Update on Apple’s noninvasive blood glucose monitor

Bloomberg technology journalist Mark Gurman reported last month that Apple is hard at work on a real-time non-invasive blood glucose monitor. I have previously described Apple’s efforts in this area anticipating that they would use a spectroscopic approach (QH): 

“The second approach is a spectroscopic technique and in particular uses near-infrared spectroscopy (NIRS). The near IR refers to the light that illuminates the sample; near infrared wavelengths (800 nm to 2500 nm) are just above the visible spectrum (390 nm to 700 nm). The key concept is that different molecules (e.g. glucose, insulin, hemoglobin) will absorb light at different wavelengths, letting a specific pattern of wavelength intensities to be transmitted through the sample. In other words, different compounds will absorb different wavelengths of light to different extents creating a distinctive spectroscopic "fingerprint." Using this technique one can identify and quantitate a specific compound.”

The spectroscopic approach was confirmed by Gurman although further details on the measurement technology were lacking. The article states optimistically that “[a]fter hitting major milestones recently, the company now believes it could eventually bring glucose monitoring to market.” But this excitement is tempered by some background on the past history of the device. Most notably, the developmental process has been long and protracted:

“Apple’s system — more than 12 years in the making — is now considered to be at a proof-of-concept stage, said the people, who asked not to be identified because the project is confidential. The company believes the technology is viable but needs to be shrunk down to a more practical size.”

Ultimately, the goal would be to make the device small enough to fit as a sensor on the Apple Watch or perhaps embedded in a specialized Apple Watch wristband. At this time, they do not appear to be close to this objective.

Another concern is that such a device would be competing with the minimally invasive

interstitial continuous glucose monitors (CGMs) that I have described previously (QH). CGMs possess a small probe (thin electrode) that is slightly invasive, i.e. it is inserted through the skin into the interstitial region between the skin and blood vessels. Although the probe penetrates the skin, it does not access or draw any blood. The probe measures glucose levels in the interstitial fluid which is related to the blood glucose concentration. The interstitial fluid is a better proxy of blood glucose than tears or sweat (which other approaches rely on), but not as good as measuring the blood itself. 

Overall the CGMs are considered to be less accurate than a blood measurement but more accurate than completely noninvasive (e.g. spectroscopic) methods. CGMs have been around for awhile, and there are several commercial models (e.g. FreeStyle Libre) that are relatively small (coin-sized) and fitted to the upper arm to be non-instrusive. They require a prescription and cost around $75/month. The probe should be replaced approximately every 2 weeks, and regular calibration of the device is needed to maintain accuracy. Because the Apple device does not have a probe, it may be longer lasting without replacement, but it too most likely would depend on calibration with a blood test for accuracy. 

It should be noted that the Apple device also estimates blood glucose from the interstitial fluid, and thus like the CGMs would suffer from inaccuracies of this approximation. In addition, it will be a challenge to make the measurement as accurately as the electrode method which is how blood glucose is measured in the lab. My guess is that the biggest hurdle will be reducing the size of the device which is currently something that sits on a table, and presumably is much bigger than the CGMs. It might be some time before they can shrink the sensor so that it can fit on the wrist.

Apple has its work cut out, but a noninvasive real-time blood glucose monitor is a moonshot worth pursuing. It could even be cheaper to maintain than the CGMs over the lifetime of the device (no need to replace electrodes). As a standalone device that is too big for the wrist, it could still communicate with the Apple Watch via Bluetooth like the FreeStyle Libre (Figure 1).

Figure 1. An Apple Watch app connected to a FreeStyle Libre CGM (Harley Turan). One day Apple's noninvasive blood glucose monitor could communicate with the Apple Watch in a similar fashion or even be installed on the watch itself.