The COR is a new consumer device that uses infrared spectroscopy to measure "how your body responds to food and fitness practices over 21 day cycles" using a prick of blood from a lancet. It usually costs $400 (I bought mine for half-off in a recent launch sale), and includes 28 blood-collection capillary tube cartridges (six months worth at weekly use). At the time of writing it is sold out.
The device has an interesting background. The founder is an ex-Apple director who worked on the Apple Watch, and it's funded by Founder's Fund (not on their portfolio page though!) and Khosla Ventures. I think these guys would probably have vetted the technology pretty well (especially post-Theranos, VCs are skeptical of devices like this). However, it does seem to have had a difficult development path on Indiegogo.
How does it work?
The COR is an infrared spectrometer with a "wavelength range of 1600-2450nm (resolution of 7-8nm)". The name COR refers to the "correlations" that can be made with these spectroscopic readings. Basically, they cannot derive readings for specific analytes (like glucose or cholesterol), but looking at the spectral data in aggregate ("ensemble spectral changes"), they can see trends that correlate with certain health outcomes ("blood response patterns").
This is interesting enough, though some of their patents (ascribed to "nueon") promise a lot more, and were part of what intrigued me about the device. For example, according to some of the data in their patents, they can measure glucose and cholesterol accurately. As far as I know, this would be a big breakthrough for infrared spectroscopy. The usual problem with replicating results like these is calibration — some technologies can achieve great results, but only if you have an accurate baseline to start off with. It reminds me a bit of the almost-non-invasive glucose-tracking GlucoWatch.
The COR website has a very informative blogpost that explains a lot more about the device and the study they are running to validate it:
Despite IR’s incredible capabilities, it is not used in today’s central blood laboratories. The reason is sensitivity. IR cannot model the trace components of blood very well, and so the labs need to use various types of fluidics, derivatization and wet chemistry to get answers. Once you have this type of equipment for the low concentration things (cancer biomarkers, hormones, etc.) you might as well use it for the high concentration biomarkers as well. Accordingly, there’s no clear place for IR in a central blood lab.
Let’s circle back to the mission stated above. We want to give people the ability to see how lifestyle changes impact their blood in real time. To achieve that, IR is a perfect technology. Instead of prospectively looking for individual biomarkers in people with essentially no insight into their lifestyle changes, we look for and analyze overarching, subtle IR pattern changes in groups of people who are all performing the same lifestyle program.
Spectroscopy
Spectroscopy is a technology that seems to be perennially almost there. The appeal of a cheap, non-invasive tool for general molecular analysis is too appealing. In particular, Raman spectroscopy has been vaunted many times over the years as a way to measure analytes of all kinds.
You can actually buy a handheld Raman spectroscope for raw material identification. It is very cool technology, but the examples show identification of bulk ingredients (e.g., a vial of lactose), which you expect to have a consistent spectroscopic pattern. Picking out specific analytes from a complex mixture is much harder.
Baffling things about the COR
The programs
The programs suggested by COR are... unusual. For example, "The Sardinia":
I get that Sardinians do the things listed more than other populations, and are regarded as having a healthy diet and lifestyle, but it is a strange mix of things.
The "Okinawa" program includes "10 minutes rajio taiso", and "one cup of reishi tea". The Pro Athlete program includes a "Polar plunge" every day(!) I think I would have to be much more dedicated to the programs to do all these things.
The time it takes
The COR takes 90 minutes to read out, continually measuring as the blood separates. By the time it's done, the capillary tube is dry and empty. It's not like I need results quickly or anything, but I am not used to a consumer device taking this long.
The lack of outputs
When you process a sample on a device, generally you expect a result? The COR seems to be very coy about results. You are supposed to get a score from 0-100, representing some measure of blood health, but I have yet to get a number. I am assuming at some point I will get something, but for now it really feels like the device is an unnecessary part of the experience.
Conclusion
The ideas behind the COR are solid. I believe ensemble patterns with spectroscopy will correlate with relevant blood markers. I believe that lifestyle interventions can change blood chemistry. I do find the product very strange though!
I find it difficult to know what I am supposed to do with a number that may or may correlate with omega 3, or glucose, or cholesterol. It could mean one is up and the other two are down, it's just difficult to know. The patents promise a lot more than the COR currently offers, so maybe this is the start of a big effort to figure out more specific results using crowd-sourced data.
I hope the COR succeeds and the product continues to improve. If not, I think it would be interesting to hack the machine to get access to raw spectroscopic data, though I suspect that will be tough.
Comment