While the Apple Watch may never be able to non-invasively measure a user’s glucose levels, an intriguing and massive new study conducted by the health startup Cardiogram and the University of California San Francisco suggests that the device can accurately detect when a wearer has diabetes. The study specifically found that the Apple Watch and other wearables were able to detect the disease in previously diagnosed patients 85% of the time.
All told, the study monitored approximately 14,000 Apple Watch and Android Wear owners over the course of many weeks. As for how the testing was done, the researchers explain that they used an avalanche of health sensor data to train a deep neural network “by presenting it with samples from people with and without diabetes, hypertension, sleep apnea, atrial fibrillation, and high cholesterol.” Incidentally, Cardiogram calls its AI-based algorithm DeepHeart.
As to how heart rate data is tied into the detection of diabetes, Cardiogram co-founder Johnson Hsieh explains: “Your heart is connected with your pancreas via the autonomic nervous system. As people develop the early stages of diabetes, their pattern of heart rate variability shifts.”
Hsieh further cites a 2015 study wherein researchers discovered that a “high resting heart rate and low heart rate variability” is capable of predicting when individuals are liable to develop diabetes “over a 12-year period.”
The research here is obviously incredibly important, especially as the number of individuals suffering from diabetes continues to grow. As the study notes, more than 100 million individuals in the U.S. alone either suffer from diabetes or are prediabetic.
“1 in 4 of those with diabetes are undiagnosed and, even worse, 88.4% of people with prediabetes don’t realize they have it,” the report further adds.
With these new research results in mind, Hsieh adds that the Cardiogram app for iOS and Android will likely incorporate DeepHeart into subsequent app updates.