The 7-month study took place in 2 consecutive phases; the study schematic is shown in Figure 1 . In Phase 1 (4 weeks), after a 14-day period on an ad-lib diet, participants received 14 days of home-delivered DASH-SRD meals. The goal of Phase 1 was to measure BP salt-sensitivity in a pragmatic way and familiarize participants with the DASH-SRD using commercially available dietary modification. In Phase 2 (6 months), all participants received 5 dietitian-led motivational interviewing (MI)-based telephone counseling sessions over 4 months and were randomly assigned to counseling alone or counseling plus the WHEELS (Women’s and Men’s Hypertension Experiences and Emerging LifestyleS) mobile application. The WHEELS intervention uses structured messaging to rebuild, reframe, and integrate cognitive representations of the DASH dietary pattern See appendix for more information about the WHEELS mobile application. The goal of study Phase 2 was to evaluate the efficacy of the remotely delivered dietary interventions.

Study schematic.

The primary outcome of this analysis is the DASH adherence score from the 3-day food records in participants assigned to the WHEELS group plus MI-based counseling vs. MI-based counseling alone. Based on national survey data and our preliminary work, we expected a baseline DASH score of 2.5 ± 1.5 points (range, 0-9). Accounting for anticipated 15% dropout rate, a final sample size of 90 would provide >80% power to detect an intragroup change of 0.6 points between baseline and the end of study by paired t-testing, and >80% power to detect a differential between group change in dietary score of 0.9 points by 2-sample t-test (both with 2-sided Type I error of 5%). Secondary outcomes of interest were the change in dietary sodium and potassium intake, urinary sodium to potassium ratio, blood pressure and metabolic biomarkers over the course of the intervention.

Analyses were completed using STATA 18 (STATACorp, College Station, TX). Two-sample t-tests and chi-square tests were used as appropriate to compare baseline characteristics between treatment groups. A generalized estimating equations model with an exchangeable correlation structure was used to analyze changes in DASH adherence scores and BP, and the effect of the WHEELS mobile application over time. Paired t-testing was used to evaluate changes between baseline and the end of study in other clinical variables. All analyses were conducted as intention-to-treat.

The study was approved by the Institutional Review Board and conducted at the VA Ann Arbor Healthcare System in Ann Arbor, Michigan. Potential participants were identified through electronic health record screening and contacted by mail or telephone. Eligible Veterans were aged >45 years with hypertension (here defined as screening systolic BP ≥ 130 and/or diastolic BP ≥ 85 mmHg, and/or current use of antihypertensive drugs) and obesity (defined as body mass index ≥30 kg/m ² and/or waist circumference >94 cm). In addition, participants were required to own a smartphone and be willing to consider using a mobile application if randomized to that group.

Key exclusion criteria included uncontrolled on-treatment BP (>160 mmHg systolic), history of heart failure, cardiomyopathy, or severe valvular heart disease, unoperated arterial aneurysm recommended for intervention, myocardial infarction or stroke within the prior 6 months, chronic kidney disease with estimated glomerular filtration rate <45 mL/min/ 1.73 m, prior hyperkalemia requiring urgent treatment, uncontrolled diabetes mellitus (hemoglobin A1c >10%), and active alcohol or drug use disorder.

After inclusion/exclusion criteria were confirmed and consent was signed, participants completed a Block Food Frequency Questionnaire (FFQ) to assess habitual dietary intake, baseline clinic BP was measured (see Measurements below), and metabolic biomarkers were obtained. Phase 1 of the study started at the end of the screening visit.

During a baseline period of 14 days, subjects consumed their usual diet and completed a 3-day food record that was reviewed by study dietitians (see Measurements below). Antihypertensive medications most likely to affect BP salt-sensitivity (diuretics, renin-angiotensin-aldosterone antagonists) were held during this phase, and participants were asked to measure home BP at least daily. Participants with symptomatic hypertension or systolic BP of > 180 mmHg on more than 1 occasion during the baseline interval were evaluated for safety and need for medication adjustment.

Clinic and 24-hour ambulatory BP were measured and 24-hour urine collection was obtained at the end of this baseline interval of 14 days (see Measurements below). All participants then began a 14-day feeding period with home-delivered, eucaloric DASH-SRD meals, remaining off of any previously held antihypertensive medications. The meals were prepared and delivered weekly by Mom’s Meals (PurFoods, LLC, Ankeny, IA) and were aligned with DASH diet principles. Each meal contained <600 mg sodium, <30% fat, and <10% saturated fat content and snacks were provided if needed to meet daily caloric needs; a sample menu is attached in the Supplement . Participants again completed a 3-day food record, and at the end of the interval clinic and 24-hour ambulatory BP were measured and 24-hour urine collection was obtained.

Study Phase 2 began immediately following the final Phase 1 study visit; any antihypertensive medications held during Phase 1 were re-started. All participants were assigned to receive 5 motivational interviewing-based nutrition counseling sessions with a study registered dietitian who had previously received training in MI techniques. These sessions were delivered remotely over the telephone and lasted up to 1 hour. The 5 sessions were completed approximately monthly, with the first session occurring within 1 to 2 weeks after the end of study Phase 1. All sessions for an individual participant were completed by the same registered dietitian for nutrition counseling continuity. Standardized templates were used to document visits. The fidelity of MI counseling was assessed and confirmed throughout the trial using the OnePass coding system by an MI-trained dietitian who was not involved in participant counseling.

A 24-hour dietary recall was collected to start each session. Participants were guided in setting 1 or 2 SMART (Specific, Measurable, Achievable, Realistic, and Time-bound) goals focused on DASH diet principles during each session. For each goal, participants were guided in discussion about its importance, and their confidence in their ability as well as anticipated barriers to meet it. Participants reflected on goals during subsequent visits.

Participants were randomized to either counseling alone, or counseling plus the WHEELS mobile application at the beginning of Phase 2 via REDCAP’s randomization feature. The WHEELS program is a cognitive-behavioral approach grounded in self-regulation model theory. The program uses daily tailored messaging and surveys to remap cognitive representations of the DASH-SRD, influenced by the user’s self-identified motivations for healthy eating. A previous email-based iteration of the WHEELS program had excellent participant engagement and was associated with increased DASH diet adherence and decreased BP

In the WHEELS program, each day of the week focuses on 1 of 7 DASH-SRD domains (increased intake of fruits, vegetables, lean proteins, low-fat dairy, and whole grains; limited intake of unhealthy fats and sodium). Users complete brief morning surveys on their knowledge, skills, and attitudes related to that day’s DASH-SRD domain, receive tailored messages based on their responses and self-identified motivations, then complete a brief evening survey on how successfully they met DASH-SRD goals that day. In this study, the WHEELS sequence repeated weekly for a total of 8 weeks/56 days (see Supplemental Material for additional details).

Clinic and 24-hour ambulatory BP were measured and 24-hour urine collection obtained at 1 month and 6 months of Phase 2. Three-day food records were obtained immediately prior to the Phase 2, 1- and 6-month visits. The FFQ was completed prior to the Phase 2, 6-month visit. Participants were invited to complete a final FFQ at 12 months of follow-up to evaluate the sustainability of dietary changes after the study intervention ended.

Resting brachial BP and heart rate measurements were collected at each in-person study visit ( Figure 1 ) using a Carescape V100 monitor with appropriately sized automated cuff (GE Healthcare, Chicago, IL). Measurements were collected unattended after the subject had been seated for >15 minutes and until 3 readings within 10 mmHg from each other were obtained.

The 24-hour ambulatory BP readings were obtained at 4 timepoints (at week 2 and week 4 of Phase 1, at 1 and 6 months of Phase 2; Figure 1 ). Participants were provided a preprogrammed SunTech Oscar II ambulatory BP system (Suntech Medical Inc, Raleigh, NC) with appropriately adjusted cuff and asked to wear the monitor for 24 hours the day prior to their next in-person visit. The monitors were programmed to take BP readings every 30 minutes during daytime hours (7:00-21:00) and every 60 minutes during nighttime hours (22:00-6:00). Participants completed a diary of cuff usage and sleep-wake times for adjustment of day-night intervals by the study team.

The acceptability of recordings for analysis was assessed according to published international norms Salt-sensitive BP was defined as a decrease of ≥4 mmHg in 24-hour mean arterial BP between weeks 2 and 4 of Phase 1 (ie, prior to and after 2 weeks of provided DASH-SRD) Direct comparisons could not be performed between Phase 1 and Phase 2 ambulatory BP readings because most participants had BP medication(s) weaned during Phase 1 that were restarted at the beginning of study Phase 2.

Antihypertensive medication use was recorded at each study visit. VA pharmacy records were reviewed to assess antihypertensive medication adherence by proportion of days covered by dispensed medications. Adequate adherence was defined as ≥80% of days covered (

Click to access PQA_Measures_Overview.pdf