Smartwatches appear to be very inaccurate in measuring energy consumption. This raises questions about the added value of this functionality.
Smartwatch and Energy Consumption
Well, that’s disappointing. You think you’re tracking your energy consumption, only to receive inaccurate data.
Researchers from Stanford University tested a total of 6 different smartwatches; the Samsung Gear 2, Apple Watch, Fitbit Surge, Mio, Alpha 2, PulseOn, and Basis Peak [1]. They had 60 participants wear up to four of the smartwatches simultaneously and compared the results with those of professional equipment for measuring energy consumption and heart rate.
All seven tested smartwatches were found to deviate by at least 20% from the energy consumption as tested by the professional equipment. In the ‘best’ case (Fitbit Surge), the smartwatch showed a deviation of 27%. The PulseOn was off by a whopping 93%! Heart rate measurement proved to be easier.
Not so Smartwatch
Why are smartwatches so far off?
This can go wrong during measurement and during ‘knowledge’. Because we’re actually asking them for information for which they receive too little input, smartwatches have to make a lot of assumptions. Moreover, they lack the resources for accurate measurements for the information they themselves have to provide.
Smartwatches have to measure how much you move and then calculate how much energy that movement costs in your specific case. The measurement itself is done by measuring movement and heart rate. So, it’s not surprising that energy consumption during sedentary activities showed a larger margin of error (average 50%). But even during activities like running and walking, the smartwatches were on average 30% off.
In addition to the measurement itself, the movement has to be converted into energy consumption. For this, smartwatches have to make calculations based on algorithms that use variables such as age, gender, height, and weight.
So, they remain estimates based on incomplete measurements.
If it Fits Your Smartwatch
In some cases, smartwatches can also show you the total number of calories burned per day, your resting metabolism plus activity consumption.
You may wonder how accurate this is if the separate activities are measured incorrectly.
The question then is what the added value is. Does this provide you with more insight into your calorie needs than, for example, already working with formulas such as Harris-Benedict and Katch-McArdle in the Fitsociety app?
And what formulas do smartwatches actually use for your resting metabolism?
You would actually want to know this to possibly verify the justification yourself. For these types of formulas, you normally enter your variables and calculate on a daily basis what your estimated consumption is. The majority of your consumption comes from your resting metabolism. The additional consumption from activity is then roughly estimated, for example in the case of the app. You indicate how active you are by choosing from different activity levels. Based on this activity level, your consumption is increased by a certain factor.
‘Not very specific,’ you could say. “Light training” and “heavy training” are quite broad terms and the definition of those varies from person to person. Moreover, you may train “heavier” one day than another. And how does that work if you’re a mail carrier? Does that count as light training? In practice, this works as trial and error. You choose an activity level and calculate your needs, you stick to your diet based on this, and then check whether the results are in line with expectations. Then you can decide whether your best estimate was correct and make adjustments if necessary.
That doesn’t sound very scientific either. Why do so many fitness professionals still work with this method? Because obtaining accurate data is very difficult without living like a lab rat. Walking around all day with a mask on to measure your oxygen consumption is not exactly practical. Handy at a Star Wars convention perhaps, but less so at other social gatherings.
We would very much like to base your personal energy consumption in the FITsociety App on specific activity. How much you burn when doing a certain form of cardio, for example, but also the number of calories burned per rep during strength training. However, we have decided that the chance of receiving incorrect data is too high. We do not want to create the illusion that this can be calculated so accurately.
Even worse; that you use incorrect data to set up your diet plan.
Strength training poses some challenges when it comes to calculating energy consumption compared to cardio. There have been several studies on this, albeit with different results [2]. However, there are too many variables. How heavy is the weight compared to your body weight and experience? How is the exercise performed? Are you using the best technique to lift the weight (weightlifting) or the best technique to stimulate your muscles to grow (bodybuilding)? Are you performing the exercise explosively, are you using a lot or little muscle mass for the exercise? We would have to conduct dozens of studies to calculate the energy consumption for a single exercise under all these different conditions. And then you’re back to estimating once those values are established; ‘does this count as an explosive repetition, is it 60 or 70% of my 1RM?’ That will not make your training run smoothly either.
Cardio has fewer variables and should therefore be more accurately estimated [3], but apparently, the smartwatches still struggle with this.
The perfect method, accurate and practical, to measure your exact energy consumption, does not (yet) exist. So, the question is whether you prefer to rely on your own estimates or those of a smartwatch. For me, that’s not a difficult question.
Motivation Tool
But I also don’t think the added value of the smartwatch lies in this, or should lie in this. When I tried out the Apple Watch for a review, I let my daughter try it for a while too. This resulted in us suddenly hearing stomping from upstairs at 11:30 because she wanted to achieve certain activity goals. A nice gadget that reminds you regularly that you want to achieve certain goals and shows this in a visually appealing way can apparently be very motivating.
Research from last year nonetheless showed that fitness trackers do not add value compared to a normal weight loss program [4]. The researchers also state that their test group is not representative for many people, so they also do not want to claim that fitness trackers do not work. Some people in the study did benefit from them.
However, when it comes to motivation, I personally think you’re better off with a ‘point system’ than an inaccurate representation of energy consumption. Linked to your agenda, for example. “Congratulations! You’ve earned 1000 points by running. Only 300 points to go and you can go to dinner with your sister”.
Consistency is important here. You have to be able to rely on whether you have actually moved more or less. For motivation, this does not necessarily need to be translated into presumed calories.
What are your experiences with smartwatches? Did they motivate you? Did they provide insight into your energy consumption and were you able to compare this with other data? I’d love to hear about it.
References
- J. Pers. Med. 2017, 7(2), 3; doi:10.3390/jpm7020003
- Robergs RA, Gordon T, Reynolds J, Walker TB. Energy expenditure during benchpress and squat exercises. J Strength Cond Res. 2007 Feb;21(1):123-30. PubMed PMID: 17313290.
- McARDLE, W.D. et al. (2000) Energy expenditure at rest and during physical activity. In: McARDLE, W.D. et al., 2nd ed. Essentials of Exercise Physiology, USA: Lippincott Williams and Wilkins
- Jakicic JM, Davis KK, Rogers RJ, King WC, Marcus MD, Helsel D, Rickman AD, Wahed AS, Belle SH. Effect of Wearable Technology Combined With a Lifestyle Intervention on Long-term Weight LossThe IDEA Randomized Clinical Trial. JAMA. 2016;316(11):1161-1171. doi:10.1001/jama.2016.12858
