Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Nov 24.
Published in final edited form as: West J Nurs Res. 2019 Aug 30;42(8):581–592. doi: 10.1177/0193945919871697

Examining Potential Psychosocial Mediators in a Physical Activity Intervention for Older Adults

Siobhan K McMahon 1, Beth Lewis 2, J Michael Oakes 3, Jean F Wyman 1, Weihua Guan 4, Alexander J Rothman 5
PMCID: PMC7684560  NIHMSID: NIHMS1635543  PMID: 31470769

Abstract

The purpose of this study was to examine psychosocial constructs targeted as potential mediators in a prior physical activity (PA) intervention study. This secondary analysis used data from 102 older adults randomized to one of four conditions—within a 2 (Interpersonal Strategies: yes, no) × 2 (Intrapersonal Strategies: yes, no) factorial design. We tested intervention effects on social support, self-efficacy, self-regulation, and goal attainment, and whether these constructs mediated intervention effects on PA. Participants who received interventions with interpersonal strategies, compared to those who did not, increased their readiness (post-intervention), the self-regulation subscale of self-assessment, and goal attainment (post-intervention, 6-months). Participants who received interventions with intrapersonal strategies, compared to those who did not, increased their social support from family (post-intervention). There was no statistically significant mediation. To understand mechanisms through which interventions increase older adults’ PA and to improve intervention effectiveness, researchers should continue to examine potential psychosocial mediators.

Clinical Trial Registry: NCT02433249.

Keywords: physical activity, psychosocial mediators, older adults

The types and doses of physical activity (PA) that are beneficial for older adults have been well-described in the current literature (Bauman, Merom, Bull, Buchner, & Fiatarone Singh, 2016), and yet 87% of adults over the age of 65 are relatively inactive (U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion, 2014). In general, the interventions developed to increase older adults’ PA have had small positive effects post intervention (Chase, 2014). However, building an evidence base regarding the psychosocial mechanisms through which interventions increase older adults’ PA will help to increase their effectiveness. To do this, intervention researchers should thoroughly examine the psychosocial constructs targeted by the intervention under investigation (Connell et al., 2019; Sheeran, Klein, & Rothman, 2017). Thus, the purpose of this paper is to describe the role of constructs targeted as potential psychosocial mediators in a recent intervention study designed to increase older adults’ PA (McMahon et al., 2017).

Evidence-based PA guidelines for older adults recommend muscle-strengthening activities at least twice per week, moderately intense aerobic activities for at least 150 minutes per week, and, for those with reduced mobility, balance-challenging movements at least three times per week (Physical Activity Guidelines Advisory Committee, 2018). To promote these guidelines and increase PA in older adults, researchers have developed various interventions with average effect sizes ranging from 0.14 to 0.18 (Chase, 2014, French, Olander, Chisholm, & McSharry, 2014). The results of meta-analyses have also shown that positive effects are associated with intervention studies guided by theory (Chase, 2014) and include certain techniques for changing one’s behavior, such as overcoming barriers to PA (French et al., 2014). This evidence has advanced our understanding of the characteristics and content of interventions that lead to small increases in older adults’ PA. However, little is known about the processes through which older adults who receive interventions are motivated to increase and maintain their PA. To design more effective interventions, it is necessary to develop a better understanding of the psychosocial mechanisms through which such interventions work.

Although studies often describe the psychosocial constructs targeted as potential mediators, they seldom thoroughly examine the links between these potential mediators, the intervention, and the PA outcome (Carey et al., 2019; Keller, Fleury, Sidani, & Ainsworth, 2009). For example, prior research has shown that interventions have positive effects on potential psychosocial mediators such as self-regulation (Olson & McAuley, 2015), self-efficacy (French et al., 2014), and social support (Yeom & Fleury, 2014). However, the few reports that also assess the effects of these mediators on PA have reported mixed results (Barrera, Toobert, Angell, Glasgow, & Mackinnon, 2006; Becofsky, Baruth, & Wilcox, 2014; van Stralen, de Vries, Mudde, Bolman, & Lechner, 2011). To improve our understanding of the psychosocial mechanisms through which interventions increase older adults’ PA, behavior change experts have recommended that researchers systematically examine the links between interventions, potential psychosocial mediators, and behavioral outcomes within their studies (Nielsen et al., 2018). This evidence will accelerate the design and development of more effective interventions by strengthening the links between their content and the psychosocial constructs they target.

Purpose

In this paper, we examine five potential psychosocial mediators—social support (i.e., provided by friends and family), readiness, self-efficacy, self-regulation (i.e., benefits, self-assessment, and integration), and goal attainment—targeted by the interventions in a recent PA study in older adults (McMahon et al., 2017). Our research questions included:

  1. What were older adults’ ratings of the five potential psychosocial mediators at baseline, post-intervention, and the six-month follow-up?

  2. Did participants who received interventions with interpersonal behavior change strategies, compared to those who did not, rate the potential psychosocial mediators higher post intervention and/or at the six-month follow-up?
    1. Is there evidence that the potential psychosocial mediators engaged by interventions with interpersonal behavior change strategies mediated intervention effects on PA, post intervention and/or at the six-month follow up?
  3. Did the participants who received interventions with intrapersonal behavior change strategies, compared to those who did not, rate the potential psychosocial mediators higher post intervention and/or at the six-month follow-up?
    1. Is there evidence that the potential psychosocial mediators engaged by interventions with intrapersonal behavior change strategies mediated intervention effects on PA, post intervention and/or at the six-month follow up?

Methods

All study procedures were reviewed and approved by the University of Minnesota Institutional Review Board. Each participant in the study provided informed consent. The study is registered with ClinicalTrials.gov (NCT02433249).

Design

This study was a secondary analysis of a 2×2 factorial intervention study (see Table 1). Given the findings of this study (McMahon et al., 2016), we assumed that the intervention would have a medium-sized effect (0.40) on the potential psychosocial mediators. Thus, we estimated that a sample size of 100 would enable us to detect medium effect sizes with 80% power and a significance level of 0.05 in two-tailed hypothesis tests.

Table 1.

Main intervention components across four study conditions.

Core Intervention Components Experimental Intervention Components
Condition Otago Exercise Protocol Physical Activity Monitor Interpersonal BCS Intrapersonal BCS
1 Yes Yes No No
2 Yes Yes Yes No
3 Yes Yes No Yes
4 Yes Yes Yes Yes
Open in a new tab

Notes: BCS = behavior change strategies.

Primary Study

The primary study was guided by the wellness motivation theory (WMT), which posits that older adults’ health-related actions (e.g., PA) are influenced by social contextual factors and behavioral change processes (Perez & Fleury, 2009). The objective was to assess whether all behavior change strategies within the intervention are necessary to produce positive effects. To do so, behavior change strategies were organized into two theoretically and empirically plausible sets—interpersonal (e.g., social support and social comparison) and intrapersonal (e.g., goal setting)—which enabled an assessment of their unique and combined effects. According to the WMT and prior research (McMahon et al., 2016; Silva-Smith, Fleury & Belyea, 2013; Yeom & Fleury, 2014), both of these sets of strategies can work by increasing perceptions of social support (i.e., friends and family), readiness, self-efficacy, self-regulation (i.e., benefits, self-assessment, and integration), and goal attainment (see Figure 1). Participants were randomized to one of four conditions as a product of the 2 (Interpersonal Strategies: yes, no) × 2 (Intrapersonal Strategies: yes, no) factorial design. The results of the primary study showed that participants who received interventions with interpersonal behavior change strategies, compared to those who did not, increased their PA up to six months post intervention (McMahon et al., 2017).

Figure 1.

Figure 1.

Open in a new tab

Conceptual model of the intervention study.

Participants

The participants (N = 102) were community-dwelling adults who were of age ≥ 70 years; able to walk; had reported no lower extremity injury, infection, or surgery in the last six weeks; reported no neurocognitive disorder; scored >21 (out of 26 points) on the telephone-based Mini-Mental State Exam (Newkirk et al., 2004); and reported PA levels below the national guidelines for older adults (Topolski et al., 2006).

Intervention Content

As in the prior report (McMahon et al., 2017), there were four intervention components: two core and two experimental (Table 1). The two core intervention components were a PA protocol and a PA monitor, both of which were utilized in all four conditions. The PA protocol was based on the evidence-based Otago Exercise Program and adapted for small groups (Kyrdalen, Moen, Røysland, & Helbostad, 2013; Shubert, Smith, Goto, Jiang, & Ory, 2017). It involves five leg-strengthening exercises, eleven balance-challenging exercises, five flexibility exercises, and encouragement to increase the time spent walking. The protocol also guides interventionists to gradually introduce exercises according to individuals’ abilities and to counsel recipients to safely perform these exercises at least three times per week (Gardner, Buchner, Robertson, & Campbell, 2001). The PA monitors (Fitbit One™) were given to participants at the beginning of the study with instructions for using the monitors and obtaining assistance to troubleshoot problems with them throughout the study.

The two experimental intervention components included two distinct sets of behavior change strategies: interpersonal and intrapersonal. The interpersonal set was provided to participants randomized to conditions two and four (Table 1). Strategies in this set included encouraging one to increase social support for PA; encouraging friendly social comparisons; engaging in group problem-solving to overcome environmental and social barriers to PA; identifying oneself as a role model; and integrating more PA into one’s social routines by, for example, using community resources. The intrapersonal set of behavior change strategies was provided to participants randomized to conditions three and four (Table 1). Strategies in this set included encouraging the development of personally meaningful, realistic, and specific PA goals; identifying what makes PA satisfying; engaging in problem-solving to overcome personal barriers to PA; integrating PA into personal routines; and developing anticipatory plans to cope with potential setbacks or disruptions.

The interventions in each condition were manualized and delivered by a trained interventionist to small groups (four–six participants) that met 90 minutes each week for eight weeks. The participants received workbooks designed specifically for the condition to which they were randomized.

Data Collection

Data collection took place from April 2014 to March 2016. Data were collected from each participant at the baseline, post intervention, and at the six-month follow-up by trained research assistants who were masked to the conditions. This process included assessments comprised of structured questionnaires and observations of functional strength and balance according to a short physical performance battery. Data were directly entered into REDCap and were hosted at the University of Minnesota (Harris et al., 2009) on secure tablet computers. PA outcome data was captured using monitors (Fitbit One™) and Fitabase (2015).

Measures

Demographic and health variables.

The self-reported demographic variables included age, sex, number of persons living in the household, annual income, educational attainment, race, and ethnicity. Self-reported chronic conditions included osteoarthritis, diabetes, cardiovascular disease, and lung disease. Self-reported fall risks included a history of one or more falls in the last year, feelings of unsteadiness while standing or walking, and worry about falls (Stevens & Phelan, 2013).

Potential psychosocial mediators

Social support.

Social support was assessed using the 26-item Social Support and Exercise Survey (Sallis, Grossman, Pinski, Patterson, & Nader, 1987), which captures the extent to which a person believes they receive support for PA from friends and family (e.g., how often a friend or family member offers to schedule outings around their PA). The response scales for each item ranged from 1 (not at all) to 5 (very often). Cronbach’s alphas for the family and friend subscales in this study were 0.88 and 0.90, respectively.

Self-efficacy.

Self-efficacy was assessed using the 13-item Barriers Self-Efficacy Scale, which captures a person’s confidence to engage in recommended PAs when facing common barriers such as pain or bad weather (McAuley, 1992). The response scales for each item ranged from 0 (no confidence) to 100 (complete confidence). Cronbach’s alpha was 0.90.

Readiness.

Readiness was assessed with three items from the Index of Readiness, which captures the frequency with which a person thinks about their ability and commitment to engage in PA, and whether they have plans for staying active (Fleury, 1994). The response scales for each item range from 1 (not at all) to 5 (very often). Cronbach’s alpha was 0.76.

Self-regulation.

Self-regulation was assessed with the nine-item Index of Self-Regulation, which is divided into three subscales (Fleury, 1998; Moore et al., 2016; Yeom & Fleury, 2011). The benefits subscale captures the frequency with which a person thinks about the benefits and importance of fall-reducing PA. The self-assessment subscale captures the frequency with which a person self-tracks their exercise, looks for signs of progress, and assesses the extent to which goals are achieved. The integration subscale captures the extent to which a person integrates fall-reducing PA into everyday life. The response ranges for each item in all subscales were 1 (not at all) to 5 (very often). Cronbach’s alphas for the benefits, self-assessment, and integration subscales were 0.85, 0.85, and 0.86, respectively.

Goal attainment.

Goal attainment was assessed using the Goal Attainment Scale, which captures the extent to which a person achieves their personal goals related to PA and fall prevention (Kiresuk & Sherman, 1968; Toto, Skidmore, Terhorst, Rosen, & Weiner, 2015). Prior to intervention, participants identified two to four such goals and refined them to be specific, measurable, attainable, and timed according to the eight-week intervention and six-month follow-up period. Additionally, participants rated their perceptions of the importance of each goal and the difficulty of attaining the goal using response scales ranging from 1 (not at all) to 5 (extremely). In line with the procedures described by Kiresuk, Smith, and Cardillo (2014) and Turner-Stokes (2009), participants created a five-point attainment scale ranging from −2 to +2 for each goal. First, they identified an expected outcome for each goal, which was designated as the indicator of attainment (0). Next, four related indicators that represent not attaining each goal (−2, −1) and exceeding each goal (+1, +2) were created. Participants used the scale to assess the extent to which they achieved their personal goals (or not). The ratings for each goal were weighted based on their perceived difficulty and importance and then aggregated into a single composite T-score for each goal at each time point. Participants’ individual T-scores ranged from <35 (much worse goal attainment than expected) to >65 (much better goal attainment than expected), with a mean of 50 and a standard deviation of 10. Gerontological researchers have described this person-centered approach to capturing goal attainment as sensitive to changes in individual behavior (Rockwood et al., 2003).

Outcome variable

Physical activity.

Research assistants assessed the duration of PA based on the average total number of active minutes each day captured by the accelerometers in participants’ PA monitors (Fitbit One™) over the course of a week. According to the recent literature, Fitbit Ones™ can accurately capture steps and estimate expended energy (Case, Burwick, Volpp, & Patel, 2015; Diaz et al., 2015; Floegel, Florez-Pregonero, Hekler, & Buman, 2014). Participants wore their PA monitors while awake for seven consecutive days at each measurement time point. They were asked to remove their monitor only for sleep or water-based activities (Cain & Geremia, 2011). Total activity duration was considered the sum of valid minutes of activity—which was classified as light, moderate (“fairly active”), or vigorous (“very active”)—excluding the minutes classified as sedentary by proprietary Fitbit algorithms, which have been shown to be valid for the Fitbit One™ among younger adults (Diaz et al., 2015). Total PA, rather than just moderate-to-vigorous PA, was chosen as the outcome metric for two reasons. First, fee Otago exercise protocol includes four types of light and moderately intense activities that can be practiced in structured or free-living contexts and can be adapted according to an individual’s abilites (Gardner et al., 2001). Second, many older adults prefer light over moderate or vigorous activities, and there is growing evidence that even light-intensity activities improve the health and well-being of older adults (Buman et al., 2010). Days were considered invalid and not included in the analysis if (a) participants reported that they did not wear the PA monitor for at least eight hours or (b) screening of weekly data revealed that data were only available for fewer than eight hours. A minimum of three days of wear time was required to estimate the weekly averages of PA at all data-collection time points (Hart, Swartz, Cashin, & Strath, 2011). On average, participants wore their Fitbits for 13 hours per day over six consecutive days during all data-collection time points.

Data Analysis

Statistical analyses were performed using Stata, version 13.1 (StataCorp, 2013). The analysis was comprised of three main steps. In the first step, we calculated frequencies, means, and standard deviations to describe the distributions, central tendencies, and variability in the demographic and clinical variables at the baseline, as well as the psychosocial and PA variables at the baseline, post intervention, and six-month follow-up.

In the second step, we performed an analysis of covariance (ANCOVA) to assess the effects of the interventions with interpersonal behavior change strategies and those with intrapersonal behavior change strategies on the potential psychosocial mediators targeted in this study: social support (i.e., friends and family), readiness, self-efficacy, self-regulation (i.e., benefits, self-assessment, and integration) and goal attainment. We controlled for the baseline values of each, as well as age and sex.

In the third step, we conducted mediation analyses for each potential psychosocial mediator that significantly increased post intervention or at the six-month follow-up. To do this, we used the causal mediation approach outlined by Valeri and VanderWeele (2013) via the Paramed command in Stata (Emsley & Liu, 2013), which fitted a linear regression model to the outcome of the treatment (i.e., intervention) and each mediator, which were included as covariates. Then, the natural direct and indirect effects were calculated based on the coefficients of these models and the standard errors and confidence intervals were calculated using bias-corrected bootstrapping with 1,000 replications. Confidence intervals that did not include zero were considered statistically significant, suggesting a mediation effect.

We used the causal mediation approach because it addresses four critical assumptions regarding mediation: (a) there is no unmeasured confounding in the treatment-outcome relationship, (b) there is no unmeasured confounding in the mediator-outcome relationship, (c) there is no unmeasured confounding in the treatment-mediator relationship, and (d) there is no mediator-outcome confounder affected by the treatment (Valeri & VanderWeele, 2013). Although randomization was controlled for assumptions (a) and (c), assumption (b) was addressed by including potential confounders of the mediator-outcome relationship as covariates in our adjusted models, including the baseline values of each mediator and outcome as well as the age and sex. Finally, although intervention-mediator interactions were not statistically significant in any of the models, we included the interaction term in all adjusted models to avoid drawing incorrect conclusions and to understand the dynamics of mediation (VanderWeele, 2015).

Results

Participants were principally white (75%) women (75%) who lived alone (53%) and had at least some college education (50%). Twenty-five percent of participants had at least one fall in the last year, and 90% reported living with at least one chronic condition. Differences in participants’ baseline demographic, health, and PA characteristics were not statistically significant across the study conditions (McMahon et al., 2017). The participants averaged 1,146 (SD = 484) minutes of total PA (of light, moderate, and vigorous intensity) per week at the baseline, 1,282 (SD = 598) one week post intervention, and 1,284 (SD = 616) six months post intervention. Ninety-seven (95%) participants completed post-intervention assessments, 95 (93%) participants completed the six-month follow-up assessment, and 91 (89%) participated in data collection at all three assessment time points. There were no significant differences in the baseline characteristics between participants with complete and incomplete data (McMahon et al., 2017). Participants attended, on average, 7.20 (SD = 1.40) of the 8 intervention sessions.

The means and standard deviations of self-rated social support, readiness, self-efficacy, self-regulation, and goal attainment at the baseline, post-intervention time point, and six-month follow-up are detailed by condition in Table 2. Participants’ ratings of social support for PA from friends and family were low, with averages ranging from 1.51 (SD = 0.54) to 2.16 (SD = 1.01) across all conditions and time points. In addition, goal attainment ratings were low at the baseline across conditions, with averages ranging from 35.54 (SD =6.50) to 38.26 (SD = 6.27). This indicated that, as expected, participants were not attaining their new goals before the intervention started. Participants’ self-rated levels of confidence in engaging in PA when faced with common barriers (i.e., self-efficacy for exercise) were moderate across conditions and time points, with averages ranging from 54.82 (SD = 16.87) to 66.36 (SD = 18.43).

Table 2.

Unadjusted mean values of potential psychosocial mediators by study condition.

Baseline (n = 101) Post-Intervention (n = 98) 6-Month Follow Up (n = 95)
Condition Mean SD Mean SD Mean SD
1. Otago + Physical Activity Monitor (n = 25)
 Social support (family)a 1.73 0.67 1.84 0.92 1.91 0.96
 Social support (friends)a 1.84 0.93 1.68 0.62 1.71 0.67
 Readinessa 3.69 0.73 3.96 0.72 3.96 0.96
 Self-efficacyb 66.36 18.43 64 17.66 57.42 25.24
 Self-regulation (benefits)a 3.97 0.82 4.11 0.89 4.26 0.78
 Self-regulation (Self-assessment)a 2.72 1.10 3.31 1.15 3.06 1.24
 Self-regulation (integration)a 2.97 1.02 3.76 0.77 3.58 1.06
 Goal attainmentc 37.78 5.78 49.62 4.79 48.94 9.63
2. Otago + Physical Activity Monitor + Interpersonal Behavior Change Strategies (n = 25)
 Social support (family)a 1.96 .92 1.88 .69 1.82 .89
 Social support (friends)a 1.74 .83 1.71 .66 1.84 .91
 Readinessa 3.83 .99 4.19 .65 4.05 .81
 Self-efficacyb 63.82 17.52 62.36 17.21 60.15 21.55
 Self-regulation (benefits)a 3.83 .78 4.23 .67 3.97 .79
 Self-regulation (Self-assessment)a 2.43 1.14 3.57 .92 3.21 .95
 Self-regulation (integration)a 3.24 1.10 3.81 .81 3.67 .77
 Goal attainmentc 38.26 6.27 52.45 10.26 56.50 10.00
3. Otago + Physical Activity Monitor + Intrapersonal Behavior Change Strategies (n = 25)
 Social support (family)a 2.08 0.97 2.36 0.83 2.09 0.72
 Social support (friends)a 1.67 0.82 1.72 0.69 1.70 0.61
 Readinessa 3.45 0.79 4.03 0.50 3.76 0.64
 Self-efficacyb 59.96 17.57 57.69 17.74 54.82 16.87
 Self-regulation (benefits)a 4.04 0.64 4.01 0.76 4.16 0.47
 Self-regulation (Self-assessment)a 2.25 1.02 3.26 0.98 2.95 1.01
 Self-regulation (integration)a 2.28 1.10 3.52 0.93 3.29 0.98
 Goal attainmentc 37.56 5.92 47.15 8.66 43.92 7.53
4. Otago + Physical Activity Monitor + Inter- and Intrapersonal Behavior Change Strategies (n = 26)
 Social support (family)a 1.88 .93 2.16 1.01 1.96 1.07
 Social support (friends)a 1.51 0.54 1.78 0.78 1.80 0.82
 Readinessa 3.89 0.97 4.40 0.53 4.15 0.85
 Self-efficacyb 67.49 20.22 69.23 19.93 66.49 16.31
 Self-regulation (benefits)a 4.13 0.72 4.36 0.89 4.37 0.57
 Self-regulation (Self-assessment)a 2.79 1.24 4.15 0.82 3.81 0.82
 Self-regulation (integration)a 3.01 1.27 4.18 0.74 3.63 1.06
 Goal attainmentc 35.54 6.50 52.62 9.47 52.62 10.87
Open in a new tab

Note: Interpersonal = interpersonal behavior change strategies; intrapersonal = intrapersonal behavior change strategies; Otago = evidence-based fall-reducing physical activity protocol.

a =

the mean scores of multiple items with response scales ranging from 1–5.

b =

the mean scores of multiple items with response scales ranging from 0–100.

c =

mean weighted goal attainment scores ranging from 30–70.

Mediation analyses

Interventions with interpersonal behavior change strategies

Effects on potential psychosocial mediators.

Participants who received interventions with interpersonal strategies, compared to those who did not, had significantly higher ratings of readiness (B [SE] = 0.29 [0.12], = 0.01), self-assessment (a subscale of self-regulation; B [SE] = 0.58 [0.19], p = 0.003), and goal attainment (B [SE] = 0.32 [0.16], p = 0.05) immediately post intervention (see Table 3). At the six-month follow-up, they had higher ratings of self-assessment (a subscale of self-regulation; B [SE] = 0.48 [0.20], p = 0.02) and goal attainment (B [SE] = 0.79 [0.19], p = 0.001; see Table 3).

Table 3.

Effects of interpersonal and intrapersonal behavior change strategies on potential psychosocial mediators based on ANCOVA.

Post Intervention (n = 98) Six-month Follow-Up (n = 91)
B SE 95% CI B SE 95% CI
Social Support (Friends)a
 Receipt of interpersonal BCS 0.33 0.08 −1.74, 2.41 0.09 0.14 −0.37, 0.19
 Receipt of intrapersonal BCS 0.23 0.12 −0.46, 00 0.16 0.14 −0.45, 0.13
Social Support (Family)a
 Receipt of interpersonal BCS −0.11 0.14 −0.16,0.39 −0.17 0.16 −0.16, 0.49
 Receipt of intrapersonal BCS 0.35 0.14 0.08, 0.62 0.01 0.16 −0.31, 0.32
Self-efficacyb
 Receipt of interpersonal BCS 3.23 3.04 −9.26, 2.80 5.31 3.51 −12.28,1.67
Receipt of intrapersonal BCS 1.69 3.04 −7.73, 4.34 2.85 3.52 −9.84,4.15
Readinessa
 Receipt of interpersonal BCS 0.29 0.12 0.06, 0.53 0.17 0.16 −0.48, 0.15
 Receipt of intrapersonal BCS 0.13 0.12 −0.37,0.11 −0.02 0.16 −0.30, 0.33
Self-regulation (Benefits)a
 Receipt of interpersonal BCS 0.26 0.16 −0.07, 0.50 0.01 0.13 −0.27, 0.25
 Receipt of intrapersonal BCS −0.03 0.16 −0.28, 0.35 0.08 0.13 −0.18, 0.34
Self-regulation (Self-assessment)a
 Receipt of interpersonal BCS 0.58 0.19 0.20, 0.96 0.48 0.20 0.07, 0.88
 Receipt of intrapersonal BCS 0.30 0.20 −0.09, 0.69 0.27 0.21 −0.14, 0.68
Self-regulation (Integration)a
 Receipt of interpersonal BCS 0.24 0.16 −0.07, 0.55 −0.05 0.18 −0.41, 0.30
 Receipt of intrapersonal BCS 0.18 0.16 −0.12,0.50 0.03 0.18 −0.32, 0.37
Goal Attainmentc
 Receipt of interpersonal BCS 4.21 1.72 0.76, 7.66 9.03 2.01 5.04, 13.02
 Receipt of intrapersonal BCS −0.31 12.07 −3.87, 3.25 −2.87 2.21 −7.26, 1.53
Open in a new tab

Note: 95% CIs that do not include zero are bolded. The covariates in the models are age, sex, and the baseline values of the potential psychosocial mediators. BCS = behavior change strategies.

a =

the mean scores of multiple items with response scales ranging from 1–5.

b =

the mean scores of multiple items with response scales ranging from 0–100.

c =

mean weighted goal attainment scores ranging from 30–70.

Indirect effects on physical activity through potential psychosocial mediators engaged by the intervention.

The results of the unadjusted and adjusted mediation analyses are presented in Table 4. Adjusted analyses of the indirect effects (i.e., mediation) of the interventions on PA based on the self-regulation subscale of self-assessment were positive, but not statistically significant, post intervention (B [SE] = 8.10 (35.80), 95% CI [−56.17, 81.65) and at the six-month follow-up (B [SE] = 11.17 [34.40], 95% CI [−34.3, 115.98]). Similarly, the adjusted indirect effects (i.e., mediation) of goal attainment were positive but not statistically significant at the six-month follow-up (B [SE] = 32.78 [37.04], 95% CI [−36.51, 110.08]).

Table 4.

Direct, indirect, and total effects of interventions with interpersonal behavior change strategies on physical activity.

Natural Direct Effect Natural Indirect Effect Total Effect
Potential Mediator Model Specifications
Effect Boot (SE) Boot 95% CI Effect Boot (SE) Boot 95% CI Effect Boot (SE) Boot 95% CI
Readiness for Physical Activity Post Intervention
 Unadjusted 235.49 126.05 −40.24, 452.22 52.87 37.58 .66, 163.20 288.36 118.51 39.13, 504.31
No interaction
 Unadjusted 258.24 122.64 10.70, 495,45 30.61 52.25 −40.12, 183.61 288.84 119.08 39.71, 506.15
Interaction
 Adjusted 211.69 71.02 64.65, 347.31 −11.81 24.71 −72.44, 30.21 199.87 68.31 59.52, 326.27
Interaction
Self-regulation of Physical Activity (Self-assessment Subscale Post Intervention)
 Unadjusted 221.73 123.86 −35.63, 447.04 68.51 42.81 7.20, 179.81 292.75 118.51 39.71, 509.14
No interaction
 Unadjusted 219.60 118.08 13.67, 472.26 70.56 65.31 −29.38, 243.42 290.16 113.51 80.06, 512.90
Interaction
 Adjusted 181.40 70.85 51.43, 327.95 8.10 35.80 −56.17,81.65 189.50 66.38 72.23, 330.14
Interaction
Self-regulation of Physical Activity (Self-assessment Subscale at 6-month Follow-up)
 Unadjusted 177.41 126.94 −52.17, 435.34 21.40 28.52 −14.37,119.80 198.91 125.06 −31.67,455.96
No interaction
 Unadjusted 109.39 131.29 −163.12, 361.96 81.30 60.26 10.44,292.67 190.69 126.07 −53.30,441.83
Interaction
 Adjusted 130.94 88.60 −46.25, 301.83 11.17 34.40 −34.31,115.98 142.10 84.30 −19.92,328.18
Interaction
Goal Attainment (Physical Activity-related Goals Post Intervention)
 Unadjusted 332.95 121.49 77.50, 555.65 −49.21 31.64 −149.77, −5,39 281.81 118.51 21.64, 485.31
No interaction
 Unadjusted 361.71 124.53 109.52, 600.96 −75.55 47.04 −228.22, −11.78 286.16 118.51 37.89, 489.72
Interaction
 Adjusted 201.57 70.72 64.06, 341.65 −13.11 21.04 −81.76, 14.88 188.46 66.03 55.17, 311.75
Interaction
Goal Attainment (Physical Activity-related Goals at 6-month Follow-up)
 Unadjusted 170.32 145.25 −108.48, 499.13 30.12 61.03 −108.48, 449.13 200.45 129.76 −53.88, 454.78
No interaction
 Unadjusted 186.24 148.24 −85.72, 486.24 3.11 55.74 −115.47, 113.57 189.35 127.30 −46.93, 444.19
Interaction
 Adjusted 102.88 92.83 −56.52, 319.86 32.78 37.04 −36.51, 110.08 135.77 82.14 −11.51, 318.36
Interaction
Open in a new tab

Note: 95% CIs that do not include zero are bolded. The covariates in the models are age, sex, and the baseline values of the potential psychosocial mediators.

BCS = behavior change strategies. Interaction = the interaction term of receipt of interpersonal BCS (No/Yes) * the mediator that was included in the model.

Interventions with intrapersonal behavior change strategies

Effects on potential psychosocial mediators.

Participants who received interventions with intrapersonal strategies, compared to those who did not, had significantly higher ratings of social support from family (B [SE] = 0.35 [0.14], p = 0 .02) immediately post intervention (see Table 3).

Indirect effects on physical activity through potential psychosocial mediators engaged by the intervention.

The results of the unadjusted and adjusted mediation analyses were not statistically significant, as shown in Table 5.

Table 5.

Direct, indirect, and total effects of interventions with intrapersonal behavior change strategies on physical activity.

Potential Mediator Model Specifications Natural Direict Effect Natural Indirect Effect Total Effect
Effect Boot (SE) Boot 95% CI Effect Boot (SE) Boot 95% CI Effect Boot (SE) Boot 95% CI
Social Support from Family Post Intervention
 Unadjusted −127.99 121.82 −385.17, 87.53 22.07 37.22 −22.05, 156.89 −105.92 113.25 −335.57, 100.64
No interaction
 Unadjusted −113.37 129.46 −367.11, 140.38 7.11 40.30 −71.87, 86.10 −106.25 124.51 −350.30, 137.79
Interaction
 Adjusted 70.14 75.27 −76.70, 218.34 −53.63 30.98 −114.36, 7.10 17.19 72.07 −124,07, 158.44
Interaction
Open in a new tab

Note: 95% CIs that do not include zero are bolded. The covariates in the models are age, sex, and the baseline values of the potential psychosocial mediators.

BCS = behavior change strategies. Interaction = the interaction term of receipt of interpersonal BCS (No/Yes) * the mediator that was included in the model.

Discussion

This is one of the first randomized PA intervention studies for older adults that thoroughly examines the links between the constructs targeted as potential psychosocial mediators, the tested interventions, and PA outcomes. We observed four main findings. First, on average, participants rated the social support of friends and family for PA as very low across all conditions and time points. Second, participants who received interventions that included interpersonal behavior change strategies provided higher ratings of their readiness, self-assessment (a subscale of self-regulation), and goal attainment compared to those who did not receive these interventions. Third, the mediation effects of self-assessment (a subscale of self-regulation) post intervention and at the six-month follow-up, as well as goal attainment at the six-month follow-up, were positive but not statistically significant. Fourth, participants who received interventions with intrapersonal behavior change strategies provided significantly higher ratings of social support for PA from family post intervention, compared to those who did not receive these interventions. However, the increases in ratings did not lead to positive or significant mediation effects.

The first finding of this study is that many participants reported rarely or never receiving social support for PA from family and friends before or after the intervention. Although this was unexpected, reports from prior studies that assessed older adults’ perceived level of social support using the Social Support and Exercise Survey (Sallis et al., 1987) also found low average ratings: 1.7–2.8 out of 5 (McMahon et al., 2016; Oka & Shibata, 2012; Park, Elavsky, & Koo, 2014) and one small group with an average rating of 3.2 post-intervention (Yeom & Fleury, 2014). Given that ratings of 1 represent no support, ratings of 2 represent rare support, and ratings of 3 represent occasional support, our finding indicates the challenge facing intervention strategies designed to elicit increased support for activity from an older adults’ family or friends. Moreover, this observation is consistent with the results of a recent meta-analysis showing that PA interventions for older adults were unable to elicit significant increases in social support (Shvedko, Whittaker, Thompson, & Greig, 2018).

One possible explanation for the low ratings of social support for PA is that the assessment tool used in this study, although valid (Sallis et al., 1987), may not capture the sources of support that older adults perceive as helpful for staying physically active, such as peers, neighbors, and members of organizations to which they belong (e.g., church and exercise programs). The future of older adults’ PA maintenance may rely on small groups; rather than this support coming from family and friends, it may be more likely for this support to come from like-minded peers and acquaintances who share a common goal. Moreover, given the positive relationship between social support and PA (Lindsay Smith, Banting, Eime, O’Sullivan, & van Uffelen, 2017), it is important to gain a better understanding of what constitutes meaningful social support for PA among older adults and investigate strategies for promoting this while examining the links between intervention strategies, the construct of social support, and PA outcomes.

The second finding of this study suggests that, for older adults, interpersonal behavior change strategies may be an innovative way to target readiness, self-regulation, and goal-attainment, which are psychosocial constructs that are typically targeted by intrapersonal behavior change strategies either alone or in combination with intrapersonal strategies. Previous research shows that interventions that combine interpersonal and intrapersonal strategies lead to increases in older adults’ readiness, intention, self-regulation, and self-efficacy (McMahon et al., 2016; Olson & McAuley, 2015; Yeom & Fleury, 2014). The current study is the first to find that these increases are explicitly tied to the inclusion of interpersonal strategies in interventions. However, the specific types of interpersonal strategies that lead to these increases are still not well understood. Those used in this study, which targeted readiness, self-regulation, and goal attainment, included peer-to-peer discussions about increasing social support for PA, including PA in social routines, and participants’ experiences performing and increasing PA. Future research must reproduce these findings and identify which interpersonal behavior change strategies, or which combination of strategies, are most effective when used in PA interventions.

Our third finding is that participants’ higher ratings of readiness, self-assessment, and goal attainment after receiving interventions with interpersonal behavior change strategies did not lead to statistically significant mediation effects. However, the mediation effects of self-assessment and goal attainment did show positive trends. Prior studies addressing self-regulation as a psychosocial mechanism of change in older adults’ PA reported mixed results. For example, one conclusion from a recent meta-analysis examining the associations between behavior change techniques and older adults’ PA is that techniques that presumably target self-regulation may not be effective in older adults (French et al., 2014). Conversely, the results of Olson and McAuley’s (2015) recent study suggest that interventions with both interpersonal and intrapersonal strategies increased self-regulation among older adults and, in turn, the increased self-regulation predicted increased PA. These results are interesting, given the strong evidence that self-regulation is an important mechanism of action among younger populations (Rhodes & Pfaeffli, 2010; Warner, Wolff, Ziegelmann, Schwarzer, & Wurm, 2016). In sum, our findings indicate that self-assessment and goal attainment may be important mechanisms leading to increased PA among older adults when targeted by interventions with interpersonal behavior change strategies.

It is possible that higher doses of interpersonal strategies (e.g., more frequent implementation and longer duration) were needed to elicit larger effects on the targeted psychosocial constructs in this study. The current literature that addresses PA among older adults does not describe minimal dosing requirements for individual behavior change strategies or the minimum levels at which the targeted psychosocial constructs become clinically meaningful. Thus, it is worthwhile for researchers to investigate the relationships between incrementally varied doses of behavior change strategies and the responses they elicit in potential psychosocial constructs.

The fourth finding in this study—that descriptions and tests of the effect of intrapersonal behavior change strategies on potential psychosocial mediators produced mainly null results—helps to explain why participants exposed to these strategies did not significantly increase their PA (McMahon et al., 2017). Although interventions with intrapersonal behavior change strategies (e.g., setting personally meaningful goals) elicited significant increases in social support from family for exercise, participants rated the frequency of such support as rare. The comparative literature in this area is scarce. In one study, Warner et al. (2016) tested an intervention based on the health action process approach that included intrapersonal strategies targeting several facets of self-regulation, similar to those used in this study (e.g., goal setting and action planning), and found that older adults did not increase their PA. It remains unclear why the intrapersonal behavior change strategies in our study did not elicit changes in most of the targeted psychosocial constructs. One possible explanation is that older adults, particularly those over the age of 70 years, might prefer interpersonal behavior change strategies over intrapersonal strategies within PA interventions (Devereux-Fitzgerald, Powell, Dewhurst, & French, 2016). Older adults most likely have some life experiences during which they developed goals according to their personal priorities, abilities, and available resources. Thus, they may believe that time spent formally creating personal goals and documenting PA-related behavior is less relevant to their performance of recommended PAs than time spent informally considering their personal goals while exchanging ideas and experiences with peers about their performance, progress, and outcomes. Future research is needed to explore which intrapersonal behavior change strategies are relevant and elicit changes in the key psychosocial determinants of older adults’ behavior.

The strengths of this study include its methods of outcome assessment and design. PA was measured objectively to overcome the biases associated with self-reports, such as recall and social desirability, thus enabling more accurate mediation analyses (Ainsworth, Cahalin, Buman, & Ross, 2015). Another strength was the factorial study design, which enabled attribution of differences between factor levels (e.g., receipt versus non-receipt of interpersonal behavior change strategies) to the relative effects of each distinct set of behavior change strategies on the potential psychosocial mediators. This, in turn, enabled identification of the active behavior change strategy set(s) within the intervention and examination of how each set does or does not work (Collins, Dziak, Kugler, & Trail, 2014). Thus, the study contributes to the evidence base needed to develop more efficacious, translatable interventions.

This study also has several limitations. First, the instruments used to assess the targeted psychosocial constructs may not be adequate. For example, we used a valid measure of perceived self-efficacy for overcoming common barriers to perform the recommended exercises, but participants may have experienced changes in other facets of this construct, such as walking self-efficacy. Additional refinement of self-efficacy into capability (e.g., “can-do”) and motivation constructs (e.g., “will do”), might be warranted, considering recent questions regarding the many factors that drive responses to self-efficacy questions (Williams & Rhodes, 2016). Second, we used a small sample size, which limited the precision with which we were able to estimate mediation effects. The sample size was calculated based on assumptions about potential changes in PA outcomes and the psychosocial constructs targeted as potential psychosocial mediators, not the indirect effects of the intervention on PA through the potential mediators. To advance intervention development, it is essential for researchers to critically analyze the relevance and sensitivity of the measures used to assess all the targeted psychosocial constructs and to include mediation effects when estimating the sample size needed for future studies (Fritz & Mackinnon, 2007).

In sum, an important strategy for addressing the public health problem of low activity levels among older adults is to build an evidence base that describes the psychosocial mechanisms of action through which interventions increase older adults’ PA. To contribute to this body of accumulating evidence, we describe and test the constructs targeted as potential psychosocial mediators in a recent intervention study designed to increase older adults’ PA (McMahon et al., 2017). Our findings highlight that social support for PA among participants was very low across conditions and time points. In addition, older adults’ self-assessment (a subscale of self-regulation) and goal attainment may be modified by interpersonal behavior change strategies, which are an innovative approach to elicit these facets of self-regulation in older adults. Moreover, self-assessment and goal attainment may be important psychosocial mediators of older adults’ PA. Together, these findings underscore the need for further research to understand the complexities of older adults’ motivation for increased and sustained PA. Future intervention research should thoroughly and systematically examine the links between the tested interventions, the potential psychosocial mediators they target, and the PA outcomes (Carey et al., 2019). The accumulation of such evidence will accelerate the design and development of interventions that precisely target relevant psychosocial mechanisms of action through which older adults can increase and maintain their PA (Nielsen et al., 2018), increasing the effects of public health efforts to promote PA among older adults.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the University of Minnesota Clinical Translational Science Institute (grant numbers KL2TR000113 and UL1TR000114).

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

  1. Ainsworth B, Cahalin L, Buman M, & Ross R (2015). The current state of physical activity assessment tools. Progress in Cardiovascular Diseases, 57(4), 387–395. doi: 10.1093/aje/kwq249 [DOI] [PubMed] [Google Scholar]
  2. Barrera M, Toobert DJ, Angell KL, Glasgow RE, & Mackinnon DP (2006). Social support and social-ecological resources as mediators of lifestyle intervention effects for type 2 diabetes. Journal of Health Psychology, 11(3), 483–495. doi: 10.1177/1359105306063321 [DOI] [PubMed] [Google Scholar]
  3. Bauman A, Merom D, Bull FC, Buchner DM, & Fiatarone Singh MA (2016). Updating the evidence for physical activity: Summative reviews of the epidemiological evidence, prevalence, and interventions to promote “active aging.” The Gerontologist, 56(Suppl. 2), S268–S280. doi: 10.1093/geront/gnw031 [DOI] [PubMed] [Google Scholar]
  4. Becofsky K, Baruth M, & Wilcox S (2014). Psychosocial mediators of two community-based physical activity programs. Annals of Behavioral Medicine, 48(1), 125–129. doi: 10.1007/S12160-013-9578-3 [DOI] [PubMed] [Google Scholar]
  5. Buman MP, Hekler EB, Haskell WL, Pruitt L, Conway TL, Cain KL, … King AC (2010). Objective light-intensity physical activity associations with rated health in older adults. American Journal of Epidemiology, 172(10), 1155–1165. doi: 10.1093/aje/kwq249 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cain KL, & Geremia CM (2011). Accelerometer data collection and scoring manual for adult & senior studies. San Diego, CA: San Diego State University. [Google Scholar]
  7. Carey RN, Connell LE, Johnston M, Rothman AJ, de Bruin M, Kelly MP, & Michie S (2019). Behavior change techniques and their mechanisms of action: A synthesis of links described in published intervention literature. Annals of Behavioral Medicine, 53(8), 693–707. doi: 10.1093/abm/kay078 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Case MA, Burwick HA, Volpp KG, & Patel MS (2015). Accuracy of smartphone applications and wearable devices for tracking physical activity data. JAMA, 313(6), 625–626. doi: 10.1001/jama.2014.17841 [DOI] [PubMed] [Google Scholar]
  9. Chase JA (2014). Interventions to increase physical activity among older adults: A meta-analysis. The Gerontologist, 55(4), 706–718. doi: 10.1093/geront/gnu090 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Collins LM, Dziak JJ, Kugler KC, & Trail JB (2014). Factorial experiments: efficient tools for evaluation of intervention components. American Journal of Preventive Medicine, 47(4), 498–504. doi: 10.1016/j.amepre.2014.06.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Connell LE, Carey RN, de Bruin M, Rothman AJ, Johnston M, Kelly MP, & Michie S (2019). Links between behavior change techniques and mechanisms of action: An expert consensus study. Annals of Behavioral Medicine, 53(8), 708–720. doi: 10.1093/abm/kay082 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Devereux-Fitzgerald A, Powell R, Dewhurst A, & French DP (2016). The acceptability of physical activity interventions to older adults: A systematic review and meta-synthesis. Social Science & Medicine, 158, 14–23. doi: 10.1016/j.socs-cimed.2016.04.006 [DOI] [PubMed] [Google Scholar]
  13. Diaz KM, Krupka DJ, Chang MJ, Peacock J, Ma Y, Goldsmith J, … Davidson KW (2015). Fitbit®: An accurate and reliable device for wireless physical activity tracking. International Journal of Cardiology, 185, 138–140. doi: 10.1016/j.ijcard.2015.03.038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Emsley R, & Liu H (2013). PARAMED: Stata module to perform causal mediation analysis using parametric regression models. Retrieved from https://econpapers.repec.org/software/bocbocode/s457581.htm
  15. Fitabase. (2015). Research platform. Retrieved from http://www.fitabase.com/
  16. Fleury J (1994). The index of readiness: Development and psychometric analysis. Journal of Nursing Measurement, 2(2), 143–154. [PubMed] [Google Scholar]
  17. Fleury J (1998). The index of self-regulation: Development and psychometric analysis. Journal of Nursing Measurement, 6(1), 3–17. [PubMed] [Google Scholar]
  18. Floegel T, Florez-Pregonero A, Hekler E, & Buman MP (2014). Simultaneous validation of five commercially-available activity monitors in older adults with varied ambulatory abilities. The Gerontologist, 54,187. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. French DP, Olander EK, Chisholm A, & McSharry J (2014). Which behaviour change techniques are most effective at increasing older adults’ self-efficacy and physical activity behaviour? A systematic review. Annals of Behavioral Medicine, 48(2), 225–234. doi: 10.1007/s12160-014-9593-z [DOI] [PubMed] [Google Scholar]
  20. Fritz MS, & Mackinnon DP (2007). Required sample size to detect the mediated effect. Psychological Science, 18(3), 233–239. doi: 10.1111/j.1467-9280.2007.01882.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gardner MM, Buchner DM, Robertson MC, & Campbell AJ (2001). Practical implementation of an exercise-based falls prevention programme. Age and Ageing, 30(1), 77–83. doi: 10.1093/ageing/30.1.77 [DOI] [PubMed] [Google Scholar]
  22. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, & Conde JG (2009). Research electronic data capture (REDCap)—A metadata-driven methodology and workflow process for providing translational research informatics support. Journal of Biomedical Informatics, 42(2), 377–381. doi: 10.1016/j.jbi.2008.08.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hart TL, Swartz AM, Cashin SE, & Strath SJ (2011). How many days of monitoring predict physical activity and sedentary behaviour in older adults. International Journal of Behavioral Nutrition and Physical Activity, 8(62). doi: 10.1186/1479-5868-8-62 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Keller C, Fleury J, Sidani S, & Ainsworth B (2009). Fidelity to theory in PA intervention research. Western Journal of Nursing Research, 31(3), 289–311. doi: 10.1177/0193945908326067 [DOI] [PubMed] [Google Scholar]
  25. Kiresuk TJ, & Sherman RE (1968). Goal attainment scaling: A general method for evaluating comprehensive community mental health programs. Community Mental Health Journal, 4(6), 443–453. [DOI] [PubMed] [Google Scholar]
  26. Kiresuk TJ, Smith A, & Cardillo JE (2014). Goal attainment scaling: Applications, theory, and measurement. New York, NY: Psychology Press, 10.4324/9781315801933 [DOI] [Google Scholar]
  27. Kyrdalen IL, Moen K, Røysland AS, & Helbostad JL (2013). The Otago exercise program performed as group training versus home training in fall-prone older people: A randomized controlled trial. Physiotherapy Research International, 19(2), 108–116. doi: 10.1002/pri.1571 [DOI] [PubMed] [Google Scholar]
  28. Lindsay Smith G, Banting L, Eime R, O’Sullivan G, & van Uffelen JGZ (2017). The association between social support and physical activity in older adults: A systematic review. International Journal of Behavioral Nutrition and Physical Activity, 14(1), 1–21. doi: 10.1186/s12966-017-0509-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. McAuley E (1992). The role of efficacy cognitions in the prediction of exercise behavior in middle-aged adults. Journal of Behavioral Medicine, 15(1), 65–88. [DOI] [PubMed] [Google Scholar]
  30. McMahon SK, Lewis B, Oakes JM, Wyman JF, Guan W, & Rothman AJ (2017). Assessing the effects of interpersonal and intrapersonal behavior change strategies on physical activity in older adults: A factorial experiment. Annals of Behavioral Medicine, 51(3), 376–390. doi: 10.1007/s12160-016-9863-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. McMahon SK, Wyman JF, Belyea MJ, Shearer N, Hekler EB, & Fleury J (2016). Combining motivational and physical intervention components to promote fall-reducing physical activity among community-dwelling older adults: A feasibility study. American Journal of Health Promotion, 30(8), 638–644. doi: 10.4278/ajhp.130522-ARB-265 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Moore SM, Schiffman R, Waldrop-Valverde D, Redeker NS, McCloskey DJ, Kim MT, … Grady P (2016). Recommendations of common data elements to advance the science of self-management of chronic conditions. Journal of Nursing Scholarship, 48(5), 437–447. doi: 10.1111/jnu.12233 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Newkirk LA, Kim JM, Thompson JM, Tinklenberg JR, Yesavage JA, & Taylor JL (2004). Validation of a 26-point telephone version of the Mini-Mental State Examination. Journal of Geriatric Psychiatry and Neurology, 17(2), 81–87. doi: 10.1177/0891988704264534 [DOI] [PubMed] [Google Scholar]
  34. Nielsen L, Riddle M, King JW, Aklin WM, Chen W, Clark D, … Green P (2018). The NIH Science of Behavior Change Program: Transforming the science through a focus on mechanisms of change. Behaviour Research and Therapy, 101, 3–11. doi: 10.1016/j.brat.2017.07.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Oka K, & Shibata A (2012). Determinants of meeting the public health recommendations for physical activity among community-dwelling elderly Japanese. Current Aging Science, 5(1), 58–65. doi: 10.2174/1874609811205010058 [DOI] [PubMed] [Google Scholar]
  36. Olson EA, & McAuley E (2015). Impact of a brief intervention on self-regulation, self-efficacy and physical activity in older adults with type 2 diabetes. Journal of Behavioral Medicine, 38(6), 886–898. doi: 10.1007/s10865-015-9660-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Park CH, Elavsky S, & Koo KM (2014). Factors influencing physical activity in older adults. Journal of Exercise Rehabilitation, 10(1), 45–52. doi: 10.12965/jer.l40089 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Perez A, & Fleury J (2009). Wellness motivation theory in practice. Geriatric nursing; (New York, NY: ), 30(2 Suppl), 15. doi: 10.1016/j.gerinurse.2009.02.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Physical Activity Guidelines Advisory Committee. (2018). 2018 Physical Activity Guidelines Advisory Committee scientific report. Washington, DC: U.S.Department of Health and Human Services; Retrieved from https://health.gov/paguidelines/second-edition/report.aspx [Google Scholar]
  40. Rhodes RE, & Pfaeffli LA (2010). Review mediators of physical activity behaviour change among adult non-clinical populations: A review update. International Journal of Behavioral Nutrition and Physical Activity, 7(1). 37–48. doi: 10.1186/1479-5868-7-37 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Rockwood K, Howlett S, Stadnyk K, Carver D, Powell C, & Stolee P (2003). Responsiveness of goal attainment scaling in a randomized controlled trial of comprehensive geriatric assessment. Journal of Clinical Epidemiology, 56(8), 736–743. doi: 10.1016/S0895-4356(03)00132-X [DOI] [PubMed] [Google Scholar]
  42. Sallis JF, Grossman RM, Pinski RB, Patterson TL, & Nader PR (1987). The development of scales to measure social support for diet and exercise behaviors. Preventive Medicine, 16(6), 825–836. doi: 10.1016/0091-7435(87)90022-3 [DOI] [PubMed] [Google Scholar]
  43. Sheeran P, Klein WM, & Rothman AJ (2017). Health behavior change: Moving from observation to intervention. Annual review of psychology,68, 573–600. doi: 10.1146/annurev-psych-010416-044007 [DOI] [PubMed] [Google Scholar]
  44. Shubert TE, Smith ML, Goto L, Jiang L, & Ory MG (2017). Otago exercise program in the United States: Comparison of 2 implementation models. Physical Therapy, 97(2), 187–197. doi: 10.2522/ptj.20160236 [DOI] [PubMed] [Google Scholar]
  45. Shvedko A, Whittaker AC, Thompson JL, & Greig CA (2018). Physical activity interventions for treatment of social isolation, loneliness or low social support in older adults: A systematic review and meta-analysis of randomised controlled trials. Psychology of Sport and Exercise, 34, 128–137. doi: 10.1016/j.psychsport.2017.10.003 [DOI] [Google Scholar]
  46. Silva-Smith AL, Fleury J, & Belyea M (2013). Effects of a physical activity and healthy eating intervention to reduce stroke risk factors in older adults. Preventive Medicine, 57(5), 708–711. doi: 10.1016/j.ypmed.2013.07.004 [DOI] [PubMed] [Google Scholar]
  47. StataCorp. (2013). Stata statistical software: Release 13 [computer software]. College Station, TX: StataCorp LP. [Google Scholar]
  48. Stevens JA, & Phelan EA (2013). Development of STEADI: A fall prevention resource for health care providers. Health Promotion Practice, 14(5), 706–714. doi: 10.1177/1524839912463576 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Topolski TD, LoGerfo J, Patrick DL, Williams B, Walwick J, & Patrick MB (2006). The Rapid Assessment of Physical Activity (RAPA) among older adults. Preventing Chronic Disease, 3(4), A118. [PMC free article] [PubMed] [Google Scholar]
  50. Toto PE, Skidmore ER, Terhorst L, Rosen J, & Weiner DK (2015). Goal Attainment Scaling (GAS) in geriatric primary care: A feasibility study. Archives of Gerontology and Geriatrics, 60(1), 16–21. doi: 10.1016/j.archger.2014.10.022 [DOI] [PubMed] [Google Scholar]
  51. Turner-Stokes L (2009). Goal attainment scaling (GAS) in rehabilitation: A practical guide. Clinical Rehabilitation, 23(4), 362–370. doi: 10.1177/0269215508101742 [DOI] [PubMed] [Google Scholar]
  52. U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion. (2014). Physical activity. In Healthy People 2020. Retrieved from https://www.healthypeople.gov/2020/topics-objectives/topic/physical-activity/objectives#5072
  53. Valeri L, & VanderWeele TJ (2013). Mediation analysis allowing for exposure-mediator interactions and causal interpretation: Theoretical assumptions and implementation with SAS and SPSS macros. Psychological Methods, 18(2), 137 10.1037/a0031034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. VanderWeele TJ (2015). Explanation in causal inference: Methods for mediation and interaction. New York, NY: Oxford University Press. [Google Scholar]
  55. van Stralen MM, de Vries H, Mudde AN, Bolman C, & Lechner L (2011). The long-term efficacy of two computer-tailored physical activity interventions for older adults: Main effects and mediators. Health Psychology, 30(4), 442–452. doi: 10.1037/a0023579 [DOI] [PubMed] [Google Scholar]
  56. Warner LM, Wolff JK, Ziegelmann JP, Schwarzer R, & Wurm S (2016). Revisiting self-regulatory techniques to promote physical activity in older adults: Null-findings from a randomised controlled trial. Psychology & Health, 31(10), 1145–1165. doi: 10.1080/08870446.2016.1185523 [DOI] [PubMed] [Google Scholar]
  57. Williams DM, & Rhodes RE (2016). The confounded self-efficacy construct: Conceptual analysis and recommendations for future research. Health Psychology Review, 10(2), 113–128. doi: 10.1080/17437199.2014.941998 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Yeom HA, & Fleury J (2011). Validity and reliability of the index of self-regulation scale for physical activity in older Korean Americans. Nursing Research and Practice, 2011. doi: 10.1155/2011/329534 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Yeom HA, & Fleury J (2014). A motivational physical activity intervention for improving mobility in older Korean Americans. Western Journal of Nursing Research, 36(6), 713–731. doi: 10.1177/0193945913511546 [DOI] [PubMed] [Google Scholar]

RESOURCES