Most sleep studies get their data from government questionnaires and subjects hooked up to wires in labs, which–duh–aren’t great at quantifying real-world shut-eye. You’re probably not going to send that work email at 1:00am when you’re in a sleep lab, or own up to it on a questionnaire. That’s why the University of Michigan decided to pull sleep data from the very thing keeping people up at night–a smartphone app.
“We wanted people to want to help us,” says Daniel Forger, one of the scientists behind the study, released in Science Advances. “So we thought, ‘If we give them a useful app then maybe they’ll want to give us data.'” With this in mind, Forger and his team launched Entrain in 2014, a free jet lag-hacking app that helps users align their circadian clock to the time zone they’ve traveled to by recommending different lighting scenarios.
Here’s the thinking: Everyone has a regular pattern of sleep and wakefulness, governed by light exposure. When you change the timing of that light exposure (say, when you travel to a new time zone), you mess with your biological clock and start to feel groggy when you should feel awake, and vice versa. Entrain helps you correct this circadian misalignment by telling you when you should seek the sun or a dark room–but only after you volunteer some details about your usual sleeping habits.
“People were motivated to give good data to the app because they wanted to get over jetlag quicker,” says Forger. And, as it happened, the same data users entered to hack jetlag was useful for characterizing larger global sleeping patterns. Within the first year of the app’s release, over 8,000 users from 128 different countries provided details on their home time zones, how much indoor and outdoor light they got each day, as well as when they went to bed and woke up. A mother lode of information, and at almost no cost to the researchers.
So Forger and his team crunched the data to figure out how age, gender, home country, and light exposure affect sleeping habits, and out came patterns that were consistent with lab studies. That’s what they were hoping for. “We’re trying to prove to the scientific community that apps can be used for data collection,” says Forger. “And we did this by validating our findings against those found in labs.”
But that doesn’t mean there weren’t valuable takeaways. This study is one of the first to quantify social influences on sleep, and even Forger was surprised by some of the results. “In the beginning, I thought wake time would be determined by society and bedtime by our biological cues to sleep,” he says. “But it’s actually the opposite.”
The study shows that people in Singapore and Brazil are night owls who wake at sunrise, while Aussies hit the hay much earlier and wake at the same time (so they get more sleep). The difference suggests there are societal forces dictating bedtime, and that bedtime determines how much sleep you’ll get–observations that were only possible thanks to the global uptake of the app. “Sleep duration being primarily dependent on bedtime is an important lesson here,” says Charles Czeisler, chief of the Division of Sleep and Circadian Disorders at Brigham and Women’s Hospital. “That’s how we get people to improve their sleep.”
The Future of Sleep
Still, the smartphone-powered study isn’t perfect. First off, it’s probably biased towards affluent jet setters. “With mobile tech, you’ve got to ask, how representative of the population is this?” says Center for Disease Control epidemiologist Anne Wheaton. “They’re people who use cell phones, travel, and are interested in their health.” That’s a problem when you consider that bad sleeping habits adversely affect the poor, including the more than 15 million American shift workers who work evenings, nights, and other irregular hours. Like jetlag, graveyard shifts can mess with your internal clock and, in the worst cases, lead to obesity, diabetes, and other health problems.
And lab tests have one big advantage over apps like this, which rely on users’ memory to report when they fell asleep. “Just like in sleep labs, people have a hard time judging when they actually went to sleep,” says Wheaton. “They’ll complain that they only got one hour of sleep when we measured seven.” Even though people are inclined to give better data to apps that benefit them, that doesn’t mean it will be more accurate.
The next step for sleep science, then, will be gathering real-time data from more objective sources: fitness wearables. The International Data Corporation predicts that over 200 million wearables will be bouncing around the planet by 2019, and a lot of them–like those from Garmin, Fitbit, and Jawbone–include accelerometers that track your movement during sleep. It’s called actigraphy, and it isn’t perfect, but it’s more reliable than your half-asleep memory when it comes to reporting data. If you’re restless from 1:00am to 3:00am, you probably won’t remember, but your tracker will.
Companies aren’t exactly keen to share this data trove with science (although Jawbone has sprinkled some data crumbs about American bedtimes), so researchers will have to do their own actigraphy studies in the meantime. Forger and his team are on top of it: The next version of their app will take data from users’ wearables to validate their self-reported data.