The BetterYou Impact: Randomized Control Trial Results with Stanford University

by | May 15, 2023

RCT

Russell Gerber, Sean Higgins, Andrew Boucher, Adreon Morgan

1. Introduction 

1.1 Background and context 

In recent years, there has been a growing interest in wellness (though there are many possible definitions, wellness typically encompasses holistic lifestyle and behavioral factors that enhance an individual’s physical, mental, and emotional life dimensions that impact their work performance, efficiency, life satisfaction, and strength of social relationships) as a support to individuals that extends beyond the absence of inhibiting medical conditions. Attendant with this rise in emphasis on wellness, various technological tools and interventions have emerged, aiming to enhance individuals’ physical and mental health by offering personalized recommendations and monitoring progress in diverse wellness goals and aspects. These applications often utilize behavior change techniques, such as positive reinforcement and goal-setting, to encourage users to adopt and maintain healthy behaviors. 

One such app is BetterYou, which utilizes passive tracking to monitor various aspects of users’ wellness, including sleep, steps, learning, reading, meditation, and spirituality. BetterYou also provides automated coaching tips to help users improve their habits and achieve their health and wellness goals. 

While wellness apps have gained popularity in recent years, their efficacy is not always proven through scientific evidence. Randomized controlled trials are necessary to rigorously evaluate the effectiveness of such apps and to determine whether they have the potential to improve health outcomes. 

This paper presents the results of a study evaluating the efficacy of BetterYou in improving users’ health behaviors. Conducted in conjunction with Stanford University, the study consists of a randomized trial of BetterYou against a placebo app. Theoretical grounding for the app comes from the behavior modification model of Stanford University Professor B.J. Fogg,1 and thus this model also provides initial hypotheses and testable implications as to the function of the app. The results of this study have the potential to inform the development of future wellness apps and to guide clinical practice in promoting healthy habits. 

1.2 Overview of the problem statement 

The prevalence of health conditions and diseases that can potentially be ameliorated by behavioral interventions, such as exercise, physical activity, sleep, and social connection, has risen significantly in recent years. These lifestyle-related diseases include obesity, diabetes, cardiovascular diseases, mental health disorders, even certain types of cancer such as colon and breast cancer which are associated with chronically low activity levels.2 

Research has shown that adopting healthy behaviors, such as getting enough sleep and engaging in regular physical activity, can help prevent and manage these chronic conditions.3 However, many people struggle to maintain these behaviors over time, often due to a lack of motivation, knowledge, or support. 

1https://behaviordesign.stanford.edu/people/bj-fogg 

2Obesity: Prevalence of obesity has increased over 50% in approximately the past two decades alone, and currently sits at over 40% and is not contained within broadly defined demographic subsets https: //www.cdc.gov/nchs/products/databriefs/db360.htm. Obesity is a major risk factor for various health issues, such as heart disease, stroke, and type 2 diabetes. 

Diabetes: The prevalence of diabetes, particularly type 2 diabetes, has also increased dramatically over the past several decades. Diabetes Research Institute estimates prevalence at 37.3 million Americans living with the disease https://diabetesresearch.org/diabetes-statistics. 

Cardiovascular diseases: Cardiovascular diseases, including heart disease and stroke, are the leading cause of death in the US https://www.cdc.gov/heartdisease/facts.htm. 

Mental health disorders: Prevalence of mental illness in the US rose to greater than 1 in 5 overall in 2021 https://www.nimh.nih.gov/health/statistics/mental-illness. 

Cancer: Some cancers, such as colon https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3915109/ and breast https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8221371/ cancer, are associated with sedentary lifestyles and poor health behaviors . 

Engaging in regular physical activity, maintaining a healthy diet, and getting enough sleep can help lower the risk of developing these cancers. Several factors contribute to this increase, including sedentary lifestyles, poor dietary habits, stress, and social isolation. 

Physical activity, regular exercise, healthy sleep patterns, and healthy diet can help reduce the risks of obesity, diabetes, cardiovascular diseases, certain cancers, and mental illness. Additionally, social connection has been shown to improve mental health and well-being, reducing the risk of developing mental health disorders and alleviating symptoms in those who have them. 

3See https://pubmed.ncbi.nlm.nih.gov/21300732/ for a meta-analysis of the relationship between sleep duration and cardiovascular outcomes; https://www.nature.com/articles/s41598-021-03997-z for a meta-analysis of the relationship between sleep duration and all-cause mortality; https: //jamanetwork.com/journals/jamainternalmedicine/article-abstract/2212267 and https://www. thelancet.com/journals/lanpub/article/PIIS2468-2667(21)00302-9/fulltext for an analysis of the relationship between physical activity and all-cause mortality; and https://www.nature.com/articles/ s41591-022-02012-w for an analysis of the relationship between daily steps and chronic disease mitigation. 

Wellness apps, such as BetterYou, have emerged as a potential behaviorally-based solution to this problem. These apps aim to motivate and support users in adopting and maintaining healthy behaviors through personalized recommendations and tracking. However, there is a lack of scientific evidence grounding the efficacy of these apps in a well-established theory of behavioral change and habit formation. 

The purpose of this study is to evaluate the effectiveness of BetterYou in improving two key health-supporting behaviors: sleep and physical activity (measured as steps). The study focuses on these behaviors as they are widely recognized as important contributors to overall health and well-being. 

By evaluating the efficacy of BetterYou in improving these behaviors, this study seeks to contribute to the scientific understanding of the potential of wellness apps to promote healthy habits. The results of this study may also inform the development of future wellness apps and guide clinical practice in promoting healthy behaviors. 

1.3 Purpose and objectives of the study 

In addition to studying health-promoting behavioral change, this study also seeks to evaluate the predictions of the behavior modification model of B.J. Fogg, which provides the theoretical basis for the app. 

The objectives of this study are: 

• To evaluate the efficacy of BetterYou in improving users’ physical activity levels, as measured by daily step count. 

• To evaluate the efficacy of BetterYou in improving users’ sleep quality and duration. 

• To evaluate the extent to which the predictions of the Fogg model are supported by the data collected in this study. 

The results of this study have the potential to provide valuable insights into the effectiveness of wellness apps in promoting healthy habits and improving health outcomes. They may also inform the development of future wellness apps and guide clinical practice in promoting healthy behaviors. 

2. Theoretical framework 

The Fogg model proposes that behavior change is most likely to occur when three factors converge: motivation, ability, and a prompt or trigger. According to the model, behavior change is facilitated by higher levels of motivation and ability (greater ease or lower barrier to undertaking an action), as well as a prompt that triggers the behavior. 

In the context of BetterYou, the app is designed to leverage these three components to encourage users to adopt and maintain healthy behaviors. The app provides personalized recommendations, tracking, and reporting to enhance users’ motivation and ability, while also providing prompts or triggers to encourage behavior change. 

In order to assess the effectiveness of the app in promoting behavior change, all participants in the study completed an intake survey designed to assess their motivation and ability with regard to physical activity (measured by steps) and sleep. By assessing participants’ motivation and ability at the outset of the study, we can determine whether the Fogg model accurately predicts behavior levels in this context. We can also evaluate the extent to which the app is effective in promoting behavior change, as well as whether these changes are consistent with predictions of the Fogg model. 

3. Methodology 

3.1 Study Design: 

This study was a single-blind randomized controlled trial conducted in conjunction with Stanford University. Participants were recruited from the Stanford undergraduate population and were screened for their ability to participate for the full six weeks of the trial. Participants were randomly assigned to either the treatment group, who were instructed to download and use the BetterYou app, or the control group, who were instructed to download a placebo app that passively tracked their steps and sleep but did not provide any notifications, coaching, challenges, progress highlights, or other features provided by BetterYou. 

3.2 Participants: 

After attrition, the study consisted of 74 undergraduate students (40 Treatment, 34 Control) from Stanford University.4 Participants were required to have an iPhone, be over the age of 18, and be able to commit to participating in the study for the full six-week duration. 

3.3 Intervention: 

Participants in the treatment group were instructed to download and use the BetterYou app. The app provides personalized recommendations and tracking to enhance users’ motivation and ability, while also providing prompts or triggers to encourage behavior change. The app includes features such as coaching tips, challenges, progress highlights, and notifications to encourage users to maintain healthy habits. 

Control group participants were instructed to download a placebo app that passively tracked their steps and sleep but did not provide any notifications, coaching, challenges, progress highlights, or other features provided by BetterYou. This app was designed to control for any placebo effects that might arise from participants simply tracking their activity levels and sleep. 

3.4 Data Collection: 

Participants completed an intake survey at the beginning of the study that included questions designed to assess motivation and ability regarding physical activity (measured by steps) and sleep. Data on sleep and activity levels were directly collected from participants’ mobile devices. 

3.5 Data Analysis: 

Data were analyzed using descriptive statistics to assess the results of the intake survey, and compare the two groups’ initial physical activity levels and sleep quality and duration. Additionally, regression analysis was conducted to assess the relationship between motivation and ability, as measured by the intake survey, and physical activity levels and sleep quality and duration. Treatment effects were analyzed via Difference-in-Differences regression to determine whether any significant divergence in the evolution of steps and sleep between the treatment and control groups developed. Finally, treatment effects were analyzed for the presence of mediating and moderating effects arising from differing initial motivation and ability levels. 

4. Results 

4.1 Descriptive statistics 

4.1.1 Intake Survey 

The intake survey consisted of 21 questions that assessed demographic factors (age, year of school, gender, race/ethnicity), and motivation and ability factors related to sleep and physical activity (steps). Examples of motivation-relevant questions included: “How important to you is getting good sleep?” and “In general – how willing are you to change or modify your habits?” Examples of ability-relevant questions included: “Thinking about your typical week, how difficult would it be to sleep an extra half hour a night?” and “What obstacles to walking/light exercise do you face, if any?” Because completion of the intake survey was required for admission into the study, the completion rate for the survey was 100%, with a minimum completion rate of 93.2% (69/74) per item. 

Responses were generally well-distributed across the support of possible options, with participants showing awareness of the importance of sleep and exercise for overall well being. The least-responded items were those that indicated the student placed either no value on steps or sleep, or those that indicated the student was uninterested in wellness overall. 

Principal Component Analysis was used to reduce responses on motivation and ability questions into single-dimension factors for each goal (one factor each for steps and sleep, for motivation and ability, totaling four reduced Fogg factors).5 

4.2 Covariate Balance 

Covariate balance is a critical aspect of RCTs because it ensures that the treatment and control groups are similar across all observed characteristics, except for the treatment itself. This similarity allows researchers to isolate the causal effect of the treatment and make more accurate inferences about the treatment’s impact on the outcome of interest. 

When covariate balance is achieved, it means that the distribution of potential confounding factors, such as demographics, baseline characteristics, and other variables that might influence the outcome, is similar between the treatment and control groups. This balance helps eliminate biases and minimize the risk of confounding, which occurs when an external factor influences both the treatment assignment and the outcome. 

In the present study, we tested for the balance of covariates by evaluating the responses to our survey items. In only two instances did we find slight statistically significant differences (after randomization) between our treatment and control groups. 

Figure 1: Covariate Balance – P-Value of Mean Difference by Treatment Condition 

5The first principal component was selected to represent the factor for each of motivation and ability respectively. The PCA was conducted by first reducing the covariate set to only contain only those elements directly related to each question domain before input into the training step (i.e. conducting the Singular Value Decomposition and obtaining the component weights) of the PCA algorithm. Factors were normalized to a support of [0, 1] 

4.3 Treatment Effects: D-i-D Regression 

Difference-in-Differences (D-i-D) is a causal treatment design commonly used in observational studies but can also be applied in RCT settings like the present study. The primary idea behind D-i-D is to estimate the treatment effect by comparing the changes in outcomes between the treatment and control groups over time, before and after the intervention. 

To apply D-i-D in an RCT setting like this one, we collected data on the outcomes of interest for both groups at two points in time: before (pre-treatment) and after (post treatment) the intervention (in this case, BetterYou). This pre/post data collection allows for the estimation of treatment effects while accounting for any time-invariant unobserved factors that may affect the outcome. 

The key assumption of the D-i-D design is that, in the absence of treatment, the treatment and control groups would follow parallel trends over time. This means that any differences in activity that develop between the two groups can be attributed to the effect of the intervention, BetterYou. 

The random assignment of participants helps ensure that potential confounding factors are balanced between the treatment and control groups, strengthening the causal inference made using the D-i-D approach. 

The regression equation is presented below: 

yi,t = βXi,t + δ Ti + ϵi,t 

In this equation, the data elements are yi,t the level of the outcome, either steps or sleep, X represents the set of motivation and ability covariates gleaned from the initial screening survey and reduced through the factor analysis,6 as well as other contextual factors such as day of the week, and ϵ is the residual term. T represents the indicator variable for the treatment condition, and can take on the values 1 (if the individual received the treatment, BetterYou) or 0 (if the individual received the placebo app). The coefficients are β, which measure and account for the influence of the motivation and ability covariate factors on steps and sleep – that is, that control for the difference in habitual activity organically arrived-at between members, and the coefficient of interest in this study is δ, the difference in steps or sleep that is due to receiving the treatment, BetterYou.7 

The main results of our study are captured in the coefficients presented in the table below. 

Participants in the treatment group achieved, on average, an additional 32 minutes of sleep per night, and walked an additional 984 steps per day. 

6This factor reduction is necessary so as not to exactly identify each individual in the survey, which would absorb the treatment indicator. 

7We will also refer to δ as the treatment effect. 

Table 1: OLS Regression Coefficients: Main Specification 

Sleep Steps 

Coeff. P-Value Coeff. P-Value 

Treatment 31.60 <0.001 984.08 <0.001 

Activity Goal 0.9653 0.022 0.7048 <0.001 

Ability Factor 29.0736 <0.001 2979.97 <0.001 

Motivation Factor 9.5311 0.072 1873.08 <0.001 

Controls Yes Yes 

Observations 2794 2575 

R-squared 0.425 0.203 

Note: ”Activity Goal” is Target Sleep Minutes for sleep 

regressions, Target Steps for steps regressions. 

4.3.1 Fogg Model Evaluation 

The Fogg Behavior Modification Model posits that behavior change is most likely to occur when three factors converge: motivation, ability, and a prompt or trigger. In the context of this study, we derived motivation and ability factors from the intake survey completed by participants, focusing on their attitudes and perceived capabilities regarding physical activity (steps) and sleep. Evaluating the basic prediction of the Fogg Model, we aimed to determine whether individuals with higher motivation and ability would exhibit greater levels of action in terms of steps and sleep. 

To assess this prediction, we incorporated the derived motivation and ability factors into a regression equation, with the outcome variables being the levels of physical activity and sleep duration. By examining the relationship between the motivation and ability factors and the observed action levels, we can determine whether the Fogg Model’s predictions hold true in the context of our study. If the regression results support the model, we would expect to see positive associations between the motivation and ability factors and the corresponding action levels, indicating that individuals with higher motivation and ability engage in healthier behaviors, such as taking more steps and experiencing better sleep quality and duration. This analysis not only helps to validate the Fogg Model’s theo retical foundation but also provides insights into the effectiveness of behavior modification strategies that leverage motivation and ability to promote healthy habits. 

For this regression, the equation was specified to be agnostic to the Treatment condition: yi,t = βXi,t + γM M + γA A + γMA M ∗ A + ϵi,t 

Here, M and A represent the motivation and ability factors,8 and M*A represents the interaction of motivation and ability. This specification allows the regression to capture linear and combined effects of motivation and ability.9 The outcome of the regression is presented below in a table of coefficients, as well as in the form of a heat map, which was created by calculating the value of the sum γM M + γA A + γMA M ∗ A at the different combinations of the motivation and ability factors. 

Table 2: OLS Regression Coefficients 

Fogg Model Initial Behavior Levels Specification 

Sleep Steps 

Coeff. P-Value Coeff. P-Value 

Ability Factor 9.6434 0.592 11334 <0.001 

Motivation Factor 2.3475 0.869 9209.36 <0.001 

Motivation*Ability 87.13 <0.001 16118 <0.001 

Controls Yes Yes 

Observations 2794 2575 

R-squared 0.036 0.108 

Note: ”Activity Goal” is Target Sleep Minutes for sleep 

regressions, Target Steps for steps regressions. 

4.4 Treatment effects moderation regressions using Motivation and Ability factors 

The Fogg Behavior Modification Model emphasizes the crucial role of well-timed and effective prompts in achieving behavioral change, particularly for individuals with lower motivation and ability levels. In the context of the model, better prompts have a more significant impact on those with lower motivation and ability factors, as they help catch these individuals at the opportune moments when their motivation and ability rise above the action line, leading to the desired behavior change. 

For individuals with higher average motivation and ability levels, they already spend more time above the action line, providing more opportunities for prompts to lead to action. As a result, the difference between great and poor prompting matters less for 

8These factors were aggregated together within Xi,t. in the prior regression equation. 9The interaction term is simply the multiplication of the motivation and ability factors together, and allows the regression equation to capture nonlinearities in the relationship between these factors and the level of activity. It allows the regression to provide evidence in response to questions such as: ’Does additional motivation matter more for activity (steps/sleep) when ability is high versus when it is low?” 

Figure 2: Heatmap of Initial Daily Sleep Minutes Contribution Due to Motivation and Ability Factors 

these individuals. In contrast, those with lower motivation and ability levels benefit more significantly from well-timed and targeted prompts, as they less frequently find themselves above the action line. Effective prompting can help them capitalize on those rare moments when motivation and ability converge, thereby facilitating behavior change. 

In light of this theoretical underpinning, the critical importance of effective prompting becomes apparent, especially for those struggling with low motivation or those whose busy lives have left them with little time for health and wellness. While services like BetterYou do not exclude highly motivated individuals, their greatest impact lies in providing support for those in greater need. By evaluating the efficacy of BetterYou and its use of prompts in the context of the Fogg Model, we can better understand how well-timed and effective prompting can lead to significant improvements in health behaviors, particularly for individuals with lower motivation and ability levels. 

We can assess the ability of BetterYou to provide such effective nudges statistically by analyzing motivation and ability as moderators of the treatment effect. Moderating effects occur when the relationship between the treatment and the outcome varies depending on the level of a third variable. 

To assess moderating effects in an econometric setting, researchers typically include an interaction term between the treatment indicator variable and the potential moderator variable in the regression model.footnote: interaction term simply means including among the explanatory variables an additional variable created by multiplying the treatment and the potential moderator together. The interaction term captures the extent to which the treatment effect varies depending on the level of the moderator variable. If the interaction term is statistically significant, it suggests that the relationship between the treatment and 

Figure 3: Heatmap of Initial Daily Steps Contribution Due to Motivation and Ability Factors 

the outcome is moderated by the third variable. In other words, the treatment effect is not constant across different levels of the moderator variable. 

For example, in the context of this RCT, the treatment variable is the indicator for receiving the BetterYou app, and a moderator variable would be the individual’s initial motivation level. The interaction term between the treatment indicator and the motivation variable would then capture how the treatment effect varies depending on the level of motivation. With the moderator included, the regression equation now becomes: 

yi,t = βXi,t+γM M +γA A+γMA M ∗A+δbase Ti+δAA∗Ti+δMM ∗Ti+δM∗AM ∗A∗Ti+ϵi,t 

The outcome of the regression is presented below as a heat map, which was created by evaluating the treatment effect coefficients (δbase + δAA ∗ Ti + δMM + δM∗AM ∗ A) at the different combinations of the motivation and ability factors. 

As you can see, the heat maps flip their orientation from those representing the relationship between motivation, ability, and initial activity levels: when moderating the treatment effect of BetterYou’s prompts and coaching, lower motivation and ability levels correspond to higher treatment effects. This inversion is precisely as the Fogg model predicts under the assumption that BetterYou’s prompts are timely and effective. These findings provide robust evidence for the effectiveness of the BetterYou app in promoting physical activity and sleep. 

Figure 4: Heatmap of Initial Daily Sleep Minutes Contribution Due to Motivation and Ability Factors 

5. Discussion 

5.1 Summary of key findings 

In this white paper, we present the results of a randomized controlled trial (RCT) conducted by BetterYou in collaboration with Stanford University to assess the effectiveness of our digital coaching app in promoting healthier lifestyles. Our study included a diverse sample of participants, representing a broad range of knowledge, interest, motivation, ability, and attitudes towards behavioral change. 

The covariate balance between the treatment and control groups was satisfactory, with only two variables (Gender and Women’s Health Interest) found to be imbalanced. However, these variables were not expected to significantly influence the results. Participants who were randomly assigned to the treatment group experienced a substantial increase in their physical activity levels, walking an average of 984 additional steps per day. Additionally, they slept nearly 32 more minutes each night compared to the control group. When evaluated against the Fogg model, our results align with the predicted patterns of an effective, timely, and accurate behavioral intervention tool. 

5.2 Comparison with prior research 

Our findings contribute to the growing body of evidence supporting the use of digital health interventions to promote behavioral change. The positive effects of the BetterYou app on physical activity and sleep duration are generally consistent with prior research demonstrating the benefits of mobile health technologies in promoting healthier lifestyles.

Figure 5: Heatmap of Initial Daily Steps Contribution Due to Motivation and Ability Factors 

However, comparing these findings to other studies of mobile app-based interventions for physical activity and sleep reveals some parallels and differences. A systematic review and meta-analysis conducted by Direito et al. (2017)[1] found that mHealth technologies employing behavior change techniques effectively increased physical activity levels. Similarly, a scoping review by Agher et al. (2020)[2] suggested that connected health interventions improved adherence to cardiovascular disease prevention measures. However, it is worth noting that the magnitude of improvements observed in our study appears to be more substantial than the effects reported in some other studies[3][4]. 

A meta-analysis by Iribarren et al. (2021)[5] examined the effectiveness of mobile apps in promoting health and managing disease, finding that app interventions yielded modest improvements in physical activity and sleep quality. Another study by Martin et al. (2015)[6] investigated an automated mHealth intervention for physical activity promotion, yielding positive results in terms of daily step count. In a randomized trial by King et al. (2016)[7], three motivationally targeted mobile device applications were found to be effective in promoting initial physical activity among midlife and older adults. However, a study by Finkelstein et al. (2016)[8] reported mixed results regarding the effectiveness of activity trackers with and without incentives to increase physical activity. 

5.3 Limitations of the study 

5.3.1 Nudges 

A limitation of our study is that we cannot examine the effect of an individual nudge. The AI system employed by BetterYou targets opportune times to nudge people selectively toward their goals, generating a correlation between activity levels and receipt of nudges that constitutes an endogenous relationship. As a result, the causal effect of individual nudges cannot be extracted using this research design. 

5.4 Future research 

Future research should focus on addressing the limitations of the current study by exploring the causal impact of individual nudges and further investigating the role of motivation and ability in driving behavioral change. Additionally, longer-term studies are needed to assess the sustainability of the observed effects and the potential for digital health interventions like BetterYou to contribute to lasting improvements in health and well-being. 

References 

[1] Direito A, Carra¸ca E, Rawstorn J, Whittaker R, Maddison R. mHealth Technologies to Influence Physical Activity and Sedentary Behaviors: Behavior Change Techniques, Systematic Review and Meta-Analysis of Randomized Controlled Trials. Ann Behav Med. 2017 Apr;51(2):226-239. doi: 10.1007/s12160-016-9846-0. PMID: 27757789. 

[2] Agher D, Sedki K, Tsopra R, Despres S, Jaulent MC. Influence of Connected Health Interventions for Adherence to Cardiovascular Disease Prevention: A Scoping Review. Appl Clin Inform. 2020 Aug;11(4):544-555. doi: 10.1055/s-0040-1715649. Epub 2020 Aug 19. PMID: 32814353; PMCID: PMC7438176. 

[3] Zahrt OH, Evans K, Murnane E, Santoro E, Baiocchi M, Landay J, Delp S, Crum A. Effects of Wearable Fitness Trackers and Activity Adequacy Mindsets on Affect, Behavior, and Health: Longitudinal Randomized Controlled Trial. J Med Internet Res. 2023 Jan 25;25:e40529. doi: 10.2196/40529. PMID: 36696172; PMCID: PMC9909519. 

[4] Zahrt OH, Evans K, Murnane E, Santoro E, Baiocchi M, Landay J, Delp S, Crum A. Effects of Wearable Fitness Trackers and Activity Adequacy Mindsets on Affect, Behavior, and Health: Longitudinal Randomized Controlled Trial. J Med Internet Res. 2023 Jan 25;25:e40529. doi: 10.2196/40529. PMID: 36696172; PMCID: PMC9909519. 

[5] Iribarren S, Akande T, Kamp K, Barry D, Kader Y, Suelzer E. Effectiveness of Mo bile Apps to Promote Health and Manage Disease: Systematic Review and Meta analysis of Randomized Controlled Trials. JMIR Mhealth Uhealth 2021;9(1):e21563. URL: https://mhealth.jmir.org/2021/1/e21563. DOI: 10.2196/21563. 

[6] Martin SS, Feldman DI, Blumenthal RS, Jones SR, Post WS, McKibben RA, Michos ED, Ndumele CE, Ratchford EV, Coresh J, Blaha MJ. mActive: A Randomized Clinical Trial of an Automated mHealth Intervention for Physical Activity Promotion. J Am Heart Assoc. 2015 Nov 9;4(11):e002239. doi: 10.1161/JAHA.115.002239. PMID: 26553211; PMCID: PMC4845232. 

[7] King AC, Hekler EB, Grieco LA, Winter SJ, Sheats JL, Buman MP, Banerjee B, Robinson TN, Cirimele J. Effects of Three Motivationally Targeted Mobile Device Applications on Initial Physical Activity and Sedentary Behavior Change in Midlife and Older Adults: A Randomized Trial. PLoS One. 2016 Jun 28;11 

[8] Finkelstein EA, Haaland BA, Bilger M, Sahasranaman A, Sloan RA, Nang EEK, Evenson KR. Effectiveness of activity trackers with and without incentives to in crease physical activity (TRIPPA): a randomised controlled trial. Lancet Diabetes Endocrinol. 2016 Dec;4(12):983-995. doi: 10.1016/S2213-8587(16)30284-4. Epub 2016 Oct 4. PMID: 27717766.

Interested in learning more about how we can help your employee population improve their steps and sleep while reducing burnout?

Related Posts