Tighter blood glucose control through medication linked to higher death risk

[Grateful]

I agree. It must be hypo comas causing death if good control on meds and insulin reduce inflammation and infections etc.

Maybe. Or perhaps taking sulfonylreas long-term is adding mortality risk in a different, unspecified way. From what I can tell, the study doesn't address that issue at all. Presumably the data were judged to be interesting enough without trying to figure out what the actual mechanism is.

 

[ringi]

@ickihun I think you have missed the main point of the study, that "length of life" is not improved by moving from "OK" BG control to "great" BG control when using most drugs/insulin. Hence trying to push BG down to "normal" levels in Type2 with insulin is not a good option, better to accept a level that is a little higher. (The ACCORD study showed the same.)

The study looks at 3 ranges of AC1 results, I expect they did not have enough data to look at more ranges. I don't have access to the full paper or dataset to see a graph of AC1 results in each range and it is very unlikely that there is anything "magic" about 7%(53), I expect it was just a nice round number.

Below 7% (53), Between 7%(53) and 8.5%(69), And over 8%(69) People on insulin etc did best when they were between 7% and 8.5%, people on metformin 53 only did best when below 7%.

The NICE guidelines already have this, with the target AC1 being higher for people who are on more drugs or insulin.

We also have the issue that the study tells us nothing about what happens when these drugs/insulin are combined with low carb....

There may also be a choice to be made between "length of life" and "not going blind" etc........ But I don't trust the research enough to know if this is the case.

Therefore maybe you should reduce your insulin a little, to allow your BG to be a little higher, but it is possible the data is not strong enough to show your AC1 of 48 is too low. You could email the people who did the study to ask them what they think.

(PS remember that high insulin increases inflammation, and I think hypos had nothing to do with the results.)

 

[Biggles2]

Thanks for posting the abstract here.

@Oldvatr you can link to the full article from the abstract, or via the following link:
http://onlinelibrary.wiley.com/doi/10.1111/dom.13155/pdf

So the obvious conclusion I would draw is that if normal HbA1c is achieved by lifestyle changes only (i.e. as if we were normals) then thats the best remedy, Doctor.

Yes, I would agree with you! Unfortunately, most studies are funded by the pharmaceutical industry, so most of the research looks at outcomes associated with medications, not outcomes associated with dietary interventions. For information on the sponsor of this study, see the acknowledgements on page 20 of the article.

 

[Biggles2]

But remember that for a drug to cause hypos it must increase insulin levels....

Agreed @ringi, Dr. Jason Fung explains the failure of the ‘glucotoxicity paradigm’ to address the real issue which is insulin toxicity in the following October 2017 article:
“Failure of the Blood Glucose Paradigm”:

“If the problem is both insulin toxicity and glucotoxicity, then increasing insulin toxicity to reduce glucotoxicity is not a winning strategy.”​

https://idmprogram.com/failure-blood-glucose-paradigm/
The last paragraph is informative.

 

[LittleGreyCat]

The algebra left me confused, but I think it is more or less saying:

If you are on Metformin then high average HbA1c can reduce your life expectancy. I assume that they assume that you will rarely if ever be skirting hypo territory. There must be other drugs which lower BG without a hypo risk as well.

If you are on insulin and/or other drugs which actively lower BG and therefore carry the constant risk of hypos, then maintaining a low average HbA1c can increase your risk of mortality.

There is no explanation of the causes. It may well be that to maintain a low average HbA1c you have to constantly skirt hypo territory and that this is more damaging than running a higher (but not very high) average HbA1c. This could be very important for out T1 friends because there is the temptation to try and maintain "normal" levels of BG with strict control. I am sure i have read debates where HCPs and T1s have disagreed about targets.

One confusing thing is that it seems to imply that if you are on insulin et. al. then running high HbA1c averages doesn't reduce your life expectancy but if you are just on Metformin then it does.

More research (or at least more detail) needed.

 

[ringi]

There must be other drugs which lower BG without a hypo risk as well.

There were none in common usage at the time the data was collected. I think they only included people who were on the same drugs for the complete time. Hence SGLT2 inhibitors were not included in the study.

One confusing thing is that it seems to imply that if you are on insulin et. al. then running high HbA1c averages doesn't reduce your life expectancy but if you are just on Metformin then it does.

More research (or at least more detail) needed.

Firstly that is not what they found, they found that on insulin et. al. that high HbA1c and low HbA1c both reduce your life expectancy, with mid-range being best.

I think we already know the answer to this, that high insulin levels in our bodies are not good, and all the drugs in the study apart from Metformin work by increasing Insulin levels. Very high BG is worse than increased insulin levels, but mid-range BG is a good tradeoff in exchange for a lower blood insulin level.

As the study was just looking at the data, the only included in the paper what they could see from the data they were working with. It is assumed that everyone will be reading all papers about Type2. After all, this is a research paper not a chapter from a textbook.

 

[ickihun]

@ickihun I think you have missed the main point of the study, that "length of life" is not improved by moving from "OK" BG control to "great" BG control when using most drugs/insulin. Hence trying to push BG down to "normal" levels in Type2 with insulin is not a good option, better to accept a level that is a little higher. (The ACCORD study showed the same.)

That's just it a little higher causes me problems with numb feet and constant thrush. Exercise cannot help. I also have atherosclerosis which can get worse with uncontrolled diabetes. Then there's my hormone imbalances PCOS which is great on lower bgs. Weight loss too.
You cannot tell me that I don't benefit from not getting these things in good control and far less worry. Too much anxiety and mood swings aren't good for lasting relationships and career prospects.
The whole picture for me is what's important.
So
Weight loss til on as little insulin has been my on-going plan for a few yrs now. Progress is being made....which will continue.
I won't be panicking with this finding but I must express the whole health picture should be considered not snippets, like this seems to cover.
I won't be risking high bgs after these findings, no one should be expected too either.

 

[ringi]

@ickihun You are being very sensible choosing a BG level based on how it affects your body.

This study stops doctor's thinking that a lower AC1 is always better when these drugs in a use.

 

[ickihun]

There were none in common usage at the time the data was collected. I think they only included people who were on the same drugs for the complete time. Hence SGLT2 inhibitors were not included in the study.



Firstly that is not what they found, they found that on insulin et. al. that high HbA1c and low HbA1c both reduce your life expectancy, with mid-range being best.

I think we already know the answer to this, that high insulin levels in our bodies are not good, and all the drugs in the study apart from Metformin work by increasing Insulin levels. Very high BG is worse than increased insulin levels, but mid-range BG is a good tradeoff in exchange for a lower blood insulin level.

As the study was just looking at the data, the only included in the paper what they could see from the data they were working with. It is assumed that everyone will be reading all papers about Type2. After all, this is a research paper not a chapter from a textbook.

I guess me losing weight on excellent bgs on insulin if I wasn't over-weight could threaten my life. If only! (Me not being heavily over-weight).

 

[Art Of Flowers]

But what happens to people where low carb and all type2 drugs don't work or intolerant?
I'm such person. Low carb helps but not solves high bgs totally. I am highly insulin resistant. Only insulin stops me rotting away. I cannot do heavy exercise, walking is in pain. Many others have other health problems too.

If you watch the video from Jason a Fung on The Two Big Lies of Type 2 Diabetes he describes how he treated many people with long term type 2 diabetes who were being treated with insulin and who were dramatically overweight. He was successful in about 90% of cases using a low carb regime and intermittent fasting. One of his patients was about to have bariatric surgery, but managed to achieve the desired weight loss using fasting instead. Typically it took about 10 weeks to bring down the level of medication and in many cases he managed to get his patients off all medication including insulin.

There do appear to be two groups of people with type 2. One has too much insulin in their system due to insulin resistance from fatty liver and being overweight. The other has not enough insulin due to poor performing beta cells. The first group with too much insulin can get huge benefits from a low carb intermittent fasting regime. The fat blocking the liver and pancreas can be burned off with a ketogenic diet and fasting resulting in less insulin resistance and better beta cell function in the pancreas.

The other group who have too little insulin need to take insulin to augment the small amount their bodies can produce.

Treating those type 2 people who already have high insulin levels with more insulin or drugs which promote insulin production may increase their insulin resistance and the excess insulin in their system may lead to heart disease, strokes, cancer and Alzheimer’s.

 

[Oldvatr]

@ickihun You are being very sensible choosing a BG level based on how it affects your body.
This study stops doctor's thinking that a lower AC1 is always better when these drugs in a use.

My GP wants me to average a higher bgl level (7.0 mmol/l) than I had shown I was capable of achieving (5,4). I am currenntly running at 6.6 mmol/l. It is true that in my case, I was having frequent excursions in the hypoland territory when I was trying to go lower, and now I rarely get such problems. Feels like a good compromise that I am happy with.

 

[LittleGreyCat]

That's just it a little higher causes me problems with numb feet and constant thrush. Exercise cannot help. I also have atherosclerosis which can get worse with uncontrolled diabetes. Then there's my hormone imbalances PCOS which is great on lower bgs. Weight loss too.
You cannot tell me that I don't benefit from not getting these things in good control and far less worry. Too much anxiety and mood swings aren't good for lasting relationships and career prospects.
The whole picture for me is what's important.
So
Weight loss til on as little insulin has been my on-going plan for a few yrs now. Progress is being made....which will continue.
I won't be panicking with this finding but I must express the whole health picture should be considered not snippets, like this seems to cover.
I won't be risking high bgs after these findings, no one should be expected too either.

You are convincingly demonstrating the difference between an average result for a large group of patients and a specific result for one special individual.

I think the lesson to be taken from this study is that all other things being equal it is better to maintain slightly higher average HbA1c to improve all cause mortality risk.

There will always be instances where an individual has special circumstances and needs a different set of targets.

As long as HCPs recognise this then all is well.

If they blindly follow the average (as was the case with "healthy carbohydrates" a while back) then it is time to ignore the mainstream advice.

This study may be a temptation towards the dark side. My last HbA1c was 6.2 up from 5.8 and I'm currently "off treats" to try and get it down below 6 again. It looks as though I could moderate my dietary restrictions and still stay within the "moderate" range.

I haven't picked up from the report what significant difference there is in all cause mortality for Metformin users who are low and medium. Just that high isn't good.

 

[Oldvatr]

There were none in common usage at the time the data was collected. I think they only included people who were on the same drugs for the complete time. Hence SGLT2 inhibitors were not included in the study.



Firstly that is not what they found, they found that on insulin et. al. that high HbA1c and low HbA1c both reduce your life expectancy, with mid-range being best.

I think we already know the answer to this, that high insulin levels in our bodies are not good, and all the drugs in the study apart from Metformin work by increasing Insulin levels. Very high BG is worse than increased insulin levels, but mid-range BG is a good tradeoff in exchange for a lower blood insulin level.

As the study was just looking at the data, the only included in the paper what they could see from the data they were working with. It is assumed that everyone will be reading all papers about Type2. After all, this is a research paper not a chapter from a textbook.

Extract from the full report:
<<<<
Two treatment cohorts comprised of glucose-lowering therapies with a low risk of hypoglycaemia: metformin monotherapy; and a composite group comprised of metformin monotherapy and acarbose, DPP-4 inhibitors, GLP-1 receptor agonists, SGLT-2 inhibitors or thiazolidinediones as monotherapy or in combination with metformin.Two treatment cohorts comprised of glucose-lowering therapies with a low risk of hypoglycaemia: metformin monotherapy; and a composite group comprised of
metformin monotherapy and acarbose, DPP-4 inhibitors, GLP-1 receptor agonists, SGLT-2 inhibitors or thiazolidinediones as monotherapy or in combination with metformin.Two treatment cohorts comprised of glucose-lowering therapies with a low
risk of hypoglycaemia: metformin monotherapy; and a composite group comprised of metformin monotherapy and acarbose, DPP-4 inhibitors, GLP-1 receptor agonists, SGLT-2 inhibitors or thiazolidinediones as monotherapy or in combination with metformin.
>>>>

So SGLT-2 was included in the study in the Low Risk group. The study had a follow up in 2015 I believe and includes that data in the conclusions. again extracted:
<<<<
Patients with type 2 diabetes were identified between January 2004 and December 2013, with follow-up to January 2015
.>>>

 

[Oldvatr]

@Oldvatr you can link to the full article from the abstract, or via the following link:
http://onlinelibrary.wiley.com/doi/10.1111/dom.13155/pdf

Yes, I would agree with you! Unfortunately, most studies are funded by the pharmaceutical industry, so most of the research looks at outcomes associated with medications, not outcomes associated with dietary interventions. For information on the sponsor of this study, see the acknowledgements on page 20 of the article.

Yes page 20 is interesting. One pharma company supplied both the funding, and most of the co-authors, so this raises the usual query about independence and conflicts of interests. This is a problem where an external interest supplies not just funding, but also employees involved in the conduction of the study and/or the writing of the report or conclusions. I did not see any list of peer reviewers either.

 

[ickihun]

My GP wants me to average a higher bgl level (7.0 mmol/l) than I had shown I was capable of achieving (5,4). I am currenntly running at 6.6 mmol/l. It is true that in my case, I was having frequent excursions in the hypoland territory when I was trying to go lower, and now I rarely get such problems. Feels like a good compromise that I am happy with.

I remember telling my older gp that endocrologist wanted me inbetween 51-xx and he was surprised since I'm on insulin. GP thought it should be lower. Seeing dn at hospital this week. Last hba1c was 48 but this next one will be higher. I think back to my common 53-56 result.
I'm not as obsessive testing and eating. Will do that tomorrow but hopefully she will get enough info from my less meter readings.

 

[Oldvatr]

One problem I have with an all mortality study is that this study uses data from the hospital, and will attempt to prove a link to mortalities where the cause of death probably has no link at all to blood sugar or any of the drugs used. I have seen a similar study that did isolate out different morbidities so that it could be shown which flavour of death is affected, such as CVE or stroke or cancer, which may indeed be linked to diabetes, and ignored RTA deaths, suicides and accidental deaths following injury where there is less obvious connection to diabetes, As said elsewhere in this thread it is not clear if any deaths were due to DKA, or hypo, or diabetic coma or infection such as MRSA for example.

Just because a body in the morgue belonged to a diabetic on certain medication does not mean that their death is a direct result of their treatment regime unless there is methodology employed to seperate to the cause of death too. There is an Australian study that does the reverse of the Cardiff study, and starts at the morgue and COD then traces back (in the AU case to LDL levels) to possible causes,

 

[ringi]

As they are comparing people with different AC1 and on different diabetes drugs, "RTA deaths, suicides and accidental deaths" should have been the same in all the groups of people. Hence the differences in death rate between the groups give us useful information.

 

[ringi]

My GP wants me to average a higher bgl level (7.0 mmol/l) than I had shown I was capable of achieving (5,4). I am currenntly running at 6.6 mmol/l. It is true that in my case, I was having frequent excursions in the hypoland territory when I was trying to go lower, and now I rarely get such problems. Feels like a good compromise that I am happy with.

This is sensible, provided the higher AC1 is being got by reducing drug/insulin usage. The aim should always be to reduce BG with diet and lifestyle, so as to force a reduction in drug doses.

 

[Oldvatr]

As they are comparing people with different AC1 and on different diabetes drugs, "RTA deaths, suicides and accidental deaths" should have been the same in all the groups of people. Hence the differences in death rate between the groups give us useful information.

Not necessarily. Someone in a hypo state could be more at risk of losing concentration or losing control, so it could contribute a skew factor.Again, are diabetics necessarily more prone to depression that could lead to suicide? The report does not consider or even identify these endgames so we are blaming the meds soley, and not the co-morbidities that may exist. The other study I saw did, and seperated these effects out as not being relevant to their analysis. My heart medications also have co-morbidity that could lead to a fatal endpoint, in their own right, and again, would my death be put down to a heart medication issue, or a diabetes treatment regime issue? Or would I just be put down?

The findings of this report do seem to corroborate the other report I saw, so there is mileage in it, but I am not 100% convinced by their conclusions wrt treatment regime. It seems to be a little simplistic in methodology, and hence I detect possible weaknesses in the conclusions.

 

[cz_dave]

Ok, this is related to Type 2 but still kind of interesting. What does everyone think?

http://www.diabetes.co.uk/news/2017...ion-linked-to-higher-death-risk-96381539.html