Impact of Push Notifications on Physical Activity and Sodium Intake Among Patients with Hypertension: Microrandomized Trial of a Just-in-Time Adaptive Intervention

TL;DR

A study found that push notifications for hypertension patients didn't boost physical activity or reduce sodium intake but increased app engagement.

Contribution

This study evaluates the short-term impact of microrandomized push notifications on lifestyle behaviors in hypertension patients.

Findings

Push notifications did not increase step count after physical activity prompts.

Dietary notifications did not lead to more low-sodium food choices.

Notifications increased mobile app engagement and searches for low-sodium options.

Abstract

Achieving adequate blood pressure control is challenging for patients and clinicians. Digital hypertension management solutions that use push notifications to promote lifestyle management have been proposed as an approach, but their effectiveness remains unknown.

This analysis was designed to interrogate the independent and short-term effects of push notifications, tailored to participant and environmental factors, and on physical activity levels and sodium intake among individuals with hypertension.

The myBPmyLife study was a 6-month randomized controlled trial of participants with self-reported hypertension recruited from an academic medical center and federally qualified health centers. A core component of the intervention consisted of microrandomized push notifications promoting lifestyle modifications that were randomly delivered at 4 daily time points and focused on physical activity and dietary sodium intake. Our primary outcome for this secondary analysis was the step count 60 minutes after a physical activity notification and lower-sodium food choices 24 hours after a dietary notification. This analysis focuses on the results of the microrandomized trial and used a centered and weighted least squares method adapted for 2 or more treatments.

A total of 298 participants were randomized to the intervention arm, of whom 287 had data available for analysis. Participants’ mean age was 59.5 (SD 13.1) years, 138 (48.1%) were women, and 206 (71.8%) were White. Participants were randomized at 187,517 time points over 6 months, which led to 0.96 (SD 0.86) push notifications per day divided between activity (50.4%; SD 0.4) and dietary (49.8%; SD 0.4) notifications. Activity notifications did not increase step count in the 60 minutes after a notification (estimate 1.01, 95% CI 0.98‐1.04; P=.40). Similarly, dietary notifications did not impact the number of lower-sodium food choices in the subsequent 24 hours (estimate 0.93, 95% CI 0.83‐1.04; P=.23), but in exploratory post hoc analyses, did increase mobile app use by 95.5% (95% CI 1.81‐2.10; P<.001), mobile app clicks or searches by 93.7% (95% CI 1.72%‐2.16%; P<.001), and low sodium searches by 113.0% (95% CI 1.73‐2.53; P<.001), all within 60 minutes.

In patients with hypertension, push notifications did not impact short-term physical activity levels or dietary sodium intake but did improve intervention engagement.

Figures