The Cause of Hyperglycemia in Type 2 Diabetes

Sixteen Post Bumps for 2016 ...  No. 1

Post Summary

  • The main role of insulin in the body is to inhibit processes that release energy substrates into circulation on an otherwise "always on" continual basis. 
  • Insulin plays a stimulatory role in the clearance of glucose from circulation, but this is secondary to its inhibitory role, and may not even be physiologically important. 
  • The generally accepted progression of "insulin resistance" -- beginning with impaired uptake of glucose in response to insulin stimulation, leading to a "backing up" of glucose into circulation, and requiring more and more insulin to clear the glucose -- is almost certainly incorrect.
  • Radiolabel studies have demonstrated that in both T1 and T2 diabetics, glucose uptake is not impaired, and indeed may even be greater than normal.
  • Hyperglycemia is due to relative insulin deficiency even if overall insulin levels are high.  This deficiency is due to failure of the pancreas to secrete insulin in both sufficient quantities and in a timely manner.
  • Endogenous glucose production is supposed to be suppressed when dietary (exogenous) glucose is added.  The failure to suppress this is responsible for the hyperglycemia.
  • Rather than hyperinsulinemia, the critical "event" is impaired or an almost complete lack of early glucose stimulated insulin secretion (GSIS).  This "defect" is accompanied by a lack of suppression of glucagon secretion.
  • Subsequent hyperinsulinemia in those capable of secreting substantial amounts (early T2s) is due to endogenously produced hyperglycemia.

Originally Published 10/3/2013

I will forever be grateful to (former?) blogger LynMarie Daye of Adipo Insights blog for bringing the following paper to my attention:

If you are a regular reader and you haven't read it yet, what are you waiting for?  If you're a new reader it's a must (unless of course you're just here for the gossip you claim is boring and inappropriate - grin).   The article is quite long, but it is written in very understandable fashion.  A lot of info, so perhaps best digested in small segments, but well worth the effort. 

The article begins with a history lesson about a physiology professor named Sir Edward Schafer, who "appears to have named insulin and described its actions" in a book published in 1916 entitled The Endocrine Organs.  Schafer, described a substance with dual and simultaneous functions: 
  • Autacoid:  excitatory or stimulatory, e.g. glucose transport, lipid synthesis 
  • Chalone:  inhibitory, e.g. inhibiting lipolysis, gluconeogenesis, ketogenesis
The authors go on to describe how scientists erroneously extrapolated some in vitro rodent data in the 1950's, leading to what they describe as a 20 year "Black Age" in endocrinology ... one in which insulin's autacoid functions of glucose transport were presumed to be of primary importance.  By the 1970's, they say, it was already known that this was not the case, and indeed these functions may not even be of significant physiological importance after all.    However this paradigm made its way into textbooks that have not been updated to this day.   
{Aside 1/1/16:  Since first encountering this paper and writing this post, I have added to my collection and perused several more physiology and biochemistry texts.  Sadly, this perpetuation of the "Black Age" paradigm holds true in recent editions of some major texts used in medical and nutrition programs.  This is something I'd like to see change.  Textbooks will always be behind in terms of new discoveries and the "current consensus", but this is not the issue here.  It is simply not true that, for example, diabetics can't utilize glucose due to a lack of insulin to get the glucose into the cells.}  
The bottom line:  It is the "chalonic" or inhibitory functions of insulin, primarily on lipolysis in fat tissue and hepatic glucose production (gluconeogenesis and glycogenolysis), that are the major physiologically important functions of insulin.

This paper, however, focused on Type 1 diabetes.  So although there are parallels, it is easy to dismiss or at least question some of the findings/assertions and how they relate to the far more common Type 2 diabetes.  Still, it boils down to this:  Hyperglycemia.  What causes it?  Circulating Glucose levels (at all times) are ultimately impacted by three factors:
  • Exogenous (dietary) glucose
  • Endogenous glucose production (EGP)
  • Glucose uptake and utilization
Any combination of increases in glucose added and/or decrease in glucose cleared from circulation would increase the circulating blood glucose levels.   This paper uneqivocally places the "blame" for hyperglycemia -- both in the fasted state and the fed state -- on abnormally high endogenous glucose production in Type 1 Diabetics.    Two summary quotes of note, first on ... :  
"... the (fallacious) concept of insulin being ‘required’ for glucose entry into cells rather than just accelerating glucose uptake. The hyperglycaemia of diabetes was interpreted as a ‘damming back’ of glucose in the blood stream as a consequence of a lack of insulin.   This became established teaching and, although the concept was shown to be erroneous in the mid‐1970s, the teaching has not changed. Consequently, therapy has been based on a flawed concept."
"It is now well established that what Schafer called insulin’s ‘chalonic’ (or inhibitory) actions are the physiologically more important. Indeed, its autacoid (excitatory) action has recently been shown to be, on the whole, physiologically unimportant. " {all emphasis mine}
While discussed in the article, most of the studies seem to relate to Type 1's and extrapolated to both types.  Have there been more studies in Type 2's?  Since they tend to have higher fasting insulin levels than non-diabetics, and it would appear that glucose clearance is impaired, does EGP play the same role, or is it more of a minor player?    As it turns out, it's the same.   Thanks to the LC tag team of Fred Hahn and Richard Feinman, I stumbled upon the following study:

Pathogenesis of Fasting and Postprandial Hyperglycemia in Type 2 Diabetes: Implications for Therapy  

First, below left, we have the glucose and insulin profiles for 5 hrs following a 50g OGTT in this study.   Here we see the altered insulin secretion already in the "mild" diabetes.  This is similar to that seen in Yokioshi below right (from this blog post).  I've shown glucose and insulin profiles (left extend 2 hrs longer) for both studies.  

The "mild" diabetics in "A" (fasting glucose ∼7.2 mmol/l  [~130 mg/dL] before glucose ingestion and increased to ∼11 mmol/l [~200 mg/dL] after glucose ingestion) correspond most closely to the IGT and group I Type 2's in "B" (IGT:  FBG under 7.8 mmol/l [~140 mg/dL] , 2 hr OGTT between 8.9 and 13.3 mmol/l [~140-240 mg/dL], Group I: FBG < 7.8 mmol/l [~140 mg/dL]), while the "severe" diabetics (fasting glucose ∼10 mmol/l [180  mg/dL], and experience marked and prolonged hyperglycemia after eating)  correspond to the Group II Type 2's  (FBG 7.8-11.0 mmol/l [~140-198 mg/dL]) in "B".
FIG. 1.
As diabetes progresses, the insulin response to a carbohydrate challenge is first delayed but similar in magnitude (even a bit higher), and later significantly diminished over all.   At right I've repeated the glucose and insulin plots and included the glucagon plot.  The suppression of glucagon in non-diabetics -- that occurs after the insulin peaks normally early on -- is absent already even in the "mild" diabetics.  

So, now a study is discussed where the OGTT glucose was labeled with radioactivity.  Here are those plots.   Glucose appearance is measured in mg/kg/min, which is why although the diabetics have higher levels, the plots don't look like those in the above plots. 

Top:  The diabetics have higher glucose appearance both before and after the glucose challenge.  

Middle:  Tracks just the radio-labeled exogenous glucose.  This plot show that the dietary glucose is absorbed and cleared in similar fashion in diabetics vs. non-diabetics.

Bottom:  Total glucose - exogenous glucose = edogenous glucose.   Here the difference between diabetics and non-diabetics is most clear.  To quote the article:
Endogenous glucose production was higher in the diabetic subjects than in the nondiabetic subjects before eating. In addition, whereas endogenous glucose production rapidly decreased in the nondiabetic subjects after glucose ingestion, suppression of endogenous glucose production was slower in the people with diabetes and required approximately 6 h for a nadir to be reached.  The excess amount of glucose released due to slower suppression of endogenous glucose production entirely accounted for the higher postprandial rates of glucose appearance
But what about "insulin resistance"?  I put that in quotations, because for quite some time now, what is commonly referred to or implied as insulin resistance -- some "backup" in circulation of ingested glucose -- does NOT appear to occur in the glucose intolerant or diabetic state.    
What about glucose disappearance? Was hyperglycemia due to lower rates in the diabetic subjects? As is evident from this side, the answer is no. As is evident from the upper panel of Fig. 4, if anything the rates of glucose disappearance were higher in the diabetic subjects than in the nondiabetic subjects both before and after glucose ingestion.

So ...

The trail that led me to this paper was again the mega Facebook discussion on Fred Hahn's page, that he followed up with a post (from a year ago, the image of which showed on the Facebook preview) on the Pinterest page of "real expert" Richard Feinman.  As we had been discussing the primary defect in diabetes and the cause of hyperglycemia, I found the plots Feinman highlighted to be most interesting.   Feinman had also written a recent post comparing side effects of LC diets vs. thiazolidinediones.  All of this on the background of a discussion on what diabetes really is, and whether or not a low carb diet can "cure" diabetes or if managing hyperglycemia with LC was the best and only way.

I can only repeat here again, that diabetes is a disease of relative insulin deficiency:  a combination of pancreatic beta-cell dysfunction and hepatic insulin resistance (there is a lot more about that in the Rizza paper, hopefully I can address that in a blog post at some point).   Pitting LC vs. an infrequently used class of drugs (due to safety issues and other options) is quite the strawman.  What is telling is what medication changes were made in response to severe carbohydrate restriction in two studies conducted by Eric Westman (16 wks , 6 months).  You would think that after these time frames on roughly 20g/day carbs, there would be a greater reduction in medications.  Generally those that come off meds were on metformin only at baseline -- with weight loss comes improvements in hepatic insulin signaling, and these would have been the milder cases to begin with.  Those taking insulin and metformin, only reduce insulin (which makes sense as without dietary carb you wouldn't want to suppress EGP quite so much!) but generally remain on metformin.  And the thiazo's?  A mixed bag at best and not much of a sample size to draw any sort of conclusion -- especially not that VLC is some sort of alternative.

Is is a good idea to compound untreated frank diabetes with a low fat, high carbohydrate diet?  Well, if that diet is not in sufficient caloric deficit to result in weight loss, no, it's not a good idea.  Caveat:  What we in developed Western nations consider "low fat" is simply not.  So any time "high carbohydrate" is discussed, it cannot be without attention to how much fat, on an absolute level, the person is consuming.   Given the documented (not just anecdotes) rapid reversals of diabetes with a few approaches, I do not see that VLC as the default approach is warranted.  It seems to me that the subjects in Westman's studies are likely in that place of being unable to tolerate such things as oat bran and legumes that Dr. "Wheat Belly" Davis formerly worried might cause hypoglycemia in those on diabetes meds!  

Food for thought:
If dietary carbohydrate is not causing the hyperglycemia, then wouldn't it be preferable to address the underlying metabolic problems -- relative insulin deficiency and hepatic insulin resistance -- that are? 
VLC seems to be a way of coping with a suboptimal metabolism rather than restoring it, if possible through non-pharmaceutical intervention.  If that is not possible, here's the question:  Is the underlying hormonal imbalance unhealthful if you don't "need" the insulin to process your dietary carbs?   Given all of insulin's other functions, I don't see how it could not be ... many animal studies seem to bear that out, particularly with the more extreme ketogenic diets.  But ultimately I don't know.  But here's the kicker:  Those people advocating LC so vehemently don't know either.  They just don't.  Yet they attach "healthy" to every aspect of their lifestyle and brag on health markers that are not known to independently improve risk of CVD or contribute to lower mortality rates (or at least not increased rates), and may even increase mortality.  All the while they cling desperately to an infinitesimally small population (in the overall scheme of things) of arctic dwelling cold-water mammal consuming people that aren't even in ketosis and who likely eat less saturated fat than your typical SAD American let alone a New Age NuttyKer, as proof that their diet is best and healthiest and ... well, did I mention best ... oh and only diet for diabetics, etc.

Finally, given the profiles shown in this article, the issues some have with protein should become a little more obvious ... hint:  it's not excess amino acids needing to be turned into glucose that are the problem.  But the "obvious" solution to the low carbers, is to limit protein as well now.  Wise?  Perhaps as a last resort if you are so averse to "medications" as to deny yourself insulin replacement therapy as an option.  But at some point you're going to be cutting into lean mass.


Kemby said…
"VLC seems to be a way of coping with a suboptimal metabolism rather than restoring it" Absolutely spot on. Just because something makes your symptoms go away for a while doesn't mean the underlying issue won't become worse months or even years down the line.
eulerandothers said…
'Is is a good idea to compound untreated frank diabetes with a high carbohydrate diet? Obviously not. '

Untreated frank diabetes doesn't seem to be the problem bringing people to low-carb boards. It's frank obesity, or the threat of it! I'm not sure that a high carbohydrate diet that is calorie restricted wouldn't be beneficial for a diabetic.
Nigel Kinbrum said…
Most* people with T2DM don't only have hyperglycaemia. They also have hypercholesterolaemia & hypertriglyceridaemia (& hyperuricaemia?) due to full adipocytes leaking too much glycerol & NEFAs into the blood. As diabetics only have BG meters, hyperglycaemia is the only parameter that they can monitor, which is why VLC diets look great for T2DM.

Fred Hahn was arguing about skinny people with T2DM on Facebook, but I didn't want to comment as there were too many conversations there. Fatness level at the onset of T2DM is merely a function of how many adipocytes someone has at the point where they are all full*. Someone with virtually zero adipocytes (i.e. someone with lipodystrophy) becomes T2DM while skinny (see ). ~85% of people increase the number of adipocytes as existing adipocytes become full, so ~85% of people get fat before they become T2DM. Thiazolidinediones increase the number of adipocytes.

*Except for those with LADA or T2DM caused by genetically broken insulin signalling.
George said…
How much of the excess endogenous glucose comes from bona fide amino acid and glycerol gluconeogenesis, and how much is simply being released from glycogen stores? The reduction of glycogen storage by VLC diets to a more manageable size may be key to VLC success here.
George said…
Diabetes has been "cured" with low-calorie diets of pure carbohydrate. The problem is that low calorie diets are hard to sustain as a permanent way of life, especially when they are high in carbs.
George said…
But if you don't make your "symptoms" go away you're likely to go blind and lose your legs.
Can it really get worse than that?
Some perspective please, we are talking about real damage and disability, vs. some hypothetical risks that speculative blogging can cook up.
It's like people who won't wear a seatbelt in case something goes wrong with the mechanism and it strangles them for some reason.
Paleo Nouveau said…
Very well said!
eulerandothers said…
'But if you don't make your "symptoms" go away you're likely to go blind and lose your legs.'

No. I think if your 'symptoms' go away and you go on to lose your legs or go blind, that is worse. You thought that by 'curing' the symptoms, you 'cured' the disease. Which you did not.
carbsane said…
Keep dreaming George.
George said…
Wot no stats? Hepatic glycogen is surely pool of first resort when defective hormonal processes seek to elevate blood glucose between meals. The smaller the pool, the more tightly regulated it will be. I raised this question with Mr Jaminet some while ago and he was able to reference animal studies that actually addressed the question.
If you are a high-carb feeder you have bigger glycogen stores that a VLC-er (water weight, anyone?) - I think I read that somewhere - so there is surely less urgency to conserve them by resort to de novo gluconeogenesis.
carbsane said…
I'm not following your point regarding the "key to VLC success". I'll have a post up on protein and gluconeogenesis today or tomorrow. I was going to have steak for breakfast, but might opt for chocolate cake cuz it's the same thing ;-) <-- joke, sarcasm.
carbsane said…
Very low fat diets are hard to stick to. Very low carb diets are hard to stick to, and so many of those end up resigned to very low calories after all.
carbsane said…
Nice links!
carbsane said…
How about making your symptoms go away with insulin or with a diet that restores full function where that is possible?
carbsane said…
EXACTLY! At least with neuropathy, it seems that insulin deficiency and/or resistance is the likely culprit rather than the hyperglycemia.
carbsane said…
Also, there are more symptoms than hyperglycemia. How are the NEFA? How is adipocyte function? Runaway LDL probably means the rarely mentioned NEFA are quite high.
prerna said…

There are safer and more effective alternatives to drugs to treat Type 2 Diabetes. One, that
can also be used with drugs such as Metformin, is Resveratrol. In two
recently published human clinical trials done by respected medical schools it
was shown that transmax resveratrol in the first study, and bioforte
resveratrol in the second one, had the effect of reducing blood glucose,
improving insulin sensitivity, lowering blood pressure and LDL cholesterol, and
even lowering body weight. The scientist who did the transmax study
stated, "Resveratrol can be an effective adjunct therapy for type 2
Diabetics currently using one of the Metformin like drugs, or who are
controlling their Diabetes without drugs."