One of the huge innovations set to change the wearables landscape is blood glucose tracking, which is being worked on by pretty much ever wearables company going.
While some years from arriving on our smartwatches, Apple is rumored to be working on glucose tracking, and we’ve reported on a host of other projects, all trying to bring glucose to the wrist.
Diabetes is a huge global health issue, with some 33% of US adults classed as pre-diabetic. Offering glucose insights from smartwatches would catapult wearables from casual wellness gadgets into essential consumer medical devices.
What’s more, established brands such as Abbott, which makes glucose monitors, as well as Nutrisense, Levels, and January, are all looking to offer glucose tracking to general, non-diabetic consumers.
But it made us curious. What’s it like to track glucose? Is there anything to learn for non-diabetics?
I set about two weeks of tracking to find out.
> Inside Afon: The UK company beating Apple to blood glucose
How glucose tracking works
We’re still years away from consumer non-invasive glucose tracking from the wrist, so there are two main ways of monitoring glucose.
The first is drawing blood from your finger, and the second is via a continuous glucose monitor (CGM).
A CGM is a “minimally invasive” sensor that’s stuck to the arm, with a flexible needle that sits under the skin. It lasts for two weeks before it needs to be replaced, and will send data to your smartphone.
The two key CGM manufacturers are Abbott and Dexcom – but plenty of third-party services will pair with those CGMs to analyze data. I used Nutrisense, a nutrition service designed to offer insights into your glucose data.
> January CEO talks AI and glucose tracking
The set-up process was pretty easy, if a little daunting. The CGM itself comes with a large applicator (above) that does look a bit scary. The CGM is pre-loaded inside, so you choose the location on your arm and press down to have it stamped onto the skin, and that flexible ‘needle’ sit under the skin.
I was assured it wouldn’t hurt. And while it *barely* hurt, it still wasn’t pleasant. I also managed to place it too low on the arm, which we’re sure is a mistake you’ll only make once. It meant it was permanently visible and easy to knock against things and next time I’d position it further up the tricep.
Once it’s on the skin, and we were set up in the Nutrisense app, it was ridiculously easy to pair the two. Rather than some laborious Bluetooth pairing, you simply tap your phone to the sensor using NFC, to load unsynced blood glucose data into the Nutrisense app.
It’s all about tapping your phone to your CGM throughout the day and examining trends in your glucose.
I was surprised by the short memory of the CGM. You need to scan at least once every 8 hours, which means that you’ll need to pair as soon as you wake up. I did start to get gaps in my data in the second week, once the novelty has worn off and I started forgetting to tap.
In the app, you’ll see a graph of your glucose level. There’s a health range band, and you’ll see any excursions outside of these ranges.
The data only really makes sense if you provide content of what you eat, so services such as Nutrisense require you to log your food. It will then rate your body’s response, and help you learn if you react to specific foods or food types. If you're really on it with food tracking, then it will also track macros.
Food logging was the bit I was skeptical about having tried it before, and it turned out to be the most laborious task.
Remembering to log meals was tough. And while it can be done retrospectively, you do need to be fairly accurate about the timings. Simple things like bananas were easy enough, but more complex meals perhaps with a side were annoying to input.
You can scan a barcode, but in the UK, most weren't recognized.
Meal logging has huge benefits in terms of accountability and forging good habits, but any system that relies on the data needs to make this as easy as possible. We didn’t find the Nutrisense system that user-friendly.
It poses a question about how consumer wearables will provide context about what causes swings and changes in our bodies in the future. If companies ask people to log food in this way, uptake will be low. This will be a key challenge to overcome.
What we learned from two weeks of tracking
I did learn a fair bit about myself during the two weeks – but far more about how glucose tracking could and should look in the future, once it lands on consumer devices.
The first thing I learned was that I wasn’t diabetic or pre-diabetic, which is great news. Some 33% of adults in the US are, and many don’t know. This is a huge benefit that widespread glucose tracking will offer.
I spent 99% of the time in the optimal range, with a couple of excursions beyond the upper range.
I also scored in the healthy range for overnight glucose levels, variability, and average glucose levels. I had some peaks out of an optimal range, but they were linked to exercise.
So perhaps the biggest thing I learned is that I don’t need to wear a CGM at all.
But it was the types of foods and activities that spikes glucose levels that surprised me.
Bananas are a big part of the morning routine here, but eating a banana and walking the dog caused huge swings in glucose. And my poorest reaction to a meal was attributed to eating ramen, which caused my largest excursion out of range. But given that 99% of my time was spent happily in range, there’s little motivation here to make any significant changes.
The Nutrisense personal nutritionist offered some sage advice over pairing carbs such as bananas with protein at breakfast to reduce the spikes.
But there were benefits away from the raw data.
Immediately I found myself to be accountable by wearing a CGM. And that changed my behavior. Having to log when I’d eaten sugary food certainly drove me to make conscious decisions about what to eat – such as not grabbing that croissant or muffin with my coffee in the morning.
The future of wearable glucose tracking
As a non-diabetic, using a CGM for two weeks didn’t yield game-changing levels of insights.
But as I mentioned, I did learn about what wearables need to do, once glucose tracking does arrive. After all, we've seen many wearables metrics that haven't yielded useful outcomes – particularly stress tracking, in my opinion.
My time with a CGM has crystalized how non-invasive glucose monitoring can deliver meaningful outcomes to pre-diabetics, and healthy people, too.
The technology will make millions of pre-diabetics aware of their condition, and help people make changes that could have huge implications for their health. It could be one of the biggest wearable innovations to date – bigger than the impact of the optical heart rate sensor in 2014.
And even for healthy people, it has the power to create better eating habits – to create accountability and promote better nutritional health.
The notion of 10,000 steps a day has changed societal mindsets around activity. Glucose trends – even if they’re not presented as scientifically or accurately as a CGM – have the power to do the same for nutrition.
These insights should be light touch, and we're going to need strong and robust AI systems if we're to make food tracking less of a chore.
That should be the goal for consumer wearables makers as they embark on this next chapter.
How we test