Over the past few months, Apple has hired a formidable team of biomedical experts with deep experience in medical sensor technologies. Notably, many of the folks now working for Apple have done impressive and groundbreaking work in the realm of continuous glucose monitoring (CGM).
Many news outlets, as a result, have reported that Apple's rumored iWatch may be able to non-invasively measure a user's glucose levels. Such a device would be a godsend for diabetics who often have to monitor their blood glucose levels multiple times a day, either by drawing blood from their finger or through an implanted sensor paired with an external monitoring device.
The rumors swirling around the iWatch have grown so unwieldly, the expectations so beyond the realms of modern science, that many are already pegging the iWatch as a revolutionary medical device that will leapfrog competing devices like the Fitbit and the Nike FuelBand by offering unprecedented medical sensor technologies to the masses. Just last week, the San Francisco Chronicle published a report claiming that Apple is researching sensor technologies capable of predicting heart attacks "by studying the sound blood makes at it flows through arteries."
It's time to jump back to reality.
Here, we will specifically focus on the the idea that Apple's iWatch will be able to measure a user's glucose levels.
Recently, well-connected 9to5Mac blogger Mark Gurman reiterated that the iWatch will, in fact, be able to monitor glucose levels.
Our knowledge is reliant upon what Apple is programming the Healthbook app to be capable of and based on the company’s recent hires. Our sources today have reiterated that Healthbook is planned to be able to read glucose-related data...
Given Gurman's impressive track record for accurately breaking Apple news, many outlets have similarly made the leap from "Apple is hiring folks with expertise in continuous glucose monitoring" to "the iWatch will monitor user glucose levels."
A deeper examination of the issue, however, strongly suggests otherwise.
Non-invasive CGM is an incredibly complex problem that presents a number of challenging medical and technological hurdles. Indeed, medical device companies have been trying to solve this problem for decades, with no real success to speak of.
Apple and C8 Medisensors
Over the past few months, Apple has hired a number of scientists and engineers from C8 Medisensors, an innovative California-based company (now defunct) that was singularly focused on developing a non-invasive CGM device called the HG1-c.
To really gauge the feasibility of an iWatch that monitors glucose levels, it's helpful to take a deeper look at the HG1-c's capabilities and limitations. Indeed, doing so brings to light a number of daunting challenges that would arise in bringing the technology to market, let alone embedding it in a wristwatch.
Employing a technology called Raman spectroscopy, the HG1-c was able to measure users' glucose levels by transmitting a pulse of light through the skin, thereby causing glucose molecules to vibrate. An optic sensor then detected the light reflected off of these molecules, whereupon the device analyzed the resulting "fingerprint" and returned a glucose reading.
Issue 1: Size
The HG1-c was a wearable device intended to be worn across the abdomen. It was impressively compact for the technology it housed, but certainly not small enough to embed inside of a wristwatch. Bear in mind that it also came with a separate battery pack. Together, the device weighed in at 5 ounces, about an ounce heavier than the iPhone 4.
Issue 2: Sunlight
Size aside, the HG1-c carried a number of other limitations that would prevent it from being wrist ready.
On account of the technology it used, the HG1-c was extremely sensitive to sunlight and needed to be cloaked in as much darkness as possible to truly be effective. This, of course, is an obvious deal-breaker for any device meant to be worn on the wrist.
I was able to chat with Charles Martin, a former C8 employee who helped work on the HG1-c's firmware, who expounded on this in greater detail.
Mr. Martin explains:
Yes, the camera sensor had to be shrouded in darkness to function. You have to understand that Raman Spectroscopy is looking for a very faint signal emitted by the glucose molecules. A rough analogy: try to pick out someone's voice in a noisy room. The sunlight was this kind of noise that the camera sensor was not calibrated against. They did try to implement algorithms to discount measurements against sunlight anomalies, but some of the anomaly criteria these algorithms were supposed to detect, overlapped. This made things hard to verify and test on the device.
Also bear in mind that the camera sensor's performance was affected by variables as innocuous as a user's body hair and skin color, limitations that certainly don't bode well for a wearable device designed for the masses.
Issue 3: Physical Activity
Further, the HG1-c wasn't designed to be worn while partaking in physical activity because the sensor, in order to work accurately, had to be nestled up directly against the skin. And to help provide better optics for the sensor, users of the device were supposed to apply a layer of gel between their skin and the device.
Again, this limitation is an absolute deal breaker for a mass consumer product intended for the wrist.
Issue 4: Battery Life
Another issue that would seemingly preclude C8's technology from working in a wristwatch, at last for now, pertains to battery life. The HG1-c, which came with a separate battery pack, featured a battery life of 30 hours when taking glucose measurements every 15 minutes and a battery life of 20 hours when taking glucose measurements every 10 minutes.
With reports that Apple is already struggling to attain battery life of "at least 4-5 days between charges," it stands to reason that battery-hungry CGM functionality isn't in the cards.
Achieving the impossible in a few months
So if we're to believe that an iWatch capable of non-invasive CGM is on the horizon, we're forced to make a number of lofty assumptions about Apple's ability to miniaturize the technology, vastly improve its battery life, and address all of the other aforementioned issues that make the device ill-suited as a mass market consumer product. Also keep in mind that the HG1-c was only tested on and intended for individuals over 18 and non-pregnant women.
What's more, many of Apple's biomedical and medical hires only joined the company in the last few months. Developing a bonafide medical breakthrough device in such a compressed timeframe runs against all notions of plausibility.
C8's CGM technology was impressive, but much work remained
For the sake of discussion, let's assume that the iWatch only measured glucose levels at the directive of the user. In other words, let's envision a non-continuous and user-initiated glucose monitor.
Even in this scenario, a multitude of serious issues remain.
To say that monitoring glucose levels via non-invasive means is an extremely complex and challenging task is an understatement on the grandest scale.
A multitude of companies over the course of many decades have tried and failed to tackle this very problem even as it pertains to invasive glucose monitoring, spending untold hundreds of millions of dollars in the process.
That said, C8's own device, which, again, relied upon Raman spectroscopy, showed much promise. The device even attained CE Mark Approval in Europe in October of 2012.
Nonetheless, the technology still had a long way to go before becoming a viable product.
On this topic, I was able to chat with C8 Medisensor CTO Rudy Hofmeister, who explained to me that while the HG1-c device was a technological feat, it was nowhere ready to being where it needed to be for a consumer device.
The HG1-c was sort of like a dancing bear; that it worked as well as it did was an absolute miracle, but it was nowhere near where it needed to be in order to be a viable medical product, and certainly not a consumer medical product where the bar is even higher.
There's a big difference between performance and efficacy. The device certainly had a level of performance that was statistically significant; it was good sometimes, random sometimes, but not anywhere good enough to be used as a diagnostic or monitoring device for a disease state.
When I inquired about the device receiving CE Mark Approval, Hofmeister explained that this approval was for an early version of the device that wasn't even manufacturable.
The CE Mark Approval allows you to sell, but it doesn't really cover efficacy. So while you can meet the device's claims, those claims can be so watered down that the product has no value.
While Hofmeister left C8 about a year before operations came to a halt, he remained abreast of what was going on at the company, as the talented folks working there were part of the team he helped assemble.
When I inquired as to how fast the C8 team was making improvements to the device before the company ran out of funding, Hofmeister explained:
All avenues to increase efficacy were being pursued, but any improvements attained were incremental, and none that would have led to the level of improvement that was needed.
What about a different approach for non-invasive CGM?
Might it be possible for Apple's team to employ a strategy other than Raman spectroscopy? Of course, but as you'll read below, no other non-invasive CGM approach - and there have been many - has ever borne fruit.
Fact: Non-invasive glucose monitoring has never been cracked
Underscoring the difficulty, frustration, and complexity associated with non-invasive glucose monitoring, John L. Smith, a leading expert on non-invasive glucose measuring technologies, wrote a book on the pursuit in 2007, updating it a few years later in 2013.
Smith's seminal book covers, in exhaustive detail, the vast number of companies (including C8) that have experimented with an extensive array of technologies and strategies in what ultimately remains a difficult puzzle that has never seen a marketable solution.
And in those seven years, I’m sad to say that no technology has yet reached the marketplace, or for that matter, been reliably reported as actually succeeding in laboratory or clinical testing. I have personally looked at perhaps another two dozen technologies (but can’t discuss many of these newer ones, due to confidentiality agreements), been intrigued by a few and disappointed in most others. Some new companies have joined the quest, and many others have reached the end of their participation.
The 2013 version of the book concludes:
As in the attempts detailed here, the horizon will continue to be clouded by spurious correlation, incomplete understanding of the sources of error, lack of rigorous evaluation of results and wishful interpretation of data. Unlike the cure for cancer, where partial success has been achieved in many areas, this one still seeks a breakthrough. It is hoped that the attempts detailed here will help to prevent others from repeating past mistakes and premature announcements, but a rational assessment would suggest that many more lie ahead.
In short, measuring glucose levels via non-invasive means has never been proven to successfully work in any product to hit the market. Solving any problem that lies at the intersection of medicine and technology is ridiculously tough. Doing so with a wearable mass-consumer mobile device only adds many more layers of complexity.
To believe that Apple, with a team that was mostly assembled in mid-2013 and onwards, will soon be able to crack this nut is patently absurd.
Apple would prefer to steer clear of FDA approval and Medical Audits
What's more, securing FDA approval would entail extensive and careful clinical trials, with an approval process that could last as long as 18 months. In essence, entering the world of medical devices would require Apple to jump through hoops it traditionally takes pains to avoid.
On this topic, I was able to get in touch with a former C8 employee (who wished to remain anonymous) who explained some of the inherent regulatory hurdles associated with developing medical devices: