Mobile health (mHealth) refers to the use of mobile and wireless technology to achieve a health objective. While there is no standard for what constitutes mHealth, common features used alone or in combination include smartphone apps, SMS text messaging, and wearable monitoring devices. mHealth interventions targeting cardiovascular health are effective in supporting self-management of glycemic control [1,2], blood pressure [3], weight [4], and physical activity behavior [5]. Despite the evidence of mHealth intervention effectiveness, few studies have included Hispanic/Latino adults [6,7], and no studies focused on rural-living Hispanic/Latino adults were found [1]. In urban-living Hispanic/Latino adults with type II diabetes, mHealth improved medication adherence, increased fruit and vegetable consumption, and increased physical activity with SMS text messages 3 times a day for 3 weeks [8]. Another study reported lowered hemoglobin A_1c (HbA_1c) levels by sending SMS text messages (informational, motivational, and prompting) over 6 months [2].

Hispanic/Latino adults, a population underrepresented in mHealth intervention studies targeting cardiovascular risk, have notable cardiovascular health disparities. In the general population, type II diabetes affects 17% of Hispanic/Latino adults compared to 8% of non-Hispanic White adults [9]. Cardiovascular morbidity and mortality outcomes are worse for the Hispanic/Latino adult population with a greater risk of developing heart failure [10] and a 50% higher likelihood of death from complications related to diabetes than non-Hispanic White adults [11]. Among Hispanic/Latino adults living in rural areas, the prevalence of obesity is higher (Hispanic/Latino adults: 36%; non-Hispanic White adults: 32%), and health status is reported as low or poor (Hispanic/Latino adults: 28%; non-Hispanic White adult: 19%) [12]. Healthy diet and physical activity behaviors, which can be self-managed using mHealth, are less common among Hispanic/Latino adults, particularly among those with type II diabetes [13,14]. Clearly, the Hispanic/Latino adult population has increased cardiovascular risk that is not adequately managed.

An estimated 19% of the total Unites States population is Hispanic with the fastest growth of this population observed in rural areas [15]. In the United States, an estimated 79% of Hispanic/Latino adults and 71% of rural-living adults own a smartphone [16]. Most adults in rural areas access the internet daily (76%); however, high-speed internet access is a major concern for rural adults (24%) compared to urban adults (13%) [17]. While Hispanic/Latino adults are noted to use smartphones to text and access email [18], it remains unclear how likely they are to use smartphone apps to actually manage their health [19]. One study in rural-living Hispanic/Latino adults indicated that 81% had the willingness to use mHealth, but few (15%) were aware of how mHealth could support the self-management of chronic health conditions like hypertension and diabetes [20]. A national survey on health app use found that Hispanic/Latino adults were significantly more likely to use a health app compared to non-Hispanic adults, and the most common feature desired was the ability to communicate with a health care provider through the health app [21].

The intervention in this pilot study was based on the concept of chronic disease self-management. Specific self-management skills include decision-making (what is my priority health problem?), problem-solving (what can I do to manage this problem?), action-taking (how do I begin managing this problem?), accessing resources, and partnering with a health care provider [22]. Goal setting theory [23] and self-efficacy, a concept of social cognitive theory [24], support the mHealth intervention to build self-management skills and establish new patterns of healthy behavior (Textbox 1).

The purpose of this study was to evaluate the feasibility, usability, and acceptability of a stand-alone mHealth cardiovascular risk self-management intervention using MyFitnessPal, a compatible smart scale, and an engaged bilingual health care provider in a rural Hispanic/Latino adult population.

Given the limited number of mHealth interventions to reduce cardiovascular risk in the Hispanic/Latino adult population, it is important to understand what aspects of mHealth interventions are most engaging and usable.

A descriptive study design using quantitative and qualitative methods was used to evaluate the feasibility, usability, and acceptability of a cardiovascular risk self-management mHealth intervention conducted with rural Hispanic/Latino adults between May 2019 and June 2020. The 12-week intervention was tested using an unblinded randomized 2-group design. Both groups were asked to use MyFitnessPal to self-monitor daily steps, weight, and calories. The intervention group received support to download, initiate, and troubleshoot technology challenges with MyFitnessPal and a smart scale, while the enhanced usual care group received only a general recommendation to use MyFitnessPal to support healthy behaviors. This report details the feasibility and acceptability of the mHealth intervention and compares mHealth usability between the 2 groups. The cardiometabolic health outcomes in both groups are reported separately [26].

A sample of 70 rural-living Hispanic/Latino adults was recruited from 2 rural communities in Nebraska. Inclusion criteria were Hispanic/Latino adults (aged 19-65 years); English or Spanish speaking; living within 50 miles of either community; self-report hypertension, obesity, type II diabetes, or dyslipidemia; having a smartphone; and using Bluetooth or Wi-Fi. Exclusion criteria were current participation in a health behavior program (eg, Weight Watchers), BMI≥45 unless a clinician granted permission, or high risk for a cardiac event during fitness testing. Pregnant women were excluded from the step fitness test, BMI, and waist measurement.

In this study, a minimum sample of 54 (27 per group) was determined by setting the desired margin of error for the mean metabolic equivalents (METs), the primary outcome of cardiorespiratory fitness, at no more than 0.80. Using a standard deviation for walking METs of 2.00, a minimum sample size of 27 per group yields a margin of error of 0.75. A margin of error assumed a 95% confidence interval. With an assumed attrition rate of 30%, the total sample goal was 70 (35 per group), which is appropriate for pilot work, allowing for descriptive statistics, estimation of effect sizes, and hypothesis generation [27].

To support intervention development and delivery, a community advisory board was established. Members of the board included two Hispanic/Latino community residents and representatives of the local federally qualified health clinic, local public health department, Hispanic community center, and educational service unit. The board had 2 in-person meetings and communicated through email to advise on community resources, recruitment strategies, and the cultural relevancy of the SMS text messages that were developed for the intervention group.

To standardize the intervention the following were used: a training manual with topic checklist, set time allowances for study activities, a script for participant visits, and role-playing participant visits prior to beginning recruitment. Participants were recruited using bilingual flyers posted in the community, zip code–targeted Facebook ads, and within the local clinic. A study coordinator took calls from those interested in the study, discussed eligibility, and scheduled study visits at either the mobile clinic or the local community center. Following informed consent and baseline measures, the participants were randomized to either the enhanced usual care group or the intervention group. Both groups had visits at baseline, 12 weeks, and 24 weeks to complete paper questionnaires, collect lab tests (HbA_1c and lipids), measure biometrics (blood pressure, BMI, and waist circumference), and complete a 2-minute step test to evaluate cardiorespiratory fitness. At a 6-week booster visit, weight and blood pressure were measured. Study visits were conducted in a reserved area of a community center, a conference room in a large family practice clinic, or in a small community clinic during off-clinic hours.

A stratified sampling approach was used to select intervention group participants for a 15-minute semistructured interview on the acceptability of mHealth upon completion of the intervention. Interviewees were selected based on gender (male/female), age (under/over 40), and level of engagement during the study (high/low). Participants were asked specifically about their experience with (1) MyFitnessPal to track steps and calories, (2) the smart scale to track daily weight, and (3) twice weekly SMS text messages. The interviews were conducted face-to-face in a private area of the data collection space (conference room and clinic room) in the language preferred by the participant.

Quantitative data were examined for missingness, errors, normality, and equality of variances. Continuous variables were analyzed using means and standard deviations. Categorical variables were analyzed using frequencies and percentages. Only participants who responded to a measure at a specific time point were included in analyses; therefore, the number reported at each time point varied. Qualitative data were collected through audiorecorded interviews that were transcribed verbatim in Spanish and then translated to English for thematic analysis [29].

During enrollment (May-December 2019), 77 people expressed interest in participating, but 7 were not enrolled because they lacked a qualifying health condition. Of the 70 participants who enrolled, 45 were referred by someone already in the study, and 25 were referred by their health care provider. Table 1 details the sample demographic characteristics by group. Group differences in the demographic data were assessed with independent t tests, chi-square tests, and Fisher exact tests. No demographic variables were significantly different except for the preferred language in the study, where 77% (26/34) of the enhanced usual care group and 94% (34/36) of the intervention group preferred to complete the study in Spanish (questionnaires, informational SMS text messages, and MyFitnessPal; Fisher exact test P=.04).

The overall attrition was 28% (n=19) at 12 weeks and 54% (n=38) at 24 weeks. Attrition by group was similar at 12 weeks (enhanced usual care: n=9, 26%; intervention: n=11, 31%) and 24 weeks (enhanced usual care: n=18, 53%; intervention: n=21, 57%). Although there was missing data, this was primarily due to physical limitations with biometric testing, a mechanical limitation of blood analyzer (result too high to read), unanswered questions, and suspension of in-person data collection due to COVID-19 during the last 2 months of the study. Instead, phone calls were made to 15 participants to collect responses to the questionnaires and a self-reported weight.

Usability was higher in the intervention group at all 3 time points and increased at each time point (6, 12, and 24 weeks). Fewer participants in the enhanced usual care group completed the usability measure compared to the intervention group at all 3 time points. This was due to only collecting the Health-ITUES from those who reported downloading the free version of MyFitnessPal (Table 2). Given the small numbers in the enhanced usual care group, descriptive statistics are reported rather than a statistical test of difference between groups.

Intervention group adherence data were obtained from the MyFitnessPal premium app, which reported a daily step total, weight by date, and a daily calorie total. A dichotomous variable was created to indicate the presence or absence of captured data for each of the self-monitoring activities: daily steps, weight, and calories/meal (breakfast, lunch, and dinner). Weekly adherence for steps, weight, and calories was determined by the number of days per week of data capture. In this study, participants were adherent to mHealth self-monitoring if there was data captured ≥4 days per week for each week during the intervention (weeks 1-12) and postintervention (weeks 13-24).

Five participants completed an interview about the intervention. Three interviewees were low engagers (2 female participants and 1 male participant) and 2 were high engagers (1 female participant and 1 male participant). Three themes were identified: (1) mHealth is useful, (2) mHealth has challenges, and (3) changes may be needed for success with a future study (Textbox 2). Participants described mHealth as useful by creating “self-awareness,” “personal accountability,” “empowerment,” and “motivation” to work on their health behaviors. The Spanish setting in MyFitnessPal and Spanish SMS text messages were appropriate and understandable. One participant remarked, “it [mHealth] works not only for you but for the other members of your family.” Participants also described exploring and using other mHealth features including macronutrient information, BMI calculation, and the visual weight trend on the scale.

Participants experienced disruptions in the automatic sync function of the app, resulting in steps and weights that were not captured. These technical issues would either be ignored, managed, or reported to the technology support person during in-person meetings. The accuracy was questioned by participants. They reported the counter was not picking up steps or they may not have had their phone with them when being active. The accuracy of calorie tracking was also questioned because portion size estimation was difficult, and the food database had limited ethnic foods. Three participants noted time as a barrier to logging food into the app.

The low mHealth engagers suggested more training or orientation for those who were not as skilled with smartphone use. Additional training was also suggested for entering recipes into MyFitnessPal so that homemade food could be accurately tracked. Most of the interviewees “liked” the SMS text messages and felt they were “good” and helpful to “remind me to do something for my health.” One participant did not find the SMS text messages “motivating” and suggested a feature to “opt-out” of receiving messages. A wearable activity monitor instead of the MyFitnessPal app was suggested to increase the accuracy of activity tracking.