• • Use this as a mental model for the evolution to MORE sensors, more complex fusion and that algos just won't stop.

New Sleep-Tracking Wearables Help Solve Real Medical Problems

 Getty Images

Every year, nearly a million exhausted people with sleep apnea—a breathing disorder caused when throat muscles relax and block the airway during sleep—get into car accidents, causing over a thousand deaths. Apnea is linked to obesity, heart disease, diabetes, an additional $3.4 billion in medical costs, and $16 billion in auto collision costs. Even though apnea has telltale signs (loud snoring, daytime fatigue), it goes undiagnosed 75 percent of the time.

Why? It’s damned expensive and horrendously inconvenient to diagnose sleep apnea. Polysomnography, the standard medical sleep study, requires a medical technician to attach 22 wires to a person’s body and monitor them all night long. The average cost is nearly $3,000. In the world of high-deductible health plans, that payment comes right out of the patient’s pocket. Follow-up testing to measure the effectiveness of treatment is financially unthinkable. The idea of doing clinical sleep studies once a month to monitor progress is a diagnostic crack-pipe fantasy.

Enter wearables, specifically the “pro-sumer” variety with FDA clearance and clinical backing. Unlike first-generation activity trackers that measure movement and sometimes heartbeat, clinical consumer wearables like the recently releasedSleepImage can measure heart rhythm (ECG), breathing volume, and snoring (through tissue vibration). They can also keep tabs on body movement as well as position—whether a person is sleeping on their back, side, or belly. Algorithms calculate the second-to-second relationship between heart rate variability and breathing variability. This relationship between heart and breathing rhythms, known as cardio-pulmonary coupling, maps to the sleep stages and breathing disruptions that previously only a polysomnogram could measure.

The SleepImage is a small, oval-shaped disc that sticks to the chest near the heart. An electrode on a short wire is also attached a little lower, on the ribs. At the end of the night, a person peels off the wearable, uploads the data to a HIPPA-compliant web site, and sees a breakdown the night’s sleep: stable versus unstable sleep stages, REM sleep, snore count, position and sleep interruptions. A prescription subscription gives a person more detailed clinical data visualizations and gives their doctor enough raw biometric data to distinguish between the cardio-pulmonary patterns of obstructive sleep apnea and other sleep-related breathing disorders.

Pro-sumer wearables—consumer devices that generate clinically relevant biometric data—are not (yet) cheap. The SleepImage is $249. Its consumer-level data service is $99 a year, and its prescription-level data service is $149 per year. But it’s an order of magnitude cheaper than polysomnography, and it’s two orders of magnitude cheaper if used four nights in a row, or once a month for half a year, to see whether measures to address the problem—losing weight, sleeping with a mouthpiece that helps keep the airway open, or a continuous positive air pressure (CPAP) mask—make a measurable difference. The real shift that pro-sumer devices enable is from one-time snapshots to a continuous, long-duration series of views that allow a patient and doctor to adjust tactics and measure the results.

Pro-sumer wearables also move measurement from the lab to the actual environment where a person lives. If the devices get cheap enough, the data makes it possible to tease apart individual sleeping issues and social sleeping problems: a partner’s movements, snores and yes, cover stealing.

In a clinical case study using wearables, researchers stacked husband-and-wife time-series data to observe that the husband’s snoring wasn’t loud enough to wake up his wife. But it was loud enough to disrupt her stable sleep, leaving her wrecked the next day. After the husband was treated for apnea, his wife’s sleep quality improved even more than his did. At long last, the Quantified Self can deliver what snorers’ partners have always wanted: evidence.

Jokes aside, the hope for pro-sumer wearables is that they can transcend the “nice to have” activity data generated by movement trackers and provide the “need to have” data that indicates major medical issues. It’s one thing for a healthy person to tinker with their caffeine intake, room temperature and the thread count of their sheets to see what happens to their sleep. It’s a completely different thing for a doctor and patient to use wearables to addresses life-threatening conditions. That shift has a potentially monumental impact on patients, their sleeping partners, and everyone who shares the road with millions of drowsy drivers.