Wearable technologies that count your steps, measure your heart rate and monitor your sleep - like Fitbit, Jawbone or the Nike Fuelband - seem to offer the possibility of a new era of health and fitness. These little devices promise to buzz and beep you to better workouts, lower blood pressure, and less stress; they can even remind you to sit up straight.
Unfortunately, activity trackers are not doing as much as you might think to fix health problems, as recent research by three doctors at the University of Pennsylvania shows.
While a lot of people are interested in the potential for wearables to transform health behaviours, there hasn't been much evidence yet that these devices do that, says Mitesh Patel, one of the authors of the research. There is also little understanding of how the health community can get wearables into the hands of the patients who need them most.
Because wearables are relatively new and expensive, current owners tend to be younger, more affluent, and more physically active than other groups.
One survey by Nielsen found that nearly half of wearable owners were under 35 years old, and 29 per cent earned more than $US100,000 annually.
"The people who really need these devices are not the ones who are using them," Patel says.
In addition, a huge number of people who own wearable devices put them away and never think about them again. One 2013 online survey by Endeavour Partners, a consulting firm, showed that half of Americans who had owned an activity tracker no longer used it, and one third had abandoned the device within six months of acquiring it.
Yet another lesson from Patel's research is that you might not need to shell out for an expensive wearable device at all.
In another study, Patel and his colleagues examined questions about the accuracy of wearable devices. They hooked test subjects up to a huge number of devices - they put three on their arm, clipped three to their waistband, and put an iPhone and an Android running fitness apps in either pocket - and then had the people walk on a treadmill for 500 or 1500 steps.
They found that most of the devices were accurate, though one, the Nike Fuelband, underestimated activity by about 20 per cent in their trial.
However, one finding of the experiment was that smartphone apps were just as accurate as wearable devices in measuring activity, at least when walking on the treadmill. "We did find in general that most of these devices are accurate, but we found that smartphones, which most [people] carry with them every day, are just as accurate," says Patel.
Patel stresses that simply buying a wearable isn't enough - people need effective strategies for using the data that will motivate them to stay active. Using findings from behavioural economics, which looks at how psychology and other hidden factors affect people's decisions, Patel and his colleagues have come up with some guidelines for how people or organizations can best use activity trackers.
For example, Patel points out that some businesses today have given their employees the devices, with the hope of improving health, increasing employee satisfaction, and reducing healthcare costs. Some organizations might create a contest or post a leaderboard that rewards those employees who are most active during the month.
According to Patel, these are terrible strategies. A contest or a leaderboard will only succeed in rewarding the employees that are the most active - who are likely those who need the least help.
So what is one to do? The doctors urge employers and exercisers alike to pay attention to some basic findings from behavioural economics.
First, he says, people are more motivated by small, immediate rewards than large, distant ones. It's better to put people in a lottery where they have a one in 10 chance of winning $10 than a one in 100 chance of winning $100. A company could, for example, give employees the chance to win a small amount of money every day if they met their fitness goal the previous day.
Secondly, people are more motivated by the prospect of losing something they have than of gaining something they don't - a branch of behavioral economics known as "prospect theory." So someone who is hoping to stay active could make a small bet with their friends or family. When they don't meet their daily exercise goals, a small amount of money might be taken from that pool and, for example, given to charity.
Thirdly, people are more motivated to do an activity when those they know are involved. So companies might make exercise goals into a team activity. Individual exercisers can involve their friends or families in their wagers, or share their progress with friends using social media.
Most importantly, Patel says, exercisers need to set up what he calls a "daily feedback loop."
"We know from previous studies that if you want to create a daily behaviour, you have to set the feedback loop to mirror that behavior." People are much more likely to exercise every day if they are monitored and rewarded every day, rather than next year.
While wearable devices aren't having a broad impact yet, the industry is still young, and behavioural economics in particular offers some good suggestions for how people and organizations might put them to use.
"There isn't much evidence at all on how we can best use wearable devices to change behaviour. I do think there is potential," Patel says.