Insulin & Weight/Fat Loss ~ The Long and the Short of It

In the comments of my recent post, On Cherry Picking & Debunking, my "Insulin Warring" partner Todd Becker posted a multi-point commentary.  I'll get to the rest over in the comments section in that post when I get a chance, but one of his points hit on something just a little different as it goes to the core of my issue with the insulin hypothesis. (And I'm reminded I'm long overdue for an installment on the IW exchange with Todd)

In any case, here was what I said:
G&K demonstrated absolutely no correlation between fasting insulin and weight change. Therefore fasting insulin is not a factor in determining weight loss.
Todd challenged:  
Your claim that there is "no correlation" between insulin levels and weight looks only at the instantaneous correlation. But fat loss may be a delayed response, as the Woodhouse paper I linked to you indicates. Figures 1 and 2 in G&K suggest that weight either dropped or plateau'd after a slight lag in the few cases where insulin levels approached 10 uU/ml. That the weight loss continued in some cases even after insulin levels begin to increase suggests that weight regain may be a delayed metabolic response.
The G&K protocol used "fasting insulin" as a surrogate for basal insulin. These are not the same thing. Fasting insulin was measured twice weekly at 7-8 a.m. after a 12 hour overnight "fast". But fat loss may depend on how much of the day insulin levels remain very low. Truly low basal insulin levels may take many weeks to establish.

Let's first address the Grey & Kipnis study with a discussion that comes full circle on one concept as regards insulin.  But first a little salute to Todd Becker for being one of the few folks Jimmy Moore queried for his blog who appears to have actually read James Krieger's articles.  He also did an excellent job, all disagreements I may have with his analysis aside, of addressing James' assertions.  So rather than sending you on a trip about the web, in the first installment of his series, Insulin…an Undeserved Bad Reputation, James discusses something that I've repeated often regarding Taubes' theories on insulin.  That being that insulin is the fattening hormone.  It's not that Taubes gets the biochemistry/metabolism wrong, it's that he takes it wholly out of context.  James constructed this graphic shown in the screenshot below:

James precedes this graphic with the following:
One misconception regarding a high carbohydrate intake is that it will lead to chronically high insulin levels, meaning you will gain fat because lipogenesis will constantly exceed lipolysis (remember that fat gain can only occur if the rate of lipogenesis exceeds the rate of lipolysis).  However, in healthy people, insulin only goes up in response to meals.  This means that lipogenesis will only exceed lipolysis during the hours after a meal (known as the postprandial period).  During times when you are fasting (such as extended times between meals, or when you are asleep), lipolysis will exceed lipogenesis (meaning you are burning fat).  Over a 24-hour period, it will all balance out (assuming your are not consuming more calories than you are expending), meaning you do not gain weight. 
I am in 100% agreement with James' analysis on this point.   Essentially one can only consider carbohydrates and insulin "fattening" if one takes the biochemistry/endocrinology out of context.   Yes, when insulin levels are high it favors net fat deposition, and low, net fatty acid mobilization.  But in normal individuals "insulin spikes" play a rather minor role in basal insulin levels and are balanced out by low basal insulin levels.  

Basically James and I point to a fundamental flaw in the insulin hypothesis:  Extrapolating the short term (instantaneous)  effects of insulin to the long term regulation of body weight.  

See where I'm going here?  Todd is now saying that in Grey & Kipnis, we're looking at the "instantaneous" picture and ignoring the *possible* delayed effect of lowered insulin.  Respectfully, I will say that you really can't have your coconut flour cake and eat it too.  Either the rather dramatic fluctuations in fasting insulin achieved by G&K in already obese young women, through dramatic variation in carb content of the diet, impact weight** or not.  Either the instantaneous action of insulin that (paraphrasing) "fixes and traps fat" in our fat tissue leads to long term weight gain, or it doesn't.  I'm often chided to look at "real life" observations more.  There again, some folks (myself included) seem to lose weight very rapidly at first (even after the initial glycogen associated water weight whoosh is over) long before the hyperinsulinemia of obesity has subsided substantially.  I contend that Todd's personal experience of delayed weight loss once his insulin finally lowered is diametrically opposed to what the majority of low carbers experience:  rapid initial losses that plateau out (often well above a normal weight) after a period of time when their fasting insulin finally comes down.

Let's look once again at the results for G&K:

To refresh, these diets were ~25% protein and either essentially 75% carb or 75% fat.  A truer head-to-head of high carb vs. high fat we couldn't get.  Todd contended in other comments that the durations of the diets may not have been sufficient either.  There are several problems with this, we see no "delayed" response to which he eludes, only continued maintenance or weight loss.  That weight loss continued even after fasting insulin was again raised, I suppose *could* be explained by some delayed effect, but here again I'm sensing some mental gymnastics.   Tautology girl here (grin) would suggest a simpler explanation:  weight loss continued because calorie deficit continued.  Where there are some anomalies for the maintainers, there's no consistency.  This was the purpose of my constructing the scatter plot of change in weight vs. change in insulin.  

I stand by my conclusion, to repeat:
G&K demonstrated absolutely no correlation between fasting insulin and weight change. 
In order to even hint at causality, one must at least identify a correlation.  Nothing to see here folks.

I remain a bit confused over what the timeframe for a delay might be.  In the weight loss group we were talking a month on each diet.  They lost weight continually throughout the three months on 1500 cal/day.  Beginning with high carb, proceeding through low carb, and returning to high carb.  I'm not sure that assessing fasting insulin and weight weekly would qualify as any sort of "instantaneous" picture.

**Before continuing, I have to address the whole weight vs. fat mass thing.  Unfortunately G&K only measured body weight, not fat mass.  This is what most of us measure on our bathroom scales as well.  I've not seen any data demonstrating that the proportion of fat/total mass lost on LC differs from that achieved with other dietary interventions.  Therefore, absent additional information, I worded my conclusion carefully, fully aware that fat and weight loss are not the same thing (and I've come across yet another study demonstrating more favorable fat loss with exercise that I'll be blogging on soon, and I'll address the Woodhouse paper at that time if I can get my hands on the full text) . 

OK, onto Todd's other point regarding fasting insulin vs. basal insulin.  Generally these are the same thing.  Perhaps he meant insulin AUC = total insulin exposure.  To discuss this I'll repeat the graphic first highlighted here:

This is the 24 hour insulin profile for overweight (avg. BMI 26) women, all consuming a 50% carb diet:  
□ = isocaloric 15% protein,  ● = isocaloric 30% protein, Δ = ad lib 30% protein (weight loss)   
Hmmmm... draw a horizontal line around the 35 line for insulin and you have .... ta da! ... James' graphic!

These subjects had fasting/basal insulin levels of around 10-15.  Even the AUC insulin exposure did not differ much.  In this study they first replaced 15% of fat calories with protein at isocaloric weight maintaining levels.  Weight was steady.  Then they offered the high protein diet ad libitum and the subjects (who were told not to expect weight loss, this wasn't a weight loss study!) and they spontaneously reduced intake by ~440 cal/day and lost an average of 10.8 lbs in 12 weeks.  As the study authors stated, the weight loss was fully explained by the reduction in caloric intake.  I would also note that the insulin levels "spiked" and remained rather high for several hours throughout the day for all three dietary interventions.

I'm not sure what sort of delayed effect of insulin I'm supposed to be waiting for here.....

So ... to all still clinging to the last vestiges of insulin theory, I implore you.  Try to forget everything you've "learned" from a certain guru.  Look at the data objectively.  

We've come full circle from James/Me addressing the error of extrapolating short term insulin action to long term accumulation of fat, to Todd now inferring some long term action of insulin we're somehow missing in these studies only looking at a snapshot of fasting insulin levels.  

Embrace the tautology!!!  It's not all that interesting, I know.  But it tells us a lot.

(Not directed at Todd)  Let the why-ning begin!!


James Krieger said…
It seems to me that Todd is not drawing his conclusions from the evidence. Instead, he is starting with his conclusions and coming up with ad hoc rationalizations as to why the evidence doesn't support his conclusions.
Sanjeev said…
you're allowed to argue opposites in law

"my client didn't do it, and if he did to it there were mitigating circumstances, and even if there were no mitigating circumstances, precedent xxx lets him off the hook, and if it doesn't ..."

but in the real world it would make sense to separate out the conflicting criticisms into streams, like so:

"I propose this theory, aaa for why they're wrong, and from there we conclude bbb"

"Another theory which would make them wrong is xxx, and following this, we can conclude yyy"

Such an analysis would seem a little clearer to me, anyway.
Muata said…
Great post CS, but what I've never been able to understand about the insulin hypo (and I was a believer for a while) is that all the supposed "evils" of insulin only happen when you eat more calories than your body needs ... oh, and this happens regardless of the macro make-up of the diet. Will there ever be a point when the main proponents of the insulin hypo simply fess up?
CarbSane said…
Y'know Muata, many revered voices have quietly shifted course, but they dare not counter this too vocally or publicly. It's an odd phenomenon, especially as one reader commented once how the LC movement prides itself on being so "science based", countering flawed/dogmatic "conventional wisdom". Sigh.
Muata said…
What's really sad CS is that following a sensible LC program IS the best approach for many obese folks--not ALL. If the same folks who spend time arguing over their LC way of life would simply count calories, spend less time arguing over the internet, and move more, I think the LC community would be a nicer and more pleasant place for those looking for a "system" that may, or may not, work for them over the long haul.

I dunno, maybe it's just human nature to want to believe that your way is THE right way for everyone. And this mentality, unfortunately, can be seen in all of these dietary tribes (LF, LC, Vegan, Paleo, etc.). Go figure ...
CarbSane said…
I think it was also you Muata who commented on a desire to be "different". In the low carb/paleo/etc. world that seems to be feeling like one is onto some cutting edge, conventional wisdom shattering epiphany but it's just not so. LC is great for many, but it is not the be all and end all and I see far too many people beating themselves upside the head because they can't seem to stick with it. Granted that's not just LC, but rather than seeking out something they can stick with, these folks keep returning and recommitting and self-prostelytizing in hopes that this time they can make that square peg fit.
CarbSane said…
Posting for Todd Becker:

James - I was not “starting with” or defending any positive conclusions; I’m mere pointed that CarbSane’s interpretation of G&K is not airtight, by raising an alternate possibility. Saying that “A is not proven” is not the same as proving A is false or not-A is true.

Carb Sane and Sanjeev – There is no contradiction in my position. I think you missed my main substantive point about relative drops in insulin levels vs. absolute insulin levels. I acknowledged that when the obese subjects reduced carb levels to 25%, their insulin levels adjusted quickly in relative terms, dropping from 38 to 19 uU/ml. My point was that 19 uU/ml as an absolute level may still be too high to facilitate of fat loss. And the quantitative B-L-D data Carb Sane shows to illustrate James’ qualitative graphic, the insulin never drops below about 20 uU/ml. It may take more time (and/or a diet with lower than 25% carbs) in order to drive insulin levels sufficiently low to inhibit LPL and upregulate HSL, facilitating fat loss.

I’m raising this as a possibility that you have not ruled out; I’m not claiming it as a fact. There is a big difference! However, my “hypothesis” is certainly consistent with observations that basal insulin levels above 10-15 uU/ml are associated with pre-diabetes and insulin resistance ( with the ultra low basal insulin levels (below 5 uU/ml insulin) of lean non-Westerners like the Kitavans and Okinawans, and with my own fat loss that came slowly as I got my basal insulin levels from 15 down to 4 uU/ml. Furthermore, I say this acknowledging that there is no problem with periodic insulin spikes; the key is that insulin may have to go below a threshold of something like 8-10 uU/ml for many hours a day to facilitate fat loss.

If you want to challenge me to go further and substantiate this hypothesis as a positive conclusion, I’d be more than happy to try. But the intent of my argument was to play skeptic to your position – not advocate of an alternative hypothesis. I’m sure that each of you, being versed in logical and legal argumentation, can well understand the difference.


Todd is having the following problem posting in comments:

I tried posting today and it asks me to create a profile in one of Google, WordPress, LIveJournal, TypePad, OpenID)etc. But when I try to create such a profile, it rejects my attempt, saying "Your Wordpress (or Google or OpenID etc.) credentials could not be verified. Or it just closes the page and my comment is deleted.

If any of my readers have experienced this or have a solution for Todd, feel free to post here or email me through my profile.
CarbSane said…
@Todd: I'm thinking that if someone looks at the 24 hr insulin profiles from that study ( and knows nothing of what they represent, would you not agree they are indistinguishable from each other? They could be 24 hr profiles for the same person 3 months apart, or for any sort of dietary intervention.

So this is where it seems that you are starting from a conclusion and trying to explain the results. If I just looked at those graphs I see no difference. This is how the graphs - that showed weight - "popped" out at me off the page of G&K even before I read so much as the abstract of that paper.

One of the three groups lost considerable weight, the others maintained on different diets. Insulin levels were "sufficient to facilitate weight loss" for one group but not the others if they were essentially the same? I'm not getting where you're coming from on this.

BTW: Unless I read incorrectly, the LC diet was basically 25% protein 75% fat, and the HC diet was 25% protein 75% carb.