A comparative effectiveness trial (CET) is being conducted to compare the effectiveness of the MobileMen app at increasing PA in African American men compared to Stridekick, a similar fitness app. The participants are assigned to one of the 2 arms: intervention (MobileMen) or comparator (Stridekick). The participants will use the assigned app on their personal smartphones for the duration of the 6-month trial. This study was registered with ClinicalTrials.gov (NCT05621044).

This protocol was approved by the Pennington Biomedical Research Center institutional review board on July 31, 2024 (approval 2021-052). All collected data will be deidentified before the analyses. All participants will complete an informed consent form and will receive US $200 compensation at the completion of their participation.

MobileMen is a mobile app tailored to the needs of African American men who aspire to increase their PA engagement [60]. MobileMen was developed using an extensive and iterative process described elsewhere [40]. The development of the MobileMen app was completed in July 2024 and is available for use on both iOS and Android platforms. See Table 1 for features of the MobileMen app.

The Stridekick app is intended to increase PA in users through a variety of functions. Stridekick is a free app available on iOS and Android platforms. See Table 1 for Stridekick features.

Eligibility criteria include self-identification as an African American male, a resting systolic blood pressure ≤159 mm Hg and a diastolic blood pressure ≤99 mm Hg (to reduce risk of exercise-induced cardiovascular events), free of significant medical problems that would impact their ability to engage in aerobic and/or resistance training, smartphone ownership, age > 30 years, and “low active” at baseline (defined as having an average daily accelerometer-assessed step count of <7500 steps) [61].

Community-based efforts, including advertising at churches, health fairs, Young Men’s Christian Association, fraternities, social media posts, and advertisements on paid, owned, and earned media, are being used to recruit study participants. Potential participants undergo a telephone screening to assess initial eligibility.

Men who meet the initial inclusion criteria are invited for an in-person or remote orientation visit with study staff. Study staff review the study protocol and facilitate an informed consent process. During this study visit, participants complete several baseline assessments, including a sociodemographic questionnaire and height, weight, pulse, and resting blood pressure measurements. We calculate BMI by using self-reported weight and height (kg/m²). Health-related quality of life is assessed using the 36-item short-form health survey, a widely used validated measure with strong internal consistency (Cronbach α > 0.80 across most domains) and robust evidence of construct and criterion validity and has been validated in African Americans [62,63]. To assess PA and dietary habits, participants complete the International Physical Activity Questionnaire and National Cancer Institute’s Automated Self-Administered 24-hour Dietary Assessment Tool, respectively [64,65]. Self-efficacy for fruit and vegetable consumption is measured with 7 items that measure participants’ confidence in their ability to consume fruits and vegetables in different contexts like when away from home or in a rush [66]. Self-efficacy for PA is measured using the Self-Efficacy for Exercise Scale [67]. SDT-based motivation for PA is measured using the Behavioral Regulations in Exercise Questionnaire v.3 [68,69]. This 24-item questionnaire assesses the 6 SDT subtypes of motivation, with each item scored on a 5-point Likert scale ranging from 0 (not true for me) to 4 (very true for me). The Behavioral Regulations in Exercise Questionnaire v.3 has strong internal consistency for each motivational subscale; it validly measures the different types of motivation and is predictive of exercise behavior [68,69]. Eligible participants receive a Fitbit to wear for 1 week to assess baseline PA levels.

At the start of this remote or in person visit, study staff evaluate the Fitbit wear compliance. Participants who did not wear the Fitbit for at least 3 weekdays and 1 weekend day are required to rewear the device. Participants who self-report <60 minutes of moderate to vigorous PA on the International Physical Activity Questionnaire are eligible. Eligible participants are randomized to one of the 2 conditions, the MobileMen intervention group or the Stridekick comparator group, as directed by the next assignment letter contained in a randomized sequence of sealed and numbered envelopes. Using a computerized pseudorandom number generator, the statistician determines the randomization order in advance. The study coordinator reveals the randomization after the participant is deemed eligible. Project staff assist each participant in downloading the designated (MobileMen or Stridekick) app onto their smartphone, provide an app orientation, and assist in app navigation.

At the conclusion of the 6-month intervention, participants complete all follow-up assessments (except for the Physical Activity Readiness Questionnaire for Everyone and demographic questionnaires) by using the same protocols as in study visit 1. Participants in both groups also complete additional measures such as a survey to assess their satisfaction with the mobile app and the System Usability Scale, a validated 10-item measure that provides a global view of the subjective usability for technology, to assess the app’s usability [70]. Participants also complete a brief interview with the study staff to collect feedback regarding suggestions for app improvement and users’ impressions of the app content.

A sample of 50 participants per trial arm was determined based on step count data from a recent meta-analysis [72]. The standard deviation in nondiseased populations on their step counts is 2000 to 3000 steps per day. When the sample size per group is 50, a 2-sided 95% CI for the difference of two means will extend to about 800 steps per day from the observed difference in means, assuming an SD of 2000 steps per day and the CI is based on the large sample z statistic. Doubling the sample reduces the half width of the CI to 560 steps per day, which reduces the half width by only 30% but disproportionately increases the recruitment time and difficulty. Within the 2 arms, we can estimate longer term adherence at 6 months with a width of a 2-sided 95% CI equal to 578 steps per day when the sample proportion adherent is 0.50, and the width shortens as the adherence increases.

Response frequency, proportions, and means will be calculated. Means and standard deviations, medians and interquartile ranges, distributions, frequencies, and percentages will characterize the data. Comparisons on demographics (ie, age, marital status) will be done using chi-square tests, t tests, and correlations, with 2-tailed P=.05. The changes from baseline to all postbaseline visits until week 24 in steps per day will be assessed. The model includes response data from all postbaseline visits with no imputation for missing data, as the model assumes missingness at random. The step counts at baseline are included as a fixed factor in the model. An unstructured covariance structure is assumed, and the denominator degrees of freedom is computed using the Kenward-Roger method. In case the model will not converge with the unstructured covariance structure, the heterogeneous compound symmetry and the heterogeneous Toeplitz structure is used instead (in that order). The least-squares mean estimates for the mean change from baseline to week 24 as well as the difference in the estimates at 12 weeks is displayed with their corresponding standard errors, P values, and 95% CI. The characteristics of the dropouts and/or noncompliant individuals are intensively examined to shed light on the potential weaknesses of the MobileMen app (differential dropout overall or within certain subgroups) and assumptions about missingness in the estimates. The evaluation of the app will include its impact using the pre-post data. The examination of the correlation matrices along with the mean changes will inform whether the variables appear to move similarly in both arms and how the interventions are working.

To identify possible mediating constructs, we plan to assess preintervention and postintervention SDT motivation and self-efficacy for PA. The selected mediation variables will be assessed using PROC CAUSALMED, which provides estimates of direct, indirect, and total effects. We will be able to judge if mediation is occurring, and if so, the strength and direction of the mediation effect. Since this study may not have an adequate sample size to power mediation effectively, this analysis will be primarily used to estimate the potential effect in a larger sample for future trials.

App utilization data will be examined to understand the acceptability and intensity of the interactions with MobileMen via several criteria, including but not limited to logins, notifications sent and received, number of activities inputted, badges earned, tips read, number of health parameters inputted, competitions competed in, and incentives earned. Complier average causal effect models will be used to examine the intensity of the intervention app utilization [73].

There are no PA promotion apps designed specifically for African American men. The MobileMen app seeks to fill this gap by becoming the first smartphone app that aims to support PA adoption and maintenance in African American men. We anticipate that app usage will be high in African American men given that it contains content that was requested, including the ability to track activity, nutrition information, goal setting, competitions, communication with other men, and rewards. We also anticipate that MobileMen will result in increased PA compared to men utilizing the Stridekick app. Finally, it is expected that the men utilizing the MobileMen app will show increases in the theoretical components targeted, which may serve as mediators of effects. If these findings are shown, it will provide evidence that a tailored app can be effective in promoting health behavior change in African American men. The next step would be to find ways of disseminating and implementing the app broadly through health systems or community-based organizations during which it will have the potential to improve the health of African American men nationwide.

To the best of our knowledge, MobileMen is a first of its kind intervention to promote PA in African American men. Existing literature demonstrates the importance of culturally tailored health interventions, particularly for populations that face unique structural and social barriers to health. Culturally tailored interventions have been shown to improve engagement, relevance, and outcomes in public health settings [56,57]. Despite the disproportionate burden of chronic disease among African American men, few health behavior change programs are specifically designed for this group [74]. African American men also face distinct barriers to PA related to gender role expectations and environmental constraints, making generalized PA programs less effective for this population [21]. The MobileMen app addresses this gap by providing a tailored solution grounded in the health behavior change theory and designed specifically for African American men. Additionally, African American people show high levels of engagement with mobile technology, supporting the use of mHealth tools as an accessible and culturally relevant strategy for promoting behavior change [75]. This focus on a high-engagement demographic with unique health needs sets MobileMen apart from existing health apps and underscores its potential value in fostering sustainable PA behavior change.

The CET, which is currently in progress, has several strengths, including a multidisciplinary research team with combined expertise in African American men’s health, mHealth app development, PA assessment, wearable technology use in interventions, and psychological mediators of PA behavior change. The app was designed using a community-engaged approach [40]. The primary outcome of CET, that is, change in daily step count, will be assessed using objective activity monitoring rather than self-report, which is a study strength due to the reliability and validity of objectively measured step counts [76]. Further, this study not only aims to assess the primary PA outcome (steps) but also the hypothesized mediating construct (autonomous motivation), addressing the question of how MobileMen might support PA behavior change.

However, no study is without limitations. Even with a sample of 100 participants, we are underpowered to conduct statistical mediation tests, leaving us unable to assert if behavioral changes associated with MobileMen use can be attributed to changes in potential mediators. At present, the mechanisms of action for PA behavior change or most health behavior changes are not well understood, as they have not been closely examined [77]. Therefore, this is a critical next step for PA research in this population. Further, the location from which African American men are recruited, the Southern United States, presents a limitation in the generalizability of the results of the study. As such, a larger study with a more geographically diverse sample of African American men should be considered. Further, the mobile app is only available to men who have use of a mobile phone; nonmobile phone users will need other methods for accessing PA-tailored interventions. Finally, choosing a commercially available comparator app may introduce bias. Stridekick was not created by our team, and although we vetted the app, it is possible that components that were unaccounted for may influence our results. Further, differences in the content, degree of personalization, and theoretical foundation between Stridekick and MobileMen may introduce potential confounding. Therefore, any observed effects may partially reflect these differences rather than the cultural tailoring or theoretical elements of MobileMen alone. Further, the participants are aware of which app they are assigned to, which may introduce expectation bias. In summary, future studies of MobileMen might consider a comparator app that is more closely matched in content and theoretical underpinnings while also considering blinding or other strategies to reduce expectation bias.

To date, no mobile PA and health app exists that is specifically tailored to the needs of African American men. MobileMen seeks to fill this gap with culturally tailored content and foundations in the health behavior change theory. This CET will assess its efficacy compared to a popular mHealth app, providing the research team with important data for app modifications before testing the final version in a large-scale effectiveness trial.