Continuous glucose monitoring for the masses

Prior to 2024, continuous glucose monitors (CGMs) have required a doctor’s prescription in the United States. I’ve been using CGMs prescribed by physicians on and off since 2018. The brand I used previously was Abbott’s Freestyle Libre, which are more affordable and more suitable for people without diabetes. I had not previously used a prescribed Dexcom CGM due to cost and fit for purpose.

When the first over the counter CGMs were made available to the public, I decided to give this one, the Dexcom Stelo, a try. I was interested both in Dexcom’s technology as well as the usefulness of a CGM that’s over the counter.

The “why” of using a CGM in a person who is not diabetic or prediabetic

I do not have diabetes or prediabetes. I have been on a low carbohydrate diet. I fit the “Lean Mass Hyperresponder” phenotype and was a subject in a well-known study of these individuals whose findings are beginning to be published now.

I’ve used CGMs to check up on my diet, and more recently, to monitor for what’s called “adaptive glucose sparing,” where a person may experience a rise in fasting glucose while on a low carbohydrate diet. There’s little information on whether this is harmful or not.

Here are some key considerations for people in this group:

• They/we don’t need alarms and other safety features since we are not using insulin or other diabetic medications that might dangerously lower our glucose levels.
• Our glucose target range is going to be tighter than someone with diabetes or prediabetes because our bodies are regulating glucose normally.

Using Dexcom Stelo – What’s good

• Easy to purchase. I don’t have to ask for a prescription and then find a pharmacy to fill it. Prescriptions also expire, so I don’t have to ask for a new prescription after refills run out. The Stelo is purchased online and shipped to your door. You can sign up for a subscription, or buy the devices a la carte, with a two devices in each box. For someone in my situation, CGM use may be episodic rather than continuous, so it’s easy to put the subscription on pause or cancel it altogether.
• Cost. The Stelo is much cheaper than its prescribed cousins. This allows me to access and use the technology at a decent price.
• Durability. The Stelo requires an overpatch, while the Freestyles do not. Besides being more visible, this initially concerned me that the Stelo would have a hard time staying in place. In my personal experience, the Stelo was much more stable, and I worried a lot less about the device being scraped off. In fact, when the sensor expires, it takes some effort to remove it. This is a good thing. I have lost a few of the Freestyle CGMs in doorways and on dance floors.
• Integration with Apple Health. This is a big plus compared to the Freestyle. All of the individual readings go into Apple Health where they can be summarized like any other health metric. The Stelo also brings in exercise data from Apple Health, allowing for at-a-glance views of the impact of workouts on blood glucose. Meals are entered in manually in the Stelo app. The Freestyle app requires manual data entry for all of this information, and glucose levels never make it into Apple Health.
• Connectivity. It seems that Stelo is less reliant on a continuous connection to my phone compared to the Freestyle models. It seems to ping the phone every 5-15 minutes only. My perception is that this uses less battery power and also makes it easier to troubleshoot connection issues. In my experience, I lost connectivity infrequently, and was able to easily restore it with time.
• Data and reports. Dexcom ships an app/website, Clarity, that summarizes data over time. The app also makes raw data for download as well as more data that’s not presented in the Stelo app. Freestyle has a similar web-based system, Libreview. I would say these are comparable in functionality. I will post images of my aggregate data in a future post.
• Accuracy. This is an ongoing issue with these devices, which do not measure blood glucose directly – they measure the fluid between cells. They are regulated and tested extensively. However, people’s perceptions vary significantly. I personally experienced fairly good concordance with fingerstick blood glucose readings. I did experience changes from sensor to sensor in terms of baseline. I was having more difficulty in this area with Freestyle sensors, though, with some baseline readings 10-30 mg/dl different sensor to sensor. That’s my experience – there is plenty of online commentary about which sensor systems are perceived to be more accurate. A key issue, again, for people without diabetes, is that they will be more sensitive to perceived shifts in baseline.

Not as Good

• Data interpretation. This is less an issue for me, more for the average consumer. I see people online struggling to understand what the data coming out of this device means. Many do not know that a CGM tells you what your glucose levels is, but not what causes your glucose level to be. Specifically, it’s not clear to some online that consultation with a physician and/or further testing will be needed to understand what’s happening. I know this will be an ongoing issue for any manufacturer of these devices. Most physicians in my experience are not trained to understand the data coming out of CGMs either.
• Support. At the beginning, Dexcom only offered support via a chatbot connected to humans over chat only. There was no phone number, no one to speak to, and calls made to Dexcom about Stelo were referred back to the chatbot. It was easy to see their support team was getting overwhelmed, as the chatbot case numbers assigned seemed to grow rapidly. A device like this, in a completely new category, is going to need a lot of support, and the online forums demonstrate this. There is a non-significant risk of sensor failure, sensor loss, and other issues. In the era of prescription-only devices, the audience of people obtaining these devices is/was much more prepared to get answers on their own, or from their prescriber.

What’s missing

This data is a completely new addition to our lives, and as such has huge potential to help people live better. It also has the potential to do the opposite, through unintended consequences, such as anxiety, confusion, and connection to person’s overall health.

I have seen some helpful guides on interpreting the data coming out of CGM’s, like this one from Levels, and there are services like this that can help people use these devices successfully.

At the current time, there doesn’t seem to a national champion / leader for metabolic health, and in that, an assessment of the place for devices like these in people’s lives. A visible leader either in the public space or as part of the companies producing these devices could help people understand where they fit.

I understand Dexcom is beginning to use artificial intelligence to help consumers interpret the data, which is great. Having this data plus other health data in the same place as it is in Apple Health also affords huge opportunities. Although a CGM cannot be used to diagnose diabetes or prediabetes, the data may connect people to care sooner than later.

Just because you can, does it mean you should?

Back to the “why.”

Is wearing a CGM for someone without diabetes or prediabetes a good idea? Is it effective or cost-effective to prevent or manage these conditions? This is a huge topic with much written about.

In my experience, I would say the information is nice to have and more confirmatory of what I already expect. Because of my participation in the LHMR study, I already know a lot about my glucose level and what makes it normal, from extensive lab testing. The average consumer will not likely be receiving an extensive workup as part of using a CGM, and they probably shouldn’t either, unless it’s deemed medically necessary. However, some of them should receive a workup and they don’t know it, based on what I’ve seen in online spaces.

The unknown part is whether the time, expense, and effort to equip people with CGMs, or whatever device comes next to measure glucose continuously will have a huge health effect. A recent study put out by Virta, Inc, looked at improvement in glucose control between using a CGM or regular fingerstick blood glucose readings. Recalling the advent of pedometers and other health tracking devices, there isn’t good evidence that this data modulates health behavior on a population level. The data may in some cases do the opposite – I have seen stories online of people becoming discouraged with CGM baseline readings that are higher based on a specific sensor device due to placement or other factors, or from misinterpretation of transiently high glucose levels. Even more cynically, it’s possible food manufacturers could tailor their production to process foods that do not raise blood sugar levels but are no healthier than the foods they replace. They did this previously with the switch to low-fat and fat-free processed foods.

One day, continuous glucose measurement may be baked into a device we already use, such as a smart watch, contact lenses, or even headphones. We have to understand whether this is more data or more health. We don’t know the answer to that yet.

My conclusion, for now

The availability of these devices without a prescription was going to happen, and it probably should have happened. A person can purchase a CGM to use one time or a few times to check up on their diet/exercise, and maybe discover issues that connect them to their physician for more care. It’s not clear that this is something every person should use, and right now the manufactures of these devices are not able to support every person who might use a CGM.

In future blog posts, I’m going to look at the published literature more carefully and offer up more insights. I’ll also do some analyzing of the many thousands of glucose measurements I now have access to.

Comments and questions always welcomed.