Abstract
Three hundred and sixty-three older adults with multiple sclerosis completed a cross-sectional study examining hierarchical correlates of physical activity using a social cognitive theory perspective within a social ecological model (i.e., built environment, social environment, and individual social cognitive theory variables). Hierarchical linear regression analyses were conducted, wherein significant associations were noted for built environment (i.e., land-use mix diversity and aesthetics) and physical activity in Step 1 (R2 = .09). Social and built environment were significant correlates in Step 2 (R2 = .15). Finally in Step 3, individual social cognitive theory variables (i.e., self-efficacy and outcome expectations) were the only significant correlates of total physical activity (R2 = .38). Results were comparable for health-promoting physical activity; however, self-efficacy was the only significant correlate in Step 3 (R2 = .36). This study provides guidance for researchers and practitioners on relevant targets for tailoring interventions for older adults with multiple sclerosis and supports an emphasis on self-efficacy as a primary predictor of health behavior change.
Keywords: built environment, psychological theory, self-efficacy, social environment, social support
There are one million adults currently living with multiple sclerosis (MS) in the United States (Wallin et al., 2019). The highest prevalence of MS occurs in adults between 55 and 64 years of age, and this underscores the critical focus on individuals aging with MS (Wallin et al., 2019). Aging with MS is associated with declines in physical and cognitive functioning (Fox et al., 2015; Rao, Leo, Bernardin, & Unverzagt, 1991; Bollaert, Sandroff, Stine-Morrow, Sutton & Motl, 2019), and health-promoting behaviors such as physical activity may be effective and safe second-line approaches for improving MS symptoms and disease progression in older adults with MS (Motl & Pilutti, 2012; Bollaert & Motl, 2019). Older adults with MS engage in less physical activity and more sedentary behavior than middle-aged and young adults with MS (Klaren et al., 2016). There is a general lack of research on factors associated with variation in physical activity levels among older adults with MS, but such information is important for promoting adoption and maintenance of changes in this behavior.
Social cognitive theory (SCT) has frequently and effectively been included as a foundation for promoting physical activity behavior among persons with MS (Bandura, 2004; Motl, Pekmezi, & Wingo, 2018). The SCT includes the primary assumption of triadic reciprocal determinism whereby the individual, environment, and behavior dynamically interact (Bandura, 2004). The SCT further identifies self-efficacy as a primary determinant of behavior change and additional factors such as outcome expectations, social support, and environmental factors as perceived barriers/facilitators (Motl, McAuley, Snook, & Gliottoni, 2009; Suh, Joshi, Olsen, & Motl, 2014; Uszynski et al., 2018). Those variables from SCT include person level through social and environmental factors and can be aligned and studied through ecological frameworks (Winett, Williams, & Davy, 2009). Social ecological models (SEMs), in particular, provide a guide for conceptualizing the unique influences of macro-through-micro-level variables as influences of behavior and behavior change (Bronfenbrenner, 1979). Such a SEM and SCT framework would suggest hierarchical associations among built environment, social environment, and individual-level variables that distinctly influence physical activity behavior.
There has been interest in studying SEM and SCT in both older adults from the general population and persons with MS. One study examined the associations among SCT variables, built environment factors, and physical activity among older adults in the general population and reported the association between built environment characteristics and levels of moderate-to-vigorous physical activity (MVPA) that interacted with supportive SCT variables (i.e., high social support, high self-efficacy, and low barriers) (Carlson et al., 2012). Another study examined SCT variables embedded within SEM as correlates of physical activity among adults with MS and reported that individual-level variables (i.e., self-efficacy), social environment (i.e., social support), and built environment (i.e., aesthetics) were significantly associated with health-promoting physical activity (Silveira & Motl, 2019a). This supports additional research examining such associations among older adults with MS that could inform appropriate strategies and context for promoting physical activity in this particular population.
The current study examined SCT correlates of physical activity among older adults with MS using a SEM framework. We hypothesized that built environment, social environment, and individual-level SCT variables (i.e., self-efficacy and outcome expectations) would be associated with total and health-promoting physical activity. We further hypothesized that the associations between variables with physical activity would differ in strength in a way that is hierarchically consistent with the SEM framework as depicted in Figure 1, wherein proximal variables (i.e., individual SCT variables) are strongly associated.
Figure 1 —
Social ecological model depicting hypothesized associations among social cognitive theory variables.
Methods
Participants
Participants were recruited through the National MS Society via an e-mail distributed to 17,500 individuals with MS. The study was described as a survey examining thoughts and behaviors associated with physical activity among adults with MS over the age of 60 years. Inclusion criteria were self-reported age over 60 years and diagnosis of MS. Overall, 561 participants accessed the survey, 531 participants met recruitment criteria and enrolled, 363 participants completed the full questionnaire battery, and 168 participants did not complete the survey. Participants who did not complete the survey (n = 168) were significantly older and reported a longer MS disease duration than those who completed the survey (n = 363). This study includes only the 363 participants with complete data for all questionnaires.
Measures
Demographic and clinical characteristics.
Participants self-reported date of birth, sex, marital status, employment status, race, education, annual household income, year of MS diagnosis, and MS clinical course. The single-item Patient Determined Disease Steps scale measured self-reported disability status (Learmonth, Motl, Sandroff, Pula, & Cadavid, 2013). Scores on the Patient Determined Disease Steps range from 0 = Normal: I may have some mild symptoms, mostly sensory due to MS but they do not limit my activity. If I do have an attack, I return to normal when the attack has passed to 8 = Bedridden: Unable to sit in a wheelchair for more than one hour.
Built environment.
The Abbreviated Neighborhood Walkability Scale (NEWS-A) was administered as a measure of neighborhood walkability, notably perceived residential density (six items; range 173–865; α = .08), land-use mix diversity (23 items; range 1–5; α = .96), land-use mix access (three items; range 1–4; α = .87), street connectivity (two items; range 1–4; α = .67), infrastructure and safety for walking (six items; range 1–4; α = .84), aesthetics (four items; range 1–4; α = .79), traffic hazards (three items; range 1–4; α = .63), and crime (three items; range 1–4; α = .80) (i.e., built environment characteristics) (Cerin, Saelens, Sallis, Frank, et al., 2006). Residential density items assessed average person-density in the respondent’s immediate neighborhood on a 5-point Likert-like scale (1 = none; 5 = all) and scores are weighted relative to the average residential density. For example, “How common are apartments or condos 1–3 stories in your immediate neighborhood?” Land-use mix diversity was assessed by the walking proximity from home to 23 locations (e.g., supermarket), with responses ranging from 1- to 5-min walking distance to >30-min walking distance. Higher scores on the land-use mix diversity subscale indicate closer average proximity of various types of stores and facilities. The NEWS-A land-use mix access, street connectivity, infrastructure and safety for walking, aesthetics, traffic hazards, and crime subscale items were rated on a 4-point Likert scale (1 = strongly disagree; 4 = strongly agree). Higher scores for land-use mix access, street connectivity, infrastructure and safety for walking, and aesthetics indicate higher perceived walkability and higher scores for traffic hazards and crime indicate lower perceived walkability. Scores from the NEWS-A have been validated among persons with MS (Silveira, & Motl, in press).
Social environment.
Social support for physical activity was measured using the six-item Social Provisions Scale (SPS; Konopack & McAuley, 2012). Items address the six domains of attachment, social integration, reassurance of worth, reliable alliance, guidance, and opportunity for nurturance regarding current relationships and support for physical activity. Each item is rated on a 4-point Likert scale (1 = strongly disagree; 4 = strongly agree) and then summed into a total score from 6 to 36; higher scores reflect more perceived support for physical activity. The SPS has been established as a valid and reliable measure among persons with MS with internal consistency α = .90 (McAuley, Jerome, Marquez, Elavsky, & Blissmer, 2003; Suh et al., 2014).
Social cognitive theory.
The Exercise Self-Efficacy Scale (EXSE) assessed self-efficacy for engaging in 30 min or more of MVPA, most days of the week in 1-month increments across the next 6 months. The six items were rated on an 11-point scale (0 = not confident at all; 100 = highly confident) and averaged for a total score. The EXSE scale has been established as a valid and reliable measure among persons with MS with internal consistency α = .99 (Motl, Snook, McAuley, Scott, & Douglass, 2006).
The Multidimensional Outcome Expectations for Exercise Scale (MOEES) measured perceptions regarding physical, social, and self-evaluative benefits of physical activity and exercise (Wojcicki, White, & McAuley, 2009). The MOEES includes 19 items rated on a 5-point Likert scale (1 = strongly disagree; 5 = strongly agree), and items are summed to yield a total score (range 33–75), where higher scores indicate higher perceived benefits of participation in regular physical activity and exercise. The MOEES has been established as a valid and reliable measure among persons with MS with internal consistency α = .76–.83 (McAuley, Motl, White, & Wojcicki, 2010).
Physical activity.
The Godin Leisure-Time Exercise Questionnaire (GLTEQ) was used to measure total and health-promoting physical activity (Godin & Shephard, 1985). Participants self-report frequency of engagement in 15 or more min per day of strenuous, moderate, or mild physical activity during the previous week. The GLTEQ total score is calculated by multiplying the frequency of strenuous, moderate, and mild physical activity by nine, five, and three metabolic equivalents respectively and summing the weighted score with higher scores indicating more physical activity. The GLTEQ health contribution score (HCS) representing health-promoting physical activity is calculated using only the strenuous and moderate activity item scores. The GLTEQ is established as a valid and reliable measure of physical activity for persons with MS (α = .74, r = .34–.60; Sikes et al., 2018).
Procedures
Study procedures were approved by the University of Alabama at Birmingham Institutional Review Board. E-mail invitations included an active link for the survey that was hosted online via Qualtrics. The survey began with a consent form wherein participants selected an option for consenting (“I consent” or “I do not consent”). Participants were then assessed for inclusion criteria by confirming that they were 60 years of age or older (“Yes” or “No”) and having an MS diagnosis (“Yes” or “No”) before accessing any questionnaires. Participants were then guided through the survey and given the opportunity to receive remuneration (10 USD) by providing a residential address for receipt of a gift card. The scientific data were downloaded and stored as de-identified, separate from participant addresses. The median time for completing the survey was 30 min.
Statistical Analysis
All analyses were performed using SPSS Statistics 24 (IBM, Inc., Armonk, NY). Baseline descriptive characteristics are reported as mean ± SD unless otherwise noted. Due to nonnormal distributions, Spearman’s Rank-Order Correlations estimated the associations among GLTEQ total, GLTEQ HCS, NEWS-A subscales, SPS, EXSE, and MOEES scores. The magnitude of correlation coefficients was interpreted using Cohen’s guidelines for small (.10), medium (.30), and large (.50) effect sizes (Cohen, 2013). The bivariate correlations identified built environment, social environment, and individual SCT variables with significant associations with physical activity for entry in the regression analyses. Hierarchical linear regression analysis was conducted for GLTEQ total and GLTEQ HCS separately, and we regressed GLTEQ scores with NEWS-A subscales (built environment) in Step 1, SPS in Step 2 (social environment), and EXSE and MOEES (SCT individual determinants) in Step 3. We examined the β coefficients for each outcome predicting GLTEQ scores as well as model fit based on R2 and change in R2 (ΔR2) per step of the regression analyses.
Results
Participant demographics, clinical characteristics, and questionnaire mean scores are presented in Table 1. Participant mean age was 66 ± 6 years and years since MS diagnosis was 19 ± 11 years. The median (IQR) Patient Determined Disease Steps score was 3.0 (4.0) indicating moderate MS disability (i.e., gait disability: MS does interfere with my activities, especially my walking. I can work a full day, but athletic or physically demanding activities are more difficult than they used to be. I usually do not need a cane or other assistance to walk, but I might need some assistance during an attack). The majority of participants had a relapsing-remitting MS clinical course (62%), followed by secondary progressive MS (26%) and primary progressive MS (12%). Participants mostly identified as female (79%), married (69%), not employed (78%), European American (94%), and college educated (72%).
Table 1.
Sample Demographics, Clinical Characteristics, and Questionnaire Means
| Age, years (n = 356), mean ± SD | 65.7 ± 5.7 |
| MS duration, years (n = 363), mean ± SD | 19.0 ± 10.9 |
| PDDS (n = 363), median (IQR) | 3.0 (4.0) |
| MS clinical course (n = 360), n (%) | |
| RRMS | 223 (61.9) |
| Primary progressive | 42 (11.7) |
| Secondary progressive | 95 (26.4) |
| Gender (n = 363), n (%) | |
| Female | 285 (78.5) |
| Male | 78 (21.5) |
| Marital status (n = 362), n (%) | |
| Married | 251 (69.3) |
| Single | 24 (6.6) |
| Divorced/separated | 61 (16.9) |
| Widow/widower | 26 (7.2) |
| Employed (n = 363), n(%) | |
| Yes | 282 (77.7) |
| No | 81 (22.3) |
| Race (n = 363), n(%) | |
| European American | 340 (93.7) |
| African American | 9 (2.5) |
| Other | 14 (3.8) |
| Education (n = 363), n(%) | |
| High school | 21 (5.8) |
| 1–3 years college | 80 (22.0) |
| College graduate | 133 (36.6) |
| Master’s degree | 99 (27.3) |
| PhD or equivalent | 30 (8.3) |
| Annual household income (n = 341), n(%) | |
| <$15,000 | 8 (2.3) |
| $15,000-$24,000 | 34 (10.0) |
| $25,000-$49,000 | 67 (19.6) |
| $50,000-$74,000 | 79 (23.2) |
| $75,000-$99,000 | 65 (19.1) |
| ≥$100,000 | 88 (25.8) |
| Physical activity (GLTEQ), mean ± SD | |
| Total | 25.2 ± 21.3 |
| HCS | 15.2 ± 18.5 |
| Self-efficacy (EXSE), mean ± SD | |
| Total | 50.5 ± 38.1 |
| Outcome expectations (MOEES), mean ± SD | |
| Total | 60.9 ± 7.5 |
| Physical | 26.6 ± 3.0 |
| Self-evaluative | 21.7 ± 3.0 |
| Social | 12.6 ± 3.2 |
| Social support (SPS), mean ± SD | |
| Total | 17.7 ± 3.25 |
| Perceived built environment (NEWS-A), mean ± SD | |
| Residential density | 209.4 ± 72.4 |
| Land-use mix diversity | 1.9 ± 0.9 |
| Land-use mix access | 2.0 ± 1.0 |
| Street connectivity | 2.6 ± 1.0 |
| Infrastructure and safety for walking | 2.6 ± 0.9 |
| Aesthetics | 3.2 ± 0.7 |
| Traffic hazards | 2.5 ± 0.5 |
| Crime | 1.4 ± 0.6 |
Note. PDDS = patient determined disease steps; RRMS = relapsing-remitting multiple sclerosis; MS = multiple sclerosis; IQR = interquartile range; GLTEQ = Godin Leisure-Time Exercise Questionnaire; HCS = health contribution score; EXSE = Exercise Self-Efficacy Scale; MOEES = Multidimensional Outcome Expectations Scale; SPS = Social Provisions Scale; NEWS-A = The Abbreviated Neighborhood Walkability Scale.
Spearman rank-order correlation analyses results are presented in Table 2. Land-use mix diversity, land-use mix access, street connectivity, and aesthetics were significantly correlated with GLTEQ total, whereas land-use mix diversity, land-use mix access, infrastructure, and safety for walking, aesthetics, and crime were significantly correlated with GLTEQ HCS. Those associations were small in magnitude. The associations between SPS and MOEES with GLTEQ total and GLTEQ HCS were medium in magnitude, whereas associations between EXSE and GLTEQ total and GLTEQ HCS were large in magnitude.
Table 2.
Correlations Among Perceived Built Environment, Social Environment, Individual Determinants, and Physical Activity (n = 363)
| GLTEQ | ||
|---|---|---|
| Variable | Total | HCS |
| Built environment | ||
| NEWS-A residential density | .07 | .05 |
| NEWS-A land-use mix diversity | .26** | .25** |
| NEWS-A land-use mix access | .23** | .21** |
| NEWS-A street connectivity | .14** | .10 |
| NEWS-A infrastructure and safety for walking | .10 | .12* |
| NEWS-A aesthetics | .20** | .19** |
| NEWS-A traffic hazards | −.02 | −.05 |
| NEWS-A crime | −.08 | −.11* |
| Social environment | ||
| SPS | .33** | .30** |
| Individual determinants from SCT | ||
| EXSE | .66** | .66** |
| MOEES | .41** | .37** |
Note. GLTEQ = Godin Leisure-Time Exercise Questionnaire; HCS = health contribution score; NEWS-A = The Abbreviated Neighborhood Walkability Scale; SPS = Social Provisions Scale; SCT = social cognitive theory; EXSE = Exercise Self-Efficacy Scale; MOEES = Multidimensional Outcome Expectations Scale.
p < .05.
p < .01.
The significant correlates identified in the bivariate analysis were then entered into the regression for explaining variance in GLTEQ total and GLTEQ HCS scores using hierarchical linear regression. The results for GLTEQ total analyses are presented in Table 3. Step 1 included built environment variables, and there were significant associations noted for land-use mix diversity (β = 0.16) and aesthetics (β = 0.16). Those variables explained 9% of the variance (R2 = .09). Social environment (SPS) was included in Step 2, and significant associations were noted for SPS (β = 0.24), land-use mix diversity (β = 0.16), and aesthetics (β = 0.13). Step 2 variables accounted for 15% of the variance in total physical activity (R2 = .15), and SPS explained an additional 6% of variance over built environment variables alone (ΔR2 = .06). The final regression model in Step 3 included self-efficacy and outcome expectations as established SCT individual determinants of physical activity among older adults. The addition of EXSE and MOEES scores explained 38% of the variance in total physical activity (R2 = .38), and these individual-level determinants explained an additional 24% of the variance over built and social environment variables (ΔR2 = .24).
Table 3.
Hierarchical Linear Regression SEM Model Predicting Total Physical Activity (n = 363)
| GLTEQ total | B | SE B | β | R2 | ΔR2 |
|---|---|---|---|---|---|
| Step 1 | .09 | ||||
| NEWS-A land-use mix diversity | 3.78 | 1.53 | 0.16* | ||
| NEWS-A land-use mix access | 1.68 | 1.47 | 0.08 | ||
| NEWS-A street connectivity | 0.61 | 1.27 | 0.03 | ||
| NEWS-A aesthetics | 2.02 | 1.68 | 0.16** | ||
| Step 2 | .15 | .06*** | |||
| NEWS-A land-use mix diversity | 3.79 | 1.50 | 0.16* | ||
| NEWS-A land-use mix access | 1.16 | 1.43 | 0.05 | ||
| NEWS-A street connectivity | 0.36 | 1.24 | 0.02 | ||
| NEWS-A aesthetics | 4.02 | 1.64 | 0.13* | ||
| SPS | 1.55 | 0.33 | 0.24*** | ||
| Step 3 | .38 | .24*** | |||
| NEWS-A land-use mix diversity | 1.42 | 1.30 | 0.06 | ||
| NEWS-A land-use mix access | 0.67 | 1.23 | 0.03 | ||
| NEWS-A street connectivity | 0.54 | 1.06 | 0.02 | ||
| NEWS-A aesthetics | 1.77 | 1.43 | 0.06 | ||
| SPS | 0.38 | 0.30 | 0.06 | ||
| EXSE | 0.27 | 0.03 | 0.49*** | ||
| MOEES | 0.30 | 0.14 | 0.10* |
Note. SEM = social ecological model; GLTEQ = Godin Leisure-Time Exercise Questionnaire; NEWS-A = The Abbreviated Neighborhood Walkability Scale; SPS = Social Provisions Scale; EXSE = Exercise Self-Efficacy Scale; MOEES = Multidimensional Outcome Expectations Scale.
p < .05.
p < .01.
p < .001.
We then conducted hierarchical regression analyses for health-promoting physical activity (GLTEQ HCS) using unique variables from the correlation analysis (Table 4). Step 1 included built environment variables and there were significant associations noted for land-use mix diversity (β = 0.18), aesthetics (β = 0.11), and crime (β = −0.13) that explained 10% of the variance in physical activity (R2 = .10). Social environment (SPS) was included in Step 2 and there were significant associations noted for SPS (β = 0.21) and only land-use mix diversity (β = 0.19); those variables accounted for 14% of the variance in health-promoting physical activity (R2 = .14). The inclusion of SPS explained an additional 4% over built environment variables alone (ΔR2 = .04) and associations between health-promoting physical activity and aesthetics and crime were attenuated when including SPS. Step 3 included self-efficacy (β = 0.49) and outcome expectations (β = 0.07), and the variables collectively and significantly explained 36% of the variance in health-promoting physical activity (R2 = .36). Self-efficacy was the only significant predictor in the model which explained an additional 22% of variance over built and social environment variables (ΔR2 = .22).
Table 4.
Hierarchical Linear Regression SEM Model Predicting Health-Promoting Physical Activity (n = 363)
| GLTEQ total | B | SE B | β | R2 | ΔR2 |
|---|---|---|---|---|---|
| Step 1 | .10 | ||||
| NEWS-A land-use mix diversity | 3.78 | 1.33 | 0.18** | ||
| NEWS-A land-use mix access | 1.64 | 1.28 | 0.09 | ||
| NEWS-A infrastructure and safety for walking | 0.13 | 1.24 | 0.01 | ||
| NEWS-A aesthetics | 3.10 | 1.50 | 0.11* | ||
| NEWS-A crime | −3.93 | 1.61 | −0.13* | ||
| Step 2 | .14 | .04*** | |||
| NEWS-A land-use mix diversity | 3.81 | 1.30 | 0.19** | ||
| NEWS-A land-use mix access | 1.00 | 1.26 | 0.05 | ||
| NEWS-A infrastructure and safety for walking | 0.04 | 1.21 | 0.002 | ||
| NEWS-A aesthetics | 2.66 | 1.47 | 0.10 | ||
| NEWS-A crime | −2.38 | 1.62 | −0.08 | ||
| SPS | 1.19 | 0.29 | 0.21*** | ||
| Step 3 | .36 | .22*** | |||
| NEWS-A land-use mix diversity | 1.74 | 1.14 | 0.09 | ||
| NEWS-A land-use mix access | 0.12 | 1.10 | 0.01 | ||
| NEWS-A infrastructure and safety for walking | 0.73 | 1.05 | 0.03 | ||
| NEWS-A aesthetics | 1.13 | 1.29 | 0.04 | ||
| NEWS-A crime | −0.51 | 1.41 | −0.02 | ||
| SPS | 0.32 | 0.27 | 0.06 | ||
| EXSE | 0.24 | 0.03 | 0.49*** | ||
| MOEES | 0.17 | 0.12 | 0.07 |
Note. SEM = social ecological model; GLTEQ = Godin Leisure-Time Exercise Questionnaire; NEWS-A = The Abbreviated Neighborhood Walkability Scale; SPS = Social Provisions Scale; EXSE = Exercise Self-Efficacy Scale; MOEES = Multidimensional Outcome Expectations Scale.
p < .05.
p < .01.
p < .001.
Discussion
This study examined SCT variables within a SEM framework as correlates of physical activity behavior, as such an approach could inform behavioral interventions for older adults with MS. We noted significant, hierarchical associations among built environment, social environment, and SCT individual-level variables as determinants of both total and health-promoting physical activity. The SCT individual-level variables, namely self-efficacy and outcome expectations, were the primary determinants of physical activity among older adults. When accounting for SCT individual-level variables, social environment and built environment variables were no longer statistically significant predictors of physical activity behavior. Overall, this study supports the hypothesis that SEM macro-through-micro-level environmental variables hierarchically influence physical activity, and this supports an emphasis on the strength of SCT individual-level variables as fundamental predictors of behavior among older adults with MS.
The theoretical basis of SCT asserts that self-efficacy is the primary determinant of behavior change as it relates to the concept of self-control and the ability to modulate behavior to reach overarching goals (Bandura, 2004). Results from this study relating to self-efficacy are consistent with a large body of evidence examining physical activity and SCT in MS (Motl et al., 2009; Silveira & Motl, 2019b; Suh et al., 2014; Uszynski et al., 2018; White, Wójcicki, & McAuley, 2011). For example, a recent study examining multilevel environmental correlates of physical activity in adults over the age of 18 with MS reported self-efficacy as the primary predictor of self-reported physical activity, using SCT embedded with the SEMs framework (Silveira & Motl, 2019a). Behavioral interventions might target self-efficacy for physical activity through the four established sources of self-efficacy: mastery experiences, vicarious experiences, social support, and physical and emotional responses (Bandura, 2004). Evidence-based behavior change strategies that increase these sources of self-efficacy include action planning, positive feedback, and reflection on experiences that could be supported by behavioral coaching in conjunction with strategies focused on positive outcome expectations.
Outcome expectations have been highlighted as a primary factor within behavioral interventions that promote lifestyle physical activity in MS (Silveira & Motl, 2019b). Our results highlight outcome expectations as a small, but significant predictor of total physical activity among older adults with MS. This may support inclusion of evidence-based behavioral strategies that guide and reinforce symbolic thinking, vicarious experiences, and incentive value. Outcome expectations predicted total, but not health-promoting physical activity (i.e., MVPA), and this warrants further inquiry given recent literature highlighting an association between MVPA and walking performance among older adults with MS (Baird et al., 2019). We believe that further research might identify appropriate strategies and targets to promote MVPA among older adults beyond self-efficacy that could aid in management of walking impairment in this population.
Social support for exercise was not a significant predictor of physical activity when accounting for individual-level determinants from SCT. This is not consistent with previous literature among persons with MS or older adults (Carlson et al., 2012; Motl et al., 2009). We posit that social support may be indirectly associated with physical activity, or that perhaps only certain types of social support apply among older adults with MS. Perceived built environment variables did not significantly explain total or health-promoting physical activity in the current study when accounting for social environment and SCT individual-level variables. There is existing literature that highlights an association between physical activity and built environment variables among older adults; however, the majority of this research has been conducted in other countries where cultural norms and accessibility for alternate forms of physical activity such as walking for transport may be more universal (Barnett, Barnett, Nathan, Van Cauwenberg, & Cerin, 2017). We posit that built and social environment should be addressed in future behavioral interventions; however, these factors must be customized to individual settings and perceptions.
This study is not without limitations. Participants self-reported all data and the cross-sectional nature of this research prevents causal inferences regarding multilevel environmental influences of physical activity. However, our study design allowed for a robust, nationally representative examination of physical activity and its potential determinants among older adults with MS. This sample primarily included females, who reported high income and education, and this limits generalizability of results broadly among persons with MS. Further consideration of individual factors, such as day-to-day functioning, mobility, pain, and MS disease state, is required given the bidirectional association between physical activity and MS-specific factors (Motl et al., 2008). For example, given the episodic nature of MS disease, variables such as self-efficacy may have limited influence on physical activity during periods of disease activity in the form of relapses/episodes. The social support questionnaire utilized in this study includes six items that do not extensively inquire about different sources of support that may uniquely influence physical activity. Further focal research examining the six provisions or sources of social support among older adults with MS is needed. Additional questionnaire considerations for future research include example activities presented in the standard GLTEQ that may not be appropriate for persons with MS, and researchers may consider additional measures of built environment such as objective measures of the built environment.
Collectively, self-efficacy and outcome expectations were the primary determinants of total physical activity behavior among older adults with MS. Self-efficacy was the only determinant of health-promoting physical activity when accounting for multilevel environmental variables, partially confirming our hypothesis. Our models were theoretically grounded in SEM and SCT for identifying targets that might guide the promotion of physical activity among older adults with MS. Overall, this study provides guidance for researchers and practitioners on the most relevant targets for tailoring physical activity interventions for older adults with MS and supports the continued emphasis on self-efficacy as a primary predictor of health behavior change and maintenance.
Acknowledgments
S.L. Silveira is supported by R.W. Motl’s National Multiple Sclerosis Society Mentor-Based Postdoctoral Fellowship (MB 0029). J.F. Baird is supported by the National Institutes of Health training grant (2T32 HD071866-06).
References
- Baird JF, Cederberg KLJ, Sikes EM, Silveira SL, Jeng B, Sasaki JE, … Motl RW (2019). Physical activity and walking performance across the lifespan among adults with multiple sclerosis. Multiple Sclerosis and Related Disorders, 35, 36–41. doi: 10.1016/j.msard.2019.07.003 [DOI] [PubMed] [Google Scholar]
- Bandura A (2004). Health promotion by social cognitive means. Health Education & Behavior, 31(2), 143–164. doi: 10.1177/1090198104263660 [DOI] [PubMed] [Google Scholar]
- Barnett DW, Barnett A, Nathan A, Van Cauwenberg J, & Cerin E (2017). Built environmental correlates of older adults’ total physical activity and walking: A systematic review and meta-analysis. International Journal of Behavioral Nutrition and Physical Activity, 14(1), 103. doi: 10.1186/s12966-017-0558-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bollaert RE, & Motl RW (2019). Physical and cognitive functions, physical activity, and sedentary behavior in older adults with multiple sclerosis. Journal of Geriatric Physical Therapy, 42(4), 304–312. doi: 10.1519/JPT.0000000000000163 [DOI] [PubMed] [Google Scholar]
- Bollaert RE, Sandroff BM, Stine-Morrow EA, Sutton BP, & Motl RW (2019). The intersection of physical function, cognitive performance, aging, and multiple sclerosis: A cross-sectional comparative study. Cognitive And Behavioral Neurology, 32(1), 1–10. doi: 10.1097/WNN.0000000000000179 [DOI] [PubMed] [Google Scholar]
- Bronfenbrenner U (1979). The ecology of human development. New York, NY: Harvard University Press. [Google Scholar]
- Carlson JA, Sallis JF, Conway TL, Saelens BE, Frank LD, Kerr J, … King AC (2012). Interactions between psychosocial and built environment factors in explaining older adults’ physical activity. Preventive Medicine, 54(1), 68–73. doi: 10.1016/j.ypmed.2011.10.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cerin E, Saelens BE, Sallis JF, Frank LD (2006). Neighborhood environment walkability scale: Validity and development of a short form. Medicine & Science in Sports & Exercise, 38(9), 1682–1691. doi: 10.1249/01.mss.0000227639.83607.4d [DOI] [PubMed] [Google Scholar]
- Cohen J (2013). Statistical power analysis for the behavioral sciences. New York: Routledge. [Google Scholar]
- Fox RJ, Bacon TE, Chamot E, Salter AR, Cutter GR, Kalina JT, & Kister I (2015). Prevalence of multiple sclerosis symptoms across lifespan: Data from the NARCOMS Registry. Neurodegenerative Disease Management, 5(6, Suppl.), 3–10. doi: 10.2217/nmt.15.55 [DOI] [PubMed] [Google Scholar]
- Godin G, & Shephard R (1985). A simple method to assess exercise behavior in the community. Canadian Journal of Applied Sport Science, 10(3), 141–146. [PubMed] [Google Scholar]
- Klaren RE, Sebastiao E, Chiu C-Y, Kinnett-Hopkins D, McAuley E, & Motl RW (2016). Levels and rates of physical activity in older adults with multiple sclerosis. Aging and Disease, 7(3), 278. doi: 10.14336/AD.2015.1025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Konopack JF, & McAuley E (2012). Efficacy-mediated effects of spirituality and physical activity on quality of life: A path analysis. Health and Quality of Life Outcomes, 10(1), 57. doi: 10.1186/1477-7525-10-57 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Learmonth YC, Motl RW, Sandroff BM, Pula JH, & Cadavid D (2013). Validation of patient determined disease steps (PDDS) scale scores in persons with multiple sclerosis. BMC Neurology, 13(1), 37. doi: 10.1186/1471-2377-13-37 [DOI] [PMC free article] [PubMed] [Google Scholar]
- McAuley E, Jerome GJ, Marquez DX, Elavsky S, & Blissmer B (2003). Exercise self-efficacy in older adults: Social, affective, and behavioral influences. Annals of Behavioral Medicine, 25(1), 1–7. doi: 10.1207/S15324796ABM2501_01 [DOI] [PubMed] [Google Scholar]
- McAuley E, Motl RW, White SM, & Wojcicki TR (2010). Validation of the multidimensional outcome expectations for exercise scale in ambulatory, symptom-free persons with multiple sclerosis. Archives of Physical Medicine and Rehabilitation, 91(1), 100–105. doi: 10.1016/j.apmr.2009.09.011 [DOI] [PubMed] [Google Scholar]
- Motl RW, Arnett PA, Smith MM, Barwick FH, Ahlstrom B, & Stover EJ (2008). Worsening of symptoms is associated with lower physical activity levels in individuals with multiple sclerosis. Multiple Sclerosis Journal, 14(1), 140–142. doi: 10.1177/1352458507079126 [DOI] [PubMed] [Google Scholar]
- Motl RW, McAuley E, Snook EM, Gliottoni RC (2009). Physical activity and quality of life in multiple sclerosis: Intermediary roles of disability, fatigue, mood, pain, self-efficacy and social support. Psychology, Health, & Medicine, 14(1), 111–124. doi: 10.1080/13548500802241902 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Motl RW, Pekmezi D, & Wingo BC (2018). Promotion of physical activity and exercise in multiple sclerosis: Importance of behavioral science and theory. Multiple Sclerosis Journal—Experimental, Translational and Clinical, 4(3), 2055217318786745. doi: 10.1177/2055217318786745 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Motl RW, & Pilutti LA (2012). The benefits of exercise training in multiple sclerosis. Nature Reviews Neurology, 8(9), 487. doi: 10.1038/nrneurol.2012.136 [DOI] [PubMed] [Google Scholar]
- Motl RW, Snook EM, McAuley E, Scott JA, & Douglass ML (2006). Correlates of physical activity among individuals with multiple sclerosis. Annals of Behavioral Medicine, 32(2), 154–161. doi: 10.1207/s15324796abm3202_13 [DOI] [PubMed] [Google Scholar]
- Rao SM, Leo GJ, Bernardin L, & Unverzagt F (1991). Cognitive dysfunction in multiple sclerosis.: I. Frequency, patterns, and prediction. Neurology, 41(5), 685–691. doi: 10.1212/WNL.41.5.685 [DOI] [PubMed] [Google Scholar]
- Sikes EM, Richardson EV, Cederberg KJ, Sasaki JE, Sandroff BM, Motl RW (2018). Use of the Godin leisure-time exercise questionnaire in multiple sclerosis research: A comprehensive narrative review. Disability and Rehabilitation, 41(11), 1243–1267. doi: 10.1080/09638288.2018.1424956 [DOI] [PubMed] [Google Scholar]
- Silveira SL, & Motl RW (2019a). Environmental correlates of health-promoting leisure physical activity in persons with multiple sclerosis using a social cognitive perspective embedded within social ecological model. Preventive Medicine Reports, 15, 100921. doi: 10.1016/j.pmedr.2019.100921 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Silveira SL, & Motl RW (2019b). Do social cognitive theory constructs explain response heterogeneity with a physical activity behavioral intervention in multiple sclerosis? Contemporary Clinical Trials Communication, 15, 100366. doi: 10.1016/j.conctc.2019.100366 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Silveira SL, & Motl RW (in press). Validation of neighborhood environment walkability scale scores in persons with multiple sclerosis. Journal of Transport and Health. [Google Scholar]
- Suh Y, Joshi I, Olsen C, & Motl RW (2014). Social cognitive predictors of physical activity in relapsing-remitting multiple sclerosis. International Journal of Behavioral Medicine, 21(6), 891–898. doi: 10.1007/s12529-013-9382-2 [DOI] [PubMed] [Google Scholar]
- Uszynski MK, Casey B, Hayes S, Gallagher S, Purtill H, Motl RW, & Coote S (2018). Social cognitive theory correlates of physical activity in inactive adults with multiple sclerosis. International Journal of MS Care, 20(3), 129–135. doi: 10.7224/1537-2073.2016-111 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wallin MT, Culpepper WJ, Campbell JD, Nelson LM, Langer-Gould A, Marrie RA, … LaRocca NG (2019). The prevalence of MS in the United States: A population-based estimate using health claims data. Neurology, 92(10), e1029–e1040. doi: 10.1212/WNL.0000000000007035 [DOI] [PMC free article] [PubMed] [Google Scholar]
- White SM, Wójcicki TR, & McAuley E (2011). Social cognitive influences on physical activity behavior in middle-aged and older adults. The Journals of Gerontology, Serics B: Psychological Sciences & Social Sciences, 67(1), 18–26. doi: 10.1093/geronb/gbr064 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Winett RA, Williams DM, & Davy BM (2009). Initiating and maintaining resistance training in older adults: A social cognitive theory-based approach. British Journal of Sports Medicine, 43(2), 114–119. doi: 10.1136/bjsm.2008.049361 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wojcicki TR, White SM, & McAuley E (2009). Assessing outcome expectations in older adults: The multidimensional outcome expectations for exercise scale. The Journals of Gerontology, Serics B: Psychological Sciences & Social Sciences, 64(1), 33–40. doi: 10.1093/geronb/gbn032 [DOI] [PMC free article] [PubMed] [Google Scholar]