Monday, 7 January 2013

It’s the Calories, Stupid.

I thought I'd mark my return to blogging by taking the piss out of a certain Diet Doctor for his post It’s the Insulin, Stupid, who takes (and tweaks) Fig 7A from Hyperinsulinemia Drives Diet-Induced Obesity Independently of Brain Insulin Production.

At first glance, Fig 7A looks like a CIH believer's dream come true (apart from the words "High Fat Diet").

Hyperinsulinemia → Obesity.

Obesity is caused by too much insulin. Game, Set and Match to insulin.

Not so fast! Let's take a look at the rest of Fig 7.

Figure 7. Revisiting the Current Model of Obesity and Type 2 Diabetes(A) The most widely accepted model of the pathogenesis of obesity and type 2 diabetes posits that a high-fat diet leads to obesity and insulin resistance (there is debate about the relative order and causality of these). In this widely held view, insulin resistance then leads to hyperinsulinemia, which is followed by β cell exhaustion, and then type 2 diabetes. The accepted model is incompatible with our results that put the insulin hypersecretion genetically upstream of obesity.(B) Our data support a model whereby insulin levels must be kept low to maintain energy expenditure in white adipose tissue via the expression of Ucp1. Our data do not address the order of subsequent events after obesity (outside the yellow box), such as insulin resistance and/or type 2 diabetes, since they were not observed in our studies. In other words, the effects of insulin gene dosage on obesity are independent of sustained changes in glucose homeostasis or insulin resistance.

↑ Peripheral Hyperinsulinemia → ↓ Uncoupling Proteins (WAT) → ↓ Energy Expenditure → ↑ Obesity.

Obesity is caused by a reduction in energy expenditure in these mice. Game, Set and Match to The Energy Balance Equation. It’s the Calories, Stupid. In these mice, energy expenditure is strongly influenced by insulin levels. In humans, not a lot. In humans, insulin can act as a stimulant or a sedative.

I'm not an insulin denier as is obvious from my other blog posts. I'm still restricting my carbohydrate intake to ~125g/day from whole foods.

I'm not a food reward denier. I've been using food reward principles to lose weight.

This post will probably annoy some people. Before wasting your time writing a comment, please read my Moderation Policy.


JJ said...

Hi Nige...

I wrote the paper you are commenting on and I have been following it around the internet as it gets discussed/interpreted/misinterpreted. Insulin had multiple effects on the adipose tissue, including regulation of lipolytic genes. I think the evidence that insulin is important in obesity is strong. I don't see any reason why insulin's effects can't include effects on expenditure as well.

As for humans, no one has any idea what CHRONIC (life long) reduction in insulin hypersecretion would do the Ucp1 in WAT. The ACUTE (minute to minute) effects of insulin on EE are less important here.

I think a lot of people are missing the fact that the experiment represents 1/3 of the mouse lifetime and starts just after weening.

ProudDaddy said...

Welcome back! Missed you!

On my end, we're expecting a son in May! I'll be 72. Methinks your clear thinking can help me enjoy this unexpected adventure.

Nigel Kinbrum said...

Hi Dr. Johnson. Thank you for your comment. You've probably noticed that insulin is a very emotive subject in the blogosphere! I am of the opinion that obesity is multi-factorial and that insulin is involved.

Humans have a much higher mass:surface area ratio than mice, so we can't increase energy expenditure by heat production by the same proportion as mice. See Of mice and men, Kleiber's Law & FIRKO.

Nigel Kinbrum said...

Hi Wayne.

What have you been getting up to? On second thoughts, don't answer that. Good luck!

Clear thinking? I wish.

P.S. I found your PM on Facebook. It was in "Other" instead of "Inbox".

JJ said...

Hi Nige,

You can just call me Jim. No need for formality - its the internet :) And yes, I have noticed that insulin can rile people up in these circles. I honestly had no idea and I still don't really understand why though.

I think your interpretation that insulin is multi-factorial is obviously the correct one. I think if you ask any serious academic in any of the related fields they would agree. I should also stress that the people in the know know that there is tons that we don't understand about insulin - which is why we do the research still :-)

The literature is a never-ending, living-breathing entity.

Regarding mice and men. I agree on the size difference :-) To me it just means that we cannot make physiologically-relevant heat in hour WAT and BAT at the same rate as mice. However, we also live longer so we have decades to accumulate and lose pounds. Mice can gain and drop 5% of their weight in days.

I think there is still a lot we don't know about 'biochemical' energy expenditure in humans, in the long term. Heck, a few years ago almost nobody believed adults have BAT, but now I think it is well accepted. Same goes for browning of WAT - we are just at the very beginning of learning what is going on.

My friend Andre Carpentier is doing some great stuff on energy burning in fat up in Quebec.

Nigel Kinbrum said...

Hi Jim,

I think that people get a strong emotional attachment to a way of eating that makes them lose weight, feel much better etc. I certainly did in 1997 when I tried the Atkins Diet. If blogs had existed back then, I would have been blogging about the Atkins Diet and trying to get everybody on it.

Simon Carter said...

Nigel, I'm glad to see that you are back! I've been reading through your blog and I have enjoyed and learned from it. As a fellow engineer I like your approach. Dr. Johnson has also been a welcome breath of fresh air to blogworld.

Unknown said...

Glad to see you back, Mr. Kinbrum.

So here's one to you and Dr. Johnson (I lack the permit to call him Jim.)

Now I know that Dr. Johnson only discusses the results of his study and doesn't endorse any dietary ideas or interventions. However, is it possible to find agreement on the idea, calories aside, that it is good to take measures that make insulin secretion and utilisation a tight and efficient process for long term metabolic health (even beyond weight gain/loss)?


JJ said...

Hi Kade,

Everyone can call me Jim :-)
No permit necessary :-)

" is good to take measures that make insulin secretion and utilisation a tight and efficient process for long term metabolic health (even beyond weight gain/loss)..."

I agree. I think that is about as accurate and succinct as it can be stated. I would always like to add that it is critical to not restrict insulin past the point of glucose-intolerance. I would not want people with diabetes to forego the amount of insulin required to maintained glucose homeostasis and reduce long-term complications.

On that note, there is emerging evidence from our Vancouver group (led by Dr. Warnock) and from others that islet transplantation can be more effective than best medical therapy (insulin injections) for reducing the rate of onset of several complications. The current thinking is that real beta-cells can respond more efficiently, subtlely and discretely than even the most advanced insulin users.



Nigel Kinbrum said...

Hi Kade,

You can call me Nige - its the internet :)

I see that Jim has replied while I was writing this...

My answer: Overeating just about anything (except perhaps pure fat) produces excessive insulin secretion. Even ItsTheWoo has stated thusly.

Therefore it's a good idea to avoid things that can cause chronic overeating. It may be foods that produce excessive reward for some people. It may be foods that result in unstable blood glucose for some people. It may be marketing for some people. And so on...

Unknown said...

Thanks much for your response, Jim. I had a feeling that you'd say the same.

And Mr. Kinbrum--, I mean Nige. Don't mine me, I jest a bit from time to time. Thanks again for your views. I asked you this question because you closed your post on the idea that you still work towards keeping an eye on calories, but also controlling your carbohydrate intake.

Now your response resonates well with people in various camps, but that is in the context of weight and overeating.

I know that what I'm about to bring up is often a rare scenario, but sometimes, one can come across people who are far from overweight but still exhibit hyperinsulinemia in response to a particular diet or food choice. Also very true that aside from pure fat, everything will instigate an insulin response, but I do think the level varies substantially between 'isoenergetic' servings of different foods (I'm guessing that you've had a look at this insulin index table: I assume that one could be a bit clever with food choice--as you are with carbohydrate control--to reduce insulin secretion, which should have a beneficial effect.

Nigel Kinbrum said...

Hi Kade,

For all of my life I've eaten too much for my activity level and have been over-fat to varying degrees. A low-carb diet helped me to reduce my food intake and lose some flab, but it didn't make me slim. I'm now tackling other factors that make me overeat.

In slim people, is symptomless postprandial hyperinsulinaemia a problem? If there are symptoms, eating in such a way that minimises insulin spikes is a good idea. I am familiar with the insulin index.

Unknown said...

"In slim people, is symptomless postprandial hyperinsulinaemia a problem?"

This is a good question, and probably what I was trying to lead into with my original post.

I would defer to the ones doing the research, such as Jim, and those who have actually torn their way through the mountain of studies, such as yourself, Nige. Although if I had to have an instinctive answer, it'd be yes. I mean, sometimes, it takes a long while for symptoms to show. I just don't think that any hormone, especially one as central as insulin, should be pushed to excess through diet or life style choice, overweight or not.

Nigel Kinbrum said...

Although I think that the occasional large insulin spike is harmless, I think that multiple large insulin spikes a day every day is probably not.

I also think that the biggest threat to diets that don't produce large insulin spikes (low-carb, low-reward, paleo/real food etc) is Big Food. Maximising consumption of crap-in-a-bag/box/bottle is their no 1 priority.

Unknown said...

Agreed, Nigel. Occasional spikes, especially postprandial, can be overlooked. It's the general state of hyperinsulinemia and consistent postprandial over secretions that I think can be problematic.

Having that said, and not trying to argue for any one diet, I will add that proper whole-foods low-fat diets can also achieve improvements with fasting glucose, lowered insulin, etc., much like low-carb, plaeo and low-reward plans. It's usually because one will find it very hard to consume chronically high levels of calories on a whole food diet and with a healthy metabolism a genuinely high carbohydrate and low fat diet would result in some level of excess calories getting oxidised for energy, even if we're talking about those 'insidious starches'.

Unknown said...


I have a quick question about your blog, do you think you could e-mail me?


Nigel Kinbrum said...

Hi Heather,
I'm away from the laptop at the moment. You can email me.

Puddleg said...

Oh, if you overate pure fat, this might happen:

To deal with the excess energy some of the glycogen is converted to glucose, blood sugar is elevated, insulin rises, fat is stored.

You can get high fasting BG on a VLC diet. When this happens, why wouldn't the CIH then apply?

"Sometimes life finds a way."

Puddleg said...

Good to have you back.

This paper gives CIH a role in context. it's not so much about DNL as the Feinman doctrine; "dietary carbohydrate determines the fate of dietary fat".
It is however about energy partitioning (what is used, what is stored) not obesity. Not exactly the same thing, even the lean need to store and burn triglycerides.

Nigel Kinbrum said...

Hi George,
"Oh, if you overate pure fat, this might happen:"
It might, but I don't think that it does. See Figs 1 and 2 in Extended effects of evening meal carbohydrate-to-fat ratio on fasting and postprandial substrate metabolism. A 40g oral fat load has virtually zero effect on serum glucose or serum insulin.

Diana said...

The takeaway that I get from this is that as a human being who wants to lose a few pounds, I should not eat a high fat diet. Whatever the role insulin plays, that's clear enough. Or is it? Nigel, help me.

Diana said...

George I hate to tell you this but J.P. Flatt said " "dietary carbohydrate determines the fate of dietary fat" first.

Nigel Kinbrum said...

Diana said...
"The takeaway that I get from this is that as a human being who wants to lose a few pounds, I should not eat a high fat diet. Whatever the role insulin plays, that's clear enough. Or is it? Nigel, help me."

Hello Diana. The takeaway that I get from this is that the best way to make C57BL6/J mice hyperinsulinaemic, sluggish and fat is to feed them with Research Diets D12492 (or whatever was used in the study - I can't read the Supplemental Information).

If a high fat diet reliably makes humans hyperinsulinaemic, sluggish and fat, I would agree with you. However, it doesn't. Many humans spontaneously either reduce energy intake or increase energy expenditure, or both.

I have no idea what happens if humans eat Research Diets D12492.

See High-Fat Diets, Obesity and Brain Damage.

Diana said...

LOL, Nigel.

Regarding high fat diets & cognitive function:

"Results from our work in rats and others findings from human epidemiologic studies demonstrate deficits in cognitive performance following chronic ingestion of high fat, high saturated fat, diets. "


If you search pubmed for "high fat diets" and "cognitive" you find a lot.

If everyone spontaneously reduced intake on a high fat diet, so many of us wouldn't get fat. I sure didn't.

Nigel Kinbrum said...

Hi Diana,

I'm out and typing this on my phone so I'll be brief. If everybody had the same response, life would be so much simpler!

Puddleg said...

Thanks Diana. Feinman often quotes the "Flatt doctrine" (then) as one of the useful principles in this area.
It is useful when gently reminding people why low-fat diets can actually work sometimes, why restricting junk fat (e.g. in deep fried food) is a worthwhile thing even for low-carbers, and so on.

Unknown said...

Epidemiological studies have their short comings, and everyone's not an insulin resistant type 2 diabetic. Still an interesting topic with regards to cognitive functionality.

"If everybody had the same response, life would be so much simpler!"

If only.

Nigel Kinbrum said...


For humans in the US/UK, "high fat diet" usually means including manufactured foods like take-aways, restaurant food, ready meals, cakes, biscuits, ice-cream etc.

If instead, "high fat diet" means including natural foods like meats, oily fish, nuts, seeds, cheese etc, would the result be the same?

A natural "high fat diet" is less likely to result in long-term positive energy balance. It's long-term positive energy balance that causes insulin resistance in most humans.

In rodents, D12492 causes insulin resistance very quickly.

Unknown said...

Nigel, do you think that the results would've been substantially different if they had used the foods you listed as part of their high-fat-low-carbohydrate diet?

Nigel Kinbrum said...

Hi Kade,

That's interesting. The brain doesn't use FFAs as a fuel source, so serum FFA level should make no difference to brain function.

Each diet lasted for 5 days. After 5 days on a low-carb, high-fat diet, serum glucose falls from ~5 to ~3.5mmol/L and serum ketones rise from 0 to ~4.5mmol/L. The brain has to adapt to a shift in fuel usage. This can take up to 14 days. This may explain the reduced cognition.

I don't know the significance of 9% lower cardiac PCr/ATP (P < 0.01). This may be due a shift in fuel usage. There was no change in cardiac function.

Puddleg said...


I don't think the paper you cited refutes my suggestion.
We're not talking about healthy volunteers, or people eating high-fat diets in general, but people eating high-fat diets who have high fasting BG and who gain, or don't lose, weight.
The first question is, in this scenario, can endogenous glucose mimic dietary glucose and store fat via CIH mechanisms (DGAT upregulation)?
The second would be, where does the glucose come from (triglyceride glycogen, dietary protein, or fasting amino acid breakdown)?

BTW, brain does use FFAs, but they have a much longer half-life than glucose or ketones.

Puddleg said...

About the "high fat diet" - the actual diets fed to mice resemble nothing so much as a fatty pudding mix.
Sucrose, maltodextrin, casein, lard, soybean oil with a few vitamin, mineral and amino acid supplements.

You could probably make icecream with that, but you couldn't make a nut or a steak.

Carbsane, in a freakish moment of lucidity, suggested that cellular (intrinsic) fats may be different from acellular (extrinsic) fats, in a similar way to cellar/acellular carbohydrates (a la Spreadbury).

This strikes me as very likely true, with dairy fats being intermediate due to their packaging in cholesterol, phospholipid and casein, and fermented dairy re-introducing cellular elements.

Nigel Kinbrum said...

George, I'm going out for the evening soon so I don't have time to discuss the first part.
"BTW, brain does use FFAs, but they have a much longer half-life than glucose or ketones." Directly? This is news to me. Source?

Puddleg said...

This is an example:

The brain has lipogenic enzymes and carnitine. It contains a large amount of fat (some synthesized from glucose, some delivered by lipoproteins) which needs to be turned over. If not, we wouldn't need to eat DHA once our brain stopped growing. Fatty acid oxidation is upregulated in schizophrenics.
(This paper makes it obvious that beta oxidation and carnitine shuttling are normal functions of brain tissue.)
Fatty acids don't supply enough energy that the brain is ever able to rely on them, but they do contribute.

Unknown said...

Hi, Nigel.

Thanks for the response. I myself don't like the length of the study. But I do wonder if long term, such a diet--with the same high fat ratio--made up of whole food sources such as meat, cheese, fish, etc., would have a different result.


Yeah, mice diets are crapolla pudding. It's what works to break their hypothalamus - get them hyperinsulinemic - etc.

Would be great to read about FFAs as fuel for the brain and having a longer half-life.

Nigel Kinbrum said...

George, if FFAs are a minor brain fuel source, would raised serum FFAs have any effect on cognitive function?

Nigel Kinbrum said...

"We're not talking about healthy volunteers, or people eating high-fat diets in general, but people eating high-fat diets who have high fasting BG and who gain, or don't lose, weight.
The first question is, in this scenario, can endogenous glucose mimic dietary glucose and store fat via CIH mechanisms (DGAT upregulation)?
The second would be, where does the glucose come from (triglyceride glycogen, dietary protein, or fasting amino acid breakdown)?"

High fasting BG on a low-carb high-fat diet may be caused by liver IR, or maybe something like this.

Stable or rising bodyweight suggests that (ingested + synthesised) fat =/> burned fat, resulting in a zero/positive net influx of fat to adipocytes.

As to where endogenous glucose production comes from, see Liver and Kidneys Synthesize Glucose, though this is about starvation rather than low-carb high-fat diets.

So, does the body synthesise excessive glucose from various substrates (including fat), then synthesise the excess back into fat? I think it's possible, but on a low-carb high-fat diet, dietary fat >> synthesised fat.

Puddleg said...

The question is whether the excess endogenous glucose can drive fat accumulation through the same CIH mechanism that is proposed for dietary glucose, or at least slow down lipolysis.

That post of yours on high BG you link to above is something I often think about. It makes sense to me. Insulin inhibits gluconeogenesis and glycogenolysis and that can come in handy, a little starch might help keep glucose down.

I don't know that serum FFAs get to the brain, though the modern idea is that the BBB is more porous than first thought. But phospholipids, lipoproteins and DNL fats turn up there and probably exist as FFAs before they are oxidized.

Cellular energy metabolism during fetal development ☆: VI. Fatty acid oxidation by developing brain

These data suggest that during periods of high fat intake in the suckling rat the brain has an increased capacity for long chain fatty acid oxidation and that in addition to ketone bodies and leucine, fatty acids may be utilized as an alternative substrate in developing brain.

Nigel Kinbrum said...


Kade posted the study which concluded "Raising plasma free fatty acids...reduced cognition...

Sounds sus to me!

Unknown said...

People aren't into anecdotals, but I've come across quite a few who seem to exhibit improved cognitive function and calmer mood states through carb-restriction with whole food choices. This doesn't mean binging on highly neurotic extreme high fat diets, but just a general mix of meat, eggs, vegetables and some tubers, which automatically turns into a substantially carb-restricted low-carb diet.

Nigel Kinbrum said...

Kade, I know of similar experiences. I am very sceptical of short-term (5-7 days) human studies on high fat diets.

RE High fat diets for rodents: Here's a D12492 data sheet. Although sucrose is only 6.7% of total calories, it's 8.89% of total weight. This makes D12492 quite sweet.

If high fat diets always make rodents fat, then "Ratty" (~70% of total calories from fat) should have been very fat. It wasn't. I'm pretty sure that Peter didn't feed his pet rat any maltodextrin or refined sucrose. Or large amounts of trans-fats (found in other high fat diets for rodents).

Unknown said...

Of course, Nigel.

I mean, it goes without saying. A mere week is too little a time frame to conclude anything. Any kind of major dietary change will come with transient shift in health markers that could very well be the result of a host of other confounding factors at play.

As for the crappy mouse & rat diet point. Generally I am in agreement, but I will also play Devil's Advocate here and say that while the diet is sweetened, and potentially palatable, it's still not high enough in my opinion to have a profoundly meaningful impact--it's something but it doesn't seem to be *the* x-factor. My opinion's also the same when it comes to arguments against PUFAs these days, which have been known to provoke people into reducing their dietary options substantially while insisting that the--still persistent--problems are largely because of those last few inconsequential grams of PUFA.

Now a general mix of nasty industrial fats with dollops of sugar can certainly spell disaster.

Puddleg said...

That study is useful to me.
Here's the fulltext:
Dietician meals so probably lots of heart-healthy oils and spreads.
It's interesting that neither group improved; they both missed their usual diet, but the high-fat group missed it more, because their diet had changed more.
They would be getting less glucose than they were used to as well as more fat.

At the rate brain burns fats, even if it could use serum FFAs there would be no possible benefit from elevating them.

It just confirms what we all know; that sudden drastic changes to the macronutrient composition of the diet have cognitive effects.
you can't just change the habits of a lifetime and expect everything to be fine right away.

It was interesting that their mood stayed stable despite the stressor of the diet switch.

Puddleg said...

Some more thoughts - why were FFAs elevated and not just triglycerides?
Lipolysis from lower insulin levels - evidence for the CIH.

What will happen after 5 days?
FFAs are ligands for PPAR-alpha. PPAR-alpha will be upregulated, FFAs will be burned, ketones will be produced at higher levels, FFAs and cognition will normalise.

Ketoadaptation, birth of the mythical fat-burning beast.