Randomized Controlled Theory-Based,E-Mail-Mediated Walking Intervention

Abstract

The purpose of this study was to evaluate the ability of two concurrent randomized controlled interventions based on social cognitive theory to increase walking. A second purpose was to compare the efficacy of the intervention between two distinct groups: dog owners and non-dog owners. Adult dog owners (n = 40) and non-dog owners (n = 65) were randomized into control or intervention groups. Intervention groups received bi-weekly emails for first 4 weeks and then weekly email for the next 8 weeks targeting self-efficacy, social support, goal setting, and benefits/barriers to walking. Dog owner messages focused on dog walking while non-dog owners received general walking messages. Control groups received a 1-time email reviewing current physical activity guidelines. At 6 months, both intervention groups reported greater increases in walking and maintained these increases at 12 months. The greatest increases were seen in the dog owner intervention group. In conclusion, dog owners accumulated more walking, which may be attributed to the dog–owner relationship.

Keywords

physical activity intervention social cognitive theory walking randomized controlled trial

Introduction

Although promotion of physical activity (PA) has been a public health priority for decades, key U.S. public health agencies, such as the Centers for Disease Control and Prevention (CDC), routinely collect data indicating that most Americans are not meeting PA guidelines (CDC, 2012a). Physical inactivity is directly related to the prevalence of obesity, and obesity is a major underlying factor in the development of many chronic diseases (CDC, 2014). Walking is an excellent way for most people to increase their PA, and it is a powerful public health strategy. Walking is an easy way to start and maintain a physically active lifestyle because walking is accessible to almost anyone, does not require specific skills or abilities to perform, can be performed alone or with others, and is adaptable (i.e., can be performed at any chosen intensity, and is inexpensive; U.S. Department of Health and Human Services, 2015). Intervention strategies are needed to increase walking across populations.

Problem

Addressing physical inactivity is a crucial public health challenge. Regular PA reduces the risks of all-cause mortality, and risks and symptoms of many chronic conditions (Physical Activity Guidelines Advisory Committee, 2008). Despite these known health benefits, as few as 4% of adults ages 20 to 59 and 3% of adults 60 years of age and older meet current PA recommendations (Troiano et al., 2008). Although several PA interventions result in increased PA pre to post intervention, maintenance of those changes after the intervention ends is not often examined (Fjeldsoe, Neuhaus, Winkler, & Eakin, 2011). When maintenance is assessed, it is often not achieved (Fjeldsoe et al., 2011). Intervention approaches are needed that effectively support both increasing PA behavior and maintaining PA behavior change.

Promotion of walking could be a feasible population-level PA strategy because it does not require special equipment, requires little planning, is low cost with a low rate of injury, and can be done year-round in various settings (U.S. Department of Health and Human Services, 2015). The current U.S. PA Guidelines recommend that adults get at least 150 min (2 hr 30 min) each week of moderate-intensity PA, such as brisk walking (Physical Activity Guidelines Advisory Committee, 2008). In addition, intensity of PA is an important factor in adoption and maintenance of PA behavior change, which further supports the importance of promoting walking (Dishman & Buckworth, 1996). Recently, the Surgeon General released the Call to Action to Promote Walking and Walkable Communities (U.S. Department of Health and Human Services, 2015). In this report, the health care sector is recognized as playing a valuable role in helping to increase walking among all Americans.

Furthermore, it has been suggested that dog walking in particular may be a population-level PA-promotion strategy with long-term adherence (Christian et al., 2013), as 47% of U.S. households own a dog (the Humane Society of the United States, 2013), and research suggests this population has the potential to increase PA via dog walking (Curl, Bibbo, & Johnson, 2016; Reeves, Rafferty, Miller, & Lyon-Callo, 2011; Richards, McDonough, Edwards, Lyle, & Troped, 2013b). Studies show that a majority of dog owners (DOs) do not walk their dog (Richards, 2015; Richards et al., 2013b) and that DOs are not necessarily more active than non-dog owners (NDOs), except when considering the activity obtained via dog walking (Richards, 2015). Further studies are needed to examine the differences in PA between DOs and NDOs to understand how walking interventions may affect these distinct populations.

Internet-mediated physical activity interventions have the potential to reach a large number of people with lower costs compared with in-person intervention delivery (van den Berg, Schoones, & Vliet Vlieland, 2007). By using email, intervention participants can have flexibility with when and where they choose to interact and receive intervention information (Napolitano & Marcus, 2002). Previous reviews on the effectiveness of Internet-mediated PA interventions have indicated email as a promising intervention delivery mode (Marcus, Ciccolo, & Sciamanna, 2009; Marcus, Nigg, Riebe, & Forsyth, 2000; van den Berg et al., 2007).

Theoretical Framework

PA is a multifaceted health behavior with no single determinant able to predict or explain PA adoption and maintenance. Social cognitive theory (SCT) proposes that behavior is influenced through reciprocal interactions between personal factors, environmental influences, and behavioral attributes (Bandura, 1997). Self-efficacy is considered a personal factor and is the central construct in SCT. Self-efficacy refers to an individual’s confidence in his or her ability to perform a behavior while overcoming barriers and exerting control over the behavior through self-regulation and goal setting (Bandura, 1997). Goal setting is effective at improving PA behavior because it directs one’s effort toward desired outcomes and sets up a positive feedback cycle of improved self-efficacy and goal attainment. Specifically, (a) attaining a goal boosts self-efficacy, (b) people with higher self-efficacy set more challenging goals, and (c) more challenging goals are associated with greater behavior changes (Dishman, Vandenberg, Motl, Wilson, & DeJoy, 2010; Shilts, Horowitz, & Townsend, 2004). Outcome expectations and outcome expectancies are also considered personal factors. Outcome expectations are the consequences an individual anticipates from making a behavior change, and outcome expectancies are the value an individual places on those particular expected outcomes (Williams, Anderson, & Winett, 2005). The environment construct is defined broadly to encompass social influences such as social support. In terms of PA, behavioral attributes include intensity and difficulty. It is thought that self-efficacy directly influences PA behavior and also indirectly influences PA behavior through other SCT constructs such as social support (Maddux, 1995). It is theorized that when an individual perceives positive outcome expectations, places a high value on these outcomes, has high self-efficacy, and perceives that the PA is not unduly difficult, PA will increase (Williams et al., 2005).

Purpose

The purpose of this study was to evaluate the ability of two concurrent randomized controlled interventions based on SCT to increase walking. One intervention targeted DOs, and one intervention targeted NDOs. We hypothesize the following:

Hypothesis 1: Participants assigned to either intervention group were expected to have a greater increase in walking when compared with the control group.

A second purpose as to compare the efficacy of the intervention between two groups: DO and NDO. The second hypothesis was as follows:

Hypothesis 2: DOs in the intervention group would have the largest increases in walking and be more likely to maintain this increase over time, which may be related to perceived dog support for walking.

Method

Design and Sample

Based on prior work (unpublished), we hypothesized that the intervention would increase the proportion of participants who were physically active 60 or more min per week to 40%. We also hypothesized that there would be a slight increase in physical activity among the control groups due to the effect of becoming aware of their activity through their responses to the questionnaires. There would be sufficient statistical power (power = 0.80, when alpha = .05) to detect an absolute difference in proportion of 30% (40% in the intervention group and 10% of control group) if 19 participants were recruited for each group for a total of 38 participants per intervention. To ensure adequate power, our goal was to recruit 21 participants per group to adjust for a 10% dropout rate.

In the summer of 2014, participants were recruited through flyers and email contacts targeted at large places of employment and through the university’s clinical and translational sciences institute. Inclusion criteria for the DO group were adults 18 years of age and older who report little (<20 min a week) or no dog walking in a typical week. Inclusion criteria for the NDO group were adults 18 years of age and older who were currently not meeting national PA guidelines. In addition, all participants needed to report regular use of email and ability to walk for at least 10 min at a time. Exclusion criteria included known cardiac or pulmonary disease, joint instability, pregnancy, and known thyroid disease. Initially, 121 DOs and 205 NDOs expressed interest in the study. A research assistant screened participants for eligibility and then obtained signed informed consent for 105 participants (40 DOs and 65 NDOs; see Figure 1). Participants were then randomly assigned to the intervention or control group using a random number generator. The lead researcher was not blinded to group assignment. However, the research assistant who collected baseline and follow-up measures was blinded to group assignment. Procedures were approved by the Purdue University Committee on the Use of Human Research Subjects.

Intervention Procedure and Structure

The DO intervention group received the Dogs Physical Activity and Walking (PAW) intervention, which is a 3-month email-mediated intervention to increase dog walking among DOs. A complete description of the Dog PAW intervention, including email contents, is published elsewhere (Richards, Ogata, & Ting, (2015)). Both the DO and NDO intervention emails were developed to be aligned with individual, behavioral, and environmental constructs of SCT (please see Table 1 for content of intervention emails for NDOs).

Table 1.

Content of Intervention Emails for Non-Dog Owners.

Time	SCT variable	Intervention content
Wk 1	Self-regulation Outcome expectations	Goal setting: how to create a specific action plan Describe anticipated result of walking, benefits of exercise (outcome expectations)
Wk 2	Self-regulation Reinforcements	Goal setting and self-monitoring Motivational messages
Wk 3	Self-regulation Self-efficacy	Self-monitoring Time management strategies
Wk 4	Self-regulation Self-efficacy Social support	Progress toward goals reviewed Overcoming barriers Social support can be provided by creating expectations, contracting with oneself or someone else, by having an exercise buddy
Wk 5	Self-regulation Outcomes Environment	Self-monitoring Reinforce the anticipated result of walking, benefits of exercise (outcome expectations) Geographic mapping tool to find, create, map walking paths
Wk 6	Self-regulation Self-efficacy	Review of goal progress and self-monitoring Time management, relapse prevention, overcoming barriers
Wk 7	Outcomes	Reinforce the anticipated result of walking, benefits of exercise (outcome expectations)
Wk 8	Self-regulation Self-efficacy	Review of goal progress and self-monitoring Time management, relapse prevention, tips to increase walking
Wk 9	Social support Environment	Encourage walking with friends, family, groups Geographic mapping tool to find, create, map walking path
Wk 10-12	Self-regulation Self-efficacy Reinforcements	Review of goal progress and self-monitoring Time management and overcoming barriers Motivational messages about health outcomes of walking, examples walking schedules to encourage maintenance

Note. SCT = Social cognitive theory.

Starting in mid-September 2014, both intervention groups received emails regarding the importance of walking and strategies to increase walking. The DO group messages were tailored to focus on dog walking while NDOs received general walking messages. Intervention emails were sent bi-weekly for the first 4 weeks of the study and then weekly for the next 8 weeks. These emails targeted principles of self-efficacy, social support, and goal setting. Both control groups received a one-time email reviewing current PA guidelines.

Measures

Measurement of variables occurred at baseline, immediately post intervention (3 months), and at 6 and 12 months through standardized online questionnaires. With the exception of dog-specific questions, measures were analogous for both intervention groups and both control groups.

Sociodemographic characteristics included age, gender, marital status, household income, race, ethnicity, and education, and were assessed at baseline. Health measures included number of poor physical health days and poor mental health days in the past 30 days using the questions from the Behavioral Risk Factor Surveillance System (CDC, 2012a). Body mass index (BMI) was calculated based on self-reported height and weight using the following formula: weight (lb) / [height (in)]² × 703 (CDC, 2015). Participants were classified as overweight if BMI was 25.0 to 29.9 and obese if BMI was ≥30.0 (CDC, 2012b).

Theoretical constructs were measured using existing measures with demonstrated reliability and validity and adapted to be specific to walking (Sallis, Grossman, Pinski, Patterson, & Nader, 1987; Sallis, Pinski, Grossman, Patterson, & Nader, 1988; Steinhardt & Dishman, 1989) or dog walking (Richards, McDonough, Edwards, Lyle, & Troped, 2013a). A full description of the measures used for DOs can be found at Richards et al. (2013a). All measures demonstrated acceptable levels of internal consistency reliability (α > .70). Measures for the NDO group included self-efficacy for walking, which was measured with two Likert-type scale subscales: Making Time (five items) and Resisting Relapse (four items; Sallis et al., 1988). Outcome expectation items were used to assess the benefits participants believe they will get from walking (five Likert-type scale items; 1 = strongly agree, 5 = strongly disagree; that is, lose weight, improve mood, improve health). Outcome expectancy items were used to assess the value placed on each specific outcome expectation (five Likert-type scale items; 1 = very unimportant, 5 = very important; Steinhardt & Dishman, 1989). Social support for walking items assessed perceived social interactions and activities aimed at supporting walking received from their family (four items) and friends (four items; Sallis et al., 1987). In addition to family and friend social support, dog-related support for walking was assessed (three items) in the DO groups (Richards et al., 2013a). Mean scores were computed across all items in each subscale.

Self-reported walking and overall PA during the past 7 days were assessed with six items from the International Physical Activity Questionnaire (Craig et al., 2003). Questions assessed the number of days and minutes per day of walking, moderate (MPA) and vigorous (VPA) performed for at least 10 min at a time. In addition, dog walking, defined as an activity in which both the dog and the owner are walking together with the dog on or off leash, assessed only in the DO groups, with three items: number of days of dog walking in a typical week, average number of dog walks per day, and the typical duration per dog walk (Richards et al., 2013a).

Data Analysis

Descriptive statistics were used to summarize the findings of participant characteristics, theoretical constructs, and PA variables. Chi-square and two-sample t tests were used to assess differences between the intervention and control groups at baseline and between baseline and post intervention. Data were analyzed using R 3.2.2 .Statistical significance was set at p < .05. However, p values < .10 are also reported due to reduced sample size in the DO intervention group post baseline. These findings will require further investigation, however; the risk of rejecting important research hypotheses was judged more important that the risk of Type I error. Model diagnostics were performed to examine normality, constant variance, and independence assumptions of each fitted model.

To examine whether theoretical construct variables changed across time, between groups, a linear mixed model, in which the group, time point, and their interaction were the independent variables, and participant ID was a random effect with no nesting structure, was used. The across time points analyses were carried out by using Tukey’s honest significant difference (HSD) test from the linear mixed modeling. Due to multicollinearity between theoretical constructs, it was not appropriate to include all theoretical constructs in one model for a multivariate analysis. Sociodemographic variables were not controlled for because there were no significant correlations between these variables and the physical activity outcomes.

Findings

On average, participants were overweight or obese (76%), middle-aged (44.1 ± 13.0 years), educated (at least 80% with a college degree), females (75%). Seventy-three percent of participants were non-Hispanic White, 11% were Black, 10% were Hispanic, and 6% were Asian. Eighteen percent of participants were single, 18% were divorced, and 64% were married or partnered. At baseline, there were no significant differences in any sociodemographics between groups.

At baseline, DOs overall reported significantly fewer poor physical health (0.9 ± 3.3 days) and mental health (1.5 ± 3.8 days) than NDOs overall (6.5 ± 8.4 poor physical health days, 7.5 ± 8.1 poor mental health days) in the past 30 days (see Tables 2 and 3). At 3 months, the reported poor physical and mental health days in the DO group increased to NDO levels, with no significant differences between groups. No significant changes were seen in 6-month or 12-month time points in NDO groups. At 6 months, the DO intervention group had a significant increase in poor physical health days compared with baseline (baseline = 0.2 ± 0.5; 6 months = 5.2 ± 7.7; p <.10). There were no significant changes seen at 12 months. There were no significant changes in BMI between groups or across time.

Table 2.

Means and Standard Errors (M ± SE) of the Theoretical Constructs Across Time in Dog Owners.

	Dog owner control				Dog owner intervention
	Baseline (n = 20)	3 months (n = 20)	6 months (n = 19)	12 months (n = 17)	Baseline (n = 20)	3 months (n = 17)	6 months (n = 16)	12 months (n = 16)
Theoretical constructs
Pet attachment	3.9 ± 0.6	NA	4.1 ± 0.8	4.0 ± 0.8	3.8 ± 0.4	NA	3.9 ± 0.6	3.9 ± 0.6
Self-efficacy
Resisting relapse	3.3 ± 0.9	3.1 ± 0.6	3.4 ± 0.5	3.3 ± 0.9	3.6 ± 1.0	3.5 ± 0.8	3.3 ± 0.8	3.5 ± 1.1
Making time	3.7 ± 0.7	3.3 ± 0.7	3.4 ± 0.7	3.4 ± 1.0	3.5 ± 1.0	3.4 ± 0.9	3.2 ± 0.8	3.4 ± 1.1
Outcome expectations
Owner-specific	4.0 ± 0.7	4 ± 0.9	4.3 ± 0.6	4.1 ± 0.7	4.1 ± 0.7	4.1 ± 0.5	4.1 ± 0.7	4.2 ± 0.6
Dog-specific	4.6 ± 0.4	4.3 ± 0.9	4.6 ± 0.5	4.5 ± 0.6	4.4 ± 0.5	4.3 ± 0.5	4.3 ± 0.5	4.4 ± 0.6
Outcome values
Owner-specific	4.0 ± 0.6	3.9 ± 0.6	4.1 ± 0.5	4.0 ± 0.6	3.9 ± 0.9	3.9 ± 0.6	3.9 ± 0.6	4.2 ± 0.6
Dog-specific	4.4 ± 0.6	4.3 ± 0.9	4.5 ± 0.4	4.4 ± 0.7	4.4 ± 0.6	4.4 ± 0.5	4.4 ± 0.5	4.4 ± 0.5
Social support
Dog support	3.6 ± 0.9	NA	3.8 ± 1.1	3.8 ± 1.1	3.4 ± 1.2	NA	3.3 ± 1.0	3.8 ± 1.2
Friend support	1.5 ± 0.6	1.3 ± 0.4	1.4 ± 0.6	1.3 ± 0.5	1.2 ± 0.3	1.2 ± 0.5	1.3 ± 0.6	1.5 ± 0.6
Family support	2.3 ± 0.8	2.1 ± 0.9	2.0 ± 1.0	2.2 ± 1.3	2.3 ± 0.9	2.1 ± 1.2	2.0 ± 1.1	2.0 ± 1.0
Poor health days
Physical health	1.7 ± 4.5	4.3 ± 6.6	5.1 ± 6.7	3.4 ± 5.4	0.2 ± 0.5	4.9 ± 9.8	5.2 ± 7.7*	3.0 ± 4.7
Mental health	2.0 ± 4.8	6.0 ± 7.7	3.8 ± 4.8	5.2 ± 6.2	1.0 ± 2.3	6.7 ± 10.3	5.2 ± 7.7	5.4 ± 7.6

Note. NA = not assessed.

Significance mark of difference between groups across time: Difference from baseline: *p < .10. **p < .05.

Table 3.

Means and Standard Errors (M ± SE) of the Theoretical Constructs Across Time in Non-Dog Owners.

	Non-dog owner control				Non-dog owner Intervention
	Baseline (n = 32)	3 months (n = 22)	6 months (n = 20)	12 months (n = 20)	Baseline (n = 33)	3 months (n = 19)	6 months (n = 19)	12 months (n = 23)
Theoretical constructs
Self-efficacy
Resisting relapse	2.9 ± 0.8	2.7 ± 0.9	2.6 ± 0.7	2.8 ± 1.1	2.7 ± 1.2	2.5 ± 1.0	2.6 ± 0.8	2.8 ± 0.9
Making time	2.9 ± 0.8	2.7 ± 0.9	2.3 ± 0.8***	2.9 ± 1.0*	2.9 ± 0.9	2.6 ± 1.0	2.4 ± 0.9	2.8 ± 0.8
Outcome expectations	4.0 ± 0.5	3.9 ± 0.6	4.2 ± 0.5	3.9 ± 0.6	4.1 ± 0.7	3.9 ± 1.0	3.9 ± 0.9	4.1 ± 0.6
Outcome values	4.1 ± 0.5	3.8 ± 0.8	4.1 ± 0.5	4.0 ± 0.5	4.2 ± 0.7	4.1 ± 0.7	3.9 ± 1.0	4.1 ± 0.6
Social support
Friend support	2.0 ± 0.8	1.8 ± 1.1	1.9 ± 0.9	1.9 ± 1.0	1.9 ± 1.0	1.6 ± 0.8	1.7 ± 0.8	1.6 ± 0.9
Family support	2.2 ± 1.0	1.9 ± 1.0	2.1 ± 0.9	2.2 ± 1.3	2.1 ± 0.9	1.8 ± 0.8	2.1 ± 1.1	2.1 ± 1.0
Poor health days
Physical health	5.8 ± 6.4	5.3 ± 4.9	3.9 ± 4.6	4.5 ± 7.9	7.2 ± 9.9	3.7 ± 5.0	3.5 ± 6.8	4.9 ± 7.3
Mental health	6.5 ± 6.8	7.1 ± 8.9	7.3 ± 9.4	5.9 ± 8.7	8.5 ± 9.1	4.1 ± 5.4	5.2 ± 8.2	8.0 ± 10.6

Significance mark of difference between groups across time: difference from previous time: *p < .10. **p < .05; difference from baseline: ***p < .05.

Theoretical Constructs

With the exception of the self-efficacy measures, there were no significant differences or changes in theoretical constructs at baseline or across follow-up measures (see Tables 2 and 3). When comparing across DO and NDO groups, DOs in both the control and intervention group reported significantly higher self-efficacy at baseline when compared with NDOs. This difference remained significant at 3 months for the DO intervention group. At 6-month follow-up, the Making Time subscale of self-efficacy slightly increased in the DO control group but significantly decreased for the NDO control group. Significant changes in self-efficacy were not seen at 12 months in any group.

Walking and Dog Walking

At baseline, there were no significant differences in any of the PA variables between control and intervention groups or between DOs and NDOs (see Figures 2 to 5). At baseline, the DO control group walked an average of 103.2 ± 99.7 min per week and the DO intervention group walked an average of 132.0 ± 138.4 min per week. In addition to overall walking, both DO intervention and control groups walked less than 20 min per week with their dog. At baseline, the NDO control group walked an average of 41.3 ± 41.6 min per week and the NDO intervention group walked an average of 49.2 ± 38.8 min per week. Immediately post intervention (3 months), the DO control group significantly increased their MPA compared with the NDO control group. This increase was not significant when compared with the DO intervention group (see Figure 3). At 6 months, the NDO control significantly increased their MPA compared with baseline measures (see Figure 3). The DO intervention group significantly increased their dog walking compared with the control group (DO intervention = 68.4 ± 84.9 vs. DO control = 28.9 ± 30.6 min; p < .05; see Figure 4). The NDO intervention group significantly increased their weekly minutes of walking compared with baseline measures (6 months = 160.0 ± 135.7 vs. baseline = 49.2 ± 38.8 min; p < .05; see Figure 5). At 12 months, the NDO intervention group maintained the significant increase in weekly minutes of walking (152.4 ± 135.7 min), and the DO intervention group continued to see significant increases in weekly minutes of dog walking (154.2 ± 152.4 min) compared with baseline measures and compared with control groups. The DO control group also significantly increased their dog walking when compared with baseline measures (12 months = 79.2 ± 83.2 vs. baseline = 8.2 ± 10.2 min; p < .05). Overall, the DO intervention group had an increase of 141.2 ± 160.1 weekly min of dog walking while the NDO intervention group had an increase of 100.4 ± 123.5 weekly min of walking.

Discussion

The primary purpose of this study was to evaluate the ability of two concurrent randomized controlled interventions to increase walking. A second purpose was to compare the efficacy of the intervention between two distinct groups: DOs and NDOs. Significant increases in walking or dog walking were not seen until 6-month measures. This suggests that physical activity is a complex behavior to change and may require longer time to incorporate into everyday life. This finding is supported by habit formation research, which has found it takes between 66 and 254 days to make a health behavior change, such as physical activity, a habit (Lally, Van Jaarsveld, Potts, & Wardle, 2010). In addition, 3-month measures were collected in December and seasonality is known to affect PA behaviors. Previous studies have shown that inclement weather affects leisure time PA (Matthews et al., 2001). However, other studies have shown that dog walking is not as strongly affected by cold or rainy weather (Temple, Rhodes, & Higgins, 2011).

The primary expected outcome of this study was supported. Participants randomly assigned to both intervention groups showed a significant increase in walking or dog walking when compared with participants in the control groups at 6 and 12 months. The second expected outcome was also supported as the DO intervention group reported the greatest increase in walking and maintained this increase at 12 months. However, analysis was not able to directly relate this increase to perceived dog support for walking. Previous studies strongly support that DOs who walk their dogs are motivated to do so because of dog-related support for walking (Johnson & Meadows, 2010; Rhodes, Murray, Temple, Tuokko, & Higgins, 2012; Richards, Ogata, & Ting, 2015).

Importantly, the increase seen in dog walking in the intervention group did not appear to be at the sacrifice of other forms of PA. For example, weekly minutes of non-dog walking, MPA, and VPA remained stable in the intervention group. Interestingly, significant relationships were not identified between changes in theoretical construct variables with changes in walking or dog walking. Therefore, we were unable to draw conclusions on which SCT constructs were most important in producing PA behavior change.

Results of this intervention indicate that a simple theory-based, email-mediated intervention was effective at increasing and maintaining an increase in walking or dog walking among both DOs and NDOs. The effectiveness of email-mediated interventions in increasing PA has been supported by other studies as well (Marcus et al., 2009; Plotnikoff, Pickering, McCargar, Loucaides, & Hugo, 2010). Findings in the DO intervention group were consistent with the Dogs PAW intervention, which previously demonstrated efficacy in increasing dog walking and maintaining this increase at 12 months (Richards et al., 2015). Current findings support that a sense of DO responsibility to care for their dog(s) is an appropriate strategy to increase and maintain increases in dog walking (Rhodes et al., 2012). Importantly, a recent study showed that this sense of DO responsibility aligns with intrinsic forms of motivation, which is associated to long-term adherence to behavior change (Teixeira, Carraca, Markland, Silva, & Ryan, 2012). Future dog walking promotion initiatives should consider the dog and owner relationship.

There are several study limitations worth noting. First, despite recruiting enough participants for the study to be adequately powered, attrition was higher than expected based on previous studies (Richards et al., 2015). Much of the attrition in the DO groups occurred because of unavoidable issues outside of the intervention (i.e., dog death, relocation, participant illness). A larger trial with more diverse participants is warranted. In addition, monitoring the number of emails viewed by participants could further strengthen the findings of this study as it is currently unknown whether each participant received the full dose of the intervention. Also, PA intervention research has been unable to determine the needed intervention dose or duration to achieve and maintain behavior change. It is possible that more frequent emails or receiving emails for longer than 3 months would intensify the dose–response effect. In addition, this study relied on self-reported PA behavior, which is prone to self-report, social desirability, testing, and recall bias. These biases may be seen in the overall high weekly minutes of self-reported PA. However, the survey items used to assess PA have been extensively tested and are shown to be reliable and valid measures (Craig et al., 2003; Richards et al., 2013a). Objective assessments of PA such as the use of accelerometers, pedometers, or smartphone-based fitness apps should be considered in future studies (Strath et al., 2013; van den Berg et al., 2007).

There are also strengths worth noting. First, the intervention messages were created based on a well-studied health behavior theory, SCT (Maddux, 1995). In the past, PA interventions have not consistently utilized health behavior theories in intervention development and evaluation (Marcus et al., 2006; Rhodes & Pfaeffli, 2010). When theories have been used, the intervention typically only utilizes one or two constructs of the theory instead of using the theory in its entirety (Rhodes & Pfaeffli, 2010). There is significant research that shows health behavior change interventions are more likely to be successful if they are based on a clear understanding of the targeted health behavior and influencing factors (Richards & Cai, 2016). In addition, the study design allowed for comparison between two groups (DO and NDO), which have not been previously examined in an intervention setting. Overall, the generalizability of this study is high as the intervention was delivered via email, the setting for the intervention is easily transferable, and implementation costs are low.

Application

Our findings show that a SCT-based email intervention among DOs and NDOs had positive effects on increasing walking. Although these messages targeted specific behaviors and specific behavior change strategies based on SCT, the messages were standardized across intervention groups. It appears that this generic messaging was effective in increasing walking behaviors and has the potential to reach a large number of people. Messages tailored to participants’ readiness to change and/or current PA level may be necessary to achieve greater effects (Marcus et al., 2006).

Given the high proportion of the population that has regular visits to primary care providers, these settings provide a valuable opportunity to influence PA behavior (Richards & Cai, 2016). Walking in particular can be an important behavior to promote because it can be undertaken without the need for special training or equipment and therefore can be a sustainable behavior change. Specific to dog walking, health care providers can ask their patients whether they own a dog and use this information to promote dog walking (Epping, 2011). More specifically, theory-based, email-mediated interventions can be valuable for patients and providers as an independent and convenient way to facilitate an increase in PA (Marcus et al., 2009). Providing clinicians with information and training on creating and delivering theory-based, email-mediated PA promotion is a promising avenue to health behavior change.

In summary, this study found two theory-based, email-mediated walking interventions to be feasible and effective in increasing walking among two distinct populations: DOs and NDOs. The email format of this intervention allows convenience, flexibility, and independence in PA behavior change. Approaches are needed that effectively support both increasing PA behavior and maintaining PA behavior change. This study addresses an often lacking finding in PA research maintenance; not only was PA increased, but this increase was also maintained at 12-month follow-up.

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.

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 a grant from the Human Animal Bond Research Initiative.

Author Biographies

Elizabeth A. Richards, PhD, RN, CHES, is an assistant professor at Purdue University School of Nursing. Her research is focused on community-based, population-level physical activity promotion.

Niwako Ogata, BVSc, PhD, is an assistant professor at Purdue University College of Veterinary Medicine. Her research is focused on companion animal behaviors.

Ching-Wei Cheng, BA, is a graduate student at Purdue University College of Science with a major in statistics.

References

Bandura

(1997). Self-efficacy: The exercise of self control. New York, NY: Freeman.

Centers for Disease Control and Prevention. (2012a). Behavioral Risk Factor Surveillance System Survey Data. Atlanta, GA: Author.

Centers for Disease Control and Prevention. (2012b). Defining overweight and obesity. Retrieved from http://www.cdc.gov/obesity/adult/defining.html

Centers for Disease Control and Prevention. (2014). Chronic disease overview: Costs of chronic disease. Retrieved from http://www.cdc.gov/nccdphp/overview.htm

Centers for Disease Control and Prevention. (2015). BMI calculation and interpretation. Retrieved from http://www.cdc.gov/healthyweight/assessing/bmi/adult_bmi/index.html#Interpreted

Christian

Wesgarth

Bauman

Richards

E. A.

Rhodes

Evenson

K. R.

. . . Thorpe

(2013). Dog ownership and physical activity: A review of the evidence. Journal of Physical Activity & Health, 10, 750-759.

Craig

C. L.

Marshall

A. L.

Sjöström

Bauman

A. E.

Booth

M. L.

Ainsworth

B. E.

. . . Oja

(2003). International Physical Activity Questionnaire: 12-country reliability and validity. Medicine & Science in Sport & Exercise, 35, 1381-1395. doi:10.1249/01.MSS.0000078924.61453.FB

Curl

A. L.

Bibbo

Johnson

R. A.

(2016). Dog walking, the human-animal bond and older adults’ physical health. Gerontologist. Advance online publication. doi:10.1093/geront/gnw051

Dishman

R. K.

Buckworth

(1996). Increasing physical activity: A quantitative synthesis. Medicine & Science in Sport & Exercise, 28, 706-719.

10.

Dishman

R. K.

Vandenberg

R. J.

Motl

R. W.

Wilson

M. G.

DeJoy

D. M.

(2010). Dose relations between goal setting, theory-based correlates of goal setting and increases in physical activity during a workplace trial. Health Education Research, 25, 620-631. doi:10.1093/her/cyp042

11.

Epping

J. N.

(2011). Dog ownership and dog walking to promote physical activity and health in patients. Current Sports Medicine Reports, 10, 224-227. doi:10.1249/JSR.0b013e318223ee41

12.

Fjeldsoe

Neuhaus

Winkler

Eakin

(2011). Systematic review of maintenance of behavior change following physical activity and dietary interventions. Health Psychology, 30(1), 99-109. doi:10.1037/a0021974

13.

The Humane Society of the United States. (2013). Pets by the numbers. Retrieved from http://www.humanesociety.org/issues/pet_overpopulation/facts/pet_ownership_statistics.html

14.

Johnson

R. A.

Meadows

R. L.

(2010). Dog-walking: Motivation for adherence to a walking program. Clinical Nursing Research, 19, 387-402. doi:10.1177/1054773810373122

15.

Lally

Van Jaarsveld

Potts

H. W.

Wardle

(2010). How are habits formed: Modelling habit formation in the real world. European Journal of Social Psychology, 40, 998-1009.

16.

Maddux

(1995). Self-efficacy, adaptation, and adjustment: Theory, research, and application. New York, NY: Plenum.

17.

Marcus

B. H.

Ciccolo

J. T.

Sciamanna

C. N.

(2009). Using electronic/computer interventions to promote physical activity. British Journal of Sports Medicine, 43, 102-105. doi:10.1136/bjsm.2008.053744

18.

Marcus

B. H.

Nigg

C. R.

Riebe

Forsyth

L. H.

(2000). Interactive communication strategies: Implications for population-based physical-activity promotion. American Journal of Preventive Medicine, 19, 121-126.

19.

Marcus

B. H.

Williams

D. M.

Dubbert

P. M.

Sallis

J. F.

King

A. C.

Yancey

A. K.

, . . . Interdisciplinary Working Group on Quality of Care and Outcomes Research. (2006). Physical activity intervention studies: What we know and what we need to know: A scientific statement from the American Heart Association Council on Nutrition, Physical Activity, and Metabolism (Subcommittee on Physical Activity); Council on Cardiovascular Disease in the Young; and the Interdisciplinary Working Group on Quality of Care and Outcomes Research. Circulation, 114, 2739-2752. doi:10.1161/CIRCULATIONAHA.106.179683.

20.

Matthews

C. E.

Freedson

P. S.

Hebert

J. R.

Stanek

E. J.

Merriam

P. A.

Rosal

M. C.

. . . Ockene

I. S.

(2001). Seasonal variation in household, occupational, and leisure time physical activity: Longitudinal analyses from the seasonal variation of blood cholesterol study. American Journal of Epidemiology, 153, 172-183.

21.

Napolitano

M. A.

Marcus

B. H.

(2002). Targeting and tailoring physical activity information using print and information technologies. Exercise and Sport Sciences Reviews, 30, 122-128.

22.

Physical Activity Guidelines Advisory Committee. (2008). Physical activity guidelines advisory committee report. Washington, DC. Retrieved from http://www.health.gov/PAGuidelines/Report

23.

Plotnikoff

R. C.

Pickering

M. A.

McCargar

L. J.

Loucaides

C. A.

Hugo

(2010). Six-month follow-up and participant use and satisfaction of an electronic mail intervention promoting physical activity and nutrition. American Journal of Health Promotion, 24, 255-259. doi:10.4278/ajhp.08032830

24.

Reeves

M. J.

Rafferty

A. P.

Miller

C. E.

Lyon-Callo

S. K.

(2011). The impact of dog walking on leisure-time physical activity: Results from a population-based survey of Michigan adults. Journal of Physical Activity & Health, 8, 436-444.

25.

Rhodes

R. E.

Murray

Temple

V. A.

Tuokko

Higgins

J. W.

(2012). Pilot study of a dog walking randomized intervention: Effects of a focus on canine exercise. Preventive Medicine, 54, 309-312. doi:10.1016/j.ypmed.2012.02.014

26.

Rhodes

R. E.

Pfaeffli

L. A.

(2010). Mediators of physical activity behaviour change among adult non-clinical populations: A review update, International Journal of Behavioral Nutrition and Physical Activity, 7, Article 37. doi:10.1186/1479-5868-7-37

27.

Richards

E. A.

(2015). Does dog walking predict physical activity participation: Results from a national survey. American Journal of Health Promotion, 30(5), 323-330. doi:10.4278/ajhp.140820-QUAN-416

28.

Richards

E. A.

Cai

(2016). Integrative review of nurse-delivered physical activity interventions in primary care. Western Journal of Nursing Research, 38, 484-507. doi:10.1177/0193945915581861

29.

Richards

E. A.

McDonough

Edwards

Lyle

Troped

(2013a). Development and psychometric testing of the Dogs and Walking Survey (DAWGS). Research Quarterly for Exercise and Sport, 84, 492-502. doi:10.1080/02701367.2013.839935

30.

Richards

E. A.

McDonough

Edwards

Lyle

Troped

(2013b). Psychosocial and environmental factors associated with dog walking. International Journal of Health Promotion and Education, 51, 198-211.

31.

Richards

E. A.

Ogata

Ting

(2015). Dogs, physical activity, and walking (Dogs PAW): Acceptability and feasibility of a pilot physical activity intervention. Health Promotion Practice, 16, 362-370. doi:10.1177/1524839914553300

32.

Sallis

J. F.

Grossman

R. M.

Pinski

R. B.

Patterson

T. L.

Nader

P. R.

(1987). The development of scales to measure social support for diet and exercise behaviors. Preventive Medicine, 16, 825-836.

33.

Sallis

J. F.

Pinski

R. B.

Grossman

R. M.

Patterson

T. L.

Nader

P. R.

(1988). Development of self-efficacy scales for health-related diet and exercise. Health Education Research, 3, 283-292.

34.

Shilts

M. K.

Horowitz

Townsend

M. S.

(2004). Goal setting as a strategy for dietary and physical activity behavior change: A review of the literature. American Journal of Health Promotion, 19, 81-93.

35.

Steinhardt

M. A.

Dishman

R. K.

(1989). Reliability and validity of expected outcomes and barriers for habitual physical activity. Journal of Occupational and Environmental Medicine, 31, 536-546.

36.

Strath

S. J.

Kaminsky

L. A.

Ainsworth

B. E.

Ekelund

Freedson

P. S.

Gary

R. A.

, . . . American Heart Association Physical Activity Committee of the Council on Lifestyle and Cardiometabolic Health and Cardiovascular, Exercise, Cardiac Rehabilitation and Prevention Committee of the Council on Clinical Cardiology, and Council on Cardiovascular and stroke Nursing. (2013). Guide to the assessment of physical activity: Clinical and research applications: A scientific statement from the American Heart Association. Circulation, 128, 2259-2279. doi:10.1161/01.cir.0000435708.67487.da

37.

Teixeira

Carraca

Markland

Silva

Ryan

(2012). Exercise, physical activity, and self-determination theory: A systematic review. International Journal of Behavioral Nutrition and Physical Activity, 9, Article 78.

38.

Temple

Rhodes

Higgins

(2011). Unleashing physical activity: An observational study of park use, dog walking, and physical activity. Journal of Physical Activity & Health, 8, 766-774.

39.

Troiano

R. P.

Berrigan

Dodd

K. W.

Mâsse

L. C.

Tilert

McDowell

(2008). Physical activity in the United States measured by accelerometer. Medicine & Science in Sports & Exercise, 40, 181-188. doi:10.1249/mss.0b013e31815a51b3

40.

U.S. Department of Health and Human Services. (2015). Step it up! The surgeon general’s call to action to promote walking and walkable communities. Washington, DC: Office of the Surgeon General, U.S. Department of Health and Human Services.

41.

van den Berg

M. H.

Schoones

J. W.

Vliet Vlieland

T. P.

(2007). Internet-based physical activity interventions: A systematic review of the literature. Journal of Medical Internet Research, 9, Article e26. doi:10.2196/jmir.9.3.e26

42.

Williams

D. M.

Anderson

E. S.

Winett

R. A.

(2005). A review of the outcome expectancy construct in physical activity research. Annals of Behavioral Medicine, 29(1), 70-79. doi:10.1207/s15324796abm2901_10