How Fatty Diets Cause Diabetes

How Fatty Diets Cause Diabetes

This was the provocative title of the Science Daily piece that made a minor splash across the LC web recently.  Perhaps overshadowed by the AHS drama and fallout, but nonetheless picked up on by a number of people.  The article begins with:
Newly diagnosed type 2 diabetics tend to have one thing in common: obesity. Exactly how diet and obesity trigger diabetes has long been the subject of intense scientific research. A new study led by Jamey D. Marth, Ph.D., director of the Center for Nanomedicine, a collaboration between the University of California, Santa Barbara and Sanford-Burnham Medical Research Institute (Sanford-Burnham), has revealed a pathway that links high-fat diets to a sequence of molecular events responsible for the onset and severity of diabetes.
The article is referring to the journal article linked below:
Pathway to diabetes through attenuation of pancreatic beta cell glycosylation and glucose transport (or request access)

A connection between diet, obesity and diabetes exists in multiple species and is the basis of an escalating human health problem. The factors responsible provoke both insulin resistance and pancreatic beta cell dysfunction but remain to be fully identified. We report a combination of molecular events in human and mouse pancreatic beta cells, induced by elevated levels of free fatty acids or by administration of a high-fat diet with associated obesity, that comprise a pathogenic pathway to diabetes. Elevated concentrations of free fatty acids caused nuclear exclusion and reduced expression of the transcription factors FOXA2 and HNF1A in beta cells. This resulted in a deficit of GnT-4a glycosyltransferase expression in beta cells that produced signs of metabolic disease, including hyperglycemia, impaired glucose tolerance, hyperinsulinemia, hepatic steatosis and diminished insulin action in muscle and adipose tissues. Protection from disease was conferred by enforced beta cell–specific GnT-4a protein glycosylation and involved the maintenance of glucose transporter expression and the preservation of glucose transport. We observed that this pathogenic process was active in human islet cells obtained from donors with type 2 diabetes; thus, illuminating a pathway to disease implicated in the diet- and obesity-associated component of type 2 diabetes mellitus.
In a nutshell, the elevated NEFA they observed in an obese mouse model led to a disruption in glucose transport in beta cells, and this same defect is seen in the islet cells donated by humans with type 2 diabetes.  The mechanism seems to be elevated NEFA →  reduced expression of certain genes → impaired glucose transport.  I'm a bit backlogged with other interests to look into this very detailed paper as regards all the exact details here, but I think the following excerpts are sufficient to get the gist.
... Pancreatic beta cell dysfunction is also a diagnostic determinant of type 2 diabetes and includes defective insulin secretion exemplified by the loss of glucose-stimulated insulin secretion (GSIS). Normally, beta cells sense elevations of blood glucose by expressing cell surface glucose transporters that enable concentration-dependent glucose transport across the plasma membrane. This is followed by glucokinase action and ultimately calcium-dependent insulin secretion. Loss of GSIS has been linked to impaired beta cell glucose transporter (Glut) expression among animal models, and failure of GSIS has been proposed to further provoke the pathogenic onset of type 2 diabetes in humans. ...
... Genetics, diet and obesity contribute to the current epidemic of human type 2 diabetes. A substantial proportion of the human population seems predisposed to a combination of these factors, perhaps analogous to different mouse strains that are resistant or susceptible to diet- and obesity-induced diabetes31. Our findings show that the extent of disruption of beta cell glycosylation and glucose transport by diet and obesity directly contributes to disease onset and severity in susceptible mice, illuminating a pathogenic pathway that encompasses lipotoxicity and glucotoxicity32. This pathway seems to be conserved in normal human islet cells and is activated in islets from donors with type 2 diabetes.  A pathogenic tipping point in this pathway may occur when elevated free fatty acid (FFA) concentrations impair the expression and function of FOXA2 and HNF1A transcription factors sufficiently in beta cells to deplete GnT-4a glycosylation and glucose transporter expression. The resulting dysfunction of beta cells leads to impaired glucose tolerance and failure of GSIS and further contributes to hyperglycemia, hepatic steatosis and systemic insulin resistance. Preservation of beta cell GnT-4a glycosylation and glucose transporter expression breaks this pathogenic cycle and its link to diet and obesity. 
This paper is actually yet another one providing evidence in support of the Fatty Acid Hypothesis for how obesity can cause insulin resistance and diabetes.   It does not go into how fat cells become IR first, but rather deals with the mechanism by which the resulting elevation in circulating NEFA leads to pancreatic beta cell dysfunction.  That being interfering with glucose transport to the beta cell so your pancreas does not properly sense hyperglycemia.

In my opinion, the title and implications of the Science Daily article are irresponsible, but so too are the knee-jerk responses of many of those favoring low carb high fat diets.  It should always be remembered that NEFA is not the equivalent of dietary fat.  A high fat diet does not necessarily impact NEFA.  However, two articles by Keith Frayn I've discussed here previously, Adipose Tissue as Buffer ... and Fatty Acid Trafficking, present the scenario whereby dietary fatty acids can contribute significantly to NEFA levels.  Additionally, VLC diets fail to suppress NEFA release in the postprandial period.  

Several are dismissing this whole thing out of hand.  The usual "well the high fat diet also has carbs" argument abounds.  Perhaps a better title for the article might be How Fattening Diets Cause Diabetes.  Clearly some are more genetically predisposed, or all obese would be diabetic which is not the case.  But perhaps instead of T2 or NIDDM or "adult onset" diabetes, we can use Diabesity or OIDM (obesity induced diabetes mellitus) as a designation.   If your high fat/carb/calorie junky SAD → obesity → elevated NEFA and you are genetically predisposed, it will lead to hyperglycemia.  The solution?  Stop eating carbs to control the hyperglycemia?  Yes, that works (for some, to some degree), but it doesn't appear to restore beta cell function without substantial weight loss.  And of all the anecdotal reports of those doing longterm VLC are worth considering it seems the reports of becoming more and more carb intolerant indicate persisting beta cell dysfunction.  Considering all of the other important functions of insulin in the body, it may be worthwhile looking at another approach long term.  JMO.


Jeff Consiglio said…
Of course most main-stream health professionals say that being overweight causes insulin-resistance/diabetes, whereas low-carbers generally reverse the causality by stating that insulin resistance (Caused by those "evil" carbs) is THE cause of obesity.

I lean toward the view that these two views are not necessarily mutually exclusive or incompatible with each other.

I agree that excess fat (Especially visceral fat)pumps out nasty inflammatory thinga-ma-jingies that can lead to insulin resistance.

But we also see that Southeast Asians seem to become diabetic while relatively LEAN.

So that indicates to me that being overweight is only one potential pathway to insulin resistance/diabetes - since some lean folks do in fact develop insulin resistance.
Sue said…
Possibly some of the lean diabetics are skinny-fat - have a lot of adipose fat in relation to lean mass. Those obese that don't get type 2 just extra good and storing the fat away protectively so don't get elevated NEFA. At a certain fatness, an individual thing, the storing of fat protectively starts to decline.
Anonymous said…
This is one mechanism by which lipotoxicity promotes MetSyn/diabetes. Combine it with glucotoxicity (via ChREBP) and you are all set.
Sue said…
Robb Wolf's take on the study:
CarbSane said…
Hi Jeff: Obesity → IR or IR → Obesity?
You say these are not mutually exclusive and I would agree to an extent. I believe that in the vast majority of cases when obesity is associated IR, however, a certain degree of obesity occurs first, then a significant level of IR/compensation exacerbates the situation leading to more weight gain. When NEFA are elevated, both glucose uptake and oxidation are impaired and this is why I believe so many folks report "problems" eating carbs in that state and why LC works quite well -- seemingly better the more obese/IR the person.

Sue, what you say is consistent with Frayn's FAH. Indeed the FAH seems consistent with pretty much every observation of non-autoimmune/injury induced beta cell function forms of diabetes. It appears Robb didn't read the full text? I'm going to post a link and a comment there. Let's see how that goes. :-)

@Lucas: I'll have a look!
Jeff Consiglio said…
I see insulin resistance as a multi-factorial problem with often numerous and overlapping causes.

- I certainly agree that NEFA plays a part.

- The "ionic theory" of insulin resistance/metabolic-syndrome tells that something as "simple" as a mere magnesium or potassium deficiency can have PROFOUND effects upon one's ability to handle glucose.

- Lack of Chromium (Sometimes referred to as "glucose tolerance factor") can also be part of the downward spiral in insulin sensitivity. Also lack of vanadium and B vitamins.

- Excess iron (More common in men and post-menopausal females) can also cause hepatic insulin resistance. But don't tell that to all those LC dieters awash in a sea of iron rich red-meats.

- One of the biggest causes of insulin resistance in skeletal muscle tissue is intra-muscular triglycerides

- Not exercising muscles also causes them to be less receptive to insulin's effects.

- Lack of D3 (VERY Common these days, particularly in darker skinned folks.)

- Sleep deprivation has a MAJOR effect upon insulin sensitivity for the worse

- "They" used to say that caffeine reduced the risk of getting diabetes, but more recent evidence strongly suggests that it actually makes one LESS insulin sensitive.

- Mental stress

- Etc. etc.

Kind of a "witch's brew" of factors all conspiring to mess up people's insulin sensitivity.
CarbSane said…
Hi Jeff, I think another problem is that "diabetes" is such a heterogenous disease. I wrote about this here:

Perhaps we should call the 80% of those diagnosed with hyperglycemia that are obese as diabesitetics. The overwhelming factor that led to their disease is obesity and the overwhelming evidence that the initiating factor in how obesity -> IR and hyperglycemia is NEFA.

I've got a post in the works regarding some of those other factors, cause/effect, etc.
Sanjeev said…
> insulin resistance as a multi-factorial problem

From all the examples you give it looks like insulin resistance is not the problem but part of the body's solution
Sanjeev said…
This comment has been removed by the author.
Sanjeev said…
point being, if the underlying cause of the IR STAYS THERE but you could magic away the IR (and only the IR) would the person be better off?

If IR is a solution I doubt it ... an obese person with good insulin sensitivity would not be able to lose fat ... in this case Taubes would be right about insulin blocking fat loss.
Sue said…
Evelyn, another write up about the study from Denise Minger at Mark's Daily Apple:
For your take on it. I'm not sure.
Sanjeev said…
> Kind of a "witch's brew" of factors all
> conspiring to mess up people's
> insulin sensitivity.

IMHO Peoples' insulin sensitivity seems to be what it needs to be for optimal health UNDER THE GIVEN CONSTRAINTs (the things you list).

It's like the body's a math student. Start by writing down the givens, do what you need to do with tools you have to get the required (optimal health).

Another way to look at it: Once we can measure insulin level and resistance easily and routinely at the moment should we develop drugs to get rid of the IR?
fredd said…
Hi CS,

I was just wondering, and I apologise if you have covered this... in general, what are the differences between obese/T1D/T2D/healthy individuals in regards to insulin response to glucose intake? e.g. if one of each of the above "categories" of folks were to eat 25g glucose, how would their insulin responses differ in terms of magnitude, length and effectiveness at reducing blood glucose? Perhaps this is basic stuff, but you seem to have an amazing knowledge of it and a great ability to explain complex things simply, so I thought you might be the best person to ask!

Thanks and love your blog!
CarbSane said…
Hi Fredd, first thanks for the compliments! I've actually got a longer post in the hopper on this but here's a short version:

T1 (and any form of acute onset diabetes resulting from disease/injury/autoimmune attack): no pp insulin response or basal insulin production, rampant hyperinsulinemia pp and fasting -> insulin therapy absolute necessity!

Normal insulin sensitive: Appropriate to keep BG in range (this changes in the context of the size and composition of a meal, time since the last meal and its size, etc.)

Mild insulin resistance/compensator: Likely elevated fasting and pp insulin to keep NEFA and BG in normal range. Apparently some folks live their whole lives in this state to no detriment! They test normal.

"Diagnosed IR" with elevated fasting BG ("prediabetic") and/or impaired glucose tolerance (IGT - pp BG spikes and/or takes longer to come down to fasting), more elevated fasting insulin, elevated pp insulin response but can no longer completely compensate. Still keeping things sort of under control ...

As things progress, in the "common T2 diobesabetic*", the pp insulin response diminishes and pp glucose spikes get worse until they reach a certain level so as to be considered frank diabetes. Hepatic IR likely progresses concurrently and FBG goes up as well.

As the disease progresses further and beta cell mass declines, the T2 essentially becomes a T1.

Hope this helps.

*I think I'm going to settle on this term for obesity induced diabetes: diobesabetic
Helen said…
Hi fredd,

I'm not CarbSane, but I think it looks a little like this:

Obese and not diabetic: insulin-resistant but produces enough insulin to maintain normal glucose: beta cells and insulin production increase to meet demand.

T1D: not insulin-resistant (at least at outset - can become so if weight gain occurs when on exogenous insulin); do not produce insulin on their own. (Caveat: There are some slow-onset autoimmune diabetics for whom insulin production is reduced for a time before complete beta cell failure.)

T2D: insulin-resistant and has a degree of beta-cell failure: beta cells produce insulin, but not enough to meet increased demand. (The degree of impairment varies and tends to deteriorate over time, requiring insulin-stimulating drugs or exogenous insulin.) (Note that T2D is probably not just one disease but many - with the insulin resistance or impaired insulin production resulting from different mutations that weaken various parts of the glucose-sensing-signalling system. (That's a very unscientific way to put it - I'm no scientist.)

Healthy individuals have neither insulin resistance nor an insulin deficit.

There are also more unusual forms of diabetes. Some are mitochondrial and highly heredity - often occurs with partial deafness; others result from monogenic mutations (these have traditionally been termed "MODY" or mature-onset diabetes of the young - a misnomer) affecting genes controlling enzymes and transcription factors controlling various aspects of the glucose/insulin system. Some types of MODY are mild and static, others can be quite severe and deteriorate over time. These more unusual types often get misdiagnosed as Type I or Type II.

There may be more overlap among different types of diabetes than has been thought. There may be an autoimmune component of at least some forms of Type II diabetes; mitochondrial dysfunction is also apparent in Type II diabetes; mutations affecting some of the same pathways are present in some forms of MODY and some Type II diabetics; weight gain might precipitate LADA in some people, a version of Type I (Latent Autoimmune Diabetes in Adulthood).

For more information on Type II, MODY, and LADA forms of diabetes, I highly recommend Jenny Ruhl's web site and blog


But I really was logging on to say this....

Carb tolerance does go down on a low-carb diet, but this is temporary. I haven't heard of its leading to permanent impairment. It's part of the body's attempt to prevent hypoglycemia - one produces less insulin if the body is used to fewer carbs - it takes a few days of higher-carb eating to increase it.

The rationale for diabetics eating a low-carb diet is that their insulin response is already blunted, so you are lowering the carb load to what the body can handle. (Or reducing the amount of exogenous insulin required if insulin-dependent.) In diabetics with significant beta-cell failure, ratcheting up the carbs won't result in compensation from the beta cells - it will just result in higher blood glucose.

There are some diabetics for whom beta cell failure perhaps isn't the primary problem - the problem is either insulin resistance or impaired glucose signalling, which can be further impaired by excess dietary fat. To make matters more confusing, it seems probable that many Type II diabetics have a problem with fat blunting insulin sensitivity or insulin signalling *and* beta cell failure.

I seem to be the type who does better with less fat and more high-fiber carbs. Apparently, there's no one solution for everyone.

I'm looking forward to the day when we can hold up our finger to a scanner and have our genes read, including the ones that are switched on or off, so we don't have to bang around in the dark wondering what treatment is best for treating what is a very heterogeneous disease.
CarbSane said…
I'm looking forward to the day when we can hold up our finger to a scanner and have our genes read, including the ones that are switched on or off, so we don't have to bang around in the dark wondering what treatment is best for treating what is a very heterogeneous disease.

I would like to see this "very heterogenous" disease at LEAST be studied and treated for it's known subclasses by current knowledge. Clearly a large chunk share a similar etiology and progression. But I used to ask folks who were diagnosed what their testing was like and some were diagnosed on a single FBG! Perhaps they are exaggerating and there was a history building up to that or whatever, but back around 2 years ago when I was doing IF diligently I had my FBG go up a bit with no rhyme or reason to well within pre-diabetic range (mid 1-teens) here and there. Most of the time it was in the comfy 80's and I think my highest pp BG (after a big carby Mexican meal) was somewhere in the high 130's, yet had I happened to go to the doc those days ... I'm also shocked to be told that it is apparently not routine for an OGTT to measure not just BG but also insulin. Again, perhaps these folks aren't telling the whole story but how can a doctor treat a "diabetic" without knowing at least the basic cause of their hyperglycemia.

This is what bothers me about metformin -- the golden child diabetes drug -- it's great for certain things but totally ineffective for others. Rather than experimenting on the patient with a bunch of drugs to see what works, why not at least gather all the info possible first?
Anonymous said…
Denise Minger writes in the Daily Apple:
'Unlike metabolically damaged humans, who tend to have rock-bottom HDL cholesterol and rising triglycerides, some mice experience higher HDL and unchanged (or reduced!) triglycerides when eating the diets that make them diseased (PDF). This points to some clear differences between how humans and mice experience diet-induced metabolic problems.'

Just today, I saw this in my e-mail from NCBI:
fredd said…
Thanks CS (and Helen)!

Should the following:

"rampant hyperinsulinemia pp and fasting -> insulin therapy absolute necessity"

Read this?

"rampant HYPOinsulinemia pp and fasting -> insulin therapy absolute necessity!"
Melchior Meijer said…
Hi CarbSane,

Have you seen these 'ancient' observations?

On their traditional very low carbohydrate diet, Inuits had excellent glucose tolerance. After fasting though, they saw the horror curves we see in many modern day low carbers. What the heck is the confounder?
Melchior Meijer said…
And then we have this observation, some 40 years later.


My guess: dairy. Introduce milk and you got glucose intolerance. Both the hospitalized Inuit and modern day low carbers were/are exposed to dairy and both were/are glucose intolerant. Free living (traditional) Inuit and modern day 'paleo's' were/are not. Any other suggestions to explain the observations in these two links?
CarbSane said…
@fredd: Actually I meant rampant hyperGLYCEMIA!
CarbSane said…
ps nice comment Helen!
CarbSane said…
Hi Melchior: Re the first paper, I note that the macro ratios were F/P/C roughly 47/45/8 so not nearly as high fat as most ZC'ers advocate. Further 135g fat on average when one looks at the compositions of seal vs. beef fat, the Eskimo diet is flooded with omega 3's and far lower in sat fat and palmitic acid. That combined with that high protein doesn't surprise me that they would be very glucose tolerant. However upon fasting their bodies are perhaps not used to the elevated NEFA moreso than others? Although the article did state the reduced OGT following prolonged fasting is seen with others.

I'll have a look at the other paper when I get a chance. :)
psychic24 said…
I'm probably wrong, but to me it seems the only reason VLC helps with any degree of improving insulin resistance is based on losing weight. Losing weight increases the insulin sensitivity of the adipocytes, which in turn prevents them from leaking NEFA's when insulin is present. This makes sense that when people refer to the good kind of insulin resistance the coincides with going VLC is actually the peripheral insulin resistance which can be easily reversed as soon as carbs are present and insulin is preventing extra NEfa's to yet again leak out. From all the presented information is seems that the worst possible dietary intervention for people with diabetes(ll) and the like is a mixed diet, as it is only bound to produce hyperglycemia. I guess the real question is whether to pursue a truly high fat or high carb diet to lose weight and reverse the adipocyte IR. During the beginning of each there would be pathological consequences, high fat would yield more lipotoxic diacylglycerides and ceramides and the associated ROS's (as mentioned on one of your previous posts), while the high carb would initially allow pathological amounts of sugar in the blood (hyperglycemia). As soon as weight loss would begin the effects would be ameliorated and soon everything would be dandy, i think... ONly problem I think is that even on a high fat diet, eventually carbs would need to be introduced due to the increased cortisol from gluconeogenesis and the fact that we have a vital organ in our body (pancreas) that is still somewhat damaged. And though it may be contrary to logical reasoning, I think only carbs can only properly revamp beta cell functioning--as long as they are introduced into an environment where the fat cells aren't IR. Sorry for the rambling...but is there any truth to any of this? =)
RRX said…
I have ScienceDaily in my Reader feed and the titles of their articles are always ridiculous and often not even consistent with what the article actually says. Clearly the headlines are written by someone other than the writer of the articles.
CarbSane said…
Hi psychic: What you say sounds about right to me, except I don't really have a problem with VLC in initial phases so long as it's producing weight loss. There are folks who have been VLC for years who lose a little or none at all and then remain significantly obese or overweight.

Melchior's first link above is interesting. The Inuit in that study had exceptional glucose tolerance. They also ate an average of almost 300g protein/day, roughly half their calories. I was just remembering that protein increases IGF-1 levels and would add this may be at play in this population as well:

I don't believe a VLC/VHF diet is optimal in maintenance. Although since lipids are less toxic w/o the hyperglycemia, I suppose eating such a diet will at least delay progression for a bit.
CarbSane said…
Forgot to say, in calorie deficit, I don't see the problems with DAG, ceramide, ROS being an issue. Elevated NEFA that are supplying fatty acids that are quickly oxidized don't seem to be a problem.
psychic24 said…
Though I don't put all my eggs in the epidemiological basket, I think i'm more certain, based on the plethora of evidence, that reducing protein consumption to 20%> is optimal; the link on IGF-1 being increased by increasing protein doesn't convince me--it says that IGF-1 is reduced in diabetics and by innervating a different mechanism (via increased protein) is the way to fix it? sounds a bit fishy. kinda like medication used to treat the symptoms without the underlying cause. the epidemiological citations from that article elicit a fair degree of skepticism on my part, mainly due to the drawing of conclusions based on observing a singular metabolic marker(seems there could be many different reasons for the amount of IGF-1 present). But aside from all that I was hoping to get your earnest opinion, based on scientific evidence or anecdotal experiences (both would be even better =)), what you consider to be optimal: a higher fat diet with moderate carbs, or a higher carb diet with moderate fats.

p.s. I think one of the most interesting aspects, though, when recommending any of these diets is considering the cognitive effects they present on the subject (i.e. lower fat associated with more depression). Any opinions in this area would also be greatly appreciated.
By the way, all your posts are superb and probably most fun to respond to =)