Part 2: Wearing a Continuous Glucose Monitor as a non-diabetic Physician: The Experience

2018.09.12 Continuous Glucose Monitoring, Washington, DC USA 05711
2018.09.12 Continuous Glucose Monitoring, Washington, DC USA 05711 (View on

In Part 1, I described why a physician without diabetes (me) would wear a continuous glucose monitor (CGM): Why I decided to wear a continuous glucose monitor as a non-diabetic physician

In this post I’m going to write about the experience. The next one, about the data.


  • It’s possible/highly likely that some of my thoughts and experiences may come across as naive or less informed because I don’t live with diabetes; if that’s the case, please accept the discomfort and educate me. I want to learn.
  • I am a non-obese human who is fat-metabolism adapted, eating a real-food, low carbohydrate, diet, absent any industrialized bean and seed oils. My assumption based on prior lab testing is that I am not diabetic or pre-diabetic, however CGM can show things that a HbA1c can’t…
  • I’m not going to mention the manufacturer of the device used, consistent with my independence, freedom from conflict, and no endorsement of products or services statements, which I’ll repeat again below
2018.05.29 Low Carbohydrate Meetup Washington, DC Presentation  453
2018.05.29 Low Carbohydrate Meetup Washington, DC Presentation 453 (View on

Acquiring the device

In the United States, a physician order is required to purchase the device, even if using personal funds (which I did – no health care premium dollars were used to subsidize this experiment). In Canada, this is not the case, as physician colleagues there have been able to purchase the device over the counter. The manufacturer will assist by providing the wording of the order that the physician needs to write to allow access at a retail pharmacy.

The device is sold as a reader and a sensor. The sensor is almost as expensive as the reader and does much of the work.

Installation and Operation

2018.09.11 Continuous Glucose Monitoring, Washington, DC USA 1285
It stays put, and it doesn’t hurt. 2018.09.11 Continuous Glucose Monitoring, Washington, DC USA 1285 (View on

The system is ingeniously put together. I assume it has to be for something as sensitive as diabetes management. The sensor comes with an applicator that pushes a close to 1″ needle into your upper arm, in my case my triceps muscle. The sensor will stay there 24/7 for 10 days. It is safe to bathe, swim, or workout with the sensor in place. I did all of these things, sometimes quite intensively, and the sensor worked for the entire time.

I didn’t mind being asked what it was wherever I went. It was a good test to see how much the population knows about this technology (not very much at all) or the experience of having diabetes. Currently 88% of the US population does not have diabetes, and I’d gather that where I live, the percentage of people with no experience with diabetes is even higher. In a future iteration, it might be nice to have a flesh colored sensor that doesn’t stand out as much.

2018.09.11 Continuous Glucose Monitoring, Washington, DC USA 1194
2018.09.11 Continuous Glucose Monitoring, Washington, DC USA 1194 (View on

The reader is lightweight and also allows for fingerstick readings. I am not very good at fingerstick readings and given the lower-stakes situation with the data, I relied on the calibration of the meter itself. This brand of meter arrives calibrated; it will be 12 hours before you get your first reading, and then readings will occur every 15 minutes as well as when you scan.

The sensor will store 8 hours of readings, which means you need to carry the reader with you most of the time, or you will lose data. This experience, of being tethered to a device like this, was a good experience for me to have. Why? Because all I needed to do was wand the sensor every 8 hours. This is a fraction of the work that a person living with diabetes must do to maintain their health and prevent life threatening complications.

When a reading is scanned wirelessly, there’s an option to include information about food (carbohydrate counts), exercise, or other custom notes. The device keeps track of pre-meal and post-prandial readings, which is nice.

2018.09.11 Continuous Glucose Monitoring, Washington, DC USA 526
2018.09.11 Continuous Glucose Monitoring, Washington, DC USA 526 (View on

The Interface

The reader has a fairly simple interface and will do some basic trending.

There are two official software options to analyze the data: A desktop software program, and web-based one. Both did a good job of importing data from the reader. A USB cable is required (and supplied) to do this – the reader does not wirelessly connect. For Macintosh users with USB-C ports, the reader was able to connect with a simple converter.

CGM_Desktop_Software_Main_Screen_annotated-531-2 (View on

The default carbohydrate load is 15 gm rather than 0 gm

2018.09.11 Continuous Glucose Monitoring, Washington, DC USA 1279
2018.09.11 Continuous Glucose Monitoring, Washington, DC USA 1279 (View on

I found it curious that when entering food information, the meter “assumes” a 15 gram carbohydrate load. The user must manually reduce this number 1 at a time until they get to the desired number. In a future iteration, the carbohydrate load would be assume to be 0 and the user would count up. I assume this is is the legacy of the high-carbohydrate regime of various national bodies created in the 20th Century…

The reader assumes a target average glucose level of 154 mg/dl

CGM_Desktop_Software_Main_Screen_annotated-531 (View on

Based on a HbA1c target of 7.0%, this is the average glucose that will exist in your bloodstream, according to this software. Tables based on the work of Richard Bernstein, MD, give an average value of 172 mg/dl.

This is where experts disagree:

Hb-A1c = 7.0% is “fine” according to the American Diabetes Association, but it is “Horrible” on my table… greater than HbA1c = 7.5% is either “Deathly” or it is “Suicidal”…Source

In any event, the design is guided by the current paradigm. The device will not accept a median goal of HbA1c below 6.0% and will go as high as 8.5%, which in the Bernstein table is a dizzying 225 mg/dl median glucose level.

CGM_Median_Goal-532 (View on
CGM_Settings_annotated-534 (View on

Afer a few days, the software will calculate an approximate HbA1c, which is nice (I’ll cover that when I go over my data)

The interesting thing is, there’s really no reference on what a glucose level should be for a non-diabetic person on a 24/7 basis. There’s this guidance from Bernstein:

“Keep Your Glucose Levels Tightly Controlled Between 70 and 100 mg/dl all day long, everyday… or else you will pay the consequences.”Source

Here’s what the meter comes preprogrammed with:

CGM_Target_Glucose_range-535 (View on

Again, very different advice than what is given by the legacy organizations.

New Insights, Much Wonderment

To be able to see how your body regulates glucose is really a thing of wonder that we haven’t been able to do until now. As I’ll share in my next post, I was in awe of what the body can do, and do it well. At the same time, I am also in awe of the impact of the body failing to do what it needs to do well, and how serious a situation that can be, especially if worsened by the nutrition/advice given by the health profession(s).

2018.09.11 Continuous Glucose Monitoring, Washington, DC USA 1198
2018.09.11 Continuous Glucose Monitoring, Washington, DC USA 1198 (View on

I understand that things like ECGs are being integrated into the wearable world, but really, they don’t hold a candle to the impact of understanding one of the most important risk factors for health and life. This is especially true in a country where the majority (88%) of people with pre-diabetes don’t know they have it, and are therefore very likely to progress to fulminant diabetes.

2018.05.29 Low Carbohydrate Meetup Washington, DC Presentation  454
2018.05.29 Low Carbohydrate Meetup Washington, DC Presentation 454 (View on

The future

It’s obvious that the work of this system happens in the sensor and the reader is more like an accessory. I can see a world where the circuitry in the reader is embedded within a mobile device and that part of the system won’t be necessary. I’m sure the engineers in Cupertino will be able to tell us what will happen to the sensor… ūüôā

Should every human wear a continuous glucose monitor? I’m not sure. Should they wear it episodically to see how they’re doing? Maybe. Will these be integrated into wearable devices eventually anyway? Probably.

I may purchase another sensor around Halloween time. I decided not to test my pancreas with a large glucose load during this 10 days because I actually enjoy a real food, healthy fat diet, and prefer a nice relaxed metabolism :).

Feel free to ask questions or add comments in the comments.


Many gay men seem obsessed about their weight vs their health. You don’t need a glucose monitor if you aren’t diabetic. BTW Didn’t Dr Atikins die over-weight and from heart diseas?


Thanks for stopping by and for the dialogue. I’ll take these one by one:

  1. Many gay men also happen to be physicians (or I should say, many physicians also happen to be gay) interested in the health of the people they serve. Your statement doesn’t have evidence to back it up and is a little judgement and bias-tinged, don’t you think?
  2. Did you read my “Why” post prior to this one? Here it is . If 55% of the population of California is insulin resistant, and the overwhelming majority don’t know it, 50% of them will progress to fulminant diabetes, what we think is needed may change. Hence this work. See this study for more information about CGM insights
  3. No. He died from an intracranial hemorrhage from a fall.

Hope this adds clarity thanks for stopping by, happy to entertain other questions,


Hi Dr Eytan,
Not sure if you will see a comment in September 2023 for a blog entry from 2018.
You might update your post to include the option of using an iPhone (or others) as the reader. That makes it less expensive because you don’t have to purchase a reader.

I am also a non-diabetic, but my A1c is creeping higher, so I have used a CGM a few times as a learning tool. I have maybe used four sensors over three years.

Thanks for your blog posts – 1 – 12.


Thank you for your comment, and yes, in later posts I referred to the fact that a reader is no longer needed – a smartphone can now be the reader. The era of CGM is still exciting!


Ted Eytan, MD