Invisible Metrics: 8 Eye-Opening Discoveries from Wearing a CGM
What my Continuous Glucose Monitor taught me about myself
I’m a data nerd. I love looking at data, finding trends, and figuring out what that means for whatever it is I’m working on. Call it hold over from my corporate years spent in marketing and product strategy or from all the research I had to do over the years getting my Masters and Doctorate.
While I’m not a fan of the term “biohacking,” I do love the data and insights I get from wearables (and labwork!). I don’t use a ton of wearables, I keep it fairly simple: Oura Ring, Apple Watch, and quarterly I wear a continuous glucose monitor (CGM) for 10 days.
The Oura Ring tells me all about sleep, cardiovascular health, and my cycle. It also tracks activity, but I prefer my Apple Watch for that. It also lets me note meals, so I can see if a heavier meal or late night dinner has an impact on my sleep – spoiler alert: they always do. Seeing the impact to my heart rate variability, overnight heart rate, respiratory rate, and, of course, sleep is actually the reason why I don’t really drink alcohol these days – even 1 drink after 6 p.m. is enough to screw it up for me, and it’s just not worth feeling like shit and super-run down the next day.
The immediately actionable insights I get are invaluable.
One of the biggest wearable MVPs for me, though, is a CGM. I do not have diabetes. And, I’ll say right off the bat here, that there are plenty of doctors out there that think CGMs just shouldn’t be worn by people that don’t have diabetes. CGMs were built to give individuals with diabetes the ability to track their glucose without having to do finger pricks after every meal (or more frequently) to help them better understand when and how to take insulin. CGMs track interstitial glucose not blood glucose, making them not as accurate as the finger-stick monitors … more on that later.
There’s some controversy around whether there’s any benefit for non-diabetics to use CGMs. I can certainly see how it would not be appropriate for folks who don’t understand the data it gives or who may be at risk for an eating disorder or over-fixation on the numbers. Believe me, I get it: the first time I wore one, it was a shock to the system. I won’t lie; it sent me into a small tailspin. But after having worn one for much of 2023 and some of 2024, I 1,000% see the value in it. These days I wear one for 10 days once per quarter.
Where the CGM Adventure Started
I have hypothyroidism. Because of that, I’m at a higher risk of developing diabetes in the future (YES! That is a concern if you have a hypothyroid!). It’s for my own benefit for me to understand my blood glucose more than just the lab work that’s done during my yearly check-up. My interest in wearing a CGM has existed for maaany years long before I finally got one to wear.
Until August of last year, the only way you could get a CGM was if a physician prescribed one for you. Even then, insurance doesn’t always cover the cost, and they can be prohibitively expensive.
In the other half of my life, I advise health and wellness startups, so I’m always keeping my eyes open for interesting startups. I have been tracking Levels for several years now. In 2023, I finally took the plunge and got a subscription; if you meet their study participation criteria, you can get regular shipments of CGMs directly from them. I started on a Dexcom 6 and then upgraded to the Dexcom 7 when it became available.
Levels takes the data from the CGM and displays it in a way that really helps you see the impact food (and sleep and exercise) has on your glucose. Dexcom has apps, but the Dexcom 6 app was pretty useless (I mean really awful); the Dexcom 7 app is a little better. I have worn a Dexcom 7 using JUST the Dexcom 7 app with no Levels integration, and it was sufficient. But, just sufficient.
These days, you can now get the Dexcom Stelo OTC for about $99, if you pay as you go, for 2 sensors that last 15 days each (they give a cost break if you do a subscription program). $100 per month can still be spendy; they are HSA/FSA eligible, though, for individuals who have those.
CGM Accuracy
I’m not going to speak to what research studies have found or what Dexcom says their accuracy is. I’m only going to speak to my personal experiences. Throughout 2023 and 2024, I would often use both the CGM and a finger-stick blood glucose monitor, because I wanted accurate data. And, I’ll say the Dexcom 6 was pretty terrible for accuracy. When I would do a QA comparison, my numbers were off by 10-20 mg/dL … on several occasions it was off by 40 mg/dL, which is insane. Sadly, from what I remember, the 10-20 mg/dL is what they consider to be acceptable variance.
I expected a difference, given that finger-stick monitors use blood glucose and CGMs go off of interstitial glucose, but I felt real empathy for diabetics that have to rely on CGMs for insulin … being 20 mg/dL off is damn near the difference between being pre-diabetic (100-125 mg/dL) and diabetic (>/= 126 mg/dL). I was frustrated a LOT. I would often use the blood measure to calibrate the CGM, though calibrating it didn’t make much of a difference, overall.
The Dexcom 7 is far better. Though it can still be off by 10-20, I’ve found that happens far less frequently – it’s often within 5 mg/dL. And, that’s a pretty acceptable margin of error to me.
Ideal Numbers and Ranges
For non-diabetics, ideal blood glucose ranges before a meal should be in the 70s-80s. While a glucose increase during and after meals is normal, even with optimal types of carbohydrates, it’s often best to keep that rise below 140 mg/dL. Adults with diabetes should have a pre-meal blood glucose of 80-130 mg/dL according to the CDC and their blood glucose should return to under 180 mg/dL within two hours of the start of a meal1.
Keep in mind targets may need to be individualized based on an individual’s needs and may not reflect optimal levels.
My Top 8 Findings
I learned a LOT from wearing a CGM – a lot of it was surprising to me. These are my most notable learnings, thus far.
A quick note: You’ll notice I talk pretty candidly about “junk” types of food (taquitos, GF oreos, GF cake, etc.), that’s because 1) I love junk food, 2) I’m aware of the tradeoffs and impacts when I have it, which is not often, 3) I believe there’s room in a well-balanced eating pattern for all foods, and 4) I don’t believe in food as a reward or punishment which is usually what happens with deprivation-based diets and approaches. So, yes, I’m a doctor of clinical nutrition who also eats fun foods and doesn’t believe in eating “perfectly.” Fuck that.
1. Foods considered generally “healthy” spike my glucose
One of the worst spikes I’ve ever experienced was seeing my glucose at 210 mg/dL. Was it from a piece of GF carrot cake? Maybe some cookies? Or a GF pizza? NO. It was a box of organic lentil kale soup. You read that right: Organic lentil kale soup.
The first time I had the soup it sent me up into the 170s. The second time, it caused the 210 spike. For the record, that’s higher than what I’ve experienced with GF oreos (180 mg/dL) and some leftover Thai BBQ pork with broccoli crunch salad (175 mg/dL). But, it’s not just the soup that did it. Several types of starchy vegetables, beans, and whole grains, which for most people are considered generally good, healthy foods to eat, would also cause me to have a large spike, especially if I didn’t eat them with sufficient protein. More on that next.
2. The order of my macros mattered
I could minimize glucose spikes by eating adequate protein and healthy fats with my carbs … that’s pretty basic knowledge. The bigger surprise to me was that I could have the same exact meal, for example chicken with 3 servings of mixed vegetables, and if I had the protein first (the chicken) and the veggies last, my glucose wouldn’t spike at all.
A snack with pumpkin seeds and brazil nuts, low fat cottage cheese, baby carrots and a plum would cause minimal rise and no spikes if I ate everything in that order. But if I started with the plum or baby carrots, there’s a good chance I’d see a shaper rise.
Depending on the meal or snack, I start with the protein or fat first. With the examples above, I start with the protein for the meal (chicken, veggies, quinoa) and fat for the snack (nuts and seeds). Not only does this help minimize glucose impacts, it also gives the fat and protein a head start on digesting; fat is the hardest of the three macros to digest. It’s also the slowest but most energy efficient macro (as it should be at 9 calories/g compared to protein and carbs’ 4 calories/g). I also know that if I’m super-hungry, while a fast-digesting carbohydrate might make me feel better sooner, it’ll be gone again in a blink of an eye. I have the fat or protein first because it’ll last longer and be more satiating, which means I won’t be reaching for a snack, again, as fast.
3. Serving size matters
Even the most glycemic-friendly meal can cause a larger-than-ideal rise if it’s more than my body needs. I’ve been known to have 3-5 cups of vegetables in one sitting after my protein. And, there’s a lot of days when that works for me. But, there are days it definitely doesn’t, and if I push past my body’s fullness cues, I’ll still see a rise. Why? Because too much is still too much, even of colorful and green vegetables.
4. Not all glucose-friendly foods are healthful foods
One of my biggest surprise learnings was that taquitos from a local fast-casual Mexican joint doesn’t really impact my glucose. Every time I’ve had them, my glucose has a pretty standard rise and fall staying below 140 mg/dL. However, they are far from a healthful meal.
But this might beg the question: What’s better, then? A less-than-healthful meal that doesn’t really impact glucose or a healthful meal with things like lentils and whole grains that spike my glucose? The answer is: Neither and both. Having fun foods, like taquitos, and having meals that spike glucose every once in a while is fine. Making fun foods the majority of my meals as a strategy to keep my glucose stable or having meals that spike my glucose (regardless of composition) every day are both harmful and problematic in different ways.
I’ll just remind you here, that plenty of people do not get a glucose spike from lentils or quinoa and plenty DO get them from taquitos and the like. I’m just special that way. What works for me, works for ME, and may not work for you.
5. Stress, exercise and saunas can impact glucose as much and more than food
A 30 to 45-minute HIIT workout followed by a steam has definitely caused a glucose spike for me. Stress whether physical stress, like from exercise and/or a sauna, and mental/emotional stress are well-documented for their impacts on glucose, both in positive and negative ways. In fact, in times of stress – we tend to notice this more during mental/emotional stress – the body becomes deficient in dopamine. Which is one reason we reach for comforting (often refined carbohydrate-rich) foods and/or doom scroll.
This isn’t a how-to article, but when that lack of dopamine hits, instead of finding it in comfort food or social media, have something with tyrosine in it: Tyrosine is the building block of dopamine.
Knowing stress can impact glucose and actually SEEING it via my CGM … those were two different realities for me. Stress from exercise is good stress. It’s hormesis in action. Same with a sauna/steam tent. You can, of course, go overboard with either. Minimizing a spike that occurs from exercise is not something I’m particularly concerned about when I don’t feel generally stressed in other ways.
In fact, the vast majority of the time, one of the best, most effective ways to lower your glucose if it starts to rise too much after a meal is to go for a brisk walk. I’ve watched my post-meal glucose drop just by getting on my treadmill for 15 minutes.
6. It’s not just about the meal; it’s also about life
Stress and life in general can really throw things for a loop. You can have the same meal and it can hit your body differently depending on what else is going on in your body, in your mind, and/or around you. Remember that organic lentil kale soup? It caused a spike both times, but at drastically different levels (170 and 210 mg/dL). Same with the taquitos. They didn’t cause the same exact response every single time.
There’s actually a new study out that backs that up. The same exact meal can impact individuals differently (meaning, how lentils affect me is different than how lentils may affect you) and it can impact the same individual in very different ways depending on the day, stressors, and for individuals with a menstrual cycle, where you are in your cycle (the lentils may have gone up to 210 because of mental/emotional stress or cycle-related hormone changes or some other reason).
7. A junk food-induced glucose spike rarely stays confined only to that meal
Like I said, I love me some junk food. But, I don’t consume it that often because the fall out, whether I physically feel it or not, lasts far longer than its worth. I’ve seen my body take days to really come back to normal glucose levels, especially if I’m experiencing stress, not exercising, and/or not getting adequate sleep.
Interestingly, there’s also a new study that shows that it can take your brain two weeks to reach its normal brain insulin levels after a 5-day junk food fest (the study actually looked at an added 1,500 calories from ultra-processed foods over 5 days). I’ll have a whole article on this soon, so no spoiler alerts.
8. Some products actually can help reduce glucose impacts
While doing some research on a startup for the Advisory side of my work life, I came across a company that makes a product called Monch Monch. It claims to act like a fiber sponge that soaks up glucose before it can hit the blood stream. Now, ordinarily, I dismiss these types of things as hocum – pure marketing fluff. Something I would cite as Exhibit A for why to hire me and business partner, Sarah, so you don’t make claims you can’t support when positioning and marketing your products.
BUT, Dr. Robert Lustig, who is a highly respected physician and a strong anti-sugar voice, is both the Co-Founder and Chief Medical Officer. And, they have a few clinical trials under their belt, which I appreciate (I would appreciate more if they linked out to all of them instead of forcing me to hunt for them). Those two factors were enough to pique my curiosity and have me do my own at-home tests.
I did two different challenges:
Eating the same cauliflower crust pizza on two days apart around similar times with similar activity patterns in my days. The first time with no Monch Monch; the second time with Monch Monch.
Having Monch Monch with a meal that had some brown rice in it but waiting to have the Monch Monch until my glucose started to rise (to see what effect it would have and how long it would take).
In both tests, my glucose dropped by about 15 mg/dL. Now, I’m an n of 1. And this was far from a lab-quality research experiment. But still. While it doesn’t make it like you never ate the carbs – which their marketing kind of claims – it does seem to effectively help minimize the impact. That said, if you have some cake and you spike above 200, even a 15 mg/dL reduction isn’t going to be enough to be meaningful. SO, take it with a grain of salt (not literally … please don’t add salt to the Monch Monch, if you try it).
Also, Monch Monch: If you want some help not over-promising but still being compelling with your marketing, we can help you with that.
For the record, I have not received any incentive of any kind from Monch Monch to review or try their product.
That’s where I’ll leave it for now. There’s still so much more I could talk about! One of these days I’ll get to all of the things I love about my Oura ring and how I use the data from it.
References
1 Manage Blood Sugar. CDC.