One of my meters is lying

[Oldvatr]

At the moment I am doing 'blue sky' research, which is akin to throwing spaghetti at a wall to see what sticks. I am open to others joining in the brainstorming and suggestions are welcome. I am definitely biassed since I have a vested interest in identifying the contaminants since it is in my blood, it's increasing slowly, and I want to take avoiding action to prevent trouble down the line.

It may be the condition I have that's causing this is specific to my body, and maybe the effect is personal to me. If anyone else suspects they are seeing this sort of error that seems to be meal-related, then please add your own story to the thread - it may occur on other meter technologies as well. I do strongly think it is enzyme technology dependant, so maybe the Navii replacement from SD also shows this effect.

 

[Oldvatr]

OK, I am having another lightbulb moment. I promise that if I run down the street shouting 'I have it' or Eureka, then I will be fully clothed. No doubt others observing this would think I am declaring a Covid-19 diagnosis and keep well away from me.

Looking into the Enzyme used I found this paper
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3208862/
I believe that the enzyme in the SD is GDH= FAD, and is the fungal-derived one. In which case it is sensitive to Xylose instead of Malitol.
Looking further into xylose, it is a common part of food.
https://www.researchgate.net/publication/287299810_Xylose_as_a_dietary_sugar_Sources_and_analysis

Now, I have in the last couple of months been increasing my carb intake. Mostly this is in the form of wholegrain bread, and also potatoes. Bingo!. Both are a source of xylose

Now I have noticed as an 'en passant' that when I buy miniature potatoes then my sugars are well behaved, but recently my daughter has been buying medium potatoes that are not babies anymore. I also know that standard older potatoes are not good for me at all. So recently I have been eating more and older (starchier) potatoes in increasing portion sizes.

At last, I can identify a mechanism associated with my meals and readings thereof. It may not be my body at fault but simply due to my gluttony. I am happier now. If this is the case then others here will probably not see this effect.

 

[Antje77]

Now I have noticed as an 'en passant' that when I buy miniature potatoes then my sugars are well behaved, but recently my daughter has been buying medium potatoes that are not babies anymore. I also know that standard older potatoes are not good for me at all.

Completely off topic, but I'm definitely going to try out miniature potatoes to see if they behave better than the regular ones! Would be a very nice surprise if they do

 

[Oldvatr]

Completely off topic, but I'm definitely going to try out miniature potatoes to see if they behave better than the regular ones! Would be a very nice surprise if they do

May I recommend Aldi as a reliable source? That's where I usually get mine, and they are labeled as miniature.

 

[LittleGreyCat]

Completely off topic, but I'm definitely going to try out miniature potatoes to see if they behave better than the regular ones! Would be a very nice surprise if they do

I always understood that new potatoes (which are naturally small) had less free starch than old potatoes which had bulked up on the starch as they grew.

I'm not sure where small old potatoes sit in the starchiness range.

Also, boil them, chill them, slice them, fry them.
Allegedly reduces the availability of the starches.

 

[Oldvatr]

I always understood that new potatoes (which are naturally small) had less free starch than old potatoes which had bulked up on the starch as they grew.

I'm not sure where small old potatoes sit in the starchiness range.

Also, boil them, chill them, slice them, fry them.
Allegedly reduces the availability of the starches.

freeze them? It is the starch, hence the Xylose

 

[Oldvatr]

For centuries now, pig farmers have given their pigs straw as both bedding and feed. Now although straw and hay contain xylose in abundance, the poor pigs do not produce the enzyme needed for metabolism, so the xylose goes in and comes straight out again unchanged. Modern pig farmers now use pellet feed containing the enzyme and the pigs get fatter and sweeter and the yield rate improves for the farmer. This process increases the hidden sugars in pork products to our detriment, I have recently been eating more pork and less beef, so again this may be part of the meal effect I am reporting here. Not sure if cattle get the same enzyme in their winter feed cake but they do get to pasture more than pigs and their hay intake is low relative to pigs straw, I think. Don't think silage is high in xylose either.
For Pigs
https://jasbsci.biomedcentral.com/articles/10.1186/s40104-017-0226-9
and beef
https://pubmed.ncbi.nlm.nih.gov/22591448/

 

[Oldvatr]

So far no one has called out the elephant in the room. My Codefree meter is the one suspected of lying, and it uses the GOD Sensor (Really! Glucose oxidase technology or GOD). My treatise so far has deviated to looking at the GDH-FAD sensor used in the other meter of mine. Not to worry. Apparently and according to some papers I have read, the GOD ones are more susceptible to xylose as well as maltose. Now the ISO standard only lists Maltose as an interferent, and only tests maltose but not xylose. Apparently, the earlier technology of FAD sensors was also susceptible, but the new FAD3 grade that is being used in recent meters is designed to filter it out and do not react to this monosaccharide. Not sure what grade the Caresense Dual uses. The new Navii from SD is claiming that it too is not as sensitive as the older Codefree, and it uses a GDH FAD sensor technology. However, it may be that SD is already aware of the effect I am seeing, and hopefully, their replacement meter is more robust.

I have pinged off an email to SDBiosensor in Korea to see if they have any answers.

 

[Oldvatr]

I was just packing my thoughts away to avoid wearing my brain cells out when I made another related discovery that may also be of interest, although I do not think it is involved in my saga here.
Apparently most meters are affected by immunoglobulins. Not hobgoblins but equally mischievous. These are antibodies released by plasma (white cells) in response to a call to arms by our immune system, Apparently, the presence of these in the blood can lead to falsely elevated bgl readings. It seems that meters using the GDH PQQ enzyme technology are most prone to this. and the FDA has issued a safety warning
https://www.sciencedirect.com/topics/nursing-and-health-professions/glucose-dehydrogenase
This may also be of interest
https://www.hpra.ie/docs/default-so...ic-monitor-list-ireland_v1-09-15.pdf?sfvrsn=2

This may explain why people getting the covid vaccination noticed high bgl results for a while, and why people in hospital may struggle to keep their levels in control. I was wondering why my consultant felt that bgl below 13 mmol/l was considered good and why when I clocked 20+ it was not a blue light event for them.

 

[Oldvatr]

So far no one has called out the elephant in the room. My Codefree meter is the one suspected of lying, and it uses the GOD Sensor (Really! Glucose oxidase technology or GOD). My treatise so far has deviated to looking at the GDH-FAD sensor used in the other meter of mine. Not to worry. Apparently and according to some papers I have read, the GOD ones are more susceptible to xylose as well as maltose. Now the ISO standard only lists Maltose as an interferent, and only tests maltose but not xylose. Apparently, the earlier technology of FAD sensors was also susceptible, but the new FAD3 grade that is being used in recent meters is designed to filter it out and do not react to this monosaccharide. Not sure what grade the Caresense Dual uses. The new Navii from SD is claiming that it too is not as sensitive as the older Codefree, and it uses a GDH FAD sensor technology. However, it may be that SD is already aware of the effect I am seeing, and hopefully, their replacement meter is more robust.

I have pinged off an email to SDBiosensor in Korea to see if they have any answers.

I have now received acknowledgment of my email requesting some further info. Looks promising

 

[LittleGreyCat]

I'm currently puzzling over my mild anaemia.

Libre 2 is reading interstitial fluid.
Freestyle Freedom Lite finger prick is reading (I assume) a mixture of venous blood and interstitial fluid.
HbA1c is reading haemoglobin from venous blood draw.

Libre 2 and finger prick seem to broadly agree.

HbA1c is a lot more pessimistic.

Which is likely to be more accurate?

 

[Oldvatr]

I'm currently puzzling over my mild anaemia.

Libre 2 is reading interstitial fluid.
Freestyle Freedom Lite finger prick is reading (I assume) a mixture of venous blood and interstitial fluid.
HbA1c is reading haemoglobin from venous blood draw.

Libre 2 and finger prick seem to broadly agree.

HbA1c is a lot more pessimistic.

Which is likely to be more accurate?

Depends on what you want the result to be. A fingerprick result is a spot check and is frozen in time. If you take enough of them you can get an average over the time period, or by day or by week so it will have the result of removing the individual peaks and troughs, so giving a more mellow result, and reducing noise. So that takeaway last night will gradually recede into the distance in terms of its overall effect on your body. The HbA1c is like an average over several months and gives a general feeling of how well your overall control is, so is useful for making long term adjustments to medication, etc. (tweaking)

The fingertip test results are the frightening ones. These are used for bolus doses and compensation doses of insulin, so need to be the most accurate.

The HbA1c is a warm fuzzy feeling if good, a tut-tut from the doc if not.

I use fingertip because I do it daily with a food log, so I can immediately identify those meals that cause a 'slap-head' moment and I can also average them myself and see the trend they are going in. I can also do a moving average that allows me to emulate the HbA1c. I think most meter Apps can do these nowadays,

The accuracy obeys the GIGO rule, garbage in - garbage out so it s the accuracy of conversion of whatever meters you are using that determines accuracy. Averaging smooths out any meter errors over time so gives a more stable figure

Bear in mind that the YSI or Beckmann blood analyzers used in the Lab are only 3% accurate themselves

The monthly average result is like HbA1c, and there is a conversion tool on this website you can use to convert between the two. But they are measuring different things. Modern meters are calibrated to give results that are closer to the lab results, but older meters used to be calibrated to 'plasma' blood and were a factor of 1.2 out.

The Libre is factory calibrated to meet the same as fingertip meters. It requires a scan every 8 hours to get the most accurate results. Longer sample periods reduce accuracy.

 

[LittleGreyCat]

Depends on what you want the result to be.

I am aware that a finger prick is a moment in time.
I use finger pricks to check the accuracy (or not) of the Libre 2 readings.
In other words, if I regularly cross check between finger prick and continuous readings I should have a more accurate picture than from either on their own.

As I have stated elsewhere, my LIbre estimates HbA1c as around 6% which is (on some measures) only just pre-diabetic.
My blood draw HbA1c was 7% which is around the threshold for starting second stage medication.

What I want to know is how good/bad my control is.

Can I rely on a blended mix of continuous monitoring cross checked with finger pricks, or should I be guided by HbA1c?
Further, can I back up any decision with my care team?

 

[Oldvatr]

I am aware that a finger prick is a moment in time.
I use finger pricks to check the accuracy (or not) of the Libre 2 readings.
In other words, if I regularly cross check between finger prick and continuous readings I should have a more accurate picture than from either on their own.

As I have stated elsewhere, my LIbre estimates HbA1c as around 6% which is (on some measures) only just pre-diabetic.
My blood draw HbA1c was 7% which is around the threshold for starting second stage medication.

What I want to know is how good/bad my control is.

Can I rely on a blended mix of continuous monitoring cross checked with finger pricks, or should I be guided by HbA1c?
Further, can I back up any decision with my care team?

You have invested in the Libre, and it makes sense to use that investment. Use the finger checks to check it is on target as advised in the user manual. You have a choice of relying on the Libre for all your decisions or just to do the graphic plotting and averaging. Since the sensors are expensive and time-limited this may prompt you to do it the other way round and rely on finger tests to make decisions. I have no personal experience of the Libre but I have seen comments about its accuracy and HbA1c calculations which indicate a possible error mechanism in the technology, but I am not able to say with any certainty.

Personally, I would go for a standard meter since if it breaks or goes off-key, then I can source a replacement easily and I am not left high and dry. The CGM is a nice toy, but the technology is new, it is single-sourced, and there will be others coming along in time that will be cheaper and better. It does cut down on the extra processing of results, that's for sure, and gives a picture without requiring a degree in spreadsheet use. But my spreadsheet ties in with data from my meals which is also very useful. Hobson's Choice. Your Choice too.

 

[Covlocks]

I found a very similar problem. My code free has been good for me, but my GP gave me a new machine 6 weeks ago, a Wavesense Jazz. The readings were way out from my code free so I called the manufacturers. We worked out the problem. The Jazz lancets are very very thin so the blood sample for that machine is very tine, whereas the code free the lance is much thicker.
So what did that mean? The slightest contamination of the blood sample for the Jazz throws the reading right out (who scrubs their hands before every test?) whereas the code free takes a larger drop of blood which dilutes any small contamination.
So scrub my hands before testing with both = same reading, but do a test during the day where I haven’t scrubbed my hands = completely different readings to the point that one afternoon my jazz gave me a 30.1mmol reading but my code free gave me 8 (I repeated the test, same reading). Conclusion is new machines that need a very tiny blood sample need scrupulously clean hands.

 

[Oldvatr]

Update. SD Biosensor has not followed up with anything so far. My Caresesense Dual started playing up (bad connection on strip port) and they replaced my meter. Limited parallel testing (3 in a bed) shows that the replacement meter tracks the original Caresense within 0.7 mmol/ l as per the ISO requirements, so I have more confidence in using this meter as a baseline. The SD is still jumping all over the shop when I eat by up to 3 pr 4 mmol/l higher than the Dual.