According to a 2015 consumer wearables report, 1 in 10 Americans has a fitness tracker. Yet there have been various media reports questioning the accuracy of these devices. Today we talk about whether the accuracy of fitness trackers matters; upcoming research trends; and, the most important question, even if they are accurate, are fitness trackers serving their purpose? Are they getting us moving by holding us accountable?
With the help of our expert, Dr. Greg Welk, we discuss the complexities involved in making a device to quantify movement. We will also explore the findings of Dr. Welk’s 2014 paper from Medicine & Science in Sports & Exercise that captured media attention by examining the accuracy of what are now earlier versions of eight mainstream monitors available on the market.
We live in a very challenging environment to be active today…monitors provide a way to keep people accountable…[People] have to build physical activity into their day; it doesn’t just happen. It has to be actively planned and prompted during the day.
– Dr. Greg Welk
Dr. Greg Welk is a Professor in the Department of Kinesiology at Iowa State University, where he oversees the Physical Activity and Health Promotion lab. With more than 50 peer-reviewed, data-based research publications to his credit, he is a national leader in developing and assessing the accuracy of physical activity measurements while promoting physical activity.
The show notes, biomarkers, and links to the apps, devices and labs and everything else mentioned are below. Enjoy the show and let me know what you think in the comments!
- Expensive accelerometers, previously used for research, were introduced to the consumer market through development of the Wii (4:58).
- University researchers have performed calibration measurements on consumer devices; the degree of accuracy they have found reflects the level of calibration research occurring at the consumer device companies (7:38).
- The integration of several types of sensors enables better inferences of activity (9:46).
- An overall error of 10% to 15% was found in the 2014 study of consumer devices; this is a reasonable level of error given that the goal of consumer monitors is behavioral change, not precision (12:39).
- Companies must balance accuracy with fashion and functionality (15:32).
- In the forthcoming study, device accuracy was measured while the subject engaged in twenty-minute time periods of sedentary activity (office work); locomotive activity (aerobic workout) and circuit training. Analyzed in this manner, monitors would overestimate activity for certain types of work and underestimate activity for other types of work, contrary to a previous study showing a reasonable overall error (17:09).
- Consumer companies develop their own equations and algorithms and work to improve them, while researchers publish, share, and exchange calibrations used (19:21).
- Currently in research there is a movement to use accelerometers that report in the same unit, g-forces. This unit allows inferences across monitors (20:15).
- Different data storage and software processing capabilities are required between consumer and research devices (21:15).
- There is a ongoing movement to use monitoring devices to integrate patient activity and other parameters into an electronic medical report for healthcare purposes (23:15).
- In the coming years, minimizing inactive behavior, while promoting active behavior, will be a desirable combination for the market. Researchers are currently pursing ways to better monitor sedentary time (28:15).
- Upcoming studies will be looking at the new Jawbone Up3 and FitBit Charge. Other studies will be looking to expand the health coaching model (36:09).
- The biomarkers Dr. Greg Welk tracks on a routine basis to monitor and improve his health, longevity and performance is to minimize sedentary time. He previously tracked heart rate variability.
- Dr. Greg Welk’s one biggest recommendation on using body data to improve your health, longevity and performance is to use devices as needed to hold oneself accountable for actively planning and prompting physical activity throughout the day.
Dr.Greg Welk & The Physical Activity Lab
- The Physical Activity Lab : Find more information on lab activities and past studies.
- PubMed Link of Dr. Greg Welk’s work.
- Medicine & Science in Sports & Exercise : The original 2014 study entitled Validity of Consumer-Based Physical Activity Monitors.
Terms & Technology Used in Activity Trackers
- Accelerometer: a device measuring body movements in terms of acceleration.
- Altimeter: measures altitude.
- Counts: a unit that lacked consistency across initial devices due to how devices were scaled or internally processed. Researchers would then take this unit and link it to an area of interest, for example a certain number of counts would equate to a number of calories.
- G-force:: the force of gravity.
- Gyroscope: a device used to measure orientation within 3D space.
- Piezoelectric: an early accelerometer type, able to be scaled and calibrated, that converted a physical force into an electrical current and produced units referred to as counts.
The list of activity and fitness tracking devices discussed in this episode:
- ActiGraph: Demonstrated a 12.6% error in the 2014 paper.
- Amiigo: Mentioned by Damien, Dr. Welk did not have plans of it in future studies.
- Apple Watch: Performance to be evaluated in upcoming spring study, with results projected for January 2016.
- Basis B1 Band: Highest error on 2014 study (23.5%) and has been discontinued.
- Body Media Fit : Top performer in the 2014 paper with a 9.3 % error.
- Directlife: Demonstrated a 12.8% error in the 2014 paper.
- Fit Bit Charge: Performance to be evaluated in upcoming studies.
- Fit Bit Flex: Performance to be evaluated in upcoming studies.
- Fit Bit One: Top second performer in the 2014 paper with 10.1% error.
- Fit Bit Zip: Top third performer in the 2014 paper with 10.4% error.
- JawBone Up: Demonstrated a 12.2% error in the 2014 paper.
- Jawbone Up 24: Performance to be evaluated in upcoming studies.
- Metria Patch: The wearable patch with monitoring technology of a device.
- Nike Fuelband: Demonstrated a 13% error in the 2014 paper.
- Heart Rate: Dr. Welk previously used heart rate monitors to track intensity of workouts in high level training.
- Heart Rate Variability (HRV): An athlete can track HRV as an indicator of whether the athlete’s body has fully recovered from previous workouts and whether the athlete’s body is ready for the next workout. Athletes can use HRV to identify symptoms of overtraining. Please check out other episodes and guests of The Quantified Body focusing on HRV: Episode 1: Andrew Flatt (resistance training), Episode 6: Ronda Collier (managing stress), and Episode 8: Todd Becker (hormesis and stress).
- Calories: Devices may overestimate or underestimate the number of calories burned by ten to fifteen percent. On the level of individual assessment, the exact number of calories burned is less important; what is important is that using the device may cause the individual to make a behavioral change.
- Steps: Individuals may use monitoring devices to track steps taken. As is the case with devices used to track calories, the use of devices measuring steps may cause positive behavioral changes.
- Leonardo De Vinci: credited with the idea of a monitoring device.
- Thomas Jefferson: was one of the first developers of a mechanical pedometer.
Full Interview Transcript
[Dr. Greg Welk ]: Hi.
[Damien Blenkinsopp]: Greg, thank you so much for being on the show.
[Dr. Greg Welk ]: Yup, my pleasure.
[Damien Blenkinsopp]: So, first of all, I wanted to dive into what the interest is in tracking exercise, or activity in general. What kind of benefits do you think this can have for us?
[Dr. Greg Welk ]: Well as researcher I’ve been using monitors for many years. We use them to study how active people are so we can quantify associations with health benefits, and evaluate whether interventions work to change behaviors.
So the recent movement to consumer monitors has been a really interesting development. So I think the research community is at least still tying to learn how best to use these devices, and whether they do work to change behavior or not.
[Damien Blenkinsopp]: Right, so it’s mostly about, is it going to change our behaviors, and thus provide benefits in terms of, say increased activity, or more focused activity, on things that are tending to work. Obviously, there’s still a lot of exercise science debates going out there. Like, is cardio better, is weight lifting better, resistance training better. There’s a lot of these kind of discussions still going on.
Maybe over time, if we’re tracking exercise on a larger level, and we have big data. Do you think it can help some of these questions?
[Dr. Greg Welk ]: Sure. I think the devices give consumers the power to monitor their own behavior. I guess the key is whether they pay attention to the cues, and tune into them. Because tracking data by itself won’t necessarily change behavior unless it becomes something that’s use in your daily life.
So what we would refer to the data as, would be helping people with self-monitoring. Learning to monitor their own behavior, and using that as cues to change their behavior. But when you talk about the broader level of the data being stored in the cloud, and being able to infer what people are doing and not doing, it opens up all kinds of new opportunities to understand human behaviors. So that’s very exciting as well.
[Damien Blenkinsopp]: Great. And so perhaps you could talk about, what were you doing before this exercise tracking market started developing? What kind of things were you doing in order to estimate and assess activity?
[Dr. Greg Welk ]: Well again, accelerometers have been around for 20 or 30 years for active research in the research community, so there’s been probably second or third or fourth generation monitors being used by researchers since the early 1990s, I’d say. Even a little bit, there were some preliminary models in the 1980s. So we’ve been using accelerometers for many years.
I guess the key that turned it into a consumer marketplace was when the Wii was developed, actually. Because when the Wii was released, the accelerometer that was used in the Wii device all of the sudden became in demand, and the price came down.
And that’s what allowed some of the monitors to develop a consumer versions that were more price competitive at sort of the price point. Because we were previously using very expensive accelerometers, and then the price dropped tremendously and allowed all the consumer development.
[Damien Blenkinsopp]: Great. For people at home, could you describe briefly how does an accelerometer work?
[Dr. Greg Welk ]: Right. Well there’s different technologies. Actually, many people aren’t aware, but Leonardo Da Vinci is credited with coming up with the early idea of a monitoring device, and Thomas Jefferson was one of the first developers of a mechanical pedometer, hundreds of years ago. So the concept of tracking steps has been around for many years.
The most early accelerometers use what’s called a piezoelectric bender moment. This type of device would allow a physical force to be converted into an electrical current, which could then be scaled and calibrated. So the way researchers use these initial devices, the devices all produce a unit that we called counts, an activity count.
And it’s a nebulous unit because a count depended on, it didn’t have a value to it, and it depended on how the device was scaled or internally processed. But researchers could take that device and then calibrate it by linking it to a certain number of calories, or oxygen consumption, to determine how many counts equates to how many calories, for example.
And now those calculations are built into most of these devices. Again, the early devices were piezoelectric, and there was some concerns about the reliability of the sensors, for example, in the early models. But then when they went to solid state devices, they became much more reliable. And now there’s the typical accelerometer that’s at least used in most research devices, it’s more of a men’s accelerometer that senses true gravitational forces, rather than bending moments.
And that opens the door for other type of sensors. So some of the accelerometers can tell, for example, by the direction of the moment whether you’re standing or sitting. So there’s interest in posture as well as movement.
[Damien Blenkinsopp]: So it sounds like there’s quite a wide variety of accelerometers that are available today, and some of them are still more expensive, and less expensive. So it sounds like it’s a technology in evolution. The ones you’re using in research currently are basically ahead of the game compared to the ones we have in consumer devices. Is that correct?
[Dr. Greg Welk ]: Well, somewhat. But actually the studies that we’ve done recently, we’ve compared some of the consumer monitors directly head-to-head with some of the established research monitors. And they’re in the ballpark. So the consumer monitors are not far behind. Which leads you to conclude that there is some calibration work going on behind the scenes at these companies.
So the various companies don’t really talk much about the research. I guess they figure the consumers don’t really care or want to know the details of how accurate it is, and that’s why researchers like myself study it. So we did compare some of the leading consumer monitors against standard research grade monitors.
And several were better than others, but in general it shows that some of the technologies are probably using pretty advanced calibration methods, or multiple sensors, to get the type of precision that they’re getting.
[Damien Blenkinsopp]: So when you say multiple sensors, is that so they have maybe more than one accelerometer inside the device, and they’re using algorithms to calibrate?
[Dr. Greg Welk ]: Right. In addition to a standard accelerometer, a lot of devices include a gyroscope, for example. That’s built into almost all of our cell phones. So, a lot of the way the field’s moving is that a lot of the technology that already exists in your cell phone can actually be an activity monitor. So instead of wearing something on your wrist, we’re actually wearing a very powerful sensing tool in our pockets.
The challenge is calibrating it to having people wear it in different body positions, and things like that. So a standard position, if it’s worn on the wrist, for example, gives you a better way to calibrate it, because everyone’s wearing it in the same spot. But the other sensors would be gyroscopes, or altimeters.
So the one limitation of accelerometers, for example, is that it can’t really tell if you’re walking up a hill , or carrying a backpack. They’re based on gait, for example. So the accelerometers get pretty good at detecting that locomotion is taking place. And then if you’re running, you’re moving faster, and there’s more counts, so your count rate goes up. And they can tell that that’s running.
And more advanced sensors are now using pattern recognition technologies that can determine what type of activity it is. The real challenge is you’re limited if you’re using just an accelerometer, because we know, again, it can’t detect if you’re carrying a backpack, or walking up a hill. So, including an altimeter or some of the devices, including heat sensors, or heart rate monitors. So the integration of several sensors allows you to make better inferences about what people are doing.
[Damien Blenkinsopp]: Great, great. So, let’s talk a little bit about some of the accuracies. Which devices did you actually look actually look at in your studies?
[Dr. Greg Welk ]: Well, that’s the challenge and some of the fun. The paper we had published even just this past year in Medicine and Science and Sports and Exercise compared a number of the leading consumer monitors, but by the time the paper goes live and is published, there’s already new monitors in use.
So, for example, FitBit is a very popular device, and in that original study we compared the FitBit One and the FitBit Zip, which were worn on the hip. Those were the early FitBit devices. And we also compared the BodyMedia Fit, which was an armband monitor, and a host of other monitors. The Groove, DirectLife, the Nike FuelBand. But then, by the time the paper is even out, the consumer models are already very dynamically changing.
So we actually just finished another study that’s submitted and in review now, that compared most of the recent wrist worn monitors. So the movement in the field was away from the waist worn monitors, which is where researchers have typically used them and where pedometers are worn, to the more wrist position, because people are used to wearing things on their wrist, and it’s more fashionable.
Our new study we used the FitBit Flex, which is the wrist worn monitor, and also the Jawbone Up24. And those two tend to be the leading performers, at least in our tests so far.
[Damien Blenkinsopp]: So it’s the Jawbone and the…?
[Dr. Greg Welk ]: The FitBit worked out reasonably well. I’ve seen better results in almost all of our work with the BodyMedia Technology. BodyMedia was the original developer of the most powerful research grade monitor, and most of their technology and development went into the JawBone. So I don’t know if consumers are aware, but JawBone is a very powerful technology company, BodyMedia, that is behind a lot of the infrastructure and the pattern recognition work that they do.
[Damien Blenkinsopp]: Yeah, I mean, because it’s interesting, because some of these companies are now even if they were start-ups are getting bought out by some of the bigger companies. Because Intel bought the Basis watch?
[Dr. Greg Welk ]: Right.
[Damien Blenkinsopp]: Well, of course Intel’s a huge company.
[Dr. Greg Welk ]: Sure.
[Damien Blenkinsopp]: I mean, you’ve got Apple moving in with the watch. Which I guess you haven’t had access to, with their watch.
[Dr. Greg Welk ]: We’re actually just starting that. We have plans to do another study in the spring with the Apple watch. But it’s not going to be out for another couple months, or maybe January or so. For us, anyway.
[Damien Blenkinsopp]: Great, great. So in general, if you’re using a JawBone or the latest FitBit, you think it’s good enough for people to be using, and not be concerned with any accuracy differences between the reality the research grade ones you’re using?
[Dr. Greg Welk ]: Right. I’m pretty comfortable with the results. Actually, our paper showed an error of 10 to 15% for overall estimates, and I have a caveat to that to add, but that’s considered an individual level assessment. So if a person is wearing it and it says that you burned 200 calories, a 10% error would mean its 220, or 180.
So there’s a buffer around that point. But that’s a reasonable amount of error for a consumer level device. Especially when the goal is behavior change. So if you’re using it for research, then precision is more important. But if your real goal is to use it to monitor your activity and to prompt you to be more active, then I think the accuracy is pretty good.
[Damien Blenkinsopp]: Yeah. And I guess the idea behind that, also, is if you’re just wearing something, it’s kind of relative. If I took 10,000 steps one day, and I take 8,000 the next day, is it safe to say that the monitor’s reporting the same kind of bias if it does have a bias and so…
[Dr. Greg Welk ]: Right. That would be constant error, and you’re right. If you have more on one day than another, that would say that you’re more active. And that’s the real goal from a behavior change standpoint, is to have it cue your behavior and prompt you to be accountable to move.
[Damien Blenkinsopp]: Right. And I guess for the moment, that’s probably one of the best uses, in terms of, let’s say, because exercise science is still trying to define what’s, say the ideal movements and the things like this.
I don’t know if you’ve looked at the Amiigo. I don’t think it’s commercially available yet, but that was looking at trying to track more different aspects of the exercises. The movements, and so on.
[Dr. Greg Welk ]: No I haven’t heard of that one. We keep pretty close on this. My students keep me tapped into all the latest and greatest out there, but we hadn’t heard of that one. But I think you’re right, that eventually the movement needs to be to integrate these sensors with other technologies, where they can provide more information.
So it’s not just the amount of movement that you’re getting, and how does that relate to your health, and does it change your behavior and change your health outcome.
[Damien Blenkinsopp]: Right. Right. I guess it’s going to be evolving for quite a while, until we get to any kind of ideas of what’s optimum. Because it’s probably going to be different for each person, as well, depending on where they’re coming from. It’s quite complex.
For the Basis watch which I used for about a year what I found, first of all, it didn’t really reflect very well changes like in the steps reporting, trail running, walking. Sometimes it would think I was doing one thing versus the other. So I think that was for sure something that needs ongoing work to… I don’t know how the Basis turned up in your work. Did you look at the Basis compared to the others?
[Dr. Greg Welk ]: Yeah, we did in our early study. And it didn’t perform too well, but we got contacted by the company after that, and they said, ‘Oh, I think we can do better than that.’ And they were surprised by the results, but we reported what we got. The Basis was one of the poorer performing monitors in our initial study, but they indicated it might have been that we had an earlier version than their latest, or something along those lines.
But you’re right, it’s a challenge for the companies to balance the need for accuracy with consumer forces. So right now, most of the emphasis in the development is on fashion, or size, and making them smaller. But accuracy is not a high priority among most of the companies. And that’s the balance that they’ll have to play, is how to keep the quality good and add new sensors, but without making them bulkier, or not as functional or fashionable anymore. So there’s this balance that they have to play.
[Damien Blenkinsopp]: Yeah, it seems like the people who are most concerned about accuracy are, potentially, some of the athletes, the people who have more going on, or quantified selfers. But for the mass market, it seems like they just kind of assume that it works. Kind of thing.
[Dr. Greg Welk ]: Yeah.
[Damien Blenkinsopp]: And that they don’t have to worry about it too much. I mean, it’s not something that they really question a lot. Because I guess it’s kind of beyond them. We’re talking about accelerometers, and the technologies are relatively advanced and complicated for just the average person to start trying to dig into.
[Dr. Greg Welk ]: Exactly. And so the consumer, like when our study came out, actually the media picked it up, and a lot of the media buzz was, ìThe monitor isn’t as accurate as you think.î And so, again, I pointed out that 10 to 15% error was decent in the paper, but when the media spun it then it became known that your monitor may not be as accurate as you think.
But I think the media and the general public, for example, don’t fully realize how difficult it is to quantify movement. So when we walk around, the accelerometers are basing most of the estimates only on this ability to detect gait: walking and running. If you’re standing, lifting weights, the monitor can’t tell what you’re doing. And if you’re walking up a hill, for example, it can’t note that and really capture the true energy cost. If you’re digging a ditch, with these monitors, you could be burning tons of calories, but you’re not moving in space. And so the monitor’s not picking up that detail.
And that’s the caveat I mentioned earlier, that I was going to get back to. Our recent study, we broke up the time periods and we looked at the accuracy of the devices for sedentary time, like typical office work. And then we looked at locomotor activity, like an aerobic workout on a treadmill. And then we looked at 20 minutes of circuit training.
So we had people for about an hour come in, and we had them first do 20 minutes of sitting around, standing around, just doing office work. 20 minutes on a treadmill, and then 20 minutes in a free weight circuit training environment. And then we looked at the accuracy for all three settings.
So, when I was pointing out earlier that the accuracy was 10 to 15% that was in a more controlled study, and also for an overall period. Like for a full hour. But then when you break out the accuracy for individual segments, the monitors were all over the place. So they overestimated for one activity and underestimated for another.
[Damien Blenkinsopp]: So I guess they’re mostly, are they standardized on jogging or walking primarily?
[Dr. Greg Welk ]: Pretty much, yeah. That’s where the JawBone BodyMedia Fit tended to be better performers, because they have other sensors that are detecting the patterns. Whereas the FitBit and the DirectLife and the Nike FuelBand had more variability in their estimates for each individual activity. So they have more error, even though the overall estimates looked comparable.
[Damien Blenkinsopp]: So what are they combining with the accelerometer in order to determine the type of activity so that they can run a different algorithm, I guess, on the count.
[Dr. Greg Welk ]: Right. A lot of the work actually looks at the patterns within, how the counts are accumulated. So if the counts are in a repeatable pattern, like in a rhythm of frequency, then we know it’s walking. If they see a chaotic pattern, then we can know it’s not walking, or running, and then other patterns can be detected. For example, the BodyMedia can detect a pattern in cycling that most other companies have not been able to do. And we’ve done studies that show it gets better at predicting the energy cost of cycling than other monitors.
So they use other sensors, again, these heat sensors, or just the pattern within the minute. And so even though the devices are producing an estimate of a single minute they’re actually sampling at 30 hertz or more to capture all this rich data.
[Damien Blenkinsopp]: And so, I guess an important thing to understand here is that they’re using proxies, right? They’re taking data which isn’t directly counting steps, calories, and then they’re using research is it standardized research, or is it their own personal estimates?
[Dr. Greg Welk ]: For the most part, all the consumer devices are doing in on their own, and that’s what researchers are now starting to test. Again, when the original calibration work that came out that most of the people in our field use; we developed the calibrations, they get published, and they get used by other researchers. But the companies are on their own to kind of develop their own calibration equations, or pattern recognition algorithms. And then they work over time to improve them.
[Damien Blenkinsopp]: So there’s no activity going towards a standardization of any of this as yet? Whether it’s the hardware, or the software?
[Dr. Greg Welk ]: Not within the consumer market. In the research world, we’re sort of converging back around the use of raw G-forces, I’ll say. So the original accelerometers were producing an output that was called counts. Again, that was you couldn’t compare one count to another count, because they were scaled differently. But if all the accelerometers could report in the same unit, which is gravity, G-forces, then we can make inferences across monitors. And that’s becoming the unit that we should have been using all along, but it wasn’t possible before. But there’s sort of a move towards using raw data, and using raw G-Forces as the unit of comparison.
[Damien Blenkinsopp]: That sounds great. So, I guess, would you envision that consumer companies will eventually be doing this, or is this a very different cost? Has it got a very different cost structure as yet? Could you give us an idea of what the difference is? Is it 10x, or…
[Dr. Greg Welk ]: No, I’d say maybe 2x or less, in terms of the sophistication of a research monitor. It’s more the processing capabilities that they built in the data storage that differentiates a consumer monitor from a research one. So, what researchers want to know is we want to be able to see the raw data in minute by minute form, or second by second. And that requires a lot larger data storage, which then makes the device bigger.
And also we need better processing software. A lot of the consumer monitors export only data every hour, for example, because it doesn’t require as much storage in the unit and keeps the data piece smaller. So it’s a factor of what’s important. So a researcher needs different things than the consumer needs, and it’s kind of that trade off.
But I’d envision, again, the consumer monitors evolving more in the direction of combining multiple sensors, and linking, eventually, to other health related indicators. There’s a big movement now towards how they could be used by physicians, or behavior change applications. So I think the real key is going to be how these devices can interface with the health care system.
Right now they’re just kind of a fun thing that people are using. But the real question is whether health care providers will start using them, and whether insurance companies will start paying for it. Because getting people more active would go a long way to preventing chronic disease and reducing obesity in the population.
[Damien Blenkinsopp]: And for something like that, that’s going to encourage the device manufacturers to move, potentially, towards some standards like the one you were just talking about, which, when it comes to government and stuff, they want something more standardized. And we can’t have just all these monitors.
Just if all the medical doctors arenít going to be able to work with all of them, obviously, that’s not the way it works. And I can imagine there would be some kind of approval process eventually. I don’t know if you’ve seen anything going on like that. I know that Apple was apparently talking with the government, and some of the health bodies I don’t know if you were looking at that about some of the markers and some of the structure that they were putting behind the watch.
[Dr. Greg Welk ]: Yeah, I’ve seen a little bit, but probably not enough to comment. I think stymieing competition would go against what America’s about, and what the companies are about. And so all companies should have a chance to compete for that marketplace, and not force health care providers to use the same device.
But the key is whether the devices can be successfully integrated into an electronic medical record, for example. So I work with a couple of companies that have already been working on the medical interface. And there’s a patch the product’s called the Metria Patch and it’s basically a piece of, version of an accelerometer that a doctor could slap on a person’s arm, tell them to wear it for a week. And then they tear it off and mail it back in, or bring it back in.
And it has all the sophistication of the best available research device, but in a form that a physicianís group could use. And they’ve started working into how that data can be incorporated directly into an electronic medical record, where it can be acted on.
[Damien Blenkinsopp]: And that’s big enough to store the type of minute by minute data you were talking about?
[Dr. Greg Welk ]: Oh sure. Yeah, it is. Yeah. And it has the adhesive capabilities, so when you put it on, it becomes like a patch that sticks to your arm for a week at a time, or even a month at a time. You could swim with it, shower with it, and eventually when it’s taken off and they’ve even figured out the breathability, like how to make sure that if you’re wearing it, that it doesn’t affect your skin underneath. So there’s a lot of work in that particular interface.
But even with that device, I think the missing piece is the health coaching that goes along with it. So for it to ever get traction in the health care community, they need to know that they can see the data, and then use the data to inform behavior change. We’ve done a lot of work lately with health coaching applications of some of these monitors.
We have a virtual health coach, for example, that can see the participant’s data, and we can comment about whether they’ve been active that week, or whether we see patterns that they could act on to improve their physical activity or weight loss behaviors. And so we’ve got a number of studies in process with this kind of health coach model.
[Damien Blenkinsopp]: That sounds very interesting. How long before you think this kind of thing becomes more available to consumers? I know there’s some health patches already out. We’ve spoken with one provider, Ronda Collier, and she’s been working with the health patch for a heart rate monitor. So there are some of these starting to creep into the market.
But they’re still a bit cost prohibitive, based on the fact that it’s a usable one, consumable. So you have to buy one every week, or every month.
[Dr. Greg Welk ]: Right, but that at least gives a cost structure that a physician’s group can budget for. It’s more challenging to use like a physical device that gets lost, or that they have to buy and upgrade. It’s sort of like disposability is an advantage, in a lot of way, because it can be built into the fee structure.
[Damien Blenkinsopp]: That’s right, and it can be updated easily, rather than having to pay a higher price and have someone have to have that for one or two years before they can upgrade.
[Dr. Greg Welk ]: Right.
[Damien Blenkinsopp]: Especially at this time when the technology’s moving very fast. So do you foresee the technology changing rapidly for the next five years, or do you think it’s going to settle down soon?
[Dr. Greg Welk ]: Oh I think it’s just getting started in dynamic phases. I think all these are new and very exciting. We saw them all out at Christmas, I’m sure a lot of people got them in their Christmas presents this year. You saw them highly advertised at the Christmas sales and things.
I think the movement will be towards these smart watches, and the integration with your cell phone. And again, in a few years, the technology already exists in cell phones to do this, so the watch really becomes the cue, or the device, that prompts you about how active you are. Whether the accelerometer is in your watch or in your pocket doesn’t really matter, they’ll basically start interacting with other aspects of your life and provide prompts about your blood pressure, and your heart rate, and other features that can inform your health.
[Damien Blenkinsopp]: Great, thank you very much. Do you have any practical tips for use, which could help to get better feedback, more accurate feedback, over time? For instance, a lot of these watches are, as you say, worn on the wrist. Does it make any difference which wrist you put it on?
[Dr. Greg Welk ]: No, I don’t think that matters. The companies may tell you it’s best if worn on the left, because that’s where they calibrated it. The right or left doesn’t really matter, but I think the key, as you pointed out earlier, is that people should use them on a relative basis, as a reminder to be active. If they want to take a day off and remove the monitor, that’s okay. The world’s not going to end if they have a gap in their data, because the real goal is to be using it to cue your behaviors.
I, for example, don’t use one every day, or I use them when I feel the need to monitor, or things like that. Other people, however, really do benefit from the daily accountability, and want to have that as a daily reminder. So it’s kind of an individual preference there.
[Damien Blenkinsopp]: Yeah, one of the main things I was using it for was just judging how much I was walking, and how far, or how many steps. Just trying to keep that at a reasonable level, and increasing it. Actually, at the time I was suffering from chronic health issues, and it was difficult for me to walk.
So it was great for me to see the progress. Like, I made 10,000 steps today, that’s awesome.
[Dr. Greg Welk ]: Right.
[Damien Blenkinsopp]: Unfortunately, it didn’t take into account the hills which we’re talking about which was a huge factor in whether I made that 10,000 or if I went up this hill. So I’m sure that’s going to come relatively soon.
In terms of the next things you see happening in the market, which you’re kind of excited about, you talked about the patch. Is there anything else you see in the consumer market which you think will be coming soon? Or in the next five years?
[Dr. Greg Welk ]: I think one of the convergences is going to be with the balance between not just promoting activity, but minimizing inactivity, or sedentary behavior. So the researcher community is really intrigued with the dual effects of both inactive lifestyle, and basically being more active and less inactive. That’s one of the things I’m personally much more interested in is minimizing my sedentary time. So I’m like a number of people that have a standing desk, and try to be on my feet more than I’m sitting. So I think the monitors that can start proving information about sedentary time and activity time, that’s going to be a desirable combination.
[Damien Blenkinsopp]: And position, as you said earlier. Because right now most of them don’t tell if you’re standing or sitting. And there’s a fair amount of research, as I understand it now, that sitting’s a really bad idea. For our backs and other aspects. Whereas, at least it sounds like you and I are trying to stand, and do the standing desk thing, more because of that, but it’s not getting tracked?
[Dr. Greg Welk ]: Right. But most devices don’t really have a good sense of how to track sedentary time. The research community is actively pursuing that, but we haven’t really seen good output from most of the consumer monitors on sedentary time yet. I mean, I think they’re working on it. So I think eventually there will be these separate indicators.
And even on the smart watch, the prototype for Apple, they have a really cool interface I don’t know if you’ve seen that, that shows these loops. There would be a circle, and when you meet the day’s activity goal, your circle would complete, or you would get a full circle. And if you also minimize the time you spend sitting, you would also get a full circle. And the way they track that is you can’t have any hours where youíre sedentary for more than an hour. So basically it forces you, every hour, to remember to stand up and move, which is a good cue.
And so they have built that into the planned smart watch. Again, I haven’t gotten to try one myself yet, but that’s what I’ve seen in the prototypes.
[Damien Blenkinsopp]: Yeah, and if you have alerts. We have calendar alerts on our IPhone, if we have alerts on our watches that say hey, you’ve been sitting for an hour, it’s time to get up. The Basis watch actually has it, so every day it would say I was sitting for too long a period. That’s one of the other things it used to say was I was asleep whenever I was watching TV, or something like that.
Which is difficult to assess, when you are asleep versus watching TV or something very low key.
[Dr. Greg Welk ]: I think yeah, that’s a movement where the research community is actively interested in both the effects of promoting activity and minimizing sedentary time. I guess you brought up before that there’s not a consensus about the right type of activity, but I think among researchers and the public health community, we’re in agreement that more physical activity is better, but a moderate amount is sufficient for health. So if we can get the population to do 30 minutes of physical activity on most days of the week, that’s what we’re after. Unfortunately a large percentage still doesnít do that.
And then the addition of resistance training is a really important piece that isn’t really captured with these monitors, unfortunately. So that’s where we need people to be adding some resistance exercise to their routine, because that adds independent benefits also.
[Damien Blenkinsopp]: And the Amiigo I brought up, which is trying to do that. I’ve actually been watching for a year. I don’t think they’ve actually delivered any of the devices yet, it was an Indiegogo project. I could be wrong, I haven’t look at it for a while. But they’ve run into some manufacturing issues, and I’m not sure what the technology was there. But someone is trying to incorporate this resistance training aspect into it.
[Dr. Greg Welk ]: Yeah, I had seen a couple. And we were planning to incorporate them into that study I mentioned, about resistance training, but we couldn’t get them in time. I forgot the names of those, but they weren’t ready. So we just decided to go forward with the study with the monitors we had.
[Damien Blenkinsopp]: Great. So this is an ongoing work project, right? You’re going to be continuously looking at the new devices coming out, and tracking. Is this something that you’re going to continue working in your assessment center?
[Dr. Greg Welk ]: Oh, definitely. My students are very interested in that, so I have a whole fleet of grad students that help me do this work. They’re very interested in keeping up on the latest technologies. Our lab is called the Physical Activity Lab. So if you went to physicalactivitylab.org, we have, that’s where we post information on our past studies, and things like that.
But when we finished testing this one consumer study, we then completed another one in the fall that will be presented at a conference this spring. And that’s, again, in review now for publication, and we’re already planning another study this spring to compare the new smart watches, like you said, and also the new JawBone Up3, and the FitBit Charge. Those are the latest devices that, again, are released. Even before we can finish testing them, there’s new monitors that come out. So we’re very interested in testing the new JawBone Up3 that has a number of new features in it.
[Damien Blenkinsopp]: Yeah, that’s great to hear you’re going to be keeping up with it, so we can follow you and keep up-to-date on all of this. Because I haven’t seen anyone else covering this in so much detail.
I’d like to ask you a couple of personal questions about your use of data. In terms of biomarkers or personal data, do you track anything on a routine basis for yourself to either monitor or improve your health, longevity, or performance?
[Dr. Greg Welk ]: Yeah, I’m kind of a dabbler, probably, more than a regular user. So I dabble in them maybe because I’m more thinking about them from a research perspective. But as I think about it, the indicator that I’m probably most intrigued with is the sedentary time. So it operates in my mind all the time about how long I’m sitting, and trying to minimize sedentary time.
So, I’ve started becoming more interested in devices and assessments that can do that; either simple cell phone apps or logs that I can track my behavior with. So I used to do a lot with high level training and using heart rate monitors to track the intensity of my workouts and things like that, but I kind of know by now my own ranges, and if I’m working hard enough. So as an exercise physiologist, I kind of know what I need to know about my own training. But this whole sedentary world is a new frontier.
[Damien Blenkinsopp]: Great, great. Did you ever look at HRV? We’ve talked about this quite a bit on this show.
[Dr. Greg Welk ]: Yeah, early on I used to use a heart rate monitor and using HRV to track. It’s a good indicator of, over training, for example, if your body’s not rested from a previous night, a lot of athletes would use the advanced heart rate variability indicators to determine if their body’s fully rested and ready for the next workout. So at one point in my life I was very focused on triathlon training and did a lot of that monitoring. But recently it’s not been as high of a priority for me.
[Damien Blenkinsopp]: Great, thank you. What would be your biggest recommendation or insight on how people could use data effectively? Because there’s more and more data coming about, and especially with these devices. How would you suggest that they use this data effectively to improve any aspect of their life?
[Dr. Greg Welk ]: Good point, and I think that’s where the goal would be for using the device to help you change your behavior, and to use it to inform decisions, not just to collect data for the sake of collecting data. So, I think as the devices become more sophisticated they’ll provide actionable insights that give you prompt, that says last week you were active in the mornings, or you weren’t active in the mornings. And it would cue you to maybe be more active at different periods of the day.
So I think that’s where it’s got to get, where the devices are basically helping people to make changes. I mean, we live in a very challenging environment to be active today, with people using cars, and the busyness of our lifestyles. And so I think the monitors provide a way to keep people accountable to their goals, and realize that they have to build physical activity into their day; it doesn’t just happen. It has to be actively planned and promoted during the day.
[Damien Blenkinsopp]: Great, thank you so much for that. I’d really love to have you back on in say a year or something. Would you mind if we reached out and, obviously, you’re going to be seeing the trends as they develop, and I’m sure you’ll have a completely different idea and research on where things have got to in a year’s time.
[Dr. Greg Welk ]: Yeah, I’d be happy to. I’m actually, we’re working on our health coaching models, we’re testing some of the new devices in the spring, and we’re also including expanding our health coaching model, where we’ll be testing this new patch device from a physical physician referral system. I think that’s where the key is. And we’ll be able to learn from these studies what type of information consumers need to change their behavior, and what level of involvement do they need.
So a lot of our goals in our health coaching studies is to figure out a very cost effective solution. For example, can we give people a monitor and a few tips every week, and even text messages. Is that enough? Or do we need to give them a one on one counseling session, three weeks in a row, and then put them on their own?
And so we’re trying to figure out that optimal dose and scope of what a behavior change component needs to be. And, you know, we’ll have a lot more to know by next year.
[Damien Blenkinsopp]: Great. Yeah, and so you bring up a lot of the challenge of using these is understanding the psychology and the behavioral implications of it. So this is the more complex area to look into, really. Analyzing, as you say, how to report the data of people on frequency, the format, and what actually influences behavior and gets us to the end goal.
Well great Greg, thank you so much for your time today, and looking forward to talk to you again as your research expands and continues to look at this area.
[Dr. Greg Welk ]: Yeah, I’d be happy to join, and I’ve been enjoying following the blogs, and I look forward to continuing the discussion.