The ubiquitous nature of “wearables” is undeniably here to stay. According to Statista the wearable market is forecast for shipments of ~560 million devices in 2024, with growth projected beyond this! Sure, there are folks who want less data just like there are folks who want a ‘dumb phone’ but the reality is these wearables are going to become all pervasive as part of the “quantified self” movement.
This is a societal trend, though, and will extend well beyond consumer wearables. I have previously written about healthcare and it’s potential use of wearables. My hunch is that we will see companies doing a better job of managing data and metric overload, perhaps using a more ‘lurking’ type of strategy whereby they only really make metrics known when they’re abnormal (this isn’t a great retention strategy though so perhaps this is naive or too rose tinted glasses of me). This approach may have advantages when it comes to managing some people’s aversion to wearables and of course the aforementioned data/metric overload.
Regardless of all of this, the ubiquity of wearables means they’ve stepped out of the niche of the technophiles of the world and into the mainstream. This has brought with it many challenges for all parties; consumers, professionals working with consumers and the companies themselves. Hopefully this piece of writing helps the reader to choose the right wearable, for the right purpose at the right time and get the most out of it (which also may mean not using a wearable).
*A quick note, I will be using the term ‘wearables’ but this article really speaks to any form of biometric tracker.
Something I have noticed in coaching, is that the ubiquity of technology has meant that the generation of younger athletes who’ve grown up with technology are less developed when it comes to the intuition. As a specific example, I grew up running (before GPS watches) with a guy who could tell distance and pace quite intuitively (I certainly couldn’t - more on this soon), this seems to be a dying skill. Athletes are often struggling to ‘feel’ things they perhaps could previously, being more reliant on technology, which is a fragile existence in case of technology failure but also less than optimal (see below information on approach to using wearables). This is, however, not about being the old man shaking his fist at the sky, it’s an acknowledgement of reality. A reality with different circumstances and metaphorical ‘roads that lead to Rome’. I mentioned before that I could never feel pace or distance whereas my training partner could. Ironically, the very technology I mention driving a disconnect between the athletes and their perceptions is what’s helped me solve this as I have aged. The feedback offered by my GPS has helped my dial in my perception (this needs to be an active process though, it doesn’t happen via osmosis).
Who Shouldn’t Use Wearables
A lesson I remember learning early in my internship, whilst on my orthopaedics rotation, was that the most important decision a surgeon can make is who NOT to operate on. It in this spirit that we need to think about wearables when it comes to personality types because there are certainly groups of people who can struggle with wearables for various reasons.
Given the duration that wearable technology has now been available to consumers, we are starting to see pathology associated with it. These generally fall under the umbrella of psychological concern and obsession over the data impairing quality of life. For example orthosomnia; the obsession about sleep metrics which can negatively impact sleep itself.
Similarly, there is research that data can do some harm, with some interesting studies showing fake ‘negative’ sleep data impaired performance! Fascinating, I know! (If this interests you, make sure you are subscribed as a two part series on the placebo effect is coming).
To answer the question of who should not use wearables specifically; those with obsessive personalities, who do (or may) struggle with not being impacted by the data. Basically, those who don’t just see it as an interesting data point. Or as some authors suggest: “a perfectionist athlete with personality traits characterized by excessively high personal standards and extreme critical self-evaluation”.
I would add to this, that in some cases, particularly in newer wearables (earlier in the product cycle), those who do not have the time or will to really understand the data may be best served to avoid these wearables until their user interface and user experience is further advanced and can engineer this need out (at least to a degree).
The Best Way to Approach Using Wearables
As I covered in this article about continuous lactate monitors with respect to advice for new users there are a few fundamental concepts that should be a starting point for consumer wearables:
Spend your initial period of use observing only, do not change anything. Try to live your normal life.
Start from the assumption that you data is normal (unless you have good reason to believe otherwise, which is NOT the new device. For example feeling like you sleep poorly and thus buying a sleep tracker).
Contextualise your data, using other data sources and understand sources of error in the data and what may influence them. *See below section on “important concepts in wearables”
Characterise the full range of your data (including the extremes, either via experimentation or experience).
Once you understand all of this, you can start to modify factors you think may warrant modification, using the data as feedback IN ADDITION to how you feel - feelings and perception are another data point and should be used in addition to other data sources.
DO NOT fall prey to Goodhart’s Law (which states: “When a measure becomes the goal it ceases to be a good measure”). Remember the goal is improvement in performance or health; not data (though hopefully the data reflects this too).
Important Concepts in Wearables
There are a few important concepts in wearables that users need to understand to ensure they are getting the most out of their wearables and not misunderstanding or misusing them.
Measured vs Estimated
Hat tip to Marco Altini on this one, check out his writing on this here (also subscribe to his Substack if you’re interested in HRV).
In essence, wearables will generally measure a number of metrics (perhaps reporting them) and then estimate others as a result.
For example a watch, band or ring will usually measure heart rate (and depending on the sensor HRV, a derivative of heart rate) and movement (via accelerometers). Some will measure skin temperature and oxygen saturation (this is a little more nuanced as a measure so let’s ignore it for the purposes of this piece - if you have questions use the comments section on substack and I will answer there).
So generally we have:
Measured: Heart Rate, Heart Rate Variability (HRV), Movement, Oxygen Saturation, Skin Temperature, GPS
Estimated: Respiratory Rate, Sleep, Steps, Calories, VO2Max
One thing to note on estimation, is that these algorithms are generally built on lines of best fit and often on white middle aged men. This is important as populations outside of this may not be well represented. Classically optical heart rate sensor struggle with darker skin tones (including tattoos). Similarly, these lines of best fit often fit middle ranges of physiology, so high or low values may not be as well estimated.
How is it Measured?
This may sound absurd, but not all measurement techniques are the same and as a result there can be different sources of error and limitations to applicability of data.
Perhaps the best example here is heart rate, and not just because people understand the metric. Heart rate is classically reported in the ‘units beats per minute’ (bpm). This is important because sampling can influence the data significantly. The instantaneous heart rate you see at any moment may be an average of a period prior to that (the last minute, 10 secs, 5 secs etc) or a truly instantaneous heart rate based on the duration between the last 2 heart beats (quite uncommon to my knowledge). Even a rudimentary understanding of mathematics means you’d understand the differences here, which are especially pronounced during phases of rapid change in heart rate (you know like during exercise). Similarly heart rate can be measured in two different ways: electrically (true heart rate) or via optical sensors (aka PPG) which is actually a pulse rate. Of course these two different measurement techniques yield potential different sources of error. PPG uses changes in light as a result of arterial wall movement to measure your pulse rate (which is the output of your heart rate), but any source of light creeping in or movement of the sensor means that there’s a potential error in the data (this is why these sensors are usually suggested to be worn quite tight to the skin). Comparatively electrical heart rate monitors (usually chest straps) don’t measure pulse and only the electrical activity of the heart which can (in rare situation) vary a little from the heart’s pumping.
Another example is step counters, they generally use a rhythmic movement as measured by the accelerometers in the device (usually on an arm). As a result, not swinging that arm means you won’t record any steps done, and rhythmic activities with that arm may yield steps you aren’t doing (you’d be surprised how far I ‘walk’ when hand grinding coffee).
This may seem extreme and pedantic, but without understanding measurement you can’t understand sources of error and thus whether there’s a reason the data may be wrong (and how it could be so). Similarly, understanding this and pairing it with the estimation considerations above shows potentially compounding sources of error.
How to Choose Which Wearable(s) to Use
This may be a little repetitive for some regular readers, but I assure you it is due to the importance of the concept. The very first question to ask yourself when thinking about any tracking technology is “why”. Why do you want it?
Then the related questions: What will it add? and How will you use the data?
Hopefully the answers reveal the type of data you want to track and how it aligns with your health and/or performance goals. From there it boils down to comparing options and their relative pros and cons. Of course, cost is one of the factors to consider, beyond this I would consider a few other factors such as research (is the device valid, accurate and reliable) and what form factor would you prefer (rings, bands, watches, etc). It may even be that you’d prefer not to have a wearable at all, for example mattresses can measure sleep and apps can measure HRV.
To be clear, sometimes these devices can be motivating themselves: “I want to exercise to use my new tracker” for example. If this is you and this is the case, great, it’s a small price to pay for a large benefit in my mind. That said, make sure you’re honest with yourself and at least try to use the data, not to mention try to build the habit because the new, shiny thing, sometimes doesn’t stay that way too long.
Common Questions About Wearables
These are some questions I have been asked multiple times, which is usually the trigger for me to write about something.
How Accurate are Calories on my Wearable?
In short: Not.
This is through no fault of the wearable device, but calorie burn estimates are exceedingly difficult because the act of training changes calorie burn and it differs vastly between individuals. There are multiple different methods that could be used to estimate calories, but unfortunately none of them particularly accurate. The only real way to measure calories burned is via gas exchange (measuring the amount of oxygen you are breathing in and carbon dioxide you’re breathing out).
That’s not to say your calorie tracker is entirely useless, a significant change is probably meaningful as an indicator of movement and perhaps energy expenditure for the day. They may even assist in trying to create a caloric deficit but again, I’d not be married to the specific number as much as trends.
If You Could Only Have One Wearable, Which Would it Be?
This is a no brainer for a runner like me, it’s my GPS watch. That said, I am super analogue and prefer to notifications, texts etc to come through on it (I need no more distractions). For most folks a watch is probably the answer, they’re currently the most versatile on the market, most can do some variety of sleep, perhaps HRV and then any amount of tracking. But it’s pretty individual really.
If you’re interested in some more of my thoughts on wearables (particularly as they relate to athletic performance) Sean and I chatted about them quite a bit early on in my episode with him on the Upside Strength Podcast (Spotify Apple)
References
Jahrami H, Trabelsi K, Vitiello MV, BaHammam AS. The Tale of Orthosomnia: I Am so Good at Sleeping that I Can Do It with My Eyes Closed and My Fitness Tracker on Me. Nat Sci Sleep. 2023 Jan 21;15:13-15. doi: 10.2147/NSS.S402694. PMID: 36713639; PMCID: PMC9875581.
Trabelsi K, BaHammam AS, Chtourou H, Jahrami H, Vitiello MV. The good, the bad, and the ugly of consumer sleep technologies use among athletes: A call for action. J Sport Health Sci. 2023 Jul;12(4):486-488. doi: 10.1016/j.jshs.2023.02.005. Epub 2023 Mar 1. PMID: 36868375; PMCID: PMC10362482.
Draganich C, Erdal K. Placebo sleep affects cognitive functioning. J Exp Psychol Learn Mem Cogn. 2014 May;40(3):857-64. doi: 10.1037/a0035546. Epub 2014 Jan 13. PMID: 24417326.
Gavriloff D, Sheaves B, Juss A, Espie CA, Miller CB, Kyle SD. Sham sleep feedback delivered via actigraphy biases daytime symptom reports in people with insomnia: Implications for insomnia disorder and wearable devices. J Sleep Res. 2018 Dec;27(6):e12726. doi: 10.1111/jsr.12726. Epub 2018 Jul 10. PMID: 29989248.
Bent B, Goldstein BA, Kibbe WA, Dunn JP. Investigating sources of inaccuracy in wearable optical heart rate sensors. NPJ Digit Med. 2020 Feb 10;3:18. doi: 10.1038/s41746-020-0226-6. PMID: 32047863; PMCID: PMC7010823.
A. Puranen, T. Halkola, O. Kirkeby and A. Vehkaoja, "Effect of skin tone and activity on the performance of wrist-worn optical beat-to-beat heart rate monitoring," 2020 IEEE SENSORS, Rotterdam, Netherlands, 2020, pp. 1-4, doi: 10.1109/SENSORS47125.2020.9278523.
The "why" in the purchasing decision that u flag is super important. I find talking with athletes (or my own thoughts) it can be surprisingly difficult to identify especially with the influence of others or marketing. Or interestingly also difficult to influence once the purchase decision has been made but not yet bought. It can be quite enlightening to speak again on purchases 4-8weeks afterwards and see just how much is "dead data" recording