Effectiveness of the fun for wellness online behavioral intervention to promote well-being and physical activity: protocol for a randomized controlled trial

Self-efficacy theory [13] provided the conceptual model that guided the creation of capability-enhancing learning opportunities (i.e., challenges) for participants to engage with in the FFW intervention. The capability-enhancing learning opportunities made available to participants are 152 interactive and scenario-based challenges that are ordered in the FFW website by the BET I CAN abbreviation [1]. The Behavior-intensive learning opportunities focus on enhancing a person’s capabilities for both effective goal setting and forming positive habits [16]. The Emotion-intensive learning opportunities focus on enhancing a person’s capabilities for both coping with negative feelings and collecting positive feelings [17]. The Thought-intensive learning opportunities focus on enhancing a person’s capabilities for both challenging negative expectations and creating a new life story [18]. The Interaction-intensive learning opportunities focus on enhancing a person’s capabilities for both communicating and connecting with other people [19]. The Context-intensive learning opportunities focus on enhancing a person’s capabilities for both reading and changing cues in the environment [20]. The Awareness-intensive learning opportunities focus on enhancing a person’s capabilities for both knowing herself/himself and knowing the issue [21]. The Next steps-intensive learning opportunities focus on enhancing a person’s capabilities for both making and sticking with a strategy [22]. To summarize, each BET I CAN challenge available in the FFW intervention was designed to provide an opportunity for an individual to increase their capabilities to organize and execute actions that may increase their well-being.

Every BET I CAN challenge in the FFW online behavioral intervention was designed in a way to expose an individual to one or more proposed sources of self-efficacy beliefs: enactive mastery experiences; vicarious experiences; verbal persuasion; and self-assessments of physiological and/or emotional states [13]. An example of exposure to an enactive mastery experience in the FFW online behavioral intervention is requiring an individual to engage with an interactive game to complete a BET I CAN challenge [1]. An example of exposure to a vicarious experience in the FFW online behavioral intervention is requiring an individual to watch a vignette of professional actors (some of whom may be classified as overweight and/or obese) modeling healthful behaviors to complete a BET I CAN challenge [1]. An example of exposure to verbal persuasion in the FFW online behavioral intervention is requiring an individual to listen to a coach deliver an animated lecture on healthful behaviors to complete a BET I CAN challenge [1]. An example of exposure to self-assessments of physiological and/or emotional states in the FFW online behavioral intervention is requiring an individual to engage in a self-reflection exercise to complete a BET I CAN challenge [1]. In summary, each of these capability enhancing-learning opportunities was guided by the substantial extant literature on the potential importance of proposed sources of self-efficacy information in exercise psychology [15, 23].

A necessary condition for valid testing of self-efficacy theory is a high degree of concordance between the domain specific self-efficacy beliefs and the proposed outcomes of the domain specific self-efficacy beliefs of interest [13]. The importance of maximizing concordance between efficacy beliefs and proposed outcomes of efficacy beliefs has been demonstrated in exercise psychology [24]. Bandura [25] advocated for the construction of study-specific self-efficacy scales as a mechanism for maximizing concordance between the domain specific self-efficacy beliefs and the proposed outcomes of the domain specific self-efficacy beliefs of interest. The conceptualization of self-efficacy beliefs within the FFW intervention was concordant with a multidimensional conceptualization of well-being [26].

As depicted in Fig. 1 the FFW intervention was hypothesized to exert both a positive direct effect, and a positive indirect effect through well-being self-efficacy, on well-being (i.e., subjective well-being and well-being actions). Within the FFW conceptual framework the well-being construct is based on both a multidimensional model of subjective well-being (e.g., how satisfied are you with your current wellness and physical health) and a multidimensional model of well-being actions (e.g., how many days per week do you engage in moderate physical activity for at least 30 min). From this point forward we generally use the succinct expression, well-being, to simultaneously refer to both subjective well-being and well-being actions as conceptualized within the FFW intervention.

The conceptual model was expanded to include well-being actions self-efficacy for two primary reasons. The first reason was based on results from previous research where some evidence was provided for the efficacy of the FFW online behavioral intervention to increase well-being self-efficacy in those who complied with the intervention [26]. We reason that if exposure to the BET I CAN challenges can increase an individual’s self-efficacy regarding his or her well-being related thoughts (i.e., subjective well-being) then exposure to the BET I CAN challenges may also increase an individual’s self-efficacy regarding his or her well-being related behavior (i.e., well-being actions). The second reason for adding well-being actions self-efficacy to the current study was to maximize concordance between well-being actions and the measured set of self-efficacy beliefs. Recall that within the FFW conceptual framework the operational definition of the well-being self-efficacy construct (i.e., the degree to which an individual perceives that they have the capability to attain a positive status in key domains of their life) is more concordant with the operational definition of the subjective well-being construct (i.e., an individual’s satisfaction with their status in key domains of their life) than with the operational definition of the well-being actions construct (i.e., an individual’s actions that may improve their status in key domains of their life). The operational definition for the well-being actions self-efficacy construct in the FFW effectiveness trial is the degree to which an individual believes that they have the capability to take actions that may improve their status in key domains of their life.

The conceptual model was expanded to include physical activity for three primary reasons. The first reason was based on previous research where some evidence was provided for the efficacy of the FFW online behavioral intervention to increase physical well-being actions [31]. Measurement of physical well-being actions in the aforementioned study, however, consisted of only two items (e.g., how often do you engage in moderate physical activity such as brisk walking for about 30 min at least five times a week?). The current study seeks to more thoroughly measure physical activity. The second reason for adding physical activity to the current study was in response to calls for effective and scalable interventions to combat the global pandemic of physical inactivity [6, 7]. The final reason for adding physical activity to the current study was existing evidence that well-designed cognitive-behavioral interventions can successfully promote physical activity in adults with obesity [5]. The operational definition for physical activity in the FFW effectiveness trial is bodily movement produced by skeletal muscles that requires energy expenditure [34].

The conceptual model was expanded to include physical activity self-efficacy for three primary reasons. The first reason was based on results from previous research where some evidence was provided for the efficacy of the FFW online behavioral intervention to increase well-being self-efficacy beliefs [26] and physical well-being actions [31] in those who complied with the intervention. The second reason for adding physical activity self-efficacy to the current study was to maximize concordance between physical activity and the measured set of self-efficacy beliefs. The final reason for adding physical activity self-efficacy to the current study was existing evidence of a positive relationship between self-efficacy and physical activity in adults [10]. The operational definition for physical activity self-efficacy in the FFW effectiveness trial is the degree to which an individual perceives that they have the capability to engage in a recommended amount of weekly physical activity for health.

The conceptual model was expanded to include self-efficacy to regulate physical activity for reasons that closely follow the rationale already provided for including physical activity self-efficacy. The reason, however, for including both a self-efficacy “level” construct (i.e., physical activity self-efficacy) and a self-regulatory efficacy construct (i.e., self-efficacy to regulate physical activity) is that the former focuses on an individual’s beliefs in his or her ability to accomplish levels of a task while the latter focuses on an individual’s beliefs to overcome possible barriers to accomplishing a task that he or she already knows how to do [25]. Self-efficacy theory [13] posits that a self-efficacy level construct may play a central role in the initiation of a behavior (e.g., engaging in a recommended amount of weekly physical activity in week 1) while a self-regulatory efficacy may play a central role in the maintenance of a behavior (e.g., engaging in a recommended amount of weekly physical activity for six consecutive months). The importance of both a self-efficacy level construct and a self-regulatory efficacy construct has been demonstrated in exercise contexts [35, 36]. The operational definition for self-efficacy to regulate physical activity in the FFW effectiveness trial is the degree to which an individual perceives that they have the capability to overcome possible barriers to engagement in a recommended amount of weekly physical activity for health.

Approximately 9 hundred participants will be enrolled in this study. Participants will be recruited through the SurveyHealth (http://​www.​surveyhealthcare​.​com/​) company’s panel recruitment based on the eligibility criteria for this study. Partnering with a panel recruitment company is consistent with recruitment in previous research on FFW [26, 27] and with a movement toward larger and smarter approaches that promote physical activity [9].

There are five eligibility criteria for participation in this study. The first eligibility criterion is the ability to access the online intervention. This criterion will be assessed by asking each participant to confirm that he or she will have access to a technological device (e.g., smartphone) that can access the online intervention via a web browser throughout the study. The second eligibility criterion is living in the USA. This criterion will be assessed by asking each participant to confirm that he or she is living in the USA. A justification for this criterion is evidence for differences in the prevalence of physical activity in adults by country [7]. The third eligibility criterion is being an adult between 18 and 64 years old. This criterion will be assessed by asking each participant the year in which he or she was born. A justification for this criterion is that our target population is adults (i.e., 18–64 years) and not older adults (i.e., 65 years and above) based on evidence-based age groupings for global recommendations on physical activity for health [38]. The fourth eligibility criterion is a body mass index (BMI) ≥ 25.00 kg/m². Our BMI criterion includes the overweight category (i.e., 25.00–29.99 kg/m²) as well as the obese category (i.e., ≥ 30.00 kg/m²) consistent with many interventions promoting physical activity in obese populations [5]. This criterion will be assessed by asking each participant his or her height and weight. A justification for this criterion is trying to promote physical activity in a population where very few individuals may meet public health guidelines for physical activity [4]. The final eligibility criterion is the absence of simultaneous enrollment in another intervention program promoting either well-being or physical activity. This criterion will be assessed by requiring each participant to confirm that he or she will not be enrolled in another formal intervention program promoting either well-being or physical activity during the FFW study period. A justification for this criterion is a reduction in the likelihood of confounding the results from the current study with results that may be due to enrollment in other programs. The short form of the international physical activity questionnaire (IPAQ; [39, 40]) and the physical activity readiness questionnaire [41] will also be included in the screener and responses to these instruments may be used for exploratory purposes.

Eligible participants will be randomly assigned to the intervention (i.e., FFW) or the wait-list control (i.e., UC) group via software code that is written to accomplish equal (i.e., balanced) allocations to the FFW and UC groups. Research staff will be blinded to participant group assignment. The outcomes assessor will be blinded to participant group assignment.

Instruments to measure demographic information, anthropometric characteristics, self-efficacy, well-being and physical activity will be included in the battery. Data on literature-based demographic covariates of multidimensional well-being [43] will be collected at Time 1 and will include participant gender, age, race, education-level, marital status, and annual income. Zip code data will be collected as a proxy for a host of built environment factors that may be related to an individual’s level of physical activity [10]. Anthropometric data will be collected at all three time points and will be assessed by asking each participant his or her height and weight. Demographic, zip code, and anthropometric variables are collectively referred to as covariates from this point forward.

The probability of falsely rejecting a true null hypothesis (i.e., α) will be set to equal .05. This approach is consistent with a majority of RCTs with multiple outcomes, where a downward adjustment to α generally has not been applied to maximize statistical power (in the event that a null hypothesis is false) [66]. To address, however, a reasonable concern with the possibility of an inflated α (in the event that a null hypothesis is true), when statistical significance is observed for a focal parameter we will emphasize an effect size estimate; provide a 95% confidence interval for the effect size estimate; and, explicitly note that caution should be exercised with regard to observed statistical significance until confirmatory studies become available in the future [66].

Effect size will be estimated in each model by dividing the mean difference by the square root of the variance pooled across the UC and FFW groups. In Model A this effect size estimate is equivalent to Cohen’s d [67]. In Model B and Model C this effect size estimate can be regarded as an extension of Cohen’s d to a latent class framework [57]. In an effort to gain some textual parsimony, we will denote the estimated effect size in each model as Cohen’s d hereto forward. Similarly, we will use heuristics put forth by Cohen [67] to describe the magnitude of the absolute value of Cohen’s d: 0.20, (small), 0.50 (medium) and 0.80 (large).

Model A will impose a regression model for each proposed outcome with measures taken at Time 2 and Time 3 as the dependent variables. Covariates, the outcome at Time 1, and group allocation (FFW =1, UC = 0) will serve as predictors of the outcome at Time 2 and Time 3 and these regression coefficients will be estimated freely. Intercepts for the outcome at Time 2 and Time 3 will be estimated freely. Covariance between the error terms for the outcome at Time 2 and Time 3 will be estimated freely. The focal parameters will be the direct effects from group allocation to the outcome at Time 2 (γ_Time2) and Time 3 (γ_Time3). A positive focal parameter value will convey that the FFW group has a higher adjusted mean for the outcome as compared to the UC group.

Model B will impose a latent class (with two classes) regression model with CACE estimation for each proposed outcome with measures taken at Time 2 and Time 3 as the dependent variables. The first class (i.e., Class 1) will be conceptualized as never-takers. The second class (i.e., Class 2) will be conceptualized as compliers. A dichotomous indicator of latent class (where 0 = non-compliers in the FFW group, 1 = compliers in the FFW group, and a missing value for participants in the UC group) will be generated. Compliance classification, modeled as a categorical latent variable, will be regressed on covariates. Covariates, the outcome at Time 1 and group allocation will serve as predictors of the outcome at Time 2 and Time 3 and these regression coefficients will be freely estimated in Class 1 and Class 2. The two direct effects from group allocation to the outcome at Time 2 and Time 3 will be fixed to 0 in Class 1 (i.e., the exclusion restrictions: γ_ntTime2 = γ_ntTime3 = 0), and will be estimated freely in Class 2 (i.e., γ_cTime2, γ_cTime3). The intercepts for the outcome at Time 2 and Time 3 will be estimated freely in each class. Covariance between the error terms for the outcome at Time 2 and Time 3 will be estimated freely in each class. The focal parameters will be the direct effects from group allocation to the outcome at Time 2 and Time 3 in Class 2 (i.e., γ_cTime2, γ_cTime3).

A positive focal parameter value will convey that compliers in the FFW group had a higher adjusted mean for the outcome as compared to potential compliers in the UC group.

Model C will estimate each parameter estimated in Model B while relaxing the exclusion restriction (i.e., freely estimate γ_ntTime2 and γ_ntTime3), making Model B nested within Model C. The change in the likelihood ratio χ² (robust) test \( , \Delta {\chi}_R^2 \)_, will formally compare the fit of these nested models. There is a substantive and a methodological rationale for evaluating the plausibility of the exclusion restriction assumption in the FFW online behavioral intervention. From a substantive standpoint the researchers may expect, based on results from the 2015 FFW efficacy trial [1], that some of the participants allocated to the FFW group may engage with the intervention at a level that yields a FFW engagement score greater than 0 (i.e., no engagement) but less than 21 (i.e., full participation). From a methodological standpoint it is important to note that it is well-known that the estimate of γ_c can be biased when the true γ_nt effect is not zero but is forced to equal zero, particularly when compliance with the intervention is less than high [45]. Therefore, the focal parameters in this model will be both the γ_cTime2 and γ_cTime3 effects and the γ_ntTime2 and γ_ntTime3 effects.

The probability of rejecting a truly false null hypothesis for every focal parameter (i.e., γ_cTime2, γ_cTime3) was estimated (N = 900) in Mplus 8.0 using Monte Carlo methods [68] under the assumption that engagement is likely to be less than 50% [1]. For each of the focal parameters in Model B the population parameter value equaled a value that corresponded to either a small (i.e., d = 0.20), moderate (i.e., d = 0.50), or large (i.e., d = 0.80) positive effect. A range of effect sizes were modeled consistent with relevant recommendations in exercise science [69]. The population model assumed a engagement rate of 25, 45%, or 65% based upon results observed in the 2015 FFW efficacy trial [1]. In the 2015 FFW efficacy trial engagement ranged from 15.6 to 54.9% with a mean of 31.6% across dimensions of well-being [1]. Our simulations assume a ~ 10% increase in engagement, which we believe may result from the new remuneration plan. Missing data (i.e., 35% at Time 2 and 40% at Time 3) were modeled based upon results observed in the 2015 FFW efficacy trial [1]. The quantity of replications requested equaled 10,000. Each replication was originally drawn from a conditionally multivariate normal distribution.

Table 2 provides the power estimation for γ_cTime2 and γ_cTime3 at Time 2 and Time 3. Power estimation for a small effect ranged from .30 (25% engagement) to .74 (65% engagement). Power estimation for a moderate effect ranged from .95 (25% engagement) to 1.00 (at least 45% engagement). Power estimation for a large effect equaled 1.00. We conclude that we are likely to have low to moderate power for small effects (depending on engagement level) and high power for moderate and large effects. Budgetary constraints preclude enrollment of more than approximately 900 participants.