Welcome all seeking refuge from low carb dogma!

“To kill an error is as good a service as, and sometimes even better than, the establishing of a new truth or fact”
~ Charles Darwin (it's evolutionary baybeee!)

Saturday, April 9, 2011

GCBC Reference Check ~ Part IV of ? ~ Kipnis

In a recent post, following up on discussion of Grey & Kipnis as  in the Fasting Insulin & Weight Loss post,  I noted that G&K was  but one study (of many) where calories were controlled that demonstrated definitively that it was calories, and only calories, and not carbohydrates or any hormonal effect thereof, that determined weight loss/stability.

Taubes made the slanderous claim on his blog:
... the value of controlling variables in a scientific experiment is something that a reasonably well-educated child supposedly understands. And what I want to know is why don’t nutritionists understand it and those researchers out there doing diet trials and studying obesity and weight regulation. Because their failure to do so — and I would argue that it may be a willful failure — has led to what may be another of the great misconceptions in modern nutrition research.

I ended my post with the following:

Now I realize 1971 is a bit too contemporary for Mr. Taubes, but since he did reference a text published in 1972 he could have found this study were he interested.

I think Mr. Taubes owes Drs. Grey and Kipnis an apology for his unprofessional, unethical, and downright below the belt assaults on their scientific abilities and integrity! Don't you?
In a "gotcha" moment, regular commenter Christian had this to say:
Hmm, the only problem is that
a) Taubes read the study and
b) also interviewed the researcher and
c) the researcher himself comes to a different conclusion regarding his own study than you do.

Let me begin by saying that, although I've read GCBC, and some sections multiple times, my many waking hours spent stalking {} have kept me from my GCBC Reference Memorization drills.  It seems to me that G&K is just the type of study Taubes laments will never happen, only it did 30 years earlier.   At this point I'm not surprised that he came across it, and even cited it.  It's been all downhill from Reshef when it comes to looking into Gary's references.

To (a):  I would point out that, by his own admission, citing a study or reference is no indication that Taubes has actually read it, let alone viewed it with a critical eye.  (Apparently one of the things I got Gary to do was actually read Keith Frayn's text!  See, I'm useful for something after all.  Phew!)

The funny thing about G&K is that it's a rather poor quality scan of an old article.   So more than with the  fancy PDF's and nice HTML texts - where you often have to click to view tables - these graphs practically leap off the pages at you.  If you even skim the article, knowing nothing of the dietary interventions you are presented with fasting insulin vs. bodyweight.  Graphs that CLEARLY show no correlation!!

So let's see what Gary Taubes had to say about this study in GCBC.  I've included the two paragraphs preceding his mention of this Grey & Kipnis study for context:  

Since none of this research is particularly controversial, it’s hard to imagine why obesity researchers would not take seriously the hypothesis that carbohydrates have a unique ability to fatten humans—or, as Thomas Hawkes Tanner put it in The Practice of Medicine almost 140 years ago, that “farinaceous and vegetable foods are fattening, and saccharine matters are especially so.” Researchers who study carbohydrate metabolism have found this science compelling. In 1991, the Belgian physiologist Henri-Géry Hers, an authority on what are known as glycogen-storage diseases, one of which is named after him, put it this way: “Eating carbohydrates will stimulate insulin secretion and cause obesity. That looks obvious to me….” But this simple chain of cause and effect has nonetheless been rejected out of hand by authority figures in the field of human obesity, who believe that the cause of the condition is manifestly obvious and beyond dispute, that the law of energy conservation dictates that obesity has to be caused by eating too much or moving too little.
George Cahill, a former professor at the Harvard Medical School, is a pedagogical example. Cahill had done some of the earliest research on the regulation of fat-cell metabolism by insulin in the late 1950s, and had coedited the 1965 Handbook of Physiology on adipose-tissue metabolism. In 1971, when Cahill gave the Banting Memorial Lecture at the annual meeting of the American Diabetes Association, he described insulin as “the overall fuel control in mammals.” “The concentration of circulating insulin,” he explained, “serves to coordinate fuel storage and fuel mobilization into and out of the various depots with the needs of the organism, and with the availability or lack of availability of fuel in the environment.” When I interviewed Cahill in 2005, he told me it was true that “carbohydrate is driving insulin is driving fat.” But Cahill did not consider this chain of cause and effect to be a sufficient reason to speculate that carbohydrates drive obesity. Nor did he consider it a possibility that avoiding carbohydrates might reverse the process. Rather, he believed unconditionally that positive caloric balance was the critical factor. When it came to weight regulation, Cahill repeatedly told me, “a calorie is a calorie is a calorie.” He acknowledged that the obese ate no more, on average, than the lean, and this is why he believed that the obese must be fundamentally lazy and this was the proximate cause of their obesity.*121 There was no reason to test competing hypotheses, Cahill said, because any competing hypothesis would contradict the laws of physics as he understood them.
When clinical investigators tried to unravel the connection between diet, insulin, and obesity in human subjects, as the University of Washington endocrinologist David Kipnis did in the early 1970s, the results were invariably analyzed in light of this same preconception. Kipnis had fed ten “grossly obese” women a series of three-and four-week diets that were either high or low in calories, and high or low in carbohydrates. The fat-rich diets lowered insulin levels, Kipnis reported in The New England Journal of Medicine in 1971, and the carbohydrate-rich diets raised them, regardless of how many calories were being consumed. Even when these women were semi-starved on fifteen hundred calories a day, a high-carbohydrate content (72 percent carbohydrates and only 1 percent fat) still increased their insulin levels, even compared with the hyperinsulinemia of these obese women on their normal diets.  
One interpretation of these results is that we could remove the carbohydrates from the diet and replace them with fat, and weight would be lost, perhaps without hunger, because insulin levels would drop, even if the total calories consumed did not. Kipnis’s results, as the University of Heidelberg clinicians Gotthard Schettler and Guenter Schlierf wrote in 1974, underlined the “necessity of restricting carbohydrates in obesity in order to restore insulin levels to normal, thus hopefully decreasing appetite and fat deposition….”
Kipnis, however, refused to believe that carbohydrates might cause obesity, or that avoiding carbohydrates might ameliorate the problem. When I interviewed him over thirty years later, he described the findings of his research as “very obvious.” “You manipulate the amount of carbohydrates you give a human,” he said, “you can manipulate his or her basal insulin level.” He also said that “insulin causes deposition of fat in fat cells.” But when it came to the cause of human obesity or weight gain, Kipnis rejected the relevance of these physiological phenomena. “Most people are obese because they eat more than they need to sustain the energy requirements that they have,” he said. “They eat too damn much.”
As an aside, the opening paragraphs and those that precede it make clear that, in this place in this book, Taubes is making the case that calories have nothing to do with weight gain/loss - it's the carbohydrates.  Skipping down to the last two paragraphs regarding Kipnis, Taubes - for a refreshing change of pace - doesn't mischaracterize Kipnis' position on this matter.  But he does, however, paint him as an ideological zealot who can't see the forest for the trees that it is insulin that makes you fat by causing fat deposition into the fat cells.  Kipnis knew, OTOH, from conducting this study, that these subjects were indeed eating "too damn much".  And Kipnis, although he and Grey weren't specifically looking at this aspect in their study, knew that the obese young ladies maintained weight when they ate an energy balance level of calories (irrespective of whether Carb/Fat/Pro was 62/18/20 or 25/53/22), and lost weight predictably on a 1500 cal CRD (again irrespective of the extremes of carb or fat content that were more extreme in this leg -- essentially zero fat v. zero carb, ~25% protein).

But Gary says "One interpretation of these results is that we could remove the carbohydrates from the diet and replace them with fat, and weight would be lost, perhaps without hunger, because insulin levels would drop, even if the total calories consumed did not."


Nope.  Gary is just pulling another one from his nether-regions in this "analysis".  Did he read G&K and conclude that when insulin levels dropped weight was lost?  I'll repeat the scatter plot of G&K's results that need not even be constructed if one looks at the graphs they did provide, but is instructive nonetheless.

Not only do the insulin levels reported in either leg of this study correlate one iota with weight changes, this is a particularly damning piece of evidence that definitively denounces Taubes' carbohydrate hypothesis.  Because when calories were reduced, subjects lost weight.  Period.  When carbohydrate was varied from as little as 135g to as much as 650g on weight stable caloric intake, the subjects maintained weight.  And weight loss was not impeded by just under 300g carbs on the 1500 cal CRD and weight loss.

In this regard, Gary misrepresents what G&K really demonstrated, and then portrayed Kipnis as one of those willful failures amongst all the other hapless scientific researchers.  I think Taubes owes Grey and especially Kipnis an apology for this.  

So as to point (b) from the outset of this post, one might want to have been a fly on the wall (or bug in the phone) for this conversation, as it would have been interesting to note what Kipnis said regarding his research.  We'll never know.  Imagine how Kipnis must feel with the amount of their conversation Taubes left on the "cutting room floor" that ultimately portrays him in such an unfavorable light in print.  But I digress ...

As to point (c), I don't see that Kipnis came to a different conclusion than I do, nor how this is relevant.  Firstly, tons of blogs and bloggers thrive on "debunking" or otherwise tearing to shreds various studies and the conclusions drawn by the hopelessly biased and paid off by Big ________ (fill-in-the-blank) researchers.  Secondly, tons of secondary research is published every day (everywhere from peer review journals to research blogs to research papers in academia) of this sort.  Kipnis seems to be a calorie guy in his interview with Gary (so we agree) and doesn't come to his "different" conclusion in a manner that has any bearing on the validity of mine.

The conclusion that fasting insulin bears no correlation to weight loss is solid.  Also solid is the conclusion that fasting insulin is not only not *the* factor that determines weight loss, but probably not even *a* factor.  

I doubt Gary read this paper, or if he did he did so with his usually biased eye:  that being one looking for evidence to support one's hypothesis while avoiding inconvenient evidence refuting it along the way.  Here in G&K Gary had a study that demonstrated that in the already "grossly obese", very dramatic swings in dietary content  - utterly inconsistent with what these obese were eating before the study (avg. 48/38/14 carb/fat/pro) - can influence basal insulin levels for relatively short periods of time.  He jumped the gun.

Too bad for him he provides the reader with evidence directly countering his hypothesis.  I'd call this extraordinary.  :-)


Stargazey said...

The conclusion that fasting insulin bears no correlation to weight loss is solid. Also solid is the conclusion that fasting insulin is not only not *the* factor that determines weight loss, but probably not even *a* factor.

For whatever reason, Grey & Kipnis chose not to measure insulin sensitivity in their subjects. We can speculate about their possible insulin resistance, but given their ages, their basal plasma glucose levels and their plasma insulin secretory responses to oral glucose, the subjects were reasonably insulin sensitive.

The high carb load in their ad lib diet produced a high basal insulin. A lower carb load lowered basal insulin, but it returned to a high level once high carbs were reintroduced.

People who are insulin sensitive are different from those who are not. It is possible that at this point in their history the subjects were able to accommodate a higher or lower basal insulin, but that with the onset of insulin resistance, a higher basal insulin would eventually preferetially push dietary carbohydrates into adipocytes to be stored as fat, while at the same time decreasing the subjects' ability to access that fat for energy.

CarbSane said...

Hi Stargazey, I'm not addressing IR/IS in this post, only the relationship to weight gain and the context in which Taubes referenced the study.

This is the underpinning of Taubes' whole theory - that carbs make us fat because they elicit an insulin response and insulin traps fat in our cells causing it to accumulate. But this doesn't square with the evidence. And, indeed, most obese have some level of both hyperinsulinemia AND elevated free fatty acids. The fat is not locked away well enough!

This study has had me rethink some thoughts regarding postprandial insulin levels and chronic hyperinsulinemia. Certainly HI under these conditions is not a good thing and LC weight loss is probably not only a viable option for weight loss, but likely the preferable route for most.

Stargazey said...

...that carbs make us fat because they elicit an insulin response and insulin traps fat in our cells causing it to accumulate.

Is that what he says? Because if he does, it would mean that no one could ever eat carbs without getting fat.

CarbSane said...

That's what Gary Taubes is all about. His basic premise is that overeating or sedentary behavior doesn't make us fat but eating carbs which drives insulin which drives fat accumulation does. A growing child doesn't grow taller from overeating they "overeat" to grow taller. Likewise Taubes asks "if vertical, why not horizontal". Fat accumulates and then you overeat. Or not. He can't seem to keep consistent on whether or not the amount the obese eat has anything to do with their state.

eulerandothers said...

A study from Kipnis the year before (1970) had this concluding paragraph:

The observation that the elevated plasma insulin levels of obese individuals decrease progressively toward normal in prolonged starvation (30), even though they are still markedly overweight, is consistent with this thesis. On the basis of the present studies, one might predict that removing carbohydrate from the diet of the obese individual and substituting a diet consisting almost exclusively of fat, even in amounts sufficient to maintain obesity, might result in a progressive fall in the basal plasma insulin level. Preliminary studies in our laboratory have demonstrated such to be the case; namely,
obese subjects exhibited a progressive fall in their basal plasma insulin levels when fed a 3500 cal diet consisting of 90% fat and 10% protein despite the maintenance of their obese state.

CarbSane said...

Ahh, the rat study that got Peter's boxers bunched.

But obese subjects exhibited a progressive fall in their basal plasma insulin levels when fed a 3500 cal diet consisting of 90% fat and 10% protein despite the maintenance of their obese state flies in the face of all those anecdotal claims of people eating more and losing weight like gangbusters on a VLC/VHF diet.

Each of us may have a "metabolic individuality" that gives an advantage to one extreme of dietary composition or another, but it's not likely more than a hundred calories absent a significant digestive issue.

Stargazey said...

So if caloric intake is unchanged, obese nondiabetic subjects are able to maintain their weight even when their insulin levels decrease substantially.

How do you explain the observation that administration of exogenous insulin often causes weight gain in diabetics? Are these people simply replacing pounds that were lost before their diagnosis of type 2 diabetes?

Sanjeev said...

I've always felt that was a terrible analogy; telling people they shouldn't run because people with broken femurs can't run.

MYTH: Since diabetics who inject insulin gain weight, this means that insulin is the reason for weight gain in non-diabetics

FACT: Amylin is co-secreted with insulin in non-diabetics; amylin has appetite suppressant and lipolytic effects

I would like to thank Dr. Stephan Guyenet for this information. I had known about amylin but hadn’t looked into it in any great detail. Amylin is a hormone that is secreted by your pancreas at the same time as insulin. Amylin decreases appetite, and also stimulates lipolysis (the breakdown of fat into fatty acids).

Type 1 diabetics do not produce amylin, and amylin secretion is impaired in type 2 diabetics. Pramlintide, a drug that mimics the effects of amylin, has been found to produce weight loss in diabetics.

This information demonstrates that the effects of insulin injection in a diabetic cannot be compared to the effects of physiological changes in insulin in a non-diabetic, yet many people erroneously make this comparison as if they are similar.

Sanjeev said...

that's one of James Krieger's series, "Insulin: An Undeserved Bad Reputation, Part 2"

should have made it CLICKABLE

Sanjeev said...

eulerandothers ...

for real?
I always wanted to ask Euler (and others)

"how do you pronounce Euclid ?"

(maybe I should have prefaced the post with "caffeine fuelled GEEK HUMOUR AHEAD !!!" )

Sanjeev said...

By the way CarbSane, I admit (in a court of law) that everything you say is true.

Now I DEMAND, repeatedly, annoyingly, repeatedly, annoyingly, with no new argument or logical justification (ie, extreme prejudice), that you stop saying it.

BECAUSE it's true (and I stipulated it's true in a court of law, so MOVE ON.)

(did I mention the court of law? ) repeatedly? Annoyingly? illogically? well then I'll repeat it annoyingly one more time ...

CarbSane said...

@Stargazey, I suspect aggressive insulin therapy results in some hypos or near hypos and increased food intake. It doesn't take much for any of us to gain 10 lbs in a year - 100 cal/day - an amount that could easily go undetected for all but those who diligently track/monitor.

RRX said...


The link you posted said that the weight gain "has been attributed to anabolic effects of insulin, appetite increases, and reduction of glycosuria". None of those things are the mechanisms that Taubes ascribes to.

The second link did not have to do with the food consumed (track amounts/kcals ingested and energy expended). Therefore, it fails to speak to the mechanism by which the weight gain occurred.

simik said...


"Anabolic effect" is the property of insulin to signal cells to store nutrients, that includes signaling adipose tissue to store fat. That's exactly the mechanism touted by Taubes.

Appetite increase may be an effect of hypoglycemia and/or of blocked fat release due to hyperinsulenemia.

How can "reduction of glycosuria" in and by itself result in weight gain is beyond me.

CarbSane said...

Welcome simik! The reduction of glucosuria means a diabetic is no longer peeing out calories in fairly significant amounts.

Nobody denies the anabolic effect of insulin. It's this notion that it's some sort of renegade hormone that locks fat away permanently so that carbs make you fat but somehow fat doesn't. Fatty acids make a beeline to the fat tissue, carbs do not. Eat too much of both and yes, the fats get socked away for later, but don't eat too much and the fats temporarily "trapped" are released later in the day.

RRX said...


Along those lines, you made me think of a question that I cannot imagine the answer to. Perhaps you or someone else on here knows. What does Taubes say about muscle growth? Anything? Does he think that muscle and fat are added just the same, at the same time? How does he explain the growth of muscle involving insulin? Has he even bothered with that area, or has he brushed it off with his anti-physical activity view?

Stargazey said...

CarbSane, my comment about insulin resistance is relevant because it appears that insulin resistance is what traps fat in fat cells. (Of course, because the body uses rheostatic mechanisms and not on/off switches, it's not really "trapped" there. It's just harder for the fat to move out of adipose than it normally would be.) Because these women were not insulin resistant, their fat stores were still freely accessible and they were still operating on a paradigm that was not restricted by insulin responsiveness.

I do wonder why there was no drop in weight when they moved from the ad lib diet to the low-carb diet. Normally when people have been eating lots of carbs, there is a significant loss of water weight when transitioning to a low-carb diet, yet none of these subjects demonstrated it.

Stargazey said...

RRX, the links I gave were not intended to prove the mechanism of weight gain subsequent to treatment with exogenous insulin. I simply wanted to point out that there appears to be an association between chronic insulin administration and weight gain. The first set of authors may have attributed the weight gain to "anabolic effects of insulin, appetite increases, and reduction of glycosuria," but as you know, attribution does not constitute proof.

Sanjeev, Dr. Richard Bernstein does give Symlin to type 1 diabetic patients who have a hard time controlling their appetite, but it does not work in all cases.

RRX said...

If not to discuss the mechanisms of a causal connection between these factors of weight gain/loss, then what are we discussing? Associations alone are worthless.

Stargazey said...

RRX, correlation is useful in suggesting relationships between variables. Once a relationship is observed, we can certainly discuss possible mechanisms for the relationship. However, the fact that a particular author or group of authors speculates something, does not thereby prove or disprove it.

Sanjeev said...

> Dr. Bernstein does give Symlin to type 1 diabetic
> patients who have a hard time controlling their

My point wasn't about any specific action of any specific chemical agent, just making the general point you made later - one cannot conclude much about the "normal obese" from the experience of diabetics.

RRX said...

1. Re. "...correlation is useful in suggesting relationships between variables. Once a relationship is observed, we can certainly discuss possible mechanisms for the relationship." I understand that perfectly well. Hence my statement, "Associations alone are worthless." And my desire to go further to discuss the possible underlying mechanisms and what the research has found, and in the case of this specific post how that compares to the claims of Taubes.

2. As Sanjeev cited and stated, we cannot necessarily say that diabetics represent "normal" obese.

simik said...

What does Taubes say about muscle growth? Anything? Does he think that muscle and fat are added just the same, at the same time? How does he explain the growth of muscle involving insulin?
I don't remember Taubes delving into muscle growth in GCBC.

Unlike adipose tissue muscles require a stimulus of overloading to signal growth, hyperinsulinemia by itself won't do much.

CarbSane said...

@Stargazey: I disagree with this assertion:
" insulin resistance is what traps fat in fat cells."

Everything I'm reading seems to indicate that pathological insulin resistance initiates with fat cells losing their insulin sensitivity. I've blogged on a study demonstrating that larger fat cells become less sensitive to insulin. This is supported by the (again can't recall the exact one OTTOMH) diabetes drug that improves the condition by prompting adipocyte differentiation. The folks tend to gain weight but gain improved metabolic control.

RRX said...


Again, does Taubes speak to that question?

From the Taubes line of reasoning and what you said, correct me if I am wrong, the logic is that the nutrient environment can be the same but with the addition of an overloading stimulus, the response of the body is to grow muscle...but not to gain fat?

Sanjeev said...

in GCBC Taubes wrote about insulin's putative effect of proliferating arterial smooth muscle cells as an initial step toward atherosclerosis.

I would have to check again, but there may be stuff about muscle insulin resistance and insulin's effects on mitochondria.

I believe those were the only insulin-to-muscle interactions he speaks about (I'm sure of the first one, not so much about the next two) ... nothing I can remember about skeletal muscle anabolism.

Sanjeev said...

> may be stuff about muscle insulin resistance

as related to obesity, not fitness / muscle growth

Post a Comment

Moderation is currently on. Thanks in advance for your patience.