How to Use a CGM to Control Glucose Spikes

TL;DR

Continuous glucose monitoring can help control eating behavior by revealing glucose responses in real time and making unwanted spikes immediately visible. Peter uses average glucose and glucose standard deviation to assess both overall levels and variability, treating low glucose with low variability as a practical proxy for low insulin when continuous insulin measurement is unavailable.

Transcript

Welcome to The Qualies, a subscriber-exclusive podcast. Qualies is just a shorthand slang for a qualification round, which is something you do prior to the race, just a little bit quicker. Qualies podcast features episodes that are short, eh, we're hoping for less than 10 minutes each, which highlight the best questions, topics, tactics, et cetera,... Read More

Key Insights

• A continuous glucose monitor is a powerful behavioral control tool for Peter because real-time feedback makes the consequences of food choices visible. His desire to avoid seeing glucose spikes can be strong enough to discourage him from eating tempting foods, including a large cookie offered during a flight.
• Continuous glucose data helps connect activity, nutrient deprivation, and treats to measurable glucose outcomes. Peter uses these relationships to calibrate when and how he eats certain foods, aiming to reduce their impact rather than claiming that he always follows an inflexible diet.
• Human behavior responds strongly to feedback, and Peter compares a CGM to the RPM tachometer in a race car. A driver may function without it, but immediate information allows more precise decisions about shifting, just as glucose feedback can guide food and activity choices.
• The Dexcom G6 is described as an FDA-approved medical device with accuracy of approximately plus or minus two or three percent. Its precision and real-time output support insulin dosing for people with diabetes, which also creates regulatory complications for making the same device broadly available as a consumer product.
• Continuous insulin monitoring is difficult because insulin cannot readily be measured through a rapid antibody or enzymatic reaction without washing steps. Peter says insulin assays have required processes such as radioimmunoassay or ELISA, preventing an immediate point-of-care measurement comparable to continuous glucose sensing.
• Insulin changes might eventually be estimated from glucose data, but Peter believes the task requires more information than CGM readings alone. A useful model would need extensive blood draws paired with glucose measurements to build a regression relationship capable of predicting later insulin levels.
• Low average glucose combined with low glucose variability is a practical proxy for low insulin when insulin cannot be measured continuously. A CGM can generate reports over seven, 14, 30, or 90 days that include both average glucose and glucose standard deviation.
• Hemoglobin A1C does not reveal glucose variability and can misestimate average glucose when red blood cell lifespan differs from its assumed 90–120 days. Peter therefore considers A1C directionally tolerable but places greater value on CGM measurements of actual average glucose and standard deviation.

Questions & Answers

Q: What does Peter want to achieve with a continuous glucose monitor?

Peter wants to use real-time glucose information to control his behavior, avoid unwanted glucose spikes, and understand how food, fasting, and exercise interact. He also monitors average glucose and glucose standard deviation. Together, low glucose and low variability serve as his practical proxy for maintaining low insulin when direct continuous insulin measurement is unavailable.

Q: How does a CGM help control eating behavior?

A CGM makes the glucose consequences of eating visible soon after they occur, creating immediate accountability. Peter says he resisted a very large cookie on a plane largely because he knew he would later see the result in his CGM data. His internal competitiveness makes glucose spikes unpleasant to observe, strengthening his motivation to avoid certain choices.

Q: How can fasting and exercise affect the glucose response to treats?

Peter uses CGM feedback to calibrate the relationship among activity levels, nutrient deprivation, and occasional treats. He describes eating fries at Fenway Park after fasting all day and working out, then observing no glucose spike. He does not present this as a universal formula, but as an example of using personal data to understand and minimize disruption.

Q: Why does Peter compare a CGM to a race car tachometer?

Peter compares glucose feedback to an RPM tachometer because both provide information needed for more precise decisions. A driver could still operate a race car without hearing the engine or seeing the tachometer, but would not drive as effectively. Likewise, real-time glucose readings help a person adjust food and activity with more precision than guessing without feedback.

Q: What CGM measurements does Peter monitor most closely?

Peter focuses on average glucose and glucose standard deviation, which represents variability. His CGM can produce reports covering seven, 14, 30, or 90 days. He emphasizes that an average glucose value alone is incomplete because two people can have the same average while having standard deviations of 10 and 30, indicating very different glucose and insulin profiles.

Q: Why is low glucose variability important for estimating insulin?

Low glucose variability adds context that average glucose cannot provide by itself. Peter explains that the same average glucose, such as 85 or 95, can occur with a standard deviation of 10 or 30. He considers these patterns representative of very different insulin profiles, so low average glucose combined with low variability is a better proxy for low insulin.

Q: Why is continuous insulin monitoring difficult to build?

Continuous insulin monitoring is difficult because insulin assays cannot currently produce an immediate result through the kind of rapid antibody or enzymatic reaction needed for a point-of-care device. Peter says insulin was initially measured with radioimmunoassay and is commonly measured using ELISA processes involving repeated application and washing steps, which cannot be completed in a moment.

Q: Why does Peter consider hemoglobin A1C less useful than CGM data?

Peter considers A1C directionally tolerable but limited because it estimates average glucose rather than measuring it directly, and it provides no information about glucose variability. Its interpretation also relies on the assumption that a red blood cell lives for 90–120 days. Departures from that range can cause A1C to overestimate or underestimate average glucose.

Summary & Key Takeaways

• Peter values the Dexcom G6 primarily as a behavioral feedback tool. Seeing glucose in real time discourages him from eating foods likely to produce unwanted spikes. His internal competitiveness makes the data especially motivating, although he acknowledges that he still sometimes departs from his preferred eating behavior.

• Continuous glucose data helps Peter coordinate food, fasting, and exercise. He compares the feedback to a race car tachometer, which helps a driver choose when to shift. By observing his responses, he can better calibrate activity levels, nutrient deprivation, and occasional treats while trying to minimize glucose disruption.

• Peter monitors average glucose and glucose standard deviation because identical averages can conceal very different variability. He considers low glucose and low variability useful proxies for low insulin. He views hemoglobin A1C as directionally informative but limited because its interpretation depends heavily on assumptions about red blood cell lifespan.