Non-Invasive Glucose Monitoring: Hope or Hype?
For years, people with diabetes have sought out an accurate way to measure glucose without the need for fingersticks or fluid samples. Here’s a look at the latest in research and development of non-invasive glucose monitoring devices.
Over the last two decades, dozens of companies worldwide have sought to develop a non-invasive glucose monitor – a sensor that could measure your glucose levels without puncturing the skin or having to insert a device into the body. However, the majority of these devices have either failed to work or have fallen short of FDA approval due to their ability to accurately measure glucose, leaving millions with diabetes frustrated.
“There is significant demand for a non-invasive glucose monitor, as most of our patients hate sticking their fingers multiple times each day to check their blood sugars,” said Dr. Farah Khan, an endocrinologist and Clinical Assistant Professor of Medicine at the University of Washington. “There are sometimes accuracy issues with CGM technology as well, so if a non-invasive monitor could somehow be more accurate than the CGMs currently available, this would be a huge step forward.” Even if it does not exceed the accuracy level of a CGM, having a device that is close in accuracy and that eliminates the need to puncture the skin could benefit many.
These sensors could make fingersticks and inserting CGMs a thing of the past; just one of the many advantages of developing a non-invasive glucose monitor. Some other advantages include avoiding the pain of a needle, eliminating false readings due to contamination by substances on the surface of your hands, and limiting waste from disposable testing supplies such as test strips or lancets. Additionally, many non-invasive monitors have no disposable components, which could wipe out recurring costs for testing supplies, only requiring a one-time purchase.
Optical Technology: Shedding Light on Your Glucose
The discussion around monitoring glucose levels without the need for blood or other bodily fluids is not new; in fact, devices have attempted to use various forms of optical technology since the 1990s. Optical technology refers to any method that can identify different substances in a fluid (such as your blood) based on the interaction of those substances with electromagnetic radiation, or light.
Physicians and scientists have already used optical techniques in many areas outside diabetes; X-ray machines can create images of broken bones based on how our bones absorb and reflect X-rays differently than the surrounding bodily tissue. Nearly all of the non-invasive glucose monitors currently in development use some form of optical technology to collect glucose measurements.
While different devices use different types of light (some use infrared waves, others radio frequency waves, etc.), the basic principle is the same. A monitor shines a beam of light through your skin, which is then reflected and scattered by the glucose in your blood. The sensor then detects signals from the light that is reflected back. Glucose exhibits a unique reflection pattern that differs from the other components of your blood; the difference in signals is how the sensor could determine your glucose level.
Of course, this is much easier said than done. One of the key challenges to developing a non-invasive glucose monitor has been the lack of accuracy, specifically in discerning between glucose and the other components in our blood.
“The glucose signal from an optical sensor is small, and other molecules in the body produce interfering signals that are similar, which is called noise,” said Dr. David Klonoff, Medical Director of the Diabetes Research Institute in San Mateo, California and editor-in-chief of the Journal of Diabetes Science and Technology. “The ratio of glucose signal to noise is often very small, and it can be difficult to isolate the glucose signal itself.”
However, our ability to isolate glucose signals alone has improved. “Newly-developed artificial intelligence (AI) and machine learning algorithms have given us greater computational power to isolate a glucose signal from the noise,” said Klonoff. “We are getting closer to seeing a commercially viable product emerge, thanks to improved sensors that are smaller and more accurate.”
There are varying degrees of optimism as to when this technology might become accurate enough to be used in an FDA-approved glucose monitor. “The FDA has never cleared an optical non-invasive glucose monitor; however, I expect that the regulatory requirements for eventual clearance will be different than what is required for currently cleared products that measure glucose in blood or interstitial fluid,” said Klonoff.
Somewhat less optimistic, Khan said, “I worry that these forms of optical technology may not have enough precision or accuracy to help our patients manage their diabetes. Time and clinical experience will tell, but I would not expect these devices to become big players in the market until at least the next 5-10 years based on the challenges with precision and accuracy.”
Other non-invasive technology in the pipeline
Outside of non-invasive optical glucose monitors, there has been no shortage of innovation in glucose monitoring in general. Several companies have sought to develop a non-invasive fluid sampling device, using bodily fluids such as saliva, sweat, tears, or vapor, from your breath.
“This would be less invasive than even a minimally invasive technology, as they would not require inserting a needle or probe into the body,” said Klonoff. Similar issues of accuracy remain, but there has been documented progress in this area.
One of these devices developed by Nemaura Medical, called sugarBEAT, has received CE clearance for the market in Europe. Sitting on your skin like a patch, sugarBEAT is a small device that measures your glucose by running a slight, imperceptible electric current through your skin, which causes glucose to move within the interstitial fluid right below your skin. Early trial data from Nemaura Medical shows that sugarBEAT is very close to the accuracy of CGMs currently available on the market.
Who is currently developing non-invasive glucose monitors?
Although people with diabetes continue to wait for this revolutionary technology to reach the market, researchers and investors have spared no expense in both time and money. Investors have raised hundreds of millions of dollars in funding for several startups around the globe, not to mention some of the efforts by larger, more established tech companies.
This review article from the Journal of Diabetes Science and Technology identifies upwards of thirty non-invasive optical and fluid-sampling glucose sensors currently either in development or awaiting approval through the regulatory process. Across all these devices are some fascinating innovations on optical technology. We will focus on a few of the most noteworthy endeavors, some of which have already been approved and released in Europe.
DiaMonTech is a biotechnology startup based in Germany, which is currently developing three different versions of its non-invasive glucose monitor.
The D-Base, which is a desktop device (roughly the size of a shoebox) intended for use by multiple people in a clinical setting
D-Pocket, a hand-held device that fits in the user’s pocket or handbag, on which you press your finger for a glucose measurement
D-sensor, an integrated sensor that can be worn as a bracelet
All three use the same technology to gather glucose measurements, which the company coined “photothermal detection.” The sensor directs a beam of light onto the skin, which warms the glucose molecules in your skin very slightly (only around 1/1000˚C.) After a few seconds, the sensor gives a glucose reading based on the absorption and reflection pattern of the light.
While each of these versions is at a different point in the approval process, there is evidence that the technology does work and is approaching the accuracy level of CGMs.
The accuracy of glucose monitors is typically described with a metric called MARD, or mean absolute relative difference. MARD is listed as a percentage, describing how far off you can expect your sensor reading to be from your actual blood glucose level. Generally, this means that the lower the percentage, the more accurate the sensor. Based on pre-clinical tests, the D-Base has a MARD between 11.3% and 12.1%. The MARD values for the Dexcom G6 and Abbott Freestyle Libre 2 CGMs are 9.0% and 9.3%, respectively, so the D-Base is slightly less accurate than some of the popular CGMs on the market today.
None of the versions are currently FDA approved, but the D-Base did receive a CE clearance in 2019 for use in Europe and was submitted by DiaMonTech as a pre-submission for FDA approval. All three devices are still in development to further improve accuracy, so they are not currently available for purchase.
The tech giant considered including a non-invasive glucose sensor in the new Apple Watch Series 7, which unfortunately did not come to fruition. However, there is still reason for optimism. Reports came about as a result of the UK-based startup Rockley Photonics, one of Apple’s suppliers, releasing a “clinic-on-the-wrist” health watch including a non-invasive monitor. The technology is quite similar to that used by DiaMonTech and could represent a huge addition to Apple’s health monitoring features in the near future.
Cnoga is a digital healthcare startup based in Israel, which has developed a hybrid non-invasive monitor called the TensorTip Combo Glucometer, or CoG. The device is referred to as “hybrid” since users must calibrate it with both invasive and non-invasive measurements for the first three days of use. After this time, however, the device is completely non-invasive. Similar to the aforementioned devices, CoG emits a small amount of light through your finger, which you insert into the device. As the light signal is absorbed and reflected by the glucose in your blood, the signal projects onto another camera, which processes the signals using an AI algorithm to produce a glucose value.
CoG is a small, portable device that has received CE clearance in Europe. According to its website, and research on the device, the CoG obtained a MARD between 14% and 17%, so the accuracy of the device still lags behind CGMs. However, the device has been, and continues to be, tested in a clinical setting, and there are plans for the company to seek FDA approval in the near future.
GWave by Hagar
The GWave is a non-invasive glucose monitor created by Hagar, also a startup based in Israel, that uses radio frequency waves to measure blood glucose continuously. This device has currently not received CE clearance nor FDA approval but has been investigated in a pilot study earlier this year.
For more information about GWave, read our article from earlier this year, “Wave of the Future: New Glucose Technology Could Revolutionize Care.”
This Seattle-based startup has developed two non-invasive glucose sensors, both of which use a patented radio frequency wave technology called Bio-RFID. The two devices are as follows:
The KnowU, a pocket-size device that measures your glucose on-demand with a simple palm placement on the device
The UBand, a continuous monitor that is worn on your wrist
Know Labs is conducting external pre-FDA clinical studies for its Bio-RFID technology, the final step before entering the FDA approval process. While this is still early in the process, data from its early human trials are quite promising. When compared to fingerstick glucose measurements, the Bio-RFID measurements had a MARD of only 6.7%. This is well within the accuracy standards established by CGMs currently on the market.
The bottom line
Of course, this list is incomplete; these are just a few of the many startups and biotechnology companies seeking to develop a non-invasive glucose monitor. See below for some of the other efforts across the globe to improve this potentially revolutionary method of glucose monitoring and bring it to market.
Over the last 100 years, we have progressed from measuring glucose in the urine, to blood testing and fingersticks, to CGM. Much has changed in that time, from improving the convenience of taking a measurement to expanding our wealth of glucose data with the amount of measurements we can take each day. However, the need for blood or other bodily fluids to take a measurement has been constant.
So is it hope or hype? Although it may seem that regulatory barriers squash any hopes of a non-invasive device reaching the market, the demand, motivation, and money are there to keep development going. It will be interesting to watch which of these small companies gets the attention of one of the big players in diabetes technology; this could be a clue about which device is furthest along. The technology is improving rapidly, so while it may still be a couple years before we see these devices on the market, it’s only a matter of time before the puzzle is complete.
Other non-invasive glucose monitoring projects
Lifeleaf, a continuous, real time glucose monitor by LifePlus
Afon Technology, which is developing a wearable device for your wrist
GlucoTrack by Integrity Applications, a device with CE clearance that measures glucose with a clip that attaches to your earlobe
Brolis Blood Analysis Sensor
GlucoBeam by RSP systems
Gluco Quantum by Genki Vantage
K'Watch Glucose smartwatch by PKVitality