How Accurate is Fitbit? Here is What We Found
How accurate is Fitbit? After a rather impressive run at the helm of wrist wearables, and great-looking bottom lines, FitBit got an egg in the face with a conspicuous lawsuit in the first week of 2016 for accuracy issues—heart rate accuracy issues of the Charge HR and Surge.
Apparently, your FitBit will make you fat, less heart friendly and anything in between, or conspire with some junk food to smack you with an outright, fatal stroke in your sleep or run.
Is FitBit Health Accuracy a Joke?
The FitBit accuracy lawsuit reminds us of the first few novel months of the pioneering generation of health bands. Back when the first bunch of health bracelets came out, everyone wanted to sport one, or two. After just a couple of weeks of use, about 1 in 2 people were losing their adulation for them. They figured out a smartphone could do a better, much better job.
At the same time, scientists proved the fancy wristbands sucked, despite most offering only step tracking. That was then, and most of the new crop have a bevy of improved sensors stuck on the still fancy frames to “accurately” sense changes in motions of the physical body. FitBit claims that its health trackers can precisely read physiologic values such as activity, calories burned, energy expenditure, heart rate, sleep, and so on.
But just before you resolve to blow the dust off that chest strap and toss out your FitBit health band, how accurate is your FitBit really? And do the trackers even work? Heck, how fit is your FitBit?
Let’s find out.
1. How Accurate is FitBit’s Heart Rate Tracking?
To start this off is the contested, wireless heart rate monitoring capability of FitBit’s two devices, the FitBit Charge HR, and the equally priced FitBit Surge.
FitBit uses an in-house PurePulse Technology to measure heart rate. PurePulse utilises Photoplethysmography, or optical heart rate monitoring. In a nutshell, what that does is use LEDs located on the bottom side of the device to light through the skin and “see” blood flow. Depending on how fast blood flows beneath, the device’s algorithm determines your pulse rate, which is not unlike the tech used on the Apple Watch.
FitBit’s proprietary heart rate tech calls the shots in a different tune—very unlike the electrocardiogram (ECG) tech used with the “proven accuracy” of the Polar H7 chest strap. That is not to mean the ECG technology can scoop all awards for accuracy either, but because it is widely considered by clinicians as the standard bar of accuracy for heart rate monitoring.
According to Jonathan Selbin, the lead partner for the complainants, when they used the two FitBit devices, they sometimes failed to record a heartbeat. Selbin and Co. claimed in this source document that FitBit’s readings erred and were off by a “wildly inaccurate” 75 beats per minute (bpm) in extreme exercising intensity and 24 bpm in less vigorous activity. The group claimed it had a board-certified cardiologist examine the claims before targeting the lawsuit at FitBit, and their doubts were confirmed.
To verify these, Consumer Reports, the US-based consumer products examiner, re-tested the Charge HR and Surge for accuracy. CR has previously vouched for the two Fit.
What Consumer Reports Found Out
Consumer Reports enrolled the help of two volunteers for testing, a male, and female. To further dig into the nitty gritty, CR had both volunteers wrap up two copies of the same model on the same arm, one right around the wrist and another a few inches above. That was probably because FitBit, in its PurePulse Tips Page, encourages users to secure their device at the right position—a few inches above the wrist.
Tests were conducted in low-intensity activity (walking and at 110 bpm) up to extreme activity (running and at 150 bpm)—CR took up to 64 heart rate measurements. The plaintiff had stated that the Charge HR and Surge erred while the users engaged in an extreme activity. That data was then compared to data recorded by an ECG based Polar H7 chest strap heart rate monitor.
According to the test, both the Charge HR and Surge are on point. However, FitBit’s jewels did crawl behind the Polar H7 by up to three beats per minute. But Consumer Reports consider this “negligible” for most people.
So there you have it, that FitBit’s heart rate sensor is not dying to kill you.
2. How accurate is FitBit Sleep Tracking?
The next hot issue spews out of discussing the accuracy of any FitBit sleep tracker.
In May 2015, soon after a James P. Brickman filed a similar class action lawsuit based on FitBit sleep tracking accuracy, here is what FitBit stated to a popular wearables media outlet: “FitBit trackers are not intended to be scientific or medical devices.” With this in mind, let’s find out if your FitBit deserves a place on your wrist.
A 2012 study from West Virginia University compared “a FitBit device” with an ActiWatch-64 and found that both “novel activity monitoring” devices showed inaccuracies in sleep tracking by overestimating both how well and how long one slept.
According to the study, FitBit sensitivity was higher than the actigraphy-based ActiWatch-64 in all sleep stages and during arousals. For one, FitBit overestimated sleep by a concerning 67 minutes. Still, both sleep tracking devices were consistently similar in good performance and slip ups. However, that was almost four years ago, and a lot may have since improved or slipped. Nevertheless, tracking sleep time and quality is a whole other ball game as evidenced by real-life use.
The Brickman complaint targeted FitBit trackers with a sleep tracking function. That wrapped almost every FitBit activity and sleep tracker in a cold blanket, right from the first-generation FitBit Flex up to the second-generation FitBit Charge HR. All except the FitBit Zip, which doesn’t support sleep tracking.
First, FitBit sleep trackers utilise a 3-axis accelerometer to determine if you are asleep or not. If your movements are still for at least an hour, the device assumes (best word) you are actually asleep. However, if you are just lying around sleepless, it would mean the sleep tracking stats collected automatically during that time will indicate the complete opposite.
Charge, Charge HR and Surge
Second-generation FitBit devices such as the Charge HR, Surge, and Charge activate sleeping mode automatically you hit the sheets. Come morning, and the Charge, or HR, or Surge may indicate a 100% total sleep time—even if it took you some 30 minutes or more to actually fall asleep. The formula used to calculate sleep quality is as follows: 100 x Time asleep/Time Asleep + Time Restless + Time Awoken during Sleep. That leaves out consideration for time lying still but sleepless. It is also virtually impractical for the device to establish at what point you fall asleep. For the FitBit Charge HR, Charge, and Surge this is a particular issue thanks to the automatic function for detecting sleeping mode.
Stats on sleep quality can get a little-to-wildly messier, though. How accurately an accelerometer detects restlessness is simply by sensing movements during the time you record as having been sleeping. Moreover, you have to tap on the FitBit Flex to let the device know that you want to sleep—and nap. If your nap does not exceed an hour, you have to input that time manually on your FitBit dashboard. Otherwise, the nap goes unrecorded.
For the analogue types, such as the FitBit Flex, one must tap on the head for about 2 seconds before the device indicates with two dimming lights that it is switching to sleep mode. Forget to do this, and the next morning your dashboard will indicate that you never did slumber the night before. Moreover, arousing in a hurry and forgetting to tap the Flex rapidly for about 2 seconds, and only remembering later that you should have immediately you woke up, means your sleep time will overshoot the reality. At arousal, the Flex has to be tapped twice, vibrate 3 times and indicate a spinning light to deactivate sleeping mode.
Fitbit One and Ultra
For the One and Ultra, one has to press the physical button until the device indicates a stopwatch counter, or the icons start to blink to indicate a switch to sleeping mode. To wake it, hold the tracker’s button for 2 seconds to accurately deactivate sleep mode.
Still, FitBit sleep tracking is among the best—if not the best—according to Consumer Reports testing.
3. How Accurate is FitBit Step Tracking?
Step tracking is enabled on the devices by design by an inbuilt altimeter. An altimeter measures steps and floors climbed based on detected changes in altitude. And things aren’t always stepping up in the right direction, according to real-life user reviews such as in here
Walking, driving, running and just about any “stepping” activity you engage in on a bumpy road can lead your FitBit to inaccurately register extra steps. Simple, inevitable activities such as opening the door, which may let in a gush of wind and destabilise indoor altitude, can lead to your FitBit tracker to register that you have climbed another floor, for example. FitBit itself admits that sometimes your device’s accuracy will not work as you’d expect.
Most step trackers on the market right now still get it wrong about 10 percent of the time. Under “normal” circumstances, though, your FitBit is fit for the job. The health wristbands maker insists that the max operating altitude for its devices is 9.144 km. And that the optimal temperature to use the Charge HR, for example, is between -20 degree Celsius and 45 degree Celsius. Of course, anywhere out of these limits will not give the accurate readings you might expect.
In a past study that was published in the journal Medicine and Science in Sports and Exercise, the lead researcher confirmed, by comparing different FitBit-style trackers, that FitBit’s devices may not be that gaping. According to Gregory Welk, the FitBit One (10.1%) error rating and Zip (10.4%) scored higher than the likes of Jawbone Up (12.2% error rating), Nike FuelBand (13%) and Basis Band (23.5%). BodyMedia’s FIT hit top marks and the overall best, showing an error rating of only 9.3%.
4. How Accurate is FitBit GPS?
The Surge has in-built GPS tracking function. The best way to utilise GPS functionality is to measure distance. Otherwise, using it for step tracking is not exactly the most accurate extravaganza you will ever speak of. GPS signals strength shifts quite a bit almost every minute. But that is largely since the function relies on constantly moving, man-made satellites orbiting millions of kilometres up around space. So your signal strength will determine how accurate your distance tracking stats out. There is probably very little FitBit can do to rectify that even if they wanted.
Tips to Improve the Accuracy of your FitBit Heart Rate tracker
- Ensure your health band is not secured too tightly around the arm. That would restrict and interfere with blood floor. Meaning you might end up registering a mistakenly low heart rate and end up pushing yourself too hard
- If engaging in strenuous exercise sessions, ensure you rest your arm a bit to ensure the FitBit device catches your pulse. Having your arms twisted bar the wrist band from correctly picking up on your blood flow.
- Wear your (Charge HR) a finger’s width away from your wrist bone. At that point, blood flows better and you can get a better heart rate reading from higher up your arm than at the wrist bone.
- The Charge HR is optimised for wrist health, activity and sleep monitoring. What that means is that it can show drastically different results for, say, step tracking than other FitBit products, especially if you happen to move your wrist quite a bit. The fresh off the garage, FitBit Blaze SmartWatch will most likely show this trait as well. The Surge, with its bevy of motion sensors, may also fall a bit on the overestimates side, if you will.
How Accurate is Fitbit? A Conclusion
FitBit’s line of activity, heart rate and sleep trackers is not perfect. Sometimes they will want to misbehave where you least expect them to. Like when you are driving and the Charge “thinks” you are hiking, simply because you trudged a water bottle somewhere along the highway. However, as tests by credible organisations such as Consumer Reports indicate, there is no much cause to worry over the accuracy of your favourite FitBit device.
See recommended fitness trackers on Amazon
Fitbit Charge 2
Garmin Vivosmart HR+
Misfit Shine 2
Xiaomi Mi Band 2