Friday, January 27, 2012

Insulin and Obesity: Another Nail in the Coffin

There are several versions of the insulin hypothesis of obesity, but the versions that are most visible to the public generally state that elevated circulating insulin (whether acute or chronic) increases body fatness.  Some versions invoke insulin's effects on fat tissue, others its effects in the brain.  This idea has been used to explain why low-carbohydrate and low-glycemic-index diets can lead to weight loss (although frankly, glycemic index per se doesn't seem to have much if any impact on body weight in controlled trials). 

I have explained in various posts why this idea does not appear to be correct (1, 2, 3), and why, after extensive research, the insulin hypothesis of obesity lost steam by the late 1980s.  However, I recently came across two experiments that tested the hypothesis as directly as it can be tested-- by chronically increasing circulating insulin in animals and measuring food intake and body weight and/or body fatness.  If the hypothesis is correct, these animals should gain fat, and perhaps eat more as well. 

In the first study, published in 1980 by Dr. Dennis A. Vanderweele and colleagues, rats were implanted with mini-pumps delivering insulin at a steady rate throughout the day and night for 7 days (4).  They tested four different doses: 0, 1, 2 and 6 units per day, and measured food intake and body weight.  This is a model of chronically elevated insulin reminiscent of what is seen in insulin-resistant people. 

They found that all doses of insulin reduced body weight gain, but only the 2 unit dose was statistically significant.  They speculated that the 6 unit dose was less effective at reducing weight gain because it was sufficiently high to cause hypoglycemia (low blood sugar), which is a potent trigger for food intake.  Total food intake was suppressed at all doses by insulin as well, but this was apparently not due to illness.  The authors conclude:
1) insulin limits meal size when blood levels are modestly elevated for prolonged periods of time in the rat, 2) this decrease in meal size is not compensated for by an increase in meal frequency and, hence, total daily food ingestion and body weight gain are reduced, and 3) this effect appears to be a heightening of satiety rather than an induction of illness.
The second study, published in 2002 in Nature Medicine by Dr. Ellen L. Air and colleagues, took a different approach (5).  The authors used a drug called Cpd2, which mimics the effect of insulin on the insulin receptor, but can be taken orally rather than by injection (6).

They placed mice on a fattening diet (HFD) with or without Cpd2 added to the food, then measured food intake, body weight and body fatness over a 6 week period.  Here is what they observed when they measured body fatness:

Consistent with the previous study, Cpd2 opposed weight and fat gain, and suppressed food intake as well, without causing illness.  Glucose tolerance and blood glucose levels were also improved in the Cpd2-treated group.

It is worthwhile to note that there are also studies showing that daily injections of long-acting insulin lead to body fat gain in rodents over time (6a, 6b, 6c).  However, these studies used high doses of insulin that resulted in hypoglycemia, which as I mentioned above is a potent trigger for food intake and causes a major hormonal response (the "counter-regulatory response").  That's the unique thing about the studies discussed above-- they investigated the impact of insulin signaling on body fatness without involving hypoglycemia, a situation that is more akin to what is observed in people with elevated fasting insulin due to insulin resistance. 

The Final Nail

Contrary to what the insulin hypothesis of obesity predicts, chronically elevated insulin if anything seems to oppose weight and fat accumulation in animal models.  This likely involves insulin's action in the brain to constrain body fat accumulation.

The evidence suggests that:
  • Experimentally preventing the increase in circulating insulin that occurs on fattening diets does not alter the course of fat gain in rodents or dogs (7).
  • Experimentally elevating circulating insulin by creating liver insulin resistance does not lead to fat gain in rodents (8).
  • Experimentally increasing circulating insulin by infusing it directly into the blood does not cause fat gain in rodents, but instead makes them leaner (above).
  • Roughly a quarter of obese humans have normal circulating insulin and normal insulin sensitivity ("metabolically healthy" obese) (9).
  • In Pima (Akimel O'odham) native Americans, one of the most obesity-prone populations in the world, and certain other populations, insulin resistance and higher insulin secretion consistently predict less body fat gain over time (10, 11, 12, 12a, 12b). This has not been observed in all populations, but the fact that it occurs in some casts further doubt on the idea that elevated insulin is a central contributor to fat gain.
Researchers have been studying insulin's impact on food intake and body fatness for a long time, and among those who are the most intimately involved in the field, the idea that elevated insulin leads to body fat accumulation seemed rather unlikely by the late 1980s (13).  Although biology is complex, and you can never be totally certain where scientific progress will lead, we have enough evidence at this point to say that this hypothesis is probably not correct.  We now know that food intake is regulated by a complex 'symphony' of signals originating from the brain, fat tissue, the gastrointestinal tract and the pancreas, and insulin seems to play little or no role in this process on a meal-to-meal basis (14).  In the long term, it may constrain food intake and body fat mass due to its actions in the brain, in a manner similar to leptin (although less potently). 

We need a better hypothesis to explain why low-carbohydrate diets cause fat loss in many obese people, and we definitely need a better hypothesis to explain why obesity arises in the first place.  An alternative hypothesis, that insulin opposes fat gain through its action in the brain, is plausible and has received support from a variety of lines of evidence.  Insulin resistance in the brain, leading to reduced insulin signaling in neurons, has been documented in animal models of obesity, and likely contributes to fat accumulation. Therefore, the 'insulin hypothesis of obesity' that makes sense is that a reduced insulin (and particularly leptin) signal in the brain contributes to fat gain.


Tucker Goodrich said...

It's not a great experiment. The hypoglycemia is what causes over eating. If you mask that with the insulin you're killing the critical part of the process, IMHO.

Marwan Daar said...

Tuck, what do you mean when you say that hypoglycemia is *masked* by insulin?

Stephan is saying that very high doses of insulin can cause hypoglycaemia (which can in turn cause increased food intake).

But most proponents of the insulin hypothesis of obesity (as I understand them) claim that high doses of insulin cause increased accumulation of fat inside cells, and prevents fat from leaving cells.

So even if high doses of insulin caused weight gain, it is through a very different mechanism than that invoked by proponents of the insulin hypothesis.

Sam Knox said...

If only we could get those pesky humans to stop getting fat every time we inject them with insulin.

Bill said...

Great post Stephen.
You make some very interesting points. But while those studies support your idea, it doesn’t seem to be a universal phenomenon (at least in the rodent world). Hamsters eat more and gain more weight in a similar paradigm:

DH said...

Hey Stephan, interesting q for you: Since food reward/taste influences setpoint, do you believe that eating while holding your nose (or sick, can't taste well) so that your taste senses are dminished would lead to a lower setpoint? In other words, if I can bypass the taste, will weight loss be expected?

Stephan Guyenet said...

Hi Sam,

The reason diabetics gain weight when they inject insulin is that they stop pissing out large quantities of glucose, among other things. Insulin is required for fat storage in fat cells but it doesn't regulate fat storage. Permissive vs. regulatory.

But in any case, not all diabetics gain weight with insulin. Jenny Ruhl of Blood Sugar 101 lost weight when she started insulin.

Hi Bill,

The hamsters were hypoglycemic. You have to give a sub-hypoglycemic dose if you want to look at the effect of hyperinsulinemia on food intake and body weight.

Steven Hamley said...

Perhaps a reason low carb diets are successful at weight loss in some/most people is because they increase mitochondrial biogenesis (through more glucagon, which increases PGC-1a).

If someone has poorly functioning mitochondria they may have difficulty burning body fat, leading to slow and steady weight gain over time.

Mary Rogge suggests high AMPK from poorly functioning mitochondria increase appetite and decrease energy expenditure, requiring higher leptin levels to balance this, thereby elevating the setpoint.

There's also the idea of triglycerides blocking leptin transport across the blood brain barrier. But as you've said it requires more work.

If mitochondria are a factor in obesity it can explain how high carb diets don't cause obesity (because carbs don't damage mitochondria), low carb diets help some lose weight (only if they have a problem with mitochondria), why people can increase carbs after low carbing without gaining weight (because they've resolved their mitochondrial dysfunction) and also things like low carb flu (more dependence on beta-oxidation that is only met when enough mitochondria grow).

Jiffry said...

Very interesting post Stephan.

I remember reading that people on low-carb diets experienced higher fasting blood sugar compared to populations with higher starch consumption and that VLC diets can result in insulin resistance.

Perhaps some low carb dieters lose weight because of increased insulin as a opposed to less of it?

What about the common notion that elevated insulin "locks up" the fat cells and elevated levels mean you lose lean mass as opposed to body fat when weight loss occurs?

FrankG said...

"In the first study...rats were implanted with mini-pumps delivering insulin at a steady rate throughout the day and night for 7 days. ... This is a model of chronically elevated insulin reminiscent of what is seen in insulin-resistant people."

Well it might be *A* model for chronically elevated insulin as seen in IR but I'd suggest that it is not a very good one.

SEVEN (7) days? A week-long trial on rats?!? Heck, at this rate we could bang out all those long term studies in under a month!

Seriously Stephan you are losing your scientific objectivity. Talk to some people with Type 2 Diabetes using insulin, talk to their Doctors.

There is no nail...

Andreas Eenfeldt said...

Hi Stephen,
Is this anti-insulin crusade running out of steam?

Just about any doctor knows treating patients with insulin leads to weight gain and just about any diabetic knows that using high doses of insulin stimulate weight gain.

People (humans!) with insulin producing tumors, insulinomas, often become ravenously hungry and massively obese if untreated. Coincidence?

To be so sure of the opposite because of a non-hypoglycemic dose one week (!) rat trial etc. or because you know of one anecdotal story of a diabetic who believes she did not gain weight on insulin... well that's not too convincing imho.

Chuck said...

forgive me if it was discussed but what was eaten in the 2 experiments referenced?

Gretchen said...

"The reason diabetics gain weight when they inject insulin is that they stop pissing out large quantities of glucose, among other things."

This is the knee-jerk explanation many doctors give to their patients, but in fact not all people with diabetes are pissing out large quantities of glucose. Dr Whittaker uses the same logic to say that all diabetics need huge quantities of his expensive supplements because they're peeing away their vitamins. At least you qualified with "among other things."

"But in any case, not all diabetics gain weight with insulin." This is true, and Levemir causes less weight gain than Lantus. Some people lose weight on insulin because without it the glucose can't get into the cells and the cells are essentially starving despite lots of glucose in the blood and send signals to eat more. With insulin, the glucose can get into the cells and they stop whining.

If you inject insulin into the brain, appetite decreases. But not many people would want to give insulin shots to the brain. They might have different levels of insulin resistance in brain and in fat cells.

Dr Richard Bernstein, who advocates LC diets for diabetics, one gave some thin patients an extra 900 calories a day of pure fat in an effort to get them to put on weight. After 6 months of this calorie excess, they hadn't gained an ounce.

This issue is complex, and it's dangerous to make black and white statements.

Gretchen said...

Correction: Bernstein didn't say his patients didn't gain an ounce; he said they didn't gain a pound in 6 months of 900 extra calories a day in the form of olive oil.

Melchior Meijer said...

Hi Andreas,

I have been interested in insulinoma for a while, because it is such a perfect model. But afaik patients only occasionally become obese if untreated. Please correct me if I'm wrong, I think it is very important we get this straight. Any idea what coud be different in the obese insulinoma patients compared to the ones who do not gain weight?

Anonymous said...

Hi, Stephan

As I mentioned before, the Internet wastes far too much time on this insulin hypothesis.

It is like asking "What cheese in the moon made of?" to learn about the moon. It would have gotten us nowhere to ask that. That is my analogy.

Real research into obesity is dealing with GENETICS, extremely complex neural circuitry ( which we have vast unknowns as I am sure you know :) ) and also the beneficial gut hormones cocktail produced by the re - routing of bariatruic surgery.

Scientsts are currently trying to indentify this cocktail of hormones and make it work as a pill. ( but again, baritaric surgery is not perfect. Patients are still obese - but less)

No one in genuine obesity science is wasting their time anymore with the unscientific and, childish, discredited "eat less, move more" or the insulin hypothesis

Sure insulin has a role ( a bit) but this hypothesis is farrrr too simplistic and immensely and dramatically incomplete.

Genetics, learning about the hellishly complex neural circuitry that controls body weight involuntarily, gut hormones and gut microbiota are the future.

Isaac Newton asked the right questions, one of THE most important things in science.

I know you're working hard on the neural circuitary part.( which is HUGE) I respect that.

Anonymous said...

The reason drugs like Prozac cause massive weight gain is because they act on the same neural circuitry that controls our body weight. Mike Tyson has obesity in his family, and was on Prozac type drugs.

It is no wonder at all that he gained massive weight.

They are not even an option because of this side effect for some poeple. I guess Poeple like Mike Tyson will REALLY get fat on those drugs.

Neural circuitry is a HUGE part. Over 10 % of morbid obesity ( and there is good reason to suspect more) can be explained by a signle defective gene

David Moss said...

So what about hypoglycemia? What id insulin didn't lock (too much) fat away directly, causing hunger, but rather caused hypoglycemia, causing hunger, causing too much fat to be stored? This is, loosely, the view of lots of people who don't know anything about insulin, but who know that carbs give them 'sugar crashes.'

Andreas Eenfeldt said...

I guess it could depend on the severity of the hyperinsulinemia?

I heard about one older lady with insulinoma at the hospital I work at, who did not want surgery. She used her alarm clock to get up and eat several times during the night and became very obese (with severe heart disease too).

Of course, according to Stephen it does not count if there could be hypoglycemia involved. That hyperinsulinemia often leads to hypoglycemia makes that a blatantly arbitrary way to ignore evidence in my mind.

Anonymous said...

I also wanted to clarify that clearly chronic major elevated insulin levels are clearly a health problem and do not mean to downplay that in anyway.They also can be thwarting with regards to obesity.

But as Dr. Linda Bacon Ph.D., points out, and most people here know, insulin is not all bad. And there are times , in normal healthy individuals who have a normal insulin spike response after eating, that it CAN be your BEST FRIEND for regulation of body weight. :)

Yes, it is one of the hormones identified by science in fat storage - but only one.So many more hormones involved

Insulin is a complex thing.

I am sorry about the Gary Taubes/Stephan fued. I respect both these guys for different reasons. To Gary's credit he spoke out against "eat less, move more" as a solution for obesity. That is the factor about him I support much more than the incorrect view that insulin is the cause of obesity etc.

I respect Stephan because he is working on a specific area of obesity- neurobiology.

Science is a group effort.

Gretchen said...

I agree that science is a group effort, and I apologize for going overboard with reaction to the "spilling glucose in urine" idea. I've just heard it so much that it sets me off.

I think one problem with anyone who has taken a firm stand on some scientific problem is that one is then required to defend it.

In my opinion, there's no single factor that causes obesity, and hence any theory that claims that there is can be criticized.

I also think it's often necessary to overstate your case in order to get any attention at all.

What Taubes was trying to do was to show that the fat increases cholesterol increases heart disease paradigm was wrong and that insulin is important.

That doesn't need to mean it's the *only* important factor.

Stephan Guyenet said...

Hi Andreas,

My "crusade" is to explain science to non-scientists, and that is what I have done. The insulin hypothesis of obesity never had much traction among scientists, and it is rapidly losing steam among non-scientists as well.

The effects of insulin in the context of diabetes, insulinoma and hypoglycemia are easily explained and bear little relevance to normal physiology.

In uncontrolled diabetes, you often have a severe relative insulin deficiency (too little insulin relative to the degree of insulin sensitivity). Insulin is required for fat storage in fat cells, so fatty acids can't be stored. But a lot of things are required for fat storage in fat cells: cell membranes, DNA polymerase, and a variety of other proteins. But that doesn't mean they regulate fat storage under normal circumstances. It's the difference between 'permissive' and 'regulatory'. Very different concepts.

There are at least three other reasons why diabetics sometimes (but not always) gain fat with insulin injections:

1) insulin stops glycosuria, which can account for a loss of hundreds of calories a day

2) insulin stops diabetic futile cycling in the liver, which is a major energy sink

3) injected insulin is poorly controlled relative to letting the pancreas do it, so many diabetics experience hypoglycemic episodes. Hypoglycemia is a potent trigger for food intake and induces major hormonal alterations as I will explain below.

Hypoglycemia also explains the case of insulinomas, which cause substantially increased insulin secretion and hypoglycemic episodes. Hypoglycemia activates the "counterregulatory response", which is a last-ditch effort by the brain to avoid hypoglycemic coma and death. It triggers epinephrine and glucagon release, and ravenous hunger in a desperate effort to bring up circulating glucose. This leads to fat gain.

This coordinated response has nothing to do with the normal response to circulating insulin and glucose in the physiological range.

John said...


Lucas Tarfur has written about low carb diets and insulin & leptin signaling in the brain.

Anonymous said...


I hear you completely. Back when I first started learning about obesity , all I heard about was insulin.

I never heard Dr. Friedman , or Dr. Sharma or Dr. Bacon mention insulin that much at all.On top of that, all those above individuals criticize ( and rightly so) the equally simplistic , unproven extrapolation that all we need to do is "eat less, move more" to cure obesity.

Dr. Friedman has spoecifically said we need to move away from that UNscientific approach.Evry single other medical condition DEMANDS a scientific approach, obesity should be no different. No nostrun is ever used to treat any other medical condition.

The insulin hypothesis advocates , as well as "eat less, move more" advocates ( the popular anti Gary Taubes critics too) are BOTH- BOTH off base Internet guru popular concept items right now.

BOTH those camps are wrong. Research from pioneering scientists is going in a completely different path than EITHER of those Internet camps.

They're both asking "what cheese is the moon made of?"

Sure energy balance itself is valid ( pure weight gain/loss not differentiating between fat and muscle) but the BODY has much more to say about that than we ever could consciously- especially over the very long term . (body weight over a decade and more)

I just wanted to share. Thanks for getting the word out about insulin.

Best Wishes,


Estalentau said...

Last time I checked insulin is known 1)to activate LPL in fat cells and have the TG from de VLDL and chilomicrons shipped into the adipocytes (increasing fat accumulation) and 2) to inhibit the PDE, which in turns inhibit lipolysis inside the adipocytes (increasing fat acumulation).

FrankG said...

"Hypoglycemia activates the "counterregulatory response", which is a last-ditch effort by the brain to avoid hypoglycemic coma and death. It triggers epinephrine and glucagon release, and ravenous hunger in a desperate effort to bring up circulating glucose. This leads to fat gain.

This coordinated response has nothing to do with the normal response to circulating insulin and glucose in the physiological range."

So what is it about insulin that leads to weight gain in Type 2 Diabetes (Metabolic Syndrome) when the Blood Glucose (BG) is well controlled using injected insulin?

Doesn't a normal BG level imply that the circulating insulin and glucose are in the physiological range..?

I have personally been using injected insulin for over 9 years now and have experienced the very occasional hypoglycemic attack when I miscalculated the insulin amount around intense exercise at the gym. I wonder how many other of the posters here have experienced this? It is just as you describe: an intense need for food... right now!

But during that same number of years I also gained significant weight (excess fat mass) while by all other measurements -- such as regular home BG testing and A1c tests at the lab -- I was not experiencing hypos.

Is it really so hard to imagine a less acute state of high insulin levels leading over a long time (many months) to hunger? Chronic hyperinsulimemia.

You cannot so easily dismiss weight gain resulting from this... when a dog is chasing it's tail, which part of the dog is leading?

When high insulin leads to lower BGs, which leads to hunger, which leads to eating (preferably fast energy foods), which leads to higher BGs, which leads to more need for insulin which leads to lower BGs, and so on... which is cause and which is effect?

And no, a 7 day trial of rats does not model this chronic high insulin. What were the rat's insulin levels during that week? If the infused/injected insulin was lowering their BGs, they would have compensated by reducing their own secretion of insulin -- this is part of a feedback loop remember... a relatively simple peripheral control mecahnism that the body relies on to remain in homeostasis

I also wonder: exactly what are the qualifications that make one person a "scientist" and another not a scientist?

Andreas Eenfeldt said...

"[Extra insulin resulting in hypoglycemia, hunger and weight gain] has nothing to do with the normal response to circulating insulin and glucose in the physiological range."

Sure, you're absolutely right. But I'm not talking about the physiologic range. I'm talking about hyperinsulinemia, circulating insulin well above the normal range.

Do you know how many of my non-insulin treated obese patients have hyperinsulinemia when I test for it? The large majority.

Are you really acknowledging that in a state of hyperinsulinemia (such as the one most obese patients are in) the result may be hunger and fat gain? Or are you not?

It seems you want it both ways.

Thomas said...


Do you mean eat less, move more doesn't work, or that it's not sustainable?

LeonRover said...


Association . . Correlation . . . Causation . . . ???


Is that it?

Anonymous said...

Hi, Thomas. I mean that it is not supported by the scientific literature.It is not a given at all like Internet gurus say.

The body has ways to the get the weight back on us that we will not even realize because it happen so slow and over the long term.

Volunatary factors such as exercising and eating well, while they should be encouraged for health, do not substantially affect body weight over the lon gterm.We have *some* control but not much.

We have to acknowledge and respect the extremely limited potency of these meaures to affect body weight.

Eat less, move more is an exptrapolation of the commercial dieting industry. It is fighting our body- not honoring it.It is a very unscientiic approach tp obesity based on beliefs and assumptions.

It is possible that if we honor our bodies eat to satiety ,move around a lot we could stay at the lower end of our setpoint,( a range of 10 to 15 pounds or so) or help our bodies sort things out a bit, but that is about it over the long term

Over a decade most people's weight remian remarkably stable. Conscious intervention can not account for this.

Nobody can match the many, many, many millions of calories taken in toi expended over a decade. Yet, research shows most poeple are pretty stable in weight.

There is a biological involuntary system whcih sorts this out and far more accurate than we could, (99.6 % precision in fact when the system is not broken.

There is an immensely strong biolgical underpinning to morbid obesity.

Our body active;ly resist weight change in EITHER diorection. Morbidly obese poeple have this system broken, The brakes are never applied when gain starts.

Most people if they gain top out at a certain weight, feel horrible and their body gets them to lose a bit. Conversely when they farcically diet, the body gets the weight back.

The fattest I personally can get from over eating is 240 pounds tops at a height of 6' 3". My body fights this ferociously and I drop back down to 220 or less sooner or later. And when I try to get below 200 my fights me tremendously.

But I carry a lot of muscle and in prime could squat 440 12 reps, so 213 is an OK weight for me I am lean at that weight.

Evrybody is different. Biology dictates, Mike Tyson will never be stringy and skinny like Lance Arnstrong and vice versa.

Mike Tyson's build has the naturaly propensity for weight gain. But he stayed at low end of set point with boxing

But then again he was on meds that affect neural circuitry and weight gain .

The whole thing is immensely complex.

Our brains are actively monitoring our body weight and fat.

My friend got into a wasting state and was seriously sick. He oculd not help his extreme skinniness. But as his healthimproved his body cooperated and he regained a healthy weight.

He was a STICK and was eating reasonably Certainly his intake could not explain THAT level of skinniness.

What I now know from Dr. Friedman ( a true expert) explains all of this much better and it makes sense.
Take care,


Sanjeev said...

> Just about any doctor knows

just about any SCIENTIST and any LAYMAN with a bit of scientific knowledge, and in fact, ANYONE who's well read, but lacks a dull axe knows about controlling variables.

> it does not count if there could be hypoglycemia

Because the popular CIH HYPTOTHESIS codified by Gary Taubes states that INSULIN on its very own, completely by itself has the special, magical ability to "lock fat" into fat cells.

It states that insulin ITSELF is the mediator and controller directly of fat cells and how much they take in and release, not through indirect mechanisms, but, again, DIRECTLY.

That is the idea that the above experiments speak to.

Sanjeev said...

interesting that every time this kind of thing comes up, it becomes about diabetics and insulin, and the results are applied to carbohydrate - released insulin.

Inconvenient facts are just swept under the rug.

How about the insulin from protein, which can be as high as any insulin from carbohydrate - doesn't count, huh?

Talk about "having it both ways".

Sanjeev said...

> the results are applied to
> carbohydrate - released insulin

the results are applied to carbohydrate - released insulin in non-diabetics.

FrankG said...

@LeonRover -- "Association . . Correlation . . . Causation . . . ???
Is that it?"

Let's see... on the one hand we have practical observations and successful, health-promoting interventions by a practising physician, based on experiences with real people; and on the other hand we have a 7 days trial with rats.

We have the observations: that people with the symptom of Obesity are more likely to also suffer with Hyperinsulinemia, Insulin Resistance (IR), Glucose Intolerance, Type 2 Diabetes, Dyslipidemia, Hypertension, Cardiovascular Disease (CVD) and the rest of the Metabolic Syndrome.

Then we have experimental evidence that by putting these same people on a diet which reduces their tendency to raised Blood Glucose (BG,) we thereby reduce their need for insulin -- both exogenous (injected) and/or endogenous (secreted naturally). This invariably leads to significant loss of excess fat mass, improved lipids, lower Blood Pressure, and improvements in every other health marker -- all this just from a dietary intervention; which leaves the person satisfied... i.e. not hungry all the time.

What conclusions and what practical applications would you suggest based on these observations?

Do we immediately leap to the fact that it must be leptin or genetics at play here? And/or that we should wait until a pharmacological solution comes along to save us all from ourselves?

Now maybe none of this applies to you -- your blogger profile mentions skiing and rock-climbing -- so perhaps you have little or no experience of significant excess fat mass over multiple decades, and the metabolic sequelae involved but please trust me when I say that: while insulin injected into a rat brain may be an appetite suppressant, when injected into the subcutaneous tissue of an human (or raised chronically by a diet high in sugars and refined starches) it has quite the opposite effect.

FrankG said...

Sanjeev -- Diabetes provides excellent working examples of dietary and insulin ineterventions in humans... as opposed to rats.

As a person with Type 2 Diabetes who also injects insulin, I am initmately aware of the short and long-term effects of food, insulin physical activity, stress etc... on my BG, weight, health etc...

I have Excel spreadsheets reaching back almost a decade tracking my progress.

Can you say the same about your own self-experimentation?

I don't see anyone denying that dietary Protein can also raise BG -- although it is a much slower effect that Carbohydrate -- that is why I have chosen (and easily sustained) a low-carb/high-fat (LCHF) diet for these past 3+ years.

I understand that Dr Andreas also recommends LCHF to his patients, and the the point he just made was that hyperinsulinemia is present in non-diabetic obese patients as well. Do we also ignore/discount them?

What is the point of this discussion if not to help those of us who have metabolic dis-orders?

CarbSane said...

Hi Stephan, I can see from the comments that more nails are needed to seal this coffin. :)

Medjoub said...


The guy that writes this blog is named STEPHAN. His name has an "A" in place of the "E". Despite the close phonic relationship, the names 'Stephen' and 'Stephan' are indeed different. Is it that difficult to show your interlocutor at least the bare minimum of respect when attempting to engage him in HIS forum?

Kris said...

Stephan, I really appreciate what you are doing here.

I used to believe in the insulin hypothesis myself, but you have very convincingly proved me wrong.

Don't let the haters get to you.

Sanjeev said...

> Do we also ignore/discount them?

We don't ignore the people, and we don't ignore or dispute that the DIET may "work"[0] (IFF it reduces caloric intake below maintenance); we dispute the proffered theory for WHY the diet works.

> Sanjeev -- Diabetes provides excellent working examples of dietary and insulin ineterventions in humans... as opposed to rats.

and real research in non diabetics beats all of those, and multiple lines of evidence there all converge here: in the "normal obese" (non-diabetics), insulin is involved but does not control ; IOW, CIH is incorrect

> I don't see anyone denying that dietary Protein can also raise BG

I didn't claim anyone was denying it.

protein raises INSULIN. Some specific amino acids come close to glucose for insulin stimulation. If insulin "locks fat into fat cells" why doesn't the insulin secreted to handle protein also do the same? glucagon doesn't count: AFAIK, in humans glucagon has minimal effects outside the liver[1]

> I have Excel spreadsheets reaching back almost a decade tracking my progress.

And I have read many reports indicating high fat worsens the situation and controlled research showing low calorie high carb can reverse it.

[0] for some definitions of work, depending on individual goals. Lex Rooker, last time I read him, can't eat any significant amount of carbs after a few years on near zero carb. does that count as "works"? If he's ever caught in an emergency and needs sustained anaerobic work ...

[1] I need updating on this, if the state of the art has moved on and proved otherwise

Sanjeev said...

David Moss said...
> So what about hypoglycemia?

So now that insulin intrinsically, on its own, is not to blame, hypoglycemia is to blame for the obesity epidemic but since insulin causes hypoglycemia ... worl, it's still the bloody insulin, innit squire[0]?

first, how common is real hypoglycemia? As common as obesity? And has it increased with time (to explain the obesity epidemic, hypoglycemia would have to occur a lot more frequently, wouldn't it? Are people in the US fainting all over the place these days? Planes falling out of the sky, car crashes, dogs & cats living together)?

Since this is about obesity and the obesity epidemic, I'll assume we're talking about the "normal obese" (probably some insulin resistance but not fully diabetic)

insulin's half life is SHORT. Remember that insulin gets dumped right into the portal vein just before that goes to the liver, the liver gets first shot at insulin, AND the liver's the organ that has the most insulin-destroying capacity: so insulin's half life is SHORT.


most of the insulin dumped in response to a meal is long gone by the time hypoglycemia rolls around.

So naturally insulin's to blame, right?

phhhhttt ... I don't buy it

come on ... any engineer in the world ... any systematic, rational problem solver will tell you the first place they'd look is the control systems and specifically the switchover point - the control systems or some sensor or some effector or the communications between these are not moving optimally from fed state to post-prandial.

is the liver too insulin sensitive (turning off GNG even when blood glucose is too low?)

Are there special circumstances extant - excessive stress? lack of sleep? pharmaceuticals? Less acceptable chemical agents?

given the facts, IMHO, these are much more reasonable lines of inquiry. The problem now is low blood glucose, and old research that proves insulin blocks fat use in vitro is inapposite. Arguing that insulin drove glucose out of the blood is irrelevant too: that's one of insulin's proper functions, and by the time this happens the insulin level is usually far below the meal's peak.

I'm not a biologist, but that's my take on your question anyway.

> People (humans!) with insulin producing tumors, insulinomas, often become ravenously hungry and massively obese if untreated. Coincidence?
thhhhhhufferin ttthhhhuccatash

[0] or should that be guvnor? Wish I had learned English in my yoot.

Medjoub said...

Sanjeev -

Did Lex Rooker resolve the issue of why he was gaining weight on zero carb?

That was a brief topic of interest on the Paleonu blog a couple years ago...

Sanjeev said...

Medjoub said...

Sanjeev -

Did Lex Rooker resolve the issue of why he was gaining weight on zero carb?

He intentionally reduced his caloric intake.

Unless he's had another episode of fat gain with a different solution, that was worked

Did you set that up?

> is the liver too insulin sensitive (turning off GNG even when blood glucose is too low?)

Which means, in the presence of that amount of insulin plus that amount of blood glucose, what normally gets the liver to produce glucose and why isn't it occasionally, with a "normal" amount of circulating insulin

IOW, where to put the research emphasis and dollars? I'd put it on the liver and brain.

Andreas Eenfeldt said...

Stephan (and Medjoub),
I apologize for my misspelling. I certainly did not mean any disrespect.

FrankG said...

@Sanjeev -- "We don't ignore the people, and we don't ignore or dispute that the DIET may "work"... we dispute the proffered theory for WHY the diet works."

Sure... there is room in science to disagree as to the mechanisms involved. That is why I remain open-minded to the discussion AND that is also why I remain staunchly opposed to statements which imply that the discussion is over and chronic high-insulin levels driven by a diet high in sugars and refined starches has been disproved as a major player in our current obesity (metabolic) epidemic.

I (and apparently many others) have yet to be convinced of any such finality as "nails in coffins", and if the studies presented here -- such as a 7 day rat trial as a model of chronic high insulin levels -- is meant to be viewed as representative of the level of evidence being presented for any such "coffin", then I remain a long way from being convinced.

Satasha said...

Dear all,
I have high insulin and have had this for 20 years at least (since I was 15). As a teen I used to have hypoglykemia, not anymore. Carb intake seems to have not much of an effect on my BG. Interestingly, I still weigh the same, although I eat a lot and had a twin pregnancy. Many docs tested me for insuline resistance or diabetes 2, but these never showed up...
Any comments? I love inputs and discussions ;-)

rodeo said...

Andreas Eenfeldt, you're obese patients are not in the normal physiological range since they are obese which leads to insulin resistance and hyperinsulinaemia. I doubt that they have hyperinsulinaemia and hypoglycemia, right?

Stephan is talking about higher insulin in normal people leading to lesser food intake due to insulin's effect (in the brain) on satiety.

The nail in the coffin is not aimed at LC per se but instead at the idea that it would have anything to do with carbs->insulin->fat-gain and that carbs would be the cause of the obesity epidemic.

Jane said...

Nobody is going to understand the role of insulin in obesity without understanding how beta cells work. Insulin secretion OSCILLATES. Damaged beta cells don't do it properly.

'Peripheral insulin concentrations oscillate because of insulin secretory bursts every 6-10 min, which are the main source of overall insulin release. ... This oscillatory pattern, which is important for insulin action, is observed at the level of the individual beta-cells and in the islets, both in vitro (in the isolated perfused pancreas) and in vivo. ... Routine screening procedures for insulin secretion yield little predictive value for later development of diabetes but more sophisticated methods using time-series analysis of diurnal, ultradian and rapid oscillatory insulin secretion reveal the presence of profound defects in glucose-intolerant individuals.'

Beta cells are very sensitive to oxidative stress, because they make large amounts of proteins for export that need to fold properly. Oxidative stress prevents proper folding.

The key to all this is iron overload and deficiencies of magnesium and manganese. So long as we go on believing that we have too little iron and not too much, we will never solve these problems.

Gretchen said...

"This coordinated response has nothing to do with the normal response to circulating insulin and glucose in the physiological range."

I think this is the problem here. People who are obese do *not* have normal responses. Something is broken.

So what may pertain to "normal" people may not pertain to those in whom something is broken. And if you're at a normal weight, you're probably not very interested in dieting.

Beth@WeightMaven said...

Gretchen, if the obese once had a "normal response to circulating insulin and glucose in the physiological range" then what caused that response to become broken?

If the response has always been broken, what caused that?

Count me in the group who thinks that neither was caused by dietary carbs per se.

Don S said...

While I agree that the complete body of evidence does indeed lead to a conclusion that elevated peripheral insulin levels as "the" cause of obesity is not tenable, these studies seem weak, unless they measured total peripheral insulin levels during the experiment. Of course if insulin, or an oral insulin receptor agonist, is given, then endogenous insulin production will decrease; these studies do not show what elevated insulin does, nor what long term elevated insulin level does.

What's so complicated about why "low carb" can induce weight loss? High protein has high satiety; ketosis makes one if not a bit nauseated, then often just not hungry; total calorie intake decreases. The concerns that many have with "low carb" is the tendency of some to use that as an excuse to eat lots of high fat red meat, often high salt level processed meats, and the lack of fiber and phytochemicals in the diets of some as they implement the plan.

Yes, Stephan's explanation of weight gain associated insulin treatment of diabetes is a bit simplistic. Conventionally treated diabetics do not have glycolsuria but are less likely to gain weight as are those intensely treated with greater insulin dosing. Current thinking is that high insulin dosing leads to insulin resistance ... perhaps in the brain as well as peripherally.

Of interest may be and

Don S said...

It is also of note that an insulin analog, detemir, does not cause the insulin associated weight gain, at least over a moderate term time course, apparently by means of reduced intake. ( Those using detemir instead if NPH for basal control actually lost weight, and had decreased leptin and higher resistin.

This would be consistent with Stephan's position that stimulation of peripheral insulin receptors is not the key issue driving weight gain. Of note detemir appears to have tissue preference for brain compared to peripheral tissues. ( Also consistent with Stephan's hypothesis. Furthermore, euglycemic infusion of insulin and detemir demonstrates ( that "While inducing comparable peripheral effects, detemir exerts stronger acute effects on brain functions than human insulin and triggers a relative decrease in food consumption, suggesting an enhanced anorexigenic impact of detemir compared with human insulin on central nervous networks that control nutrient uptake."

Stephan,are you aware of any direct evidence that chronically high levels of insulin induce central insulin resistance? Also, can you suggest a reason why detemir does seem to induce such a central resistance at least in studies that have followed it out for over 12 months? (see for example)


Melchior Meijer said...

Thanks, Andreas.

FrankG said...

@Don S -- all of the Detemir studies you cite, used either Type 1 Diabetics or "Healthy" subjects.

Neither of these groups is generally characterised as having either Insulin Resistance (IR) or Excess Fat Mass -- let alone both... so for the purpose of this discussion could be considered "normal".

As Gretchen points out above, we have a disjoint between lean subjects and obese.

I for one, see little point in discussing the normal, healthy role of insulin in a normal*, healthy person.

Let's please be sure we are comparing apples to apples.

Isn't the topic "Insulin and Obesity" -- as the title says?


*You might argue that Type 1 Diabetes is not "normal" but in this case (at its simplest) the insulin is used as an hormone replacement in what is otherwise [generally] an healthy, lean and younger person. Arguably the "default" situation for Type 1 is too little insulin, as they tend to err on the side of safety... as opposed to the high levels seen with IR in Obesity and Type 2

David Moss said...


“first, how common is real hypoglycemia? As common as obesity? And has it increased with time...?”

Depends what you mean by 'real' hypoglycemia. If you mean, as your examples suggest, the sort of hypoglycemia largely limited to diabetics, that leads to “fainting all over the place” and “planes falling out of the sky” then this sort of 'real hypoglycemia' isn't necessary for the hypoglycemia-insulin hypothesis. All you need is slightly depressed blood glucose that leads to increased peckishness. Anecdotal reports that people do feel hungrier than before, as well as tired, shaky etc. after sugar, is certainly very common indeed, among people who have no axe to grind about insulin.
As this article makes clear, this sort of hypoglycemia is common among non-diabetics and can occur even at levels of blood sugarout that wouldn't usually attract attention and classification as 'real hypoglycemia.' A study discussed on Evolutionary Psychiatry blog also shows that reactive hypoglycemia is very common among PCOS suffers (which applies to about 10% of women).

Sanjeev: “How about the insulin from protein, which can be as high as any insulin from carbohydrate - doesn't count, huh?”

Because, as every-one knows that protein raises glucagon as well, which raises BG.

“so insulin's half life is SHORT. Very. most of the insulin dumped in response to a meal is long gone by the time hypoglycemia rolls around.”

This contradicts every graph of postprandial insulin and blood glucose, I've ever seen- normally insulin is elevated for at least a couple of hours when people eat a meal, while BG remains depressed.

“Are there special circumstances extant - excessive stress? lack of sleep? pharmaceuticals? Less acceptable chemical agents?”

Maybe there are, but this wouldn't threaten the carbohydrate-insulin hypothesis. Every-one recognises that the Kitavans et al can eat lots of insulin-stimulating carbs and be fine. What's most important is whether hunger/over-eating is generally increased following an insulinemic meal, for whatever reason. Even if the problem is partly caused by the fact that the liver shuts off GNG too readily, if the problem requires an insulinemic meal of carbohydrate to manifest, then something along the lines of the CI hypothesis remains alive and kicking.

Don S said...

Frank, the issue is determining normal function and how it is effected. The euglycemic study ("Euglycemic Infusion of Insulin Detemir Compared With Human Insulin Appears to Increase Direct Current Brain Potential Response and Reduces Food Intake While Inducing Similar Systemic Effects"
Hallschmid et al Diabetes April 2010 vol. 59 no. 4 1101-1107; full article at ) uses healthy normal weight males but references the findings in type 2 diabetics as well.

From that article:

"Insulin therapy using detemir has been frequently found to induce weight-sparing effects in comparison with other insulins, curtailing body weight gain in patients with type 2 diabetes (11,12) and maintaining stable body weight in type 1 diabetic patients (9,13,14). The mechanisms behind this favorable effect of detemir are unclear. Because of its pharmacologic properties, detemir might cross the blood-brain barrier faster and in higher quantities than other insulins and induce stronger effects on brain functions (15,16). Supporting this assumption, detemir in comparison with human insulin has been found to amplify the central nervous impact of hypoglycemia (16,17) and to exert stronger magnetoencephalographic effects in overweight humans (15) who display relative central nervous insulin resistance (18–20). ...
... while eliciting comparable peripheral effects, euglycemic infusion of insulin detemir compared with human insulin triggers a distinct negative shift in EEG DC-potential recordings and reduces calorie uptake in healthy men, supporting our hypothesis that detemir affects brain functions to a greater extent than human insulin and induces stronger anorexigenic effects on central nervous networks that control food intake. This outcome suggests that enhanced catabolic insulin signaling to the brain may be an important mechanism behind the limitation of weight gain observed in diabetic patients receiving detemir treatment (9,11–14) ... A most remarkable finding of our study is the reduction of ad libitum food intake by around 300 kcal in the detemir compared with the human insulin condition in the presence of identical peripheral actions of both insulins. This pattern renders the contribution of systemic mediators to this effect highly unlikely, rather suggesting that enhanced central nervous insulin signaling in the detemir condition resulted in decreased caloric intake."

One would hypothesize a lesser effect in obese individuals who already have neuronal insulin resistance.

Don S said...


I for one would very much appreciate a post putting together how insulin centrally seems to increase impulsivity (one presumes through reward mechanisms) in various models (for example ) but decreases food reward salience. Is there other evidence that the reward salience of food and those associated with drug behaviors are subserved by different and possibly conflicting mechanisms?

Don S said...

Sorry for multiposting but this article is also of much relevance:
"we have demonstrated that the insulin-sensitive component of brain glucose metabolism comprises 15–20% of the resting brain glucose metabolic rate in insulin-sensitive individuals. However, in insulin-resistant individuals, insulin-evoked increases in whole-brain CMRglc are less than half of this for the same insulin increment. It is of crucial significance that these effects are most marked in brain regions involved in appetite control and motivational behaviors. ... The present novel demonstration of insulin responsiveness in the ventral striatum, which is reduced in people characterized by insulin resistance (not necessarily with obesity), suggests the possibility that these changes are intrinsic to insulin resistance and may contribute to the evolution of insulin resistance into obesity and metabolic syndrome. We hypothesize that people with insulin resistance will need to generate a higher circulating insulin in order to experience the reward sensations of eating. ... In conclusion, our data show that the effect of low circulating levels of insulin is to stimulate cerebral metabolism within appetite and reward networks, focusing on the ventral striatum in the healthy human brain. For a given insulin concentration, this effect is reduced in insulin-resistant individuals, most markedly in the ventral striatum, consistent with an emerging picture of reduced function within cerebral reward networks in those at risk of overeating and obesity. These data support an etiological model of progressive dysregulation of reward networks on repeated exposure to palatable foods in people with insulin resistance. ..."

FrankG said...

As I see it: in the normal healthy person (yes even including a well-managed Type 1 Diabetic) Insulin (an anabolic hormone) has multiple peripheral roles including: the facilitation of Glucose uptake by cells for use as energy, the storage of Fatty Acids for medium to longer-term energy needs, and the uptake of Amino Acids (Protein building blocks) for the growth/repair of lean tssue, buidling hormones etc...

I can also accept a central role in signalling that the body is busy digesting the last meal so we can dampen down the hunger signals (or increase satiety if you prefer).

After a relatively short time (an hour or two after eating) insulin levels drop to a background (basal) level, Fatty Acids are used as energy between meals and all can remain in balance indefinitely... all is well.

Now take that same healthy, normal, balanced person and feed them foods rich in sugars and refined carbohydrates -- not just for a day, or a week, or a month but for years. And top them up with this same "rocket-fuel" as mid-morning, mid-afternoon and evening snacks... rarely if ever allowing the insulin levels to drop down to basal. In this case we gradually get a different picture emerging: Fatty Acids are mostly being stored and rarely used, the body craves foods rich in Glucose (and/or becomes less and less active) to make up for the energy shortfall. This creeps up slowly on the person over many weeks, months and years. As the excess fat mass increases, so does Insluin Resistance (IR)at different rates in different tissues, and so does the level of insulin required to compensate for the IR -- locked in a vicious cycle.

Even if the brain is being told by the insulin to dampen down the hunger, it is being overwhelmed by the cells which are starved for energy... then at some stage the brain also becomes resistant to the insulin signalling.

Same insulin, trying to do the same job... but now we are locked in an energy storage mode as opposed to an energy expenditure mode.

The solution? Cut out the sugars and refined carbs, then allow the body to re-establish its own equilibrium... hoping that not to much metabolic damage has occurred along the way.

This is why a 7 day trial on rats does not convince me.

Gretchen said...

@Beth. If I could answer your questions, I might get a Nobel Prize. But I can't, which is too bad, because I'd love a free trip to Sweden.

Sanjeev said...

> as well as tired, shaky etc. after sugar

Interesting ... how much sugar and how soon after does the hunger manifest?

Don S said...


Thing is that science is not about what you can accept; it is about figuring out models that more accurately reflect what is, whether it fits what we a priori accept or not.

None of these studies bring anything to bear on the whether or not any particular nutrition plan is healthy or effective at inducing or maintaining fat loss; it instead helps flesh out one layer more the complexities of real biological systems in all their nonlinear dynamic beauty.

James W. Hill, MD said...

Before one dismisses insulin's role in causing or worsening obesity, consider this hypothesis:

When insulin resistance in muscle cells is greater than in fat cells, as occurs in many cases of obesity, elevated insulin drives fat storage and worsens obesity. When insulin resistance in fat cells is at least equal to that of muscle cells, as may occur in some obese people, and certainly occurs in thin people who are fully insulin sensitive (i.e., both muscle and fat cells have zero insulin resistance), then insulin-driven preferential storage of energy in fat cells rather than in muscle cells may not be evident.

This hypothesis seems to fit the studies Stephan cited here. For example, the failure of rats in the studies of Vanderweele (1980) and Air (2002) to gain weight may be at least partly explained by the rats' muscle cells being insulin sensitive. The same would hold true for rodents with insulin resistance only in their liver cells (e.g., LIRKO mice).

And the effect of this caloric partitioning by differential insulin resistance between fat and muscle would be compounded by resistance to both insulin and leptin in the brain, both of which lead to greater food intake. The extra calories may be shunted more to fat cells than to muscle cells as long as fat cells are relatively more insulin sensitive than are muscle cells.

Of course, Stephan is correct in observing that "food intake is regulated by a complex 'symphony' of signals originating from the brain, fat tissue, the gastrointestinal tract and the pancreas ...." But it seems these studies don't yet disprove that insulin may be playing second violin in that symphony.

Sanjeev said...

I wasn't going to reply because this was SO BAD ...

In answering this question:
>> If insulin "locks fat into fat cells" why
>> doesn't the insulin secreted to handle
>> protein also do the same?

this is a non-sequitur:
> Because, as every-one knows that protein
> raises glucagon as well, which raises BG.

>> “so insulin's half life is
>> SHORT. Very. most of the insulin dumped in
>> response to a meal is long gone by the
>> time hypoglycemia rolls around.”

> This contradicts every graph of
> postprandial insulin and blood glucose,
> I've ever seen- normally insulin is
> elevated for at least a couple of hours
> when people eat a meal

that settles it then. Insulin's half life is quite long. Must be.

have you actually researched human hypoglycemia? What the concensus scientific/medical opinion is? Here's one hit of many - the bulk of the non-layman hits echo this

> However, it turns out that most of these
> patients suspected of postprandial
> hypoglycemia probably do not have it. If they
> do have hypoglycemia, it is transient and
> often not related to their symptoms. Some
> have argued against this diagnosis for the
> following reasons: Many healthy individuals
> have a very low glucose level at some point
> during an OGTT without symptom

And that's one part of what you want to rely on to "rescue" N_CIH_AA? [2] Moving from a rock solid CIH "locks fat into fat cells" to a diagnosis that may not exist ... that's called a retreat, not a reinforcement.

> something along the lines of the CI hypothesis remains alive and kicking

you MUST be trolling. MUST BE.

"hypoglycemic peckishness makes you obese" is an animal of a different Clade than the standard CIH, which claims (paraphrased) "insulin released to handle carbohydrate locks fat into fat cells and makes you UNCONTROLLABLY hungry"[1]

Sanjeev said...

>> “Are there special circumstances extant -
>> excessive stress? lack of sleep?
>> pharmaceuticals? Less acceptable chemical
>> agents?”

> Maybe there are, but this wouldn't threaten
> the carbohydrate-insulin hypothesis

So the extremely well studied(some of these studies being well randomized and controlled), well documented effects of sleep on hunger, of adequate sleep reducing hunger, of antidepressants on obesity, of quitting addictive drugs on obesity, all give way, are all of an order of magnitude lower than, hypoglycemic peckishness, whose existence is disputed by the relevant professionals in the field.

> What's most important is whether
> hunger/over-eating is generally increased
> following an insulinemic meal, for whatever
> reason.

"for whatever reason" ... meaning any thing that coincidentally causes hunger between some vague points in time supports CIH.

this MUST be a joke, right?

By revising what was once a somewhat testable hypothesis and turning it into a new hypothesis based on dubious phenomena, with open-ended, vague criteria for proof, you've made non-standard CIH unfalsifiable.

If you want to claim this IS CIH "in some form", fine ... you just killed both.

> this wouldn't threaten the
> carbohydrate-insulin hypothesis

You'r right - it doesn't threaten

[0] "hypoglycemic peckishness". Not only are we revising CIH beyond recognition, we're revising hypoglycemia even further from the already suspect diagnosis well into untestable, ufalsifiable waters

[1] IOW " teh insulins p0Wx0rz j00R @553z d00bz"

[2] not CIH at all

> This contradicts every graph of
> postprandial insulin and blood glucose,
> I've ever seen- normally insulin is
> elevated for at least a couple of hours
> when people eat a meal

that settles it then. Insulin's half life is quite long. Must be.
Really. You saw a few pictures and that proves insulin half-life is long


or copy anbd paste:

one more thing about hypoglycemia: here are some insulin and glucose curves for insulin resistant people (the "fructose" group is the resistant one)

Where's the insulin concentration high when the glucose is near zero?

by the way ... these curves destroy standard CIH too. there's that "Multiple independent lines of evidence" thing again

Sanjeev said...

The curves are in figure 5

Jimmy Gee said...


I think you need to step away from the trees for a while and look at the forest. It might improve your perspective and reasoning ability. It would do you good to get out of "your world" for a while.

CarbSane said...

@James MD: "When insulin resistance in muscle cells is greater than in fat cells, as occurs in many cases of obesity, elevated insulin drives fat storage and worsens obesity."

There is considerable evidence that fat cells become IR before muscle cells. Peripheral IR when in conjunction with obesity, does not pre-date the obesity it is the result, primarily, of chronic overnutrition. The carb-> insulin-> obesity hypothesis cannot be rescued by this sub-hypothesis.

CarbSane said...

@FrankG: I doubt it's worthwhile to point this out, but you keep comparing all of Andreas' experience with one study. As if that's all Stephan is basing his opinion on. The fact of the matter is that Stephan and others like myself, have looked at the enormous body of the scientific evidence, not a few cherry picked studies and outdated texts, and demonstrated time and again that the carb->insulin->obesity hypothesis is not a workable one.

Diazoxide would be a best selling diet drug if it were.

Gretchen said...

"Peripheral IR when in conjunction with obesity, does not pre-date the obesity it is the result, primarily, of chronic overnutrition."

Have you considered that it might be both?

I'd be interested in the "considerable evidence" that fat IR precedes muscle IR.

Nigel Kinbrum said...

Inactive muscles down-regulate uptake of glucose and FFAs, so are IR.
Full fat cells "leak" FFAs and glycerol, so are IR.
Which happens first?
I'd quote Lyle McDonald and say "It depends".

James W. Hill, MD said...


"There is considerable evidence that fat cells become IR before muscle cells."

Thank you. A citation would be welcome, for further investigation.

"Peripheral IR when in conjunction with obesity, does not pre-date the obesity it is the result, primarily, of chronic overnutrition. The carb-> insulin-> obesity hypothesis cannot be rescued by this sub-hypothesis."

I'm not arguing for the carb -> insulin -> obesity hypothesis, but rather overnutrition -> peripheral IR (muscle IR > fat IR) and central IR -> elevated insulin and greater food intake -> more fat accumulation

From a 2005 review by Elvira Isganaitis and Robert Lustig:

Because insulin resistance and hypersecretion often coexist and are partly interdependent, it can be difficult to tease out the relative contributions of each to the genesis of obesity. Still, insulin resistance appears to contribute to weight gain in adults and children, particularly with regard to the development of abdominal obesity. This may occur because of heterogeneity in insulin resistance between tissues. Adipose tissue tends to retain its sensitivity to insulin in the face of hepatic and skeletal muscle resistance. In experimental models, adipose tissue-specific and muscle insulin receptor knockout animals remain lean, whereas liver and CNS knockout animals become obese and have type 2 diabetes develop. Chronic insulin administration leads to muscle insulin resistance, whereas adipose insulin sensitivity remains high. (emphasis added)

Nigel Kinbrum said...

"Because insulin resistance and hypersecretion often coexist and are partly interdependent..."
Insulin resistance causes insulin hypersecretion, not the other way round.

A fat but active person can have IR fat cells and sensitive muscle. A slim but sedentary person can have sensitive fat cells and IR muscle cells.

David Moss said...

The fact that protein raises glucagon isn't a non sequiter: we're discussing hypoglycemia caused byty insulin and glucagon raises blood sugar, stopping hypoglycemia. Also glucagon causes more free fatty acids to be released into the blood- the opposite of locking fat away into fat cells.

The very graph which you link to below regarding insulin shows exactly what I said- namely that insulin remains elevated for about 4 hours and that's after a drink of glucose, rather than actual food.

You cite one review which supports your position, but here's another review which takes the opposite view (

“you MUST be trolling. MUST BE.”

I just asked a simple question asking for more info from Stephan, you're the one posting ad hominems and straw men left, right and centre.

“So the extremely well studied(some of these studies being well randomized and controlled), well documented effects of sleep on hunger, of adequate sleep reducing hunger, of antidepressants on obesity, of quitting addictive drugs on obesity, all give way, are all of an order of magnitude lower than, hypoglycemic peckishness, whose existence is disputed by the relevant professionals in the field.”

The question you asked wasn't whether sleep/stress/pharmaceuticals might be better studied factors in obesity, you asked whether they might be “special circumstances” in hypoglycemia. If they were (I replied) it wouldn't undermined the CIH, because the ingestion of carbohydrate and subsequent would still be the problem, even if they were only problematic because of sleep/disordered GNG etc.

“"for whatever reason" ... meaning any thing that coincidentally causes hunger between some vague points in time supports CIH.”

No. “For whatever reason” means the same as above. Even if a cause of the problem (hunger/over-eating) is disordered GNG or a stress disorder or too little sleep, if the problems still arise only where there is carbohydrate and insulin, then that would support a CIH. Similarly, it is not important whether problems are cause by hypoglycemia, insulin directly locking excess fat away, poor mitochondrial flexibility or something else, if the problems are still resultant from the insulinemic properties of carbohydrate

Paul Hagerty said...
This comment has been removed by the author.
CarbSane said...

@Gretchen & James: Here's one blog post discussing a review by Keith Frayn on the subject:

Frayn seems to have changed his tune a bit on this, but more so looking at the postprandial side of adipocyte function. As cells grow, they seem to lose their ability to effectively clear fats from the bloodstream as well.

James, I'll have a look at that citation. However if they're talking the various IRKO's, the muscle cells retain their insulin sensitivity in the LIRKO (liver knock out) as do the fat cells. These are genetic mutations, however, they do not develop resistance "normally" I think you would agree.

Sanjeev said...

my points remain unanswered, just one example -

this was the exchange

> insulin dumped in response to a meal
> is long gone by the time hypoglycemia rolls
> around.”
>This contradicts every graph of postprandial
> insulin and blood glucose, I've ever seen-
> normally insulin is elevated for at least a
> couple of hours when people eat a meal,
> while BG remains depressed.

And you report the exchange this way:

> The very graph which you link to below
> regarding insulin shows exactly what I
> said- namely that insulin remains elevated
> for about 4 hours and that's after a drink
> of glucose, rather than actual food.

what happened to the "while BG remains depressed"? [6]

revised away to suit your needs. Redefine claims until "some version of it is true" ... sound familiar? Done it lately elsewhere?

it's a little like what you're trying to do with "CIH" ... redefine it and make the criteria for proof vague and open-ended. Keep doing it until it cannot be incorrect.

it's called "widening the goal posts".

as regards "CIH" or more accurately the "CI guess" you've already widened the goal posts to encompass Alpha Centauri (diagnoses that may or may not exist prove it) and include Betelgeuse ("low"[1] blood glucose sometimes[1] happens around[1] the same time as hunger sometime[1] after a meal[2]), and include Sirius (sleep can't disprove but can prove, if my proposed coincidence happens) and a couple of quasars(sleep can't disprove but can prove, if my proposed coincidence happens). Someday you'll be able to drop-kick Andromeda through there and "prove" CI guess is true.

What's the draw to CI guess that makes all this necessary?

> I just asked a simple question asking for more info from Stephan

That was your first post.
Later you went off the rains into the deep end, to wit:

> something along the lines of the CI hypothesis remains alive and kicking.

Nothing that you proposed looks anything like the "CI hypothesis"[4]. You've used correlations and concocted specious arguments and linguistic legerdemain to convince yourself disproving it proves it [5]

Say hi to Douglas Adams next time you're at a zebra crossing.

[0] oh, but wait, that "while" went missing above, so maybe low blood glucose is no longer needed for "hypoglycemic peckishness" ... more goal widening

[1] I'm getting shivers ... ooohhhhh ... like a York Peppermint Patty. Exacting rigor always excites me to no end

[2] and never mind the documented cases of asymptomatic low blood glucose ... mre of that "nothing disproves, everything proves"

[3] >you asked whether they might be “special circumstances
you really didn't get that part of it at all, huh?

[4] it was a guess, not a hypothesis, and now it's a disproven guess

[5] ... oh, my bad "something like it". There's those vague criteria again ... you seem quite skilled at those.

[6] re define the BG level at 30 minutes in those curves as depressed[7] you didn't try it yet but it's in keeping with what's gone before.

[7] my bad ... you would phrase it "something like depressed"

Sanjeev said...

> Later you went off the rains

that's referring to Claude Rains, "hobo Sheriff" fame

phhhhhtttt ... should be "off the rails"

Sanjeev said...

responding to my observation of loose criteria

> No. “For whatever reason” means the same as
> above. Even if a cause of the problem
> (hunger/over-eating) is disordered GNG or a
> stress disorder or too little sleep, if the
> problems still arise only where there is
> carbohydrate and insulin

you loosen them further
> problems still arise only where there is
> carbohydrate and insulin

instead of tightenin up the diagnosis and the times that it might apply and the levels of insulin and glucose, you leave the time and the diagnosis completely vague and you elide "low " from the definition "hypoglycemia is low blood glucose".

CarbSane said...

BTW James, The LIRKO mouse does not become obese, while MIRKO does. Both retain insulin sensitivity in the fat tissue. The LIRKO mouse becomes profoundly hyperinsulinemic as well.

Gretchen said...

@CarbSane, I was hoping for primary sources, not a blog discussing a review that discusses primary sources.

CarbSane said...

Sorry Gretchen, but I'm not one to dump lists of PubMed abstract links on people. I have several other studies I've blogged on. I almost always provide a direct link to full text or try to share relevant data and/or the PDF via Google docs. With all the peer review researching I do, and the enormous time investment in the blog itself, it is impossible for me to provide a list of original studies. If this is inadequate for your needs, I guess you can just ignore my comments on others' blogs.

nothing91 said...

"@CarbSane, I was hoping for primary sources, not a blog discussing a review that discusses primary sources."

You've got a lot of nerve asking for the evidence when someone says there's "considerable evidence" of something. Outrageous. :-)

James W. Hill, MD said...


“The LIRKO mouse does not become obese, while MIRKO does. Both retain insulin sensitivity in the fat tissue. The LIRKO mouse becomes profoundly hyperinsulinemic as well.”

Exactly. A good hypothesis to explain this is that hyperinsulinemia does not drive excess fat storage as long as muscle cell are insulin sensitive relative to fat cells. That is the implication of Lustig’s article I linked above.

On this view, LIRKO mice are hyperinsulinemic because of the insulin resistance in their liver cells, but because their muscle cells are relatively insulin sensitive, they don’t get fat. MIRKO mice have insulin-resistant muscle cells, so they get fat.

In the presence of relatively insulin-resistant muscle cells, peripherally acting insulin can apparently provide a strong fat accumulation signal.

When the brain is also insulin and leptin resistant, appetite is greater, satiety may be lessened, and caloric intake increases. This worsens the fat gain established by hyperinsulinemia in the presence of relative insulin resistance in muscle cells.

Stephan Guyenet said...

Hi Dr. Hill,

I think it's worth pointing out that the MIRKO mice are not much fatter than normal mice. Their body fat percentage is only 38% higher than controls (and they also have less lean mass than controls, so the body fat % measurement may overestimate the absolute increase in fat mass). That would be like me going from 10% body fat to 14%, and that's from a complete lack of insulin signaling in muscle tissue, not a partial resistance as you see in insulin-resistant people. So even in this extreme case, the increase in fat mass is modest. Compare that with leptin receptor knockout mice, which are so obese they can barely walk.

I have yet to see any evidence that fat tissue remains insulin sensitive while muscle becomes insulin resistant in humans. To the contrary, the evidence seems more consistent with the idea that they become resistant in concert, and that is at least partially due to muscle getting excess fatty acids dropped on it because adipose tissue isn't trapping fatty acids efficiently well enough because it's insulin resistant.

DH said...

I feel like I had a pretty good question above. Since palatability is a central part of your hypothesis on food reward, I asked whether clipping/covering your nose while eating (which would render foods tasteless or less tasty) would lead to weight loss and a lower set point? Please address my question if you get a chance, even if just an opinion or hunch. Thanks Stephan.

CarbSane said...

@James, the MIRKO muscle cells put out myokines that enhance glucose uptake and DNL in the adipose tissue of these mice. This is not analogous to any human model of peripheral insulin resistance that I am aware of -- and does not appear to be some excessive trapping of fatty acids.

I have an excellent review of the IRKO's if you're interested. Email me carbsane at gmail dot com.

FrankG said...

@Dr Hill -- you may have read this Editorial in the American Journal of Clinical Nutrition, December 2005

"...A growing body of studies has pointed to the presence of heterogeneity regarding insulin resistance and insulin sensitivity among different tissues (4). Thus, there are indications that, within a person, the lipolytic response of adipose tissue may maintain insulin responsiveness even as skeletal muscle insulin resistance is detectable. In recent years, several studies independently reported that intermuscular fat deposition as well as increased fat content of muscle cells is associated with insulin resistance (5). The previously held view of adipose tissue as a relatively passive, energy-storing organ has developed into an appreciation of the important role of this tissue in regulating energy metabolism and serving endocrine functions (6). In particular, a dissociation of fatty acid release from adipose tissue from energy requirements in other organs may lead to fat deposition in nonadipose tissue, including skeletal muscle and liver (7). Using elegant, nuclear magnetic resonance spectroscopy techniques, Shulman et al (8) conducted studies suggesting that skeletal muscle mitochondrial function may be increasingly impaired with age, resulting in insulin resistance. Their findings, published in a series of articles, point to the importance of linking adipose tissue, skeletal muscle, and pancreatic endocrine function in a systems biology approach. ..."

I realise that the above states "the lipolytic response of adipose tissue..." but it also states "...adipose tissue may maintain insulin responsiveness..."

I think that differing rates of IR in different tissues (and I have read the same from multiple sources) makes sense of the observation that not all Type 2 Diabetics (up to 20%) are obese at the time of diagnosis.


I must voice my confusion over posts above where on the one hand we are told "IRKO" mice are genetic mutations that do not model normal human IR and yet they are also held up as examples to show how we can model the development of [normal] human IR?

FrankG said...

Even supposing that all the tissues (liver, muscle, adipose, CNS) do become Insulin Resistant at the same time, does it necessarily follow that the IR is at the same level or has the same level of effect in all those tissues? In other words: is an IR human just the same as a non-IR human but needing extra insulin to do the same job?

FrankG said...

More discussion here Dr Hill... although this one does seem to rely heavily on mouse models, so please note my confusion just above...

"These reports indicate that insulin resistance can no longer be considered as only a uniform systemic alteration. Conceptually, it is better to consider insulin resistance as being tissue-specific, and the ultimate phenotype of the individual depends on the sum total of the effects of the differing sensitivities of the differing tissues."

Don S said...


A very interesting piece that editorial. So it does seem increasingly well established that different tissues do not necessarily "become resistant in concert". Some may start out relatively more IR than others from before birth, some may be more prone to become more resistant given certain inputs that others, and these patterns may vary between individuals.

I would think that brain receptors, even varying by region, may end up in that list. It may end up being that individuals have a variety genetic predispositions that develop certain insulin resistance patterns that are influenced by environments from prenatal exposures (placental insufficiency for example) on.

This divergent pattern of IR in different tissues and perhaps even within different tissues and how they relate to each other in concert given certain exposures is going to be a very interesting story to follow. Certainly it is more complicated than the simple model that Taubes promotes to the general public.

FrankG said...

@Don S -- I can agree that it certainly does appear to be more complex than the simplified way that which Taubes and others have written/said is often paraphrased or misquoted by others (some prime examples of this are apparent even in the comments above) but then you have to start somewhere, and for me the biggest paradigm shift has to be the move away from the idea that behaviour drives biochemistry ..I am convinced it is the other way around.

On the other hand we can sometime get bogged down in the minute/exquisite details of biochemical processes when many of these may be downstream of what might prove to be a better focus for our efforts...

I recall the analogy of hitting a person on the head repeatedly with an hammer, then examining the complex set of biochemical reactions which ultimately eliicit the pain response. An interesting line of research to be sure, with the distinct possibility of coming up with a pharmaceutical solution aimed at blocking the pain signals... but so far as the test subject is concerned the common-sense solution to not having the pain, is to avoid being repeatedly hit on the head!

In my case the "hammer" was a diet high in sugars and refined starches... by avoiding those I have improved my health immeasurably. If it turns out to be as a result of a peripheral and/or central effect of insulin and/or leptin, or indeed something else altogether does not change the fact that a simple and easily sustainable dietary change has reaped huge rewards for me.

Don S said...

And I mostly agree. Although I think that behavior and biochemistry (and neurobiology) are neither one driving the other as much as different levels of analysis.

There are many nutrition plans that can be very healthy, especially in comparison to standard Western fare. Much of course depends on the individual, not only their biochemistry and neurobiology, but what they can live with for a wide variety of reasons ... culturally, economically, and so on. I personally read the evidence such as it is right now and see much to support a diet with 25 to 30% of energy needs from protein, enough plant material, inclusive of a wide variety of vegetables, root vegetables, fruits, nuts, and seeds to get an adequate amount both sorts of fibers, low SFA, high PUFA (especially n-3s), high MUFAs, and low in sodium and in highly processed foods in general. I am convinced by the data on the benefits of dairy, especially those with active cultures (fermented dairy). There are likely many contributing reasons for such to be a healthy diet and it is by no means the only path to a healthy diet. We are an omnivorous species and have adapted to a variety of niches.

Understanding how a hit with a hammer is perceived as pain is important, but of course it does not need to be known in order to know that one should duck if a hammer is being swung at your head.

Stephan Guyenet said...

Hi FrankG,

I followed the links you provided, followed the references they contained and saw nothing that supported the idea that insulin resistance occurs in muscle before fat in humans. The editorial on mouse studies reported on selective insulin resistance in mice, but that was because they were genetically engineered to lack insulin receptors in specific tissues. These papers suggest that insulin resistance in muscle (or the whole body) is the result of insulin resistance in adipose tissue and the resulting failure of adipose to trap fatty acids:

There are also the findings of Nir Barzilai's group that surgical removal of visceral fat prevents insulin resistance in aged mice, consistent with the idea that adipose is required for muscle insulin resistance to develop.

Suppression of circulating free fatty acids in insulin-resistant and diabetic humans greatly improves insulin sensitivity:

Here's how a recent review put it (
"Recent studies, however, suggest that the sequence of events may be the following: in states of positive net energy balance, triglyceride accumulation in “fat-buffering” adipose tissue is limited by the development of adipose tissue insulin resistance. This results in diversion of energy substrates to nonadipose tissue, which in turn leads to a complex array of metabolic abnormalities characteristic of insulin-resistant states and type 2 diabetes."

As I pointed out above, a complete lack of insulin signaling in muscle tissue only leads to modest fat accumulation in mice, so it's hard to see how that could account for human obesity. Also, if muscle tissue insulin resistance is the problem in obesity, then it implies that exercise should be the most effective solution.

Hi DH,

I would expect nose clipping to lead to fat loss, and anecdotally that's what people have reported on Seth Roberts' Shangri-La forum. But it has not been tested systematically to my knowledge.

Stephan Guyenet said...

Just to clarify, I don't think a failure of adipose tissue fatty acid trapping is the only cause of muscle insulin resistance. Lack of exercise is a major factor, and probably other factors as well such as inflammation.

Gretchen said...

Try this:

Diabetes. 1987 May;36(5):626-31.
Development of obesity in Zucker rats. Early insulin resistance in muscles but normal sensitivity in white adipose tissue.

It's a bit elderly, but anyone interested might be able to track forward to see if it's been refuted.

These were genetically obese mice, but not those IRKO mice, which I think are sufficiently pathological to make any simplistic conclusions from them difficult.

Don S said...

Okay. I am a pediatrician and my focus is of course biased towards understanding the antecedents of obesity and its prevention. I would appreciate some help trying to integrate the discussion of the development of insulin resistance to the specific circumstance of children who were born small for gestational age (SGA), and hence how that then may inform on IR development in others:

As Stephan noted in his insulin resistance series having been born SGA (presumably secondary to undernutrition in utero due to placental insufficiency) is a strong risk factor for later obesity and for insulin resistance indenpendent of obesity and at a level greater than would expected for the degree of adiposity.

The development of the IR and obesity in these kids has been studied. See

At first SGA kids have greater insulin sensitivity than do AGA kids. But "[c]onsequent to catch-up weight gain between birth and 2 yr, SGA children showed a dramatic transition toward central adiposity and insulin resistance between ages 2 and 4 yr" even with similar BMI's.

The authors' take leans to muscle IR before the adiposity (bolding mine):

"Accumulation of central and specifically visceral fat will lead to insulin resistance, possibly due to increased lipolysis and release of free fatty acids (23). Alternatively, the early development hyperinsulinemia resulting from a β-cell compensatory response to muscle-specific insulin resistance could positively feed back to promote further peripheral and central fat deposition. In support of this concept, experimental studies in low-birth-weight children and adolescents with insulin resistance, show that insulin sensitization with metformin leads to a gradual reduction of total and central fat mass (24, 25). In those studies, metformin also increased lean body mass while reducing circulating IGF-I levels, possibly indicating increased IGF-I sensitivity (24, 25). The opposite changes that we have now observed in 2- to 4-yr-old SGA infants (i.e. decreasing lean body mass despite increasing IGF-I levels) might therefore indicate that they have a relative IGF-I resistance. Indeed, defects in insulin and/or IGF-I signaling that are specific to muscle tissue could explain why SGA infants show lesser gains in lean mass but rather divert nutrients toward fat accumulation (26). Such metabolic abnormalities may be more marked in those SGA infants who showed rapid catch-up between 0 and 2 yr (7), and this might explain the link between earlier weight gain and subsequent gains in adiposity."

Growth hormone has been used btw to help achieve catch up height growth in some of these kids without, it must be noted, any effect on long term body composition compared to those with natural catch up growth - both have less LBM and greater percent body fat than do those born AGA on average.


CarbSane said...

@Don: I was surprised to discover that low birthweight babies are actually "fatter" than normal. This predisposes them to IR. You may find this article of interest to you:
Viewpoints on the Way to the Consensus Session. Where does insulin resistance start? The adipose tissue

@Gretchen: Odd how someone who criticized me for providing a blog post link -- that lead to a full text link to a relevant review paper -- would cite a study w/o even providing the PubMed abstract link. Unfortunately I don't find a full text even being available online.

In any case, this demonstrates the problem with such discussions. In human obesity we're talking insulin acting to stimulate esterification and limit lypolysis. This study looks at glucose transport into fat cells. So I'll bow out and work on that blog post on the many faces of IR. I'll post a link back for any who might be interested.

Stephan Guyenet said...

Hi Gretchen,

Thanks for posting that. Zucker rats are genetically leptin receptor deficient and thus are deficient in leptin signaling and obese for that reason. That demonstrates clearly the ability of leptin resistance to lead to obesity and insulin resistance.

It has been demonstrated in leptin deficient animals that leptin deficiency leads to insulin resistance both directly and indirectly. Indirectly, because it causes obesity, and directly, because a deficiency of leptin signaling in the brain directly causes peripheral insulin resistance. When you give these animals leptin in the brain, their insulin sensitivity improves immediately (actually, even in normal rodents insulin sensitivity improves when you give leptin into the brain) . The reason Zucker are insulin resistant is because their brains are leptin resistant and they are obese.

Also, in the paper you linked to, they didn't examine the ability of insulin to suppress free fatty acid release from adipose tissue specifically, which is what matters for peripheral insulin sensitivity. There is such a thing as 'selective insulin resistance', whereby some tissues become resistant to some of insulin's effects but not others (e.g. the liver). It would be interesting to see how well insulin suppressed lipolysis in those rats.

Don S said...


Not in this study anyway. From the article:

"At baseline (age 2 yr), there were no differences in body composition between SGA and AGA children. ... However, between ages 2 and 4 yr, the gains in abdominal fat and body adiposity were strikingly higher in SGA than AGA children; and the SGA children also showed lower gains in lean body mass [abdominal fat mass at age 4 yr, ... Therefore, by ages 3–4 yr, SGA children had greater total and abdominal fat and lower lean body mass than AGA children ..."

Again, also in this study at age 2 the formerly SGA children had greater insulin sensitivity than did those who were AGA and of similar BMI and body composition. (The authors do note that other studies have shown some formerly SGA children develop IR even by age 1.)

I appreciate the Viewpoint article however I lose the trail for the original source as that viewpoint cites another review as its source and that review is not available to me on-line other than as an abstract. The source may be this 1998 article that used skin calipers beginning at 2 months, and honestly I am not so comfortable using that approach as definitive. Yes there is loose skin, and lack of muscle mass in service of preserving organs and in particular brain growth, but I have not seen an original source that documents that there is relative preservation of fat tissue at the same time.

That said the search for such a source got me to this review which is very interesting. The effects on different glucose transporters and the different effects on them in different regions is fascinating. Also "expression of the insulin receptor is increased in IUGR skeletal muscle and tends to be increased at both the α- and β-mRNA transcript and alpha and beta chain protein level in the liver". BUT, there appear to be fewer myocytes AND those there are relatively resistant to growth. Thus increased insulin sensitivity leads to greater FFA even before high levels of adiposity and the limited number of myocytes get impaired quickly leading for them to becoming relatively IR within a few years after birth and even before high levels of adiposity. I think.

Not sure yet if this informs on anything else in this discussion!

FrankG said...

Here is another to add to the mix...

"[This] study demonstrates for the first time that insulin resistance already develops during weight gain within the normal range of body weight."

I read this as IR being present before significant excess fat mass.

I wonder if we are chasing our tails, or if our tails are leading us in circles?

I note in the "Viewpoints on the Way to the Consensus Session" subtitled "Where does insulin resistance start? The adipose tissue" that they go on to say "Adipose tissue releases a variety of factors known to modulate insulin sensitivity" so does this suggest that IR is present first in the Adipose tissue or that the Adipose tissue plays a role in causing IR in the other tissues? Perhaps subtle but it seems to me that the phrasing is open to interpretation.

There there is that word "consensus"... generally a red-flag for me ;-)

p.s. sorry I am not posting these as hyperlinks -- I am not sure of the syntax on this board... but hopefully you are internet savvy enough to simply select the text and then R-click launch the link or search on it ;-)

Stephan Guyenet said...

Hi Frank,

I acknowledge that there are many things yet to be learned about insulin resistance and obesity, and I don't claim to know all the details. However, that study again supports the idea that insulin resistance is caused by fat tissue accumulation and/or excess calorie intake. They deliberately overfed people, causing them to gain fat, and as a result they developed insulin resistance.

It's an interesting paper, thanks for posting it.

Gretchen said...

I think there's no question that obesity causes insulin resistance. However, that doesn't mean that IR could also cause (or contribute to) the obesity.

It's a vicious circle.

BTW, for a project I'm working on, I've been rereading several books on IR that I read long ago, and that Zucker rat study I came across was referred to as an "elegant study" by the chapter authors, which is why I thought it might be trustworthy even though full text isn't available online.

CarbSane said...

Nothing against you Gretchen, but whenever I hear the term "elegant study" I wanna run away! I don't know what that means, really. I find many such described studies are actually quite horrible. Keckwick & Pawan has been described as such ;)

FrankG said...

Thanks Stephan. Perhaps I could suggest a way to rephrase your above statement...

They deliberately overfed people, as a result they [simultaneously] developed insulin resistance and gained fat

As I keep alluding to with my dog chasing its tail analogy, and I understand that Gretchen is suggesting the same: it is a vicious cycle with the IR and fat mass feeding into each other... as we get closer and closer to when it actually starts (just like the big bang) it seems to become very hard to see exactly which happens first.

Even if we all agree that is starts with overfeeding, the question still remains "so why are we overfeeding?" It is not a natural thing to do.

FrankG said...

To be clear: I am not suggesting that we all agree that IR does [only] start with overfeeding... I just put it out there as a "for instance" to highlight how difficult it seems to be to pin down the start of what becomes this pathological process.

Don S said...

Y'all understand why the "What comes first the chicken or the egg?" question is so silly, right? When it comes to biologic processes asking "what comes first?" is as silly as asking where the first spot is on a circle.

What is clear both from the example of SGA kids, and from the study that Frank links to, is that non-adipose tissue IR can and does occur at levels of adiposity that are not considered overweight or obese.

So IR in some tissues, at least in baseline lean adult males without a FH of diabetes etc., occurs before overweight or obesity, BUT occurs in response to caloric excess driving some product, possibly FFA from adipocytes. Note that in this population of normally lean males without a FH of diabetes, etc., lean body mass increased as well as did fat mass.

In SGA kids it seems to be a larger than normal insulin sensitivity driving those adipocyte products that effect a smaller lean body mass and literally fewer myocytes, driving those cells from high insulin sensitivity to IR in short order, and also at a point of normal BMI and body composition.

This clarifies one aspect of the process for at least these two groups. It does not mean that the same process applies to those who carry genes that predispose to T2 DM, or to those who otherwise have a genetic predisposition to becoming obese in an obesogenic environment. It does not mean that even within those populations that insulin resistance changes are the most important thing going on let alone the only important thing going on.

Still, as for insulin and obesity having another nail in the coffin, perhaps this link is appropriate.;
Not dead yet. The simple view of it, sure, but as Dr. Hill puts it, some role in the symphony.

CarbSane said...

@Don: Near as I can tell, the nail in the coffin here is regarding the hyperinsulinemia (chronic high insulin levels) causes obesity part of Taubes hypothesis. The IR was a tangent introduced (to distract?). The KO mice were brought up, well the LIRKO is severely hyperinsulinemic yet remains lean.

As to IR and obesity, there is a theory out there of some sort of critical fat threshold. It is consistent with why some IR are thin. They have insufficient fat tissue to buffer much excess fat. This is similar to what Frayn describes in his review (the previous blog post I linked too).

I agree that the discussion can be meaningless to a degree, especially since it is impossible to ferret out exactly what some initiating event might be.

However, in GCBC, Taubes laid out a scenario in which IR led to obesity -- this one was based on excess carbs -> glucose -> fat stored in fat cells when muscles refuse it. Now with fructose, he lays out a different scenario where fructose causes IR -> hyperinsulinemia -> trapped fatty acids. It doesn't help matters that Gary will never set the record straight in writing so that his acolytes know what it is he's actually saying these days...

Don S said...

I suspect that there are many pathways to IR and to obesity depending on the exact circumstances that each individual is born with and into. The guru media approach lumps all of "obesity" as one thing with one cause but the states of obesity and of IR are more likely end states of multiple different pathways. That's the "systems biology" reference made by one author quoted earlier. ("Attractor basin" is another phrase that pops to mind to those of us who are fans of Chaos Theory.) That complexity sells few books however. Simple sound bites do.

Nevertheless we perhaps owe Taubes a debt of gratitude. His concept of GCBC moved the popular mindset away from thinking exclusively in that First Law of Thermodynamics, just take in fewer calories, mantra. Indeed it is true that all calories are not the same. From there we can better appreciate that not all fats are equal, not all PUFAs even, that not all carbs are equal, that some food sources satiate more than others, that some build or spare LBM more than others, and so on.

Those who are interested in dodging the hammer can continue to do what works for them, which for many is likely to be as it is for Frank, less refined carbs, less highly processed (also often with high salt and nitrates) food, and adequate protein (with many many variations about that) along with adequate exercise, and some of us will continue to read with interest as people like Stephan discover the roles that the brains reward centers (and individual variation of those centers) play and how they respond to different exposures; to others as they learn more about whether or not the relative resistance between different tissues and regions (including in the brain, and balances between various inputs at different points play roles in what states emerge and how what came before influences that; and to still others who study a wide variety of other possible contributing factors such as intestinal microbiome enterotypes.

The media gurus can argue over which one of them possess "the truth" that they are gracing the rest of us by revealing. The internet can continue to be full of acolytes who worship the book of Taubes, or the various books of "Paleo", or of Atkins. I would hope that those of us who appreciate science (including those who actually do science) will be more open to the reality that one state can have many different antecedents and enjoy the changing story as it slowly gets told.

Gretchen said...

@Don Amen.

FrankG said...

Well said Don - Thanks.

I personally take great exception to the use of disparaging terms like "acolyte" and "guru", which carry the clear implication that I, and/or others here, who may also have learned a great deal by reading GCBC etc.. are mindlessly following what someone else writes or says, based only on belief and trust in that person. Not to mnention the clear religious overtones.

Nothing could be further from the truth.

An huge lesson from reading GCBC for me was: to no longer accept the word of any expert, no matter how well accredited, without first examining their evidence-based proof for myself and applying critical thinking to their conclusions. Does it stand up to scrutiny? Was their methodology sound? Are there alternate interpretations for what they observed?

This actually came as quite a shock to me: having grown up in the UK in an era when "white-coated experts" on the BBC could be trusted implicitly.

I have learned to keep asking questions, to keep reading and to keep learning for myself. And yes that even includes the application of critical thinking when reading Gary Taubes, Stephan Guyenet etc...

Applying this to of "coffins" and "final nails" I was told above that...

"The fact of the matter is that Stephan and others like myself, have looked at the enormous body of the scientific evidence, not a few cherry picked studies and outdated texts, and demonstrated time and again that the carb->insulin->obesity hypothesis is not a workable one."

Does this stand up to scrutiny? Straight away I am on guard with the obvious appeals to authority. But beyond that Don, as you[1] and I [2] had already pointed out in previous comments: the studies presented in this blog post do not qualify as their are claimed to be, as models of chronic hyperinsulinemia in humans with IR.

Which leaves me wondering... with access to an "enormous body of the scientific evidence" how is it that these easily dismissed studies are offered as "final nails"? Surely there must be more rigorous studies that do stand up to the scrutiny of even an amateur scientist such as myself? What am I to make of this?

No doubt the constant nagging questions of someone like myself could be irritating to some scientists but I think that real science is driven by questions. Isn't it?

"That is the essence of science: ask an impertinent question, and you are on the way to a pertinent answer." - Jacob Bronowski



Don S said...


Jacob Bronowski!! He's my personal hero!

The Ascent of Man on PBS when I was a kid was probable the single most important influence on me outside of my family. The book of the show is still available, and so are many episodes on YouTube - this one "Knowledge or Certainty" may be of particular interest to you. So are many books of his collected essays and lectures. Good stuff. Mathematician, scientist, Blake scholar, poet, historian ... probably one of the last great Renaissance men, true intellectuals who knew about ... everything.

You know how the CrossFit and multiple other fitness programs promote intense training in multiple modalities - Olympic lifts, gymnastic moves, jumping rope, rowing, running, so on? Bronowski was the poster boy for CrossFit of the mind:

"Man is unique not because he does science, and his is unique not because he does art, but because science and art equally are expressions of his marvelous plasticity of mind."

FrankG said...

Thanks Don -- he was an incredible man with an incredible mind. He and Einstein seem alike in that despite their immense intellects they never lost their childlike wonder about everything!

I have the entire series of The Ascent of Man, a classic that has more than stood the test of time.

So many quotes but one of my favourites, that I repeat to my son at University...

"It is important that students bring a certain ragamuffin, barefoot irreverence to their studies; they are not here to worship what is known, but to question it."

Gretchen said...

I'm trying to learn more about leptin and came across an interesting paper

Yin Yang.

that is a little more current than the 1980 paper cited at the beginning of this post.

Reading it gave me an idea. I think part of the controversy on this issue is that there are two different actions of these hormones. One is on fat metabolism, and the other is on appetite, and they aren't always linked.

Thus, insulin might inhibit appetite while at the same time increasing fat synthesis. In fact, the authors cite experiments in which "chronic ICV infusion can increase fat mass in mice without affecting food intake."

This is the problem many obese people face. If they gain weight, their doctors assume they're eating more, and it's not always true.

It's another cause vs effect situation. Let's say insulin increases fat. Simultaneously, because there's more fat available, it decreases appetite.

That would make sense to me.

I think Taubes is talking about fat metabolism and Stephan is talking about appetite. No wonder they can't agree.

Sanjeev said...

I doubt JB ever meant

"ask leading questions meant to push your agenda, not with a genuine desire to learn, and after all is said and done, smugly pat yourself on the back for asking questions"

Just my interpretation, but who am I to say ... as someone who asks impertinent questions of my own beliefs, I'm not likely to get a sore elbow from patting myself on the back.

Sanjeev said...

Maybe the low carb proponents should take control back from Taubes.

What is Taubes' stance? Personally I don't know.

read the piece of GC,BC that motivates the entire obesity discussion and Taubes writes:

GT> ... they eat no more than the lean—surprising as it seems, the
GT> evidence backs this up ...

GCBC chapter 14 paragraph 3

GT> The more closely we look at the evidence and at obesity itself,
GT> the more problematic the science becomes. Lean people will often
GT> insist that the secret to their success is eating in moderation,
GT> but many fat people insist that they eat no more than the
GT> lean—surprising as it seems, the evidence backs this up—and yet
GT> are fat nonetheless. As the National Academy of Sciences report
GT> Diet and Health phrased it, “Most studies comparing normal and
GT> overweight people suggest that those who are overweight eat fewer
GT> calories than those of normal weight.” Researchers and
GT> public-health officials nonetheless insist that obesity is caused
GT> by overeating, without attempting to explain how these two notions
GT> can be reconciled. This situation is not improved by the
GT> prevailing attitude of many nutritionists, obesity researchers,
GT> and public-health authorities that it is evidence of untoward
GT> skepticism to raise such issues, or to ask questions that lead
GT> others into contemplating the contradictions themselves.

sounds more like fantasy physics than metabolism, specifically, sounds like he's claiming the obese create mass out of of vacuum.

Not even Pons and Fleischmann went that far. They thought they were turning mass into energy, not creating new mass.

And elsewhere (not in the place that sets the stage for him to make his argument ... like any propagandist, he tries to hook people using masses of picked cherries before introducing or admitting any kind of doubt) he writes that CICO must be true.

Sanjeev said...

So I must again wonder, what is Taubes' stance? The most accurate description might be
"anything to make myself right"

"anything to sell my fruity ideas ... tell scientists CICO must be true (but diminish it using any specious argument I can think of) and tell obese people insulin is creating mass out of vacuum"

five years of research and he never came across doubly labeled water studies.

I've noted this before ... and I later found out James Krieger wrote about the above quote long before I did - James' comment:

JK> When Taubes quoted that piece from the NAS report, he left out
JK> some important adjacent sentences about underreporting
JK> (underreporting was suspected back in 1980 until it was confirmed
JK> using doubly labeled water in the 1990's). This is consistent with
JK> CarbSane's observations of how Taubes consistently cherry picks
JK> from his own references.

I've yet to read or hear the words "doubly labeled water" from Gary Taubes.

I've yet to read or hear Taubes retract or clarify that mangling of the NAS statement.

But why talk about old stuff (even though he's still pushing it sometimes) ...

what does Taubes believe now? I for one don't know.

He said recently in a radio interview

GT> "... no sugar you might get away with it like the Asians do ..."

so ... what IS taubes' claim, exactly?

Why would any experimenter set out to test Taubes' ideas when he's so willing to take a new stance that contradicts the old stance yet maintain the facade of the old stance?

Others can't see eye to eye with Taubes because others are curious and open and applying scientific methods and skepticism but he's talking out of his ass.

this is my conclusion ... nothing to do with Stephan or James of course ...

> I think Taubes is talking about fat metabolism

This does not comport with the available evidence. He's spewing his own fevered fantasy.

If you think you know what Taubes stands for,

check for yourself

Around 4:50: The worst is sugar if you never ate sugar at all you could probably eat these other foods ... the rice ... the pasta ... just like the SouthEast Asians can stay lean eating rice heavy diets they do it because they eat virtually no sugar compared to what we eat in the US

in his own words, mp3 format listen for yourself the actual mp3 is here

> I think Taubes is talking about fat metabolism


Extremely so.

Sanjeev said...

> I think Taubes is talking about fat metabolism

doubtful that that's what he's talking about.

robrob said...

very interesting article, it makes me less afraid of insulin. lol.

I think the reason low carbs works is not because of effect on insulin at all. but rather the fact that a person takes in more saturated fats (thus taking in more vita d, k, a, calcium etc that require fat for absorption) which does two things. the fat relieves the fat cells of the job of making fat for the cells to use since they can't use glucose very well. there job is to make fat out of glucose when glucose is always high so the cells have sufficient fat to fuel the.

we all know how type one can still burn fat. they just can't use glucose very well and their livers don't have insulin to shut down the glucose release. so they pour out glucose despite a already high glucose level and this glucose is available for fat cells to make fat. osmosis as it were.

second these vita/min are required for proper glucose tolerance. if the cells cannot get the gluocse or enough on time then it is as if your suffering low blood sugar. hence your survival kicks in and you get the hungries, the jitters etc especially at night or a long fast.

you do know your body doesn't need insulin to use glucose or to make fat do you? the hyperglycemia is due to the liver pouring out glucose due ot lack of insulin or resistance to it. insulin just speeds up the uptake.

insulin turns the breaks on the liver. if the cells are resistant to glucose uptake the brain controls that amount of glucose keeping it higher than normal so as to use high concentration to get the glucose in via osmosis. the cells including the brain, if they are cholesterol deficient, calcium vita d and othe vita/min defecient they will not handle gluocose and hence don't want it as gluocse damages cytoplasm without their lipid rafts. tho depseration forces them to do so like in exercise and needing some glucose to burn fat better.

fat cells job is to recycle cholesterol, to store it when dietary sources are unreliable due to low fat high carb diets, and in order to store it it must store fat too to buffer the choelserol.they also hoard calcium too when dietary sources are low. thus the more malnorished you are the fatter you become unless you can't get sufficient calories at all. obesity is simply a malnorishment problem. part of that is lack of sun exposure for vita d production due to sunscreen use and low sulfer low cholesterol diets. the more fat/cholesterol/glucose you need and have the more fat cells you need to manage it all especially if the fat cells are malnorished themselve, you know the work is lighter with more hands.

low carb diets also releive the body of damaging effects of excess glucose ot the body from glycation. this alone would decrease inflammation and make the body work better.

but low carb diets are not meant to be followed forever they are just for restoring the nutrient status of the body and relieving if of the damaging affects of glucose and fructose that stays around and sticks to everything and oxidizing it.

so I am finding you do needs carbs one for the vita/min phytochemicals etc but also for the insulin surge you need to signal the brain your nutritional status.

you can't get alot of the nutrients you need from meat and dairy and nuts and seeds only. I don't know anyone who can eat enough vegetables to compensate not eating fruits and whole grains.

ketosis may also be a signal of winter time and that the food enviorment is not optimal.ketosis is good for short term but I wouldnt want to rely on long term it just might send the wrong message to the brain and cause problems in the long run. of course this is just my opinion.


Puddleg said...

Some nail. Some coffin.
Carbohydrate is fattening for the susceptible, vege PUFA and fructose prepare the terrain, so that there are more susceptible people than ever.
Insulin plays a pivotal role in the scenario I have just described, which is the everyday familar scenario in modern humans.
How closely does the rat experiment model the human scenario? Very little.
Insulin gives us a convenient way to explain the effects of the fattening carbohydrate, but we can leave it out altogether (as Dr Riichard Mackarness did in Eat Fat to Get Slim) without altering the facts of the case.
All hormones are dependent on context; in the fattening context, insulin has a fattening role.

SM said...

This experiment does not strike me as valid. The high-insulin condition of interest in humans (in metabolic syndrome) is:

1) high insulin normal BG (early stages of MS)

2) high insulin slightly elevated BG (pre-diabetes), and

3) high insulin and high BG (frank diabetes)

This tested none of those. When you simply introduce extra insulin into a working system, you will have some extra insulin and LESS than normal BG so this does not mimic the condition of metabolic syndrome, does it?

Roger L. Cauvin said...

Petro Dobromylskyj has posted his analysis of the Vanderweele study on his Hyperlipid blog: