Journal of Physical Activity and Health, 2015, 12, 924 -930 http://dx.doi.org/10.1123/jpah.2014-0018 © 2015 Human Kinetics, Inc.
ORIGINAL RESEARCH
Pilot Trial of a Social Cognitive Theory-Based Physical Activity Intervention Delivered by Nonsupervised Technology in Persons With Multiple Sclerosis Yoojin Suh, Robert W. Motl, Connor Olsen, and Ina Joshi Background: Physical inactivity is prevalent in people with multiple sclerosis (MS) and this highlights the importance of developing behavioral interventions for increasing physical activity (PA) in MS. This pilot trial examined the efficacy of a 6-week, behavioral intervention based on social cognitive theory (SCT) delivered by newsletters and phone calls for increasing PA in persons with MS who were physically inactive and had middle levels of self-efficacy. Methods: The sample included 68 persons with relapsing-remitting MS who were randomly assigned into intervention and control groups. The intervention group received SCT-based information by newsletters and phone calls, whereas the controls received information regarding topics such as stress management over 6 weeks. Participants completed self-report of PA and social cognitive variables. Results: The intervention group had a significant increase in self-reported PA (d = 0.56, P = .02) over the 6 weeks, but the controls had a nonsignificant change (d = –0.13, P = .45). Goal setting was changed in the intervention group (d = 0.68, P ≤ .01) and identified as a significant mediator of change in self-reported PA. Conclusions: This study provides initial evidence for the benefit of a theory-based behavioral intervention for increasing PA in MS. Keywords: exercise, rehabilitation, neurological disease, self-management
Physical activity has been associated with improvements in fitness outcomes, symptoms, and quality of life among persons with multiple sclerosis (MS),1 yet physical inactivity is particularly common in this neurological disease.2 Those observations underscore the importance of designing behavioral interventions for increasing physical activity among persons with MS.3 Behavioral interventions involve teaching persons the skills, techniques, and strategies that support successful health behavior change,4 and have been recognized as approaches for overcoming physical inactivity in those with chronic disease conditions.3 To be effective, behavioral interventions should be based on theoretical models that guide the identification and selection of variables as targets for changing the outcome of interest.4 Social cognitive theory (SCT)5 has been a prominent theoretical model for effectively explaining and predicting physical activity in the general population6 and those with Parkinson’s disease and MS.3 Recently, Bandura has operationalized a SCT-based model of associations among self-efficacy (ie, beliefs about confidence in undertaking a specific health behavior), multidimensional outcome expectations (ie, beliefs about the physical, social, and self-evaluative benefits of a health behavior), sociostructural factors (ie, facilitators and impediments of undertaking a health behavior), goal-setting (ie, self-management strategy for undertaking a health behavior), and health behavior and its change. This model indicates that self-efficacy has direct and indirect effects on physical activity as a health behavior, and the indirect effects operate via multidimensional outcome expectations, impediments and facilitators, and goal-setting.5 There has been research supporting Bandura’s model5 and its variables as correlates of physical activity in cross-sectional7 and longitudinal8 examinations of persons with MS. The authors are with the Dept of Kinesiology and Community Health, University of Illinois at Urbana-Champaign. Motl ([email protected]) is corresponding author. 924
Beyond theory, behavioral interventions can be delivered through non-face-to-face methods. We are aware of research on delivering behavioral interventions for improving physical activity in MS via print materials and telephone calls. For example, one recent paper demonstrated that print media delivered materials based on the transtheoretical model (TTM) resulted in a possible improvement of physical activity levels and function in women with MS.9 Another study delivered motivational interviewing through telephone calls as a method of improving health promoting activities, including physical activity, among persons with MS.10 Such interventions have been associated with increases in physical activity, but we believe that there are several ways for extending this line of research on behavioral interventions in MS. Bandura’s 3-fold stepwise implementation model should be considered when developing and delivering behavior interventions for increasing physical activity.5 This implementation model explains that the type of intervention materials can be tailored to 3 levels of a person’s self-efficacy and motivational readiness to achieve behavior change.5 For example, the implementation model suggests that tailored-print materials and telephone interactions should be ideally suited for those who have middle levels of selfefficacy (ie, persons meeting the second level of the 3-fold model). Those within the first level of the 3-fold model require very little assistance with behavioral change, whereas those within the third level of the 3-fold model require intensive intervention for behavior change. Another consideration involves examining mediators of behavioral interventions that can inform the design and tailoring of future research by streamlining the content. One last consideration involves inclusion of a control condition that receives print materials and social contact on the telephone, but does not receive information on physical activity change. This pilot, randomized controlled trial (RCT) examined the efficacy of a SCT-based intervention delivered by newsletters and telephone calls for increasing physical activity in persons with MS who had middle levels of self-efficacy, but were not regularly
Physical Activity Intervention in MS 925
physically active. This study further examined mediators of intervention effects on physical activity change, particularly self-efficacy, outcome expectations, functional limitations, social support, and goal-setting.
Methods
Downloaded by Purdue Univ on 09/17/16, Volume 12, Article Number 7
Recruitment We distributed a flyer through e-mail among those who participated in a recently- completed longitudinal study of physical activity in persons with relapsing-remitting MS (RRMS).11 Briefly, the sample in that previous study consisted of 223 women and 46 men. The mean age was 45.9 years (SD = 9.6), and the duration of MS (time since diagnosis) was 8.8 years (SD = 7.0). The median Patient Determined Disease Scale (PDDS) score was 2 (range = 0–6) and corresponded with moderate disability (ie, no limitations in walking, but significant problems that limit daily activities).12 Persons who were interested in the current study contacted our laboratory for information and screening for inclusion. The inclusion criteria were definite diagnosis of RRMS; independently ambulatory or ambulatory with single point assistance (eg, cane); relapse free in the past 30 days; being nonactive defined as not engaging in regular physical activity (ie, 30 minute-accumulated per day) on more than 2 days of the week during the previous 6 months; free of contraindications for physical activity (eg, no underlying cardiovascular disease) based on Physical Activity Readiness Questionnaire;13 having the visual ability necessary to read 14 point font; meeting the age requirement (ie, 18 to 64 years of age). We focused on RRMS because it is the most common type of MS and persons are seemingly more able to engage in physical activity via behavioral intervention than other types of MS. One final screening criterion was to have middle levels
of exercise self-efficacy. The criterion for middle level of efficacy was determined based on the distribution of Exercise Self-Efficacy Scale (EXSE) scores from our previous longitudinal study.11 Based on a frequency analysis, the middle category of overall scores on the EXSE ranged between 50 and 70, and this was operationalized as the middle level of exercise self-efficacy. Power Analysis. The proposed sample size for this RCT was 60 persons with RRMS. The sample size was based on published self-reported physical activity mean scores (ie, GLTEQ) from a behavioral intervention for detecting a moderate effect size for the interaction term (d = 0.50)14 with assumptions of α = .05, β = .10 for 90% power, and reliability of .90. The power analysis indicated a sample of 50 participants, and we opted for a 20% increase in sample size considering the attrition rate from other nonsupervised technology interventions.15 The final sample for adequate power considering drop-out and replacement of missing follow-up data were 60. We did not perform a power analysis for self-efficacy or other SCT variables. The flow of participants is presented using CONSORT diagram in Figure 1.
Measures Physical Activity. Physical activity was measured by the Godin Leisure-Time Exercise Questionnaire (GLTEQ).16 Researchers have provided evidence for the validity of this measure in persons with MS.17,18 The GLTEQ is a self-administered measure of usual physical activity. The GLTEQ has 3 items that measure the frequency of strenuous (eg, jogging), moderate (eg, fast walking), and mild (eg, easy walking) exercise for periods of more than 15 minutes during one’s free time in a typical week. The weekly frequencies of strenuous, moderate, and mild activities are multiplied by 9, 5, and 3 metabolic equivalents, respectively, and summed to form a
Figure 1 — Consort diagram. JPAH Vol. 12, No. 7, 2015
926 Suh et al
measure of total leisure activity. The scores range between 0 and 119 in arbitrary units.
Downloaded by Purdue Univ on 09/17/16, Volume 12, Article Number 7
Self-Efficacy. Self-efficacy was assessed by the EXSE.19 This scale has 6 items that assess an individual’s belief in his or her ability to engage in 20+ minutes of moderate physical activity 3 times per week, in 1-month increments, across the next 6 months. The items are rated on a scale from 0 (Not at all confident) to 100 (Completely confident) and averaged into a composite score that ranges between 0 and 100. Higher scores reflect greater confidence in one’s ability to regularly engage in exercise. Outcome Expectations. Outcome expectations for exercise were measured by the Multidimensional Outcomes Expectations for Exercise Scale (MOEES).20,21 This scale contains 15 items that reflect 3 subdomains of outcome expectations. Six items reflect physical outcome expectations, 4 items assess social outcome expectations, and 5 items measure self-evaluative outcome expectations. The 15 items were rated on a 5-point scale from 1 (Strongly disagree) to 5 (Strongly agree) and summed to form the subscale measures of outcome expectations. Functional Limitations. Functional limitations represented
impediments for physical activity and was measured using the Functional Limitations component of the abbreviated Late-Life Function and Disability Instrument (LL-FDI).22 The Functional Limitations component contains 15 items that correspond with advanced lower extremity function, basic lower extremity function, and upper extremity function. The 15-items are rated on a 5-point scale ranging from 1 (none) to 5 (cannot do). Goal-Setting. Exercise goals were measured by the Exercise
Goal-setting Scale (EGS).23 This contains 10 items that reflect goalsetting for exercise behavior and the items are rated on a 5-point scale from 1 (Does not describe) to 5 (Describes completely). The item scores are summed into an overall score that ranges between 10 and 50. Higher scores reflect a stronger tendency for setting goals for exercise.
Social Support. Social support for physical activity was measured
by the 12-item Social Support and Exercise Survey (SSES).24 The items on the SESS are rated on a 5-point scale ranging from 1 (None) to 5 (Very often), and family and friends scales are scored by summing total of items 1 to 12. In addition, perceptions of social support were measured by the 6-item modified Social Provisions Scale (SPS).25 The items on the SPS are rated on a 4-point scale ranging from 1 (strongly disagree) to 4 (strongly agree), and item scores are summed to form an overall score that ranges between 6 and 24. Disability. Disability was measured using the Patient Determined
Disease Steps (PDDS) scale.12 The PDDS is a self-report questionnaire that contains a single item for measuring disability using an ordinal scale from 0 (Normal) through 8 (Bedridden). This scale was developed as an inexpensive surrogate for the Expanded Disability Status Scale (EDSS) and scores from the PDDS are linearly and strongly related with physician-administered EDSS scores.26
Intervention Materials Printed Newsletters. Each newsletter targeted a specific agent of SCT for increasing physical activity in persons with MS. The purpose of the newsletters was to offer information about key determinants of physical activity that motivate physical activity change for persons with MS. Newsletters were delivered through United States Postal Service (USPS) and e-mail. The first newsletter focused on
the physical and psychological benefits of physical activity as well as self-monitoring among persons with MS. The second newsletter was then given for description of goal-setting for physical activity. The third newsletter contained how to improve exercise self-efficacy as well as description of 4 sources in enhancing exercise self-efficacy (ie, mastery experiences, social modeling, social persuasion, and interpretations of physiological and affective responses). The fourth newsletter included information that developed appropriate outcome expectations of being physically active. The fifth and sixth newsletters targeted barriers/impediments to physical activity and social support/facilitators for physical activity by providing guidance on overcoming barriers and identifying facilitators of physical activity. Telephone Calls. Telephone calls were made to all participants
in this study for a 15-minute discussion after the delivery of each newsletter. The same research assistants delivered the intervention materials through the telephone calls during this study (ie, research assistants were yoked with participants throughout the study). The main purpose of telephone calls was to reinforce information on the newsletter and help participants understand the content, and apply it to their daily routines. This occurred through a semiscripted phone interview. In addition, the phone calls included coaching on how to set-up a goal with a pedometer as well as monitoring average step counts per participant for meeting the goal.
Pedometer. Pedometers (Yamax SW-200, Yamasa Tokei Keiki
Co., Ltd, Tokyo, Japan) and log-books were delivered to the intervention group for the purpose of self-monitoring and tracking physical activity. The pedometer and log-book further assisted in setting-up goals as a self-regulatory strategy for increasing physical activity.
Control Materials Participants in the control group received a total of 6 newsletters designed to provide information that was not related to physical activity (eg, managing stress, nutrition, allergies, blood pressure, alcohol use, and cholesterol); these were selected based on resources provided by on the website of the National MS Society. Telephone calls were further followed up to check that the participants received the newsletters as well as that the content of the newsletters were easy to follow. The topic of physical activity was not discussed during phone calls. This condition served as a control for the attention and social contact that occurred in the intervention arm of the pilot study.
Procedures The researchers contacted potential participants who expressed interest in the study and conducted the screening procedure to evaluate inclusion criteria of this study. After initial telephone contact and screening, all participants who volunteered provided a signed informed consent document. Upon receipt of informed consent, participants were sent a battery of questionnaires along with an accelerometer through USPS. The participants were further provided prestamped and preaddressed envelopes for return postal service. Participants were asked to wear the accelerometer for a 7-day period along with completing the questionnaires as baseline assessment. Upon return, all questionnaires were checked for completeness, and if any has missing data, researchers made a phone call to collect the data. Importantly, the data were collected by a person who did not deliver the intervention and control conditions (ie, blinding), and the interventionists were blinded regarding the focal outcomes of
JPAH Vol. 12, No. 7, 2015
Downloaded by Purdue Univ on 09/17/16, Volume 12, Article Number 7
Physical Activity Intervention in MS 927
the pilot trial. Participants were matched in pairs based GLTEQ and PDDS scores after completion of the baseline assessment. Participants within the matched pairs were then randomly assigned by a researcher who was uninvolved in data collection and intervention delivery to either the intervention or control conditions using a random number generator. The intervention materials were delivered in a single wave across a 6-week period and the participants in the intervention received a total of 6 newsletters developed based on active agents (eg, self-efficacy) from the SCT. The control group received 6 newsletters that included information unrelated with physical activity change. Print newsletters were mailed by USPS on Thursday each week to ensure that the newsletters arrived at the beginning of the next week. Newsletters were further emailed to participants in case of missing mail or delivery failure. After the newsletter was delivered, telephone calls were made on Monday and Tuesday for the intervention group, and the control group received calls on Wednesday and Thursday. We proceeded with follow-up assessments using the same procedures for baseline testing after delivery of the intervention and control conditions. Participants received $10 for completing the measures on each of the measurement periods.
Data Analyses All analyses were conducted in PASW Statistics, version 18.0 for Windows. We included all persons who provided baseline data and were randomized into a condition into the data analysis (ie, intent-to-treat whereby once randomized always analyzed). Any missing follow-up data were replaced with the corresponding baseline value, thereby allowing for the intent-to-treat analysis. We compared the intervention and control groups for initial differences on demographics and MS-related variables (eg, type of MS) using independent samples t tests and chi-square analyses. We further compared initial differences in physical activity and SCT variables using independent samples t tests. The effects of the intervention on physical activity and SCT variables was examined using 2 (Condition: Intervention & Control) × 2 (Time: Pre & Post) mixed factor analysis of variance (ANOVA) based on univariate F-statistic. Condition was a between-subjects factor and time was
a within-subjects factor. Interactions were decomposed using paired and independent samples t-tests. Effects sizes for F-statistic were expressed as partial eta-squared (η2p). Effects sizes based on differences in mean scores were expressed as Cohen’s d.27 Multiple linear regression analyses were conducted to examine the SCT variables as mediators. This analysis involved regressing change in physical activity on condition in the first analysis and then regressing change in physical activity on condition plus change in putative mediators in the second analysis. Mediation was based on the effects of condition on physical activity becoming reduced in magnitude and nonsignificant after controlling for the mediator variable(s). The proportion of explained variance in the outcome from the regression analysis was based on the adjusted R2.
Results Sample Characteristics 97 participants were recruited and assessed for eligibility, and 72 participants satisfied the study inclusion criteria. Of note, the mean (SD) of the EXSE scores based on the telephone screening for inclusion was 56.6 (3.9); this satisfied our definition of middlelevel of self-efficacy (ie, EXSE score between 50 and 70). Of the 72 participants who satisfied inclusion criteria, 68 participants provided informed consent and completed the questionnaires and returned the study packet at baseline; this surpassed our sample from the power analysis. The resulting 68 persons were randomly assigned into either intervention or control groups; the 4 persons who did not provide baseline data were not randomized into condition. Two people did not provide follow-up data and we replaced the missing follow-up data with baseline scores (ie, intent-to-treat analysis whereby all persons who provided baseline data and were randomized into condition were included in data analyses). Independent samples t tests and chi square statistics (χ2) indicated no initial differences in sex (P = .20), age (P = .35), weight (P = .61), height (P = .09), education (P = 1.00), race (P = .08), employment (P = .63), MS type (P = 1.00), disease severity (P = .64), and disease duration (P = .60) (see Table 1). The mean age of the total sample
Table 1 Demographic and Clinical Data at Baseline for the Intervention and Control Groups Total (n = 68)
Characteristic Sex (male/female)
Control (n = 34)
Intervention (n = 34)
12/56
8/26
4/30
49.0 (8.8)
48.0 (9.4)
50.1 (8.1)
Education (college graduate/less than college graduate)
48/20
24/10
24/10
Race (Caucasian/non-Caucasian)
65/3
31/3
34/0
Employment (unemployed/employed)
36/32
17/17
19/15
MS type (RRMS/Other)
66/2
33/1
33/1
Weight (kg)
78.9 (22)
79.8 (20.3)
77.5 (23.7)
Height (cm)
167.1 (7.1)
167.4 (5.1)
165.2 (6.0)
Time since diagnosis (years)
12.1 (7.9)
12.7 (8.8)
11.6 (7.1)
Disease severity (arbitrary units)
2.0 (1.8)
2.2 (1.8)
2.0 (1.8)
Age (years)
Note. Values represent mean (standard deviation) or number of cases for dichotomous variables; Disease severity = Patient Determined Disease Steps Scale score (PDDS). JPAH Vol. 12, No. 7, 2015
928 Suh et al
at baseline was 49 years (SD = 8.8; range 18–64). The majority of the sample was female (81%), Caucasian (96%), well-educated (47% > were college graduates), and were employed (47.1%). The sample consisted primarily of individuals with RRMS (n = 66), with the remaining individuals having SPMS (n = 1) or benign MS (n = 1). The mean time since MS diagnosis was 12.1 years (SD = 7.9, range 3–33 years) and the mean PDDS score was 2.0 (SD = 1.8, range 0–7). There were no major or obvious differences in demographic and clinical characteristics between this sample and the original cohort from which it was drawn.11 The EXSE score in this sample was considerably lower than the parent cohort (mean = 47.3 vs. mean = 77.7, respectively) by design considering that we focused on those with middle levels of efficacy when designing this intervention.
significant differences between the 2 groups on physical outcome expectations (P = .00), and self-evaluative outcome expectations (P = .02) indicating that the control group reported higher values compared with the intervention group.
Effects of Intervention on Physical Activity and SCT variables There was a statistically significant condition-by-time interaction on GLTEQ scores, F(1, 66) = 5.47, P = .02, η2p = .08. The intervention group reported a statistically significant (P = .02) and medium increase in physical activity over time (d = 0.56), whereas the control group had a small (d = –0.13) and nonsignificant (P = .45) reduction in physical activity over time. One additional analysis examined change in step counts per day from pedometers over the intervention period. The 1-way repeated-measures ANOVA indicated a statistically significant time main effect on steps per day, F(1, 32) = 8.03, P ≤ .01, η2p = .20. Overall, steps per day increased by 1368 steps per day from the first to last week of the 6-week intervention period (d = .46). The same analytic model (mixed factor ANOVA) was conducted to determine changes in social cognitive variables. There was a statistically significant condition-by-time interaction on goal setting, F(1, 66) = 16.88, P ≤ .01, η2p = .20. The intervention group reported a statistically significant and large increase in goal setting over time (d = 0.68, P ≤ .01), whereas the control group had a nonsignificant, small reduction in goal setting over time (d = –0.19, P = .14). The ANOVA did not indicate statistically significant condition-by-time interactions on exercise self-efficacy [F(1, 66) = 3.19, P = .08, η2p = .05], physical outcome expectations [F(1, 66) = 3.01, P = .09, η2p = .04], social outcome expectations [F(1, 66) = .83, P = .36, η2p = .01], self-evaluative outcome expectations [F(1, 66) = .59, P = .44, η2p = .00], functional limitations [F(1, 66) = .11, P = .73, η2p = .00], family support for exercise [F(1, 66) = 3.40, P = .07, η2p = .04], friends support for exercise [F(1, 66) = .24, P = .62, η2p = .00], and social provisions scale [F(1, 66) = .01, P = .91, η2p = .00].
Downloaded by Purdue Univ on 09/17/16, Volume 12, Article Number 7
Intervention Fidelity All 6 newsletters, regardless of intervention or control condition, were mailed on time as scheduled, and no newsletters were returned for an incorrect postal address. Telephone calls further were made as scheduled, but 9.6% and 14.6% of all phone calls were missed for intervention and control groups, respectively. Of the calls that were successfully, 87.8% were completed on the first attempt and 12.1% were completed on the second attempt. There were no third attempts for phone calls.
Group Differences in Physical Activity and SCT variables Independent samples t-tests compared possible group differences in physical activity and SCT variables at baseline (see Table 2). There were no significant differences between the intervention and control groups on pretrial values for GLTEQ scores (P = .38), activity counts (P = .21), self-efficacy (P = .68), goal setting (P = .20), social outcome expectations (P = .48), functional limitations (P = .76), family support for exercise (P = .06), friends support for exercise (P = .24), and social provisions (P = .74). There were statistically
Table 2 Physical Activity and Social Cognitive Theory Mediator Variables in Pretrial and Posttrial for Intervention and Control Groups Control Variables
Intervention
Pretrial
Posttrial
Pretrial
Posttrial
GLTEQ
22.7 (19.4)
20.3(21.9)
19.1 (14.8)
27.4 (20.6)
EXSE
47.3 (15.3)
41.9(18.1)
48.7 (11.9)
48.9 (14.1)
EGS
22.5 (9.7)
20.6 (8.9)
19.6 (8.7)
25.4 (9.2)
MOEES-Physical
33.0 (3.7)
33.6 (4.1)
35.9 (3.2)
35.1 (4.3)
MOEES-Social
23.3 (3.6)
24.2 (5.1)
24.0 (4.4)
24.2 (4.4)
MOEES-Self-evaluative
16.0 (2.1)
16.2 (2.3)
17.3 (2.4)
17.3 (2.4)
LL-FDI
33.0 (13.4)
33.2(13.5)
32.0 (11.3)
32.7 (11.2)
SSES-Family
30.7 (16.9)
24.2(16.5)
23.3 (14.2)
22.4 (13.0)
SSES-Friends
27.7 (20.3)
23.9(19.2)
22.3 (17.1)
20.7 (14.0)
SPS
15.0 (2.7)
15.5 (1.8)
15.2 (2.3)
15.6 (1.6)
Note. Values represent mean (standard deviation); n = 34 per condition for all variables. Abbreviations: GLTEQ, Godin Leisure-Time Exercise Questionnaire; EXSE, Exercise Self-efficacy Scale; EGS, Exercise Goal Setting Scale; MOEES, Multidimensional Outcome Expectations for Exercise Scale; LL-FDI, LateLife Function and Disability Inventory; SSES, Social support and Exercise Survey; SPS, Social Provisions Scale. JPAH Vol. 12, No. 7, 2015
Physical Activity Intervention in MS 929
Downloaded by Purdue Univ on 09/17/16, Volume 12, Article Number 7
Mediator Analysis The mediator analysis involved 3 preconditions: 1) condition was related with physical activity, 2) condition was related with the presumed mediator, and 3) the presumed mediator was associated with physical activity. The aforementioned mixed factor ANOVAs supported preconditions 1 and 2 for considering goal setting as a mediator of change in GLTEQ scores. Bivariate correlation analysis indicated that there was a statistically significant relationship between changes in goal setting and GLTEQ scores (r = .40, P ≤ .01) supporting precondition 3. The initial analysis regressed change in GLTEQ on condition (0 = control, 1 = intervention) and the model was significant, F(1, 66) = 5.47, P = .02, adjusted R2 = .06. Condition was a significant predictor of change in GLTEQ (B = 10.85, SE B = 4.63, β = .28, P = .02). The next analysis regressed change in GLTEQ scores on condition and change in goal setting. The model was significant, F(2, 65) = 6.84, P = .002, adjusted R2 = .15. Only change in goal setting was a significant predictor of change in GLTEQ (B = 0.80, SE B = 0.29, β = 0.35, P = .007). The effect of condition on GLTEQ was no longer statistically significant (P = .35).
Discussion The primary objective of this pilot RCT was to examine the efficacy of a SCT-based behavioral intervention delivered by newsletters and telephone calls for increasing physical activity in persons with MS who were physically inactive and had middle levels of self-efficacy. The results indicated that the intervention group, who received SCTbased intervention materials, appeared to become more physically active based on self-reported physical activity over the 6 weeks compared with the control group. Only goal setting was changed by the intervention, and it was identified as a significant mediator of changing in self-reported physical activity. To that end, the major accomplishment of this pilot, RCT was the provision of evidence for the efficacy of theory-based, behavioral intervention delivered by newsletters and phone calls for promoting self-reported physical activity by way of goal-setting among inactive persons with MS who had middle levels of self-efficacy. The efficacy of the intervention on social cognitive variables (eg, self-efficacy and goal setting) was further examined in the current study. Only goal setting significantly changed over the 6 weeks in the intervention group and it was identified as a significant mediator of change in self-reported physical activity. Such results are consistent with a previous Internet-delivered behavioral intervention in persons with MS whereby only goal setting was significantly changed by the intervention and served as a mediator of the intervention.14 This might be explained by goal setting theory28 such that goal setting is associated with performance outcomes (ie, physical activity), and this might be especially true in those who have middle levels of self-efficacy (ie, those who are ready to commit to behavior change but tend to give up in the face of barriers because of moderate efficacy). Such persons can achieve desired changes by learning self-regulatory strategies (ie, goal setting) through interactive guidance (ie, print materials or telephone counseling).5 Another possible explanation might be the emphasis that was placed on goal setting, instead of other social cognitive variables, throughout the intervention period. The phone calls for the intervention group played a role in reinforcing the intervention materials, plus tracking weekly step counts from pedometer and discussing setting/modifying goals based on the step counts for increasing physical activity. Overall, goal setting may represent an important target variable that should be considered in developing
behavioral interventions for increasing physical activity in persons with MS who are physically inactive. One observation of this study is the lack of significant changes in exercise self-efficacy over time in the intervention group. This indicates that levels of exercise self-efficacy were maintained throughout the study period for the behavioral intervention. This is inconsistent with the findings of previous research whereby selfefficacy decreased over time in the intervention group, and this decline is explained by overestimated at baseline and recalibration over time in both persons with MS and healthy adults.14,29 Perhaps our focus on persons with moderate or middle self-efficacy resulted in a rating that was more realistic and therefore maintained over the 6 weeks. Importantly, we did not power the study for detecting a change in self-efficacy, and this represents a plausible alternative explanation for not detecting a differential change in self-efficacy between intervention and control groups. There are novel features and notable findings of the current study. One of the novel features was that the control group received nonphysical activity information through newsletters and phone calls over the intervention period. Indeed, the newsletters contained MS-related information (ie, managing stress, nutrition) from the National MS Society website and this condition served as control for attention and social contact. This highlights that the current study had a credible control condition as opposed to the condition receiving standardized, generic information or being a waitlist.10,14 Another notable feature of the current study is the focus on middle levels of exercise self-efficacy in persons with MS based on Bandura’s 3-fold stepwise implementation model, and this provides the initial evidence that newsletters and phone calls can benefit physical activity promotion, especially for those with middle levels of selfefficacy of physical activity. One additional notable finding of this study is that average step counts from pedometer per day increased by 26.2% over the intervention period. This was consistent with a systematic review of pedometer-based physical activity intervention in the general population indicating that pedometer users who had step goals significantly increased physical activity by 26.9%.30 There are some limitations of the current study. We did not include time-spent in light, moderate, and vigorous physical activity based on the accelerometer in the study because of data processing problems that altered the data files from the accelerometers in a format that did not permit further analyses beyond total activity counts. Another limitation is that the sample had minimal disability and the results should not be generalized among those with more severe mobility impairments. An additional limitation is the representativeness of the sample based on the clinical and demographic characteristics of the participants. The sample of the current study primarily consisted of women with RRMS, and although this is the primary demographic and clinical course of MS, we cannot be certain that our sample is representative of the larger population of those with MS. This sample was recruited from a larger pool of participants enrolled in a previous physical activity study,11 and this might represent a significant source of bias among those with a strong desire for participation in research. This may have changed expectations that may have primed the current intervention study. Importantly, the demographic and clinical characteristics of the current sample were not different from the parent cohort study.11
Conclusions The current pilot RCT indicated that a newsletter and telephone call delivered behavioral intervention based on SCT resulted in a statistically significant and large increase in self-reported physical
JPAH Vol. 12, No. 7, 2015
930 Suh et al
activity over 6 weeks in persons with MS who were inactive and had middle levels of self-efficacy. The effect of the intervention was mediated by only change in goal setting, and this is consistent with other research in persons with MS.14 Overall, nonsupervised technology (ie, newsletters and telephone calls) could have the potential for promotion of physical activity among persons with MS. This approach further has low cost, ease of delivery and time management, and large sample reach. We await future trials of behavioral interventions developed with theory and delivered through non-face-to-face approaches (ie, newsletters and telephone calls) for increasing physical activity and improving outcomes in persons with MS.
Downloaded by Purdue Univ on 09/17/16, Volume 12, Article Number 7
References 1. Latimer-Chung AE, Pilutti LA, Hicks AL, et al. The effects of exercise training on fitness, mobility, fatigue, and quality of life among adults with multiple sclerosis: a systematic review to inform guideline development. Arch Phys Med Rehabil. 2013;94(9):1800–1823. doi:10.1016/j.apmr.2013.04.020 2. Motl RW, McAuley E, Snook EM. Physical activity and multiple sclerosis: a meta analysis. Mult Scler. 2005;11:459–463. PubMed doi:10.1191/1352458505ms1188oa 3. 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(2):85–90. PubMed doi:10.1097/NPT.0b013e31829157c0 4. Glanz K, Bishop DB. The role of behavioral science theory in development and implementation of public health interventions. Annu Rev Public Health. 2010;31:399–418. PubMed doi:10.1146/annurev. publhealth.012809.103604 5. Bandura A. Health promotion by social cognitive means. Health Educ Behav. 2004;31:143–164. PubMed doi:10.1177/1090198104263660 6. Baranowski T, Anderson C, Carmack C. Mediating variable framework in physical activity intervention. Am J Prev Med. 1998;15(4):266–297. PubMed doi:10.1016/S0749-3797(98)00080-4 7. 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 sclerosis. J Phys Act Health. 2011;8:626–635. PubMed 8. 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. PubMed doi:10.1080/08964289.2011.6 03768 9. Plow M, Bethoux F, McDaniel C, McGlynn M, Marcus B. Randomized controlled pilot study of customized pamphlets to promote physical activity and symptom self-management in women with multiple sclerosis. Clin Rehabil. 2014;28(2):139–48. PubMed 10. Bombardier CH, Cunniffe M, Wadhwani R, Gibbons LE, Blake KD, Kraft GH. The efficacy of telephone counseling for health promotion in people with multiple sclerosis: a randomized controlled trial. Arch Phys Med Rehabil. 2008;89(10):1849–1856. PubMed doi:10.1016/j. apmr.2008.03.021 11. Motl RW, McAuley E, Sandroff BM. Longitudinal change in physical activity and its correlates in relapsing-remitting multiple sclerosis. Phys Ther. 2013;93(8):1037–1048. PubMed doi:10.2522/ ptj.20120479 12. Hadjimichael O, Kerns RB, Rizzo MA, Cutter G, Vollmer T. Persistent pain and uncomfortable sensations in persons with multiple sclerosis. Pain. 2007;127:35–41. PubMed doi:10.1016/j.pain.2006.07.015
13. Thomas S, Reading J, Shephard RJ. 1992). Revision of the physical-activity readiness questionnaire (PAR-Q). Can J Sport Sci. 1992;17:338–345. PubMed 14. Motl RW, Dlugonski D, Wójcicki TR, McAuley E, Mohr DC. Internet intervention for increasing physical activity in persons with multiple sclerosis. Mult Scler. 2011;17(1):116–128. PubMed doi:10.1177/1352458510383148 15. Marcus BH, Napolitano MA, King AC, et al. Telephone versus print delivery of an individualized motivationally tailored physical activity intervention: project STRIDE. Health Psychol. 2007;26(4):401–409. PubMed doi:10.1037/0278-6133.26.4.401 16. Godin G, Shephard RJ. A simple method to assess exercise behavior in the community. Can J Appl Sport Sci. 1985;10:141–146. PubMed 17. Gosney JL, Scott JA, Snook EM, Motl RW. Physical activity and multiple sclerosis: validity of self-report and objective measures. Fam Community Health. 2007;3:144–150. PubMed doi:10.1097/01. FCH.0000264411.20766.0c 18. Motl RW, Snook EM, McAuley E, Scott JA, Douglass ML. Correlates of physical activity among individuals with multiple sclerosis. Ann Behav Med. 2006;32(2):154–161. PubMed doi:10.1207/ s15324796abm3202_13 19. McAuley E. Self-efficacy and the maintenance of exercise participation in older adults. J Behav Med. 1993;16:103–113. PubMed doi:10.1007/ BF00844757 20. McAuley E, Motl RW, White SM, Wójcicki TR. Validation of the multidimensional outcome expectations for exercise scale in individuals with multiple sclerosis. Arch Phys Med Rehabil. 2009;91:100–105. PubMed doi:10.1016/j.apmr.2009.09.011 21. Wojcicki TR, White SM, McAuley E. Assessing outcome expectations in older adults: the multidimensional outcome expectations for exercise scale. J Gerontol B Psychol Sci Soc Sci. 2009;64(1):33–40. PubMed doi:10.1093/geronb/gbn032 22. McAuley E, Konopack JF, Motl RW, Rosengren K, Morris KS. Measuring disability and function in older women: psychometric properties of the late life function and disability instrument. J Gerontol A Biol Sci Med Sci. 2005;60:901–909. PubMed doi:10.1093/gerona/60.7.901 23. Rovniak LS, Anderson ES, Winett RA, Stephens RS. Social cognitive determinants of physical activity in young adults: a prospective structural equation analysis. Ann. Behav. 2002;24:149–156. PubMed doi:10.1207/S15324796ABM2402_12 24. Sallis JF, Grossman RM, Pinski RB, Patterson TL, Nader PR. The development of scales to measure social support for diet and exercise behaviors. Prev Med. 1987;16(6):825–836. PubMed doi:10.1016/0091-7435(87)90022-3 25. Cutrona CE, Russell D. In: Jones WH, Perlman D, eds., Advances in personal relationships The provisions of social relationships and adaptation to stress. Greenwich, Conn.: JAI Press; 1987:37–67. 26. Learmonth YC, Motl RW, Sandroff BM, Pula JH, Cadavid D. Validation of patient determined disease steps (PDDS) scale scores in persons with multiple sclerosis. BMC Neurol. 2013;13(1):37. PubMed doi:10.1186/1471-2377-13-37 27. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale, NJ: Lawrence Earlbaum Associates; 1988. 28. Locke EA, Latham GP. Building a practically useful theory of goal setting and task motivation: a 35-year odyssey. Am Psychol. 2002;57(9):705–717. PubMed doi:10.1037/0003-066X.57.9.705 29. Klamm EL, Wo’jcicki TR, White SM, Szabo AN, Kramer AF, McAuley E. Differential effects of physical activity intervention on self-efficacy in older adults. Ann Behav Med. 2009;36:S82. 30. Bravata DM, Smith-Spangler C, Sundaram V, et al. Using pedometers to increase physical activity and improve health: a systematic review. J Ame Medi Associ. 2007;298(19):2296–2304. PubMed doi:10.1001/ jama.298.19.2296
JPAH Vol. 12, No. 7, 2015
ORIGINAL RESEARCH
Pilot Trial of a Social Cognitive Theory-Based Physical Activity Intervention Delivered by Nonsupervised Technology in Persons With Multiple Sclerosis Yoojin Suh, Robert W. Motl, Connor Olsen, and Ina Joshi Background: Physical inactivity is prevalent in people with multiple sclerosis (MS) and this highlights the importance of developing behavioral interventions for increasing physical activity (PA) in MS. This pilot trial examined the efficacy of a 6-week, behavioral intervention based on social cognitive theory (SCT) delivered by newsletters and phone calls for increasing PA in persons with MS who were physically inactive and had middle levels of self-efficacy. Methods: The sample included 68 persons with relapsing-remitting MS who were randomly assigned into intervention and control groups. The intervention group received SCT-based information by newsletters and phone calls, whereas the controls received information regarding topics such as stress management over 6 weeks. Participants completed self-report of PA and social cognitive variables. Results: The intervention group had a significant increase in self-reported PA (d = 0.56, P = .02) over the 6 weeks, but the controls had a nonsignificant change (d = –0.13, P = .45). Goal setting was changed in the intervention group (d = 0.68, P ≤ .01) and identified as a significant mediator of change in self-reported PA. Conclusions: This study provides initial evidence for the benefit of a theory-based behavioral intervention for increasing PA in MS. Keywords: exercise, rehabilitation, neurological disease, self-management
Physical activity has been associated with improvements in fitness outcomes, symptoms, and quality of life among persons with multiple sclerosis (MS),1 yet physical inactivity is particularly common in this neurological disease.2 Those observations underscore the importance of designing behavioral interventions for increasing physical activity among persons with MS.3 Behavioral interventions involve teaching persons the skills, techniques, and strategies that support successful health behavior change,4 and have been recognized as approaches for overcoming physical inactivity in those with chronic disease conditions.3 To be effective, behavioral interventions should be based on theoretical models that guide the identification and selection of variables as targets for changing the outcome of interest.4 Social cognitive theory (SCT)5 has been a prominent theoretical model for effectively explaining and predicting physical activity in the general population6 and those with Parkinson’s disease and MS.3 Recently, Bandura has operationalized a SCT-based model of associations among self-efficacy (ie, beliefs about confidence in undertaking a specific health behavior), multidimensional outcome expectations (ie, beliefs about the physical, social, and self-evaluative benefits of a health behavior), sociostructural factors (ie, facilitators and impediments of undertaking a health behavior), goal-setting (ie, self-management strategy for undertaking a health behavior), and health behavior and its change. This model indicates that self-efficacy has direct and indirect effects on physical activity as a health behavior, and the indirect effects operate via multidimensional outcome expectations, impediments and facilitators, and goal-setting.5 There has been research supporting Bandura’s model5 and its variables as correlates of physical activity in cross-sectional7 and longitudinal8 examinations of persons with MS. The authors are with the Dept of Kinesiology and Community Health, University of Illinois at Urbana-Champaign. Motl ([email protected]) is corresponding author. 924
Beyond theory, behavioral interventions can be delivered through non-face-to-face methods. We are aware of research on delivering behavioral interventions for improving physical activity in MS via print materials and telephone calls. For example, one recent paper demonstrated that print media delivered materials based on the transtheoretical model (TTM) resulted in a possible improvement of physical activity levels and function in women with MS.9 Another study delivered motivational interviewing through telephone calls as a method of improving health promoting activities, including physical activity, among persons with MS.10 Such interventions have been associated with increases in physical activity, but we believe that there are several ways for extending this line of research on behavioral interventions in MS. Bandura’s 3-fold stepwise implementation model should be considered when developing and delivering behavior interventions for increasing physical activity.5 This implementation model explains that the type of intervention materials can be tailored to 3 levels of a person’s self-efficacy and motivational readiness to achieve behavior change.5 For example, the implementation model suggests that tailored-print materials and telephone interactions should be ideally suited for those who have middle levels of selfefficacy (ie, persons meeting the second level of the 3-fold model). Those within the first level of the 3-fold model require very little assistance with behavioral change, whereas those within the third level of the 3-fold model require intensive intervention for behavior change. Another consideration involves examining mediators of behavioral interventions that can inform the design and tailoring of future research by streamlining the content. One last consideration involves inclusion of a control condition that receives print materials and social contact on the telephone, but does not receive information on physical activity change. This pilot, randomized controlled trial (RCT) examined the efficacy of a SCT-based intervention delivered by newsletters and telephone calls for increasing physical activity in persons with MS who had middle levels of self-efficacy, but were not regularly
Physical Activity Intervention in MS 925
physically active. This study further examined mediators of intervention effects on physical activity change, particularly self-efficacy, outcome expectations, functional limitations, social support, and goal-setting.
Methods
Downloaded by Purdue Univ on 09/17/16, Volume 12, Article Number 7
Recruitment We distributed a flyer through e-mail among those who participated in a recently- completed longitudinal study of physical activity in persons with relapsing-remitting MS (RRMS).11 Briefly, the sample in that previous study consisted of 223 women and 46 men. The mean age was 45.9 years (SD = 9.6), and the duration of MS (time since diagnosis) was 8.8 years (SD = 7.0). The median Patient Determined Disease Scale (PDDS) score was 2 (range = 0–6) and corresponded with moderate disability (ie, no limitations in walking, but significant problems that limit daily activities).12 Persons who were interested in the current study contacted our laboratory for information and screening for inclusion. The inclusion criteria were definite diagnosis of RRMS; independently ambulatory or ambulatory with single point assistance (eg, cane); relapse free in the past 30 days; being nonactive defined as not engaging in regular physical activity (ie, 30 minute-accumulated per day) on more than 2 days of the week during the previous 6 months; free of contraindications for physical activity (eg, no underlying cardiovascular disease) based on Physical Activity Readiness Questionnaire;13 having the visual ability necessary to read 14 point font; meeting the age requirement (ie, 18 to 64 years of age). We focused on RRMS because it is the most common type of MS and persons are seemingly more able to engage in physical activity via behavioral intervention than other types of MS. One final screening criterion was to have middle levels
of exercise self-efficacy. The criterion for middle level of efficacy was determined based on the distribution of Exercise Self-Efficacy Scale (EXSE) scores from our previous longitudinal study.11 Based on a frequency analysis, the middle category of overall scores on the EXSE ranged between 50 and 70, and this was operationalized as the middle level of exercise self-efficacy. Power Analysis. The proposed sample size for this RCT was 60 persons with RRMS. The sample size was based on published self-reported physical activity mean scores (ie, GLTEQ) from a behavioral intervention for detecting a moderate effect size for the interaction term (d = 0.50)14 with assumptions of α = .05, β = .10 for 90% power, and reliability of .90. The power analysis indicated a sample of 50 participants, and we opted for a 20% increase in sample size considering the attrition rate from other nonsupervised technology interventions.15 The final sample for adequate power considering drop-out and replacement of missing follow-up data were 60. We did not perform a power analysis for self-efficacy or other SCT variables. The flow of participants is presented using CONSORT diagram in Figure 1.
Measures Physical Activity. Physical activity was measured by the Godin Leisure-Time Exercise Questionnaire (GLTEQ).16 Researchers have provided evidence for the validity of this measure in persons with MS.17,18 The GLTEQ is a self-administered measure of usual physical activity. The GLTEQ has 3 items that measure the frequency of strenuous (eg, jogging), moderate (eg, fast walking), and mild (eg, easy walking) exercise for periods of more than 15 minutes during one’s free time in a typical week. The weekly frequencies of strenuous, moderate, and mild activities are multiplied by 9, 5, and 3 metabolic equivalents, respectively, and summed to form a
Figure 1 — Consort diagram. JPAH Vol. 12, No. 7, 2015
926 Suh et al
measure of total leisure activity. The scores range between 0 and 119 in arbitrary units.
Downloaded by Purdue Univ on 09/17/16, Volume 12, Article Number 7
Self-Efficacy. Self-efficacy was assessed by the EXSE.19 This scale has 6 items that assess an individual’s belief in his or her ability to engage in 20+ minutes of moderate physical activity 3 times per week, in 1-month increments, across the next 6 months. The items are rated on a scale from 0 (Not at all confident) to 100 (Completely confident) and averaged into a composite score that ranges between 0 and 100. Higher scores reflect greater confidence in one’s ability to regularly engage in exercise. Outcome Expectations. Outcome expectations for exercise were measured by the Multidimensional Outcomes Expectations for Exercise Scale (MOEES).20,21 This scale contains 15 items that reflect 3 subdomains of outcome expectations. Six items reflect physical outcome expectations, 4 items assess social outcome expectations, and 5 items measure self-evaluative outcome expectations. The 15 items were rated on a 5-point scale from 1 (Strongly disagree) to 5 (Strongly agree) and summed to form the subscale measures of outcome expectations. Functional Limitations. Functional limitations represented
impediments for physical activity and was measured using the Functional Limitations component of the abbreviated Late-Life Function and Disability Instrument (LL-FDI).22 The Functional Limitations component contains 15 items that correspond with advanced lower extremity function, basic lower extremity function, and upper extremity function. The 15-items are rated on a 5-point scale ranging from 1 (none) to 5 (cannot do). Goal-Setting. Exercise goals were measured by the Exercise
Goal-setting Scale (EGS).23 This contains 10 items that reflect goalsetting for exercise behavior and the items are rated on a 5-point scale from 1 (Does not describe) to 5 (Describes completely). The item scores are summed into an overall score that ranges between 10 and 50. Higher scores reflect a stronger tendency for setting goals for exercise.
Social Support. Social support for physical activity was measured
by the 12-item Social Support and Exercise Survey (SSES).24 The items on the SESS are rated on a 5-point scale ranging from 1 (None) to 5 (Very often), and family and friends scales are scored by summing total of items 1 to 12. In addition, perceptions of social support were measured by the 6-item modified Social Provisions Scale (SPS).25 The items on the SPS are rated on a 4-point scale ranging from 1 (strongly disagree) to 4 (strongly agree), and item scores are summed to form an overall score that ranges between 6 and 24. Disability. Disability was measured using the Patient Determined
Disease Steps (PDDS) scale.12 The PDDS is a self-report questionnaire that contains a single item for measuring disability using an ordinal scale from 0 (Normal) through 8 (Bedridden). This scale was developed as an inexpensive surrogate for the Expanded Disability Status Scale (EDSS) and scores from the PDDS are linearly and strongly related with physician-administered EDSS scores.26
Intervention Materials Printed Newsletters. Each newsletter targeted a specific agent of SCT for increasing physical activity in persons with MS. The purpose of the newsletters was to offer information about key determinants of physical activity that motivate physical activity change for persons with MS. Newsletters were delivered through United States Postal Service (USPS) and e-mail. The first newsletter focused on
the physical and psychological benefits of physical activity as well as self-monitoring among persons with MS. The second newsletter was then given for description of goal-setting for physical activity. The third newsletter contained how to improve exercise self-efficacy as well as description of 4 sources in enhancing exercise self-efficacy (ie, mastery experiences, social modeling, social persuasion, and interpretations of physiological and affective responses). The fourth newsletter included information that developed appropriate outcome expectations of being physically active. The fifth and sixth newsletters targeted barriers/impediments to physical activity and social support/facilitators for physical activity by providing guidance on overcoming barriers and identifying facilitators of physical activity. Telephone Calls. Telephone calls were made to all participants
in this study for a 15-minute discussion after the delivery of each newsletter. The same research assistants delivered the intervention materials through the telephone calls during this study (ie, research assistants were yoked with participants throughout the study). The main purpose of telephone calls was to reinforce information on the newsletter and help participants understand the content, and apply it to their daily routines. This occurred through a semiscripted phone interview. In addition, the phone calls included coaching on how to set-up a goal with a pedometer as well as monitoring average step counts per participant for meeting the goal.
Pedometer. Pedometers (Yamax SW-200, Yamasa Tokei Keiki
Co., Ltd, Tokyo, Japan) and log-books were delivered to the intervention group for the purpose of self-monitoring and tracking physical activity. The pedometer and log-book further assisted in setting-up goals as a self-regulatory strategy for increasing physical activity.
Control Materials Participants in the control group received a total of 6 newsletters designed to provide information that was not related to physical activity (eg, managing stress, nutrition, allergies, blood pressure, alcohol use, and cholesterol); these were selected based on resources provided by on the website of the National MS Society. Telephone calls were further followed up to check that the participants received the newsletters as well as that the content of the newsletters were easy to follow. The topic of physical activity was not discussed during phone calls. This condition served as a control for the attention and social contact that occurred in the intervention arm of the pilot study.
Procedures The researchers contacted potential participants who expressed interest in the study and conducted the screening procedure to evaluate inclusion criteria of this study. After initial telephone contact and screening, all participants who volunteered provided a signed informed consent document. Upon receipt of informed consent, participants were sent a battery of questionnaires along with an accelerometer through USPS. The participants were further provided prestamped and preaddressed envelopes for return postal service. Participants were asked to wear the accelerometer for a 7-day period along with completing the questionnaires as baseline assessment. Upon return, all questionnaires were checked for completeness, and if any has missing data, researchers made a phone call to collect the data. Importantly, the data were collected by a person who did not deliver the intervention and control conditions (ie, blinding), and the interventionists were blinded regarding the focal outcomes of
JPAH Vol. 12, No. 7, 2015
Downloaded by Purdue Univ on 09/17/16, Volume 12, Article Number 7
Physical Activity Intervention in MS 927
the pilot trial. Participants were matched in pairs based GLTEQ and PDDS scores after completion of the baseline assessment. Participants within the matched pairs were then randomly assigned by a researcher who was uninvolved in data collection and intervention delivery to either the intervention or control conditions using a random number generator. The intervention materials were delivered in a single wave across a 6-week period and the participants in the intervention received a total of 6 newsletters developed based on active agents (eg, self-efficacy) from the SCT. The control group received 6 newsletters that included information unrelated with physical activity change. Print newsletters were mailed by USPS on Thursday each week to ensure that the newsletters arrived at the beginning of the next week. Newsletters were further emailed to participants in case of missing mail or delivery failure. After the newsletter was delivered, telephone calls were made on Monday and Tuesday for the intervention group, and the control group received calls on Wednesday and Thursday. We proceeded with follow-up assessments using the same procedures for baseline testing after delivery of the intervention and control conditions. Participants received $10 for completing the measures on each of the measurement periods.
Data Analyses All analyses were conducted in PASW Statistics, version 18.0 for Windows. We included all persons who provided baseline data and were randomized into a condition into the data analysis (ie, intent-to-treat whereby once randomized always analyzed). Any missing follow-up data were replaced with the corresponding baseline value, thereby allowing for the intent-to-treat analysis. We compared the intervention and control groups for initial differences on demographics and MS-related variables (eg, type of MS) using independent samples t tests and chi-square analyses. We further compared initial differences in physical activity and SCT variables using independent samples t tests. The effects of the intervention on physical activity and SCT variables was examined using 2 (Condition: Intervention & Control) × 2 (Time: Pre & Post) mixed factor analysis of variance (ANOVA) based on univariate F-statistic. Condition was a between-subjects factor and time was
a within-subjects factor. Interactions were decomposed using paired and independent samples t-tests. Effects sizes for F-statistic were expressed as partial eta-squared (η2p). Effects sizes based on differences in mean scores were expressed as Cohen’s d.27 Multiple linear regression analyses were conducted to examine the SCT variables as mediators. This analysis involved regressing change in physical activity on condition in the first analysis and then regressing change in physical activity on condition plus change in putative mediators in the second analysis. Mediation was based on the effects of condition on physical activity becoming reduced in magnitude and nonsignificant after controlling for the mediator variable(s). The proportion of explained variance in the outcome from the regression analysis was based on the adjusted R2.
Results Sample Characteristics 97 participants were recruited and assessed for eligibility, and 72 participants satisfied the study inclusion criteria. Of note, the mean (SD) of the EXSE scores based on the telephone screening for inclusion was 56.6 (3.9); this satisfied our definition of middlelevel of self-efficacy (ie, EXSE score between 50 and 70). Of the 72 participants who satisfied inclusion criteria, 68 participants provided informed consent and completed the questionnaires and returned the study packet at baseline; this surpassed our sample from the power analysis. The resulting 68 persons were randomly assigned into either intervention or control groups; the 4 persons who did not provide baseline data were not randomized into condition. Two people did not provide follow-up data and we replaced the missing follow-up data with baseline scores (ie, intent-to-treat analysis whereby all persons who provided baseline data and were randomized into condition were included in data analyses). Independent samples t tests and chi square statistics (χ2) indicated no initial differences in sex (P = .20), age (P = .35), weight (P = .61), height (P = .09), education (P = 1.00), race (P = .08), employment (P = .63), MS type (P = 1.00), disease severity (P = .64), and disease duration (P = .60) (see Table 1). The mean age of the total sample
Table 1 Demographic and Clinical Data at Baseline for the Intervention and Control Groups Total (n = 68)
Characteristic Sex (male/female)
Control (n = 34)
Intervention (n = 34)
12/56
8/26
4/30
49.0 (8.8)
48.0 (9.4)
50.1 (8.1)
Education (college graduate/less than college graduate)
48/20
24/10
24/10
Race (Caucasian/non-Caucasian)
65/3
31/3
34/0
Employment (unemployed/employed)
36/32
17/17
19/15
MS type (RRMS/Other)
66/2
33/1
33/1
Weight (kg)
78.9 (22)
79.8 (20.3)
77.5 (23.7)
Height (cm)
167.1 (7.1)
167.4 (5.1)
165.2 (6.0)
Time since diagnosis (years)
12.1 (7.9)
12.7 (8.8)
11.6 (7.1)
Disease severity (arbitrary units)
2.0 (1.8)
2.2 (1.8)
2.0 (1.8)
Age (years)
Note. Values represent mean (standard deviation) or number of cases for dichotomous variables; Disease severity = Patient Determined Disease Steps Scale score (PDDS). JPAH Vol. 12, No. 7, 2015
928 Suh et al
at baseline was 49 years (SD = 8.8; range 18–64). The majority of the sample was female (81%), Caucasian (96%), well-educated (47% > were college graduates), and were employed (47.1%). The sample consisted primarily of individuals with RRMS (n = 66), with the remaining individuals having SPMS (n = 1) or benign MS (n = 1). The mean time since MS diagnosis was 12.1 years (SD = 7.9, range 3–33 years) and the mean PDDS score was 2.0 (SD = 1.8, range 0–7). There were no major or obvious differences in demographic and clinical characteristics between this sample and the original cohort from which it was drawn.11 The EXSE score in this sample was considerably lower than the parent cohort (mean = 47.3 vs. mean = 77.7, respectively) by design considering that we focused on those with middle levels of efficacy when designing this intervention.
significant differences between the 2 groups on physical outcome expectations (P = .00), and self-evaluative outcome expectations (P = .02) indicating that the control group reported higher values compared with the intervention group.
Effects of Intervention on Physical Activity and SCT variables There was a statistically significant condition-by-time interaction on GLTEQ scores, F(1, 66) = 5.47, P = .02, η2p = .08. The intervention group reported a statistically significant (P = .02) and medium increase in physical activity over time (d = 0.56), whereas the control group had a small (d = –0.13) and nonsignificant (P = .45) reduction in physical activity over time. One additional analysis examined change in step counts per day from pedometers over the intervention period. The 1-way repeated-measures ANOVA indicated a statistically significant time main effect on steps per day, F(1, 32) = 8.03, P ≤ .01, η2p = .20. Overall, steps per day increased by 1368 steps per day from the first to last week of the 6-week intervention period (d = .46). The same analytic model (mixed factor ANOVA) was conducted to determine changes in social cognitive variables. There was a statistically significant condition-by-time interaction on goal setting, F(1, 66) = 16.88, P ≤ .01, η2p = .20. The intervention group reported a statistically significant and large increase in goal setting over time (d = 0.68, P ≤ .01), whereas the control group had a nonsignificant, small reduction in goal setting over time (d = –0.19, P = .14). The ANOVA did not indicate statistically significant condition-by-time interactions on exercise self-efficacy [F(1, 66) = 3.19, P = .08, η2p = .05], physical outcome expectations [F(1, 66) = 3.01, P = .09, η2p = .04], social outcome expectations [F(1, 66) = .83, P = .36, η2p = .01], self-evaluative outcome expectations [F(1, 66) = .59, P = .44, η2p = .00], functional limitations [F(1, 66) = .11, P = .73, η2p = .00], family support for exercise [F(1, 66) = 3.40, P = .07, η2p = .04], friends support for exercise [F(1, 66) = .24, P = .62, η2p = .00], and social provisions scale [F(1, 66) = .01, P = .91, η2p = .00].
Downloaded by Purdue Univ on 09/17/16, Volume 12, Article Number 7
Intervention Fidelity All 6 newsletters, regardless of intervention or control condition, were mailed on time as scheduled, and no newsletters were returned for an incorrect postal address. Telephone calls further were made as scheduled, but 9.6% and 14.6% of all phone calls were missed for intervention and control groups, respectively. Of the calls that were successfully, 87.8% were completed on the first attempt and 12.1% were completed on the second attempt. There were no third attempts for phone calls.
Group Differences in Physical Activity and SCT variables Independent samples t-tests compared possible group differences in physical activity and SCT variables at baseline (see Table 2). There were no significant differences between the intervention and control groups on pretrial values for GLTEQ scores (P = .38), activity counts (P = .21), self-efficacy (P = .68), goal setting (P = .20), social outcome expectations (P = .48), functional limitations (P = .76), family support for exercise (P = .06), friends support for exercise (P = .24), and social provisions (P = .74). There were statistically
Table 2 Physical Activity and Social Cognitive Theory Mediator Variables in Pretrial and Posttrial for Intervention and Control Groups Control Variables
Intervention
Pretrial
Posttrial
Pretrial
Posttrial
GLTEQ
22.7 (19.4)
20.3(21.9)
19.1 (14.8)
27.4 (20.6)
EXSE
47.3 (15.3)
41.9(18.1)
48.7 (11.9)
48.9 (14.1)
EGS
22.5 (9.7)
20.6 (8.9)
19.6 (8.7)
25.4 (9.2)
MOEES-Physical
33.0 (3.7)
33.6 (4.1)
35.9 (3.2)
35.1 (4.3)
MOEES-Social
23.3 (3.6)
24.2 (5.1)
24.0 (4.4)
24.2 (4.4)
MOEES-Self-evaluative
16.0 (2.1)
16.2 (2.3)
17.3 (2.4)
17.3 (2.4)
LL-FDI
33.0 (13.4)
33.2(13.5)
32.0 (11.3)
32.7 (11.2)
SSES-Family
30.7 (16.9)
24.2(16.5)
23.3 (14.2)
22.4 (13.0)
SSES-Friends
27.7 (20.3)
23.9(19.2)
22.3 (17.1)
20.7 (14.0)
SPS
15.0 (2.7)
15.5 (1.8)
15.2 (2.3)
15.6 (1.6)
Note. Values represent mean (standard deviation); n = 34 per condition for all variables. Abbreviations: GLTEQ, Godin Leisure-Time Exercise Questionnaire; EXSE, Exercise Self-efficacy Scale; EGS, Exercise Goal Setting Scale; MOEES, Multidimensional Outcome Expectations for Exercise Scale; LL-FDI, LateLife Function and Disability Inventory; SSES, Social support and Exercise Survey; SPS, Social Provisions Scale. JPAH Vol. 12, No. 7, 2015
Physical Activity Intervention in MS 929
Downloaded by Purdue Univ on 09/17/16, Volume 12, Article Number 7
Mediator Analysis The mediator analysis involved 3 preconditions: 1) condition was related with physical activity, 2) condition was related with the presumed mediator, and 3) the presumed mediator was associated with physical activity. The aforementioned mixed factor ANOVAs supported preconditions 1 and 2 for considering goal setting as a mediator of change in GLTEQ scores. Bivariate correlation analysis indicated that there was a statistically significant relationship between changes in goal setting and GLTEQ scores (r = .40, P ≤ .01) supporting precondition 3. The initial analysis regressed change in GLTEQ on condition (0 = control, 1 = intervention) and the model was significant, F(1, 66) = 5.47, P = .02, adjusted R2 = .06. Condition was a significant predictor of change in GLTEQ (B = 10.85, SE B = 4.63, β = .28, P = .02). The next analysis regressed change in GLTEQ scores on condition and change in goal setting. The model was significant, F(2, 65) = 6.84, P = .002, adjusted R2 = .15. Only change in goal setting was a significant predictor of change in GLTEQ (B = 0.80, SE B = 0.29, β = 0.35, P = .007). The effect of condition on GLTEQ was no longer statistically significant (P = .35).
Discussion The primary objective of this pilot RCT was to examine the efficacy of a SCT-based behavioral intervention delivered by newsletters and telephone calls for increasing physical activity in persons with MS who were physically inactive and had middle levels of self-efficacy. The results indicated that the intervention group, who received SCTbased intervention materials, appeared to become more physically active based on self-reported physical activity over the 6 weeks compared with the control group. Only goal setting was changed by the intervention, and it was identified as a significant mediator of changing in self-reported physical activity. To that end, the major accomplishment of this pilot, RCT was the provision of evidence for the efficacy of theory-based, behavioral intervention delivered by newsletters and phone calls for promoting self-reported physical activity by way of goal-setting among inactive persons with MS who had middle levels of self-efficacy. The efficacy of the intervention on social cognitive variables (eg, self-efficacy and goal setting) was further examined in the current study. Only goal setting significantly changed over the 6 weeks in the intervention group and it was identified as a significant mediator of change in self-reported physical activity. Such results are consistent with a previous Internet-delivered behavioral intervention in persons with MS whereby only goal setting was significantly changed by the intervention and served as a mediator of the intervention.14 This might be explained by goal setting theory28 such that goal setting is associated with performance outcomes (ie, physical activity), and this might be especially true in those who have middle levels of self-efficacy (ie, those who are ready to commit to behavior change but tend to give up in the face of barriers because of moderate efficacy). Such persons can achieve desired changes by learning self-regulatory strategies (ie, goal setting) through interactive guidance (ie, print materials or telephone counseling).5 Another possible explanation might be the emphasis that was placed on goal setting, instead of other social cognitive variables, throughout the intervention period. The phone calls for the intervention group played a role in reinforcing the intervention materials, plus tracking weekly step counts from pedometer and discussing setting/modifying goals based on the step counts for increasing physical activity. Overall, goal setting may represent an important target variable that should be considered in developing
behavioral interventions for increasing physical activity in persons with MS who are physically inactive. One observation of this study is the lack of significant changes in exercise self-efficacy over time in the intervention group. This indicates that levels of exercise self-efficacy were maintained throughout the study period for the behavioral intervention. This is inconsistent with the findings of previous research whereby selfefficacy decreased over time in the intervention group, and this decline is explained by overestimated at baseline and recalibration over time in both persons with MS and healthy adults.14,29 Perhaps our focus on persons with moderate or middle self-efficacy resulted in a rating that was more realistic and therefore maintained over the 6 weeks. Importantly, we did not power the study for detecting a change in self-efficacy, and this represents a plausible alternative explanation for not detecting a differential change in self-efficacy between intervention and control groups. There are novel features and notable findings of the current study. One of the novel features was that the control group received nonphysical activity information through newsletters and phone calls over the intervention period. Indeed, the newsletters contained MS-related information (ie, managing stress, nutrition) from the National MS Society website and this condition served as control for attention and social contact. This highlights that the current study had a credible control condition as opposed to the condition receiving standardized, generic information or being a waitlist.10,14 Another notable feature of the current study is the focus on middle levels of exercise self-efficacy in persons with MS based on Bandura’s 3-fold stepwise implementation model, and this provides the initial evidence that newsletters and phone calls can benefit physical activity promotion, especially for those with middle levels of selfefficacy of physical activity. One additional notable finding of this study is that average step counts from pedometer per day increased by 26.2% over the intervention period. This was consistent with a systematic review of pedometer-based physical activity intervention in the general population indicating that pedometer users who had step goals significantly increased physical activity by 26.9%.30 There are some limitations of the current study. We did not include time-spent in light, moderate, and vigorous physical activity based on the accelerometer in the study because of data processing problems that altered the data files from the accelerometers in a format that did not permit further analyses beyond total activity counts. Another limitation is that the sample had minimal disability and the results should not be generalized among those with more severe mobility impairments. An additional limitation is the representativeness of the sample based on the clinical and demographic characteristics of the participants. The sample of the current study primarily consisted of women with RRMS, and although this is the primary demographic and clinical course of MS, we cannot be certain that our sample is representative of the larger population of those with MS. This sample was recruited from a larger pool of participants enrolled in a previous physical activity study,11 and this might represent a significant source of bias among those with a strong desire for participation in research. This may have changed expectations that may have primed the current intervention study. Importantly, the demographic and clinical characteristics of the current sample were not different from the parent cohort study.11
Conclusions The current pilot RCT indicated that a newsletter and telephone call delivered behavioral intervention based on SCT resulted in a statistically significant and large increase in self-reported physical
JPAH Vol. 12, No. 7, 2015
930 Suh et al
activity over 6 weeks in persons with MS who were inactive and had middle levels of self-efficacy. The effect of the intervention was mediated by only change in goal setting, and this is consistent with other research in persons with MS.14 Overall, nonsupervised technology (ie, newsletters and telephone calls) could have the potential for promotion of physical activity among persons with MS. This approach further has low cost, ease of delivery and time management, and large sample reach. We await future trials of behavioral interventions developed with theory and delivered through non-face-to-face approaches (ie, newsletters and telephone calls) for increasing physical activity and improving outcomes in persons with MS.
Downloaded by Purdue Univ on 09/17/16, Volume 12, Article Number 7
References 1. Latimer-Chung AE, Pilutti LA, Hicks AL, et al. The effects of exercise training on fitness, mobility, fatigue, and quality of life among adults with multiple sclerosis: a systematic review to inform guideline development. Arch Phys Med Rehabil. 2013;94(9):1800–1823. doi:10.1016/j.apmr.2013.04.020 2. Motl RW, McAuley E, Snook EM. Physical activity and multiple sclerosis: a meta analysis. Mult Scler. 2005;11:459–463. PubMed doi:10.1191/1352458505ms1188oa 3. 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(2):85–90. PubMed doi:10.1097/NPT.0b013e31829157c0 4. Glanz K, Bishop DB. The role of behavioral science theory in development and implementation of public health interventions. Annu Rev Public Health. 2010;31:399–418. PubMed doi:10.1146/annurev. publhealth.012809.103604 5. Bandura A. Health promotion by social cognitive means. Health Educ Behav. 2004;31:143–164. PubMed doi:10.1177/1090198104263660 6. Baranowski T, Anderson C, Carmack C. Mediating variable framework in physical activity intervention. Am J Prev Med. 1998;15(4):266–297. PubMed doi:10.1016/S0749-3797(98)00080-4 7. 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 sclerosis. J Phys Act Health. 2011;8:626–635. PubMed 8. 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. PubMed doi:10.1080/08964289.2011.6 03768 9. Plow M, Bethoux F, McDaniel C, McGlynn M, Marcus B. Randomized controlled pilot study of customized pamphlets to promote physical activity and symptom self-management in women with multiple sclerosis. Clin Rehabil. 2014;28(2):139–48. PubMed 10. Bombardier CH, Cunniffe M, Wadhwani R, Gibbons LE, Blake KD, Kraft GH. The efficacy of telephone counseling for health promotion in people with multiple sclerosis: a randomized controlled trial. Arch Phys Med Rehabil. 2008;89(10):1849–1856. PubMed doi:10.1016/j. apmr.2008.03.021 11. Motl RW, McAuley E, Sandroff BM. Longitudinal change in physical activity and its correlates in relapsing-remitting multiple sclerosis. Phys Ther. 2013;93(8):1037–1048. PubMed doi:10.2522/ ptj.20120479 12. Hadjimichael O, Kerns RB, Rizzo MA, Cutter G, Vollmer T. Persistent pain and uncomfortable sensations in persons with multiple sclerosis. Pain. 2007;127:35–41. PubMed doi:10.1016/j.pain.2006.07.015
13. Thomas S, Reading J, Shephard RJ. 1992). Revision of the physical-activity readiness questionnaire (PAR-Q). Can J Sport Sci. 1992;17:338–345. PubMed 14. Motl RW, Dlugonski D, Wójcicki TR, McAuley E, Mohr DC. Internet intervention for increasing physical activity in persons with multiple sclerosis. Mult Scler. 2011;17(1):116–128. PubMed doi:10.1177/1352458510383148 15. Marcus BH, Napolitano MA, King AC, et al. Telephone versus print delivery of an individualized motivationally tailored physical activity intervention: project STRIDE. Health Psychol. 2007;26(4):401–409. PubMed doi:10.1037/0278-6133.26.4.401 16. Godin G, Shephard RJ. A simple method to assess exercise behavior in the community. Can J Appl Sport Sci. 1985;10:141–146. PubMed 17. Gosney JL, Scott JA, Snook EM, Motl RW. Physical activity and multiple sclerosis: validity of self-report and objective measures. Fam Community Health. 2007;3:144–150. PubMed doi:10.1097/01. FCH.0000264411.20766.0c 18. Motl RW, Snook EM, McAuley E, Scott JA, Douglass ML. Correlates of physical activity among individuals with multiple sclerosis. Ann Behav Med. 2006;32(2):154–161. PubMed doi:10.1207/ s15324796abm3202_13 19. McAuley E. Self-efficacy and the maintenance of exercise participation in older adults. J Behav Med. 1993;16:103–113. PubMed doi:10.1007/ BF00844757 20. McAuley E, Motl RW, White SM, Wójcicki TR. Validation of the multidimensional outcome expectations for exercise scale in individuals with multiple sclerosis. Arch Phys Med Rehabil. 2009;91:100–105. PubMed doi:10.1016/j.apmr.2009.09.011 21. Wojcicki TR, White SM, McAuley E. Assessing outcome expectations in older adults: the multidimensional outcome expectations for exercise scale. J Gerontol B Psychol Sci Soc Sci. 2009;64(1):33–40. PubMed doi:10.1093/geronb/gbn032 22. McAuley E, Konopack JF, Motl RW, Rosengren K, Morris KS. Measuring disability and function in older women: psychometric properties of the late life function and disability instrument. J Gerontol A Biol Sci Med Sci. 2005;60:901–909. PubMed doi:10.1093/gerona/60.7.901 23. Rovniak LS, Anderson ES, Winett RA, Stephens RS. Social cognitive determinants of physical activity in young adults: a prospective structural equation analysis. Ann. Behav. 2002;24:149–156. PubMed doi:10.1207/S15324796ABM2402_12 24. Sallis JF, Grossman RM, Pinski RB, Patterson TL, Nader PR. The development of scales to measure social support for diet and exercise behaviors. Prev Med. 1987;16(6):825–836. PubMed doi:10.1016/0091-7435(87)90022-3 25. Cutrona CE, Russell D. In: Jones WH, Perlman D, eds., Advances in personal relationships The provisions of social relationships and adaptation to stress. Greenwich, Conn.: JAI Press; 1987:37–67. 26. Learmonth YC, Motl RW, Sandroff BM, Pula JH, Cadavid D. Validation of patient determined disease steps (PDDS) scale scores in persons with multiple sclerosis. BMC Neurol. 2013;13(1):37. PubMed doi:10.1186/1471-2377-13-37 27. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale, NJ: Lawrence Earlbaum Associates; 1988. 28. Locke EA, Latham GP. Building a practically useful theory of goal setting and task motivation: a 35-year odyssey. Am Psychol. 2002;57(9):705–717. PubMed doi:10.1037/0003-066X.57.9.705 29. Klamm EL, Wo’jcicki TR, White SM, Szabo AN, Kramer AF, McAuley E. Differential effects of physical activity intervention on self-efficacy in older adults. Ann Behav Med. 2009;36:S82. 30. Bravata DM, Smith-Spangler C, Sundaram V, et al. Using pedometers to increase physical activity and improve health: a systematic review. J Ame Medi Associ. 2007;298(19):2296–2304. PubMed doi:10.1001/ jama.298.19.2296
JPAH Vol. 12, No. 7, 2015
