Levels & Stelo: What a CGM Actually Does for Wellness

Generic nutrition advice feels abstract and disconnected because it lacks a feedback loop. We eat something, and hours or days later, we might feel energetic or sluggish, but it's hard to draw a straight line from cause to effect. The noise of daily life obscures the signal. This is the core workflow problem that a continuous glucose monitor (CGM) begins to solve: it provides a near-real-time data stream, closing the gap between action and metabolic reaction.

What It Actually Does

A continuous glucose monitor is a small sensor, typically worn on the back of the arm, that measures the glucose in your interstitial fluid. Services like Levels and the upcoming Stelo from Dexcom build a user-friendly software layer on top of this hardware, making CGMs available without a prescription and helping you interpret the data. It's best understood as a metabolic literacy tool.

• It reveals your personal glucose response to specific foods, moving you beyond generic carb-counting to precise, N-of-1 understanding.
• It quantifies the impact of non-food variables—like sleep, stress, and exercise—on your metabolic health.
• It allows you to see the effect of stacking habits, like the glucose-blunting effect of a post-meal walk.
• It provides a tangible data point that can be correlated with other wellness markers like sleep scores or heart rate variability (HRV).

How I Use It for Personal Wellness

I treat a CGM not as a permanent wearable but as a two-week diagnostic sprint. For 14 days, I wear the sensor and live my life, but with one addition: a simple digital ledger. I use a plain text file, but you could use any note-taking app. I log my meals, my workouts, my sleep quality, and my subjective stress levels. This becomes the 'Ledger' layer in my personal health stack.

At the end of the first week, I export the CGM data and combine it with my journal. Then, I turn to an AI model. I feed it the combined data with a prompt designed to synthesize the information, asking it to identify the top three food culprits for glucose spikes and to correlate poor sleep nights with next-day glucose variability. The goal isn't a perfect score, but a handful of actionable insights. For me, the surprise wasn't that oatmeal spiked my glucose, but the *degree* to which a single night of poor sleep dysregulated my blood sugar for the entire following day. This insight is far more powerful than a generic 'get more sleep' recommendation.

How Practitioners Use It

For health coaches and functional medicine practitioners, a CGM is a powerful tool for client education and adherence. Instead of providing a generic meal plan, a coach can have a client run a 14-day CGM experiment. The resulting data provides an objective foundation for collaborative conversation.

The practitioner can use the data to create highly personalized protocols. They can say, 'Your data shows that quinoa works well for you, but brown rice seems to create a significant spike. Let's build your carb sources around that.' For the client, this feels tailored and specific, increasing their motivation to adhere to the plan. The practitioner can also use an LLM to generate client-friendly summaries of the data, translating the charts and numbers into encouraging, actionable advice that reinforces the 'why' behind their recommendations.

Where It Falls Short

A CGM is a powerful tool, but its limitations are significant. Honesty here is critical for its responsible use in a private wellness practice.

• It is not a medical device for diagnosing or managing diabetes. This must be done under the supervision of a qualified clinician.
• The data can be noisy. 'Compression lows'—artificially low readings caused by lying on the sensor during sleep—are common and can cause needless alarm.
• It can create or exacerbate food anxiety and obsessive tracking. For some personality types, it can be a gateway to orthorexia, and a practitioner must screen for this.
• Glucose is only one marker of metabolic health. A CGM tells you nothing about insulin levels (the more important upstream signal), inflammation, or nutrient deficiencies.
• You are sending sensitive health data to a corporation. Always review the privacy policy and understand how your data is stored and used.

“The data is not a judgment. It is simply information, a starting point for a more intelligent conversation with your own body.”

— Wellness & AI

The Point: A Temporary Teacher

A CGM earns its place in your AI health stack not as a permanent fixture, but as a temporary teacher. The ultimate goal is not to stare at a graph for the rest of your life, but to internalize the lessons it provides. After a few well-structured experiments, you no longer need the device to know how a meal will make you feel. You've gained the literacy. You've learned the language of your metabolism, and now you can navigate your health with more wisdom and agency than before.