HYPER-PERSONALISED NUTRITION
Once reserved for diabetics, continuous glucose monitoring (CGM) is the bio-hacking tool of the future. Start-ups are harnessing the technology to help you fine-tune your diet, boost your energy and supercharge your performance
I’ve got a needle in my arm, and I’m craving cake. The cake in question is my mum’s, a rich, flourless chocolate cake that she has just handed me at brunch. The needle – more like a filament, really – is attached to a sensor called a continuous glucose monitor (CGM), which measures my blood-glucose level in real time.
I know two things. First, this cake is delicious. Second, it will probably raise my glucose reading sky high, which has all sorts of potential unwanted consequences, ranging from inducing me to fall asleep on the couch in an hour to, over the long term, premature cardiovascular disease. I’ve already had pancakes and a bagel this morning, but what the hell. I eat the cake.
Later, I check an app called Levels, which receives data from my CGM and puts it through a proprietary algorithm to gauge my metabolic health – one aspect of this being how efficiently my body processes carbohydrates. Every meal is assigned a score based on how much my glucose rises and how long it stays high. On the app’s 10-point scale, Mum’s brunch rates a zero. Not good.
It’s my first week of beta-testing Levels’ CGM-based metabolic-health programme and, so far, it has been… interesting. A CGM is normally available by prescription to those who have been diagnosed with either type 1 or type 2 diabetes. These patients rely on CGM readings to gauge when they need to inject insulin (the hormone that helps to bring glucose levels down) and when they need to eat to prevent hypoglycaemia (low blood sugar).
Levels and a few other start-ups have begun to make CGMs available to non-diabetics like me who want to understand how our metabolism responds to what we eat and how that affects our sleep, energy and athletic performance. The idea is that by keeping on top of our blood glucose levels, we might change our behaviour and improve our quality of life.
I tested two of the new apps: Levels, which is orientated towards diet and optimising metabolism, and Supersapiens, aimed at athletes. After at least a month of using both, I can report that brunch has been all but ruined for me. But I have a new appreciation for the importance of controlling my glucose level and how much better it makes me feel, both when exercising and in everyday life.
Judgement Days
Levels was founded by a former SpaceX engineer named Josh Clemente, who, to his surprise, hit a wall at the age of
28. Many afternoons, his energy levels would crash. (Sound familiar?) The part-time CrossFit trainer looked fit but, inside, he says, “I felt like I had some sort of fatal illness.” He bought a glucose meter and began testing his blood multiple times per day, keeping a spreadsheet of the results. He then got a prescription
CGM from a doctor friend, and he was astonished by what it revealed. “It basically told me I fit the criteria for pre-diabetes,” he says.
Current estimates suggest that one in 10 UK men over the age of 40 has diabetes. Millions more are at risk of pre-diabetes, which means they have higher-than-normal blood sugar levels – and the vast majority are unaware. More alarmingly, research from the US estimates that only 12% of adults are fully metabolically healthy, defined as having optimal triglycerides, HDL cholesterol, blood pressure and blood glucose levels and waist circumference. A similar crisis faces Britain.
I was sort of in the same boat. My last fasting glucose test, a few weeks before I tried a CGM, had put my level at a reassuring 87mg/dL. (At or below 99mg/dL is considered to be ideal.)
A few weeks later, my Levels kit arrived. I applied the coin-sized disc to my upper arm, inserting a filament painlessly into my skin. The sensor took minute-byminute readings, storing the data on my phone. Watching my glucose level rise and fall through the day told a different, more troubling, story than the static reading at my doctor’s office.
The Levels app took my CGM data, as well as my entries for meals and exercise, and computed a daily metabolic score, which it expressed as a percentage. This score was based on my average glucose level each day, the number and steepness of post-meal spikes and the amount of time I spent outside the Levels-defined optimal glucose range of 70-110mg/dL.
For most of a given week, the app scored my metabolic performance in the high fifties, which I assumed was good until I talked to Dr Casey Means, a co-founder of Levels and its chief medical officer. “Well,” she said diplomatically, “the lowest it goes is 50, so there’s that.” Oh.
The Levels app seemed rather judgemental, to be honest, but with good reason. High levels of blood glucose are correlated with chronic disease, says Dr Peter Attia, whose practice focuses on longevity. They can damage blood vessels and arteries, leading to kidney disease, erectile dysfunction, blindness and even dementia. High levels of insulin are strongly linked to obesity, and the hormone has anabolic effects that could fuel certain cancers.
The basic theory behind Levels is that by lowering your average glucose level – and especially the spikes after eating – you can reduce the amount of insulin that you need to produce and thus lower your risk of chronic disease. “Improving glucose is low-hanging fruit for improving health and well-being,” says Means.
Weaning myself off carbs proved to be difficult. Many of my go-to meals – yogurt with granola and fruit for breakfast, a sandwich for lunch, pasta or risotto for dinner – hiked up my glucose levels. And certain foods triggered extreme reactions. A “healthy” lunch of vegetarian tacos set my all-time record of 208.
On the other hand, my CGM seemed to be OK with eggs and bacon for breakfast, steak or grilled salmon at dinner and nuts instead of crisps at snack time. I began making modest changes to my diet and generally eating smaller portions. Levels prompts users to photograph their meals with each log entry, and that made me pay more attention to what I was eating. My daily score crept upward, into the sixties, seventies, even the high eighties. I’ve always loved carbs and have never been a fan of any overly restrictive diet, but my CGM was carb-shaming me into refuelling more reasonably.
“CGM becomes a tool to change behaviour,” says Dr Dominic D’Agostino, an associate professor at the University of South Florida, who experimented with diets and foods while beta-testing Levels. “I considered myself hyperaware of what food does to the body, but CGM takes that to another level.”
Fuel Your Performance
At the same time, I was using another, very different CGM-based platform, Supersapiens, looking for insights into the link between blood sugar and athletic performance. Co-founded by Phil Southerland, a former professional cyclist, Supersapiens is designed to teach elite and recreational athletes about the importance of glucose levels.
As a young cyclist with type 1 diabetes (his pancreas produces little-to-no insulin), Southerland noted that on days when his blood sugar was low, he would race poorly, while a more stable glucose level often translated into better performance. But there was no way to keep track of how his level changed during an event. When he began using a CGM about a decade ago, he realised how important tracking levels while active could be.
“My knowledge of everything that happened in my body – food, insulin, nutrition, sleep – increased exponentially in those first 20 days of using CGM,” he says. His fellow athletes on Team Novo Nordisk began using CGM to help with their training and racing, too, and their results improved. “It was a phenomenal tool,” he says.
As CGM technology got better and its use spread, Southerland started looking for ways to bring this data to non-diabetic athletes. Supersapiens, which launched in September 2020, is the result. The app uses a new sensor made by Abbott called the Libre Sense
“Improving glucose is lowhanging fruit for improving your health”
Glucose Sport Biosensor, which is geared towards athletes. The main difference is that it measures a narrower range than sensors for diabetics, between 55mg/dL and 200mg/dL.
When my sensor arrived, I implanted it in my other arm, then went for a bike ride. Every minute, the sensor sent a glucose reading to the phone mounted to my handlebar. As I headed up a long climb, I noticed something strange: my glucose number started rising when I had expected it to go down. As the climb got harder and my liver and kidneys cranked out more fuel for my muscles via gluconeogenesis, the number jumped again, up to about 140mg/dL (similar to where I land after a bowl of cereal). But two hours into the ride, the number plunged down to 70mg/dL. My energy level was tanking as well, and by the time I made it home, I was hungry enough to eat a bike tyre.
Supersapiens is meant to help you prevent such disasters. When I noticed my glucose level dropping, I should have eaten something. The point is to learn these lessons, says Southerland: how do I feel after eating muesli for breakfast versus, say, an Egg McMuffin? Should I eat energy bars on the ride, or just a banana? “It’s like becoming your own scientist,” he says. On my next ride, I nibbled on an energy bar whenever my glucose level began to dip. I rode home feeling strong.
I spent more than a month using various CGM systems. The most important rule I learned is that, when it comes to blood sugar, there are few hard and fast rules. On days when my glucose level jumped up and down like a kangaroo, I felt tired and cranky. On days when my number was more stable, so was my energy level. My sleep tended to be better, too. “What we’ve seen is that even people who think they don’t have problems with diabetes or glucose control have massive spikes and massive crashes, which have significant consequences on performance,” Southerland says.
A Learning Curve
One thing I did notice is that the various CGM sensors I was wearing almost never agreed with one another, each giving a different number. This is not a big deal for patients with diabetes (who are more concerned with high glucose levels, above 180mg/dL), but it could be critical for healthy people who are trying to avoid sliding into metabolic dysfunction. “The difference between 100mg/dL and 120mg/dL is huge for my patients,” says Attia, one of the few physicians who prescribe a CGM for non-diabetic patients as a preventative measure.
A patient whose average glucose level is 120mg/dL is likely further along the road to pre-diabetes or diabetes than one who hovers at 100mg/dL. “Diabetes and metabolic dysfunction are on a continuum. It’s not binary,” he says.
In general, though, all of my CGM readings moved in the same direction. They all went up after a bagel and down after a dog walk. More interesting was the daily variation in my glucose response. For example, I succumbed to a bacon, egg and cheese croissant. I was dreading the CGM scolding, but it was fine. My glucose didn’t react much, and Levels rated the croissant a six out of 10. But a few days later, the same croissant eaten at the same time scored one out of 10.
Which brings us to the other big lesson: how we handle glucose can change from day to day. “A lot of factors can influence your glucose levels,” says exercise physiologist Federico Fontana, who is the head of performance for
Team Novo Nordisk. “Some individuals react with more variability in the course of the day, while others are more stable. This will depend on a lot of things, such as the amount of exercise you are doing, diet, how your insulin sensitivity changes – even temperature.”
Mental stress also affects insulin response, says Means. She adds that our insulin sensitivity shifts over the course of the day. “We tend to be more insulin sensitive in the mornings, so I tend to front-load my carbs in the first part of the day.”
With her words in mind, I decided to take on one more glucose challenge: my mum’s chocolate cake, which had done me in that morning at brunch. I had a big piece. I had been riding and hiking more, I’d had a good but lowish-carb lunch, and I felt like my insulin game was on point. A few CGM experiments had taught me that dark chocolate doesn’t seem to affect my glucose level nearly as much as, say, flour tortillas.
So, I ate the cake with confidence. And it barely moved the needle.