Donate to Science & Enterprise

S&E on Mastodon

S&E on LinkedIn

S&E on Flipboard

Please share Science & Enterprise

Apps Shown to Reduce Blood Pressure, Detect Diabetes

Woman with phone and laptop

(FirmBee, Pixabay)

7 Mar. 2019. Popular health apps on smartphones are shown in reviews of data from app users to help reduce blood pressure and screen for type 2 diabetes. Results of these studies are scheduled for 2 separate presentations on Sunday, 17 March at the annual scientific meeting of American College of Cardiology in New Orleans.

One team led by cardiologist Bimal Shah at Duke Health in Durham, North Carolina reports on a study using a blood pressure monitoring app made by health technology company Livongo, in Mountain View, California, where Shah is also chief medical officer. Livongo is a developer of mobile health monitoring systems for diabetes and hypertension, or high blood pressure. Hypertension is a common condition affecting 1 in 3 adults in the U.S., or about 75 million people, according to Centers for Disease Control and Prevention. But barely more than half of people with the condition (54%) have their blood pressure under control, despite hypertension increasing the risk for heart disease and stroke.

The Livongo technology includes a home blood pressure cuff that connects wirelessly to the user’s smartphone app. The cuff reads the individual’s blood pressure and provides instant feedback, along with tips from the app on reducing high blood pressure and gaining a healthy lifestyle. The Livongo program also provides live human health coaches, as well as data sharing with physicians and family members. In addition, Livongo participants can track changes in blood pressure over time.

Shah and colleagues tracked 708 Livongo program participants with an average age of 54, all of whom had type 2 diabetes. At the start of the study, more than 6 in 10 of these individuals (63%) had blood pressure greater than 130/80 millimeters, systolic over diastolic measures, considered the threshold for high blood pressure by American College of Cardiology guidelines. Nearly 3 in 10 participants (29%) had blood pressure readings of 140/90 millimeters, the threshold for level 2 high blood pressure.

After 6 weeks, 41 percent of participants recorded blood pressure lower than the 130/80 threshold, while 24 percent still showed blood pressure greater than 140/90 millimeters. The findings suggest the program of home monitoring, online tips, and live coaching can help many individuals reduce their blood pressure. But Shah notes that the results still need to be validated in a clinical trial measuring the program participants against a comparison group receiving the usual care.

Diabetes screening with heart rate monitor

Heart disease is a condition often found in people with type 2 diabetes, and in both cases symptoms may not be readily apparent to individuals. University of California in San Francisco is conducting a large-scale online project called the Health eHeart Study, that tracks overall and heart health with mobile devices, including smartphone apps.

One of those apps is Instant Heart Rate, offered by the health technology company Azumio in Redwood City, California. Instant Heart Rate uses the smartphone’s flashlight and camera to read photoplethysmography, or PPG, signals emitted by blood vessels through the skin. PPG sends low-intensity infrared light waves that penetrate skin layers to blood vessels, where blood absorbs rather than reflects back the light. The app then detects changes in blood volume and calculates the individual’s heart rate at any particular time, such as at rest or after exercise.

Instant Heart Rate is a very popular app, and as a result a team led by UC-San Francisco postdoctoral researcher Robert Avram was able to analyze results from more than 54,000 Health eHeart Study participants, of which 7 percent self-reported their diabetes. Avram and colleagues developed a deep-learning algorithm, a form of artificial intelligence, to analyze the Instant Heart Rate data and detect participants with diabetes, based only on those PPG data.

The results show the analysis of PPG data correctly identifies people with diabetes in 72 percent of the cases. When adding in risk factors for type 2 diabetes such as age, gender, body mass index, and race or ethnicity, that rate rises to 81 percent. The analytics are particularly effective at indicating people without diabetes, successful in 97 percent of cases.

The researchers are validating the algorithm in 2 cardiovascular prevention clinics, and plan to test the model across different ethnic groups, such as African- and Asian-Americans that tend to be under-represented in these studies. Once successfully validated, say the researchers, the algorithm could be available as a smartphone app within 2 years.

More from Science & Enterprise:

*     *     *

Comments are closed.