E500 - can people please check if it does measure blood glucose or just makes it up

I thought I would post this as a new topic. Those who are following this know I have been using two CGMs Dexcom and E500/hBand. I also visited a lab yesterday at 11am. At that point the lab said my blood glucose was 5 mmol/l, the Dexcom said 7.4 mmol/l and E500/hBand said 6.15.

Today, however, I have started with a fast and not had my usual breakfast. I have just eaten some chia seeds and also taken my second 2mg dose of rapamycin (I intend being quite severely calorie restricted just short of a fast, but I wanted some food to go with my daily supplements). However, E500/hBand is coming up with a peak in blood glucose with similar timings to the previous days although with a very slightly different maximum. Dexcom, however, is showing a slight downtrend in blood sugar which is the sort of thing you would expect.

Hence I wonder if E500 is actually making it up on the basis of 3 meals a day. The peak of the meals appears to be around the same time 10.20, 14.15, 20.10 each day. I am quite routine orientated so it would not be surprising normally.

If other E500 users could check their glucose charts and see if they appear dubious that would be kind.

1 Like

To start;

Did you adjust the fasting default to the 5 mmol/l in the H Band app?

I have not adjusted anything. If it is always 1mmol/L too low I don’t mind. My issue is that it does not seem to be actually measuring glucose.

In any event I don’t actually fast for blood tests. Because I do them every week I see no reason to fast. Hence I would expect a fed level of glucose based upon eating at 9am.

In a strict sense it read a higher level than the lab. You could argue for an interstitial fluid delay of up to 20 minutes, but that does not explain the situation.

My reading from midnight 01/07/2023 to 5:00am
H Band software using E500 watch

The shape of the chart looks similar

You have a missing 30 minutes from 09:00 to 09:30.

Did you have the watch off?

Then the reading increased.

I had the watch charging, there are, however, still values on the chart.

The above is dexcom for today. It may overstate the absolute value by up to 2.4 mmol/l, but it does recognise that I did not eat breakfast.

Wonder if we have any programmer type members, who can reverse-engineer the software.

I am a programmer and have written mobile software. However, if everyone has the same shaped chart it is not measuring anything.

The H Band software is used by many different models of watches. Before the watches had a “glucose monitor” module.

A better idea is to get these glucose watches to work with Levels and NutriSense.

The the market will explode in size.

How many people could afford $300.00 per month for the CGM currently used by those{Levels and NutriSense] program’s?

How many could afford a $50.00 “watch” and a very small monthly or yearly fee to use Levels or NutriSense?

1 Like

If you reverse engineer, you can confirm your suspicion.

H Band collects data and sends to their algorithm/software. Like all other app.
Write to the person/company who developed H Band, jiafu zhang. His point of contact in in the program.

Because I am running two CGMs simultaneously it is pretty obvious.

It is probably one to check out with the watch manufacturer. I cannot see contact details for them.

The Developer contact is on the app downloads page, it in the Google app store.

Also I just(01/07/2023) went to the Google app store, to update the H Band app to latest version.

Looking a bit further into this there is a function on the watch to measure glucose. It takes a while to run, asks you to stay still and comes out with values which differ by up to 1 mmol/l on the trot.

I will try this overnight.

FWIW

Saturday the 7th of January-

Had 10oz black coffee started at 8:45am finished 9:08am

Then had a ciabatta roll and 8oz/225g of burrata filling “stracciatella”
Started at 9:08am finished at 9:18am

There are functions on the watch for spot measurements of several parameters and you can also control/run the measurements parameters from the software.

With the chest band on your chest the ECG records continuous.

I have superimposed my chart on yours. It does seem to be trying to fit any responses from the phone to a three meal day.

When you run the glucose tests from the watch they do vary a lot in sequence. Hence it may be that the software tries to smooth out the variations and fit things to a particular curve, but I would prefer it to report the actual test results and then if needs be create a curve from them.

Dexcom is a bit erratic in its reporting, but it is reporting something close to reality.

FWIW

The measured glucose number are recorded/stored every 5 minute’s. Then recorded/stored in 30 minutes block showing highest and lowest glucose in the 30 minutes time frame.