Live experiment part 4: pure erythritol

[Alecmcq]

When was your last meal before you started the experiment? Have you considered doing an EF until your blood glucose is low and stable and then trying the experiment? I’m wondering if your body might still have been processing residual glucose when you started

Karen
Good thinking. My previous meal was eggs Benedict at 9am, and I started this testing at 12.30pm. So 3.5hr gap. I am not T2D and I think my IR has significantly reduced during the past year on fasting/keto, so I think any residual effect from the meal would have gone. But to be fair I don’t know that, I am guessing. I will add fasting to my list of tests.

how did you come to settle on the 6 g amounts for testing? Second, what is the accuracy of your meter? Third, how many times do you repeat each test?

1. The 6g was my normal “dose” that I have on my raspberries.
2. I don’t know, and I have listed monitor testing on my list of tests I am going to perform
3. At the moment I don’t repeat, they are all single tests. I am going to repeat this test, as the results seem strange to me.

All good thinking. I now have a list of tests to do!! I try to do 2 tests each weekend, but as they each need 2hrs some time away from a meal, this does limit things.

But I am having fun here, so it’s no dramas doing it. And the pricks don’t hurt me. Hardly feel them!!

Thanks for everyone’s feedback and ideas.

[Duncan_K]

Have you done a control - testing over those time periods without eating anything to see what your natural variability is?

[mtncntrykid]

In reference to @PaulL comment.
https://www.healthline.com/diabetesmine/why-meters-cant-tell-us-our-blood-sugar-levels

[Alecmcq]

In reference to @PaulL comment.

Healthline

Why Glucose Meters Can’t Tell Us Our Blood Sugar Levels

Given that many diabetics rely heavily on blood glucose meters to ensure their blood sugar levels are stable, we explore meter accuracy and why results vary.

Ron
This is now 6 years old, so although some of the points made in here are valid, I am expecting/assuming the technology to have come along a bit in that 6 years. The focus in this article is on absolute accuracy, and I am not that fussed with absolute accuracy. What I am concerned about though is relative accuracy/consistency. That’s what matters to me. I don’t really care whether I start at 4.5 or 5.5. What I care about is moving up or down from the starting point, and thus relative accuracy/consistency is important.

I am going to test my monitor in various ways for consistency, and we’ll see how that goes. What I am going to test is:

1. Same finger, same drop of blood, 3 tests one after the other, no time gaps.
2. Different fingers, same hand, 3 tests, no time gaps
3. Different hands, 2 tests per hand, no time gaps
4. Same finger, 3 tests, 5 minutes apart

Anybody got any other ideas for testing monitor consistency? After all that testing, I think my fingers will be wanting a rest!

[Alecmcq]

Have you done a control - testing over those time periods without eating anything to see what your natural variability is?

Duncan
Good suggestion. The straight answer to your question is no I haven’t, but it is a good idea. I will do it.
Cheers
A

[mtncntrykid]

This is an excellent idea and actually what I was thinking when Paul addressed the inaccuracy of meters. Not so much the reading being off but the inconsistency between readings and potentially the cause of increase and decrease in your test. I know from experience that I have followed up with a back to back test and had variations. It is a good point though about advancements in technology as my meter is about 8 years old.

[PaulL]

The point of a margin of error is that if your reading is 4.8 +/- 0.2, then the actual level in your blood could be anywhere from 4.6 to 5.0. This means that readings of 4.8, 4.9, and 4.7 are for all intents and purposes the same blood value, because there is no way to distinguish between readings whose margins of error overlap. Depending on the accuracy of Alec’s meter, the readings he got tonight could possibly all be within the margin of error and therefore essentially the same.

[Alecmcq]

Depending on the accuracy of Alec’s meter, the readings he got tonight could possibly all be within the margin of error and therefore essentially the same.

Absolutely, and I was assuming a 0.2 margin being just margin of error noise. But of course that is total guesswork. So I think a few tests to see what the variation really is are in order. I will do some tonight if I remember.

I am singing in a concert this afternoon, and I always sing fasted (do much better on an empty stomach), so I will be reasonably fasted (12 hrs) when I get home this evening.

[TGAMEL62]

Todd
Thanks for your thoughts. And you are right, the changes are not high, they are quite subtle.

I think your conversions are slightly off. 5.0 = 90.
6g = 0.212oz
Cheers
Alec

Sorry, I did not do the calculations, just looked at a chart and copied the information from the wrong rows, actually all of my numbers were incorrect, but I have rectified that. Sorry bout that, guess I shouldn’t post first thing in the morning before having my cup of tea. slight_smile:

[KetoLikeaLady]

This is indeed what I am testing here. The theory is that erythritol should not raise BG (although it did raise mine a bit), and if the sweetener causes an insulin response, then BG will go down.

For clarity, I meant that Fung says that with the cephalic insulin response, you may not see a change in blood sugar at all, even though you are actually having an insulin response.

[Alecmcq]

For clarity, I meant that Fung says that with the cephalic insulin response, you may not see a change in blood sugar at all, even though you are actually having an insulin response.

Ah, in which case, I don’t understand that, and I need to go and do some research. Thanks for pointing this out.
Cheers
A

[PaulL]

I was assuming a 0.2 margin being just margin of error noise. But of course that is total guesswork.

Does the manual for your meter say anything about its accuracy?

[Alecmcq]

Does the manual for your meter say anything about its accuracy?

I will have to try to find it. My memory says something like 5% which was where my 0.2 came from… but maybe I am remembering badly…

[Duncan_K]

Have you done a control - testing over those time periods without eating anything to see what your natural variability is?

Duncan
Good suggestion. The straight answer to your question is no I haven’t, but it is a good idea. I will do it.
Cheers
A

Another thought - if this is to test the best sweetener to have on your raspberries, it might be an idea to test under the circumstances you intend to use the sweetener - with your raspberries. You might actually find that the raspberries spike your insulin more than the sweetener would anyway.

[Alecmcq]

Another thought - if this is to test the best sweetener to have on your raspberries, it might be an idea to test under the circumstances you intend to use the sweetener - with your raspberries. You might actually find that the raspberries spike your insulin more than the sweetener would anyway.

I am not entirely sure how to interpret the results if I did that. The raspberries will undoubtedly raise my BG levels, which will then release insulin and my BG would come down. I am trying to see if the sweetener drives an insulin response, and I am not sure how I could disentangle the effects of the raspberries and the sweetener.

[Duncan_K]

Eat raspberries without sweetener as the control, then raspberries with the various sweeteners. If there’s no significant difference then the sweetener can’t be having an effect larger than the raspberries.

[mtncntrykid]

I think just as your doing.Test for sweeteners, then you could test the raspberries, then you could test them together. I would be interesting if you got a sweetener rise and also a raspberry rise and how those numbers would correlate with the number of the combined test.

[Alecmcq]

Duncan, Ron
I’ll have to sleep on those ideas.
Cheers
Alec

[Alecmcq]

Does the manual for your meter say anything about its accuracy?

From the website for my monitor:

Accurate—reliable results you can trust1

1Data on file. [10/10 accuracy: 95% of measured glucose results shall fall within ±10 mg/dL of lab reference value for blood glucose concentrations <100 mg/dL and within 10% for blood glucose concentrations ≥100 mg/dL.]

So, essentially, they are promising +/-10%. Not great. So, a reading of 5.0 could be anywhere between 5.5 and 4.5. So all the readings I have taken so far are within the bounds of the declared accuracy.

So, why bother? I reckon the monitor is more accurate than that, and that is what I am about to test.

First test: 3 readings, same drop of blood: 4.3, 4.2, 4.2. Now that is not a bad start!!

2nd test: 3 readings, 3 different fingers, same hand: little finger 3.9; middle finger 4.0; index finger 4.2. Not so good, especially as this test was immediately after test 1!!

3rd test: 3 tests from the other hand: 4.0, 4.3(new strip pack), 4.1

4th test: 3 tests, same finger, 5 mins apart: 4.3, 4.4, 4.2

Analysis
Total recordings: 12
Average: 4.175
Std dev: 0.15

Conclusions:

1. My estimated 0.2 variation “noise” looks ok to me (compared to a StdDev of 0.15)
2. This outcome is significantly better than the declared accuracy from the makers. However, this could well be the absolute variation from a lab test result, not a measure of internal consistency.
3. I am pretty happy with this level of consistency. Although there was a 3.9 outlier, 11/12 results were very close to within 0.2 of the average.

What do you lot think? My fingers deserve a holiday!
Cheers
Alec

[PaulL]

Welcome to the wild and wacky world of science!

Actually, it seems as though you are on the right track.

I also like the suggestion of testing yourself with berries alone, and then with sweetener, just to nail all the possibilities down. Suppose the berries’ effect on BG and insulin is great enough that it swamps the effect from a sweetener? Then you wouldn’t have to worry about it. Would be good to know, eh?