You are here


The smart watch: telemetry for humans?

Technology disruption will be a game changer in our fight against diabetes, obesity and other health problems

The Apple Watch Series 6 can measure the wearer's oxygen level and ECG.

The Abbott Libre measures blood sugar every 15 minutes during a two-week period.

I RECENTLY bought a new Apple watch, now in its sixth edition since it first launched in 2015. I noted with amusement its pulse oximeter function: this is where you put a device on your finger tip with a light source that measures the oxygen in your blood stream. The light source sends out red light and infra-red light and the red light is absorbed by deoxygenated blood while the oxygenated blood absorbs more infra-red light.

In the latest Apple watch it measures the oxygen level from your wrist and that is somewhat technically challenging. As a result, there could be errors as it would be better off measuring it from the fingertip.

There is also another application in the watch that measures the ECG, which started from its fourth edition. The ECG measurement is quite good, and it managed to pick up an abnormal rhythm called atrial fibrillation in three of my patients. On the other hand, it also made one of my patients very anxious and scurrying to the cardiologist as it picked up benign occasional dropped beats.

You can see where Apple is going with its watches. It is eventually going to end up as a telemetry device for humans. I think in the next iteration of the watch, it will likely measure blood sugar probably in a proxy manner.

We know Apple is serious as it is acquiring a lot of companies with the necessary technology in this space.

Your feedback is important to us

Tell us what you think. Email us at

On the matter of measuring blood glucose, there are devices that can measure your blood glucose continually without a finger prick to draw blood. The device I am familiar with is the Abbott Libre device where one attaches a sensor tag with a microfilament into the back of your arm; it stays put for the next fortnight and measures blood sugar every 15 minutes throughout the period. The sensor works by near field communication and there is a phone app that can download the data.

Installing the sensor is not painful; the sensor is quite waterproof and you can carry out your daily activities including exercise, bath and sleep. Of course, the interesting data that it measures is your dynamic blood sugar level and it is your glycemic response to food. I have done this for a number of patients and the data is fascinating.

I get the patients to log their food intake in an app that helps me capture the food and its carbohydrate, fat and protein content. We then correlate it with the blood sugar data that captures all the peaks and troughs of the blood sugar.

The food logging app and analytics that a Singapore wellness startup, Cura-Maker, had developed to provide an interpretation with the food intake is interesting. Combining pictorially highlighted spikes with statistical measures, the daily trend of the blood sugar levels is shown. We have found that in healthy non-diabetic patients certain foods that were universally thought to be healthy could actually spike the blood sugars for these patients. Some of these spikes were pretty high and we use a statistical measure called standard deviation to measure it. Usually one standard deviation could be calibrated as a norm but some of these blood sugar spikes deviated from the norm by two to three times.

An example would be cooked oats and blueberries and we found that this really spiked blood sugars in a number of patients. Some were very sensitive to bread including whole meal bread. Refined carbohydrates like white rice, white bread, potato and yellow noodles were a given when it came to spiking blood sugars. What was even more fascinating was how some of my patients developed low blood sugar after the spikes and some had low sugars in the middle of their sleep. There is actually a word for this and it is called hypoglycaemia.

Now these patients were not diabetic and in fact some of them were very fit and healthy individuals. We realised that they had to change their food intake to manage the spikes and the hypoglycaemia. Some interventions included eating a snack later after dinner that included carbohydrates with some fat, protein and fibre to correct the hypoglycaemia.

One of the problems is that low blood sugar during the sleep triggers quite high levels in the morning also known as Dawn effect. A number of patients also tend to have higher sugar levels in the morning and this could be the release of the stress hormones as we wake up to go to work. Other patients found that the sinful foods that they loved did not appear to spike their sugars. These included ice cream or desserts. This really goes to show that our sugar response to food is individual, pretty much like a thumbprint or a scan of the iris in the eye.

Realising all these patterns is interesting but what is the impact of correcting them? My patients corrected their diet habits by removing offending foods that spiked their sugars, changing combinations of food such as having high fat and low carbohydrate breakfast and eating protein first before carbohydrates. Some of them moved their exercise timings and factored in snacks at strategic times that allowed their sugar peaks and troughs to even out to produce a nice "sine wave" pattern. Subjectively, some of them felt better with less fatigue and more energy but interestingly a few of them lost weight or managed to drop their weight through a long stagnant resistant plateau level of weight.

So what is the science behind this? There is a lot of recognition that the traditional way of measuring average blood sugar levels over three months in diabetics called haemoglobin A1c did not manage to predict 89 per cent of the eye complications. There is a lot of buzz with the advent of continuous glucose monitoring devices that the many peaks and troughs or glycemic variability (GV) could be the answer.

The issue is that the medical world cannot agree on the best way to measure such peaks and troughs and to quantify it in a way that one can run clinical trials. Nevertheless there is a growing body of evidence that seems to point in the direction of GV. There is some evidence that this variability could impact heart attacks, strokes and mortality. In diabetic patients, it could hasten the development of nerve and retina complications. There is even implication of cognitive loss and depressive symptoms in certain cohorts of patients.

Today we have four continuous glucose monitoring devices approved by the Food and Drug Administration; in the tech world there are inroads made into the non-invasive measurement of blood glucose. The pharma companies are also rising to the challenge of non-invasive blood glucose monitoring. In a strange way this is exactly the disruption that Tesla has brought to the automotive world. We are all waiting for the arrival of such a device be it in a patch or in a watch. I suspect this will be a game changer in our fight against diabetes, obesity and all the attendant evils that accompany them.

  • This article is produced on alternate Saturdays in collaboration with Singapore Medical Specialists Centre

BT is now on Telegram!

For daily updates on weekdays and specially selected content for the weekend. Subscribe to