Three weeks ago I pressed a small white disc onto the back of my arm. It was the Stelo continuous glucose monitor by Dexterity (the consumer CGM from Dexcom), and it came with a spring-loaded applicator, a tiny filament that slips painlessly under the skin, and a Bluetooth connection to my phone. I am not diabetic. I have no clinical reason to wear one. I put it on because I wanted to see what was actually happening inside my body every time I ate, exercised, slept, or sat at my desk for three hours straight.
What followed was one of the most eye-opening health experiences of my life. Within days I was making different decisions about food, movement, and timing. Not because a doctor told me to. Because the data told me to. I watched my glucose spike to 203 after a bowl of pasta. I watched it barely move after a steak. I watched "sugar-free" gum push my numbers higher than a handful of nuts ever did. The monitor does not lie, and it does not wait for your quarterly bloodwork to deliver the news.
I also discovered something unexpected along the way: Claude, Anthropic's AI, can read the glucose data directly from Apple Health on my phone. I started asking it to pull my numbers, chart my trends, and explain what I was seeing. It became my real-time metabolic coach. Between the CGM, a connected blood pressure monitor, a smart scale, and Claude interpreting all of it through Apple Health, I built a personal health dashboard that would have been science fiction five years ago.
Here are the 10 biggest lessons from my first three weeks.
1. Pasta Is a Glucose Bomb, and Now I Have the Receipt
On January 30th, I ate pasta for dinner. My glucose rocketed to 203 mg/dL. For context, a healthy non-diabetic person should stay under 140 after a meal, and ideally under 120. I hit 203. The next day I hit 186. The day after, 185. Three consecutive days of what I now call "pasta rockets."
The CGM showed me in real time what no nutrition label could communicate: white pasta converts to sugar fast, and my body takes a long time to bring it back down. Two weeks later, after changing my approach, my post-meal peaks dropped to the 128-148 range. I did not need a doctor to tell me to cut back on pasta. I saw the 203 on my phone, felt fine physically (that is the insidious part), and made a different choice the next day. Behavioral change through immediate feedback is the most powerful diet intervention I have ever experienced.
2. "Sugar-Free" Is a Lie (and Maltitol Is the Villain)
I chewed a piece of Ice Breakers gum. Sugar-free. Said so right on the package. I was sitting at my desk, had not eaten in hours, and my glucose was stable. Then it jumped to 126.
The culprit: maltitol. It is a sugar alcohol that food companies love because it is cheap and tastes like real sugar. The label says "sugar-free" and that is technically correct. But maltitol has a glycemic index of 35-50, which means it absolutely raises blood sugar. It is not in the same universe as erythritol (near zero impact) or stevia (near zero impact). Maltitol is the worst of the so-called sugar-free sweeteners, and it is in everything: sugar-free gummy bears, "diabetic-friendly" chocolate, protein bars marketed as keto, sugar-free cookies, cough drops. The "sugar-free" label is doing real damage to people who trust it. Flip the package. If maltitol is on the list, it is not sugar-free in any way that matters to your blood.
3. A 15-Minute Walk Is the Most Powerful Tool You Own
On Valentine's Day I had sushi. California rolls, which are basically vehicles for sugar-vinegared white rice. My glucose peaked at 148. Without the 15-minute walk I took immediately after dinner, that number would have been 165-170. Research shows a post-meal walk shaves 15-25 mg/dL off a spike. I have confirmed this repeatedly on my own body.
The mechanism is elegant: walking activates your leg muscles (the largest muscles in your body), and those muscles pull glucose out of your bloodstream through a pathway that does not even require insulin. It is called non-insulin-mediated glucose uptake. You do not need intensity. A relaxed conversational pace works. You do not need a gym. You need 15 minutes and a pair of shoes. On a separate occasion, I ate lasagna, did a post-meal walk, and hit 100% time in range for the entire day with an average glucose of 102. The walk turned a carb-heavy meal into a non-event.
4. Food Sequencing Changes Everything
On sushi night, I did not just eat California rolls. I ate edamame first, then chicken, then the rolls last. This is called protein pre-loading, and the data was striking. The glucose curve rose gradually rather than sharply, the peak was contained, and the descent was smooth with no second bump.
When protein and fat hit your stomach before carbohydrates, they slow gastric emptying and blunt the glucose spike. I predicted a peak of 130-140 with this strategy. The actual peak was 148, which is still dramatically better than eating rolls first would have produced. Eating the same food in a different order changes your glucose response by 20-30%. That is free. It costs nothing. It requires no supplements, no willpower, no special products. Just eat the protein before the carbs.
5. "Healthy" Foods Are Not Always What They Seem
Greek yogurt is supposed to be a health food. High protein, probiotics, the works. But many flavored Greek yogurts are loaded with added sugar, and even the ones marketed as "healthy" can spike glucose more than you would expect. The CGM revealed that foods I had filed under "always fine" were sometimes causing meaningful rises that I would never have noticed without continuous monitoring.
The same lesson applied to "super green" gummies. The label was covered in wholesome language: K2, D3, methylfolate, a greens blend. Great nutrient profile. But the delivery vehicle was cane sugar, 8 grams of it. At 8:45 PM, after my glucose was coming down nicely from dinner, popping those gummies would have sent me right back up to 125-130 and restarted the cycle. Health food marketing and actual glucose impact are often two completely different stories. The CGM is the only way to know which story is true for your body.
6. Morning Workouts Set the Metabolic Tone for the Entire Day
Strength training in the morning improves insulin sensitivity for the rest of the day. When I did a morning workout, my muscles were primed to absorb glucose more effectively by dinnertime. I ate lasagna after a morning lifting session and stayed at 100% time in range. On days without a morning workout, the same types of meals produced higher and longer spikes.
The best pattern I found was bookending the day: strength training in the morning to set the metabolic foundation, then a 15-minute walk after dinner to handle the biggest meal. This combination consistently produced my cleanest glucose days. One interesting discovery: during intense exercise, glucose actually rises temporarily because the liver dumps glycogen to fuel the effort. My post-workout smoothie peaked at 128, but this is a fundamentally different kind of spike than a dietary one. The body is using that glucose efficiently, and it returns to baseline quickly.
7. Stress Spikes Glucose Without a Single Bite of Food
I was sitting at my desk. Had not eaten in hours. Glucose was stable. Then I noticed it climbing to 126. The likely culprit: cortisol. When you are mentally stressed, your body releases cortisol, which signals the liver to dump glycogen into the bloodstream. Your body thinks you need fuel to fight or flee. You are just answering email.
This was one of the most humbling discoveries. I can eat perfectly all day and still see elevated glucose if I am wound up about a work problem. Some people also experience a secondary cortisol bump in the early-to-mid afternoon that mimics the dawn phenomenon. The CGM does not just reveal food responses. It reveals stress responses, sleep debt, and the metabolic cost of a high-pressure job. It is a window into your autonomic nervous system, not just your diet.
8. You Are Slowly Caramelizing Your Own Proteins (and You Can Slow It Down)
This is the lesson that changed my entire perspective. Glucose in your bloodstream spontaneously attaches to proteins through a process called glycation. It is the same chemical reaction that browns a steak or makes toast golden. The Maillard reaction. It is happening inside you right now, and the rate is directly proportional to your average glucose.
Your A1C test measures how much glucose has attached to hemoglobin in your red blood cells over the past 90 days. Those cells get replaced, so the damage resets. But the proteins in your eyes, kidneys, nerves, blood vessels, and collagen also accumulate glucose damage, and some of those proteins do not turn over for years or ever. Glycation stiffens arteries, clouds lenses, scars kidneys, and degrades nerves. It is one of the established hallmarks of aging. Every day you keep glucose in range, you slow this process to the normal baseline rate. Every day you do not, you accelerate it. My A1C has been around 5.5%. After seeing what the CGM revealed, I am targeting 5.0-5.2%. That difference represents measurably less protein damage accumulating in my body year over year.
9. AI Turned My Health Data Into a Conversation
Here is where the experience went from interesting to transformative. Claude can access Apple Health data on my phone. The Stelo CGM writes glucose readings to Apple Health every five minutes. My Withings blood pressure monitor writes blood pressure readings there. My scale writes weight data there. Suddenly I had an AI that could pull all of this data, chart trends, calculate daily averages, and explain what the patterns meant.
I started asking Claude questions in real time. "What is my average glucose today?" "Can I eat sushi tonight and stay in range?" "My blood pressure is elevated, could it be the alcohol from last night?" Claude pulled my data, analyzed it, and gave me context I could act on immediately. It predicted my sushi peak within 10 points. It explained why my glucose rose during exercise. It caught that my "super green" gummies were essentially candy with vitamins. This is not a futuristic concept. This is working today, on my phone, with consumer hardware and an AI chat interface. The combination of continuous glucose monitoring plus continuous AI interpretation is a fundamentally new way to understand your body.
10. The Boring Stuff Is the Anti-Aging Protocol
After three weeks, the playbook is simple. Morning strength training. Post-meal walks. Eat protein before carbs. Read labels. Manage stress. Protect sleep (one bad night can make a CGM look prediabetic). Track the data. Ask questions.
None of this is exotic. None of it requires a prescription or a supplement stack or a biohacking conference ticket. But the CGM makes it real in a way that generic health advice never can. I watched my body go from 203 peaks to 128 peaks in two weeks. I watched a 15-minute walk turn a carb-heavy dinner into a 100% time-in-range day. I watched "sugar-free" gum spike me harder than a handful of nuts. I watched my glucose rise from stress alone. I am not guessing anymore. I have the data.
The sensor falls off in another week. I will put a new one on.
