Reminder: When it comes to heart-rate tracking, consumer fitness trackers are not medically accurate

Every few months we see studies about how fitness trackers aren't accurate, or how they won't help you lose weight. The latest comes from The Journal of the American Medical Association, or JAMA as it's more commonly referred to.

The peer-reviewed medical journal published a research letter questioning the accuracy of wrist-worn heart-rate monitors. These studies aren't necessarily groundbreaking, but they tend to get significant media coverage. We know fitness trackers aren't perfect. None of these are medical devices and they aren't designed to replace a visit to the doctor. But the data they collect (even if it isn't 100 percent accurate) can still be useful.

Understanding heart-rate data

Fitbit , Garmin , Mio, Polar, Apple and many others use a similar technique when it comes to measuring heart rate -- flickering green LED lights illuminate the capillaries while a sensor measures the blood as it flows past.

Accuracy for these optical sensors will vary from person to person. This is due to factors such as fit, position, sweat, hair density and sometimes even skin color. The tracker should be positioned a finger width away (two fingers when working out) from the wrist bone. It should be tight, but not that tight where it's going to leave an indentation on your skin. A tracker that is too tight could affect blood flow and actually have a negative impact on accuracy.

Many of the top-selling fitness trackers with optical heart-rate sensors are accurate when it comes to resting heart rate. Results can vary during workouts, though, and that's exactly what the JAMA study focused on. In my testing, the Garmin Vivoactive HR and Fitbit Surge were fairly accurate on an easy run, but both struggled a bit (the Surge more than the Garmin) during a hard interval run when compared to data collected from a Polar chest strap. You can view the results below (click here to view an interactive version of the chart):

Garmin, Polar and Apple allow you to pair a chest strap for more accurate heart rate readings, which I recommend when doing a hard interval workout. Fitbit doesn't provide this option.

While you can run with the Fitbit Charge 2 , most owners don't. It doesn't have GPS built-in and you are required to carry your smartphone. A majority of Charge 2 owners use the device to track all-day activities like steps and heart rate. For that reason, I decided to test the heart-rate sensor based on real-world activities. I took my dog for a walk, ate lunch and relaxed on the couch, all while also wearing a chest strap.

The results were relatively accurate. There were times it was off slightly, but the Charge 2 performed well. The tracker averaged my heart rate at 65 bpm, compared to 67 bpm recorded on the chest trap. You can view the full results below (click here to view an interactive version of the chart):

Why 100 percent accuracy isn't a deal breaker

Heart-rate tracking is useful on fitness trackers for two reasons: it helps the tracker better estimate calorie burn and improves sleep tracking measurements. And, when used properly, it can help you gain a better understanding about your overall health. Optical heart-rate sensors also provide a convenience factor you don't get with a chest strap: they can be used continuously for days, months, or even years (with charging), compared to a chest strap that you would only wear for the length of your workout.

One benefit is being able to track resting heart rate, which is the number of times your heart beats per minute when you're at rest. (Note that your heart rate will dip even lower when sleeping). Trackers from Fitbit, Garmin and Apple are all able to determine resting heart rate.

But keep in mind that different trackers and smartwatches sample your heart rate at different rates. Fitbit and Mio offer continuous tracking on 1-second intervals, while Garmin varies its sampling rate depending on your activity. The Apple Watch measures heart rate every 10 minutes (unless working out, when it's measured one a 1-second interval). Despite these varied sampling rates, resting heart-rate measurements on most of the top trackers remain pretty consistent. I have found the resting heart-rate results from the Apple Watch, and various Garmin and Fitbit trackers to be on par with those results from my doctor.

Here's why resting heart rate is helpful. Mine is between 42 bpm and 46 bpm. It's not something to worry about when my resting heart rate is off by a few beats, but when it's at 60 bpm, I know something is wrong. Either I am overtraining and need to take a day off from running, or I could be stressed from work and it may be time for some relaxation.

Now as we've said: these aren't medical devices, and as the JAMA study found, optical heart-rate sensors aren't 100 percent accurate. But that shouldn't worry you. A tracker that is off by 5 to 10 bpm is still good enough for tracking resting heart-rate trends, and even for everyday fitness tracking.

Having the ability to view heart-rate trends can help you detect abnormalities. It could even save your life. That's what happened to an Apple Watch user in the Massachusetts and a Fitbit owner in New Jersey. But none of that is a replacement for a cardiologist or an annual physical exam.

"We believe optimal heart rate monitoring for those striving to reach their health and fitness goals is best achieved by taking heart rate across an extended time frame, such as during the entire task of interest, to get more reliable and meaningful outcomes," a Fitbit spokesperson explained. "Instantaneous heart rate readings at a pre-determined time point can be more subject to error (e.g., human error and random mechanical disturbances through unwanted movement)."

When asked about Fitbit's internal accuracy testing, the spokesperson added that the company "conducted extensive internal studies with more than 60 subjects, which show that Fitbit's PurePulse technology performs to industry standard expectations for optical heart rate on the wrist with an average absolute error of less than 6 bpm and an average percent error of less than 6%."

The same is true for steps

So what about step counter accuracy? The same information can be applied. Each company and each device will give you a different reading. You may get false reads, such as when doing the dishes or even when sitting on the bus, but the accuracy of the device is less important than is consistency.

If you stick with any one device, and it's 5 percent off every day, you'll at least have a record of your daily progress relative to your previous performance. (Your bathroom scale may not be providing your "actual" weight either, but if it's consistent, it should be letting you know if you're gaining or losing weight with each weigh-in.) The idea behind using these devices is simply to motivate you to walk more each day. It's not a big deal if it's off by 100 or 150 steps.

Of course, heart-rate accuracy can matter a lot more for intense workouts or specific cardio training, or in instances where real medical advice or monitoring is needed. But it's a reminder that many fitness trackers aren't so much about being fully accurate as they are about being relative guides to daily progress.

Which trackers are accurate?

All of these devices are subject to an endless array of software and firmware updates, in which the manufacturers are constantly tweaking and modifying algorithms and measurement methods. As the charts above show, heart rate accuracy varies from device to device. When we review fitness trackers and smartwatches, one of the things we focus on is heart-rate accuracy. If there is anything out of the ordinary, you can bet it will show up in our review.

For now, I am comfortable recommending all of the trackers mentioned in this article. The Fitbit Charge 2, the Apple Watch and various Garmin watches ( Forerunner 235 , 735XT, Fenix 3 HR and Vivoactive HR ), but be sure to keep an eye on our wearable reviews section.