Unlocking Preventative Healthcare with Continuous Metabolic Measurement

Person checking a smart watch
June 14th, 2020

Medical data is currently in the dark ages.

You probably get a checkup once per year.

You probably occasionally get prescribed blood tests by your doctor.

You may even have a wearable that can tell you what your heart rate is and how many steps you took.

But you don't know at any given moment what your blood sugar is, what your hormone levels are, or anything else about your body.

As a result, you and your doctor are flying blind when it comes to decision making.

Your limitation to taking annual snapshots of the body in the form of blood tests is akin to a pilot seeing the radar once every few minutes.

But what if that were different?

What if instead of having extremely infrequent information about the body, we could have much more regular information about the body?

If we had a continuous stream of bodily measurements, we would be able to unlock an entirely new level of healthcare.

The Power of Continuous Measurement

First, continuous measurement would mean that we could catch signs of disease way earlier than before. Instead of waiting until your annual checkup, you would be able to have an alert that an anomaly was detected and you could be instructed to go to your physician for further screening and treatment.

Second, continuous measurement would mean that predictions and diagnoses could be made more effectively. With more data, machine learning algorithms could be trained to make diagnoses better than doctors themselves. This may sound out of the realm of what is reasonable but we're already seeing machine learning systems surpass doctors in their ability to diagnose cancer from x-ray images. This isn't to say that doctors would be replaced. Instead, in an ideal system, a machine would deliver a diagnosis and a doctor would inspect the diagnosis, consider the machine's reasoning, and take action based on that information.

Third, continuous measurement would mean that daily interventions could be measured for their efficacy in moving biomarkers. It would mean that you could know exactly what makes a difference for your body and exactly what doesn't. For example, when you have real-time glucose monitoring, you can see precisely which foods spike your blood sugar and which do not. Some people have blood sugar spikes from cookies but not bananas, while others have blood sugar spikes from bananas but not cookies. There are some continuous glucose monitors on the market today, but imagine this for all biomarkers. With such technology, you'd be able to see how today's exercise, smoothie, and stressful meeting each impacted your body in concrete ways.

Continuous Measurement Today

Let's take a look at what exists in continuous measurement technology today.

In the consumer market, smart watches and fitness bands allow everyday individuals to monitor their heart rate, monitor their sleep, and take other simple measurements. Unfortunately, these devices are limited to a few metrics that have limited predictive power for individual health.

In the medical market, meanwhile, continuous glucose monitors allow diabetic individuals to monitor their blood sugar in real time. Unfortunately, these monitors are limited to just glucose, and they are expensive, invasive, and difficult for regular individuals to acquire.

There are plenty of other medical devices out there that do continuous measurement, but they are typically constrained to a single metabolic measurement and are very invasive.

Clearly some new technology and new products need to be developed to build upon our existing capabilities.

The Future of Continuous Measurement

There are several ways we could perform continuous measurement of the body in the future.

One way is to have patients wear one or more intravenous monitors on their arms to monitor a variety of metabolites. This would be taking the continuous glucose monitor to the extreme. The challenge here of course is getting the device to measure more than just glucose. It's not impossible, but it would be difficult to get the device to be compact and economical and as non-invasive as possible. Levels Health is a company that is looking to bring continuous glucose monitoring to the masses, and it is yet to be seen whether they have plans to develop a device that can monitor other metabolites.

A second way is to have patients take bodily measurements once a week with test kits provided to the patient. The measurements could be in the form of peeing on or spitting on a strip, or in the form of mailing in urine and/or saliva samples. The downside of this method is it requires quite a bit of patient involvement, and this could mean it is out of the question for the every day consumer. Vivoo is a company that allows patients to monitor their health on a regular basis by peeing onto test strips and scanning the strips with their phones.

Yet another way would be to retrofit toilets of patients with an attachment under the toilet seat that could measure bodily fluids with every use of the toilet. The device would fade into the background, and wouldn't require the patient to do anything at all. It could perform measurements about once a day, and it could measure anything present in urine or feces. The downside of this method, of course, is the challenge in manufacturing the device in a reliable and cost effective way, but if it could be done, it would be even more powerful than the previously mentioned methods. A group of Stanford researchers built a prototype of such a device.

The last category of measurement, and perhaps the most ideal, would be to develop a way to extend the capabilities of standard smart watches and fitness bands and allow them to make measurements of all kinds of metabolites. This could be superior because it would be non-invasive, it would be inexpensive, and it would be accessible to everyday people. At the same time, the early models would likely have lower accuracy, akin to the sleep monitors of today. Such a device is currently out of the range of our existing technology, but with sufficient R&D and advancements in engineering it should be feasible. GlucoWise is a company that is building a non-invasive glucose monitor that uses infrared measurements to estimate blood glucose without penetrating the skin. And several groups of researchers have published papers on exactly this topic.


In summary, there are three main benefits of continuous body measurement:

  1. Catching signs of disease early
  2. Making diagnoses with machine learning
  3. Learning the daily interventions that work for you

And there are several approaches one could take:

  1. Expanded capabilities of continuous glucose monitors
  2. Weekly measurements with test strips or mail-in tests
  3. Toilet attachments that analyze bodily fluids
  4. Non-invasive wearables

It's clear that continuous body measurement, with all of its power for predictions and interventions, is the future of medicine.

The next question is how to make it all a reality.