-
 416 MOTL AND MCAULEY
planned, and accumulated during daily living in the context of
one’s choice of environment. This type of physical activity can be
further changed through large-scale, cost-effective behavioral in-
terventions. The focus on free-living physical activity and its
association with domains of HRQOL further reflects the possibil-
ity that the promotion of this behavior might become part of the
clinical armamentarium of those who treat MS patients (Benito-
León, 2011). Of note, some domains of HRQOL, such as Physical
Functioning or Vitality, might be more strongly associated with
free-living physical activity than other domains, such as Bodily
Pain, thereby informing the focus of future randomized controlled
trials (RCTs; Baumstarck et al., 2013). This latter observation is
based on a hierarchical model whereby HRQOL, and its domains
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
in particular, may be more proximal outcomes of physical activity
This document is copyrighted by the American Psychological Association or one of its allied publishers.
compared with global QOL (McAuley et al., 2006, 2008; Stewart
& King, 1991).
This study involved a secondary analysis of previously pub-
lished data (Motl & McAuley, 2009) and the purpose involved
examining the expected association between intraindividual
changes in free-living physical activity and domains of HRQOL
across a 6-month period in persons with MS, while controlling for
possible covariates of disability status, MS clinical course and
duration, age, and sex (Benito-León et al., 2003; Mitchell et al.,
2005; Krokavcova et al., 2009). We expected that intraindividual
change in physical activity would be associated with intraindi-
vidual changes in domains of HRQOL, and the associations might
be stronger for domains associated with the physical component of
health status that are largely compromised in MS, such as Physical
Functioning and Vitality (i.e., energy or fatigue; Rejeski & Mi-
halko, 2001). The associations were specified a priori in a panel
model that was directly tested using panel analysis (Kessler &
Greenberg, 1981). This investigation is important, considering
calls for research efforts on predictors (Baumstarck et al., 2013)
that can inform behavioral interventions for improving HRQOL in
MS (Benito-León, 2011); this is particularly salient for clinicians,
as free-living physical activity is changeable through intervention.
Method
Participants
The recruitment, inclusion criteria, and characteristics of the
sample in this prospective panel study have been extensively
described by previous researchers (Motl & McAuley, 2009).
Briefly, persons with MS were recruited from three midwest
chapters of the National MS Society primarily through (a) adver-
tisements placed in MS Connection quarterly publications and (b)
e-mail messages distributed to registered members of the chapters.
Those who were interested in participation contacted the research
coordinator, who described the study and its procedures, answered
all questions, and conducted a screening for inclusion criteria. The
inclusion criteria were (a) definite diagnosis of MS, (b) relapse-
free in the last 30 days, and (c) ambulatory with minimal assis-
tance. There were 511 individuals who expressed interest in par-
ticipation, and 387 of them underwent screening. Twenty-seven
individuals did not satisfy inclusion criteria and 16 persons de-
clined participation. The research coordinator sent an informed
consent document and MS verification letter (completed and re-
turned by the patient’s physician) to the remaining 344 persons.
Both forms were returned for 300 persons. Of those who returned
the forms, eight did not continue with participation for unknown
reasons (i.e., those persons did not complete the initial or baseline
study measures). There were 292 persons who provided baseline
data, and 276 of them provided follow-up data 6 months later (95%
retention). The 5% attrition involved a change in the residential
addresses of 16 participants, but all participants who provided
baseline data were included in the data analysis.
Measures
Physical activity. Physical activity was measured by the Go-
din Leisure-Time Exercise Questionnaire (GLTEQ; Godin &
Shephard, 1985). The GLTEQ is a self-administered measure of
physical activity that has been validated in persons with MS
(Gosney, Scott, Snook, & Motl, 2007; Motl, McAuley, Snook, &
Scott, 2006) and has captured the effects of behavioral interven-
tions in MS (Motl, Dlugonski, Wójcicki, McAuley, & Mohr,
2011). The GLTEQ has three items that measure the frequency of
strenuous (e.g., jogging), moderate (e.g., fast walking), and mild
(e.g., easy walking) physical activities for periods of more than 15
min in a typical week. The weekly frequencies of strenuous,
moderate, and mild physical activities are multiplied by 9, 5, and
3 metabolic equivalents, respectively, and summed into a measure
of physical activity that ranges between 0 and 119. Higher scores
indicate higher levels of physical activity.
Health-related quality of life. HRQOL was measured using
the original, unmodified Medical Outcomes Study 36-Item Short-
Form Health Survey (SF-36; Ware & Sherbourne, 1992). The
SF-36 is a standard and generic measure of HRQOL that has
commonly been adopted for research in persons with MS (Mitchell
et al., 2005; Krokavcova et al., 2009). The SF-36 has been iden-
tified as the most widely used and carefully validated measure of
HRQOL for inclusion in research on physical activity (Rejeski &
Mihalko, 2001). The SF-36 includes 36 items with scores that are
transformed into a 0 –100 scale, and then averaged into eight
subscales or domains of Physical Functioning, Role–Physical,
Bodily Pain, General Health, Vitality, Social Functioning, Role–
Emotional, and Mental Health. The scores per subscale range
between 0 and 100 and higher scores reflect better domain-specific
HRQOL.
Disability. Disability was measured using the Patient Deter-
mined Disease Steps (PDDS) scale (Hadjimichael, Kerns, Rizzo,
Cutter, & Vollmer, 2007). This scale was developed as a surrogate
of the Expanded Disability Status Scale (EDSS; Kurtzke, 1983)
and when the assessment of disability is not possible through
face-to-face interactions; the PDSS is therefore an appropriate
assessment of disability in the current study, as all data were
collected through the United States Postal Service (USPS). The
PDDS is a self-report questionnaire that contains a single item for
measuring self-reported disability using an eight-level ordinal
scale. Scores from the PDDS are linearly and strongly related to
physician-administered EDSS (Kurtzke, 1983) scores (r ⫽ .78;
Learmonth, Motl, Sandroff, Pula, & Cadavid, 2013).
Demographic and clinical outcomes. Sex (0 ⫽ female, 1 ⫽
male) and age (years) were measured using a standard laboratory
scale. MS clinical course (0 ⫽ relapsing–remitting MS, 1 ⫽
progressive types of MS) and duration since confirmed MS diag-
nosis (years) were measured using the same scale, and the clinical
 QUALITY OF LIFE AND MS
course was confirmed via a form letter completed by the partici-
pant’s neurologist.
Procedure
The procedure was approved by our university’s institutional
review board. Upon return of signed informed consent and MS
verification documentation, participants were sent a battery of
questionnaires through the USPS. We further provided prestamped
and preaddressed envelopes for return postal service. The project
coordinator called to make sure the participants received the ma-
terials and understood the instructions. The participants then com-
pleted the battery of questionnaires that included the GLTEQ
(Godin & Shephard, 1985) and the SF-36 (Ware & Sherbourne,
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
1992), as well as PDDS (Hadjimichael, Kerns, Rizzo, Cutter, &
This document is copyrighted by the American Psychological Association or one of its allied publishers.
Vollmer, 2007) and demographic/clinical measures. Participants
returned the completed study materials through the USPS. We
contacted participants by telephone and e-mail as a reminder to
return the study materials up to three times. We further collected
any missing questionnaire data through follow-up telephone calls.
This same procedure was completed 6 months later. All partici-
pants received $40 remuneration, which was prorated to be $20 per
period of assessment.
Data Analysis
We used linear panel analysis and covariance modeling with the
full-information maximum likelihood (FIML) estimator in Mplus
6.0 (Muthén & Muthén, 1998 –2011). Panel analysis is a useful
analytic procedure for testing directional relationships among vari-
ables across time (Kessler & Greenberg, 1981). Covariance mod-
eling is a family of techniques that allow for testing the fit of
models that describe the nature of relationships among multiple
predictor and/or outcome variables simultaneously. The FIML
estimator was selected because there were missing data and this
estimator is an optimal method for the treatment of missing data in
covariance modeling that has yielded accurate fit indices with
simulated missing data (Enders & Bandalos, 2001).
The panel model that was tested specified direct associations
between physical activity and all eight domains of HRQOL for
both baseline and follow-up. The panel model further included
cross-lagged path coefficients between baseline and follow-up for
physical activity and all eight domains of HRQOL, respectively,
and vice versa, and stability coefficients between baseline and
follow-up for physical activity and HRQOL (this allowed for
follow-up measurements to reflect residual change over time). This
panel model provided an indication of the cross-sectional, prospec-
tive, and contemporaneous longitudinal associations between in-
traindividual covariation in physical activity and HRQOL; Figure
1 provides a simplified illustration of the full panel model we
tested for physical activity and the eight domains of HRQOL. An
important feature of this approach is that the panel model does not
require group-level changes in mean scores over time for estimat-
ing associations involving change. We further added disability
status, MS clinical course and duration, age, and sex into the model
as possible covariates. We saturated the model with direct associ-
ations between the covariates and physical activity and HRQOL
domains for baseline and follow-up.
Model fit was described using the ␹2, standardized root mean
squared residual (SRMR), and comparative fit index (CFI). Good-
417
Figure 1. Illustration of basic panel model for examining the associations
between physical activity and health-related quality of life over a 6-month
period tested in 292 persons with multiple sclerosis.
model data fit was based on the combinatory rule of SRMR ⱕ .08
and CFI ⱖ .95 (Hu & Bentler, 1999).
Results
Sample Characteristics
The baseline sample (N ⫽ 292) consisted of 245 women and 47
men. There were 246 individuals who were diagnosed with
relapsing–remitting MS and 46 persons who were diagnosed with
either primary progressive or secondary progressive MS (84%
relapsing–remitting MS). The mean age was 48.0 years (SD ⫽
10.3, range ⫽ 20 – 69 years) and the mean duration of MS (time
since definite diagnosis) was 10.3 years (SD ⫽ 7.9, range ⫽ 1–35
years). The median PDDS scale score (Hadjimichael, Kerns,
Rizzo, Cutter, & Vollmer, 2007) was 3.0 (range ⫽ 0 – 6) and this
corresponded with an EDSS score (Kurtzke, 1983) of 4.0 – 4.5 (i.e.,
gait disability without the need for an assistive device).
Descriptive Statistics
Descriptive statistics for the GLTEQ (Godin & Shephard, 1985)
and the eight domains of the SF-36 (Ware & Sherbourne, 1992) are
provided in Table 1. Paired t tests indicated that there were no
statistically significant mean changes in any of the variables across
the 6-month period (i.e., no group level changes), but there was
individual variability in change over time (i.e., intraindividual
changes) based on the standard deviation of change scores in Table
1. This indicated, for example, that some persons increased phys-
ical activity, whereas others had decreases in physical activity over
time, and is important for the subsequent panel analysis.
Panel Analysis
Model fit. The panel model represented an acceptable fit for
the data (␹2 ⫽ 140.72, df ⫽ 56, SRMR ⫽ 0.06, CFI ⫽ 0.98). The
␹2 value was statistically significant (p ⬍ .001), but the combina-
tory rule for the SRMR (i.e., ⱕ .08) and CFI (i.e., ⱖ .95; Hu &
Bentler, 1999) indicated a good fit of the panel model for the data.
There were not large modification indices that suggested mean-
ingful model modifications for improving model-data fit. The
standardized path coefficients for the panel model are provided in
Table 2.
Cross-sectional associations. The standardized path coeffi-
cients were statistically significant between baseline physical ac-
 418 MOTL AND MCAULEY

Table 1
Descriptive Statistics and Changes Across Time for the Measures in the Covariance Modeling

Measure Baseline 6-month follow-up 6-month change t p

GLTEQ 26.8 (22.2) 26.0 (23.9) 0.8 (22.7) 0.53 0.59

Physical function 59.0 (26.3) 58.2 (26.9) 0.8 (13.3) 0.95 0.34
Role–physical 43.5 (39.6) 44.9 (40.7) ⫺1.4 (36.6) ⫺0.64 0.52
Role–emotional 65.7 (40.7) 62.4 (41.9) 3.3 (42.4) 1.25 0.21
Vitality 42.8 (22.6) 43.5 (22.7) ⫺0.7 (14.1) ⫺0.81 0.42
Mental health 71.5 (18.9) 72.0 (19.3) ⫺0.5 (12.8) ⫺0.59 0.56
Bodily pain 69.0 (24.1) 68.8 (22.8) 0.2 (18.7) 0.17 0.87
Social function 55.4 (21.9) 55.0 (23.0) 0.4 (15.9) 0.41 0.68
General health 54.7 (22.5) 56.0 (21.86) ⫺1.3 (15.4) ⫺1.40 0.16
Note. GLTEQ ⫽ Godin Leisure-Time Exercise Questionnaire; values represent mean (SD); change ⫽ Baseline
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

minus follow-up scores.

This document is copyrighted by the American Psychological Association or one of its allied publishers.

tivity and baseline Physical Function (␥ ⫽ .16, p ⬍ .0001),
Role–Physical (␥ ⫽ .14, p ⬍ .005), Role–Emotional (␥ ⫽ .13, p ⬍
.05), Vitality (␥ ⫽ .20, p ⬍ .0001), Mental Health (␥ ⫽ .12, p ⬍
.05), Social Function (␥ ⫽ .18, p ⬍ .001), and General Health (␥ ⫽
.20, p ⬍ .0001), but not Bodily Pain (␥ ⫽ .07, p ⫽ .21). Those who
reported higher initial levels of physical activity reported better
initial HRQOL on seven of eight domains on the SF-36 (Ware &
Sherbourne, 1992).
Cross-lagged associations. There were no statistically signif-
icant cross-lagged path coefficients between baseline physical
activity and follow-up Physical Function (␥ ⫽ .01, p ⫽ .86),
Role–Physical (␥ ⫽ ⫺.01, p ⫽ .94), Role–Emotional (␥ ⫽ ⫺.09,
p ⫽ .16), Vitality (␥ ⫽ ⫺.05, p ⫽ .33), Mental Health (␥ ⫽ .01,
p ⫽ .83), Bodily Pain (␥ ⫽ ⫺.03, p ⫽ .66), Social Function (␥ ⫽
.00, p ⫽ .95), and General Health (␥ ⫽ ⫺.02, p ⫽ .66). There were
no statistically significant cross-lagged path coefficients between
baseline Physical Function (␥ ⫽ .18, p ⫽ .09), Role–Physical (␥ ⫽
.08, p ⫽ .26), Role–Emotional (␥ ⫽ ⫺.06, p ⫽ .39), Vitality (␥ ⫽
.18, p ⫽ .06), Mental Health (␥ ⫽ .00, p ⫽ .97), Bodily Pain
(␥ ⫽ ⫺.01, p ⫽ .93), Social Function (␥ ⫽ ⫺.12, p ⫽ .26), and
General Health (␥ ⫽ ⫺.11, p ⫽ .20) with follow-up physical
activity. This indicated that baseline physical activity did not
predict intraindividual change of HRQOL over time, and vice
versa.
Contemporaneous, longitudinal associations. The standard-
ized path coefficients were statistically significant between
follow-up physical activity and follow-up Physical Function (␤ ⫽
.12, p ⬍ .005), Role–Emotional (␤ ⫽ .16, p ⬍ .01), Vitality (␤ ⫽
.13, p ⬍ .001), and Social Function (␤ ⫽ .12, p ⬍ .05), but not
Role–Physical (␤ ⫽ .07, p ⫽ .27), Bodily Pain (␤ ⫽ .03, p ⫽ .64),
Mental Health (␤ ⫽ .03, p ⫽ .64), and General Health (␤ ⫽ .03,
p ⫽ .60). This indicated that those persons who reported a change
(increased or decreased) in levels of physical activity over six
months reported concurrent change (improved or worsened, re-
spectively) in HRQOL on four of eight domains on the SF-36
(Ware & Sherbourne, 1992).
Covariates. The only variable that had significant effects on
physical activity and HRQOL was disability status; the coeffi-
cients for all other variables were nonsignificant. Disability status
had significant effects on baseline physical activity (␥ ⫽ ⫺.21,
p ⬍ .0001) and baseline Physical Function (␥ ⫽ ⫺.72, p ⬍ .0001),
Role–Physical (␥ ⫽ ⫺.51, p ⬍ .0001), Role–Emotional
(␥ ⫽ ⫺.16, p ⬍ .01), Vitality (␥ ⫽ ⫺.39, p ⬍ .0001), Mental
Health (␥ ⫽ ⫺.13, p ⬍ .05), Bodily Pain (␥ ⫽ ⫺.42, p ⬍ .0001),
Social Function (␥ ⫽ ⫺.42, p ⬍ .0001), and General Health
(␥ ⫽ ⫺.34, p ⬍ .0001). Those with worse disability had lower
levels of physical activity and worse HRQOL across all eight
domains.
Disability status further had statistically significant effects on
follow-up Physical Function (␥ ⫽ ⫺.26, p ⬍ .0001), Role–
Physical (␥ ⫽ ⫺.41, p ⬍ .0001), Role–Emotional (␥ ⫽ ⫺.22, p ⬍
.01), Vitality (␥ ⫽ ⫺.10, p ⬍ .05), Bodily Pain (␥ ⫽ ⫺.13, p ⬍
.05), and Social Function (␥ ⫽ ⫺.21, p ⬍ .0001), but not General
Health (␥ ⫽ ⫺.02, p ⫽ .74) and Mental Health (␥ ⫽ .01, p ⫽ .82).

Table 2
Path Coefficients for the Associations Between Physical Activity and Domains of Health-Related Quality of Life (HRQOL) Tested
Using Panel Analysis on Two Time Points Separated by 6 Months in 292 Persons With Multiple Sclerosis

Cross-sectional path coefficient Cross-lagged path coefficient Cross-lagged path coefficient Contemporaneous, longitudinal path
HRQOL (baseline physical activity ¡ (baseline physical activity ¡ (baseline HRQOL ¡ follow- coefficient (follow-up physical
domain baseline HRQOL) follow-up HRQOL) up physical activity) activity ¡ follow-up HRQOL)

Physical function .16 (p ⬍ .0001) .01 (p ⫽ .86) .18 (p ⫽ .09) .12 (p ⬍ .005)
Role–physical .14 (p ⬍ .005) –.01 (p ⫽ .94) .08 (p ⫽ .26) .07 (p ⫽ .27)
Role–emotional .13 (p ⬍ .05) –.09 (p ⫽ .16) –.06 (p ⫽ .39) .16 (p ⬍ .01)
Vitality .20 (p ⬍ .0001) –.05 (p ⫽ .33) .18 (p ⫽ .06) .13 (p ⬍ .001)
Mental health .12 (p ⬍ .05) .01 (p ⫽ .83) .00 (p ⫽ .97) .03 (p ⫽ .64)
Bodily pain .07 (p ⫽ .21) –.03 (p ⫽ .66) –.01 (p ⫽ .93) .03 (p ⫽ .64)
Social function .18 (p ⬍ .001) .00 (p ⫽ .95) –.12 (p ⫽ .26) .12 (p ⬍ .05)
General health .20 (p ⬍ .0001) –.02 (p ⫽ .66) –.11 (p ⫽ .20) .03 (p ⫽ .60)
Note. Values are standardized path coefficients (p value indicating significance of path coefficient).
 QUALITY OF LIFE AND MS
Those with worse disability had a greater reduction in six of eight
domains of HRQOL over time. There was no statistically signif-
icant association between disability status and follow-up physical
activity (␥ ⫽ .12, p ⫽ .17).
Discussion
The present panel study examined the cross-sectional, prospec-
tive, and longitudinal associations between free-living physical
activity and domains of HRQOL over a 6-month period in persons
with MS. This panel study was based on previous research in older
adults (e.g., Rejeski & Mihalko, 2001) and the need for additional
research on physical activity and domains of HRQOL for inform-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ing decisions regarding management of MS patients in clinical
This document is copyrighted by the American Psychological Association or one of its allied publishers.
practice (Benito-León, 2011). To that end, we observed both
cross-sectional and longitudinal, but not prospective, associations
between physical activity and four of eight domains of HRQOL,
even after controlling for covariates, including disability status, a
variable that has consistently predicted HRQOL in MS (Benito-
León et al., 2003; Mitchell et al., 2005). Of note, the longitudinal
associations indicated that change in physical activity yielded a
corresponding change in four specific domains of HRQOL over 6
months, and three of the four domains were associated with the
mental component (i.e., Role–Emotional, Vitality, and Social
Function) of HRQOL; the other domain (i.e., General Health) was
associated with the physical component. This might suggest that
free-living physical activity and its change over time has greater
benefits for the mental components of HRQOL. Such a possibility
is consistent with evidence in older adults (Netz, Wu, Becker, &
Tenenbaum, 2005), and perhaps this occurs through changes in
perceptions of control brought about by physical activity (McAu-
ley et al., 2006, 2008; Motl, McAuley, Snook, & Gliottoni, 2008;
Motl & Snook, 2008; Netz et al., 2005; Rejeski & Mihalko, 2001).
The possibility of stronger benefits for domains corresponding
with mental components of HRQOL could be further examined
with measures of depression and anxiety, and requires verification
in RCTs that change free-living physical activity and examine its
influence on domains of HRQOL as well as mediator variables
(e.g., self-efficacy) in MS. If physical activity does influence
domains of HRQOL, this would provide evidence for the promo-
tion of free-living physical activity as an approach for managing
reduced HRQOL in MS patients by rehabilitation psychologists.
This is critical given that the promotion of this behavior might
become part of the clinical armamentarium of clinicians who treat
MS patients (Benito-León, 2011), including rehabilitation psychol-
ogists.
This study has extended previous research on physical activity
and quality of life (QOL) in MS. Indeed, previous research has
largely focused on free-living physical activity in association with
global QOL (i.e., satisfaction with life; e.g., Motl & McAuley,
2009), or summary physical and mental components of HRQOL
(Motl et al., 2008), rather than specific domains of HRQOL. For
example, we previously reported a small association between
change in physical activity and change in global QOL (standard-
ized path coefficient ⫽ .07) in the same sample of 292 persons
with MS (Motl & McAuley, 2009). Herein, we report slightly
larger associations between changes in physical activity and four
domains of HRQOL with standardized path coefficienst that
ranges between .12 and .16. This greater magnitude of associations
419
for HRQOL is logical considering that global QOL is a more distal
outcome, whereas HRQOL is a more proximal outcome associated
with the possible benefits of physical activity (McAuley et al.,
2006, 2008; Stewart & King, 1991). This argument has been
supported empirically in previous research among older adults
(McAuley et al., 2006, 2008), and assists with interpreting the
results reported herein when compared with previous research. To
that end, future studies might consider HRQOL and its specific
domains as more proximal outcomes of change in free-living
physical activity within the context of RCTs, and compare such
changes against changes in global QOL or summary components
of HRQOL. This is particularly relevant for rehabilitation psychol-
ogists who might focus on the remediation of mental health prob-
lems in MS.
This study focused on free-living physical activity and its asso-
ciation with domains of HRQOL over time. We note here that the
correlations between changes in physical activity and HRQOL,
when statistically significant, ranged between .12 and .16 in this
study; these correlations can be converted into Cohen’s effect size
d (resulting range of d between 0.24 and 0.32) for comparison with
a previous meta-analysis of exercise training on QOL outcomes in
MS (Motl & Gosney, 2008). That meta-analysis reported a mean
effect size across 13 studies (both RCTs and quasi-experimental
designs) of 0.24 for an improvement in QOL outcomes, particu-
larly after supervised exercise training. In addition, the association
between change in physical activity and Vitality approximated a
Cohen’s effect size d value of .26 and this is consistent with the
meta-analysis results for the effects of exercise training on fatigue
as a domain of QOL (d ⫽ 0.19). Accordingly, there seemed to be
some congruity regarding the association between exercise train-
ing and free-living physical activity with aspects of QOL in MS,
although this would require direct, head-to-head comparison in an
RCT. If there are congruent associations, this would open the door
for a range of options when promoting physical activity and
exercise training as approaches for improving HRQOL in MS.
Regarding the covariates, we observed that disability status
had cross-sectional and prospective associations with domains
of HRQOL, and this is consistent with previous research. In-
deed, disability is a primary cause of reduced QOL in persons
with MS (Amato et al., 2001; Benedict et al., 2005; Lobentanz
et al., 2004) and is characterized by a restriction or inability to
perform activities in the manner considered normal. Disability
is frequently measured among those with MS by the EDSS
(Kurtzke, 1983). For example, EDSS scores were negatively
correlated with overall QOL and physical aspects of health-
related QOL in cross-sectional studies of patients with MS
(Amato et al., 2001; Benedict et al., 2005; Lobentanz et al.,
2004). The effect of disability on QOL has typically remained
statistically significant, even when controlling for other factors
such as fatigue, cognitive impairment, anxiety, depression, and
social support (Amato et al., 2001; Benedict et al., 2005;
Lobentanz et al., 2004). We note that the other covariates of MS
duration and clinical status, sex, and age did not exhibit asso-
ciations, either cross-sectional or prospective, with domains of
HRQOL in the present study. This conflicts with previous
research, but can be explained by the simultaneous consider-
ation of all covariates in the panel analysis. This indicates that
disability status is a prominent predictor of HRQOL in MS, but
disability status is not easily modified with intervention, and
 420 MOTL AND MCAULEY
this observation highlights the importance of our results regard-
ing physical activity. Physical activity is modifiable, and has
been changed in persons with MS through behavioral interven-
tions that teach self-management skills, strategies, and tech-
niques for behavior change delivered using the Internet (Motl et
al., 2011), telephone (Bombardier et al., 2013) and pamphlets
(Plow, Bethoux, McDaniel, McGlynn, & Marcus, 2014). For
example, behavioral interventions could teach and promote the
practice of self-monitoring using a pedometer along with goal
setting for increasing physical activity behavior in MS (Motl, in
press). Such an approach could be incorporated into the clinical
practices of rehabilitation psychologists for increasing physical
activity in persons with MS.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
There are some important limitations of the present prospec-
This document is copyrighted by the American Psychological Association or one of its allied publishers.
tive panel study. We only included a generic measure for
capturing HRQOL domains, and the results might have differed
based on a disease-targeted measure, although this has not been
supported in previous research on physical activity and QOL in
MS (Motl et al., 2008). We included a self-report measure of
disability status rather than a measure based on a neurological
examination performed by a clinician. In addition, we only
measured physical activity and HRQOL over a relatively short
period (6 months); this could be lengthened in future research,
thereby yielding larger changes in the focal variables for ex-
amining associations over time. The sample primarily consisted
of Caucasian women with relapsing–remitting MS; this is gen-
erally consistent with the demographics of MS. We fully ac-
knowledge that future researchers should consider using a more
diverse sample. This would allow for a broader generalization
of the findings among less representative groups of individuals
with MS, such as men and persons who are not Caucasian. One
possible confound might be the overlap in item content between
the GLTEQ (Godin & Shephard, 1985) and SF-36 (Ware &
Sherbourne, 1992). Indeed, the SF-36 contains items regarding
physical activity, but this differs from the GLTEQ in the nature
of the questions and prompt. The items and prompt on the
SF-36 are focused toward how one’s health status influences or
limits participation in activities of daily living that range from
bathing and dressing through running, rather than on actual
leisure-time physical activity participation. This suggests that
the GLTEQ and SF-36 have limited overlap, and the SF-36 has
been recognized as an ideal measure of HRQOL for physical
activity research (Rejeski & Mihalko, 2001).
The evidence of a positive relationship between physical
activity and domains of HRQOL in those with MS has impli-
cations for health-care professionals including clinicians, reha-
bilitation psychologists, physical therapists, and exercise spe-
cialists. Those professionals might consider focusing on
promoting free-living physical activity as a method for improv-
ing domains of HRQOL among those with MS. This is partic-
ularly important, given that persons with MS are less physically
active than the general population (Motl et al., 2005) and have
more compromised HRQOL (Benito-León et al., 2003; Mitchell
et al., 2005). Such continued research will advance our under-
standing of methods of increasing physical activity and maxi-
mizing the associated HRQOL benefits in those with MS. This
is critical, given calls for understanding modifiable targets of
interventions for improving HRQOL as a patient-reported out-
come in MS (Baumstarck et al., 2009), and physical activity has
been identified as a primary, but largely understudied target
(Benito-León, 2011).
References
Amato, M. P., Ponziani, G., Rossi, F., Liedl, C. L., Stefanile, C., & Rossi,
L. (2001). Quality of life in multiple sclerosis: The impact of depression,
fatigue and disability. Multiple Sclerosis, 7, 340 –344. doi:10.1177/
135245850100700511
Baumstarck, K., Boyer, L., Boucekine, M., Michel, P., Pelletier, J., &
Auquier, P. (2013). Measuring quality of life in patients with multiple
sclerosis in clinical practice: A necessary challenge. Multiple Sclerosis
International, 2013, Article ID 524894 (8 pp.). doi:10.1155/2013/
524894
Benedict, R. H., Wahlig, E., Bakshi, R., Fishman, I. Munschauer, F.,
Zivadinov, R., & Weinstock-Guttman, B. (2005). Predicting quality of
life in multiple sclerosis: Accounting for physical disability, fatigue,
cognition, mood disorder, personality, and behavior change. Journal of
the Neurological Sciences, 231, 29 –34. doi:10.1016/j.jns.2004.12.009
Benito-León, J. (2011). Physical activity in multiple sclerosis: The missing
prescription. Neuroepidemiology, 36, 192–193. doi:10.1159/000328276
Benito-León, J., Morales, J. M., Rivera-Navarro, J., & Mitchell, A. J.
(2003). A review about the impact of multiple sclerosis on health-related
quality of life. Disability and Rehabilitation, 25, 1291–1303. doi:
10.1080/09638280310001608591
Bombardier, C. H., Edhe, D. M., Gibbons, L. E., Wadhwani, R., Sullivan,
M. D., Rosenberg, D. E., . . . Kraft, G. E. (2013). Telephone-based
physical activity counseling for major depression in people with multiple
sclerosis. Journal of Consulting and Clinical Psychology, 81, 89 –99.
doi:10.1037/a0031242
Bouchard, C., & Shephard, R. J. (1994). Physical activity, fitness, and
health: The model and key concepts. In C. Bouchard, R. J. Shephard, &
T. Stephens (Eds.), Physical activity, fitness, and health: International
proceedings and consensus statement (pp. 11–20). Champaign, IL: Hu-
man Kinetics Publishers.
Enders, C. K., & Bandalos, D. L. (2001). The relative performance of full
information maximum likelihood estimation for missing data in struc-
tural equation models. Structural Equation Modeling, 8, 430 – 457. doi:
10.1207/S15328007SEM0803_5
Godin, G., & Shephard, R. J. (1985). A simple method to assess exercise
behavior in the community. Canadian Journal of Applied Sport Sci-
ences, 10, 141–146.
Gosney, J. L., Scott, J. A., Snook, E. M., & Motl, R. W. (2007). Physical
activity and multiple sclerosis: Validity of self-report and objective
measures. Family and Community Health, 30, 144 –150. doi:10.1097/01
.FCH.0000264411.20766.0c
Hadjimichael, O., Kerns, R. D., Rizzo, M. A., Cutter, G., & Vollmer, T.
(2007). Persistent pain and uncomfortable sensations in persons with
multiple sclerosis. Pain, 127, 35– 41. doi:10.1016/j.pain.2006.07.015
Hu, L., & Bentler, P. M: (1999). Cutoff criteria for fit indices in covariance
structure analysis: Conventional criteria versus new alternatives. Struc-
tural Equation Modeling, 6, 1–55. doi:10.1080/10705519909540118
Kessler, R. C., & Greenberg, D. F. (1981). Linear panel analysis: Models
of quantitative change. New York, NY: Academic Press, Inc.
Krokavcova, M., van Dijk, J. P., Nagyova, I., Rosenberger, J., Gavelova,
M., Gdovinova, Z., & Groothoff, J. W. (2009). Perceived health status as
measured by the SF-36 in patients with multiple sclerosis: A review.
Scandinavian Journal of Caring Sciences, 23, 529 –538. doi:10.1111/j
.1471-6712.2008.00633.x
Kurtzke, J. F. (1983). Rating neurologic impairment in multiple sclerosis:
An Expanded Disability Status Scale (EDSS). Neurology, 33, 1444 –
1452. doi:10.1212/WNL.33.11.1444
Lankhorst, G. J., Jelles, F., Smits, R. C., Polman, C. H., Kuik, D. J.,
Pfennings, L. E., . . . Vleugels, L. (1996). Quality of life in multiple
 QUALITY OF LIFE AND MS
sclerosis: The disability and impact profile (DIP). Journal of Neurology,
243, 469 – 474. doi:10.1007/BF00900502
Learmonth, Y. C., Motl, R. W., Sandroff, B. M., Pula, J. H., & Cadavid, D.
(2013). Validation of Patient Determined Disease Steps (PDDS) scale
scores in persons with multiple sclerosis. BMC Neurology, 13, 37– 44.
doi:10.1186/1471-2377-13-37
Lobentanz, I. S., Asenbaum, S., Vass, K., Sauter, C., Klosch, G., Kristofer-
itsch, W., . . . Zeitlhofer, J. (2004). Factors influencing quality of life in
multiple sclerosis patients: Disability, depressive mood, fatigue and
sleep quality. Acta Neurologica Scandinavica, 110, 6 –13. doi:10.1111/
j.1600-0404.2004.00257.x
McAuley, E., Doerksen, S. E., Morris, K. S., Motl, R. W., Hu, L.,
Wojcicki, T. R., . . . Rosengren, K. R. (2008). Pathways from physical
activity to quality of life in older women. Annals of Behavioral Medi-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
cine, 36, 13–20. doi:10.1007/s12160-008-9036-9
McAuley, E., Konopack, J. F., Motl, R. W., Morris, K. S., Doerksen, S. E.,
This document is copyrighted by the American Psychological Association or one of its allied publishers.
& Rosengren, K. R. (2006). Physical activity and quality of life in older
adults: Influence of health status and self-efficacy. Annals of Behavioral
Medicine, 31, 99 –103. doi:10.1207/s15324796abm3101_14
Mitchell, A. J., Benito-León, J., González, J. M., & Rivera-Navarro, J.
(2005). Quality of life and its assessment in multiple sclerosis: Integrat-
ing physical and psychological components of well-being. Lancet Neu-
rology, 4, 556 –566. doi:10.1016/S1474-4422(05)70166-6
Motl, R. W. (in press). Lifestyle physical activity in persons with multiple
sclerosis: The new kid on the MS block. Multiple Sclerosis.
Motl, R. W., Dlugonski, D., Wójcicki, T. R., McAuley, E., & Mohr, D. C.
(2011). Internet intervention for increasing physical activity in persons
with multiple sclerosis. Multiple Sclerosis, 17, 116 –128. doi:10.1177/
1352458510383148
Motl, R. W., & Gosney, J. L. (2008). Effect of exercise training on quality
of life in multiple sclerosis: A meta-analysis. Multiple Sclerosis, 14,
129 –135. doi:10.1177/1352458507080464
Motl, R. W., & McAuley, E. (2009). Pathways between physical activity
and quality of life in adults with multiple sclerosis. Health Psychology,
28, 682– 689. doi:10.1037/a0015985
Motl, R. W., McAuley, E., & Sandroff, B. M. (2013). Longitudinal change
in physical activity and its correlates in relapsing–remitting multiple
sclerosis. Physical Therapy, 93, 1037–1048. doi:10.2522/ptj.20120479
Motl, R. W., McAuley, E., & Snook, E. M. (2005). Physical activity and
multiple sclerosis: A meta-analysis. Multiple Sclerosis, 11, 459 – 463.
doi:10.1191/1352458505ms1188oa
Motl, R. W., McAuley, E., Snook, E. M., & Gliottoni, R. C. (2008). Does
the relationship between physical activity and quality of life differ based
421
on generic versus disease-targeted instruments? Annals of Behavioral
Medicine, 36, 93–99. doi:10.1007/s12160-008-9049-4
Motl, R. W., McAuley, E., Snook, E. M., & Scott, J. A. (2006). Validity of
physical activity measures in ambulatory individuals with multiple scle-
rosis. Disability and Rehabilitation, 28, 1151–1156. doi:10.1080/
09638280600551476
Motl, R. W., & Snook, E. M. (2008). Physical activity, self-efficacy, and
quality of life in multiple sclerosis. Annals of Behavioral Medicine, 35,
111–115. doi:10.1007/s12160-007-9006-7
Muthén, L. K., & Muthén, B. O. (1998 –2011). Mplus. [Computer soft-
ware]. Los Angeles, CA: Author.
Naess, H., Beiske, A. G., & Myhr, K. M. (2008). Quality of life among
young patients with ischaemic stroke compared with patients with mul-
tiple sclerosis. Acta Neurologica Scandinavica, 117, 181–185. doi:
10.1111/j.1600-0404.2007.00923.x
Netz, Y., Wu, M., Becker, B. J., & Tenenbaum, G. (2005). Physical activity
and psychological well-being in advanced age: A meta-analysis of
intervention studies. Psychology and Aging, 20, 272–284. doi:10.1037/
0882-7974.20.2.272
Plow, M., Bethoux, F., McDaniel, C., McGlynn, M., & Marcus, B. (2014).
Randomized controlled pilot study of customized pamphlets to promote
physical activity and symptom self-management in women with multiple
sclerosis. Clinical Rehabilitation, 28, 139 –148. doi:10.1177/
0269215513494229
Rejeski, W. J., & Mihalko, S. L. (2001). Physical activity and quality of life
in older adults. Journals of Gerontology: Series A. Biological Sciences
and Medical Sciences, 56, 23–35. doi:10.1093/gerona/56.suppl_2.23
Rudick, R. A., Miller, D., Clough, J. D., Gragg, L. A., & Farmer, R. G.
(1992). Quality of life in multiple sclerosis. Comparison with inflam-
matory bowel disease and rheumatoid arthritis. Archives of Neurology,
49, 1237–1242. doi:10.1001/archneur.1992.00530360035014
Stewart, A. L., & King, A. C. (1991). Evaluating the efficacy of physical
activity for influencing quality-of-life outcomes in older adults. Annals
of Behavioral Medicine, 13, 108 –116.
Ware, J. E., Jr., & Sherbourne, C. (1992). The MOS 36-item Short-Form
Health Survey (SF-36). Medical Care, 30, 473– 483. doi:10.1097/
00005650-199206000-00002
Received October 24, 2013
Revision received July 11, 2014
Accepted July 21, 2014 䡲