Time in Range references

Jasmin2000

Well-Known Member
Messages
107
Type of diabetes
Type 1
Treatment type
Insulin
Looking up the NICE / NHS and FDA guidance on TIR we find 3.9-10 mmol/L is recommended for diabetics, with a comment that <3.9 mmol/Lis dangerous.
Does anyone know or have references for the clinical studies / peer reviewed papers from which this range was takes? I mean 3.9 is pretty specific - why not just say 4.0?
 

TheSecretCarbAddict

Well-Known Member
Messages
146
Type of diabetes
Type 2
Treatment type
Tablets (oral)
I'm not sure about the lower limit, but I recently listened to The Diabetes Code by Dr Jason Fung and he mentioned that 10mmol/L is renal threshold for blood glucose, i.e., when body starts to expel glucose through urine.
 
  • Like
Reactions: Jasmin2000

Jasmin2000

Well-Known Member
Messages
107
Type of diabetes
Type 1
Treatment type
Insulin
I'm not sure about the lower limit, but I recently listened to The Diabetes Code by Dr Jason Fung and he mentioned that 10mmol/L is renal threshold for blood glucose, i.e., when body starts to expel glucose through urine.
Thanks @TheSecretCarbAddict, that's exactly the type of biological rationale I'm looking for.
I'm thinking the 3.9 mmol/L is due to induction of counter-regulatory hormones and a side-salad of insulin resistance, but getting this in black and white is really difficult.
 

TheSecretCarbAddict

Well-Known Member
Messages
146
Type of diabetes
Type 2
Treatment type
Tablets (oral)
Just did a search on Google out of curiosity, but no specific research comes up. Nothing apart from the fact that 3.9mmol/L maps to a nice and round number of 70mg/dL. WHO resource I found at the top of my search talks about mean fasting blood glucose metric with normal range of 70mg/dL and 100mg/dL. I guess the 70mg/dL is your nice and round number, but in different measurement systems, and we pick it because anything below it is no longer normal.
 

Lamont D

Oracle
Messages
16,114
Type of diabetes
Reactive hypoglycemia
Treatment type
I do not have diabetes
Thanks @TheSecretCarbAddict, that's exactly the type of biological rationale I'm looking for.
I'm thinking the 3.9 mmol/L is due to induction of counter-regulatory hormones and a side-salad of insulin resistance, but getting this in black and white is really difficult.
Many advisors argue about the lower limit target.
These are what I have researched. Though it is individual and take in account other conditions.
T1 is to be just above normal levels.
T2 is no lower than 4. However many do and cope without symptoms lower than that, the same can be said of 'false hypos' sugar crashes, are usually above normal levels (4-6mmols).
However overall, and from my specialist endocrinologist who is very special to me, informed myself that any reading under 3.5mmols, is definitely a hypo.

I hope that doesn't muddy the water much more!
 

Jasmin2000

Well-Known Member
Messages
107
Type of diabetes
Type 1
Treatment type
Insulin
Just did a search on Google out of curiosity, but no specific research comes up. Nothing apart from the fact that 3.9mmol/L maps to a nice and round number of 70mg/dL. WHO resource I found at the top of my search talks about mean fasting blood glucose metric with normal range of 70mg/dL and 100mg/dL. I guess the 70mg/dL is your nice and round number, but in different measurement systems, and we pick it because anything below it is no longer normal.
But that's defining the normal by saying anything lower is not normal? It leaves the question of why 70 mg/dL is defined as normal in the first place. There must be a biological rationale for why these figures were chosen.

So far I've heard of a) brain cells die below 3.9, b) it gives us a nice buffer before serious hypos, c) it minimizes development of hypo unawareness and d) BG below 3.9 is likely to trigger counter-regulatory responses. c) and d) are credible reasons - a) might happen at a much lower BG, and b) is a bonus consequence - these are what I'm gong to look for in the literature.
 

KennyA

Moderator
Staff Member
Messages
3,007
Type of diabetes
Treatment type
Diet only
But that's defining the normal by saying anything lower is not normal? It leaves the question of why 70 mg/dL is defined as normal in the first place. There must be a biological rationale for why these figures were chosen.

So far I've heard of a) brain cells die below 3.9, b) it gives us a nice buffer before serious hypos, c) it minimizes development of hypo unawareness and d) BG below 3.9 is likely to trigger counter-regulatory responses. c) and d) are credible reasons - a) might happen at a much lower BG, and b) is a bonus consequence - these are what I'm gong to look for in the literature.
Bilous and Donnelly's Handbook of Diabetes might be a good place to start. For instance, B&D say that 48 mmol/l was chosen as the automatic diagnosis point because moderate diabetic retinopathy is rare at levels beneath that figure, but incidence rises sharply at higher levels.


PDF here
https://onlinelibrary.wiley.com/doi/book/10.1002/9781118976074

I've seen various population cohort studies of non-diabetic people showing a marked clustering around 38mmol/mol - graph from the Dutch Lifeline study attached as example.
 

Attachments

  • HB copy.jpg
    HB copy.jpg
    140.7 KB · Views: 60
  • Like
Reactions: Jasmin2000

In Response

Well-Known Member
Messages
3,549
Type of diabetes
Type 1
Treatment type
Pump
Not sure about the 3.9 but my endo once told me our brains need 2.8 mmol/l in order to work properly.
So, I guess the lower limit needs to be higher than this.
I had also assumed that the reason for 3.9 was that it was "anything below 4" but no idea why 4 is relevant apart from being a round number.
 
  • Like
Reactions: Jasmin2000

Jaylee

Oracle
Retired Moderator
Messages
18,253
Type of diabetes
Type 1
Treatment type
Insulin
Looking up the NICE / NHS and FDA guidance on TIR we find 3.9-10 mmol/L is recommended for diabetics, with a comment that <3.9 mmol/Lis dangerous.
Does anyone know or have references for the clinical studies / peer reviewed papers from which this range was takes? I mean 3.9 is pretty specific - why not just say 4.0?
The DVLA guidance regarding fitness to drive regarding BG levels might throw some sort of light on it? (Edit, other countries may differ?)


I’m confident I can still function @3.9.

But sub 3.5 my eyes start to “strobe” just before the symptoms if dropping lower?
 

Jasmin2000

Well-Known Member
Messages
107
Type of diabetes
Type 1
Treatment type
Insulin
Bilous and Donnelly's Handbook of Diabetes might be a good place to start. For instance, B&D say that 48 mmol/l was chosen as the automatic diagnosis point because moderate diabetic retinopathy is rare at levels beneath that figure, but incidence rises sharply at higher levels.


PDF here
https://onlinelibrary.wiley.com/doi/book/10.1002/9781118976074

I've seen various population cohort studies of non-diabetic people showing a marked clustering around 38mmol/mol - graph from the Dutch Lifeline study attached as example.
Doubtless, the consensus recommendations for the range itself and the TIR come from many clinical studies and I was hoping to delve into the NICE/FDA guideline references, and the references cited in those references, to find the original data, but I'll probably end up with some fragment of the Rosetta Stone that I can't understand. So publications like B&D that cite the range and provide rationale are just as good for exemplification and we can add retinopathy to the list with a causal BG (presumably out of range) and a time-frame.

For events happening at certain BGs, I found this in one of the summary references as a guide to some well-known symptoms of hypos and clearly 70 mg/dL (3.9 mmol/L) is depicted as the average BG where glucagon and epinephrine secretion is increased and I assume there's a paper somewhere with the data for this.

main-qimg-2cf57a4ad6e2e550a4d22b88c0ad3bd9-lq.jpg


This gives me a biological rational for the 3.9 mmol/L border and why it is "dangerous" to go below. I'll go back to the consensus recommendations docs to find out how that "danger" is defined.
 

Jasmin2000

Well-Known Member
Messages
107
Type of diabetes
Type 1
Treatment type
Insulin
Not sure about the 3.9 but my endo once told me our brains need 2.8 mmol/l in order to work properly.
So, I guess the lower limit needs to be higher than this.
I had also assumed that the reason for 3.9 was that it was "anything below 4" but no idea why 4 is relevant apart from being a round number.
Your endo cites what appears to be a curious number at first - but see the figure in my post above; 2.8 mmol/L corresponds to "decreased cognition, aberrant behavior."
 
  • Like
Reactions: In Response

KennyA

Moderator
Staff Member
Messages
3,007
Type of diabetes
Treatment type
Diet only
This is from B&D and it shows the same kind of thing.
 

Attachments

  • Fig.13.4.jpg
    Fig.13.4.jpg
    45.8 KB · Views: 52
  • Like
Reactions: SimonP78

Jasmin2000

Well-Known Member
Messages
107
Type of diabetes
Type 1
Treatment type
Insulin
This is from B&D and it shows the same kind of thing.
Wow - there's physiological changes to hypo unawareness?! I always though it meant you didn't notice your confusion or lack of concentration, which are taking place anyway - but no, counter-regulatory responses are actually at a lower BG. This makes a mockery of the hypo awareness questionnaires - they should be measuring blood values for glucagon.

My counter-regulatory spikes begin at 5-6 mmol/L and go up to 10 mmol/L - does that make me super hypo aware? :D
 

EllieM

Moderator
Staff Member
Messages
9,388
Type of diabetes
Type 1
Treatment type
Pump
Dislikes
forum bugs
I like this study on what happened when 153 healthy non diabetics wore a dexcom cgm for ten days.

A median time of 15 minutes a day was spent under 3.9mmol/L or 70mg/dL (this was a US study so mg/dL were the primary units used).
28% of the cohort had at least 1 hypo event. A hypo event is defined as at least two sensor values <54 mg/dL that are ≥15 min apart with no intervening values. (Having said that, the study admitted that hypo events were much more common during the first day of sensor wear, and I personally would want to discount those, as I know dexcom under reads for me in the first 24 hours.)